sign in sign up
<<
Home , Sulfur dioxide

Chemical and Biological Engineering Publications Chemical and Biological Engineering

1960 Reductive of Gypsum by Monoxide Thomas D. Wheelock Iowa State University, [email protected]

D.R. Boylan Iowa State University

Follow this and additional works at: http://lib.dr.iastate.edu/cbe_pubs Part of the and Reaction Engineering Commons, Complex Fluids Commons, and the Other Chemical Engineering Commons The ompc lete bibliographic information for this item can be found at http://lib.dr.iastate.edu/ cbe_pubs/273. For information on how to cite this item, please visit http://lib.dr.iastate.edu/ howtocite.html.

This Article is brought to you for free and open access by the Chemical and Biological Engineering at Iowa State University Digital Repository. It has been accepted for inclusion in Chemical and Biological Engineering Publications by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. Reductive Decomposition of Gypsum by

Abstract Tremendous domestic reserves of gypsum and constitute a potential source of raw material f i x -based chemicals. As in Europe today, may become one of our principal raw materials for . Several European acid plants are based on a process in which sulfur dioxide is freed from anhydrite by heating the latter with and to a sintering temperature (4). The uls fur dioxide is converted into acid and the clinker is used for portland .

Disciplines Catalysis and Reaction Engineering | Complex Fluids | Other Chemical Engineering

Comments Reprinted (adapted) with permission from Ind. Eng. Chem., 1960, 52 (3), pp 215–218. Copyright 1960 American Chemical Society.

This article is available at Iowa State University Digital Repository: http://lib.dr.iastate.edu/cbe_pubs/273 T. D. WHEELOCK and D. R. BOYLAN I Department of Chemical Engineering, Iowa State University of Science and Technology, Ames, Iowa Reductive Decomposition of Gypsum by Carbon Monoxide

Sulfur dioxide and lime can be produced from gypsum under newly determined conditions. Process provides a new route for manufacturing sulfuric acid from gypsum

TREMExDOcS domestic reserves of gyp- carbon monoxide at elevated tempera- ditions at atmospheric pressure. calcium sum and anhydrite constitute a potential tures to produce sulfur dioxide and lime. cannot exist in the presence of source of raw material fix sulfur-based The solids do not sinter and the lime may calcium sulfate because of Reaction 3. chemicals. As in Europe today, calcium be a by-product of value. Because this Of course, the kinetics of this reaction sulfate may become one of our principal by-product can be disposed of in more may be unfavorable. raw materials for sulfuric acid. Several ways than portland cement, the process To show that a is European acid plants are based on a should be more flexible than the Euro- needed for decomposing calcium sul- process in which sulfur dioxide is freed pean process. fate, equilbrium constants for Reaction from anhydrite by heating the latter with 4 are included. coke and shale to a sintering tempera- Thermodynamics ture (4). The sulfur dioxide is converted The principal reaction (Reaction 1) is Experimental into acid and the clinker is used for endothermic and therefore favored by portland cement. higher temperatures. Reaction 2, which For this study natural gypsum having In the work described here a simpler is undesirable, is exothermic and is the following composition was used. process for freeing sulfur dioxide \vas favored by lower temperatures and high The gypsum was crushed and then sep- investigated. Calcium sulfate reacts carbon monoxide partial pressures. arated by Tyler standard sieves into with a gaseous reducing agent such as Above 2100' F., under equilibrium con- narrow-size fractions. The -7f8-mesh fraction was used for most runs. Both commercial and chemically pure grades of carbon monoxide were used. Calculated Equilibrium Constants and Heats of Reaction for Reactions Involved in the Process The was specified as 99.9791" pure, and the sulfur dioxide was AHR, grade. The liquid , Cal. ' Log,oK Mole which after vaporization served as a Reaction 1200' K. 1400' K. 1600' K. 1400' K. 1. CaS04 + CO = CaO + SO2 + CO? 0.31 1.48 2.28 43,400 ,~_~_- 2. CaSOa + 4CO = Cas + 4CO: 7.92 6.69 5.66 -48,400 - 3. 3CaSO4 CaS = 4Ca0 4.502 -6.68 -0.77 3.44 222,200 + + - rubber tubing 4. CaSO, = CaO + SOZ + 1/2 02 -7.42 -4.51 -2.38 110,600

lnsulollng firebrick

Crushed inrulotfng firebrick

Leco zircon lube

Alumina bolls -3+4 mesh Rotameter SO2 Gypsum particlar Crromel-olumel !hermOCOUPle

I! Liqusd N2 The reactor (diameters, 0.75 to 1.13 Carbon monoxide and dioxide, sulfur dioxide, and nitrogen, either alone or in inches) was suspended inside the - combination, were continuously metered, mixed, and passed through the reactor fired muffle furnace from the triple- containing gypsum beam balance

VOL. 52, NO. 3 MARCH 1960 215 Results Effect of Temperature. The effect Average Gypsum" Composition (1) of temperature on the desulfurization Calcium sulfate in the gypsum could be Constituent Weight To rate and on the formation of calcium sul- quantitatively decomposed by passing a fide was studied between 2100' and Hz0 19.6 stream of nitrogen over the gypsum CaO 30.9 2300 ' F. using various gas compositions. MgO 0.1 heated to about 2200' F. However, When the gas mixture fed to the reactor adding as little as 1% SOa 45.1 of the decomposi- contained 3y0 carbon monoxide, results c02 0.7 tion products, sulfur dioxide and , were obtained as shown in Figure 2. At R208 0 to the nitrogen prevented the decomposi- Si02 3.3 2110" F. the gypsum passed through an NaCl -0.3 tion. If several per cent of carbon mon- initial induction period where little or Total 100.0 were also present, calcium sulfate no decomposition occurred. The reac- decomposed in the presence of as much as a Cnited States Gj-psum Co., Fort Dodge, tion rate soon increased, and a relatively Iowa. 7y0 sulfur dioxide. constant but rapid desulfurization rate Generally, calcium sulfate was con- was established. At the end the gypsum verted to . At times cal- was 8770 desulfurized, and the solids con- cium sulfide was produced. Conse- diluent for the gaseous reactants, con- tained 11yo . Increasing tained less than 0.1% oxygen. quently the criteria chosen for comparing the temperature to 2200" F. increased The reactor, usually charged with a the effect of operating conditions were the the initial desulfurization rate: but the bed of gypsum 1 inch thick, was sus- rate of desulfurization and the concentra- maximum rate was unaffected. The tion of calcium sulfide in the residual pended inside the preheated furnace, and total desulfurization was increased to al- solids. The conversion to sulfide was ap- as its temperature rose, a mixture of sul- most looyo, and no calcium sulfide was proximately proportional to the percent- fur dioxide and air was passed through to found in the solids. When the tempera- age of calcium sulfide in the residue. prevent the gypsum from decomposing. ture was raised to 2310' F., the initial As operating temperature was ap- The total desulfurization or conversion rate was the same as for 2200' F.? but of calcium sulfate to calcium oxide was proached, nitrogen and carbon dioxide after 8 minutes the rate fell to a much calculated from the composition of the were added. When the reactor tempera- lower but constant value. final solids. By assuming that the ture had leveled out, the flow of air was i2'hen the gas fed contained 47, car- instantaneous conversion was propor- stopped and the flow of carbon monoxide bon monoxide, the initial and maximum tional to the weight lost by the gypsum was started. This marked the beginning rates (Figure 3) and the per cent calcium of a run. At regular intervals of 1 to 10 charge, desulfurization curves such as sulfide in the residue (Figure 4) varied minutes, depending on the rate of de- those in Figures 1 and 2 were plotted. ivith temperature. composition, the reactor weight was These curves usually had either one or two LVhen sulfur dioxide was excluded from noted. Operating conditions were kept constant rate periods for which the rates constant. After the reactor reached a could be reasonably correlated with the gas fed, the maximum desulfuriza- constant weight, it was slowly withdrawn reaction conditions. Generally the tion rate reached a peak value at about from the furnace, while nitrogen and a greater part of the desulfurization took 2150' F. with 4% carbon monoxide and small amount of carbon monoxide Lvere place at a constant rate which corre- 2250' F. with 2%. The variation in the passed through it. sponded to the maximum rate. rate kvith temperature was much greater Cndecomposed sulfate in the residual solids was determined gravimetrically (7). Sulfide and calcium were deter- mined iodometrically and by Venenate (2), respectively. loon90

'oo--m2400

IDewIfurization s F =- 60 2240 2 0 L c 0 2 50 2200 a E 1 E 40 2160 f D

Gor Cornpoiition

t -~ 2040

Figure 1. The desulfurization pro- Time, min. ceeded at a constant rate for much of the run when mixtures of carbon mon- Figure 2. When sulfur dioxide was present in feed gas containing less than oxide and nitrogen were fed 5y0 carbon monoxide, S-shaped desulfurization curves were obtained

2 16 INDUSTRIAL AND ENGINEERING CHEMISTRY GYPSUM DECOMPOSITION

Figure 3. The maxi- mum desulfurization rate reached a peak value at about 2200' F.

0' 2100 220'3 2300 Temperature during maximum

desulfurization rate period. OF. Oi 2100 2300 2200 Figure 4. Production of calcium SUI- Temperature, OF: fide can be minimized by using high temperatures for 470 carbon monoxide than for 2y0. mum desulfurization rates varied linearly 20 -r - _-- Again an increase in temperature from with carbon monoxide concentration 2100 to 2200' F. sharpl:; reduced the (Figure 5). For these conditions the rate calcium sulfide produced. of decomposition was negligible with less Gcr Composll on 2 3% c 0,. 5 e/. s 0, Although the physicJ size and shape of than 1.8YG carbon monoxide. With 16 the final solids were about the same as for more than 5% carbon monoxide the Timperoluro 2140- 22ZO'F the untreated gypsirm, the surface of the data points probably fell below the treated particles was different in appear- plotted straight linv because some cal- ance. Gypsum has a relatively smooth cium sulfide was produced. surface, while sdids recovered from a iun When sulfur dioxide was left out of the made at 2200' F. had a porous surface gas fed to the reactor, the desulfurization E 8 and solids from a run made at 2300' F. rate was again a linear function of carbon c had a glassy surface. The latter must monoxide concentration, but the rate of have reached a state of incipient fusion, increase in the rate of desulfurization was which may explain the smaller de- smaller and the point of origin was sulfurization rate at this temperature. different. Thus decomposition took The glazed surface might offer increased place at an appreciable rate eken with no resistance to the flow of the gaseous carbon monoxide. With 47~carbon reactants to the particle interior. On monoxide the rate was about the same as the other hand, the change in rate the maximum rate observed for the condi-

night have been due to the change in tions of Figure 5. '01234567 structure which reportedly occurs in this When 470 or more carbon monoxide CO in feed gas, */* temperature range (3). was present, the desulfurization rate was In correlating the effect of temperature, not affected much by sulfur dioxide in Figure 5. The reaction was first-order the gypsum bed outlet tempercture was concentrations up to 77,. For smaller with respect to carbon monoxide con- used because most of the reactors were concentrations of carbon monoxide, sul- centration equipped with only a single thermo- fur dioxide reduced the initial desulfuriza- couple. In obtaining the data plotted tion rate. Its effect on the maximum in Figure 1 a re;:ctor equipped to measure rate was not fully determined. than the carbon monoxide concentration both bed inlet and outlet temperatures The desulfurization rate was influenced to prevent production of calcium sulfide. was used. This provides some measure only to a small degree by carbon dioxide When sulfur dioxide was excluded from of the error incurrrd by not using a mean in concentrations up to 3070. the gas fed, a much lower concentration bed temperature. The per cent calcium sulfide in the of carbon dioxide was effective in sup- Effect of Gas Composition. The residual solids increased with carbon pressing calcium sulfide. effect of gas composition was explored monoxide concentration but decreased In correlating the results, the composi- a1 a temperature he1 of 2200' F. with carbon dioxide concentration tion of the gas fed to the reactor was used, When a mixture of carbon monoxide, (Figure 6). In the presence of 5% sul- as it approximated the composition carbon dioxide, sulfur dioxide, and fur dioxide the concentration of carbon in the reaction zone and the latter could nitrogen was fed, the initial and maxi- dioxide must be five to six times greater not be accurately determined.

VOL. 52, NO. 3 c MARCH 1960 217 4 Gas Composition I

~ I 3°~oco,20%co,,5~/~so, ,

I Tern p er oture

-a2190 to 222OOF: ~

- ---

I I 0.I 0.2 0.3 Mass velocity, Ib./sec.xftz AFigure 7. The decrease in the initial desulfurization rate when the mass velocity was increased was unexpected

4 Figure 6. Production of calcium sulfide can be minimized in feed gas, o/o CO, by using small concentrations of carbon monoxide and large Effect of Mass Velocity. Most of concentrations of carbon dioxide the runs were carried out using a gas mass velocity of 0.2 Ib.)sec. X sq. ft. temperature, gas composition, mass When a four-component gas mixture velocity, or particle size, it is apparent was fed, the initial and maximum desul- that more than one mechanism is rate- i furization rates varied with mass velocity controlling during the desulfurization of (Figure 7). The increase in the maxi- a single batch of gypsum. mum rate with mass velocity was prob- ably due to an increase in the average literature Cited carbon monoxide concentration within (1) Am. SOC.Testing Materials, "ASTM the reaction zone rather than to an in- Standards Including Tentatives," Part 3, pp. 265-77, 1952. crease in the rate of mass transfer. (2) Diehl, H., Goetz, C. A., Hach, C. C., The decomposition rate increased to a J. Am. Works Assoc. 42, 40-8 greater extent for the same increase in (1950). mass velocity when the feed gas con- (3) Gruver, R. M., J. Am. Ceram. SOG.34, C 353-7 (1951). c tained 2% carbon monoxide and 98% (4) Hull, W. Q., Schon, F., Zirngibl, H., L nitrogen. For this case the increase in a Gas Composition IND.ENG. CHEM. 49, 1204-14 (1957). 'c rate seemed due to an increase in the 'c 3%C0,2O%C0*,5%SOp - $ '1 Temperature mass transfer rate rather than to an in- Reductive decomposition of gyp- a 2160 to2220'F crease in the average carbon monoxide sum using carbon monoxide in low concentration. concentration is both thermodynami- Effect of Particle Size. The effect of cally and kinetically feasible. The particle size on the desulfurization rate reaction appears to be first-order was determined by using three size with respect to carbon monoxide 1 I fractions of gypsum: -3l/2+4, -7+8, 005 010 015 020 025 concentration, and the optimum tem- Particle size, in. and -12f14-mesh. The initial and perature is 2200" to 2250" F. maximum rates are plotted against the Desulfurization rate is a function Figure 8. The maximum desulfuriza- average screen opening in Figure 8. of temperature, gas composition, tion rate can be increased by reducing Because the maximum rate increased with gas mass velocity, and particle size. the particle size decreasing particle size, and mass trans- Side reactions yielding calcium sul- port in the gas phase did not seem to be fide can be limited by employing a RECEIVEDfor review July 13, 1959 ACCEPTEDDecember 7, 1959 limiting, the rate of internal diffusion high temperature, a low concentra- might have been the rate-controlling tion of carbon monoxide, and a con- Division of Industrial and Engineering Chemistry, 136th Meeting, ACS, Atlantic mechanism during the maximum de- centration of carbon dioxide five to sulfurization rate period for the condi- City, N. J., September 1959. Investiga- six times greater than the carbon tion carried out at the Iowa Engineering tions common to both Figures 7 and 8. monoxide concentration in the re- Experiment Station. Financial assistance Because the initial and maximum action zone. received from E. I. du Pont de Nemours rates were not affected in the same way by & Co., Inc., and the Ethyl Corp.

2 1 8 INDUSTRIAL AND ENGINEERING CHEMISTRY