Environ. Sci. Technol. 2001, 35, 4120-4125
Vacuum Method for Carbonation of Cementitious Wasteforms MICHAEL A. VENHUIS* AND ERIC J. REARDON Department of Earth Sciences, University of Waterloo, Waterloo, Ontario, Canada, N2L 3G1
The application of carbonation treatment to cement-based wasteforms as a means to reduce the leachability of entrained undesirable substances, both radioactive and nonradioactive, has been the subject of much study over the past decade. Upon carbonation, hydrated cement phases release their water of hydration and are converted into carbonate minerals. The carbonation process has been shown to reduce the pore size distribution and permeability of the cementitious materials since portions of the original pore network become sites for precipitation of secondary carbonate minerals. As a result, the leachability of entrained contaminants can be markedly reduced. Current methods to carbonate cement-based wasteforms after they have cured rely on exposure to high pressure or supercritical CO2 pressures. In this study, a new lowpressure technique is presented. The method provides more complete carbonation than high-pressure techniques. The principle is to remove the water of reaction as it is produced, thereby maintaining an open pore network to facilitate the transfer of CO2 into the specimen. This is accomplished by conducting the reaction at near-vaccum pressures in the presence of a desiccant. The nearvacuum conditions lower the impedance of water transport from the carbonating specimen to the desiccant due to the large mean free path of the water vapor molecules. The technique was applied to a series of wasteform samples with entrained cationic and anionic waste components. Carbonation penetration depths of up to 11 mm were attained within 45 h of reaction for cylindrical wasteform samples prepared with OPC at a water/cement ratio of 0.60. A carbonation penetration depth of 15 mm was attained in a 6 d reaction of blended cement (OPC and 30% Class ‘F’ fly ash). In standardized leach tests, cationic waste constituents showed lower leachabilities from carbonated samples than from uncarbonated samples. Anionic waste constituents, however, showed greater leachabilities, and anion leachability increased with the degree of carbonation.
1. Introduction Public concern over nuclear waste disposal in the past few decades has spawned considerable research into disposal methods. For low-level radioactive wastes, focus has centered on the use of cementitious materials for their solidification and stabilization. The U.S. Environmental Protection Agency cites it as one of the best-demonstrated available technologies * Corresponding author phone: (613)224-5864; fax: (613)224-9928; e-mail:
[email protected]. Current address: Golder Associates Ltd., 1796 Courtwood Cres., Ottawa, Ontario, Canada K2C 2B5. 4120
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 20, 2001
for the land disposal of most radioactive elements (1). The use of cement in stabilization/solidification technology is advantageous because of its low cost, availability and adaptability, tolerance to wet materials (e.g. needs water to set), nonflammability, and durability in the natural environment. The disadvantage of cement-based wasteforms is that the constituent minerals of hydrated cement have moderate to high solubility at the near-neutral pH conditions of most natural waters. Thus the reaction of the wasteform with carbon dioxide (carbonation) to create a protective rind or layer around the cemented wasteform has been widely investigated (2-10). The principal reactions involved in the carbonation of hydrated cements are the transformation of CSH gel (xCaO‚SiO2‚nH2O), portlandite (Ca(OH)2), ettringite (Ca6Al2O6(SO4)3‚32H2O), and AFm (Ca4Al2O6(SO4)•12H2O) to calcium carbonate, gypsum, and an assemblage of amorphous aluminum and silica oxyhydroxides (12, 13). Substantial water is released as a result of these reactions as shown below:
xCaO‚SiO2‚nH2O + CO2(g) f xCaCO3 + SiO2 + nH2O
Ca(OH)2 + CO2(g) f CaCO3 + H2O
(1) (2)
Ca6Al2O6(SO4)3•32H2O + 3CO2(g) f 3CaCO3 + 3CaSO4•2H2O + Al2O3•3H2O + 23H2O (3) Ca4Al2O6(SO4)•12H2O + 3CO2(g) f 3CaCO3 + CaSO4•2H2O + Al2O3•3H2O + 7H2O (4) The reaction of carbon dioxide and Ordinary Portland Cement (OPC) causes portlandite and other hydrated cement phases to release their water of hydration and convert to calcium carbonate minerals. Since the pore network provides sites for precipitation of these secondary minerals, this process reduces the overall pore size distribution and permeability of the concrete (2), desirable properties for any matrix intended to immobilize waste materials. Carbonation treatment can thus markedly alter the microstructure of a hydrated cement and potentially reduce the leachability of entrained contaminants. The formation of calcium carbonate solid solutions may also reduce the leachability of some ions (3). Many of the studies reporting results of leach tests on carbonated cements have shown improvement in the leach resistance for a number of metal ions and radionuclides (36). Enhanced leaching, however, of nitrate from carbonated wasteforms in deionized water and nitrate, cobalt, cadmium, and lead from carbonated wasteforms in acetic acid have been reported by Bin-Shafique et al. (7) and Gutierrez et al. (8). Leachability is perhaps the most important aspect to consider for immobilization methods of radioactive wastes, and the extent to which it has been reduced for particular waste components must be demonstrated for any proposed carbonation treatment. The problem in producing extensive carbonation of cementitious material is that water produced from the reaction accumulates in the pores. Eventually the pores become blocked, preventing further migration of CO2 into the sample. Thus when wasteform samples are subjected to CO2, carbonation typically yields an exterior rind of fully carbonated material with a largely unreacted interior. The use of high pressures (up to 5.5 MPa) of CO2 to overcome this problem was tried by Reardon et al. (9). Their rationale was that high amounts of CO2 would be introduced throughout 10.1021/es0105156 CCC: $20.00
2001 American Chemical Society Published on Web 09/12/2001
the sample before pore closure would occur. They attained a greater degree of carbonation, but pore closure still stopped the reaction before carbonation was complete. Recently, CO2 under supercritical conditions has been used to carbonate cemented wasteforms (5, 10, 11). These authors argue that supercritical CO2 is an effective solvent for water and that substantive carbonation of wasteforms can be attained. They report carbonation penetration depths of 4 mm after 2 h of reaction at 8.4 MPa and 35 °C (5). In a recent study, Venhuis (14) compared the efficiency of high pressure and supercritical CO2 carbonation treatments. The results showed similar CO2 mass uptake results for the two methods, but, based on petrographic examination, the reaction rind appeared to penetrate deeper for the supercritical treatment. The maximum reported penetration depth attained with the supercritical CO2 treatment was 7 mm after 72 h of reaction (14). It appears that to enable deeper and more pervasive carbonation of cementitious wasteforms, some means to remove or prevent the secondary water of reaction from accumulating must be found. This study describes an alternative, vacuum technique to carbonate cementitious wasteforms. The method addresses the problem of pore closure accompanying carbonation by continuous removal of water from the sample as it is produced. In the treatment design presented here, the sample is placed in an evacuated reaction cell in contact with a desiccant. Near-vacuum conditions are maintained throughout the carbonation reaction. This results in a large mean free path for gaseous water molecules, which enhances their transport from the accumulating liquid water in the pores of the sample wasteform to the desiccant as the reaction proceeds. This study also presents the results of leach tests of various contaminants from wasteforms carbonated with this technique and compares them to uncarbonated samples.
2. Experimental Methods 2.1. Sample Preparation. Cemented wasteform samples were prepared by mixing OPC (St. Mary’s Type 10 OPC; CSA CAN3A5) with a simulated waste solution. The solution was prepared to contain nickel (500 ppm), arsenic (500 ppm), strontium (600 ppm), cesium (1000 ppm), nitrate (1000 ppm), and chloride (1356 ppm) using appropriate quantities of the following reagent-grade chemicals: NiCl2•6H2O, Na2HAsO4• 7H2O, SrCl2•6H2O, CsCl, and NaNO3. Upon combination of these salts, the solution turned greenish, and a white precipitate, identified later as nickel arsenate (Ni3(AsO4)2• 8H2O), formed. The solution was vigorously stirred and mixed with Type 10 OPC powder that had been passed through a number 10 sieve. The resulting grout mixture was stirred several minutes until a consistent slurry was obtained and then poured into low density polyethylene (LDPE) molds producing cylindrical grout specimens approximately 2.7 cm in length and 2.2 cm in diameter. Prior to casting each specimen, the cement slurry was briefly restirred to ensure a homogeneous mixture. After casting, the grout specimens were capped and cured for various times in a 100% RH environment before reacting with CO2. Grout specimens were prepared at a water to cement mass ratio (w/c) of 0.45 (a standard used in the construction industry and commonly used in other carbonation studies) and at 0.60. A w/c ratio of 0.60 has been reported to give favorable porosity and strength properties to cemented wasteforms (8). This ratio is also being considered by the nuclear industry for encapsulation of hazardous liquid wastes (2). Samples were cured for either 14 or 60 days at room temperature (23 ( 1 °C) in a 100% RH, CO2-free environment. A total of 21 samples were prepared: seven samples for each 0.60 w/c ratio preparation condition (14- and 60-day cure times) and seven samples for the 0.45 w/c ratio condition with a 14-day cure time. Two samples from each preparation
FIGURE 1. Schematic representation of vacuum carbonation setup. condition were kept at 100% RH and left uncarbonated to serve as controls in later leach tests. 2.2. Experimental Procedure. After curing and before exposure to carbon dioxide, the specimens were demolded, and the outer layer of each specimen was scraped to remove potentially carbonated surface material. Samples were then weighed and dried under vacuum for 48 h to remove excess water and create an open pore network. Each set of five specimens was placed in the reaction cells above a saturated barium chloride dihydrate (BaCl2•2H2O) solution with excess solid present. The desiccant maintained an 88% RH or 2.8 kPa partial pressure of H2O in the cell. Under this condition, free water cannot accumulate within the grout specimens, and the pore network should remain open. The reaction vessel was a stainless steel reaction cell (500cc Zipper Closure, Autoclave Engineers). A pressure transducer connected to a data acquisition system was used to monitor and log pressure readings throughout each carbonation experiment. Before introducing CO2, the samples were allowed to equilibrate with the desiccant for approximately 3 h. After equilibration, metered quantities of CO2 were periodically added to the vessel from a CO2 reservoir through a series of computeractuated solenoid valves. By maintaining the injection reservoir at a constant pressure and controlling the opening of solenoid valves, similar volumes of CO2 were delivered upon each injection. Figure 1 shows a diagram of the CO2 injection system. A purge valve was needed because the high pressures in the CO2 supply tank caused too large a quantity of CO2 to enter the injection reservoir. Each addition of CO2 to the reaction cell caused an immediate increase in pressure (ca. 13 kPa). As the CO2 was consumed by reaction with the samples, the pressure in the reaction cell decreased. Once the pressure had dropped below a certain value (usually controlled at 5 kPa), 2 additional minutes were allowed for equilibration before the next addition of CO2 was introduced into the reaction vessel. The amount of CO2 introduced with each injection was known, and the frequency of the injections was logged using the data acquisition software. After 45 h, the frequency of injections had decreased to about once per hour for the 14day cured 0.45 and 0.60 w/c ratio samples. It took an additional 25 h, however, for the 60-day cured, 0.60 w/c ratio samples to reach this uptake rate. 2.3. Leach Test and Analysis. Duplicate samples of each wasteform formulation and treatment condition and the VOL. 35, NO. 20, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4121
FIGURE 2. Thin section photomicrographs under crossed polars of vacuum carbonated samples: (A) 0.60 w/c, 14-day cure time and carbonated 45 h (100% carbonated); (B) 0.60 w/c, 60-day cure time and carbonated 70 h (75% carbonated); and (C) 0.45 w/c, 14-day cure time and carbonated 45 h (55% carbonated). Carbonated regions are identified by the whitish buff matrix. Width of field of view is 1.19 cm for all samples. duplicate control samples were evaluated for their leaching characteristics using the American Nuclear Society’s ANSI/ ANS 16.1 procedure (15), a semidynamic leach test. In this procedure, the amount of leachate water used is determined by the geometrical properties of the wasteform (approximately 280 mL for this study). The leachate water was replaced after 2, 7, 24, 48, 72, 96, 120, 456, and 1128 h. Leachate concentrations of Ca, Sr, and Ni were analyzed using a TJA Inductively-Coupled Plasma Spectrophotometer (ICP), and nitrate and chloride were analyzed using a Dionex Ion Chromatograph (IC). Cs and As were analyzed using a Varian 1475-series Atomic Absorption Spectrophotometer (AAS). Samples not used for leaching experiments were sectioned to provide material for investigation by optical microscopy and X-ray diffraction (XRD). For XRD analysis, portions of both the carbonated (“rim”) and uncarbonated (“core”) regions from reacted samples were ground into a fine powder using a mortar and pestle and scanned at 0.05° increments over a 2θ range of 5-55° using a Seimens D500 diffractometer (Cu KR radiation and 1 s count time).
3. Results 3.1. CO2 Uptake. The amount of CO2 uptake by samples was obtained from the computer-logged pressure readings of the reaction cell and injection reservoir, which indicated the number and volume of CO2 injections during the experiment. CO2 uptake could also be approximated from the mass difference of specimens before and after carbonation, after correction for the mass loss of water. Overall, wasteforms prepared with a w/c ratio of 0.60 showed greater CO2 uptake than those prepared with a w/c ratio of 0.45. A cure time of 14 days enabled greater uptake of CO2 compared to a cure time of 60 days. 3.2. Microscopic Examination. Due to the characteristic, high-order, interference colors of CaCO3 under cross polarized light, carbonated material can be easily differentiated from uncarbonated or partially carbonated material by a color change: a buff color matrix for carbonated material and a dark gray to black matrix for largely uncarbonated material (16). Figure 2 shows representative photomicrographs of thin sections made from the various vacuum-carbonated specimens, and this difference is readily apparent. The thin sections were made from circular sections sliced from the center of the cylindrical specimens. Microscopic examination revealed complete carbonation of the 14-day cured 0.60 w/c ratio samples. Carbonation in the 14-day cured 0.45 w/c ratio and 60-day 0.60 w/c ratio samples was incomplete. The visual extent or depth of carbonation was 5 mm for the 14-day cured 0.45 w/c ratio 4122
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 20, 2001
FIGURE 3. X-ray diffraction patterns for various carbonated specimens (Cu Kr radiation): P, portlandite; E, ettringite; C3S, tricalcium silicate; C, calcite; A, aragonite; Al, aluminum (from aluminum sample holder). samples and 7 mm for the 60-day 0.60 w/c ratio samples, which corresponds to approximately 60% and 87% carbonation, respectively (Figure 2). This “visual” degree of carbonation evident in the photomicrographs is consistent with the calculated percentages of carbonation (refer to supporting text). 3.3. XRD Analyses. XRD patterns for both carbonated and uncarbonated samples are given in Figure 3. The similarity between the XRD patterns of uncarbonated samples with core regions of the 14-day 0.45 w/c and 60-day 0.60 w/c ratio samples indicates incomplete carbonation of these samples. Portlandite and ettringite are evident in the XRD patterns of uncarbonated core material. CSH is gellike material and does not display strong diffraction lines. It is
FIGURE 4. Cumulative fraction of waste elements released from vacuum carbonated wasteforms over the first 456 h of the leach test. Solid symbols represent carbonated samples and outlined symbols uncarbonated samples. Note the different scales for each element. Leach tests were completed according to ANSI/ANS 16.1 (15). usually revealed only by a slight increase in the baseline between 25° and 35° 2θ. So CSH’s presence or absence in carbonated specimens cannot be confidently established on the basis of XRD evidence. There is XRD evidence of calcite in the uncarbonated material, which is likely due to small amounts in the original cement powder. The aluminum peaks present are due to the Al-metal sample holder used in the XRD analysis. The core and rim XRD patterns for the 14-day cured 0.60 w/c ratio sample were indistinguishable, supporting the other lines of evidence indicating complete carbonation of the sample. Calcite, aragonite, and portlandite were the dominant phases identified in all rim samples with calcite being the predominant CaCO3 polymorph. Greater proportions of aragonite were present in the 14-day cured 0.60 w/c ratio samples than in other carbonated rim samples. With the exception of portlandite, little evidence of uncarbonated material was found. The peak found at 18.1 degrees 2θ is assigned to portlandite, suggesting incomplete conversion of this mineral during carbonation. However, this peak may be due, at least partly, to gibbsite (Al(OH)3), an expected
product from the carbonation of calcium aluminate minerals (AFm and ettringite) in the original cement. Gibbsite shows a major peak at a similar 2θ as the second major peak for portlandite. Arguing against the presence of gibbsite, however, is that no other peaks could be associated with this mineral. 3.4. Leach Testing. The ANS 16.1 (15) leach test results for the carbonated and uncarbonated grout specimens are displayed in Figure 4. The results are plotted as cumulative fraction of waste component released versus xtime. No plot is provided for nickel because all leachate concentrations for this element were below detection. Figure 4 shows the uncarbonated wasteform samples exhibited substantially greater leaching of calcium, strontium, and cesium than the carbonated samples. The percent recoveries of these cationic elements from the samples during the 45-day leach test procedure were all relatively low even for the uncarbonated samples. Cesium was an exception with close to 100% recovery observed for two of the three uncarbonated sample mixes (Figure 4). Carbonation, it seems, has an important sequestering effect on all cationic components. In contrast, VOL. 35, NO. 20, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4123
TABLE 1. Leachability Indices for Carbonated and Uncarbonated Waste Forms leachability index uncarbonated waste constituent
0.60 w/c 14 day
0.60 w/c 60 day
0.45 w/c 14 day
0.60 w/c 14 day
0.45 w/c 14 day
As Ca Cl Cs NO3 Sr
8.8 8.0 4.2 6.3 5.4 8.4
9.0 8.7 4.6 6.3 5.5 8.6
9.3 8.7 4.9 6.0 6.0 8.7
9.8 7.2 5.2 4.5 6.6 5.7
9.9 7.5 5.6 4.5 6.7 5.9
Walton et al. (17) carbonated uncarbonated 10.5
11.7
7.1 10.6
8.6 9.6
Results of Walton et al. (17) for selected waste constituents are provided for comparison.
all anionic elements, chloride, nitrate, and arsenate, showed greater leaching with carbonation (Figure 4), and samples that showed greater carbonation showed a greater degree of anion leaching. The calculated leachability indices based on the ANS 16.1 procedure (15) are given in Table 1. A lower value to the index indicates higher leachability. The results of this study are compared with those of Walton et al. (17) for selected waste ions. These authors found slightly higher leachability indices for strontium and nitrate than ours, and they found calcium leachability to be greater in carbonated than in uncarbonated wasteforms, which is opposite to that seen in this study.
4. Discussion 4.1. Sample Carbonation. In the initial stages of cement carbonation, CO2 invades the entire sample, causing an initial degree of pervasive carbonation. During this stage, the pore network is open, facilitating CO2 ingress and water removal from the interior of the sample. Subsequently, if the production of carbonation water exceeds the water removal by the desiccant, carbonation water will eventually block off CO2 entry to the interior of the sample, and carbonation will then proceed as a front, progressively moving toward the interior of the sample. From the petrographic photos, this may have occurred for the 14-day 0.45 w/c and the 60-day 0.60 w/c samples (Figure 2). Pore closure can only be prevented if H2O removal keeps pace with CO2 uptake. This can be accomplished if the amount of water removed by the desiccant is monitored during the reaction, and the delivery rate of the CO2 is made on that basis. This was not done in the current study. The only sample in which complete carbonation occurred was the 14-day cured 0.60 w/c sample (Figure 2A). Presumably, the high initial water content and short cure time of this sample provided a large and highly interconnected pore network that facilitated the transfer of water out of the sample and into the desiccant during carbonation. 4.2. Wasteform Mineralogy. Examination of the XRD patterns for the 14-day cured 0.60 w/c sample indicates near complete conversion of the uncarbonated materials to calcium carbonates, and because XRD patterns for both core and rim portions were indistinguishable, a pervasive carbonation was achieved. However, the principal XRD peak of portlandite is still evident in the rim material taken from all samples suggesting that carbonation is not 100% complete, even for the most carbonated sample. This is supported by the CO2 uptake rates measured after 45 h of carbonation, which showed that carbon dioxide was still being consumed by the samples, albeit very slowly (refer to supporting text). The 14-day cured 0.60 w/c sample showed greater proportions of aragonite to calcite than the other carbonated samples (Figure 3). This is likely due to faster carbonate precipitation reactions, which generally favor the formation 4124
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 20, 2001
of metastable phases such as aragonite (18). Aragonite is expected to convert to the more stable polymorph calcite with time (19). The particular CaCO3 polymorph that forms with carbonation can be an important factor in applying the method for waste immobilization. Due to lattice size restrictions, certain waste elements could be displaced from the structure of metastable calcium carbonate polymorphs during their phase transition to calcite and into the cement pore water, where they may be readily leached. An example of this is the commonly observed expulsion of relatively large Sr2+ ion (1.2 Å) from the aragonite structure, where it is in 9-fold coordination with O2-, during its conversion to calcite, where it would have to be in 6-fold coordination. 4.3. Waste Element Leaching. Changes induced by carbonation can be both physical and chemical, affecting both the pore structure and the element fixation characteristics of the cemented wasteform. On the physical side, precipitation of carbonates in the pore network reduces the leaching of waste elements by providing greater physical impedance to diffusion out of the wasteform. Although no supporting porosimetry analyses were conducted in this study, it is likely that there was a greater reduction in the pore network size and connectivity associated with the carbonation of the 0.45 w/c samples as compared to the 0.6 w/c samples. This would account for the generally lower leachability of the 0.45 w/c samples for all waste components except cesium. On the chemical side, changes in mineralogy and pore solution composition induced by the lower pHs of the carbonation environment can result in changes in the leachability of waste components. Some of these chemical contributions to anion and cation leachability are considered below. Anions. The anion leaching behavior in this study is an example of waste components being removed from a phase of low solubility and redistributed into a phase of high solubility during carbonation. The major anions added to the grout mixes, arsenate, nitrate, and chloride, are likely originally incorporated in the lattice of ettringite or in the interlayer regions of the AFm phase. Their association with these phases is well documented (e.g. refs 20-22). Both ettringite and AFm have low solubility at high pH conditions but are unstable at the low pHs produced by the carbonation reaction and dissolve, releasing their entrained anions. Thus anionic waste components would accumulate in the developing porewater as these phases dissolve during carbonation and transform into calcium carbonate and aluminum oxyhydroxides. As the developing porewater itself becomes “evapo-transported” to the desiccant, these anionic waste components would ultimately precipitate as highly soluble calcium or alkali arsenate, nitrate and chloride salts, which are highly susceptible to leaching by water. This likely explains why all carbonated samples showed much greater leaching of anions than uncarbonated samples. A similar enhanced leaching of nitrate upon carbonation of cement samples was
lower Ca(OH)2 content of the IP cement. Ca(OH)2 undergoes a volume reduction of 20% when it converts to calcite. In contrast, carbonation of CSH produces two solid phases: calcite and silica gel. Thus less carbonation shrinkage may occur, the lower the Ca(OH)2 content of the original hydrated cement.
Acknowledgments Funding in support of this research was provided by the National Science and Engineering Research Council of Canada (NSERC), the Center for Research in Earth and Space Technology (CRESTech), and Kinectrics (formerly Ontario Power Technologies). The authors would like to thank Randy Fagan and Sean Andreou for analytical and laboratory assistance and for reviewing an early draft of the manuscript.
Supporting Information Available
FIGURE 5. Cross-section of a 6-day vacuum carbonation treatment of a Type IP cement mix sample (14-day cure time, w/c ) 0.6) at the end of the experiment showing an average carbonation penetration depth of 15 mm and absence of shrinkage cracks. observed by Gutierrez et al. (8). Thus, carbonation of cemented wasteforms may not be suitable for containment of anionic waste elements unless a cement composition can be designed in some way to sequester the anions released during the carbonation reaction. Cations. Carbonation of cemented wasteforms is efficient in promoting the immobilization of cationic waste elements. There was a significant reduction in the leachability of calcium, strontium, and cesium compared to uncarbonated wasteforms. Strontium leaching was similar but lower than that observed by Gutierrez et al. (8). Cesium, like other alkali elements, should be heavily partitioned into the porewater during the initial setting of the cemented wasteform. Upon carbonation, however, cesium likely precipitates as pollucite, CsAlSi2O6•xH2O, due to the higher silica content of porewater that results from carbonation (23, 24). Pollucite is very insoluble, and its precipitation would account for lower Cs leachability with increased carbonation (see Figure 4). 4.4. Blended Cement. Portlandite, or Ca(OH)2, is a common product of the hydration of OPC. In leaching environments, portlandite is dissolved, leaving a more open porosity within the wasteform, which can lead to even greater leaching. Because of this, many cement applications today use blended cements, i.e., OPC with an added reactive source of silica, such as fly ash, silica fume, or blast furnace slag. These reactive sources of silica provide for the conversion of any formed Ca(OH)2 to CSH (3). To examine the possible difference in style of carbonation of a blended cement compared to OPC, one carbonation experiment was conducted on a sample of a Type IP cement. The cement was prepared by mixing OPC from this study with 20 mass % Class F fly ash. The sample (6 cm length × 5 cm diameter) was prepared at a water/cement of 0.6 and cured at 100% RH for 14 days. The vacuum carbonation treatment was conducted over 6 d, involving just under 800 separate injections of 40 mg of CO2 gas. Each injection caused the cell pressure to increase to 9 kPa, and the next injection was delivered once the pressure fell below 4 kPa. Figure 5 shows a cross-section through the specimen after the 6 d reaction. The depth of carbonation averaged 15 mm, which is similar to that attained in a separate test with a similar-sized OPC sample over the same period. An important feature of the carbonated product from this experiment compared to OPC mixes was that there was no evidence of any shrinkage cracks. This is likely due to the
Experimental details and tables of chemical composition of cemented wasteforms prior to carbonation (S-1) and results of vacuum carbonation experiments (S-2). This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Gougar, M.; Scheetz, B. E.; Roy, B. E. Waste Manage. 1996, 16, 295. (2) Dayal, R.; Klein, R. Evaluation of Cement-Based Backfill Materials; Report No. 87-223-K; Ontario Hydro Research Division: Toronto, Ontario, 1987. (3) Smith, R. W.; Walton, J. C. Mater. Res. Soc. Symp. Proc. 1991, 403. (4) Macias, A.; Kindness, A.; Glasser, F. P. Cement Concr. Res. 1997, 27, 215. (5) Hartmann, T.; Paviet-Hartmann, P.; Rubin, J. B.; Fitzsimmons, M. R.; Sickafus, K. E. Waste Management. 1999, 19, 355. (6) Lange, L. C.; Hills, C. D.; Poole, A. B. J. Hazard. Mater. 1997, 52, 193. (7) Bin-Shafique, M. D. S.; Walton, J. C.; Gutierrez, N.; Smith, R.; Tarquin, A. J. Environ. Eng. 1998, 124, 463. (8) Gutierrez, N.; Bin-Shafique, M. D. S.; Walton, J. C.; Tarquin, A.; Smith, R.; Sheely, P.; Rodriguez, M.; Andrede, R. Proceed. 1996 HSRC/WERC Joint Conference on The Environment; Albuquerque, NM, 1996; May 21-23. (9) Reardon, E. J.; James, B. R.; Abouchar, J. Cement Concr. Res. 1989, 19, 385. (10) Jones, R. Jr. U.S. Patent No. 5,518,540, 1996. (11) Rubin, J. B.; Taylor, C.; Sivils, L. D.; Carey, J. W. 1997 International Ash Utilization Symposium; Lexington, KY, October 20-22, 1997. (12) Cole, W. F.; Kroone, B. J. Am. Conc. Inst. 1960, 31, 1275. (13) Venuat, M.; Alexandre, J. Rev. Mater. Const. 1968, 638, 421. (14) Venhuis, M. A. M. Sc. Thesis, University of Waterloo, 2000. (15) Measurement of the Leachability of Solidified Low-Level Radioactive Wastes Short-term Procedure ANSI/ANS 16.1; American Nuclear Society: Chicago, IL, 1986. (16) Pitts, J. Concrete 1987, 21, 5. (17) Walton, J. C.; Bin-Shafique, S.; Smith, R. W.; Gutierrez, N.; Tarquin, A. Environ. Sci. Technol. 1997, 31, 2345. (18) Brecevic, L.; Nielsen, A. E. J. Crystal Growth. 1989, 98, 504-510. (19) Richardson, M. G. Carbonation of Reinforced Concrete; CITIS: Dublin, Ireland, 1988; p 15. (20) Kumarathasan, P.; McCarthy, G. J.; Hassett, D. J.; PflughoeftHassett, D. F. Mater. Res. Soc. Symp. Proc. 1990, 178, 83. (21) Myneni, S. C. B.; Traina, S. J.; Logan, T. J.; Waychunas, G. A. Environ. Sci. Technol. 1997, 31, 1760. (22) Perkins, R. B.; Palmer, C. D. Appl. Geochem. 2000, 15, 1203. (23) Reardon, E. J.; Dewaele, P. J. J. Am. Ceram. Soc. 1990, 1681. (24) Hoyle, S.; Grutzeck M. W. J. Am. Ceram. Soc. 1989, 72, 1938.
Received for review January 8, 2001. Revised manuscript received June 15, 2001. Accepted July 5, 2001. ES0105156 VOL. 35, NO. 20, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4125
