Trace Na2CO3 Addition to Limestone Inducing High-Capacity SO2CaptureRui Han, Fei Sun, Jihui Gao,* Siyu Wei, Yanlin Su, and Yukun Qin

School of Energy Science and Engineering, Harbin Institute of Technology, Harbin 150001, China

*S Supporting Information

ABSTRACT: Although the literature has reported enhancedindirect sulfation of limestone by adding Na2CO3, the amountof Na2CO3 additive required to achieve high CaO conversionis typically high (∼4.0 mol %), which commonly results inadverse effects in fluidized-bed combustion boiler systems andincreased cost of sorbents. In this work, we demonstrate forthe first time that trace Na2CO3 (0.1 mol %) can significantlyenhance the sulfate conversion of limestone. This enhancedsulfation is attributed to the increased surface area andoptimized pore size distribution. The trace Na2CO3 additivesplits the pores of the original sorbents peaking at ∼70 nminto pores peaking at ∼4 nm and ∼140 nm due to the slightpromotion of sintering. This well-developed pore structure results in a relatively high reactivity for sulfation. Thus, the Na2CO3additive influences the sorbent reactivity in two ways: (1) at less than 0.5 mol %, tuning its pore structure; (2) at more than 0.5mol %, promoting the product layer diffusion. We also find that trace amount of other metal salts, such as CaCl2 and NaCl,clearly enhance the sulfation of limestone. The strategy of enhancing limestone sulfation by the addition of trace amount of metalsalts offers evident engineering and economic advantage.

1. INTRODUCTION

With the growing concerns regarding environmental issues,various pollutant gas emission regulations, especially thoseaddressing SO2 emissions, have been reinforced. In July 2011,China enacted a new coal-fired power plant air pollutantsemission standard (GB 13223−2011) that limits SO2 emissionsto less than 200 mg/m3. This standard is stricter than thoseemployed in the United States, the European Union, and Japan.In September 2014, the action plan for coal power trans-formation and upgrading for energy conservation and emissionsreduction (2014−2020) was published, furthering limiting theSO2 emissions value to 35 mg/m3 for newly built coal-firedgenerating units in the eastern region. These increasinglystringent emission standards have presented a significantchallenge for coal-fired power plants.Limestone, which is characterized by low cost and high

availability, has been widely used to control SO2 emissions fromcoal power plants. Compared with the traditional wet-desulfurization technology, the direct injection of limestoneinto the circulating fluidized bed (CFB) for the in situ captureof SO2 at high temperatures is cost efficient and flexible withlow water consumption.However, the main problem associated with this technology

is the sharply decreased sulfation rate of limestone, particularlywhen the utilization rate of calcium reaches 30−40%.1 Thus, inpractical application, more limestone must be injected into thefurnace to satisfy the emission standard. This rapid decrease inthe sulfation rate of limestone is mainly attributed to the

increased molar volume of the sulfation product CaSO4 relativeto that of CaO, which causes small pores in CaO to becomeplugged, producing CaO particles with a product layer and anunsulfated core.1−3 Diffusion in the product layer involvessolid-state ion diffusion, whose rate is much slower than therate of chemical reactions.4

To improve the sulfation kinetics, inorganic salts, forexample, Na2CO3,

5−11 NaCl,5−7,12−14 and CaCl2,7,15−17 have

been added to limestone to enhance the sulfation performance.A significant body of relevant literature reports the addition oflow melting point inorganic salts to limestone to improve thesulfation process. The current understanding of the effects oflow-melting-point inorganic salts on the Ca-based sulfationprocess can be summarized by the following three points: (1)changing the pore structure of CaO, which alters the porediffusion resistance;5,8,11,13−16 (2) doping metal cations into thelattice of the product, which improves the solid-state iondiffusion;6,7,9,10,12 and (3) comelting with the reactant (CaO)and product (CaSO4) to form a eutectic salt, which reducesdiffusion limitation and leads to morphological changes of thereaction product.17−19

Table S1 (Supporting Information) summarizes the testconditions applied, the optimum metal salt additive dose and

Received: August 12, 2017Revised: October 7, 2017Accepted: October 9, 2017Published: October 9, 2017

Article

pubs.acs.org/est

© XXXX American Chemical Society A DOI: 10.1021/acs.est.7b04141Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Cite This: Environ. Sci. Technol. XXXX, XXX, XXX-XXX

the sulfation type for the studies described above. As can beseen, sorbents have previously been tested under significantlydifferent conditions, which makes it difficult to summarize theresults of previous work regarding the effects of metal salts onthe sulfation of limestone. Additionally, in previous works, thedose of the metal salt additive used to modify the limestoneranges from 0.6 mol % to 74 mol %, and the optimum additivedose is usually higher than 1.0 mol %. Such a high metal saltdose may lead to undesirable side effects in fluidized-bedboilers, such as ash deposition, bed agglomeration or corrosionand an increase in the cost of the sorbents compared with thatof raw limestone.In this work, an interesting phenomenon has been observed

that the sulfation conversion of limestone was greatly enhancedby treatment with trace amounts of Na2CO3 (∼0.1 mol %).Based on this phenomenon, we systematically investigated theinfluence of trace amounts of metal salt on the SO2 capturebehavior of limestone under various operating conditions,including different metal salt types, sorbent particle sizes, andlimestone types. Additionally, the underlying mechanism of theeffect of the trace metal salt additive on limestone sulfation waselucidated. This work promotes a new strategy for enhancingthe sulfation of limestone by the addition of trace amounts ofNa2CO3 and hence efficiently avoids the undesirable side effectscaused by the addition of high levels of inorganic salts.

2. EXPERIMENTAL SECTION

2.1. Sample Preparations. Two different Chinese lime-stones were used in these experiments. Table S2 summarizesthe X-ray fluorescence (XRF) analysis results of these materials.Limestone was crushed and sieved to three particle sizesranging 45−75 μm, 150−180 μm, and 425−500 μm for furtheruses. The Zhejiang limestone particles sized in the range of150−180 μm, which have a relatively higher residence time inCFB,20 were utilized in all tests except for investigating theeffects of geographical origins or particle size on the sulfation ofNa2CO3 treated sorbents.Na2CO3, CaCl2, and NaCl were used to modify the

limestones. The modified limestone with various metal saltdoses were synthesized by a wet impregnation method. Detailsof the synthesis protocol are provided in the Text S1. Themodified limestone was subsequently calcined for 0.5 h in N2 at850 °C, and characterized with regards to (i) surfacemorphology using scanning electron microscopy (SEM), (ii)surface area and pore volume using N2 adsorption apparatusand mercury injection apparatus, (iii) chemical composition

and crystal structure using X-ray powder diffraction (XRD)(the crystal structure was obtained by the whole pattern fittingand Rietveld (WPFR) refinement method with the Jade 6.0software.21,22), and (iv) thermal decomposition and SO2capture performance using a thermogravimetric analyzer(TGA). Details of the equipment and method used for particlecharacterization are given in the Text S2.

2.2. Sulfation Procedures. A thermogravimetric analyzer(TGA, Mettler Toledo TGA/DSC 1, America) and a speciallydesigned fixed-bed reactor (Figure S2) were used to investigatethe sulfation reaction performance of the samples. The detailsof the fixed-bed reactor are provided in the Text S3.All TGA tests followed the same procedure. Approximately

10 mg of the sample was loaded into the TGA in a corundumsample pan (5 mm diameters, 5 mm deep). A constant N2 flowof 100 mL/min was used as a purge flow over the microbalancefor 30 min before heating was conducted. The limestonesample was then heated to 850 °C, which is usually theoptimum sulfation temperature,23,24 at a rate of 10 °C /minunder N2 environment. Once the preset temperature wasreached, the TGA gas supply was switched to a synthetic fluegas containing SO2, CO2, O2, and N2. The flow rate wasmaintained at 100 mL·min−1 throughout the test, a flow rate ofwhich it has been verified that the mass transfer is not thelimiting step (i.e., a further increase of the gas flow rate resultedin no observable change in the measured weight vs time curvefor the given sample size of ∼10 mg). Table S3 provides theexperimental conditions. Samples were sulfated for 60 min, andthe sample mass and temperature were recorded continuously.The calculation formula of CaO conversion was shown in TextS4.

3. RESULTS AND DISCUSSION

3.1. SO2 Capture Behavior of Trace Na2CO3-TreatedLimestone. Many previous studies have investigated the effectof Na2CO3 dose on the sulfation of limestone. Wang et al.10

found that the sulfation conversion of limestone first increasesand then decreases as the Na2CO3 dose increases. Theconversion attains its highest value at the Na2CO3 dose ofapproximately 3.8 mol %. Figure 1a shows the CaO conversionof sorbents after 60 min of sulfation versus the Na2CO3 dose.The conversion ratio of CaO to CaSO4 was approximately 35%without the Na2CO3 additive after 60 min of sulfation, but itrapidly reached a higher value (42%) when 0.1 mol % Na2CO3was added to the limestone for the same duration. Upon afurther increase of the Na2CO3 dose to 0.5 mol %, the CaO

Figure 1. Effect of the Na2CO3 dose on the sulfation conversion of limestone: (a) calcined Zhejiang limestone, China, and (b) calcined limestones ofvarious geographical origins.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b04141Environ. Sci. Technol. XXXX, XXX, XXX−XXX

B

conversion of the sorbents decreased sharply to 26%. Then, theCaO conversion of the sorbents gradually increased as theNa2CO3 dose increased further from 0.5 to 4.0 mol %. Uponthe addition of 4.0 mol % Na2CO3, the CaO conversion of thesorbents reached a still higher value (41%), as reported by aprevious study.10 Besides, the same sulfation procedure wasperformed in a self-made fixed-bed reactor, and the enhance-ment effect of trace Na2CO3 on limestone sulfation was alsoobserved (Figure S3).Limestones from different geographical areas contain

significantly different impurities, which could function ascatalysts owing to their influence on the crystallinestructure.25−27 To verify, whether the effect of trace Na2CO3on sulfation is a phenomenon specific to one type of limestone,two kinds of limestone with different impurity contents werecompared, as shown in Figure 1b. As the Na2CO3 doseincreased, the CaO conversion levels for Zhejiang limestoneand Hunan limestone showed similar variation trends. Bothlimestones achieved a maximum sulfation conversion with theaddition of 0.1 mol % Na2CO3. This result means that thedifference in impurity content does not impact the enhance-ment effect of trace Na2CO3.The effect of particle size on the conversion ratio of Zhejiang

limestone treated with Na2CO3 was also investigated, and theresults are shown in Figure S4. As expected, the CaOconversion ratio increased as the particle size decreased.Moreover, the sorbents with different particle size all achieveda relatively higher CaO conversion by adding 0.1 mol %Na2CO3.3.2. CaO Crystal Structure Analyses. The sulfation

reaction is assumed to consist of two processes: an initialsurface reaction-controlled stage and a second diffusion-controlled stage with a CaSO4 layer coating the outer surfaceof the unreacted CaO particles.28 Hsia et al.27,29 designed inertmarker and reactive isotope experiments for the first time toillustrate that Ca2+ ions coupled with O2− ions move outward tothe CaSO4/gas interface to react with SO2 during the secondstage. This conclusion was also confirmed by Fan et al.30 andYang et al.25,31 Therefore, many researchers proposed that theenhanced sulfation by adding metal salt additive can beattributed to the promoted product layer diffusion byheterogeneous ion doping.10,32 However, such effects com-monly originate from relatively high Na2CO3 doses in previousstudies and the high CaO conversion induced by trace Na2CO3may suggest a different explanation.XRD measurements were performed to characterize the

chemical composition and the lattice distortion of the Na2CO3-

treated samples. The diffractograms are presented in Figure S5.Only the lime phase is present in the calcined samples. Thelattice distortion of Na2CO3-treated sorbents after calcinationwere obtained by the WPFR refinement method based on theXRD results to determine if the enhancement of limestonesulfation by trace Na2CO3 is due to heterogeneous ion doping.As shown in Figure 2a, the lattice distortion increases as theNa2CO3 dose increases. This phenomenon can be explained asfollows: as the Na2CO3 dose increases, more Na

+ replace Ca2+

in the lattice, and more oxygen vacancies are created, whichresults in the shrinkage of the crystal structure (Figure 2b).When the Na2CO3 dose was 0.1 mol %, the crystal structure ofCaO showed no noticeable change compared to that of theuntreated sample. The change in the crystal structure becameremarkable only when the Na2CO3 dose was greater than 0.5mol %. Thus, it can be concluded that the enhanced sulfation oftrace Na2CO3-treated sorbent is not caused by heterogeneousion doping.

3.3. Morphology and Pore Structure Analyses.3.3.1. Surface Micromorphology. Scanning electron micros-copy (SEM) images of untreated sorbents and sorbents treatedwith different Na2CO3 doses are shown in Figure 3. Figure 3a−d show the single-particle morphology of sorbents, while Figure3e−h show the enlarged local morphology of single-particlesorbents. The particles of untreated sorbent after calcinationconsist of aggregates of coral-like structures (Figure 3e).Comparatively, treating sorbents with Na2CO3 enlarges theirpore sizes, and as the Na2CO3 dose increases, the pore sizes ofthe sorbents gradually increase, as observed in Figure 3f−g,respectively. Moreover, the untreated sorbents appear to have acoarse surface, while the surface of the sorbents treated withNa2CO3 seem smoother.

3.3.2. Change in the Pore Structure Parameters. Both themercury intrusion porosimetry (MIP) and N2 adsorption wereused to determine the pore structure parameters of the sorbentstreated with different Na2CO3 doses. As small pores couldcollapse due to the injection of high-pressure mercury into thepores during MIP testing, only pores larger than 100 nm wereconsidered for the mercury intrusion method. In contrast, theN2 adsorption method has higher accuracy for micropores andmesopores, with measured pore sizes ranging from 2 to 100nm. Here, Smercury (100 nm) and Vmercury (100 nm) refer to thespecific surface area and specific pore volume derived from porediameters larger than 100 nm. SBET and VN2 refer to the specificsurface area, and specific pore volume derived from porediameters ranging from 2 to 100 nm. Figure 4 shows thechanges in the pore structure parameters of sorbents treated

Figure 2. (a) Effect of the Na2CO3 dose on the lattice distortion of the sorbents. (b) Schematic of lattice distortion caused by vacancy.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b04141Environ. Sci. Technol. XXXX, XXX, XXX−XXX

C

with different Na2CO3 doses. The trend for Smercury (100 nm)shows that pores larger than 100 nm increased with 0.1 mol %Na2CO3 treatment, and that, as the Na2CO3 dose increased, thepores larger than 100 nm rapidly decreased to almost zero. The

trend of SBET is similar to that of Smercury (100 nm), but SBET isslightly larger than Smercury (100 nm). Additionally, Vmercury (100nm) increases with Na2CO3 dose, while the VN2 shows theinverse trend. The value of Vmercury (100 nm) is much higherthan that of VN2, which means that the specific pore volume isdecided by macropores.Typical results in the previous literature show that Na2CO3

treatment leads to a decrease in surface area and the growth ofmacropores, usually attributed to sintering.5,8,10 In contrast, ourexperimental results show that 0.1 mol % Na2CO3 additivesignificantly augments the surface area of sorbents, which maybe associated with the enhancement of sulfation by traceNa2CO3 addition.

3.3.3. Change in the Pore Size Distribution. To furtherreveal the pore structure evolution of Na2CO3-treated sorbents,pore size distributions (PSDs) were obtained by both MIP andN2 adsorption analyses, as shown in Figure 5. The untreatedsorbent exhibits a bimodal PSD in the N2 adsorption method,revealing two peaks associated with pores ∼4 nm and ∼70 nmin diameter. Compared with untreated sorbents, the sorbenttreated with 0.1 mol % Na2CO3 shows an increasing populationof smaller pores associated with the ∼4 nm peak and adecreasing population of large pores associated with the ∼70nm peak. When the Na2CO3 dose was increased from 0.1 mol% to 4.0 mol %, only a few pores smaller than 3 nm remained,while the other pores almost disappeared. As the pore sizemeasuring range (<100 nm) of the N2 adsorption method islimited, the MIP method is more effective for analyzing large-scale macropores. Figure 5b illustrates the MIP method-basedPSD, in which the most probable pore sizes of the sorbentsfollow an increasing trend with increasing Na2CO3 dose.Specifically, the most probable pore size increases from 70 nm(untreated) to 140 nm (0.1 mol % Na2CO3) or even larger,reaching nearly 2000 nm at 4.0 mol % Na2CO3. From acomprehensive analysis of the N2 adsorption and MIP data, weconclude that at the 0.1 mol % Na2CO3, the population ofpores peaking at ∼70 nm is converted to those pores peaking at∼4 nm and ∼140 nm.Based on the above analysis, the formation and evolution of

pores during Na2CO3-treated limestone calcination is describedas follows. CaCO3 first decomposes on the surface of CaCO3particles. Then, a porous CaO layer forms and the CaCO3−CaO interface migrates to the interior of the particles.34 Withthe phase transformation from CaCO3 to CaO, the Na2CO3additive diffuses to the CaCO3−CaO surface. The sintering ofnewly formed CaO grains is enhanced by the action of Na2CO3.A model of the pore structure evolution during enhancedsintering is shown in Figure 5c. For the untreated sorbents, thepores with a diameter of approximately 70 nm presumablyoriginate from the release of CO2 during calcination of thelimestone.33 For the sorbents treated with 0.1 mol % Na2CO3,the sintering of the sorbent grains is slightly accelerated. A fewgrains stick together to form grain clusters, and a sintering neckis formed. Inside the grain clusters, small pores of ∼4 nm areproduced, whereas macropores of ∼140 nm are generatedbetween the grain clusters. For the sorbents treated with moreNa2CO3, a harsher sintering occurred. Small pores inside thegrain clusters disappear because of the growth of the sinteringneck, and macropores between the grain clusters are furtherenlarged. This model successfully explains the pore structureevolution of the Na2CO3-treated sorbents.

3.4. Mechanism Discussion. Figure 6 shows thecorrelations between CaO conversion, surface area and lattice

Figure 3. Effect of the Na2CO3 dose on the micromorphology of thesorbents: (a,e) untreated sorbents after calcination, (b,f) 0.1 mol %Na2CO3-treated sorbents after calcination, (c,g) 0.5 mol % Na2CO3-treated sorbents and (d,h) 4.0 mol % Na2CO3-treated sorbents. Theright column is local enlarge the image of the left column.

Figure 4. Effect of the Na2CO3 dose on the pore structure parameterof the sorbents.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b04141Environ. Sci. Technol. XXXX, XXX, XXX−XXX

D

distortion with different Na2CO3 dose. When the Na2CO3 doseis in the range of 0−0.5 mol %, the change trend of sulfationconversion is highly consistent with the change trend of theBET surface area. When the Na2CO3 dose is greater than 0.5

mol %, the change trend of the sulfation conversion becomesconsistent with the change trend of the lattice distortion degree.A potential explanation of the results is presented in Figure 7.

Part of this figure is based on a schematic in the literature (35).For the untreated sorbents, the pores ∼70 nm in diameter,which were obtained by calcination, are susceptible to poreblockage/plugging.35 For sorbents treated with 0.1 mol %Na2CO3, partial pores ∼70 nm in diameter can transform intopores ∼4 and 140 nm in diameter due to slight sintering. Thelarger pores are less susceptible to blockage, and the smallerpores provide a large reaction area, which enables higherconversion. For sorbents treated with 0.5 mol % Na2CO3, thelarger pores are further enlarged to approximately 400 nm, andthe small pores disappear due to severe sintering. The availablesurface area for the reaction decreases sharply, resulting inlower conversion. For sorbents treated with 4.0 mol % Na2CO3,the conversion is increased again, possibly because of thedramatically lowered product layer diffusion resistance providedby abundant Na+ doping, as confirmed in Figure 2.

Figure 5. Effect of the Na2CO3 dose on the pore size distribution of the sorbents: (a) N2 adsorption method, (b) MIP method and (c) Schematic ofthe pore structure evolution with Na2CO3 treatment.

Figure 6. Correlation between CaO conversion, surface area andlattice distortion with different Na2CO3 dose.

Figure 7. Schematic representation of the relationship between the Na2CO3 dose and conversion.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b04141Environ. Sci. Technol. XXXX, XXX, XXX−XXX

E

The measured sulfation reaction rates for the Na2CO3‑treatedsorbents (Figure S6) further support our explanation regardingthe relationship between the Na2CO3 dose and CaOconversion. For the sorbents treated with 0.1 mol % Na2CO3,the reaction rate in the chemical reaction-controlled stagedecays more slowly than that of untreated sorbents, which maybe interpreted as a reduction in pore blockage. Compared with0.1 mol % Na2CO3-treated sorbents, sorbents treated with 0.5mol % Na2CO3 have a lower reaction rate in the chemicalreaction stage, due to its low surface area, and a higher reactionrate in the diffusion stage. The higher reaction rate in thediffusion stage may be attributed to the almost uniformcoverage of the sorbent surface by the product layer due to theopen pore structure, which reduces the thickness and thediffusion resistance of the product layer. When the sorbents aretreated with 4.0 mol % Na2CO3, the reaction rate in thechemical reaction exceed that of the sorbents treated with 0.5mol % Na2CO3. The faster reaction rate confirms the conceptof a “catalytic effect”, and the higher reaction rates areconsistent with the finding of Li et al.,32 who reported anacceleration of the solid-state diffusion mechanism when thesorbents were treated with Na2CO3.3.5. Agglomeration of Sorbent Particles after Sulfa-

tion. Researchers have proposed that sodium salt could causesevere bed agglomeration. The agglomeration tendency ofZhejiang limestone treated with Na2CO3 was tested using themethod employed by Manovic et al.36 The Na2CO3-treatedsorbents were sulfated for 60 min in the fixed-bed reactor(Figure S2). The sulfated sorbents were directly taken out fromthe reactor, and shown in Figure S7. The sorbent particlestreated by 0.1 mol % Na2CO3 exhibited good dispersion aftersulfation. As the Na2CO3 dose increased, the agglomeration ofthe sorbent particles gradually deteriorated. When the Na2CO3dose is increased to 4.0 mol %, the sorbent particlesagglomerated into a pellet. This result directly demonstratesthat high doses of Na2CO3 trigger severe bed agglomeration.3.6. SO2 Capture Behavior of Limestone Treated with

Other Metal Salts. Metal salts of different chemical naturesbehave differently, and sulfate conversion is reported to be afunction of their presence in limestone.5,6 Herein, we furtherdetermine if other trace metal salts have the same enhancementeffect as trace Na2CO3 on limestone sulfation. Figure 8 showsthe effects of the CaCl2, NaCl and Na2CO3 dose on the

sulfation conversion of limestone. When the metal salt dose isincreased from 0.5 to 6.0 mol %, the change trend of the CaOconversion depends on the type of metal salt. However, as themetal salt dose increases from 0.0 to 0.5 mol %, the impact ofdifferent metal salts on the CaO conversion is almost the same.The limestone samples treated with 0.1 mol % Na2CO3, CaCl2,and NaCl all exhibit higher conversion, which means that theenhancement of limestone sulfation using trace additives hasrelatively broad applicability.Furthermore, we explore the reasons for the different effects

of CaCl2 and NaCl on the sulfation of limestone as the metalsalt dose ranges from 0.5 to 6.0 mol %. The apparentmorphology of the calcined sorbents treated with differentdoses of CaCl2 or NaCl is shown in Figure S8 and Figure S9. Asthe CaCl2 or NaCl dose increases, the pore sizes of the sorbentsgradually increase, which is similar to the SEM results for thesorbents treated with Na2CO3. The pore structure cannotexplain why the high dose CaCl2 and NaCl have different effectfor the sulfation of limestone. We further investigated the high-temperature stability of the sorbents treated with CaCl2 orNaCl, as shown in Figure S10. For the limestone treated withCaCl2 (Figure S10a), only one weight loss step, caused bydecomposition of the limestone, is observed. In contrast, for thelimestone treated with NaCl (Figure S10b), we observe twoweight loss steps. The second may be attributed to theevaporation of NaCl. This speculation is further confirmed bythe weight loss characteristics of pure NaCl (Figure S10c). Thevolatilization of NaCl leads to only a small amount of Na+

replacing Ca2+ in the sorbent lattice, and the product layerdiffusion is not improved significantly. Accordingly, theaddition of high amounts of NaCl results in low sulfateconversion of limestone.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.est.7b04141.

Information about preparation process of treatedsorbents, equipment and method used for particlecharacterization, experimental process of fixed-bedreaction (FBR), processes of calculation formulas, energydispersive spectrometer analysis of treated sorbents,curves of sulfation conversion with reaction time in FBR,sulfation conversion for different particle sizes of theNa2CO3-treated sorbents, XRD diffractograms ofNa2CO3-treated sorbents, sulfation kinetics of theNa2CO3-treated sorbents, photograph of the Na2CO3-treated sorbent powders after sulfation, micromorphol-ogy of the CaCl2-treated sorbents, micromorphology ofthe NaCl-treated sorbents, thermal decompositionperformance of the CaCl2-treated sorbents and NaCl-treated sorbents (PDF)

■ AUTHOR INFORMATIONCorresponding Author*Phone: (86)-0451-8641 3231; fax: (86)-0451-8641 2528; e-mail: gaojh@hit.edu.cn.

ORCIDJihui Gao: 0000-0003-4833-7836NotesThe authors declare no competing financial interest.

Figure 8. Effect of the CaCl2, NaCl and Na2CO3 dose on the sulfationconversion of limestone.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b04141Environ. Sci. Technol. XXXX, XXX, XXX−XXX

F

■ ACKNOWLEDGMENTS

This work was supported by the National Natural ScienceFoundation of China under Grant < No.51176043,No.91434134, No.51421063>.

■ REFERENCES(1) Anthony, E. J.; Granatstein, D. L. Sulfation phenomena influidized bed combustion systems. Prog. Energy Combust. Sci. 2001, 27,215−236.(2) Anthony, E. J.; Bulewicz, E. M.; Jia, L. Reactivation of limestonesorbents in FBC for SO2 capture. Prog. Energy Combust. Sci. 2007, 33,171−210.(3) Hu, G. L.; Dam-Johansen, K.; Wedel, S.; Hansen, J. P. Review ofthe direct sulfation reaction of limestone. Prog. Energy Combust. Sci.2006, 32, 386−407.(4) Jeong, S.; Kang, S. L.; Sang, I. K.; Jin, H. Y.; Yong, J. K.; Sang, S.K. Mechanisms of direct and in-direct sulfation of limestone. Fuel2015, 161, 1−11.(5) Laursen, K.; Grace, J. R.; Lim, C. J. Enhancement of the sulfurcapture capacity of limestones by the addition of Na2CO3 and NaCl.Environ. Sci. Technol. 2001, 35 (35), 4384−4389.(6) Fuertes, A. B.; Fernandez, M. J. The effect of metallic saltadditives on direct sulfation of calcium carbonate and ondecomposition of sulfated samples. Thermochim. Acta 1996, 276 (1),257−269.(7) Hu, G.; Dam-Johansen, K.; Wedel, S.; Hansen, J. P. Enhancementof the Direct Sulfation of Limestone by Alkali Metal Salts, CalciumChloride, and Hydrogen Chloride. Ind. Eng. Chem. Res. 2007, 46 (16),5295−5303.(8) Laursen, K.; Kern, A. A.; Grace, J. R.; Lim, C. J. Characterizationof the enhancement effect of Na2CO3 on the sulfur capture capacity oflimestones. Environ. Sci. Technol. 2003, 37 (16), 3709−3715.(9) Chen, C.; Ye, Z.; Wang, C. Enhancement of direct sulfation oflimestone by Na2CO3 addition. Fuel Process. Technol. 2009, 90 (7−8),889−894.(10) Wang, C.; Shen, X.; Xu, Y. Investigation on sulfation of modifiedCa-based sorbent. Fuel Process. Technol. 2002, 79 (2), 121−133.(11) Stouffer, M. R.; Yoon, H. An investigation of CaO sulfationmechanisms in boiler sorbent injection. AIChE J. 1989, 35 (8), 1253−1262.(12) Adanez, J.; Fierro, V.; García-Labiano, F.; Palacios, J. Study ofModified Calcium Hydroxides for Enhancing SO2 Removal DryingSorbent Injection in Pulverized Coal Boilers. Fuel 1997, 76 (3), 257−265.(13) Shearer, J. A.; Johnson, I.; Turner, C. B. Effects of sodiumchloride on limestone calcination and sulfation in fluidized-bedcombustion. Environ. Sci. Technol. 1979, 13 (9), 1113−1118.(14) Davini, P.; Demichele, G.; Ghetti, P. An investigation of theinfluence of sodium chloride on the desulphurization properties oflimestone. Fuel 1992, 71 (7), 831−834.(15) Houte, G. V.; Rodrique, L.; Genet, M.; Delmon, B. Kinetics ofthe reaction of calcium sulfite and calcium carbonate with sulfurdioxide and oxygen in the presence of calcium chloride. Environ. Sci.Technol. 1981, 15 (3), 327−332.(16) Houte, G. V.; Delmon, B.; Maon, J. C.; Dumont, P.Desulfurization of flue gases in a fluidized bed of modified limestone.J. Air Pollut. Control Assoc. 1978, 28 (10), 1030−1033.(17) Zhao, Y.; Lin, W. C. Multi-functional sorbents for thesimultaneous removal of sulfur and lead compounds from hot fluegases. J. Hazard. Mater. 2003, 103 (1−2), 43−63.(18) Zhang, K.; Li, X. S.; Li, W. Z.; Rohatgi, A.; Duan, Y.; Singh, P.;Li, L.; King, D. L. Phase transfer-catalyzed fast CO2 absorption byMgO-based absorbents with high cycling capacity. Adv. Mater.Interfaces 2014, 1 (3).140003010.1002/admi.201400030(19) Yang, X.; Zhao, L.; Xiao, Y. Effect of NaNO3 on MgO-CaCO3

Absorbent for CO2 Capture at Warm Temperature. Energy Fuels 2013,27 (12), 7645−7653.

(20) Mattisson, T.; Lyngfelt, A. A sulphur capture model forcirculating fluidized-bed boilers. Chem. Eng. Sci. 1998, 53 (6), 1163−1173.(21) Gong, L.; Yang, J.; Li, Y.; Mo, M.; Xu, C.; Lin, J. Solvothermalsynthesis and upconversion emission of monodisperse ultrasmallSrYbF nanocrystals. J. Mater. Sci. 2013, 48 (10), 3672−3678.(22) Drazin, J. W.; Castro, R. H. R. Phase Stability in Calcia-DopedZirconia Nanocrystals. J. Am. Ceram. Soc. 2016, 99 (5), 1778−1785.(23) O’Neil, E. P.; Ulerich, N. H.; Newby, R. A.; Keairns, D. L.Criteria for the selection of SO2 sorbents for atmospheric pressurefluidized-bed combustors. Electric Power Research Institute report,FP-1307, 1, 721−1.(24) Leckner, B.; Amand, L. E. Emissions from a circulating and astationary fluidized bed boiler: a comparison. Proceedings of the ninthinternational conference on FBC. ASME, Boston, MA; 1987. p 891−7.(25) Yang, H.; Wu, Z.; Qiu, X.; Wang, J.; Liu, J.; Guo, F. Ca2+ cationdiffusion through CaSO4 product layer during sulfur retained reactionwith CaO. Acta Chim. Sinica 2003, 61 (9), 1410−1415.(26) Duo, W.; Laursen, K.; Lim, J.; Grace, J. Crystallization andfracture: formation of product layers in sulfation of calcined limestone.Powder Technol. 2000, 111 (1−2), 154−167.(27) Hsia, C.; St. Pierre, G. R.; Raghunathan, K.; Fan, L. S. Diffusionthrough CaSO4 formed during the reaction of CaO with SO2 and O2.AIChE J. 1993, 39 (4), 698−700.(28) Duo, W.; Laursen, K.; Lim, J.; Grace, J. Crystallization andfracture: product layer diffusion in sulfation of calcined limestone.Powder Technol. 2004, 111 (1), 154−167.(29) Hsia, C.; St Pierre, G. R.; Fan, L. S. Isotope study on diffusion inCaSO4 formed during sorbent-flue-gas reaction. AIChE J. 1995, 41(10), 2337−2340.(30) Fan, L. S.; Jiang, P.; Agnihotri, R.; Mahuli, S. K.; Zhang, J.;Chauk, S.; Ghosh-Dastidar, A. Dispersion and ultra-fast reaction ofcalcium-based sorbent powders for SO2 and air toxics removal in coalcombustion. Chem. Eng. Sci. 1999, 54 (22), 5585−5597.(31) Yang, H.; Wu, Z.; Qiu, X. Development of mechanism of CaOsulfation reaction. J. Fuel Chem. Technol. 2003, 31 (1), 92−96.(32) Li, Z.; Liang, P.; Cai, N. Understanding the enhancement effectof Na2CO3 additive on the direct sulfation of limestone. Energy Fuels2015, 29 (1), 278−286.(33) Alvarez, D.; Abanades, J. C. Pore-size and shape effects on therecarbonation performance of calcium oxide submitted to repeatedcalcination/recarbonation cycles. Energy Fuels 2005, 19 (1), 270−278.(34) Li, Z. S.; Liang, P. T.; Cai, N. S. A rate equation theory for thepore size distribution of calcined CaCO3 in calcium looping. FaradayDiscuss. 2016, 192, 197−216.(35) Donat, F.; Florin, N. H.; Anthony, E. J.; Fennell, P. S. Influenceof high-temperature steam on the reactivity of CaO sorbent for CO2capture. Environ. Sci. Technol. 2012, 46 (2), 1262−1269.(36) Manovic, V.; Stewart, M. C.; Macchi, A. Agglomeration ofsorbent particles during sulfation of lime in the presence of steam.Energy Fuels 2010, 24 (12), 6442−6448.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b04141Environ. Sci. Technol. XXXX, XXX, XXX−XXX

G