Long term study of hardened cement pastes containing silica fume and fly ash

Long term study of hardened cement pastes containing silica fume and fly ash

Construction and Building Materials 60 (2014) 48–56 Contents lists available at ScienceDirect Construction and Building Materials journal homepage: ...

2MB Sizes 0 Downloads 35 Views

Construction and Building Materials 60 (2014) 48–56

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Long term study of hardened cement pastes containing silica fume and fly ash Vili Lilkov a, Ivan Rostovsky b, Ognyan Petrov c,⇑, Yana Tzvetanova c, Plamen Savov a a

University of Minig and Geology ‘‘St. Ivan Rilski’’, Sofia, Bulgaria University of Architecture, Civil Engineering and Geodesy, Sofia, Bulgaria c Institute of Mineralogy and Crystallography ‘‘Akad. Ivan Kostov’’, Sofia, Bulgaria b

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 Fly ash and silica fume participate

actively in the hardening of cement paste.  The used additives stimulate hydrate products with temperature of dehydration >600 °C.  Polymerization of calcium hydrosilicates is observed after 600 days of hardening.  The portion of products dehydrating below 340–350 °C is increased.  Carbonation and polymerization in the hardening cement are enhanced by the additives.

a r t i c l e

i n f o

Article history: Received 7 June 2013 Received in revised form 17 February 2014 Accepted 21 February 2014 Available online 20 March 2014 Keywords: Silica fume Fly ash Pozzolit Cement paste Hydration products

a b s t r a c t The hydration products in cements with individual addition of SF and two types of FA from TEPS and mixtures of FA and SF are studied. The phase transformations are followed in a 4-year period. Fly ashes from TEPS ‘‘Bobov Dol’’ and TEPS ‘‘Pernik’’, silica fume from the metallurgical plant ‘‘Kremikovtsi’’, and their mixed product Pozzolit are suitable pozzolanic additions to cement as they participate actively in the processes of formation of hardened cement paste. The above additives stimulate the formation of hydrate products with temperature of dehydration above 600 °C, more clearly expressed after the 48th day of hydration. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction One of the most widely used active mineral additions are silica fume [1–6] and fly ash from TEPS [7–11]. They exert positive influence on a number of important features of hardened cement paste

⇑ Corresponding author. Tel.: +359 2 9797055; fax: +359 2 9797056. E-mail address: [email protected] (O. Petrov). http://dx.doi.org/10.1016/j.conbuildmat.2014.02.045 0950-0618/Ó 2014 Elsevier Ltd. All rights reserved.

and concrete – physical–chemical, non-permeability and corrosion resistance. The addition of a mixture of silica fume and fly ash to cement is one of the ways to obtain concrete with high exploitation characteristics (high performance concrete) [12–20]. By use of fly ash there can be solved some problems concerned with the workability, which occur in case of addition of greater quantities of silica fume and at the same time silica fume compensates for the relatively low initial strength of the cement composites with added fly ash.

V. Lilkov et al. / Construction and Building Materials 60 (2014) 48–56

The hydration processes in cement are most intensive in the initial periods of hardening and then the type and quantity of the hydration products change slower and the degree of hydration is usually evaluated by inspection of the chemically bound water, the quantity of the calcium hydroxide, and the lowering of the quantity of the tri-calcium silicate in the clinker minerals. Important part in the hydration of cement is the interaction of the tri-calcium silicate (alite, C3S) with water to form hydrosilicate gel (C–S–H) with ratio CaO/SiO2 = 1.6–1.7 [21,22].

3CaO  SiO2 þ zH2 O ! Cax SiðOHÞy  nH2 O þ ð3  xÞCaðOHÞ2

ð1Þ

The active mineral additives, known also as pozzolana, react with portlandite according to the following scheme

49

of diffusion of the hardened cement paste [33–35]. The volume changes, which accompany the processes of carbonation lead to filling of the empty pore volumes with Ca-carbonates and densify the structure of the hardened cement paste. The carbonation of portlandite, Ca-hydrosilicates, and ettringite proceeds according the following scheme [33,36]

CaðOHÞ2 þ CO2 ! CaCO3 þ H2 O;

ð4Þ

3CaO  2SiO2  3H2 O þ 3CO2 ! 3CaCO3 þ 2SiO2 þ 3H2 O;

ð5Þ

3CaO  Al2 O3  3CaSO4  32H2 O þ 3CO2 ! 3CaCO3 þ 3ðCaSO4  2H2 OÞ þ Al2 O  xH2 O þ ð26  xÞH2 O: ð6Þ

4

SiO4 þ xCaðOHÞ2 þ ðy  2xÞOH þ H2 O ! Cax SiðOHÞy  nH2 O:

ð2Þ

As a result, C–S–H gel is formed with lower CaO/SiO2 ratio. The above described reaction is known as pozzolanic reaction. The silica fume (SF) accelerates the early hydration of C3S ensuring enough quantity of active centres as a basis for the formation of C–S–H. At this stage the intensity of the process is determined by the specific surface of the SF grains [23]. At different stages of hydration of cement with presence of silica fume there are formed three types of hydrosilicates: – From the hydration of tricalcium silicate – reaction (1). – From the pozzolanic reaction between SF and portlandite, resulting in formation of Ca hydrosilicates similar to the ones obtained during hydration of C3S but with lower CaO/SiO2 ratio. The intensity of portlandite formation in the early stages is higher than the speed of the reaction of its consumption by silica fume. – When the quantity of portlandite significantly lowers then SF reacts with C–S–H to form hydrosilicates with low CaO/SiO2 ratio, which at later stages of hardening are characterized with high degree of polymerization [23–26]. The fly ashes, which contain alumina also react with portlandite (reaction (3)) to form Ca alumo-hydrosilicates 4

2AlðOHÞ4 þ 3Ca2þ þ SiO4 ! Ca3 Al2 SiðOHÞ8 O4

ð3Þ

With inclusion of pozzolanic additives in cement the quantity of portlandite cannot be used as a reliable criterion for the degree of hydration due to the continuing pozzolanic reactions in the hydrating cement. In case of high concentrations of the mineral additions it may be very low even missing irrespective of the fact that the hydration is significant [24,27,28]. From another side, the addition of silica fume and fly ash stimulates the polymerization of the hydrosilicates, which are formed during hydration of cement accompanied by evolution of water. Therefore, the quantity of chemically bound water is not a correct indicator for determining the degree of hydration of cements with pozzolanic additions [18,24,29–32]. Portlandite and the Ca-hydrosilicates, which comprise about 85% of the weight of the hardened cement paste, are capable of forming carbonate in a longer time period. When the cement composites are subjected to the action of carbon dioxide, the latter is dissolved in the pore-liquid of the cement pastes ensuring CO2– 3 ions, which react with Ca2+ ions of portlandite and the Ca-hydrosilicate gel to form calcite – CaCO3. In case of prolonged contact of the cement pastes with air and water it is possible to reach complete carbonation with small residue of ettringite, carbo-aluminates and hydrosilicates. The depth of the formed carbonate layer depends on the time of contact with CO2 and its concentration in the surrounding environment as well as on the coefficient

The carbonation in the presence of SF and fly ash is a more complicated process due to the pozzolanic reactions. The products, subjected to carbonation (CH and C–S–H) participate in pozzolanic reactions or they themselves are obtained as a result of these reactions [33,36]. The carbonation of CH and C–S–H proceeds simultaneously [37–39]. With the development of the process and the accompanying lowering of the portlandite quantity the mean length of the polymer chains of C–S–H progressively increases, the excess of water is released [38] and with those processes one can explain the carbonation shrinkage [40]. The process is markedly expressed in cements with pozzolanic additives in which there are formed Ca-hydrosilicates with low CaO/SiO2 – ratio as well as in aggressive carbonating media [41]. Thus the pozzolanic additions modify the cement matrix, alterations occur in it, which develop with time. The changes depend on many factors – chemical composition and quantity of the additions, specificity of the interactions with the hydration products of cement, the quantity of water expressed by the water to cement ratio. Therefore, to fully clarify the above dependencies there is need to conduct a longterm study of the type and quantity of the hydration products and their transformations in respect to any additive alone or in combination. The studies on cements with separate addition of silica fume or fly ash are numerous but the studies on the structure of hydration products in cements for cases of combined addition SF and FA are restricted [19,42]. The recent paper reports results on studies on the type and quantity of the hydration products in cements with individual addition of SF and two types of FA from TEPS and mixtures of FA and SF, the latter being a product according the trade mark ‘‘Pozzolit’’ and their transformations are followed in a 4-year period. 2. Materials and methods 2.1. Materials The cement (PC) is produced by the cement plant ‘‘Titan’’ (Zlatna Panega, Bulgaria) – CEM I 42.5 N with mineral composition: C3S – 57.52%; C2S – 23.48%; C3A – 5.54%; C4AF – 11.7%, with addition of 1.76% gypsum and density of 3 g/cm3. Two types of FA are used (Table 1), both with the particle size 100% below 63 lm, respectively, from ‘‘TEPS ‘‘Pernik’’ (Pernik, Bulgaria) (sample FA1) and TEPS ‘‘Bobov Dol’’ (Bulgaria) (FA2) as well as silica fume (SF) with particle size 90% below 10 lm and 60% below 1 lm. The specific surface of the particles of the mineral additives (determined by BET method) is: SF – 18.6 m2/g; FA1 – 1.76 m2/g; FA2 – 2.67 m2/g. Applying mechanical mixing of FA and SF in weight ratio 50%:50% there have been prepared two mineral additives Pozzolit 1 (Pz1) and Pozzolit 2 (Pz2), respectively from FA1 and SF and from FA2 and SF.

2.2. Samples The samples were prepared with water to binder ratio W/C = 0.5 (Table 2). The cement and the mineral addition were homogenized in a dry state. The dry materials were added into water and the resulting mixture was stirred for 5 min. The hydrating pastes were hermetically closed in plastic containers up to the 24-th hour and then the material was released from the containers and kept in water at 20 °C up to the age of testing.

50

V. Lilkov et al. / Construction and Building Materials 60 (2014) 48–56

Table 1 Chemical composition of Portland cement and the mineral additives. Chemical content

PC (%)

FA1 (%)

FA2 (%)

SF (%)

SiO2 Fe2O3 (t) Fe2O3 FeO TiO2 Al2O3 MnO CaO MgO Na2O K2O P2O5 SO3 LOI

21.59 3.78 3.65 0.12 0.15 4.29 0.04 6.41 0.99 0.35 0.95 0.07 2.38 <0.05

58.50 8.32 5.05 2.95 0.73 25.31 0.05 2.13 1.78 0.86 2.02 0.05 0.25 0.88

50.40 7.79 5.81 1.78 0.71 23.07 0.06 6.69 2.09 1.08 3.13 0.23 2.38 2.37

89.50 2.88 2.31 0.51 <0.05 1.13 0.02 0.98 1.55 0.42 0.60 0.12 0.40 2.40

Table 2 Composition of the studied samples of Portland cement with mineral additives. Sample

Composition

Sample

Composition

PC PCFA1 PCFA2

PC 100% 90% PC, 10% FA1 90% PC, 10% FA2

PCSF PCPz1 PCPz2

90% PC, 10% SF 90% PC, 10% Pz1 90% PC, 10% Pz2

2.3. Methods of investigation The pozzolanic activity of the mineral additions was determined according to Bulgarian Standard (BDS 16720-87) by measuring the quantity of CaO, which is bound by the addition from a solution of calcium hydroxide after a contact lasting 30 days. In 20% solution of calcium hydroxide with concentration of 1.15 mg/cm3 there is added 0.5 g mineral additive. The content of CaO in the filtrate, obtained after filtering of the solution was determined by titrating of complexone III in the presence of indicator Fluorexon (mixture of Thymolphtalein and potassium chloride). The quantity of the reacted CaO with the addition is calculated by estimation of the difference between the initial concentration of the lime solution and that, obtained after titration of the sample (expressed in mg of CaO for 1 g of addition). The complex thermal analysis after 48 days and after 4 years of hardening was performed with apparatus type 3427 MOM (Hungary), heating from 20 to 1000 °C with a speed of 10 °C/min in air and using Al2O3 as inert substance. The mass of the samples is 800 mg. At an age of 600 days the thermal analysis was performed with apparatus Stanton Rederoff (UK) on samples with mass of 16 mg preserving the above mentioned conditions. The IR spectra of the initial materials and the hardened cement pastes were recorded on a Bruker Tensor 27 FTIR Spectrometer in a regime of absorption using tablets of 300 mg KBr and 3 mg of the studied substance. Pure KBr served as a reference spectrum.

3. Experimental results and discussion 3.1. Determination of the pozzolanic activity of the mineral additions Up to the 15th day highest pozzolanic activity among the used additives was manifested by SF and later its activity increased slower. After 25 days the activity is 328 mg CaO/g and is lower than the activity of FA from TEPS ‘‘Bobov Dol’’ (FA2) with a value of 335.6 mg CaO/g (Fig. 1). Lowest pozzolanic activity was showed by FA from TEPS ‘‘Pernik’’ (FA1) and highest pozzolanic activity after the 20th day was registered for the mineral additives Pozzolit 1(Pz1) and Pozzolit 2 (Pz2). The addition of SF to FA from TEPS ‘‘Pernik’’ increases more than 2.3 times its pozzolanic activity up to the 30th day and the addition of SF to the more active FA2 increased its activity with about 25% for the same age. 3.2. Complex thermal analysis The thermal analysis showed the presence of three groups of endothermal effects connected with disintegration of the hydrate products and evolution of the chemically bound water

Fig. 1. Pozzolanic activity of the mineral additives.

[28,42–45]: ettringite, Ca-hydrosilicates and Ca-hydroaluminates with temperature of dehydration about 450–480 °C, denoted later as products with low temperature of dehydration (PI); portlandite (Pt) with temperature of the endothermic peak in the range 500– 525 °C and hydrate products (plus calcite) with temperature of dehydration higher than 530 °C denoted as products with high temperature of dehydration (PII). The positions of the peaks reflecting the endothermic effects are changing in respect to time of hydration and composition of the samples. The compounds in the hardened cement paste, which contain CO2, decompose when heated in the temperature range 530– 950 °C displaying three characteristic endothermic peaks at temperatures, which depend on the age of the cement paste and its composition and which correspond to more or less temperature stable Ca carbonates [45,46]. Table 3 contains the total quantity of the hydration product and of portlandite for ages of hardening of 48 and 600 days and 4 years. When calculating the quantity of the hydrate products in mg per gram cement in each studied sample then the total quantity of the hydration products in the samples with mineral additives for the three ages is higher from the respective one in the pure paste and the portlandite quantity in the pure sample is highest (Table 3), which confirms the pozzolanic properties of the used fly ashes and silica fume. Figs. 2–5 display the curves of changes with time of the total quantity of the hydrate products of all the three main groups of separate compositions, calculated in mg per gram solid bounding substance.

Table 3 Total quantity of the hydrate products (HP) and quantity of portlandite (Pt) in mg per 1 g binder. Composition

PC PCSF PCFA1 PCFA2 PCPz1 PCPz2

48 days

600 days

4 years

HP

Pt

HP

Pt

HP

Pt

447.4 438.3 401 424 410.7 426.6

75.2 49.6 54.2 52.8 52.6 51.2

471.4 487.8 511.5 520 502.5 508.5

77.8 33.8 30.2 32.5 30.2 43.2

495 504.7 510.6 515.8 493.4 501.3

85 26.1 27.5 27 26.9 30.1

V. Lilkov et al. / Construction and Building Materials 60 (2014) 48–56

Fig. 2. Change of the total quantity of the hydrate products (HP) up to the 4th year of hydration (mg/g binding substance).

Fig. 3. Changes in the quantity of portlandite (Pt) up to the 4th year of hydration (mg/g binding substance).

48th day: Highest quantity of hydrate products was registered in the compositions without any addition (PC) and with silica fume (PCSF), while lowest one was exhibited by the compositions with fly ash from TEPS ‘‘Pernik’’ (PCFA1 b PCPz1) (Fig. 2). The quantity of portlandite in the pure cement paste (PC) is 16.8% from the total quantity of hydrate products – 75.2 mg/g solid substance and in the compositions with mineral additives it is in the range 11.3– 13.5% (49.6–54.2 mg/g solid substance) (Fig. 3). The quantity of hydrate products with low temperature of dehydration (PI) is highest in the pure cement paste and lowest in the sample with addition of fly ash FA1 (Fig. 4). The quantity of the products with high temperature of dehydration (PII) is highest in the samples with fly ash from TEPS ‘‘Bobov Dol’’ (PCFA2 b PCPz2) and lowest in the plain cement (Fig. 5). The quantitative ratio between the hydrate products from groups PI and PII for the pure cement paste is 5.9 and for all the rest compositions with mineral additives it is in the range 2.7–3.3. Figs. 6 and 7 present, respectively, the DTA and DTG curves obtained for the studied samples. In the cases of compositions with mineral additives the peaks and the ranges of the thermal effects

51

Fig. 4. Changes in the quantity of hydrate products with low temperature of dehydration (HPI) up to the 4th year of hydration (mg/g binding substance).

Fig. 5. Changes in the quantity of hydrate products with high temperature of dehydration (HPII) up to the 4th year of hydration (mg/g binding substance).

are shifted towards higher temperatures compared with those of the plain cement paste. This is markedly expressed for the compositions with addition of fly ash from TEPS ‘‘Bobov Dol’’. From the presented results follows the conclusion that up to the age of 48 days silica fume and fly ashes actively participate in the process of hydration reacting with portlandite, which was formed during the cement hydration. The fly ash from TEPS ‘‘Pernik’’ is with lower activity than the one from TEPS ‘‘Bobov Dol’’ and highest initial activity is demonstrated by silica fume. During the hydration of the pure cement paste, portlandite and hydration products with temperature of dehydration lower than 435 °C are predominantly formed. In the compositions with mineral additions as a result of pozzolanic reactions there are formed CA-hydrosilicates and Ca-carbonates, which thermally decompose at temperatures above 620–640 °C. A conclusion can be made that the carbonation and polymerization of the Ca-hydrosilicates is accelerated in the cement pastes with pozzolanic mineral additives. 600 Days: The total quantity of hydrate products for all compositions is higher after the 48th day (Fig. 2) with highest increment in the cement pastes with fly ash from Thermal plant ‘‘Pernik’’ and

52

V. Lilkov et al. / Construction and Building Materials 60 (2014) 48–56

Fig. 7. DTG-curves – 48 days of hydration.

Fig. 6. DTA-curves – 48 days of hydration.

lowest in the pure cement paste. In comparison with the 48th day the relative content of portlandite compared to the total quantity of the hydrate products in the plain cement paste is almost the same (16.5% or 77.8 mg/g cement) and for the compositions with mineral additives it had become lower of about 1.2–1.8 times and is in the ranges 30.2–43.2 mg/g solid substance (Fig. 3). In all compositions there was registered lowering of the quantity of the products, which thermally decompose at about 450–460 °C. The relative portion of the hydration products from group PII (Fig. 5) was drastically increased compared to those from group PI (Fig. 4) and this ratio is highest in the compositions with Pozzolit and addition of fly ash from TEPS ‘‘Bobov Dol’’, while in the cement paste without addition it is lowest. From the DTG curves (Fig. 8) it is seen that there are present new thermal effects in the temperature range 530–700 °C – more clearly expressed in the compositions with mineral additives. It is obvious from the presented results that between the 48th and 600th days there were simultaneously realized several processes in the cement pastes: continuing hydration of the clinker minerals of cement; pozzolanic reactions of the fly ashes and silica fume with portlandite from the hydration of cement; and mainly carbonation of portlandite and the Ca hydrosilicates. The carbonation of C–S–H leads to formation of calcium carbonates with a low degree of crystallization – vaterite and aragonite, which

Fig. 8. DTG-curve in the interval 530–1000 °C – 600 days of hydration.

dissociate in the temperature interval 530–700 °C [47,48]. The stable calcium carbonate (calcite), which is obtained from the direct carbonation of portlandite dissociates between 750 and 800 °C.

V. Lilkov et al. / Construction and Building Materials 60 (2014) 48–56

4 years: At an age of 4 years the total quantity of hydration products in the samples with addition of fly ashes and Pozzolit (evaluated according to the total quantity of released water at heating up to 1000 °C) is a little bit smaller than the one at the 600th day and in the plain cement paste and the cement paste with silica fume there was registered an increase, respectively with 5% and 3.5% (Fig. 2). The quantity of portlandite in the plain cement paste was increased and in all other compositions it was lowered (Fig. 3). In the compositions with mineral additions the relative portion and the total quantity of the hydrate products PII was lowered and the portion of the hydrate products PI was increased, while in the case of the plain cement paste the tendency is reversed (Figs. 4 and 5). Figs. 9 and 10 present, respectively, the DTA and DTG curves of the studied samples after 4 years of hydration. It is obvious from

53

the results that between the 600th day and the 4th year the process of hydration of the clinker minerals of cement as well as the pozzolanic reaction between the mineral additives and portlandite continues with low intensity. In this time period the dominating processes in the hardened cement paste are re-structuring of the different groups of hydration products, carbonation, and polymerization of the hydrosilicates connected with release of water. Thus the integral quantity of the chemically bound water in the cement stone cannot be used as a criterion to judge the intensity of the processes of hydration and structural reorganization in the cement pastes, especially for these with mineral additives. The increase of the quantity of the hydrate products in group PI is connected with polymerization of the hydrosilicates and reactions of fly ashes with calcium hydroxide resulting in formation of calcium alumohydrosilicates. The weight losses during heating of the samples in the temperature range from 340–350 °C to 450–480 °C at ages of 40 and 600 days and 4 years are graphically expressed at Fig. 11. Most probably they are due to the presence of calcium hydroaluminates and calcium alumohydrosilicates with complex composition. They are obtained basically during hydration of the aluminate phases of cement and also probably during carbonation of ettringite and reaction of the fly ashes with the calcium hydroxide. Their quantity in the compositions with mineral additives is significantly higher than that in the pure cement paste in the cases of all three ages of hardening. In the compositions PCFA1 and PCPz1, in which FA1 with lower pozzolanic activity was added the quantity of these products does not change after the 48th day, while in the compositions with silica fume and fly ash from TEPS ‘‘Bobov Dol’’ the respective quantity increases up to the 600th day. Up to the 4th year of hydration their quantity does not change or alters slightly, which means that this group of hydrate products is stable in respect to restructuring of the newly formed hydrates in the cement stone. 3.3. IR spectrometry Figs. 12 and 13 show, respectively, the IR spectra of the initial materials and the cement pastes after hardening for 2 years. The

Fig. 9. DTA-curves – 4 years of hydration.

Fig. 10. DTG-curves – 4 years of hydration.

54

V. Lilkov et al. / Construction and Building Materials 60 (2014) 48–56

Fig. 11. Weight-loss during heating in the interval from 340–350 °C to 450–480 °C at ages of 48 and 600 days and 4 years.

Fig. 13. IR spectra of cement pastes after 2 years of hardening. Fig. 12. IR spectra of plain cement and the mineral additives.

spectrum of plain cement reveals absorption band at 3640– 3645 cm1, related to OH stretching vibrations of the portlandite molecules while the cements with mineral additives do not display

this band due to the lower portlandite content in these samples. The band at around 3430 cm1 characteristic for calcium hydrosilicates in the hardened cement pastes is difficult to be analyzed

V. Lilkov et al. / Construction and Building Materials 60 (2014) 48–56

because the initial material display such well expressed band. The band at 1120 cm1 characterizing ettringite (asymmetric stretching vibrations of SO2 4 groups) is weakly expressed only in the IR spectrum of the plain cement paste. The main differences in the spectra of plain cement paste and cement pastes with mineral additions are concerned with the presence of calcium carbonate phases. Principle bands characterizing calcite are at 875 cm–1 and 1415–1418 cm–1 (stretching vibrations of CO2 3 anions) and in the spectra of cement with mineral additives there are additional bands at 713 cm1 and 745 cm1, characteristic for vaterite and aragonite. The spectral band of vaterite and aragonite at around 856 cm–1 is manifested in the spectra of cement pastes with addition of silica fume (samples PCSF, PCPz1, and PCPz2). Vaterite is with lowest thermo-dynamical stability among the three phases of calcium carbonate it displays its bands in the IR spectra being a result of carbonation of the Ca-silicates during contact with air as well as in the case of carbonation of imperfectly crystalline portlandite.

4. Conclusions The fly ashes from TEPS ‘‘Bobov Dol’’ and TEPS ‘‘Pernik’’, the silica fume from the metallurgical plant ‘‘Kremikovtsi’’, and their mixtures in the form of mineral additive Pozzolit are suitable pozzolanic additions to cement. They participate actively in the processes of formation of the structure of the hardened cement paste. The pozzolanic activity of the fly ash from TEPS ‘‘Bobov Dol’’ during the first several months is comparable with that of the silica fume, while the fly ash from TEPS ‘‘Pernik’’ is with lower pozzolanic activity. The silica fume, the fly ashes and the mineral addition ‘‘Pozzolit’’ participate in the process of hydration of cement and stimulate the formation of hydrate products with temperature of dehydration above 600 °C, more clearly expressed after the 48th day of hydration. After 600 days of hardening when the activity of the additions and cement itself lower there is observed polymerization of the calcium hydrosilicates, accompanied by increase of the portion of products with temperature of dehydration lower than 340–350 °C. The quantity of hydration products per 1 gram of cement in the samples with mineral additions after 4 years of hydration is higher than the respective one in plain cement. In the case of water to binder ratio 0.5 in the samples and quantities of mineral additives of 10% the portlandite in the cement stone is not consumed fully even after 4 years of hydration. The calcium hydrosilicates and calcium hydroaluminates, which decompose in the temperature interval from 340–350 °C to 450–480 °C are formed in the cement pastes up to the 48th day of hardening and up to the 600th day their quantity increases only in the cases of compositions with silica fume and fly ash from TEPS ‘‘Bobov Dol’’. They are stable up to the 4th year and their quantity increase steadily. The carbonation of portlandite and the carbonation and polymerization of the hydrosilicate gel in the hardening cement pastes is accelerated when using active mineral additives, in which additional quantities of vaterite and aragonite is formed.

References [1] Persson B. Hydration and strength of high performance concrete. Adv Cem Mater 1996;3(3–4):107–23. [2] Persson B. Seven-year study on the effect of silica fume in concrete. Adv Cem Mater 1998;7(3–4):139–55. [3] Almusallam Abdullah A, Beshr Hamoud, Maslehuddin Mohammed, Al-Amoudi Omar SB. Effect of silica fume on the mechanical properties of low quality coarse aggregate concrete. Cem Concr Compd 2004;26(7):891–900.

55

[4] Song Ha-Won, Packa Seung-Woo, Nam Sang-Hyeok, Janga Jong-Chul, Saraswathya Velu. Estimation of the permeability of silica fume cement concrete. Constr Build Mater 2010;24(3):315–21. [5] Song Ha-Won, Jang Jong-Chul, Saraswathy Velu, Byun Keun-Joo. An estimation of the diffusivity of silica fume concrete. Build Environ 2007;42(3):1358–67. [6] Mahmoud N, Salehi AM. Assessing the effectiveness of pozzolans in massive high-strength concrete. Constr Build Mater 2010;24(11):2108–16. [7] Berry EE, Malhotra VM. Fly ash for use in concrete – a critical review. ACI Mater J 1980;77(8):59–73. [8] Bilodeau A, Sivasundaram V, Painter KE, Malhotra VM. Durability of concrete incorporating high volumes of fly ash from sources in USA. ACI Mater J 1994;91(1):3–12. [9] Wang Qiang, Feng Jingjing, Yan Peiyu. The microstructure of 4-year old hardened cement-fly ash paste. Constr Build Mater 2012;29:114–9. [10] Joshi RC, Lohtia RP. Fly ash in concrete: production, properties and uses in Netherlands. Gordon and Breach Science Publishers; 1999. [11] Han SH, Kim JK, Park YD. Prediction of compressive strength of fly ash concrete. Cem Concr Res 2003;33(7):965–71. [12] Stoitchkov V, Abadjiev P, Lilkov V, Vasileva V. Effect of the ‘‘Pozzolit’’ active mineral admixture on the properties of cement mortars and concretes Part 1: Physical and mechanical properties. Cem Concr Res 1996;26(7):1066–72. [13] Lilkov V, Stoitchkov V. Efect of the ‘‘Pozzolit’’ active mineral admixture on the properties of cement mortars and concretes Part 2: Pozzolanic activity. Cem Concr Res 1996;26(7):1073–81. [14] Thomas MDA, Shehata MH, Shashiprakash SG, Hopkins DS, Cail K. Use of ternary cementitious systems containing silica fume and fly ash in concrete. Cem Concr Res 1999;29(8):1207–14. [15] Poon Chi Sun, Lam Lik, Wong Yuk Lung. Effects of fly ash and silica fume on interfacial porosity of concrete. J Mater Civ Eng 1999;11(3):197–205. [16] 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. [17] Barbhuiya SA, Gbagbo JK, Russell MI, Basheer PAM. Properties of fly ash concrete modified with hydrated lime and silica fume. Constr Build Mater 2009;23:3233–9. [18] Jaturapitakkul Chai, Kiattikomol Kraiwood, Sata Vanchai, Leekeeratikul Theerarach. Use of ground coarse fly ash as a replacement of condensed silica fume in producing high-strength concrete. Cem Concr Res 2004;34(4):549–55. [19] Kocak Y. A study on the effect of fly ash and silica fume substituted cement paste and mortars. Sci Res Essays 2010;5(9):990–8. [20] Thomas MDA, Shehata MH, Shashiprakash SG, Hopkins DS, Cail K. Use of ternary cementitious systems containing silica fume and fly ash in concrete. Cem Concr Res 1999;29(8):1207–14. [21] Taylor HFW. Cement chemistry. 2nd ed. London: Thomas Telford; 1997. [22] Livingston RB, Bumrongjaroen W. Optimization of silica fume, fly ash and cement mixes for high performance concrete. In: 2005 World of Coal Ash (WOCA), April 11–15, 2005, Lexington, Kentucky, USA; 2005. [23] Zhao Qi Wu. Young JF. The hydration of tricalcium silicate in the presence of colloidal silica. J Mater Sci 1984;19:3477–86. [24] Ramachandran VS. Hydration reactions in cement containing condensed silica fume. In: Int Workshop on Condensed Silica Fume in Concrete, Montreal, May 1987, IRC Paper N 149. [25] Groves GW, Rodger SA. The Hydration of C3S and Ordinary Portland Cement with Relatively Large Additions of Microsilica. Adv Cem Res 1989;2:135–40. [26] Justnes H, Meland I, Bjoergum JO, Krane TA. 29Si MAS NMR study of the pozzolanic activity of condensed silica fume and the hydration of di and tricalcium silicates. Adv Cem Res 1990;3:111–6. [27] Mitchell DG, Hinczak I, Day RA. Interaction of silica fume with calcium hydroxide solutions and hydrated cement pastes. Cem Concr Res 1998;28(11):1571–84. [28] Xu A, Sarkar SL, Nilsson LO. Effect of fly ash on the microstructure of cement mortar. Mater Struct 1993;26(7):414–24. [29] Francis Young J. Investigations of calcium silicate hydrate structure using silicon-29 nuclear magnetic resonance spectroscopy. J Am Ceram Soc 1988;71(3). C-118–C-120. [30] Zhang Min-Hong, Gjørv OddE. Effect of silica fume on cement hydration in low porosity cement pastes. Cem Concr Res 1991;21(5):800–8. [31] Lu Ping, Sun Guokuang, Youn J Francis. Phase composition of hydrated DSP cement pastes. J Am Ceram Soc 1993;76(4):1003–7. [32] Georgescu M, Badanoiu A. Hydration process in 3CaOSiO2-silica fume mixtures. Cem Concr Compos 1997;19(4):295–300. [33] Wang Xiao-Yong, Lee Han-Seung. A model predicting carbonation depth of concrete containing silica fume. Mater Struct 2009;42(6):691–704. [34] Ukrainczyk N, Ukrainczyk M, Sipusic J, Matusinovic T. XRD and DTA investigation of hardened cement paste degradation. In: Proc of the 11th Conference on Materials, Processes, Friction and Wear, MATRIB’06, Vela Luka, 22–24.06.2006, Hrvatska; 2006. p. 243–9. [35] Chi Jack M, Huang Ran, Yang CC. Effects of carbonation on mechanical properties and durability of concrete using accelerated testing method. J Mar Sci Technol 2002;10(1):14–20. [36] Malviya R, Chaudhary R. Factors affecting hazardous waste solidi-fication/ stabilization: a review. J Hazard Mater 2006;137:267–76. [37] Groves GW, Rodway DK, Richardson IG. The carbonation of hardened cement pastes. Adv Cem Res 1990;3(11):117–25.

56

V. Lilkov et al. / Construction and Building Materials 60 (2014) 48–56

[38] Groves GW, Brough AR, Richardson IG, Dobson CM. Progressive changes in the structure of hardened C3S cement pastes due to carbonation. J Am Ceram Soc 1991;74(11):2891–6. [39] Dunster AM. An investigation of the carbonation of cement paste using Trimethylsilylation. Adv Cem Res 1989;2(7):99–106. [40] Swenson EG, Sereda PJ. Mechanism of the carbonation shrinkage of lime and hydrated cement. J Appl Chem 1968;18:111–7. [41] Chen Jeffrey J, Thomas JJ, Jennings HM. Decalcification shrinkage of cement paste. Cem Concr Res 2006;36:801–9. [42] Chaipanich A, Nochaiya T. Thermal analysis and microstructure of Portland cement-fly ash-silica fume pastes. J Therm Anal Calorim 2010;99(2):487–93. [43] El-Shimy E, Abo-El-Enein SA, El-Didamony H, Osman TA. Physico-chemical and thermal characteristics of lime-silica fume pastes. J Therm Anal Calorim 2000;60(2):549–56.

[44] Alarcon-Ruiz L, Platret G, Massieu E, Ehrlacher A. The use of thermal analysis in assessing the effect of temperature on a cement paste. Cem Concr Res 2005;35(3):609–13. [45] Villain G, Thiery M, Platret G. Measurement methods of carbonation profiles in concrete: thermogravimetry, chemical analysis and gammadensimetry. Cem Concr Res 2007;37:1182–92. [46] Chi Jack M, Huang Ran, Yang CC. Effects of carbonation on mechanical properties and durability of concrete using accelerated testing method. J Mar Sci Technol 2002;10(1):14–20. [47] Sauman Z. Carbonization of porous concrete and its main binding components. Cem Concr Res 1971;1(6):645–62. [48] Grandet J. Contribution à l’étude de la prise et de la carbonatation des mortiers au contact des matériaux poreux’’, PhD thesis (in French), Université Paul Sabatier, Toulouse; 1975. p. 287.