Monitoring of Stainless-Steel Slag Carbonation Using X-ray Computed

Nov 27, 2013 - ACS eBooks; C&EN Global Enterprise ... Air Pollution and Industrial Hygiene · Apparatus and Plant Equipment · Cement, Concrete, and Rel...
5 downloads 0 Views 5MB Size
Download PDF
Article pubs.acs.org/est

Monitoring of Stainless-Steel Slag Carbonation Using X‑ray Computed Microtomography Marijn A. Boone,*,†,‡ Peter Nielsen,† Tim De Kock,‡ Matthieu N. Boone,§ Mieke Quaghebeur,† and Veerle Cnudde‡ †

Unit Sustainable Materials Management, VITO, Mol 2400, Belgium Department of Geology and Soil Science, and §Department of Physics and Astronomy, Ghent University, Gent 9000, Belgium

‡

ABSTRACT: Steel production is one of the largest contributors to industrial CO2 emissions. This industry also generates large amounts of solid byproducts, such as slag and sludge. In this study, fine grained stainless-steel slag (SSS) is valorized to produce compacts with high compressive strength without the use of a hydraulic binder. This carbonation process is investigated on a pore-scale level to identify how the mineral phases in the SSS react with CO2, where carbonates are formed, and what the impact of these changes is on the pore network of the carbonated SSS compact. In addition to conventional research techniques, high-resolution X-ray computed tomography (HRXCT) is applied to visualize and quantify the changes in situ during the carbonation process. The results show that carbonates mainly precipitate at grain contacts and in capillary pores and this precipitation has little effect on the connectivity of the pore space. This paper also demonstrates the use of a custom-designed polymer reaction cell that allows in situ HRXCT analysis of the carbonation process. This shows the distribution and influence of water and CO2 in the pore network on the carbonate precipitation and, thus, the influence on the compressive strength development of the waste material.

■

INTRODUCTION Anthropogenic CO2 emissions into the atmosphere have been shown to contribute directly to global warming.1,2 In 2006, over one-third of the carbon emissions came from industry, of which more than half can be attributed to the production of steel, cement, plastic, paper, and aluminum.3 Steel production, the largest contributor to industrial CO2 emissions, also generates large amounts of solid byproducts in the form of slag and sludge.4 Generally, these byproducts are considered as environmentally non-hazardous waste materials. The most conventional disposal method is landfilling,5 although steel slags can also be used as material for cement production, road construction, civil engineering work, fertilizers, etc.6,7 Steel slag has also been considered as an alternative mineral resource for ex situ carbonation and can be considered a stable and safe option for CO2 sequestration.8 Using steel slag as a feedstock for mineral carbonation has some advantages over primary Ca/ Mg silicate ores, which are more commonly proposed as source material for mineral sequestration. Steel slag is readily available near sources of CO2 and is more susceptible to carbonation because of its chemical instability.9 Furthermore, carbonation of these materials can reduce the leaching of potentially hazardous metals and, thus, improve the environmental quality.10−12,4 This is especially important for stainless-steel slag (SSS), because these could potentially be considered as a hazardous waste material as a result of the noticeable amount of Cr present, which can potentially leach into the environment.7,13 Apart from CO2 sequestration and mineral stabilization, the formation of carbonates by accelerated carbonation of steel slag © XXXX American Chemical Society

can also serve as a consolidation for the slag material. Johnson et al.14 produced compacts from SSS and CO2 at a pressure of 3 bar with a compressive strength up to 9 MPa without the addition of a cementitious binder. Quaghebeur et al.15 produced compacts with compressive strengths up to 55 MPa from SSS and CO2 at a moderate pressure (20 bar) and temperature (140 °C). The carbonation of the compacts resulted in a reduction of the leachability of chromium and molybdenum and complied with the limit values for the reuse of waste in Belgium and Netherlands. These studies show that valorization of fine-grained steel slag is possible without the use of a hydraulic binder. The processes that determine how and where carbonate is formed and how the strength of the compact is increased are however less clear. This paper therefore focuses on the carbonation process on a pore-scale level to identify how the mineral phases in the SSS react with CO2, where carbonates are formed, and what the impact of these changes is on the pore network of the carbonated SSS compact. In addition to conventionally used research techniques, high-resolution X-ray computed tomography (HRXCT) is applied to visualize and quantify the changes that occur throughout the carbonation process. Received: July 2, 2013 Revised: November 26, 2013 Accepted: November 27, 2013

A

dx.doi.org/10.1021/es402767q | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

■

Article

MATERIALS AND METHODS Materials. Fine-grained SSS, supplied by the Belgian waste processing company Recmix, was dried at a temperature of 40 °C and sieved into five different grain-size classes: 500 μm. After sieving, 35% of grain-size class 25−100 μm was mixed with 35% of grain-size class 100−200 μm and 30% of grain-size class 200−500 μm. Afterward, 10 wt % demineralized water was added, and the mixture was compacted at a pressure of 350 kg/ cm2. Two types of compacts were generated using, on one hand, a mold of 42 mm diameter for conventional analysis techniques and, on the other hand, a smaller mold with a diameter of 6 mm for visualizing and analyzing the carbonation process in situ using HRXCT. Carbonation Experiments. The compacts with a diameter of 42 mm were carbonated in an automated carbonation unit consisting of a stainless-steel autoclave (Premex). The autoclave was heated to a constant temperature of 80 °C and pressurized with CO2 at a pressure of 20 bar. The CO2 pressure was kept constant by replenishment of the consumed CO2. Compacted SSS samples were carbonated at different time periods, 0.5, 2, and 8 h, to monitor the progress of the carbonation reaction. The 6 mm compacts were carbonated in a polymer reactor cell at a CO2 pressure of 20 bar and a temperature of 50 °C. This polymer cell is custom-designed and made to fit the HRXCT scanner at the Centre for X-ray Tomography of Ghent University (UGCT). With this polymer cell, three-dimensional (3D) images of a sample in situ, at pressures up to 120 bar and temperatures up to 60 °C, can be obtained in a non-destructive manner. Conventional Analysis. The chemical composition of the different grain sizes of the fine-grained steel slag before carbonation was determined with energy-dispersive X-ray fluorescence analysis (ED-XRF) under a He atmosphere. The mineralogy of the fine-grained steel slag and the carbonated 6 mm samples was determined using a X’Pert Pro MPD diffractometer (XRD) from Phillips. Uniaxial compression tests were performed to determine the compressive strength of the carbonated samples. The total amount of carbonate formed was determined by analyzing the weight loss because of the CO2 release with a Netzsch STA 449 C thermogravimetric analysis (TGA) system. Analyses were conducted over a temperature range of 25−1000 °C with a constant heating range of 10 °C/min. The structure and composition of the carbonate phase was visualized using a scanning electron microscope (SEM, JEOL JSM-6340F) in combination with energy-dispersive X-ray spectroscopy (EDS) measurements. HRXCT Analysis. HRXCT is applied to visualize the internal structure of the SSS in a non-destructive manner. Regular CT scanning was performed on subsamples taken from a non-carbonated and an 8 h carbonated 42 mm diameter sample. In situ scanning through time is applied on the 6 mm diameter compact to monitor the changes of the internal structure throughout the carbonation process. In total, four CT scans in the reactor cell were obtained at different points in time. Initially, the sample was scanned in the cell at ambient conditions, after which the cell was heated to 50 °C and CO2 was gradually added to the cell until a pressure of 20 bar was attained. The second and third scans were performed during a pCO2 of 20 bar and a temperature 50 °C after 0.5 and 8 h of exposure, respectively. After the third scan, the reactor cell was cooled and depressurized to ambient conditions and the sample

was dried. After drying, the fourth scan was performed. Figure 1 shows the schematic overview of the scanner setup in combination with the custom-designed polymer reactor cell. The temporal resolution per scan was just below 1 h.

Figure 1. Schematic overview of the polymer reactor cell in the HRXCT setup.

HRXCT was performed at the UGCT using a custom-built CT scanner setup with a FeinFocus FXE160.51 transmissiontype X-ray tube combined with a Varian PaxScan 2520 V a-Si flat panel detector for image acquisition.16 The X-ray source was operated at 120 kV with an effective target power of 9 W. Beam hardening artifacts were diminished using a 1 mm thick aluminum filter. The tomographic data sets were reconstructed using the Octopus software package.17 The reconstructed 3D volumes of the sample in the polymer reactor were matched using the visualization software package VGStudio MAX, to obtain an overview of the temporal and spatial changes in the carbonating sample. The software package Morpho+ is used for further analysis on the reconstructed 3D image to obtain quantitative information about the porosity, pore size distribution, and grain abundance. Detailed information about Morpho+ and the workflow thereof is described by Brabant et al.18 X-ray CT images provide information on the linear attenuation coefficient of the different phases in the material under investigation.19 This linear attenuation coefficient is a function of the density and the atomic number of the different elements in the sample and, therefore, gives no direct chemical information of the material.

■

RESULTS AND DISCUSSION Conventional Analysis. The SSS compacts with a diameter of 42 mm were exposed to a pCO2 of 20 bar and a temperature of 80 °C for different time intervals: 0.5, 2, and 8 h. The results from the uniaxial compression test (figure 2) show the increase in strength as a function of the carbonation time. Before carbonation, the average strength of the compacts was 1.8 ± 0.4 MPa. After 0.5, 2, and 8 h of exposure, the samples have an average strength of 39 ± 4, 52 ± 1, and 55 ± 4 MPa, respectively. Initially, there is a fast increase in strength. After 2 h of carbonation, the increase in strength is less pronounced. The results furthermore show that, in merely 2 h of B

dx.doi.org/10.1021/es402767q | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

hand, has appeared, a phase which could not clearly be identified before carbonation. The position of the calcite diffraction peak at 29.66 (2θ) and superimposed peaks at 2θ values of 29.78 and 29.86 indicate that elements, such as Mg, are incorporated in the calcite lattice. On the basis of the position of these calcite diffraction peak(s), the Mg incorporation can be estimated to be between 8 and 15 mol % MgCO3.20−22 TGA (Figure 2) of the SSS before carbonation indicated a loss in mass of 0.2 wt % around 430 °C and a second loss in mass of 0.5 wt % between 550 and 800 °C. These losses correspond to 0.7 wt % portlandite and 1.2 wt % calcite, respectively. After 0.5 h, all portlandite in the material has been carbonated and 8 wt % calcite is present. After 2 and 8 h of carbonation, there is 10 and 13 wt % calcite present, respectively. TGA shows an initial rapid decrease of portlandite and a slower increase of calcite with progressing carbonation, data which is consistent with the XRD analysis. In Figure 2, the increase in strength matches well with the amount of CO2 lost from the material. After 2 h of carbonation, a compact is created with a strength higher than 50 MPa and 44 g of CO2 is bound in carbonate per kg of SSS. After 8 h, the compressive strength of the compacts increases even further and 59 g of CO2 is bound in carbonate per kg of SSS. From the previous measurements, it is clear that the carbonate phases formed because pressurized CO2 exposure is responsible for the strength increase of the SSS compacts. Structural information about the formed carbonate and the location of the carbonate are determined using SEM. Figure 4 represents an internal surface of a SSS compact carbonated for a period of 8 h. SEM imaging showed the complexity in mineral constitution in the SSS compacts. SEM−EDS point measurements confirm this mineral variety between different grains and even within some individual grains. In Figure 4, a detail of a pore is given with mappings for calcium (B, red), silicon (C, blue), and magnesium (D, white). The combination of these elemental maps illustrates a carbonate phase that mainly consists of calcium with incorporation of magnesium. The carbonate crystals are preferentially located at the grain contacts

Figure 2. Evolution of compressive strength and amount of carbonate as a function of the carbonation time.

carbonation, samples with strengths above 50 MPa can be generated, making this material suitable for the construction industry in terms of strength. ED-XRF elemental analysis of the SSS before carbonation shows that the most abundant elements present are calcium (33 wt %), silicon (16 wt %), and magnesium (6.6 wt %). Other common elements are Cr (2.2 wt %), Al (1.7 wt %), Fe (1.0 wt %), Mn (1.0 wt %), Ti (0.5 wt %), and P (0.25 wt %). The XRD analysis in Figure 3 on SSS from three different grain-size classes resulted in seven clearly distinguishable crystalline mineral phases: merwinite [Ca3Mg(SiO4)2], bredigite [Ca14Mg2(SiO4)8], portlandite [Ca(OH)2], cuspidine (Ca 4Si 2 O 7F 2 ), akermanite−gehlenite [Ca 2 Mg(Si2O7)−Ca2Al(AlSiO7)], periclase (MgO), and magnesium− iron−chromium oxide [(Mg,Fe)Cr2O4]. Although all mentioned mineral phases are present in every grain size fraction, portlandite and periclase are more abundant in the fine-grained fraction (25−100 μm). Bredigite, on the other hand, was enriched in the coarser fractions (100−500 μm). When the XRD data are compared before and after 8 h of carbonation, the portlandite phase is no longer present and the intensity of the diffraction peaks of mineral phases, such as merwinite, cuspidine, bredegite, periclase, and akermanite− gehlenite, have substantially decreased. Calcite, on the other

Figure 3. XRD analysis on SSS before and after carbonation (Po, portlandite; M, merwinite; Ca, calcite; Cr, Mg/Fe/Cr oxide; G, akermanite/ gehlenite; Cu, cuspidine; Pe, periclase; and B, bredigite). C

dx.doi.org/10.1021/es402767q | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

gray values and the high attenuating phases are represented by more white values. The pore space, the phase with almost no attenuation, is represented in dark gray to black. The mineral grains are represented by a wide range of bright gray values because of their variety in elements and densities, representing the complexity of the minerals present in the SSS. The 2D slice in Figure 5a represents the non-carbonated sample. The carbonated sample in Figure 5b shows the presence of a carbonate phase located between the grains and in the smaller pores, similar to the SEM image (Figure 4). The scanner setup and reconstruction parameters were identical to obtain the HRXCT images of the non-carbonated and carbonated samples. For the 3D image analysis on these images, the same attenuation interval was chosen to compare the porosity in the different scans. The main limitation of the porosity determination on the HRXCT images in comparison to other techniques is however the resolution of the scanned sample. The porosity below the resolution of 2.9 μm in the HRXCT image cannot be distinguished, which results in an underestimation of the porosity with respect to other techniques, such as mercury intrusion porosimetry.27 The total porosity before carbonation, determined with 3D image analysis, is 21.7%. The closed porosity is only 0.3%, showing that almost all of the pores are connected. The 8 h carbonated sample has a total porosity of 16.7%, of which 0.5% is calculated as closed porosity. The larger part of the porosity in the samples is open porosity, and only a few pores are closed. However, it must be stressed that, in reality, these pores can still be connected with pore throats below the resolution of the scan, i.e., below 2.9 μm. A carbonation process of 8 h of SSS compacts with a grain-size range from 25 to 500 μm causes a limited decrease of the pore space of 5% of the compact and significantly increases the compressive strength of that compact. Furthermore, the decrease does not cause a closing or clogging of the pore system. The change in porosity is consistent with TGA in Figure 2, where an additional 12 wt % carbonate is observed after 8 h of carbonation. The pore size distribution is obtained by digitally dividing the pore network into individual pore bodies using a watershedbased separation. After separation, individual pores are described using their maximum opening, i.e., the diameter of the biggest sphere that fits inside the pore. In Figure 6, the pore size distribution in terms of maximum opening is given for the non-carbonated and carbonated compacts. The graph shows that the amount of pores with a maximum opening larger than 50 μm is the same before and after carbonation. The amount of pores with a maximum opening between 50 and 18 μm decreases after carbonation; however, the amount of pores with a maximum opening below 18 μm increases. The volume of these pores as a function of their maximum opening shows that the porosity change between the carbonated and noncarbonated samples is occurring in the pores with a maximum opening below 50 μm. These results show that the carbonation in the SSS compacts has little effect on the larger pores. Carbonates precipitate in the small pores with a maximum opening below 50 μm, causing a decrease of the amount of these pores and an increase in smaller pores because of incomplete filling of the pore by the precipitation. The previous measurements show that carbonation causes a drop in porosity mainly in the intermediate pores, without closing of the pore network. However, to clearly quantify the change in the pore network of the SSS compact and determine how the individual mineral grains react during the carbonation

Figure 4. (A) SEM image of the internal surface of a compact carbonated for 8 h and SEM−EDS mappings of (B, red) Ca, (C, blue) Si, and (D, white) Mg.

and to a lesser extent at the pore walls. The carbonate phase itself is not texturally homogeneous and is quite rich in small pores. In many cases, calcium (and magnesium) have been dissolved from the outer layer of the silicate minerals, leaving a cation-depleted silica-rich layer behind, indicating an incongruent dissolution of the silicate minerals. The XRD data show no increase in any crystalline silicon phase, indicating that these silica-rich zones on the original silicate minerals most likely consist of amorphous silica. The reaction process is comparable to the aqueous carbonation reaction observed for wollastonite (CaSiO3).23−26 However, from the results, it is unclear if the formed silica layer is formed by an interfacecoupled dissolution precipitation model25 or a preferential leaching of cations.26 It must be noted that, in some localized pores, zones of precipitated amorphous silica are interlaced with the carbonate, indicating that some locations in the pore network have other chemical conditions that allow for silica dissolution and precipitation. HRXCT Analysis. HRXCT is used to obtain information about the porosity, pore connectivity, and pore size distribution of the SSS before and after carbonation. The subsamples of the 42 mm diameter compacts were scanned, and the resulting 3D images had an isotropic voxel size of 2.9 μm3. A detail of a twodimensional (2D) slice through the scanned non-carbonated and carbonated samples is given in Figure 5. In these internal 2D slices, the low attenuating phases are represented by dark

Figure 5. Detail of a 2D slice through the scanned (a) non-carbonated and (b) carbonated SSS. D

dx.doi.org/10.1021/es402767q | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

the 42 mm compacts, and the amount of pores with a maximum opening below 18 μm increases again. The graph of the amount of pores during the carbonation process is quite different from the other graphs below a maximum opening of 50 μm. The amount of smaller pores decreases more drastically in the interval between 50 and 18 μm, and there is no increase in the amount of very small pores below a maximum opening of 18 μm. The change in porosity occurs mainly in the pores with a maximum opening below 60 μm, especially during the carbonation reaction. The apparent difference in porosity and pore size distribution between the scan during carbonation and the scan after carbonation and drying is most likely caused by the presence of water saturated with ions. The presence of water in the pores increases the attenuation value of those pores, making discrimination with the surrounding material sometimes difficult. The water-filled pores can therefore be allocated as material in the CT image, lowering the porosity in the scanned image. In the scan during exposure, this underestimation of the pore space is present in the smallest pores. This indicates that the water is limited to the smaller pores and capillaries, while the larger pores are filled with the CO2 gas phase. From these results, it can be concluded that carbonate precipitation mainly occurs in the smaller pores and at the grain contacts where water is present, while the reaction in the large pores is limited. Apart from precipitation of carbonate phases in the smaller pores and capillaries, dissolution also occurs during the carbonation process. To assess the amount of dissolution, the CT images before and after carbonation were matched and divided, highlighting the differences. A 2D slice through the divided 3D images is represented in Figure 7. The black spots

Figure 6. Pore size distribution before and after carbonation in the 42 mm diameter compacts and the pore size distribution before, during, and after carbonation in the 6 mm diameter compacts.

process, it is crucial to monitor the same sample during the carbonation process. Therefore, smaller 6 mm diameter SSS compacts are carbonated in a polymer reaction cell, and the internal changes are monitored through time using HRXCT. The porosity determined by 3D image analysis of the samples before CO2 exposure and at ambient conditions was 20.7%, with 20.4% open porosity. After 8 h of exposure and temperature and pressure conditions of 50 °C and 20 bar, the porosity drops to a value of 16.2%, with an open porosity of 16.0%. The open porosity in the sample after carbonation at ambient conditions and after drying of the sample is 18.0% with a total porosity of 18.5%. The values of the 6 mm diameter samples differ slightly from the subsamples taken from the 42 mm compacts. The initial porosity before carbonation is 1% lower in the 6 mm sample, and the porosity after carbonation is 1.3% higher in the 6 mm sample. Although the measurements in the 42 mm subsamples were performed on two different samples, the porosity change because of carbonation is less pronounced in the 6 mm samples. This effect is most likely caused by the difference in the temperature in the two experiments, because of limitations in the temperature of the reactor cell for in situ HRXCT analysis. A temperature difference of 30 °C has an impact on the dissolution of Ca and Mg from the silicates and, therefore, on the progression of the carbonation reaction. In reality, the temperature difference between the small and larger compacts will be more pronounced because of the exothermic nature of the carbonation reaction, causing an even higher temperature in the 42 mm diameter compacts. Figure 6 represents digitally determined pore size distribution in the 6 mm diameter samples, where the individual pores are described using their maximum opening. The graph indicates that the larger pores with a maximum opening larger than 50 μm remain similar before, during, and after carbonation. The smaller pores however change significantly throughout the carbonation process. In the graphs of the 6 mm diameter compact before and after carbonation, a trend similar to the 42 mm samples is present. The amount of pores with a maximum opening between 50 and 18 μm is diminishes after carbonation and consistent with the data from the CT scans on

Figure 7. Two-dimensional slice of a matched and divided 3D image, with dissolution spots in black (white arrow) and precipitation in bright white.

represent the zones where dissolution occurred, whereas the white zones represent the newly formed precipitates because of carbonation. The amount of dissolved phases represents about 1% of the total volume of the sample, which is however a considerable part of the difference in porosity between the scan before and after carbonation. The individual dissolution zones were compared to the 3D CT image of the sample before carbonation to analyze which grains or which type of material had been dissolved. This showed that the dissolved zones do not represent entire grains but only parts of the SSS grains. This indicates that certain parts of the SSS grains preferentially dissolve, which is most likely related to the multimineralic E

dx.doi.org/10.1021/es402767q | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

Climate Change; Core Writing Team, Pachauri, R. K., Reisinger, A., Eds.; IPCC: Geneva, Switzerland, 2007; p 104. (3) Allwood, J. M.; Cullen, J. M.; Milford, R. L. Options for achieving a 50% cut in industrial carbon emissions by 2050. Environ. Sci. Technol. 2010, 44 (6), 1888−1894. (4) Navarro, C.; Diaz, M.; Villa-Garcia, M. A. Physico-chemical characterization of steel slag. Study of its behavior under simulated environmental conditions. Environ. Sci. Technol. 2010, 44 (14), 5383− 5388. (5) Proctor, D. M.; Fehling, K. A.; Shay, E. C.; Wittenborn, J. L.; Green, J. J.; Avent, C.; Bigham, R. D.; Connolly, M.; Lee, B.; Shepker, T. O.; Zak, M. A. Physical and chemical characteristics of blast furnace, basic oxygen furnace, and electric arc furnace steel industry slags. Environ. Sci. Technol. 2000, 34 (8), 1576−1582. (6) Geiseler, J. Use of steelworks slag in Europe. Waste Manage. 1996, 16 (1−3), 59−63. (7) Shen, H. T.; Forssberg, E. An overview of recovery of metals from slags. Waste Manage. 2003, 23 (10), 933−949. (8) Gerdemann, S. J.; O’Connor, W. K.; Dahlin, D. C.; Penner, L. R.; Rush, H. Ex situ aqueous mineral carbonation. Environ. Sci. Technol. 2007, 41 (7), 2587−2593. (9) Huijgen, W. J. J.; Witkamp, G. J.; Comans, R. N. J. Mineral CO2 sequestration by steel slag carbonation. Environ. Sci. Technol. 2005, 39 (24), 9676−9682. (10) Reddy, K. J.; Gloss, S. P.; Wang, L. Reaction of CO2 with alkaline solid wastes to reduce contaminant mobility. Water Res. 1994, 28 (6), 1377−1382. (11) Meima, J. A.; van der Weijden, R. D.; Eighmy, T. T.; Comans, R. N. J. Carbonation processes in municipal solid waste incinerator bottom ash and their effect on the leaching of copper and molybdenum. Appl. Geochem. 2002, 17 (12), 1503−1513. (12) Huijgen, W. J. J.; Comans, R. N. J. Carbonation of steel slag for CO2 sequestration: Leaching of products and reaction mechanisms. Environ. Sci. Technol. 2006, 40 (8), 2790−2796. (13) Baciocchi, R.; Costa, G.; Di Bartolomeo, E.; Polettini, A.; Pomi, R. Carbonation of stainless steel slag as a process for CO2 storage and slag valorization. Waste Biomass Valorization 2010, 1 (4), 467−477. (14) Johnson, D. C.; MacLeod, C. L.; Carey, P. J.; Hills, C. D. Solidification of stainless steel slag by accelerated carbonation. Environ. Technol. 2003, 24 (6), 671−678. (15) Quaghebeur, M.; Nielsen, P.; Laenen, B.; Nguyen, E.; Van Mechelen, D. Carbstone: Sustainable valorisation technology for fine grained steel slags and CO2. Refract. Worldforum 2010, 2, 75−79. (16) Masschaele, B.; Cnudde, V.; Dierick, M.; Jacobs, P.; Van Hoorebeke, L.; Vlassenbroeck, J. UGCT: New X-ray radiography and tomography facility. Nucl. Instrum. Methods Phys. Res., Sect. A 2007, 580 (1), 266−269. (17) Vlassenbroeck, J.; Dierick, M.; Masschaele, B.; Cnudde, V.; Hoorebeke, L.; Jacobs, P. Software tools for quantification of X-ray microtomography. Nucl. Instrum. Methods Phys. Res., Sect. A 2007, 580 (1), 442−445. (18) Brabant, L.; Vlassenbroeck, J.; De Witte, Y.; Cnudde, V.; Boone, M. N.; Dewanckele, J.; Van Hoorebeke, L. Three-dimensional analysis of high-resolution X-ray computed tomography data with Morpho+. Microsc. Microanal. 2011, 17 (2), 252−263. (19) Cnudde, V.; Boone, M. N. High-resolution X-ray computed tomography in geosciences: A review of the current technology and applications. Earth-Sci. Rev. 2013, 123, 1−17. (20) Goldsmith, J. R.; Graf, D. L. Relation between lattice constants and composition of the Ca−Mg carbonate. Am. Mineral. 1958, 43, 84− 101. (21) Goldsmith, J. R.; Graf, D. L.; Heard, H. C. Lattice constants of the calcium−magnesium carbonates. Am. Mineral. 1961, 46, 453−459. (22) Milliman, J. D.; Gastner, M.; Muller, J. Utilization of magnesium in coralline algae. Geol. Soc. Am. Bull. 1971, 82, 573−580. (23) Huijgen, W. J. J.; Witkamp, G. J.; Comans, R. N. J. Mechanisms of aqueous wollastonite carbonation as a possible CO2 sequestration process. Chem. Eng. Sci. 2006, 61 (13), 4242−4251.

nature of the SSS grains determined in the SEM images. When the differential image is compared to the 3D CT image after 0.5 h of carbonation, the parts of the SSS grains that are dissolved after 8 h of carbonation are still present. Dissolution is therefore still taking place after 0.5 h of carbonation in selective zones in the compacted SSS, even though the first carbonates have already formed near the grain contacts. The dissolved minerals have an intermediate gray value in the CT image, which makes it difficult to deduce their mineral composition. From the TGA measurements, it is clear that portlandite has already reacted 0.5 h in the carbonation process. The SEM images demonstrated an incongruent dissolution of Ca and Mg from the silicates, indicating that the completely dissolved phases are most likely (hydrated) magnesium oxide. XRD analysis in Figure 3 shows that the intensity of the diffraction peak of MgO has decreased after 8 h of carbonation. When the XRD, TGA, and HRXCT data are combined, it is clear that, at the used reaction conditions, calcium hydroxide reacts fast, giving an initial strength to the SSS compact, while CaMg silicates react much slower through time. The dissolved calcium and part of the silicon from the silicates will precipitate in the pore network of the SSS, further strengthening the compact with limited influence on the permeability of the pore network. This paper shows the strength development because of carbonation in SSS waste compacts on a pore-scale level. Detailed analysis using conventional analytical techniques in combination with HRXCT has shown that the carbonates formed during carbonation mainly precipitate at the grain contacts and in the capillary pores and that this carbonate precipitation has little effect on the connectivity of the pore spaces. This paper also demonstrates the use of a customdesigned polymer reaction cell that allows for in situ HRXCT analysis of the carbonation process under a CO2 pressure of 20 bar and through time. In situ HRXCT analysis has made it possible to deduce the distribution of water and CO2 in the pore network of the compact and has shown where and how carbonate precipitation occurs during carbonation.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Bo Peeraer, Raymond Kemps, Annemie De Wilde, and Myrjam Mertens for the assistance during this work and the company Recmix for supplying the SSS. This research was funded by a VITO Ph.D. grant.

■

NOMENCLATURE SSS stainless-steel slag HRXCT high-resolution X-ray computed tomography UGCT Centre for X-ray Tomography of Ghent University

■

REFERENCES

(1) Intergovernmental Panel on Climate Change (IPCC). IPCC Special Report on Carbon Dioxide Capture and Storage. Prepared by Working Group III of the Intergovernmental Panel on Climate Change; Cambridge University Press: New York, 2005. (2) Intergovernmental Panel on Climate Change (IPCC). Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on F

dx.doi.org/10.1021/es402767q | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

(24) Daval, D.; Martinez, I.; Corvisier, J.; Findling, N.; Goffé, B.; Guyot, F. Carbonation of Ca-bearing silicates, the case of wollastonite: Experimental investigations and kinetic modeling. Chem. Geol. 2009, 265, 63−78. (25) Ruiz-Agudo, E.; Putnis, C. V.; Rodriguez-Navarro, C.; Putnis, A. Mechanism of leached layer formation during chemical weathering of silicate minerals. Geology 2012, 40 (10), 947−950. (26) Schott, J.; Pokrovsky, O. S.; Spalla, O.; Devreux, F.; Gloter, A.; Mielczarski, J. A. Formation, growth and transformation of leached layers during silicate minerals dissolution: The example of wollastonite. Geochim. Cosmochim. Acta 2012, 98, 259−281. (27) Cnudde, V.; Cwirzen, A.; Masschaele, B.; Jacobs, P. Porosity and microstructure characterization of building stones and concretes. Eng. Geol. 2009, 103 (3−4), 76−83.

G

dx.doi.org/10.1021/es402767q | Environ. Sci. Technol. XXXX, XXX, XXX−XXX