(4) The oxidation mother liquor is recycled without any special purification, leading to the negligible consumption of the catalyst. Acknowledgment
The authors wish t o thank R . Yokouchi, K. Hagihara, T. Kato, Y. Takehisa, and 31. Sat0 for their advice and encouragement. Also thanks are due h l . Shoji, K. Ozaki, and H. Yoshihara for their cooperation.
Literature Cited
Ardis, A. E., Nasti, F. L., Vaitekunas, A. A. (to Olin Mathieson Chemical Corp.), U.S. Patent 3,036,122 (May 22, 1962). Bryant, H. S., Duval, C. A., ?.lchIakin, L. E., Savoca, J. I., Chem. Eng. Progr., 67 (9), 69 (1971). Saffer, A., Barker, R. S. (to Mid-Century Corp.), U. S. Patent 2,833,816 (May 6, 1958). Thompson, B., Neely, S. D. (to Eastman Kodak Co.), U. S. Patent 3,240,803 (March 15, 1966). Toray Industries, Inc., British Patent 1,043,426 (Sept 21, 1966). RECEIVED for review November 7, 1972 ACCEPTEDJanuary 29, 1973
Promises for Ultrasonic Waves on Activity of Silica Gel and Some Supported Catalysts Ramaswami Ranganathan, lndresh Mathur, Narendra N. Bakhshi," and Joseph F. Mathews Department of Chemistry and Chemical Engineering, University of Saskatchewan, Saskatoon, Sash., 871V O W O , Canada The effect of various types of irradiation, such as @-particles, neutrons, and y-rays on the activity of catalysts (Coekelbergs et al., 1962; Mikovsky and Weisz, 1962; Taylor, 1965; Taylor, 1968; Weisz and Swegler, 1955) is reported. In these studies the catalytic activity increased significantly with the radiation treatment. I n all cases the commercially available catalyst was irradiated. Usually, fairly large dosages of irradiation were required to produce significant changes in the catalytic activity (Llikovsky and Weisz, 1962). Recently, another type of irradiation, ultrasonics, was used to increase the activity of a catalyst (Li et al., 1964; Ranganathan et al., 1971). The ultrasonic treatment (insonation) was carried out during the manufacture of the catalyst as opposed to the irradiation of the commercially available catalyst. The information reported in the literature is as yet too scant to afford a clear picture of the subject. The situation is made complex by the several variables in the case of irradiation with ultrasonics-the frequency, intensity, duration of insonation, and the atmosphere over the catalyst system during irradiation. The literature on insonation of catalysts can be divided in three categories: (I) insonation of the catalyst during the preparation stage; (11) insonation of the already prepared catalyst; and (111) insonation of the reaction mixture (including the catalyst). Two of these systems have been discussed in detail in one of our earlier papers (Ranganathan et al., 1971). A brief introduction is in order. Case I: lnsonation of Catalyst during Preparation Stage
The catalyst is insonated during the preparation stage. As the ultrasonics produce cavitation and mixing a t the microlevel, insonation affects the catalyst particle size, crystallite size, pore size, and the pore-size distribution (Kapustin, 1962; Li et al., 1964; Slaczka, 1964). Case II: lnsonation of Already Prepared Catalyst
The insonation is carried out while the catalyst particles are suspended in a liquid medium. This method is useful in re-
generating spent catalysts (Graves et al., 1966; llertes, 1962). It appears that ultrasonically produced cavitation helps to remove the deposits on the spent catalyst and so to regenerate its activity. This is an example of the cavitation and micromixing acting a t various points on the catalyst surface (both inner and outer). Case 111: lnsonation of Reaction Mixture Including Catalyst
This procedure was not considered in detail in our earlier publication. It is now expanded to include the recent developments. The reaction mixture containing both the catalyst and the reactants is insonated. The insonation done in this manner could affect both the catalyst and the reactants. This is a much more complex case (compared t o Cases I and 11) and may have commercial possibilities. Kiener and Young (1958) studied the effect of stationary sound tvaves on the decomposition of formic acid and of ammonia and the hydrogenation of ethylene with nickel filament as catalyst. The reaction was carried out in the filament temperature range of 12O-19O0C. With this type of ultrasonic treatment, the decomposition rate of formic acid increased by about 50% and that of ammonia by about 1570, whereas hydrogenation of ethylene remained unaffected. The increase in reaction rate was attributed to an increase in mass transfer owing to insonation. Kikolaev and Askadskii (1958) used this method in studying the effect of insonation (20 kHz) on the decomposition rate of 0 . l X hydrogen peroxide with silica gel catalyst. The increase in the rate of decomposition was attributed to breaking of the silica gel particles. The sound waves did not change the concentration of hydrogen peroxide without the catalyst. Greguss and Greguss (1960) catalyzed the decomposition of hydrogen peroxide with manganese dioxide gels and suspensions (at 25OC) and insonated the reaction mixture a t 875 kHz. There was a faster initial evolution of oxygen compared with no insonation, but the time required t o complete the decomposition was the same in both cases. The initial increase in rate was attributed to the lesser thickness of the adsorption Ind. Eng. Chem. Prod. Res. Develop., Vol. 12, No. 2, 1973
155
Table I. Corrected Rate Constants and Arrhenius Parameters for Decomposition of Hydrogen Peroxide (Ranganathan et al., 1971) Temp,
37 44 54 63
O C
Blank, hr-' X 10'
2.97 7.74 18.51 60.00 Arrheniur parameters
E , kcal Log A , A in hr-1 g-1 cat
Rate constant, hr-' g - ' (cat) X 1 0' Uninsonated gel lnsonated gel
29.5 52.7 127.0 364.0
Rate constant, hr-' m--2 X 104 Uninsonated gel
lnronated gel
0.089 0.156 0.377 1.08
0.109 0.175 0.578 2.18
29.9 48.1 158.4 597.0
Blank
Uninsanated gel
lnronated gel
23.2 12.9
21.5 12.5
28.0 17.0
layer on the iiisonated catalyst compared with the uninsonated one. Aero Projects (1965) reports coupling a buridle of wires coated with a catalytic material (Pt, S i , or Fe) to a transducer. With wires immersed in a reaction mixture aiid the transducer driven in a frequency range of 0.5-300 kHz, there was a n increase in reaction rate for reactions such as formation of ammonia and sulfur trioxide and fat hardening and oxidation of methyl alcohol, ethyl alcohol, and carbon monoxide. Gulyalv and Levinson (1968) studied the increase in activity of catalase in a fresh suspension of "chlorella" owing to insoilation a t a frequency of 800 kHz, an intensity of 7.6 W cm-2, arid a duration of 5 min. The activity increase was 200250%. This increase was stimulated by the presence of K-, Fez+, and Mg2+ and was inhibited by the presence of HzlW-, YOz-, and YO*-, all in physiological concentrations. Cook aiid Cho (1969) were able t o induce the catalytic (FeS04) oxidation of glucose to glucuronic acid in dilute, aerated solution by irisonating the solution at' 96.84 kHz. Iyoda et al., (1970) studied the effect of ultrasonics on the Ziegler polymerization of ethylene in n-heptane. Acceleration of polymerization was observed for lower energy inputs, and depression was observed for higher energy inputs. Lfal'tsev and Solov'era (1970) synthesized ammonia in a n aqueous solutioii of a nitrogen-hydrogen mixture by insonation a t 550 kHz aiid 4-5 W cnir2.Catalysts used for the reaction were Pt black, Rh black, and Pd black. No reaction took place in t'he absence of insonation. The activities of the catalysts decreased in the order Pd < R h < Pt. Smit'h et al. (1971) investigated the catalytic oxidation or modification of certain water pollutants (aniline, stilbestrol, o-chloronitrobenzene, phenol, aiid iodide ion) with ultrasound. X synergistic effect between ultrasound arid certain heterogeneous cat'alysts (1InO2, Pt/C, Pd/C, aiid RhjC) was demonstrated quantitatively for the iodide ion and qualitatively for the other compounds. The authors have not carried out aiiy work in this area (Case 111). Review of the pertinent literature suggests that this field is fertile for elucidation. There are many examples in each area where ultrasonics have been used with advantage either to increase the activit'y of a catalyst or to regeiierate a spent catalyst or simply to bring about a n increased rate of reaction when the reaction system was insonated along with t'he catalyst. However, of the methods described, the application of sound waves during the preparation stage (Case I) of the catalyst appears to be most fruitful as far as the production of a n active catalyst is concerned. The results reported in the present investigation are on the catalysts prepared in the preseiice aiid absence of ultrasonic waves. The decomposition of hydrogen peroxide was used t o 156 Ind. Eng. Chem. Prod. Res. Develop., Vol. 12, No. 2, 1973
measure the activity of various catalysts. This reaction was selected mainly because the analytical technique and the reaction procedure are simple, and there are 110 side reactions (Ranganathan et al., 1971). The results of a number of other investigators for various unsupported catalysts prepared without insonation have also been included. Silica Gel Catalysi
The kinetic studies (Ranganathan e t al., 1971) were carried out in a batch reactor with silica gel catalyst (prepared with aiid without insoiiatiori). Surface areas and pore volumes were also measured. The ultrasonic equipment consisted of a generator capable of producing ultrasonic rvaves at, 90 kHz a t an average output of 80 W. The silica gel catalyst was so affected by irisonation that both the catalytic activity as well as the activation energy were increased (Table I). Supported Catalysts
-1fter having looked a t the effect of ultrasonics on unsupported catalysts (Kapustin, 1962; Li et al., 1964; Ranganathan et al., 1971; Slaczka, 1964) and finding that insonation increased the activity of the catalyst (silica-gel), we decided to study the effect of insonation on supported catalysts. Since a literature search revealed that little is known about the effect of ultrasonic waves on the supported catalyst systems, we studied chromium oxide, manganese oxide, cobalt oxide, and nickel oxide on 7-alumina support. Though the reaction, decomposition of hydrogen peroxide, was used as a measure of the activity of a particular supported catalyst, the technique was somewhat different from that used iii the case of silica-gel catalyst (see the discussion under Compensat,ion Effect). The experimental technique and the insonation procedure have been described in detail (Rlathur et al., 1972). The activation energies E , frequency factors A , pore volumes, and BET surface areas of the supported catalysts are shown in Table 11. Some consistent effects of ultrasonics have been observed. I n all cases the activation energy and frequency factor are higher for the insonated catalyst. Evidently, insoiiatioii during the catalyst preparation stage has affected the active sites of the catalyst. Studies are currently being carried out to further elucidate this phenomeiioii. The surEace areas (13ET) aiid pore volumes (between 120 and lo6 A) are higher for the insonated catalysts. One possible explaiiation for these results is the scouring action of the ultrasonic waves during support impregnation which could have kept crystallites of the active metal oxides from blocking the exterior pores. The surface of the impregnated catalyst pellets is smooth in the uninsoiiated case and rough (scoured) in the insonated case.
Table It. Arrhenius Parameters, BET Surface Areas, and Pore Volumes of Supported Catalysts E, Co to lyst
Alumina support Chromium oxide-alumina Insonated Uninsonated Manganese oxide-alumina Insonated Uninsonated Cobalt oxide-alumina Insonated Uninsonated Nickel oxide-alumina Insonated Uninsonated
kcol mol-’
Log A, A in hr-’ g-’
Surface area (BET), m2 g-’
Pore vol, pores between 120 and 10‘ A
Rote constant ot 50°, hr-1 g-1
...
...
220
0.314
...
14.80 13.30
10.73 9.50
135 108
0.347 0.290
4.17 3.37
7.0 5.64
5.85 4.84
120 107
0.306 0.267
7.9 7.0
5.60 4.94
125 105
0.359 0,284
1.80 1.60
9.45 8.09
6.62 5.62
143 114
0.275 0.255
1.75 1.45
13.0 10.6
The rate constants (hr-l g-l) a t 50°C are also higher for the insonated catalyst in all cases. There are probably a t least two factors causing this difference in the supported catalyst case: differences in the surface areas and differences in reaction kinetics. At present the authors have no satisfactory technique to measure the active metal area (as opposed to the BET surface area); therefore, the effect of the differences in surface areas is unresolved. It is possible that insonation has produced a catalyst where the active metal oxide is better dispersed on alumina, hence creating more active sites. I n chromia on alumina, electron microprobe analyzer scans were made to measure qualitatively the depth of penetration of the active metal into the alumina support. X a t h u r et al. (1972) found that the chromium concentration is highest a t the periphery and decreases as the center of the pellet is approached. This might be expected as the diffusion of chromia controls the chromium concentration in the pellet. The concentration profile was approximately a parabola in both cases. Chromia had a better chalice to diffuse into the pores and disperse on the support in the insonated case. Hence, the ultrasonically produced catalyst will be more active (per unit mass of catalyst). Compensation Effect
The calculation of activation energy alone does not permit a complete evaluation of the activity of a catalyst since the frequency factor -1 is also catalyst dependent. This type of behavior, where both A and E seem to be linked together so that A increases as E increases, has been called the “compensation effect” by Cremer (1955) and the “theta effect” by Schwab (1950). A compensation effect plot is shown in Figure 1. I n the bottom line, similar data for hydrogen peroxide decomposition by other catalysts as gleaned from the literature, along with the insonated and uninsonated silica gel, were also plotted. h close examination of the data of Figure 1 shows that the bottom line which correlates 33 points corresponds to quite diverse catalysts used in the study of the decomposition of hydrogen peroxide. They are either unsupported or alloy catalysts. Because of the diversity of catalysts on the compensation line, the Arrhenius parameters of any other catalyst for the decomposition of hydrogen peroxide should lie on the bottom line (Ranganathan et al., 1971). However, the present results on four supported catalysts indicate that a separate
-101
1
0
10
I
20
I
I
30
I
40
activation energy, kcol Figure 1. Compensation effect for decomposition of hydrogen peroxide on various supported and unsupported catalysts 0 Unirradiated rare eorth oxides
A
0
4
Irradiated rore eorth oxides Irradiated and annealed rore earth oxides Bismuth alloys Indium alloys
-+ Gallium alloys
-0- Uninsonoted chromium oxide-alumina 4- Insonoted chromium oxide-alumina
& Uninsonoted cobalt oxide-alumina 4 lnsonated cobalt oxide-alumina
0
Uninsonated manganese oxide-alumina
#
lnsonated manganese oxide-alumina
0 4
4
X
Uninsonoted nickel oxide-alumina Insonated nickel oxide-alumina Uninsonated silica gel Insonoted silica gel Uninsonoted manganese oxide- lumina, KMnOd titrotion method
Ind. Eng. Chem. Prod. Res. Develop., Vol. 12, No. 2, 1973
157
Table 111. Comparisons of Methods with Uninsonated Manganese Oxide on q-Alumina Method
Permanganate titration Gasometric, oxygen evolution
Activation energy, kcal/mol
Log A, A in hr-* g-’
4.23
3.30
5.64
4.84
(approximately parallel) compensation line is formed, which might be due t o the following: With all catalysts in the bottom line, the permanganate titration method had been used to follow the kinetic studies, whereas with the four supported catalysts (because of their high activity), a gasometric technique was used for analysis. All catalysts on the bottom line are either unsupported or alloy catalysts, whereas the catalysts on the top line are supported catalysts. Differences in basic metal properties of the four supported catalysts as compared t o the other 33 catalysts. To compare the differences in analytical techniques, the Arrhenius parameters for manganese oxide-4-alumina catalyst were evaluated by use of both permanganate titration and gasometric methods. Even though the results in Table I11 show a deviation in parameters of around 25%, the Arrhenius parameters still lie on the same compensation line (Figure I). Hence, the variation in analytical techniques does not explain the difference in compensation lines. Either item 2 or 3 may have relevance to the formation of a separate line for the supported catalysts. Furthermore, this line is approximately parallel to the one obtained earlier with 33 catalysts. At present, we have no explanation for this unusual behavior, but more work is continuing t o elucidate this observation.
158 Ind. Eng. Chem. Prod. Res. Develop., Vol. 12, No. 2, 1973
literature Cited
I
Aero Projects Inc., British Patent 991,739 (C1 BOlj) (1965); Chem. Abstr., 63, 4994a (1965). Coekelbergs, R., Crucq, A., Frennet, A., Advan. Catal. Relat. Subi.. 13. 55 (1962). Cook,”C. H:, Cho, Y.,’Yakhak Hoeji, 13, 16 (1969); Chem. Abstr., 73, 45709m (1971). Cremer, E., Advan. Catal. Relat. Subj., 7, 75 (1955). Graves, C. A., Steiner, D. F., Hirdler, F. C., U.S. Patent 3,231,513 (C1 252-413) (1966); Chem. Abstr., 64, 10453a (1966). Greguss, A., Greguss, P., Akust. Zh., 6, 441 (1960); Chem. Abstr., 55. 16094e (1961). Guly’alv, L. S.’, LeGnson, M.S., Zzv. Sab. Otd. Akad. iVauk SSR, Ser. 13201. M e d . N a u k , 1968, 63-8; Chem. Abstr., 71, 106633. (1969). Iyoda, J., Nagata, A., Hayaski, K., Shukara, I., Osaka Kogyo Gujutsu Shakensho Kiho, 21, 67 (1970); Chem. Abstr., 73, 121054-5 (1971). Kapustin, A. P., in “Effect of Ultrasound on the Kinetics of Crystallization,” p 24, Consultants Bureau, New York, N.Y., 1962
Li,-W&-Chou, Mal’tsev, A. N., Kobozev, K. I., Russ. J . Phys. Chem., 38, 41 (1964). Mal’tsev, A. N., Solov’era, I. V., ibid., 44, 1092 (1970). Mathur, I., Bakhshi, N. N., Mathews, J. F., Can. J . Chem. Eng., 50. 344 (19721. Meries,-T.‘S., U.S. Patent 2,968,652 (1961); Chem. Abstr., 57, 2406c (1962). Mikovsky, R. J., Weisz, P. B., J. Catal., 1, 345 (1962). Nikolaev, L. A., Askadskii, A. A., K h i m . S a u k a Prom., 3, 131 (1958); Chem. Abstr., 52, ll543b (1958). Ranganathan, R., Bakhshi, N. N., hlathews, J. F., J . Catal., 21, 186 (1971). Schwab,G. h., Advan. Catal. Relat. Subj., 2, 251 (1950). Slaczka, A., Znt. Chem. Eng. Process. Ind., 9 (l),63 (1964). Smith, G. V., Patil, F., Pavlov, Y., Chen, J. V., U.S. Clearinghouse Fed. Sci. Tech. Inform.. P.B. Reu. No. 197733. 1971. Taylor, E. H., in 5th Intern. Symp. on t6e Reactivity of Solids, D 409. Butterworths. Londor,. Eneland. 1965. Taylori‘E. H., Advan.’Catal. R k / a t . h b j . ; 18, 111 (1968). Weisz, P. B., Swegler, E. W., J . Chem. Phys., 23, 1567 (1955). Wiener, H. B., Young, P. W., J . A p p l . Chem., 8 , 336 (1958). Presented at the Division of Industrial and Engineering Chemistry, 164th Meeting, ACS, New York, N.Y., August 27 to September 1, 1972. This work was supported by the National Research Council of Canada.