- Email: [email protected]
Assessing the suitability of three Australian fly ashes as an aluminosilicate source for geopolymers in high temperature applications
Materials Science and Engineering A 528 (2011) 3390–3397
Contents lists available at ScienceDirect
Materials Science and Engineering A journal homepage: www.elsevier.com/locate/msea
Assessing the suitability of three Australian fly ashes as an aluminosilicate source for geopolymers in high temperature applications夽 William D.A. Rickard ∗ , Ross Williams, Jadambaa Temuujin, Arie van Riessen Centre for Materials Research, Curtin University, PO Box U 1987, Perth, WA 6845, Australia
a r t i c l e
i n f o
Article history: Received 27 October 2010 Received in revised form 4 January 2011 Accepted 5 January 2011 Available online 19 January 2011 Keywords: Geopolymer Fly ash Thermal properties
a b s t r a c t Fly ash characteristics cannot be assumed to be constant between power stations as they are highly dependent on the coal source and burning conditions. It is critical to understand the characteristics of fly ash in order to produce geopolymers suitable for high temperature applications. We report on the characterisation of fly ash from three Australian power stations in terms of elemental composition, phase composition, particle size, density and morphology. Geopolymers were synthesised from each of the fly ashes using sodium silicate and sodium aluminate solutions to achieve a range of Si:Al compositional ratios. Mechanical properties of geopolymer binders are presented and the effect of the source fly ash characteristics on the hardened product is discussed, as well as implications for high temperature applications. It was found that the twenty eight day strength of geopolymers is largely dependent on the sub 20 m size fraction of the fly ash. Strength loss after high temperature exposure was found to be dependent on the concentration of iron in the fly ash precursor and the Si:Al ratio of the geopolymer mixture. Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.
1. Introduction Geopolymers are a synthetic inorganic polymer produced by the alkali activation of aluminosilicates [1]. Geopolymers have gained a lot of interest in recent years amongst scientists and engineers because of their potential as an environmentally friendly replacement for ordinary Portland cement (OPC). Geopolymers have also
been considered for use in high temperature applications, such as fire proof coatings, making use of their intrinsic thermal stability [2]. The geopolymerisation reaction that creates a polymeric backbone of aluminium and silicon atoms is described in Eq. (1) [3]. Simplistically, geopolymerisation involves the dissolution of aluminosilicates from the source material followed by short ranged polymerisation to form a Si–Al based geopolymer backbone.
→
Si-Al source + Silicate + Water + Alkaline liquid
Geopolymer precursor
n(Si2O5,Al2O2) + 2nSiO2 + 4nH2O + NaOH (or KOH) → (Na+,K+) + n(OH)3-Si-O-Al --O-Si-(OH) 3 ⎜
(OH)2 Geopolymer precursor +
Alkaline ions
→
Geopolymer backbone ⎜
n(OH)3-Si-O-Al-O-Si-(OH) 3 + NaOH (or KOH)
⎜
(OH)2
夽 This project is carried out under the auspice and with the financial support of the Centre for Sustainable Resource Processing (CSRP), which is established and supported under the Australian government’s Cooperative Research Centres Program. ∗ Corresponding author. Tel.: +61 8 9266 4219; fax: +61 8 9266 2377. E-mail address: [email protected] (W.D.A. Rickard).
⎜
⎜
(Na+,K+) - (-Si-O-Al --O-Si-O-) + 4nH2O ⎜
⎜
⎜
O
O
O
(1)
The degree of dissolution of aluminosilicates in high pH alkaline solutions is largely dependent on the particle size, morphology and composition of the source material [4–8], in particular the amorphous aluminosilicates. Previous research has shown that geopolymers synthesised from aluminosilicate sources with suitable overall chemical composition but lacking an appropriate reactive component will result in incomplete dissolution and con-
0921-5093/$ – see front matter. Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2011.01.005
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
sequently be of lower strength [7,8]. The potential impact of not properly characterising the source material is more significant when using fly ash as an aluminosilicate source for synthesising geopolymers as it is known to have highly variable physical and chemical properties [9,10]. Geopolymers can be synthesised from many materials with a high concentration of aluminosilicates such as metakaolin or fly ash. This paper concentrates on the use of fly ash as a source material. Fly ash is produced from the coalescence of non-combustible material in the flue gases of coal fired power stations. The use of fly ash in the synthesis of geopolymers is not to be mistaken with the use of fly ash in ordinary Portland cement (OPC). In OPC, fly ash is used to improve specific characteristics, such as workability and reduction of temperature during curing, but does not add significantly to the strength of the binder [11]. In geopolymers, fly ash is the source of the aluminosilicates for the binder and is thus the critical component for strength development. Fly ash is a powdery material made up of small glass spheres, consisting primarily of silicon, aluminium, iron, and calcium oxides [12]. The typical particle size range is between 1 and 150 m [13]. Crystalline phases commonly identified in fly ashes are quartz, mullite and various iron rich phases such as hematite [14]. Class F fly ash, defined as having a (SiO2 + Al2 O3 + Fe2 O3 ) ≥ 70 wt%, is readily available in Australia and is produced from burning anthracite or bituminous coal [9]. Fernandez-Jimenez et al. [6] have proposed a model for the dissolution of fly ash in alkaline environments. This model is only loosely used as a guide in this study as it does not address the role of crystalline phases present in the fly ash. Bakharev [7] observed that fly ashes with a low amorphous content and a relatively large average particle size produced low strength geopolymers. It can be gathered from the literature that geopolymer source material must be carefully selected and accurately characterised in order to optimise its effectiveness in producing usable geopolymers. The desired characteristics of source material for geopolymers will depend on the final application. Thus geopolymers designed for use in high temperature applications will require different source material to geopolymers designed for use in structural applications. For example, previous research has shown that high levels of iron in fly ash have a negative effect on the high temperature performance of geopolymers [5,15]. The effect of other fly ash characteristics on high temperature geopolymers such as the propensity to crystallise and the effect on melting temperature have not been published in the literature. The objective of this paper is to convey an understanding of the important characteristics of fly ash and how these characteristics affect the formation of geopolymer cements. The measured data on the three Australian fly ashes used in this study will be useful for researchers and engineers looking to synthesise geopolymers from these sources in the future.
2. Experimental 2.1. Materials Fly ash was sourced from Collie power station in Western Australia, Eraring power station in New South Wales and Tarong power station in Queensland. The fly ash was provided by Fly Ash Australia (FAA). Sodium silicate solutions were prepared by dissolving sodium hydroxide pellets from Univar Pty Ltd. and fumed silica from Cabosil in deionised water. The solutions were allowed to dissolve for 24 h at 70 ◦ C prior to use. Sodium aluminate solutions were supplied by Coogee Chemicals who specified that the solution contained 19 wt% Al2 O3 , 25.5 wt% NaOH and 55.5 wt% H2 O.
3391
Table 1 Elemental ratios of the geopolymers prepared in this study. Fly ash
Si:Al
Na:Al
H:Si
Collie Collie Collie Eraring Eraring Eraring Tarong Tarong Tarong
2.0 2.5 3.0 2.0 2.5 3.0 2.0 2.5 3.0
1.25 1.25 1.25 1.25 1.25 1.25 1.32 1.25 1.25
5.0 4.5 6.0 5.0 4.0 4.5 6.4 5.5 5.4
2.2. Geopolymer synthesis Samples were prepared with a range of compositional ratios as shown in Table 1. Only the amorphous component of the aluminosilicates in the fly ashes was used in the calculation as the crystalline material was considered inert during the geopolymerisation reaction. Geopolymers were synthesised by mixing the fly ash and the alkaline solution in a planetary centrifugal mixer (ARE 250, Thinky, Japan) for 5 min at 1300 RPM. A further 30 s at 2100 RPM was used for defoaming. The slurry was then cast into various sized moulds for testing. The moulds were sealed and samples cured for 24 h at 70 ◦ C. Samples were stored at ambient temperature prior to testing. 2.3. XRF X-ray fluorescence (XRF) was outsourced to a commercial laboratory (Ultra Trace Pty Ltd.). The samples were fused in a glass disc prior to analysis. Loss on ignition (LOI) of fly ash was conducted up to 1000 ◦ C. 2.4. XRD analysis X-ray diffraction (XRD) patterns were collected on a Bruker D8 Advance diffractometer operated at 40 kV and 40 mA using Cu-K␣ radiation. The data was collected using a nominal 2 step size of 0.01◦ , count time of 0.5 s and a 2 range of 10–120◦ . Crystalline phases were identified using EVA version 11. Rietveld modelling of the data was performed using Bruker AXS TOPAS version 4.1. A fluorite (CaF2 ) internal standard (10 wt%) was used to facilitate quantitative analysis of the amorphous phase. The fundamental parameters approach was used to model the instrumental peak shape function, allowing the crystallite size to be determined [16]. 2.5. SEM Scanning electron microscopy (SEM) was conducted on an Evo 40XVP (Zeiss, Germany) and a Neon 40EsB (Zeiss, Germany) using secondary electrons (SE) as well as backscattered electrons (BSE). The accelerating voltage used was either 5 kV for SE imaging or 20 kV for BSE imaging. Flat samples were polished to 1 m using diamond paste. Samples were coated with carbon prior to imaging in the SEM. Elemental analysis was performed using an Oxford Instruments energy dispersive X-ray spectrometer (EDS). Analysis of X-ray spectra was performed using Inca-Analyser software (Oxford Instruments, England). 2.6. Particle sizing Particle size analysis was performed with a Malvern laser diffraction system at a commercial laboratory. The dispersing solution was deionised water with a sodium hexametaphosphate dispersing agent. The solution and approximately 1 wt% of fly ash
3392
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
was sonicated for 10 min to break up aggregates prior to size analysis.
Table 2 Oxide composition of each fly ash as determined by XRF. Uncertainties in brackets.
2.7. Density Density was calculated by measuring the displacement of fly ash in a fixed volume containing a fluid. A known mass of fly ash was placed in a 100 ml specific gravity (SG) bottle. The rest of the volume was filled with ethanol and the mixture was sonicated to remove any entrained air. The bottle was then weighed and the average particle density of the fly ash was calculated using the following formula: Flyash =
mflyash 100 − Vethanol
where Vethanol =
(mtotal − mflyash − mbottle ) ethanol
Oxide
Collie (wt%)
Eraring (wt%)
Tarong (wt%)
SiO2 Al2 O3 Fe2 O3 CaO K2 O TiO2 MgO Na2 O P2 O5 SrO BaO Other
51.38(8) 26.90(10) 13.20(2) 1.74(5) 0.90(4) 1.47(1) 1.41(3) 0.41(5) 1.09(2) 0.23(1) 0.38(1) 1.15(6)
65.47(8) 23.00(10) 4.03(2) 1.59(5) 1.68(4) 0.84(1) 0.51(3) 0.56(5) 0.27(2) 0.05(1) 0.06(1) 1.67(6)
73.68(8) 22.40(10) 0.64(2) 0.08(5) 0.53(4) 1.28(1) 0.17(3) 0.09(5) 0.08(2) 0.01(1) 0.03(1) 0.96(6)
LOI (1000 ◦ C) Sum of aluminosilicates Sum of alkali SiO2 /Al2 O3 Si/Al (molar)
0.44 78.3(13) 1.31(6) 1.91(1) 1.62(1)
1.37 88.47(13) 2.24(6) 2.85(1) 2.42(1)
0.79 96.08(13) 0.62(6) 3.29(2) 2.79(1)
All reported results are the average of 5 separate measurements. 2.8. Mechanical testing Cylinders of 25 mm diameter and 50 mm height were prepared for mechanical testing. Geopolymer samples were tested 28 days after synthesis. Compressive strength testing was conducted on a Lloyds universal tester EZ50. A load rate of 0.25 MPa/s was used to closely comply with ASTM C39 (a test method for concrete specimens) [17]. The stated strength values are the average of 4 repeat tests. 3. Results and discussion 3.1. Fly ash composition XRF was performed to assess the chemical composition of each fly ash (Table 2). As expected each fly ash was composed of aluminosilicates, iron oxides and a number of minor oxides. Each fly ash had a similar amount of Al2 O3 , though the amount of SiO2 varied from 51 wt% to 74 wt%. The molar silicon to aluminium ratio, varied from 1.6 (Collie) to 2.8 (Tarong). A significant difference between the fly ashes is the amount of iron oxide, which varied from 0.6 wt% (Tarong) to 13.2 wt% (Collie). Previous research has shown that volume changes caused by oxidation of the iron detrimentally affect the performance of fly ash geopolymers at elevated temperatures [15]. Iron rich fly ash particles have also been observed to inhibit the dissolution of aluminosilicates during geopolymerisation [5]. The sum of the alkalis initially present in the fly ashes is listed in Table 1. Alkalis are known to be excellent fluxing agents [18] and as such their added presence, as there will be alkali in the geopolymer, may reduce the melting point of the geopolymer (unless a more thermally stable crystalline phase is formed). The XRF data provides a preliminary indication as to the suitability of each of the fly ashes for production of fire resistant geopolymers although it is limited by the fact that there is no
information about the reactivity of the aluminosilicate component. Research by Chen-Tan et al. [5] showed that only the amorphous aluminosilicates in the fly ash are reactive in the geopolymerisation reaction that forms a geopolymer. Thus in order to properly quantify the reactive component of each of the fly ashes, the phase composition was determined. A fluorite internal standard was used to facilitate quantitative analysis of the amorphous phase. The phase composition of each of the fly ashes is presented in Table 3. It was observed that there were two populations of quartz of differing crystallite size. It is thought that the large crystallite size quartz (>100 nm) is primary quartz, i.e. it is present prior to the coal combustion process and exists as discreet particles amongst the fly ash. The second quartz population that was identified had a much smaller crystallite size (<100 nm) and slightly larger lattice parameters, and is thought to have formed during or after combustion, thus is secondary quartz, and is believed to be present within fly ash spheres. Previous research by Williams and van Riessen has also found a similar distribution of the quartz crystallite size in Collie fly ash [19]. The presence of quartz in a source material is undesirable for geopolymers designed for high temperature applications due to differential expansion upon heating. This can cause micro cracking which reduces the strength of the material. This problem is more significant where the particle size of the quartz is larger. Kong et al. [20] observed that the presence of quartz based aggregates significantly reduced the compressive strength of fly ash geopolymers upon heating to 800 ◦ C. Similar results have been observed with quartz based OPC mortars [21]. Differential thermal expansion between the primary quartz in the fly ash and geopolymer will reduce the strength of the composite at high temperature. It is the opinion of the authors that the secondary quartz detected in the studied fly ashes will not adversely affect the high temperature performance of the resulting geopolymer due to its small particle size and location within the fly ash spheres (see Section 3.2).
Table 3 Phase composition of Collie fly ash. Uncertainties in brackets. Phase
Formula
Amorphous content Mullite (ICSD 66452) Mullite (ICSD 66449) Quartz low (ICSD 83849) Quartz low Primary (ICSD 83849) Magnetite (ICSD 43001) Hematite (ICSD 88417)
Al4.56 Si1.44 O9.72 Al4.59 Si1.41 O9.7 SiO2 SiO2 Fe3 O4 Fe2 O3
Collie fly ash (wt%)
Eraring fly ash (wt%)
Tarong fly ash (wt%)
54.00 (45) 15.80 (18)
62.74 (31)
50.82 (28) 25.1 (11)
11.14 (18) 15.05 (21) 2.51 (83) 1.50 (64)
20.88 (14) 8.08 (16) 6.81 (14) 1.491 (52)
10.31 (14) 13.77 (13)
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
3393
Table 4 Amorphous composition of each of the fly ashes. Uncertainties in brackets. Oxide
Collie fly ash (wt%)
Eraring fly ash (wt%)
Tarong fly ash (wt%)
SiO2 Al2 O3 Fe2 O3 CaO K2 O TiO2 MgO Na2 O P2 O5 SrO BaO Other
20.90(65) 15.39(41) 9.11(32) 1.74(5) 0.90(4) 1.47(1) 1.41(3) 0.41(5) 1.09(2) 0.23(1) 0.38(1) 1.15(6)
45.03(65) 7.67(32) 2.49(16) 1.59(5) 1.68(4) 0.84(1) 0.51(3) 0.56(5) 0.27(2) 0.05(1) 0.06(1) 1.67(6)
42.79(59) 4.11(26) 0.64(2) 0.08(5) 0.53(4) 1.28(1) 0.17(3) 0.09(5) 0.08(2) 0.01(1) 0.03(1) 0.96(6)
Sum of amorphous aluminosilicates SiO2 /Al2 O3 Si/Al (molar)
36.29(77) 1.36(6) 1.15(4)
52.70(73) 5.87(26) 4.98(19)
46.90(64) 10.42(68) 8.84(49)
Quantitative XRD used to determine the amorphous content of each fly ash resulted in the following values; Collie 54.0 wt%, Eraring 62.7 wt% and Tarong 50.8 wt%. The elemental composition of the amorphous phase is calculated by subtracting the contribution of the crystalline phases from the XRF data. Table 4 details the amorphous content of each of the fly ashes. It should be noted that any crystalline phases in the fly ashes that are below the detection limit of the XRD are included in the amorphous content. The Si:Al ratio of the amorphous content was measured to be significantly different to the bulk (Table 2), especially for Eraring and Tarong fly ashes. Geopolymer made from these fly ashes should be synthesised using the Si:Al ratios in Table 4. The low concentration of reactive aluminium in Eraring and Tarong fly ashes requires the use of alkali–aluminate solutions rather than the commonly used alkali–silicate solution to achieve typical Si:Al ratios. The bulk of the iron in each of the fly ashes is amorphous; the highest proportion of crystalline to amorphous iron oxide was measured to be 0.38 (Collie fly ash). The low concentration of crystalline iron oxide may be in part due to a very small crystallite size and X-ray absorption by the covering material, as iron phases are commonly found within glass spheres [22]. The amorphous iron is known to order to phases such as hematite during thermal treatment with concomitant volume changes which have been observed to adversely affect the thermal performance of a fly ash based geopolymer [15]. 3.2. Fly ash morphology
the surface area can be calculated from the particle size data. Collie fly ash was calculated to have a specific surface area of 1.56 m2 /cm3 (1.56 × 106 m2 /m3 ). This value is much higher than Eraring and Tarong fly ashes, having specific surface areas of 0.92 m2 /cm3 and 0.99 m2 /cm3 , respectively. Based on specific surface area it would be expected that Collie fly ash would dissolve faster than Eraring and Tarong fly ashes once placed in an alkaline solution due to a smaller particle size. However, the phase and inter-particle location of the aluminosilicates will also affect the dissolution rate. Particle size is not a complete assessment of the reactivity of the amorphous material in a fly ash during geopolymerisation or its workability during casting. SEM analysis was done to assess fly ash morphology, looking specifically at the particle shape and the location of the aluminosilicate glass within the particles. If the glass is encapsulated within an un-reactive crystalline material then it would not be available for dissolution during geopolymerisation. Fly ash was mounted in resin and polished to 1 m surface finish in order to observe the interior structure of the particles. The cross section of the fly ash particles in all samples showed a typical spherical morphology with a large concentration of particles appearing to be glassy. A high degree of interparticle and intraparticle heterogeneity was also observed. This has been observed previously [10] and is due to local variations in temperature and composition during the coal combustion and fly ash capture process. Figs. 2, 4 and 7 give an indication of the typical particle morphology of Collie, Eraring and Tarong fly ashes, respectively. Tarong fly ash was observed to have more irregular shaped parti-
Fly ash morphology is known to affect bulk characteristics of the subsequent geopolymer [4]. Spherical morphology is beneficial to the synthesis of geopolymers as it allows for good workability at low liquids to solids mix ratios [23]. Low water content is often desirable in high temperature applications as it reduces dehydration shrinkage when heated. Particle size determines the surface area that is initially available for dissolution by the alkaline solution. It is also known that smaller fly ash particles (<20 m) are more likely to have a highly glassy composition (ideal for geopolymerisation), as small particles quench faster than large particles [10]. Fig. 1 compares the particle size distribution for each of the fly ashes in this study. Each fly ash was found to have a fineness better than the Australian standard for fly ash in cement (AS3582.1), where at least 75% of all particles must be smaller than 45 m. The percentage of particles smaller than 45 m was similar for each of the fly ashes. The major differences in particle size distribution occurred below 10 m. Collie fly ash had 43% of its particles smaller than 10 m whereas Eraring fly ash only had 27%. If the bulk of the fly ash particles are assumed to be spherical, an estimation of
Fig. 1. Particle size distribution for each fly ash. 45 m fineness: Collie (83%), Eraring (79%), Tarong (84%). 10 m fineness: Collie (43%), Eraring (27%), Tarong (36%).
3394
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
Fig. 2. SEM micrograph showing the particle distribution for Collie fly ash. The arrow indicates a porous, irregular shaped particle.
Fig. 3. SEM micrograph showing different particles observed in Collie fly ash. (a) Fly ash particle containing blocky crystallites likely to be quartz; (b) quartz particle; (c) porous fly ash particle; (d) fly ash particle containing iron. Operating conditions: Accelerating voltage = 5 kV and signal = secondary electron.
cles than the other fly ashes. This has the propensity to reduce the workability of geopolymer slurries made from this fly ash. Figs. 3, 5 and 6 show examples of the various types of particles observed in the fly ashes. Many of the crystalline phases, detected in the XRD analysis, were observed in the SEM images by searching for characteristic morphology and using EDS for elemental identification. Quartz was typically observed as separate
Fig. 4. SEM micrograph showing the particle distribution for Eraring fly ash.
Fig. 5. SEM micrograph of Eraring fly ash showing a fine mullite structure in a particle that appeared glassy at low magnification.
particles, whereas iron and mullite phases were only observed within glassy particles. There were, however, some blocky crystalline structures observed in some glassy particles (Fig. 3a) which were likely to be secondary quartz phases formed during the coal combustion process. This observation supports the finding of a secondary quartz population of smaller crystallite size detected by XRD analysis (Table 2). Iron structures were observed with a range of morphologies (Figs. 3d and 6). This is also in good agreement with the XRD results where a range of iron phases, including an amorphous phase, were identified. Fig. 5 shows an example of a particle that appeared glassy at low magnification, only to find the bulk of it was crystalline at higher magnification. The needle shaped crystals are approximately 1 m long and 100–200 nm wide, characteristic of mullite. The authors of this study were also aware that inter-particle porosity affects the density of the fly ash and resultant geopolymer. Inter-particle porosity is caused by gases released by combusting material and clays during the formation of the particle [10]. Closed porosity was observed in all of the fly ashes. Larger, irregular shaped particles contained a high degree of porosity (Figs. 2, 3c, and 6b). This was most evident in Tarong fly ash which was observed to have a higher degree of irregular shaped particles. Cenospheres and plerospheres were also observed in all fly ashes.
Fig. 6. SEM micrographs illustrating the variation of iron structures in Eraring fly ash. (a) Brain like iron structure; (b) fine cubic iron structure; (c) dispersed iron structure; (d) near solid iron structure. Operating conditions: Accelerating voltage = 20 kV and signal = backscattered electron.
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
3395
Fig. 7. SEM micrograph showing the particle distribution for Tarong fly ash.
SEM investigations indicated that the average particle density of the fly ashes was potentially different due to variations in the observed internal porosity and the amount of iron. Particle density was measured with the assumption that the particles had closed porosity, as supported by SEM analysis. Table 5 details the density results. The density results in this study were in good agreement with results in the literature where the density of other ashes was measured between 2 and 2.5 g/cm3 [14,24]. As expected, the concentration of iron had a large influence on the density. Collie fly ash, with the most iron, had the highest density. However, iron is not the only influence on density as the variations in measured density were greater than what is expected by the differences in the iron concentration between fly ashes. 3.3. Geopolymer synthesis Geopolymeric materials designed for high temperature applications do not necessarily require high mechanical strength. For instance, a high compressive strength is not essential for applications, such as fireproof coatings, whereas it is required for structural applications such as columns and tunnels. A more significant characteristic of a high temperature material is its stability during and after high temperature exposure. To get an indication of their suitability for high temperature applications, the fly ashes in this study were used to synthesise a range of geopolymers and subjected them to elevated temperatures. Table 1 details the elemental ratios of the geopolymer samples made in this study. The main compositional variable was the Si:Al ratio, which varied from 2.0 to 3.0. The other ratios were varied by a small amount in certain cases to achieve workable slurries during synthesis. Only the amorphous aluminosilicates, as determined by quantitative XRD (Table 4), were used in the compositional calculations. Each of the fly ashes required the activating solution to contribute a significant proportion of either the aluminium (Eraring and Tarong) or silicon (Collie) to the mixture to obtain the desired Si:Al ratio. The geopolymers were mechanically strength tested before and after exposure to 1000 ◦ C in a furnace to assess their strength retenTable 5 Particle density of each of the fly ashes. Oxide wt% taken from XRF results. Fly ash
Iron oxides (wt%)
Density (g/cm3 )
Collie Eraring Tarong
13.2 4.03 0.64
2.40 ± 0.04 2.02 ± 0.03 2.00 ± 0.03
Fig. 8. SEM micrographs comparing the microstructure of the geopolymers made from each fly ash before and after heating to 1000 ◦ C (Si:Al 2.0). (a, b and c) Collie, Eraring and Tarong fly ash geopolymers, respectively. (d, e and f) Collie, Eraring and Tarong fly ash geopolymers after exposure, respectively.
tion after exposure. Table 6 compares the compressive strength before and after elevated temperature exposure. From Table 6 it can be seen that the initial compressive strength varied greatly between the fly ashes. It is believed by the authors that the variation in compressive strength between the geopolymer mixes is largely due to differing levels of geopolymerisation between mixes. The greater the conversion of the amorphous aluminosilicates from the fly ash into geopolymer gel, the stronger the sample. Collie fly ash geopolymers with low Si:Al exhibited the highest initial compressive strength whereas the rest of the mixes produces geopolymers of at least 25 MPa, sufficient strength for most of the intended applications. Post firing compressive strengths were observed to be dependent on Si:Al ratio and iron content in the fly ash. Collie fly ash geopolymers exhibited the greatest strength loss after exposure. Visual inspection of the Collie fly ash samples after firing revealed a colour change from grey to red and a high degree of surface cracking. This has been observed previously [15] and is characteristic of the oxidation of the iron from the source fly ash. This effect was not observed in Eraring and Tarong fly ash geopolymers due to the low iron content. Consistently for each of the samples, the higher the Si:Al ratio, the better the strength retention (or increase) after high temperature exposure. Eraring geopolymers in particular exhibited exceptional strength increase after firing. Tarong fly ash geopolymers with Si:Al of 2.5 and 3.0 also increased in strength after exposure. Strength increases are believed to be due to sintering of the aluminosilicates of the geopolymer and the unreacted fly ash resulting in greater inter-particle connectivity. It was noted that the fired geopolymers had wider strength variability and thus a corresponding larger uncertainty than the un-fired geopolymers. This suggests that the geopolymers are more brittle after firing. SEM was performed on the geopolymers used in this study to analyse the microstructure (Fig. 8). Collie fly ash based geopolymers
3396
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
Table 6 Compressive strength of geopolymer made from each of the fly ashes. Note: The sample listed as ‘1’ indicates that it was too weak to be tested. Fly ash
Si:Al
28 day compressive strength (MPa)
Compressive strength after firing to 1000 ◦ C (MPa)
Percentage of room temperature strength
Collie Collie Collie Eraring Eraring Eraring Tarong Tarong Tarong
2.0 2.5 3.0 2.0 2.5 3.0 2.0 2.5 3.0
128(9) 53(10) 29(3) 31(2) 33(8) 28(5) 26(2) 26(4) 25(2)
24(9) 15(4) 1 78(11) 132(19) 126(20) 13(8) 73(17) 99(24)
19% 29% – 249% 396% 457% 49% 277% 396%
were observed to have the highest proportion of geopolymer gel indicating a higher degree of fly ash was dissolved into the matrix. This is likely due to Collie fly ash having the finest particle size distribution of the fly ashes. Eraring and Tarong geopolymers were observed to be primarily composed of partially reacted fly ash particles bonded by the geopolymer gel on their surface. The bulk of the completely un-reacted fly ash particles was larger than 20 m in all geopolymer samples. The morphology of the geopolymers changed significantly after firing to 1000 ◦ C. Sintering of the unreacted fly ash particles and the geopolymer gel resulted in a more homogenous and better connected microstructure in all samples. This however does not explain the strength loss in the Collie fly ash samples. It is believed by the authors that sintering caused by localised strength increases, though bulk cracking caused by crystallisation of the iron oxides resulted in the general strength loss. There was also an apparent increase in porosity. This is likely caused by densification of the geopolymer gel during sintering. The size of the macro pores was observed to vary between samples with Collie fly ash geopolymers having the largest. 4. Conclusions Results from the characterisation of three Australian fly ashes of varying compositions have been presented. Quantitative phase analysis determined that only a portion of each of the fly ashes is amorphous and as such reactive with the alkaline solution during geopolymerisation. SEM investigations supported the XRD analysis as identified crystalline phases were observed in the fly ash particles. An understanding of the location and morphology of each of the phases was also obtained by observing the interior structure of the fly ash particles in the SEM. The average particle density of fly ash was found to be largely dependent on the concentration of iron. However, internal porosity was also identified as a contributing factor. Collie fly ash appears to be the most reactive during geopolymerisation due to its finer particle size. This was supported by the high compressive strength in Collie fly ash geopolymers. Eraring and Tarong fly ash geopolymers only achieved moderate compressive strengths in the mixes synthesised in this study. Collie fly ash geopolymers exhibited significant mechanical strength loss after exposure to 1000 ◦ C, whereas most of the Eraring and Tarong fly ash geopolymers exhibited strength gains of up to 5-fold. 4.1. Implications for high temperature applications Fly ash, in general, is known to be suitable for use as a source material for geopolymers. However, additional considerations need to be made when assessing a source material for geopolymers that are designed to be used in high temperature applications. In this paper, the implications of certain characteristics of three fly ashes have been explored with reference to previous studies in the lit-
erature. Collie fly ash is the least suitable of the three fly ashes because of its high iron content. Eraring and Tarong fly ashes both have low iron content, making them more suitable for use in high temperature applications. Additional fly ash characteristics have also been identified as important. Fly ash particle size, morphology and presence of crystalline phases will greatly influence the characteristic of the resulting geopolymer. Finer fly ash particle size is preferable for increased ambient compressive strength. A spherical morphology is preferable for low water content geopolymer mixes (due to ease of workability) which is beneficial for reduced shrinkage at elevated temperatures. The presence of free quartz particles in the fly ashes may also reduce the workability and has the potential to induce expansion cracking at elevated temperatures. Acknowledgements The authors would like to thank Craig Heidrich from the Ash Development Association of Australia (ADAA) and Nick Cameron from Fly Ash Australia (FAA) for their assistance in acquiring the fly ash used in this study. The authors acknowledge the facilities, scientific and technical assistance of the Curtin University Electron Microscopy Laboratory, a facility funded by the university, state and Commonwealth governments. References [1] P. Duxson, A. Fernandez-Jimenez, J.L. Provis, G.C. Lukey, A. Palomo, J.S.J. van Deventer, J. Mater. Sci. 42 (2007) 2917–2933. [2] V.F. Barbosa, K.J. MacKenzie, Mater. Res. Bull. 38 (2003) 319–331. [3] J. Davidovits, J. Therm. Anal. 37 (1991) 1633–1656. [4] J.G.S. van Jaarsveld, J.S.J. van Deventer, G.C. Lukey, Mater. Lett. 57 (2003) 1272–1280. [5] N.W. Chen-Tan, A. van Riessen, V.L.Y. Chi, D.C. Southam, J. Am. Ceram. Soc. 92 (2009) 881–887. [6] A. Fernandez-Jimenez, A. Palomo, M. Criado, Cem. Concr. Res. 35 (2005) 1204–1209. [7] T. Bakharev, Cem. Concr. Res. 36 (2006) 1134–1147. [8] H. Rahier, J.F. Denayer, B. van Mele, J. Mater. Sci. 38 (2003) 3131–3136. [9] O.E. Manz, Fuel 78 (1999) 133–136. [10] R.T. Hemmings, E.E. Berry, Fly Ash and Coal Conversion By-Products: Characterization, Utilization, and Disposal V: Symposium, 1987, pp. 3–38. [11] A. Neville, Properties of Concrete, fourth ed., Longman Group Ltd., England, 1995. [12] R.W. Goodwin, Combustion Ash/Residue Management: An Engineering Perspective, Noyes Publications, 1993. [13] E.E. Berry, V.M. Malhotra, J. Am. Conc. Inst. 77 (1980) 59–73. [14] T. Matsunaga, J.K. Kim, S. Hardcastle, P.K. Rohatgi, Mater. Sci. Eng. A 325 (2002) 333–343. [15] W.D.A. Rickard, A. van Riessen, P. Walls, Int. J. Appl. Ceram. Technol. 7 (2010) 81–88. [16] D. Balzar, Voigt-function model in diffraction line-broadening analysis, in: Microstructure Analysis from Diffraction, International Union of Crystallography, 1999. [17] ASTM C39 (2005). [18] G. Kovalchuk, P.V. Krivenko, Producing fire and heat-resistant geopolymers, in: J.L. Provis, J.S.J. van Deventer, et al. (Eds.), Geopolymers: Structure, Processing, Properties and Industrial Applications, Woodhead Publishing Ltd., 2009, pp. 227–266. [19] R.P. Williams, A. van Riessen, Fuel 89 (2010) 3683–3692.
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397 [20] D. Kong, J.G. Sanjayan, K. Sagoe-Crentsil, Proceeding of the International Conference on Pozzolan, Concrete and Geoplymer, Khon Kaen, 2005, pp. 105–119. [21] Subaer, A. van Riessen, J. Mater. Sci. 42 (2006) 3117–3123.
3397
[22] G.A. Norton, R. Markuszewskl, H.R. Shanks, Environ. Sci. Technol. 20 (1986) 409–413. [23] D. Kong, J. Sanjayan, K. Sagoe-Crentsil, Cem. Concr. Res. 37 (2007) 1583–1589. [24] S.H. Lee, E. Sakai, M. Daimon, W.K. Bang, Cem. Concr. Res. 29 (1999) 1791–1797.
Contents lists available at ScienceDirect
Materials Science and Engineering A journal homepage: www.elsevier.com/locate/msea
Assessing the suitability of three Australian fly ashes as an aluminosilicate source for geopolymers in high temperature applications夽 William D.A. Rickard ∗ , Ross Williams, Jadambaa Temuujin, Arie van Riessen Centre for Materials Research, Curtin University, PO Box U 1987, Perth, WA 6845, Australia
a r t i c l e
i n f o
Article history: Received 27 October 2010 Received in revised form 4 January 2011 Accepted 5 January 2011 Available online 19 January 2011 Keywords: Geopolymer Fly ash Thermal properties
a b s t r a c t Fly ash characteristics cannot be assumed to be constant between power stations as they are highly dependent on the coal source and burning conditions. It is critical to understand the characteristics of fly ash in order to produce geopolymers suitable for high temperature applications. We report on the characterisation of fly ash from three Australian power stations in terms of elemental composition, phase composition, particle size, density and morphology. Geopolymers were synthesised from each of the fly ashes using sodium silicate and sodium aluminate solutions to achieve a range of Si:Al compositional ratios. Mechanical properties of geopolymer binders are presented and the effect of the source fly ash characteristics on the hardened product is discussed, as well as implications for high temperature applications. It was found that the twenty eight day strength of geopolymers is largely dependent on the sub 20 m size fraction of the fly ash. Strength loss after high temperature exposure was found to be dependent on the concentration of iron in the fly ash precursor and the Si:Al ratio of the geopolymer mixture. Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.
1. Introduction Geopolymers are a synthetic inorganic polymer produced by the alkali activation of aluminosilicates [1]. Geopolymers have gained a lot of interest in recent years amongst scientists and engineers because of their potential as an environmentally friendly replacement for ordinary Portland cement (OPC). Geopolymers have also
been considered for use in high temperature applications, such as fire proof coatings, making use of their intrinsic thermal stability [2]. The geopolymerisation reaction that creates a polymeric backbone of aluminium and silicon atoms is described in Eq. (1) [3]. Simplistically, geopolymerisation involves the dissolution of aluminosilicates from the source material followed by short ranged polymerisation to form a Si–Al based geopolymer backbone.
→
Si-Al source + Silicate + Water + Alkaline liquid
Geopolymer precursor
n(Si2O5,Al2O2) + 2nSiO2 + 4nH2O + NaOH (or KOH) → (Na+,K+) + n(OH)3-Si-O-Al --O-Si-(OH) 3 ⎜
(OH)2 Geopolymer precursor +
Alkaline ions
→
Geopolymer backbone ⎜
n(OH)3-Si-O-Al-O-Si-(OH) 3 + NaOH (or KOH)
⎜
(OH)2
夽 This project is carried out under the auspice and with the financial support of the Centre for Sustainable Resource Processing (CSRP), which is established and supported under the Australian government’s Cooperative Research Centres Program. ∗ Corresponding author. Tel.: +61 8 9266 4219; fax: +61 8 9266 2377. E-mail address: [email protected] (W.D.A. Rickard).
⎜
⎜
(Na+,K+) - (-Si-O-Al --O-Si-O-) + 4nH2O ⎜
⎜
⎜
O
O
O
(1)
The degree of dissolution of aluminosilicates in high pH alkaline solutions is largely dependent on the particle size, morphology and composition of the source material [4–8], in particular the amorphous aluminosilicates. Previous research has shown that geopolymers synthesised from aluminosilicate sources with suitable overall chemical composition but lacking an appropriate reactive component will result in incomplete dissolution and con-
0921-5093/$ – see front matter. Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2011.01.005
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
sequently be of lower strength [7,8]. The potential impact of not properly characterising the source material is more significant when using fly ash as an aluminosilicate source for synthesising geopolymers as it is known to have highly variable physical and chemical properties [9,10]. Geopolymers can be synthesised from many materials with a high concentration of aluminosilicates such as metakaolin or fly ash. This paper concentrates on the use of fly ash as a source material. Fly ash is produced from the coalescence of non-combustible material in the flue gases of coal fired power stations. The use of fly ash in the synthesis of geopolymers is not to be mistaken with the use of fly ash in ordinary Portland cement (OPC). In OPC, fly ash is used to improve specific characteristics, such as workability and reduction of temperature during curing, but does not add significantly to the strength of the binder [11]. In geopolymers, fly ash is the source of the aluminosilicates for the binder and is thus the critical component for strength development. Fly ash is a powdery material made up of small glass spheres, consisting primarily of silicon, aluminium, iron, and calcium oxides [12]. The typical particle size range is between 1 and 150 m [13]. Crystalline phases commonly identified in fly ashes are quartz, mullite and various iron rich phases such as hematite [14]. Class F fly ash, defined as having a (SiO2 + Al2 O3 + Fe2 O3 ) ≥ 70 wt%, is readily available in Australia and is produced from burning anthracite or bituminous coal [9]. Fernandez-Jimenez et al. [6] have proposed a model for the dissolution of fly ash in alkaline environments. This model is only loosely used as a guide in this study as it does not address the role of crystalline phases present in the fly ash. Bakharev [7] observed that fly ashes with a low amorphous content and a relatively large average particle size produced low strength geopolymers. It can be gathered from the literature that geopolymer source material must be carefully selected and accurately characterised in order to optimise its effectiveness in producing usable geopolymers. The desired characteristics of source material for geopolymers will depend on the final application. Thus geopolymers designed for use in high temperature applications will require different source material to geopolymers designed for use in structural applications. For example, previous research has shown that high levels of iron in fly ash have a negative effect on the high temperature performance of geopolymers [5,15]. The effect of other fly ash characteristics on high temperature geopolymers such as the propensity to crystallise and the effect on melting temperature have not been published in the literature. The objective of this paper is to convey an understanding of the important characteristics of fly ash and how these characteristics affect the formation of geopolymer cements. The measured data on the three Australian fly ashes used in this study will be useful for researchers and engineers looking to synthesise geopolymers from these sources in the future.
2. Experimental 2.1. Materials Fly ash was sourced from Collie power station in Western Australia, Eraring power station in New South Wales and Tarong power station in Queensland. The fly ash was provided by Fly Ash Australia (FAA). Sodium silicate solutions were prepared by dissolving sodium hydroxide pellets from Univar Pty Ltd. and fumed silica from Cabosil in deionised water. The solutions were allowed to dissolve for 24 h at 70 ◦ C prior to use. Sodium aluminate solutions were supplied by Coogee Chemicals who specified that the solution contained 19 wt% Al2 O3 , 25.5 wt% NaOH and 55.5 wt% H2 O.
3391
Table 1 Elemental ratios of the geopolymers prepared in this study. Fly ash
Si:Al
Na:Al
H:Si
Collie Collie Collie Eraring Eraring Eraring Tarong Tarong Tarong
2.0 2.5 3.0 2.0 2.5 3.0 2.0 2.5 3.0
1.25 1.25 1.25 1.25 1.25 1.25 1.32 1.25 1.25
5.0 4.5 6.0 5.0 4.0 4.5 6.4 5.5 5.4
2.2. Geopolymer synthesis Samples were prepared with a range of compositional ratios as shown in Table 1. Only the amorphous component of the aluminosilicates in the fly ashes was used in the calculation as the crystalline material was considered inert during the geopolymerisation reaction. Geopolymers were synthesised by mixing the fly ash and the alkaline solution in a planetary centrifugal mixer (ARE 250, Thinky, Japan) for 5 min at 1300 RPM. A further 30 s at 2100 RPM was used for defoaming. The slurry was then cast into various sized moulds for testing. The moulds were sealed and samples cured for 24 h at 70 ◦ C. Samples were stored at ambient temperature prior to testing. 2.3. XRF X-ray fluorescence (XRF) was outsourced to a commercial laboratory (Ultra Trace Pty Ltd.). The samples were fused in a glass disc prior to analysis. Loss on ignition (LOI) of fly ash was conducted up to 1000 ◦ C. 2.4. XRD analysis X-ray diffraction (XRD) patterns were collected on a Bruker D8 Advance diffractometer operated at 40 kV and 40 mA using Cu-K␣ radiation. The data was collected using a nominal 2 step size of 0.01◦ , count time of 0.5 s and a 2 range of 10–120◦ . Crystalline phases were identified using EVA version 11. Rietveld modelling of the data was performed using Bruker AXS TOPAS version 4.1. A fluorite (CaF2 ) internal standard (10 wt%) was used to facilitate quantitative analysis of the amorphous phase. The fundamental parameters approach was used to model the instrumental peak shape function, allowing the crystallite size to be determined [16]. 2.5. SEM Scanning electron microscopy (SEM) was conducted on an Evo 40XVP (Zeiss, Germany) and a Neon 40EsB (Zeiss, Germany) using secondary electrons (SE) as well as backscattered electrons (BSE). The accelerating voltage used was either 5 kV for SE imaging or 20 kV for BSE imaging. Flat samples were polished to 1 m using diamond paste. Samples were coated with carbon prior to imaging in the SEM. Elemental analysis was performed using an Oxford Instruments energy dispersive X-ray spectrometer (EDS). Analysis of X-ray spectra was performed using Inca-Analyser software (Oxford Instruments, England). 2.6. Particle sizing Particle size analysis was performed with a Malvern laser diffraction system at a commercial laboratory. The dispersing solution was deionised water with a sodium hexametaphosphate dispersing agent. The solution and approximately 1 wt% of fly ash
3392
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
was sonicated for 10 min to break up aggregates prior to size analysis.
Table 2 Oxide composition of each fly ash as determined by XRF. Uncertainties in brackets.
2.7. Density Density was calculated by measuring the displacement of fly ash in a fixed volume containing a fluid. A known mass of fly ash was placed in a 100 ml specific gravity (SG) bottle. The rest of the volume was filled with ethanol and the mixture was sonicated to remove any entrained air. The bottle was then weighed and the average particle density of the fly ash was calculated using the following formula: Flyash =
mflyash 100 − Vethanol
where Vethanol =
(mtotal − mflyash − mbottle ) ethanol
Oxide
Collie (wt%)
Eraring (wt%)
Tarong (wt%)
SiO2 Al2 O3 Fe2 O3 CaO K2 O TiO2 MgO Na2 O P2 O5 SrO BaO Other
51.38(8) 26.90(10) 13.20(2) 1.74(5) 0.90(4) 1.47(1) 1.41(3) 0.41(5) 1.09(2) 0.23(1) 0.38(1) 1.15(6)
65.47(8) 23.00(10) 4.03(2) 1.59(5) 1.68(4) 0.84(1) 0.51(3) 0.56(5) 0.27(2) 0.05(1) 0.06(1) 1.67(6)
73.68(8) 22.40(10) 0.64(2) 0.08(5) 0.53(4) 1.28(1) 0.17(3) 0.09(5) 0.08(2) 0.01(1) 0.03(1) 0.96(6)
LOI (1000 ◦ C) Sum of aluminosilicates Sum of alkali SiO2 /Al2 O3 Si/Al (molar)
0.44 78.3(13) 1.31(6) 1.91(1) 1.62(1)
1.37 88.47(13) 2.24(6) 2.85(1) 2.42(1)
0.79 96.08(13) 0.62(6) 3.29(2) 2.79(1)
All reported results are the average of 5 separate measurements. 2.8. Mechanical testing Cylinders of 25 mm diameter and 50 mm height were prepared for mechanical testing. Geopolymer samples were tested 28 days after synthesis. Compressive strength testing was conducted on a Lloyds universal tester EZ50. A load rate of 0.25 MPa/s was used to closely comply with ASTM C39 (a test method for concrete specimens) [17]. The stated strength values are the average of 4 repeat tests. 3. Results and discussion 3.1. Fly ash composition XRF was performed to assess the chemical composition of each fly ash (Table 2). As expected each fly ash was composed of aluminosilicates, iron oxides and a number of minor oxides. Each fly ash had a similar amount of Al2 O3 , though the amount of SiO2 varied from 51 wt% to 74 wt%. The molar silicon to aluminium ratio, varied from 1.6 (Collie) to 2.8 (Tarong). A significant difference between the fly ashes is the amount of iron oxide, which varied from 0.6 wt% (Tarong) to 13.2 wt% (Collie). Previous research has shown that volume changes caused by oxidation of the iron detrimentally affect the performance of fly ash geopolymers at elevated temperatures [15]. Iron rich fly ash particles have also been observed to inhibit the dissolution of aluminosilicates during geopolymerisation [5]. The sum of the alkalis initially present in the fly ashes is listed in Table 1. Alkalis are known to be excellent fluxing agents [18] and as such their added presence, as there will be alkali in the geopolymer, may reduce the melting point of the geopolymer (unless a more thermally stable crystalline phase is formed). The XRF data provides a preliminary indication as to the suitability of each of the fly ashes for production of fire resistant geopolymers although it is limited by the fact that there is no
information about the reactivity of the aluminosilicate component. Research by Chen-Tan et al. [5] showed that only the amorphous aluminosilicates in the fly ash are reactive in the geopolymerisation reaction that forms a geopolymer. Thus in order to properly quantify the reactive component of each of the fly ashes, the phase composition was determined. A fluorite internal standard was used to facilitate quantitative analysis of the amorphous phase. The phase composition of each of the fly ashes is presented in Table 3. It was observed that there were two populations of quartz of differing crystallite size. It is thought that the large crystallite size quartz (>100 nm) is primary quartz, i.e. it is present prior to the coal combustion process and exists as discreet particles amongst the fly ash. The second quartz population that was identified had a much smaller crystallite size (<100 nm) and slightly larger lattice parameters, and is thought to have formed during or after combustion, thus is secondary quartz, and is believed to be present within fly ash spheres. Previous research by Williams and van Riessen has also found a similar distribution of the quartz crystallite size in Collie fly ash [19]. The presence of quartz in a source material is undesirable for geopolymers designed for high temperature applications due to differential expansion upon heating. This can cause micro cracking which reduces the strength of the material. This problem is more significant where the particle size of the quartz is larger. Kong et al. [20] observed that the presence of quartz based aggregates significantly reduced the compressive strength of fly ash geopolymers upon heating to 800 ◦ C. Similar results have been observed with quartz based OPC mortars [21]. Differential thermal expansion between the primary quartz in the fly ash and geopolymer will reduce the strength of the composite at high temperature. It is the opinion of the authors that the secondary quartz detected in the studied fly ashes will not adversely affect the high temperature performance of the resulting geopolymer due to its small particle size and location within the fly ash spheres (see Section 3.2).
Table 3 Phase composition of Collie fly ash. Uncertainties in brackets. Phase
Formula
Amorphous content Mullite (ICSD 66452) Mullite (ICSD 66449) Quartz low (ICSD 83849) Quartz low Primary (ICSD 83849) Magnetite (ICSD 43001) Hematite (ICSD 88417)
Al4.56 Si1.44 O9.72 Al4.59 Si1.41 O9.7 SiO2 SiO2 Fe3 O4 Fe2 O3
Collie fly ash (wt%)
Eraring fly ash (wt%)
Tarong fly ash (wt%)
54.00 (45) 15.80 (18)
62.74 (31)
50.82 (28) 25.1 (11)
11.14 (18) 15.05 (21) 2.51 (83) 1.50 (64)
20.88 (14) 8.08 (16) 6.81 (14) 1.491 (52)
10.31 (14) 13.77 (13)
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
3393
Table 4 Amorphous composition of each of the fly ashes. Uncertainties in brackets. Oxide
Collie fly ash (wt%)
Eraring fly ash (wt%)
Tarong fly ash (wt%)
SiO2 Al2 O3 Fe2 O3 CaO K2 O TiO2 MgO Na2 O P2 O5 SrO BaO Other
20.90(65) 15.39(41) 9.11(32) 1.74(5) 0.90(4) 1.47(1) 1.41(3) 0.41(5) 1.09(2) 0.23(1) 0.38(1) 1.15(6)
45.03(65) 7.67(32) 2.49(16) 1.59(5) 1.68(4) 0.84(1) 0.51(3) 0.56(5) 0.27(2) 0.05(1) 0.06(1) 1.67(6)
42.79(59) 4.11(26) 0.64(2) 0.08(5) 0.53(4) 1.28(1) 0.17(3) 0.09(5) 0.08(2) 0.01(1) 0.03(1) 0.96(6)
Sum of amorphous aluminosilicates SiO2 /Al2 O3 Si/Al (molar)
36.29(77) 1.36(6) 1.15(4)
52.70(73) 5.87(26) 4.98(19)
46.90(64) 10.42(68) 8.84(49)
Quantitative XRD used to determine the amorphous content of each fly ash resulted in the following values; Collie 54.0 wt%, Eraring 62.7 wt% and Tarong 50.8 wt%. The elemental composition of the amorphous phase is calculated by subtracting the contribution of the crystalline phases from the XRF data. Table 4 details the amorphous content of each of the fly ashes. It should be noted that any crystalline phases in the fly ashes that are below the detection limit of the XRD are included in the amorphous content. The Si:Al ratio of the amorphous content was measured to be significantly different to the bulk (Table 2), especially for Eraring and Tarong fly ashes. Geopolymer made from these fly ashes should be synthesised using the Si:Al ratios in Table 4. The low concentration of reactive aluminium in Eraring and Tarong fly ashes requires the use of alkali–aluminate solutions rather than the commonly used alkali–silicate solution to achieve typical Si:Al ratios. The bulk of the iron in each of the fly ashes is amorphous; the highest proportion of crystalline to amorphous iron oxide was measured to be 0.38 (Collie fly ash). The low concentration of crystalline iron oxide may be in part due to a very small crystallite size and X-ray absorption by the covering material, as iron phases are commonly found within glass spheres [22]. The amorphous iron is known to order to phases such as hematite during thermal treatment with concomitant volume changes which have been observed to adversely affect the thermal performance of a fly ash based geopolymer [15]. 3.2. Fly ash morphology
the surface area can be calculated from the particle size data. Collie fly ash was calculated to have a specific surface area of 1.56 m2 /cm3 (1.56 × 106 m2 /m3 ). This value is much higher than Eraring and Tarong fly ashes, having specific surface areas of 0.92 m2 /cm3 and 0.99 m2 /cm3 , respectively. Based on specific surface area it would be expected that Collie fly ash would dissolve faster than Eraring and Tarong fly ashes once placed in an alkaline solution due to a smaller particle size. However, the phase and inter-particle location of the aluminosilicates will also affect the dissolution rate. Particle size is not a complete assessment of the reactivity of the amorphous material in a fly ash during geopolymerisation or its workability during casting. SEM analysis was done to assess fly ash morphology, looking specifically at the particle shape and the location of the aluminosilicate glass within the particles. If the glass is encapsulated within an un-reactive crystalline material then it would not be available for dissolution during geopolymerisation. Fly ash was mounted in resin and polished to 1 m surface finish in order to observe the interior structure of the particles. The cross section of the fly ash particles in all samples showed a typical spherical morphology with a large concentration of particles appearing to be glassy. A high degree of interparticle and intraparticle heterogeneity was also observed. This has been observed previously [10] and is due to local variations in temperature and composition during the coal combustion and fly ash capture process. Figs. 2, 4 and 7 give an indication of the typical particle morphology of Collie, Eraring and Tarong fly ashes, respectively. Tarong fly ash was observed to have more irregular shaped parti-
Fly ash morphology is known to affect bulk characteristics of the subsequent geopolymer [4]. Spherical morphology is beneficial to the synthesis of geopolymers as it allows for good workability at low liquids to solids mix ratios [23]. Low water content is often desirable in high temperature applications as it reduces dehydration shrinkage when heated. Particle size determines the surface area that is initially available for dissolution by the alkaline solution. It is also known that smaller fly ash particles (<20 m) are more likely to have a highly glassy composition (ideal for geopolymerisation), as small particles quench faster than large particles [10]. Fig. 1 compares the particle size distribution for each of the fly ashes in this study. Each fly ash was found to have a fineness better than the Australian standard for fly ash in cement (AS3582.1), where at least 75% of all particles must be smaller than 45 m. The percentage of particles smaller than 45 m was similar for each of the fly ashes. The major differences in particle size distribution occurred below 10 m. Collie fly ash had 43% of its particles smaller than 10 m whereas Eraring fly ash only had 27%. If the bulk of the fly ash particles are assumed to be spherical, an estimation of
Fig. 1. Particle size distribution for each fly ash. 45 m fineness: Collie (83%), Eraring (79%), Tarong (84%). 10 m fineness: Collie (43%), Eraring (27%), Tarong (36%).
3394
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
Fig. 2. SEM micrograph showing the particle distribution for Collie fly ash. The arrow indicates a porous, irregular shaped particle.
Fig. 3. SEM micrograph showing different particles observed in Collie fly ash. (a) Fly ash particle containing blocky crystallites likely to be quartz; (b) quartz particle; (c) porous fly ash particle; (d) fly ash particle containing iron. Operating conditions: Accelerating voltage = 5 kV and signal = secondary electron.
cles than the other fly ashes. This has the propensity to reduce the workability of geopolymer slurries made from this fly ash. Figs. 3, 5 and 6 show examples of the various types of particles observed in the fly ashes. Many of the crystalline phases, detected in the XRD analysis, were observed in the SEM images by searching for characteristic morphology and using EDS for elemental identification. Quartz was typically observed as separate
Fig. 4. SEM micrograph showing the particle distribution for Eraring fly ash.
Fig. 5. SEM micrograph of Eraring fly ash showing a fine mullite structure in a particle that appeared glassy at low magnification.
particles, whereas iron and mullite phases were only observed within glassy particles. There were, however, some blocky crystalline structures observed in some glassy particles (Fig. 3a) which were likely to be secondary quartz phases formed during the coal combustion process. This observation supports the finding of a secondary quartz population of smaller crystallite size detected by XRD analysis (Table 2). Iron structures were observed with a range of morphologies (Figs. 3d and 6). This is also in good agreement with the XRD results where a range of iron phases, including an amorphous phase, were identified. Fig. 5 shows an example of a particle that appeared glassy at low magnification, only to find the bulk of it was crystalline at higher magnification. The needle shaped crystals are approximately 1 m long and 100–200 nm wide, characteristic of mullite. The authors of this study were also aware that inter-particle porosity affects the density of the fly ash and resultant geopolymer. Inter-particle porosity is caused by gases released by combusting material and clays during the formation of the particle [10]. Closed porosity was observed in all of the fly ashes. Larger, irregular shaped particles contained a high degree of porosity (Figs. 2, 3c, and 6b). This was most evident in Tarong fly ash which was observed to have a higher degree of irregular shaped particles. Cenospheres and plerospheres were also observed in all fly ashes.
Fig. 6. SEM micrographs illustrating the variation of iron structures in Eraring fly ash. (a) Brain like iron structure; (b) fine cubic iron structure; (c) dispersed iron structure; (d) near solid iron structure. Operating conditions: Accelerating voltage = 20 kV and signal = backscattered electron.
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
3395
Fig. 7. SEM micrograph showing the particle distribution for Tarong fly ash.
SEM investigations indicated that the average particle density of the fly ashes was potentially different due to variations in the observed internal porosity and the amount of iron. Particle density was measured with the assumption that the particles had closed porosity, as supported by SEM analysis. Table 5 details the density results. The density results in this study were in good agreement with results in the literature where the density of other ashes was measured between 2 and 2.5 g/cm3 [14,24]. As expected, the concentration of iron had a large influence on the density. Collie fly ash, with the most iron, had the highest density. However, iron is not the only influence on density as the variations in measured density were greater than what is expected by the differences in the iron concentration between fly ashes. 3.3. Geopolymer synthesis Geopolymeric materials designed for high temperature applications do not necessarily require high mechanical strength. For instance, a high compressive strength is not essential for applications, such as fireproof coatings, whereas it is required for structural applications such as columns and tunnels. A more significant characteristic of a high temperature material is its stability during and after high temperature exposure. To get an indication of their suitability for high temperature applications, the fly ashes in this study were used to synthesise a range of geopolymers and subjected them to elevated temperatures. Table 1 details the elemental ratios of the geopolymer samples made in this study. The main compositional variable was the Si:Al ratio, which varied from 2.0 to 3.0. The other ratios were varied by a small amount in certain cases to achieve workable slurries during synthesis. Only the amorphous aluminosilicates, as determined by quantitative XRD (Table 4), were used in the compositional calculations. Each of the fly ashes required the activating solution to contribute a significant proportion of either the aluminium (Eraring and Tarong) or silicon (Collie) to the mixture to obtain the desired Si:Al ratio. The geopolymers were mechanically strength tested before and after exposure to 1000 ◦ C in a furnace to assess their strength retenTable 5 Particle density of each of the fly ashes. Oxide wt% taken from XRF results. Fly ash
Iron oxides (wt%)
Density (g/cm3 )
Collie Eraring Tarong
13.2 4.03 0.64
2.40 ± 0.04 2.02 ± 0.03 2.00 ± 0.03
Fig. 8. SEM micrographs comparing the microstructure of the geopolymers made from each fly ash before and after heating to 1000 ◦ C (Si:Al 2.0). (a, b and c) Collie, Eraring and Tarong fly ash geopolymers, respectively. (d, e and f) Collie, Eraring and Tarong fly ash geopolymers after exposure, respectively.
tion after exposure. Table 6 compares the compressive strength before and after elevated temperature exposure. From Table 6 it can be seen that the initial compressive strength varied greatly between the fly ashes. It is believed by the authors that the variation in compressive strength between the geopolymer mixes is largely due to differing levels of geopolymerisation between mixes. The greater the conversion of the amorphous aluminosilicates from the fly ash into geopolymer gel, the stronger the sample. Collie fly ash geopolymers with low Si:Al exhibited the highest initial compressive strength whereas the rest of the mixes produces geopolymers of at least 25 MPa, sufficient strength for most of the intended applications. Post firing compressive strengths were observed to be dependent on Si:Al ratio and iron content in the fly ash. Collie fly ash geopolymers exhibited the greatest strength loss after exposure. Visual inspection of the Collie fly ash samples after firing revealed a colour change from grey to red and a high degree of surface cracking. This has been observed previously [15] and is characteristic of the oxidation of the iron from the source fly ash. This effect was not observed in Eraring and Tarong fly ash geopolymers due to the low iron content. Consistently for each of the samples, the higher the Si:Al ratio, the better the strength retention (or increase) after high temperature exposure. Eraring geopolymers in particular exhibited exceptional strength increase after firing. Tarong fly ash geopolymers with Si:Al of 2.5 and 3.0 also increased in strength after exposure. Strength increases are believed to be due to sintering of the aluminosilicates of the geopolymer and the unreacted fly ash resulting in greater inter-particle connectivity. It was noted that the fired geopolymers had wider strength variability and thus a corresponding larger uncertainty than the un-fired geopolymers. This suggests that the geopolymers are more brittle after firing. SEM was performed on the geopolymers used in this study to analyse the microstructure (Fig. 8). Collie fly ash based geopolymers
3396
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397
Table 6 Compressive strength of geopolymer made from each of the fly ashes. Note: The sample listed as ‘1’ indicates that it was too weak to be tested. Fly ash
Si:Al
28 day compressive strength (MPa)
Compressive strength after firing to 1000 ◦ C (MPa)
Percentage of room temperature strength
Collie Collie Collie Eraring Eraring Eraring Tarong Tarong Tarong
2.0 2.5 3.0 2.0 2.5 3.0 2.0 2.5 3.0
128(9) 53(10) 29(3) 31(2) 33(8) 28(5) 26(2) 26(4) 25(2)
24(9) 15(4) 1 78(11) 132(19) 126(20) 13(8) 73(17) 99(24)
19% 29% – 249% 396% 457% 49% 277% 396%
were observed to have the highest proportion of geopolymer gel indicating a higher degree of fly ash was dissolved into the matrix. This is likely due to Collie fly ash having the finest particle size distribution of the fly ashes. Eraring and Tarong geopolymers were observed to be primarily composed of partially reacted fly ash particles bonded by the geopolymer gel on their surface. The bulk of the completely un-reacted fly ash particles was larger than 20 m in all geopolymer samples. The morphology of the geopolymers changed significantly after firing to 1000 ◦ C. Sintering of the unreacted fly ash particles and the geopolymer gel resulted in a more homogenous and better connected microstructure in all samples. This however does not explain the strength loss in the Collie fly ash samples. It is believed by the authors that sintering caused by localised strength increases, though bulk cracking caused by crystallisation of the iron oxides resulted in the general strength loss. There was also an apparent increase in porosity. This is likely caused by densification of the geopolymer gel during sintering. The size of the macro pores was observed to vary between samples with Collie fly ash geopolymers having the largest. 4. Conclusions Results from the characterisation of three Australian fly ashes of varying compositions have been presented. Quantitative phase analysis determined that only a portion of each of the fly ashes is amorphous and as such reactive with the alkaline solution during geopolymerisation. SEM investigations supported the XRD analysis as identified crystalline phases were observed in the fly ash particles. An understanding of the location and morphology of each of the phases was also obtained by observing the interior structure of the fly ash particles in the SEM. The average particle density of fly ash was found to be largely dependent on the concentration of iron. However, internal porosity was also identified as a contributing factor. Collie fly ash appears to be the most reactive during geopolymerisation due to its finer particle size. This was supported by the high compressive strength in Collie fly ash geopolymers. Eraring and Tarong fly ash geopolymers only achieved moderate compressive strengths in the mixes synthesised in this study. Collie fly ash geopolymers exhibited significant mechanical strength loss after exposure to 1000 ◦ C, whereas most of the Eraring and Tarong fly ash geopolymers exhibited strength gains of up to 5-fold. 4.1. Implications for high temperature applications Fly ash, in general, is known to be suitable for use as a source material for geopolymers. However, additional considerations need to be made when assessing a source material for geopolymers that are designed to be used in high temperature applications. In this paper, the implications of certain characteristics of three fly ashes have been explored with reference to previous studies in the lit-
erature. Collie fly ash is the least suitable of the three fly ashes because of its high iron content. Eraring and Tarong fly ashes both have low iron content, making them more suitable for use in high temperature applications. Additional fly ash characteristics have also been identified as important. Fly ash particle size, morphology and presence of crystalline phases will greatly influence the characteristic of the resulting geopolymer. Finer fly ash particle size is preferable for increased ambient compressive strength. A spherical morphology is preferable for low water content geopolymer mixes (due to ease of workability) which is beneficial for reduced shrinkage at elevated temperatures. The presence of free quartz particles in the fly ashes may also reduce the workability and has the potential to induce expansion cracking at elevated temperatures. Acknowledgements The authors would like to thank Craig Heidrich from the Ash Development Association of Australia (ADAA) and Nick Cameron from Fly Ash Australia (FAA) for their assistance in acquiring the fly ash used in this study. The authors acknowledge the facilities, scientific and technical assistance of the Curtin University Electron Microscopy Laboratory, a facility funded by the university, state and Commonwealth governments. References [1] P. Duxson, A. Fernandez-Jimenez, J.L. Provis, G.C. Lukey, A. Palomo, J.S.J. van Deventer, J. Mater. Sci. 42 (2007) 2917–2933. [2] V.F. Barbosa, K.J. MacKenzie, Mater. Res. Bull. 38 (2003) 319–331. [3] J. Davidovits, J. Therm. Anal. 37 (1991) 1633–1656. [4] J.G.S. van Jaarsveld, J.S.J. van Deventer, G.C. Lukey, Mater. Lett. 57 (2003) 1272–1280. [5] N.W. Chen-Tan, A. van Riessen, V.L.Y. Chi, D.C. Southam, J. Am. Ceram. Soc. 92 (2009) 881–887. [6] A. Fernandez-Jimenez, A. Palomo, M. Criado, Cem. Concr. Res. 35 (2005) 1204–1209. [7] T. Bakharev, Cem. Concr. Res. 36 (2006) 1134–1147. [8] H. Rahier, J.F. Denayer, B. van Mele, J. Mater. Sci. 38 (2003) 3131–3136. [9] O.E. Manz, Fuel 78 (1999) 133–136. [10] R.T. Hemmings, E.E. Berry, Fly Ash and Coal Conversion By-Products: Characterization, Utilization, and Disposal V: Symposium, 1987, pp. 3–38. [11] A. Neville, Properties of Concrete, fourth ed., Longman Group Ltd., England, 1995. [12] R.W. Goodwin, Combustion Ash/Residue Management: An Engineering Perspective, Noyes Publications, 1993. [13] E.E. Berry, V.M. Malhotra, J. Am. Conc. Inst. 77 (1980) 59–73. [14] T. Matsunaga, J.K. Kim, S. Hardcastle, P.K. Rohatgi, Mater. Sci. Eng. A 325 (2002) 333–343. [15] W.D.A. Rickard, A. van Riessen, P. Walls, Int. J. Appl. Ceram. Technol. 7 (2010) 81–88. [16] D. Balzar, Voigt-function model in diffraction line-broadening analysis, in: Microstructure Analysis from Diffraction, International Union of Crystallography, 1999. [17] ASTM C39 (2005). [18] G. Kovalchuk, P.V. Krivenko, Producing fire and heat-resistant geopolymers, in: J.L. Provis, J.S.J. van Deventer, et al. (Eds.), Geopolymers: Structure, Processing, Properties and Industrial Applications, Woodhead Publishing Ltd., 2009, pp. 227–266. [19] R.P. Williams, A. van Riessen, Fuel 89 (2010) 3683–3692.
W.D.A. Rickard et al. / Materials Science and Engineering A 528 (2011) 3390–3397 [20] D. Kong, J.G. Sanjayan, K. Sagoe-Crentsil, Proceeding of the International Conference on Pozzolan, Concrete and Geoplymer, Khon Kaen, 2005, pp. 105–119. [21] Subaer, A. van Riessen, J. Mater. Sci. 42 (2006) 3117–3123.
3397
[22] G.A. Norton, R. Markuszewskl, H.R. Shanks, Environ. Sci. Technol. 20 (1986) 409–413. [23] D. Kong, J. Sanjayan, K. Sagoe-Crentsil, Cem. Concr. Res. 37 (2007) 1583–1589. [24] S.H. Lee, E. Sakai, M. Daimon, W.K. Bang, Cem. Concr. Res. 29 (1999) 1791–1797.