-3,782
Ir, r-I,
PODGURSKI,
J. T. KUMMER, T. w.DRWITTAND P. H. EMMETT
4. Solubility curves have been obtained for Xa$i205 from undersaturation but they could not i)e checked from supersaturation. *j. The solubility ciirve for fi3Naz0.13Si0?. 1lH20 changes but slightly with temperature. This coinposition Imt fits the exprrimcntal data.
I CONrRIRUTION
Vol. 72
6 . The solid phases observed a t the temperatures used were : SNa20.13Si02.11H20, Na2Si206, Na,SiO,, NazSi08.Hz0, aNazSi03.6Hz0, /3Na&i03. 613s0,Na8iOr5Hz0, Na3HSi04.5H20, NasHSiOi. %II!#, NaaHSiOrHpO, ?J:tOH and NaOH.H20. ~'HILAI)I.)LPIIIA ti,
PA.
RECEIVED APRIL16, 1950
FROM THE bfEI,LOU INSTITUTE AYD THE UNIVERSITY O F PITTSBURGH]
Preparation, Stability and Adsorptive Properties of the Carbides of Iron' 1 %H.~ H. PODGURSKI, J. T. KUMMER, T. W. DEWITTAND P. H. EMMETT furnace control similar to that described by Benedict.3 A thermocouple in a well down the center of the catalyst tube enabled one to detect any temperature rise in the catalyst during carbiding or adsorption ruqs. Liquid nitrogen, Dry Ice-alcohol, and ordinary ice were used for the -195, -78.5 and 0 " baths, respectively. The bath a t -46" was maintained by keeping chlorobenzene a t its melting point. Carbon monoxide was prepared from formic acid and phosphoric acid. Just before use, it was passed over hot copper a t 300" to remove traces of oxygen; through soda-lime to remove carbon dioxide; and through a Dry Ice trap to remove water vapor. Helium and tank nitrogen were purified by standard procedures. Hydrocarbons (Phillips hydrocarbon gases, pure grade) were employed for some of the carbiding in place of carbon monoxide. The hydrocarbons (usually either propane, butane or pentane) were further purified by being passed over hot reduced copper, through a tube containing ascarite and finally through a tube containing freshly-activated alumina. The pentane was carried over the iron catalyst n ith pure dry helium. The iron catalysts used for preparing the adsorbents in the present work were fused synthetic ammonia-type catalysts. Except for catalyst 910, which was free of added components, promoted catalysts were used in this investigation. Most of the work was done on promoted catalyst 123 which, in addition to oxygen, contained 72.26% Fe, 2.26% Al2Oa, 0.6270 Si02 and 0.21% Z r 0 2 . The other catalysts employed were 435, with 72.2% Fe, 0.8% A1203 and 0.25% K20; and 422, with 73.26% Fe, 1 .st?% A1203and 0.58% ZrO2. These fused oxide catalysts were initially reduced with hydrogen a t space velocities near 1000 and a t successively higher temperatures starting at 3.50 and extending up to 500" for the promoted C d talysts . The samples carbided with carbon monoxide were preThe apparatus employed both for carbiding runs and for pared by passing carbon monoxide over the reduced adsorption measurements was an adaptation of a standard adsorption apparatus such as has been used in surface catalysts a t temperatures starting a t 200 and increasing iron under area measurements by low-temperature adsorption iso- gradually to 276". As is well known,1fl,11.12 therms and has been described on numerous occasio~is.~~the5e conditions will form a carbide that can be identified Surface area measurements with nitrogen a t - 195" by its X-ray structure and that corresponds approximately were made on all samples of carbides m d mcfals u w l 111 in composition to Fe2C. In the present work, the perccntage carbiding was determined in two ways. During the present work. The catalyst tube was held a t a constant temperature the cnrhiding process a sample of the exit gas was analyzed for carbon dioxide a t various time intervals. An intega(to within h0.25') in the range 100 to 300° by using ii tion of this time V S . carbon dioxide curve gave values for (1) Joint contribution from the Gulf Research & Development the total carbon dioxide yield from which the extent of Company's Multiple Fellowship, Mellori Institute and t h r I rnivwcarbiding was calculated by assuming that the carbon sity of Pittsburgh, Pittsburgh, Pa. monoxide reacted with the iron t o form Fe2C and carbon (2) Fischer and Tropsch BvennstofChcm , 7 , 97 (1920j dioxide. The amount of carbide formed was checked by (3) (a) Kummer, DeWitt and Emmett, THISJOURNAL 70, 3632 noting the increase in the weight of the sample. The two (1948); (b) Browning, Kiimmer and Emmett, .I Chcin P h y s , methods always agreed t o within about f0.170 carbide 16, 739 (1948). carbon when correction was made for a trace of oxygen (4) Weller. Hofer and Andersoo, THISJOURNAL, 70, 799 (1948).
Introduction The possible importance of metallic carbides in the synthesis of hydrocarbons from hydrogen and carbon monoxide over catalytic Co, Ni or Fe was first suggested by Fischer and Tropsch.2 They proposed that the carbides might be intermediates in the Fischer-Tropsch process by being formed by the reaction of carbon monoxide with the metals and reduced by hydrogen to form hydrocarbons. In recent years i t seems to have been well established that for i r ~ n and ~ ~prob, ~ ably for cobalt48jthe bulk carbides are not intermediates. Nevertheless, the carbides are known to be formed during synthesis. It seemed well, therefore, to obtain some information relative to the adsorptive properties of the carbides toward carbon monoxide and hydrogen with a view to assessing more accurately their part, if any, in the Fischer-Tropsch synthesis. The present paper reports such adsorption studies for the various iron carbides and also gives additional information relative to the preparation and properties of the iron carbides including the use of hydrocarbons as carbiding agents, the rate of carbiding as a function of surface area, and the relative stability of samples of Fe3C, Hagg F e K , and hexagonal FezC E Experimental
(5) Anderson, Hall, Krieg and Seligman, i b i d . , 71, 183 (1949) (6) Hofer, Cohn a i d Peebles, ibzd.. 71, 18R (1949). (7) Emmett, A .S T Lf 4 1 (1941) ( 8 ) Emmett, '4dvances in Colloid Science," Iuterwience P u h Itshers, N e w Vork V Y l f i l l pp 1 'W ~
f'4) Renedict, Kcs Scr Insfvumcnls, 8, 252 (1937); 18, 36 (1941), 13, 21 ( l w 2 ) 1 0 11.itir a n d Irs,
