- Email: [email protected]
Precipitation of calcium carbonate during direct aqueous carbonation of flue gas desulfurization gypsum
Chemical Engineering Journal 213 (2012) 251–258
Contents lists available at SciVerse ScienceDirect
Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej
Precipitation of calcium carbonate during direct aqueous carbonation of flue gas desulfurization gypsum Kyungsun Song a,b,1, Young-Nam Jang a,1, Wonbaek Kim a,⇑, Myung Gyu Lee c, Dongbok Shin b, Jun-Hwan Bang a, Chi Wan Jeon a, Soo Chun Chae a a b c
Korea Institute of Geoscience & Mineral Resources (KIGAM), Gwahang-no 124, Yuseong-gu, Daejeon 305-350, Republic of Korea Department of Geoenvironmental Sciences, Kongju National University, Kongju, Chungnam 314-701, Republic of Korea Resources Recycling Engineering Department, University of Science and Technology of Korea, Gwahang-no 113, Yuseong-gu, Daejeon 305-333, Republic of Korea
h i g h l i g h t s " Direct carbonation of FGD gypsum was studied. " Direct carbonation produced a mixture of vaterite and calcite. " Vaterite/calcite increased with carbonation time. " Impurity-free calcite was obtained in direct carbonation.
a r t i c l e
i n f o
Article history: Received 26 July 2012 Received in revised form 16 October 2012 Accepted 18 October 2012 Available online 25 October 2012 Keywords: Flue gas desulfurization gypsum Calcium carbonate Calcite Vaterite Carbonation
a b s t r a c t The precipitation of calcium carbonate during the direct aqueous carbonation of flue gas desulfurization (FGD) gypsum, an industrial waste product, was investigated. For comparison, two-step carbonation was also attempted. Calcite was the dominating phase produced in the two-step carbonation, while the direct carbonation produced a mixture of vaterite and calcite phases under all conditions. The relative amounts of each phase were determined by comparing their X-ray diffraction peak intensities. The amount of vaterite phase in the mixture was found to increase with carbonation time for a fixed CO2 flow rate. An induction period before the precipitation of physically detectable calcium carbonate crystals was assessed using X-ray photoelectron spectroscopy (XPS) and field-emission scanning electron microscopy (FE-SEM). For the CO2 flow rate of 1 L/min, the C1s peak of CaCO3 became clear after carbonation took place for longer than 15 min. Virtually impurity-free, single-phase calcite crystals were precipitated from the solution extracted during the induction period. The amount of calcite obtained was typically 5.3% of the amount that can be obtained in the case of the complete reaction. However, further elaboration of this method may allow for the preparation of pure calcite crystals from industrial by-product FGD gypsum. Ó 2012 Elsevier B.V. All rights reserved.
1. Introduction
CO2 ðgÞ þ 2NH4 OH ! ðNH4 Þ2 CO3 ðaqÞ
Flue gas desulfurization (FGD) gypsum is the main product of the desulfurization system used to remove SOx from coal combustion products. One established utilization of waste gypsum has been the production of the fertilizer ammonium sulfate [1–3]. At the same time, calcium carbonate is produced as a byproduct which might a have commercial value depending on the purity, morphology, etc. Its application ranges from the conventional fields such as paper or cosmetics to drug industries [4]. The process conventionally consists of the two following consecutive reactions:
CaSO4 2H2 O þ ðNH4 Þ2 CO3 ðaqÞ
⇑ Corresponding author. Tel.: +82 42 868 3623. 1
E-mail address: [email protected] (W. Kim). These authors equally contributed to this study.
1385-8947/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cej.2012.10.010
! CaCO3 ðsÞ þ ðNH4 Þ2 SO4 þ 2H2 O
ð1Þ
ð2Þ
The above reactions are characterized by the high yield and purity (up to 99%) of the ammonium sulfate produced [2]. Burnett et al. [3] clearly demonstrated that the purity of the ammonium sulfate could be guaranteed because most contaminants of FGD gypsum remained in the solid residue. If impurities exist in solid form, they can be easily separated from the ammonium sulfate solution by filtration. Considering the solubilities of ammonium sulfate and calcium carbonate, the former, with high solubility (74 g/100 mL of H2O at 20 °C), should exist in solution and most of the comparatively insoluble latter (0.15 g/100 mL H2O at
252
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
25 °C) should precipitate [5]. Thus, pure ammonium sulfate crystals can be precipitated after removing the insoluble impurities by filtration. On the other hand, the solubility of calcium carbonate is low and it cannot be purified in a similar manner. Recently, we evaluated the feasibility of a single-step direct carbonation reaction of FGD gypsum [6]. The process can be expressed as:
was used for the determination of trace elements following acid digestion (HNO3/HClO4). Particle size distribution was measured using a laser scattering particle size analyzer (HELOS/RODOS & SUCELL, Sympatec GmbH). The elemental mapping was carried out using an electron probe microanalyzer (EPMA, EPMA-1600, SHIMADZU).
CaSO4 2H2 O þ 2NH4 OHðaqÞ þ CO2 ðgÞ
2.2. Carbonation experiments
! CaCO3 ðsÞ þ ðNH4 Þ2 SO4 þ 2H2 O
ð3Þ
The reaction (1) is exothermic and the ammonia is usually required to be chilled below 20 °C to reduce the vaporization of ammonia [7]. Therefore, a single-step direct carbonation of FGD gypsum would have an obvious advantage over the conventional two-step process. The effects of the ammonia content, CO2 flow rate, solid-to-solution ratio, and CO2/N2 gas mixture ratio were analyzed by systematically monitoring the temperature, pH of the slurry, and the carbonation conversion rate of products [6]. As expected, comparatively pure ammonium sulfate could be obtained in the process. In the meantime, the calcium carbonate that precipitated in the direct carbonation process always produced a mixture of the thermodynamically unstable vaterite and the stable calcite phases under all conditions. In this paper, we focus on the precipitation of calcium carbonate and its polymorphism. Additionally, our attempts to produce calcium carbonate of high purity from FGD gypsum in a manner similar to that used for the production of ammonium sulfate are discussed. The purity, phase, and other physical characteristics of the calcium carbonate produced were studied using a variety of spectroscopic and microscopy techniques. A comparative study of calcium carbonate produced by the conventional two-step process is also presented. 2. Experimental 2.1. Materials FGD gypsum was obtained from Yeongheung Thermal Power Plants, Incheon, Korea and dried at 45 °C overnight to remove the surface water. Wet chemical analysis was performed for determination of major elements and inductively coupled plasma-optical emission spectrometry (ICP-OES; Optima 5300DV, PerkinElmer)
Direct aqueous FGD gypsum carbonation was carried out by injecting CO2 gas (99.99%) into an ammonia solution of FGD gypsum as shown in Fig. 1. Our preliminary study showed that the solid-to-solution ratio in the range of 15–60% had no notable effect on the carbonation rate and efficiency [6]. Here, the solid-to-solution ratio (%) was calculated on a mass basis. Based on this result, the solid-to-solution ratio was set at 15%. Thus, 200 g of FGD gypsum and aqueous ammonia of 3.9% (v/v) were used. The concentration of ammonia was adjusted using a commercial 25 wt% ammonia solution (with a density of 0.9 g/mL). An amount of ammonia in slight excess (120%) of the stoichiometric ratio 2 was used (Eq. (3)). The ammonia solution containing FGD-gypsum was stirred with a mechanical impeller for 5 min at 400 rpm before injection of CO2 gas (1–3 L/min). The suspension was sampled (in 20 mL portions) at predetermined intervals. The samples were filtered for the quantification of dissolved sulfate ions and calcium species. After the filtered solution was allowed to settle for 10 h, dissolved calcium ions were determined after refiltering. The concentration of CaCO3 was estimated from the difference between the total and dissolved concentrations of calcium. All experiments were performed at room temperature and atmospheric pressure with monitoring of temperature and pH (Orion 410A, Thermo Scientific). Filtration was performed using a 0.2 lm membrane filter (cellulose acetate, Sartorius). Cations including calcium ions were determined using ICP-OES after acidification of the sample to pH 2 with instrumental grade HNO3. Dissolved sulfate ions were quantified by ion chromatography (ICS-3000, Dionex) after filtering samples. Filtered solid samples were washed with deionized water purified using a Milli-Q 18 MX cm system and dried overnight at 60 °C. The samples were then examined by X-ray diffraction (XRD; X’pert MPD, Philips Analytical) and field-emission scanning electron microscopy (FE-SEM; S-4700, Hitachi) equipped with an energy dispersive X-ray spectroscopy (EDX; EMAX 7200-H, Horiba). X-ray photoelectron
Fig. 1. Schematic diagram of the direct aqueous carbonation reactor.
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
253
spectroscopy (XPS; MultiLab 2000, Thermo Scientific) was employed to identify the formation of CaCO3 at the surface. The C1s peaks were compared and binding energies were calibrated based on the contaminant hydrocarbon C1s peak at 284.8 eV. In the two-step FGD gypsum carbonation, which was conducted for comparison, the following reactions were consecutively performed: (1) formation of ammonium (bi)carbonate, and (2) the reaction between FGD gypsum and the ammonium (bi)carbonate thus formed. In the first reaction to prepare ammonium (bi)carbonate, the same ammonia concentration of 3.9% (v/v) was employed with the CO2 gas flow of 1 L/min. This reaction was terminated with the indication of the stabilization of pH. The second reaction of the resultant ammonium (bi)carbonate with FGD gypsum was initiated after the ammonium (bi)carbonate solution cooled to room temperature. All other analytical processes were identical to those in the direct aqueous carbonation. Fig. 2. XRD pattern of FGD gypsum.
2.3. Determination of phase composition Calcium carbonate has three anhydrous crystalline forms of calcite, aragonite, and vaterite. Calcite is the thermodynamically most stable form under ambient conditions and vaterite is the least. In our experiments, only the vaterite and calcite polymorphs of calcium carbonate crystals precipitated; the aragonite phase was not detected. The phases were identified by XRD conducted with a Rigaku D/MAX 2200 equipped with a graphite monochromator operating at 40 mA and 40 kV using Cu Ka radiation and a continuous scan at diffraction angles between 10° and 80°. To estimate the phase composition, a step scan was typically carried out for the diffraction angles between 22° and 34° for the (1 1 0), (1 1 2), and (1 1 4) peaks of vaterite, and the (1 0 4) peak of calcite, with the step width of 0.01° and counting time of 2 s. The peak profile fitting was carried out using the MDI Jade 6.5 program furnished with the XRD instrument. The data profiles were fit with a pseudo-Voigt profile function. The relative amounts of the vaterite was determined using Eq. (4) as proposed by Rao [8]:
fv ¼ ½ðI110v þ I112v þ I114v Þ=ðI110v þ I112v þ I114v þ I104c Þ
ð4Þ
where I represents intensities of XRD peaks and the subscripts v and c stand for vaterite and calcite phases. 3. Results and discussion 3.1. Characterization of the FGD-gypsum The XRD pattern of the FGD gypsum shows that it is calcium sulfate dihydrate containing muscovite (KAl2Si3AlO10(OH)2) and dolomite (CaMg(CO3)2) as minor phases (Fig. 2). The formula weight of FGD gypsum was calculated as CaSO41.9H2O based on the wet chemistry data (32.5% CaO, 44.8% SO3, and 20.2% combined water). The nominal purity of calcium sulfate dihydrate estimated from the composition results was approximately 97.5%. Minor impurities were determined as 0.7% SiO2, 0.4% Al2O3, 0.2% Fe2O3, and 0.1% K2O. However, heavy metal impurities Pb, As, Hg, Zn, Mn, and Cd were not detected. The particle size ranged from 1 to 100 lm with volume mean diameter (VMD) of 32.9 lm, but most particles (over 80%) were smaller than 74 lm. EPMA mapping indicates that trace cations Si, Al, and K are associated with the muscovite (Fig. 3), since they appear together. 3.2. Carbonation of FGD gypsum FGD gypsum showed high carbonation reactivity at room temperature and atmospheric pressure [6], producing ammonium sulfate and calcium carbonate. The calcium carbonate was a mixture
of calcite and vaterite under all conditions. The ammonium sulfate was very pure since most of the insoluble impurities could be easily filtered out. The filtered solution contained representative metallic impurities Si, K, Na, Al, and Mg in concentrations less than 10 mg/L, and Fe was under the detection limit (0.01 mg/L). This result suggests that most of the contaminants originated from relatively insoluble minerals such as muscovite and dolomite. Accordingly, the purity of ammonium sulfate was typically over 98%, which is pure enough for commercial applications. It is expected that the direct carbonation reaction would be very complicated because gaseous, liquid, and solid phases are involved, leading to a number of other chemical interactions: (1) absorption of CO2 gas in the ammonia solution, (2) formation of ammonium (bi)carbonate, (3) chemical reaction between ammonium (bi)carbonate and FGD gypsum, and (4) precipitation of calcium carbonate. The relevant reactions are:
NH4 OHðaqÞ ! NHþ4 þ OH
ð5Þ
CO2 ðgÞ ! CO2 ðaqÞ
ð6Þ
CO2 ðaqÞ þ H2 O ! H2 CO3 ðaqÞ
ð7Þ
H2 CO3 ðaqÞ ! Hþ þ HCO3
ð8Þ
HCO3 þ OH ! CO2 3 þ H2 O
ð9Þ
HCO3 þ NHþ4 ! NH4 HCO3 ðaqÞ
ð10Þ
þ CO2 3 þ 2NH4 ! ðNH4 Þ2 CO3 ðaqÞ
ð11Þ
CaSO4 2H2 O ! Ca2þ þ SO2 4 þ 2H2 O
ð12Þ
2NHþ4 þ SO2 4 ! ðNH4 Þ2 SO4 ðaqÞ
ð13Þ
Ca2þ þ CO2 3 ! CaCO3 ðsÞ
ð14Þ
The initial pH of the aqueous ammonia solution after charging it with FGD gypsum was 12.0. At this stage, before the injection of CO2, the initial concentrations of dissolved calcium and sulfate were determined as 560 mg/L (or 14.0 mM) and 1350 mg/L (or 14.0 mM), respectively. These amounts correspond to only 1.8% of the total gypsum added. On the contrary, CO2 gas can be dissolved easily as CO2 3 (pKa2 = 10.3 in Eq. (9)) at this high pH. The progress of the carbonation reaction was monitored by measuring the pH and sulfate ion concentration of the solution
254
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
Fig. 3. EPMA-mapped images of FGD gypsum.
(Fig. 4). When the carbonation was complete, the sulfate ion concentration and pH approached constant values of approximately 800 mM and 7.0, respectively. Also, it can be seen that the dissolution rate of FGD gypsum increased with CO2 flow rate. This suggests that the increase in CO2 flow rate facilitates the production of carbonate ions in the solution that react with calcium ions, thereby enhancing the carbonation. In addition, it would also be possible that the enhanced agitation by increased CO2 flow rate may increase the dissolution of FGD gypsum. Fig. 5 shows the FE-SEM images of the solid residues produced during carbonation, which show the morphological changes of the reaction products with time. The lower magnification micrographs (500) indicate that there was a significant size reduction of FGD gypsum between 10 min and 15 min. The carbonation is likely to proceed via the abrasion of FGD gypsum particles, as was previously observed in the direct aqueous carbonation process [9]. Higher magnification micrographs (10 000) show that calcium carbonate precipitates were formed separately from the FGD gypsum and were not coated on the surface of FGD gypsum. This fact can be further substantiated by the EDX mapped images of the sample obtained after 15 min of the direct carbonation (Fig. 6). The discrepancy of images from C-Ka (CaCO3) and S-Ka (CaSO4) suggests that CaCO3 is not grown from the surface of FGD gypsum.
Fig. 4. Variation in pH and concentration of dissolved sulfate ions during the direct aqueous carbonation of FGD gypsum at various CO2 flow rates.
Also, it can be seen that the sponge-like spherical vaterite phase increased after carbonation for 45 min. For comparison purposes, calcium carbonate was also prepared by the two-step carbonation method. The first step (formation of ammonium (bi)carbonate) is exothermic and the change in temperature is depicted in Fig. 7. Again, the completion of the reactions was indicated by the equilibrium pH at around 8. This pH value corresponds to the production of ammonium bicarbonate (NH4HCO3) [10]. Like direct carbonation, two-step carbonation was similarly facilitated by increased CO2 flow rate. Bonenfant et al. [11] also demonstrated that the CO2 absorption rate in amine solutions increases with increased CO2 flow rate. During the early stage of the second step, the concentration of dissolved sulfate, and thus, that of calcium, became much higher than that in the direct carbonation; after carbonation for 5 min, the concentration of dissolved sulfate in the two-step carbonation was 524 mM (66%), while it was only 28 mM (3.5%) in the direct carbonation. The XRD study revealed that the dominating phase produced over the entire period of two-step carbonation was calcite (Fig. 8).
3.3. Polymorphs of calcium carbonate Vaterite is the thermodynamically least stable phase among three polymorphic forms of calcium carbonate but this metastable form is frequently found under specific conditions [12–15]. In our experiments, the polymorphs of calcium carbonate were found to depend on the carbonation sequence, carbonation time, etc. The change in phase with carbonation time was monitored at the CO2 flow rate of 1 L/min. Fig. 9 shows the XRD patterns of products after direct carbonation for various times. Calcium carbonate was not detected by XRD when the carbonation time was shorter than 15 min. Fig. 10 shows the variation in the amount of vaterite phase with time for direct carbonation and two-step carbonation. In direct carbonation, the vaterite phase increased with time, as was previously confirmed by FE-SEM (Fig. 5). It is seen that the vaterite phase increased rapidly, reaching a plateau after 45 min. On the contrary, calcite was the main phase in the two-step carbonation throughout the reaction period (Fig. 8). Dickinson et al. [16] suggested that the deficiency of calcium ions leads to the production of thermodynamically stable calcite. According to them, pCO2 influences the crystallization and vaterite formation is kinetically favored at high pCO2 when calcium ion concentration is higher than 80 mM. Han et al. [17] also observed
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
255
Fig. 5. FE-SEM images of solid residues formed during direct aqueous carbonation for various durations.
Fig. 6. (A) SEM image (10 000 magnification), (B) EDS mapping (C-Ka), (C) EDS mapping (Al-Ka), (D) EDS mapping (S-Ka) after 15 min of the direct carbonation.
the same tendency for increasing proportions of vaterite with increasing CO2 flow rate owing to higher CO2 3 concentration. These
observations are consistent with our results. As described in Section 3.2 (Fig. 4), the concentration of sulfate ions is relatively
256
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
Fig. 7. Variation in pH and temperature during reaction (1) with CO2 flow rate (without adding FGD gypsum).
Fig. 10. Variation in vaterite fraction in solid residues of direct aqueous carbonation with carbonation time. Here, fv represents the relative amounts of the vaterite in the calcium carbonate precipitate.
Fig. 8. XRD patterns of the solid residues of two-step carbonation for various durations. Here, g, v, and c represent the gypsum, vaterite, and calcite phases, respectively.
Fig. 11. XPS C1s spectra of the solid residues of direct aqueous carbonation for various durations.
Fig. 9. XRD patterns of the solid residues of one-step carbonation for various durations. Here, g, v, and c represent the gypsum, vaterite, and calcite phases, respectively.
low (120 mM, 15% of the total amount) at 20 min and increases sharply to 89% of the total amount (720 mM) at 30 min. The concentration of dissolved sulfate ions could be converted to the corresponding amount of dissolved calcium ions. Consequently, during the initial stage when the calcium ions are dissolved insufficiently, the more thermodynamically stable calcite phase may precipitate through slow nucleation and growth. On the other hand, when the calcium concentration increases after 20 min, the tendency for the vaterite phase to precipitate also increases. In comparison, the two-step carbonation produced calcite as the major phase (Fig. 8). The dominant formation of thermodynamically stable calcite may be due to the slower precipitation of calcium carbonate when bicarbonate ions were used as a carbonate source instead of carbonate ions [18].
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
257
because high-purity calcium carbonate may precipitate from the solution during the induction period, as discussed in Section 3.5. 3.5. Precipitation of high-purity calcite
Fig. 12. Calcite concentration in filtered solutions after direct carbonation for various durations.
3.4. Induction period Calcium carbonate is known to exist as a dissolved ion pair during the induction period before it begins to precipitate [19]. An induction period was generally observed in a slow precipitation process (i.e., when the concentration ratio of calcium to carbonate ions is far from the stoichiometric composition) [20]. The induction period of direct carbonation was assessed using XPS. Fig. 11 shows a series of C1s XPS photoelectron lines from the solid residue after direct carbonation for various times. The peak at 289 eV is assigned to the C1s peak of CaCO3 and the peak at 284.8 eV represents the adventitious contamination C1s peak (the calibration peak) [21]. The intensity of the C1s peak of CaCO3 increases as the carbonation time increases. It is distinguishable after carbonation for 15 min and becomes clear after 20 min. The shift of the peak toward lower energy (15 min) suggests that complete crystallization of CaCO3 had not been achieved yet. This agrees with the XRD data shown in Fig. 9. The induction period becomes shorter as CO2 gas flow increases. This is because the rapid chemical absorption of CO2 gas promotes not only the dissolution of FGD gypsum (increasing calcium ion concentration), but also the precipitation of calcium carbonate (Fig. 12). On the contrary, in the case of the conventional two-step reaction, the initial concentrations of the calcium and bicarbonate ions were high enough to precipitate CaCO3 as early as 5 min. Accordingly, the identification or detection of the induction period is not possible in two-step carbonation. This difference is notable
The induction period during direct carbonation can be utilized to precipitate high-purity calcium carbonate crystals in a manner similar to that used to obtain ammonium sulfate. Most of the impurities in FGD gypsum were insoluble and did not leach into the solution as previously confirmed by ICP analysis. Therefore, high-purity calcium carbonate crystals can be precipitated from the solution extracted during the induction period. To confirm this, we allowed the solution to stand for about 10 h and filtered off the solid precipitates. FE-SEM and XRD analyses showed that the precipitates were single-phase calcite crystals. Fig. 13 shows FE-SEM images of the calcite crystals precipitated from the filtered solution after 5, 15, and 20 min. The calcite crystals were a few lm in size and were extremely pure, as revealed by ICP analysis, which indicated that most impurities were below the detection limit. This method obviously cannot be applied to the conventional two-step carbonation reaction since it lacks an induction period. The total amount of pure calcite crystals obtained in this study was approximately 5% of the FGD gypsum. To increase the yield, it would be essential to retard the crystallization process of calcium carbonate and/or lengthen the induction time. The precipitation rate of calcium carbonate is known to decrease by the small amount of additives such as organophosphorus compounds [22]. In addition, a continuous process designed for the automatic extraction and replenishment of induction-period solutions would be worth to endeavor. 4. Conclusions The precipitation of calcium carbonate crystals during the direct aqueous carbonation of flue gas desulfurization (FGD) gypsum was investigated. FGD gypsum revealed high carbonation reactivity at room temperature and atmospheric pressure. The formation of ammonium (bi)carbonate was the rate limiting step in the direct aqueous carbonation of FGD gypsum; thus, increased CO2 flow rate enhanced carbonation. The calcium carbonate polymorphs produced by direct carbonation were rhombic calcite and spherical vaterite, while aragonite was not observed. The amount of vaterite phase in the mixture was found to increase with carbonation time for a fixed CO2 flow rate. In contrast, twostep carbonation produced mostly calcite crystals for all carbonation times. Virtually impurity-free calcium carbonate was crystallized from the solution which was extracted during the induction period of direct carbonation. The amount of pure calcite crystals
Fig. 13. FE-SEM images of calcite precipitated from the filtered solution after direct aqueous carbonation for (A) 5 min, (B) 15 min, and (C) 20 min.
258
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
was typically 5.3% of the amount that can be obtained in the case of the complete reaction. The process described above may enable us to prepare pure calcite crystals from industrial by-product FGD gypsum. Although the amount of pure calcite crystals obtained in this study was rather small (about 5% of FGD gypsum), if the process can be elaborated further in a cycled manner, it may add to the economic benefits of the direct carbonation of FGD gypsum. Acknowledgements This research was supported by the Basic Research Project of the Korea Institute of Geoscience and Mineral Resources (KIGAM) funded by the Ministry of Knowledge Economy of Korea. References [1] G.B. Cordell, Reaction kinetics of the production of ammonium sulfate from anhydrite, Ind. Eng. Chem. Process Des. Dev. 7 (1968) 278–285. [2] M.-I.M. Chou, J.A. Bruinius, V. Benig, S.-F.J. Chou, R.H. Carty, Producing ammonium sulfate from flue gas desulfurization by-products, Energy Sources Part A 27 (2005) 1061–1071. [3] W.C. Burnett, M.K. Schultz, C.D. Hull, Radionuclide flow during the conversion of phosphogypsum to ammonium sulfate, J. Environ. Radioact. 32 (1996) 33– 51. [4] W. Wei, G.-H. Ma, G. Hu, D. Yu, T. McLeish, Z.-G. Su, Z.-Y. Shen, Preparation of hierarchical hollow CaCO3 particles and the application as anticancer drug carrier, J. Am. Chem. Soc. 130 (2008) 15808–15810. [5] D.R. Lide, CRC Handbook of Chemistry and Physics, 86th ed., CRC Press, Boca Raton, 2005. [6] Korea Ministry of Knowledge Economy, Utilization and sequestration of CO2 using industrial minerals, 2010, GP2010-018-2010(1). [7] F. Kozak, A. Petig, E. Morris, R. Rhudy, D. Thimsen, Chilled ammonia process for CO2 capture, Energy Procedia 1 (2009) 1419–1426. [8] M.S. Rao, Kinetics and mechanism of the transformation of vaterite to calcite, Bull. Chem. Soc. Jpn. 46 (1973) 1414–1417.
[9] H. Béarat, M.J. McKelvy, A.V.G. Chizmeshya, D. Gormley, R. Nunez, R.W. Carpenter, K. Squires, G.H. Wolf, Carbon sequestration via aqueous olivine mineral carbonation: role of passivating layer formation, Environ. Sci. Technol. 40 (2006) 4802–4808. [10] H. Bai, A.C. Yeh, Removal of CO2 greenhouse gas by ammonia scrubbing, Ind. Eng. Chem. Res. 36 (1997) 2490–2493. [11] D. Bonenfant, M. Mimeault, R. Hausler, Determination of the structural features of distinct amines important for the absorption of CO2 and regeneration in aqueous solution, Ind. Eng. Chem. Res. 42 (2003) 3179–3184. [12] D. Kralj, L. Brecˇevic´, J. Kontrec, Vaterite growth and dissolution in aqueous solution III. Kinetics of transformation, J. Cryst. Growth 177 (1997) 248–257. [13] D. Kralj, L. Brecˇevic´, A.E. Nielsen, Vaterite growth and dissolution in aqueous solution I. Kinetics of crystal growth, J. Cryst. Growth 104 (1990) 793–800. [14] Y. Sheng Han, G. Hadiko, M. Fuji, M. Takahashi, Crystallization and transformation of vaterite at controlled pH, J. Cryst. Growth 289 (2006) 269– 274. [15] N. Spanos, P.G. Koutsoukos, The transformation of vaterite to calcite: effect of the conditions of the solutions in contact with the mineral phase, J. Cryst. Growth 191 (1998) 783–790. [16] S.R. Dickinson, G.E. Henderson, K.M. McGrath, Controlling the kinetic versus thermodynamic crystallisation of calcium carbonate, J. Cryst. Growth 244 (2002) 369–378. [17] Y.S. Han, G. Hadiko, M. Fuji, M. Takahashi, Effect of flow rate and CO2 content on the phase and morphology of CaCO3 prepared by bubbling method, J. Cryst. Growth 276 (2005) 541–548. [18] J.-H. Bang, Y.N. Jang, K.S. Song, C.W. Jeon, W. Kim, M.G. Lee, S.-J. Park, Effects of sodium laurylsulfate on crystal structure of calcite formed from mixed solutions, J. Colloid Interface Sci. 356 (2011) 311–315. [19] M. Vucˇak, J. Peric´, A. Zˇmikic´, M.N. Pons, A study of carbon dioxide absorption into aqueous monoethanolamine solution containing calcium nitrate in the gas–liquid reactive precipitation of calcium carbonate, Chem. Eng. J. 87 (2002) 171–179. [20] J. Gómez-Morales, J. Torrent-Burgués, A. López-Macipe, R. RodríguezClemente, Precipitation of calcium carbonate from solutions with varying Ca2+/carbonate ratios, J. Cryst. Growth 166 (1996) 1020–1026. [21] S.L. Stipp, M.F. Hochella Jr., G.A. Parks, J.O. Leckie, Cd2+ uptake by calcite, solid– state diffusion, and the formation of solid–solution: interface processes observed with near-surface sensitive techniques (XPS, LEED, and AES), Geochim. Cosmochim. Acta 56 (1992) 1941–1954. [22] A.G. Xyla, J. Mikroyannidis, P.G. Koutsoukos, The inhibition of calcium carbonate precipitation in aqueous media by organophosphorus compounds, J. Colloid Interface Sci. 153 (1992) 537–551.
Contents lists available at SciVerse ScienceDirect
Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej
Precipitation of calcium carbonate during direct aqueous carbonation of flue gas desulfurization gypsum Kyungsun Song a,b,1, Young-Nam Jang a,1, Wonbaek Kim a,⇑, Myung Gyu Lee c, Dongbok Shin b, Jun-Hwan Bang a, Chi Wan Jeon a, Soo Chun Chae a a b c
Korea Institute of Geoscience & Mineral Resources (KIGAM), Gwahang-no 124, Yuseong-gu, Daejeon 305-350, Republic of Korea Department of Geoenvironmental Sciences, Kongju National University, Kongju, Chungnam 314-701, Republic of Korea Resources Recycling Engineering Department, University of Science and Technology of Korea, Gwahang-no 113, Yuseong-gu, Daejeon 305-333, Republic of Korea
h i g h l i g h t s " Direct carbonation of FGD gypsum was studied. " Direct carbonation produced a mixture of vaterite and calcite. " Vaterite/calcite increased with carbonation time. " Impurity-free calcite was obtained in direct carbonation.
a r t i c l e
i n f o
Article history: Received 26 July 2012 Received in revised form 16 October 2012 Accepted 18 October 2012 Available online 25 October 2012 Keywords: Flue gas desulfurization gypsum Calcium carbonate Calcite Vaterite Carbonation
a b s t r a c t The precipitation of calcium carbonate during the direct aqueous carbonation of flue gas desulfurization (FGD) gypsum, an industrial waste product, was investigated. For comparison, two-step carbonation was also attempted. Calcite was the dominating phase produced in the two-step carbonation, while the direct carbonation produced a mixture of vaterite and calcite phases under all conditions. The relative amounts of each phase were determined by comparing their X-ray diffraction peak intensities. The amount of vaterite phase in the mixture was found to increase with carbonation time for a fixed CO2 flow rate. An induction period before the precipitation of physically detectable calcium carbonate crystals was assessed using X-ray photoelectron spectroscopy (XPS) and field-emission scanning electron microscopy (FE-SEM). For the CO2 flow rate of 1 L/min, the C1s peak of CaCO3 became clear after carbonation took place for longer than 15 min. Virtually impurity-free, single-phase calcite crystals were precipitated from the solution extracted during the induction period. The amount of calcite obtained was typically 5.3% of the amount that can be obtained in the case of the complete reaction. However, further elaboration of this method may allow for the preparation of pure calcite crystals from industrial by-product FGD gypsum. Ó 2012 Elsevier B.V. All rights reserved.
1. Introduction
CO2 ðgÞ þ 2NH4 OH ! ðNH4 Þ2 CO3 ðaqÞ
Flue gas desulfurization (FGD) gypsum is the main product of the desulfurization system used to remove SOx from coal combustion products. One established utilization of waste gypsum has been the production of the fertilizer ammonium sulfate [1–3]. At the same time, calcium carbonate is produced as a byproduct which might a have commercial value depending on the purity, morphology, etc. Its application ranges from the conventional fields such as paper or cosmetics to drug industries [4]. The process conventionally consists of the two following consecutive reactions:
CaSO4 2H2 O þ ðNH4 Þ2 CO3 ðaqÞ
⇑ Corresponding author. Tel.: +82 42 868 3623. 1
E-mail address: [email protected] (W. Kim). These authors equally contributed to this study.
1385-8947/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cej.2012.10.010
! CaCO3 ðsÞ þ ðNH4 Þ2 SO4 þ 2H2 O
ð1Þ
ð2Þ
The above reactions are characterized by the high yield and purity (up to 99%) of the ammonium sulfate produced [2]. Burnett et al. [3] clearly demonstrated that the purity of the ammonium sulfate could be guaranteed because most contaminants of FGD gypsum remained in the solid residue. If impurities exist in solid form, they can be easily separated from the ammonium sulfate solution by filtration. Considering the solubilities of ammonium sulfate and calcium carbonate, the former, with high solubility (74 g/100 mL of H2O at 20 °C), should exist in solution and most of the comparatively insoluble latter (0.15 g/100 mL H2O at
252
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
25 °C) should precipitate [5]. Thus, pure ammonium sulfate crystals can be precipitated after removing the insoluble impurities by filtration. On the other hand, the solubility of calcium carbonate is low and it cannot be purified in a similar manner. Recently, we evaluated the feasibility of a single-step direct carbonation reaction of FGD gypsum [6]. The process can be expressed as:
was used for the determination of trace elements following acid digestion (HNO3/HClO4). Particle size distribution was measured using a laser scattering particle size analyzer (HELOS/RODOS & SUCELL, Sympatec GmbH). The elemental mapping was carried out using an electron probe microanalyzer (EPMA, EPMA-1600, SHIMADZU).
CaSO4 2H2 O þ 2NH4 OHðaqÞ þ CO2 ðgÞ
2.2. Carbonation experiments
! CaCO3 ðsÞ þ ðNH4 Þ2 SO4 þ 2H2 O
ð3Þ
The reaction (1) is exothermic and the ammonia is usually required to be chilled below 20 °C to reduce the vaporization of ammonia [7]. Therefore, a single-step direct carbonation of FGD gypsum would have an obvious advantage over the conventional two-step process. The effects of the ammonia content, CO2 flow rate, solid-to-solution ratio, and CO2/N2 gas mixture ratio were analyzed by systematically monitoring the temperature, pH of the slurry, and the carbonation conversion rate of products [6]. As expected, comparatively pure ammonium sulfate could be obtained in the process. In the meantime, the calcium carbonate that precipitated in the direct carbonation process always produced a mixture of the thermodynamically unstable vaterite and the stable calcite phases under all conditions. In this paper, we focus on the precipitation of calcium carbonate and its polymorphism. Additionally, our attempts to produce calcium carbonate of high purity from FGD gypsum in a manner similar to that used for the production of ammonium sulfate are discussed. The purity, phase, and other physical characteristics of the calcium carbonate produced were studied using a variety of spectroscopic and microscopy techniques. A comparative study of calcium carbonate produced by the conventional two-step process is also presented. 2. Experimental 2.1. Materials FGD gypsum was obtained from Yeongheung Thermal Power Plants, Incheon, Korea and dried at 45 °C overnight to remove the surface water. Wet chemical analysis was performed for determination of major elements and inductively coupled plasma-optical emission spectrometry (ICP-OES; Optima 5300DV, PerkinElmer)
Direct aqueous FGD gypsum carbonation was carried out by injecting CO2 gas (99.99%) into an ammonia solution of FGD gypsum as shown in Fig. 1. Our preliminary study showed that the solid-to-solution ratio in the range of 15–60% had no notable effect on the carbonation rate and efficiency [6]. Here, the solid-to-solution ratio (%) was calculated on a mass basis. Based on this result, the solid-to-solution ratio was set at 15%. Thus, 200 g of FGD gypsum and aqueous ammonia of 3.9% (v/v) were used. The concentration of ammonia was adjusted using a commercial 25 wt% ammonia solution (with a density of 0.9 g/mL). An amount of ammonia in slight excess (120%) of the stoichiometric ratio 2 was used (Eq. (3)). The ammonia solution containing FGD-gypsum was stirred with a mechanical impeller for 5 min at 400 rpm before injection of CO2 gas (1–3 L/min). The suspension was sampled (in 20 mL portions) at predetermined intervals. The samples were filtered for the quantification of dissolved sulfate ions and calcium species. After the filtered solution was allowed to settle for 10 h, dissolved calcium ions were determined after refiltering. The concentration of CaCO3 was estimated from the difference between the total and dissolved concentrations of calcium. All experiments were performed at room temperature and atmospheric pressure with monitoring of temperature and pH (Orion 410A, Thermo Scientific). Filtration was performed using a 0.2 lm membrane filter (cellulose acetate, Sartorius). Cations including calcium ions were determined using ICP-OES after acidification of the sample to pH 2 with instrumental grade HNO3. Dissolved sulfate ions were quantified by ion chromatography (ICS-3000, Dionex) after filtering samples. Filtered solid samples were washed with deionized water purified using a Milli-Q 18 MX cm system and dried overnight at 60 °C. The samples were then examined by X-ray diffraction (XRD; X’pert MPD, Philips Analytical) and field-emission scanning electron microscopy (FE-SEM; S-4700, Hitachi) equipped with an energy dispersive X-ray spectroscopy (EDX; EMAX 7200-H, Horiba). X-ray photoelectron
Fig. 1. Schematic diagram of the direct aqueous carbonation reactor.
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
253
spectroscopy (XPS; MultiLab 2000, Thermo Scientific) was employed to identify the formation of CaCO3 at the surface. The C1s peaks were compared and binding energies were calibrated based on the contaminant hydrocarbon C1s peak at 284.8 eV. In the two-step FGD gypsum carbonation, which was conducted for comparison, the following reactions were consecutively performed: (1) formation of ammonium (bi)carbonate, and (2) the reaction between FGD gypsum and the ammonium (bi)carbonate thus formed. In the first reaction to prepare ammonium (bi)carbonate, the same ammonia concentration of 3.9% (v/v) was employed with the CO2 gas flow of 1 L/min. This reaction was terminated with the indication of the stabilization of pH. The second reaction of the resultant ammonium (bi)carbonate with FGD gypsum was initiated after the ammonium (bi)carbonate solution cooled to room temperature. All other analytical processes were identical to those in the direct aqueous carbonation. Fig. 2. XRD pattern of FGD gypsum.
2.3. Determination of phase composition Calcium carbonate has three anhydrous crystalline forms of calcite, aragonite, and vaterite. Calcite is the thermodynamically most stable form under ambient conditions and vaterite is the least. In our experiments, only the vaterite and calcite polymorphs of calcium carbonate crystals precipitated; the aragonite phase was not detected. The phases were identified by XRD conducted with a Rigaku D/MAX 2200 equipped with a graphite monochromator operating at 40 mA and 40 kV using Cu Ka radiation and a continuous scan at diffraction angles between 10° and 80°. To estimate the phase composition, a step scan was typically carried out for the diffraction angles between 22° and 34° for the (1 1 0), (1 1 2), and (1 1 4) peaks of vaterite, and the (1 0 4) peak of calcite, with the step width of 0.01° and counting time of 2 s. The peak profile fitting was carried out using the MDI Jade 6.5 program furnished with the XRD instrument. The data profiles were fit with a pseudo-Voigt profile function. The relative amounts of the vaterite was determined using Eq. (4) as proposed by Rao [8]:
fv ¼ ½ðI110v þ I112v þ I114v Þ=ðI110v þ I112v þ I114v þ I104c Þ
ð4Þ
where I represents intensities of XRD peaks and the subscripts v and c stand for vaterite and calcite phases. 3. Results and discussion 3.1. Characterization of the FGD-gypsum The XRD pattern of the FGD gypsum shows that it is calcium sulfate dihydrate containing muscovite (KAl2Si3AlO10(OH)2) and dolomite (CaMg(CO3)2) as minor phases (Fig. 2). The formula weight of FGD gypsum was calculated as CaSO41.9H2O based on the wet chemistry data (32.5% CaO, 44.8% SO3, and 20.2% combined water). The nominal purity of calcium sulfate dihydrate estimated from the composition results was approximately 97.5%. Minor impurities were determined as 0.7% SiO2, 0.4% Al2O3, 0.2% Fe2O3, and 0.1% K2O. However, heavy metal impurities Pb, As, Hg, Zn, Mn, and Cd were not detected. The particle size ranged from 1 to 100 lm with volume mean diameter (VMD) of 32.9 lm, but most particles (over 80%) were smaller than 74 lm. EPMA mapping indicates that trace cations Si, Al, and K are associated with the muscovite (Fig. 3), since they appear together. 3.2. Carbonation of FGD gypsum FGD gypsum showed high carbonation reactivity at room temperature and atmospheric pressure [6], producing ammonium sulfate and calcium carbonate. The calcium carbonate was a mixture
of calcite and vaterite under all conditions. The ammonium sulfate was very pure since most of the insoluble impurities could be easily filtered out. The filtered solution contained representative metallic impurities Si, K, Na, Al, and Mg in concentrations less than 10 mg/L, and Fe was under the detection limit (0.01 mg/L). This result suggests that most of the contaminants originated from relatively insoluble minerals such as muscovite and dolomite. Accordingly, the purity of ammonium sulfate was typically over 98%, which is pure enough for commercial applications. It is expected that the direct carbonation reaction would be very complicated because gaseous, liquid, and solid phases are involved, leading to a number of other chemical interactions: (1) absorption of CO2 gas in the ammonia solution, (2) formation of ammonium (bi)carbonate, (3) chemical reaction between ammonium (bi)carbonate and FGD gypsum, and (4) precipitation of calcium carbonate. The relevant reactions are:
NH4 OHðaqÞ ! NHþ4 þ OH
ð5Þ
CO2 ðgÞ ! CO2 ðaqÞ
ð6Þ
CO2 ðaqÞ þ H2 O ! H2 CO3 ðaqÞ
ð7Þ
H2 CO3 ðaqÞ ! Hþ þ HCO3
ð8Þ
HCO3 þ OH ! CO2 3 þ H2 O
ð9Þ
HCO3 þ NHþ4 ! NH4 HCO3 ðaqÞ
ð10Þ
þ CO2 3 þ 2NH4 ! ðNH4 Þ2 CO3 ðaqÞ
ð11Þ
CaSO4 2H2 O ! Ca2þ þ SO2 4 þ 2H2 O
ð12Þ
2NHþ4 þ SO2 4 ! ðNH4 Þ2 SO4 ðaqÞ
ð13Þ
Ca2þ þ CO2 3 ! CaCO3 ðsÞ
ð14Þ
The initial pH of the aqueous ammonia solution after charging it with FGD gypsum was 12.0. At this stage, before the injection of CO2, the initial concentrations of dissolved calcium and sulfate were determined as 560 mg/L (or 14.0 mM) and 1350 mg/L (or 14.0 mM), respectively. These amounts correspond to only 1.8% of the total gypsum added. On the contrary, CO2 gas can be dissolved easily as CO2 3 (pKa2 = 10.3 in Eq. (9)) at this high pH. The progress of the carbonation reaction was monitored by measuring the pH and sulfate ion concentration of the solution
254
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
Fig. 3. EPMA-mapped images of FGD gypsum.
(Fig. 4). When the carbonation was complete, the sulfate ion concentration and pH approached constant values of approximately 800 mM and 7.0, respectively. Also, it can be seen that the dissolution rate of FGD gypsum increased with CO2 flow rate. This suggests that the increase in CO2 flow rate facilitates the production of carbonate ions in the solution that react with calcium ions, thereby enhancing the carbonation. In addition, it would also be possible that the enhanced agitation by increased CO2 flow rate may increase the dissolution of FGD gypsum. Fig. 5 shows the FE-SEM images of the solid residues produced during carbonation, which show the morphological changes of the reaction products with time. The lower magnification micrographs (500) indicate that there was a significant size reduction of FGD gypsum between 10 min and 15 min. The carbonation is likely to proceed via the abrasion of FGD gypsum particles, as was previously observed in the direct aqueous carbonation process [9]. Higher magnification micrographs (10 000) show that calcium carbonate precipitates were formed separately from the FGD gypsum and were not coated on the surface of FGD gypsum. This fact can be further substantiated by the EDX mapped images of the sample obtained after 15 min of the direct carbonation (Fig. 6). The discrepancy of images from C-Ka (CaCO3) and S-Ka (CaSO4) suggests that CaCO3 is not grown from the surface of FGD gypsum.
Fig. 4. Variation in pH and concentration of dissolved sulfate ions during the direct aqueous carbonation of FGD gypsum at various CO2 flow rates.
Also, it can be seen that the sponge-like spherical vaterite phase increased after carbonation for 45 min. For comparison purposes, calcium carbonate was also prepared by the two-step carbonation method. The first step (formation of ammonium (bi)carbonate) is exothermic and the change in temperature is depicted in Fig. 7. Again, the completion of the reactions was indicated by the equilibrium pH at around 8. This pH value corresponds to the production of ammonium bicarbonate (NH4HCO3) [10]. Like direct carbonation, two-step carbonation was similarly facilitated by increased CO2 flow rate. Bonenfant et al. [11] also demonstrated that the CO2 absorption rate in amine solutions increases with increased CO2 flow rate. During the early stage of the second step, the concentration of dissolved sulfate, and thus, that of calcium, became much higher than that in the direct carbonation; after carbonation for 5 min, the concentration of dissolved sulfate in the two-step carbonation was 524 mM (66%), while it was only 28 mM (3.5%) in the direct carbonation. The XRD study revealed that the dominating phase produced over the entire period of two-step carbonation was calcite (Fig. 8).
3.3. Polymorphs of calcium carbonate Vaterite is the thermodynamically least stable phase among three polymorphic forms of calcium carbonate but this metastable form is frequently found under specific conditions [12–15]. In our experiments, the polymorphs of calcium carbonate were found to depend on the carbonation sequence, carbonation time, etc. The change in phase with carbonation time was monitored at the CO2 flow rate of 1 L/min. Fig. 9 shows the XRD patterns of products after direct carbonation for various times. Calcium carbonate was not detected by XRD when the carbonation time was shorter than 15 min. Fig. 10 shows the variation in the amount of vaterite phase with time for direct carbonation and two-step carbonation. In direct carbonation, the vaterite phase increased with time, as was previously confirmed by FE-SEM (Fig. 5). It is seen that the vaterite phase increased rapidly, reaching a plateau after 45 min. On the contrary, calcite was the main phase in the two-step carbonation throughout the reaction period (Fig. 8). Dickinson et al. [16] suggested that the deficiency of calcium ions leads to the production of thermodynamically stable calcite. According to them, pCO2 influences the crystallization and vaterite formation is kinetically favored at high pCO2 when calcium ion concentration is higher than 80 mM. Han et al. [17] also observed
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
255
Fig. 5. FE-SEM images of solid residues formed during direct aqueous carbonation for various durations.
Fig. 6. (A) SEM image (10 000 magnification), (B) EDS mapping (C-Ka), (C) EDS mapping (Al-Ka), (D) EDS mapping (S-Ka) after 15 min of the direct carbonation.
the same tendency for increasing proportions of vaterite with increasing CO2 flow rate owing to higher CO2 3 concentration. These
observations are consistent with our results. As described in Section 3.2 (Fig. 4), the concentration of sulfate ions is relatively
256
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
Fig. 7. Variation in pH and temperature during reaction (1) with CO2 flow rate (without adding FGD gypsum).
Fig. 10. Variation in vaterite fraction in solid residues of direct aqueous carbonation with carbonation time. Here, fv represents the relative amounts of the vaterite in the calcium carbonate precipitate.
Fig. 8. XRD patterns of the solid residues of two-step carbonation for various durations. Here, g, v, and c represent the gypsum, vaterite, and calcite phases, respectively.
Fig. 11. XPS C1s spectra of the solid residues of direct aqueous carbonation for various durations.
Fig. 9. XRD patterns of the solid residues of one-step carbonation for various durations. Here, g, v, and c represent the gypsum, vaterite, and calcite phases, respectively.
low (120 mM, 15% of the total amount) at 20 min and increases sharply to 89% of the total amount (720 mM) at 30 min. The concentration of dissolved sulfate ions could be converted to the corresponding amount of dissolved calcium ions. Consequently, during the initial stage when the calcium ions are dissolved insufficiently, the more thermodynamically stable calcite phase may precipitate through slow nucleation and growth. On the other hand, when the calcium concentration increases after 20 min, the tendency for the vaterite phase to precipitate also increases. In comparison, the two-step carbonation produced calcite as the major phase (Fig. 8). The dominant formation of thermodynamically stable calcite may be due to the slower precipitation of calcium carbonate when bicarbonate ions were used as a carbonate source instead of carbonate ions [18].
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
257
because high-purity calcium carbonate may precipitate from the solution during the induction period, as discussed in Section 3.5. 3.5. Precipitation of high-purity calcite
Fig. 12. Calcite concentration in filtered solutions after direct carbonation for various durations.
3.4. Induction period Calcium carbonate is known to exist as a dissolved ion pair during the induction period before it begins to precipitate [19]. An induction period was generally observed in a slow precipitation process (i.e., when the concentration ratio of calcium to carbonate ions is far from the stoichiometric composition) [20]. The induction period of direct carbonation was assessed using XPS. Fig. 11 shows a series of C1s XPS photoelectron lines from the solid residue after direct carbonation for various times. The peak at 289 eV is assigned to the C1s peak of CaCO3 and the peak at 284.8 eV represents the adventitious contamination C1s peak (the calibration peak) [21]. The intensity of the C1s peak of CaCO3 increases as the carbonation time increases. It is distinguishable after carbonation for 15 min and becomes clear after 20 min. The shift of the peak toward lower energy (15 min) suggests that complete crystallization of CaCO3 had not been achieved yet. This agrees with the XRD data shown in Fig. 9. The induction period becomes shorter as CO2 gas flow increases. This is because the rapid chemical absorption of CO2 gas promotes not only the dissolution of FGD gypsum (increasing calcium ion concentration), but also the precipitation of calcium carbonate (Fig. 12). On the contrary, in the case of the conventional two-step reaction, the initial concentrations of the calcium and bicarbonate ions were high enough to precipitate CaCO3 as early as 5 min. Accordingly, the identification or detection of the induction period is not possible in two-step carbonation. This difference is notable
The induction period during direct carbonation can be utilized to precipitate high-purity calcium carbonate crystals in a manner similar to that used to obtain ammonium sulfate. Most of the impurities in FGD gypsum were insoluble and did not leach into the solution as previously confirmed by ICP analysis. Therefore, high-purity calcium carbonate crystals can be precipitated from the solution extracted during the induction period. To confirm this, we allowed the solution to stand for about 10 h and filtered off the solid precipitates. FE-SEM and XRD analyses showed that the precipitates were single-phase calcite crystals. Fig. 13 shows FE-SEM images of the calcite crystals precipitated from the filtered solution after 5, 15, and 20 min. The calcite crystals were a few lm in size and were extremely pure, as revealed by ICP analysis, which indicated that most impurities were below the detection limit. This method obviously cannot be applied to the conventional two-step carbonation reaction since it lacks an induction period. The total amount of pure calcite crystals obtained in this study was approximately 5% of the FGD gypsum. To increase the yield, it would be essential to retard the crystallization process of calcium carbonate and/or lengthen the induction time. The precipitation rate of calcium carbonate is known to decrease by the small amount of additives such as organophosphorus compounds [22]. In addition, a continuous process designed for the automatic extraction and replenishment of induction-period solutions would be worth to endeavor. 4. Conclusions The precipitation of calcium carbonate crystals during the direct aqueous carbonation of flue gas desulfurization (FGD) gypsum was investigated. FGD gypsum revealed high carbonation reactivity at room temperature and atmospheric pressure. The formation of ammonium (bi)carbonate was the rate limiting step in the direct aqueous carbonation of FGD gypsum; thus, increased CO2 flow rate enhanced carbonation. The calcium carbonate polymorphs produced by direct carbonation were rhombic calcite and spherical vaterite, while aragonite was not observed. The amount of vaterite phase in the mixture was found to increase with carbonation time for a fixed CO2 flow rate. In contrast, twostep carbonation produced mostly calcite crystals for all carbonation times. Virtually impurity-free calcium carbonate was crystallized from the solution which was extracted during the induction period of direct carbonation. The amount of pure calcite crystals
Fig. 13. FE-SEM images of calcite precipitated from the filtered solution after direct aqueous carbonation for (A) 5 min, (B) 15 min, and (C) 20 min.
258
K. Song et al. / Chemical Engineering Journal 213 (2012) 251–258
was typically 5.3% of the amount that can be obtained in the case of the complete reaction. The process described above may enable us to prepare pure calcite crystals from industrial by-product FGD gypsum. Although the amount of pure calcite crystals obtained in this study was rather small (about 5% of FGD gypsum), if the process can be elaborated further in a cycled manner, it may add to the economic benefits of the direct carbonation of FGD gypsum. Acknowledgements This research was supported by the Basic Research Project of the Korea Institute of Geoscience and Mineral Resources (KIGAM) funded by the Ministry of Knowledge Economy of Korea. References [1] G.B. Cordell, Reaction kinetics of the production of ammonium sulfate from anhydrite, Ind. Eng. Chem. Process Des. Dev. 7 (1968) 278–285. [2] M.-I.M. Chou, J.A. Bruinius, V. Benig, S.-F.J. Chou, R.H. Carty, Producing ammonium sulfate from flue gas desulfurization by-products, Energy Sources Part A 27 (2005) 1061–1071. [3] W.C. Burnett, M.K. Schultz, C.D. Hull, Radionuclide flow during the conversion of phosphogypsum to ammonium sulfate, J. Environ. Radioact. 32 (1996) 33– 51. [4] W. Wei, G.-H. Ma, G. Hu, D. Yu, T. McLeish, Z.-G. Su, Z.-Y. Shen, Preparation of hierarchical hollow CaCO3 particles and the application as anticancer drug carrier, J. Am. Chem. Soc. 130 (2008) 15808–15810. [5] D.R. Lide, CRC Handbook of Chemistry and Physics, 86th ed., CRC Press, Boca Raton, 2005. [6] Korea Ministry of Knowledge Economy, Utilization and sequestration of CO2 using industrial minerals, 2010, GP2010-018-2010(1). [7] F. Kozak, A. Petig, E. Morris, R. Rhudy, D. Thimsen, Chilled ammonia process for CO2 capture, Energy Procedia 1 (2009) 1419–1426. [8] M.S. Rao, Kinetics and mechanism of the transformation of vaterite to calcite, Bull. Chem. Soc. Jpn. 46 (1973) 1414–1417.
[9] H. Béarat, M.J. McKelvy, A.V.G. Chizmeshya, D. Gormley, R. Nunez, R.W. Carpenter, K. Squires, G.H. Wolf, Carbon sequestration via aqueous olivine mineral carbonation: role of passivating layer formation, Environ. Sci. Technol. 40 (2006) 4802–4808. [10] H. Bai, A.C. Yeh, Removal of CO2 greenhouse gas by ammonia scrubbing, Ind. Eng. Chem. Res. 36 (1997) 2490–2493. [11] D. Bonenfant, M. Mimeault, R. Hausler, Determination of the structural features of distinct amines important for the absorption of CO2 and regeneration in aqueous solution, Ind. Eng. Chem. Res. 42 (2003) 3179–3184. [12] D. Kralj, L. Brecˇevic´, J. Kontrec, Vaterite growth and dissolution in aqueous solution III. Kinetics of transformation, J. Cryst. Growth 177 (1997) 248–257. [13] D. Kralj, L. Brecˇevic´, A.E. Nielsen, Vaterite growth and dissolution in aqueous solution I. Kinetics of crystal growth, J. Cryst. Growth 104 (1990) 793–800. [14] Y. Sheng Han, G. Hadiko, M. Fuji, M. Takahashi, Crystallization and transformation of vaterite at controlled pH, J. Cryst. Growth 289 (2006) 269– 274. [15] N. Spanos, P.G. Koutsoukos, The transformation of vaterite to calcite: effect of the conditions of the solutions in contact with the mineral phase, J. Cryst. Growth 191 (1998) 783–790. [16] S.R. Dickinson, G.E. Henderson, K.M. McGrath, Controlling the kinetic versus thermodynamic crystallisation of calcium carbonate, J. Cryst. Growth 244 (2002) 369–378. [17] Y.S. Han, G. Hadiko, M. Fuji, M. Takahashi, Effect of flow rate and CO2 content on the phase and morphology of CaCO3 prepared by bubbling method, J. Cryst. Growth 276 (2005) 541–548. [18] J.-H. Bang, Y.N. Jang, K.S. Song, C.W. Jeon, W. Kim, M.G. Lee, S.-J. Park, Effects of sodium laurylsulfate on crystal structure of calcite formed from mixed solutions, J. Colloid Interface Sci. 356 (2011) 311–315. [19] M. Vucˇak, J. Peric´, A. Zˇmikic´, M.N. Pons, A study of carbon dioxide absorption into aqueous monoethanolamine solution containing calcium nitrate in the gas–liquid reactive precipitation of calcium carbonate, Chem. Eng. J. 87 (2002) 171–179. [20] J. Gómez-Morales, J. Torrent-Burgués, A. López-Macipe, R. RodríguezClemente, Precipitation of calcium carbonate from solutions with varying Ca2+/carbonate ratios, J. Cryst. Growth 166 (1996) 1020–1026. [21] S.L. Stipp, M.F. Hochella Jr., G.A. Parks, J.O. Leckie, Cd2+ uptake by calcite, solid– state diffusion, and the formation of solid–solution: interface processes observed with near-surface sensitive techniques (XPS, LEED, and AES), Geochim. Cosmochim. Acta 56 (1992) 1941–1954. [22] A.G. Xyla, J. Mikroyannidis, P.G. Koutsoukos, The inhibition of calcium carbonate precipitation in aqueous media by organophosphorus compounds, J. Colloid Interface Sci. 153 (1992) 537–551.