NOTES
Sept., 1956
acteristics of Raney cobalt catalyst have been made and are reported here. Experimental
3.2
e; *
2.8
E 2.4 e
9
&4
1339
2.0
5 1.6 1.2 0.2
0.4
0.6
0.8
v /.v m .
1.0
1.2
1.4
1.6
~~
Fig. l.-He;Lts
of adsorption on acetylene black.
sions and particle diameters (Table 11) indicates that L-2817 has attained a higher degree of internal ordering than Graphon. One might expect a corresponding increase in surface homogeneity. This indeed is the conclusion that may be drawn from the heat curve for L2817. The general shape of the curve is similar to that reported for Graphon. However, the maximum associated with intermolecular attraction of adsorbate molecules, has been displaced from approximately V / V , = 0.6 in the case of Graphon to V / V , = 0.9 in the present example. Thus, in L2817 there is a much sharper distinction between the region of monolayer completion and subsequent adsorption. TABLE IT Graphitized end Parent blacks products L2463 Graphon L-2817 Spheron 6 All values in &ngatroms
Particle Size 250 500 La" 51 76 92 L," 23 46 74 a L. = crystallite dimension parallel to carbon layers within crystal. * Lo = crystallite dimension perpendicular to parallel layer groups.
A comparisone has been made between the height of the maximum in the heat curve for argon on Graphon with the theoretical interaction energy for an argon molecule on a graphite surface. It was suggested that the experimentally observed mauima may be lower than expected because of residual disorder on the graphon surface. It is interesting that the maxima for nitrogen on graphon and L2817 both occur a t about 3 kcal./mole in spite of differences in degree of graphitization. Acknowledgment.-The author is indebted to Prof. R. A. Beebe for helpful discussions and suggestions on calorimetric technique.
Rnney cobalt catalyst was prepared in a manner similar to that used for nickel. Fifty grams of the alloy2 containing 40% cobalt and 60% aluminum was dissolved in a solution of 200 g. of sodium hydroxide in 1 1. of distilled water a t 50". The digestion time with stirring was one hour. The catalyst was washed in a continuous-type washer with 8 1. of distilled water which had had hydrogen bubbled through it for 15 minutes prior to use and also during the washing. The catalyst was then washed with 1 1. of 95% ethanol in the washer. The catalyst was removed and washed five times b y centrifugation and decantation with 100% ethanol. The ethanol washings were cloudy. The catalyst wa,a stored in the refrigerator. It is interesting to note that the Raney cobalt is easier to wash free of alkali than is the Raney nickel. The wash solution was neutral to phenolphthalein after 1 1. of water had passed through the washer, whereas 10 to 12 1. is sometimes re uired with the nickel. %he palmitic acid was Eastman Kodak Company white label grade. The behenic acid was a sample which had been prepared by hydrogenation of erucic acid and subsequent recrystallization. Fatty acid adsorption studies were made as previously d e ~ c r i b e d . ~Inconclusive results were obtained with palmitic acid. A cobalt compound, which is slightly soluble in the benzene, is formed during the shaking. As a result, evaporations of the supernatant benzene solutions are contaminated, giving erroneous readings. This benzene-soluble compound may be a cobalt palmitate formed when oxygen in the adsorption tube oxidizes a small amount of the cobalt, allowing a salt to form. Several adsorption measurements were carried out with the benzene-acid solutions covered with argon instead of nitrogen and oxygen from the air. The blue coloration was present in the benzene solution a t the completion of the shaking, although the intensity of the color was decreased somewhat. A smooth adsorption isotherm was not obtained using palmitic acid as the adsorbate. Behenic acid, CHr( CH&-COOH, was substituted for palmitic acid in an attempt to reduce the solubility of the cobalt salt. Behenic acid is less soluble than palmitic acid in benzene. Also, the oxygen was swept out of the adsorption tube with argon saturated with benzene prior to shaking. In contrast to the results obtained with almitic acid, little blue coloration was observed in the befenic acid-benzene solution following the adsorption. As in the case of palmitic acid, the color was most intense in the strongest acid solutions. The r e d t s of the adsorption studies are shown in Fig. 1.
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Equilibrium concn. of acid in benzene (g./l,). STUDIES WITH RANEY COBALT CATALYST Fig. 1.-Adsorption of long-chain fatty acids on Raney as a function of the concentration of the acid in BY ANDREW J. CHADWELL, JR., A N D HILTONA. SMITH cobalt benzene, 25': 0,behenic acid; A, palmitic acid. Depzrtment of Chemzstry, The University of Tennessee, Knozuille, Tennessee Received March $9,1966
In a recent article' studies dealing with certain characteristics of Raney nickel catalyst were reported. Raney cobalt alloy is now commercially available, and studies dealing with similar char-
Total aluminum determinations were made by the method of Willard and Tang.4 Hydrogen content determinations and catalytic activity toward hydrogenation of benzene were determined in the same manner as for Raney nickel.'
(2) The alloy was obtained froni the Raney Catalyst Company, Chattanooga, Tennessee. (3) H. A. Smith and J. F. Fuzek. J . A m . Chem. Soc., 68, 229 (1946). ( 1 ) H. A. Smith, A. J. Chadwell, Jr., and 8.5,Kirelia, THIS JOURNAL. (4) H. H. Willard and N . K. Tang, Ind. E n y . Chem., A n a l . E d . , 69, 820 (1955). 357 (1937).
1.340
NOTES
Experimental Results and Discussion As seen from Fig. 1, fatty acids are adsorbed on Raney cobalt in a manner somewhat similar to that found for nickel, although a higher concentration is necessary to form a monolayer. The number of moles of behenic (G2) acid adsorbed per unit weight of nickel is the same within experimental error 51s that for palmitic acid (Cle). Experiments with fatty acids of shorter chain length were even more subject to errors due to formation of soluble cobalt soaps than were those with palmitic acid. The surface area of the cobalt catalyst calculated by assuming a cross section of 20.5 for the adsorbed acid is 15.6 square meters per gram, which is a little less than the value usually obtained for Raney nickel. The total aluminum content of the cobalt catalyst was found to vary between 3.3 and 2.7% depending on whether a single batch of alkali was employed or whether the aluminuin was removed from the alloy by treatment with several fresh batches of alkali a t 10-minute intervals. Raney nickel prepared under similar conditions analyzed around 11% total aluminum. Hydrogen content studies with Raney cobalt indicated that hydrogen was present in the freshly prepared catalyst, but not to the extent found in Ratiey nickel. Values obtained on several samples ranged from 40 to 70 standard milliliters of hydrogen per gram of cobalt. The hydrogen is held rather tightly by the cobalt. Pumping a t room temperature for 0.5 hour removed only about 6% of the total hydrogen. Upon flaming the decomposition bulb with a Bunsen flame, the remaining hydrogen was given off. A rapid explosion was not obtained as in the case with Raney nickel. Residual hydrogen determinations showed that 99% of the total hydrogen was removed in this manner. The hydrogenation of benzene over freshly prepared Raney cobalt was found to proceed slowly a t 120' with the rate becoming fairly rapid a t 150" or above. The rate was found to be first order in hydrogen pressure and independent of the quantity of benzene placed in the bomb. Plots of the logarithm of hydrogen pressure against time gave good straight lines with the exception of slight curvature during the first minute or two of the reaction when freshly prepared catalyst was employed. As the catalyst aged, the kinetic behavior deviated from strictly first-order behavior. The activation energy was found to be approximately 23 kcal. per mole in the range 155-180'. At lGO", = 4.65 X sec.-l for 1 1. hydrogen volume and 1 g. of catalyst. A Comparison of Raney Nickel and Raney Cobalt Catalysts.-Raney cobalt is somewhat easier to prepare than Raney nickel in that the removal of alkali by washing can be accomplished rather quickly with the cobalt while a lengthy washing procedure is required for nickel. The aluminum content is some 10-12% for nickel but only around 3% for cobalt. The surface area as measured by fatty acid adsorption is somewhat smaller than for nickel; the acid is more strongly held in the nickel, while benzene-soluble compounds are formed with the cobalt. This difEculty can be fairly well overcome
Vol. 60
by the use of very long chain acids. The hydrogen content of the cobalt is less than for nickel, but the hydrogen appears to be more strongly bound. The hydrogenation activity of the cobalt catalyst as measured by its ability t o effect the addition of hydrogen to benzene is less than that of nickel. However, the activation energy is greater for the cobalt so that the differences in activity become less as the temperature increases. The lower activity of the cobalt may well be caused by the firmer bonding of hydrogen to the cobalt. As the temperature is raised, the hydrogen becomes more readily available, and hence the activity is increased. Acknowledgment.-The authors are grateful t o the Office of Ordnance Research, United States Army, for the sponsorship of this research.
THE KINETICS OF THE DECOMPOSITION OF MALONIC ACID IN AROMATIC AMINES BY LOUISWATTS CLARK Contn'bUion from Lhe Departmenf of Chemistry, Saint Joseph CoZZepe Emmitsburg, Margland Recdved March $0, 1966
Kinetic studies have been reported on the decarboxylation of malonic acid in aqueous solution,' as well as in several non-aqueous solvents including glycerol,2 dimethyl sulfoxide,2 triethyl phosphate,3 quinoline4 and aniline.5 Apparently, no kinetic s h d y of the decomposition of malonic acid in pyridine has been reported. Preliminary experiments in this Laboratory having revealed that malonic acid is smoothly decarboxylated in pyridine solution a t temperatures near lOO", the kinetics of the reaction were carefully investigated for the sake of comparison with other aromatic amines. Since the decomposition of malonic acid in aniline proceeds quite slowly at 80 and 100' (the temperatures a t which the experiments reported6 were performed), further studies were made at higher temperatures. Results of these investigations are reported herein. Experimental Reagents.-The reagents used were (1) reagent grade malonic acid; (2) reagent grade aniline, boiling range 183186'; (3) reagent grade pyridine, boiling range 114-116'. Apparatus and Technique.-The apparatus and technique in this study were the same as those usedin studying the decomposition of malonic acid in triethyl phosphate.s The Decomposition of Malonic Acid in Pyridine and Aniline.-At the beginning of each experiment 100 ml. of the amine (saturated with dry COZ gas) was placed in the reaction flask in the thermostat oil-bath. A sample of malonic acid weighing either 0.1764 g. in the case of pyridine or 0.1810 g. in the case of aniline (the amount requlred to produce 38.0 and 39.0 ml. of COS.,respectively, on complete reaction) was placed in a thin glass capsule and introduced at the proper moment into the solvent in the usual manner. The rapidly rotating mercury sealed stirrer immediately crushed the capsule, the contents dissolved and mixed in the solvent, and reaction began. The course of the reaction was followed in the usual manner. ( 1 ) G. A. Hall, Jr.. J . Am. Chem. SOC.,71, 2691 (1949). (2) L. W. Clark, THIEJOUBNAL,60, 825 (1956). (3) L. W. Clark, {bid., 60, 1150 (1956). (4) P. E. Yankwioh and R. L. Belford, J . Am. Cham. Soc., 76, 4178 (1953). (5) Y. Ogata and R. Oda, Bull. Inst. Plus. Chsm. Rea. (Tokyo), Cham. Ed., 13, 217 (1944); C. A., 4S, 7904 (1940).
