SEPTEMBER, 1940
INDUSTRIAL AND ENGINEERING CHEMISTRY
corresponds to that which would be shown by the decomposition of NiHz to leave nickel and drive off hydrogen. The particle size is about 167 mesh. It is pyrophoric a t normal temperatures. Figure 4 shows that catalysts having different physical characteristics may have practically equal catalytic characteristics. Duplicate samples of cottonseed oil were hydrogenated. I n one sample 0.5 per cent of nickel by weight of reduced nickel was used, in the other, 0.5 per cent of nickel b y weight of Raney nickel. The reduced nickel mas finer than 325 mesh. The Raney nickel was about 167 mesh. The reduced nickel was not pyrophoric, the Raney nickel was. The maximum hydrogenating temperature was 385390 O F., and the pressure 15 pounds per square inch. The graphs show that the saturation of the oil proceeded a t practically the same rate. Characteristics of the hydrogenated oil a t critical points on the curves were substantially the same.
Uses Nickel catalysts from alloys usually referred to as Raney nickel find use in the many reactions for which nickel is a catalyst. Adkins, in his comprehensive work with copper-
1203
chromium oxide and nickel catalysts ( I ) , gives numerous examples of the adaptability, selectivity, and general use of Raney nickel. Recent patents describe its utilization in the hydrogenation of rosin ( 3 ) and in the hydrogenation of aliphatic nitro compounds (8). Its use in the hydrogenation of indene-coumarone resins was recently described (4). An important use is the synthesis of sex hormones (IO).
Literature Cited (1)
Adkins, Homer, “Reactions of Hydrogen”, Madison, Univ.
Wis., 1937. (2) Bradley and Taylor, Proc. Roy.SOC.(London), 159, 56 (1937). (3) Byrkit, R. J., U. S.Patent 2,174,651 (Oct. 3, 1939). (4) Carmody, W. H., Kelly, H. E., and Sheehan, William, IND. ENG.CHEM.,32, 684-92 (1940). (5) Covert and Adkins, J . A m . Chem. Soc., 54, 4116-17 (1932). (6) Fraser, 0. B. J., Trans. Electrochem. Soc., 71, 425-86 (1937). (7) Gwyer, 2. anorg. Chem., 57, 113 (1908). U. S.Patents 2,139,122-4 (8) Ham, H . B., and Vanderbilt, B. M,, (Dee. 6, 1938); Johnson, Kenneth, Ibid., 2,157,386 (May 9 , 1939); Vanderbilt, B. M., Ibid., 2,157,391 (May 9, 1939). (9) Raney, Murray, Ibid., 1,563,587 (Dec. 1 . 1925); 1,628,190 ( M a y 10, 1927); 1,915,473 (June 27, 1933). (10) Ruzicka, L., and Goldberg, M. W., Helv.Chim. Acta, 19, 1407-10 (1936).
HYDROGENATION OF PETROLEUM The present status of high-pressure hydrogenation of petroleum is discussed as to theory and commercial application. Particular reference is made to the production of aviation gasoline, motor fuel, aviation blending agents, and Diesel fuel, and the conversion of heavy asphaltic fractions. Yields and inspection data on these operations are included. Economics are discussed, and it is shown that for a 30 cent spread between crude and fuel oil price, hydrogenation in competition with thermal cracking for production of motor fuel will show a 20 per cent return on the added investment for hydrogenation.
F
ROM the time of Sabatier until the middle of the third decade of this century, commercial hydrogenation was practiced to a limited extent under very restricted conditions. Effective catalysts, composed mainly of reduced nickel, were available, but impurities in the feed or hydrogen, such as sulfur and arsenic, tended to deactivate them rapidly. Since coal and oil usually contain such substances, these limitations made it impossible to hydrogenate carbonaceous materials of mineral origin. The industrial application of hydrogenation was therefore limited for many years to the treatment of fats and oils of animal and vegetable origin. Early attempts to convert coal into liquid hydrocarbons by subjecting it to high hydrogen pressures were not too satisfactory. However, when the research organization of the I. G. Farbenindustrie discovered sulfur-resistant catalysts, hydrogenation of carbonaceous materials of mineral origin assumed commercial possibilities, although the use of pressures of the order of 3000 to 4000 pounds per square inch was
E. V. MURPHREE Standard Oil Development Company, New York, N. Y
C. L. BROWN Esso Laboratories, Standard Oil Company of Louisiana,
Baton Rouge, La.
E. J. GOHR Esso Laboratories, Standard Oil Development Company, Elizabeth, N. J.
required. I n 1927 the Standard Oil Development Company joined in the further development of the hydrogenation process. As a result of this work the hydrogenation process has been adapted to the following applications : High-octane-number aviation gasoline production from kerosene and gas oil fractions. Motor gasoline production from gas oils. Aviation blending agent production by saturation of branchedchain polymers. Production of high-grade Diesel fuels from low-quality gas oils. Production of water-white paraffinic kerosenes from inferiorquality distillates. Production of high-viscosity-index lubricating oils from poorquality lubricating distillates. Refining (or “hydrofining”) of gasolines to low sulfur content and high stability. Conversion of asphaltic crudes and refinery residues into lower boiling gas oils of increased paraffinicity. Preparation of low-aniline-point high-solvency naphthas. Preparation of high-flash high-octane-number safety fuels. The theoretical background of the above adaptations is discussed briefly, and the commercial applications of the more important are presented in some detail. At present, commercial hydrogenation for lubricating oils, solvents, and kerosenes has been superseded by solvent extrsction methods. No commercial application of hydrofining has been made. These
1204
INDUSTRIAL AND ENGINEERING CHEMISTRY
VOL. 32, NO. 9'
GENERAL VIEW OF BATON ROUGEHYDROGENATION PLANT Hydrogenation production units a t right, compressor house a t left center, pump house and reactor stalls at left.
adaptations of hydrogenation, therefore, will not be discussed further in this paper. Technical information covering these methods has been presented in previous papers (3, 5, 7, 10).
Theory of Hydrogenation of Petroleum Derivatives
bearing on simple hydrogen addition to olefinic compounds are well covered in existing data. The data of Parks and Todd (9) give the following standard free-energy change for the hydrogenation of gaseous diisobutylene t o gaseous isooctane:
AFO = - 28,570 f 27.6T Two general types of hydrogenation are found in these processes-namely, simple and destructive hydrogenation. Simple Increasing temperature favors less hydrogenated product; hydrogenation involves merely the addition of hydrogen to a however, a t low temperatures (0-150' C.) the rate of reaction carbon-carbon double bond; hydrogenation of olefinic polymers is low except with extremely active and very sensitive catalysts. Hydrogenation over a nickel catalyst (employing reis the most clear-cut example. Destructive hydrogenation generation of the catalyst) may be considered as a combination of simple hydrogenation - . is successfully conducted commercially a t temperatures of 180and other catalytic reactions, such 190" C. and pressures in the range as cracking, dehydrogenation, ring of 1 to 4 atmospheres (4). Howclosure, polymerization, decomposiever, nickel catalysts are sensitive tion to the elements, etc. Destructo sulfur poisoning, and their cative hydrogenation is exemplified in pacity between regenerations for the conversion of gas oils and heavy hydrogenating diisobutylene i s residues to light distillates. limited by the sulfur content of t h e The course of the hydrogenation feed stock (4). Catalysts resistant operation is influenced by choice of to sulfur poisoning require higher catalyst, total pressure, hydrogen temperatures for the realization of partial pressure, temperature, feed economic rates of hydrogenation, rate, and type of feed stock. The and higher pressures are required operating variables are, of course, to assist in repressing competing reinterrelated and are in turn influa actions. enced by the type of catalyst and m Equilibrium relations derived feed stock employed and the char4 3 .5 from the above equation for freeacter and extent of reaction de8 energy change in the reaction are sired. Hydrogen partial pressure t shown graphically in Figure 1, reinfluences the rate and extent of U w lating the temperature t o the hydroconversion and also affects catalyst 3 gen partial pressure required to life. Lower hydrogen pressures I obtain saturation of diisobutylene tend to increase the rate of catat o the extent of 99.0 and 99.95 per lyst degradation. The choice of .os cent. However, a competitive o p e r a t i n g temperature depends series of reactions involving depolyupon the type and age of the catamerization of the diisobutylene and lyst, the nature of the reaction subsequent hydrogenation of the promoted, and the feed rate which resulting isobutylene can occur. will give the desired conversion. The standard free-energy change (9) 5 .01 For a given operation the temperafor depolymerization of gaseous diture is chosen so as to permit operaisobutylene to gaseous isobutylene tion a t a feed rate of reasonable .005 250 WO 550 400 450 500 550 is magnitude. The type of feed stock TEMPERATURE OC. influences the composition of the AFO = 17,920 - 42.2T FIGURE 1. EQUILIBRICM RELATION OF TEMhydrogenated product under given PERATURE, HYDROGEN PARTIAL PRESSURE, operating conditions. AND PER CENTOLEFINSIN PRODUCT FOR THE and calculated equilibrium relaSIMPLE HYDROQENATION. REACTION, tionships are shown in Figure 2. T h e t h e r m o d y n a m i c relations CEHM(gas) HZF! CEHIS(gas)
+
SEPTEMBER, 1940
INDUSTRIAL AND ENGINEERING CHEhlISTRY
The subsequent reaction of hydrogenation of gaseous isobutylene to gaseous isobutane is AFO = -28,140
+ 30.8T
According to Figure 2, depolymerization can be appreciable as low as 200' C. If depolymerization takes place, the subsequent reaction involving hydrogenation of the isobutylene produced favors further depolymerization in view of the removal of isobutylene as isobutane. The relative yields of iso.octane and isobutane are, however, governed by the relative rates of hydrogenation and depolymerization, and fortunately several catalysts of a sulfur-resistant character are available which give practically theoretical yields of isooctane and megligible decomposition to isobutylene.
1203
somewhat, however, and an examination of some of them with regard to the preferred reactions and means for obtaining them is instructive. Thus, motor gasolines and aviation fuels prepared by moderate temperature destructive hydrogenation contain negligible amounts of olefins; these distillates, in order that they may possess satisfactory octane numbers, should therefore contain large amounts of isoparaffins and
FIGCRE 3. FLOWSHEETOF HYDROGEN PRODUCTION
TEMPERATURE OC.
FIGURE 2. EQUILIBRIUM RELATION OF TEMPERATURE AND PRESSURE FOR THE REACTION, Iso-CsHle (gas) e 2 Iso-C4Hs(gas)
DESTRUCTIVE HYDROGENATION. The problem of relative rates of hydrogen addition and cracking is important also in destructive hydrogenation. Sweeney and Voorhies (18) made an analysis from a thermodynamic standpoint of the destructive hydrogenation process by studying the analogous reactions:
+ CiiHia + 3Hz
+ CaHs + CsHio
CioHs 4H2 = CEHS C ~ H I O naphthalene benzene butane anthracene
benzene m-xylene
It was shown that at temperatures between 400' and 538" C. both reactions tend to proceed with a rupture of the polycyclic rings when the hydrogen partial pressure is sufficiently high (10 to 50 atmospheres). It was inferred that naphthalene, anthracene, or similar condensed aromatics could not be formed from benzene and butane, benzene and m-xylene, or other similar compounds in the presence of adequate hydrogen partial pressure. Since polynuclear aromatic compounds are typical components found in cracked tars and gas oils, this study served to provide a thermodynamic reason for the absence of these materials in destructively hydrogenated petroleum fractions. Of the applications of hydrogenation listed above, all but the hydrogenation of polymers are examples of destructive hydrogenation, carried out, in general, roughly a t 200 atmospheres pressure. The objectives of each application differ
cyclic compounds. The production of these desirable highoctane constituents can be accomplished by the choice of catalysts. Selective catalysts may be applied which have relatively low hydrogenating capacity, and which crack polycyclic compounds boiling beyond the gasoline range to substituted monocyclic compounds within the gasoline range; hydrogenation in this case is limited to saturating the paraffinic side chains of the cyclic compound and the molecular fragments which result. Nonselective catalysts tend to hydrogenate the cyclic feed stock strongly, if aromatic in type, and simultaneously crack and hydrogenate the lower boiling products to acyclic compounds. The relative activities of selective and nonselective catalysts in the hydrogenation of benzene and the difference in their action in the preparation of hydrogenated gasoline are shown in Table I. Catalyst A (selective)
TABLEI. COMPARISON OF SELECTIVE AND SONSELECTIVE HYDROGENATION CATALYSTS UNDER SIMILAR CONDITIONS A,
Selective
E,
Nonselective
Hydrogenation of Benzene a t Conditions of Increasing Temp. Vol. yocyclohexane in hydro product Temp. condition 1 1 100 2 4 100 3 21 100 4 28 100 Hydrogenation of Cracked, Light Cycle Gas Oil Motor gasoline Vol. 7 0 67 72 A. S. T. RI. octane No. 77.5 70.5 Hydro gas oil 40 40 Vol. 70 Aniline point, F. (" C.) 90(32) 138(59)
gave only 28 per cent cyclohexane under conditions where catalyst B (nonselective) gave 100 per cent. Although they have about the same activity in converting gas oils to gasoline, the products from catalyst A are less paraffinic than from catalyst B, the gasoline having higher octane number and the gas oil lower aniline point. High-quality Diesel fuels, on the other hand, are usually prepared in the presence of a nonselective catalyst. Here the
INDU81‘11IAL AND E N G I N E 1 . ~ l l I N G j C H E M I S ~ l i ~
VOI.. :32, NO. 9
is given in the following discussion. Typical views from tlie Baton Rouge hydrogenation plant a.re included. Hydrogen is generated i n a cootinuous taostage process as sliown in Figure 3. 111 the first or re-forming stage, natural gas and steam reaci as follows, CHI
+ Ha0 = CO + 3Hs
using a catalyst. heated to liigii temperatnrc. The gaseous products from the re-iorrning furnace are mixed with additional steam and ~ R S Sthrough anotlier cata.lytic mass held at a lower temperature. In this converter step the s t e m and carbon rrionoxide react. to furni carbon dioxide ai111 additional hydrogen according to t h e reaction,
