Pozzolanas and Pozzolanic Materials

9

Pozzolanas and Pozzolanic Materials Michael John McCarthy☆ and Thomas Daniel Dyer☆

9.1 INTRODUCTION The term ‘pozzolana’ can have two distinct meanings. The first refers to pyroclastic rocks, essentially glassy and sometimes zeolitised, which occur either in the neighbourhood of Pozzuoli or around Rome.1 The second meaning includes all those inorganic materials, either natural or artificial, which harden in water when mixed with calcium hydroxide (hydrated lime), or with materials that can release calcium hydroxide, such as Portland cement (PC). However, in this chapter, the term ‘pozzolana’ will be used to refer to naturally occurring materials, whilst the wider group of materials will be referred to as ‘pozzolanic materials’. ‘Blended cement’ is taken to mean a combination of these materials with PC. For a long time, the use of pozzolanas has been mostly restricted to Italy—where considerable reserves of natural pozzolanas are found—and to Greece (Santorin earth). In other countries, the interest in these materials is relatively recent and has arisen from the need for reusing industrial by-products such as fly ash and silica fume. This historical background can help explain why so many countries have long distrusted pozzolana-containing cements, despite the historical use of lime– pozzolana mortars and the almost 100-year experience with blended cements. Results of many studies have substantially confirmed that blended cements can yield concrete showing a high ultimate strength and enhanced resistance to the attack of aggressive agents. In addition to the previous edition of this book,1 the properties and applications have been exhaustively covered in proceedings devoted, either partly or completely, to pozzolanic materials, and in many technical papers. Establishing a precise classification of pozzolanic materials proves difficult since this common name includes those which are very different in terms of chemical composition, mineralogical nature and geological origin and which are related only by the general property they have, to react and harden when mixed with lime and water. The more commonly accepted classification concerns the origin of pozzolanic materials and, therefore, one common subdivision is between natural and artificial materials. Natural materials do not require any further treatment apart from grinding; artificial pozzolanic materials result from chemical and/or structural modifications of materials originally having no or only weak pozzolanic properties. The latter can be residues of certain production methods or products manufactured from selected raw materials. The division between natural and artificial pozzolanic materials is not well defined, since there are materials, such as Danish moler, French gaize and some rhyolitic tuffs from the United States which, besides typically pozzolanic constituents, also contain variable amounts of clay minerals which only take on a clear pozzolanic character by firing. A proposed classification of pozzolanic materials is shown in Fig. 9.1.2

9.2 TYPES OF POZZOLANIC MATERIALS 9.2.1 Natural Pozzolanas 9.2.1.1 Materials of Volcanic Origin (Pyroclastic Rocks) Pyroclastic rocks result from explosive volcanic eruptions, which project droplets of molten magma into the atmosphere. The rapid pressure decrease occurring during the eruption causes the gases originally dissolved in the liquid magma to be released. As a consequence, each particle will contain a number of bubbles forming a microporous structure.3 Simultaneously, the particles are subject to a quenching (either in air or water) process which is responsible for their glassy state. The microstructure of three typical Italian volcanic pozzolanas is shown in Figs 9.2–9.4. Non-explosive eruptions produce volcanic ashes with little or no pozzolanic activity since quenching is not sufficiently rapid to prevent crystallisation.4 Incoherent Materials ‘Incoherent’ in this context means particles which are not strongly attached to one another. They include Italian pozzolanas from Campania (Naples) and Latium (Rome); the so-called Santorin earth from Greece; the incoherent glassy rhyolites, to be ☆

We wish to acknowledge that this chapter draws upon that of Professor F. Massazza in edition 4 coupled with current revisions.

Lea’s Chemistry of Cement and Concrete. https://doi.org/10.1016/B978-0-08-100773-0.00009-5 © 2019 Elsevier Ltd. All rights reserved.

363

364 Lea’s Chemistry of Cement and Concrete

FIG. 9.1 Classification of pozzolanic materials. (From: Massazza F. Chemistry of pozzolanic additions and mixed cements. Il Cemento 1976;1:3–38.)

FIG. 9.2 SEM image of Bacoli pozzolana (Italy) (600
).

Pozzolanas and Pozzolanic Materials

FIG. 9.3 SEM image of Salone pozzolana (Italy) (300
).

FIG. 9.4 SEM image of Vizzini pozzolana (Italy) (300
).

365

366 Lea’s Chemistry of Cement and Concrete

found in the United States,5 India (Bombay)6 and Turkey.7 Rhine trass is more commonly included among the tuffs, that is, compact, coherent materials, but its deposits also contain incoherent layers mainly made up of glass.8 The Japanese Furue Shirasu and Higashi Matsuyama pozzolanas also belong to this group of glassy volcanic pozzolanas.9 Bavarian trass is strictly not a volcanic pozzolana, being formed as a result of the impact (shock) of a large meteorite that produced the Ries crater.10 Table 9.1 shows that the chemical composition of incoherent volcanic pozzolanas varies within wide limits and that silica and alumina prevail over other constituents. The alkali content (Na and K) can potentially be high—approaching 10% by mass.8 Loss-on-ignition (LOI) also varies significantly, from very low values to as much as 10%. This would appear to mainly reflect the loss of water of crystallisation from clay minerals, as well as decomposition of carbonate minerals. The mineralogical composition of some volcanic pozzolanas is shown in Table 9.2. The materials consist of a glassy matrix in which crystalline mineral inclusions are embedded. The table describes the glass as the active phase since it will normally be the main phase to undergo pozzolanic reaction. The crystalline phases can comprise silicate minerals including tectosilicates, such as quartz and feldspars; phyllosilicates, including clays and micas; nesosilicates, including olivines; inosilicates including augite and diopside and feldspathoids. Additionally, quantities of carbonate minerals, iron oxides (such as magnetite) and fluorite have been identified.5,12,16 Compact Materials (Tuffs) The deposits of volcanic pozzolanas are often associated with compact layers (tuffs) which originate from weathering and cementation of loose particles by diagenetic or other natural processes. Weathering can cause zeolitisation (transformation into zeolite minerals, often under pressure and/or elevated temperature, in the presence of water) and, probably, argillation (the formation of clay minerals in a similar manner to zeolitisation)17: the volcanic glass can be transformed into zeolitic (and feldspathoid) minerals or clay minerals. Zeolitisation of volcanic glass has been demonstrated unambiguously in laboratory experiments,18–20 although the feasibility of argillation

TABLE 9.1

Chemical Analyses of Some Incoherent Volcanic Pozzolanas (%)

Pozzolana

Country

SiO2

Al2O3

Fe2O3

CaO

MgO

Na2O

K2O

TiO2

SO3

LOI

Ref.

Bacoli Barile Salone Vizzini Volvic Santorin earth Rhine tuffash Rhyolite pumicite Furue shirasu Higashi Matsuyama

Italy Italy Italy Italy France Greece Germany USA Japan Japan

53.08 44.07 46.84 50.48 54.30 65.1 58.91 65.74 69.34 71.77

17.89 19.18 18.44 16.77 16.80 14.5 19.53b 15.89 14.56 11.46

4.29 9.81 10.25 11.59

9.05 12.25 8.52 4.86

1.23 6.66 4.75 5.24

3.08 1.64 1.02 0.87 4.47

7.61 1.12 6.35 0.17 2.60

0.31 0.97 0.06 0.08

0.65 tr.a tr. 0.15

3.05 4.42 3.82 9.68

5.5 2.85 2.54 1.02 1.14

3.0 2.48 3.35 2.61 1.10

1.1 1.33 1.33

6.12 4.97 3.00 1.53

4.53 1.92 2.39 2.55

11 11 11 11 12 13 8 5 9 9

0.54

0.25 0.14

3.5 2.21 3.43 1.85 6.50

tr. ¼ trace. Al2O3 + TiO2.

a

b

TABLE 9.2

Minerals in Some Volcanic Pozzolanas

Pozzolana

Country

Active Phases

Inert Phases

Ref.

Bacoli Barile Salone Vizzini Volvic Santorin earth Rhine trass Bavarian trass Rhyolite pumicite Furue Shirasu Higashi Matsujama

Italy Italy Italy Italy France Greece Germany Germany USA Japan Japan

Glass Partially decomposed glass Glass, analcime Glass Glass Glass Glass (55%–60%) Glass (62%–67%), chabazite (3%), analcime (5%) Glass (80%) Glass (95%) Glass (97%)

Quartz, feldspars, augite Pyroxenes, olivine, mica, analcime Leucite, pyroxenes, alkali feldspars, mica Feldspars, quartz, olivine, clay minerals Andesine, quartz, diopside Quartz, anorthite, labradorite Quartz (9%), feldspar (15%) Quartz (19%), feldspar (15%) Clay (5%), calcite, quartz, feldspar, etc. (15%) Quartz (1%), anorthite (3%) Quartz (1%), anorthite (1%)

11,12,14 11 11,15 11 12 13 10 10 5 9 9

TABLE 9.3

Pozzolanas and Pozzolanic Materials

367

LOI

Ref.

11.10 7.41 16.33 6.34 9.11 7.27 12.06

10 10 23 23 21 24 6 25 26 26 27

Chemical Analyses of Pozzolanic Tuffs (%)

Pozzolana

Country

SiO2

Al2O3

Fe2O3

CaO

MgO

Na2O

K2O

Rhine trass Bavarian trass Selyp trass Ratka trass Yellow tuff Dacite tuff Gujarat tuff Trass K Zeolite (clinoptilolite) Zeolite (mordenite) Opoka

Germany Germany Hungary Hungary Italy Romania India Bulgaria Japan Japan Lithuania

52.12 62.45 55.69 73.01 54.68 67.70 40.9 71.63 71.65 71.11 54.1

18.29 16.47 15.18 12.28 17.70 11.32 12.0 10.03 11.77 11.79 2.7

5.81 4.41 6.43 2.71 3.82 2.66 14.0 4.01 0.81 2.57 1.3

4.94 3.39 2.83 2.76 3.66 3.73 14.6 1.93 0.88 2.07 23.2

1.20 0.94 1.01 0.41 0.95 1.64 1.45 1.22 0.52 0.15 0.62

1.48 1.91

5.06 2.06

SO3

0.26 0.10 3.43

6.38 0.18

2.35 1.80 1.66 0.24

3.44 1.33 0.87

3.05 0.34 0.27

9.04 9.50 16.97

has not been similarly demonstrated. The degree of transformation reached by the original deposit depends on the intensity of the diagenetic actions as well as on their duration. Zeolitisation improves pozzolanic properties, whereas argillation reduces them.21,22 The chemical compositions of some pozzolanic tuffs are shown in Table 9.3. The silica and alumina contents are, as would be expected, comparable to the incoherent materials described previously. LOI is generally higher, mainly the result of the incorporation of water of crystallisation during the zeolitisation and argillation processes. In contrast, the mineralogical composition of tuffs is more complex15 since the volcanic glass of the original material is transformed by an autometamorphic process28 to zeolite compounds such as herschelite ((Na,Ca,K)AlSi2O63H2O), chabazite ((Ca,Na2,K2,Mg)Al2Si4O126H2O) and phillipsite ((Ca,Na2,K2)3Al6Si10O3212H2O).8,29 Glass often still makes up a substantial proportion of compact tuffs. For instance, the main minerals in Rhine trass and Bavaria trass are glass (55%–60%), quartz (9%), feldspar (15%) and glass (62%–67%), quartz (19%), feldspar (15%), respectively, alongside variable amounts of zeolite and clay minerals.10

9.2.1.2 Materials of Sedimentary Origin Some sedimentary rocks are capable of combining with lime. These include certain clays and the so-called diatomaceous earths. The former originates from the alteration of igneous rocks, whereas the latter form from the siliceous skeletons of microorganisms (diatoms) deposited in fresh or sea waters. It is not uncommon for diatoms and clay minerals to occur together. Clay minerals, especially those belonging to the montmorillonite group, can react with lime giving calcium silicate and aluminate hydrates,30–33 but they cannot normally be used as pozzolanas, since they increase the water demand and tend to lower the strength of mortar and concrete. The largest deposit of diatomites occurs in California, USA. Other important deposits are found in the former USSR, Canada, Algeria and Germany1 and in other countries. A diatomaceous earth, the so-called moler, which consists mainly of a mixture of montmorillonite and amorphous opal, is found in Denmark, where it has been used either as-received, or else calcined. The pozzolanic properties are remarkably improved if moler is burnt in order to decompose the clay minerals.34 Table 9.4 shows the chemical composition of some diatomaceous earths. The pozzolanic constituents of the materials are opal and clay minerals. The opal content in diatomites ranges from 25% to 100%.5 Other minerals include quartz and feldspars. The clay content of diatomaceous earths is reflected in the alumina content, with a consequent decrease in silica. Diatomaceous earths are highly reactive towards lime owing to their high content of amorphous silica and high specific surface area. In spite of the strongly pozzolanic behaviour, the use of diatoms in blended cements is hampered by their high specific surface area, which causes the water demand of cement to increase. Small additions of diatomites to concrete can improve plasticity and reduce bleeding. 9.2.1.3 Materials of Mixed Origin (Hybrid Rocks) North of Rome there are stratified deposits of a crumbly rock (Sacrofano earth; Fig. 9.5) composed of materials of different origin (volcanic, sedimentary and organic).11,35,37–39 The upper layers show a silica content up to 90%, considerable LOI and small amounts of other oxides. As far as the innermost deeper layers are concerned, the silica content is appreciably less; alumina can reach about 20%, but other oxides are present only in very small amounts.35 The low iron content explains the light colour of these materials, generally known as ‘white earths’.

368 Lea’s Chemistry of Cement and Concrete

TABLE 9.4

Chemical Analyses of Silica-Rich Pozzolanas of Different Origin (%)

Pozzolana Diatomaceous earths Moler Diatomite Diatomite Mixed origin Sacrofano White hearth (a) White hearth (b) White hearth (c) White hearth (d) Beppu white clay Gaize

Country

SiO2

Al2O3

Fe2O3

Denmark USA USA

75.60 85.97 60.04

8.62 2.30 16.30

6.72 1.84 5.80

Italy Italy Italy Italy Italy Japan France

85.50 90.00 84.25 78.40 56.80 87.75 79.55

3.02 2.70 4.50 12.20 21.40 2.44 7.10

0.44 0.70 1.55 1.50 1.70 0.41 3.20

TiO2

1.22

1.10

CaO

MgO

Na2O

K2O

SO3

LOI

Ref.

1.10 trace 1.92

1.34 0.61 2.29

0.43 0.21

1.42 0.21

1.38

2.15 8.29 11.93

34 5 5

0.16

0.26

0.77

7.94 6.10 8.40 8.60 7.50

0.11

0.11 0.86

5.90

11 35 35 35 35 9 36

0.58 0.20 2.40 1.55 2.35 0.19 2.40

0.23 1.04

FIG. 9.5 SEM image of Sacrofano pozzolana (Italy) (300
).

For the layers that are richer in silica, X-ray diffraction (XRD) analysis shows a band at around 0.405 nm that is typical of dried silica gel, which is attributed to the ‘groundmass’—the matrix in which larger crystals are embedded in igneous rock. In the patterns of the high-alumina layers, this band fails to occur and peaks of montmorillonite-type clays attacked by acid solution appear.37 The crystalline minerals, sometimes clear but often altered, consist of feldspar (sanidine or plagioclase), biotite, quartz, calcite, zeolites and diatom skeletons.34 The presence of diatoms with fragments of volcanic rocks shows that these deposits probably originated from the deposition of materials of different origin in stagnant water, followed by acid attack. For the minerals of the upper layers, the destruction of the crystalline structure occurred with the formation of silica gel; for the minerals of the lower layers the transformation was only partial, and zeolite and clay minerals formed. Within the Oita prefecture (Japan) there is a high-silica deposit (Beppu white clay), essentially made of opal. Other minor minerals are quartz, cristobalite and opaque constituents.9 The silica gel deposit originated from the neighbouring rocks made of hornblende and andesite which were decomposed by hot springs.

Pozzolanas and Pozzolanic Materials

369

In Central Asia, Jurassic shales are found which were calcined by natural subsurface coal fires. This material, called gliezh, is rich in SiO2 and Al2O3 due to its clay origin and exhibits pozzolanic activity.40

9.2.2 Artificial Pozzolanic Materials 9.2.2.1 Fly Ash Fly ashes consist of finely divided particles produced by burning pulverised coal (and more recently in combination with other organic materials or co-fuels41,42) in the generation of electricity at power stations. Given the high temperatures reached during the instantaneous burning of the fuel, most of the mineral components melt and form small fused drops. The subsequent sudden cooling transforms them partly or entirely into spherical glass particles. Recognition of the reaction potential of fly ash from electricity generation can be traced to around 100 years ago,43 with some of the initial work in this area carried out in the United States in the 1930s.44,45 Standards and other guidance documents covering the material’s requirements were introduced thereafter.46 Early applications with fly ash, mainly in dams, are reviewed in a paper from the mid-1950s.47 Experience gained in the period since, and the technical, environmental and economic benefits that the material offers, have seen increasing use. Recent developments, with moves to other fuels/means of generating electricity in some places, may affect regional availability of fly ash. The characteristics of fly ash depend on a range of factors including the type of coal (anthracite, bituminous, subbituminous and lignite) burning conditions and the collection system.48 A summary of recent developments, with regard to technology at coal-fired power stations, and which may influence fly ash and its behaviour is given in Table 9.5. These are aimed at reducing emissions and increasing efficiency in the electricity generation process.49–52 Temporary wet-storage and recovery/processing are other approaches to sourcing material, for use as an addition in concrete, that are receiving increasing interest.53,54 Fly ash can be characterised chemically according to the coal type used and major oxides present. In the United States, according to ASTM C618,55 Class F fly ash (siliceous) is commonly produced using anthracite or bituminous coals, with the sum of SiO2, Al2O3 and Fe2O3 exceeding 70%. Class C (calcareous) fly ash is usually derived from subbituminous coals or lignite, with the sum of the above oxides exceeding 50% (Note: There may not always be correspondence between coal type/fly ash classes). It is also mentioned that Class C fly ash typically has a higher total calcium content than Class F. In Europe, EN 450-156 covers siliceous fly ashes and also makes reference to processing (e.g. drying, grinding, etc.) and includes provisions for co-combustion. The chemical composition of bituminous fly ash can vary within certain limits. Silica and alumina are the main compounds, as with natural pozzolanas. Silica usually ranges from about 40% to 60% and alumina from 20% to 30%; the iron content which can be around 5%–10% may in some cases be higher. The lime content is generally between 2% and 5%. Alkalis are present in appreciable amounts, with potassium prevailing over sodium (the limit in EN 450-156, as Na2Oeq, is 5%), while sulfate levels are typically up to about 2%. TiO2, is found at around 1%, together with minor oxides. Carbon is also present to a lesser or greater degree, depending on the coal type, burning process, and the use of carbon-removal techniques, for example, electrostatic, which have been installed at some power stations (limits in standards, as LOI, are 6%55 and 5% to 9% for different categories56). Since coal mineral matter does not contain more than five or six components; clay minerals, pyrite, quartz and one or two calcium, iron and magnesium carbonates,57 this is reflected in the fly ash produced. Moreover, considering the mineral distribution in the coal and the burning conditions in the boilers, particles are probably produced by coal fragments containing only one or two mineral species.57 Microscopic examination and electron probe microanalysis confirm these effects. Indeed, besides the prevailing vitreous ground mass, only four compounds are present to any appreciable extent: quartz, mullite, hematite and magnetite.57–59 TABLE 9.5

Examples of Developments in Electricity Generation/Fly Ash Production

Development

Process/Operation

Low NOx technology

Changes in combustion conditions (air/fuel, temperature) Injection of additives, with or without a catalyst, following combustion (e.g. SCR/SNCR)a Coal burnt with other material (e.g. wood-based, meat and bone meal, etc.) Replacement of combustion chamber air with oxygen Increase in pressure/temperature of steam used in the generation process Removal of carbon and concentration of fine particles from stored material

Co-combustion Oxy-fuel combustion Supercritical technology Processing of wet-stored fly ash a

SCR, selective catalytic reduction; SNCR, selective non-catalytic reduction.

370 Lea’s Chemistry of Cement and Concrete

Magnetite may possibly be a mixture of ferrites.60 Although the overall chemical composition of fly ash does not vary appreciably, microprobe analysis indicates chemical heterogeneity of the particles.61 Fly ash particles are typically spherical and glassy, but they also exhibit other morphologies, as reflected in an early fly ash particle classification, based on various criteria including, colour, crystallinity and texture.57 The finest glassy particles are generally thick but many are hollow. In some cases, the largest ones look like empty spheres filled with smaller, spherical particles. Other types of particles are irregular and may contain variable quantities of bubbles, which makes them take on a spongy appearance, or a high content of crystalline minerals. The particle diameters of fly ash range from <1 to >150 mm. Specific surface area can be variable, typically ranging from <0.2 to >0.8 m2/g (Blaine).57 It has been noted that while the mineral part of fly ash has a relatively uniform specific surface area of around 0.7–0.8 m2/g (BET nitrogen adsorption),62 differences in the property between materials, with this type of measurement, correspond to the porosity of the carbon present. As noted above, sub-bituminous coal and lignite combustion generally results in high-calcium fly ashes (Class C).55 These fly ashes are typically low in alkalis, but sometimes they may show high contents. In some cases, increased sulfate levels have also been noted. The crystalline phases occurring in low lime fly ashes are few, whereas many phases can be found in high lime fly ashes as a consequence of a more variable chemical composition. Some of these also exhibit hydraulic properties. Analysis by XRD has shown the following crystalline phases in lignite fly ashes: quartz, lime (free), periclase, anhydrite, ferrite spinel, merwinite, alkali sulfates, melilite, mullite, sodalite and hematite.63 The presence of tricalcium aluminate (C3A) has been detected by XRD analysis64 but in some cases C3A and bredigite (C2S) were only assumed to be present, as their occurrence was not clear, with the presence of other compounds.65 Differences between low- and high-lime fly ashes relate to the chemical and mineral composition and structure of the glass. These differences are highlighted by changes occurring in the XRD background generated by the glass. For fly ashes containing up to 20% CaO, a statistical relationship exists between the CaO content and the position of the X-ray maxima in the background (Fig. 9.6).66 For fly ashes with about 20%–27% CaO, the relationship is no longer valid and invariably maxima appear at 32.0–32.5°2y (Cu radiation).66 The position of maxima of low-calcium fly ashes is indicative of a siliceous glass structure, whereas that of high-calcium fly ashes (CaO > 20%) is typical of a calcium aluminate glass structure.66 IR, NMR and TMS (trimethylsilylation) investigations suggest that the degree of condensation of silicate ions increases as the lime content in the glass decreases. Thus the monomer + dimer content is as high as 7%–8% in 27.3% CaO fly ash, but it is <1% in 2.8% CaO fly ash. Silica fume and rice husk ash contain only polymers.67

9.2.2.2 Burned Clay and Shale Clay minerals gain a distinct pozzolanic activity when burned at temperatures between 600°C and 900°C. Owing to the chemical composition of clay and related materials, these artificial pozzolanic materials are mostly composed of silica and alumina. The thermal treatment adopted causes the crystalline network of the clay minerals to be destroyed, while silica and alumina remain in a disordered, unstable, amorphous state. Heating does not affect anhydrous minerals such as quartz or plagioclase, so that pozzolanic activity depends only on the clay mineral content as well as on the thermal treatment conditions.5,68–70

FIG. 9.6 Positions of glass X-ray maxima versus analytical CaO contents for 16 fly ashes. (From: Diamond S. On the glass present in low-calcium and in high-calcium fly ashes. Cem Concr Res 1983;13:459–64.)

Pozzolanas and Pozzolanic Materials

371

An example of a material in this grouping is metakaolin, whose use as an addition in concrete has developed since the 1990s.71 This is produced through the thermal treatment of kaolin, with its properties dependent on the feed (clay with at least 85%–90% kaolin is appropriate for improved concrete performance)72 usually at temperatures between 650°C and 800°C.73 The reactivity of metakaolin has been found to depend on the crystallinity of the original kaolinite (with well-ordered less reactive)74 and degree of dehydroxylation achieved (>95% for highest reactivity).75 It is normally milled to a fine powder during manufacture, with a specific surface area by BET nitrogen adsorption in the range 10–25 m2/g,76 which has also been noted to influence its performance in cementitious systems.77 The burning or retorting of certain oil shales produce ashes which harden when mixed with water. Their chemical composition varies largely according to their origin. The silica content can range between 22%78 and 42%79 and conversely lime can vary between 55%78 and 22%.79 Hardening results from the presence of cementitious compounds such as C2S, CA and CS.78–80 Burned shales have a more complicated mineralogical composition than burned clays depending on their composition, temperature and duration of burning. As an example, shale burned at temperatures ranging between 750°C and 840°C contains b-quartz, b-cristobalite, calcite, a-Fe2O3 and muscovite, which are already present in shale, and gehlenite, anorthite, wollastonite, orthoclase, anhydrite, b-C2S, CA and CaO formed during the burning process.79 Oil shale ashes should possess pozzolanic properties as they consume lime when they are mixed with water and hydrated lime or PC.80

9.2.2.3 Silica Fume The manufacturing processes of silicon metal and ferrosilicon alloys in an electric arc furnace occur at temperatures up to 2000°C. They generate fumes containing spherical microparticles of amorphous silicon dioxide. This is the reason why the product is called ‘silica fume’ or, owing to its form and chemical composition, ‘microsilica’, ‘condensed silica fume’ and ‘volatilised silica’.81 It should be noted that the term ‘microsilica’ can also refer to naturally formed particles of silica and fine particulate materials formed by other chemical processes (such as the hydrolysis of silicon tetrachloride (SiCl4)). The reduction of quartz to silicon releases gaseous SiO. This is transported by combustion gases to lower temperature zones where it is oxidised by air and condenses in the form of tiny particles of silicon dioxide. The main features of silica fume are a high silica content, high specific surface area and amorphous structure. These characteristics mean the material is strongly pozzolanic both in terms of its capacity for binding lime and rate of reaction. The chemical composition of silica fume varies with the origin of the material and lies in the ranges shown in Table 9.6.82 The silicon metal process gives purer products, whereas the production of silicon alloy results in more complex compositions, with the minor element content being as high as 30%.83 It may also contain traces of quartz.67 Low-lime silica fume shows a high degree of condensation of silicate ions since it is formed only by polymeric species.67 Silica fume particles are spherical and have an average diameter of 0.1 mm. The BET specific surface area ranges from 15 to 35 m2/g, which influence how it is supplied for concrete production (often as densified powder or slurry).84 Silica fume is commonly used at levels of around 10% in cement, with superplasticising admixtures normally required to achieve appropriate fresh properties in concrete. It frequently finds application in high strength/performance applications.

TABLE 9.6 Chemical Analyses of Silica Fume From the Production of Silicon Metal and 75% Ferro-Silicon Alloy (%)82

SiO2 C Fe2O3 Al2O3 CaO MgO Na2O K2O S LOI

Si Metal

75% FeSi

94–98 0.2–1.3 0.02–0.15 0.1–0.4 0.08–0.3 0.3–0.9 0.1–0.4 0.2–0.7 0.1–0.3 0.8–1.5

86–90 0.8–2.3 0.3–1.0 0.2–0.6 0.2–0.6 1.0–3.5 0.8–1.8 1.5–3.5 0.2–0.4 2.0–4.0

372 Lea’s Chemistry of Cement and Concrete

9.2.2.4 Other Materials Ash exhibiting a marked pozzolanic character can be obtained by burning rice husk within certain temperature ranges. Rice husk ash contains >80% silica,85 with a range of relatively high surface areas reported.64,85–87 Crystalline silica, such as quartz and cristobalite, can be present in large amounts depending on the burning conditions.67 The pozzolanic activity of rice husk ash depends on the firing temperature and the retention period. It has been noted that controlled burning between 550°C and 700°C for 1 h converts silica into an amorphous phase.86

9.3 MIXTURES OF POZZOLANIC MATERIALS WITH LIME Until the 19th century, lime–pozzolanic material mixes were the only hydraulic mortars capable of hardening in water and, at the same time, of resisting the attack of aggressive solutions, including sea water. Due to their slow rate of hardening, limepozzolanic material mixes have gradually been replaced by blended cements. Since the lime-pozzolanic material–water system is simpler than the clinker-pozzolanic material-water system, its examination helps to provide an understanding of the behaviour and properties of cements containing pozzolanic materials.

9.3.1 Pozzolanic Reaction The term ‘pozzolanic activity’ covers all reactions occurring among the active constituents of pozzolanic materials, lime and water. This definition, although approximate, is nonetheless acceptable from a technical and practical viewpoint. As a result of the difficulty in following the quantities of a pozzolanic material’s active phases throughout the hydration process, the progress of pozzolanic reaction is commonly evaluated in terms of the reduction of free lime in the system or increase in soluble silica and alumina in acid88 using the Feret–Florentino method. The term ‘pozzolanic activity’ includes two parameters, namely the maximum amount of lime with which a pozzolanic material can combine and the rate at which the process of combination occurs. Both factors depend on the nature of the pozzolanic material and, more specifically, on the quality and quantity of the active phases. The heterogeneity of the family of pozzolanic materials, as well as the complex phenomena occurring during hydration, do not allow a model of pozzolanic ‘activity’ to be defined, and only general effects can be identified. Fig. 9.7 shows that, should water be in excess, the amount of combined lime may vary appreciably according to the type of pozzolanic material.11 After 180 days of reaction, pozzolanic materials are able to combine with 45%–75% of lime with respect to their mass. In pastes, lime combination is lower, since the development of hydration products hinders the attainment of equilibrium conditions.89 There is general agreement that the overall amount of combined lime depends on 1. 2. 3. 4. 5.

the nature of the active phases; their content in the pozzolanic material; their SiO2 content; the lime/pozzolanic material ratio of the mix; the length of curing;

whereas the combination rate depends on 6. the specific surface area of the pozzolanic material; 7. the water/solids mix ratio; 8. temperature. These influences are briefly reviewed below: 1. Within the zeolite family, herschelite and clinoptilotite are considered to be more active than analcime.90,91 Zeolitic pozzolanas are regarded as more reactive than glassy ones.29 Glass contained in Bavarian trass combines with less lime than that of Rhine trass.10 The glasses of various pozzolanas have different capacities for combining with lime. As an example, glasses of Rhine and Bavarian trasses, as well as those of an obsidian, combine, respectively, 0.364, 0.272 and 0.176 CaO g/g of pozzolana. Table 9.7 shows that the phases present in volcanic pozzolanas bind with lime to different extents.92 2. It is evident that, other properties being equal, the larger the amount of combined lime, the higher the content of active phases in the pozzolanic material, and the lower the content of inert or crystalline phases (quartz, sanidine, mullite, magnetite, etc.).

Pozzolanas and Pozzolanic Materials

373

FIG. 9.7 Combined calcium hydroxide versus specific surface area of pozzolanic material. Lime/pozzolanic material ratio ¼ 80:100, water/binder ratio ¼ 2:1. (From: Costa U, Massazza F. Factors affecting the reaction with lime of Italian pozzolanas. In: Proceedings of the sixth international congress on the chemistry of cement, Moscow; Sep. 1974, Supplementary paper, Section III. p. 2–18.)

TABLE 9.7

Lime-Binding Capabilities of the Principal Trass Minerals and Their Contributions to Trass-Lime Binding92 Free Alkali

Mineral Component Rhenish trass Quartz Feldspar Leucite Analcime Kaolin Glass phase Total Bavarian trass (glass phase only) Obsidian glass

Lime Reaction (mg CaO/g)

Na2O (mg/g)

K2O (mg/g)

Average Amount in Trass (%)

Calculated Lime Reaction (mg CaO/g Trass)

43 117 90 190 34 364 — 272

1.5 1.1 1.3 10.7 0.3 18.0 — 6.0

0.4 0.2 1.8 3.0 2.1 24.0 — 6.0

13 15 6 7 2 55 98 66

5.5 17.5 5.4 13.3 0.7 200.0 242.5 179.0

176

3.7

3.1

—

—

3. The amount of combined lime is related to the SiO2 content in the active phases, which ranges between 45% and 75% in volcanic glass89 and in fly ash,55 but reaches and sometimes exceeds 95% in very active amorphous microsilica, such as natural silica gels89 or silica fume.82 Fig. 9.7 illustrates this: pozzolana 6, containing around 85% silica, combines with most of the available lime within 28 days, whereas other natural pozzolanas and fly ashes, with about 50%–60% silica, combine with 31%–51% of the lime contained in the mix.11 However, the glass of Rhine trass combines with more lime than Bavarian trass, in spite of the fact that their silica contents are about 55% and 67%, respectively.92 Similarly, the

374 Lea’s Chemistry of Cement and Concrete

FIG. 9.8 Ca(OH)2 reacted with pozzolanic materials F, V and R estimated by X-ray diffraction analysis varying the mixing ratio and curing temperature. Water/binder ¼ 0.56. (From: Takemoto K, Uchikawa H. Hydratation des ciments pouzzolaniques. In: Proceedings of the seventh international congress on the chemistry of cement, Paris; 1980, vol. I. p. IV-2/1–21.)

4. 5.

6.

7. 8.

glass phase of fly ash varies in composition and structure, which can give a difference in reactivity.66 Thus, other chemical and structural factors also play an important role in determining pozzolanic activity. Within certain limits, the amount of combined lime increases as the lime/pozzolanic material ratio increases (Fig. 9.8).9,11,93 Combined lime also depends on the curing time (Fig. 9.9), but the rate of this process varies widely between pozzolanic materials. Fig. 9.9 shows that after 90 days of curing, the reaction of fly ashes is far from complete, whereas natural pozzolanas are essentially fully reacted by this time.94 The behaviour of sample 6, a very active natural microsilica, appears to be complete after only 28 days. Fig. 9.7 shows that the short-term activity largely depends on the specific surface area of the pozzolanic material, whilst long-term activity is related to chemical and mineralogical composition.9 The reaction rate of pozzolanic materials is reportedly proportional to the square of the specific surface area.9 The role played by the specific surface was also evident in pastes containing calcium hydroxide and two samples of microsilica (obtained by hydrolysis of SiCl4) and silica fume having specific surface areas of 200 and 20 m2/g, respectively. The calcium hydroxide was fully consumed by the finer silica within 1 day, but was still partially uncombined after 28 days when coarser silica was used.95 The larger the water content of the mix, the higher the rate of lime combination. Thus, the pozzolanic reaction is slower in a paste than in a dispersion, and may be incomplete after many years. The rate of pozzolanic reaction increases with temperature9,93,96 (Fig. 9.8). Between 50°C and 90°C, 1:3 lime/natural pozzolana mixes, compacted with 10% water under a compressive load of 130 N/mm2, react quickly, so that most lime is already fixed after 1 day of reaction.96 Above 70°C, however, combined lime tends to stop increasing or to decrease.93,96 As shown in Fig. 9.10, evidence of this inversion is also found in fly ashes at around 60°C.93 This appears to be due to changes in the composition of the hydrated phases at higher temperatures.

Pozzolanas and Pozzolanic Materials

375

FIG. 9.9 Ca(OH)2 combined with pozzolanic materials; pozzolanic material/hydrated lime ¼ 70:30; w/s ¼ 0.6. 1–6 are natural pozzolanas; M, G and V are fly ashes. (From: Costa J, Massazza F. Natural pozzolanas and fly ashes: analogies and differences. In: Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 134–44.)

FIG. 9.10 Lime reacted at various times and temperatures. CaO/fly ash ¼ 1:4; water/binder ¼ 0.5. (From: Buttler FG, Walker EJ. The rate and extent of reaction between calcium hydroxide and pulverised fuel ash. In: Proceedings of the use of PFA in concrete, Leeds, 14–16 Apr. 1982, vol. I. p. 71–81.)

In the case of silica fume—and depending on the lime/silica ratio—after 2.5 h at 55°C, combined lime is as high as 25%–55% of the added CaO and at 90°C this reaches 68%–90%. After 24 h of hydration, unreacted lime stabilises at 3%–8%, independently of both temperature and C/S ratio of the mix.97 Conversely, the amount of combined pozzolanic material, expressed in terms of acid-soluble silica, always increases with the temperature of hydration.93 The pozzolanic reaction is also influenced by other parameters. The addition of gypsum to the pozzolanic material–lime– water system improves the rate of lime combination.98,99 Some natural pozzolanas display an initial reaction rate higher than that of some siliceous fly ashes. With time, the rate in the natural materials slows down, whereas that in fly ashes accelerates (Fig. 9.9). This different behaviour can be attributed to many factors, one being the higher BET specific surface area of natural pozzolanas, which favours a higher initial rate of lime combination.94

9.3.2 Thermal Treatment of Natural Pozzolanas When heated, many pozzolanic materials undergo chemical and structural transformations which may alter, either beneficially or detrimentally, their reactivity to lime. The beneficial effects result from the loss of water in glassy or zeolitic phases and the destruction of the crystal structure in clay minerals. Detrimental effects are the result of reduction in specific surface area, devitrification and crystallisation. The outcome of thermal treatment depends on the nature of the pozzolanic material, the temperature and the duration of heating.100 The conflicting effects induced by temperature explain the apparent contradictions that sometimes occur in the same material. For example, it was reported that combined lime decreases101 or increases100 by heating a Latium pozzolana

376 Lea’s Chemistry of Cement and Concrete

FIG. 9.11 Combined lime versus specific surface area of pozzolana no. 2 dried at 110°C and then fired at the indicated temperature. Lime/pozzolana ¼ 0.8; w/s ¼ 0.6; temperature ¼ 20°C. (From: Costa U, Massazza F. Influenza del trattamento termico sulla reattivita’ con la calce di alcune pozzolane naturali. Il Cemento 1977;3:105–22.)

at 700°C. If the temperature of calcination is increased step by step, combined lime initially increases and later decreases (Fig. 9.11).100 The same figure shows that heating is followed by a decrease in specific surface area of the pozzolana. This means that for every pozzolana, the optimum thermal treatment has to be established by appropriate testing. For several natural pozzolanas the optimum temperature is about 700°C–800°C. Above this range there is a tendency for devitrification and densification and, generally, the formation of more stable phases.100 This fact is evidenced by a decrease in the amount of acid-soluble silica and alumina.101 Microstructural changes induced by calcination are evident in natural pozzolanas as changes in the refraction index. In glassy rhyolitic pozzolanas the index decreases with increasing temperature up to 550°C–650°C and then increases.5 The amount of lime combined by pozzolanas containing mainly zeolites gradually decreases with increasing temperature.5 At up to 700°C, heating does not modify the pozzolanic behaviour of natural microsilica, which essentially consists of opal,102 sometimes slightly contaminated by clay.5 In contrast, calcination improves the reactivity of diatomites containing high quantities of clay minerals.5 In this case, the pozzolanic activity of burned clays is associated with the activity of diatoms.

9.3.3 Reaction Products The reaction of mixtures of pozzolanic materials and lime produces the same compounds as those which are found upon hydration of PC, since the chemical compositions are similar. For this reason, different types of pozzolanic material produce similar aluminate and silicate hydrates. Differences are minor and, in general, affect the amount rather than the nature of the hydrated phases. Natural pozzolanas, dispersed in a saturated or almost saturated solution of lime react to form calcium silicate hydrate (C-S-H) and the hexagonal aluminate, C4AH13.69,103,104 Similar results are obtained using zeolitic compounds90,91 and natural pozzolana-lime pastes.14,15,105 With excess water, reaction with lime is accelerated by the addition of gypsum.98 Where pozzolanas are rich in alumina, the formation of poorly crystalline C-S-H containing aluminium (C-A-S-H) occurs.106 There exists some debate with regard to the extent to which aluminium ions are substituted into the structure of C-S-H and how much are present as other discrete phases within the mass of C-S-H. After some time, crystalline calcium aluminium silicate hydrates such as str€atlingite (C2ASH8) and katoite (C3AS1.5H3) may form.107 Table 9.8108 shows that hydration products from pastes potentially differ somewhat from those formed by hydration in excess water. This is probably due to the fact that excess water accelerates the attainment of the final stage of reaction. Moreover, the hydration products formed in pastes are smaller in size and more irregular.109 By extending the duration of reaction between pozzolanic materials and lime solutions, besides calcium aluminate and calcium silicate hydrates, other compounds are formed: carboaluminate, gehlenite hydrate and hydrogarnet.9,110 Hydrogarnet appears after 70–150 days when combined lime represents 40%–60% of the initial mass of pozzolana.33 Table 9.9 shows that the nature of hydrated compounds also depends on the chemical composition of the pozzolanic material. Thus, a high-opal pozzolana (Beppu White Clay) can only give C-S-H.9 When gypsum is present in pastes of pozzolanic materials and lime, ettringite also forms.111 When the content of calcium sulfate exceeds specific values, ettringite formation can cause pastes to crumble.112 C-S-H gel is a non-stoichiometric phase with a Ca/Si ratio that varies depending on the type of pozzolanic material, the duration and temperature of curing, the lime/pozzolanic material ratio, as well as the analytical method used. As an example, an opal-based pozzolana yields ratios in the range 0.75–0.87, whilst glassy natural pozzolanas produce C-S-H ratios that are

Pozzolanas and Pozzolanic Materials

377

TABLE 9.8 Hydrates Formed Between Pozzolanic Materials and Lime (a) Using Saturated Lime Solutionsa and (b) in Pastes108,b Bacoli Pozzolana

Segni Pozzolana

Dehydrated Kaolin

Neapolitan Yellow Tuff

Rhine Trass

Hydrate Phase

Name

(a)

(b)

(a)

(b)

(a)

(b)

(a)

(b)

(a)

(b)

C-S-H C2ASH8 C4AH13 C3 AcH12 C3AS2H2–C3AH6

Calcium silicate hydrate Gehlenite hydrate — Carboaluminate Hydrogarnet

+ + – – –

+ + – + –

+ + – + +

+ + – + –

+ + – – –

+ + – – –

+ + + – –

+ + – + –

+ + + – –

+ + + – –

a

Reaction time: 90 days. Water/solids ratio ¼ 0.40; pastes cured for 5 years.

b

TABLE 9.9

Hydrates Formed in Natural Pozzolana–Ca(OH)2 Mixes9

Pozzolana

Curing Temperature (°C)

Age (Days)

C-S-H

C3AH6–C3AS2H2

C3 AcH12 C4 AH13

C2ASH8

Furue shirasu (F)

20, 40, 60 20, 40, 60 20 20 40 40 60 60 20 20 40 40 60 60 20, 40, 60 20, 40, 60 20 20 40 40 60 60 20, 40, 60 20, 40, 60

7 180 7 180 7 180 7 180 7 180 7 180 7 180 7 180 7 180 7 180 7 180 7 180

+ +++ – +++ (+) +++ + +++ – (+) (+) (+) (+) + + +++ – – – – – (+) + ++

– (+) – – – – – – (+) – +++ +++++ ++++ +++++ – – (+) – (+) ++ + +++ – –

+ +++ (+) + (+) (+) – – +++ +++ + (+) (+) – – – +++ ++ + + + (+) – +

– (+) – – – – – – + +++ (+) (+) – – – – – – – – – – – (+)

Higashi Matsuyama tuff (G)

Kanto (Hachio˜ji) loam (R)

Beppu white clay (V)

Tominaga masa soil (M)

Takehara fly ash (T)

Number of ‘+’ signs denotes relative quantity of each phase present, estimated from XRD traces.

substantially higher—between 1.35 and 1.75.9 In other cases, electron microprobe analysis has given values ranging between 0.75 (opal-based) and 0.85 (glass-based), whilst values calculated by chemical analysis were between 1.2 and 1.7.89 The formation and development of C-S-H during the pozzolanic reaction is marked by an increase in the extent to which silicate ions become polymerised. Initially, a large number of isolated silicate ions (monomers) are present. As the reaction progresses the monomer content fluctuates around a certain level, the dimer (silicate ion pair) content increases up to a maximum value and then decreases, while the content of polymer species (chains with more than two silicate ions) increases.113 At temperatures between 50°C and 90°C, the main product of the pozzolanic reaction is amorphous C-S-H, similar to that obtained upon hydration of PC, and poorly crystallised tobermorite.96 SO4 2 contained in siliceous fly ashes dissolves in lime water and after a certain period causes ettringite and gypsum to precipitate. The rate of ettringite formation depends on the rate of dissolution of alumina. If ashes are washed with water, the two compounds do not form and this means that sulfate occurs in a soluble form.114

378 Lea’s Chemistry of Cement and Concrete

Siliceous fly ashes mixed with lime and water form C-S-H, C4AH13 and C2SAH8,9 and sometimes carboaluminate (C4 AcH11 ) as well.115 If fly ashes contain sulfates, ettringite also appears.12,113 C4AH13 decreases with time, whereas the C4 AcH11 content increases.115 Calcareous fly ashes may contain variable amounts of free lime which, upon mixing, are transformed into Ca(OH)2 and can, thus, harden without any further addition of lime.116 From a practical viewpoint, these fly ashes correspond to the artificial hydraulic limes obtained by mixing hydrated lime with a pozzolanic material.117 If fly ash releases lime and sulfate, the formation of ettringite occurs without external additions.65 Where sulfate levels are high, this may lead to swelling.118 Depending on the chemical composition of the fly ashes and the burning temperature of coal, calcareous fly ashes may also contain C2S.115,118 Where this is the case, the ash may be capable of developing strength in the absence of lime through hydraulic reactions. In any case, C-S-H, tetracalcium aluminate hydrate, carboaluminate, gehlenite hydrate and ettringite are formed.119 However, if lime is entirely or mainly combined with Al2O3 and SiO2, the glass is either not or only slightly reactive. As a consequence, no prominent pozzolanic reaction is observed and, for at least 3 months, no C-S-H is formed.115 Little data concerning the reaction occurring between silica fume and Ca(OH)2 is available. The reaction with Ca(OH)2 solutions is very rapid and causes a phase to precipitate on the silicon dioxide particles as a high-silica hydrated layer. This layer is unstable and turns into C-S-H very quickly.120 In pastes of normal consistency, owing to the high reactivity of silica fume, free lime disappears generally between 7 and 28 days112 and in some cases even sooner.57 The reaction product (C-S-H) is more crystalline than the calcium silicate hydrate found in PC paste.121 In 1:1 mixes of Ca(OH)2 and silica fume, the C/S ratio of C-S-H increases in the first days up to 1.30 and then decreases, reaching 1.10 after 70 days of reaction.57 Other types of pozzolanic material can also give the reaction products mentioned above. The products from the reaction of burned kaolin (metakaolin) with lime are mainly calcium silicate hydrate (C-S-H), gehlenite hydrate (C2ASH8), and small quantities of tetracalcium aluminate hydrate (C4AH3).69,122,123 At higher temperatures and suitable lime concentrations, the tetracalcium aluminate hydrate turns into C3AH6,69 but traces of the cubic aluminate are also observed after reaction at normal temperature.122 The presence of both the tetracalcium aluminate hydrate and gehlenite hydrate is considered to be in contrast with the phase relations existing in the CaO–Al2O3–SiO2–H2O system.124 This coexistence might, however, result from some barrier to attaining final equilibrium. This view is supported by the transformation of gehlenite hydrate into hydrogarnet when the former is shaken with a calcium hydroxide saturated solution.125

9.3.4 Porosity and Microstructure The specific surface area of lime–natural pozzolana pastes increases gradually with time up to 35–100 m2/g after 90 days of curing.89 In bottle hydration (water/binder ¼ 12.5), the specific surface area of the hydrates is higher and can be greater than 150 m2/g after 1 year of curing, depending on the type of material and lime/pozzolanic material ratio.126 For a lime/pozzolana ratio up to at least 0.6, combined lime progressively increases. The specific surface area initially increases too, but, once beyond a peak value, it often decreases dramatically. The peaks of two Italian pozzolanas and two fly ashes have been found to occur at a lime/pozzolanic material ratio of 0.45,126 although the peaks of other materials may differ. Whatever the type of pozzolanic material, a direct relationship exists between the porosity and the specific surface area of the paste when mixes of pozzolanic material and lime are hydrated in a sealed bottle.126 Between 3 and 90 days of curing, paste porosity does not change much, but some differences are still found among different pozzolanic materials.89 For pozzolanic material/lime ratios between 2:1 and 1:2, the porosity of the paste increases with lime content. Experiments carried out with siliceous fly ashes have given the same result. A difference has, however, been found between a typical natural pozzolana and fly ash: natural pozzolana-lime mixes show a well-defined discontinuity in pore radii between 1.5 and 2.0 nm, whereas fly ash–lime mixes show only a slight discontinuity.126

9.3.5 Strength of Mixes of Pozzolanic Materials and Lime A practical consequence of the pozzolanic reaction is the gradual hardening of pastes containing pozzolanic materials and lime. Strength increases as the amount of combined lime increases.9,89 However, as shown in Fig. 9.12, there is no general relationship between the two parameters, although there is a correlation for a given type of pozzolanic material. The lack of correlation is also found when combined lime is compared with strength of PC–pozzolanic material blends.94 This is particularly true of materials containing unaltered clays, which often possess a high capacity to fix lime, but do not harden appreciably.2 For this reason the technical assessment of pozzolanic materials requires the measurement of strength development in the presence of lime or PC, rather than determination of the amount of fixed lime present.

Pozzolanas and Pozzolanic Materials

379

FIG. 9.12 Compressive strength versus combined calcium hydroxide. Pozzolana/calcium hydroxide ¼ 100:40; w/s ¼ 0.6. Samples stored in water at 40°C and cured for 3, 7, 28 and 90 days. (From: Massazza F, Costa U. Factors determining the development of mechanical strength in lime-pozzolana pastes. In: Proceedings of the XXII conference on silicate industry and silicate science, Budapest; 6–11 Jun. 1977, vol. I. p. 537–52.)

FIG. 9.13 Influence of lime content on the compressive strength of hydrated lime–pozzolana mixes. Water/binder ¼ 0.08. (From: Fournier M, Geoffray J-M. Le liant pouzzolanes-chaux. Bulletin de Liaison des Laboratoires des Ponts et Chaussees 1978;93:70–8.)

The strength of lime–natural pozzolana pastes initially increases with the lime/pozzolana ratio, but eventually decreases127,128 (Fig. 9.13). Hardening of a pozzolana–lime mortar is slow but continues to progress over long periods: compressive strength at an age of 2 years can be as high as three times the 28-day strength.128 The addition of gypsum has been found to have no effect on this behavior,99 and also to increase strength111 (Table 9.10). However, excessive quantities may lead to the formation of large amounts of ettringite123 which can cause disintegration.99,112

380 Lea’s Chemistry of Cement and Concrete

TABLE 9.10 Compressive Strength (kg/cm2) of Two Lime-Segni Pozzolana Mixes of Different Gypsum Content111 CaSO4 (%) Curing time Ca(OH)2 5 15% 7 days 14 days 28 days 3 months 6 months 1 year 2 years Ca(OH)2 5 25% 7 days 14 days 28 days 3 months 6 months 1 year 2 years

0

2.5

5

7.5

10

16 24 51 87 122 135 150

31 47 85 120 138 156 165

28 66 108 170 175 180 187

56 102 120 163 170 181 186

65 118 140 160 180 163 148

8 20 49 110 168 190 212

22 35 55 105 175 208 230

30 58 125 198 233 266 280

28 62 144 208 239 268 275

51 68 150 181 215 202 187

A good linear correlation has been found between the Blaine fineness of a natural pozzolana and the compressive strength of 20:80 lime–pozzolana mixes. The benefits of higher fineness are more pronounced at early ages.129 A similar relationship has been observed in mixes made from fourteen different fly ashes and lime.130 Paste strength typically attains the highest level at a 10%–15% lime content and increases with fineness of fly ash.116 The addition of gypsum is, again, effective in improving strength.131 The strength of lime–natural pozzolana mixes is improved by adding KOH or NaOH, as well as Na2SO4 or CaCl2.129,132 The 28-day paste strength can be as high as 30 MPa and rapid setting can occur.118,133 High-lime fly ashes can behave like hydraulic binders. In fact, if they have a suitable composition, they need neither lime nor cement to harden.118,133,134,135 After 28 days the compressive strength of concrete made with plain high-lime fly ash can be as high as 15–25 MPa.134 The addition of gypsum (typically around 6% by mass) improves compressive strength development.118 Metakaolin,123 as well as other clay materials,136 harden gradually when they are mixed with lime and water. The strength attained after 28 days’ curing strongly depends on the burned clay/lime ratio and water/solids ratio. The strength of metakaolin and burned clay shows a distinct peak for mix ratios ranging between two and three. The duration and temperature of the burning process affect the strength, making it necessary to select the correct thermal treatment in order to achieve optimal results.68,127,137 Strength development also depends on the nature of the clay minerals. Highest strengths are typically obtained with burned kaolin, with burned illite producing very low strengths.136 Mechanical properties of calcined clays can be improved by incorporating admixtures prior to calcination, such as ZnO.138 The presence of Zn has the effect of retarding early hydration, but produces higher compressive strengths at ages of 28 and 90 days.139 The mechanism leading to enhanced strength is currently unclear. The compressive strength achieved by rice husk ash/lime mixes depends on the ash:lime ratio, with optimum performance typically observed between ratios of 1:1 and 1.5:1. Unlike other types of pozzolanic material, rice husk ash–lime mixtures develop little strength beyond 28 days.85 Strength development is dependent on firing temperature and duration (Fig. 9.14).140 Thus, a material with good and consistent pozzolanic properties can be obtained only by burning rice husk under closely controlled conditions. The practical difficulty in assuring such conditions is the primary obstacle to the production and use of rice husk ash on a large scale.140 Strength values between 10 and 14 MPa have been reached after curing mortar made of lime and diatomaceous earth for 28 days. Due to their high specific surface area, diatomite pozzolanas have a high water demand.141 Hardening of burned shales occurs without the presence of lime. Nevertheless, the addition of up to 10% Ca(OH)2 by mass increases strength in proportion to the lime content.79 Hydraulic oil shale ashes give the best performance when they are burned at a certain temperature, with highest strengths typically obtained at temperatures of between 800°C and 850°C.79

Pozzolanas and Pozzolanic Materials

381

FIG. 9.14 Lime reactivity of rice husk ash produced at different temperatures and firing times, according to Standard IS 1727. (From: Dass A. Pozzolanic behaviour of rice husk ash. CIB Build Res Practice 1984:307–11.)

9.4 CEMENT CONTAINING POZZOLANIC MATERIALS 9.4.1 Cement Types When mixed with PC and water, pozzolanic materials react with the calcium hydroxide (portlandite) formed during hydration of the clinker. As a result of this reaction, the final portlandite content in the hydration products is always lower than that found in the control PC. The residual portlandite content in a hardened paste depends on the activity of the pozzolanic material, the amount of lime released by the hydration of clinker, as well as the pozzolanic material/cement ratio. The simultaneous presence of PC and pozzolanic material modifies the respective reactions of hydration. This mutual influence needs to be thoroughly characterised to determine the conditions which allow the best technical performance to be obtained. Cements are classified by the European Committee for Standardisation (CEN), in the standard EN 197-1, as shown in Table 9.11.142 (Note: cements containing non-pozzolanic materials—CEM I (Portland cement) and CEM III (blastfurnace cement) are not given in the table). Materials included in this standard which are deemed appropriate for combination with Portland clinker to manufacture cement are blastfurnace slag, silica fume, natural pozzolanas, fly ash, burnt shale and limestone. All are pozzolanic, with the exception of blastfurnace slag and limestone. ‘Pozzolanas’ in the context of the standard are naturally occurring materials with pozzolanic characteristics (natural pozzolanas) or natural materials rendered pozzolanic by thermal treatment—‘natural calcined pozzolanas’. Fly ash is subdivided into siliceous and calcareous varieties, with the siliceous ash containing less than 10% reactive calcium oxide. EN 197-1 contains three cement classifications that may contain pozzolanic materials—CEM II, CEM IV and CEM V. CEM II cements can contain Portland clinker in combination with one individual pozzolanic material, or can take the form of a composite cement containing clinker, plus a combination of any of the other materials, from 12% up to a total of 35% by mass. CEM IV cements (pozzolanic cements) can contain any combination of pozzolanic materials, with the exception of burnt shale, from 11% up to 55% by mass. CEM V cements— ‘composite cements’—can contain combinations of pozzolanas, siliceous fly ash or blastfurnace slag between 18% and 49% by mass. All cements can contain up to 5% of a minor additional constituent, which can be any appropriate inorganic material. This is often limestone or fly ash. Not shown in Table 9.11 is gypsum that is included to control the setting time. This is added at levels established as appropriate by the manufacturer. Up to 1% by mass of other additives can be included to enhance the manufacturing process or cement performance.

Cement Types Containing Pozzolanic Material According to EN 197-1a,142 Constituent

Main Cement Type CEM II

Pozzolana

Portland fly ash cement

CEM IV CEM V a

Portlandburnt shale cement Portland composite cementc Pozzolanic cementc Composite cementc

Limestone

Clinker K

Blastfurnace Slag S

Silica Fume Db

Natural P

Natural Calcined Q

Siliceous V

Calcareous W

Burnt Shale T

L

LL

0–5

II/A-D

90–94

—

6–10

—

—

—

—

—

—

—

0–5

II/A-P II/B-P II/A-Q II/B-Q II/A-V II/B-V II/A-W II/B-W II/A-T II/B-T

80–94 65–79 80–94 65–79 80–94 65–79 80–94 65–79 80–94 65–79

— — — — — — — — — —

— — — — — — — — — —

6–20 21–35 — — — — — — — —

— — 6–20 21–35 — — — — — —

— — — — 6–20 21–35 — — — —

— — — — — — 6–20 21–35 — —

— — — — — — — — 6–20 21–35

— — — — — — — — — —

— — — — — — — — — —

0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5 0–5

II/A-M II/B-M

80–88 65–79









































12–20








































!








































21–35








































!

0–5 0–5

IV/A IV/B V/A V/B

65–89 45–64 40–64 20–38

Notation Portland silica fume cement Portland pozzolana cement

Fly Ashes

Minor Additional Constituents

— — 18–30 31–49

























11–35























!























36–55























! — —











18–30











! — —











31–49











!

— — — —

— — — —

0–5 0–5 0–5 0–5

The values of the table refer to the sum of the main and minor additional constituents. The proportion of silica fume is limited to 10%. c In portland-composite cement CEM II/A-M and CEM II/B-M, in pozzolanic cements CEM IV/A and CEM IV/B and in Composite cements CEM V/A and CEM V/B the nature and kind of the constituents besides clinker shall be declared by designation of cement. b

382 Lea’s Chemistry of Cement and Concrete

TABLE 9.11

Pozzolanas and Pozzolanic Materials

383

In principle, pozzolanic materials can be added to PC either at the cement plant or at the construction site. In the first case, the pozzolanic material undergoes either simultaneous grinding with clinker and gypsum or separate grinding, followed by mixing and homogenisation with suitable equipment. In the latter case, pozzolanic material is introduced with PC into the concrete mixer.

9.4.2 Hydration of Clinker Phases With Pozzolanic Materials The presence of pozzolanic material modifies to some extent the reactions of hydration of PC. Any chemical and microstructural modifications can potentially affect the engineering properties of the hardened material.

9.4.2.1 Kinetics of Hydration The kinetics of early hydration can be investigated by isothermal conduction calorimetry, whereas the progress of hydration at advanced ages is best monitored through XRD, differential thermal analysis (DTA), differential scanning calorimetry (DSC) and optical and electron microscopy. For determining the portlandite content in the paste, chemical methods such as the extraction of free lime with appropriate solvents may be used. Knowledge of the systems formed by pozzolanic materials and clinker compounds makes it easier to study more complex systems containing PC. Clinker compounds have the advantage of being pure and can be tested one at a time, thus limiting interference arising from simultaneous reactions such as those occurring when PC hydrates. The findings of studies utilising pure cement phases are summarised below. Tricalcium Aluminate (C3A) Studies of the influence of pozzolanic materials on the hydration of the anhydrous aluminate phases have mainly focused on C3A, largely because this phase reacts approximately six times more rapidly than C4AF.143 The combination of natural pozzolana with C3A changes the initial rate of heat evolution during hydration, since it causes the intensity of the second peak to decrease. The change in magnitude of this peak is seen as corresponding to a reduction in the rate of hydration of the aluminate.144,145 The decrease of the second peak on the heat evolution curve is influenced by the type of pozzolanic material employed, and this appears to be related to the material’s specific surface area. For example, the delay brought about by Sacrofano pozzolana (BET specific surface area of around 60 m2/g) is greater than that of Segni pozzolana (BET specific surface area of around 30 m2/g).144 The delay may be due to other causes, such as the dissolution of alkalis and differences in surface activity,143 but these factors were not investigated. The retarding effect of fly ashes is quite similar to that displayed by natural pozzolanas. The addition of fly ash to C3A (mass ratio 30:70) suppresses hydration during the first few minutes and then lowers the rate constants by approximately seven times.143 The addition of an inert material such as ground quartz also reduces the hydration rate and delays the beginning of C3A hydration, although its effect is less marked than that caused by fly ash.143 Fly ash reduces the hydration rate of both C3A produced in the laboratory and that extracted from ground cement clinker by selective dissolution, the former hydrating more quickly than the latter.143 The reasons for the delay have not been resolved. Gypsum, or gypsum and lime, decrease the rate of C3A hydration, but the addition of pozzolanic material to these systems appears to counteract this, along with the subsequent conversion of ettringite into monosulfate.144,145 The degree of hydration of C3A reaches around 90% after 7 days of curing at ambient temperatures, irrespective of the presence of pozzolanic material.145 Tricalcium Silicate (C3S) The presence of pozzolanic materials affects many aspects of C3S hydration, including the kinetics of reaction, formation of portlandite and composition of the hydrates. The determination of the heat evolution rate is a sensitive and useful tool to characterise the way in which pozzolanic materials influence the early hydration of C3S. Very fine silica fume (Aerosil, with a BET specific surface area of 207 m2/g) reduces the length of the dormant period and increases the intensity of the main peak.146 If the C3S/silica fume ratio decreases beyond a certain level (around 0.67), the dormant period and the second peak will disappear, the Ca2+ concentration in solution will increase147 and the curve of heat evolution will show only an initial peak whose height increases as the ratio decreases.147 When C3S and amorphous silica with the same specific surface area (20 m2/g) are mixed, the silica has the same effect as Aerosil.148 The slight differences observed can be attributed to the composition of the tricalcium silicate and the different C3S/silica ratios used.

384 Lea’s Chemistry of Cement and Concrete

FIG. 9.15 Heat evolution curve in 6:4 C3S–pozzolana blends. V, F, R and S are natural pozzolanas, T is fly ash. w/s ¼ 0.4. (From: Ogawa K, Uchikawa H, Takemoto K, Yasui I. The mechanism of the hydration in the system C3S-pozzolana. Cem Concr Res 1980;10(5):683–96.)

Fig. 9.15 shows that natural pozzolanas can also modify rate of heat evolution curves to an appreciable extent. On the whole, natural pozzolanas have an accelerating effect on the hydration of C3S. More precisely, the dormant period does not vary: the second peak is only slightly delayed, but its height is noticeably increased.149 Fly ash tends to lengthen the dormant period and increase the height of the second peak,149 though a slight decrease has also been reported in some cases.150 These results can be interpreted as an initial retarding effect followed by an acceleration. However, since the overall heat of hydration evolved in the first 48 h is higher for blends containing fly ash than for plain C3S, it can be concluded that, overall, fly ash has an accelerating effect on the hydration of tricalcium silicate. A reduction in the dormant period has been observed in blends with amorphous microsilica having a BET specific surface area in the range 50–380 m2/g,147 but not when the material has a lower specific surface area (19 m2/g).151 The height of the second peak initially increases as fineness increases, but then decreases, as shown in Fig. 9.16.148 The length of the C3S dormant period shortens with increasing amounts of silica fume.57

FIG. 9.16 Calorimetric curves from the hydration of C3S with 20% by weight of amorphous silica of various surface areas (w/C3S ¼ 1). (From: Beedle SS, Groves GW, Rodger SA. The effect of fine pozzolanic and other particles on hydration of C3S. Adv Cem Res 1989;2(5):3–8.)

Pozzolanas and Pozzolanic Materials

385

The acceleration of the early hydration of C3S caused by pozzolanic materials has been related to the fineness of the particles, which provides C-S-H with a large surface for precipitation from the pore solution.55,149 This is most likely to be due to the rapid adsorption of Ca2+ ions onto the high-silica surface of pozzolanic materials. Whilst there is much debate regarding the precise mechanism leading to the dormant period, it is normally attributed to the formation of a protective high-lime calcium silicate hydrate layer on C3S surfaces. This layer is unstable and gradually dissolves to provide material for the formation of stable C-S-H gel. The layer’s dissolution leads to the end of the dormant period, since the C3S is again exposed to water. The removal of calcium and silicate ions from the solution would result in a reduction in the thickness of the layer.151 The partial removal of Ca2+ from the vicinity of C3S grains may also accelerate the transformation of the layer to C-S-H.146 This view is supported by the following arguments:  acceleration does not result from the pozzolanic reaction, as this becomes evident in pastes only after some days;  in spite of the acceleration of C3S hydrolysis, the Ca2+ concentration in the pore solution does not significantly change in the presence of pozzolanic material149;  the formation of Ca(OH)2 and C-S-H on the pozzolana particles occurs through a dissolution and precipitation mechanism149 which requires Ca2+ to move from the C3S to the pozzolanic material’s surface;  unlike pozzolanic materials, combining PC with ground silica sand,146 graphite or titanium148 produces either no, or only minor changes to the heat evolution curve. Amorphous microsilica accelerates the hydration of C3S more than natural pozzolanas since, depending on the C3S/silica ratio, this causes an earlier decrease in Ca2+ concentrations146,147 in the mix water. The decrease occurs in water dispersions,146,147 as well as in pastes.146 As shown in Table 9.12, the higher pozzolanic activity of silica fume is in agreement with its calcium adsorption capacity (CAC), which is higher than that of fly ashes. CAC is the difference in the Ca2+ concentration measured before and after dispersing a pozzolanic material in a lime-saturated solution for 3 h.152 The transfer of calcium ions from the C3S grains to the silica fume particles could also be accelerated by an early pozzolanic reaction, resulting from the high specific surface area and high silica content of the silica fume. This view is supported by the roughness which appears on particles of silica fume after only 1 h of hydration.148 Adsorption of calcium ions onto the surface of silica fume grains before the beginning of the pozzolanic reaction could also explain why the initial rate of crystallisation of portlandite is greater than that of calcium hydroxide consumption with silica fume.57 In this case Ca2+ ions adsorbed onto the silica fume could act as nuclei of crystallisation for portlandite. The delay to the early hydration of C3S caused by fly ashes has been tentatively attributed to the release of aluminate into solution from the fly ash.150 In fact, when the solution contains aluminate ions, the transformation of the initially formed protective layer (Ca/Si ¼ 3) into C-S-H (Ca/Si ¼ 0.8–1.5) is delayed.153 However, aluminate ions are not present in the pore solution of fly ash–C3S mixes, except when NaOH solution is used as mix water.150 Another theory proposes that the readily soluble fraction of fly ash (1%–5%) interferes with the hydration of C3S. In fact, lime mortars prepared from ashes washed with water have reportedly given strengths two to three times greater than those of mortars made with raw ashes.57 However, these results were attributed to the increase in specific surface area caused by

TABLE 9.12 Calcium Adsorption Capacity (CAC) and Zeta Potential of Cementitious Materials Dispersed in Ca(OH)2-Saturated Solution152 Specific Surface Area Cementitious Material

Content (%)

BET (m2/g)

Blaine (m2/g)

Calcium Adsorption Capacity (mmol Ca/g)

Zeta Potential (mV, Water Suspension)

Coarse slag Fine slag Ordinary fly ash Ordinary fly ash ground High calcium fly ash Rice husk ash Silica fume Aerosil

20 20 20 20 20 20 20 5

2.67 2.86 0.44 1.91 0.57 0.75 20.91 167

0.40 0.59 0.38 0.84 0.37 — — —

0.055 0.053 0.026 0.005 0.005 0.027 0.203 0.326

18 9 8 14 10 20 33 —

386 Lea’s Chemistry of Cement and Concrete

FIG. 9.17 Concentration of (A) Ca2+ and (B) Si4+ in water. 6:4 blends of C3S and natural pozzolanas (V, F, R) fly ash (T) and blastfurnace slag (S). Samples dispersed in water and then shaken for fixed times. (From: Ogawa K, Uchikawa H, Takemoto K, Yasui I. The mechanism of the hydration in the system C3S-pozzolana. Cem Concr Res 1980;10(5):683–96.)

breaking up agglomerates of fine particles in the ash. The prolonged dormant period observed in fly ash mixes could also be attributed to the presence of organic compounds, which may act as retarders. The evolution of the composition of water in suspensions (solids/water ratio 1:10) containing only C3S and with mixtures of C3S and pozzolanic material is similar in both cases (Fig. 9.17).149 The Ca2+ concentration first increases up to a certain level of supersaturation and then decreases with time. Natural pozzolanas and fly ashes behave similarly. Fig. 9.16 shows that the initial concentration of Si4+ in the liquid phase depends on the type of pozzolanic material, but after 3 h no differences can be detected.149 In the presence of silica fume the Ca2+ concentration first increases and subsequently decreases, but it always remains lower than that observed in plain C3S. If the C3S/silica ratio drops to 0.4, the solution never attains saturation with respect to Ca2+ (Fig. 9.18).147 Other criteria have also been used for evaluating the medium and long-term influence of pozzolanic materials on the hydration of C3S. These include the determination of the integral heat of hydration, the diminution with time of unreacted C3S, as well as the uncombined pozzolanic material content. XRD analysis reveals that after only 1 day, natural pozzolanas,149 fly ashes149,154 and silica fume152 cause a reduction in unreacted C3S content. The degree of hydration of tricalcium silicate strongly depends on the type of pozzolanic material initially149, but any difference becomes negligible after around 91 days (Fig. 9.19).149,154 The pozzolanic reaction starts slowly. Thus, the decrease in unreacted pozzolanic material becomes apparent only after 3–5 days.149,154 The degree of pozzolanic reaction (Fig. 9.20)154 reaches an ultimate level which depends, other conditions being equal, on the type of pozzolanic material.149

9.4.2.2 Structure and Composition of Hydrates C3A Pozzolanic materials do not change the types of hydrates that form from the hydration of C3A. However, if the mixes contain calcium hydroxide, C-S-H forms as a consequence of the pozzolanic reaction.

FIG. 9.18 Blends of C3S and ‘Aerosil’ silica fume dispersed in water. Ca2+ ions concentration in the liquid phase as a function of time for different C3S/S ratios; w/s ¼ 10. (From: Kurdowski W, Nocun-Wczelik W. The tricalcium silicate hydration in the presence of active silica. Cem Concr Res 1983;13:341–8.)

Pozzolanas and Pozzolanic Materials

387

FIG. 9.19 Degree of hydration of C3S as determined by XRD analysis. Composition and symbols as in Fig. 9.15. (From: Ogawa K, Uchikawa H, Takemoto K, Yasui I. The mechanism of the hydration in the system C3S-pozzolana. Cem Concr Res 1980;10(5):683–96.)

FIG. 9.20 Quantities of fly ash reacted. C3S/fly ash ¼ 3, w/s ¼ 0.5. (From: Mohan K, Taylor HFW. Pastes of tricalcium silicate with fly ash—analytical electron microscopy, trimethylsilylation and other studies. In: Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 54–9.)

In the presence of calcium sulfate and pozzolanic material, C3A hydration gives the AFt phase ettringite, and solid solutions of the AFm monosulfate hydrate and calcium aluminate hydrate phases, although at different ages.145 Where calcite is present in the pozzolanic material, the monocarbonate AFm phase (Ca4Al2O6CO311H2O) will also be formed.155 The same appears to be true of C4AF hydration.156

C3S The paste hydration of C3S results in the formation of C-S-H on the surface of unhydrated material. After some months of hydration, two distinct morphologies of C-S-H are evident—an ‘inner’ and ‘outer’ product. The inner product is found in close proximity to any remaining C3S and possesses very little by way of microstructural features. The outer product takes the form of fibres extending outwards into the pore space of the paste. C-S-H is a non-stoichiometric compound and, as a result, compositions of the substance obtained from different locations within a cement paste sample will show some variation. This will, however, also be compounded by that of the analytical technique used. The composition of C-S-H can be determined by direct and indirect methods. Microanalysis carried out by electron probe micro-analysis (EPMA), scanning electron microscope (SEM) and transmission electron microscope (TEM) equipped with suitable analytical devices belong to the first methods. Chemical determinations, based on selective dissolution, form the second group.

388 Lea’s Chemistry of Cement and Concrete

Indirect chemical determination of the Ca/Si ratio of C-S-H deriving from C3S hydration gives a value in the region of 1.7–1.8.157 Ca/Si ratios ranging between 1.5 and 1.9 have been found using direct analysis.149,158–163 The variation in composition between the particles within a paste is significant and certainly higher than that found in crystalline substances.158 No significant differences, possibly associated with time, degree of reaction, or morphological type of C-S-H, are noted.158 Where C3S and a pozzolanic material are present together, C-S-H with a high Ca/Si ratio initially forms on the C3S surface, whilst a porous layer of low Ca/Si C-S-H forms on the pozzolanic material. After a certain period of time, which depends on the characteristics of the material, the water/solids ratio, the temperature, etc., the surfaces of the grains of pozzolanic material are attacked by H3O+ protons contained in the basic solution resulting from the hydrolysis of the calcium silicates. The attack brings about a gradual dissolution of Na+ and K+ and produces an amorphous layer rich in both Si and Al on the grain surface. This layer reacts with the Ca2+ ions present in the solution and transforms into calcium silicate and calcium aluminate hydrates. Dissolved alkalis intensify the protonic attack of water.149 During the first month, the alumina, iron and SO4 2 contained in fly ashes are to be found in the form of small amounts of AFt and AFm phases.154,164 Within 90 days, however, the three elements are incorporated in C-S-H154 or hydrogarnet.164 In mature C3S/fly ash pastes, the morphologies of the hydration products of tricalcium silicate are not markedly different from those of pure C3S pastes. The hydration products surrounding the residual unreacted C3S particle core form an ‘outer’ and an ‘inner’ layer of C-S-H. The ‘outer’ product often appears as fibrillar under the TEM.164 No foreign elements are found in the inner region of C3S hydration products, whereas the outer region can contain other elements, especially potassium, deriving from the fly ash.164 After a curing period of 1 year, a rim of dense C-S-H is formed along the edges of the reacted fly ash particles, from which radial fibres of lower density C-S-H branch off. In both regions the Ca/Si ratio is roughly the same, although it is lower than in the C-S-H formed in plain C3S pastes.164 The existence of different concentric regions or shells of dense C-S-H around a fly ash particle is attributed to rhythmic precipitation processes.164 In other parts of the reaction zone, dense plates made up of hydrogarnets occur. This phase contains both Fe and Al released by the fly ash.164 In the presence of pozzolanic materials the composition of C-S-H is altered. Electron microscope analyses have shown an average Ca/Si ratio of 1.43 after 2 weeks, in the presence of fly ash (compared to 1.51 in control C3S pastes).154 In C3S/fly ash pastes, the Ca/Si ratio of the inner product around anhydrous C3S was 1.56  0.05 after 2 weeks, and 1.45  0.01 after 1 year of hydration. At the latter age, the outer product had a Ca/Si ratio of 1.6  0.16, but contained some aluminium and potassium.164 Field emission scanning electron microscopy (FESEM), has been used to compare C-S-H composition in pastes containing C3S and combinations of C3S and natural pozzolana. In plain C3S paste, the Ca/Si ratio of the hydrated mass extending between adjacent C3S grains reached a constant value of 2 after 3 days of curing.149 This value almost coincides with data reported by some,160,161,165 but is higher than the value found by others.159 However, the possible presence of portlandite crystals may locally increase the apparent Ca/Si ratio up to around 3.2. The Ca/Si ratio changes in the presence of natural pozzolanas. After 3 days the hydrates filling the space between C3S and the pozzolana grains were found to have a Ca/Si ratio as high as 2.5–3.0. However, near the pozzolana edge, the ratio sometimes reached values as high as 4, indicating the probable presence of portlandite.149 After 91 days the Ca distribution in the paste was more uniform, and outside the C3S surface the Ca/Si ratio decreases gradually, reaching a mean value of 1.7 (Fig. 9.21).149 In the presence of silica fume, the hydration of C3S also forms an outer and inner product. The reaction of silica fume is evident in the fracture surface of pastes, because the boundary between the silica particles and the outer C-S-H becomes less distinct. The smaller particles appear to be consumed.151 The silica grains can be observed in the outer regions of C-S-H, but not in the inner product that has formed on C3S grains.151 Generally, paste fracture takes place between the inner and the outer hydration products formed on the C3S grains.151 TEM studies have revealed that the composition of C-S-H which forms in the presence of silica fume is different from that formed in plain C3S paste. The inner product of plain C3S paste had a Ca/Si ratio equal to 1.68  0.07, whilst the ratio was 1.60  0.15 when silica fume was present. However, in the presence of silica fume, there were some areas in the paste consisting entirely of C-S-H having a ratio of 0.92  0.09. It was presumed that these areas had contained silica fume particles which had subsequently reacted fully with Ca2+ ions.151 This Ca/Si ratio is lower than that occurring normally, but corresponds to the low Ca/Si found in low-Ca2+ environments.166 These results suggest that the overall Ca/Si ratio of C-S-H decreases as the pozzolanic reaction proceeds and most of the silica reacts.151 Analysis by 29Si magic angle spinning nuclear magnetic resonance (MAS-NMR) of 2 year old mixtures of silica fume, lime and water, have suggested the existence of two structurally distinct forms of calcium silicate hydrates having a Ca/Si ratio ranging between 0.65 and 1.00 (low-lime C-S-H) and between 1.1 and 1.3 (high-lime C-S-H).167 Whilst these results were obtained in dispersions (water/solids ¼ 200) and, thus, may not apply directly to compounds formed in cement pastes,

Pozzolanas and Pozzolanic Materials

389

FIG. 9.21 Blends of C3S and pozzolana (V). Mix ratio ¼ 6:4, w/s ¼ 0.4, curing ¼ 91 days. Distribution of Ca/Si ratio between two C3S and pozzolana grains is shown. (From: Ogawa K, Uchikawa H, Takemoto K, Yasui I. The mechanism of the hydration in the system C3S-pozzolana. Cem Concr Res 1980;10(5):683–96.)

the implication is that C-S-H of normal composition forms initially, but when all the locally available Ca(OH)2 is consumed, the excess silica reacts with the hydrate already formed and produces the second form of C-S-H with a lower Ca/Si ratio. 29 Si MAS-NMR analysis conducted on paste made of 80% C3S and 20% silica fume at an age of one day showed that the inner product developed on the surface of C3S grains had a Ca/Si ratio of 1.6  0.1, but after 28 days the inner and the outer product had the same ratio of 1.2  0.1. This indicated that some C3S remained unreacted, since the overall Ca/Si ratio of the mixture was 1.5.148 The different compositions of C-S-H reflect the degree of silica polymerisation, calculated by determining the trimethylsilyl derivatives of C-S-H. As the degree of polymerisation increases, the Ca/Si ratio decreases. As can be seen from Fig. 9.22, pastes made up of C3S and natural pozzolana show a higher content of polymers relative to dimers than for pure C3S pastes.113,154,168 The degree of polymerisation of the polymer, expressed by the average number of silicon atoms in the anion, depends on the type of pozzolanic material. However, the variation is relatively small, ranging between 11.0 and 12.9 after 180 days of curing.168 The dimer content begins to decrease after a certain period of time, varying between 3 and 7,168 90154 and 270113 days. These very different results should be viewed with some caution, and could be attributed to different experimental conditions, water/solids ratios, C3S/pozzolanic material ratios, curing temperatures, etc. The influence of the C3S/pozzolanic material ratio on the polysilicate content of C-S-H is evident in C3S-silica fume hardened pastes. For a S/C3S ratio of up to 0.16, the polysilicate content rises slowly with the degree of hydration to around 40%, but at a ratio of 0.5 and a degree of hydration of 60%, it grows rapidly and reaches about 80%, the remainder being dimeric in nature. After 28 days of hydration in the presence of silica fume, the average length of the silicate chain (3.6) is

FIG. 9.22 Proportions of total Si present as monomer, dimer, polymer and insoluble in C3S pastes (A) and C3S—natural pozzolana pastes (B) as determined by trimethylsilylation. Specimens cured for 365 days at 20°C; w/s ¼ 0.5. (From: Massazza F, Testolin M. Trimethylsilylation in the study of pozzolanacontaining pastes. Il Cemento 1983;1:49–62.)

390 Lea’s Chemistry of Cement and Concrete

FIG. 9.23 Relative proportions of silicate species present during the progressive hydration of (A) doped C3S and (B) doped C3S and silica fume calculated from NMR spectrum peaks of 29Si nucleus. Q0 ¼ monomer units (hydrated), Q1 ¼ end units, Q2 ¼ middle units, Q4 ¼ silica network. (From: Dobson CM, Goberdhan DGC, Ramsay JDF, Rodger SA. 29Si MAS NMR study of the hydration of tricalcium silicate in the presence of finely divided silica. J Mater Sci 1988;23:4108–14.)

greater than that of the plain tricalcium silicate paste and the Ca/Si ratio drops from 1.7 to 1.5.151 The addition of silica gel modified with 1%–10% of Fe2O3 is claimed to increase the degree of polymerisation of silicate anions of C-S-H.169 29 Si MAS-NMR has established that in the silica–lime–water system the higher-silica C-S-H (Ca/Si ¼ 0.65–1.0) predominantly consists of long chains of silica tetrahedra, formed by two end units (Q1) and middle units (Q2), whereas the less silicarich C-S-H (Ca/Si ¼ 1.1–1.3) consists of dimers comprising two joined Q1 end units and short chains (consisting of Q1 end units and small numbers of Q2 middle units).167 After 24 h of hydration, the intensity of the peak Q0 (corresponding to C3S) is 60% of the original value in the plain C3S paste and <20% in the presence of silica fume. In agreement with this, Fig. 9.23 shows that the height of Q1 and Q2 peaks is higher in the presence of silica fume than in plain C3S.151 A study of the hydration of a mixture of C3S and silica fume by solid-state 29Si NMR and selective isotopic enrichment has shown that silicon atoms from silica fume and from C3S occupy different positions in the C-S-H structure. Silicon from both sources initially forms dimeric C-S-H, but, later on, silicon from the silica fume takes part in the formation of C-S-H with a longer chain length and a slightly more ordered structure than that deriving from C3S hydration.170 Portlandite is generally present in hardened pastes containing combinations of C3S and pozzolanic materials, even where pozzolanic material is present at a level which should be more than sufficient to combine with all the lime released by C3S hydration. By means of a TEM, portlandite crystals surrounding fly ash particles which show no sign of pozzolanic reaction have been observed.164 This occurrence is presumably due to restricted access of the pore solution, that is, to very low local porosity.164 In the presence of 20% silica fume, hydrated C3S paste also shows large lamellar crystals of portlandite running between masses of C-S-H.151

9.4.3 Hydration of Cements Containing Pozzolanic Materials 9.4.3.1 Kinetics of Hydration Portland cement (Portland clinker plus gypsum) and pozzolanic materials follow different reaction processes and react at different rates. However, when PC and pozzolanic materials are combined, they will each influence the reaction kinetics of the other. It is generally agreed that pozzolanic reaction becomes apparent—at least with the most common pozzolanic materials—3–14 days after mixing with water, when around 70%–80% of the C3S in the PC has reacted.171,172 The rate of pozzolanic reaction depends on the properties of the pozzolanic material, mix proportions, and temperature. Thus, in the case of silica fume, which has a BET specific surface area higher than that of other pozzolanic materials, the reaction starts earlier. The dormant period of the pozzolanic reaction is also partly explained by the strong dependence of the solubility of the reactive portion of pozzolanic materials on the alkalinity of the pore solution. Thus, the reaction will not start as long as the pH has not reached the required value.173 The kinetics of PC hydration are modified by the presence of a pozzolanic material and its influence is revealed by changes occurring in    

heat of hydration combined water the degree of hydration of C3S portlandite content

Pozzolanas and Pozzolanic Materials

391

Other phenomena act as indicators of the influence of pozzolanic materials on hydration, such as the degree of silicate polymerisation, but the above parameters are those most frequently considered. Heat of Hydration The hydration process can be monitored by recording the rate of heat evolution and the total heat of hydration released by pastes. The determination of the rate of heat evolution is a sensitive method for detecting and recording the changes induced by pozzolanic materials in the early hydration of cement up to 3–7 days. The cumulative heat of hydration is the algebraic sum of all the enthalpy variations related to every reaction occurring in the hydrating cement. It increases with time up to a certain asymptotic value following a typical path. The effect of pozzolanic materials on the heat evolved by hydrating cement becomes clearer when the measured parameters, namely the rate of heat evolution and the cumulative heat of hydration, are referred not to the whole system comprising clinker, gypsum and pozzolanic material, but only to the PC fraction. Such normalisation modifies neither the duration of the dormant period nor the time at which the second peak reaches its maximum value. In cements containing natural pozzolanas, the dormant period is shortened, the height of the first peak is increased, and the normalised height of the second peak is greater than that of the control PC (Fig. 9.24). This is most likely the result of acceleration of C3S hydration, as previously discussed. With fly ashes, calorimetric parameters prove to be very variable, since the duration of the dormant period can be lengthened9,174–176 or remain unchanged177,178 and the intensity of the first peak can be increased9 or decreased.174–178 The influence of fly ash on the second peak (when normalised) varies considerably, with different studies observing it to be either lower,175,178 or higher9,171,174,177 than that of the parent PC. In the presence of fly ash the length of the induction period increases with water/cement ratio.174 Extension of the dormant period in pastes containing 70% PC by mass and 30% of various fly ashes has also been observed using a quasi-adiabatic calorimeter. The temperature increase obtained with two low-lime and two high-lime fly ashes was lower than that of the control PC, but that of a third high-lime fly ash was higher, probably owing to the dissolution of other compounds.179 Water-leached fly-ashes have been shown to have less of a retarding effect on the second peak than that shown by as-received fly ashes.178 The cumulative heat of hydration released by PC accounts for the overall reactions occurring during hydration. It is derived from the area under the rate of heat evolution curve. The total heat released from blended cements in the first 2 days is always lower than that of the PC control, but it is generally higher175,180 when normalised to the PC fraction (Fig. 9.25). This means that fly ashes, similarly to other pozzolanic materials, accelerate the hydration of PC. Studies of the effect of silica fume on heat evolution have reported a slight extension of the dormant period and a marked increase in the height of the second peak,180 no change in the dormant period and a small reduction in the height of the first peak,175 or acceleration.181 In fact, the influence of silica fume on the length of the dormant period appears to be sensitive to the water/cement ratio, with a lower ratio leading to an extension of the dormant period, but higher ratios reducing its duration.182 A 15%183 or 20%175 level of silica fume in cement results in a lower total heat of hydration released in the first 2 days, compared to the heat evolved by plain PC paste. However, when normalised, the heat evolved is usually roughly the same175,183 or slightly higher.180 During the first 24 h the cumulative heat of hydration remains unchanged or increases

FIG. 9.24 Calorimetric curves from the hydration of some blended cements. w/s ¼ 0.4, temperature ¼ 20°C. (From: Takemoto K, Uchikawa H. Hydratation des ciments pouzzolaniques. In: Proceedings of the seventh international congress on the chemistry of cement, Paris; 1980, vol. I. p. IV-2/1–21.)

392 Lea’s Chemistry of Cement and Concrete

FIG. 9.25 Total heat evolved by various cement blends during hydration. Heat referred to the PC fraction. (From: Huang C-Y, Feldman RF. Hydration reactions in Portland cement-silica fume blends. Cem Concr Res 1985;15:585–95.)

for small silica fume contents (5%) and decreases for higher (10%–15%) contents.183 Cumulative heat evolved is strongly affected by the water/cement ratio. In the presence of silica fume at these higher levels, the rate of heat evolution of clinker increases when the water/cement ratio is between 0.5 and 0.6,180 remains unaffected when the water/cement ratio is in the range 0.35–0.50,175,183,184 and is reduced when the water/cement ratio is 0.35–0.28.184 Ground silica sand also increases the total heat evolved by the PC fraction, but the effect is small (Fig. 9.25). The influence of metakaolin on heat evolution is highly dependent on the type of PC it is used with. Where the C3A content of the PC is high, metakaolin generally has the effect of causing the second heat evolution peak in pastes to occur slightly earlier. This appears to be largely the result of an acceleration of the conversion of ettringite to monosulfate.185 With some PCs, this reaction produces quite pronounced peaks on the rate of heat evolution curve. However, where C3A levels are lower, the effect is either negligible,186 or slightly retarding.185 As for silica fume, the total heat evolved is reduced relative to a PC control, but very similar or higher when normalised.186 Combined Water The degree of hydration of PC is often estimated by determining the non-evaporable water content of the paste, that is the weight loss occurring in dried specimens heated through the temperature range between 105°C and 1000°C. Non-evaporable water includes inputs from all products of hydration, and thus also from calcium hydroxide. By subtracting the water bound in Ca(OH)2 from non-evaporable water, the content of water that is chemically combined in silicate and aluminate hydrates is obtained. This water will be referred to as ‘chemically bound’ water in the subsequent discussion. Owing to the changes caused by pozzolanic materials in the amount of the different hydrates and the variability in their composition, plus potential loss of water during sample preparation, the determination of non-evaporable water is only indicative of the degree of hydration of blended cements.187 Nevertheless, comparison of the results obtained from the blended and parent PCs can help to reveal differences in the hydration processes. The non-evaporable and bound water in fly ash cement pastes are lower than in PC,177 but are greater if normalised to the PC fraction136,177,188–190 (Table 9.13). The data show that the non-evaporable water content is already higher in blended cement than in PC after 3 days, but a difference may be observed even after 1 day.136 At early ages this difference can be interpreted as an acceleration of the hydration of PC and, at later ages, as a consequence of the progress of the pozzolanic reaction, which causes an increase in C-S-H content. In general, high-lime fly ashes (around 30% CaO) give cement paste with a higher non-evaporable water content.179 This effect is also observed in pastes containing silica fume191 and metakaolin.192 The increase in non-evaporable water content will, in part, be due to an increase in chemically bound water in C-S-H. Often, assuming that the non-evaporable water content is exclusively, the result of hydration of the PC fraction leads to an

Pozzolanas and Pozzolanic Materials

TABLE 9.13

Portlandite and Bound Water Contents for PC-Fly Ash Pastes From TGA-DTG Data188 Ca(OH)2 Content

Bound Water Content

Paste

Age

Per 100 g Total Binder

Per 100 g PC

Per l00 g Total Binder

Per 100 g PC

Control PC

2h 5h 1 day 3 days 7 days 28 days 2h 5h 1 day 3 days 7 days 28 days 2h 5h 1 day 3 days 7 days 28 days 2h 5h 1 day 3 days 7 days 28 days 2h 5h 1 day 3 days 7 days 28 days

0.94 6.05 15.05 18.12 19.16 23.94 0.40 1.53 11.16 17.76 18.57 18.94 0 2.45 8.70 12.84 12.94 15.06 1.08 1.34 11.02 16.47 16.35 20.73 0 0.96 8.91 12.77 12.53 19.42

0.94 6.05 15.05 18.12 19.16 23.94 0.46 1.76 12.84 20.44 21.37 21.79 0 3.35 11.89 17.54 17.68 20.57 1.24 1.55 12.70 18.98 18.84 23.89 0 1.32 12.21 17.49 17.16 26.60

2.82 9.00 14.75 19.53 21.39 26.85 2.34 4.13 13.56 19.55 20.95 24.34 2.44 4.54 12.17 17.21 19.25 23.14 2.80 4.72 13.82 19.93 20.60 27.91 2.23 3.54 14.13 18.94 19.50 22.38

2.82 9.00 14.75 19.53 21.39 26.85 2.69 4.75 15.61 22.49 24.11 28.01 3.33 6.20 16.62 23.51 26.29 31.60 3.23 5.43 15.93 22.97 23.74 32.16 3.05 4.85 19.36 25.95 26.72 30.66

4515a

4530

1015

1030

393

Pastes are identified by the shorthand forms 4515, 4530, 1015 and 1030 representing 15% 45 mm ash, 30% 45 mm ash, 15% 10 mm ash and 30% 10 mm ash, respectively.

a

estimated degree of hydration of about 100% to be calculated. This is highly unlikely, and indicates that the water content of the hydrates in the blended cement is higher than that of the PC paste. This conclusion is more evident in the case of cements containing high-lime fly ash, for which the calculated degree of hydration of the PC fraction may exceed 100%.136 Despite this, it should be noted that the hydration products are essentially similar. Most of the chemically bound water is released below 450°C, which is the temperature at which Ca(OH)2 decomposes.188 By comparing the weight loss between 100°C–200°C and 200°C–300°C, a straight line is obtained, suggesting that the hydrates are substantially alike, irrespective of whether or not fly ash is present in the paste.171 The amount of non-evaporable water (without normalisation) in a hardened paste decreases with increasing silica fume content (at least, up to a level of 25% by mass of solids), whereas the amount of chemically bound water increases. The results are strongly influenced by the water content of the paste, since non-evaporable water decreases with water/cement ratio.180,184,187 Degree of Hydration of C3S and Other Clinker Compounds The determination of unreacted C3S in the cement paste is a useful method for comparing the progress of hydration in a blend with PC alone. The degree of hydration of C3S at a given time increases when the cement contains pozzolanic materials, compared to PC alone.9,171,177,193 In one study it was found that after 28 days of curing, the degree of hydration of C3S in four cements containing 25% fly ash was 3%–9% greater than plain PC paste. This accelerating effect depends on the type of fly ash177 and increases with the proportion of this in the paste.171 The change in the rate of hydration of C3S has been observed at ages as early as less than one day,171,177 but it is normally seen only after several days.171

394 Lea’s Chemistry of Cement and Concrete

The determination of the residual C3A content in hydrated pastes is not a particularly useful means of evaluating the influence of pozzolanic materials on cement hydration kinetics, because of the low content of C3A in PC, the uncertainties associated with the analytical determination of this phase by XRD, and its rapid reaction. Nevertheless, results obtained suggest that fly ashes may slightly increase the phase’s rate of hydration,171 as has already been discussed for pure C3A in the presence of gypsum.144,145 The rate of hydration of C2S does not change in the presence of fly ash until 14171 or 28193 days, but at later ages it is reduced.193

Degree of Reaction of Pozzolanic Materials The unreacted content of pozzolanic material can be quantitatively determined in cementitious systems by using selective dissolution methods employing salicylic acid–methanol,154 salicylic acid–methanol and potassium hydroxide–sugar– water,171 or picric acid–methanol–water–solutions.194 With these methods the reacted fraction of pozzolanic material is calculated as the difference between the insoluble residue determined before and after a given hydration period. The glass fraction of fly ash contained in a blended cement reacts gradually (Fig. 9.26). However, after 1 year the unreacted fraction can still be as high as 50%.171 Table 9.14 shows that silica fume is more reactive than low-calcium fly ash, but after 3 days, the amount of reacted silica rises very slowly, in spite of the fact that the water/cement ratio of the silica fume cement is higher than that of fly ash.194 Table 9.14 suggests that the reactivity of a pozzolanic material can be underestimated if calcium hydroxide cannot migrate and react with particles of pozzolanic material owing to hindrances presented by the developing hydrate microstructure.195 The rate of reaction of metakaolin is relatively rapid, meaning that, where low levels are used it will have reacted to its fullest extent after relatively short periods of time. This can be deduced from measurements of Ca(OH)2 levels in metakaolin pastes, which indicate full reaction at a 10% metakaolin level after around 100 days.196

FIG. 9.26 Fraction of FA glass reacted as a function of age. Fly ash blended with PC (filled circles154) or C3S (open circles171). (From: Dalziel JA, Gutteridge WA. The influence of pulverized-fuel ash upon the hydration characteristics and certain physical properties of a Portland cement paste. Cement and concrete association technical report 560; 1986. 28 pp.)

TABLE 9.14 Degree of Hydration of Pozzolanic Materials in Cement Pastes, for Blends Containing 30% Fly Ash (FA) or 10% Silica Fume (SF) Hydration Time Specimen

1 Day

3 Days

7 Days

28 Days

90 Days

C + FA C + SF

7.0 53.1

8.5 67.9

14.3 68.7

36.3 69.0

61.6 78.5

Pozzolanas and Pozzolanic Materials

395

9.4.3.2 Compounds Occurring in Pastes of Cements Containing Pozzolanic Materials The reaction products resulting from the hydration of blended cements are the same as those occurring in PC pastes. The differences solely involve the relative proportions of the various compounds as well as their morphology. The main hydrates found in the hardened pastes are:      

ettringite monosulfate tetracalcium aluminate hydrate (often carbonated) C-S-H C2ASH8 Ca(OH)2

These compounds can be detected by DTA, DSC, XRD, optical and electron microscopy, and by selective dissolution. Ettringite and Monosulfate The AFt phase ettringite forms rapidly in cements containing natural pozzolanas197 or fly ash.188,198 Ettringite has been observed from 5 h up to 28 days in cements containing fly ash,188 and up to 1 year in cements containing natural pozzolanas.197 Ettringite can disappear after 3 days,9 being transformed into monosulfate. Its conversion was noted in lowSO3, but not in high-SO3 fly ashes.177 The conversion of ettringite into monosulfate depends on the amount of SO3 available and on the CO2 content of the cement paste. In fact, carbon dioxide reacts with excess calcium aluminate hydrate and gives monocarboaluminate, thus preventing it from reacting with ettringite to form monosulfate.199 For this reason, ettringite is often found co-existing with monocarbonate. Similar results have been obtained with a cement containing 15% silica fume cured for 10 years. AFt was associated with a medium content of calcium carbonate, whereas monosulfate was associated with a low calcium carbonate content.200 Analysis using analytical electron microscopy revealed that this AFt phase significantly deviated from the composition of pure ettringite. Table 9.15 shows that the atom ratio of plain PC pastes and those with 15% silica substitution deviates from the theoretical values.200 In particular, the CaO=SO4 2 ratio is significantly higher than the theoretical value of 2.0. These figures imply that a substitution in the SO4 sites by silica and, presumably, CO3 2 anion has occurred.200 Tetracalcium Aluminate Hydrate Tetracalcium aluminate hydrate can be present in pastes, depending on the Al2O3/SO3 ratio of the cement, but it is generally carbonated, either because it has been contaminated by the CO2 of the atmosphere,197 or because of the presence of calcite as filler in cement or as a constituent of the pozzolanic material.188 C-S-H In the presence of pozzolanic materials, C-S-H is already visible after 24 h, because of the accelerated reaction of C3S.152 Uncertainties regarding the composition of C-S-H, as observed previously for the products of C3S hydration, are again found when considering PC and cements containing pozzolanic materials. C-S-H formed by hydration of PC shows some differences with respect to that formed by C3S hydration. The variation in composition within specimens is comparable to that found in C3S pastes, but the range of Ca/Si ratios is greater, extreme values being 1.0 and 2.8.201 The reasons for this variability in composition are the same as those concerning the formation of C-S-H by the hydration of C3S, but in the case of cements, the influence of other elements such as Al, Mg, Fe and S has to be

TABLE 9.15

Pastes of PC and PC + 15% Microsilica Produced by a Flame Hydrolysis Process200 Atomic Ratios

Sample

Ca/Al

Ca=SO4 2

  Ca= SO4 2 + Si

PC PC + Si Theoretical pure phase

2.73  0.15 3.06  0.39 3.00

2.66  0.86 2.73  0.69 2.00

2.55  0.78 2.39  0.58 2.00

Atom ratios of the AFt phase for PC and PC + Si, as determined by analytical electron microscopy.

396 Lea’s Chemistry of Cement and Concrete

taken into account.202 Moreover, in PC pastes, the Ca/Si ratio has been found to increase with voltage of the electron microprobe analyses163 and curing time.201 The latter effect has not always been observed.203 However, where it is observed, it may be attributable to changes in the C-S-H sulfate content.204 Several workers have highlighted some compositional differences between ‘inner’ and ‘outer’ products. The inner product appears to be pure C-S-H with a median Ca/Si ratio of 1.5,205 1.7,206,207 2.1208 or 1.75.209 The Ca/Si ratio of the outer product seems to be higher than that of the inner one as it ranges between 1.6205 and 2.7.208 Such a high ratio perhaps reflects a mixture of C-S-H with Ca(OH)2 and AFm phases, or changes in composition due to the presence of foreign elements coming from the interstitial clinker phases. If the inner product generally appears to be pure or almost pure calcium silicate hydrate, the outer product contains several other elements such as Al, K and sulfate.205 In any case, the outer and inner products have been shown to have a different composition only when the cement is partially hydrated, that is, when chemical equilibrium is far from being attained. In the C-S-H of a 10-year-old concrete, presumably nearly fully hydrated, a single composition (Ca/Si ¼ 1.82–1.97) has been found to predominate.208 The Ca/Si ratio increases with decreasing water/cement ratio.209 This could be one of the major causes of variation in C-S-H composition and the reason why the Ca/Si ratio is lower in concrete than in pastes.209,210 Using EPMA, other workers have identified the outer and inner products in a 23-year-old paste without any significant deviation in the Ca/Si ratio of the two layers, the mean value being estimated at 1.67.207 Minor elements contained in C-S-H vary to a large extent,204 and this may account for the high variability of the Ca/Si ratio. This is supported by the observation that the (Ca + Mg)/(Si + Al + Fe + S) ratio is less variable than Ca/Si, suggesting substitution of Ca and Si with other elements within the C-S-H structure.208 EPMA carried out on concrete made of Portland and blended cement has shown that the Ca/Si molar ratio in the hydrates varies irregularly from one point to another, but this irregularity is normally distributed.210 In cements containing pozzolanic material, besides the C-S-H deriving from the hydration of clinker silicates, C-S-H from the reaction between the pozzolanic material and calcium hydroxide is also found. Table 9.16 shows cementitious materials (in this case including slag as well as pozzolanic materials), change the composition of C-S-H.152 Similar reductions in the Ca/Si and aggregate/cement ratios are also found in concrete.210 The composition of C-S-H formed from pozzolanic reaction is different from that originating from alite and belite, resulting from the different conditions of formation. After an 8-day hydration period, the presence of fly ash causes the Ca/Si ratio of the C-S-H around the alite grains to decrease from 1.71 to 1.55.206 The lower Ca/Si ratio of the C-S-H is associated with a higher potassium content.206 In mature pastes of cement containing 30% fly ash, inner and outer products having a slightly lower Ca/Si ratio than those of PC alone have been found (1.49–1.50 and 1.45–1.60, respectively).205 However, greater differences have also been reported. For instance, after 4 years of curing, a 40% fly ash cement and PC contain C-S-H with Ca/Si ratios of 1.01 and 2.03, respectively. This difference persists in the Ca/(Si + Al) ratio.152 The Ca/Si molar ratio decreases with fly ash content.205,211 Reported values of C-S-H composition are averages since the Ca/Si ratio decreases with increasing distance from the surface of alite particles or portlandite crystals.211 The presence of silica fume results in a Ca/Si ratio of C-S-H of around 1.1 which is lower than the overall ratio of PC paste (around 1.2). The high Si content of silica fume means that the Ca/Si ratio of C-S-H markedly decreases as the silica fume content increases.212 Typically, levels 13% and 28% give Ca/Si ratios of 1.3 and 0.9, respectively.213 Electron microscope analysis, carried out on 10-year-old pastes, has shown that the mean Ca/Si ratio is higher in PC (1.54  0.21) than in cement with 15% silica fume (1.36  0.19). The analytical values are rather scattered, indicating a degree of chemical imbalance throughout silica fume pastes, with both high and low Ca/Si ratio forms of C-S-H potentially co-existing.200 The presence of unusually large (35–80 mm), rounded silica particles in a silica fume-containing cement has given the opportunity to examine a simple, but significant, example of pozzolanic reaction. A cement containing 10% silica fume of this type was cured for 1 year and the polished surfaces of the paste were examined by SEM and analysed by EDXA. SEM revealed that the silica particles had reacted totally or partially, forming C-S-H while retaining their original outline.

TABLE 9.16 Composition of C-S-H in Cement Pastes With and Without 40% Blending Component (w/c 5 0.40, t 5 293 K, age 4 years) by EMPA152

PC Fly ash cement Slag cement

Ca/Si

Al/Ca

Ca/(Si + Al)

Na2O (%)

K2O (%)

2.03 1.01 1.62

0.06 0.21 0.44

1.81 0.84 0.96

0.03 0.24 0.23

0.11 0.33 0.30

Pozzolanas and Pozzolanic Materials

397

FIG. 9.27 Schematic representation of a coarse silica fume particle in a cement paste with Ca penetration to the centre. (A) Ca/Si mole ratio determined by EDXA along traverse A–B. (B) Weight % K2O determined by EDXA along traverse A–B. (From: Bonen D, Diamond S. Occurrence of large silica fume-derived particles in hydrated cement paste. Cem Concr Res 1992;22:1059–66.)

The results of EDXA showed that the Ca/Si ratio of C-S-H was more or less the same (1.7–2.0) around the silica particles, although it decreased linearly from the periphery to the centre (0.3–0.6).214 Fig. 9.27 shows a plot of Ca/Si molar ratio, along the cross-section of a silica particle. The central core can be completely calcium-free or may contain some calcium. The rate at which the Ca/Si ratio decreased ranged from 0.036 to 0.095 (Ca/Si)/mm.214 EDXA data indicate that calcium silicate hydrates of different composition can co-exist for a long period, possibly on account of their similar chemical potential and the difficulty for Ca2+ ions to travel through the microstructure of a mature paste. Morphological differences between the C-S-H formed around clinker grains and natural pozzolana particles have also been observed by means of optical microscopy in pastes containing natural pozzolanas.215 The lower Ca/Si ratio of C-S-H found in cement containing pozzolanic materials may be related to the different degree of polymerisation of the silicate anion, which is greater in blended cement than in PC pastes.113 29Si NMR investigations have shown that the Q2/Q1 ratio is higher in silica fume-containing cement than in PC. This means that the average chain length of silicate anion is increased by the presence of silica fume.216 In theory, it might be anticipated that the conversion from pozzolanic material to C-S-H leads to expansion, since the density of the reaction product is less than that of most pozzolanic materials. C-S-H typically has a density of between 1900 and 2100 kg/m3,212 whilst silica fume, for instance, possesses a density of around 2200 kg/m3.82 However, expansion is not observed.214 C2ASH8 In cements containing high-lime fly ashes, besides C-S-H, C2ASH8 (gehlenite hydrate, or str€atlingite) also forms.198 If gypsum is added to cements, gehlenite hydrate gradually decreases and eventually disappears after 60 days. C2ASH8 is also observed in pastes containing metakaolin, with a notable absence of tetracalcium aluminate hydrate.217

9.4.3.3 Pore Solution The solution contained in the pores of the cement paste can be partially extracted under high-pressure188,218,219 and then analysed by atomic absorption spectroscopy, plasma photometry, ion chromatography and other means of analysing solution chemistry. The analytical data obtained from extracted solutions must be considered with care for the following reasons:  extraction removes 10%–20% of the total free water, that is, only the water contained in the larger pores;  the uniaxial external pressure applied to the sample is not equally transferred to every point of the cement paste.190

398 Lea’s Chemistry of Cement and Concrete

In spite of the above reservations, this approach is, at present, the only method available for separating and analysing pore solutions. The presence of pozzolanic material in a cement paste will change pore solution composition, and this may have consequences with regard to concrete durability, particularly in the case of reinforced concrete where pore solution pH plays an important role in protecting steel. Up to an age of 28 days, the OH– concentration in the pore solution of cements containing 20% fly ash is lower than that of PC, but higher than that of the cement containing inert ground sand173,188 (Fig. 9.28). This reduction occurs even if the alkali content in fly ash is markedly higher than that of the clinker. This may be because of the low dissolution rate of alkali contained in the fly ash glass. Nevertheless, the difference between the curves of fly ash and ground sand suggests that after about 10 days, fly ash has contributed to increasing the pH, whereas ground sand has had a diluting effect. At a later stage, the hydroxide concentration decreases in the blended cement paste and increases in the PC paste until it reaches an asymptotic value (Fig. 9.29).220 The reduction in alkalinity in the presence of fly ash has been recorded at up to 1 year.210,220 Since the alkali content of the four fly ashes included in Fig. 9.28 was higher than that of the control of PC, the effect may be due to the incorporation of alkalis in the C-S-H gel formed by pozzolanic reaction.220 Detailed results of pore fluid extraction experiments are shown in Table 9.17. It is worth noting that, after 1 day of hydration, the pH is higher in blended cement than in the parent PC pastes.188 This may be due to an accelerated early

FIG. 9.28 OH– concentration as a function of time in the pore solution of plain PC paste and blended cement paste. 20% fly ash or fine quartz sand in cement. Temperature ¼ 20°C, water/binder ¼ 0.45. (From: Fraay ALA, Bijen JM, de Haan YM. The reaction of fly ash in concrete: a critical examination. Cem Concr Res 1989;19:235–46.)

FIG. 9.29 OH– ion concentrations in pore solutions of PC paste (C) and blended pastes containing 40% of various pulverised-fuel ashes (1,2,3,4) or 40% of hypothetical cement of zero alkali content (C0 ). Alkali content of the parent PC and four fly ashes is indicated. (From: Canham I, Page CL, Nixon PJ. Aspects of the pore solution chemistry of blended cements related to the control of alkali silica reaction. Cem Concr Res 1987;17(5):839–14.)

Pozzolanas and Pozzolanic Materials

TABLE 9.17

Chemical Data for Pore Fluids Extracted From PC-Fly Ash Pastes188 Fly Ash

Fineness (mm)

Age

pH

Si

Ca

Al

Na

K

SO422

OH–

Cations/Anionsa

Control

2h 5h 1 day 3 days 7 days 28 days 2h 5h 1 day 3 days 7 days 28 days 2h 5h 1 day 3 days 7 days 28 days 2h 5h 1 day 3 days 7 days 28 days 2h 5h 1 day 3 days 7 days 28 days

13.24 13.26 13.61 13.79 13.80 13.83 13.25 13.24 13.68 13.77 13.78 13.80 13.23 13.26 13.69 13.72 13.73 13.70 13.28 13.27 13.73 13.78 13.81 13.81 13.23 13.26 13.71 13.72 13.74 13.73

0.3 0.4 0.9 1.0 1.0 0.8 0.4 0.4 0.8 0.8 1.1 1.0 0.3 0.3 0.7 1.0 0.9 0.9 0.3 0.3 0.8 1.0 1.5 1.0 0.2 0.3 0.4 0.8 0.5 1.6

3.5 0.5 3.1 5.5 3.9 3.7 0.7 4.7 4.7 5.4 5.4 2.9 14.1 4.4 1.8 3.4 4.2 2.9 16.8 2.7 4.7 4.2 5.5 3.7 12.1 3.4 3.0 3.1 3.1 2.9

<0.1 <0.1 <0.1 0.1 0.1 0.2 <0.1 <0.1 <0.1 0.1 0.1 0.2 <0.1 <0.1 <0.1 0.1 0.1 0.2 <0.1 <0.1 <0.1 0.1 0.1 0.2 <0.1 <0.1 <0.1 0.1 0.1 0.3

46 45 61 68 66 58 48 51 63 69 75 69 55 54 73 78 77 63 48 56 69 81 90 82 60 63 74 79 82 102

426 473 549 581 605 607 424 406 492 472 558 473 335 339 434 445 444 364 386 396 492 516 510 486 313 336 393 428 434 435

177.8 166.6 88.0 8.4 8.9 9.5 150.9 152.2 32.8 6.3 7.3 9.5 144.0 124.0 10.4 5.9 5.2 4.0 166.6 161.4 15.4 6.2 7.3 8.1 143.1 156.2 1.3 1.9 3.6 6.9

175 180 403 617 629 677 176 172 474 589 599 627 168 184 487 523 539 497 190 184 540 607 641 645 168 183 510 528 546 537

0.91 1.02 1.07 1.05 1.06 0.97 0.99 0.98 1.05 1.09 1.06 0.86 0.89 0.93 1.01 1.00 0.97 0.86 0.90 0.91 1.00 0.98 0.94 0.88 0.88 0.82 0.93 0.97 0.95 1.00

15

<45

30

<10

15

<10

30

2+

Ion balance ¼ 2Ca 2SO

4

Ionic Concentrations (mmol/L)

Content (%)

<45

a

399

+ Na + + K + . 2 + OH

hydration of PC caused by the presence of fly ash, or to the dissolution of some alkaline compounds from the fly ash. Beyond one day, pH values are lower where fly ash is present, with a higher fly ash content yielding lower pH values at later ages.210,220 Finer fly ash actually appears to increase pH at later ages, possibly as a result of the release of sodium from the ash. Indeed, Table 9.17 and Fig. 9.30 both show that the increase in OH– is matched by a decrease in Ca2+, and so the alkalinity of the pore solution is mainly due to alkalis released by the PC and fly ash.173 The Na+ concentration in the pore solution of the blended cement paste is slightly higher than that of PC while the K+ concentration is lower (Table 9.17).188 Taking into account that fly ash has a higher alkali content than PC, one can reasonably conclude that the decrease in K+ indicates a take-up of potassium in the hydrated solid phases. Moreover, since the free water available in the pores decreases with time, an unchanged concentration also means that increasing amounts of Na+ are incorporated into the solid hydrate phases. Concentrations of Si and Al in the pore solution are very low. Unlike fly ash, natural pozzolanas do not change the pH of the pore solution when the PC has a high alkali content, but increase it if the PC has a moderate alkali level.221 In PC pastes OH, Na+ and K+ increase with time until constant values are reached, but in a cement containing a considerable amount of silica fume (20%) the development of concentrations of these ions is different. Figs 9.31–9.33 show that plots of concentration versus time are similar for all three ions, and that a peak appears after 5 days of hydration.222 The height of the peak depends on the water content of the mix,222,223 but after 24 h the influence of the water/cement ratio on the concentrations is negligible. The rapid development in concentration can be explained in terms of an acceleration of PC hydration caused by the fine silica particles, whereas the decrease is likely to be due to the incorporation of alkalis into the solid hydrate phases.222 The height of the concentration peak decreases with increasing silica fume content.220,223 It is worth noting that the decrease in alkali concentration caused by silica fume is not counterbalanced by an increase in Ca2+.

400 Lea’s Chemistry of Cement and Concrete

FIG. 9.30 Ca2+ concentration in the pore solution of plain PC paste and blended cement paste. 20% fly ash or fine quartz sand in cement. Temperature ¼ 20°C; water/binder ¼ 0.45. (From: Fraay ALA, Bijen JM, de Haan YM. The reaction of fly ash in concrete: a critical examination. Cem Concr Res 1989;19:235–46.)

FIG. 9.31 OH– concentrations in the pore solution of pastes as a function of age and water/(PC + silica fume) ratio. Silica fume content ¼ 20%. (From: Larbi JA, Fraay ALA, Bijen JMJM. The chemistry of the pore fluid of silica fume-blended systems. Cem Concr Res 1990;20:506–16.)

FIG. 9.32 Na+ concentrations in the pore solution of pastes. Samples as in Fig. 9.31. (From: Larbi JA, Fraay ALA, Bijen JMJM. The chemistry of the pore fluid of silica fume-blended systems. Cem Concr Res 1990;20:506–16.)

Pozzolanas and Pozzolanic Materials

401

FIG. 9.33 K+ concentration in the pore solution of pastes. Samples as in Fig. 9.31. (From: Larbi JA, Fraay ALA, Bijen JMJM. The chemistry of the pore fluid of silica fume-blended systems. Cem Concr Res 1990;20:506–16.)

The reduction in OH– concentration must be taken into account by engineers when designing concrete mixes.224 Fig. 9.34 shows the consequence of a high silica fume level, with the pH falling below 12.5, which is the value of a saturated calcium hydroxide solution.218 The decrease in pH has been shown to persist in pastes at ages of 10 years (Table 9.18).200 In the same 10-year-old pastes, the alkali concentration is lower than in a 3-month-old paste, although the decrease is only slight (Table 9.19).200 Taking into account the reduction in free water that will have occurred after a 10-year period of hydration, the same table confirms that the presence of silica fume causes a marked decrease in the concentration of Na+, K+ and OH– dissolved in the pore solution of the paste, most probably as a result of the incorporation of alkalis, especially K+, in the hydrates. Al and Si concentrations are, again, very low. Metakaolin also substantially reduces OH concentrations in the pore solution. In a study which employed metakaolin with both high- and low-alkali PC, 20% metakaolin reduced OH concentrations to comparable levels for both cements, but when used at a level of 10% was only able to reduce concentrations slightly in the high-alkali paste.225

9.4.3.4 Portlandite The progress of hydration in PCs has often been followed by measuring the increase in portlandite content with time. However, the values must be used carefully when expressing the degree of hydration of PC, since after 28–91 days the portlandite content decreases somewhat, in spite of the progress of hydration.136,226

FIG. 9.34 Influence of the silica fume content on pH values of the pore solution from cement pastes (PC and blended cements). Water/(PC + silica fume) ¼ 0.50. (From: Page CL, Vennesland O. Pore solution composition and chloride binding capacity of silica-fume cement pastes. Materiaux et Constructions 1983;16(91):19–25.)

402 Lea’s Chemistry of Cement and Concrete

TABLE 9.18 Calculated Total OH2 Concentrations and pH Values of Pore Solution Measured in 10-Year-Old Pastes200 Cement Paste

Total OH2 (m/ML)

pH

PC PC + 15% FHS HAPC HAPC + 15% FHS

202 96 510 198

13.17 12.87 13.55 13.15

PC, Portland cement; HAPC, high-alkali Portland cement; FHS, flame hydrolysis silica.

TABLE 9.19 Measured Composition (mM/L) of Pore Solution From 10-Year-Old Pastes (Figures in Parentheses are for 90 Days Curing, Symbols as in Table 9.18)200 Solution

Na

K

Ca

SO4

Al

Si

Mg

PC

47.8 (64) 46.7 (70) 78.8 (66) 59.8 —

155.4 (200) 50.0 (63) 432.7 (551) 139.4 —

2.06 (2.75) 0.18 (3.0) 0.75 (1.2) 1.22 (1.45)

1.04

0.087

0.131

<0.004

0.716

0.168

1.025

—

7.55

0.240

0.369

<0.004

1.46

<0.15

0.142

—

PC + 15% FHS HAPC HAPC + 15% FHS

The presence of pozzolanic material modifies the ordinary course of the hydration of cements such that the portlandite content depends not only on the degree of hydration of clinker, but also on the progress of the pozzolanic reaction, the type of pozzolanic material and the ratio of pozzolanic material to PC. Despite the added complexities, the determination of free lime provides useful information concerning the hydration process, when it is carried out simultaneously on blended cement and the corresponding plain PC. As shown in Fig. 9.35, after 1 hour of hydration the portlandite content in fly ash cements made with both calcareous (Class C) and siliceous (Class F) is higher than that calculated purely from dilution.178 At later stages it decreases, but after 8 h it increases again until, after 24 h, it reaches the theoretical value that would be expected if the presence of fly ash were only having a diluting effect.178 This brief movement towards negative values relative to the PC paste could be due to a retarding effect on cement hydration or early pozzolanic reactions. Taking into account the heat evolution curves from these pastes, the second explanation is more likely.

FIG. 9.35 Ca(OH)2 in normalised fly ash/cement (40:60) pastes minus Ca(OH)2 in neat cement pastes. (From: Wei F, Grutzeck MW, Roy DM. The retarding effects of fly ash upon the hydration of cement pastes: the first 24 hours. Cem Concr Res 1985;15:174–84.)

Pozzolanas and Pozzolanic Materials

403

At a later stage the portlandite content is expected to decrease gradually as a consequence of the progress of the pozzolanic reaction. Actually, the overall Ca(OH)2 content of pastes containing pozzolanic material is always lower than that of control Portland cement pastes, but at 28 days it may be higher (see Table 9.13),136,177,188 lower94,177,188,227 or even the same136,228 when the content is normalised to the PC fraction. In practical terms, in the first month the portlandite content of paste is more or less equivalent to that formed in the control PC multiplied by the dilution factor. The behaviour of natural pozzolanas and fly ashes with regard to their influence on portlandite content is quite similar, but, on comparing various cements containing 30% of different pozzolanas, fly ashes and ground limestone, it has been found that natural pozzolanas typically fix more lime than fly ashes.94 Silica fume reacts with substantial quantities of portlandite: one study examining the pozzolanic reaction of the material found that a paste containing 20% silica fume had a portlandite content of around 5% by volume at an age of 28 days.220 If the reduction were exclusively the result of dilution, this value would have been just under 20%. Similarly, a paste containing 25% metakaolin had a portlandite content of around 5% by mass after a year,197 which would have been just under 20% if only dilution was responsible. In 28-day-old blended cement pastes, Ca(OH)2 has been detected by etching polished sections with appropriate micrographic reagents and by examining them under the reflected light microscope.215 In many cements, the amount of pozzolanic material present is insufficient to combine with all the Ca(OH)2 released by clinker hydration, and so the occurrence of portlandite even after 10 years is possible.200 Even where the level of pozzolanic material in a cement is sufficiently high to potentially consume all Ca(OH)2 in pozzolanic reaction, Ca(OH)2 may still be present after long periods of time. This can only be attributed to the often slow rate of the pozzolanic reaction, and to the resistance to Ca2+ ion movement to reaction sites in a mature cement paste. Table 9.20229 shows that, although free Ca(OH)2 decreases with time, it is still detectable after 4 years of curing. The long-term persistence of large quantities of portlandite in hardened blended cements may seem improbable, since common volcanic pozzolanas, as well as fly ashes, can combine with as much as 50% of their weight of Ca(OH)2 (Fig. 9.7).11 By way of example, a blended cement having a natural pozzolana content of 35%–40% and taking into account that the portlandite content in the mature PC paste was around 22% CaO, the amount of pozzolana present should be sufficient to combine with all of the portlandite formed. In actual fact, the portlandite content in blended cement pastes made with different volcanic pozzolanas ranges from between 3% and 8%94,230 after 90 days. Of course the actual calcium hydroxide content depends on many factors, and particularly on the nature of the pozzolanic material. In cement pastes, the calcium hydroxide content decreases with increasing silica fume content,183,187,226,231,232 but small additions (around 1.3%) increase the 1-day free lime content.232 This further supports the idea that silica fume has an accelerating effect on the rate of hydration of PC. Extrapolation of experimental data indicates that 20%–22% silica fume is sufficient to combine all the calcium hydroxide released by PC hydration,227,231 provided that the paste is cured for more than 28 days. This result infers the formation of C-S-H gel with a Ca/Si ratio of around 1.10, a ratio which has indeed been found in mixes of C3S and silica fume.95 With lower water/cement ratios (0.2–0.4), 16% silica fume seems to be sufficient to combine with all the free lime contained in pastes cured for 550 days.187 This difference from the previous result is possibly due to the low water content, which does not permit complete hydration of the clinker. In pastes made with a relatively high water/cement ratio (0.6), the Ca(OH)2 content does not change after 91 days of hydration, at least up to 10 years,200 whereas with a lower water content, it decreases after 90 days. This dependence of the pozzolanic reaction on water/cement ratio can explain the occurrence of portlandite in cement paste containing 15% silica fume cured for 10 years.200

TABLE 9.20 Free Lime Content in Cement Pastes Cured for 4 Years, for Cements Containing Different Amounts of a Natural Pozzolana229 Ca(OH)2 Pozzolana Content (%)

ATD Method (%)

Franke Method (%)

0 10 20 30 40 50

10.2 7.6 5.3 2.9 1–1.5 1

9.7 7.2 5.4 3.1 1.3 0.9

404 Lea’s Chemistry of Cement and Concrete

FIG. 9.36 Portlandite in cement pastes. (From: Costa J, Massazza F. Natural pozzolanas and fly ashes: analogies and differences. In: Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 134–44.)

After a period of curing—depending on many factors such as the type of pozzolanic material and temperature—the Ca (OH)2 contained in blended cement pastes reaches a maximum. In Fig. 9.36, the lower two curves represent the average values with cements containing three different natural pozzolanas, and a similar number of fly ashes.94 At even greater ages, this decrease in portlandite can be followed by an increase.230 This fluctuation of Ca(OH)2 concentration has also been observed in cements containing low-lime fly ashes136, high-lime fly ashes226 and metakaolin.73 Increase in portlandite content is clearly driven by PC hydration, whilst the decrease is the result of pozzolanic reaction. Fluctuations in the portlandite content of pastes are likely to be caused by the simultaneous occurrence of different phenomena including acceleration or slowing of the hydration of the clinker; pozzolanic reaction; and changes in the composition of the hydrated phases. The persistence of portlandite also in mature pastes must be ascribed to the difficulties encountered by both portlandite crystals and particles of pozzolanic material in reacting. It is likely that in a mature paste the yet-uncombined grains of pozzolanic material will be unable to react with the distant calcium hydroxide crystals as they are encased in reaction products. This possible explanation is supported by the following observations:  free lime is progressively less in pastes, mortars and concrete—that is in mixes tending to have increasing w/c ratios and, thus, greater porosity and permeability (Fig. 9.37)230;  as PC fineness increases, combined lime first increases, as a consequence of a larger reacting surface, and then decreases owing to the increased difficulty with which dissolved species are able to diffuse (Fig. 9.38)230;  the combined water content, when it is normalised to the PC fraction, is always higher in blended cement pastes than in control PC paste (Table 9.13)188;  it has been observed in some pastes that, despite having a significant Ca(OH)2 content, the pozzolanic material has reacted substantially after 28 days171,194;  portlandite crystals are sometimes observed in contact with fly ash without any sign of reaction.136

FIG. 9.37 Free portlandite content of (1) paste, (2) mortar (1:3) and (3) concrete (1:6) samples cured at 40°C (w/c ¼ 0.5); two types of natural pozzolana used at 35% in cement. (From: Massazza F, Costa U. Aspetti dell’attivita’ pozzolanica e proprieta’ dei cementi pozzolanici. Il Cemento 1979;1:3–18.)

Pozzolanas and Pozzolanic Materials

405

FIG. 9.38 Ca(OH)2 combined by 100 g of fly ash M versus the specific surface area (Blaine) of control PC. (From: Massazza F, Costa U. Aspetti dell’attivita’ pozzolanica e proprieta’ dei cementi pozzolanici. Il Cemento 1979;1:3–18.)

The difficulties encountered by ionic species when moving through a mass of limited permeability may cause the pozzolanic material to extract lime from the adjacent C-S-H, as in the case of C3S—natural pozzolana mixes.149

9.4.3.5 Mechanisms of Reaction in Cements Containing Pozzolanic Materials The formation of the hydration products, as well as the evolution of microstructure as a result of their formation, depend on many factors, such as the nature of the PC and the pozzolanic material, the fineness of the materials, the water/cement ratio, and temperature. For this reason, it is impossible to describe the processes of clinker hydration and pozzolanic reaction exactly. Nevertheless, the hydration of cements containing pozzolanic materials follows a common trend which enables some general behaviour to be identified. The process of hydration of cements containing pozzolanic materials can be divided into steps of different duration, depending on the individual characteristics of materials and environmental conditions. The steps show that clinker hydration and pozzolanic reactions practically overlap throughout the entire reaction period. In the case of fly ash, the sequence can be summarised as follows:  After 1 hour of hydration, there are already signs of reaction of clinker grains in the form of needles of ettringite and some poorly defined, small granular products. Fly ash particles often appear to have some hydration product on their surface.  After 4 h, in addition to ettringite, Ca(OH)2 appears on the surface of fly ash particles. Some of the ash particles show definite signs of pitting, which become more pronounced over the next few hours.  After 8 h the clinker grains are covered with ettringite, C-S-H and some large Ca(OH)2 crystals.  After 12 h these compounds can be observed even on particles of fly ash.  After 18 h the paste acquires some cohesion, AFt rods are better crystallised, Hadley grains (voids describing the shape of an anhydrous clinker grain that has since dissolved) are more numerous, and Ca(OH)2 crystals continue to grow.  After 1 day, AFt crystals appear to be longer and interlocking in some of the void spaces. Some of the fly ash particles appear to have reacted.  After 3 days some fly ash particles are partially covered with a coating of reaction products. It is assumed that the reacted part is rich in reactive glass, whereas the unreacted one is rich in relatively non-reactive crystalline compounds.  At up to 14 days the structure becomes denser with continuous infilling from Ca(OH)2 and C-S-H.  At later ages, while the reaction of clinker declines, the pozzolanic reaction goes on gradually until ions in the hardened pastes become essentially immobile.  In long-cured pastes, fly ash particles may be in one of three conditions:  encapsulated by the reaction product;  partially etched or reacted;  completely dissolved.

9.4.3.6 Paste Microstructure Morphology of the Paste Apart from the presence of unreacted particles of pozzolanic material, the morphology of cement pastes containing pozzolanic materials is very similar to that of PCs. However, some differences exist, with the microstructure of hydrated blended cement pastes positioned somewhere between that of PC pastes and mortars consisting of mixtures of lime and pozzolanic materials.

406 Lea’s Chemistry of Cement and Concrete

The first microstructural consequence of hydration of blended cements is the early formation of shells of hydration products around clinker grains. The shells are typically around 1 mm thick and are composed mainly of C-S-H. Ettringite appears as thin needles, whilst thin plates of monosulfate and portlandite crystals can also be observed. After a while, C-S-H fibres project outwards, coating the shells and making them appear fuzzy.233 After 1 day of hydration, electron microscopic examination of a fly ash-containing cement paste has shown that the reaction has involved only clinker. Plenty of ettringite crystals as well as fibrillar C-S-H cover both the clinker grains and fly ash microspheres.188,233 This microstructure confirms that fly ash particles initially perform as adequate precipitation and nucleation sites for hydrates.234 In blended cement pastes, ettringite needles are longer and thinner than those found in PC pastes.188 SEM examination of fly ash-containing cement paste has suggested that, after a day, fly ash particles are covered with a continuous and uniform layer of Ca(OH)2 which is, in turn, covered with a layer of C-S-H.235 Nevertheless, the occurrence of this duplex film around the fly ash particles has not been confirmed conclusively205 and it cannot necessarily be considered to be a typical feature of fly ash cement pastes. The C-S-H formed on the surface of clinker and pozzolanic particles often differs in morphology. C-S-H in the form of outer and inner products around alite surfaces has been observed in 1-week-old samples made with a cement containing 20% fly ash.206 The outer product is very thin, while the inner one was 3–4 nm thick. In contrast, the fly ash particles were covered by shells formed by radial growth of C-S-H and ettringite crystals. These shells were observed on both reacted and unreacted fly ash particles.234 The morphology of the hydrates and the slowness of the pozzolanic reaction suggest that these shells must have formed according to a process of dissolution of clinker minerals and precipitation of hydration products.234 A space between the fly ash particles and the shells of precipitated hydration products has been observed in some cases, and has been attributed to the dissolution of the outer glass layer.236 However, similar clearance has been observed along unhydrated fly ash grain boundaries.211 Thus, it may be the result of the relative weakness of the interface between the unaltered fly ash particles and the rim of the hydrates. This is supported by studies on hardened pastes of C3S blended with silica fume which have shown that cracks occur between the inner and outer product when samples are broken during the preparation of the paste surface for SEM examination.151 Furthermore, the gap may form due to shrinkage of the hydrate gel under the drying conditions experienced during specimen preparation.237 Shrinkage-induced rings have also been observed in hardened pastes containing natural pozzolanas.215 At later ages, some shells have been observed to be surrounded by portlandite.234 The reactivity of fly ash microspheres seems to depend on the surface conditions rather than on the size, since it has been observed that small particles can remain unreacted, whilst larger ones show signs of reaction.234 After 28 days, fibres of C-S-H have filled the majority of the pore space and the grains of pozzolanic material will have been attacked and surrounded by a thin layer of newly formed products.188,215 The microstructures of cement pastes containing natural pozzolana were examined using optical microscopy, which revealed the formation of a layer of hydration products around the pozzolana grains at this age.215 After 17–21 months of hydration, C-S-H occurs as an inner and outer product around unhydrated clinker particles, both in the fly ash and PC pastes, the outer product having a typically fibrillar appearance.205 In addition, fly ash particles are covered with a radial fibrillar layer of C-S-H formed within the original boundaries of the particle, which results from the pozzolanic reaction.205 The fly ash particles have a clear outer boundary delineated by fibrillar C-S-H gel and some particles show the presence of reaction products within their boundaries.205 In some fly ash cement pastes, well-crystallised hydrogarnet within the boundaries of the fly ash particle occurs.205 The addition of silica fume to PC does not change the paste microstructure significantly. Sections of a 7-day-old blended cement—prepared by ion-beam thinning and observed under a TEM—have shown silica fume particles confined in locations solely within the outer hydration product of the clinker. This supports the theory that the inner hydration product does not precipitate from solution. Rather, it develops within the boundaries of the clinker particles with the interface travelling forward between hydrated gel and unreacted clinker. At 7 days, the composition of the inner hydrate is not affected by the addition of silica fume.237 Porosity of Pastes Containing Pozzolanic Materials Porosity is an intrinsic property of cement paste. From an engineering perspective, it is normally desirable to limit the volume of porosity in concrete and mortar, but it cannot be wholly eliminated. It influences the strength and mass transport processes within cement pastes, mortars and concretes with implications for structural performance and durability. However, the connection between porosity and these two properties is different. Strength is essentially related to the total porosity, whereas mass transport processes depend on the structure and size distribution of the pores.

Pozzolanas and Pozzolanic Materials

407

The porosity of cement paste depends on many factors and typically decreases with water/cement ratio and increasing curing period. Porosity also increases with curing temperature; the difference mostly concerns the volume of large pores.238 The type of cement also influences porosity, although it is more important in pastes than in mortar and concrete.239 The determination of porosity and pore size distribution is an important means of characterising cement pastes, but presents a number of difficulties. These derive mainly from the microstructural changes that methods of specimen preparation and measurement induce in samples. Porosity measurements require the preliminary removal of water from the pores of the paste, and then filling the same pores with a suitable fluid. Water removal can be obtained by a range of different procedures, including oven drying in a CO2free environment, vacuum drying over a dry ice trap (D-drying) or over magnesium perchlorate hydrate (P-drying), drying over liquid nitrogen (sublimation), and replacement of water with a solvent. The porosity of dried specimens can be determined by mercury intrusion porosimetry (MIP), helium or methanol displacement at low pressure, water intrusion, or BET nitrogen adsorption. Backscattered electron imaging of polished cement paste sections has been used to determine porosity, but the method is limited to pores larger than around 500 nm.240 In addition to total accessible porosity, MIP also gives the pore size distribution. Depending on the pressure applied, mercury can penetrate pores having diameters ranging between 3.5 nm and 200 mm.241 The pore distribution curve allows calculation of  the threshold diameter (TD): the pore diameter at which the continuous mercury intrusion begins;  the critical pore diameter (CPD): the pore diameter corresponding to the modal pore frequency. On a plot of pore size distribution it is normally the starting point of the first major peak encountered moving from coarse to fine pore sizes;  the total volume of critically sized pores (VCP). The use of so many different procedures accounts for the range of values obtained in the determination of porosity and pore size distribution, and why caution is required in comparing the results from various types of cement. When the total porosity of cement pastes containing pozzolanic materials is assessed by MIP, the value obtained is generally higher than that of comparable plain PC pastes.188,227,234,242–244 This result is true for natural pozzolanas and fly ashcontaining cements (Table 9.21).245 It is also valid for silica fume-containing cements.246,247 Regardless of the type of pozzolanic material, porosity decreases with time, but continues to be higher than in PC paste.242,243 As stated earlier, porosity values depend on the drying procedure adopted.248 Table 9.21245 shows that more severe drying leads to higher porosities than milder methods. This effect is wholly apparent for PC in both a relatively young and mature condition, but for cements containing pozzolanic materials the change in behaviour with age becomes more complex, and dependent on the specific material. Unlike PC, the porosity of blended cement paste depends on the method of determination across the range of porosity volume fractions. Porosity measured by methanol or helium pycnometry appears to be lower than by MIP.226 This is illustrated in Figs 9.39 and 9.40.248 The few points representing fly ash pastes that lie close to the line of equality correspond to samples which show little or no pozzolanicity.249 The figures also show that the mercury porosity/helium porosity ratio is higher for oven-dried than solvent-replaced specimens, as is the case for PC paste.248 TABLE 9.21

Total Porosity (%) by MIP of Cement Pastes; w/c Ratio 5 0.32245 Porosity 28 Days Drying

7 Months Drying

Sample

PC

Blended Component

Rapida

Slowb

Rapida

Slowb

CEM 1 Filler Fly ash Vizzini pozzolana Qualiano pozzolana Casteggio pozzolana Barile pozzolana Segni pozzolana Bacoli pozzolana

100 70 70 70 70 70 70 70 70

30 30 30 30 30 30 30 30

17.00 17.80 21.80 18.70 20.00 17.80 17.70 17.50 17.80

14.70 21.40 21.30 21.60 18.80 16.30 17.90 19.30 18.70

13.10 15.30 17.40 14.30 11.30 13.60 13.50 13.40 13.30

10.60 13.80 16.40 12.90 10.70 11.60 12.70 34.20 11.40

a

Dried at 70°C for 16 h under vacuum at the residual pressure of 5 mbar. Dried at 20°C in four successive stages (relative humidity of 55%, 33%, 10% and 0.01%).

b

408 Lea’s Chemistry of Cement and Concrete

FIG. 9.39 Comparison of mercury and helium porosities; specimens prepared by direct oven drying at 105°C. Samples cured at temperatures from 20°C to 80°C and hydrated for periods up to 1 year. (From: Marsh BK, Day RL. Some difficulties in the assessment of pore-structure of high performance blended cement pastes. In: Very high strength cement-based materials, Boston, 1984. Materials Research Society symposium proceedings; 1985, vol. 42. p. 113–21.)

FIG. 9.40 Comparison of mercury and helium porosities; specimens prepared by solvent-replacement method. Samples cured at temperatures from 20°C to 80°C and hydrated for periods up to 1 year. (From: Marsh BK, Day RL. Some difficulties in the assessment of pore-structure of high performance blended cement pastes. In: Very high strength cement-based materials, Boston, 1984. Materials Research Society symposium proceedings; 1985, vol. 42. p. 113–21.)

MIP also gives higher porosity in the case of cements containing up to 30% silica fume.249 It is reasonable to suppose that, since the smallest diameter of the pores intruded by mercury is larger than the diameter of the helium molecule and the pressure imposed is higher for greater MIP porosity values, the pore entrances are closed by thin deposits of hydrates which are impervious to helium or methanol, but which cannot resist mercury under a pressure as high as hundreds of MPa. The difference in microstructure of blended cement pastes compared to PC pastes is reflected in parameters frequently derived from MIP data, namely the CPD and the VCP (Table 9.22).250 After 28 days’ curing, the CPD of a blended cement is lower than that of PC. After longer periods of curing (in this case 7 months), the critical pore size is further reduced by 10%–40% depending on the type of pozzolanic material.250 The rate of CPD reduction is higher for blended cement pastes than for PC paste. Between 1 and 28 days of hydration the differential pore size distribution moves to lower values: from around l.0 to 0.1 mm. In cements containing either natural pozzolanas or fly ash, the total volume of critically-sized pores—the magnitude of the peak on a pore size distribution plot which corresponds to the critical pore diameter—is generally slightly lower than in PC pastes after 28 days’ curing, but noticeably lower after 7 months.242 The differences between, and development of, these

Pozzolanas and Pozzolanic Materials

409

TABLE 9.22 Critical Pore Diameter (CPD) and Corresponding Pore Volume (VCP) of Hardened Cement Pastes; Blending Component 5 30%, w/c Ratio 5 0.32250 Age 28 Days Sample

Drying

PC

R L R L R L R L R L R L R L R L R L

Cement + filler Cement + fly ash Cement + Vizzini tuff Cement + Quagliano pozzolana Cement + Casteggio pozzolana Cement + Barile pozzolana Cement + Segni pozzolana Cement + Bacoli pozzolana

a

7 Months

CPD (nm)

VCP (%)

CPD (nm)

VCP (%)

60 52 50 55 50 54 40 40 40 37 44 44 35 46 30 35 35 40

20.0 19.9 25.0 18.6 16.2 22.8 16.4 21.7 19.5 23.4 15.1 28.8 19.8 22.2 17.6 16.1 20.8 18.3

52 44 40 46 30 32 24 24 26 19 38 34 28 30 28 36 24 24

18.6 27.0 26.7 23.1 14.1 17.0 13.2 11.4 5.7 7.3 12.4 19.2 9.4 13.9 8.0 10.3 10.0 12.6

a

L, slow drying; R, rapid drying.

porosity parameters show that the pore size distribution of blended cement pastes is typically shifted towards smaller dimensions compared to PC pastes. The combination of silica fume with PC has the effect of refining the porosity of pastes. MIP measurements conducted on pastes containing 5% and 10% silica fume display lower TD values than the PC control, although this only becomes pronounced at later ages (90 days).251 Metakaolin appears to refine pore structures to a greater extent than silica fume.251 Even at very early ages (around 3 days) the TD value obtained using MIP is considerably lower than the PC control, indicating that the fine particles of metakaolin play an important role in filling in capillary space between cement particles.252 Over a period of a year, TD continued to decline and remained below that of the control. Typically, however, total porosity is higher in metakaolin pastes compared to the control.196 As hydration progresses, there is a decrease in the volume of capillary porosity, but an increase in the volume of gel porosity.

9.4.4 Mass Transport Through Paste The entry of gases, liquids and solutions into concrete puts durability at risk. For example, sulfate can cause expansion and cracking of concrete, while carbonation and chloride penetration may lead to corrosion of reinforcement in the presence of oxygen and moisture. By assuming that there are no cracks and with the use of dense aggregate, the passage of media into concrete will be through the capillary pores of the cement paste phase. The transportation processes for fluids can normally be defined in terms of the following:  permeability (entry of fluids under a differential pressure);  sorptivity (entry of liquids by capillary suction);  diffusion (entry of gases or ions dissolved in water). Permeability, sorptivity and diffusion depend on the interconnectivity of pores; thus when this is limited, penetration of liquid or gases can be very low, in spite of a not insignificant porosity.

9.4.4.1 Permeability Permeability of blended cement-based systems is initially higher than that of PC, but tends to become lower as the curing time increases. In contrast, the porosity of blended cements remains higher than that of PC, with both water227,253 or air242,254 as the

410 Lea’s Chemistry of Cement and Concrete

TABLE 9.23 Permeability and Porosity of Cement Pastes Hardened for up to 90 Days254 Permeability (m2 × 10–17) 1

Curing (Days) w/c ratio ¼ 0.32 PC paste FAC paste BPC paste w/c ratio ¼ 0.40 PC paste FAC paste BPC paste w/c ratio ¼ 0.50 PC paste FAC paste BPC paste

3

7

Porosity (%)

28

90

1

3

7

28

90

5.60 17.40 1.94

0.30 3.01 0.70

0.12 0.18 0.12

0.00 0.07 0.06

0.00 0.00 0.00

20.8 34.9 29.5

19.7 32.5 26.3

14.4 25.0 21.3

9.8 20.6 14.7

5.9 17.4 7.1

18.70 68.30 14.30

0.59 8.30 1.80

0.07 0.91 0.42

0.07 0.09 0.02

0.00 0.00 0.00

33.3 37.7 39.3

28.6 37.3 29.8

20.9 31.6 24.7

16.8 26.2 20.6

11.1 16.8 10.9

214.00 284.00 218.00

14.70 178.00 22.30

2.35 12.70 3.74

0.19 0.09 0.06

0.00 0.00 0.00

43.6 46.6 44.4

37.8 40.3 42.4

32.2 34.0 40.0

20.8 30.6 24.4

14.5 24.7 22.4

BPC, Bacoli pozzolana cement; FAC, fly ash cement; PC, Portland cement.

measurement fluid (Table 9.23). Although there have been variations,188 permeability has mainly been found to depend on the content of pozzolanic material and the curing time.13,243 Table 9.24 shows some typical permeability values of pastes having different pozzolanic material contents.13 While permeability and porosity vary with time according to similar equations,254 there is no general correlation between the two properties, possibly owing to differences in the microstructures of Portland and blended cement pastes. The apparent contradiction in terms of a higher porosity and lower permeability of pozzolanic systems, compared to those of PC, can be explained through the following model.255  With most blended cements, hydration occurring during the first 7–15 days almost exclusively involves the clinker + gypsum fraction. By that time, 70%–80% of alite has reacted.9  A proportion of the hydration products of clinker and gypsum form on the clinker grains, whereas the remainder is deposited elsewhere on particles of pozzolanic material and in the pores. When the pozzolanic material starts reacting, its particles are surrounded by a porous, but already stiff, structure. Therefore, much of the products of the pozzolanic reaction form some distance from each reacting particle.  Since there is no evidence of a growing pressure, part of these products must have been formed elsewhere and later precipitated in the available pores.  The volume of the precipitating mass is small and is therefore unable to fill the larger pores, but it is sufficient to obstruct the thinner channels connecting the large pores, or at least to reduce their size.  As a consequence, the porosity of blended cement paste is still higher, even after the completion of pozzolanic reaction, but permeability is lower compared to that of PC. This model is supported by the lower porosity values found by helium and methanol pycnometry, compared to the value given by mercury intrusion. Since helium and methanol porosimetry indicate only the porosity accessible from the surface of the specimen,253 the higher porosity and lower permeability of blended cement paste is reasonable.

TABLE 9.24

Relative Depth of Penetration of Water into Hydrated Cement Pastes (mm). PC Blended With Santorin Earth13

Hydration Age

PC

10% Pozzolana

20% Pozzolana

30% Pozzolana

28 days 90 days 1 year

26 25 25

24 23 23

25 23 18

25 22 15

Pozzolanas and Pozzolanic Materials

411

9.4.4.2 Sorption Sorption is due to capillary forces which are active in semi-permeable systems subjected to natural wetting/drying cycles. Sorption is a more widely occurring phenomenon than permeability since it takes place when unsaturated pastes, mortars or concretes come into contact with water or moisture in air. Sorptivity, like permeability, decreases with water/cement ratio and increasing curing time. Sorptivity has been found to be lower in blended cement pastes than those with PC.256,257 An example of this is shown in Table 9.25, where mortar data is also included. The table also indicates that between 28 and 90 days, sorptivity reduces significantly, while only a slight variation in total porosity occurs. 9.4.4.3 Diffusion of Ions Water penetrating into concrete can also transport dissolved ions, which as noted above can have importance with regard to durability. Of these, chloride represents one of the most significant with regard to concrete performance. Diffusion associated with this type of ion is mainly covered in the following section. Diffusion of chloride depends on the cement used, but can vary appreciably even between materials of the same type. Typically, sulfate-resisting Portland cements (SRPCs) show higher diffusion coefficients than conventional PCs258,259 (Table 9.26). Cements containing pozzolanic material typically give pastes having lower diffusion coefficients than PC pastes.258–262 This effect is illustrated in Table 9.27 and in Fig. 9.41. In pastes with a water/cement ratio of 0.60 and 40% fly ash in cement, a reduction in Cl– diffusivity by an order of magnitude was obtained. Similar results have been found with lower levels (10%) of silica fume.261,262 With cement combinations, the effective chloride diffusion coefficient (Deff) of Cl– depends on the PC as well as on the type and level of pozzolanic material used.263 TABLE 9.25 Porosity by Volume P and Sorptivity S of Pastes and Mortars Made From Blended Cement (CEM IV) and PC (CEM I) Curing Time (Days) 28 Days

90 Days

P (%) Sample

Note

a CEM I paste b CEM IV paste c CEM I mortar d CEM IV mortar 1 CEM I paste 2 CEM IV paste 3 CEM I mortar 4 CEM IV mortar 5 CEM I mortar 6 CEM IV mortar

1

A

B

P (%) 0.5

S (mm/m )

A

B

S (mm/m0.5)

2 3 4 5

21.02 24.94 14.48 14.31 13.21 15.51

13.69 13.89 12.32 11.43 11.62 13.41

240 Days

0.225 0.153 0.157 0.119 0.122 0.139

21.14 25.93 14.70 14.85 13.11 15.74

13.13 14.51 11.31 10.64 9.75 12.29

P (%)

S (mm/m0.5)

28.07 30.64 12.34 14.94

0.632 0.228 0.110 0.610

0.147 0.104 0.132 0.090 0.102 0.097

Ref. 256

257

A ¼ Water-accessible porosity; B ¼ Mercury-accessible porosity. (l) w/c ¼ 0.5; (2) CEN mortar; (3) w/c ¼ 0.25; (4) 0.5–2 mm sand, s/c ¼ 2.25, w/c ¼ 0.335; (5) CEN sand, s/c ¼ 3, w/c ¼ 0.425.

TABLE 9.26

Effective Diffusivity of Chloride Ions at 25°C in Various Cement Pastes258

Type of Cement

D (cm2/s × 1029)

PC PC + 30% fly ash PC + 65% GGBS SRPC

44.7 14.7 4.1 100.0

412 Lea’s Chemistry of Cement and Concrete

TABLE 9.27

Coefficients for Diffusion of Chloride Ion into Cement Pastes and Concretes260

Sample

D (cm2/s × 1028)

Temperature (°C)

PC paste PC paste PC paste Blended cement paste Blended cement paste Blended cement paste PC concrete (vibrated) PC concrete (nonvibrated) Blended cement concrete (vibrated) Blended cement concrete (nonvibrated)

1.23 2.51 4.85 0.83 0.90 0.97 1.65 1.05 1.05 2.26

10 25 40 10 25 40 25 25 25 25

FIG. 9.41 Actual diffusion coefficient of Cl– ion versus water/cement ratio for cement paste samples exposed to NaCl solution in diffusion cells. (From: Hjorth L. Cement specifications for concrete exposed to chlorides and sulfates. In: CEB-RILEM international workshop on durability of concrete structures, Copenhagen; 1983. p. 229–35.)

As far as PCs are concerned, the higher diffusivity in the paste of SRPC compared to that of conventional PC,259 is attributed to its lower C3A content and thus to a reduced capacity to bind chloride in the form of a complex salt such as calcium monochloroaluminate hydrate (C3ACaCl212H2O). However, since paste made with PC may have a different porosity,258 it is difficult to separate chemical and physical effects and to identify which has greatest influence. Whilst blended cements contain less C3A than PC alone, fly ash, metakaolin and some natural pozzolanas are rich in aluminates which lead to the formation of hydration products capable of chloride binding. The lower diffusivity of chloride observed in the pastes containing pozzolanic materials may be related to their transportation properties, which are lower than those of PC. Indeed, the chloride diffusion coefficient ratio between PC and blended cement pastes is between 3 and 10258,261,263,264 that is, more or less the same ratio as for permeability coefficients.188,242 The lower values Deff of Cl– observed in pastes containing pozzolanic materials have also been attributed to a certain interaction between Cl– and the pore walls, which should be greater with cements containing pozzolanic materials than in the those containing only PC.265,266 Tortuosity of capillary pores in a cement paste should give rise to the same geometrical resistance to the movement of all dissolved species, but Fig. 9.42 shows that the ratio of oxygen to chloride diffusion increases with a decrease in the Cl– diffusion coefficient.263 This suggests that the ionic diffusion is affected by an electric double layer at the interface between the pore walls and the pore solution. Moreover, since chloride diffusion in a dense paste is more retarded than oxygen diffusion, it seems that the surface charge mainly affects the ionic transport through the micropores.267 The activation energy for diffusion of chloride is 17.6 kJ/mol in an electrolyte solution but goes up 40–50 kJ/mol in mature PC pastes with a water/cement ratio of 0.4.258,266 In contrast, the activation energy of Cl– through quartzite is only

Pozzolanas and Pozzolanic Materials

413

FIG. 9.42 Variation of the ratio of oxygen and chloride diffusion coefficients with chloride diffusion coefficient. (From: Byfors K. Influence of silica fume and fly ash on chloride diffusion and pH values in cement paste. Cem Concr Res 1987;17:115–30.)

4–8 kJ/mol.266 From these results it was inferred that the pore walls of the cement paste interact with chloride ions, whereas those of quartzite affect ion movement only because of geometrical factors. Measurements of the resistance to the transport of chloride ions through cement paste under the potential of an electric field (Federal Highway Administration test268) have demonstrated a proportionality between the charge passed and the water permeability of the paste. At the same water permeability, the electrical charge transported through the PC paste was about three times that passed through fly-ash-containing paste. Silica fume diverges from this behaviour since the electrical charge transported through a 10% silica fume cement paste was more or less the same as that measured through PC paste.269 This result is inconsistent with diffusion measurements which have shown that the diffusivity of chloride in silica-containing paste is lower than in Portland and fly ash cement.262 This apparent discrepancy seems to depend on the changes caused by the presence of silica fume in the composition of the pore solution and of the C-S-H. By increasing the silica fume content in cement, the C/S ratio of C-S-H decreases, but a lowering of C/S causes a reduction in the amount of so-called chemisorbed Cl.270 As a consequence, the Cl– content dissolved in the capillary pore solution increases (see Fig. 9.43218). Thus, the lower Cl– binding capacity of cement causes an increase in the Cl– concentration of the solution and an improvement in the electrical transport capacity.

FIG. 9.43 Percentages of total chloride ion content remaining dissolved in capillary pore solution of hydrated cement pastes with total chloride additions of 1.0% by weight of (cement + silica). (From: Page CL, Vennesland O. Pore solution composition and chloride binding capacity of silica-fume cement pastes. Materiaux et Constructions 1983;16(91):19–25.)

414 Lea’s Chemistry of Cement and Concrete

The apparent diffusion coefficient of monovalent cations increases with the atomic number.271 The activation energies for the diffusion of Na+ and Cl– are similar in an infinitely dilute solution (18.0 and 17.5 kJ/mol, respectively) but are quite different when they are calculated from the diffusion tests through cement paste. In this case, the values are 83.74 and 50.24 kJ/mol for a water/cement ratio of 0.4.266 These results could mean that Na+ interacts with the cement paste more strongly than Cl–266; however, as stated earlier, the discrepancy could be attributed to the co-diffusion of Na+ and Ca2+. Indeed, in the absence of any calcium ion supply, as in the case of diffusion through a quartzite plate, the diffusion coefficients of both alkalis and chloride are the same.266 The migration of Ca2+ through the cement paste is not in agreement with the results of chemical analysis of the pore solution derived from hardened cement pastes. The pore solution is practically free of calcium ions, thus the transported calcium ions must come from the calcium hydrates forming the cement paste. The leaching of Ca2+ from cement paste has been evidenced by a decrease in the portlandite content during the diffusion process.258

9.5 FRESH AND MECHANICAL PROPERTIES OF CONCRETE 9.5.1 Workability and Bleeding The use of pozzolanic materials, whose fineness is similar to that of PC, is unlikely to change the workability of mortar or concrete significantly. Generally, at fixed water content, fine/low-carbon fly ashes improve workability, while some of the higher fineness pozzolanic materials tend to reduce it. The benefits achievable with fly ash in relation to flow properties, are considered to reflect several physical and chemical influences, including their particle shape and texture, dispersion of PC flocs and influences on distribution of mix water, and have been reviewed previously in detail.272,273 In contrast, coarse/high-carbon fly ash can increase water demand, affecting concrete workability. Tests for water requirement (using flow methods on mortar) are included in fly ash standards. In ASTM C618,55 fly ash mortars should not exceed 105% that of the PC reference. In EN 450-156 for Category S fly ashes (higher fineness) the water requirement should be less than 95% of the PC. With this type of material, water savings should be achievable (compared to PC) enabling reduced water/cement ratio for equal workability in concrete, or for this to be increased if the water content is maintained.273 The adoption of pozzolanic materials can affect the use of superplasticisers in mortar and concrete, with compatibility influenced by the physical (specific surface) and chemical (surface charge and reactivity) characteristics of the particular addition.274 It has been noted that in terms of superplasticiser adsorption and flow characteristics, there is similar behaviour between PC and blended fly ash cement pastes (21% fly ash; 9.6% CaO).275 In contrast to pozzolanic materials with comparable fineness to that of PC, high fineness materials such as diatomaceous earth, silica fume and metakaolin, noticeably increase the water requirement in cementitious systems.276–278 As a result, concretes containing these materials may require an appropriate addition of superplasticisers to reach the required workability without further water. The dosage of admixtures needed for a certain workability tends to increase with the quantity of high fineness pozzolanic material used.251,279,280 Bleeding of concrete occurs as a result of the tendency for the particles present to settle and is influenced by various factors associated with the mix. In general, given the properties/surface area of pozzolanic materials and their effects on workability, it may be expected that reduced bleeding will occur when these are used. Recent developments have seen the introduction of self-compacting concretes in construction. These high fluidity, self-consolidation materials normally have high powder contents and use superplasticising admixtures (and in some cases viscosity modifying admixtures for stability). Pozzolanic materials have been considered as options for this type of concrete.281–284 These have also been used successfully with regard to the pumping of concrete,285,286 where high fluidity/stability are required.

9.5.2 Compressive Strength As some time is required for pozzolanic reactions to initiate, in the early stages these materials tend to behave like inert components, diluting the PC. However, as mentioned in Section 9.4.2, pozzolanic materials usually accelerate the early hydration of clinker compounds from around 8 h after mixing with water. The combination of pozzolanic materials with PC can reduce the initial rate of hardening, but reverse the situation at later ages. This has been noted with natural pozzolanas,13,177,230,287 and fly ashes177,188,189,227,288,289 and silica fume,232,290,291 an example of which is shown for mortar in Fig. 9.44.230 While this applies to some of the finer pozzolanic materials, similar, or higher strengths from early ages have also been reported. 73,251,292

Pozzolanas and Pozzolanic Materials

415

FIG. 9.44 Effect of substituting PC for pozzolanic material on the compressive strength of ISO mortar. Values expressed as percentage of the 28-day strength of reference cement. (From: Massazza F, Costa U. Aspetti dell’attivita’ pozzolanica e proprieta’ dei cementi pozzolanici. Il Cemento 1979;1:3–18.)

As shown in Fig. 9.45, the combination of 30% fly ash with PC can reduce the early strength of the control cement by 50%, thereby exerting a greater effect than that of dilution.228 The difference in strength between blended cements and PCs decreases with age and can eventually disappear or reverse. The stage of recovery depends on the fineness of both the PC,228 and the pozzolanic material,231 as well as on the latter’s reactivity. Guidance indicates that the curing of pozzolanic mortars and concretes needs more care than that of mixes made with PC.84

FIG. 9.45 Mean values of relative mortar compressive strength with respect to the control PCs. Series 1, 2 and 3 formed by a PC and a fly ash, both of increasing fineness. (From: Costa U, Massazza F. Some properties of pozzolanic cements containing fly ashes. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag, and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. I. p. 235–54.)

416 Lea’s Chemistry of Cement and Concrete

The reduced rate of hardening through the combination of PC with pozzolanic materials does not create significant problems when blended cements are prepared in cement works, since cement manufacturers can take steps to ensure that these have strength values conforming to standard specifications. When pozzolanic material is added to cement at the concrete mixer, where the properties of the pozzolanic material and the PC cannot be modified further, measures to optimise the properties of the mix are then required. The effect of pozzolanic materials on strength depends on a number of factors, the most significant of which are    

the pozzolanic material content of the cement; the type of pozzolanic material; grading and specific surface area of the pozzolanic material; the characteristics of the blending PC.

9.5.2.1 Type and Content of Pozzolanic Material An example demonstrating the effect of pozzolanic material content in cement on strength of mortar is given in Fig. 9.44.230 This shows that there is an optimum level to achieve maximum strength which can change with test age. The strength of cements containing pozzolanic material depends on the characteristics of the material being used.94,289,293,294 This is shown in Fig. 9.46, which includes mixes containing 20% natural pozzolana. Given the contribution of the pozzolanic reaction, the strength of the mixes is always higher than that of the control sample containing 20% ground quartz.293 The behaviour of Sacrofano pozzolana illustrated in Fig. 9.46 gives lower strength than the control despite its high amorphous silica content and fineness,94 similar to that of other materials such as silica fume and ground fly ash. A high specific surface area can possibly slow down the rate of hydration in mixes without water reducers. Common natural pozzolanas and low-lime fly ashes, having fineness similar to that of PC, give greatest 28-day strength when replacing about one-third of the PC.94,188 The loss in strength with a pozzolanic material of PC-size fineness can sooner or later be recovered, depending both on the type189 and level used. For example, the strength of a cement containing 15% of rhyolitic glass slightly exceeds that of the control PC after 7 days,290 but generally the strengths of the two materials are similar by 28 days. Conversely, with certain lower-quality pozzolanic materials, the strength of the control PC may always be higher than that of the blended cements,287 at least for 3295 or 6 months.296 Concrete made with fly ash cement, cured in water for long periods, typically gains strength beyond 28 days to a greater extent than that of a PC control. Table 9.28 gives six concretes of different composition which were tested from 2 weeks up to 362 weeks curing. Table 9.29 shows that the ratio of compressive strength between 362 and 4 week old samples increases with

FIG. 9.46 Strength development of different 20% natural pozzolana–cement mixes. Paste with w/c ¼ 0.5; 40
40
160 mm specimens cured for 72 h in water and then in 65% relative humidity at 20°C.

Pozzolanas and Pozzolanic Materials

TABLE 9.28

TABLE 9.29

417

Concrete Mix Proportions (kg/m3)297

Series No.

Free Water

Cement

Fly Ash

Sand < 4 mm

Gravel < 4 mm

0 1 2 3 4 5

163 154 154 154 154 154

181 167 125 107 94 83

0 146 146 146 146 146

736 570 604 619 633 644

1280 1322 1331 1334 1333 1333

Concrete Compressive Strength as a Function of Curing Time in Water at 20°C297 Compressive Strength (MPa) After

Series No.

2 Weeks

4 Weeks

16 Weeks

181 Weeks

362 Weeks

R 362/R 4

0 1 2 3 4 5

13.4 19.0 12.0 8.8 6.9 5.8

14.8 25.4 17.1 13.4 10.5 8.4

20.8 39.4 29.2 25.2 22.1 20.6

23.8 55.1 46.4 39.8 36.0 34.2

26.3 56.1 45.9 42.2 37.9 33.9

1.8 2.2 2.7 3.1 3.6 4.0

decreasing PC/fly ash ratios, ranging from 1.8 for the control PC concrete to 4 for the richest fly ash concrete. In other terms, the greater the ratio of fly ash to PC content in the mix, the higher the relative increase in strength.297 The relative strength gain is more marked in lean than rich mixes.298 There has been a growing interest over the last few years in using higher levels of fly ash in cement.299,300 However, one of the issues associated with this in concrete is reduced early strength. This can have implications with regard to removal of formwork or the time of loading on structural elements. It has been possible to overcome this by combining rapid setting cements or lower energy clinkers with fly ash (at 45% in cement) with longer term strength benefits thereafter.299 Initial results from a study301 which considered fly ash produced from some of the modern combustion techniques mentioned in Section 9.2.2, used at a level of 30% in cement, suggest that the behaviour in concrete is similar to that of material tested in earlier research (during the 1990s). After 28 days’ curing, the strength of mortars containing 5%–25% silica fume was found to be higher than that of plain cement mortar.291 A maximum strength was recorded at around 15% silica fume, but this result was obtained by adding an increasing amount of a superplasticiser to the mix and by changing the water/cement ratio. It has been noted at very low w/c ratios that strength decreased irrespective of the silica fume content.184 Silica fume at levels ranging between 5% and 15% in cement, generally increases the compressive strength of concrete for curing between 1 day and 2 years. However, silica fume reduces the workability of the mixes, thus strength may also be influenced by the addition of superplasticisers to the mix.302 Rheological properties of silica fume concretes strongly depend on the type of silica fume used.303 Some workers have found a gradual loss in compressive strength between 90 days and 2 years with 10% silica fumecontaining concrete cured in air, while no loss occurred in water-cured samples.302,304 Air curing adversely affects the long-term compressive strength of both silica fume and control concretes,304 but the effect was more marked in silica fume concrete.305 This phenomenon has been attributed to self-stresses caused by concrete drying.304 The effects of metakaolin on the strength of mortar and concrete have been considered in several studies.192,278,306,307 In most cases (at equal w/c ratio), increases in strength beyond that of a reference PC are achieved from early ages and are higher thereafter in tests beyond 28 days. Results indicate that strength enhancements are generally obtained with increasing levels up to around 15% or so in cement. The use of superplasticisers with high fineness/reactivity pozzolanic materials has contributed to the development of high strength concretes. Fig. 9.47 shows that with a silica fume-containing concrete, strength values of more than 110 MPa were obtained after 28 days curing and around 130 MPa after 91 days. The figure also shows that similar strength can be achieved

418 Lea’s Chemistry of Cement and Concrete

FIG. 9.47 Compressive strength of concrete made with different blends. Samples: 1 ¼ 10% silica fume; 2 and 2A ¼ 25% fly ash; 3, 4 and 4A ¼ ground blastfurnace slag; 5 and 6 are normal strength comparison concretes. Variable water and superplasticizer contents used. (From: Penttala V. Mechanical properties of high strength concretes based on different binder combinations. In: Proceedings of the symposium on utilization of high strength concrete, Stavanger, Norway; 1987. p. 123–34.)

by using fly ashes (sample 2), provided that the cement content is high enough and the water/cement ratio is low.308 It has been possible to obtain high strengths by combining 5%–15% of very fine pozzolanic material such as silica fume, ground fly ash or a combination of the two.309 High strength concretes can also been achieved through the use of metakaolin.292,310

9.5.2.2 Particle Size Distribution of Pozzolanic Materials The strength of mortar decreases to varying degrees when PC is partially replaced with fly ash fractions of different sizes.288,289 Fig. 9.48 shows that the negative effect increases with the particle size and decreases with time only for the finest fraction.288 The effects of fly ash fineness covering the EN 450-1 range (up to 40% retained on a 45-mm sieve) have been investigated in concrete.311 As shown in Fig. 9.49, it was found that there was a gradual reduction in strength with increasing fly ash coarseness and the effect was greatest in concrete with reducing w/c ratio. Similar behaviour was noted for material from single sources classified to different fractions, or collected from a power station over a period of time. In order to achieve equal 28 day strength, minor adjustments to the w/c ratio of the concrete mix were made. Coarser fractions of high-lime fly ashes may behave slightly better than low-lime fly ash, probably because they contain some hydraulic constituents.288 Grinding promotes the activity of fly ashes and it has been possible by carrying this out to achieve 90-day strengths higher than that of PC, but if the fineness of fly ashes exceeds certain limits, compressive strength can decrease instead.288 Similarly, work examining the effects of separating the ultrafine fraction of fly ash312 using an air–cyclone separation technique, with almost all particles <10 mm, indicated that mortar strengths were greatest at 15% in cement, rather than 30%, with 90-day

FIG. 9.48 Reduction of the compressive strength of binders containing 30% of low-calcium fly ash, relative to pure PC. (From: Giergiczny Z, Werynska A. Influence of fineness of fly ashes on their hydraulic activity. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. I. p. 97–115.)

Pozzolanas and Pozzolanic Materials

419

80

PC 11 / 30% PFA 1-6 70

Binder content, kg/m³

28 day cube strength, N/mm²

550 60 500 450

50

40

Mean = 59.5 N/mm² Mean = 54.0 N/mm² Mean = 48.0 N/mm²

400

Mean = 40.0 N/mm²

350

Mean = 33.5 N/mm²

300

Mean = 25.0 N/mm²

200

Mean = 16.5 N/mm²

30

20

10 0

5

10 15 20 25 30 35 PFA fineness, % 45 µm sieve retention

40

45

FIG. 9.49 Effect of fly ash fineness (45 mm sieve retention; multiple sources) on 28 day compressive strength of concrete. (Reproduced with permission from: Dhir RK, McCarthy MJ and Magee BJ Impact of BS EN 450 PFA on concrete construction in the UK. Constr Build Mater 1998;12(1):59–74.)

strength exceeding the PC reference. The mechanical properties of concretes incorporating silica fumes with varying characteristics, that is, having a silica content ranging from 79% to 95% and specific surface from 20 to 26 m2/g, did not show any significant difference.313 The lower strength resulting from too great a fineness of pozzolanic material corresponds to decreases in combined lime caused by an increase in PC fineness (Fig. 9.38). The fact that both strength and combined lime decrease is possibly due to the densifying of the paste, which hinders the mobility of ions.

9.5.2.3 PC Characteristics The strength of mortars containing pozzolanic materials also depends on the characteristics of the PC used in the mix. Fig. 9.50289 shows the influence of eight fly ash types and four PCs on the relative compressive strength of standard mortars. The strength of blended cement mortars can be higher or lower than that of corresponding PC mortars. Differences in strength have been attributed to the alkali content of the PCs289 but other factors play a role. As an example, differences in the rate of strength development caused by fly ashes have been attributed to a higher rate of formation of calcium hydroxide and this, in turn, to the different C3S contents in the PC, with the effects of fineness also playing a role. The strength of mortars made with cements containing up to 10% silica fume decreases with fineness of the parent PC, but it is always higher, at least up to 28 days, than the strength of that with PC.183 The high alkali content of clinker seems to increase the strength and rate of hardening of cements containing 15% silica fume.314 Partial replacement of fly ash for slag in cement reduces the strength to a greater extent than that which usually happens with PCs.289 Fly ashes have different effects on mortar strength, depending on the strength class of the PC being used. Eight fly ashes were blended with two different class PCs made with the same clinker. Fly ash at a level of 25% in cement caused a general decrease in the strength of the mortars. However, the blended samples from only one of the PCs recovered the strength of the plain mortar after about 90 days.228 The results obtained from this and work on concrete combining two fly ashes with 22 PCs (mainly originating from the United Kingdom)311 suggest that the behaviour of a blended cement depends both on the individual PC and the pozzolanic material it is combined with. For this reason the optimum pozzolanic material content should be selected on a case-by-case basis, taking into account the properties of the individual components.

420 Lea’s Chemistry of Cement and Concrete

FIG. 9.50 Effect of PC on relative compressive strength fc(f)/fe(0) of mortar with fly ash. Age ¼ 90 days, f/c ¼ 0.25, w/(c + f) ¼ 0.50. (From: Sybertz F. Comparison of different methods for testing the pozzolanic activity of fly ashes. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. I. p. 477–97.)

9.5.3 Tensile Strength Given the nature of concrete, the tensile strength is important with regard to the formation of cracks, whether they are a result of external loading or internal processes occurring within the material. Information on tensile strength for design can be obtained from appropriate standards.315 In a study investigating a natural pozzolana,316 of similar surface area to PC, at a level of 45% in cement and 0.6 w/c ratio concrete, where the compressive strength reduced from 41.0 to 21.5 MPa (compared to the reference concrete), the corresponding flexural strength reduced from 5.0 to 3.5 MPa. Tests for splitting strength317 with fine and coarse natural pozzolana in cement at levels of 19% and 29% in concrete of 30 MPa (specified concrete compressive cylinder strength), gave small differences for the property using a range of different aggregate combinations. Other work318 has shown potential for enhancing compressive and splitting tensile strength of natural pozzolana concrete by combining with silica fume in cement. The effects of fly ashes covering a range of properties including those produced by co-combustion319 indicate general agreement between splitting tensile/flexural and compressive strengths. The results from a study320 investigating tensile and flexural strength for concretes including, fly ash, silica fume and metakaolin, proportioned using different techniques and covering a range of w/c ratios, indicates similar behaviour for the range of materials with respect to their compressive strength. Relations between splitting tensile/flexural and compressive strengths have also been noted with silica fume, in a study investigating different levels in concrete.321 For split cylinder tests, the optimum for tensile strength at 28 days was 5%–10% silica fume, while this was higher for flexural strength. Another study322 has shown that silica fume enhanced the compressive and split cylinder strengths of concrete (w/c ratio 0.35) with low quality aggregates. Metakaolin has been noted in a review323 to enhance tensile and flexural strength of concrete compared to that of PC at levels in cement up to 15%. Other work described, on mortar, suggests small effects of metakaolin on flexural strength from 7 days.

9.5.4 Modulus of Elasticity The modulus of elasticity is influenced by the characteristics of cement paste and aggregate in concrete, the relative quantities present and their response to the application of load. As for tensile strength, details about the property for design are included in standards.315 Concrete containing natural pozzolana, coarser than that of PC, used at 20% in cement in a 0.57 w/c ratio concrete324 was found to have similar modulus of elasticity to a PC reference concrete at 60 days. The inclusion of pumice or diatomite at low levels in cement (1%, 2% and 4%)325 reduced both compressive strength and modulus of elasticity at test ages up to 28 days (with less effect at higher levels). In other work referred to above, the modulus of elasticity reduced by 2.5 GPa for each 15% natural pozzolana used to replace cement in concretes of equal w/c ratio,316 while at approximately equal strength (with 19% and 29% natural pozzolana), variations in modulus values, within a narrow range, were obtained.317 A study considering relatively high levels (40%–50%) of low lime fly ash in cement (in almost equal w/c ratio concrete)326 found that in line with compressive strength, the modulus of elasticity reduced progressively with increasing fly

Pozzolanas and Pozzolanic Materials

421

45 Series A

Elastic modulus (kN/mm2)

40

PC PC/FA PC/GGBS

Ecm= 9.5 (0.8 fcc + 8)1/3 (Table 1)

Normal MDLM

35

Mixsim

30 25 20

0.7 Ecm

15 Series B

10

PC/SF PC/MK PC/LS PC/FA/SF PC/FA/MK

5 0 0

10

20

30

40

50

60

70

80

Cube strength (N/mm2) FIG. 9.51 Effect of mix proportioning method (Series A) and cement type (Series B) on modulus of elasticity of concrete. (Reproduced with permission from: Dhir RK, McCarthy MJ, Paine KA. Engineering property and structural design relationships for new and developing concretes. Mater Struct 2005;38(1):1–9.)

ash level and this occurred at test ages up to 365 days. In other work327 considering fine and coarse fly ashes, similar modulus of elasticity values were obtained between these in concretes designed for equal compressive strength (28 days). Results from tests on concrete containing various cement combinations320 and proportioned by different means are shown in Fig. 9.51 and give agreement with compressive strength. Research investigating silica fume in concrete of w/c ratio 0.6316 indicates that with 5% inclusion of silica fume in cement, the modulus of elasticity increased from 30 to 33 GPa (compressive strength increased from 41.0 to 46.5 MPa). Further increases up to 20% silica fume gave compressive strength increases of 7.5 MPa, with modulus of elasticity only changing by 1.0 GPa. In the same study metakaolin was found to show similar type behaviour for levels in cement up to 25%. Other work322 covering a range of aggregates found that the modulus of elasticity increased on average by 16% and 32% with the inclusion of 10% and 15% silica fume in 0.35 w/c ratio concrete. In relatively high strength, equal w/c ratio concretes,328 small increases in the property were found with metakaolin levels up to 15%, while noticeable increases in compressive strength were obtained.

9.5.5 Shrinkage and Creep Shrinkage of concrete is relatively insensitive to the type of cement, since it depends mainly on the percentage by volume of water in concrete329 and the type of aggregate.330 In concretes containing a range of different aggregates and of similar strength,317 natural pozzolana cement gave slightly higher drying shrinkage than the PC reference. This seemed to reflect the cement content and w/c ratios required to achieve strength. It is suggested that between the two natural pozzolana concretes considered, behaviour was influenced by the paste content, reactivity of the materials and nature of hydration products. The shrinkage of 12 concretes made with 11 cements containing 11 types of fly ash with the reference PC was between 280 and 500
106 after 224 days of exposure to a dry environment. They had previously been water-cured for 91 days. The shrinkage of the plain cement sample was 453
106. The range was only slightly greater when the specimens were cured in water for 7 days (Table 9.30).331 These results are comparable to those obtained with PC concretes using cement from a range of different sources. The mixes, which had a constant water/cement ratio and mix composition, gave shrinkage values ranging from 150
106 to 420
106 after 14 days of curing.332 Shrinkage has been found to depend on the type of fly ash used, with concretes made with material from different sources, using the same PC and of approximately similar 365-day strength giving values ranging from 497 to 678
106.333 Other work testing a fine and coarse fly ash (3.3% and 29.7% retained on a 45 mm sieve) in equal strength concrete (20, 40 and 60 MPa) found ultimate shrinkage to be lower with the coarser material by 60–85
106.327 It was suggested that this could be related to the lower w/c ratio used with this fly ash to achieve equal 28 day strength. Shrinkage is sensitive to the level of fly ash334,335 when initial curing of concrete is limited to 3 days, with increases noted with fly ash content in this case. In contrast, shrinkage is not affected with up to 40% fly ash when concrete is cured for 28 days prior to drying.335

422 Lea’s Chemistry of Cement and Concrete

TABLE 9.30 Drying Shrinkage of Fly Ash Concretes331 Shrinkage Measurements Initially Cured for 7 Days in Water

Initially Cured for 91 Days in Water

Mixture No.

Duration of Drying (Days)

Moisture Loss (%)a

Drying Shrinkage (×10–6)

Moisture Loss (%)a

Drying Shrinkage (×10–6)

Control 2 F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11

224 224 224 224 224 224 224 224 224 224 224 224

55.0 57.5 57.3 56.9 54.7 58.8 60.6 64.3 56.3 58.2 58.4 49.5

422 447 364 411 379 404 475 397 400 390 642 454

53.7 47.9 45.4 56.2 49.2 51.1 56.4 54.1 – 49.3 55.2 48.9

453 365 280 405 387 403 454 433 327 361 500 362

a

As a percentage of total original water.

Tests on concrete incorporating silica fume and metakaolin at various levels indicate reduced drying shrinkage with these materials.328,336–338 In general, the effects increased with the level of pozzolanic material. This appears to reflect a denser microstructure, increased water uptake internally and less moisture loss compared to PC concrete. Creep is related to concrete strength at the time of loading and so, all other conditions being equal, it depends on the strength of cement, the water/cement ratio and the curing period. Since the use of pozzolanic material in cement can lower the early strength of concrete, the specific creep of cement may be greater than that of PC concrete if loading has been applied too early (Fig. 9.52).335 The difference, however, tends to reduce as curing proceeds, analogous to that which happens with differences in strength. For these reasons, creep of concretes containing blended cements can be lower or higher than that of the control concrete, depending on the strength of concrete and the level of pozzolanic material. Results of several studies tend to suggest that wellcured fly ash concrete exhibits lower creep than PC concrete.339 Benefits associated with silica fume and metakaolin, which had higher compressive strengths than their PC references, have also been noted.328,339

FIG. 9.52 Effect of loading age and pozzolanic material on creep. (From: Sri Ravindraraja R, Tam CT. Properties of concrete containing low-calcium fly ash under hot and humid climate. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. I. p. 139–55.)

Pozzolanas and Pozzolanic Materials

423

9.6 TRANSPORTATION PROPERTIES OF CONCRETE The transportation of fluids into concrete has direct relevance with regard to most aspects of concrete durability. As noted in Section 9.4.4, the principle mechanisms by which this occurs can be divided into absorption, permeability and diffusion. Apart from the presence of aggregate grains, the microstructure of mortar and concrete containing pozzolanic material differs only slightly from that of pastes. Differences have been noted in the characteristics of material at the interface between cement paste and aggregates than in the bulk paste,340 which can affect the processes. Similarly, the absorption properties of aggregates and the quantity of these materials in concrete can also have an influence.341 SEM observations have shown that the consumption of calcium hydroxide by pozzolanic materials results in a reduction of the number and size of the portlandite crystals at the aggregate-paste interface, as well as generally in the bulk paste. A reduction in free lime is associated with a decrease in the number of flaws in the mortar and improvement in the homogeneity of the matrix. At the interface, the c-axes of the calcium hydroxide crystals are more or less parallel to the aggregate surface and some fly ash particles appear to be encapsulated by these crystals.342 Silica fume has been found to introduce some differences in the microstructure of the interfacial zone of mortars, which becomes dense and uniform, lacking in rims of portlandite or gaps. Practically no border line can be drawn between the bulk cement matrix and the interfacial zone.343 A dense structure has been ascribed to the presence of 15% silica fume in cement; however, contributions from superplasticiser used at relatively high dosage compared to control concrete are likely. Superplasticisers reduce the initial water film around aggregate grains irrespective of the type of cement344 and this can account for the dense structure of the interfacial zone observed in silica fume-containing concrete. Enhancements in the pore structure of transition zones have also been found in other work on silica fume and metakaolin in concrete, with improved microhardness and reduced width of this region noted.345 It has been suggested that the structure of the interface is influenced by the SiO2/CaO contents present in the pozzolana (higher/lower generally giving benefits).346 A range of techniques have been used to carry out measurements of the transportation properties.347 Tests for water permeability can take time to carry out, leading to changes in the pore structure of the material being considered. Similarly, tests measuring oxygen or air permeability or involve absorption require pre-conditioning by drying, which can lead to microcracking of concrete. They, however, provide a means of assessing its structure and their use with site concrete has been described.347 Air permeability and oxygen diffusion coefficients tend to decrease with time and increase with water/cement ratio.348 In concrete, water cured for 28 days, the coefficient of water permeability decreases with increasing strength according to a linear relationship.349 A general relationship between oxygen permeability, as well as the diffusion coefficient, and compressive strength of concretes has been found which is independent of the type of cement, the water/cement ratio, the cement content and the curing time.348 The relationships are of the form: y ¼ AxB

(9.1)

where y is coefficient of permeability or diffusion, x is compressive strength and A and B are constants. Research indicates348 that a point is reached for both oxygen permeability (see Fig. 9.53) and diffusion, where they increase rapidly when the concrete strength reduces below a certain point. As far as the influence of pozzolanic materials on the permeability of concrete is concerned, tests at 60 days investigating natural pozzolana (at 20% in cement and to partially replace sand)324 found minor differences in gas permeability (AFPC-AFREM procedure, Cembureau) between these and the reference when the concretes were of approximately similar strength. Fig. 9.54 shows that regardless of the drying method, fly ash concrete (50% level in cement) has a significantly lower oxygen permeability than a comparable PC concrete of the same age (30 days).350 Other studies investigating the effect of fly ash level in cement on air and water permeability for concretes of equal strength at 28 days, established enhanced properties as these increased up to 30%.351 Benefits were also noted with the use of a rapid hardening cement/fly ash combination, particularly with poor curing. Fly ash levels of 45% in equal 28 day strength concretes containing rapid setting cements, have also been found to reduce intrinsic (air) permeability, compared to that of PC concrete.299 Improvements associated with the use of silica fume in concrete have also been noted.352 Tests in concretes covering a range of w/c ratios found that 10% silica fume reduced oxygen permeability, compared to that of PC and this was observed consistently at test ages up to 180 days. Similarly, a water penetration method found that with 10% silica fume in 0.35 w/c ratio concrete (both with normal and lightweight aggregate) water depths measured were reduced by about half to two-thirds that of the reference.353 Increasing silica fume contents beyond 10% to 12% appear to give little further benefits.354 The same type of effect has been noted with metakaolin in concrete at 0.38 w/c ratio, where increasing reductions in water penetration and gas permeability (nitrogen) were noted up to about 15% in cement.355

424 Lea’s Chemistry of Cement and Concrete

FIG. 9.53 Relationship between compressive strength and oxygen permeability coefficient of concretes. (From: Costa U, Facoetti M, Massazza F. Permeability and diffusion of gases in concrete. In: Proceedings of the ninth international congress on the chemistry of cement, New Delhi; 1992, vol. V. p. 107–14.)

FIG. 9.54 Effect of drying method on oxygen permeability results. (From: Day RL, Konecny L. Relationships between permeability and microstructural characteristics of fly ash mortars. In: Materials Research Society symposium proceedings; 1989, vol. 137. p. 391–2.)

In addition to w/c ratio and age, sorptivity depends on the type of cement used. Table 9.25 shows that the presence of pozzolanic material reduces the sorptivity of both pastes and mortars and increases porosity. Water sorptivity has also been noted to decrease with concrete strength,356 with a relationship noted with strength, as follows: pffiffiffi (9.2) S ¼ S°  a F where S is sorptivity, F is compressive strength and S° and a are constants. Eq. (9.2) is applicable to concrete cured from 1 to 28 days. S° and a depend on the materials used. Sorptivity tests on natural pozzolanas (capillary rise and initial surface absorption (ISA, measured at 10 min)) suggest both improvements,357 and similar/higher values316 compared to PC concretes at equal w/c ratio. In the latter case, this corresponded to reductions in compressive strength (with increasing natural pozzolana level in cement). The ISA of concrete reduces with increases in the fly ash content in cement. However, the differences are relatively small.358 On the other hand, according to Fig. 9.55 ISA systematically decreases by increasing the strength of concrete.358

Pozzolanas and Pozzolanic Materials

425

FIG. 9.55 Effect of water curing (E2) and air curing (E7) on initial surface absorption at 10 min (ISA-10) of fly ash concretes: (A) 28 days, (B) 90 days, (C) 180 days. (From: Dhir RK, Byars EA. PFA concrete: near surface absorption properties. Mag Concr Res 1991;43(157):219–32.)

The decrease, which is small in water-cured concrete, is significantly greater in air-cured concretes (55% relative humidity (RH)), so that curing appears to be an important factor influencing the sorptivity of concrete. Studies investigating silica fume and metakaolin in concrete, using various absorption/sorptivity type tests316,359,360 indicate that the inclusion of these materials gives reduced absorption and this tends to increase with the level in cement. These effects also generally correspond to the influence of the materials on compressive strength. Differences have been noted for the effects of metakaolin on the trends between capillary rise and total immersion absorption tests.360 Overall, the results concerning the transportation properties suggest that the type of cement may have less of an influence on these properties than might be expected, given the differences found among pastes from different types of cement (Section 9.4.3). A possible reason is that the fluid flow is influenced by the porosity of the interface between the paste and aggregate, rather than the bulk paste. Fig. 9.56 shows how the porosity of concrete noticeably increases as the interface is approached, and this occurs with both cements containing pozzolanic material (in this case, silica fume) and PCs.361 Although, other work has questioned the impact of this effect on transportation processes.362

FIG. 9.56 Variation in paste porosity with distance from aggregate. (From: Scrivener KL, Gartner EM. Microstructural gradients in cement paste around aggregate particles. J Am Ceram Soc 1988;114:77–85.)

426 Lea’s Chemistry of Cement and Concrete

9.7 DURABILITY PROPERTIES OF CONCRETE Durability is the ability of concrete to maintain a required level of serviceability over its intended design life. It depends on both concrete properties and environmental conditions. The chemical composition of the cement and aggregate, the microstructure of the hardened cement matrix, the chemical composition of air, water and soil surrounding the concrete, temperature and its variation are all factors which may play a role in influencing durability. In the following sections the behaviour of cements containing pozzolanic materials is compared with that of PCs. It should be remembered that durability will usually depend on the entire set of properties of hardened concrete and its constituents rather than solely those of the cement. Durability is normally considered at the design stage of a construction project, with the specifications usually given in terms of parameters including cement type, minimum cement content, maximum w/c ratio, minimum strength and cover to reinforcement for the particular exposure conditions.363 In some cases, other factors, e.g. aggregate properties, airentrainment, are also covered.

9.7.1 Carbonation and Chloride-Induced Corrosion Steel reinforcement will normally be protected against corrosion when embedded in concrete through a combination of effects. These include the high pH environment at the steel surface which allows the formation of a passive oxide layer protecting the steel, the physical barrier provided by the structure of concrete, and high electrical resistivity (and hence ion mobility) of the zone around the steel. However, the protective nature of concrete can be compromised through interaction with substances in the environment. The substances of most concern are carbon dioxide (leading to carbonation) and the ingress of chloride ions.

9.7.1.1 Carbonation Carbon dioxide is present in air at levels of around 0.04% by volume and is potentially an issue for concrete durability inasmuch as it can react with the hydration products in the hardened cement paste. Studies of partially carbonated PC paste, using MAS-NMR, have revealed that the CO2 from air occurs in the paste mainly as crystalline CaCO3. However, a significant part is also present in a non-crystalline form and is probably incorporated in the C-S-H structure.364 The process has also been noted (with high CO2 concentrations) to change the density and mechanical properties of concrete.365 The depth of carbonation attains maximum values at RH levels of between 50% and 70%. In practice, with the climatic conditions in many countries, the degree of carbonation reduces with increasing humidities.366 Similarly, atmospheric CO2 concentrations and temperature also influence the process.367 Carbonation is a risk for reinforced concrete because when the front penetrates the concrete cover, it reduces the alkalinity of the pore fluids to an extent that the passive conditions break down and corrosion may occur. The development of this can lead to the cracking and spalling of the concrete cover, with damage to elements or structures. The depth of carbonation in concrete increases with time, but at a reducing rate, and in addition to the environmental factors, noted above, depends on:    

microcracks cement content water/cement ratio curing period

In general, the higher the air permeability of concrete, the greater the depth of carbonation (accelerated) at a given time and this has been found to apply irrespective of mix proportions and curing.368 Prolonged curing before exposure to carbon dioxide reduces the depth of carbonation of mortar and concrete. Indoor exposure of concrete can result in a depth of carbonation between two and four times greater than that recorded for outdoor exposures,369 which relates to differences in concentrations of CO2 and ambient humidities. Fig. 9.57 shows that carbonation increases when samples are sheltered from rain.370 Blended cements are considered to have reduced resistance to carbonation compared to PC owing to their lower portlandite content (cf. Section 9.4.3), and this has been noted in concretes or mortars of equal or similar w/c ratio.316,371 However, studies have shown the depth of carbonation to be statistically the same in concretes made of Portland or natural pozzolana- and fly ash-containing cements, provided the comparison is made between concretes having the same 28-day strength.369,372–374 Fig. 9.58 shows an example of this, comparing 1-year carbonation depth of concrete containing Portland and fly ash cements, which decrease linearly with increasing 28-day strength.375 The degree of carbonation in concretes containing 15%–30% fly ash in cement is approximately the same as that found in PC concretes.376 However, the rate of carbonation of concretes containing 50% fly ash has been found to be higher than in that of PC of the same strength.377

Pozzolanas and Pozzolanic Materials

427

FIG. 9.57 Influence of concrete compressive strength on the expected final carbonation depth (Outdoor carbonation). (From: Nischer P. Einfluss der Betongute auf die Karbonatisierung. Mitteilungen aus FI VDZ; 33, Dauerhaftigkeit von Beton; 1984.)

FIG. 9.58 Relationship between 28 day water stored strength and carbonation depth of concrete after 1 year, internal exposure, cured for 3 days. r ¼ –0.71 for series A only and for both series combined. (From: Barker AB, Matthews JD. Concrete durability specification by water/cement or compressive strength for European cement types. In: Malhotra VM, editor. Proceedings of the third international conference on durability of concrete, Nice. American Concrete Institute Special Publication 145; 1994. p. 1135–59.)

For a given compressive strength (measured at the time of carbonation depth tests), carbonation appears to be higher in mortars and concretes containing pozzolanic materials than in those with PC.378 Fig. 9.59 shows how in this case the relationship between depth of carbonation and strength is better expressed by two straight lines, the upper relating to cement containing pozzolanic materials and the lower to PCs.378 However, it should be recognised that the differences shown in Fig. 9.59 (1–2 mm) are minor compared to those caused by different exposure conditions. Owing to the slow rate of pozzolanic reaction, carbonation of cements containing pozzolanic materials is greater when the curing period is short,379 less than 7 days for example,296 when the property development of blended cement pastes is less than that of the corresponding PC. The rate and depth of carbonation depend on the fineness of the cement since this affects the degree of hydration and transportation through the paste. The simultaneous grinding of fly ash and PC proved to be a suitable method for improving the carbonation resistance of mortars and concretes, whereas the addition of fly ash or other pozzolanic materials to the mix at the batch plant does not give such good results.380 Carbonation studies of cements containing pozzolanic materials have also used accelerated techniques.381 A standard accelerated procedure involves water curing (28 days), drying (14 days in laboratory air) and carbonation of specimens in a CO2-enriched atmosphere (4.0% CO2, 20°C and 55% RH) for at least 70 days.382 It has been demonstrated that this type of testing is a suitable approach for ranking concretes, including those with various cements, in terms of their carbonation resistance.383 Tests on natural pozzolana using the AFPC-AFREM method (50% CO2) found similar carbonation depths at equal w/c ratio for PC concrete and that with 20% natural pozzolana, exposed after 60 days curing. Higher depths were noted where this was also used to replace part of the sand.324 Work examining fly ash at levels of 15%–30% indicates little difference in

428 Lea’s Chemistry of Cement and Concrete

FIG. 9.59 Relationship between compressive strength and depth of neutralisation measured on samples having the same age (2 years). (From: Tsukayama R, Abe H, Nagataki S. Long-term experiments on the neutralisation of concrete mixed with fly ash and the corrosion of reinforcement. In: Malhotra VM, editor. Proceedings of the seventh international congress on the chemistry of cement, Paris; 30 Jun.–5 Jul. 1980, vol. Ill: IV-30–35.)

carbonation depth over a range of strengths compared to PC at the low (15%) level. Slightly higher carbonation depths were noted at the higher fly ash level at low design strength (up to about 35 MPa) but little difference thereafter up to 70 MPa.383 Similar effects have been noted in other work investigating 45% fly ash in concrete, with differences tending to be less with increasing strength.299 The effect of silica fume on carbonation also depends on the properties of concrete. The addition of 10% silica fume to a PC slightly decreased the depth of carbonation of lean mixes and increased that of rich ones.384 However, when mixing water was reduced by a superplasticiser in order to keep workability constant, an addition of 10%–20% silica fume reduced carbonation depth.384 The depth of carbonation of both mortars and concretes has been found to increase with the percentage of silica fume in cement.385 The relationship between 28-day strength and carbonation depth is better expressed by a family of curves, each of them corresponding to a specified silica fume level in cement.385 Studies of natural pozzolana, silica fume and metakaolin have noted different relationships between strength and carbonation depth depending on the level of addition used in concrete.316 The corrosion associated with carbonation has been investigated in a few studies. These indicate that the environmental conditions, in particular the RH has an important influence on the process.386,387 Conditions tending to promote carbonation (optimum around 50%–70% RH) and those for high corrosion rates (>90% RH) are therefore different. Increases in corrosion rate with fly ash and silica fume have been noted in some studies compared to those containing PC in concretes of equal w/c ratio (for the latter at >10% silica fume levels), following carbonation.388,389 These may correspond to different residual alkali levels between concretes following carbonation.

9.7.1.2 Chloride Ingress Chloride can harmfully affect the durability of concrete in structures, mainly due to its corrosion-causing effects on reinforcement. Therefore, the Cl– contents of the concrete constituents are usually controlled and measures taken to minimise transportation from the external environment. Of these sources, it is the latter, normally in coastal or highway exposures, that is likely to represent the main threat to concrete structures in service. Chloride can enter concrete by several mechanisms, depending on the local conditions,390,391 mainly by absorption, permeability and diffusion, and the ability of concrete to resist these depends on its microstructure and capacity to adsorb or chemically bind ingressing ions.392–394 Factors including the fineness and densifying effects, and alumina phases have been identified as contributing to this,395 with the balance of these effects likely to change between materials. Much research has been concerned with establishing rates of chloride ingress for a range of different materials including blended cements, into concrete. Studies have used various exposure conditions and test techniques, with many adopting methods of acceleration, in order to reduce test durations, and providing various measures of chloride transportation into concrete. Other work has concentrated on the modelling of this process towards estimating effects on concrete structures and service life.396–398

Pozzolanas and Pozzolanic Materials

429

FIG. 9.60 Chloride penetration in (A) Portland and (B) blended cement concretes. (From: Diamond S. Chloride concentrations in concrete pore solutions resulting from calcium and sodium chloride admixtures. Cem Concr Aggreg 1986;8:97–102.)

Investigations of natural pozzolanas,399 fly ash,400–402 silica fume403–405 and metakaolin406,407 have found these to reduce the depth of chloride penetration or rate of chloride diffusion in mortar or concrete. Fig. 9.60 shows that the depth of penetration of chloride into blended cement concrete is about 15% lower than into PC concrete.408 The difference is smaller than in pastes (around 40%)409 since the interfacial transition zone between the aggregate and cement paste phase is more permeable than the bulk material. In studies considering PC and fly ash concretes in coastal exposures, the latter gave slight reductions in chloride levels at the surface, which increased with fly ash level. While the profiles were relatively flat for PC, see Fig. 9.61,410 noticeable reductions in chloride level were obtained with depth for fly ash concretes. It appears that chloride penetration from sea water near the surface depends mainly on porosity and transportation properties, while in the bulk material it is controlled by other factors, such as chloride binding or ion exchange. These depend on the type of cement,411 and may explain the observed behaviour. 5.0

Chloride (% cement)

4.0

3.0 PC 15% P1 2.0

30% P1 50% P1

1.0

0.0 0

5

10

15

20

25

30

Depth (mm) FIG. 9.61 Comparison of chloride profiles for PC and fly ash concretes (C45 Concrete, with 15, 30 and 50% fly ash levels in cement, 10 years sea water (tidal) exposure). (Reproduced with permission from: Thomas MDA and Matthews JD. Performance of PFA concrete in a marine environment––10-year results. Cem Concr Compos 2004;26:5–20.)

430 Lea’s Chemistry of Cement and Concrete

With increasing fly ash fineness reduced chloride transportation rates412,413 have been noted in concretes compared at equal or similar strengths. Where the effects of the pozzolanic material level have been studied, benefits as the quantity of these in cement are increased (in some cases up to a point) have generally been found. Chloride is a potential threat to durability if it penetrates the concrete cover depth and reaches the location of steel reinforcement in sufficient quantities, that is, at a threshold level, with the process initiating thereafter. This can be expressed by several means, for example, Cl/OH ratio, Cl content by mass cement, or mass of concrete. It has been suggested414 that the total chloride ion content with respect to the potential inhibitor (cement) content is the best approach. In a review examining chloride threshold levels,415 it is has been noted that there are often contradictary findings between studies for the effect of fly ash and silica fume. It is considered that the transfer of available information from tests carried out for this, to practical situations is not possible. Benefits have generally been noted with blended cements in mortar and concrete with regard to the reinforcement corrosion process and rate at which this occurs,416–421 with the level of pozzolana/pozzolanic material in cement having an effect in some cases. In addition to the quantity of chloride present, the behaviour appears to relate to factors associated with the blended cement used including their influence on the microstructure, chloride migration, and electrical resistivity. Fig. 9.62 shows that the increase in the electrical resistance of mortar caused by exposure to sea water is smaller in PC mortars than in fly ashcontaining mortars and that the overall resistance given by fly ash cement was five times greater than that of PC.422

FIG. 9.62 Effect of sea water upon mortar electrical resistance for (A) PC and (B) PC + 35% fly ash. (From: Buenfeld NR, Newman JB. The permeability of concrete in a marine environment. Mag Concr Res 1984;36(127):67–80.)

Pozzolanas and Pozzolanic Materials

431

In addition to the effects noted on steel reinforcement, chloride dissolved in waters tends to increase the rate of leaching of portlandite,399,423–425 thus increasing the porosity of mortar and concrete.426 As a consequence of this, concrete can swell, lose stiffness and strength, and become more sensitive to other environmental attacks (e.g. sulfate, frost, etc.). After 100 weeks of exposure to a 4% NaCl or 3.2% NaCl + 0.8% CaCl2 solution, the strength loss of a PC mortar was about 15% and 20%, respectively, while swelling increased by 0.6%.423 However, PC concrete broke down in a concentrated CaCl2 solution at a temperature of 20°C or lower. The breakdown has been attributed to the formation and crystallisation of complex salts containing CaCl2, Ca(OH)2 and/or CaCO3.424 In mortars containing 30% fly ash, 10% silica fume and 50% GBBS in cement, improvements in compressive strength were noted following water-curing for 28 days and then storage in a 30% calcium chloride solution at temperatures of 5°C, 20°C and 40°C for 91 days. Strength increases were found in all cases, except for fly ash mortar stored at 5°C which was possibly due to the slower rate of pozzolanic reaction at low temperature. This is supported by the strength recovery observed in the same samples stored in calcium chloride solution at 40°C.426

9.7.2 Chemical Deterioration Chemical attack of concrete itself can take a number of forms, and may involve substances originating outside or within the material. The three most significant types of chemical attack are sulfate attack (caused by water containing dissolved sulfate compounds coming into contact with concrete), alkali–aggregate reaction (usually caused by internal sources of sodium and potassium reacting with certain aggregate constituents), and attack from acidic substances.

9.7.2.1 Sulfate Attack Sulfate salts are harmful to concrete,427 as they can cause expansion, loss of strength and eventually transform the material into a mushy mass. CaSO4 reacts with calcium aluminate hydrates, forming expansive ettringite (3CaO Al2O33CaSO432H2O). Na2SO4 reacts with calcium hydroxide and forms expansive gypsum (CaSO42H2O) which in the presence of aluminates, may in turn give ettringite. MgSO4 reacts with all cement hydration products, including C-S-H, forming Mg(OH)2 (brucite) and gypsum which, at a later stage, can give ettringite. Additionally, MgSO4 attack results in the replacement of Ca in C-S-H with Mg, leading to a loss in strength.428 With any source of sulfate, where carbonate or bicarbonate and silicate ions are also present, and typically at temperatures less than 10°C, thaumasite (CaSiO3CaCO3CaSO415H2O) can form, with a consequent loss in strength. Given its important role in sulfate attack, a reduction in the level of aluminate in cement tends to reduce susceptibility.429 However, there is only a loose correlation between aluminate content and deterioration rates resulting from sulfate exposure. There are a number or reasons for this:  Aluminates already present as ettringite in the cement paste will not contribute towards sulfate attack. This has been clearly demonstrated in experiments where additional gypsum is added to PC, thus ensuring the presence of residual ettringite in the mature cement paste.430  Whilst some pozzolanic materials (for instance, fly ash) will actually increase the alumina content of a cement relative to the PC, the pozzolanic reaction normally has the effect of incorporating aluminium into the C-S-H structure (see Section 9.3.3), thus preventing its involvement in the formation of ettringite. Reduction of portlandite in the hydrated cement matrix will also tend to enhance sulfate resistance, by limiting calcium available for ettringite and gypsum formation. Inclusion of pozzolanic materials will achieve this, both through dilution of PC and the consumption of portlandite through pozzolanic reaction. The sulfate resistance of pastes, mortars and concrete is generally quantified by measuring changes in length, weight and strength occurring in specimens stored in appropriate solutions. Figs 9.63 and 9.64 illustrate the beneficial effect of natural pozzolanas and fly ash in reducing expansion resulting from sulfate attack.230,431 The rate of sulfate attack is also strongly dependent on the strength and porosity characteristics of concrete. Porosity is considered to affect the rate of deterioration to a larger extent than portlandite and aluminate contents.432 It should be stressed that whilst pozzolanic materials tend to improve resistance to sulfate attack, this takes the form of a reduction in the rate of deterioration, rather than a total prevention of failure. Nonetheless, their use can considerably extend the service life of concrete structures in sulfate-bearing environments.433,434 The mechanisms of deterioration differ quite substantially depending on the type of sulfate salt solution. The two most common are sodium and magnesium sulfate. Therefore, these two forms of sulfate attack are discussed separately below.

432 Lea’s Chemistry of Cement and Concrete

FIG. 9.63 Effect of substituting PC for natural pozzolana on the expansion of 1:3 mortar. Samples 20
40
250 mm stored in 1% MgSO4 solution. (From: Massazza F, Costa U. Aspetti dell’attivita’ pozzolanica e proprieta’ dei cementi pozzolanici. Il Cemento 1979;1:3–18.)

FIG. 9.64 Expansion of mortars stored in 0.31 N sulfate solution, with different types of fly ashes and quartz powder replacing 40% of PC. (From: Schiessel P, Haerdtl R. Relationship between durability and pore structure properties of concrete containing fly ash. In: Khayat KH, Aitcin PC, editors. Proceedings of the P. K. Mehta symposium on durability of concrete: part of the proceedings of the third international conference on durability of concrete, Nice; May 1994. p. 99–118.)

Na2SO4 Attack Natural pozzolanas have been known for some time to increase sulfate resistance of cement. In one study, three PCs with C3A contents ranging between 9.4% and 14.6% were combined with 30% volcanic glass and used in mortars exposed to a 0.353 M Na2SO4 solution. All three cements displayed reduced expansion, with the mortars surviving 1 year of exposure with <0.1% expansion—a value which is often considered a threshold above which unacceptable damage is caused.435 Generally, the strength of natural pozzolana mortars exposed to sodium sulfate (0.15%–0.30%) solutions increases in the first week, but declines thereafter.435 Fly ash also improves the resistance of concrete to Na2SO4 attack.436 This effect is not simply the consequence of dilution of PC, since replacement with sand increases expansion.432 Unlike PC, whose sulfate resistance is relatively insensitive to the duration of curing, the effectiveness of fly ash is highly dependent on curing time, with longer curing periods yielding higher resistance.437,438 Exposure of fly ash concrete and mortar to sulfate solutions will often result initially in an increase in strength, followed by a decrease.433,439 This is shown in Fig. 9.65 for flexural strength and Fig. 9.66 for compressive strength. It is worth noting

Pozzolanas and Pozzolanic Materials

433

FIG. 9.65 Flexural strength of mortar prisms cured in water for 21 days and then stored in sodium sulfate solution. PC with 7.7% C3A, low-lime fly ash 2800 (FAI) and 4200 (FA2) cm2/g (Blaine) fineness. (From: Irassar F, Batic O. Effects of low calcium fly ash on sulfate resistance of OPC cement. Cem Concr Res 1989;19:194–202.)

FIG. 9.66 Changes in compressive strength of concretes for cement contents of (A) 300 and (B) 400 kg/m3 and different levels of fly ash; cylindrical specimens stored for 14 days in sealed condition at 20°C and then immersed in a 10% Na2SO4 solution. (From: Stark D. Longtime study of concrete durability in sulfate soils. In: George Verbeck symposium on sulfate resistance of concrete. American Concrete Institute Special Publication 77; 1982. p. 21–40.)

that in the second figure, a number of higher strength concrete mixes show no sign of deterioration even after 24 months exposure to sulfates. However, it should be stressed that deterioration will ultimately occur. In the presence of low-calcium fly ash, the formation of ettringite and gypsum is delayed, with a resulting delay in crack formation. This is because pozzolanic reactions reduce both the portlandite content and the rate of mass transport through the paste.433 SEM examination of fly ash cement mortars stored in a sodium sulfate solution reveal that ettringite crystals develop in voids and that initial filling is accompanied by an increase in flexural strength of the mortars. Ettringite crystals continue to grow and eventually cause cracking and strength loss. After exposure for 1 year, deterioration had reached the point of failure, and SEM examination revealed ettringite crystals filling the voids with dimensions of 15–25 mm by 2–3 mm. Gypsum crystals were localised in cracks where they had formed blocks with diameters of between 30 and 150 mm.433 Comparative sulfate attack studies of low-calcium fly ashes from different sources indicate only minor differences in expansion values.431 One factor which appears to influence the relative performance of fly ash is particle size, with finer fly ashes performing better. This can be attributed to a refinement of the pore structure of the cement matrix by the original particles and further refinement as a result of the formation of products of pozzolanic reaction.440 It has been reported that concrete exposed to 10% Na2SO4 solution generally expands to a lesser extent when fly ash is interground with clinker and gypsum, rather than added to cement at the time of batching, presumably leading to finer, more reactive ash particles.441

434 Lea’s Chemistry of Cement and Concrete

TABLE 9.31 Physical Properties of Fly Ash and Silica Fume Mortars Before (28 Day) and Following Immersion in Water and Sulfate Solutions for 1 Year450 28 Day Water/ Cement Ratio Plain mortar Fly ash

Silica fume

10% 30% 50% 70% 5% 10% 20% 30%

Compressive Strength (MPa)

Total Pore Volumea (10–3 cm3/g)

1-Year Total Pore Volume (10–3 cm3/g)

Water b

10% Na2SO4

10% MgSO4

0.55

45.3

39.8 (17.6)

33.8 (22.1)

94.8 (83.3)

25.2 (36.1)

0.54 0.52 0.50 0.48 0.55 0.54 0.54 0.53

37.1 35.8 23.8 11.7 44.4 45.1 46.6 50.2

22.4 (12.5) 44.5 (15.1) 42.8 (16.5) 62.1 (34.6) 50.7 (13.8) 36.1 (11.6) 29.0 (20.7) 28.4 (21.1)

26.3 (34.6) 33.2 (15.4) 38.6 (13.2) 58.5 (10.6) 30.1 (11.6) 44.7 (22.6) 33.8 (17.8) 17.8 (36.5)

— 28.7 (20.9) — — — 46.8 (24.6) — 27.2 (31.2)

25.8 (32.9) 22.9 (38.4) 19.9 (42.7) 43.5 (19.5) 45.6 (28.9) 22.4 (21.4) 55.7 (57.6) 68.6 (87.5)

a

Measured by MIP, expressed as cm3/g of mortar. Figures in parentheses are ratios of pore volume larger than 0.1 mm to total pore volume (%).

b

In hardened concretes exposed to the attack of a 5% Na2SO4 solution, the depth of the sample into which ettringite crystals are observed decreases with increasing content of fly ash up to 50%.442 Although a number of high-calcium fly ashes have been found not to improve the sodium sulfate resistance of cement significantly,443 exceptions have been reported.444 Silica fume also increases the resistance of PC to sodium sulfate attack, with higher levels in cement leading to reduced expansion.435,445 Low (7%) silica fume levels have been deemed insufficiently effective at reducing expansion of mortars.446 However, higher levels (10%–15%) have been noted to improve the resistance of mortars to sulfate attack.427,447 Generally, higher levels of silica fume are necessary to adequately enhance resistance to higher Na2SO4 concentrations.448 The physical and chemical characteristics of silica fume play an important role. A 10% level of silica fume, with a high fineness and silica content, has been shown to be sufficient to keep expansion below 0.1% after 365 days of exposure, whilst 15% was required for a material with a low silica content (80%) and a relatively low specific surface area (8.75 m2/g).449 Similar results are seen in Table 9.31, which shows the development of porosity in mortars stored in water, 10% Na2SO4 or 10% MgSO4 solutions. Storage for 1 year in water generally reduces the total porosity of mortars. In contrast, storage in 10% Na2SO4 leads to an increase in porosity in the case of the PC and that containing 10% silica fume. A higher level of silica fume leads to a much smaller change in porosity.450

MgSO4 Attack Magnesium sulfate attack is considered to be more severe than that of sodium sulfate. One feature of this form of attack is the deposition of a double layer composed of brucite and gypsum, followed by several internal layers of gypsum. The formation of the double layer—and particularly the presence of brucite—is not necessarily detrimental, since it appears to play a role in protecting the surface of concrete and mortar from further ingress of sulfates, potentially limiting deterioration. Pozzolanic materials increase the resistance of mortar and concrete containing conventional PC but, unlike sodium sulfate attack, they sometimes lessen the performance of SRPCs. Moreover, as will be considered below, performance is very much more dependent on the type of pozzolanic material used. Both metakaolin and fly ash have been shown to be effective in improving resistance to magnesium sulfate attack. After 180 days’ exposure, a 5% MgSO4 solution caused a strength loss of 45% in a PC paste, but only a 5% and 20% loss in cements containing 30% and 40% metakaolin, respectively.425 A blend of 75% PC/25% fly ash displayed enhanced sulfate resistance relative to PC in a weak (0.75%) MgSO4 solution, but performed worse than the PC control in a stronger 3% solution.451 A 90% PC/10% silica fume blend displayed improved resistance of mortars exposed for around 3 years to a solution containing 3% MgSO4 relative to a PC control.451 However, an 85%/15% blend dramatically reduced the strength of mortars exposed to a 4.2% MgSO4 solution for 2 years (Table 9.32).427 A similarly proportioned blend containing SRPC also performed poorly.

Pozzolanas and Pozzolanic Materials

TABLE 9.32

435

Compressive Strength of Mortar Cubes 20 × 20 × 20 mm Immersed in 4.2% MgSO4 Solution427 Compressive Strength MPa 90

Type of Mortar

Type of Cementa

Days

Plain

PC SPC SRPC PC SPC SRPC

37.8 36.7 40.7 29.1 36.3 46.8

15% silica fume

180

Percentage of Water-Cured Samples 1

2

Years 12.8 20.2 36.6 22.1 20.4 17.7

17.4 8.5 33.9 12.0 12.5 11.7

90

180

Days 7.0 — 32.2 — — —

107 116 110 63 99 103

1

2

Years 61 61 93 46 48 39

40 23 83 21 29 23

16 — 72 — — —

a

PC, Portland cement; SPC, slag cement; SRPC, sulfate-resisting Portland cement.

Similar results are seen for silica fume in terms of the development of porosity as a result of sulfate attack. Immersion for 1 year in a magnesium sulfate solution has been shown to reduce porosity in plain and fly ash mortars (Table 9.31). However, whilst a 10% silica fume level provides protection, at lower and higher levels, the magnitude of deterioration is notable. This may be explained by results of studies examining the nature of reaction products in silica fume mortars compared to those made with PC, after exposure to magnesium sulfate.452,453 The observations included a thinner surface double layer of brucite and gypsum, a lower depth of sulfate penetration, a reduction in levels of precipitated gypsum, and extensive microscopic polygonal cracking at the surface. It is likely that the thinner double layer formed by brucite and gypsum accounts for the lower sulfate penetration depth, while the network of cracks reveals a general weakening of the bond among the particles and, thus, a higher mass loss.452 Where both magnesium and sodium sulfate are present simultaneously, the resulting deterioration process becomes more complex. However, it would appear that fly ash is capable of imparting resistance to attack in such conditions. For example, 20%–40% low-lime fly ash combined with PC containing 12.0% C3A has been shown to decrease the expansion of mortars exposed to a solution of 4.3% magnesium sulfate and 2.5% sodium sulfate. Moreover, the strength of the fly ash cement mortars increased with time (at least up to 6 months), whereas that of the reference PC mortar began to decrease after 2 months.454 The expansion of mortars was also reduced by combining a medium-lime (13.3% CaO) fly ash with a 12.8% C3A PC in proportions of 35% and 65%, respectively. However, expansion remained significantly higher than that of a high sulfateresistance PC (1.3% C3A).455 (NH4)2SO4 Attack There is little data available concerning ammonium sulfate attack. However, Table 9.33 427 indicates that this salt is a serious threat to concrete durability. The improvement in the ammonium sulfate resistance with pozzolanic additions does not seem to be effective when the C3A content of the PC is low—that is, when SRPC is used. Even where this is not used, Table 9.33 and the results from other studies456 indicate that the use of silica fume only improves performance to a very small degree. A 30% natural pozzolana cement has been shown to display greater resistance than a 5.45% C3A PC control when mortar specimens were immersed in 0.15%–0.30% Na2SO4 solutions, but performed worse when exposed to a 0.14% (NH4)2SO4 solution.436 Thaumasite Attack Like ettringite formation, thaumasite requires a source of calcium to be present, and portlandite is likely to be the most available form in hydrated cements.457 There also appears to be the requirement that a source of aluminium is present, despite its absence from thaumasite’s composition.458 Since the addition of a pozzolanic material will reduce the portlandite content of a hydrated paste, fly ash, silica fume and metakaolin will all limit the formation of thaumasite.

436 Lea’s Chemistry of Cement and Concrete

TABLE 9.33 Compressive Strength of Mortar Cubes 20 × 20 × 20 mm Immersed in 4.6% (NH4)2SO4 Solution427 Compressive Strength MPa

% of Water-Cured Sample

Age (Days) Mortar Control

15% silica fume

Cement typea PC SPC SRPC PC SPC SRPC

60 13.0 17.9 21.5 22.4 19.3 16.7

Age (Days) 90 10.6 6.6 15.5 15.7 9.3 14.4

150 — — 11.1 10.0 — —

180 — — 9.4 — — —

60 37 57 59 49 53 43

90 29 20 42 31 24 36

150 — — 30 19 — —

180 — — 28 — — —

a

PC, Portland cement; SPC, slag cement; SRPC, sulfate-resisting Portland cement.

It has been proposed that thaumasite formation can theoretically be controlled by reducing the Ca/Si ratio of the cement fraction of a concrete or mortar mix to below 1.0.459 This is, in reality, somewhat impractical, since this means very high levels of pozzolanic material will be required. Using this rule, the theoretical proportion of low-calcium fly ash required to achieve this ratio might well be over 55%, putting it outside the range of a pozzolanic cement as defined in EN 197-1. Similarly, metakaolin and silica fume might require levels in excess of 30% and 45%, respectively, which lie wholly outside the levels of normal use. However, in reality, effective thaumasite control has been achieved with fly ash levels of above 40%,460 whilst a level of 10% for silica fume and metakaolin appears to be adequate.461 This is presumably the result of portlandite being consumed by the pozzolanic reaction.

Sea Water Sulfate-induced deterioration of concrete in contact with sea water is revealed by expansion, followed by cracking and scaling, and a progressive decrease in the mass and volume of concrete. All of these phenomena cause mechanical properties to decline. Sea water-induced deterioration is less damaging than would be assumed from the HCO3  and Mg2+ concentrations present.462,463 The reason for this is that HCO3  and Mg2+ contained in the sea water react with the cement hydrates forming a protective layer comprising both aragonite (CaCO3) and brucite.464 Aragonite seems to play an important role, since when HCO3  is absent, the same amount of brucite precipitates, but its protective effect is reduced.465 Moreover, ettringite and gypsum are more soluble in chloride-bearing solutions466 and less stable in the lower pH conditions of the brucite layer. The combination of pozzolanic material with PC reduces expansion of mortars stored in artificial sea water, provided that the level in cement is sufficiently high. For this reason, 20% natural pozzolanas and fly ash in cement had only a slight influence on mortar expansion467 and silica fume proved to be effective only at levels >10%.468 Mortars incorporating pozzolanic materials such as fly ash have been shown to expand less than plain PC mortars of comparable strength.469

9.7.2.2 Alkali–Aggregate Reaction Since the 1940s, harmful chemical reactions between aggregates and PCs have been identified. The most common reaction is that occurring between certain types of silica (opal, chalcedony and tridymite) and cement alkalis. This reaction forms a viscous gel comprising alkali and alkaline-earth silicates which absorbs water from the environment, leading to expansion and consequent crack development. Concrete deterioration manifests itself at the surface as extensive map cracks, often with gel exuding from these, and sometimes pop-outs and spalling. A short summary of the different mechanisms suggested for explaining causes and effects of alkali–silica reaction (ASR) can be found in Ref. 470. Other similar alkali–aggregate reactions (AARs) include alkali-carbonate reaction and alkali–silicate reaction, although these occur much less widely. Three conditions seem to be essential to start and sustain AAR:  the alkali content of concrete must be high;  only part of the aggregate must be reactive;  the humidity of the environment must be high.

Pozzolanas and Pozzolanic Materials

437

Furthermore, expansion and damage increases with cement content471 and water/cement ratio.472 A number of methods have been proposed to evaluate the effectiveness of different types of cement in preventing damage caused by alkali–aggregate reaction473 but, owing to their more or less empirical nature,474 they do not always agree with or reproduce field behaviour of concrete.475,476 The expansive reaction does not occur with PCs containing <0.6% Na2O alkali equivalent.477 However, modern cement technology and pollution-prevention limits in force in many countries often make this target prohibitively expensive.478 Again, for economic reasons, the problem cannot always be solved by sourcing non-reactive aggregates. Alkalis can come from external sources (sea water or de-icing salts), and in these cases the use of a low-alkali cement may not prevent deterioration if free alkali levels are replenished from such sources.479,480 Soon after the discovery of alkali–silica reaction, it was found that the addition of fine pozzolanic material to PC could control the expansion.481 Expansion reduction has been observed when using natural pozzolanas and volcanic tuffs,482,483 both in the raw state and after thermal activation,484 fly ash, calcined shale, silica fume, kieselguhr, moler 485,486 as well as metakaolin and other burned clays.487 In Italy, despite the presence of reactive aggregates in some districts and their use in marine environments, only a few examples of alkali–aggregate reaction have been found. This could partly be due to the widespread use of blended cements in that country. The addition of pozzolanic materials as a solution to AAR is counter-intuitive, since natural pozzolanas can contain up to 11% alkalis and fly ashes up to 5%. However, when certain conditions are met, expansion is reduced and brought to within acceptable levels.485,488,489 Influence of Pozzolanic Materials on Alkali–Silica Expansion It is well known that the expansion of mortar bars containing reactive aggregates varies with the proportion of reactive aggregate. Expansion usually reaches a maximum at a certain content of reactive mineral (a ‘pessimum’ effect).471,480 Whilst the addition of pozzolanic material reduces this effect490; it may also display a similar pessimum. Fig. 9.67 shows that at a level of 10%–20% of high-alkali fly ash, expansion caused by sand containing reactive Beltane opal is increased, whereas 30% reduces expansion considerably.491 When using low-alkali fly ash, a lower level (20%) may be sufficient to suppress harmful expansion.492,493

FIG. 9.67 Mortar bars containing high-alkali PC/high-alkali fly ash with Beltane opal aggregate at 20°C. (From: Nixon PJ, Gaze ME. The effectiveness of fly ashes and granulated blastfurnace slags in preventing AAR. In: Proceedings of the sixth international congress on alkali in concrete: research and practice, Technical University of Denmark, Copenhagen; 22–25 Jun. 1983. p. 61–8.)

438 Lea’s Chemistry of Cement and Concrete

FIG. 9.68 Effect of fly ash level in cement on expansion of mortar. (From: Nagataki S, Ohga H, Inoue T. Evaluation of fly ash for controlling alkali– aggregate reaction. In: Malhotra VM, editor. Proceedings of the second international congress on durability of concrete, Montreal. American Concrete Institute Special Publication 126; 1991. p. 955–72.)

This pessimum effect for pozzolanic materials has been observed in other instances,483,494 although the effect is not a general rule, since other fly ashes combined with the same PC do not display a peak (Fig. 9.68).494 Natural pozzolanas, such as moler, have been found to also display a pessimum effect.486 Again, this is only observed in some cases: trass has been found to cause increasing expansion at greater levels in cement,495 whilst other natural pozzolanas have been shown to yield a continuous decrease in expansion with increasing level.484 Owing to the high reactivity of silica fume, the quantity necessary to prevent expansion is lower than that required by other pozzolanic materials.479 A pessimum effect for silica fume appears with certain types of reactive aggregate, but not with others.474 Metakaolin has also been shown to be highly effective at controlling expansion at low levels.225,487,496,497 Whether a pessimum level exists for this material is uncertain, since above certain levels (around 25%), their use becomes impractical in terms of the rheology of the resulting mix. The existence of a pessimum effect, with regards to the level of pozzolanic material in cement, indicates that the total alkali content (including both PC and pozzolanic material) cannot be used as a general criteria to identify expansive mixes. However, there is typically a critical molar ratio of Na2O/SiO2 for a given pairing of PC and pozzolanic material, where maximum expansion is observed. Fig. 9.69 shows that the critical Na2O/SiO2 ratio changes depending on the fly ash/PC combination.498 As a practical consequence, a risk may be introduced by the arbitrary use of fly ash with no consideration of its composition, since combining low alkali PC with high alkali fly ash, or high alkali PC with low alkali fly ash, may push the chemistry of the blended cement towards the critical ratio.498 As well as the alkali metal content of pozzolanic materials, the calcium content also appears to be important. Higher lime levels in fly ashes have been demonstrated to limit the reduction in expansion.474,499,500 Alkalis may also derive from external sources. Use of 30% fly ash in cement has been shown to prevent expansion of mortar exposed to a NaCl solution.501 Generally, expansion is reduced as the fineness of a given pozzolanic material increases (Fig. 9.70).502,503 However, whilst a number of studies have found correlations between fly ash fineness and the ability to limit ASR expansion,494,504–506 it should be noted that other studies have found very little by way of a relationship.489,499,507–510 In many cases, this may be attributable to differences in fly ash chemistry between different size fractions from the same source. The addition of pozzolanic material generally increases the total alkali content of mortar and concrete. However, the availability of alkalis depends on the progress of pozzolanic reaction and particularly on the rate at which the glass component of the material is attacked. For this reason, it has been suggested that fly ash acts like a PC with an alkali content of around 0.2%.507 This assumption means that the addition of fly ash, or of other pozzolanic materials, should reduce available alkalis and thus expansion, but it does not explain the pessimum effect. Fig. 9.71 shows that the reduction in alkali–silica expansion depends on the source of pozzolanic material.507 Whilst the obvious conclusion to draw is that this is the result of variation in the alkali content of the pozzolanic material, it has been shown that the use of two fly ashes having 3.35% and 1.32% Na2Oeq, respectively, resulted in almost the same expansion.488 Thus, other factors, such as the fineness of pozzolanic material, the available alkali content, and the rate of pozzolanic reaction, play an important role.

Pozzolanas and Pozzolanic Materials

439

FIG. 9.69 Effects of N/S mole ratio on mortar bar expansion. DAV, PC with 0.85% Na2Oeq; ND, PC with 0.49% Na2Oeq; ONT, fly ash with 7.35% Na2Oeq; OTT, fly ash with 3.39% Na2Oeq; NE4, fly ash with 2.26% Na2Oeq. (From: Lee C. Effects of alkalies in class C fly ash on alkali–aggregate reaction. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114, 1989; vol. I. p. 417–30.)

FIG. 9.70 Decrease of expansion by several kinds of additions according to their specific surface area (determined using the Blaine technique) and quantity in concrete. (From: Sprung S, Adabian M. The effect of admixture on the alkali-aggregate reaction in concrete. In: Proceedings of the symposium on the effect of alkalies on the properties of concrete, London; 1976. p. 125–37.)

440 Lea’s Chemistry of Cement and Concrete

FIG. 9.71 Variation of expansion of mortar with age. w/(c + fa) ¼ 0.53; a/(c + fa) ¼ 3.5; fa/(c + fa) ¼ 0.3 PC A containing 1.04% Na2Oeq blended with 7 fly ashes having different specific surfaces. (From: Hobbs DW. Influence of pulverized-fuel ash and granulated blastfurnace slag upon expansion caused by the alkali-silica reaction. Mag Concr Res 1982;34:83–93.)

The reduction in intensity of alkali–silica attack due to the addition of pozzolanic material has been observed directly in reactive silica glass tubes embedded in cement paste. Stereoscopic microscope observations and EDXA showed that after 28 days curing at 40°C, quartz glass was strongly corroded with the formation of a calcium potassium silicate hydrate gel. In contrast, it was not attacked by the paste containing 50% fly ash. An intermediate degree of corrosion was observed with a 30% fly ash cement.511 The use of 30%–60% fly ash in cement reduces the thickness of the reaction rim observed on the glass particles used as reactive aggregate, as well as expansion in mortar samples.512 There are cases in the literature in which fly ash and natural pozzolana have not been effective in preventing cracking and expansion due to alkali–silica reaction513 and cases in which they had the effect of increasing expansion.495,514 Both can be ascribed to an insufficient quantity, or poor quality, of pozzolanic material. Factors Reducing Expansion Some typical properties of cements containing pozzolanic materials have been associated with the reduction of expansion of mortars and concretes:     

reduced permeability and consequently lower ion mobility; lower alkalinity and pH of the pore solution; higher alkali content of hydrates; lower portlandite content; lower CaO/SiO2 ratio of C-S-H.

Permeability The reduced permeability of mortars and concrete containing pozzolanic materials, compared to that of plain PC mixes, can contribute to slowing down alkali–silica reaction, but cannot prevent it. This factor has the potential to play an important role where alkali ions enter concrete from an external source.

Pozzolanas and Pozzolanic Materials

441

Alkalinity The alkalinity of the pore solutions in PC pastes is mainly provided by sodium and potassium hydroxide,515 since the Ca2+ concentration is very low (around 0.001 M).516 The use of pozzolanic materials in cement generally reduces the alkali and OH– concentration in the pore solution.220,515,517–519 Table 9.17188 shows the effect of fly ash on the composition of the pore solution, while Fig. 9.72 shows that of silica fume.518 In some cases, high alkali fly ash516,520 and natural pozzolanas519 blended with low-alkali cements have been reported to increase pore solution alkalinity. However, high-alkali fly ashes have generally been found to reduce alkalinity when they are blended with high-alkali PC.520 Chemical equilibrium effects between the pore solution and solid phases may account for the different influence of the same pozzolana on pH in combination with different cements. The reduction of alkali concentration in pore solutions is accompanied by a reduction of expansion,515,516,518 and if the alkali concentration falls below a certain safe level, deleterious expansion does not occur. The threshold alkali or hydroxide concentration below which deleterious expansion is avoided is not yet well established. Proposed values range from 0.25518 or 0.3521 to 0.65 M.515 In concrete pore solutions, the concentrations of Na+, K+ and OH– decrease to a lesser extent than in mortar.518 Silica fume potentially possesses a large capacity for removal of alkali hydroxides from pore solutions. As an example, the OH– concentration, which was 0.5 M in a plain PC paste pore solution after 76 days of hydration, was reduced to 0.3, 0.15 and 0.075 M with 5%, 10% and 20% silica fume, respectively.522 In the presence of silica fume, the concentration of Na+ and K+ reaches a maximum after around 4 days and then decreases,518,522 seemingly stabilising after some time. However, the amount of alkalis removed from solution depends on the alkali content of the silica fume. A 5% level of a low-alkali (0.77% Na2Oeq) silica fume has been shown to reduce the alkali concentration in the pore solution, whereas the same amount of a high-alkali (3.63%) silica fume increased it. With low-alkali silica fume, the reduction of alkali and OH– concentrations is proportional to the silica fume content (Fig. 9.73).518 In general, the reduction in the alkali concentration in solution caused by pozzolanic materials is small compared to the magnitude of the reduction in expansion observed.517,521

Alkalis in the Solid Hydrates The decrease in the alkali content of the pore solution corresponds to an increase in the alkali content of the solid phases (Fig. 9.73).518 Pozzolanic materials reduce the CaO/SiO2 ratio of C-S-H and this allows more alkalis to be incorporated into its structure (Fig. 9.74).523 About 95% of the total alkali content can be immobilised in hydrates by blended cement pastes, compared to around 15% in those containing only PC.523 Analysis of solid phases has shown that potassium tends to be present in higher concentrations than sodium. The expansion caused by alkali–silica reaction appears to depend on the basicity of the resulting cement matrix, that is, on the (CaO)/(SiO2 + Al2O3 + Fe2O3) ratio,511,524 with a higher value leading to a greater magnitude of expansion. The use of pozzolanic material in cement reduces the lime content and increases the SiO2—and possibly the Al2O3 and Fe2O3—content of the blend, thus reducing basicity. This also explains why, as discussed previously, pozzolanic materials with a higher lime content are generally less effective in controlling expansion.

FIG. 9.72 Ion concentrations in pore solution at 6 months with 0%, 5%, 10% and 15% silica fume. (From: Durand B, Berard J, Roux R, Soles JA. Alkali-silica reaction: the relation between pore solution characteristics and expansion test results. Cem Concr Res 1990;20:419–28.)

442 Lea’s Chemistry of Cement and Concrete

FIG. 9.73 Alkali content at 6 months in solid phases of pastes with 0%, 5%, 10% and 15% silica fume. (From: Durand B, Berard J, Roux R, Soles JA. Alkali-silica reaction: the relation between pore solution characteristics and expansion test results. Cem Concr Res 1990;20:419–28.)

FIG. 9.74 Alkali retained in C-S-H formed in PC and blended cement pastes. (From: Bhatty MSY, Greening NR. Some long time studies of blended cements with emphasis on alkali-aggregate reaction. In: Proceedings of the seventh international conference on alkali-aggregate reaction, Ottawa; 1986. p. 85–92.)

Portlandite Content Both alkalis and Ca(OH)2 take part in alkali–silica reaction, with the presence of crystalline portlandite considered to be a necessary condition for deterioration.486,525 Thus, reduced expansion can be partly attributed to the dilution of PC with pozzolanic materials and the consumption of portlandite as a result of the pozzolanic reaction. However, it does not account for the pessimum effect. The importance of this factor has been demonstrated in experiments in which lime was added to mortar mixes containing pozzolanic materials which would have otherwise controlled harmful expansion.524,526 25% fly ash or 45% calcined clay were found to adequately control expansion in mortars containing a 1.22% Na2Oeq PC. However, the addition of 10% finely ground lime (calcined at a low temperature) caused harmful expansion (Fig. 9.75). Similarly, the addition of 9% Ca(OH)2 to mixes containing a reactive aggregate and a high-alkali (1% Na2Oeq) cement increased expansion of mortar bars by around 30%.470 20% silica fume with the same PC prevented expansion, but with the inclusion of 9% Ca(OH)2 expansion was no longer adequately controlled.470 These results have been confirmed by field observation: concrete slabs containing a reactive aggregate, but lacking free lime, have shown no sign of harmful expansion when exposed to NaCl used as a de-icing salt.525 Laboratory tests have subsequently confirmed this: by removing free lime from mortar by either combining it with a highly active natural pozzolana (moler) or leaching it with 30% CaCl2 solution, expansion was prevented.525 Despite the widely observed inverse relationship between pozzolanic material content and expansion,479,503,527 when considering results from several different pozzolanic materials, no wholly satisfactory global relationship emerges relating concrete expansion and portlandite content.528 The lack of such a general relationship is possibly due to a number of other factors.

Pozzolanas and Pozzolanic Materials

443

FIG. 9.75 Expansion of mortar bars made with aggregate containing 6% by weight of opal. 1%, 3%, 5%: 30% tuff, 30% slag and 30% fly ash, respectively. 2, 4, 6 are the same as 1, 3, 5, but with 10% CaO added. (From: Mins-shu T, Yu-feng Y, Mei-qi Y, Shi-hua Z. The preventive effect of mineral admixtures on alkali-silica reaction and its mechanism. Cem Concr Res 1983;13:171–6.)

One important point to remember is that the available portlandite content does not correspond to the total content, since a portion will be isolated by other hydrates, thus rendering it unable to take part either in pozzolanic or alkali–aggregate reactions. In this case, the porosity of mortar or concrete is likely to play a role in influencing expansion, since it will affect Ca2+ migration.

Competition Between Pozzolanic and Alkali–Aggregate Reaction The specific surface area of pozzolanic constituents affects the magnitude of expansion induced by alkali–aggregate reaction. Fig. 9.76 shows that expansion of mortar bars decreases with increasing fineness of fly ash.529 Indeed, this observation identifies an important aspect of alkali–silica reaction and pozzolanic reactions—the reaction is essentially the same, but with different outcomes depending on the particle size involved. Many reactive aggregates will display a pessimum particle size at which expansion is maximum and either side of which the magnitude of expansion declines.530 As particles become increasingly finer than the pessimum particle size, the outcome of the reaction between the aggregate and the alkalis becomes less damaging and more beneficial, through contribution towards strength development. This can be attributed to a parallel

0.12 Expansion (% after 14 days)

C65 0.10 0.08 BA

0.06

G12

0.04

G8

F22

0.02

F11 0.00 0

3000 Specific surface area

6000

9000

(cm2/g)

FIG. 9.76 Fourteen-day expansion results in accelerated mortar-bar tests as a function of fly ash fineness (Blaine specific surface area). The Na2O content of the fly ash samples used in this research ranged between 1.29% and 1.47% and the level in cement was 30%. (From: B erub e MA, Carles-Gibergues A, Duchesne J, Naproux P. Influence of particle size distribution on the effectiveness of type-F fly ash in suppressing expansion due to alkali–silica reactivity. In: Malhotra VM, editor. Proceedings of the fifth international conference on fly ash, silica fume, slag, and natural pozzolanas, Milwaukee. American Concrete Institute Special Publication 153; Jun. 1995, vol. I. p. 177–99.)

444 Lea’s Chemistry of Cement and Concrete

process of alkali–silica gel formation and gel dissolution, with silicate dissolved from this gel forming C-S-H. Where particle sizes are small, the rate of alkali–silica gel formation is outstripped by dissolution after a very short period of time.496 A number of experiments have demonstrated the similarities between pozzolanic reaction and alkali–silica reaction. As a typical example, 10% of finely dispersed silica fume combined with a high-alkali PC prevented expansion of a 28-day-old mortar immersed for 2 months in a saturated Ca(OH)2 + 1 M NaCl solution. In contrast, mortars containing the same amount of coarser granulated silica fume exhibited map cracking, typical of alkali–silica reaction.531 Similarly, a reactive opal-based aggregate was transformed into a very active natural pozzolana by finely grinding it before blending with PC.532 Opal and silica fume pastes behave similarly in terms of expansivity and chemical reactivity when prepared with Ca(OH)2 and/or NaOH solutions.532 However, in a mortar bar test, fine, dispersed silica fume prevented expansion caused by a highalkali cement, whereas agglomerated opal caused expansion.533 Moreover, it has been found that very high silica fume levels (45%) can cause expansion with non-reactive aggregate. Where pozzolanic materials are present, the rim of alkali–silica gel which forms at the surface of the reactive aggregate is thinner512 or, in some cases, does not form.511 The theory that pozzolanic reaction and alkali–silica reaction are essentially the same process is supported by the similar composition of the products of the two reactions. It is also worth noting that TEM analyses show that the rim of silicate hydrate formed at the surface of reactive aggregate particles has a CaO/SiO2 ratio of around 1.0512 whilst that of the rim formed on silica fume particles varies from 1.7 to 0.8 in moving from the bulk cement paste to the core of the particles.214 Thus, pozzolanic materials reduce the risk of alkali–silica reaction partly by their ability to react with the alkalis in the pore solution more rapidly than coarser aggregate particles.516 This ability will depend in part on the fineness of the particles and, thus, coarse pozzolanic materials may potentially contribute to harmful expansion. Mechanisms of Expansion Reduction Through the Use of Pozzolanic Materials Whilst the experimental results described above indicate some of the mechanisms through which pozzolanic materials control expansion caused by alkali–silica reaction, the complete mechanism is not yet fully understood. However, the key elements are likely to be the formation of hydration products (particularly C-S-H) in which alkali ions are immobilised, and a reduction in pH within pore solutions. This model requires that the pozzolanic material combines with all of the portlandite that is available for dissolution (i.e. that which is not protected by hydrates). This means that higher levels of pozzolanic material are required. The pessimum effect caused by lower levels of pozzolanic material may be due to the rise in pore solution alkali concentration and pH deriving from the constituents of the addition, being inadequately counteracted by the pozzolanic reaction itself.

9.7.2.3 Acid Attack Acid attack of concrete takes three forms. First, hydration products react with the acid to form dissolved ions, leading to loss of solid material—acidolysis. Second, in the case of some acids, insoluble salts form, some of which precipitate to cause expansion and cracking. Finally, some acids give complexes with calcium, aluminium, iron and silicate ions which produce much higher concentrations of these ions in solution than would otherwise be the case, again leading to dissolution. This can potentially occur at pH conditions under which cement would normally be relatively stable (complexolysis). Depending on the acid present, more than one of these deterioration mechanisms may be effective. Where acidolysis is the predominant mechanism, an acid solution penetrating the pores of concrete will start to cause some of the constituents to dissolve. Calcium is usually the first cation to be dissolved, since portlandite becomes soluble below relatively high pH values. AFm and AFt phases typically dissolve at slightly lower pH values, although solid aluminium and iron hydroxides are precipitated, which persist until the solution is relatively acidic. Exposure to acidic solution also causes loss of calcium—decalcification—of C-S-H gel, leaving relatively weak silica gel behind. The significance of calcium in this form of deterioration means that adjusting the composition of the cement matrix of concrete to obtain a lower calcium content will potentially impart greater resistance. Thus, pozzolanic materials may be effective in enhancing resistance to this type of acid attack, since their combination with PC will produce such an adjustment. However, reduced rates of mass transport and enhanced strength achieved through the use of pozzolanic materials are also likely to be reasons for improved resistance. Table 9.34 summarises the results of experiments carried out to evaluate the acid resistance of mixtures containing pozzolanic materials. Of the acids included in this list, nitric, hydrochloric and acetic acid can be considered compounds whose main deterioration mechanism is acidolysis. Generally (with the possible exception of nitric acid, for which less data exist), presence of fly ash, silica fume and metakaolin has the effect of improving resistance against these acids. There appears to be less benefit in using natural pozzolanas, although it should be recognised that, given their wide ranging chemical compositions, the results included in Table 9.34 are unlikely to be universally applicable to such materials.

TABLE 9.34 Effect of Pozzolanic Materials on Resistance of Paste, Mortar and Concrete Mixtures to Acid Attack Acid

Study Fly ash De Belie et al. (1996/ 1997)534,535 Bertron et al. (2005)536 Roy et al. (2001)537 Chang et al. (2005)538 Tamimi (1997)539 Torii and Kawamura (1994)450 Siad et al. (2010)540 Silica fume Kehdr and Abou-Zeid (1994)541

Pozzolanic Material

Acid Concentration, Sulfuric by Massa

Nitric

Hydrochloric

Phosphoric

Acetic

10%–38% siliceous fly 3% of each acid ash 1.7%

7.5%–22.5% siliceous fly ash 33% siliceous fly ash, 7% silica fume Optimum at 50% siliceous fly ash, 10% silica fume 10%–70% siliceous fly ash

All 1% except H3PO4 ¼ 5% 1% replenished weekly 1% replenished weekly

Reduction

2% with periodic additions to maintain pH 8.7% H2SO4, 2.2% HCl

Enhanced performance at >10% Minor enhancement

Reduction

Reduction

Reduction

Enhancement Enhancement

0.1

0.1

3% of each acid

5%–30%

2% with periodic additions to maintain pH All 1% except H3PO4 and acetic ¼ 5%

Enhanced performance at >5% Negligible change

Kim et al. (2007)278

5%–15%

2%

Enhancement

10%–20%

‘moderate to high concentration’ cyclic

Enhancement

Enhancement Enhancement

Enhancement

Pozzolanas and Pozzolanic Materials

Bertron et al. (2005)536 De Belie et al. (1996/ 1997)534,535 Torii and Kawamura (1994)450 Roy et al. (2001)537

7.5%–30%

Formic

Enhancement

Concentrated and 20% Enhancement HNO3, Concentrated and 20% H2SO4, 50% HCl 1.7%

0.2

Lactic

Enhancement

50% siliceous fly ash

13.5%–50% siliceous fly ash

LacticAcetic Mixture

Enhancement

Negligible change

Possible enhancement

Result unclear— mass gain observed

Reduction

445

Enhancement Continued

Acid

Study

Pozzolanic Material

Hanayneh et al. (2012)542 7.5%–30% Durning and Hicks (1991)543 15% Mehta (1985)544 Natural pozzolanas 13.5%–50% Algerian Siad et al. natural pozzolana (2010)540 40% ‘true pozzolana’ Sersale et al. (1998)545

Kim et al. (2007)278 Others Adesanya and Raheem (2010)547 a

Nitric

Hydrochloric

Phosphoric

Acetic

LacticAcetic Mixture

Lactic

Formic

wetting (16 h) and drying (8 h) All 5%, except sulfuric ¼ 1 and 5%

Enhancement

All 1%, except acetic ¼ 5%

Enhancement

Enhancement

8.7% H2SO4, 2.2% HCl 2:1 sulfuric/nitric mixture with pH of 3.5, with periodic additions to maintain pH 0.5% to 1.2% H2SO4; 0.3% to 0.8% HNO3; 0.2% to 0.5% HCl renewed every 2 weeks

Minor enhancement Reduction

Enhancement Reduction

Reduction

Reduction

Reduction

7.5%–30%

All 1% except H3PO4 and acetic ¼ 5%

Negligible change

Negligible change

Reduction— Result possible unclear enhancement at highest level of inclusion

5%–15%

2%

Enhancement

2.5%–20% corn cob ash

35% H2SO4, 9% HCl

Enhancement up to 10%

T€ urkel et al. 28% trass + 6% calcareous fly ash (2007)546

Metakaolin Roy et al. (2001)537

Acid Concentration, by Mass Sulfuric

Enhancement Enhancement

Enhancement

Enhancement

Enhancement

Negligible change/ reduction

Enhancement up to 15%

Where concentrations have been given in % volume, conversion to % mass assumes that this is in terms of concentrated solutions where H2SO4, HNO3 and HCl were used and that concentrations of these solutions were 95%, 70% and 37%, respectively.

446 Lea’s Chemistry of Cement and Concrete

TABLE 9.34 Effect of Pozzolanic Materials on Resistance of Paste, Mortar and Concrete Mixtures to Acid Attack—cont’d

Pozzolanas and Pozzolanic Materials

447

The precipitation of expansive reaction products is the predominant mechanism in the case of sulfuric acid—indeed, sulfuric acid attack is best viewed as being sulfate attack under acidic conditions. The products are calcium-bearing salts, and so it is perhaps not surprising that Table 9.34 indicates mainly enhanced resistance resulting from the use of pozzolanic materials. It should be noted that combinations of PC, fly ash and silica fume often prove to be optimal in resisting sulfuric acid attack.538,539 Such an approach limits both the aluminate and lime content of the cement matrix, which is likely to influence the extent to which ettringite and gypsum may be formed. Other acids known to have expansive products are citric and (in high concentrations, at least) tartaric acid. No studies on the effect of pozzolanic materials on resistance to these acids have been identified in the literature, but the expansive products are all calcium-bearing, and so this has the potential to be viable. Some of the insoluble salts precipitated as a result of reaction between hardened cement paste and acids are non-expansive and may, in fact, produce greater resistance. These acids include hydrofluoric, oxalic and phosphoric,463 although it should be stressed that deterioration resulting from phosphoric acid exposure still appears to be observed. Table 9.34 indicates mixed results regarding the effect of pozzolanic materials on resistance to phosphoric acid, although there is some evidence of improved performance resulting from the use of silica fume.542,543 This matter is complicated by the use of mass-loss measurements as a means of evaluating acid attack—the precipitation of calcium and aluminium phosphates may lead to a gain in mass, which may mask detrimental effects on the integrity of concrete. Some organic acids are capable of forming stable complexes with the constituent elements of cement. For instance, lactic acid, glycolic acid and pyrocatechol form strong complexes with aluminium, iron (III) and silicate ions, respectively. In such cases, it may not be calcium which is most vulnerable to dissolution, and the effects of including pozzolanic materials may differ from those of other acids. However, there is very limited data in the literature on this. Formic acid forms stable complexes with aluminium ions, and Table 9.34 shows that silica fume has the effect of enhancing resistance to this acid.543 This might be expected, since the use of silica fume would reduce the aluminate content of the resulting cement matrix. However, whether this is the result of a reduction in aluminate content, or simply a reduction in rates of mass transport is uncertain. Lactic acid also forms strong complexes with aluminium ions, and resistance is again enhanced by the inclusion of silica fume.544 However, it may also be enhanced by the introduction of siliceous fly ash,534,535 which will increase the aluminate content of the cement matrix. Thus, it is likely that the improved resistance is partly a physical effect, rather than a chemical one.

9.7.3 Physical Deterioration The two most common ways in which concrete may be attacked physically are through freeze/thaw action (where the formation of ice and associated water pressures in pores causes expansion and cracking) and abrasion. Abrasion can take several forms, some of which involve contact with moving water, others of which occur under dry conditions.

9.7.3.1 Freeze/Thaw Action If water freezes in near saturated concrete, increases in volume occur and the material tends to expand. The opposite happens when the temperature rises above 0°C. Repeated exposure to freeze–thaw cycles can lead to the deterioration of concrete, which can be intensified in the presence of de-icing salt. A survey suggests548 that this process is responsible for about 10% of the damage in structures in the United Kingdom and elsewhere. The freeze/thaw resistance of concrete can be enhanced by increasing concrete strength and the period of moist curing before exposure. This can also be achieved by introducing air bubbles into concrete during production to create an air-void system that can accommodate the pressures occurring during freezing as ice formation occurs. Laboratory tests indicate that the latter is most effective in achieving this.548 Both approaches are covered in concrete specifications for freeze/thaw exposure conditions, where the properties of the aggregate are also included with regard to the process.363 Investigations to examine freeze/thaw resistance of concrete have tended to adopt a range of accelerated tests (e.g. internal type methods; scaling with water or salt water as the freezing medium). It has been noted that the outcomes of these may not always reflect behaviour in practice548 and there can be difficulties selecting appropriate test conditions and performance criteria.549 According to some authors (using various test methods), fly ashes and silica fume reduce the frost resistance of non-airentrained concretes.550,551 Similar or poorer performance with pozzolanic materials, including natural pozzolana, silica fume and metakaolin, with the effect depending on the level in cement and the strength of concrete, has also been found.316 Other work indicates that non-air-entrained high strength concrete with various silica fume levels gave good resistance.445 This was believed to relate to less permeable concrete and self-desiccation, influencing water movements and saturation levels. It has also been suggested that with low w/c ratio concrete (0.32), including those with pozzolanic materials, enhanced performance can be achieved with the entrainment of air.552

448 Lea’s Chemistry of Cement and Concrete

With air-entrainment of concrete, the specific surface area or fineness of PC has been noted to influence admixture dose requirements to achieve a target air content553 and this has also been found with pozzolanic materials. Due to the porosity of residual carbon in fly ash from incomplete combustion, air-entraining admixture may be adsorbed by this component, affecting the air-water/cement interface stability and increasing the dose required to achieve a specified air content in concrete.554,555 As a result, admixtures have been formulated to be compatible with the material. It has also been noted that the effects of fly ash on air-entrainment tend to reduce when its surface area and that of the PC it is combined with are similar.556 Work investigating other pozzolanic materials including diatomaceous earth,279 rice husk ash,557silica fume557,558 and metakaolin558 indicate increasing admixture dose requirements with these materials for a target air content. Studies have investigated the effects of blended cements on the air-void system of hardened air-entrained concrete. This is normally carried out microscopically and can be related to freeze/thaw resistance. It has been suggested that the applicability of the Spacing Factor determined from this type of test may differ between internal and scaling type damage to concrete.559 Research considering fly ash560 and silica fume561 have noted little influence of using these materials in concrete on the airvoid system. Studies covering natural pozzolanas,562 fly ashes,550,563 and silica fume550 in cement (using various test methods) suggest that they do not increase deterioration, providing that concretes have a suitable entrained-air content and strength. The influence of pozzolanic material on the freeze/thaw resistance of concrete also depends on the level of addition. Fig. 9.77 suggests that only high levels of fly ash in cement ultimately decrease resistance.334 Fly ash has been noted to have no significant effect on the freezing and thawing resistance of concrete at levels of 15%–40% in cement.564,565 Levels of up to 35% are permitted within specification standards.363 The use of silica fume in cement at levels exceeding 15%, for example 20%–30%, also causes the freeze thaw resistance to decrease in the presence of entrained air.566 The reason perhaps lies in the changes occurring in the specific surface area, as well as the spacing of air voids in concretes.567 An example548 of the use of air-entrainment in a blend with fly ash and silica fume in a major bridge construction (250,000 m3 of concrete), demonstrates how this combination of materials could be used effectively on a large scale. In the Alpine regions of Italy, many dams built with blended cements are in an excellent state of preservation after decades of service, despite the large number of freeze–thaw cycles they are annually subjected to.568 Owing to the slow initial

FIG. 9.77 Weight loss versus number of freeze–thaw cycles for air-entrained concrete with various fly ash levels in cement after 14 days curing. (From: Yuan RL, Cook JE. Study of a class C fly ash concrete. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. I. p. 307–19.)

Pozzolanas and Pozzolanic Materials

449

hardening of blended cements, the frost exposure of concretes containing pozzolanic material should be somewhat delayed compared to that of PCs. Insufficient curing may be the reason why several standard methods for measuring the freeze–thaw resistance of PC concretes often give inferior results when applied to those made with blended cement.569 Freeze–thaw tests carried out after long curing periods have shown that blended cements develop equivalent or higher resistance than that of PCs, provided the concretes have the same strength.570 Combining fly ash571 or limited amounts of silica fume572 with PC has not resulted in a clear increase of scaling caused by de-icing agents. However, in spite of the fact that no particular benefits are realised when cements containing pozzolanic materials are used in the presence of freezing conditions and de-icing salts, they reduce the risk of penetration of chloride into concretes and thus of reinforcement corrosion.

9.7.3.2 Abrasion Resistance The resistance of concrete to surface abrasion is largely a function of the strength of the cement matrix and the aggregate used. With this in mind, it is not surprising that pozzolanic materials can enhance abrasion resistance only if strength can be increased through their use. Most abrasion resistance studies involving pozzolanic materials have been conducted using siliceous fly ash. Where fly ash is used with PC and water/cement ratios (water/PC + FA) are maintained, compressive strength can usually be expected to be less, and with this comes a proportional decrease in abrasion resistance.573 However, if the enhanced rheological properties imparted by fly ash are exploited to reduce the water/cement ratio, improved abrasion resistance is possible. Interestingly, when abrasion depth is plotted against flexural strength, the increase in abrasion resistance for a given improvement in flexural strength appears more marked for concrete mixes containing fly ash.574 This effect also appears to be evident in a study where siliceous fly ash was used to replace fine aggregate.575 In this case, an increase in strength is observed up to a level of 30% in fine aggregate, after which it declines. However, abrasion resistance continues to increase beyond this point, at least up to a fly ash level of 40%. Since silica fume tends to improve strength without modification of water/cement ratio, enhanced abrasion resistance is usually observed in concrete containing this. Again the improvement in resistance to abrasion where silica fume is used with PC is proportional to the increase in strength realised.576 Similar enhancement has been observed where silica fume has been used to replace fine aggregate (up to 10%).577 Metakaolin used as a replacement for fine aggregate improved abrasion resistance up to a level of 40%, after which resistance reduced.578 It should be noted that a metakaolin level of 40% is outside normal levels of usage. Again, the correlation between compressive strength and abrasion depth was strong. There is only a limited amount of research conducted into the influence of natural pozzolanas on abrasion resistance. One such material has been investigated in composite cements alongside PC, slag, silica fume, fly ash and limestone, with successful results for abrasion resistance.579 All of the cements investigated displayed enhanced abrasion resistance compared to a PC concrete, with the exception of CEM V cements containing combinations of slag, natural pozzolana and fly ash. It should be noted that the complex nature of the combination employed meant that the contribution of the natural pozzolana was unclear. However, since many natural pozzolanas do not beneficially enhance the rheological properties of concrete and are not as reactive as silica fume or metakaolin, it is likely that improved abrasion resistance can only be expected where actions are taken to reduce the w/c ratio, such as through the use of water-reducing admixtures. The results of this study also highlight an important point about abrasion resistance of concrete mixes containing pozzolanic materials, namely that the development of resistance is very sensitive to curing—more so than mixes made with PC only.578,580 This is because the slower pozzolanic reaction (at least in the case of fly ash and natural pozzolanas) requires an adequate supply of water for longer periods to allow the full potential strength of the concrete to be achieved.

9.8 CONCLUDING REMARKS Cements incorporating pozzolanic materials provide an option for engineers during the concrete specification stage for meeting particular requirements. However, it should be recognised that behaviour in a given situation can vary, even between similar types of pozzolanic materials. Furthermore, this can also change depending on the level used with PC. In many cases, there can be several requirements for concrete and as suggested previously for fly ash,581 it may be necessary to balance the selection of a pozzolanic material and its level in order to meet these and optimise performance. In general, pozzolanic materials of similar fineness to PC should have little influence on fresh properties of concrete. Fly ash, depending on its characteristics, can reduce the water requirement of concrete. For some of the higher fineness materials, the opposite may occur and the use of superplasticizing admixtures may be a necessary part of their adoption. In some cases, pozzolanic materials may also influence admixture dose requirements with regard to air-entrainment. Increased cohesiveness and reduced bleeding may also be obtained with these materials.

450 Lea’s Chemistry of Cement and Concrete

With some pozzolanic materials, reductions in early strength may be obtained, particularly when used at high levels. However, longer-term strength enhancements are also achievable. In general, if concretes are designed for equal 28 day strength, pozzolanic materials can be used up to a certain level and match the early age strength profile of PC concrete. High strength concretes are achievable with these materials, and in some cases combinations of more than one have been used for this. In terms of tensile strength and modulus of elasticity behaviour for pozzolanic materials tends to follow compressive strength. Shrinkage appears to depend on the particular pozzolanic material and basis of mix proportioning used, while for creep the strength when loading is applied is important. The transportation properties of concrete can be enhanced through the use of pozzolanic materials. These effects have been noted in comparisons made at equal w/c ratio and at 28 day strength and depend on the type of pozzolanic material. It is apparent that these properties are sensitive to the curing adopted. For many aspects of concrete durability, including chloride ingress, sulfate attack and alkali–aggregate reaction, the use of pozzolanic materials can enhance performance. However, properties such as carbonation resistance may, depending on the concrete mix and exposure conditions, give poorer performance. Under cyclic freezing and thawing conditions, air-entrainment and the air-void system appear to be important in achieving acceptable properties. For abrasion resistance, there seems to be little direct effect of pozzolanic materials on performance, although curing of concrete is important.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

19. 20. 21. 22. 23. 24. 25.

Lea FM. The chemistry of cement and concrete. London: Edward Arnold; 1970. Massazza F. Chemistry of pozzolanic additions and mixed cements. Il Cemento 1976;1:3–38. Penta F. Sulle pozzolane del Lazio. Ann Chim 1954;572–83. Parravano N, Caglioti V. Research on pozzolanas. Ric Sci 1937;271–92. Mielenz RC, Witte LP, Glantz OJ. Effect of calcination on natural pozzolanas. In: Symposium on use of pozzolanic materials in mortars and concretes. American Society for Testing and Materials Special Technical Publication, no. 99; 1950. p. 43–92. Central Board of Irrigation and Power Research. Scheme applied to River Valley Projects: survey of work done on pozzolana in India. Status report; 1971, no. 2, 177 pp. Akman MS, Mazlum F, Esenli F. A comparative study of natural pozzolans used in blended cement production. In: Fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication, 132; May 1992, vol. I. p. 471–94. Sersale R, Aiello R. Richerche sulla genesi, sulla constituzione e sulla reattivita’del ‘trass’ Renano. Silic Ind 1965;30(1):13–23. Takemoto K, Uchikawa H. Hydratation des ciments pouzzolaniques. In: Proceedings of the seventh international congress on the chemistry of cement, Paris; 1980, vol. I. p. IV-2/1–21. Ludwig U, Schwiete HE. Untersuchungen an Deutschen Trassen. Silic Ind 1963;28(10):439–47. Costa U, Massazza F. Factors affecting the reaction with lime of Italian pozzolanas. In: Proceedings of the sixth international congress on the chemistry of cement, Moscow; Sep. 1974, Supplementary paper, Section III. p. 2–18. Mortureux B, Hornain H, Gautier E, Regourd M. Comparison de la reactivite de differentes pouzzolanes. In: Proceedings of the seventh international congress on the chemistry of cement, Paris; 1980, vol. Ill: IV-110–15. Mehta PK. Studies on blended Portland cements containing Santorin earth. Cem Concr Res 1981;11:507–18. Tavasci B. Struttura della malta di calce e pozzolana di Bacoli. Il Cemento 1948;45(8–9):114–20. Tavasci B. Struttura della pozzolana di Segni. Il Cemento 1946;42(1):4–7. II Cimento 1946;42(2):25–9. Sersale R. Problemi di mineralogia applicata. Rendiconti della Società Mineralogica Italiana 1962;18:215–58. Sersale R. Structure and characterization of pozzolanas and fly ashes. In: Proceedings of the seventh international congress on the chemistry of cement, Paris; 1980, Principal report, Sub-Theme IV-I. Sersale S. Ricerche sulla zeolitizzazione dei vetri vulcanici per trattamento idrotermale: note I and II. Rendiconto dell’Accademia delle Scienze Fisiche e Matematiche della Società Nazionale di Scienze, Lettere ed Arti, Napoli 1959;XXVI:154–66. Rendiconto dell’Accademia delle Scienze Fisiche e Matematiche della Società Nazionale di Scienze, Lettere ed Arti, Napoli 1961;XXVII:317–66. Sersale R. Sulla ‘pseudomorfosi’ della leucite in analcime. Periodico di Mineralogia 1962;31(2–3):337–59. Drzaj B, Hocevar S, Slokan M, Zaic A. Kinetics and mechanism of reaction in the zeolitic tuff–CaO–H2O systems at increased temperature. Cem Concr Res 1978;8:711–20. Sersale R. Genesi e costituzione del tufo giallo Napoletano. Rendiconti dell’Accademia delle Scienze Fisiche e Matematiche della Società Nazionale di Scienze, Lettere ed Arti, Napoli 1958;25(4):181–207. Malquori G. Portland-pozzolana cement In: Proceedings of the fourth international symposium on the chemistry of cement, Washington, 1960. National Bureau of Standards, 1962; Monograph 43(2):983–1006. Wessely I. Der trass als hydraulischer zusatzstoff. Zement-Kalk-Gips 1961;14(7):284–93. Jambor J. Relation between phase composition, over-all porosity and strength of hardened lime-pozzolana pastes. Mag Concr Res 1963;15(45):131–42. Babachev G, Penchev P St. Resistance a la corrosion des ciments mixtes. In: RILEM international symposium. Durability of concrete, Prague, 1961, Final report. Prague: Publishing House of the Czech Academy of Sciences; 1962. p. 79–92.

Pozzolanas and Pozzolanic Materials

451

26. Kasai Y, Tobinai K, Asakura E, Feng N. Comparative study of natural zeolites and other inorganic admixtures in terms of characterization and properties of mortars. In: Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication, 132; May 1992, vol. I. p. 615–34. 27. Kaminskas R, Drapanauskaite D, Barauskas I. Influence of carbonated pozzolana on sulfate attack of cement stone at low temperatures. Adv Cem Res 2014;26:85–92. 28. Scherillo A. Petrografia chimica dei tufi Flegrei—Nota II. Rendiconti dell’Accademia delle Scienze Fisiche e Matematiche della Società Nazionale di Scienze, Lettere ed Arti, Napoli 1955;XXII(4):345–71. 29. Sersale R. Ricerche sperimentali sulla costituzione, sulla genesi e sul comportamento chimico dei tufi vulcanici. Rendiconti della Società Mineralogica Italiana 1961;17:499–536. 30. Cussino L, Dallatorre R. Ricerche su materiali naturali a comportamento pozzolanico. Il Cemento 1958;55(12):17–21. 31. Eades JL, Grim RE. Reactions of hydrated lime with pure clay minerals in soil stabilization. American Concrete Institute Special Publication, Research council; 1960, Publication no. 771. p. 51–63. 32. Sersale R, Aiello R, Frigione G. Indagini sui cosiddetti ‘tufi’ del Casalese. Industria Italiana del Cemento 1964;34(2):75–82. 33. Amicarelli V, Sersale R, Sabatelli V. Attivita’ pozzolanica dei prodotti piroclastici ‘argillificati’. Rendiconti dell’Accademia delle Scienze Fisiche e Matematiche della Società Nazionale di Scienze, Lettere ed Arti, Napoli 1966;33(4):257–82. 34. Johansson S, Andersen PJ. Pozzolanic activity of calcined moler clay. Cem Concr Res 1990;20:447–52. 35. Turriziani T, Corradini G. Materiali pozzolanici ad alto contenuto in silice. Industria Italiana del Cemento 1961;31(10):493–8. 36. Baire G. Le ciment-gaize pour travaux a la mer. Revue des Materiaux de Construction et des Travaux Publics 1930;248:168–70. 37. Turriziani R, Corradini G. Materiali pozzolanici ad alto contenuto in silice. Industria Italiana Cemento 1959;29(6):146–50. 38. Feret R. Recherches sur la nature et la progression de Taction pouzzolanique. Revue des Materiaux de Construction et des Travaux Publics, 1933; 281; 41–4. Revue des Materiaux de Construction et des Travaux Publics 1933;282:85–92. 39. Negro A, Stafferi L. Sulla costituzione chimico-mineralogica di un materiale naturale ad alta reattivita’ pozzolanica del Viterbese. Il Cemento 1965;62 (3):17–22. 40. Kantsepol’skii IS, Zhabitskii MS, Ragozina TA. Intensification of hardening process of Portland–pozzolana cement containing gliezh (naturally burned clay). Seriya Kimicheskikh Nauk: Isvestya Akademii Nauk UrSSR; 1957. 41. Van Den Berg JW, Vissers JLJ, Hohberg I, Wiens U. Fly ash obtained from co-combustion—state of the art on the situation in Europe. In: Federation Internationale du Beton symposium, concrete and environment, Berlin; Oct. 2001. 42. McCarthy MJ, Brindle JH, Dhir RK. A review of co-combustion fly ash for use in concrete construction. In: Dhir RK, Newlands MD, Halliday JE, editors. Proceedings of international symposium—recycling and reuse of waste materials. Thomas Telford; 2003. p. 447–58. 43. Dhir RK. In: Swamy RN, editor. Pulverized-fuel ash. Cement replacement materials. 3. Glasgow, UK: Surrey University Press; 1986. 44. Davis RE, Kelly JW, Troxell GE, Davis HE. Proportions of mortars and concretes containing Portland pozzolan cements. ACI J 1935;32:80–114. 45. Davis RE, Carlston RW, Kelly JW, Davis HE. Properties of cements and concretes containing fly ash. ACI J 1937;37:281–96. 46. McCarthy MJ, Sadiqul Islam GM, Csetenyi LJ, Jones MR. Investigating techniques for rapidly assessing fly ash reactivity for use in concrete. Lexington, Kentucky: World of Coal Ash; 2013. 47. Fulton AA, Marshall WT. The use of fly ash and similar materials in concrete. In: Proceedings of the institution of civil engineers; 1956, part 1, vol. 5. p. 714–30. 48. Thompson A. The current and future nature of combustion ashes. In: Cox M, Nugteren H, Janssen-Jurkovicova M, editors. Combustion residues— current, novel and renewable applications. Chichester, UK: Wiley; 2008. 430 pp. 49. Beer JM. High efficiency electric power generation: The environmental role. Prog Energy Combust Sci 2007;33:107–34. 50. Franco A, Diaz AR. The future challenges for “clean coal technologies”: joining efficiency increase and pollutant emission control. Energy 2009;34:348–54. 51. Buhre BJP, Elliott LK, Sheng CD, Gupta RP, Wall TF. Oxy-fuel combustion technology for coal-fired power generation. Prog Energy Combust Sci 2005;31:283–307. 52. Nussbaumer T. Combustion and co-combustion of biomass: fundamentals, technologies and primary measures for emission reduction. Energy Fuel 2003;17:1510–21. 53. Bech N, Feuerborn H-J. Ash quality in Europe—primary and secondary measures. Denver, Colorado: World of Coal Ash; 2011. 54. McCarthy MJ, Jones MR, Zheng L, Robl TL, Groppo JG. Characterizing long-term wet-stored fly ash following carbon and particle size separation. Fuel J 2013;111:430. 55. American Society for Testing and Materials. ASTM C618. Standard specification for coal fly ash and raw or calcined natural pozzolan for use in concrete. West Conshohocken, Pennsylvania: American Society for Testing and Materials; 2015. 56. British Standards Institution. BS EN 450-1. Fly ash for concrete: definition, specifications and conformity criteria. London, UK: British Standards Institution; 2012. 57. Watt JD, Thorne DJ. Composition and pozzolanic properties of pulverized fuel ashes. III. Pozzolanic properties of fly ashes as determined by chemical methods. J Appl Chem 1965;15:585–604. J Appl Chem, 1965;16:33–9. 58. Terrier P, Moreau M. Recherche sur le mecanisme de Taction pouzzolanique des cendres volantes dans le ciment. Revue de Mat eriaux de Construction et des Travaux Publics 1996;613:279–396. Revue de Mat eriaux de Construction et des Travaux Publics 1996;614:440–51. 59. Minnick LJ. Reactions of hydrated lime with pulverized coal fly ash. Miner Process 1968;9(2):15–20. Miner Process 1968;9(3):12–19. 60. Forest J, Demoulian E. Recherche d’une methode d’appreciation rapide de la reactivite des cendres volantes et des pouzzolanes additionnees au ciment. Silic Ind 1964;29(7–8):265–78.

452 Lea’s Chemistry of Cement and Concrete 61. Terrier P. Contribution of analysis by means of an electron microprobe to the cement chemistry. In: Proceedings of the fifth international congress on the chemistry of cement, Tokyo; 1968, part II. p. 278–87. 62. K€ ulaots I, Hurt RH, Suuberg EM. Size distribution of unburned carbon in coal fly ash and its implications. Fuel 2004;3(22):223–30. 63. McCarthy GJ, Swanson KD, Keller LP, Blatter WC. Mineralogy of western fly ash. Cem Concr Res 1984;14:471–8. 64. Mehta KP. Pozzolanic and cementitious byproducts as mineral admixtures for concrete—a critical review. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag, and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. I. p. 1–46. 65. Hubbard FH, Dhir RK, Ellis MS. Pulverized-fuel ash for concrete: compositional characterization of United Kingdom PFA. Cem Concr Res 1985;15:185–98. 66. Diamond S. On the glass present in low-calcium and in high-calcium fly ashes. Cem Concr Res 1983;13:459–64. 67. Uchikawa H, Uchida S, Hanehara S. Effect of character of glass phase in blending components on their reactivity in calcium hydroxide mixture. In: Proceedings of the eighth international congress on the chemistry of cement, Rio de Janeiro; 1986, vol. IV. p. 245–50. 68. Ambroise J, Martin-Calle S, Pera J. Pozzolanic behavior of thermally activated kaolin. In: Malhotra VM, editor. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication 132; 1992, vol. I. p. 731–48. 69. Malquori G, Cirilli V. Azione della calce sul caolino disidratato e sulle pozzolane naturali. La Ricerca Scientifica ed il Progresso Tecnico 1943;14(2– 3):85–93. 70. Ambroise J, Murat M, Pera J. Hydration reaction and hardening of calcined clays and related minerals: V—extension of the research and general conclusions. Cem Concr Res 1985;15:261–8. 71. Pera J. Metakaolin and calcined clays. Cem Concr Compos 2001;23(6):iii. Guest Editorial. 72. Geological Society. In: Reeves GM, Sims I, Cripps JC, editors. Clay minerals used in construction. London, UK: The Geological Society; 2006. 73. Sabir BB, Wild S, Bai J. Metakaolin and calcined clays as pozzolans for concrete: a review. Cem Concr Compos 2001;23:441–54. 74. Kakali G, Perraki T, Tsivilis S, Badogiannis E. Thermal treatment of kaolin: the effect of mineralogy on the pozzolanic activity. Appl Clay Sci 2001;20:72–80. 75. Bich C, Ambroise J, Pera J. Influence of degree of dehydroxylation on the pozzolanic activity of metakaolin. Appl Clay Sci 2009;44:194–200. 76. Jones TR. Metakaolin as a pozzolanic addition to concrete. In: Bensted J, Barnes P, editors. Structure and performance of cements. 2nd ed. London, UK: Spon Press; 2001. 584 pp. 77. Justice JM, Kurtis KE. Influence of metakaolin surface area on properties of cement-based materials. J Mater Civil Eng 2007;19(9):762–71. 78. Ish-Shalom M, Bentur A, Grinberg T. Cementing properties of oil-shale ash: I—effect of burning method and temperature. Cem Concr Res 1980;10:799–807. 79. Laubenheimer HJ. Technologische untersuchungen uber die Verwendung von Posidonienschiefer Lias epsilon aus dem raum BraunschweingHondelage zur Herstellung von Hydraulischen Bindemitteln und von Gasbeton. Tonind Zeitung 1968;92(8):296–305. 80. Khedaywi T, Yeginobali A, Smadi M, Cabrera J. Pozzolanic activity of jordanian oil shale ash. Cem Concr Res 1990;20:843–52. 81. Sellevold EJ, Nilsen T. Condensed silica fume in concrete: a world review. In: Malhotra VM, editor. Supplementary cementing materials for concrete. Minister of Supply and Services Canada; 1987. p. 165–243. 82. Hjorth L. Microsilica in concrete. Oslo, Norway: Nordic Concrete Federation, Publication no. 1, Nordic Concrete Research; Dec. 1982. 83. Batrakov VG, Kaprielov SS, Sheinfeld AV. Influence of different types of silica fume having varying silica content on the microstructure and properties of concrete. In: Malhotra VM, editor. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication 132; 1992, vol. II. p. 943–63. 84. Concrete Society. Cementitious materials: the effect of GGBS, fly ash, silica fume and limestone fines on the properties of concrete. Technical report 74. Camberley, UK: The Concrete Society; 2011. 85. Cook DJ, Suwanvitaya P. Properties and behaviour of lime-rice husk ash cements. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. II. p. 831–45. 86. Ganesan K, Rajagopal K, Thangavel K. Rice husk ash blended cement: Assessment of optimal level of replacement for strength and permeability properties of concrete. Constr Build Mater 2008;22:1675–83. 87. Rodriguez de Sensale G. Strength development of concrete with rice husk ash. Cem Concr Compos 2006;28:158–60. 88. Vittori C, Cereseto A. Solubilizzazione progressiva della silice e R2O3 dei materiali pozzolanici sotto l’azione della calce per valutazione del valore idraulico dei materiali stessi. La Chimica e L’Industria 1935;17:646–50. 89. Massazza F, Costa U. Factors determining the development of mechanical strength in lime-pozzolana pastes. In: Proceedings of the XXII conference on silicate industry and silicate science, Budapest; 6–11 Jun. 1977, vol. I. p. 537–52. 90. Sersale R, Sabatelli V. Sull’attivita’ pozzolanica delle zeoliti: I—Attivita’ pozzolanica dell’herschelite. Rendiconti dell’Accademia delle Scienze Fisiche e Matematiche della Società Nazionale di Scienze, Lettere ed Arti, Napoli 1960;27(4):263–86. 91. Sersale R. Sull’attivita’ pozzolanica delle zeoliti: II—Reattivita’ dell’analcime con soluzione d’idrossido di calcio. Rendiconti dell’Accademia delle Scienze Fisiche e Matematiche della Società Nazionale di Scienze, Lettere ed Arti, Napoli 1961;28(4):45–64. 92. Ludwig U, Schwiete HE. Researches on the hydration of trass cements. In: Proceedings of the fourth international congress on the chemistry of cement, Washington; 1960. US Monograph 1962;43(2):1093–100. 93. Buttler FG, Walker EJ. The rate and extent of reaction between calcium hydroxide and pulverised fuel ash. In: Proceedings of the use of PFA in concrete, Leeds, 14–16 Apr. 1982, vol. I. p. 71–81.

Pozzolanas and Pozzolanic Materials

453

94. Costa J, Massazza F. Natural pozzolanas and fly ashes: analogies and differences. In: Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 134–44. 95. Zhao-qi W, Young JF. The hydration of tricalcium silicate in the presence of colloidal silica. J Mater Sci 1984;79:3477–86. 96. Collepardi M, Marcialis A, Massidda L, Sanna U. Low pressure steam curing of compacted lime-pozzolana mixtures. Cem Concr Res 1976;6:497–506. 97. Kurbus B, Bakula F, Gabrovsek R. Reactivity of SiO2 fume from ferrosilicon production with Ca(OH)2 under hydrothermal conditions. Cem Concr Res 1985;15:134–40. 98. Malquori G, Spadano A. Azione combinata del gesso e della calce sui materiali pozzolanici. Ric Sci 1936;185–91. 99. Chapelle J. Attaque sulfo-calcique des laitiers et des pouzzolanes. Revue des Materiaux de Construction et des Travaux Publics 1958;511–6. 100. Costa U, Massazza F. Influenza del trattamento termico sulla reattivita’ con la calce di alcune pozzolane naturali. Il Cemento 1977;3:105–22. 101. Turriziani R, Schippa G. Influenza dei trattamenti termici sulle proprieta’ delle pozzolane Laziali. Ric Sci 1956;24:600–6. 102. Rossi G, Forchielli L. Porous structure and reactivity with lime of some natural Italian pozzolanas. Il Cemento 1976;4:215–21. 103. Turriziani R. Prodotti di reazione dell’idrato di calcio con la pozzolana: I. Ric Sci 1954;8:1709–17. 104. Ludwig U, Schwiete HE. Lime combination and new formations in the trass–lime reactions. Zement-Kalk-Gips 1963;10:421–31. 105. Millet J, Hommey R. Etude mineralogique des pates pouzzolanes-chaux. Bulletin de Liaison des Laboratoires des Ponts et Chaussees 1974;74:59–63. 106. Celik K, Jackson MD, Mancio M, Meral C, Emwas A-H, Mehta PK, Monteiro PJM. High-volume natural volcanic pozzolan and limestone powder as partial replacements for portland cement in self-compacting and sustainable concrete. Cem Concr Compos 2014;45:136–47. 107. Jackson MD, Landis EN, Brune PF, Vitti M, Chen H, Li Q, Kunz M, Wenk H-R, Monteiro PJM, Ingraffe AR. Mechanical resilience and cementitious processes in Imperial Roman architectural mortar. Proc Natl Acad Sci 2014;111:18484–9. 108. Sersale R, Rebuffat P. Microscopic investigations of hardened lime-pozzolana pastes. Zement-Kalk-Gips 1970;182–4. 109. Jambor J. Hydratationsprodukte der Kalk-Puzzolan-Bindemittel. Zement-Kalk-Gips 1963;5:177–86. 110. Sabatelli V, Sersale R, Amicarelli V. Richerche sulla costituzione di paste calce-pozzolana lungamente stagionate in acqua dolce. Rendiconti dell’Accademia delle Scienze Fisiche e Matematiche della Società Nazionale di Scienze, Lettere ed Arti, Napoli di Serie 1967;4:XXXIV. 111. Turriziani R, Schippa G. Malte di pozzolana, calce e solfato di calcio: I. Ric Sci 1954;9:1895–903. 112. Parissi F. Miscele pozzolana-calce-gesso. Le Industrie del Cemento 1932;4:32. 113. Massazza F, Testolin M. Trimethylsilylation in the study of pozzolana-containing pastes. Il Cemento 1983;1:49–62. 114. Vaquier A, Carles-Gibergues A. Sur l’importance des sulfates dans le caracte`re pouzzolanique d’une cendre volante silicoalumineuse de centrale thermique. Revue des Materiaux de Construction et des Travaux Publics 1970;662:331–7. 115. Aitcin PC, Autefage F, Carles-Gibergues F, Vaquier A. Comparative study of the cementitious properties of different fly ashes. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, vol. I. p. 91–114. 116. Massazza F, Cannas M. Ricerche sulla preparazione di calci idrauliche artificiali utilizzando le ceneri volanti del carbone sulcis. II Calore 1962;9– 15:58–64. 117. Italian Government, Norme sui requisiti accettazione e modalita’ di prova degli agglomerati cementizi e delle calci idrauliche. Italian Government Decree of 31 August 1972. 118. Liu H, Lu Z, Lin S. Composition and hydration of high calcium fly-ash. In: Proceedings of the seventh international congress on the chemistry of cement, Paris; 1980, vol. Ill: IV-7–12. 119. Carles-Gibergues A, Thenoz B, Vaquier A. Mecanisme d’hydratation d’une cendre de lignite calcaire. In: Proceedings of the seventh international congress on the chemistry of cement, Paris; 1980, vol. Ill: IV-53–9. 120. Grutzeck MW, Atkinson S, Roy DM. Mechanism of hydration of condensed silica fume in calcium hydroxide solutions. In: Malhotra VM, editor. In: Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. II. p. 643–64. 121. Buck AD, Burkes JP. Characterization and reactivity of silica fume. In: Proceedings of the third international congress on cement microscopy, Houston; 16–19 Mar. 1981. p. 279–85. 122. Malquori G, Sasso F. Azione della calce sul caolino disidratato. Ric Sci 1937;144–7. 123. Murat M. Hydration reaction and hardening of calcined clays and related minerals: I—preliminary investigation on metakaolinite. Cem Concr Res 1983;13:259–66. 124. Dron R. Experimental and theoretical study of the CaO–Al2O3–SiO2–H2O system. In: Proceedings of the sixth international congress on the chemistry of cement, Moscow; Sep. 1974, Supplementary Paper Section II. p. 16. 125. Zur Strassen H. Die Chemischen Reaktionen bei der Zementerhartung. Zement-Kalk-Gips 1958;4:137–43. 126. Tognon G, Ursella P. Combined lime and specific surface area of the hydration products of lime–pozzolana and lime–fly ash mixes. In: Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 145–54. 127. Fournier M, Geoffray J-M. Le liant pouzzolanes-chaux. Bulletin de Liaison des Laboratoires des Ponts et Chaussees 1978;93:70–8. 128. Cereseto A, Parissi F. II controllo delle pozzolane. Il Cemento Armato Amato—Le Industrie del Cemento 1944;41(8,9,10):45–8. 129. Day RL, Shi C. Influence of the fineness of pozzolan on the strength of lime–natural pozzolan cement paste. Cem Concr Res 1994;24:1485–91. 130. Ranganath RV, Sharma RC, Krishnamoorthy S. Influence of fineness and soluble silica content of fly ashes on their strength development with respect to age. In: Malhotra VM, editor. Proceedings of the fifth international conference on fly ash, silica fume, slag and natural pozzolans in concrete, Milwaukee. American Concrete Institute Special Publication 153; Jun. 1995, vol. I. p. 355–66.

454 Lea’s Chemistry of Cement and Concrete 131. Davidson DT, Mateos M, Katti RK. Activation of the lime-fly ash reaction by trace chemicals. In: 38th annual meeting of the highway research board, Washington, DC; 1959. 132. Shi C, Day RL. Chemical activation of blended cements made lime and natural pozzolans. Cem Concr Res 1993;23:1389–96. 133. Sullentrup MG, Baldwin JW Jr. High lime fly ash as a cementing agent. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. II. p. 321–31. 134. Papayianni J. Performance of a high-calcium fly ash in roller-compacted concrete. In: Malhotra VM, editor. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication 132; May 1992, vol. I. p. 367–86. 135. Joshi RC, Ward MA. Cementitious fly ashes—structural and hydration mechanism. In: Proceedings of the seventh international congress on the chemistry of cement, Paris; 1980, vol. Ill: IV-78–82. 136. Diamond S, Sheng Q, Olek J. Evidence for minimal pozzolanic reaction in a fly ash cement during the period of major strength development, pore structure and permeability of cementitious materials. In: Materials Research Society symposium proceedings, Boston, 28–30 Nov. 1988. Warrendale, PA: Materials Research Society; 1989. p. 437–46. 137. Sayanam RA, Kalsotra AK, Mehta SK, Singh RS, Mandal G. Studies on thermal transformations and pozzolanic activities of clay from Jammu region (India). J Therm Anal 1989;35:99–106. 138. Chatterji AK, Phatak TC, Dhariyal KD, Banerji AC. Pozzolanic activity of kaolin fired with admixtures of zinc oxide. J Sci Ind Res, India 1960;19b:453. 139. Taylor-Lange SC, Rajabali F, Holsomback NA, Riding K, Juenger MCG. The effect of zinc oxide additions on the performance of calcined sodium montmorillonite and illite shale supplementary cementitious materials. Cem Concr Compos 2014;53:127–35. 140. Dass A. Pozzolanic behaviour of rice husk ash. CIB Build Res Practice 1984;307–11. 141. Urrutia C, Gomez E, Rojas A. Effect of diatomaceous earth on blended-cement properties. In: Proceedings of the ninth international conference on the chemistry of cement, New Delhi; 1992, vol. V. p. 78–83. 142. British Standards Institution. BS EN 197-1. Cement, Part 1: composition, specifications and conformity criteria for common cements. London, UK: British Standards Institution; 2011. 143. Plowman C, Cabrera JG. The influence of pulverised fuel ash on the hydration reactions of calcium aluminates. In: Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 71–81. 144. Collepardi M, Baldini G, Pauri M, Corradi M. The effect of pozzolanas on the tricalcium aluminate hydration. Cem Concr Res 1978;8(6):741–52. 145. Uchikawa H, Uchida S. Influence of pozzolana on the hydration of C3A. In: Proceedings of the seventh international congress on the chemistry of cement, Paris; 1980; vol. Ill: IV-24–29. 146. Stein HN, Stevels JM. Influence of silica on the hydration of 3CaO.SiO2. J Appl Chem 1964;14:338–46. 147. Kurdowski W, Nocun-Wczelik W. The tricalcium silicate hydration in the presence of active silica. Cem Concr Res 1983;13:341–8. 148. Beedle SS, Groves GW, Rodger SA. The effect of fine pozzolanic and other particles on hydration of C3S. Adv Cem Res 1989;2(5):3–8. 149. Ogawa K, Uchikawa H, Takemoto K, Yasui I. The mechanism of the hydration in the system C3S-pozzolana. Cem Concr Res 1980;10(5):683–96. 150. Jawed I, Skalny J. Hydration of tricalcium silicate in the presence of fly ash. In: Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 60–70. 151. Dobson CM, Goberdhan DGC, Ramsay JDF, Rodger SA. 29Si MAS NMR study of the hydration of tricalcium silicate in the presence of finely divided silica. J Mater Sci 1988;23:4108–14. 152. Uchikawa H. Effect of blending components on hydration and structure formation. In: Proceedings of the eighth international congress on the chemistry of cement, Rio de Janeiro; 1986, vol. I. p. 249–80. 153. De Jong JGM, Stein HN, Stevels JM. Mutual interaction of C3A and C3S during hydration. In: Proceedings of the fifth international symposium on the chemistry of cement, Tokyo; 1968, vol. II. p. 311. 154. Mohan K, Taylor HFW. Pastes of tricalcium silicate with fly ash—analytical electron microscopy, trimethylsilylation and other studies. In: Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 54–9. 155. Kaminskas R, Barauskas I. The influence of pozzolana to the tricalcium aluminate hydration and the effects of chloride ions to formed hydrates. Ceramics-Silika´ty 2010;54:47–52. 156. Kaminskas R, Barauskas I. Influence of pozzolana on C4AF hydration and the effects of chloride and sulfate ions on the hydrates formed. CeramicsSilika´ty 2011;55:261–7. 157. Taylor HFW. Chemistry of cement hydration. In: Proceedings of the eighth international congress on the chemistry of cement, Rio de Janeiro; 1986, vol. I. p. 82–110. 158. Gard JA, Mohan K, Taylor HFW, Cliff G. Analytical electron microscopy of cement pastes: I—tricalcium silicate pastes. J Am Ceram Soc 1980;63:336–7. 159. Grutzeck MW, Roy DM. Electron microprobe studies of the hydration of 3CaO.SiO2 (tricalcium silicate). Nature 1969;223(5205):492–4. 160. Diamond S. C/S mole ratio of C-S-H gel in nature C3S paste as determined by EDXA. Cem Concr Res 1976;6:413–6. 161. Stucke MS, Majumdar AJ. The morphology and composition of an immature C3S paste. Cem Concr Res 1977;7:711–8. 162. Groves GW, Le Sueur PJ, Sinclair W. Transmission electron microscopy and microanalytical studies of ion-beam thinned sections of tricalcium silicate paste. J Am Ceram Soc 1986;69(4):353–6. 163. Rayment DL, Majumdar AJ. The composition of C-S-H phase in hydrated C3S paste. In: Proceedings of the seveth international congress on the chemistry of cement, Paris; 1980, vol. II. p. 11-64–70. 164. Rodger SA, Groves GW. The microstructure of tricalcium silicate/pulverized fuel ash blended pastes. Adv Cem Res 1988;1:84–91.

Pozzolanas and Pozzolanic Materials

455

165. Goto S, Daimon M, Hosaka G, Kondo P. Composition and morphology of hydrated tricalcium silicate. J Am Ceram Soc 1976;59(7–8):281–4. 166. Jennings HM. Aqueous solubility relationship for two types of calcium silicate hydrate. J Am Ceram Soc 1986;69(8):614–8. 167. Grutzeck M, Benesi A, Fanning B. Silicon-29 magic angle spinning nuclear magnetic resonance study of calcium silicate hydrates. J Am Ceram Soc 1989;72(4):665–8. 168. Uchikawa H, Furuta R. Hydration of C3S-pozzolana paste estimated by trimethylsilylation. Cem Concr Res 1981;11:65–78. 169. Kuznetsova TV, Dobronravova LA, Sovetnikova NK. Effect of a modified silica-containing additive on the hardening of cement systems. Zhurnal Prikladnoi Khimii (Leningrad) 1987;60(10):2351–4. 170. Brough AR, Dobson CM, Richardson IG, Groves GW. A study of the pozzolanic reaction by solid-state 29Si nuclear magnetic resonance using selective isotopic enrichment. J Mater Res 1995;30:1671–8. 171. Dalziel JA, Gutteridge WA. The influence of pulverized-fuel ash upon the hydration characteristics and certain physical properties of a Portland cement paste. Cement and concrete association technical report 560; 1986. 28 pp. 172. Copeland LE, Kantro DL. Hydration of Portland cement. In: Proceedings of the fifth international symposium on the chemistry of cement, Tokyo; 1968, vol. II. p. 387–421. 173. Fraay ALA, Bijen JM, de Haan YM. The reaction of fly ash in concrete: a critical examination. Cem Concr Res 1989;19:235–46. 174. Ghose A, Pratt PL. Studies of the hydration reactions and microstructure of cement-fly ash pastes. In: Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 82–91. 175. Meland I. Influence of condensed silica fume and fly ash on the heat evolution in cement pastes. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral. By-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. I. p. 665–73. 176. Hwang CL, Shen DH. The effects of blast-furnace slag and fly ash on the hydration of Portland cement. Cem Concr Res 1991;21:410–26. 177. Abdul-Maula S, Odler I. Hydration reactions in fly-ash-Portland cements. In: Malhotra VM, editor. Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 102–11. 178. Wei F, Grutzeck MW, Roy DM. The retarding effects of fly ash upon the hydration of cement pastes: the first 24 hours. Cem Concr Res 1985;15:174–84. 179. Diamond S, Lopez Flores F. Comparative studies of the effects of lignite and bituminous fly ashes in hydrated cement system. In: Malhotra VM, editor. Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 112–23. 180. Huang C-Y, Feldman RF. Hydration reactions in Portland cement-silica fume blends. Cem Concr Res 1985;15:585–95. 181. Ukrainczyk N, Koenders EAB. Hydration process of Portland cement blended with silica fume. Adv Mater Res 2013;699:578–83. 182. Langan BW, Weng K, Ward MA. Effect of silica fume and fly ash on heat of hydration of Portland cement. Cem Concr Res 2002;32:1045–51. 183. Kumar A, Roy DM. A study of silica fume modified cements of varied fineness. J Am Ceram Soc 1984;67:61–4. 184. Yogendran V, Langan BW, Ward MA. Utilization of silica fume in high strength concrete. In: Proceedings of the Symposium on Utilization of High Strength Concrete, Stavanger, Norway; 1987, p. 85–97. 185. Talero R, Rahhal V. Calorimetric comparison of Portland cements containing silica fume and metakaolin. J Therm Anal Calorim 2009;96:383–93. 186. Janotka I, Puertas F, Palacios M, Kuliffayova´ M, Varga C. Metakaolin sand-blended-cement pastes: Rheology, hydration process and mechanical properties. Constr Build Mater 2010;24(5):791–802. 187. Zhang M-H, Gjorv E. Effect of silica fume on cement hydration in low porosity cement pastes. Cem Concr Res 1991;21:800–8. 188. Berry EE, Hemmings RT, Langley WS, Carette GG. Beneficiated fly ash: hydration, microstructure, and strength development in Portland cement systems. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. I. p. 241–73. 189. Berry EE, Hemmings RT, Zhang M-H, Cornelius BJ, Golden DM. Hydration in high-volume fly ash concrete binders. Am Concr Inst Mater J 1994;90 (4):382–9. 190. Marsh BK, Day RL. Pozzolanic and cementitious reactions of fly ash in blended cement pastes. Cem Concr Res 1988;18:301–10. 191. Papadakis VG. Experimental investigation and theoretical modeling of silica fume activity in concrete. Cem Concr Res 1999;29:79–86. 192. Badogiannis E, Kakali G, Dimopoulou G, Chaniotakis E, Isivilis S. Metakaolin as a main cement constituent: exploitation of poor Greek kaolins. Cem Concr Compos 2005;27:197–203. 193. Lukas W. The influence of an Austrian fly ash on the reaction processes in the clinker phases of Portland cements. Mater Struct Res Test 1976;9 (53):331–7. 194. Li S, Roy DM, Kumar A. Quantitative determination of pozzolanas in hydrated systems of cement or Ca(OH)2 with fly ash or silica fume. Cem Concr Res 1985;15:1079–86. 195. Chatterji S. Reactivity of high specific surface silicas and mobility of Ca(OH)2 in cement paste. Il Cemento 1986;3:223–5. 196. Frias M, Cabrera J. Pore size distribution and degree of hydration of MK-cement pastes. Cem Concr Res 2000;30:561–9. 197. Turriziani R. Les pouzzolanes Italiennes et les ciments pouzzolaniques: II. Silic Ind 1958;23:265–70. 198. Grutzeck MW, Roy DM, Scheetz BE. Hydration mechanisms of high-lime fly ash in Portland-cement composites. In: Proceedings of symposium N on effects of fly ash incorporation in cement and concrete, Boston. Materials Research Society; 16–18 Nov. 1981. p. 92–101. 199. Massazza F, Daimon M. Conference report: chemistry of hydration of cements and cementitious systems. In: Proceedings of the ninth international congress on the chemistry of cement, New Delhi; 1992. p. 383–46. 200. Atkins M, Lachowski EE, Glasser FP. Investigation of solid and aqueous chemistry of 10-year-old Portland cement pastes; with and without silica modifier. Adv Cem Res 1993;5(19):97–102. 201. Lachowski EE, Mohan K, Taylor HFW, Lawrence CD, Moore AE. Analytical electron microscopy of cement pastes: III—pastes hydrated for long times. J Am Ceram Soc 1981;64:319–21.

456 Lea’s Chemistry of Cement and Concrete 202. Lachowski EE, Diamond S. Investigation of the composition and morphology of individual particles of Portland cement paste: I—C-S-H gel and calcium hydroxide particles. Cem Concr Res 1983;13:177–85. 203. Taylor HFW, Mohan K, Moir GK. Analytical study of pure and extended Portland cement paste: I—pure Portland cement pastes. J Am Ceram Soc 1985;68:680–5. 204. Harrison AM, Winter NB, Taylor HFW. An examination of some pure and composite Portland cement pastes using scanning electron microscopy with X-ray analytical capability. In: Proceedings of the eighth international congress on the chemistry of cement, Rio de Janeiro; 1986, vol. IV. p. 170–5. 205. Rodger SA, Groves GW. Electron microscopy study of ordinary Portland cement-pulverised fuel ash blended pastes. J Am Ceram Soc 1989;72 (6):1037–9. 206. Rayment PL. The effect of pulverised-fuel ash and the C/S molar ratio and alkali content of calcium silicate hydrates in cement. Cem Concr Res 1982;12:133–40. 207. Taylor HF, Newbury DE. An electron microprobe study of a mature cement paste. Cem Concr Res 1984;14:565–73. 208. Rayment DL, Majumdar AJ. The composition of the C-S-H phase in Portland cement pastes. Cem Concr Res 1982;12:753–64. 209. Rayment DL, Lachowski EE. The analysis of OPC pastes: a comparison between analytical electron microscopy and electron probe microanalysis. Cem Concr Res 1984;14:43–8. 210. Duchesne J, Berube MA. Effect of supplementary cementing materials on the composition of cement hydration products. Adv Cem Based Mater 1995;2:43–52. 211. Uchikawa H, Uchida S, Hanehara S. Element distribution on the polished surface of hardened cement pastes estimated by EPMA with ultra high speed wide area multianalyser. Il Cemento 1987;2:117–30. 212. Taylor HFW. Cement Chemistry. London: Academic Press; 1990. 213. Regourd M, Mortureux B. Microstructure of field concretes containing silica fume. In: Proceedings of the fourth international congress on cement microscopy, Las Vegas; 1982. p. 249–60. 214. Bonen D, Diamond S. Occurrence of large silica fume-derived particles in hydrated cement paste. Cem Concr Res 1992;22:1059–66. 215. Tavasci B, Cereseto A. Struttura del cemento pozzolanico idrato. La Chimica e L’Industria 1949;31:392–8. 216. Pietersen HS, Kentgens APM, Nachtegaal GH, Veeman WS, Bijen JMJM. The reaction mechanism of blended cements: a 29Si NMR study. In: Malhotra VM, editor. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication 132; May 1992, vol. I. p. 795–812. 217. Frı´as M, Cabrera J. Influence of MK on the reaction kinetics in MK/lime and MK-blended cement systems at 20°C. Cem Concr Res 2001;31:519–27. 218. Page CL, Vennesland O. Pore solution composition and chloride binding capacity of silica-fume cement pastes. Mat eriaux et Constructions 1983;16 (91):19–25. 219. Barneyback RS, Diamond S. Expression and analysis of pore fluids from hardened cement pastes and mortars. Cem Concr Res 1981;11:279–85. 220. Canham I, Page CL, Nixon PJ. Aspects of the pore solution chemistry of blended cements related to the control of alkali silica reaction. Cem Concr Res 1987;17(5):839–44. 221. Glasser FP, Marr J. The effect of mineral additives on the composition of cement pore fluids. Brit Ceram Proc 1984;35:419–29. 222. Larbi JA, Fraay ALA, Bijen JMJM. The chemistry of the pore fluid of silica fume-blended systems. Cem Concr Res 1990;20:506–16. 223. Diamond S. Effects of microsilica (silica fume) on pore solution chemistry of cement pastes. J Am Ceram Soc 1983;66(5). C-82-4. 224. Rasheeduzzafar SE, Hussain SE, Al-Gahtani AS. Pore solution composition and reinforcement corrosion characteristics of microsilica blended cement concrete. Cem Concr Res 1991;21:1035–48. 225. Ramlochan T, Thomas M, Gruber KA. The effect of metakaolin on alkali-silica reaction in concrete. Cem Concr Res 2000;30:339–44. 226. Feldman RF. Significance of porosity measurements on blended cement performance. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. I. p. 415–33. 227. Hooton RD. Permeability and pore structure of cement pastes containing fly ash, slag, and silica fume. In: Blended cements, Denver, 1984. American Society for Testing and Materials Special Technical Publication 897; 1986. p. 128–43. 228. Costa U, Massazza F. Some properties of pozzolanic cements containing fly ashes. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag, and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. I. p. 235–54. 229. Turriziani R, Rio A, et al. Ann Chim Appl 1954;XLIV:787–96. 230. Massazza F, Costa U. Aspetti dell’attivita’ pozzolanica e proprieta’ dei cementi pozzolanici. Il Cemento 1979;1:3–18. 231. Sellevold EJ, Bager DH, Klitgaard Jense K, Knudsen T. Silica fume—cement pastes; hydration and pore structure. Report BML 82.610. Division of Building Materials, The Norwegian Institute of Technology, Trondheim; 1982. p. 19–50. 232. Nelson JA, Young JF. Additions of colloidal silicas and silicates to Portland cement pastes. Cem Concr Res 1977;7(3):277–82. 233. Diamond S. Cement paste microstructure in concrete. Mater Res Soc Symp Proc 1987;85:21–31. 234. Cook DJ, Cao HT. An investigation of the pore structure in fly ash/OPC blends. In: Proceedings of the first international RILEM congress on pore structure and materials properties, Versailles; 1987, vol. 1. p. 69–76. 235. Diamond S, Ravina D, Lovell J. The occurrence of duplex films on fly ash surfaces. Cem Concr Res 1980;10:297–300. 236. Halse Y, Pratt PL. Development of microstructure and other properties in fly ash OPC systems. Cem Concr Res 1984;14:491–8. 237. Sinclair W, Groves GW. The microstructure of Portland cement-silica fume pastes. J Mater Sci Lett 1986;5(1):101–2. 238. Kjellsen KO, Detwiler RJ, Gjorv OE. Pore structure of plain cement pastes hydrated at different temperatures. Cem Concr Res 1990;20:927–33. 239. Massazza F. Blended cements. Special Lecture, Fourth NCB International Seminar on Cement and Building Materials, New Delhi; 1994.

Pozzolanas and Pozzolanic Materials

457

240. Scrivener KL, Patel HH, Pratt PL, Parrott LJ. Analysis of phases in cement paste using backscattered electron images, methanol adsorption and thermogravimetric analysis. In: Micro structural development during hydration of cement. Materials Research Society symposium proceedings; 1987, vol. 85. p. 67–76. 241. Abdel-Jawad Y, Hansen W. Pore structure of hydrated cement determined by mercury porosimetry and nitrogen sorption techniques. In: Roberts LR, Skalny JP, editors. Pore structure and permeability of cementitious materials. Materials Research Society symposium proceedings; 1989, vol. 137. p. 105–18. 242. Massazza F, Costa U. Permeabilita’ di paste di cemento Portland e pozzolanico: relazione con la tessitura porosa. In: Convegno AITEC: La Durabilita delle Opere in Calcestruzzo, Padova; 1987. p. 65–73. 243. Manmohan D, Mehta PK. Influence of pozzolanic, slag, and chemical admixtures on pore size distribution and permeability of hardened cement pastes. Cem Concr Aggreg 1981;63–7. 244. Feldman RF. Pore structure formation during hydration of fly-ash and slag cement blends. In: Materials research society annual meeting, Boston; 1981. 245. Massazza F. Microstructure of hydrated pozzolanic cements. In: First international workshop on hydration and setting, Dijon, 1991. London: E.&F.N. Spon; 1992. 246. Huang C-Y, Feldman RF. Influence of silica fume on the microstructural development in cement mortars. Cem Concr Res 1985;15:285–94. 247. Zhang M-H, Gjorv OE. Effect of silica fume on pore structure and chloride diffusivity of low porosity cement pastes. Cem Concr Res 1991;21:1006–14. 248. Marsh BK, Day RL. Some difficulties in the assessment of pore-structure of high performance blended cement pastes. In: Very high strength cementbased materials, Boston, 1984. Materials Research Society symposium proceedings; 1985, vol. 42. p. 113–21. 249. Feldman RF, Cheng-yi H. Properties of Portland cement-silica fume pastes: I—porosity and surface properties. Cem Concr Res 1987;15:765–74. 250. Costa U, Massazza F. Remarks on the determination of the pore distribution of Portland and pozzolanic cement pastes. In: Maso JC, editor. Proceedings of the first international congress on the pore structure and construction materials properties, Versailles; 1987. p. 159–66. 251. Poon CS, Kou SC, Lam L. Compressive strength, chloride diffusivity and pore structure of high performance metakaolin and silica fume concrete. Constr Build Mater 2006;20:858–65. 252. Khatib JM, Wild S. Pore size distribution of metakaolin paste. Cem Concr Res 1996;26:1545–53. 253. Marsh BK, Day RL, Bonner DG. Pore structure characteristics affecting the permeability of cement paste containing fly ash. Cem Concr Res 1985;15:1027–38. 254. Costa U, Massazza F. Permeability and pore structure of cement pastes. Mater Eng 1990;1(2):459–66. 255. Massazza F. Microscopic structure and concrete durability. In: Conference Europeenne, La Fissuration des Betons et la Durabilite des Constructions, Saint-Remy-les-Chevreuse; 31 Aug.–2 Sep. 1988. 256. Marchese B, D’Amore F. Sorptivity and constitution of mature cementitious mortars. Il Cemento 1993;90:25–37. 257. Massazza F, Marchese B, Costa U. Capillary absorption and composition of pastes and mortars made with Portland and pozzolanic cements. In: Proceedings of the sixth international congress on durability of building materials and components, Tokyo; 1993. p. 10. 258. Page CL, Short NR, El Tarras A. Diffusion of chloride ions in hardened cement pastes. Cem Concr Res 1981;11:395–406. 259. Hansson CM, Strunge H, Markussen JB, Frolund T. The effect of cement type on the diffusion of chloride. Nordic Concrete Research Publication no. 4; 1985. p. 70. 260. Collepardi M, Marcialis A, Turriziani R. Penetration of chloride ions into cement pastes and concretes. J Am Ceram Soc 1972;55:534–5. 261. Kumar A, Roy DM. Retardation of Cs+ and CP diffusion using blended cement admixtures. J Am Ceram Soc 1986;69:356–60. 262. Gautefall O, Havdahl J. Effect of condensed silica fume on the mechanism of chloride diffusion into hardened cement paste. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114. 1989, vol. II. p. 849–60. 263. Byfors K. Influence of silica fume and fly ash on chloride diffusion and pH values in cement paste. Cem Concr Res 1987;17:115–30. 264. Chatterji S, Kawamura M. A critical reappraisal of ion diffusion through cement based materials: I—sample preparation, measurement technique and interpretation of results. Cem Concr Res 1992;22:525–30. 265. Nagataki S, Ohga H, Kim EK. Effect of curing conditions on the carbonation of concrete with fly ash and the corrosion of reinforcement in long-term test. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, vol. I. p. 521–40. 266. Goto S, Roy DM. Diffusion of ions through hardened cement pastes. Cem Concr Res 1981;11:751–7. 267. Yu SW, Page CL. Diffusion in cementitious materials: I—comparative study of chloride and oxygen diffusion in hydrated cement pastes. Cem Concr Res 1991;21:581–8. 268. Whiting D, Mitchell TM. History of the rapid chloride permeability test. Transport Research Record 1335. Washington, DC: Transportation Research Board; 1992. p. 55–62. 269. Li S, Roy DM. Investigation of relations between porosity, pore structure, and Cl– diffusion of fly ash blended cement pastes. Cem Concr Res 1986;16:749–59. 270. Beaudoin JJ, Ramachandran VS, Feldman RF. Interaction of chloride and C-S-H. Cem Concr Res 1990;20:875–83. 271. Kondo R, Satake M, Ushiyama H. Diffusion of various ions in hardened Portland cement. In: 28th general meeting of the technical session of the cement association of Japan, Tokyo; 1974. p. 41–3. 272. Helmuth R. Fly ash in cement and concrete. Illinois, USA: Portland Cement Association; 1987. 273. Joshi RC, Lohtia RP. Fly ash in concrete: production properties and use. Amsterdam, the Netherlands: Gordon and Breach Science Publishers; 1997. 128 pp. 274. Burgos-Montes O, Palacios M, Rivilla P, Puertas F. Compatibility between superplasticizer admixtures and cements with mineral additions. Constr Build Mater 2012;31:300–9.

458 Lea’s Chemistry of Cement and Concrete 275. Alonso MM, Palacios M, Puertas F. Compatibility between polycarboxylate-based admixtures and blended-cement pastes. Cem Concr Compos 2013;35:151–62. 276. Gallias JL, Kara-Ali R, Bigas JP. The effect of fine mineral admixtures on water requirement of cement pastes. Cem Concr Res 2000;30:1543–9. 277. Ferraris CF, Obla KH, Hill R. The influence of mineral admixtures on the rheology of cement paste and concrete. Cem Concr Res 2001;31:245–55. 278. Kim H-S, Lee S-H, Moon H-Y. Strength properties and durability aspects of high strength concrete using Korean metakaolin. Constr Build Mater 2007;21:1229–37. 279. Tagnit-Hamou A, Petrov N, Luke K. Properties of concrete containing diatomaceous earth. ACI Mater J 2003;100(1):73–8. 280. Badogiannis E, Papadakis VG, Chaniotakis E, Tsivilisa S. Exploitation of poor Greek kaolins: strength development of metakaolin concrete and evaluation by means of k-value. Cem Concr Res 2004;34:1035–41. 281. Belaidi ASE, Azzouz L, Kadri E, Kenai S. Effect of natural pozzolana and marble powder on the properties of self-compacting concrete. Constr Build Mater 2012;31:251–7. 282. Khatib JM. Performance of self-compacting concrete containing fly ash. Constr Build Mater 2008;22:1963–71. 283. Persson B. A comparison between mechanical properties of self-compacting concrete and the corresponding properties of normal concrete. Cem Concr Res 2001;31:193–8. 284. Hassan AAA, Lachemi M, Hossain KMA. Effect of metakaolin and silica fume on the durability of self-consolidating concrete. Cem Concr Compos 2012;34(6):801–7. 285. Aldred J. Burj Khalifa—a new high for high performance concrete. Proc ICE Civil Eng 2010;163:66–73. 286. Cook TH. Concrete pumping and spraying—a practical guide. London, UK: Thomas Telford; 1990. 287. Delvasto S. Pozzolanic activity and characteristics of Colombian materials. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, vol. I, p. 77–89. 288. Giergiczny Z, Werynska A. Influence of fineness of fly ashes on their hydraulic activity. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. I. p. 97–115. 289. Sybertz F. Comparison of different methods for testing the pozzolanic activity of fly ashes. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. I. p. 477–97. 290. Asgeirsson H, Gudmundsson G. Pozzolanic activity of silica dust. Cem Concr Res 1979;9:249–52. 291. Kohno K, Aihara F, Ohno K. Relative durability properties and strengths of mortars containing finely ground silica and silica fume. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. I. p. 815–26. 292. Megat Johari MA, Brooks JJ, Kabir S, Rivard P. Influence of supplementary cementitious materials on engineering properties of high strength concrete. Constr Build Mater 2011;25:2639–48. 293. Chatterji S, Collepardi M, Moriconi G. Pozzolanic property of natural and synthetic pozzolans: a comparative study. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983. vol. I. p. 221–33. 294. Stuart KD, Anderson DA, Cady PD. Compressive strength studies on Portland cement mortars containing fly ash and superplasticizer. Cem Concr Res 1980;10:823–32. 295. Regourd M, Mortureux B, Hornain H. Use of condensed silica fume as filler in blended cements. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. I, p. 847–65. 296. Kasay Y, Matsui I, Fukushima Y, Kamohara H. Air permeability and carbonation of blended cement mortars. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. I. p. 435–51. 297. Hansen TC. Long-term strength of high fly ash concrete. Cem Concr Res 1990;20:193–6. 298. Swamy RN, Mahmud HB. Mix proportions and strength characteristics of concrete containing 50 per cent low-calcium fly ash. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986; vol. I. p. 413–32. 299. McCarthy MJ, Dhir RK. Development of high fly ash volume cements for use in concrete construction. Fuel J 2005;84(11):1423–32. 300. Bentz DP, Ferraris CF. Rheology and setting of high volume fly ash mixtures. Cem Concr Compos 2010;32(4):265–70. 301. McCarthy MJ, Yakub H, Csetenyi LJ, Jones MR. Initial studies on modern fly ashes for use in concrete construction. In: EuroCoalAsh 2014. Munich, Germany: VGB PowerTech, Essen; 2014. p. 79–94. 302. Carette GG, Malhotra VM, Aitcin PC. Preliminary data on long term development of condensed silica fume concrete. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication SP 114; 1989; Supplementary papers. p. 597–617. 303. Munn CJ. Compositional variations between different silica fumes and their effect on early structure in cement and concrete. Individual Project for the ACT Diploma course; 1985–86, New Zealand. 304. De Larrard F, Bostivironnois JL. On the long-term strength losses of silica-fume high-strength concretes. Mag Concr Res 1991;43(155):109–19.

Pozzolanas and Pozzolanic Materials

459

305. Carette GG, Malhotra VM. Long-term strength development of silica fume concrete. In: Malhotra VM, editor. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication 132; 1992, vol. II. p. 1017–44. 306. Curcio F, DeAngelis BA, Pagliolico S. Metakaolin as a pozzolanic microfiller for high performance mortars. Cem Concr Res 1998;28(6):803–9. 307. Wild S, Khatib JM, Jones A. Relative strength, pozzolanic activity and cement hydration in superplasticised metakaolin concrete. Cem Concr Res 1996;26(10):1537–44. 308. Penttala V. Mechanical properties of high strength concretes based on different binder combinations. In: Proceedings of the symposium on utilization of high strength concrete, Stavanger, Norway; 1987. p. 123–34. 309. Cornelissen HAW, Hellewaard RE, Vissers JLJ. Processed fly ash for high performance. In: Malhotra VM, editor. Proceedings of the fifth international conference on fly ash, silica fume, slag and natural pozzolans in concrete, Milwaukee. American Concrete Institute Special Publication 153; Jun. 1995, vol. I. p. 67–79. 310. Abdul Razak H, Wong HS. Strength estimation model for high-strength concrete incorporating metakaolin and silica fume. Cem Concr Res 2005;35:688–95. 311. Dhir RK, McCarthy MJ, Magee BJ. Impact of BS EN 450 PFA on concrete construction in the UK. Constr Build Mater 1998;12(1):59–74. 312. Jones MR, McCarthy A, Booth APPG. Characteristics of the ultrafine component of fly ash. Fuel 2006;85:2250–9. 313. Wolsiefer J, Sivasundaram Sr, V, Malhotra VM, Carette GG. Performance of concretes incorporating various forms of silica fume. In: Malhotra VM, editor. Proceedings of the fifth international conference on fly ash, silica fume, slag and natural pozzolans in concrete, Milwaukee. American Concrete Institute Special Publication 153; Jun. 1995, vol. II. p. 591–655. 314. Trattenberg A. Silica fumes as a pozzolanic material. Il Cemento 1978;75:369–76. 315. British Standards Institution. BS EN 1992-1-1. Eurocode 2: Design of concrete structures - Part 1-1: General rules and rules for buildings, London, 2004. 316. Dhir RK, Limbachiya MC, Henderson NA, Chaipanich A, Williamson A. Use of unfamiliar cements to EN 197-1 in concrete. DETR Research Contract No 39/3/238; 1999. 293 pp. 317. Videla C, Aguilar C. An updated look at drying shrinkage of Portland and blended Portland cement. Mag Concr Res 2006;58(7):459–76. 318. Shannag MJ. High strength concrete containing natural pozzolan and silica fume. Cem Concr Compos 2000;22:399–406. 319. Dhir RK, McCarthy MJ, Csetenyi LJ, Brindle JH. Co-combustion in electricity generation and the properties of fly ash for use in concrete construction. DTI Research Contract No 39/3/678 CC2255; 2005. 148 pp. 320. Dhir RK, McCarthy MJ, Paine KA. Engineering property and structural design relationships for new and developing concretes. Mater Struct 2005;38 (1):1–9. 321. Bhanja SB, Sengupta B. Influence of silica fume on the tensile strength of concrete. Cem Concr Res 2005;35:743–7. 322. Almusallam AA, Beshr H, Maslehuddin M, Al-Amoudi OSB. Effect of silica fume on the mechanical properties of low quality coarse aggregate concrete. Cem Concr Compos 2004;26:891–900. 323. Siddique R, Klaus J. Influence of metakaolin on the properties of mortar and concrete: a review. Appl Clay 2009;43:392–400. 324. Kaid N, Cyr M, Julien S, Khelafi H. Durability of concrete containing a natural pozzolan as defined by a performance-based approach. Constr Build Mater 2009;23:3457–67. 325. Aydin AC, Gu R. Influence of volcanic originated natural materials as additives on the setting time and some mechanical properties of concrete. Constr Build Mater 2007;21:1277–81. 326. Siddique R. Performance characteristics of high volume class F fly ash concrete. Cem Concr Res 2004;34:487–93. 327. Dhir RK, McCarthy MJ, Paine KA, editors. Use of BS EN 450 PFA in concrete construction. Technology digest 1, The Concrete Society; 2002. 20 pp. 328. Brooks JJ, Megat Johari MA. Effect of metakaolin on creep and shrinkage of concrete. Cem Concr Compos 2001;23:495–502. 329. ACI Committee 225. Guide to selection and use of hydraulic cements. Farmington Hills, MI, USA: American Concrete Institute; 1999. 330. Troxell GE, Raphael JM, Davis RE. Long-time creep and shrinkage tests of plain and reinforced concrete. Proc Am Soc Test Mater 1958;58:1101–20. 331. Carette GG, Malhotra VM. Characterization of Canadian fly ashes and their performance in concrete. Division Report MRP/MSL 84-137, CANMET—Energy Mines and Resources Canada; 1984. 332. Blaine RL, Arni HT. Section 10, Shrinkage and expansion of concrete—interrelations between cement and concrete properties, part 4—science series 15; 1969. US Department of Commerce, National Bureau of Standards. 333. Bilodeau A, Malhotra VM. Properties of high-volume fly ash concrete made with high-early strength ASTM Type III cement. In: Malhotra VM, editor. Proceedings of the fifth international conference on fly ash, silica fume, slag and natural pozzolans in concrete, Milwaukee. American Concrete Institute Special Publication 153; Jun. 1995, vol. I. p. 1–23. 334. Yuan RL, Cook JE. Study of a class C fly ash concrete. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. I. p. 307–19. 335. Sri Ravindraraja R, Tam CT. Properties of concrete containing low-calcium fly ash under hot and humid climate. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. I. p. 139–55. 336. Mazloom M, Ramezanianpour AA, Brooks JJ. Effect of silica fume on mechanical properties of high-strength concrete. Cem Concr Compos 2004;26:347–57. 337. Ding J-T, Li ZJ. Effect of metakaolin and silica fume on properties of concrete. ACI Mater J 2003;99(4):393–8. 338. Guneyisi E, Mehmet Gesoglu M, Mermerdas K. Improving strength, drying shrinkage, and pore structure of concrete using metakaolin. Mater Struct 2008;41:937–49. 339. Khatri RP, Sirivivathnanon V. Effect of different supplementary cementitious materials on mechanical properties of high performance concrete. Cem Concr Res 1995;25(1):209–20.

460 Lea’s Chemistry of Cement and Concrete 340. Kreijger PC. Inhomogeneity in concrete and its effect on degradation: a review of technology. Protection of concrete. E & F N Spon, Cambridge, UK, pp 31–52. 341. Dhir RK, McCarthy MJ, Tittle PAJ, Zhou S. Role of cement content in specifications for concrete durability: aggregate type influences. Proc ICE Struct Build 2006;159(4):229–42. 342. Xu A, Sarkar S, Nilsson LO. Effect of fly ash on the microstructure of cement mortar. Mater Struct 1993;26:414–24. 343. Bentur A, Cohen MD. Effect of condensed silica fume on the microstructure of the interfacial zone in Portland cement mortars. J Am Ceram Soc 1987;70(10):738–43. 344. Costa U, Facoetti M, Massazza F. Permeability of the cement-aggregate interface: influence of the type of cement, water/cement ratio and superplasticizer. In: Proceedings of the international symposium admixture for concrete improvement of properties, Barcelona; 1990. p. 392–401. 345. Duan P, Shui Z, Chen W, Shen C. Effects of metakaolin, silica fume and slag on pore structure, interfacial transition zone and compressive strength of concrete. Constr Build Mater 2013;44:1–6. 346. Kuroda M, Watanabe T, Terashi N. Increase of bond strength at interfacial transition zone by the use of fly ash. Cem Concr Res 2000;30(2):253–8. 347. Concrete Society. Permeability testing of site concrete: a view of methods and experience. Concrete society technical report no 31, Camberley, UK; 2008. 80 pp. 348. Costa U, Facoetti M, Massazza F. Permeability and diffusion of gases in concrete. In: Proceedings of the ninth international congress on the chemistry of cement, New Delhi; 1992, vol. V. p. 107–14. 349. Bamforth PB. The water permeability of concrete and its relationship with strength. Mag Concr Res 1991;43(157):233–41. 350. Day RL, Konecny L. Relationships between permeability and microstructural characteristics of fly ash mortars. In: Materials Research Society symposium proceedings; 1989, vol. 137. p. 391–2. 351. Dhir RK, Byars EA. PFA concrete: intrinsic permeability. ACI Mater J 1993;90(6):571–80. 352. Khan MI, Lynsdale CJ. Strength, permeability, and carbonation of high-performance concrete. Cem Concr Res 2002;32:123–31. 353. Chia KS, Zhang M-H. Water permeability and chloride penetrability of high-strength lightweight aggregate concrete. Cem Concr Res 2002;32:639–45. 354. Song H-W, Pack S-W, Nam S-H, Jang J-C, Saraswathy V. Estimation of the permeability of silica fume cement concrete. Constr Build Mater 2010;24:315–21. 355. Shekarchi M, Bonakdar A, Bakhshi M, Mirdamadi A, Mobasher B. Transport properties in metakaolin blended concrete. Constr Build Mater 2010;24:2217–23. 356. Ho DWS, Lewis RK. The water sorptivity of concretes: the influence of constituents under continuous curing. Durab Build Mater 1987;4:241–52. 357. Ghrici M, Kenai S, Said-Mansour M. Mechanical properties and durability of mortar and concrete containing natural pozzolana and limestone blended cements. Cem Concr Compos 2007;29:542–9. 358. Dhir RK, Byars EA. PFA concrete: near surface absorption properties. Mag Concr Res 1991;43(157):219–32. 359. Alexander MG, Magee BJ. Durability performance of concrete containing condensed silica fume. Cem Concr Res 1999;29:917–22. 360. Khatib JM, Clay RM. Absorption characteristics of metakaolin concrete. Cem Concr Res 2003;34(1):19–29. 361. Scrivener KL, Gartner EM. Microstructural gradients in cement paste around aggregate particles. J Am Ceram Soc 1988;114:77–85. 362. Scrivener KL, Crumbie AK, Laugesen P. The interfacial transition zone (ITZ) between cement paste and aggregate in concrete. Interface Science 2004;12:411–21. 363. British Standards Institution. BS 8500. Concrete—complementary British standard to BS EN 206-1: Method of specifying and guidance for the specifier. London, UK: British Standards Institution; 2006. 364. Thomas S, Meise-Gresch K, Muller-Warmuth W. MAS NMR studies of partially carbonated Portland cement and tricalcium silicate pastes. J Am Ceram Soc 1993;76(8):1998–2001. 365. Jerga J. Physico-mechanical properties of carbonated concrete. Constr Build Mater 2004;18:645–52. 366. Wierig H-J. Longtime studies on the carbonation of concrete under normal outdoor exposure. In: Proceedings of RILEM seminar on the durability of concrete structures under normal outdoor exposure, Hannover; 26–29 Mar. 1984. p. 239–49. 367. Stewart MG, Wang XM, Nguyen MN. Climate change impact and risks of concrete infrastructure deterioration. Eng Struct 2011;33:1326–37. 368. Dhir RK, Hewlett PC, Chan YN. Near surface characteristics of concrete: prediction of carbonation resistance. Mag Concr Res Sep. 1989;41 (148):137–43. 369. Kokubu M, Nagataki S. Carbonation of concrete with fly ash and corrosion of reinforcements in 20-year tests. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. I. p. 315–29. 370. Nischer P. Einfluss der Betongute auf die Karbonatisierung. Mitteilungen aus FI VDZ; 33, Dauerhaftigkeit von Beton; 1984. 371. Sideris KK, Savva AE and Papayianni J. Sulfate resistance and carbonation of plain and blended cements. Cem Concr Compos Vol 28, 2006, pp47-56. 372. Matthews JD. Carbonation of ten-year old concretes with and without pulverised fuel ash. In: Proceedings of the second international congress on ash technology and marketing, London, 1984; Abstract 398A, p. 12. 373. Massazza F. Pozzolanic cements: an answer to durability problems. In: Proceedings of the third NCB international seminar on cement and building materials, New Delhi; 1991, vol. 4, VIII-13/VI-27. 374. Jortzick H, Schubert P. Eigenschaften von Flugaschezement. Betonwerk + Fertigteil-Technik 1985;2:115–20. 375. Barker AB, Matthews JD. Concrete durability specification by water/cement or compressive strength for European cement types. In: Malhotra VM, editor. Proceedings of the third international conference on durability of concrete, Nice. American Concrete Institute Special Publication 145; 1994. p. 1135–59.

Pozzolanas and Pozzolanic Materials

461

376. Paillere AM, Raverdy M, Grimaldi G. Carbonation of concrete with low-calcium fly ash and granulated blast furnace slag: influence of air-entraining agents and freezing-and-thawing cycles. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, vol. I. p. 541–62. 377. Thomas MDA, Matthews JD. Carbonation of fly ash concrete. Mag Concr Res 1992;44(160):217–28. 378. Tsukayama R, Abe H, Nagataki S. Long-term experiments on the neutralisation of concrete mixed with fly ash and the corrosion of reinforcement. In: Malhotra VM, editor. Proceedings of the seventh international congress on the chemistry of cement, Paris; 30 Jun.–5 Jul. 1980, vol. Ill: IV-30–35. 379. Dhir RK, Matthews JS. Durability of pulverised fuel ash concrete. International Cement Review May 1992;75–81. 380. Meland I. Carbonation in hardened fly ash cements. In: Malhotra VM, editor. Proceedings of the international conference on blended cements in construction, Sheffield; 1991. p. 329. 381. Ho DWS, Lewis RK. Carbonation of concrete incorporating fly ash or a chemical admixture. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag, and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. I. p. 333–46. 382. British Standards Institution. BS 1881-210. Testing hardened concrete: Determination of the potential carbonation resistance of concrete. Accelerated carbonation method. London, UK: British Standards Institution; 2013. 383. Harrison TA, Jones MR, Newlands MD, Kandasami K, Khanna G. Experience of using the prTS12390-12 accelerated carbonation test to assess the relative performance of concrete. Mag Concr Res 2012;64(8):737–44. 384. Vennesland O, Gjorv OE. Silica concrete—protection against corrosion of embedded steel. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. II. p. 719–29. 385. Skjolsvold O. Carbonation depths of concrete with and without condensed silica fume. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, vol. II. p. 1031–48. 386. Gonzalez JA, Alonso C, Andrade C. Corrosion rate of reinforcements during accelerated carbonation in mortar made with different types of cement. In: Crane AP, editor. Corrosion of reinforcement in concrete construction. Society of Chemical Industry; 1983. Chapter 11. 387. Tuutti K. Corrosion of steel in concrete. Stockholm: Swedish Cement and Concrete Research Institute, S-100 44; 1982. 469 pp. 388. Arachchige ADM. Influence of cement content on corrosion resistance. In: Proceedings of the institution of civil engineers, construction materials; 2008, vol. 161, No. CM1. p. 31–9. 389. Kulakowski MP, Pereira FM, Dal Molin DCC. Carbonation-induced reinforcement corrosion in silica fume concrete. Constr Build Mater 2009;23:1189–95. 390. Boddy A, Bentz E, Thomas MDA, Hooton RD. An overview and sensitivity study of a multimechanistic chloride transport model. Cem Concr Res 1999;29:827–37. 391. Basheer L, Kropp J, Cleland DJ. Assessment of the durability of concrete from its permeation properties: a review. Constr Build Mater 2001;15:93–103. 392. Loser R, Lothenbach B, Leemann A, Tuchschmid M. Chloride resistance of concrete and its binding capacity—comparison between experimental results and thermodynamic modeling. Cem Concr Compos 2010;32:34–42. 393. Arya C, Buenfeld NR, Newman JB. Factors influencing chloride-binding in concrete. Cem Concr Res 1990;20(2):291–300. 394. Mangat PS, Molloy BT. Chloride binding in concrete containing PFA, GBS or silica fume under sea-water exposure. Mag Concr Res 1995;47 (171):129–41. 395. Jones MR, McCarthy MJ, Dhir RK. Chloride-resistant concrete. In: Dhir RK, Jones MR, editors. Proceedings of the international conference concrete, 2000. London: E & F N Spon; 1993. p. 1429–44. 396. Bamforth PB, Price WF, Emerson M. An international review of chloride ingress into structural concrete. Transport Research Laboratory, Contractor Report 359, Crowthorne, Berkshire; 1997. 163 pp. 397. Andrade C, Kropp J, editors. Testing and modelling chloride ingress into concrete. In: Proceedings of the third international workshop, Madrid, 2002, Proceedings PRO 38. Bagneux: RILEM; 2004. 449 pp. 398. Luping T, Nilsson L-O, Basheer PAM, editors. Resistance of concrete to chloride ingress. Testing and modelling. Florida, USA: CRC Press; 2011. 246 pp. 399. Collepardi M, Marcialis A, Turriziani R. La cinetica di penetrazione degli ioni cloruro nel calcestruzo. Il Cemento 1970;4:157–64. 400. Lukas W. Chloride penetration in standard concrete, water-reduced concrete, and superplasticized concrete. In: Malhotra VM, editor. Proceedings of the second international conference on developments in the use of superplasticizers, Ottawa. American Concrete Institute Special Publication 68; 1981. p. 253–67. 401. Papadakis VG. Effect of supplementary cementing materials on concrete resistance against carbonation and chloride ingress. Cem Concr Res 2000;30:291–9. 402. Thomas MDA, Bamforth PB. Modelling chloride diffusion in concrete. Effect of fly ash and slag. Cem Concr Res 1999;29:487–95. 403. Marusin SL. Chloride ion penetration in conventional concrete and concrete containing condensed silica fume. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, vol. II. p. 1119–33. 404. Detwiler RJ, Whiting DA, Lagergren ES. Statistical approach to ingress of chloride ions in silica fume concrete for bridge decks. ACI Mater J 1999;96 (6):670–5. 405. Shekarchi M, Rafiee A, Layssi H. Long-term chloride diffusion in silica fume concrete in harsh marine climates. Cem Concr Compos 2009;31:769–75. 406. Boddy A, Hooton RD, Gruber KA. Long-term testing of chloride-penetration resistance of concrete containing high-reactivity metakaolin. Cem Concr Res 2001;31:759–65.

462 Lea’s Chemistry of Cement and Concrete 407. Guneyisi E, Mermerdas K. Comparative study on strength, sorptivity, and chloride ingress characteristics of air-cured and water-cured concretes modified with metakaolin. Mater Struct 2007;40:1161–71. 408. Diamond S. Chloride concentrations in concrete pore solutions resulting from calcium and sodium chloride admixtures. Cem Concr Aggreg 1986;8:97–102. 409. Collepardi M, Marcialis A, Turriziani R. La penetrazione degli agenti anti-gelo nelle paste di cemento. Il Cemento 1972;3:143–50. 388. 410. Thomas MDA, Matthews JD. Performance of PFA concrete in a marine environment––10-year results. Cem Concr Compos 2004;26:5–20. 411. Parrott LJ. Contribution on paper ‘Marine performance of PFA concrete’. Mag Concr Res 1991;43:171–86. Mag Concr Res 1992;44:141–2. 412. Dhir RK, Jones MR, Seneviratne AMG. Diffusion of chlorides into concrete: Influence of PFA quality. Cem Concr Res 1991;21(6):1092–102. 413. Chindaprasirt P, Chotithanorm C, Cao HT, Sirivivatnanon V. Influence of fly ash fineness on the chloride penetration of concrete. Constr Build Mater 2007;21:356–61. 414. Glass GK, Buenfeld NR. The presentation of the chloride threshold level for corrosion of steel in concrete. Corros Sci 1997;39(5):1001–13. 415. Angst U, Elsener B, Larsen CK, Vennesland O. Critical chloride content in reinforced concrete—a review. Cem Concr Res 2009;39:1122–38. 416. Fajardo G, Valdez P, Pacheco J. Corrosion of steel rebar embedded in natural pozzolan based mortars exposed to chlorides. Constr Build Mater 2009;23:768–74. 417. Dhir RK, Jones MR, McCarthy MJ. PFA concrete: chloride-induced reinforcement corrosion. Mag Concr Res 1994;46(169):269–78. 418. Montemor MF, Simoes AMP, Salta MM. Effect of fly ash on concrete reinforcement corrosion studied by EIS. Cem Concr Compos 2000;22:175–85. 419. Rasheeduzzafar SE, Al-Saadoun SS, Al-Gahtani AS. Reinforcement corrosion-resisting characteristics of silica-fume blended-cement concrete. ACI Mater J 1992;89(4):337–44. 420. Batis G, Pantazopoulou P, Tsivilis S, Badogiannis E. The effect of metakaolin on the corrosion behavior of cement mortars. Cem Concr Compos 2005;27:125–30. 421. Parande AK, Babu BR, Karthik MA, Kumaar DKK, Palaniswamy N. Study on strength and corrosion performance for steel embedded in metakaolin blended concrete/mortar. Constr Build Mater 2008;22:127–34. 422. Buenfeld NR, Newman JB. The permeability of concrete in a marine environment. Mag Concr Res 1984;36(127):67–80. 423. Maultzsch M. Vorgange beim Angriff von Chloridlosungen auf Zementstein und Beton. Mater Tech 1984;3:83–90. 424. Chatterji S. Mechanism of the CaCl2 attack on Portland cement concrete. Cem Concr Res 1978;8:461–8. 425. Araibia A, Chabannet M, Bosc JL, Pera J. Role du metakaolin dans la resistance des betons de ciment Portland aux sulfates et chlorures. In: Malhotra VM, editor. Proceedings of the second international conference on durability of concrete, Montreal. American Concrete Institute Special Publication 126; 1991, Supplementary papers. p. 385–406. 426. Feldman RF, Chen-yi H. Resistance of mortars containing silica fume to attack by a solution containing chlorides. Cem Concr Res 1985;15:411–20. 427. Madej J. Corrosion resistance of normal and silica fume modified mortars made from different types of cements. In: Malhotra VM, editor. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication 132; 1992, vol. II. p. 1189–207. 428. Bonen D, Cohen MD. Magnesium sulfate attack on Portland cement paste: II—chemical and mineralogical analysis. Cem Concr Res 1992;22:707–18. 429. Monteiro PJM, Kurtis KE. Time to failure for concrete exposed to severe sulphate attack. Cem Concr Res 2003;33:987–93. 430. van Aardt JHP, Visser S. Influence of alkali on the sulphate resistance of ordinary Portland cement mortars. Cem Concr Res 1985;15:485–94. 431. Schiessel P, Haerdtl R. Relationship between durability and pore structure properties of concrete containing fly ash. In: Khayat KH, Aitcin PC, editors. Proceedings of the P. K. Mehta symposium on durability of concrete: part of the proceedings of the third international conference on durability of concrete, Nice; May 1994. p. 99–118. 432. Fraay A, Reigersman A, de Peel J. Sulfate resistance of mortars with pulverized fuel ash. In: Scanion JM, editor. Katharine and Bryant mather international conference on concrete durability. American Concrete Institute Special Publication 100; 1987, vol. II. p. 2042–58. 433. Irassar F, Batic O. Effects of low calcium fly ash on sulfate resistance of OPC cement. Cem Concr Res 1989;19:194–202. 434. Kalousek GL, Porter LC, Benton EJ. Concrete for long-time service in sulfate environment. Cem Concr Res 1972;2:79–89. 435. Mather K. Current research in sulfate resistance at the waterways experiment station. In: George Verbeck Symposium on sulfate resistance of concrete. American Concrete Institute Special Publication 77. p. 63–74. 436. Ducic V, Miletic S. Sulphate corrosion resistance of blended cement mortars. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, Supplementary paper no. 12. p. 26. 437. Hughes DC. Sulphate resistance of OPC, OPC/fly ash and SRPC pastes: pore structure and permeability. Cem Concr Res 1985;15:1003–12. 438. Mangat PS, Khatib JM. Influence of initial curing on sulphate resistance blended cement concrete. Cem Concr Res 1992;22:1089–100. 439. Stark D. Longtime study of concrete durability in sulfate soils. In: George Verbeck symposium on sulfate resistance of concrete. American Concrete Institute Special Publication 77; 1982. p. 21–40. 440. Chindaprasirt P, Homwuttiwong S, Sirivivatnanon V. Influence of fly ash fineness on strength, drying shrinkage and sulfate resistance of blended cement mortar. Cem Concr Res 2004;34:1087–92. 441. Freeman RB, Carrasquillo RL. Effects of intergrinding fly ash on the sulfate resistance of fly ash concrete. In: Malhotra VM, editor. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolanas in concrete, Istanbul. American Concrete Institute Special Publication 132; 1992, vol. I. 281–97. 442. Nasser KW, Lai PSH. Effect of sulfates on concrete containing fly ash. In: Malhotra VM, editor. Proceedings of the second international conference on durability of concrete, Montreal. American Concrete Institute Special Publication 126; 1991, Supplementary papers. p. 817–38. 443. Elmhirst NR. An examination of the properties of western coal fly ashes [MEng thesis]. University of Toronto; 1985. 141 pp.

Pozzolanas and Pozzolanic Materials

463

444. Manz OE, McCarthy GJ. Effectiveness of western U.S. high-lime fly ash for use in concrete. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, vol. I. p. 347–65. 445. Hooton RD. Influence of silica fume replacement of cement on physical properties and resistance to sulfate attack, freezing and thawing, and alkalisilica reactivity. ACI Mater J 1993;90:143–51. 446. Ping X, Beaudoin JJ. Mechanism of sulphate expansion II: validation of thermodynamic theory. Cem Concr Res 1992;22:845–54. 447. Nasser KW, Ghosh S. Durability properties of high strength concrete containing silica fume and lignite fly ash. In: Malhotra VM, editor. Proceedings of the third international conference on durability of concrete, Nice; 1994. American Concrete Institute Special Publication 145. p. 191–214. 448. Yamato T, Soeda M, Emoto Y. Chemical resistance of concrete containing condensed silica fume. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. II. p. 897–913. 449. Buck A. Use of cementitious materials other than Portland cement. In: Scanlon JM, editor. Katharine and Bryant Mather international conference on concrete durability. American Concrete Institute Special Publication 100; 1987, vol. II. p. 1863–81. 450. Torii K, Kawamura M. Effects of fly ash and silica fume on the resistance of mortar to sulfuric acid and sulfate attack. Cem Concr Res 1994;24 (2):361–70. 451. Lawrence CD. The influence of binder type on sulfate resistance. Cem Concr Res 1992;22:1047–58. 452. Bonen D. A microstructural study of the effect produced by magnesium sulfate on plain and silica fume-bearing Portland cement mortars. Cem Concr Res 1993;23:541–53. 453. Bonen D, Cohen MD. Magnesium sulfate attack on Portland cement paste: I—microstructural analysis. Cem Concr Res 1992;22:169–80. 454. Ouyang C, Nanni A, Chang WF. Internal and external sources of sulfate ions in Portland cement mortar: two types of chemical attack. Cem Concr Res 1988;18:699–709. 455. Hooton RD. Properties of a high-alkali lignite fly ash in concrete. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, vol. I. p. 333–45. 456. Durekovic A, Popovic K. Effect of ammonium nitrate solution on the structure and strength of cement pastes containing silica fume. Adv Cem Res 1988;1(3):138–46. 457. Bellman F, Stark J. The role of calcium hydroxide in the formation of thaumasite. Cem Concr Res 2008;38:1154–61. 458. K€ ohler S, Heinz D, Urbonas L. Effect of ettringite on thaumasite formation. Cem Concr Res 2006;36:697–706. 459. Bellman F, Stark J. Prevention of thaumasite formation in concrete exposed to sulphate attack. Cem Concr Res 2007;37:1215–22. 460. Mulenga DM, Stark J, Nobst P. Thaumasite formation in concrete and mortars containing fly ash. Cem Concr Compos 2003;25:907–12. 461. Tsivilis S, Kakali G, Skaropoulou A, Sharp JH, Swamy RN. Use of mineral admixtures to prevent thaumasite formation in limestone cement mortar. Cem Concr Compos 2003;25:969–76. 462. Mehta PK. Durability of concrete in marine environment—a review. In: Malhotra VM, editor. Proceedings of the international conference on performance of concrete in marine environment, St Andrews by the Sea. American Concrete Institute Special Publication 65; 1980. p. 1–20. 463. Biczok I. Concrete corrosion, concrete protection. Budapest: Akademiai Kiado; 1964. 543. 464. Locher FW. Influence of chloride and hydrocarbonate on the sulphate attack. In: Proceedings of the fifth international symposium on the chemistry of cement, Tokyo; 1968, vol. Ill. p. 328–35. 465. Tenoutasse M, Van Bemst A, De Keyser WL. Contribution a la determination de la nature des phases hydrates dans des mortiers immerges durant 30 ans. In: Proceedings of the international symposium behaviour of concretes exposed to sea water, Palermo; 1965. p. 57–88. 466. Seidell A, Linke WF. Solubilities of inorganic and metal organic compounds. I:Washington: American Chemical Society; 1958. p. 1487. 467. Regourd M. Physico-chemical studies of cement pastes, mortars, and concretes exposed to sea water. In: Malhotra VM, editor. Proceedings of the international conference on performance of concrete in marine environment, St. Andrews by the Sea. American Concrete Institute Special Publication 65; 1980. p. 63–82. 468. Paille`re AM, Platret G, Roussel P, Gawsewitch J. Influence of curing time on behaviour in seawater of high-strength mortar with silica fume. In: Malhotra VM, editor. Proceedings of the second international conference on durability of concrete, Montreal. American Concrete Institute Special Publication 126; 1991, vol. I. p. 559–75. 469. Pallie`re AM, Platret G, Roussel P, Gawsewitch J. Influence of mechanical strength and curing methods on sea water durability of mortars containing fly ashes and slag. In: Malhotra VM, editor. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication 132; May 1992, vol. I. p. 179–204. 470. Wang H, Gillott JE. Mechanism of alkali-silica reaction and the significance of calcium hydroxide. Cem Concr Res 1991;21:647–54. 471. Sprung S. Influences on the alkali-aggregate reaction in concrete. In: Proceedings of the symposium on alkali-aggregate reaction—preventive measures, Reykjavik; 1975. p. 231–44. 472. Krell J. Influence of mix design on alkali-silica reaction in concrete. In: Proceedings of the seventh international conference on alkali-aggregate reaction, Ottawa; 1986. p. 452–5. ^ 473. Ranc R, Sorrentino D, Cariou B. Reactions silico-alcalines mise au point de methodes fiables. Annales de I’lnstitut Technique du Batiment et des Travaux Publics 1990;107–31. 474. Berube MA, Duchesne J. Evaluation of testing methods used for assessing the effectiveness of mineral admixtures in suppressing expansion due to alkali–aggregate reaction. In: Malhotra VM, editor. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication 132; 1992, vol. I. p. 549–75. 475. Lerch W. A cement-aggregate reaction that occurs with certain sand-gravel aggregates. J Portland Cem Assoc Res Dev Lab Sep. 1959;42–50.

464 Lea’s Chemistry of Cement and Concrete 476. Alasali MM, Malhotra VM, Soles JA. Performance of various test methods for assessing the potential alkali reactivity of some Canadian aggregates. CANMET Mineral Science Laboratories, Division Report MSL 1989;89–128. 477. Stanton TE. Expansion of concrete through reaction between cement and aggregate. Proc Am Soc Civ Eng 1940;1781–811. 478. Smith RL, Raba CF Jr. Recent developments in the use of fly ash to reduce alkali–aggregate reaction. In: Dodson editor. Symposium on alkalis in concrete, Los Angeles. American Society for Testing and Materials Special Technical Publication 930; 1985. 479. Geiker M, Thaulow N. The mitigating effect of pozzolans on alkali silica reactions. In: Malhotra VM, ed. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication 132; May 1992. vol. I, p. 533–48. 480. Berra M, Di Maggio R, Mangialardi T, Paolini AE. Fused quartz as a reference reactive aggregate for alkali-silica reaction studies. Adv Cem Res 1995;7 (25):21–32. 481. Stanton TE. Studies of the use of pozzolanas for counteracting excessive concrete expansion resulting from reaction between aggregates and the alkali in cement. In: Symposium on use of pozzolanic materials in mortars and concretes, American Concrete Institute Special Publication San Francisco. American Society for Testing and Materials Special Technical Publication; 1949. 482. Gudmundsson G. Investigation on Icelandic pozzolans. In: Proceedings of the symposium on alkali-aggregate reaction-preventive measures, Reykjavik; 1975. p. 65–76. 483. Kordina K, Schwick W. Investigation on additives to concrete for the prevention of alkali-aggregate reaction. Betonwerk + Fertigteil-Technik 1981;66:328–31. 484. Sersale R, Frigione G. Portland–zeolite-cement for minimizing alkali-aggregate expansion. Cem Concr Res 1987;17:404–10. 485. Oberholster RE, Westra BW. The effectiveness of mineral admixtures in reducing expansion due to alkali-aggregate reaction with Malmesbury group aggregates. In: Proceedings of the fifth international conference on alkali–aggregate reaction in concrete, Cape Town; 1981. p. S252–31. 486. Chatterji S, Thaulow N, Christensen P, Jensen AD. Studies of alkali-silica reaction with special reference to prevention of damage to concrete—a preliminary study. In: Proceedings of the sixth international conference on alkali in concrete: research and practice, Technical University of Denmark, Copenhagen; 22–25 Jun. 1983. p. 253–60. 487. Walters SV, Jones TR. Effect of metakaolin on alkali-silica reaction (ASR) in concrete manufactured with reactive aggregate. In: Malhotra VM, editor. Proceedings of the second international conference on durability of concrete, Montreal. American Concrete Institute Special Publication 126; 1991, vol. II. p. 941–53. 488. Nixon PJ, Gaze ME. The use of fly ash and granulated blastfurnace slag to reduce expansion due to alkali–aggregate reaction. In: Proceedings of the fifth international conference on alkali–aggregate reaction in concrete, Cape Town; 1981. S252–32. 489. Hobbs DW. Expansion due to alkali-silica reaction and the influence of pulverised fuel ash. In: Proceedings of the fifth international conference on alkali–aggregate reaction in concrete, Cape Town; 1981. p. S252–30. 490. Meland I. Use of fly ash in cement to reduce alkali-silica reactions. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, vol. I. 591–608. 491. Nixon PJ, Gaze ME. The effectiveness of fly ashes and granulated blastfurnace slags in preventing AAR. In: Proceedings of the sixth international congress on alkali in concrete: research and practice, Technical University of Denmark, Copenhagen; 22–25 Jun. 1983. p. 61–8. 492. Ramachandran S, Ramakrishnan V, Johnston D. The role of high volume fly ash in controlling alkali-aggregate reactivity. In: Malhotra VM, editor. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication 132; 1992, vol. I. p. 591–613. 493. Fournier B, Langley WS, Malhotra VM. Effectiveness of fly ash in reducing expansion of concrete made with reactive aggregates from New Brunswick. ACI Special Publication 1995;153:561–90. 494. Nagataki S, Ohga H, Inoue T. Evaluation of fly ash for controlling alkali–aggregate reaction. In: Malhotra VM, editor. Proceedings of the second international congress on durability of concrete, Montreal. American Concrete Institute Special Publication 126; 1991. p. 955–72. 495. Sprung S. Einfluss von Zement und Zusatzen auf die Alkalireaktion. Schriftenreihe der Zementindustrie 1973;40:69–78. 496. Dhir RK, Dyer TD, Tang M. Alkali-silica reaction in concrete containing glass. Mater Struct 2009;42:1451–62. 497. Aquino W, Lange DA, Olek J. The influence of metakaolin and silica fume on the chemistry of alkali–silica reaction products. Cem Concr Compos 2001;23:485–93. 498. Lee C. Effects of alkalies in class C fly ash on alkali–aggregate reaction. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114, 1989; vol. I. p. 417–30. 499. Klieger P, Gebler S. Fly ash and concrete durability. In: Scanlon JM, editor. Concrete durability. Katharine and Bryant Mather international conference, Detroit, USA. Publication no. ACI SP-100. American Concrete Institute, Detroit; 1987, vol. 1, p 1043–69. [Discussion: ACI Materials Journal, May–Jun. 1988;85(3):214–s15]. 500. Lane DS. Comparisons of results from C441 and C1293 with implications for establishing criteria for ASR-resistant concrete. Cem Concr Aggreg Dec. 1999;21(2):149–56. 501. Ohga H, Nagataki S. Effect of fly ash on alkali-aggregate reaction in marine environment. In: Malhotra VM, editor. Proceedings of the fourth international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Istanbul. American Concrete Institute Special Publication; 1992, vol. I, p. 577–90. 502. Sprung S, Adabian M. The effect of admixture on the alkali-aggregate reaction in concrete. In: Proceedings of the symposium on the effect of alkalies on the properties of concrete, London; 1976. p. 125–37. 503. Ukita K, Shigematsu S-i, Ishii M, Yamamoto K, Azuma K, Moteki M. Effect of classified fly ash on alkali aggregate reaction (AAR). In: Proceedings of the eighth international conference on alkali-aggregate reaction, Kyoto; 1989. p. 259–64.

Pozzolanas and Pozzolanic Materials

465

504. Soles JA, Malhotra VM, Suderman RW. The role of supplementary cementing materials in reducing the effects of alkali-aggregate reactivity: CANMET investigations. In: Grattean-Bellew PE, editor. Proceedings of seveth international conference on alkali-aggregate reaction. National Research Council of Canada, Ottowa, Canada. Park Ridge, New Jersey, USA: Noyes Publications; 1986. p. 79–84. 505. Perry C, Day RL, Joshi RC, Langan BW, Gillot JE. The effectiveness of twelve Canadian fly ashes in suppressing expansion due to alkali-silica reaction. In: Grattean-Bellew PE, editor. Proceedings of seventh international conference on alkali-aggregate reaction. National Research Council of Canada, Ottowa, Canada. Park Ridge, New Jersey, USA: Noyes Publications; 1986. p. 93–7. 506. Shayan A, Diggins R, Ivanusec I. Effectiveness of fly ash in preventing deleterious expansion due to alkali-aggregate reaction in normal and steamcured concrete. Cem Concr Res 1996;26(1):153–64. 507. Hobbs DW. Influence of pulverized-fuel ash and granulated blastfurnace slag upon expansion caused by the alkali-silica reaction. Mag Concr Res 1982;34:83–93. 508. Kobayashi S, Hozumi Y, Nakano T, Yanagida T. Study on the effect of fly ash quality for controlling alkali-aggregate reactions. In: Malhotra VM, editor. Fly ash, silica fume, slag and natural pozzolans. proceedings third international conference, Trondheim, Norway. Detroit: American Concrete Institute; 1989, Publication no. ACI SP-114, vol. 1. p. 403–15. 509. Farbiarz J, Carrasquillo RL. Alkali-aggregate reaction in concrete containing fly ash. In: Scanlon JM, editor. Concrete durability. Katharine and Bryant Mather international conference, Detroit, USA. Detroit: American Concrete Institute; 1987. Publication no. ACI SP-100, vol. 2. p. 1787–808. 510. Durand B, Berard J, Soles JA. Comparison of the effectiveness of four mineral admixtures to counteract alkali-aggregate reaction. In: Grattean-Bellew PE, editor. Proceedings of seventh international conference on alkali-aggregate reaction. National Research Council of Canada, Ottowa, Canada, 1986. Park Ridge, New Jersey, USA: Noyes Publications; 1987. p. 30–5. 511. Chunxiang Q, Hongding G, Mingshu T. Mechanism of mineral admixture suppressing alkali-silica reaction: I—corrosion degree of reactive aggregate in blended cement pastes and its correlations with expansion value and electric resistance change. Cem Concr Res 1994;24(6):1111–20. 512. Zhang X, Groves GW. Microstructural studies of alkali-silica reaction products. Mater Sci Technol 1989;5(7):714. 513. Hobbs DW. Effect of mineral and chemical admixtures on alkali aggregate reaction. In: Proceedings of the eighth international conference on alkali– aggregate reaction, Kyoto; 1989. p. 173–86. 514. Porter LC. Small proportions of pozzolan may produce detrimental reactive expansion in mortar. Denver, US Department of the Interior, Bureau of Reclamation; 1964, Report no C-113. 22 pp. 515. Duchesne J, Berube MA. The effectiveness of supplementary cementing materials in suppressing expansion due to ASR: another look at the reaction mechanisms: II—pore solution chemistry. Cem Concr Res 1994;24(2):221–30. 516. Diamond S. Alkali reactions in concrete-pore solution effects. In: Proceedings of the sixth international congress on alkali in concrete: research and practice, Technical University of Denmark, Copenhagen; 22–25 Jun. 1983. p. 155–66. 517. Kawamura M, Takemoto K. Correlation between pore solution composition and alkali silica expansion in mortars containing various fly ashes and blastfurnace slags. Int J Cem Compos Light Concr 1988;10(4):215–23. 518. Durand B, Berard J, Roux R, Soles JA. Alkali-silica reaction: the relation between pore solution characteristics and expansion test results. Cem Concr Res 1990;20:419–28. 519. Kollek JJ, Varma SP, Zaris C. Measurement of OH ion concentrations of pore fluids and expansion due to alkali-silica reaction in composite cement mortars. In: Proceedings of the eighth international conference on the chemistry of cement, Rio de Janeiro; 1986, vol. IV. p. 183–9. 520. Nixon PJ, Page CL, Bollinghaus R, Canham I. The effect of a PFA with a high total alkali content on pore solution composition and alkali silica reaction. Mag Concr Res 1986;38(134):30–5. 521. Chunxiang Q, Hongding G, Xianghui L, Mingshu T. Mechanism of mineral admixture suppressing alkali-silica reaction: II—retardation of the transport of Na, K and OH ions in pore structure caused by acidic action of mineral admixture particles in matrix. Cem Concr Res 1994;24(7):1327–34. 522. Koelliker E. Some observations of the carbonation of solutions of Ca(OH)2 and liquid phase of cement paste and the mechanism of carbonation of concrete. In: RILEM international symposium on carbonation of concrete, Fulmer; 1976. 523. Bhatty MSY, Greening NR. Some long time studies of blended cements with emphasis on alkali-aggregate reaction. In: Proceedings of the seventh international conference on alkali-aggregate reaction, Ottawa; 1986. p. 85–92. 524. Mins-shu T, Yu-feng Y, Mei-qi Y, Shi-hua Z. The preventive effect of mineral admixtures on alkali-silica reaction and its mechanism. Cem Concr Res 1983;13:171–6. 525. Chatterji S. The role of Ca(OH)2 in the breakdown of Portland cement concrete due to alkali–silica reaction. Cem Concr Res 1979;9:185–8. 526. Ming-shu T, Su-Fen H. Effect of Ca(OH)2 on alkali silica reaction. In: Proceedings of the seventh international conference on the chemistry of cement, Paris; 30 June–5 July 1980, vol. II. p. 11-94–9. 527. Mullick AK, Wason RC, Rajkumar C. Performance of commercial blended cements in alleviating ASR. In: Proceedings of the eighth international conference on alkali–aggregate reaction, Kyoto; 1989. p. 217–22. 528. Duchesne J, Berube MA. The effectiveness of supplementary cementing materials in suppressing expansion due to ASR: another look at the reaction mechanisms: I—concrete expansion and Portlandite depletion. Cem Concr Res 1994;24:73–82. 529. Berube MA, Carles-Gibergues A, Duchesne J, Naproux P. Influence of particle size distribution on the effectiveness pf type-F fly ash in suppressing expansion due to alkali–silica reactivity. In: Malhotra VM, editor. Proceedings of the fifth international conference on fly ash, silica fume, slag, and natural pozzolanas, Milwaukee. American Concrete Institute Special Publication 153; Jun. 1995, vol. I. p. 177–99. 530. Dyer TD. Concrete durability. Boca Raton, FL, USA: CRC Press-Taylor & Francis Group; 2014. 531. Pettersson K. Effects of silica fume on alkali-silica expansion in mortar specimens. Cem Concr Res 1992;22:15–22. 532. Wang H, Gillott JE. Competitive nature of alkali-silica fume and alkali–aggregate (silica) reaction. Mag Concr Res 1992;44(161):235–9.

466 Lea’s Chemistry of Cement and Concrete 533. Farbiarz J, Carrasquillo RL, Snow PC. Alkali aggregate reaction in concrete containing fly ash. In: Proceedings of the seventh international conference on alkali–aggregate reaction, Ottawa; 1986. p. 55–9. 534. De Belie N, Verselder HJ, De Blaere B, Van Nieuwenburg D, Verschoore R. Influence of cement type on the resistance of concrete to feed acids. Cem Concr Res 1996;26:1717–25. 535. De Belie N, De Coster V, Van Nieuwenburg D. Mag Concr Res 1997;49:337–44. 536. Bertron A, Duchesne J, Escadeillas G. Attack of cement pastes exposed to organic acids in manure. Cem Concr Compos 2005;27:898–909. 537. Roy DM, Arjunan P, Silsbee MR. Effect of silica fume, metakaolin, and low-calcium fly ash on chemical resistance of concrete. Cem Concr Res 2001;31:1809–13. 538. Chang Z, Song X, Munn R, Marosszeky M. Using limestone aggregates and different cements for enhancing resistance of concrete to sulphuric acid attack. Cem Concr Res 2005;35:1486–94. 539. Tamimi AK. High performance concrete mix for an optimum protection in acidic conditions. Mater Struct 1997;30:188–91. 540. Siad H, Mesbah HA, Khelafi H, Kamali-Bernard S, Mouli M. Effect of mineral admixture on resistance to sulphuric and hydrochloric acid attacks in self-compacting concrete. Can J Civ Eng 2010;37:441–9. 541. Khedr SA, Abou-Zeid MN. Characteristics of silica-fume concrete. J Mater Civ Eng 1994;6:357–75. 542. Hanayneh B, Shatarat N, Katkhuda H. Improving durability of concrete to phosphoric acid attack. Jordan J Civ Eng 2012;6:68–82. 543. Durning TA, Hicks C. Using microsilica to increase concrete’s resistance to aggressive chemicals. Concr Int 1991;13:42–8. 544. Mehta PK. Studies on chemical resistance of low water/cement ratio concretes. Cem Concr Res 1985;15:969–78. 545. Sersale R, Frigione G, Bonavita L. Acid depositions and concrete attack: main influences. Cem Concr Res 1998;28:19–24. 546. T€ urkel S, Felekog˘lu B, Dulluc¸ S. Influence of various acids on the physico–mechanical properties of pozzolanic cement mortars. Sadhana 2007;32:683–91. 547. Adesanya DA, Raheem AA. A study of the permeability and acid attack of corn cob ash blended cements. Constr Build Mater 2010;24:403–9. 548. Construction Industry Research and Information Association. Freeze/thaw resisting concrete—its achievement in the UK. London, UK: CIRIA; 2001. 549. Aitcin PC. The durability characteristics of high performance concrete: a review. Cem Concr Compos 2003;25:409–20. 550. Virtanen J. Freeze-thaw resistance of concrete containing blast-furnace slag, fly ash or condensed silica fume. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79; 1983, vol. II. p. 923–94. 551. Malhotra VM, Mehta PK. Pozzolanic and cementitious materials. Amsterdam, the Netherlands: Gordon and Breach Publishers; 1996. 191 pp. 552. Peng G-F, Ma Q, Hu H-M, Gao R, Yao Q-F, Liu Y-F. The effects of air entrainment and pozzolans on frost resistance of 50–60 MPa grade concrete. Constr Build Mater 2007;21:1034–9. 553. Du L, Folliard KJ. Mechanisms of air entrainment in concrete. Cem Concr Res 2005;35(8):1463–71. 554. Pedersen KH, Jenson AD, Skjøth-Rasmusson MS, Dam-Johansen K. A review of the interference of carbon containing fly ash with air-entrainment in concrete. Prog Energy Combust Sci 2008;34:135–54. 555. Sturrup VR, Hooton RD, Clendenning TG. Durability of fly ash concrete. In: Malhotra VM, editor. Proceedings of the first international conference on the use of fly ash, silica fume, slag and other mineral by-products in concrete, Montebello, Canada. American Concrete Institute Special Publication 79;1983, vol. I. p. 71–8. 556. McCarthy MJ, Strompinis N, Csetenyi LJ, Sadiqul Islam GMS. Influence of Portland cement characteristics on air-entrainment in fly ash concrete. Mag Concr Res 2015;67(14):786–97. 557. Nehdi M, Duquette J, El Damatty A. Performance of rice husk ash produced using a new technology as a mineral admixture in concrete. Cem Concr Res 2003;33:1203–10. 558. Christodoulou G. A comparative study of the effects of silica fume, metakaolin and PFA on the air content of fresh concrete. In: Presented at the Young researchers forum, organised by the SCI construction materials group. London, UK: SCI; Held 27 Apr. 2000. p. 1/17–17/17. 559. Hasholt MT. Air void structure and frost resistance: a challenge to Powers’ spacing factor. Mater Struct 2013;47:911–23. 560. Dhir RK, McCarthy MJ, Limbachiya MC, Zhang DS, El-Sayed HI. PFA concrete: air entrainment and freeze/thaw durability. Mag Concr Res Feb. 1999;51(1):53–64. 561. Pigeon M, Plante P, Plante M. Air void stability, Part I: Influence of silica fume and other parameters. ACI Mater J 1989;86(5):482–90. 562. Mehta PK. Natural pozzolans. In: Malhotra VM, editor. Supplementary cementing materials for concrete. Minister of Supply and Services Canada; 1987. p. 1–33. 563. Gebler SH, Klieger P. Effect of fly ash on the durability of air-entrained concrete. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, vol. I. p. 483–519. 564. Soretz S. Influence of cement on the frost resistance. In: International Colloquium Frostbestandigkeit € von Beton, Vienna. Mitteilungen aus Forschung€ institut der Vereins der Osterreichen Zementfabrikanten; Jun. 1980. p. 159–67. 565. Naik TR, Singh SS, Hossain MM. Properties of high performance concrete systems incorporating large amounts of high-lime fly ash. Constr Build Mater 1995;9(4):195–204. 566. Yamato T, Emoto Y, Soeda M. Strength and freezing-and-thawing resistance of concrete incorporating condensed silica fume. In: Malhotra VM, editor. Proceedings of the second international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Madrid. American Concrete Institute Special Publication 91; 1986, vol. I. p. 1095–119.

Pozzolanas and Pozzolanic Materials

467

567. Galeota D, Giammatteo MM, Marino R, Volta V. Freezing and thawing resistance of non air-entrained and air-entrained concretes containing a high percentage of condensed silica fume. In: Malhotra VM, ed. Proceedings of the second international conference on durability of concrete, Montreal. American Concrete Institute Special Publication 126; 1991, vol. I. p. 249–616. 568. Massazza F, Oberti G. Durability of pozzolanic cements and Italian experience in mass concrete. In: Malhotra VM, editor. Proceedings of the second international conference on durability of concrete, Montreal. American Concrete Institute Special Publication 126; 1991, vol. II. p. 1259–83. 569. Brown PW, Clifton JR, Frohnsdorff G, Berger RL. Limitations to fly ash use in blended cements. In: Fourth international ash utilisation symposium, St Louis; 1976. ERDA MERC/SP-76/4. p. 518–29. 570. Davis RE, Davis HE, Kelly JW. Weathering resistance of concretes containing fly-ash cements. J Am Concr Inst 1941;37:281–93. 571. Barrow RS, Hadchiti KM, Carrasquillo PM, Carrasquillo RL. Temperature rise and durability of concrete containing fly ash. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. I. p. 331–47. 572. Langlois M, Beaupre D, Pigeon M, Foy C. The influence of curing on the salt scaling resistance of concrete with and without silica fume. In: Malhotra VM, editor. Proceedings of the third international conference on fly ash, silica fume, slag, and natural pozzolans in concrete, Trondheim. American Concrete Institute Special Publication 114; 1989, vol. II. p. 971–90. 573. Yen T, Hsu T-H, Liu Y-W, Chen S-H. Influence of class F fly ash on the abrasion-erosion resistance of high strength concrete. Constr Build Mater 2007;21:458–63. 574. Atiş CD, C ¸ elik ON. Relation between abrasion resistance and flexural strength of high volume fly ash concrete. Mater Struct 2002;35:257–60. 575. Siddique R. Effect of fine aggregate replacement with Class F fly ash on the abrasion resistance of concrete. Cem Concr Res 2003;33:1877–81. 576. Liu Y-W. Improving the abrasion resistance of hydraulic-concrete containing surface crack by adding silica fume. Constr Build Mater 2007;21:972–7. 577. Ghafoori N, Diawara H. Abrasion resistance of fine aggregate replaced silica fume concrete. ACI Mater J 1999;96:559–67. 578. Rashad AM. A preliminary study on the effect of fine aggregate replacement with metakaolin on strength and abrasion resistance of concrete. Constr Build Mater 2013;44:487–95. 579. Yetgin Ş, C¸avdar A. Abrasion resistance of cement mortar with different pozzolanic compositions and matrices. J Mater Civ Eng 2011;23:138–45. 580. Naik TR, Singh SS, Ramme BW. Effect of source of fly ash on abrasion resistance of concrete. J Mater Civ Eng 2002;14:417–26. 581. Thomas MDA. Optimising the use of fly ash in concrete. Skokie, Illinois: Portland Cement Association; 2007. 24 pp.

FURTHER READING 582. Hjorth L. Cement specifications for concrete exposed to chlorides and sulphates. In: CEB-RILEM international workshop on durability of concrete structures, Copenhagen; 1983. p. 229–35. 583. Pavlı´k V, Uncˇik S. The rate of corrosion of hardened cement pastes and mortars with additive of silica fume in acids. Cem Concr Res 1997;27:1731–45.