Ind. Eng. Chem. Process Des. Dev. 1983, 22, 323-329
323
Absorption of Nitrogen Monoxide into Aqueous KMnO,/NaOH and Na,SO,/FeSO, Solutions Shlgeo Uchlda,' Tsuguo Kobayashl,' and Shlzuo Kageyama' Department of Chemical Engineerlng, Shizuoka University, 3-5-1 Johoku, Hamamatsu 432, Japan
An experimental study on the absorption of low concentration nitrogen monoxide into aqueous KMnO,/NaOH and Na,SO,/FeSO, solutions has been conducted with two types of absorbers, a stirred tank and a bubble chamber. The absorption rates under various experimental conditions were measured. In the case of NO absorption into KMnO,/NaOH solutions, the rate is proportional to both the interfacial concentration of NO and either that of a complex formed in the solution or that of KMnO, depending on the pH values of the solutions. In the case of Na,SO,/FeSO, solutions, the rate shows the maximum at a certain concentration of Na2S03. The reaction mechanism involved seems not to be simple since the reaction orders with respect to Na,SO, and FeSO, are not integers. The experimental data have been analyzed by the (m ,n)th-order reaction model and equations for NO absorption rates have been presented.
Introduction Nitrogen oxides (NO,) comprise one of the main air pollutants in flue gases as well as sulfur oxides (SO,) and are considered to be a material that causes photochemical smog. Therefore, to reduce the amount of nitrogen oxides emitted from various sources such as chemical plants, power stations, and automobiles is a very urgent social problem to be solved in these years. Confining the problem to the removal of nitrogen oxides in flue gases from stationary sources, the scrubbing processes so far developed are classified into two groups, dry processes and wet processes. As for the dry processes, the nitrogen oxides are decomposed or reduced by contacting with various reducing agents, while in most wet processes nitrogen monoxide (which is the main component in the flue gases) is oxidized to more reactive nitrogen dioxide in the liquid phase. Although many processes have been proposed to remove the nitrogen oxides in flue gases, only a few processes are in practical operation. Since the wet processes to remove nitrogen oxides have several advantages over dry processes (e.g., ease of combination with already present wet desulfurization processes with some auxiliary attachments and the ability to operate at the normal temperature), considerable effort has been devoted to the development of more efficient wet processes. It is, however, very difficult to develop such processes due to the low reactivity of nitrogen monoxide. At present, two methods are considered to remove nitrogen monoxide by wet processes. One is the oxidation of NO to NOz by strong oxidizers such as ozone in the gas phase prior to the absorption into the liquid phase. Another is the oxidation of NO in the liquid phase as soon as it is absorbed into the solution containing strong oxidizers. Aqueous KMn04/NaOH mixed solution and aqueous Na2S03solution catalyzed with FeS04 are known to be very effective NO oxidizing solutions. The absorption mechanisms of NO into these solutions have not been cleared yet. In this study, the absorption of NO into the aqueous KMn04/NaOH solution in a stirred tank and aqueous Na2S03/FeS04solution in a bubble chamber has been performed. The obtained data have been analyzed and Ibiden Co. Ltd., 2-1 Kanda-machi, Ogaki-shi, Gifu 503. Department of Industrial Chemistry, Junior College of Technology, Shizuoka University. 0196-43051831 1 122-0323$0 1.50J0
the absorption mechanisms have been elucidated. Previous Studies In the absorption of NO into a solution, it is important what kind of reactant is used because of the low reactivity of NO. There are various types of solvents being promised to react with NO in a certain degree and some studies have been reported. The review here is limited to the studies on (i) oxidation of NO in the liquid phase, and (ii) complex reaction of NO with the reactant in the solution. NO Absorption into Solution Accompanied by Oxidation Reaction. Teramoto et al. (1976a) and Sada et al. (1977) performed experiments of NO absorption into KMn04/NaOH solutions and obtained the results that the absorption reaction is first order with respect to NO concentration in the gas phase and KMn04 concentration in the liquid phase, that is, (1,l)st-order reaction. They also found that the reaction rate constant is a function of temperature and NaOH concentration in the solution. Teramoto et al. (1976b) and Sada et al. (1978a) performed the experiments for NO absorption into NaC102/NaOH solution in stirred tank absorbers. Teramoto et al. concluded that the effects of NO concentration in the gas and NaClO, concentration in the solution on the absorption rate were complex and found no simple expressions for them. On the other hand, Sada et al. obtained the result that the rate is second order with respect to NO concentration and first order with respect to NaC102 concentration if ita concentration is higher than 0.8 X mol/cm3. NO Absorption with Complex Reaction in Liquid Phase. Ganz and Mamon (1953) reported that the solutions of FeS04and FeC1, absorb NO. From an experiment performed under the high NO concentration, they showed that the rate of NO absorption into the FeS04solution was lowered by the addition of some salts and organic materials. Kustin et al. (1966) reported that the NO absorption into FeS04 solution involves an equilibrium reaction, NO + FeS04 = Fe(NO)S04,and its forward reaction is fist order with respect to both NO and FeS04; the reverse reaction is first order with respect to Fe(NO)SO,. Hikita et al. (1977) and Sada et al. (1978b) also performed experiments for the same system using different types of absorbers to investigate the reaction mechanism. Hikita et al. (1978) performed a series of experiments with a stirred tank absorber for the NO absorption into 0 1983 American Chemical Society
324
Ind. Eng. Chem. Process Des. Dev., Vol. 22, No. 2, 1983
Table 1. Experimental Conditions for Absorption of CO,, SO,, and NO into Aqueous Solutions in a Stirred Tank Absorber CO, partial pressure, P A G , kPa SO, concn, ppm NO concn. w m NaOH con&: mol/cm3 X lo3 KMnO, concn, mol/cm3 x lo3 temperature, T,K stirrer speed, min- '
5.07 1850,4990 399. 900.1790 0.1,'0.5,1.0,2.0 0.04,1.00, 2.00,3.00 288,303, 318 91,150,241,391
Table 11. Experimental Conditions for NO Absorption and Measurement of Gas Holdup and Bubble Size at 291.303. and 318 Kin Bubble Chamber
NO absorption @Thermometer @Heater @Gas outlet @Gas stirrer @Baffle @Liquid stirrer
@ Thermister @Gas inlet @)Flange @Gas-Liquid interface @Sampling tap
FeSO, concn X mol/cm3
Figure 1. Details of stirred tank absorber.
NazSOs/Fe"'.EDTA-Na solutions and NazS03/Fe"'. EDTA-Na/Fez(S04)3 solutions and obtained the reaction rate constants. The rate constants are, according to their results, independent of NaZSO3concentration for both solutions, and dependent on Fem.EDTA-Na concentration, ionic strengths of ions present, and the temperature for the first solution. In addition to the above factors, the Fez(S04)3concentration also affects the rate of NO absorption. Koizumi et al. (1974) performed an experiment of NO absorption into NazS03/NaHS03solutions containing various kinds of metal salts and found that the solutions of organic and inorganic ferrous salts gave high NO absorption rate.
Experimental Apparatus In this study, two types of absorbers were used. A stirred tank absorber was used for the absorption of NO into KMn04/NaOH solutions and a bubble chamber was used for the absorption of NO into Na2S03/FeS04solutions. Absorption of NO into KMn04/NaOH Solutions. The details of the stirred tank absorber is shown in Figure 1. The main part of the tank consisted of a Pyrex glass bottle of 133 mm diameter and 220 mm height, equipped inside with 8 baffles of 10 mm width (5). Two 8-blade impellers of 70 mm diameter, (4) and (6), were used for mixing the gas and the liquid at the same speed. The gas to be absorbed was supplied from cylinders of COz or Nz-balanced NO or SO2 and was diluted with nitrogen from the other in a gas mixer to a required concentration. It was fed to the stirred tank absorber (which contained a 1500 cm3 KMn04/NaOH solution) through an inlet pipe (8) and passed through the tank after contacting with the solution for a certain time at the plane interface between the gas and the liquid (10). The contacted gas was led to a Toshiba-Beckmann NO, analyzer to measure NO concentration. Teflon pipes were used everywhere to avoid the adsorption or desorption of the diluted gases of NO and SOz. The liquid temperature was controlled by a thermistor (7) and a heater (2). The whole system was placed in the air bath to maintain the temperatures of the gas and the liquid within f0.5 "C. Absorption of NO into NazS03/FeS04Solutions. In this experiment, a bubble chamber consisting of a rectangular box of 410 mm height, 140 mm length, and 30 mm
NO concn X lo3, PPm Na,SO, concn X lo3, mol/cm3
gas holdup and bubble size
lo3,
NaCl concn X lo3, mol/cm3 NaOH concn X lo3, mol/cm' Na,SO, concn X lo3, mol/cm3 H,SO, concn X l o 3 , mol/cm3
399,900,1790 0.010,0.025, 0.040,0.050, 0.075,0.100, 0.25,0.50, 0.75,1.00, 1.50 0.001,0.005, 0.010,0.020, 0.050,0.10 0,0.10,0.50, 1.0,1.5, 2.0 0,0.10,0.50, 1.0,1.5,2.0 0,0.10,0.50, 1.0,1.5 0,0.04
width was used. It was made of acrylic plates so that the flow conditions inside the chamber could be observed and photographs of bubbles taken to determine the bubble size. Bubbles were formed from 25 stainless steel needles with holes of 0.49 mm diameter and 50 mm length which were placed in a double line at the bottom of the chamber. The gas flow rate and the amount of the solution were the same as those for the stirred tank absorber. The gas holdup was determined by measuring the static pressure of the solution at the bottom. The bubble sizes were determined by taking pictures with strobe flash. However, since the solution of NazS03and FeS04used for NO absorption was opaque, which made it impossible to take pictures of the bubbles, some other electrolyte solutions were used to estimate the bubble size for Na2S03/ FeS04 solutions. Determination of Concentrations of COz,SO2, and NO in Liquid and Gas Phases. In the case of the absorption of carbon dioxide and sulfur dioxide, samples were taken from the tank by using a 5-mL hole pipet and analyzed as follows. For the concentration of carbon dioxide in water, an excess amount of 20 mmol/L Ba(OH)zsolution was added to the sample and back-titrated with 20 mmol/L HC1 solution. The concentration of COz in NaOH solution was determined by the Warder's method (Iwasaki, 1953); that for sulfur dioxide in NaOH solution was determined by the iodometry method (Iwasaki, 1953). To determine the concentration of NO in the gas phase, a Toshiba-Beckmann NO-NO, Analyzer (Model-1951A) was used. All experimental conditions are shown in Tables I and 11. Theoretical Section Interfacial Area between Gas and Liquid. The gas-liquid interfacial area in a stirred tank absorber, uT, increases with the stirrer speed. In this study, the area was determined by using the gas absorption accompanied by a fast pseudo-fiist-order reaction (Yoshida and Miura, 1963; Kataoka and Miyauchi, 1966).
Ind. Eng. Chem. Process Des. Dev., Vol. 22, No. 2, 1983 325
In the case of the bubble chamber, it is necessary to know the gas holdup, t, and the volume-surface mean diameter, d, to determine the gas-liquid interfacial area. The gas holdup is obtained by measuring the depths of the static and the bubbling liquid in the chamber as follows e = h/z (1)
If a bubble is assumed to be a spheroid with the major diameter, dl, and the minor diameter, d2, and its volume is considered to be equal to the volume of a sphere with diameter, do, the volume-surface mean diameter, d, is given by d, = do/f(E)
(2)
where E = d l / d 2and
f(E) = 2 ~ 1 /{3E +
In [E
+ (E2- 1)1/2]
(E2- l)'/'
The interfacial area of bubbles, aB,is then calculated by (3) This is statistically shown to be representative if calculated by using values for 50 bubbles (Kalra, 1969). Gas Film Mass Transfer Coefficient and Solubility of NO in Mixed Salt Solutions. The gas film mass transfer coefficient for NO gas, kca0, in the stirred tank absorber is estimated from that for SO2gas obtained by performing an experiment of lean SO2 absorption into NaOH solution as follows (4)
The diffusivities of SO2 and NO in the gas phase are estimated by the Chapman-Enskog equation (Bird et al., 1960). Those for different temperatures and compositions are obtained by the relation, D@/T = constant (Nijsing et al., 1959; Hitchcock and McIlhenny, 1935). In the case of the bubble chamber, the gas-phase resistance is generally considered to be negligible (Yoshida and Akita, 1965). The solubilities of NO in KMn04/NaOH and N G 0 3 / F e S 0 4solutions are calculated by the Danckwerts method (Danckwerts, 1970) using data found in the literature (Armor, 1974; Seidell and Lineke, 1965; Teramoto et al., 1976a; Onda et al., 1970). Rate of Gas Absorption with (mp )th-OrderReaction. When a gas A is absorbed into a liquid containing a reactant B and reacts as follows A + uB -,product (5) with the reaction rate given by
the rate of the absorption of gas A is, according to Hikita and Asai (1963), given by
Reaction Mechanism of NO in Liquid Phase. According to Obuchi et al. (19741, reaction schemes of NO in KMn04/NaOH solutions are classified as follows depending on the conditions.
When NO is absorbed into strong alkali solutions (e.g., containing such as KOH), the following reaction is considered to occur NO
+ KMn04 + 2KOH -,KN02 + K2Mn04+ H 2 0 (a)
The absorbed NO is oxidized to NO2-, while KMnO, still remains in the form of Mn0:-. When NO is absorbed into neutral solution containing KMn04 as an oxidizer, the possible reaction in the liquid phase is NO
+ KMn0, -,KNOB+ Mn02
(b)
In this case and in the solution of low pH, NO is oxidized into NO3-, and KMn0, is reduced to Mn02, which is a fragment-like product floating on the surface of the solution and covers the effective surface area to reduce the absorption rate. It is, therefore, not practical to operate the absorption process under the condition of such a low pH value. In the present study, the absorption of NO into KMnO, solutions strongly alkalized with NaOH is performed and the absorption mechanism presented by reaction a is analyzed. On the other hand, the reactions of NO in Na2S03solution have long been known (Gmelin Handbuch, 1936)
+ Na2S03= Na2N2S05 Na2N2S06= N 2 0 + Na2S04
2N0
(4 (4
Reaction c is slow in the solution with a pH value of about 8. Kustin et al. (1966) reported the following reaction for NO in FeS04 solution FeS04 + NO
Fe(NO)SO,
-,
(e)
When NO is absorbed into an Na2S03/FeS04solution, the reactions are reported to be (Saito, 1976) FeS0, + NO -,Fe(NO)SO, (f) Fe(NO)S04 + 2Na2S03+ 2H20 -, Fe(OH), + Na2S04+ NH(S03Na)2 (g) Reaction g is slow while reaction f is in equilibrium. The reaction mechanism of NO in Na2S03/FeS04seems complex and is affected by the pH of the solution. Since the experiments of the absorption of NO into Na2S03 solution and into FeS0, solution performed prior to the absorption of NO into Na2S03/FeS04solutions showed that the absorption rate is considerably low, the main experiment is planned to be performed under the conditions where reactions f and g occur.
Experimental Results and Discussion NO Absorption into KMnO,/NaOH Solution in Stirred Tank Absorber. (1) Characteristics of Stirred Tank Absorber. Prior to the main experiment, preliminary experiments were performed to obtain the fundamental data on the tank characteristics such as the interfacial area and the gas film mass transfer coefficient. All data are summarized in Table 111. (2) Effect of Interfacial Concentration of NO. The rate of NO absorption was obtained from the difference of NO concentrations in the feed gas and in the effluent gas. The interfacial concentration was estimated with the known gas film mass transfer coefficient. The rate of NO absorption, RA, vs. the interfacial concentration, CAI,is shown in Figure 2. I t is seen that RAis proportional to CAi. That is, the power of CAi in eq 7, m,is equal to unity. Therefore, the reaction involved is first order with respect
NOconc .1790 ppm KMn~~onc..~2xlO~~mol/c~
031
008 NUXI
I
I 10
03 conc
XI@
35
,mo~cm’
Figure 4. Effect of NaOH concentration on NO absorption rate. No conc = 1790 ppm KMn04 conc = O Z X ~ O ‘mollcm3 ~ Stirrer speed = 391 rpm
E
0
04
10
04 I
I
2
1 31
32
33
3L
1 I T x103 , K-’
Figure 5. Effect of temperature on reaction rate constant.
NaOH conc = l ~ l O ‘ ~ m o l / c m ~ Stirrer speeddgl rpm n/ Temperature =303 K
,-
’I 1 002
Key,NO conc ppm
,
:::
1790
005 01 02 KMn04 conc x103, mol/cm3
for the reaction rate constant at 1X mol/cm3 of NaOH concentration as follows k2 = 8.51 X 1014 exp(-708/RT) (9)
I
(6) Reaction Mechanism. In general, the reaction rate constant is independent of the concentration of reactants. The reaction rate in this case varies, however, with NaOH concentration. As mentioned previously, Obuchi et al. (1974) reported that the reaction of NO in the liquid phase is either reaction a or b. When NO, KMn04, and NaOH are represented by A, K, and N, respectively, the following reaction scheme is considered
1
kn
K + N Z F
05
Figure 3. Effect of KMn04concentration on NO absorption rate.
A
+F
to the concentration of NO. (3) Effect of Concentration of KMnO,. Changing the concentration of KMn04, the rate of NO absorption, RA, was observed. The results are shown in Figure 3. The slopes of the resulting lines are 1/2 and from eq 7, the reaction is also frst order with respect to the concentration of KMn04. Equation 7 is, therefore, reduced to
A
+K
RA = (k2CKDA)”2CAi
-
k,l
(8)
(4) Effect of Concentration of NaOH. A plot of [RA/(DA)1/2CAi]vs. NaOH concentration a t constant KMn04 concentration is shown in Figure 4. The effect of the NaOH concentration on the rate of NO absorption is not observed when the NaOH concentration is lower than 1.0 mol/L. When the concentration becomes higher, the effect is significant. The discussion will be given later. (5) Temperature Effect. The rates of NO absorption were measured a t three different temperatures, and the Arrhenius plot is shown in Figure 5. The activation energy is given as 0.708 kJ/mol, and the experimental correlation
k,‘
k?
(h)
product
(i)
product
ci)
If reaction h is assumed to be an equilibrium reaction, the equilibrium constant, K1, is given by (10)
Therefore, the absorption rate of A is represented by
When k,’KICN > k 7 , that is, the concentration of NaOH is higher than 1 X mol/cm3, the second term in the right-hand side of eq 11 can be neglected and the rate equation becomes
In this case, KMn04 and NaOH probably form a complex and the reaction between the complex and NO becomes
Ind. Eng. Chem. Process Des. Dev., Vol. 22, No. 2, 1983 327 NO conc. = 1790 ppm FeSQ conc.1 0.05~ mollcm3 Temperature i303 K
0 0.01
, S 0.05 0.1 0.5 1 N q Q conc.x103,mol/cm3
h 2
Figure 8. Effect of Na2S03concentration on NO absorption rate. 10 50 R~~~~ x io9 , mol/cm25-l
05
100
Figure 6. Comparison of calculated absorption rate, R A , & with ~ experimental value, RkeX,,d.
1000
I
500 -
FeSQconc =005x105mollcm3 Temperature. 303 K
NO conc.=1790
$ ‘
’ /
a’
20
2
-50.0
wm
FeQ conc = 0.05x10~3molicm’ Key , Temperature, K 0 318
t
0.1
uey, 0 0
0.5
N ~ S O ,conc XIO’ moIlcm3
004 05
1
i 3
c,, x109, mol/cm3 Figure 7. Effect of interfacial concentration of NO on NO absorption rate.
a (1,l)st-order reaction. The absorption rate is then given by RA = [ ( ~ ~ ’ K ~ ) C K C N D A ] ~ / ~ C A(13) ~ where k3 = k kl’KICN, that is, the concentration of NaOH is lower mol/cm3, eq 11 is modified to than 1 X
In this case, the reaction between NO and KMn04 is a (1,l)st-order reaction. The absorption rate is given by (15) RA = (k,”DACK)’WAj where k? = 4.4 X lo9 cm3/(mol s) at 303 K. The comparison of the calculated values with the experimental data is given in Figure 6. NO Absorption into Na2S03/FeS04Solution in Bubble Chamber. (1) Characteristics of Bubble Chamber. Since the interfacial area of the bubble column has been reported to vary with the types of electrolytes, their concentrations, the surface tension of the solution, etc. (Yoshida and Akita, 1965; Stefan and Whitemore, 1971), the gas holdup and the bubble diameter were measured by preliminary experiments using various electrolytes solutions such as those shown in Table 11. The gas holdup, e, and the gas-liquid interfacial area, uB,for the bubble chamber used in this study at 50 cm3/s of the gas flow rate are 6.02 X and 1.23 cm2/cm3, respectively.
0 01
005
01
6 ~ concxld, 4 mol/cm3 Figure 9. Effect of Na2S03concentration on NO absorption rate.
(2) Effect of Interfacial Concentration of NO. As seen in Figure 7, the rate of NO absorption is proportional to the interfacial concentration of NO, C,, and the slopes of the resulting lines are unity. Therefore, the reaction involved is first order with respect to the NO concentration. (3) Effect of Na2S03Concentration. With the concentration of FeS04 held constant, the rate of NO absorption was measured for various Na2S03concentrations. As shown in Figure 8, the rate increases with Na2S03 concentration and reaches the maximum at a concentration of about 1 X lo-‘ mol/cm3. When the Na2S03concentration increases further, the pH of the solution becomes higher and the product, Fe(OH)? precipitates. Thus,the concentration of Fez+m the solution decreases and so does the rate. Since a solid is contained in the solution when Na8O3concentration is high, the theoretical analysis becomes difficult. The discussion here is thus restricted to the condition that concentration of Na2S03 is lower than 1 X lo-‘ mol/cm3. The rate is proportional to the 0.65 power of Na2S03concentration as shown in Figure 9. (4) Effect of FeS04Concentration. The rate of NO absorption is also proportional to the FeS04 concentration with the power of 0.65 as shown in Figure 10. (5) Reaction Mechanism. The results shown above are explained as follows. When NO and the complex of Na2S03and FeS04 are represented by A and F’, respectively, the reaction scheme is considered as FeS04 + Na2S03 F’
+A
ki
kn
product
F’
(k) (1)
328 Ind. Eng. Chem. Process Des. Dev., Vol. 22, No. 2, 1983 loo0
I
I
I
/
,a/,
50
0,001
0.005 0.01
0.05 01
F e q conc.x103,mol/cm3
Figure 10. Effect of FeSOl concentration on NO absorption rate.
Key.
0.2 0
0.1
Temperature. K 318 303 291 I
If reaction k is an equilibrium reaction, the equilibrium constant K2 is given by
KZ= kE/kr2
(16)
Then, the rate of NO absorption is represented by RA = (~,'"E''~C~''~DA)'/~CA~
(17)
where kiK2 = 1.37 X lp exp(-1920/RT). The comparison of the calculated results with the experimental data is shown in Figure 11.
Conclusions and Recommendation The experiments were performed on the NO absorption into KMn04/NaOH solutions and Na2S03/FeS04solutions using two types of absorbers, a stirred tank absorber and a bubble chamber, and we found the following. (1)The rate of NO absorption into KMn04 solution decreases with time due to the production of Mn02 which covers the gas-liquid interface and prevents the NO transfer. When NaOH is added in the solution, the production of Mn02is prohibited and the absorption rate does not decrease. (2) The mechanism of the absorption of NO is classified into two types depending on the concentration of NaOH. The absorption rates are represented by eq 13 and 15. (3) The absorption rate of NO i n t ~Na$03/FeS04 shows the maximum at a certain concentration of Na2S03. When
ita concentration is too high, Fe(OH), precipitates and the absorption rate decreases. (4) Although the absorption rate of NO into Na2S03/ FeS04 solution is represented by eq 17, the reaction mechanism involved may not be so simple since the powers of the concentrations of Na$03 and FeS0, in the equation are not integers. Acknowledgment The authors thank Messrs. K. Taniguchi and Y. Takiguchi for their contribution to the experimental work. Nomenclature A = gas-liquid interfacial area, cm2 aB = gas-liquid interfacial area in bubble chamber, cm2/cm3 aT = gas-liquid interfacial area in stirred tank absorber, cm2/cm3 CA = concentration of component A in liquid phase, mol/cm3 CAi = concentration of component A at interface, mol/cm3 CB = concentration of component B in liquid phase, mol/cm3 Cm = concentration of component B in bulk liquid, mol/cm3 Cj = concentration of componentj in liquid phase 0' = N for NaOH; j = K for KMn04;j = E for FeSO,; j = M for Na2S03),mol/cm3 DA = diffusivity of component A in liquid phase, cm2/s DG,k= diffusivity of component k in gas phase, cm2/s dl, d2 = diameters of ellipsoid, cm do = equivalent spherical diameter of bubble, cm d,, = Sauter's volume-surface mean bubble diameter, cm E = bubble's eccentricity h = height of bubbling liquid, cm K1,K2 = equilibrium constants for reactions h and k k f l ,krl 7 reaction rate Constants for forward and backward reactions of reaction h, (mol/cm3)-' s-' kE, kr2 = reaction rate constants for forward and backward reactions of reaction k, (mol/cm3)-' s-l kl', k 7 = reaction rate constants for reactions i and j, (mol/cm3)-' s-l k 2 = reaction rate constant for second-order reaction, (mol/cm3)-' s-' k ; = reaction rate constant for reaction 1 (mo1/cm3)-' s-' k G k = gas film mass transfer coefficient for component k , (mol/cm2) s-1 Pa-' km," = reaction rate constant for (m,n)th-order reaction, (mol/~m~)'-~-" s-' R = gas constant, (J/mol) K-' RA = absorption rate of component A, mol/cm2 s-' T = temperature, K t = time, s z = static height of liquid in bubble chamber, cm Greek Letters e = gas holdup = viscosity, (g/cm) s-' Y = stoichiometric factor Registry No. NO, 10102-43-9;KMnO,, 7722-64-7; NaOH, 1310-73-2;Na2S03,7757-83-7;FeSO,, 7720-78-7. Literature Cited Armor, J. N. J . Chem. Eng. Data 1974, 79, 62. Bird. R. 6.; Stewart, W. E.; Llghtfoot. E. N. "Transport Phenomena"; Wiley: New York, I 9 6 0 Chapter 16. Danckwerts, P. V. "Gas-Liquid Reactbns"; McGraw-Hili: New York, 1970; Chapter 1. Ganz, S. N.; Mamon, L. I.J . Appl. Chem. (USSR)1953, 26, 927. "Gmelins Handbuch der Anorganischen Chemie"; Springer-Verlag: Berlin, 1936; Syst.-Nr.YN), 5.730. Hlklta, H.; Ami, S. Kagaku Kogaku 1963, 2 7 , 823. Hlkita, H.; Asal, S.; Ishikawa, H.; Hirano, S. J . Chem. Eng. Jpn. 1977, 70, 120. Hikka, H.; Asal, S.; Ishkawa, H.; Sakamoto, S.;Kkagawa. M. J . Chem. Eng. Jpn. 1978, 1 1 , 360. Hlchcock, L. B.; McIlhenny, J. S. I d . Eng. Chem. 1935, 27, 461. Iwasaki. I. "Bunsekl Kagaku Gaisetsu"; Oakujutsu Tosho Pub.: Tokyo, 1953; Chapter 9. Kaka. H. M. S. Thesis, University of Alberta, Edmonton, Canada, 1969. Kataoka, H.; Mlyauchi, T. Kagaku Kogaku 1966, 30, 409. Koizuml, M.; Tanaka. T.; Ishihara, Y. Dentyoku Chuo Kenkyujo Gljulsu Daiichi Kenkyu Hokoku, No. 7402, No. 74031, 1974.
Ind. Eng, Chem. Process Des. Dev. 1983, 22, 329-334 Kwtln, K.; Taub, 1. A,; Weinstock, E. Inorg. Chem. 1966, 5 , 1079. Nljslng, R. A. T. 0.; Hendrlksz, R. H.; Kramers, H. Chem. €ng. Scl. 1969, 10, 88. Obuchl, A.; Hanel, T.; Okugekl, A.; Okabe, T. Nbpon Kagaku K a M l 1974, 1425. onde,K.; Sada,E.; KobayasM, T.; KRo, S.; Ito, K. J . Chem. €ng. Jpn. 1970, 3, 18. Sada, E.; Kumazawa, H.; Tsubol, N.; Kudo, I.: Kondo, T. Chem. Eng. Scl. 1977, 32,1171. Sad, E.; Kumazawa, H.; Kudo, I.; Kondo, T. Chem. €ng. Scl. 1976a, 33, 315. Sada, E.; Kumazawa, H.; Tsubol, N.; Kudo, I.; Kondo, T. Ind. Eng. Chem. Rocwss Des. Dev. 1978b, 17, 321. Sab, S. Tokkyo Koho (Japan), S.51-83883, 1976.
329
Seklell, A.; Lineke, W. F. "Solubllltles of Inorganic and Metal Organic Compounds", 4th ed.; American Chemical Society: Washington, DC, 1965. Stefan, A. 2.; Whitemore, R. C. Chem. €ng. Scl. 1971, 26, 509. Teramoto, M.; Ikeda, M.; Teranlshl, H. Kagaku Kogaku Ronbunshu 1976a. 2 . 88. Teramoto, M.; Ikeda, M.; Teranlshl, H. Kagaku Kogaku Ronbunshu 1976b, 2. 837. Yoshida, F.; Akka. K. A I C N J . 1965, 1 1 , 9. Yoshida, F.; Miura. H. A I C M J . 1963, 9 . 331.
Received for review August 21, 1981 Accepted September 15, 1982
Increase of the Gas Conversion in a Fluidized Bed by Enlarging the Cross Section of the Upper Zone of the Bed J d Corelia' and Rafael Bilbao Departamento de Qdmhx Thnlca, Facultad de Clenclas, Universkled de Zaragoza, Zaragoza. Spaln
To improve the gas-solid contact and in order to increase the gas conversion in a fluidized bed, a new contactor Is proposed. Thls contactor consists of two dlfferent beds contained, without any separating device, in the same vessel. These two beds are produced by a considerable increase of the cross section of the vessel on the upper zone. The gas conversion at the outlet has been experimentally determined for three different types of reactors: an isothermal flxed bed with gas piston flow, a small cylindrical fluidized bed, and the present fluidized bed with varying geometry. The gas conversions for this new type of contactor are greater than those obtained wlth a cylindrical fluidized bed and somewhat lower than those corresponding to a gas piston flow. Correlations between the efficiency of the contact and the parameters of the system (soiM weight fraction in the various zones and gas velocity at the inlet) are obtained.
Introduction One of the main objections to fluidized beds is their deficient gas-solid contact owing to the bubbling. This deficient contact means that, when they are used as chemical reactors, the gas conversion obtained is smaller than that corresponding to a gas piston flow. Several methods are known for improving the solid-gas contact and for achieving the increase of the gas conversion in a fluidized bed for a given space time and without eliminating any of its advantages. Some of these are aa follows: (a) the installation of devices inside the fluidized bed for breaking the bubbles; these devices can consist of horizontal or vertical internals (screens, tubes, rods, perforated plates) (Jodra et al., 1979a, 1979b; Harrison and Grace, 1971; Furusaki, 1973; Rooney and Harrison, 1976; Fujikawa et al., 1976; Claus et al., 1976; Guigon et al., 1978), or floating bubble breakers (Keillor and Bergougnou 1976); (b) the introduction of a tube with fixed packing in the fluidized bed with which Laguerie et al. (1973) increased the efficiency of butane oxidation in a fluidized bed; (c) the use of gas velocities very close to the minimum fluidizing velocity or, on the other hand, sufficiently high for the existence of parallel entrainment of solids, with which one may enable the gas flow to be close to the piston type, ~ F Iin the fluid catalytic cracking in which the reactor first of all consists of a riser in which the greater part of the gas conversion takes place (Venuto and Habib, 1978). In a previous work, the improvement of solid-gas contact in a fluidized bed was achieved with internals consisting of arrangements of horizontal tubes. Methods for quantitatively correlating the design parameters of the internals 01964305/83/1122-0329$01.50/0
with the effects produced, breakage of the bubbles (Jodra et al., 1979a), or increase of the conversion (Jodra et al., 1979b), have been previously described. To increase the efficiency of solid-gas contact maintaining the advantages of a fluidized bed, a new type of contactor, which may be designated fluidized/ (fixed or fluidized bed), is hereby proposed. It is basically constituted by two beds in the same vessel, without any separating device, produced by a considerable increase of the cross section in its upper zone, Figure 1. According to the fluidodynamic study of this contactor (Corella and Bilbao, 1982),when the system is discontinuous for the solid, the contador can operate: (a) as a fixed bed, when u1 < u, (u, = bed breaking velocity); (b) as a fluidized/ fixed bed with intermediate gas chamber, when u, < u1 < u, (u, = velocity at which discontinuity disappears); (c) as a fluidizedffixed bed without intermediate gas chamber, when u, < u1 < ud,= (umf,== minimum fluidizing velocity in the upper zone); or (d) as a fluidized/(fluidized bed with different bubbling), when u1 > Umf,m*
From the gas pressure drop aspect, the contactor shows its greatest advantages for u1 > u,. But if u1 >> ude the bubbling in the upper zone of the contador is very vigorous and the gas flow in this zone also deviates from the piston type. Therefore, in this work, the experiments with chemical reaction have been carried out at 1'5umf,, > u1
> u,.
The gas conversions obtained here with this type of reactor are compared with those obtained in an isothermal fixed bed with gas piston flow and in a conventional 0 1983 American
Chemical Society
