obtain a t the same time resins which are very tough physically arid have a reasonable osmotic volume change. For applications other than demineralization: a n entirely diff'erent set of resin properties may be required. I n decolorization cycles, a high \vet-volume capacity may be disadvantageous because the resin may show such a high selectivity for the anionic color bodies that it in time picks u p more than it can accomrriodate volume-wise and is ruptured. t\gain> in sonic difficult separations, as in the anionic complexes of the transuranium elements, a delicate balance among the factors of selectivities. diffusion rates. and mechanical properties of the resin must be obtained. At present, the ideal resin properties for any such application can be found only by experimentation. Ho\vevrr! Figures 5 to 7 provide a convenient tool for correlating such data and setting meaningful specifications on the resin needed. Other Implications of Methylene Bridging
A number of variables are involved in the chloromethylation reaction. Among these are: time and temperature. amount and compo>ition of catalyst, amount and purity of chloromethylmethyl ether, and presence of other solvents. Juggling these interrelated variables and the copolymer cornposi tion to obtain a desired point on Figure 2 is still an empirical process. There certainly is not a unique set of conditions for each point of the map, even starting \vith a given copolymer composition. 'This possibility of variability places a heav)- responsibility on the resin manufacturer to produce resin reproducibly within well defined spccifications. T\vo of the three gross properties discussed here must be specified. A chart such as Figure 5 presents a convenient quality control tool in that placing limits on t ~ v oof the properties defines a n area of acceptability.
Literature Cited
l ' h e fact that resins of identical Qtc: Qc: and Tt' can be made from tvidely differing copolymer compositions raises interesting questions as to ho\v other properties of the resin will reflect this difference in polymer topology. Ion selectivities correlate roughly \vith water content. Osmotic volume changes are extremely sensitive to small changes in resin topology to the point of appearing almost erratic. No other Lvork on such variations are reported in the literature. Certainly attempts to d o prrcise physical-chemical measurements on quaternary ammonium resins must note these topological complications.
(1) C h u , B.. IYhitney, D. C., Diamond, R. M., J . Inorg. N u c l . Chem. 24, 1405 (1962). (2) ,Helfferich, F., "Ion Exchange," p. 75, McGraw-Hill, New Eork, 1962. (3) Jones. G. D.. Ind. Eng. Chem. 44,2686 (1952). (4) Lloyd. \V,G.: Xlfrey, T., Jr., J . Poly. Scz. 62, 301 (1962). (5) Pepper. K. LY., Paisley, H. M.. Young. M. A . . J . Chem. Soc. 1953, p. 4097. (6) \$'heaton, K. M., Bauman, \Y. C., Ind. Eng. Chem. 43, 1088 (1951).
RECEIVED for review September 30, 1963 ACCEPTED February 3, 1964
A M M O X I D A T I O N OF ISOBUTYLENE I N A COATED T U B E W I L L I A M
F . B R I L L A N D JOSEPH
H . F I N L E Y
Petro-Tex Chemical Corf , F.WC Chemical Research and Deuelojment Center, Prtnceton, 'Y J . Isobutylene, oxygen, and ammonia react at
500" C. on passing through
a tube coated with a catalytic metal
oxide. Under the most favorable conditions, 80% o f the isobutylene reacted forms methacrylonitrile, methacrolein, and acetonitrile. Ammoxidation in a coated tube i s characterized b y good ammonia utilization and tolerance to high reactant concentrations. Over molybdenum oxide, the reaction rate i s dependent upon oxygen concentration and independent o f isobutylene and ammonia concentrations.
The ammoxida-
tion of methacrolein was investigated and the possible role of aldehyde intermediates in the ammoxidation of olefins considered.
HE first reported reaction of olefins and ammonia in the Tpresence of oxygen t o yield unsaturated nitriles ( 4 ) described the oxidation of isobutylene (2-methylpropene). Subsequent Lvork on this reaction, now generally known as amnioxidation, has been concerned almost exclusively with propylene and has led to the development of some excellent catalysts (9) for this olefin. Furthermore. information concerning the possible mechanism of the reaction has been obtained ( 7 ) 7). Initial attempts in our laboratories to develop a process for thr production of methacrylonitrile led to the discovery that amnioxidation proceeded readily on passing isobutylene through a n unpacked tube: the inside surface of \yhich had catalytic activity. Good methacrylonitrile selectivities and efficient utilization of ammonia, under the proper conditions:
coupled with the absence of local overheating, made the coated tube reactor appear to be superior to a fixed bed for studying reaction variables. Experimental
Reactor. T h e reactors consisted of 24-inch tubes of 304 stainless steel, Vycor, or ceramic, 22-mm. internal diameter, heated by a Hevi-Duty Rlultiple unit furnace. Gases were metered from calibrated rotameters through a manifold to the head of the reactor. Temperatures recorded in this report are the maximum temperatures measured in a thermowell: placed down the center of the reactor. T h e effluent was sampled immediately belo\v the heated portion of the reactor. through a sampling port lagged with heating tape, using an insulated syringe kept a t 160' C. VOL. 3
NO. 2
JUNE
1964
89
Residence times were calculated using the volume of the reactant gases a t the reported temperature and the reactor volume for the entire coated length of the tube (24-inch). Reaction times are somewhat less than the residence time, since the iritial portion of the reactor served to bring the reactants to temperature. Analysis. Product gases were analyzed by gas chromatography using a l?-foot, ','d-inch copper column packed rcith 5% Octoil-S on Fluoropak. 'The temperaturetcas programmed from 35' using a n I: 8( M Model 300 chromatograph. Oxygen, nitroger. and carbon monoxide did not resolve. while carbon dioxide \cas poorly resolved under these conditions. Thew components \$'ere analyzed separately using a silica gel (30- to 60-mesh) column and an activated carbon column (40- to 80-mesh) connected in parallel. Factors for relating peak arcas to mole per cent for each compound \cere determined experimentally. Conversions were calculated on the isobutylerlr reacted as determined on a n equivalent carbon basir by totaling all the carbon-containing products formed. Ammonia could Lot be determined chromatographically, since it was retained on the column. Therefore, in some experiments. thr reaction off-gas was collected in 0.5.V HC1 and the solution back-titrated to the methyl red end point. T h e simplest method of preparing a usable molybdenum oxide coating was to place the oxide powder in a horizontally positioned steel tube and heat the metal surface to redness with a hl6ker burner. Another procedure, which allo\ved the arnount of catalyst deposited to be more conveniently determined, was to place a platinum boat containing about 1 gram of molybdenum oxide near the end of the horizontally positioned tube, heat the tube under the boat to red heat, the remainder of the tube being kept at 200' to 300°, and pass air through the tube under partial vacuum. Within 30 minutes, 0.3 gram of oxide is deposited, which may be observed, \Then a Vycor tube is used: as a fine \chite film throughout the length of the tube. Stainless steel reactor tubes coated with molybdenum oxide performed \vel1 and gave reproducible results for several months before conversion and selectivity decreased (from 39 to 2 7 7 , and from 5s to 40%, respectively: under the adopted
test conditions-530', 5.1 seconds. 6.9% isobutylene, 13.8% oxygen, 13.87, ammonia). Cleaning the tube Lvith scouring po\vder and steel wool and treatment \vith concentrated ammonia restored original activity, apparently by removing accumulated iron oxide. Tubes Lvith various other catalyst compositions icere prepared by coating the tube lvith a slurry of alpha-alumina poicder (2) (United Mineral 8: Chemical Corp.) impregnated with solutions of the appropriate salts. The coating \cas dried a t room temperature and calcined at 200'. The active lives of tubes prepared in this manner were not fully- investigated, but usually decreased selectivity to methacrylonitrile \vas observed in standard runs conducted after completion of variable studies in Lchich steam \vas used as a diluent. Fixed bed catalysts \$-ere prepared in the usual manner by impregnating :g-inch Alundum pellets rvith solutions of metal salts and phosphoric acid, evaporating the solvent in a Rinco evaporator. and forming the oxides by heating to 450' to 500' in a stream of air. Attempts to prepare active Alundum tubes, allo\ving unequivocal comparison with fixed beds, by sublimation of molybdenum oxide onto the walls were unsuccessful. Direct impregnation of Alundum tubes with solutions of other metal salts in the manner used mith pellets gave tubes Lchich initially performed \cell but lost activity during a single experiment. I t appears that commercially produced illundum tubes have an inner surface which is difficult to coat. Results
.4molybdenum oxide-coated tube was chosen for the major part of the study of reaction variables, since it yielded reproducible results over long periods of time. Isobutylene, ammonia, and oxygen concentrations, temperatures, residence times, and the choice of diluent influence either the over-all reaction rate or the relative rate of production of the major products-methacrylonitrile, acetonitrile, methacrolein, and the oxides
Table 1. Ammoxidation of Isobutylene in a Molybdenum Oxide-Coated Tubea Reactant c IC yo Selecfzzzty - ~ _ _ Concn., Val. 7 0 Iso-CnHs 0 2 A-HJ Conrersion .\iA.\.b 'MAC MeC.V
-~
co, co2
Effect of Oxygen 0 0.7 3.4 6.9 13.8 17.2 20.7
13.8 13.8 13.8 13.8 13.8 13.8 13.8
6.9
13.8 13.8 13.8 13.8 13 8 13 8 13 8
0 1.6 3.5 6 9 11 0 13 8 17 9
4.6 6.9 10.4 13.8
13.8 13.8 13.8 13.8
13.8 13.8 13.8 13.8
6.9
0 1 6 15 32 37 62
n High 75(est.) 59 58 60 56
0 0 3 0 1 8 5
0 0 3 14 17 13 11
0 0 19 225 19 28
40 30 24 12 1 1 2
0 4 4 9 12 17 16
60 39 43 25 24 25 19
0 7 3 7
21 12 14 8
28 33 33 38
7 6 4 10 2
15 11 13 10 10
26 2: 32 30 36
Effect of Ammonia 34 29 30 32 31 32 27
0 24 28 55 64 58 63
Effect of Isobutylene 41 27 22 16
51 48 47 45
Effect of Dilution
a
90
9.5 4.8 9.5 6.0 13.8 13.8 10.0 20.0 20.0 25.0 12 5 25.0 33.3 16.7 33.3 Residence lime 5.7 sec., temp. 5,?0" C., nitrogen diluent.
27 36 33 27 72 .Wethacrylonitrile.
l&EC PRODUCT RESEARCH A N D DEVELOPMENT
52 56 51 49 52 Metharrolein.
Table II.
Reactant Concn
6.6
7 7 7 7 12
2 2 0 0 4
Val.
:;
SHa 14.4 14.4
T. ' C 545 545
13.2 13.2
13.2 13.2
560
14 14 14 14 18
14 4 14 4 14 0 14 0 24 8
Iso-C~H~ 0 2 7.2 14.4 7.2 14.4
6.6
,
4 4
0 0 7 .Idelhacr~lonitrile. Experinierital).
' .Vfethacrolein.
Res. Time,
Ammoxidation of Isobutylene
Selectiriiy
52
Concerszon
.Vl t 12 lb; on .VHR
2.5 2.5
28 28
MA,Va 56 57
5.1 3.3
14 11
9 28
16
580
4
23
. .
560 575 560 560 560
2.0 2.0 4.0 4.0 4.0
19 28
60 52 45 41 41
11 14 16
. . . .
82 19 45 17
Sec.
25
25 44
AYr diluent used unless steam
of carbon. I n studying concentration effects, a fixed residence time of 5 seconds, a temperature of 530°, and nitrogen diluent were generally used. 17ariation of the concentration of a single reactant while holding other concentrations constant? and preferably in excess: indicates that the rate of reaction for isobutylene is dependent on the oxygen concentration and not influenced by either isobutylene or ammonia concentration. As shown in Table I. selectivity to methacrylonitrile is not significantly harmed at the increased conversions resulting from the use of higher oxygen concentrations. In the absence of oxygen, no nitrile forms: but about 25yGof the ammonia reacts and large amounts of hydrogen are formed. \Vhile ammonia concentration has no effect on the over-all reaction rate, as demonstrated by the almost constant isobutylene conversion at various ammonia feed concentrations, shoivn in Table I, the relative amounts of nitriles and methacrolein are strongly affected. Ft'hen ammonia is present in excess of isobutylene, the production of methacrolein is very low, I n the absence of ammonia, when methacrolein is the sole useful product, the selectivity is only 40y0,, but as increasing amounts of ammonia are added, the combined selectivity to nitriles and methacrolein rises to above 80%. A manifestation of the same effect is the accompanying large decrease in combustion products, CO and CO,. Over a threefold range of isobutylene concentrations a t equal fixed concentrations of ammonia and oxygen, no change in the rate of reaction of isobutylene or the rate of formation of methacrylonitrile was observed. T h e favorable effect of low isobutylene concentration on the formation of methacrylonitrile and the absence of methacrolein was observed at the lowest isobutylene concentration shown in Table I. No trend in the production of methacrolein a t high olefin feed concentrations was apparent because of the lack of precision in determining products present in low concentrations. T h e outstanding advantages of the coated tube reactor over the fixed bed, in addition to the obviously small amounts of catalyst required, were the high concentration of reactants which are allobved and excellent utilization of ammonia. Heat transfer does not become a problem until the total concentration of olefin, oxygen, and ammonia (in the ratios of 1 : 2 :2) is increased to above 80'%.. Such high concentrations appear to give a temperature rise similar to that found when a cool flame occurs during hydrocarbon combustion, but nevertheless selectivities to methacry-lonitrile of 52Tc are obtained, with 727' of the isobutylene reacting. T h e fast reaction rate and resultant temperature rise are due entirely to the high oxygen concentration. Lowering the oxygen concentration from
.14eC.\
UAb
14
. .
, ,
, .
4 10
9
12
specijied.
Remarksc
AfoO,-coated tube llo03-coated tube packed with Vycor Raschig riiigs Empty uncoated steel tube Uncoated tube, 150 cc. of 10% ;2foO;1 catalyst. Flow rates as abol-e lloO;i-roated tube. Steam diluent MiuO.i-coated tube Coated tube.d Steam diluent Fixed bed.d Steam diluent Coated tube.d Steam diluent
17
7 93 V-.ifo-P ( 73) on ' j , - l n c h p p l l f f s or porcde,ed alpha-alumina (see
3 3 7 , to 16% eliminates the temperature rise and lowers the conversion to 537,. At the resulting 57Yc selectivity, the of methacrylonitrile, a desirable effluent contains over 3 mole 7, condition for product recovery. Efficient utilization of ammonia depended strongly on the presence of steam in the feed. T h e favorable effect of steam and the improved performance of a coated tube, compared to a fixed bed catalyst of similar composition. in converting ammonia to nitrile are illustrated in Table 11. Steam was not used extensively in the present study to avoid the increased difficulty in obtaining reliable product analyses. T h e determination of the water formed was especial1)- facilitated by using a nitrogen diluent. 'The Lvater formed in the reaction was close to that expected from the indicated stoichiometry of the ammoxidation and from the combustion of ammonia to nitrogen and water. O n varying residence times from 2.2 to 7.6 seconds, the expected increase in isobutylene reacted was observed (Figure 1). A plot of the log of oxygen concentration. obtained by assuming oxygen consumption to be four times that of iso-
500 22
475
5
10
15
20
530 51
33
25
30
35
595
'
OC(5
76
40
1s.c 1
SeC(5309
45
V0 Co nve r s i o n
Figure 1 . Dependency of selectivities on changes i n conversion, showing the effect of time and temperature Solid symbols represent changes produced b y varying temperatures Hollow symbols designote chonges produced b y varying residence times 0 Mefhacrylonitrile Acetonitrile C CO Cog
+
A
VOL. 3
NO. 2
JUNE
1964
91
butylene, against time shoi\ed that dependency of rate on oxygen concentration was first-order. T h e ratio of oxygen to isobutylene consumed generally fell between 3 and 4 in the absence of steam, but direct oxygen measurements appeared less suitable for rate tredtmeiit. T h e first-order rate constant, kl = 0.054 set.? at 530', could also be obtained from the experiments in which oxygen concentrations \vere varied a t a fixed residence time by plotting initial oxygen concentration against isobutylene reacted. T h e minor importance of consecutive reactions in destroying products \vas indicated by a drop of only 10% in selectivity to methacrylonitrile on tripling the reaction time. No well defined trend in the production of rnethacrolein and acetonitrile was observed. Estimations of the Reynolds number for the conditions studied gave values of below 50 for the fastest flow rates, indicating that in no case are the results influenced b) the onset of tilrbulence. I t was difficult to distinguish any temperature effect between 475' and 595' other than that which Lvould be anticipated from the expected increase in rate and conversion levels. Generally, changes in selectivities are similar to those produced on varying conversions by changes in residence time as shown in Figure 1. An exception was the single increase in carbon dioxide and carbon monoxide and decrease in methacrylonitrile observed, relative to those expected for the conversions obtained, a t 595'. At this temperature complete consumption of oxygen occurred. Another possible effect of temperature, which must be regarded with suspicion because of the erratic nature of the data, was a decrease in the combined selectivity to by-product methacrolein and acetonitrile. T h e production of methacrylonitrile from methacrolein in the coated tube reactor was examined to determine the possible
Table 111.
Ammoxidation of Methacrolein in Molybdenum Oxide-Coated Tube 5% Selecticzty MethacReactant Concn., V d . % % ConryloAcetoCO, Oxygen Ammonia version nitrile nzlrile COz
0
1.6 3.2 6.3 9.4 12.5
12.5 12.5 12.5 12 5 12 5 12.5
Effect of OxygenQ 25 72 39 83 74 64 97 57 96 54 99 53
24 8 12 10 11 15
3.2 3.2 3.2
Effect of Ammonia" (LOW0 0 39 0 1.6 56 57 12.5 74 64
8.0 8.0 8.0 8.0 8.0 8.0
Effect of Ammoniaa (High 0 57 0 1.3 71 32 2.5 85 53 3.8 87 52 5.6 88 53 11 2 88 55
0
10 24 33 35 31
2 )
0 3
12
92 40 24
0 2 )
0 4 2 8 8 9
91 58 44 39 40 35
Effect of Temperatureb Temp.,
c.
Res. T i m e 0 0 4 8 0 0 0 25 4 5 14 75 9 18 4 3 60 73 10 27 4 1 81 63 15 20 4 1 95 64 15 31 3 9 99 53 8 57 3 7 95 35 7.6% Mefhocrolein, 3.9-sec. res. time, 530" C., nitrogen diluent, 7.67; .Methacrolein, 72.57; 0 2 , 72.5rc .VHa, nitrogen diluent. O
380 41 5 450 480 495 530 575 a
92
I & E C PRODUCT RESEARCH A N D DEVELOPMENT
role of aldehyde as a reaction intermediate in the ammoxidation of isobutylene. At a constant methacrolein concentration, the dependency of rate on both oxygen and ammonia concentration reported for the production of acrylonitrile from acrolein was found to be applicable (7). T h e specific influence of ammonia depended somewhat on the oxygen concentration. At low oxygen concentrations, added ammonia decreased combustion and increased the rate of conversion to nitrile, even when added in very large excesses. At higher oxygen concentrations, the effectiveness of ammonia in eliminating carbon monoxide and dioxide leveled off a t low ammonia concentrations. This effect of oxygen was believed to be caused by the favored oxidation of ammonia at high oxygen concentrations. Analysis of the products for unreacted ammonia tended to confirm ammonia combustion at high oxygen concentrations; the conversion of ammonia increased and selectivity to nitrile decreased. In agreement with the Russian literature (75), unsaturated nitrile was obtained in the absence of oxygen with good selectivities, as shoivn in Table 111. The relationship between temperature and conversion of methacrolein to various products a t approximately equal residence times is shown in Table 111. In the absence of other rate data, it is not clear whether the increased formation of carbon dioxide and monoxide a t the almost complete conversion obtained above 480' is due to the oxidation of nitrile or the increased probability of aldehyde combustion at higher temperatures. ii-hile the patent literature reports, almost entirely. ammoxidation catalysts and conditions which give optimum results with propylene. the catalytic tubes used in the present study were far more effective for the conversion of isobutylene to unsaturated nitrile. iVhile olefin structure and catalyst effectiveness must be definitely related. the magnitude of the differences found in olefin performance was unexpectedly large. The decrease in selectivity of from 317, to 11% on raising the temperature from 540' to 585' to increase conversion from 8% to 30'%, indicates that the temperatures required to obtain appreciable activity with propylene are detrimental. Discussion
T h e simplest reaction pathway producin'g methacrylonitrile which is consistent with our observations involves the prior oxidation of isobutylene to methacrolein. The aldehyde formed may react rapidly and reversibly with ammonia; the addition product then loses water and is oxidatively dehydrogenated to nitrile. Higher ammonia concentrations shift the equilibrium reaction and. by removing reactive aldehyde, prevent combustion. Thus, by favoring the formation of the more stable nitrile, conversion to products other than carbon dioxide and carbon monoxide is optimized. Unfortunately, in the present system the high temperatures at which isobutylene is most efficiently oxidized favor combustion when methacrolein reacts with ammonia and oxygen in the absence of olefin, perhaps by shifting the postulated equilibrium toward aldehyde. Consequently, the postulate of a methacrolein intermediate would be compatible with good nitrile selectivities from isobutylene only if limited amounts of carbon dioxide are produced directly from the olefin-i.e., the primary yield of methacrolein from isobutylene is very high. Popova has interpreted the correspondence between carbon dioxide formation and aldehyde reactivity in the oxidation of isobutylene and other olefins. over a copper catalyst. as indicating that carbon dioxide forms by the sequential oxidation of
methacrolein (73). For the more carefully studied oxidation of propylene, the sequential oxidation of acrolein (5, 70, 77, 78) and the parallel oxidation of propylene or parallel and sequential routes which depend on temperature (3, 77, 79) have all been stressed as the main source of by-product carbon dioxide. From the observed conversions, the rate of reaction of methacrolein must be many times that of isobutylene, but it is not clear whether it is quantitatively sufficiently rapid to account for the low concentration of methacrolein by-product in the ammoxidation of isobutylene. Other more complicated pathways, not apparent from the kinds of information yet obtained, are not only possible but likely to occur. For example, the removal of an alphahydrogen from the olefin by oxygen on the catalyst, followed by reaction of the dehydrogenated product with an active nitrogen species, is an attractive possibility. Such a mechanism would explain the simultaneous formation of combustion products and acetonitrile as acceptably as one involving the prior formation of aldehyde and would meet the requirement of an allyl intermediate suggested by experiments with deuterated propylene (7). The formation of nitrile by the reaction of active nitrogen with olefins has been reported (76), and the formation of nitrogen from ammonia in the present study supports the argument that an active species may be involved. The site of the initial attack on olefin is ambiguous. For the oxidation of propylene over a copper catalyst, which presumably is initiated similarly, both addition to the double bond (8). followed by elimination from the allyl position, and initial removal of hydrogen from the methyl group ( 7 ) have been proposed. Acetonitrile is a by-product from both isobutylene and methacrolein and cannot arise simply by oxidative scission of the double bond, as has been suggested for propylene and acrolein (7). None of the reaction variables studied seemed to influence the relative amount of acetonitrile produced, perhaps indicating that its origin is intimately connected with the mode of formation of methacrylonitrile. However. catalytic surfaces of different compositions differed widely in their ability to produce acetonitrile. O n the molybdenum oxide coating, no acrylonitrile and only trace amounts of H C N were produced. Although diffusion times to the tube wall are of the same order of magnitude as reaction times, the dependency of the rate on oxygen concentration alone indicates the reaction on the wall to be the rate-controlling process. Isobutylene, which has the poorest diffusivity, has no effect on rate. Also, modification of the reactant flow with inert packing, such as Alundum pellets, has no observable effect on the reaction in the coated tube. Thus, while the potential ability of differing diffusion rates to affect selectivity is present in the reaction system, the experimentally determined results do not indicate how or if they are involved. The failure of transfer processes to the external catalyst surface 10 affect the oxidation rate of propylene on supported copper oxide has also been
observed (6). Another possible difference between reactions in coated tubes and in packed beds is the small probability of desorbed product reaching the catalyst surface again in a coated tube. The problem of comparing the behavior of coated tubes and fixed beds is complicated by the difficulty of preparing and examining similar catalytic surfaces in both systems. For molybdenum oxide, which is a poor oxidation catalyst for olefins (7, 72) but was the only oxide volatile enough for deposition directly on the surface of a tube, performance was improved in every respect-e.g., nitrile production as well as efficient utilization of ammonia and tolerance to high reactant concentrations. In view of experimental difficulties, especially the failure to impregnate an Alundum tube directly to give a catalyst with sufficiently long life, the safest comparison with other metal oxide catalysts involves the coatings made with a powdered Alundum slurry. Such comparisons indicate that good catalyst compositions do not necessarily improve nitrile production but always show better ammonia utilizations (Table 11). Conclusion
Coated tubes present a good system for the study of ammoxidations and for the utilization of the ammoxidation reaction for the production of nitrile. As for all catalytic oxidations, little is known about the mechanistic details of the reaction, but it appears that further studies can be facilitated by avoiding the complication of packed bed reactors. Ac knowledgmenl
The authors acknowledge helpful discussions with Alfio J. Besozzi. literature Cited
(1) Adams, C. R., Jennings, T. J.. J . Catalysis 2, 63 (1963). (2) Adey, W. M., Calvert, W. R.. Adrsnn. Catalysis 9, 764 (1957). (3) Belousov, V. M., Gorokhovatskil, Ya. B., Rubanik, M. Ya., Kinetika i K Q ~ Q 3, ~ Z221 Z (1962). (4) Cosby, J . N. (to .4llied Chemical & Dye Corp.), U. S . Patent 2,481,826 (Sept. 13, 1949). (5) Enikeev, E. Kh., Isaev. 0. V., Margolis, L. Ya.. Kinetika z’ KQtQlzZ2, 431 (1960). (6) GorokhovatskiI, Ya. B., Popova, E. N., Rubanik, M. Ya., Zbid., 3, 230 (1962). (7) Hadley, D. J., Chem. Ind. (London) 1961, 238.
(8) Hearne, G. W., Adarns, M. L. (to Shell Development Co.), U. S . Patent 2,486,842 (Nov. 1, 1949). (9) Idol, J. D. (to Standard Oil Co., Ohio), Zbtd., 2,904,580 (Sept. 15, 1959). (10) Isaev, 0. v.,Margolis, L. Ya.: Kinetika i KQtdiZ 1, 237 (1960). (11) Isaev, 0. B., Margolis, L. Ya., Rogenskii, S. 2 . : Z h . Obshch. Khim. 29, 1522 (1959). (12) Kutseva, L. h’.,Margolis, L. Ya., Zbid., 32, 102 (1962). (13) Popova, N. I.: Vermel, E. E., Milrnan, F. A . , Kinetika i Kataliz 3, 241 (1962). (14) Porter, Frank (to Solvay Process Co.), U. S . Patent 2,294.130 . , ’ (Xug. 25. 1942). ’ 115) Rafikov. S. R.. Sernbaev, D. Kh.. Survorov, B. V.. Z h . ’ bbshch. Khz’m. 32, 839 (1962). (16) Tsukamoto, A , , Lichtin. N. N., J . A m . Chem. Soc. 82, 3798 (1960). (17) Veatch. F., Callahan. J. L., Milberger, E. C., Forrnan, R. W., in “Proceedings of 2nd International Congress on Catalysis.” Vol. 11, p. 2656, Editions Technip, Paris, 1960. (18) Voge, H. H., LVagner. C. D., Stevenson, D. P.: J . C~talVsis 2, 58 (1963). (19) LVoodham, J. F., Holland, C. D., Ind. Eng. Chem. 52, 985 (1960).
RECEIVED for review March 3, 1964 ACCEPTED April 20, 1964 Symposium on New Catalytic Reactions in Organic Chemical Processes. Joint with Division of Petroleum Chemistry, Division of Industrial and Engineering Chemistry, 147th Meeting, ACS? Philadelphia, Pa., April 1964. VOL. 3
NO. 2
JUNE 1 9 6 4
93
