HYDROGENATION
OF NICKEL CARBONYL Nickel carbonyl was decomposed and converted into carbon dioxide, methane, and water in a dynamic pressure hydrogenation apparatus. Data, taken from 25" to 400" C. and from 1 to 75 atmospheres pressure, were treated according to the methods employed by Randall and Gerard (23),Kelley ( 9 ) , and Chipman (3). The resulting equilibrium constants obtained by this treatment showed that the reaction between water and carbon to form carbon dioxide and methane gave constant values over the whole reaction range whereas other possible reactions did not.
E. E. LITKENHOUSI AND CHARLES A. M A " University of Minnesota, Minneapolis, Minn.
N
ICKEL carbonyl was first recognized by Mond (12-20)
as an intermediate product in the carbon monoxide reduction of nickel ores. This same compound was believed by Sabatier and Senderens (%'-SO) t o exist as an intermediate product in the synthesis of methane b y the hydrogenation of both carbon monoxide and carbon dioxide over a finely divided nickel catalyst at temperatures below 400" C. Probable evidence of both the formation and decomposition of carbonyl was observed by Hightower and White (7) in the formation of a nickel mirror a t the end of their reaction has led t o the belief that chamber. Later work ( I , 6,8,3i,SS) nickel-gas complexes as well as nickel carbonyl are responsible for the catalytic effect of nickel in hydrogenation reactions. As a method of determining the hydrogenation reaction mechanism in the methane synthesis, nickel carbonyl was reduced with hydrogen over a temperature range of 25 o t o 400 o C. and a pressure range of 1 t o 7 5 atmospheres. Equilibrium constants determined over this range were compared with those based upon equations and constants given by other investigators. The principal equations are compared in Table I. The equilibrium constants are listed in terms of partial pressures of the reacting gases and also in terms of the constants of Equations 1 to 5 .
M , M were lnstalled to prevent mercury from entering the system. The hydrogen passed from the orifice through the needle valve, 0, into a T-connector, P, where it was mixed with nickel carbonyl before entering the reaction chamber. Nickel carbonyl was taken from a supply container, Q, through a needle valve, R, and T-connector X into the single-acting piston pump, T, on the backward stroke of the pump. When the piston chamber was filled with nickel carbonyl, the carbonyl was forced through a T-connector, X, and needle valve, U , to T-connector P where it was mixed with the hydrogen. The mixed nickel carbonyl and hydrogen then passed into a nickel reaction tube, A A . This hollow tube was 2 feet (61 cm.) long, 1.125 inches (2.9 cm.) 0. d., and 0.364 inch (0.925 om.) i. d. Co per cobling coils, Y , were wrapped around both ends of the niciel tube to prevent the hydrogen and nickel carbonyl from reacting before they reached the heated tube inside the resistance furnace, 2. Over both threaded ends of the reaction tube were screwed two large hexagonal nuts, BB, into which were built the iron-constantan couples, W , X , used in measuring the reaction tqmperature. These couples extended 9 inches (22.9 cm.) into the bore of the reaction tube or within 6 inches (16.2 cm.) of each other. One element of the couple was a nickel-plated Shelby seamless steel tubing, 1/4 inch (0.635 cm.) 0 . d., and inch (0.318 cm.) i. d. Through this tube was placed a ' I s inch 0. d. quartz tube to insulate the steel element from the constantan wire in the center of the quartz tubing. The steel tube was drawn around the constantan wire, and the two were welded together to make the couple. Thermocouple readings for record were made by a potentiometer, even though a millivoltmeter was also used as a check and guide during the intervals when readings were not taken. The pipe connecting hexagonal nut BB of the reaction tube with the mixing T-connector, P , was bent in the shape of a question mark to permit connection to the reaction tube, no matter where the inlet opening in the nut happened to be laced after both hexagonal nuts were drawn tight upon the nickertube. The product gases from the reaction tube were expanded through a specially designed needle valve, CC, in which any small adjustment could be made, since the needle was exceptionally long and tapered. A 1-foot (30.5-cm.) handle bqlted t o the metal handle of the valve allowed for even h e r adJustment. The reaction products were then run through Pyrex glass condensers, H H , 11, t o either a gas holder or a flowmeter. Samples of the gases were taken during th8 run and analyzed in the direct-connected gas analysis apparatus with explosion pipet LL. An aspirator bottle, N N , was lowered t o draw the gas into the glass sampler, JJ, and then raised to force the gas into the gas analysis apparatus.
Experimental Method The apparatus (Figure 1) was designed for high-temperature, high-pressure, and dynamic operation between liquid and gas reactants: Hydrogen, taken from a commercial cylinder, A , by adjusting a long-handled needle valve, B, was passed to a Bourdon pressure gage, C, two cylinders, D, E, a check valve, F, and a needle valve, G, t o a specially constructed Bourdon gage, H . Cylinders D and E acted as cushions for the hydrogen by taking up small surges in pressure. Pressure gage Is, which had a mirror arm, I, welded direct to the extreme tip of the Bourdon ring, reflected a light beam from a source of light, J, to a calibrated meter scale, K . By this means, pressure was measured within 4 pounds per square inch over the range of 1 to 75 atmospheres. The hydrogen next passed through the orifice flow manometer, L, L. This orifice consisted of a small steel disk with two female openings, one on either side, into which were forced two male blocks L, L. These blocks were held tightly compressed over the orifice disk by a threaded steel pin in a cast-iron base. By the use of various-sized orifice holes in several steel disks, the size of the orifice was rapidly changed. The difference in head on either side of the orifice was measured by electrical contact with mercury in the manometer well, N . The electrical contact points were stationary, and the height of the mercury on either side of the well was measured by forcing the mercury up to the contact points by means of a plunger attached to a graduated micrometer scale. Contact was made on either arm a t will by switching a current first through one side arm and then through the other. Thus, the difference in head could be measured directly on the one micrometer scale. A 6-volt battery supplied the current for a small pilot light mounted on the base block of the manometer well and showed when contact between the mercury and insulated contact points was made. Mercury traps 'Present address, University of T,ouisville, Louisville, I
