Ind. Eng. Chem. Res. 1988,27, 1951-1954 Salame, M. Am, Chem. SOC.,Diu. Org. Coat. Plast. Chem. 1976,36, 488-97. Spencer, M. S.;Stanley, W. L. Ag. Food Chem. 1984,2,1113-1118. Stevens, M. A.; Lindsay, R. C.; Libbey, L. M.; Frazier, W. A. Proc. Am. SOC.Hort. Sci. 1967, 91, 833-845. Walles, W. E. US Patent 3 613 957, 1971.
Walles, W. E. US Patent 3 740 258, 1973. Yasuda, H. Contemporary Topics in Polymer Science; Plenum: New York, 1979; Vol. 3, pp 103-123.
Received for review April 18, 1988 Accepted June 20, 1988
Reduction of Sulfur Dioxide with Methane over Activated Alumina John Sarlis and Dimitrios Berk* Department of Chemical Engineering, McGill University, Montreal, Quebec H3A 2A7, Canada
The reduction of SO2by CHI was studied in a gradientless quartz reactor a t a temperature range of 650-750 "C and at S02/CH4molar feed ratios of 0.5-2.5. The product mixture was found to contain H2S,COS, CO, C02, H20,and elemental sulfur. The rates of formation of the various participating species were found to increase with increasing temperature and methane content of the feed. Byproduct formation is minimized when molar feed ratios of S02/CH4are maintained above 2. Experimental yields were found to be higher than the ones obtained by thermodynamic analysis. Sulfur dioxide, which is widely accepted as the most important precursor of acid rain, is a sulfur compound that is produced in vast quantities by large fuel-consuming installations, such as power plants and non-ferrous metal smelters. A very large percentage of SO2 that is removed is converted to sulfuric acid, a product which is difficult to store and transport. Methods that involve reducing SOz to elemental sulfur appear promising since this product can be easily handled, shipped, and stored. Reducing agents that can be used for the reduction of SO2 include hydrocarbons, hydrogen, carbon monoxide, and some form of elemental carbon (Rosenberg et al., 1975). The reduction of SO2 with methane and other hydrocarbons has been known since the beginning of this century (Young, 1915). By the early 1940s, a pilot plant using natural gas as the reducing agent and producing 5 tons of sulfur/day was built at Garfield, UT. In 1970, Allied Chemical Corporation started the operation of a plant to reduce SO2in the waste gas from a sulfide roasting facility belonging to Falconbridge Nickel Mines Ltd. near Sudbury, Ontario, Canada (Hunter and Wright, 1972). The unit was designed to recover 450 tons/day of sulfur from 12% roaster gas. The catalysts used were developed by Allied specifically for SO2 reduction. Toward the middle of the 19709, the price of sulfur dropped while that of natural gas increased until this process was no longer economical. Since World War 11, a large number of papers and patents describing the use of methane and natural gas as reducing agents in both pilot plant and laboratory studies have appeared in the literature. The catalyst that was used in most of these studies was activated bauxite. A common feature of all these processes is that, in addition to elemental sulfur, the reaction products include large amounts of H2S and COS, necessitating the use of additional reactors for their treatment. Therefore, the success of a process for the reduction of SOzwill partly depend on the development of a reactor which will maximize the selectivity of elemental sulfur over HzS and COS. Although a large number of articles on the reduction of SOz with methane have been published, there is very little information concerning the kinetics of the reactions. The principal reaction for the production of elemental sulfur is CHI + 2S02 2[S] + COZ + 2HzO (1) where [SIrepresents the various sulfur species in the gas
phase. Helmstrom (1975) reported the rate law for this reaction over a bauxite catalyst at a temperature range of 500-600 OC. Although the selectivity of elemental sulfur at this temperature range is high, the rate of the reaction is relatively low, resulting in low overall conversion. A study of the thermodynamic equilibrium for this reaction shows that at higher temperatures, in addition to elemental sulphur, the product mixture may contain H2S,COS, CS2, CO, H2, HzO, and C02. The objective of this paper is to discuss the effect of the feed composition and temperature on the rates of reaction and on the selectivity of the high-temperature catalytic reduction of SO2with CHI over an activated alumina catalyst.
Materials and Methods Experimental System. The experimental setup consists of three separate sections: the gas supply section, the main reactor, and detection and analysis section. Highpurity gases are supplied from compressed gas cylinders to a multiple gas flowmeter-mixer unit. The tubular reactor is fabricated from a 2.50-cm-0.d. quartz tube. The entire reactor is mounted inside a Lindberg Heavy Duty 1500 single-zone tubular furnace. The reactor, which is 63 cm long, consists of three zones. The inlet or the preheating zone (35 cm long) is packed with quartz chips and quartz wool, the reaction zone (7 cm long) is packed with 1.20-2.40-mm activated alumina catalyst particles, and the discharge or outlet zone is packed with quartz wool and quartz chips mainly for the purpose of shortening the residence time of the gaseous products in the reactor. Three thermocouples, fitted along the axis of the reactor reaching the beginning, middle, and end of the catalytic packing, provide measurement of the temperature profile along the catalytic bed. A fourth thermocouple is fitted at the exit of the reactor. A t the end of this last section, discharge lines bring the gaseous products to a heated six-port sampling valve which is used to inject the products of the chemical reactions into the gas chromatograph. Between the reactor exit and the sampling valve, the gases are stripped from any sulfur and water as they pass through an on-line trap cooled in an ice bath. Manual sample withdrawal and subsequent injection into the gas chromatograph is also possible through sampling ports which are installed at the exit of the reactor before and after the on-line trap. Finally, the exit gases from the valve pass into a scrubber containing concentrated NaOH, with an option to be bypassed to a nearby hood where the exit
0888-5885/88/2627-1951$01.50/0 0 1988 American Chemical Society
1952 Ind. Eng. Chem. Res., Vol. 27, No. 10, 1988
gas flow rate can be measured by using a soap bubble flowmeter. The inlet and exit gases are analyzed by using a gas chromatograph equipped with a column switching valve and a thermal conductivity detector. The carrier gas is helium. The detector was calibrated for each gas by determining the response area of different volumes of the pure gases which were injected manually by means of a gas-tight syringe. Water that was condensed in the sulfur trap prevented an accurate weight determination of the sulfur. Even though it is not presently possible to analyze quantitatively for water and elemental sulfur, it is still possible to calculate their molar flow rates through material balances as shown in the following section. Calculation of Rates. When SO2is reduced with CH4, the reaction mixture at the exit of the reactor may contain elemental sulfur, H2S, COS, C02, CO, H20, Hz, 02,and carbon in addition to the unconverted reactants, SO2 and CH4. Irrespective of any chemical mechanisms that may govern the reacting system, the following is a set of independent stoichiometric equations describing the formation of each of the above species: 1. so2 [SI + 0 2 2. CHI C + 2H2 3 H2 + S+ H2S
5. 6. 7.
H2 + 7 2 0 2 ---* H20 c + 0 2 c02
c + 7202 c + 7 2 0 2 + s cos
Equations relating the extents of the above reactions to the inlet and outlet molar flow rates of each species i can be written as follows: Fi
= Fi, +
where Fiis the exit molar flow rate of i (mole/second), Fi0 is the inlet molar flow rate of i (mole/second), vi, is the stoichiometric coefficient of the ith species in the jth stoichiometric equation, and ~j is the reaction extent of the jth reaction. The expression relating the rate of production of species i to the molar flow rate for a differential reactor is ri = AFi/AW (3) where AW is the weight of the catalyst. Hence, the rate in terms of the extent of each reaction is given by (4) Our analytical system is calibrated to measure the concentrations of SO2, CH4, H2S, COS, C02, and CO. Furthermore, as no carbon, hydrogen, and oxygen are found among the reaction products, the extents of the above reactions can be calculated from eq 2 and the hydrogen balance. The rates of production then can be calculated from eq 4.
Results and Discussion The reduction experiments were carried out at five temperature levels (650,675,700,725, and 750 "C) and five different molar feed ratios (SOZ/CH4= 0.5, 1.0, 1.5, 2.0, and 2.5). The feed mixture consisted of 45% argon with the appropriate amounts of SO2 and CHI. The inlet composition for the different ratios is shown in Table I.
Table I. Inlet Gas Composition at Different Molar Feed Ratios R = SOJCH, Ar. % so,. % CH,. % 0.50 45 18.33 36.67 1.00 45 27.50 27.50 1.50 45 33.00 22.00 2.00 45 36.67 18.33 2.50 45 39.33 15.67 8
1 .o I .6 2.0 2.6 MOLRR FEED RATIO Figure 1. Effect of the molar feed ratio (S02/CH,) on the rate of depletion of SO2 at different temperatures. 0.6
The volumetric flow rate was about 5.7 mL/s (at STP), well above the experimentally determined threshold for external film diffusion interferences. The total mass of the activated alumina catalyst was varied between 1and 17 g in order to maintain differential behavior in the reactor. Because of the low reactant conversions obtained in differential reactors, the variation in the initial feed composition reflects the effect of the average reactant concentrations on the rates of the reaction. The initial specific surface area of the supplied catalyst (measured by the BET method) was 125 m2/g. After 6 h of thermal treatment at 750 "C in the presence of flowing argon, the specific surface area stabilized at 93 m2/g and showed no further change during the course of the chemical reactions. In order to verify that no homogeneous reduction was taking place in the reactor, the reactants were introduced to the empty reactor at (S02/CH4)feed ratios of 0.5 and 2.0 and at 750 and 800 "C. The same procedure was repeated with the reactor filed with quartz chips at the same temperatures and molar ratios. The maximum observed conversion of SO2was 5% at 800 "C and at a ratio of 0.5. At 750 OC any homogeneous reactions taking place were below the detection limit. Thus, the contribution of the homogeneous reactions to the overall conversion was neglected in the analysis of the results. The experimental results for the rates of the production of the main species (as calculated from eq 4) are shown in Figures 1-7. Figures 1 and 2 show the rates of disappearance of the reactants, SO2 and CHI. In this set of experiments, the overall conversions of both reactants were below 20%, maintaining differential behavior. Below 675 "C, the rates of both reactants are relatively insensitive to the molar feed ratio; however, the rates at the small ratios (Le., when methane is in excess) show a large increase with increasing temperature, while they are less sensitive to temperature at the higher ratios. Figures 3,4, and 5 show the rates of production of elemental sulfur (expressed as the monoatomic species), H2S, and COS. The main sulfur-containingspecies is elemental sulfur, whereas H2S is the major byproduct under all conditions. Significant amounts of COS are only produced
Ind. Eng. Chem. Res., Vol. 27, No. 10, 1988 1953
T-700°C T-725'C o T-750"C
1 .o I .6 2.0 2.6 MOLAR FEED RATIO Figure 5. Effect of the molar feed ratio (S02/CH4) on the rate of production of COS a t different temperatures.
MOLAR FEED RATIO
Figure 2. Effect of the molar feed ratio (S02/CH4)on the rate of depletion of CHI a t different temperatures.
T-700% T=7E°C o T-750°C
.o I .6 2.0 2.6 MOLW FEED RATIO Figure 3. Effect of the molar feed ratio (S02/CH4)on the rate of production of elemental sulfur a t different temperatures. 0.0
T-700OC T-725'C o T-7500C
Figure 6. Effect of the molar feed ratio (S02/CH4)on the rate of production of COP at different temperatures.
MOLAR FEED RATIO
8 8 2.0-
MOLAR FEED RATIO
MOLRR FEE0 RATIO
Figure 4. Effect of the molar ratio (S02/CH4) on the rate of production of H,S a t different temperatures.
Figure 7. Effect of the molar feed ratio (S02/CH4)on the rate of production of CO a t different temperatures.
above 675 "C at a lower rate than that of HzS. Byproduct formation is consistently greater at higher temperatures and low molar feed ratios. Figures 6 and 7 show the rates of COz and CO. According to eq 1,carbon dioxide is the main carbon-bearing product in the system. The rates of production of COz exhibit the same characteristic features as the rates of CHI disappearance, while the rates of CO production are appreciable only above 725 OC. The hydrogen and oxygen content of the inlet gases appears primarily as water vapor, as no hydrogen or oxygen was found among the products. Analysis of the reaction products before the sulfur traps has also shown that no CSz is produced in the reactor. Furthermore, elemental balances for carbon closed at about
99%; thus, any solid carbon formation due to the cracking of CH4 is insignificant. Tables I1 and 111show the equilibrium compositions obtained by a thermodynamic analysis of typical feed mixtures containing 45% argon and appropriate amounts of SOz and CH4 to make the corresponding molar feed ratios (SOZ/CH4)at 627 and 727 "C. The mole fraction of elemental sulfur at equilibrium was found to be significant only at a feed ratio greater than 1.5 at both temperatures, while that of H2Speaks at a feed ratio of 1. The other important sulfur-containing compounds, COS and CSz, are at much lower concentrations than HzS. Below a molar ratio of 1, significant amounts of hydrogen and carbon monoxide are found in the equilibrium mixture.
1954 Ind. Eng. Chem. Res., Vol. 27, No. 10, 1988 Table 11. Mole Fraction of Product SDecies at Eauilibrium (T= 627 OC, Pressure = 1 atm) R species 0.50 1.00 1.50 2.00 0.086 0.173 0.165 X ' 0 1 0.197 X [SI 0.283 X 0.046 0.530 X 10'' 0.882 X SO2 0.188 0.092 0.201 0.134 H2S cos 0.1460 X lo-' 0.607 X lo-' 0.262 X 0.902 X 0.923 X lo4 0.774 X ' 0 1 0.849 X 0.118 X CS2 0.109 0.186 0.158 0.528 X 10-1 CO2 0.212 x 10-3 0.854 X 0.139 co 0.963 X lo-' 0.225 0.187 0.506 X lo-' 0.681 X lo-' HzO 0.199 x 10-2 0.696 X loe3 0.148 H2 0.287 0.387 0.391 0.339 Ar 0.333 0.265 X lo-' 0.709 X 0.283 X 0.251 X CH4 0.117 X 0.890 X lo-'' 0.105 X 0.564 X 0 2 0.797 X 10" 0.308 0.637 yield 0.145 X 10'' 100 99.0 85.45 XSO, 100 ~
2.50 0.178 0.100 0.061 0.548 X 0.500 X lo4 0.138 0.127 X 0.215 0.457 X 0.397 0.447 X 0.222 x 1 0 4 7 0.720 71.35
Table 111. Mole Fraction of Product Species at Equilibrium ( T = 727 OC, Pressure = 1 atm) species
0.50 0.137 X 10'' 0.133 X lo4 0.122 0.217 X 0.555 X 10'' 0.019 0.184 0.024 0.331 0.306 0.011 0.407 X 0.110 x 10-3 100
1.00 0.303 X 0.806 X 0.210 0.530 X lo-' 0.561 X '0 1 0.115 0.095 0.101 0.120 0.353 0.646 X 10'' 0.526 X 0.141 X 100
Table IV. Comparison of Equilibrium and Experimental Yields exptl yields a t equilibrium yields a t R 627 OC 727 OC 65OOC 725 OC 0.50 0.145 X 10'' 0.110 X 0.84 0.84 1.00 0.797 X 10" 0.141 X lo-' 0.90 0.86 1.50 0.308 0.341 0.90 0.84 2.00 0.637 0.682 0.92 0.88 2.50 0.720 0.762 0.96
The mole fraction of these compounds shows an increase with increasing temperature. Elemental oxygen is absent from the system. Finally, elemental carbon is produced at a feed ratio less than 1.0 and 0.5 at 627 and 727 "C, respectively. The equilibrium compositions described above indicate that the yield of elemental sulfur (moles of [SIproduced per mole of SOz utilized) increases with increasing feed ratio. As shown in Table IV, the experimentally obtained yields are always greater than the ones predicted at equilibrium. When the feed ratio is below 1.0, the experimental yield is above 84%, whereas that predicted at equilibrium is negligible. Clearly, since the reador operates differentially, the multiple reactions taking place are not at equilibrium. In fact, from the point of view of implementing the process in industry, the reactions should not be allowed to reach equilibrium as the selectivity for elemental sulfur is improved when the process is controlled by the reaction rates. Summary The rates of formation of the various species participating in the reduction process were found to decrease with
1.50 0.096 0.214 X 0.178 0.373 X 0.177 X lo-' 0.181 0.384 X low2 0.191 0.587 X lo-' 0.385 0.159 X lo-' 0.799 X 0.341 99.24
2.00 0.188 0.042 0.081 0.120 x 10-2 0.211 x 10-5 0.156 0.886 X 0.233 0.191 x 10-2 0.389 0.103 X lo-" 0.112 x 10-16 0.682 86.82
2.50 0.190 0.095 0.052 0.703 X 0.830 X 10" 0.137 0.516 X 0.222 0.121 0.395 0.162 X 0.253 X 0.762 72.52
increasing the SOz to CHI molar feed ratio. The distribution of byproducts, namely HzS, COS, and CO, can be minimized by keeping the SO2to CHI feed ratio above the stoichiometric value of 2. No CSz, H2, or O2 was found among the products. The yield for elemental sulfur decreases with increasing temperature and decreasing SO2 to CH, feed ratio. Finally, experimental yields were found to be consistently higher than ones obtained by thermodynamic analysis, because the chemical reactions taking place when the reactor operates at steady state are not in equilibrium. Acknowledgment The authors express their appreciation to the Natural Sciences and Engineering Research Council of Canada for providing financial support for this project. Registry No. H2S, 7783-06-4; COS, 463-58-1; CO, 630-08-0; COZ, 124-38-9; HzO, 7732-18-5; S, 7704-34-9; SOZ, 7446-09-5; A l 2 0 3 , 1344-28-1; CHI, 74-82-8.
Literature Cited Helmstrom, J. L. "Kinetics of the Reduction of Sulfur Dioxide by Methane over an Activated Bauxite Catalyst". Ph.D. Dissertation, The University of Akron, Akron, OH, 1975. Hunter, W. D.; Wright, J. P. 'Sulfur Dioxide Converted to Sulfur in Stack gas Cleanup Route". Chem. Eng. 1972, 79(4), 50-51. Rosenberg, H. L.; Engdahl, R. B.; Oxley, J. H.; Genco, J. H. "The Status of SO2 Control Systems". Chem. Eng. Prog. 1975, 71(5), 66. Young, S. W. "The Thiogen Process for Removing Sulfur Fumes". Trans. AIChE 1915,8, 81. Received for review January 4, 1988 Accepted July 5, 1988