Simultaneous Removal of SO2 and Trace As2

Mar 16, 2007 - Sulfur dioxide (SO2) and trace elements are pollutants derived from coal combustion. This study focuses on the simultaneous removal of ...
4 downloads 0 Views 265KB Size
Download PDF
Environ. Sci. Technol. 2007, 41, 2894-2900

Simultaneous Removal of SO2 and Trace As2O3 from Flue Gas: Mechanism, Kinetics Study, and Effect of Main Gases on Arsenic Capture YUZHONG LI,* HUILING TONG, YUQUN ZHUO, YAN LI, AND XUCHANG XU Key Laboratory for Thermal Science and Power Engineering of Ministry of Education, Department of Thermal Engineering, Tsinghua University, Beijing 100084, China

Sulfur dioxide (SO2) and trace elements are pollutants derived from coal combustion. This study focuses on the simultaneous removal of SO2 and trace arsenic oxide (As2O3) from flue gas by calcium oxide (CaO) adsorption in the moderate temperature range. Experiments have been performed on a thermogravimetric analyzer (TGA). The interaction mechanism between As2O3 and CaO is studied via XRD detection. Calcium arsenate [Ca3(AsO4)2] is found to be the reaction product in the range of 6001000 °C. The ability of CaO to absorb As2O3 increases with the increasing temperature over the range of 400-1000 °C. Through kinetics analysis, it has been found that the rate constant of arsenate reaction is much higher than that of sulfate reaction. SO2 presence does not affect the trace arsenic capture either in the initial reaction stage when CaO conversion is relatively low or in the later stage when CaO conversion is very high. The product of sulfate reaction, CaSO4, is proven to be able to absorb As2O3. The coexisting CO2 does not weaken the trace arsenic capture either.

Introduction This study attempts to understand the fundamentals of simultaneous removal of sulfur and trace arsenic by CaO from coal combustion flue gas in the temperature range of 600-1000 °C. Previous literature (1-5) has reported that arsenic exists as As2O3(g) in combustion environment when the temperature is higher than 900 K. Its high volatility makes arsenic emission control difficult. Mahuli et al. (6) found through experiments that lime could capture arsenic oxide by forming calcium arsenate [Ca3(AsO4)2] at 600 and 1000 °C. A subsequent study reported by Jadhav and Fan (7) showed that the interaction mechanism between As2O3 and CaO depended on the temperature. Tricalcium orthoarsenate (Ca3As2O8) is found to be the product of the reaction bellow 600 °C, whereas dicalcium pyroarsenate (Ca2As2O7) is found to be the product in the range of 700-900 °C. Experiments conducted by Sterling and Helble (8) on As2O3 vapor in contact with CaO, di-calcium silicate (2CaO‚SiO2), and mono-calcium silicate (CaO‚SiO2), respectively, over the temperature range 600-1000 °C indicated that these solids were capable of reaction with * Corresponding author email: [email protected]. 2894

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 2007

arsenic vapor species. Ca3(AsO4)2 was the reaction product in all of the samples analyzed and Ca2As2O7 was not observed in their experiments using 2CaO‚SiO2 and CaO‚SiO2 at 800 °C. Therefore, the interaction mechanism between As2O3 and CaO at 700-900 °C is not determined according to above conclusions. As a product of As2O3-CaO reaction, Ca2As2O7 was only found by Jadhav and Fan. This conclusion has not been confirmed by other researchers, which suggests further studies are needed. Also, the former researchers’ conclusions are not identical in the temperature dependence on arsenic capture by CaO. Both Mahuli et al. (6) and Jadhav et al. (7) claimed that the ability of CaO to absorb As2O3 increased with the increase of temperature in the range of 400-600 °C. However, in the range of 600-1000 °C, the conclusions are not the same: Mahuli et al. (6) and Sterling et al. (8) supported that the arsenic capture also increased with the increasing temperature in this range, while Jadhav et al. (7) gave the opposite conclusion. They explained that the formation of the unstable product, Ca2As2O7, was the probable reason for the decrease of arsenic capture in this temperature range. These divergent opinions also need to be clarified. Mahuli et al. (6), Jadhav et al. (7), and Sterling et al. (8) have developed kinetics studies of As2O3-CaO reaction. On this basis, the kinetic study of As2O3-CaO reaction and that of SO2-CaO reaction are studied, and their rate constants are analyzed in comparison in this research. As for the simultaneous removal of sulfur and trace elements, the interaction with each other is important since there may be competition between them for the active sites of CaO. As a fact, the concentration of trace elements in flue gas is so small that their presence can hardly affect the sulfur capture. Whereas, as one of the main gases in flue gas, the concentration of SO2 is several magnitudes higher than that of trace elements, the sulfate reaction shall be the main reaction in the simultaneous sorption process. Therefore, it is necessary to make clear how the high concentration of SO2 affects the trace arsenic capture. This problem has also been a concern to other researchers. Jadhav et al. (7) observed that there was no drastic reduction in the amount of arsenic captured by CaO in the presence of SO2 at 500 °C. The slower rate of the sulfate reaction at 500 °C was thought to be responsible for this noncompetitive effect. Agnihotri et al. (9) reported that SO2 could decrease the ability of CaO to capture trace selenium at 600 °C. The possible explanation was the plugging/blocking of sorbent pores due to the formation of a high molar volume CaSO4 product. The above two conclusions were presented as subsidiary contents at the end of their papers, and more detailed studies have been suggested. The effect of SO2 on selenium capture has been studied in detail by us previously (10, 11), and the following conclusions have been given: There are two disadvantageous factors for trace selenium capture with high concentration SO2 coexisting in the flue gas: (1) the formation of CaSO4 product layer on the surface of CaO will reduce the amount of available active sites; and (2) the high concentration of SO2 may compete against trace SeO2 for active sites. It has been proven that (10) the selenium capture is not affected by the single factor (1) because the mass transfer mechanism is in “outward growth mode” (12, 13) and the active sites may be abundant for trace selenium. It has also been proven that (11) as a single factor, factor (2) does not affect selenium capture either because the active sites are more than sufficient so that no competition will happen between sulfur and trace selenium before the formation of a continuous dense product 10.1021/es0618494 CCC: $37.00

 2007 American Chemical Society Published on Web 03/16/2007

FIGURE 1. Schematic of thermogravimetric analyzer (TGA) reactor system: (1) As2O3 vapor generator; (2) main TGA reactor; (3) scrubbers; (x) inlet of TGA protection gas (N2); (y) inlet of carrier gas (N2); (z) inlet of SO2 standard gas or pure CO2 gas and diluent gases. layer. It has also been proven that (11) when these two factors combine together, SO2 presence and the product layer diffusion resistance can jointly reduce the selenium capture when simultaneous sorption is carried out. The above conclusions about the effect of SO2 on selenium capture was studied in two different sulfate reaction stages: one was the initial stage in which the CaO conversion was relatively low and reaction rate was controlled by chemical kinetics (“CK stage” for short), the other was the later stage in which the CaO conversion was high and the reaction rate was controlled by product layer diffusion (“PD stage” for short). In this research, the effect of SO2 presence on trace arsenic capture is also studied in the CK stage and the PD stage respectively. CO2 is another major component in flue gas with about 12-15% concentration, which is much higher than the concentrations of SO2 and trace elements. Hou et al. (14) have performed a series of experiments on a pilot-scale CFBFGD system to study the effect of CO2 on SO2 removal in the moderate temperature range. It was found that the presence of CO2 could badly inhibit the sulfur capture by CaO at 450650 °C. They suggested that the suitable desulfurization temperature was above 700 °C to avoid the negative effect of the carbonate reaction. Concerning the effect of CO2 on trace elements capture, Ghosh-Dastidar et al. (15) claimed that the CO2 presence could reduce trace selenium capture after comparing the result obtained from the experiments using nitrogen as dilution and those using air as dilution. No literature has been found to report the effect of CO2 on trace arsenic capture. Therefore, this problem has been attempted in this research.

Experimental Section Apparatus. The reaction rate and the capability of CaO adsorbing SO2 and As2O3 are obtained by measuring the mass change of a fixed amount of solid reactant in a gas-solid reactor system for a specific experimental time. The schematic of the experimental assemblies is shown in Figure 1. It consists of three parts: As2O3 vapor generator, TGA reactor, and gas scrubber. The As2O3 vapor generator, with reference to Sterling et al.’s method (8), consists of a 24 mm o.d. vaporization quartz tube housed in a horizontal furnace. A boat is used to hold the solid arsenic (As2O3) inside the heated quartz tube. A 9 mm o.d. quartz pipe wrapped by heating tapes connects the

FIGURE 2. XRD patterns for products obtained from CaO-As2O3 reactions: (a) CaO exposed to 45 ppm As2O3 at 600 °C for 5 h [Ca3(AsO4)2 File No. 01-0993]; (b) CaO exposed to 45 ppm As2O3 at 800 °C for 1.5 h [Ca3(AsO4)2 File No. 73-1298]; (c) CaO exposed to 45 ppm As2O3 at 1000 °C for 1.5 h [Ca3(AsO4)2 File No. 73-1298]; and (d-g) pictograms published in the JCPDS files. outlet of the quartz tube to the main TGA reactor. The temperature of the heating tapes is controlled to be higher than that of the vaporization tube to avoid As2O3(g) condensation. The main reactor is a Dupont 951 TGA with a 24 mm o.d. quartz tube in its horizontal furnace. A platinum crucible is used to hold the sorbent. The weight signal can be recorded every 6 s. VOL. 41, NO. 8, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2895

FIGURE 3. Initial reaction rate of arsenate reaction depending on temperature. As2O3 concentration: 17 ppm. Diluent gases: 5% O2 and balanced N2. At the exit of the TGA reactor, gases pass through a latex pipe to a scrubbing assembly in which all residual toxic gases are removed by 7% HNO3 solution. Then the clean gases are vented to the atmosphere. The three gas inlets are labeled x, y, and z. Pure N2 (50 mL/min) is introduced through the inlet x as the TGA protection gas and 200 mL/min pure N2 is introduced through the inlet y as the carrier of As2O3. The inlet z is for other gases such as 20 mL/min pure O2, SO2 standard gas, and balanced pure N2. The total flow rate is 400 mL/min which is proven to be able to nagate the effect of bulk diffusion. Materials. Solid As2O3 of analytic purity is used as the vapor source. The concentration of As2O3 is quantified by As2O3 weight changed over a certain time period with a certain gas stream flow rate. The desired As2O3 concentration in the stream can be attained by regulating the temperature of the vapor generator. As calibrated previously, the temperature has been set as 160-195 °C to give the desired As2O3 concentrations. The CaO sample is obtained from the calcinations of analytic pure Ca(OH)2 at 600 °C for 30 min. Its BET specific area is 39 m2/g and mean particle size is 47 µm. The conversions of CaO in the experiments are calculated from the weight signals of TGA. Post-Sorption Sorbent Analyses. The amounts of arsenic captured, except those that can be obtained by the TGA signals, are determined by measuring the arsenic content in sorbent after experiments. The post-sorption samples are dissolved in 1:1 hydrochloric acid which is found to be able to leach out all the arsenic. The arsenic contents of the solutions are determined by inductively coupled plasmaatomic emission spectrometry (ICP-AES).

Results and Discussion Mechanism of Interaction between As2O3 and CaO. To understand the interaction mechanism between As2O3 and CaO, the product samples are analyzed by XRD. The reaction of CaO with higher concentration of As2O3 in carrier gas is carried out for longer duration at a specific temperature to get sufficient conversion of CaO. The products XRD patterns obtained at temperatures of 600, 800, and 1000 °C are shown in Figure 2 . The comparable XRD patterns provided in JCPDS files are also included. The diffraction patterns (Figure 2a-c) show the presence of calcium arsenate and unreacted CaO or Ca(OH)2. The product formed at 600 °C is consistent with the XRD pattern of the monoclinic crystalline phase of Ca3(AsO4)2 (JCPDS File No. 01-0933), whereas, at 800 and 1000 °C , the reaction product is rhombohedral crystal structure of Ca3(AsO4)2 2896

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 2007

FIGURE 4. Estimation of reaction order, m, with respect to As2O3 concentration and rate constant kAs.

FIGURE 5. Comparison of rate constants of arsenate reaction, selenite reaction, and sulfate reaction (kAs, kSe, and kS). (JCPDS File No. 73-1298). A comparison of these patterns with other Ca-As-O including calcium pyroarsenate (Ca2As2O7) diffraction patterns in the XRD database files indicates that only Ca3(AsO4)2 is present. The results at 600 and 1000 °C are almost consistent with those reported by Mahuli et al. (6), Jadhav et al. (7), and Sterling et al. (8). However, Jadhav et al. claimed that the product from the As2O3-CaO reaction at 700-900 °C was Ca2As2O7, while our result at 800 °C is still Ca3(AsO4)2. Unfortunately, Mahuli et al. (6) and Sterling et al. (8) did not specify the reaction product obtained from As2O3-CaO reaction at 700-900 °C. Neither our findings nor Jadhav et al.’s could be confirmed by any other previous researchers. The difference may arise from different reaction system configurations and need to be further investigated. In this study, the following equation is proposed as the mechanism of the interaction between As2O3 and CaO over the range of 600-1000 °C.

3CaO(s) + As2O3(g) + O2(g) f Ca3(AsO4)2(s)

(1)

Effect of Temperature on Arsenate Reaction Rate. Figure 3 gives the initial reaction rate of arsenate reaction at the temperature range of 400-1000 °C when CaO (3 mg) is exposed to 17 ppm of As2O3 in the diluent gas. It can be seen from the figure that the reaction rate increases with the increasing temperature. The rule of temperature effect on arsenic capture shown in Figure 3 can be confirmed by the

FIGURE 6. Comparison of arsenic capture by CaO sorbent in different cases: Case 1, individual sorption of As2O3 using fresh CaO (Blank value); Case 2, simultaneous sorption of As2O3 and SO2 using fresh CaO performed in the CK stage (stage A); Case 3, simultaneous sorption of As2O3 and SO2 using used CaO with the sulfate conversion of about 58% performed in the PD stage (stage B); Case 4, individual sorption of As2O3 using used CaO with the sulfate conversion of about 58% performed in the PD stage (stage B). Experimental conditions: temperature 800 °C; As2O3 concentration 13 ppm; SO2 concentration 700 ppm; sorption time for each case 30 min; diluent stream, O2 5% and N2 balance; flow rate 400 mL/min. conclusions of both Mahuli et al. (6) and Sterling et al. (8) . However, in the range of 600-1000 °C, Jadhav et al. (7) gave the different conclusion that the arsenic capture decreases with temperature increase. The possible reason they suggested was that the formation of the unstable product, Ca2As2O7, decreased the arsenic capture in this temperature range. Kinetics Study. According to the kinetic model developed by Agnihotri et al. (9), the initial reaction rate with regard to the As2O3-CaO reaction can be given by the following equation:

- rinit )

1 dx 3 dt

( )

init

) kAsCAs2O3m

(2)

Based on the above kinetic model, kAs is calculated according to experimental data in the range of 600-900 °C. The flow rate, 400 mL/min, has been proved to be high enough to eliminate the external mass transfer. A very small amount of raw sorbent sample (about 3 mg of CaO with diameter of 47 µm) is evenly scattered on the surface of the platinum crucible of TGA to form a very thin sorbent layer, in order to minimize the effect of mass transfer resistance among sorbent particles and keep the bed temperature as uniform as possible. The product layer diffusion resistance can be neglected for very small conversions as this is the case analyzed in initial rate studies. Under appropriate conditions, initial reaction rate can be applied to calculate the kinetic parameters. For a given reaction temperature, the linear relationship between ln(dx/dt) and ln(CAs2O3) can be represented as

(dxdt)|

ln

tf0

) ln(3kAs) + m ln(CAs2O3)

(3)

Many experimental results obtained under different gas concentration and temperature conditions are gathered in Figure 4. The slope of the line fitted to the data represents the reaction order for the arsenate reaction with respect to As2O3 concentration, m. The average slope of these three lines is 0.86, therefore, m ) 0.86.

The exponential value of the intercept of each line in Figure 4 can be used to calculate the rate constant, kAs, according to eq 3. The values of kAs are shown in Figure 5. The rate constants of selenite reaction and sulfate reaction (kSe and kAs), which have been calculated before (11), are also shown here for comparison. It can be found that the rate constant of arsenate reaction is much higher than those of selenite and sulfate reactions. Effect of SO2 Presence on Arsenic Capture. As mentioned in the Introduction, the effect of SO2 on arsenic capture is studied in two different sulfate reaction stages: the CK stage and the PD stage. An experimental scheme including four cases is designed as follows. The concentration of As2O3 is set as 13 ppm and that of SO2 is set as 700 ppm, which provides a wide concentration gap. Four cases are performed at 800 °C respectively: case 1, individual sorption of As2O3 by fresh CaO; case 2, simultaneous sorption of SO2 and As2O3 by fresh CaO; case 3, simultaneous sorption of SO2 and As2O3 by used CaO with about 58% conversion caused by sulfate reaction; and case 4, individual sorption of As2O3 by the used CaO with about 58% conversion caused by sulfate reaction. The sorption time for each case is 30 min. Case 1 is performed to get the blank value. In order to explain cases 2-4 more clearly, Figure 6a is provided. As a fact, case 2 is carried out in the stage A of the sulfate reaction which is approximately regarded as a part of the CK stage and cases 3 and 4 are performed in the stage B which is thought as a part of the PD stage. The amounts of arsenic captured in cases 1-4 are determined by ICP-AES and the results are given in Figure 6b. Through comparison, it can be found that the amounts of arsenic absorbed in these four cases are almost equal, which indicates that the presence of SO2 does not affect the ability of CaO to adsorb trace As2O3 in these cases. It has been mentioned in the Introduction that there may be two possible disadvantageous factors for trace elements capture in the presence of high concentration SO2: (1) CaSO4 product layer diffusion resistance; and (2) competition from high concentration of SO2. As a matter of fact, in above experiments, case 2 is designed to verify the effect of factor (2); case 4 is designed to verify that of factor (1); and case 3 is VOL. 41, NO. 8, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2897

FIGURE 7. TG curve of pure CaSO4 absorbing As2O3 at 800 °C. As2O3 concentration 63 ppm. Diluent gases 5% O2 and balance N2.

FIGURE 8. TG curve indicating the weight of CaO sample exposed to different gases: (1) 15% CO2 in diluent gases; (2, 3) only diluent gases: 5% O2 and balance N2. designed to verify that of the combination of factors (1) and (2). Among these cases, only case 3 gives the different conclusion from that of selenium capture. In case 3, the selenium capture will be reduced because of the joint effect of factors (1) and (2) (11), while the arsenic capture is not affected by these two factors. These two disadvantages may be counteracted by other factors which are beneficial to arsenic adsorption. First of all, it has been indicated by Figure 5 that the rate constant of arsenate reaction is much higher than that of sulfate reaction. This factor may indicate the priority of As2O3 molecules to combine with the limited CaO active sites and results in no obvious effect of SO2 presence on arsenic capture. So far, it is not clear whether the product of sulfate reaction, CaSO4, is capable adsorbing arsenic. An experiment using pure CaSO4 to absorb As2O3 is performed to investigate this problem. The pure CaSO4 is obtained from the calcinations of analytic pure CaSO4‚2H2O. The result is shown in Figure 7. It can be found that the weight of CaSO4 sample increases rapidly in the first few minutes, which indicates that As2O3 can be absorbed by CaSO4. However, CaSO4 does not appear to have any ability to absorb selenium in the similar experiment using CaSO4 to absorb SeO2 at 700 °C. It seems that CaSO4 can absorb As2O3 selectively, which may be an advantageous factor for the arsenic capture in the simultaneous sorption of SO2 and As2O3 by CaO. Effect of CO2 Presence on Arsenic Capture. Typically, the concentration of CO2 contained in coal combustion flue gas is about 12-15%. The first 30-min period in Figure 8 presents the TG curve of CO2-CaO reaction at 700 °C. The initial reaction rate is very fast and the reaction comes into 2898

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 2007

FIGURE 9. TG curve indicating the weight of CaO sample exposed to different gases: (1) 15% CO2; (2) 15% CO2 and 24 ppm As2O3; (3) 15% CO2; (4) only diluent gases, 5% O2 and balance N2. Temperature 700 °C.

the PD stage very soon. It suggests that the surface of CaO particles will be immediately covered by CaCO3 product layer if CO2 is present, which induces our suspicion whether the carbonate reaction may drastically stop the sorption of arsenic. A special experiment shown in Figure 9 is carried out to reveal the possibility of trace arsenic capture with the presence of CO2. CaO (8.18 mg) is exposed to different gases in the following four steps: (1) reaction between CaO and 15% CO2 for 2 h; (2) maintaining 15% CO2 and introducing 24 ppm As2O3 into the stream for 1 h; (3) shutting off As2O3 from the stream and maintaining 15% CO2 for 30 min; (4) removing CO2 from the stream and just keeping the diluent gases. The temperature of steps 1-4 is 700 °C. It can be determined from Figure 9 that the CaO conversion is as high as 81% at the end of step (1). The CaO particles have been covered by CaCO3 product layer, which results in a very slow reaction rate. When trace As2O3 is introduced in step (2), the slope of the TG curve increases obviously comparing with that in step (1), which suggests that some As2O3 is absorbed in step (2). Thus, the conclusion can be drawn that As2O3 can be absorbed by the sorbent even with the presence of high concentration CO2. This conclusion can also be reached when comparing the curve slope in step (2) with that in step (3). There is another evidence to support this conclusion: It can be seen from the second 30-min period shown in Figure 8 that, when CO2 is removed from the feed-in stream, the signal of changed weight reduces to zero rapidly, suggesting that all the absorbed CO2 decomposed. However, when CO2 is removed in step (4) as shown in Figure 9, the signal reduces too, but the final value is not as low as zero, which means that the arsenic being absorbed in step (2) still remains in the sorbent. That is, As2O3 still can be absorbed when most of CaO has been converted to CaCO3 and the coexisting high concentration of CO2 may not prevent trace As2O3 from accessing the limited active sites. In step (2) of Figure 9, the sorbent with such a high carbonate conversion should be unlikely to have unreacted CaO on the surface, therefore arsenic should not be absorbed by unreacted CaO. The possible sorption mechanism is presented as

CaCO3(s) T CaO(s) + CO2(g)

(4)

3CaO(s) + As2O3(g) + O2(g) f Ca3(AsO4)2(s)

(5)

ature range. It is an irreversible chemical sorption. Second, the reaction rate of arsenate reaction increases with the increase of temperature at 600-1000 °C, therefore, there is a common temperature window for the effective removal of SO2 and As2O3. Third, the presence of main gases, SO2 and CO2, does not appear to have any competitive effect on trace arsenic capture. Finally, the rate constant of arsenate reaction is much higher than that of sulfate reaction. According to the conclusion deduced before (11), a higher rate constant may result in a higher removal efficiency of the corresponding pollutant, it also can be deduced that the removal efficiency of arsenic should be higher than that of sulfur when the simultaneous removal of sulfur and trace arsenic by CaO is carried out by the moderate temperature dry flue gas desulfurization (MTD-FGD) technology.

Acknowledgments FIGURE 10. Amounts of arsenic captured by CaO exposed to the stream with presence and absence of CO2 for 30 min over the range of 600-1000 °C. (The arsenic contents are calculated according to the masses of original sorbents.) Gases: As2O3 13 ppm; CO215% (if present); diluent stream: O2 5% and N2 balance; flow rate 400 mL/ min. Equation (4) denotes that the carbonate reaction at 700 °C is a reversible reaction in a dynamic equilibrium. Although the conversion of CaO is as high as 81%, the active sites of CaO which have combined with CO2 molecules are not occupied by CO2 molecules forever. They are in a dynamic equilibrium process: combinative f dissociative f recombinative f re-dissociative. That is, not all CaO active sites combining with CO2 are unavailable for As2O3. Once one of them meets an As2O3 molecule at an interval of CaOCO2 combination, they will combine via the reaction shown as eq 5, and this CaO active site will withdraw from above dynamic equilibrium because eq 5 is an irreversible reaction. Namely, arsenic is absorbed by contesting the temporarily available CaO active site. At this rate, although the CaO particle is covered by a layer of CaCO3, it still can provide temporarily available CaO active sites for arsenic capture. The total reaction scheme which combines eqs 4 and 5 can be postulated as

3CaCO3(s) + As2O3(g) + O2(g) f Ca3(AsO4)2(s) + 3CO2(g) (6) Equation 6 indicates that As2O3 can substitute the absorbed CO2 to combine with CaO. To study the effect of CO2 on arsenic capture, experiments of exposing CaO for 30 min to the stream containing trace arsenic vapor with and without CO2 presence are carried out over the temperature range of 600-1000 °C. The concentrations of As2O3 and CO2 in the stream are 13 ppm and 15% respectively. The amounts of absorbed arsenic measured by ICP-AES are shown in Figure 10. It can be found that the arsenic capture does not decrease in the presence of CO2, which suggests that As2O3 can inevitably win the chances to combine with the temporarily available CaO active sites even though the coexisting CO2 has much higher concentration. A little synergistic effect of CO2 presence on arsenic capture can also be seen from Figure 10. For now, the reason is not clear yet for this phenomenon. General Analysis. Through the above studies, the following conclusions which are beneficial for trace arsenic capture in the simultaneous removal process have been obtained. First, in the moderate temperature range above 600 °C, the product of the As2O3-CaO reaction is calcium arsenate [Ca3(AsO4)2] which can exist stably in this temper-

This work is supported by the State Key Development Program for Basic Research of China (2006CB200301). The experiments are funded by Open Fund of the Laboratory Administration of Tsinghua University (LF20050489 and LF20060797).

Nomenclature CK stage sulfate reaction stage in which the reaction rate is controlled by chemical kinetics PD stage sulfate (or carbonate) reaction stage in which the reaction rate is controlled by product layer diffusion CAs2O3

concentration of As2O3

kAs

rate constant of arsenate reaction

kS

rate constant of sulfate reaction

kSe

rate constant of selenite reaction

m

reaction order with respect to As2O3(g) concentration

rinit

initial arsenate reaction rate

t

reaction time

x

CaO conversion

Literature Cited (1) Frandsen, F.; Dam-Johansen, K. Rasmussen, L. Trace elements from combustion and gasification of coal - an equilibrium approach. Prog. Energy Combust. Sci. 1994, 20, 115-138. (2) Chesworth, S.; Yang, G.; Chang, P. Y.; Jones, A. G.; Kelly, P. B. Kennedy, I. M. The fate of arsenic in a laminar diffusion flame. Combust. Flame 1994, 98, 259-266. (3) Ratafia-Brown, J. A. Overview of trace element partitioning in flames and furnaces of utility coal-fired boilers. Fuel Process Technol. 1994, 39, 139-157. (4) Bool, L. E. Helble, J. J. A laboratory study of the partitioning of trace elements during pulverized coal combustion. Energy Fuels 1995, 9, 880-887. (5) Hirsch, M. E.; Sterling, R. O.; Huggins, F. E. Helble, J. J. Speciation of combustion derived particulate phase arsenic. Environ. Eng. Sci. 2000, 17, 315-327. (6) Mahuli, S.; Agnihotri, R.; Chauk, S.; Ghosh-Dastidar, A.; Fan, L.-S. Mechanism of arsenic sorption by hydrated lime. Environ. Sci. Technol. 1997, 31, 3226-3231. (7) Jadhav, R. A. Fan, L.-S. Capture of gas-phase arsenic oxide by lime: Kinetic and mechanistic studies. Environ. Sci. Technol. 2001, 35, 794-799. (8) Sterling, R. O. Helble, J. J. Reaction of arsenic vapor species with fly ash compounds: kinetics and speciation of the reaction with calcium silicates. Chemosphere 2003, 51, 1111-1119. (9) Agnihotri, R.; Chauk, S.; Muhuli, S. Fan, L.-S. Selenium removal using Ca-based sorbent: reaction kinetics. Environ. Sci. Technol. 1998, 32, 1841-1846. VOL. 41, NO. 8, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2899

(10) Li, Y.; Tong, H.; Zhuo, Y.; Chen, C. Xu, X. Simultaneous removal of SO2 and trace SeO2 from flue gas: effect of product layer on mass transfer. Environ. Sci. Technol. 2006, 40, 4306-4311. (11) Li, Y.; Tong, H.; Zhuo, Y.; Wang, S.; Xu, X. Simultaneous removal of SO2 and trace SeO2 from flue gas: effect of SO2 on selenium capture and kinetics study. Environ. Sci. Technol. 2006, 40 (24), 7919-7924. (12) Hsia, C.; St. Pierre, G. R.; Raghunathan, K.; Fan, L.-S. Diffusion through CaSO4 formed during the reaction of CaO with SO2 and O2. AIChE J. 1993, 39, 698-700. (13) Hsia, C.; St. Pierre, G. R.; Fan, L.-S. Isotope study on diffusion in CaSO4 formed during sorbent-flue-gas reaction. AIChE J. 1995, 41, 2337-2340.

2900

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 2007

(14) Hou, B.; Qi, H.; You, C. Xu, X. CO2 effect on medium temperature flue gas desulfurization. J. Tsinghua Univ. (Sci & Tech) 2004, 44, 1571-1574. (15) Ghosh-Dastidar, A.; Muhuli, S.; Agnihotri, R. Fan, L.-S. Selenium capture using sorbent powders: mechanism of sorption by hydrated lime. Environ. Sci. Technol. 1996, 30, 447-452.

Received for review August 2, 2006. Revised manuscript received December 11, 2006. Accepted February 8, 2007. ES0618494