SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. I. Introduction
  4. II. Experimental Procedure
  5. III. Results and Discussion
  6. IV. Conclusion
  7. Acknowledgments
  8. References

Geopolymers (GP) represent a fascinating alternative to typical structural ceramics, but all prior work has demonstrated substantial mesoporosity. This porosity is detrimental to mechanical properties as well as drying failure resistance. Here, mesoporosity in metakaolin Na-based GP has been reduced by an order of magnitude through the addition of acrylate-functional silane coupling agents, and the effect of this modification was extensively analyzed using a variety of characterization techniques. It was found that a critical concentration of approximately 0.06 mol methacryloxypropyltrimethoxysilane/mol GP was necessary for extensive property modification. Above this concentration, membranes and tubes of organic material could be observed, suggesting that only this surplus additive contributed to mesoporosity reduction. The source of this mesoporosity reduction is shown to be segregation of free water via hydrogel-like interaction caused by the formation of sodium acrylate groups.

I. Introduction

  1. Top of page
  2. Abstract
  3. I. Introduction
  4. II. Experimental Procedure
  5. III. Results and Discussion
  6. IV. Conclusion
  7. Acknowledgments
  8. References

An alkali-activated condensate aluminosilicate or geopolymer (GP),[1] as defined by Joseph Davidovits, has been intensively investigated as a structural ceramic due to its ability to cure at room temperature and the low cost of typical precursors.[2-4] Currently, the use of GPs is limited to applications using substantial filler because of the material's tendency to crack under various drying conditions. This failure is typically attributed to the formation of approximately 40 vol% porosity during the curing process[5-7] with much of that porosity being mesopores that form between precipitates.[8, 9]

Thus, a major weakness of GP gels remains in that the large amounts of water required for both curing[10] and rheological reasons[6] create mesoporous structures which are sensitive to hydration and dehydration and will crack and fail if fillers are not adequately dispersed in the GP matrix.[11, 12] A reduction in postcuring drying shrinkage would greatly expand the applications of the GP, especially in molded uses.

No reliable method of decreasing slurry viscosity of GPs for a given amount of water content is known in the literature, despite there being evidence that the amount of water typically used in GPs is excessive. Recent investigations show that shrinkage occurs when the total water in the system decreases to the point where the hydration of the alkali cations begins to decrease,[13] and that the majority of water used in GP synthesis exists merely as free water in the final product. This would be expected to have detrimental effects on mechanical properties.

It has been observed that water reducers designed for portland cement do not impart increased workability or final strength when added to a GP mixture. This is unsurprising, as the GP is seen to precipitate from monomers or short-chain oligomers, such that there is minimal solid binder to disperse.[14] Thus, any structural changes that could impart increased strength for a given workability must apply to the latter stages of precipitation.

One method by which these structural changes can be made is through the cocondensation of alkoxysilanes[15] within the GP precipitates. Functionalization on the alkoxysilanes can alter the GP network,[16] resulting in nanostructural changes. These nanostructural changes would be expected to have predictable results on the microstructure and properties of the material. In particular, it is reasonable to expect that the addition of acrylic functionalization within the GP might have long-range effects on the water transport and porosity, as the highly basic environment would create sodium acrylate functional groups within the structure. Such groups are typically used in polymer covalent networks to produce superabsorbing polymers, which deform in shape or “swell” to accommodate vast quantities of water within the covalent matrix.[17] Without a network, the radius of hydration is severely limited.[18] Solvent swelling has been observed in certain acrylate-functionalized alkoxysilanes networked onto a polystyrene particle, but in that case the system is obviously quite different.[19] In this case, the covalent network that creates the swelling volume would be supplied by the SiO4 and AlO4 tetrahedra amorphous network itself. Accordingly, the alkoxysilanes hydrolyze, yielding alcohols, and then either polycondense or condense onto the growing GP precipitates.

No previous investigation has considered the possibility of utilizing this swelling to accommodate the excess GP water. This investigation involves the synthesis of a variety of different sodium metakaolin GPs based on this premise, and their characterization.

II. Experimental Procedure

  1. Top of page
  2. Abstract
  3. I. Introduction
  4. II. Experimental Procedure
  5. III. Results and Discussion
  6. IV. Conclusion
  7. Acknowledgments
  8. References

(1) Synthesis

Geopolymers were synthesized from as-received metakaolin (Metamax HRM, BASF Corp., Florham Park, NJ) and sodium silicate solution. Sodium silicate solution (water glass) was prepared as follows: 160.0 g sodium hydroxide (Sigma-Aldrich, St. Louis, MO) was added to 360.0 g DI water in a stainless steel container and stirred using a polytetrafluoroethane (PTFE) stir bar with a magnetic stir plate until fully dissolved. The container was then sealed with plastic film, and the temperature of the stir plate was set to approximately 45°C. A total of 240.0 g fumed silica (Cab-o-sil; Cabot Corp., Boston, MA) was then added in small batches of approximately 10 g, each time allowing for complete or almost complete dissolution of the silica before adding the next batch, with the container remaining sealed when not adding silica. After all the silica was added, the container was resealed and the contents were allowed to stir at the elevated temperature (≈45°C) for 24 h. After 24 h, the solution of sodium silicate was poured into a polyethylene container, and allowed to equilibrate sealed at ambient conditions for at least 2 weeks.

Samples were then synthesized in accordance with previous work,[20] except with the addition of one or more silane coupling agents (Gelest, Morristown, PA), or other organic modifiers (Sigma-Aldrich Co., LLC) each used as received. The additives tested are listed in Table 1. Sample compositions consisted of 11.1 g metakaolin and 19.0 g (Formulation “A”) or 18.5 g (Formulation “B”) water glass, producing a nominal Na2O·Al2O3·4SiO2·11H2O) formulation (precise ratio for “A” 1.06·1·4.23·11.58, and for “B” 1.03·1·4.17·11.28 when impurities in the metakaolin are considered), plus the prescribed quantity of coupling agent. The reagents were mixed in a planetary centrifugal mixer (ARE-250; Thinky Corp., Tokyo, Japan) at 1000 rpm for 180 s, with a 60 s debubbling at 1200 rpm. The samples were then cured sealed at 25°C for at least 2 d, before being removed from their molds and stored in plastic zipper storage bags. The bags sat in ambient conditions for a week or more before tests were conducted.

Table 1. List of Additives and Their Abbreviations
AdditiveAbbreviationSummary of results
Acrylic acidAAPhase separated during mixing
Methacrylic acidMAAPhase separated during mixing
MethacryloxymethyltrimethoxysilaneMEMO-MeLimited microstructure modification
MethacryloxypropylmethyldimethoxysilaneMEMO-2,1Minimal observed effect
Methacryloxypropyltrimethoxysilane MEMOImportant microstructure modification
PhenylaminopropyltrimethoxysilanePAMSMinimal observed effect
Poly(acrylic acid)PAAPhase separated during mixing

(2) Macropore-Size Distribution

Macropore-size distribution was measured using mercury intrusion porosimetry (MIP). MIP was conducted using both traditional intrusion and modern cycling intrusion-extrusion methods,[21] and experiments were conducted using purpose-built equipment (Autopore II 9220; Micromeritics, Norcross, GA). Samples tested were cut with a diamond-edge wafer blade into cubes with 1 cm sides, and were degassed to 6.7 Pa prior to mercury filling at 3.5 kPa. Effective contact angle was determined using intrusion in slits cut with a diamond-edge wafer blade, and comparing the results to slit widths measured with an optical microscope.

(3) Mesopore-Size Distribution

Mesopore-size distribution was measured along with surface area using nitrogen adsorption. Gas physisorption tests were conducted using ultrahigh-purity (UHP) N2 and H2 in a commercial, purpose-built instrument (ASAP 2020; Micromeritics, Norcross, GA). Before analysis, samples were degassed using the included degassing system. All degassing was conducted at an evacuation rate of 670 Pa/s, with unrestricted evacuation beginning when a pressure of 670 Pa was reached. During this evacuation, the temperature was increased to 90°C. When the chamber pressure of 1.3 Pa was reached, the temperature was further increased to 200°C and the sample was degassed for 24 h, or longer if outgassing was still observed. In accordance with previous GP research,[22] some noteworthy results were verified with samples degassed at a maximum temperature of 50°C for the several days required, and the predictable small increase in adsorption was observed.

BET-specific surface area (SSA)[23] analysis was conducted using a 9-point fit of evenly spaced pressures between 0.075 and 0.275 P/P0, inclusive. Nonlocal density functional theory (NLDFT) pore size distribution analysis was conducted over the range 0.001–0.9 P/P0, inclusive, using the Jaroniec cylindrical model[24] for zeolites and silica. This range corresponds to pores of sizes 1.3–25 nm. Fitting was conducted using ridge regression, with a regularization number rr of 0.0316.[25] Comparisons showed that the samples were not unusually sensitive to the choice of assumed pore shape model or regularization number. The regularization and fit is achieved by finding the pore distribution vector inline image that minimizes the residual [Eq. (1)], where qe is a vector of the experimental isotherm as a function of relative pressure P/P0 and Qm is a matrix of model isotherms of individual pores of various sizes, each described as a function of relative pressure.

  • display math(1)

For pore sizes from ≈20 to 100 nm, the BJH method[26] of gas adsorption porosimetry analysis is considered to describe an approximation of the pore size distribution.

(4) Chemical and Reactivity Analysis

Analysis of the interaction between the condensed GP and the coupling agent was conducted using X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FTIR) analysis. These techniques allow for quantitative evaluation of the extent of reaction of a GP. In particular, as the metakaolin precursor gradually forms GP, the amorphous peak at ≈22° 2θ moves to ≈28°.[1] Likewise, as the Si–O–Si symmetric stretch is augmented by the Si–O–Al asymmetric stretch, FTIR results show a shift in a prominent hump from ≈1080 to ≈1000 cm−1.[27] Cu-alpha X-ray powder diffraction testing was conducted with a Siemens-Bruker (Madison, WI) D5000 diffractometer. Tests were conducted at 40 kV, 30 mA, a step size of 0.02° 2θ, and a measurement time of 8 s/step. Diffuse reflection infrared spectra were taken using a Nicolet Nexus 670 spectrometer (Thermo Scientific, Waltham, MA) on powdered samples mixed with KBr (1:19 wt ratio). Powder samples for both instruments were generated by dry grinding in an alumina mortar and pestle, and were collected after being passed through a 44-μm mesh.

(5) Microstructural Analysis

Scanning and transmission electron microscopy (SEM and TEM) were used to investigate the microstructure and nanostructure of the material. SEM data were obtained on sample cross-sections cut with a diamond-edge wafer blade, mounted in epoxide, and smoothed to a flat surface with a diamond-colloid polisher–grinder. The sample mount was Au–Pd sputter coated to a thickness of 5 nm. SEM Micrographs were captured at 25 kV using a typical instrument (JSM-6060LV; JEOL USA, Inc., Peabody, MA). Vacuum impregnation of the epoxide was necessary to clearly image the porosity of the modified samples, and was conducted at ≈5 kPa.

TEM micrographs were obtained on powder samples prepared as above and placed on a carbon film. Samples were imaged at room temperature using an excitation voltage of 200 kV and a lanthanum hexaboride filament (JEOL 2010, JEOL USA, Inc.). Care was taken to minimize beam exposure prior to image capture, especially for images of the organic phase.

(6) Drying Analysis and Shrinkage

Drying analysis was conducted on both GPs and GP composites. GP experiments were conducted at ambient conditions and using “B” GP composition samples modified with silanes as specified. They were cured at 50°C for 1 d in a cylindrical mold of 1 cm height and 1 cm diameter. For greater precision, larger samples and constant environmental conditions were used for similar tests using GP composites. GP composite experiments were conducted using 25°C, 30% humidity conditions with cylindrical molds of 5 cm height and 2.5 cm diameter. Shrinkage was measured by the final equilibrium width of formulation “A” GPs cast in 10-mm-width rectangular prism molds. The equilibrium widths of each were measured by calipers in the same ambient conditions after more than 2 months, and compared to the original 10 mm molded width.

Differential scanning calorimetry data were acquired at a rate of 1°C/min from room temperature to 400°C, and then immediately tested from 400°C to 1100°C at 5°C/min. The testing chamber used (STA 409 CD; Netzsch Group, Burlington, MA) was a continuous airflow chamber and the gasses used were Ultra Zero Air and UHP Helium at a 15:8 volume ratio.

(7) Compression Testing

Compression tests were conducted on GP and GP samples with and without the addition of 0.1 mol MEMO/mol GP. GP slurry was poured into a cylindrical mold of 30 cm height and 1.27 cm diameter and vibrated on a vibration table until no further debubbling was observed. These samples were cured and the resulting GP rods were cut into 2.5 cm high cylinders. Compression tests were conducted using an Instron 5882 load frame (Instron Corp., Norwood, MA) according to ASTM C39, except that drying shrinkage meaningfully altered the final diameters of the MEMO samples (see below) and in some cases caused eccentricity in the cylinder cross section. Height–width correction factors were applied using a cubic fit. The chamotte (Ceske Lupkove Zavody, Pecinov, Czech Republic) was used as received and had mean particle diameter ≈26 μm and median particle diameter ≈19 μm, measured using laser diffraction (LA-950; Horiba Scientific, Irving, CA).

III. Results and Discussion

  1. Top of page
  2. Abstract
  3. I. Introduction
  4. II. Experimental Procedure
  5. III. Results and Discussion
  6. IV. Conclusion
  7. Acknowledgments
  8. References

It was found that meaningful microstructural and property changes occurred as a result of adding MEMO to the GP. Alterations in mechanical properties, water retention properties and drying shrinkage all directly result from the simple addition of this silane to the GP slurry. It was shown that the effectiveness of the modifier on these properties depends primarily on the functional ligand, as property modification could be substantially reduced or eliminated by changing the ligand without changing the alkoxysilane functionality. The addition of the silanes reduced the reaction completion, but the reduction was minor with the quantities measured.

(1) Water Retention

Drying under various conditions revealed that the MEMO-modified GP, filled or unfilled, possessed greatly improved water retention characteristics. As shown in Fig. 1, while both sets of samples had flux-limited drying (rather than having a constant-rate evaporation-limited drying period), drying of the control was substantially faster, where the control GP lost 20% of its theoretical water mass within 9 h, whereas the GP modified with MEMO required 68 h. It was further observed that the MEMO GP did not approach equilibration within 2 weeks. The additional mass loss at very long times may have been due to loss of the silane itself or may suggest a higher equilibrium water loss. The latter might suggest the gradual depletion of hydrated sodium ions as they formed sodium methacrylate groups, which would be consistent with previous GP research.[13]

image

Figure 1. A comparison of mass loss rates as a function of time held at 25°C and 30% relative humidity. Formulation “A” geopolymers cured at 50°C showed rapid mass loss, while the same geopolymer modified with MEMO dried substantially more slowly.

Download figure to PowerPoint

A direct comparison of the drying of several MEMO-modified, PAMS-modified, mixed MEMO-PAMS-modified, and unmodified formulation “B” GPs at ambient conditions found that drying rate depended on the presence and concentration of MEMO, and that the addition of PAMS had no effect on the drying rate. The results were observed for both as-cast samples and samples prepared by removal of the surface using a file. This indicated that drying effects of MEMO occurred within the bulk of the GP and were not merely a surface-sealing effect.

DSC-TGA data showed a delay in the onset of mass loss during heating, with control and 0.05 mol MEMO/mol GP samples exhibiting an immediate broad exotherm beginning at room temperature, while 0.10 mol MEMO/mol GP delayed the exotherm until ≈62°C. The 0.15 mol MEMO/mol GP sample showed a sharp exotherm beginning at ≈69°C. At the boiling point of MEMO (≈190°C at atmospheric pressure), a small exotherm was observed for the 0.15 mol MEMO/mol GP sample but not for lower additive concentrations.

(2) Mechanical Properties and Shrinkage

Compression and shrinkage properties of MEMO-modified GPs and GP composites (Formulation “A”) were compared with controls. As shown in Table 2, the effect of the MEMO additive on the compressive strength was minimal, but its effect on the Weibull modulus was considerable. This increased modulus is consistent with the smaller standard deviation of compressive strength in implying a more reliable failure, suggesting that the MEMO was able to reduce crack propagation or other sources of brittle failure. Interestingly, MEMO additive reduced the strength benefit of the filler while the filler reduced the reliability benefit of the MEMO additive.

Table 2. Compressive Strengths of Geopolymer with and Without MEMO Additive
Geopolymer formulationStrength (MPa)Weibull scale factor (MPa), modulus
“A” 25°C15.5 ± 9.317.8, 1.80
“A” 25°C with 0.10 mol/mol MEMO18.8 ± 5.320.8, 3.84
“A” 25°C with 25 wt% chamotte filler44.6 ± 21.951.5, 2.11
“A” 25°C with 25 wt% chamotte filler and 0.10 mol/mol MEMO41.6 ± 13.746.9, 3.14

Figure 2 shows the approximate drying shrinkage incurred by GPs as a function of MEMO additive. It is clear that significant shrinkage occurred upon addition of sufficient amounts of additive. The densification of the GP that occurred during this shrinkage is even more directly related to amount of additive. The bulk densities observed, measured using mercury pychometry, are extremely high compared to other metakaolin Na- GPs described in the literature,[28] and resemble those achieved for wet metakaolin GPs cured at low temperatures.[29] As the MEMO additive is not massive and could not be expected to form a particularly dense structure, these effects were due to changes in the porous GP structure. The linear shrinkage was observed to be nonlinear in concentration while the bulk density showed a clear linear trend. This could have been caused to the anisotropy of the sample shapes used in the linear shrinkage tests (1 cm × 1 cm × 10 cm bars), or to the additional vacuum drying of the samples required for the mercury pycnometry test. Mercury intrusion found the porosity to be only minimally open, which is consistent with the micrographs observed. The lack of intruded volume in the MEMO-modified GPs made meaningless the macropore-size distribution observed with MIP.

image

Figure 2. Linear shrinkage measurements of formulation “A” geopolymers cured at 25°C. After the addition of a critical amount of MEMO additive (approximately 0.06 mol/mol), further additive causes substantial linear shrinkage. Bulk density measurements using mercury intrusion porosimetry did not show such a critical change but instead increased smoothly with increased composition.

Download figure to PowerPoint

(3) Microstructure: Imaging and Composition

Fourier transform infrared spectroscopy results (Fig. 3) clearly showed the interaction of the methacrylate groups with sodium cations, with the COONa asymmetric stretch at 1566 cm−1 and the symmetric stretch at 1394 cm−1 clearly visible,[30] whereas the bare C=O stretch observed at 1717 cm−1 in MEMO[31] was not observed. It was impossible to exclude the possibility of MEMO polymerization at the carbon double bonds because its characteristic stretch at 1638 cm−1 overlaps the large, wide H2O bend signal in the vicinity. GP reactivity, as measured by the Si–O–(Si, Al) stretch peak movement, decreased with increasing MEMO concentration, as shown in Table 3. The XRD diffraction results in Table 3 also show decreasing reactivity with increasing concentration as the major amorphous peak shifts to smaller angles.

Table 3. Reactivity Data for Formulation “A” Geopolymers
MEMO concentration (mol/mol GP) FTIR Si–O–(Si, Al) stretch peak maximum (cm−1)Approximate XRD amorphous hump maximum (°2θ)
01008[38]27.2
0.05102027.0
0.10103525.3
0.15104324.6
image

Figure 3. FTIR results of formulation “A” geopolymers with varying mol MEMO/mol GP. The formation of sodium acrylate groups is visible. Integration of SiOAl into the SiOSi stretch decreases with increased additive, suggesting decreased reactivity.

Download figure to PowerPoint

Scanning electron microscopy results revealed a dense microstructure with closed pores of μm to tens of μm in diameter interspersed among tightly bound precipitate aggregates. Some microcracking was evident among these pores, but the overall structure was quite dense by GP standards. A direct comparison of the microstructures of formulation “A” 25°C GP with the addition of 0.1 mol MEMO/mol GP (Fig. 4) shows substantially less porosity than is typical for sodium GPs alone.[32, 33] Closer images of the higher additive compositions illustrated the presence of curled organic films on a scale of hundreds of nanometers to micrometers (Fig. 5). These films were embedded into the bulk of the GP as a separate phase and appeared to show no particular affinity for the pore surfaces. This strongly suggests that their formation occurred prior to the complete hardening of the GP gel as otherwise a bulk organic phase such as the observed might be expected to be excluded by the growing GP. From the images, it cannot be determined whether complete phase separation of the MEMO additive occurred, or alternatively whether the observed structures represent only the additive that was not strongly bound within the matrix. Within the images are numerous cracks, which occurred as a result of the vacuum impregnation of the samples with epoxide. It was found to be necessary to conduct epoxide infiltration to get useful porosity contrast in the modified samples.

image

Figure 4. A scanning electron micrograph of a formulation “A” geopolymer with 0 (a) and 0.1 (b) mol MEMO/mol GP. Epoxide has been infiltrated into the modified samples to highlight the macroporosity. The decrease in macroporosity is evident.

Download figure to PowerPoint

image

Figure 5. Scanning electron micrographs of formulation “A” geopolymers with 0 (a), 0.05 (b), 0.10 (c), and 0.15 (d) mol MEMO/mol GP. As MEMO concentration increases, microstructural complexity and unreacted phases begin to appear. Membranes formed by the organic material are seen with high MEMO concentrations (c, d) where they appear curled up and bonded tightly into the geopolymer matrix.

Download figure to PowerPoint

(4) Nanostructure: Specific Surface Area, Precipitate Interaction, and Organic Phase

Gas adsorption porosimetry observed a marked decrease in SSA, as shown in Fig. 6. From these data, it is clear that the extended propyl linker of MEMO allowed for greater nanostructural changes than did the methyl linker of MEMO-Me. The network modification capabilities of the methacrylate groups are likely individually quite short range. Altering the alkoxysilane functionality of the additive or the acrylic nature of the functional group had an even more meaningful effect. Use of MEMO-2,1 as an additive rather than MEMO resulted in no meaningful SSA reduction. Use of PAMS also resulted in a smaller change in SSA than did similar quantities of MEMO. NLDFT results shown in Fig. 7 demonstrate that mesopores between 5 and 20 nm were reduced by approximately an order of magnitude above the critical concentration of 0.06 mol MEMO/mol GP.

image

Figure 6. BET-specific surface area as a function of alkoxysilane additive concentration in formulation “A” geopolymers. The addition of MEMO clearly causes decreased specific surface area. After the addition of a critical amount of MEMO additive (approximately 0.06 mol/mol) further additive causes substantial microstructural change, echoing the shrinkage results.

Download figure to PowerPoint

image

Figure 7. Nonlocal density functional theory modeling of mesoporosity from nitrogen adsorption results, as a function of MEMO concentration. Mesopores above approximately 5 nm are substantially eliminated when more than 0.06 mol MEMO/mol GP is added. The data are quite noisy due to the tiny amounts of porosity observed at this length scale.

Download figure to PowerPoint

This result suggests a substantial integration of precipitates, as pores of this size range are typically considered to be the voids that separate incompletely unified precipitates,[5, 34] while smaller pores represent precipitate roughness. TEM micrographs of precipitate aggregates appear to reflect this hypothesis, showing a qualitatively smooth surface [Fig. 8(a)] with structural features of a much larger length scale [Fig. 8(b)] than was observed in micrographs of controls observed here and elsewhere.[32] It is not clear from TEM imaging whether any organic material was bonded within precipitates and/or aggregates.

image

Figure 8. Transmission electron micrographs of a formulation “A” geopolymer with 0.1 mol MEMO/mol GP. The precipitates are seen to be less jagged than unmodified geopolymers (a). Large mesopores are visible within some agglomerates (b). The organic material forms easily distinguishable tube structures that in some cases connect agglomerates (c) and are encapsulated within agglomerates (d).

Download figure to PowerPoint

Instead, the organic material formed a separate phase of tube-like structures on the order of 10 s of nm in diameter and 100 s of nm long [Figs. 8(b)–(d)]. These structures appear imbedded within the GP aggregates and clearly bonded some GP material strongly enough to survive drying and vacuum. A substantial number of tubes are found completely detached from the GP, either removed during grinding to powder or during evacuation of the sample.

(5) Mechanism

These microstructural and property alterations imply a significant alteration of the GP chemistry, which in turn suggests substantial room for manipulation and property improvement. In particular, there exist substantial similarities between this system and an organic hydrogel system, including the predominance of cross-linked sodium methacrylate as waterlocking agents. Previous investigations of these materials have demonstrated increased hydrogel-forming capability with increased structure size, as measured mass per mass.[17] In accordance with previous GP research, the drying shrinkage results suggest a system in which insufficient water existed to preserve hydration spheres around the cations. In this case, the addition of the MEMO additive would be expected to greatly increase the hydration radius of the cations through the formation of sodium methacrylate groups within the GP. The densification observed is then directly a result of this process.

Gas adsorption results suggest that the mesoporous water in these modified GP s was waterlocked into pores at a length scale of approximately 40–100 nm. In contrast, the silane coupling agent tube structures occurred on a length scale of hundreds of nm in width and μm in length, implying that each tube affected many pores rather than forming a single massive film encapsulating a single large pore. This is consistent with the TEM observation that some GP material appeared to be trapped within the tubes, and that SEM images showed some curled structures at the surface of the cut cross section. It is possible that within the GP matrix some portion of the organic material formed membranes or films. These might have rolled up during drying when allowed freedom of movement due to being exposed when the same was cut. Evidence of multiwalled tubes in some of the images supports this theory.

Thus, the available evidence suggests that during the condensation and gelling stage, the organic material existed in membranes, which both attached to some GP precipitates and bound significant amounts of the free water. This in turn depleted much of the free water elsewhere, resulting in reduced 5–20 nm porosity. This is consistent with the observation that GP mesoporosity is typically attributed to surplus free water.[35-37] The pores which did form in this system were observed to be only 1–5 nm in diameter, and could have been formed with water released from the geopolymerization condensation reaction itself. This water appeared very late in the reaction process, and was presumably isolated from the organic membranes by the GP network.

IV. Conclusion

  1. Top of page
  2. Abstract
  3. I. Introduction
  4. II. Experimental Procedure
  5. III. Results and Discussion
  6. IV. Conclusion
  7. Acknowledgments
  8. References

Geopolymer densification and porosity reduction were achieved through the use of hydrogel chemistry. While traditional hydrogels were difficult to integrate into the GP matrix during the curing process, alkoxysilanes functionalized with methacrylate groups mixed into the GP slurry and formed micrometer-sized membranes and rolled tubes. It was found that a critical concentration of approximately 0.06 mol methacrylate/mol GP was required for substantial surface area reduction, increased linear shrinkage and increased water retention characteristics, and that smaller amounts of the additive may even reverse the trends. It is suspected that the critical composition observed may occur at the concentration where the GP network is fully saturated with organic additive, so that only additional additive causes the observed porosity reduction. With 0.10 mol methacrylate/mol GP, the natural GP mesoporosity was reduced by approximately an order of magnitude, and the Weibull modulus was increased by almost 50% even for reinforced GP composites.

Acknowledgments

  1. Top of page
  2. Abstract
  3. I. Introduction
  4. II. Experimental Procedure
  5. III. Results and Discussion
  6. IV. Conclusion
  7. Acknowledgments
  8. References

The authors acknowledge the use of facilities at the Center for Microanalysis of Materials, in the Frederick Seitz Research Laboratory at the University of Illinois at Urbana-Champaign. The authors greatly appreciate the assistance of Thomas Carlson for transmission electron microscopy assistance, and Pathikumar Sellappan for differential scanning calorimetry assistance.

References

  1. Top of page
  2. Abstract
  3. I. Introduction
  4. II. Experimental Procedure
  5. III. Results and Discussion
  6. IV. Conclusion
  7. Acknowledgments
  8. References
  • 1
    J. Davidovits, “Inorganic Polymeric New Materials,” J. Therm. Anal. Calorim., 37, 163356 (1991).
  • 2
    J. G. S. van Jaarsveld, J. S. J. van Deventer, and G. C. Lukey, “The Effect of Composition and Temperature on the Properties of Fly Ash- and Kaolinite-Based Geopolymers,” Chem. Eng. J., 89 [1–3] 6373 (2002).
  • 3
    S. A. Bernal, E. D. Rodriguez, R. M. de Gutierrez, M. Gordillo, and J. L. Provis, “Mechanical and Thermal Characterisation of Geopolymers Based on Silicate-Activated Metakaolin/Slag Blends,” J. Mater. Sci., 46 [16] 547786 (2011).
  • 4
    S. Detphan and P. Chindaprasirt, “Preparation of Fly Ash and Rice Husk Ash Geopolymer,” Int. J. Min. Met. Mater., 16 [6] 7206 (2009).
  • 5
    W. M. Kriven, J. L. Bell, and M. Gordon, “Microstructure and Nanoporosity of as-Set Geopolymers”; pp. 491503 in Mechanical Properties and Performance of Engineering Ceramics II: Ceramic Engineering and Science Proceedings, Vol. 27, Issue 2, Edited by R. Tandon, A. Wereszczak and E. Lara-Curzio. John Wiley & Sons, Inc., Hoboken, NJ, 2006
  • 6
    J. Davidovits, Geopolymer Chemistry and Applications. Institut Geopolymere, St. Quentin, France, 2008.
  • 7
    W. M. Kriven, J. L. Bell, M. Gordon, and G. Wen, “Geopolymers: More Than Just Cements”; pp. 17983 in Geopolymer, Green Chemistry and Sustainable Development Solutions, Edited by J. Davidovits. World Congress Geopolymer, Saint-Quentin, France, 2005
  • 8
    T. L. Metroke, M. V. Henley, and M. I. Hammons, “Effect of Curing Conditions on the Porosity Characteristics of Metakaolin–Fly Ash Geopolymers”; pp. 115 in Strategic Materials and Computational Design: Ceramic Engineering and Science Proceedings, Vol. 31, Issue 10, Edited by W. M. Kriven, Y. Zhou and M. Radovic. John Wiley & Sons, Inc., Hoboken, NJ, 2010
  • 9
    M. G. Blackford, J. V. Hanna, K. J. Pike, E. R. Vance, and D. S. Perera, “Transmission Electron Microscopy and Nuclear Magnetic Resonance Studies of Geopolymers for Radioactive Waste Immobilization,” J. Am. Ceram. Soc., 90 [4] 11939 (2007).
  • 10
    J. L. Provis and J. S. J. van Deventer, “Geopolymerisation Kinetics. 2. Reaction Kinetic Modelling,” Chem. Eng. Sci., 62 [9] 231829 (2007).
  • 11
    P. Duxson, G. C. Lukey, and J. S. J. van Deventer, “Thermal Conductivity of Metakaolin Geopolymers Used as a First Approximation for Determining Gel Interconnectivity,” Ind. Eng. Chem. Res., 45 [23] 77818 (2006).
  • 12
    S. S. Musil, G. Kutyla, and W. Kriven, “The Effect of Basalt Chopped Fiber Reinforcement on the Mechanical Properties of Potassium Based Geopolymer”; pp. 3142 in Developments in Strategic Materials and Computational Design III: Ceramic Engineering and Science Proceedings, Vol. 33, Issue 10, Edited by W. M. Kriven, A. L. Gyekenyesi, G. Westin and J. Wang. American Ceramics Society, Daytona Beach, FL, 2012
  • 13
    C. Kuenzel, L. J. Vandeperre, S. Donatello, A. R. Boccaccini, and C. Cheeseman, “Ambient Temperature Drying Shrinkage and Cracking in Metakaolin-Based Geopolymers,” J. Am. Ceram. Soc., 95 [10] 32707 (2012).
  • 14
    J. L. Provis and J. S. J. Deventer, “Direct Measurement of the Kinetics of Geopolymerisation by in-Situ Energy Dispersive X-Ray Diffractometry,” J. Mater. Sci., 42 [9] 297481 (2006).
  • 15
    E. P. Plueddemann, Silane Coupling Agents. Plenum Press, New York, NY, 1991.
  • 16
    B. E. Glad, C. Han, and W. M. Kriven, “Polymer Adhesion to Geopolymer Via Silane Coupling Agent Additives,” J. Am. Ceram. Soc., 95 [12] 375862 (2012).
  • 17
    R. Schweins, J. Hollmann, and K. Huber, “Dilute Solution Behaviour of Sodium Polyacrylate Chains in Aqueous NaCl Solutions,” Polymer, 44 [23] 713141 (2003).
  • 18
    K. R. Rogan, “The Variations of the Configurational and Solvency Properties of Low Molecular Weight Sodium Polyacrylate with Ionic Strength,” Colloid Polym. Sci., 273 [4] 3649 (1995).
  • 19
    S. Sacanna and D. J. Pine, “Shape-Anisotropic Colloids: Building Blocks for Complex Assemblies,” Curr. Opin. Colloid Interface Sci., 16 [2] 96105 (2011).
  • 20
    W. M. Kriven, J. L. Bell, and M. Gordon, “Microstructure and Microchemistry of Fully-Reacted Geopolymers and Geopolymer Matrix Composites,” Ceram. Trans., 153, 22750 (2004).
  • 21
    J. Zhou, G. Ye, and K. van Breugel, “Characterization of Pore Structure in Cement-Based Materials Using Pressurization–Depressurization Cycling Mercury Intrusion Porosimetry (PDC-MIP),” Cem. Concr. Res., 40 [7] 11208 (2010).
  • 22
    B. E. Glad and W. M. Kriven, “Optimization of Gas Adsorption Porosimetry for Geopolymer Analysis,” J. Am. Ceram. Soc. [in press] (2013).
  • 23
    S. Brunauer, P. H. Emmett, and E. Teller, “Adsorption of Gases in Multimolecular Layers,” J. Am. Chem. Soc., 60 [2] 30919 (1938).
  • 24
    M. Jaroniec, M. Kruk, J. P. Oliver, and S. Koch, “A New Method for the Accurate Pore Size Analysis of MCM-41 and Other Silica Based Mesoporous Materials”; pp. 7180 in Studies in Surface Science and Catalysis, Vol. 128, Edited by K. K. Unger, G. Kreysa, and J. P. Baselt. Elsevier, Waltham, MA, 2000
  • 25
    Micromeritics, Micro Active, 1.01 ed. Micromeritics Instrument Corp., Norcross, GA, 2012.
  • 26
    E. P. Barrett, L. G. Joyner, and P. P. Halenda, “The Determination of Pore Volume and Area Distributions in Porous Substances. I. Computations from Nitrogen Isotherms,” J. Am. Chem. Soc., 73 [1] 37380 (1951).
  • 27
    V. F. F. Barbosa, K. J. D. MacKenzie, and C. Thaumaturgo, “Synthesis and Characterisation of Materials Based on Inorganic Polymers of Alumina and Silica: Sodium Polysialate Polymers,” Int. J. Inorg. Mater., 2 [4] 30917 (2000).
  • 28
    K. Okada, A. Ooyama, T. Isobe, Y. Kameshima, A. Nakajima, and K. J. D. MacKenzie, “Water Retention Properties of Porous Geopolymers for Use in Cooling Applications,” J. Eur. Ceram. Soc., 29 [10] 191723 (2009).
  • 29
    P. Rovnanik, “Effect of Curing Temperature on the Development of Hard Structure of Metakaolin-Based Geopolymer,” Constr. Build. Mater., 24 [7] 117683 (2010).
  • 30
    E. S. Rufino and E. E. C. Monteiro, “Infrared Study on Methyl Methacrylate–Methacrylic Acid Copolymers and Their Sodium Salts,” Polymer, 44 [23] 718998 (2003).
  • 31
    M. Pantoja, B. Díaz-Benito, F. Velasco, J. Abenojar, and J. C. del Real, “Analysis of Hydrolysis Process of Γ-Methacryloxypropyltrimethoxysilane and Its Influence on the Formation of Silane Coatings on 6063 Aluminum Alloy,” Appl. Surf. Sci., 255 [12] 638690 (2009).
  • 32
    P. Duxson, J. L. Provis, G. C. Lukey, S. W. Mallicoat, W. M. Kriven, and J. S. J. van Deventer, “Understanding the Relationship Between Geopolymer Composition, Microstructure and Mechanical Properties,” Colloid. Surface. A, 269 [1–3] 4758 (2005).
  • 33
    C. F. Maitland, C. E. Buckley, B. H. O'Connor, P. D. Butler, and R. D. Hart, “Characterization of the Pore Structure of Metakaolin-Derived Geopolymers by Neutron Scattering and Electron Microscopy,” J. Appl. Crystallogr., 44, 697707 (2011).
  • 34
    W. M. Kriven, J. Bell, S. W. Mallicoat, and M. Gordon, “Intrinsic Microstructure and Properties of Metakaolin-Based Geopolymers”; pp. 7186 in Proceedings of the International Workshop on Geopolymer Binders - Interdependence of Composition, Structure and Properties, Edited by A. Buchwald, K. Dombrowski and M. Weil. Bauhaus-Universität, Weimar, Germany, 2006
  • 35
    J. L. Bell, P. E. Driemeyer, and W. M. Kriven, “Formation of Ceramics from Metakaolin-Based Geopolymers: Part I-Cs-Based Geopolymer,” J. Am. Ceram. Soc., 92 [1] 18 (2009).
  • 36
    J. L. Bell, P. E. Driemeyer, and W. M. Kriven, “Formation of Ceramics from Metakaolin-Based Geopolymers. Part II: K-Based Geopolymer,” J. Am. Ceram. Soc., 92 [3] 60715 (2009).
  • 37
    P. Duxson, G. C. Lukey, and J. S. J. van Deventer, “Thermal Evolution of Metakaolin Geopolymers: Part 1 - Physical Evolution,” J. Non-Cryst. Solids, 352 [52–54] 554155 (2006).
  • 38
    B. E. Glad and W. M. Kriven, “Humidity Effects on the Completion of Geopolymerization in Dilute Evaporative Slurries”; pp. 2530 in Developments in Strategic Materials and Computational Design III: Ceramic Engineering and Science Proceedings, Vol. 33, Issue 10, Edited by W. M. Kriven, A. L. Gyekenyesi, G. Westin, and J. Wang. American Ceramics Society, Daytona Beach, FL, 2012