Effect of Curing Temperature and Silicate Concentration on Fly-Ash

Apr 14, 2006 - the extent and rate of reaction, shown through an increase in mesopore volume, surface area, and an accelerated setting time. The kinet...
0 downloads 0 Views 824KB Size
Ind. Eng. Chem. Res. 2006, 45, 3559-3568

3559

Effect of Curing Temperature and Silicate Concentration on Fly-Ash-Based Geopolymerization Sindhunata, J. S. J. van Deventer,* G. C. Lukey, and H. Xu Department of Chemical and Biomolecular Engineering, UniVersity of Melbourne, Melbourne, Victoria 3010, Australia

The development of the pore structure of geopolymers synthesized from class F fly ash was studied using electron microscopy and porosimetry. Fly-ash-based geopolymer can be classified as a mesoporous aluminosilicate material, with a Si/Al composition varying from 1.51 to 2.24. The Si/Al composition and pore structure of fly-ash-based geopolymer vary depending on the curing temperature and the silicate ratio of the activating solutions (SiO2/M2O, M ) Na or K). A higher Si/Al ratio and finer pores are obtained in geopolymers synthesized at higher temperature and silicate ratios. Elevating the curing temperature increases the extent and rate of reaction, shown through an increase in mesopore volume, surface area, and an accelerated setting time. The kinetics appears to be temperature-controlled only before the material is hardened. Very high silicate ratios (SiO2/M2O g 2.0) are also believed to slow the reactions. The pore structure of K-based geopolymer is more susceptible to change in temperature than that of Na-based geopolymer. Introduction Geopolymers are amorphous aluminosilicate binders that can be synthesized utilizing sol-gel chemistry at room temperature. The process of geopolymerization starts with hydrolysis on the solid surface through exchange of H+ with monovalent cations (K+ or Na+) from the bulk solution.1 The subsequent step is believed to be the continuous dissolution of aluminosilicate precursors by the breaking of Si-O-Si or Si-O-Al bonds from the sol particles to form the reactive precursors Si(OH)4 and Al(OH)4- in solution.2 The dissolution step occurs concomitantly with precipitation on the solid surface, which is known as the reorganization of silicates and aluminates.3 Then, polymerization occurs through condensation of Si and Al, expelling water and leaving unreacted excess alkali in the liquid phase.1 The syntheses of geopolymer materials from natural aluminosilicate minerals, namely, kaolinite, feldspar, and albite have been implemented successfully.2 The use of calcined kaolinite (metakaolinite) to synthesize geopolymers is applied widely, mainly because of the high surface area and amorphous aluminosilicate structure of metakaolinite.1,4,5 Industrial byproducts, such as slags and fly ashes, have also been used because these materials have high contents of reactive glassy aluminosilicate phases.6,7 Fly-ash-based geopolymers have potential applications in the immobilization of toxic heavy metals8 and the manufacturing of construction products.6 Nonetheless, fly ashes have been primarily used as pozzolanic admixtures in cement. The replacement of cement with fly ash (10-30 wt %) could reduce global cement consumption, which would result in the reduction of CO2 emissions associated with cement manufacturing.9 The addition of fly ash to cement, particularly class F fly ash (low in calcium content), is found to reduce porosity and fluid permeability in the cement paste.10 Chang et al.11 showed that class F fly ash can be converted to mesoporous aluminosilicate materials through the fusion of NaOH and fly ash at elevated temperatures. The Si/Al composition of this fly-ash-derived material is 13.4, and its pore structure * To whom correspondence should be addressed. Tel.: + 61 3 8344 6619. Fax: + 61 3 8344 7707. E-mail: [email protected].

resembles that of MCM-41, a mesoporous material with higher Si/Al composition that is commonly used as molecular sieves. Nonetheless, the pore structure of geopolymeric materials, particularly those derived from fly ash, has not yet been investigated in detail. The purpose of this article is to characterize the pore microstructure of fly-ash-based geopolymers using various pore characterization techniques, i.e., MIP (mercury intrusion porosimetry), N2 adsorption, and electron microscopy. The characterization of the microstructure of the pores in geopolymers is essential to explaining their observed physical and chemical properties.12 It is believed that porous geopolymers can also be tailored to produce specific material properties, such as high mechanical strength, flexibility, and durability and low fluid permeability. Second, the formation of pores in aluminosilicate gel has been found to be dependent on pH, temperature, and surfactant concentration and type.13,14 Therefore, the effects of elevated curing temperature and different activators (alkaline silicate solutions) on the formation of geopolymeric gels are expected to be significant. The porosity of fly-ash-based geopolymers results predominantly from the inherent porous nature of geopolymer gel. The pore structure in the gel is formed during the drying process, when the aqueous pore solution evaporates, leaving empty voids within the geopolymer gel matrixes.15 In the synthesis of porous silica gel, this type of gel is known as dried gel or xerogel.16 Second, in an unreacted or poorly reactive system, the geopolymer gel is rarely formed. Instead, the dissolved ionic species, such as Si, Al, Ca, and Fe, which are leached out from various glassy phases in the fly ash, precipitate on the surface of the fly ash particles. Therefore, there is insufficient geopolymer gel to fill the gaps between two or more unreacted fly ash particles. These voids are generally large (micron scale) and can be distinguished easily from the porous structure of the geopolymer gel. Finally, the porosity could originate from within the partially reacted fly ash particles. It has been found that at least four types of particle morphology are commonly present in fly ash, i.e., solid spheres, plerospheres, clathrospheres, and cenospheres (hollow spheres).17,18 The different types of fly ash particles are separated according to their densities, with the density of a solid sphere being the highest (2.4-2.5 g/cm3) and that of a cenosphere being the lowest (e1.4 g/cm3). Upon reaction, only

10.1021/ie051251p CCC: $33.50 © 2006 American Chemical Society Published on Web 04/14/2006

3560

Ind. Eng. Chem. Res., Vol. 45, No. 10, 2006

Table 1. Oxide Composition of Gladstone Fly Ash component

composition (mass %)

SiO2 Al2O3 Fe2O3 TiO2 MnO CaO MgO K2O Na2O P2O5 SO3 loss on ignition total

50.01 27.97 11.95 1.30 0.19 3.48 1.32 0.69 0.19 1.01 0.50 1.39 100.00

a partially reacted cenosphere is likely to leave a significant hollow space, which is accessible to fluid penetration (an open pore) and affects the pore structure analysis. The density of Gladstone fly ash (GFA), the class F fly ash used in this study, is 2.34 g/cm3. Therefore, judging from the measured density, the amount of cenospheres is believed to be minimal. Furthermore, in a well-reacted geopolymer, it is likely that the hollow spaces originating from partly reacted cenospheres are filled with forming geopolymer gel. Experimental Methods Materials. Gladstone fly ash (GFA) was obtained from Queensland Cement Ltd. (QCL). The oxide composition of GFA was determined using a Siemens SRS3000 X-ray fluorescence (XRF) spectrometer (Table 1). The composition of the major components was 50.01 mass % SiO2, 27.97 mass % Al2O3, and 11.95 mass % Fe2O3. X-ray diffraction analysis using a Philips PW 1800 instrument revealed that the major crystalline phases of the fly ash were R-quartz (SiO2), mullite (3.2 Al2O3 SiO2), hematite (Fe2O3), and magnetite (Fe3O4). These crystalline phases are probably not reactive when fly ash is activated in a concentrated acidic or basic solution.18 Separate quantitative XRD analysis using SiroQuant software showed that GFA contains 17 mass % of Al phases and 39.1 mass % of Si phases that are amorphous. These primary amorphous phases can be fully hydrolyzed in basic solution. Other authors18,19 have reported that the glassy phases could also be associated with calcium, iron, and alkali elements, particularly sodium in a class F fly ash. A detailed glassy phase characterization is, however, beyond the scope of this paper. Major, but highly localized fluctuations in chemical composition of the glassy phases were also noted for different fly ash particles, as well as from point to point within any one particle.19 Furthermore, using nitrogen adsorption on a Tristar instrument, the BET surface area of GFA was determined to be 1.337 m2/ g, and the mean particle density was found to be 2.34 g/cm3. The particle size distribution of fly ash particles was 2.1 µm (d10), 12.4 µm (d50), and 70.0 µm (d90), as determined by a Coultier particle size analyzer. Sodium silicate solution (Vitrosol N40) and potassium silicate solution (KASIL 2040) were obtained from PQ Australia. Potassium hydroxide and sodium hydroxide (Orica Australia) were mixed with alkaline silicate solutions and used as activating solutions. Synthesis. Geopolymer samples were synthesized by mixing fly ash with specific activating solutions. The weight ratio of water to fly ash was kept constant at 0.3. This gave a constant solid/solution ratio for each geopolymer sample. If the amorphous contents of Si- and Al-associated phases are assumed to be fully reacted and the crystalline phases are assumed not to be reactive, the elemental gel composition of the geopolymeric

Table 2. Compositions of Geopolymer Samples Synthesized with a Constant H2O/M2O Ratio of 14.85, M ) Na or K sample

SiO2/M2O ratio in solution

theoretical Si/Ala

0.0K 0.79K 1.40K 2.0K 0.0Na 0.79Na 1.40Na 2.0Na

0 0.79 1.40 2.0 0 0.79 1.40 2.0

1.95 2.21 2.42 2.62 1.95 2.21 2.42 2.62

a Si/Al ratio calculated assuming that all amorphous silicon and aluminum in the fly ash are fully reacted.

gel can be calculated. Clearly, this is an approximation, as some glassy phases are known to be more reactive than others. Also, any secondary reaction products such as zeolites or silica gel that might form, even in minute quantities, are not considered here. Therefore, the estimated Si/Al ratio gives a relative value that is used here to indicate trends rather than values of absolute accuracy. The final composition of each geopolymer sample and its estimated Si/Al ratio are given in Table 2. After being mixed, the resulting slurry was cast in 35 × 75 mm cylindrical vials, sealed, and vibrated for 2 min to eliminate the air entrapped during mixing, which would affect the pore structure analysis. All samples were cured at three different curing temperatures, namely, 30, 50 and 75 °C, for 24 h. Some samples were cured for as long as 48 h. The approximate setting time of each geopolymer, cured at 30, 50, and 75 °C was determined by tilting the clear plastic vials containing the geopolymers. The setting time is defined as the time when geopolymer paste stops flowing. The setting time is a more appropriate terminology than gelation time or precipitation time because the dissolution and polycondensation are believed to occur simultaneously in the course of geopolymerization; thus, there is no clear time distinction between the two processes. MIP and N2 Adsorption. After 24 h, samples were removed from the molds and cut in cross section into four pieces with a diamond saw. The two middle pieces were dried in a vacuum oven at 105 °C for 24 h or until there was no reduction in the sample mass. Although this drying step changed the pore structure, it was essential to obtain consistent results. The dried geopolymers were then subjected to MIP (Autopore II 9220), with the pressure applied being 0.5-60 000 psia after a 15-s equilibration time, based on the standard practice used for aluminosilicate materials.20 The pore size distribution obtained from MIP tests, ranging from 2 nm to 100 µm, was calculated using the Washburn equation21 and assuming an interconnected tubular pore network model.20 A Tristar instrument was used to obtain N2 adsorption data after an equilibration time of 15 s, from which the pore size distribution22 was calculated using the Barret, Joyner, and Halenda (BJH)23 modification of the Kelvin equation.22 The mesopore volume (Vp) and surface area (Sp) obtained from the BJH method were calculated using Win 3000 software. Microstructure and Geopolymer Gel Composition. Micrographs of fly-ash-based geopolymers were obtained using a Philips XL30 FEG field-emission scanning electron microscope (FESEM) operated at 2 kV. The fly-ash-based geopolymer was crushed, then a piece of approximately 0.1 cm3 was mounted with epoxy resin, and further polishing was done. The elemental composition of the geopolymer samples was obtained using EDAX (energy-dispersive X-ray) spectrometry. The elements routinely determined were Al, Si, K, Na, Ca, and Fe, but only the Si/Al ratio is presented here. Twenty measurements were

Ind. Eng. Chem. Res., Vol. 45, No. 10, 2006 3561

Figure 1. Electron micrographs of various fly-ash-based geopolymers: (A) Gladstone fly ash; (B) geopolymer 0.0K, cured at 75 °C; (C) geopolymer 0.79K, cured at 30 °C; (D) geopolymer 0.79K, cured at 75 °C; (E) geopolymer 2.0K, cured at 75 °C; (F) geopolymer 2.0K at higher magnification.

taken across the corresponding gel matrixes to obtain a representation of the overall geopolymer gel composition. The spot size for the analyses was normally 2-4 µm, but it could be greater than 4 µm, depending on the scale of the microstructure. The powder samples were also subjected to transmission electron microscopy (TEM) using a Philips CM120 BioTWIN TEM instrument at 100 kV. Results and Discussion Microstructure and Elemental Composition of Fly-AshBased Geopolymer Gels. The microstructure of fly-ash-based geopolymers is presented as a series of electron micrographs in Figure 1A-F, showing the coexistence of the geopolymer gel and partially or completely unreacted fly ash particles. Figure 1A displays the original GFA particles, showing variations in size and composition. The geopolymer 0.0K, which is synthesized with KOH solution (H2O/K2O ) 14.85) at 75 °C (Figure 1B), presents a typical microstructure of a nonreactive system

in which geopolymer gel is rarely formed. Instead, the dissolved ion precursors, such as silicon, aluminum, and traces of calcium and iron, precipitate on the surface of the spherical fly ash particles. All unreacted fly ash particles are loosely connected by “necks”, which are believed to be precipitates of dissolved species. If any geopolymer gel is formed, the gel growth would concentrate on the surface of the spherical fly ash particles, which function as the initial nucleation points.24 Micrographs C and D are of geopolymer 0.79K cured at 30 and 75 °C, respectively. They reveal gel morphologies different from that of geopolymer 0.0K, whereby a greater degree of reaction has taken place and the gaps between fly ash particles have been filled with the formed geopolymer gel. However, geopolymer 0.79K, when cured at 30 °C, shows more porous spaces localized on the interface between the fly ash particles and the formed gel. This is not observed in geopolymer 0.79K cured at 75 °C. Instead, the latter shows a denser microstructure as a result of a greater degree of gel formation. Figure 1E displays a geopolymer synthesized with a solution of a higher silicate

3562

Ind. Eng. Chem. Res., Vol. 45, No. 10, 2006

Figure 2. Transmission electron micrographs of various fly-ash-based geopolymers: (A) geopolymer 2.0K, cured at 75 °C; (B) geopolymer 0.79K, cured at 75 °C; (C) geopolymer 2.0K, cured at 75 °C (thinner section and higher magnification compared to Figure 2A).

concentration (SiO2/M2O ) 2.0). At a higher magnification (Figure 1F), the fly-ash-based geopolymer gel resembles aluminosilicate particulates that are connected and form nanochannels and pores, as has been reported elsewhere.25 The size of the aluminosilicate particulates in a geopolymer gel determines the pore parameters observed. An array of larger particles results in a larger pore area and mean pore width. The opposite is true for a gel network consisting of smaller particles. TEM results show a dense gel (darker, at the top) and another porous gel (lighter, at the bottom) of geopolymer 2.0K cured at 75 °C (Figure 2A). The porous gel clearly presents the typical pore structure of a geopolymer gel. The pore microstructures of geopolymers 0.79K and 2.0K at a higher magnification are also presented in images B and C, respectively, of Figure 2. Similarly, both systems reveal nonspherical particles that are homogeneous in size, with a nanoscale characteristic. Electron diffraction from both samples gives patterns of broad circular rings without traces of bright speckles around the rings. This indicates a short-range ordering between neighboring atoms, which is the characteristic of an amorphous-like material. The distinct pore structure of the two systems can be clearly seen. Geopolymer 0.79K, which was synthesized at a lower silicate ratio, contains larger particles than geopolymer 2.0K, which was synthesized at a higher silicate ratio. The growth of controlled monodisperse silica particles was observed initially by Stober

and Fink.26 Presumably, the silica particles stop growing after reaching a critical size because they are electrostatically stabilized (mutually repulsive).27 Stabilization prevents particles from condensing with other particles, but it does not preclude continued condensation between particles and monomeric silicate species. The formation of a uniform pore structure in aluminosilicate particles through sol-gel chemistry is also commonly observed.28 The particle size can then be easily modified by changing the reaction temperature, curing time, and silicate ratio, as has also been observed in fly-ash-based geopolymer gels. The microstructural observations reveal that there are phase inhomogeneities showing regions of high silicon concentration, indicated by X in Figure 1D and E. According to the elemental analysis, the average Si/Al ratio of geopolymeric gel is ∼2 (Figure 1E), whereas the region marked by X was found to have a significantly high silica content (Si/Al ≈ 4-5). Regions of high silica content are frequently found in fly-ash-based geopolymers that are synthesized at higher silicate concentrations. They are easily spotted because of their microstructure, which is distinctively less porous than that of a typical geopolymer gel. It is important to note that, although this finding applies to fly-ash-based geopolymers, more reactive geopolymers, such as those based on metakaolin, do not show much variation in gel composition.29 Therefore, a more accurate and

Ind. Eng. Chem. Res., Vol. 45, No. 10, 2006 3563 Table 3. Si/Al Ratios of the Gels of Various Fly-Ash-Based Geopolymers Cured at Elevated Temperatures sample 0.0K 0.79K

1.4K 2.0K 0.0Na 0.79Na

1.4Na 2.0Na

temperature (°C)

curing time (h)

Si/Al molar ratio

75 30 50 50 75 50 75 50 75 75 30 50 50 75 50 75 50 75

24 24 24 48 24 24 24 24 24 24 24 24 48 24 24 24 24 24

1.51 ( 0.34 1.62 ( 0.16 1.66 ( 0.12 1.64 ( 0.11 1.75 ( 0.15 1.63 ( 0.20 1.97 ( 0.25 2.14 ( 0.19 2.06 ( 0.14 1.53 ( 0.07 1.61 ( 0.16 1.78 ( 0.19 1.71 ( 0.12 1.60 ( 0.12 2.24 ( 0.32 1.79 ( 0.30 1.90 ( 0.15 2.03 ( 0.21

advanced technique, such as the magic-angle-spinning (MAS) NMR spectroscopy, which uses a smaller sampling volume, might not be suitable for the characterization of fly-ash-based geopolymers if phase separation is a problem. The phase inhomogeneities in the gel are likely caused by the different release rates of silicon and aluminum from fly ash particles. The rates are affected by the reaction temperature, silicate concentration, and local variations in fly ash composition, as shown in Table 3. It is likely that homogenization cannot be achieved before complete dissolution takes place. At the onset of gelation (polycondensation), the viscosity increases rapidly; hence, the gel formed will further hinder homogenization. Thus, species coming into solution after gelation will react locally, without much possibility for homogenization. The gel composition is presented as the Si/Al ratio in Table 3. The elemental analysis took the variations in composition due to the regions of high silica concentration into consideration by taking average readings at various spots in the microstructure. Overall, the geopolymer gel gives a Si/Al molar ratio ranging from 1.51 to 2.24. Davidovits1 classified this Si-O-Alstructure in geopolymer materials as a polysialate with Si/Al composition of 1.0-2.0. Minor elements such as calcium and iron constitute 2-4 mol % of the total composition and are found to be homogeneously distributed across the gel matrix. Calcium- and iron-associated glassy phases are likely to dissolve at high pH,17 which means that these phases are likely to participate in the overall reaction. However, the effect of these cations on geopolymerization has not yet been fully elucidated and is assumed herein to be minimal.17,19 Rahier et al.29 showed through 27Al NMR studies that, with increasing silicate ratio (SiO2/M2O ) 0.0-1.9), the average Al substitution around Si in a geopolymer gel decreases. This observation is in accordance with the EDAX results, which show increasing Si/Al ratios as geopolymers are synthesized with solutions of higher silicate ratios (Table 3). There are some exceptions, especially at higher silicate concentrations, where the increasing Si/Al ratio trend is not fully followed. As the temperature is increased, the overall rate of reaction is also increased. As a result, a greater Si/Al ratio in the gel is generally observed, which is caused by the different dissolution rates of precursors from fly ash.17 The aluminum dissolves at a higher rate than silicon initially, and then the rates are decreased when the system approaches a supersaturated condition, i.e., when polycondensation is predominantly taking place. Although the dissolution of silicon is also slowed, it is still occurring at a

Figure 3. Experimental and theoretical Si/Al ratios for the gels from flyash-based geopolymers, cured at 75 °C. Table 4. Deviation of the Experimental Si/Al Ratio from Its Theoretical Ratio deviation (%) theoretical Si/Al

K geopolymer

Na geopolymer

1.95 2.21 2.42 2.62 average

22 21 19 21 21

22 28 26 23 25

significantly higher rate than the dissolution of the other elements, which results in a higher Si/Al ratio. The correlation between experimental and theoretical Si/Al ratios for geopolymers cured at 75 °C is shown in Figure 3, and the percentage deviation of the experimental Si/Al ratio from its theoretical value was calculated using eq 1 and is presented in Table 4.

deviation )

(Si/Al)theoretical - (Si/Al)experimental (Si/Al)theoretical

× 100% (1)

The experimental Si/Al ratios are consistently lower than the theoretical ratios at any silicate concentration. There are two possible reasons for this: (1) the actual Si/Al ratios of the amorphous phases are less than estimated and/or (2) the amorphous silica and alumina phases from fly ash are not fully dissolved. Although this issue is not investigated here, Ferna´ndez-Jime´nez et al.30 showed that the reactive alumina content plays a dominant role in the kinetics of gel formation. As a result, a complete reaction might never have taken place, even at higher temperatures, where the rate and extent of reaction are supposed to be higher. This means that there is a direct correlation between the extent of dissolution and the observed Si/Al ratio. The average deviations of K-based geopolymers and Na-based geopolymers are 21% and 25%, respectively. Although it is not feasible to assume that all dissolved precursors will polymerize to form aluminosilicate gel, the deviation still partly reflects the reactivities of geopolymers of different alkali cations. It is speculated that K+-containing solutions are more efficient in either the dissolution or/and polycondensation than Na+containing solutions are, because the Si/Al compositions of K-based geopolymers are closer to the predicted ratios. Pore Structure of Fly-Ash-Based Geopolymers. The volumes of Hg intruded into fly ash particles and fly-ash-based geopolymers are presented in Figure 4. The intrusion loop from

3564

Ind. Eng. Chem. Res., Vol. 45, No. 10, 2006

Figure 4. Cumulative intruded Hg volume of original Gladstone fly ash and various fly-ash-based geopolymers. 0K-75 refers to geopolymer 0.0K cured at 75 °C, while 0.79K-30, 0.79K-50, and 0.79K-75 refer to geopolymer 0.79K cured at 30, 50, and 75 °C, respectively.

Figure 5. Pore volume distributions of respective fly-ash-based geopolymers.

original fly ash reveals a “breakthrough pressure” starting from 80 psia. This indicates the pressure at which liquid Hg starts filling the interstitial spaces of the fly ash particles (interparticle pores). After the gaps are fully filled, the intruded volume ceases to increase, which shows that fly ash particles do not have any accessible pores. It can be seen that geopolymer 0.0K, which was cured at 75 °C, and geopolymer 0.79K, cured at 30 °C, have similar breakthrough pressures, but the intrusion curve is shifted slightly to higher pressure regions for the former. Geopolymers that are cured at higher temperatures, i.e., 50 and 75 °C, resulted in intrusion in higher pressure regions with greater cumulative pore volume. The intruded pressure is related to the equivalent pore size by the Washburn equation;21 hence, the pore size distribution can be plotted from the above data (Figure 5). Fly ash particles reveal a narrow pore size distribution, only within 2-4 µm. Geopolymer 0.0K displays a wider distribution with a greater total pore volume than the fly ash particles. Adding soluble silicate to the activating solution results in a smaller pore size (geopolymer 0.79K). Increasing temperature also results in a smaller pore size, with bimodal and wide pore size distributions shown at 50 and 75 °C. It is possible to designate three distinct regions of pores in geopolymer materials derived from fly ash on the basis of the pore size distribution. The following hierarchical pore structure can be used: mesopores (3.6-50 nm), macropores (50-200 nm), and pores larger than 200 nm. This classification is slightly adjusted from the IUPAC definition31 for the sake of analysis. The demarcation of meso- and macropores (50 and 200 nm) is arbitrary, yet it is also derived

Figure 6. Total pore volume distributions based on MIP of various flyash-based geopolymers, cured at 30, 50, and 75 °C.

Figure 7. Pore surface area distributions based on MIP of various flyash-based geopolymers, cured at 30, 50, and 75 °C.

from the results presented in Figure 5. Each designated pore size range is believed to originate from a different pore formation mechanism. The mesopores represent the main porous structure, which is the cluster(s) of aluminosilicate particles that constitute the Si-O-Al- network (polysialate),1 as was shown by the TEM results. The macropores, as defined (50-200 nm), are the larger pores that are formed during the early stage of reaction. As gel continues to grow, larger pores are being filled, and longer chains and branches are being formed; mesopores are thus obtained, and the macropore volume decreases. Macropores are apparent in fly-ash-based geopolymers that are cured at medium temperature (50 °C). Pore sizes larger than 200 nm are present in geopolymers formed in mixtures with low reactivity, such as (i) geopolymers that are synthesized only with alkali solutions (geopolymers 0.0K and 0.0Na) and (ii) geopolymers that are cured at low temperature (30 °C). Comparison between MIP and BJH Methods. More geopolymers than listed in Table 2 were synthesized and intruded with mercury. The pore volume distributions for the three designated pore size regions are presented in Figure 6, and the associated surface area distributions are presented in Figure 7. According to the MIP method, it is clear that the pore structure of well-cured fly-ash-based geopolymers is mainly mesoporous (Figure 6). N2 adsorption is capable of measuring pores smaller than 3.6 nm, which are commonly referred to as micropores. However, it was found that the volume and area of micropores in fly-ash-based geopolymers prepared at any reaction conditions are relatively low. To get a fair comparison between the MIP and BJH methods, only the mesopore parameters are presented, with the mesopore volume (Vp) in

Ind. Eng. Chem. Res., Vol. 45, No. 10, 2006 3565 Table 5. Apparent Setting Times, ts, of Various Fly-Ash-Based Geopolymers, Cured at 30, 50, and 75 °C curing time (min) sample

30 °C

50 °C

75 °C

0.0K 0.79K 1.4K 2.0K 0.0Na 0.79Na 1.4Na 2.0Na

240-250a ∼40 ∼44 ∼40 >1440 ∼45 ∼72 ∼68

720-780a ∼18 ∼18 ∼21 180-192a ∼16 ∼24 ∼25

125-130 ∼17 ∼18 ∼20 67-71 ∼14 ∼22 ∼24

a Recorded t is the time when geopolymeric paste stops flowing. The s samples, however, are weak because of a lack of gel content.

Figure 8. Effect of elevated curing temperature (50 and 75 °C) on the total mesopore volume (Vp) of fly-ash-based geopolymers.

Figure 9. Effect of elevated curing temperature (50 and 75 °C) on the mesopore surface area (Sp) of fly-ash-based geopolymers.

Figure 8 and the mesopore area (Sp) in Figure 9. All samples that were cured at 30 °C, as well as geopolymers 0.0K and 0.0Na, which were cured at 50 °C, are not presented, as the mesopore volumes in these samples are negligible. It can be seen that the Vp and Sp values from the MIP method are consistently greater than the Vp and Sp values obtained from the BJH method (Figures 8 and 9). For example, Vp and Sp of geopolymer 0.79K, which was cured at 75 °C, are 0.212 cm3/g and 60.9 m2/g, respectively, from the MIP measurements, but are only 0.131 cm3/g and 55.1 m2/g, respectively, from the BJH method. These quantitative discrepancies are also apparent in Na-based geopolymers and vary depending on the SiO2/M2O ratios used. Several reasons can explain the differences in pore parameters obtained using the two techniques. The sizes and adsorptive behaviors of the two adsorbents (Hg and N2) are significantly different. The physical adsorption of a nonwetting liquid (Hg) on a solid surface is very different from N2 adsorption, which involves a complex mechanism between gas and liquid condensation on the solid surface. It is also likely that the “true” pore structure of the geopolymers affects the results obtained. As described earlier, an interconnected tubular pore model must be assumed to derive the V and S values. The real pore network and shape are undetermined from these two methods alone. Diamond32 reported that, in concrete, long percolative chains of intermediate pores with varying sizes and shapes, leading to larger pores, will give false volumes and areas by mercury porosimetry. This phenomenon might apply to any porous aluminosilicate materials, including geopolymers. The more porous the geopolymer gel is, the more complex the pore structure is.

There are several other techniques that can be used to complement and provide a better understanding of the pore structure of geopolymers. The techniques that have been applied in the pore structure analysis of porous aluminosilicates include quantitative analysis by transmission electron microscopy (TEM)33 and small-angle neutron/X-ray scattering (SANS and SAXS).33 Despite the apparent differences in Vp and Sp values from the two methods, the trends in pore structure evolution with respect to the variations in curing temperature and SiO2/ M2O ratio agree with each other quite well. The MIP and BJH methods cannot give a precisely quantitative determination of the pore parameters unless an accurate pore network model is applied. Yet, the effect of various reaction conditions on the pore structure of geopolymers can still be studied using these two techniques. Effect of Curing Temperature on Fly-Ash-Based Geopolymerization. Generally, there is an increase in total pore volume (V) and surface area (S) as the curing temperature is elevated. Also, a greater degree of reaction is indicated, particularly by the increase in mesopore volume and area as the curing temperature is raised (Figures 6 and 7). The geopolymers that were cured at 30 °C and those synthesized with alkali solution only (geopolymers 0.0K and 0.0Na) are very porous. Yet, a low level of mesopores is observed. As the reaction temperature is raised from 30 to 50 and 75 °C, the Vp and Sp values of the fly-ash-based geopolymers increase significantly (Figures 6 and 7). This trend applies to both Kand Na-based geopolymers at any SiO2/M2O ratio. The extent of the increase in Vp and Sp, however, varies depending on the silicate ratio used. Elevating the reaction temperature is known to increase the extent of dissolution of precursors (primarily Al and Si) from the glassy phases in fly ash.17 Increasing the amounts of precursors leads to an increase in nucleation rates, and thus polycondensation, as it can be seen that the apparent setting times (ts) of geopolymers reacted at higher temperatures (50 and 75 °C) are generally shorter than those of geopolymers reacted at lower temperature (30 °C, Table 5), except for geopolymer 0.0K, which shows an unexpected gelation behavior. The ts value of geopolymer 0.0K increases when the temperature is raised from 30 to 50 °C and decreases again when the temperature is increased further to 75 °C. It is proposed that the anomalous gelation behavior of K-based geopolymer at high alkalinity results from a reduction in the number of nuclei at 50 °C because of increased solubility, causing the reaction to proceed longer.35 At 30 °C, the solubility is low, so that setting is faster because of precipitation of dissolved species, rather than polycondensation of monomeric silicate with aluminate. Therefore, although the samples appear to have set, they are also weak because of insufficient geopolymer gel, which functions as a binder to existing fly ash particles. At the higher temperature of 75 °C, the increased rate of gel growth might

3566

Ind. Eng. Chem. Res., Vol. 45, No. 10, 2006

compensate for the reduction in number of nuclei, thus causing gelation times to decrease again. The “unexpected” gelation behavior of particular K+ solutions was also noted by Dent Glasser and Harvey.36 It was mentioned earlier that the final Si/Al ratio correlates with the rate of dissolution of silicon and aluminum from the fly ash particles. Increasing the curing temperature of geopolymer 0.79K from 50 to 75 °C altered the Si/Al ratio from 1.66 to 1.75 (Table 3). However, increasing the curing time from 24 to 48 h while maintaining the temperature at 50 °C changed the Si/Al ratio only slightly from 1.66 to 1.64. The same observation is also valid for geopolymer 0.79Na when the curing time was extended from 24 to 48 h at 50 °C. When the temperature for geopolymer 0.79Na was raised from 50 to 75 °C at 24 h, the Si/Al ratio decreased from 1.78 to 1.60, which appears to be anomalous behavior. The lower mesopore volume of 0.79Na at 75 °C compared to 50 °C (Figure 8) seems to confirm the lower Si/Al ratio, suggesting lower reactivity at 75 °C. The prolonged curing time beyond 24 h does not seem to change the Si/Al ratio significantly, although a higher reactivity might be obtained. This leads to the understanding that there is an instance in the course of geopolymerization when the temperature-controlled kinetics is inhibited. Before ts, significant dissolution usually occurs because of the existing alkaline and silicate species in the solution, which are needed for nucleation and gel growth to proceed.15 Toward the point of supersaturation, however, dissolution is replaced by the predominant polycondensation process, which results in water being expelled from the solidified gel matrixes. (Although it is believed that the dissolution and polycondensation processes occur simultaneously, they are separated conceptually here.) In this instance, some of the alkali cations are bonded into the aluminosilicate gel network, which leads to a decrease of the alkali concentration in the pore solution. The reaction kinetics is thus slowed. Further reactions beyond ts are likely to be controlled by the diffusion of remaining ion precursors to reach equilibrium with partly reacted fly ash, whereby dissolution and condensation might still take place, but to a lower extent and at a lower rate. It appears that the pore structure of K-based geopolymers is more susceptible to variations in curing temperature. The Vp and Sp values of K-based geopolymers increase systematically as the reaction temperature is raised from 30 to 50 and 75 °C (Figures 6 and 7). On the other hand, for Na-based geopolymers, there is a significant increase in both Vp and Sp from 30 to 50 °C, but a plateau from 50 to 75 °C. The same trend applies to geopolymers synthesized at all SiO2/M2O ratios. It has been mentioned that the effect of temperature on the kinetics of different geopolymers can be elucidated from the setting behaviors of different systems. However, the ts values of both K-based and Na-based geopolymers decreased only slightly as the temperature was raised from 50 to 75 °C (Table 5), despite the significant difference in the pore structures of the two types of geopolymers. It is believed that a large decrease in ts was not observed because of the highly concentrated sol-gel system. Effect of SiO2/M2O Ratio on Fly-Ash-Based Geopolymerization. The curing temperature has been shown to have a substantial effect on fly-ash-based geopolymerization by increasing both the dissolution and polycondensation rates at high temperature (75 °C). It is also clear that the effect of temperature on the kinetics of reaction is related to the silicate ratio used in the solution. The effect of the SiO2/M2O ratio on fly ash geopolymerization can be understood by considering two aspects: (1) the alkali cation interaction with the silicate (and

aluminate) anions in the solution and (2) the distribution of the respective aluminosilicate anions at different SiO2/M2O ratios. The Vp and Sp values of geopolymers synthesized using a SiO2/M2O ratio of 0.0 are significantly lower than those of geopolymers synthesized using higher SiO2/M2O ratios (Figures 8 and 9). The extents of dissolution of aluminum and silica in alkali-containing solution are known to be less than those in solutions containing soluble silicate, because depolymerization of Si-O-Al and Si-O-Si bonds from the solid surface is enhanced in the presence of more soluble silicate species.16 A lower extent of dissolution leads to a decrease in nucleation rate. Hence, geopolymers 0.0K and 0.0Na, cured at low temperatures (30 and 50 °C), hardly set, even after 24 h, and the geopolymers that were cured at 75 °C took longer to set than the geopolymers at higher silicate ratios. At SiO2/M2O ) 0.0, the Vp and Sp values of Na-based geopolymer are significantly greater than those of K-based geopolymers (Figures 8 and 9). It is likely that the significantly higher porosity in geopolymer 0.0Na is caused by the presence of zeolites (the characterization of zeolites is not presented in the current work). Other authors37-41 have reported the formation of zeolitic phases, namely, faujasite, hydroxysodalite, and NaA, upon activation of class F fly ash with only NaOH and/or KOH solution (no soluble silicate). To form zeolitic crystals from fly ash, a significant hydrolysis of glassy phases needs to be achieved, which usually involves reactions at high solutionto-solid ratios and elevated temperatures, at least above 50 °C, preferably in NaOH solutions. A slight addition of soluble silicate results in a more mesoporous structure, which means that more geopolymer gel is being formed (Figures 8 and 9). For example, the Vp and Sp values of geopolymer 0.79K are larger than those of geopolymer 0.0K. This is also the case for their Na counterpart, although the difference is not as strikingly large as for the K-based geopolymers. A slightly higher soluble silicate concentration in the solution accelerates the reaction of K-based geopolymerization significantly, as can be seen from the reduction of ts to ∼17 min. In Na-based geopolymerization, the kinetics of the reaction is accelerated slightly more, with the ts being shortened to ∼14 min at 75 °C. Although curing at higher temperature causes faster dissolution, the faster gel growth resulting from the coagulating ability of Na+ with monomeric silicate42,43 will cause faster setting. This is not the case for K-based geopolymers, which will not induce the gelation step to proceed unless there is a sufficient supply of nutrients (larger aluminosilicate anions) for gel growth.43 At 75 °C, as the silicate ratio is increased from 0.79 to 2.0, there is a general increase in Vp, Sp, and ts, with Vp and Sp showing optima at SiO2/M2O ) 1.4 (Figures 8 and 9, and Table 5). This means that addition of soluble silicate generally enhances geopolymerization, but an overly high silicate concentration (SiO2/M2O g 2) is likely to reduce reactivity. It is speculated that, in a very high silicate solution, the high concentration of cyclic silicate species inhibits further condensation. This can be seen from the delayed setting time as the silicate ratio is increased from 0.79 to 2. In condensation, only the uncomplexed tetrahedral aluminate Al(OH)4- reacts with monomeric Si(OH)4 or larger linear silicate anions.44 Linear silicate species can easily be depolymerized as the bulk conditions are changed, allowing polymerization to take place easily through the polycondensation of silicate and aluminate monomers. On the other hand, the large cyclic structures, such as cage or ring silicate species, slow the kinetics and only serve as reservoirs for small, acyclic species that are more responsible for gel growth.45

Ind. Eng. Chem. Res., Vol. 45, No. 10, 2006 3567

An anomaly is observed in the pore structure of the K-based geopolymers as the silicate ratio in the solution changes from 0.79 to 1.4 and 2. Geopolymer 0.79K shows a larger Vp but a smaller Sp than geopolymers 1.4K and 2.0K. This is not observed in Na-based geopolymers (Figures 8 and 9). From TEM and MIP results, it can be established that geopolymer 0.79K comprises aluminosilicate particles that are larger than those in geopolymer 2.0K. The former might have a larger porosity (greater pore volume), but a gel network consisting of larger particles leads to a significant decrease in pore surface area. Furthermore, the Vp values of the K-based geopolymers at SiO2/M2O ) 1.4 and 2.0 are only slightly larger than those of their Na counterparts when both samples are cured at 75 °C (Figure 8). Yet, the Sp values of K-based geopolymers are strikingly larger (Figure 9). Larger Vp and Sp values can be associated with a greater extent of reaction because of the increasing amount of porous gel in the matrix. A larger Sp value can also be contributed from a more uniform, smaller particle array that constitutes an extended pore network. In this instance, it is not yet feasible to establish that K-based geopolymers are more reactive than Na-based geopolymers. It was speculated earlier that K-based geopolymers are more efficient in the dissolution step because the Si/Al ratios of K-based geopolymers are closer to their predicted values. Wijnen et al.42 suggested that the larger hydration shell of K+ results in a weaker interaction between K+ and water molecules, thus giving a better dissolution, which is essential for nucleation to proceed. Numerous authors43,46 have suggested that the larger K+ cation size contributes to the increase in the number of Al-O-Si bonds because of its association with larger aluminosilicate anions, which results in extended gel growth. These two factors seem to explain the differences in pore structure between Na and K geopolymers. Nonetheless, the effect of cation size on the mechanism of formation of aluminosilicate gels, such as geopolymer gels, is still ambiguous.45,47 The increase in soluble cations is definitely known to affect the distribution of anionic species and the structure of porous aluminosilicate by increasing the depolymerization of larger silicate anions, thus enhancing nucleation and gel growth. Yet, the effects of cation size are complex and do not correspond to simple extension of the reasoning for alkali metal cations; rather, they are interrelated with various variables, such as the silicate and alkali concentrations. Therefore, a definitive assessment of the distinct pore structure characteristics of K-based and Na-based geopolymers is not yet feasible based on the current results. Conclusions The pore microstructure of geopolymers derived from class F fly ash can be studied qualitatively using MIP, N2 adsorption techniques, and electron microscopy. The microstructural observations (using FESEM and TEM) reveal that fly-ash-based geopolymer is an amorphous material, with nanosize pore characteristics. The EDAX compositional analysis revealed that fly-ash-based geopolymers are aluminosilicate gels with Si/Al ratios ranging from 1.51 to 2.24. A well-reacted fly-ash-based geopolymer shows a mesoporous structure (3.6-50 nm) that develops with increasing curing temperature and silicate ratio (SiO2/M2O ) 0.0-2.0, where M ) Na or K). An increase in curing temperature from 30 to 75 °C leads to a greater extent and higher rate of reaction, but the temperature-controlled kinetics is likely to be inhibited after the time of setting. Increasing the silicate ratio (SiO2/M2O ) 0.0-2.0) also increases the degree of reaction, although a very high silicate concentration (SiO2/M2O g 2.0) proved to be detrimental and to reduce

reactivity. The pore structure of K-based geopolymer is more susceptible to a change in temperature than is that of Na-based geopolymer. Yet, a definitive assessment of the effect of each cation on fly ash geopolymerization is not yet feasible based on the current results. Acknowledgment Financial support from the Australian Research Council (ARC) and the Particulate Fluids Processing Centre (PFPC), a Special Research Centre of the ARC, is gratefully acknowledged. Rachel Caruso is thanked for the access to the Tristar instrument, and Janine Hulston is thanked for valuable discussions on MIP. Literature Cited (1) Davidovits, J. Chemistry of Geopolymeric Systems, Terminology. In Geopolymer ‘99 Proceedings; Saint-Quentin, France, 1999. (2) Xu, H.; van Deventer, J. S. J. The Geopolymerisation of AluminoSilicate Minerals. Int. J. Min. Process. 2000, 59, 247. (3) Lee, W. K. W.; van Deventer, J. S. J. Structural Reorganisation of Class F Fly Ash in Alkaline Silicate Solutions. Colloids Surf. A: Physicochem. Eng. Aspects 2002, 211, 49. (4) Palomo, A.; Glasser, F. P. Chemically-Bonded Cementitious Materials Based on Metakaolin. Br. Ceram. Trans. J. 1992, 91, 107. (5) Barbosa, V. F. F.; Mackenzie, K. J. D.; Thaumaturgo, C. Synthesis and Characterisation of Materials Based on Inorganic Polymers of Alumina and Silica: Sodium Polysialate Polymers. Int. J. Inorg. Mater. 2000, 2, 309. (6) Palomo, A.; Grutzeck, M. W.; Blanco, M. T. Alkali-Activated Fly Ashes: A Cement for the Future. Cem. Concr. Res. 1999, 29, 1323. (7) Cheng, T. W.; Chiu, J. P. Fire-Resistant Geopolymer Produced by Granulated Blast Furnace Slag. Min. Eng. 2003, 16, 205. (8) van Jaarsveld, J. G. S.; van Deventer, J. S. J.; Lorenzen, L. The Potential Use of Geopolymeric Materials to Immobilise Toxic Metals: Part I. Theory and Applications. Min. Eng. 1997, 10, 659. (9) Gartner, E. Industrially interesting approaches to “low CO2” cements. Cem. Concr. Res. 2004, 34, 1489. (10) Dhir, R. K.; Byars, E. A. Pulverized Fuel-Ash ConcretesIntrinsic Permeability. Aci Mater. J. 1993, 90, 571. (11) Chang, H. L.; Chun, C. M.; Aksay, I. A.; Shih, W. H. Conversion of Fly Ash into Mesoporous Aluminosilicate. Ind. Eng. Chem. Res. 1999, 38, 973. (12) Sindhunata; Xu, H.; Lukey, G.; van Deventer, J. S. J. The Effect of Curing Conditions on the Properties of Geopolymeric Materials Derived from Fly Ash. Presented at the International Symposium of Advances in Concrete through Science and Engineering, IL, Mar 21-24, 2004. (13) Nakanishi, K.; Takahashi, R.; Nagakane, T.; Kitayama, K.; Koheiya, N.; Shikata, H.; Soga, N. Formation of Hierarchical Pore Structure in Silica Gel. J. Sol-Gel Sci. Technol. 2000, 17, 191. (14) Witte, B. D.; Aernouts, K.; Uytterhoeven, J. B. Aging of Aluminosilicate and Silica Gels in Aqeous Solutions of Various Ph and Al Content, a Textural and Structural Evaluation. Microporous Mater. 1996, 7, 97. (15) Hench, L. L.; West, J. K. The Sol-Gel Processes. Chem. ReV. 1990, 90, 33. (16) Iler, R. K. The Chemistry of Silica: Solubility, Polymerization, Colloid and Surface Properties, and Biochemistry; Wiley: New York, 1979. (17) Pietersen, H. S.; Fraay, A. L. A.; Bijen, J. M. In Fly Ash and Coal ConVersion By-products: Characterization, Utilization, and Disposal VI; Day, R. L., Ed.; MRS Proceedings 178; Materials Research Society: Warrendale, PA, 1988; p 139. (18) Hemmings, R. T.; Berry, E. E. In Fly Ash and Coal ConVersion By-products: Characterization, Utilization, and Disposal IV; McCarthy, G. J., Ed.; MRS Proceedings 113; Materials Research Society: Warrendale, PA, 1988; p 3. (19) Qian, J. C.; Lachowski, E. E.; Glasser, F. P. In Fly Ash and Coal ConVersion By-products: Characterization, Utilization, and Disposal IV; McCarthy, G. J., Glasser, F. P., Roy, D. M., Hemmings, R. T., Eds.; MRS Proceedings 113; Materials Research Society: Warrendale, PA, 1988; p 45. (20) Webb, P. A.; Orr, C. Analytical Methods in Fine Particle Technology; Micromeritics Instrument Corporation: Norcross, GA, 1997. (21) Washburn, E. W. Note on a Method of Determining the Distribution of Pore Sizes in Porous Materials. Proc. Nat. Acad. Sci. U.S.A. 1921, 7, 115.

3568

Ind. Eng. Chem. Res., Vol. 45, No. 10, 2006

(22) Gregg, S. J.; Sing, K. S. W. Adsorption, Surface Area, and Porosity; Academic Press: London, 1982. (23) Barret, E. P.; Joyner, L. G.; Halenda, P. P. The Determination of Pore Volume and Area Distributions in Porous Substances. I. Computations from Nitrogen Isotherms. J. Am. Chem. Soc. 1951, 73, 373. (24) McCormick, A. V.; Bell, A. T. The Solution Chemistry of Zeolite Precursors. Catal. ReV.-Sci. Eng. 1989, 31, 97. (25) Kriven, W. M.; Bell, J. L.; Gordon, M. Microstructure and Microchemistry of Fully-Reacted Geopolymers and Geopolymer Matrix Composites. Ceram. Trans. 2003, 153, 227. (26) Stober, W.; Fink, A. Controlled Growth of Monodisperse Silica Spheres in the Micron Size Range. J. Colloid Interface Sci. 1968, 26, 62. (27) Brinker, C. J. Hydrolysis and Condensation of Silicates: Effects on Structure. J. Non-Cryst. Solids 1988, 100, 31. (28) Setzer, C.; van Essche, G.; Pryor, N. In Handbook of Porous Solids; Schuth, F., Sing, K. S. W., Weitkamp, J., Eds.; Wiley-VCH: Weinheim, Germany, 2002; Vol. 3, p 1543. (29) Rahier, H.; Simons, W.; Mele, B. V.; Biesemans, M. LowTemperature Synthesized Aluminosilicate Glasses. Part III. Influence of the Composition of the Silicate Solution on Production, Structure and Properties. J. Mater. Sci. 1997, 32, 2237. (30) Ferna´ndez-Jime´nez, A.; Palomo, A.; Sobrados, I.; Sanz, J. The Role Played by the Reactive Alumina Content in the Alkaline Activation of Fly Ashes. Microporous Mesoporous Mater. 2006, 91, 111. (31) Giesche, H. In Handbook of Porous Solids; Schuth, F., Sing, K. S. W., Weitkamp, J., Eds.; Wiley-VCH: Weinheim, Germany, 2002; Vol. 1, p 1000. (32) Diamond, S. Mercury Porosimetry: An Inappropriate Method for the Measurement of Pore Size Distributions in Cement-Based Materials. Cem. Concr. Res. 2000, 30, 1517. (33) Kerch, H. M.; Gerhardt, R. A.; Grazul, J. L. Quantitative Electron Microscopic Investigation of the Pore Structure in 10:90 Colloidal Silica/ Potassium Silicate Sol-Gels. J. Am. Ceram. Soc. 1990, 73, 2228. (34) Ramsay, J. D. F. In Handbook of Porous Solids; Schuth, F., Sing, K. S. W., Weitkamp, J., Eds.; Wiley-VCH: Weinheim, Germany, 2002; Vol. 1, p 135. (35) North, M. R.; Fleischer, M. A.; Swaddle, T. W. Precipitation from Alkaline Aqueous Aluminosilicate Solutions. Can. J. Chem.-ReV. Can. Chim. 2001, 79, 75. (36) Dent Glasser, L. S. The Unexpected Behaviour of Potassium Aluminosilicate Solutions. J. Chem. Soc., Chem. Commun. 1984, 664.

(37) Inada, M.; Eguchi, Y.; Enomoto, N.; Hojo, J. Synthesis of Zeolite from Coal Fly Ashes with Different Silica-Alumina Composition. Fuel 2005, 84 (2-3), 299. (38) Murayama, N.; Yamamoto, H.; Shibata, J. Mechanism of Zeolite Synthesis from Coal Fly Ash by Alkali Hydrothermal Reaction. Int. J. Min. Process. 2002, 64, 1. (39) Querol, X.; Alastuey, A.; Fernandez-Turiel, J. L.; Lopez-Soler, A. Synthesis of Zeolites by Alkaline Activation of Ferro-Aluminous Fly Ash. Fuel 1995, 74, 1226. (40) Querol, X.; Alastuey, A.; Lopez-Soler, A.; Plana, F.; Andres, J. M.; Juan, R.; Ferrer, P.; Ruiz, C. R. A Fast Method for Recycling Fly Ash: Microwave-Assisted Zeolite Synthesis. EnViron. Sci. Technol. 1997, 31, 2527. (41) Zhao, X. S.; Lu, G. Q.; Zhu, H. Y. Effects of Ageing and Seeding on the Formation of Zeolite Y from Coal Fly Ash. J. Porous Mater. 1997, 4, 245. (42) Wijnen, P. W. J. G.; Beelen, T. P. M.; de Haan, J. W.; van de Ven, L. J. M.; van Santen, R. A. The Structure Directing Effect of Cations in Aqueous Silicate Solutions. A 29Si NMR Study. Colloids Surf. 1990, 45, 255. (43) Dent Glasser, L. S. The Gelation Behaviour of Aluminosilicate Solutions Containing Na+, K+, Cs+, Me4n+. J. Chem. Soc., Chem. Commun. 1984, 1250. (44) Harris, R. K.; Samadi-Maybodi, A.; Smith, W. The Incorporation of Aluminum into Silicate Ions in Alkaline Aqueous Solutions, Studied by Aluminum-27 NMR. Zeolites 1997, 19, 147. (45) Swaddle, T. W. Silicate Complexes of Aluminium(III) in Aqueous Systems. Coord. Chem. ReV. 2001, 219-221, 665. (46) McCormick, A. V.; Bell, A. T.; Radke, C. J. Multinuclear NMR Investigation of the Formation of Aluminosilicate Anions. J. Phys. Chem. 1989, 93, 1741. (47) Azizi, S. N.; Harris, R. K.; Samadi-Maybodi, A. Aluminium-27 NMR Investigation of the Influence of Cation Type on Aluminosilicate Solutions. Magn. Reson. Chem. 2002, 40, 635.

ReceiVed for reView November 11, 2005 ReVised manuscript receiVed March 19, 2006 Accepted March 20, 2006 IE051251P