(20) Lane, D. A., Katz, M., “The Photomodification of Benzo(a)pyrene, Benzo(b)fluoranthene and Benzo(k)fluoranthene Under Simulated Atmospheric Conditions”, in Vol9, “Advances in Environmental Science and Technology, Fate of Pollutants in the Air and Water Environments”, I. H. Suffet, Ed., pp 137-54, WileyInterscience, New York, N.Y., 1977. (21) Salamone, M. F., Heddle, J. A., Katz, M., “Mutagenesis by Polynuclear Aromatic Hydrocarbons”, unpublished data, Centre for Research on Environmental Quality, York University, Toronto, Canada, 1977.
(22) Dong, M., Locke, D. C., Ferrand, E., Anal. Chem., 48, 368 (1976). (23) Hoffman, D., Wynder, E. L., in “Air Pollution”, 2nd ed., A. C. Stern, Ed., Vol 11, p 187, Academic Press, New York, N.Y., 1968. (24) Cautreels, W., Van Cauwenberghe, K., Atmos. Enuiron., 10,447 (1976). Receiued for review September 7,1977. Accepted January 24,1978. Work partially supported by a grant from the Ontario Ministry of the Environment.
Fluidized-Bed Combustion of Coal with Lime Additives: Catalytic Sulfation of Lime with Iron Compounds and Coal Ash Ralph T. Yang”, Ming-Shing Shen, and Meyer Steinberg Department of Energy and Environment, Brookhaven National Laboratory, Upton, N.Y. 11973
Iron oxide (Fez031 is catalytic to the sorption of SO2 by CaO in a chemical environment similar to that in fluidized-bed combustion of coal. Four percent by weight of Fez03 physically mixed with CaO approximately doubles the sorption (sulfation) rate. The kinetic mechanisms of the catalytic reaction are discussed based on the view that SO3 formation is the rate-limiting step. Coal ash also catalyzes, the sulfation reaction, and Fe203 appears to be the active constituent in such catalytic effects. Other iron-containing compounds existent in coal and fluidized-bed combustion systems are oxidized rapidly in the combustion gases to form Fez03 and in turn catalyze the sulfation reaction. H
In recent years, fluidized-bed combustion has been recognized as a promising and versatile technology for clean combustion of coal (1-3). One of the crucial factors in effecting this technology is a relatively high reactivity of the lime additive toward SO2 to form calcium sulfate in the environment of the combustion gases. A significant amount of work has been done in the past to determine the effects of the reactivity of the variables under combustion conditions, such as temperature and concentration of SO2 ( 4 , 5 ) ,total pressure and rapidly changing atmospheres (6, 7), water vapor and oxygen partial pressures (8,9),and the generic factor of the limestone (10).
Little attention has been paid to the effect, if any, of the solids in the fluidized bed on the reactivity. An exception is the effect of the added NaC1, which has been known as a catalyst (11-14). Coal ash generated in the combustion process is mostly of fine sizes and is entrained in the combustion gas. However, significant amounts of coal ash do remain in the fluidized bed a t any instant. For example, in the pilot-scale fluidized-bed combustor a t Argonne National Laboratory, about 5 wt % of the bed material was found to be coal ash (13). Moreover, the lime particles remaining in the bed contain significant amounts of Fez03 in the sulfated layer. For example, the work at Argonne showed that, based on electron microprobe analysis, the spent (recycled) lime sorbent contained up to 8%Fe in the sulfated layer (14). Consequently, the spent lime appeared reddish and brownish in color. Coal ash, which contained about 20% FezO3, was obviously the source of the iron contaminant. Being aware that Fez03 is a known high-temperature catalyst for the oxidation of SO2 to SO3 ( 1 5 ) ,we were prompted to examine the effects of Fe20B and the coal ash on the reactivity of lime with the combustion 0013-936X/78/0912-0915$01.00/0
gases. It was also hoped that the results would shed light on the mechanism of the sulfation reaction of CaO as well as the sulfation kinetics in fluidized-bed combustion.
Experimental A Du Pont thermoanalyzer Model 951 was used for the rate measurements. A small quartz boat with an area of about 0.6 cm2 was used as the sample holder. The thermoanalyzer contained a time-derivative computer which took the derivative of the weight change with respect to time. The derivatives were used to calculate the rates of sulfation. A quartz tube packed with alumina chips and housed in a tubular furnace served as the preheater for the reactant gases. The steam in the reactant gas was generated by bubbling the inert carrier gas (N2) through a water bath prior to entering the preheater. The bubbler was jacketed, and water was circulated in the jacket from a constant temperature water bath. In a typical experiment, about 20 mg of calcium oxide sample was spread into a thin layer on the holder as the solid reactant. The preheated gas mixture flowed over the sample surface at a velocity of about 10 cm/s. This velocity was predetermined to be high enough to maximize the gas film diffusion rate. The initial rate, Le., the rate in the initial 5 min, was employed to calculate all the kinetic parameters, because it was close to the intrinsic chemical rate. The calculation procedures were straightforward and were outlined previously (8).
All the solid samples were of reagent grades. The gases were supplied by Matheson Co. as custom-made premixed SO2 in N2 a t various specified concentrations.
Results and Discussion The rates of sulfation of the reagent-grade CaO were compared with those mixed with 4 wt % of Fe203. As shown in Figure 1,4%of Fez03 almost doubled the rate. For a preliminary understanding of such catalytic effects, we compared the kinetic parameters of the reactions involved based on our experimental results as follows. For the sulfation of CaO, regardless of the true mechanisms of the reaction, the valence of sulfur is changed from +4 to +6, or SO2 is oxidized to SO3. The following reactions can be written to represent the process: CaO + SO^
+ ~ 2 0 2 A CaO - SO^ -+-CaSO4 rb
(1)
For the sulfation of CaO being catalyzed by Fe2O3, which is physically separated from the CaO, the following reactions may describe the sequences:
@ 1978 American Chemical Society
Volume 12, Number 8, August 1978
915
CaO + ~ e 2 0 3+
SO^ + ~ 2 0 2 Cas04
-
CaO + ~ e z 0 3 SOB
+ ~ez03
I
(2)
&
rz(=rc) = kzP26;e-w
-
rc = k , P & e - p
Environmental Science & Technology
I
I
I
I
I
I
I
4 % Fe203
30t
5 % STEAM 4 % v205
I O 20 30 40 50 60 7 0 8 0 90 100 110 T I M E , rnin
120
Figure 1. Sulfation rate of powdered reagent CaO (Maliinckrodt)with Fez03 (Baker,Tyler 200/325),V205(Fisher,200/325),steam and FepO3 plus steam at 850 O C (0.25% SO2 and 5% O2 in N2)
(D
Slope n = 0.59
(5)
Here, the simplifications were made based on our experimental conditions; we assumed that Po2was constant and that Pso3was small and also constant. While it is difficult, if not meaningless, to compare directly r1 and rc on a per mass of solid basis from Boreskov’s and our data, the fact that Fez03 does catalyze the sulfation of CaO indicates that rc is greater than r,. The differences in the kinetic parameters in Equations 3 and 4 are too small to draw inferences on the mechanisms of the catalysis by FezO3. However, it appears that rc is not much greater than r l ; hence, both Reactions 1and 2 are operative in the reaction system that contains FezO3. From the results on the effects of the size (Figure 6) and of the amount (Figure 7 ) of FezO3, the overall rate increases only slightly with the increasing total surface area of the FezO3: not at all proportional. These results agree well with the foregoing discussion. Note that CaO indeed is also a good high-temperature “catalyst” for the oxidation of SO2 to SO3as is Fez03. It may be predicted that certain salts involving CaO, and which do not bond SO3 in this high-temperature range, are good high-temperature catalysts for the oxidation of SOz. Figure 1 also shows results on the sulfation reactions with 5% water vapor in the gas phase. The catalysis of sulfation of CaO with water vapor has been discussed previously (8, 9). The intriguing result here is that while both water vapor and Fen03 catalyze the sulfation, the synergistic effect is lower than that with Fez03 alone. However, this fact can also be explained in line with the foregoing discussion. Water vapor combines very rapidly with SO3 to form sulfuric acid. Thus, water vapor here simply acts as a scavenger and reduces the catalytic effect by FezO3. Of course, sulfuric acid also reacts with CaO, but perhaps at a lower rate than SO3. A note should be made here on the rate dependence on the partial pressure of SOZ. Sulfation of lime was thought to be a first-order reaction with respect to SO2 by various workers, e.g., Borgwardt ( 4 ) . However, their experiments were performed in the lower range of the partial pressure of SO2. For example, Borgwmdt’s data were for SO2 concentrations below 6.4 x 10-sgmol/cm3 (total pressure = 1atm). In Figures 2 and 5, a straight line does not seem to fit the data at higher partial pressures; the data tend to follow a quadratic form. A complicated rate expression similar to that of Boreskov and Sokolova (16) is likely to exist for the sulfation reaction of 916
I
(4)
The work of Boreskov and Sokolova (16) and of Kawaguchi (17) showed that for the SO2 l/Z02 SO3 reaction with Fez03 as the catalyst, the rate can be expressed, after simplifications, as the following:
+
I
40
(3)
The slopes were outlined by using the least-squares linear regression method. Similarly, from Figures 4 and 5, the overall rate of the catalyzed sulfation, with 4%Fez03 in CaO, can be expressed by:
I
4 % Fe203
Reactions 1 and 2 are both rate-limited by the oxidation step, or limited by ra and rc, respectively. This is so because it is believed that the reaction between CaO and SO3 is very fast. This reaction is also being studied in our laboratory, and the preliminary results show that it is indeed so (19). From Figures 2 and 3, the rate of Reaction 1,r l , can be expressed by:
rl(=r,) = k l P @ i e - w
I
i
a /
c
.-
t/l l
,
