United States Patent (19) 11 3,968,133 Knifton (45) July 6, 1976 54 CARBOXYLATION PROCESS FOR 3,723,486 3/1973 Kajimoto...................... 260/410.9 R PREPARING LINEAR FATTY ACDS OR 3,776,929 12/1973 Mrowca....................... 260/410.9 R ESTERS 3,819,669 6/1974 Knifton........................ 260/410.9 R 3,832,391 8/1974 Parshall........................ 260/410.9 R (75) Inventor: John F. Knifton, Poughouag, N.Y. Primary Examiner-Patrick P. Garvin 73 Assignee: Texaco Inc., New York, N.Y. Assistant Examiner-John F. Niebling 22) Filed: Nov. 25, 1974 Attorney, Agent, or Firm-T. H. Whaley; C. G. Ries; (21) Appl. No.: 526,867 Bernard Marlowe 52) U.S. C. ......................... 260/410.9 R; 260/413; 57 ABSTRACT 260/497 A; 260/488 K; 260/491; 260/.532; This invention concerns a process for preparing linear 260/533 A fatty (carboxylic) acids and esters from the reaction of 51 Int. Cl’............................................ C11C 3/02 olefins, alcohols (or water) and carbon monoxide. The (58) Field of Search......... 260/410.9 R, 413, 497 A, process is catalyzed by catalytic amounts of disper 260/488 K, 491, 532,533 A sions of ligand-stabilized, palladium(II) halide com plexes, in quaternary ammonium, phosphonium or ar 56 References Cited sonium salts of trihalostannate(II) or trihaloger UNITED STATES PATENTS manate(II). These empirically selected palladium cata 3,437,676 4/1969 Kutepow...................... 260/410.9 R lysts can be recycled with fresh olefin feed several 3,530, 155 9/1970 Fenton......................... 260/410.9 R times before substantial loss of catalytic activity is 3,530,168 9/1970 Biale...... ... 260/410.9 R encountered. 3,657,368 4, 1972 Parshall...... ... 260/410.9 R 6 Claims, No Drawings 3,700,706 10/1972. Butter.......................... 260/410.9 R 3,968, 133 2 While these homogeneous catalysts are a vast im CARBOXYLATION PROCESS FOR PREPARING provement over the more easily poisoned and less se LNEAR FATTY ACDS OR ESTERS lective heterogeneous catalysts, and the widely used highly toxic metal carbonyls such as nickel, cobalt and STATEMENT OF INVENTION 5 iron cabonyls. They also have certain drawbacks. This invention relates to the catalytic conversion of These include difficulties in maintaining high conver alpha-olefins to linear fatty acids or esters. sions, high selectivities and high yields upon recycling More particularly, this invention concerns the car the catalyst, due to catalyst degradation, and the prob boxylation of alpha-olefins with carbon monoxide in lems of mechanical losses, and further catalysts decom the presence of hydroxyl-containing co-reactant, under 10 position, during the separation of the carboxylated moderate conditions of temperature and pressure, products from the homogeneous catalysts and solvents. using dispersions of certain ligand-stabilized palladium Catalyst instability is a particular problem with the less (II) halides in quaternary ammonium, phosphonium thermally stable palladium-containing homogeneous and/or arsonium salts of trihalostannate (II) or trihalo catalysts. germanate(II). 15 Carboxylation, as used througout this disclosure and CLOSEST PRIOR ART AND DISTINCTIONS OF claims, refers to the process of preparing saturated, THIS INVENTION OVER SAID ART preferably linear carboxylic acids or their esters from The closest prior art that the applicant is aware of is alpha-olefin substrates. Linear paraffinic carboyxlic U.S. Pat. No. 3,657,368 (G. W. Parshall). This patent esters are of the type: RCOOR", wherein R is a satu 20 describes the use of certain transition metal halides in rated alkyl radical containing 3 to 40 carbon atoms, combination with molten quaternary ammonium or wherein R' is an aliphatic or aromatic radical. phosphonium trihalostannate(II) and trihaloger BACKGROUND OF THE INVENTION manate(II) salts as catalysts for the carboxylation, hy Carboxylic acids are characterized by the presence of ? droformulation, hydrogenation and isomerization of one of more carboxyl groups in an organic molecule. olefins, and the hydrogenation of nitriles. The acids are usually written as RCOOH. The hydro The following differences of substance are believed gen atom of this group may be displaced and appear as to patentably distinguish applicant's claimed invention a hydronium ion thereby justifying by theory the term from the disclosure of Parshall. 1. While Parshall osten acid. 30 sibly discloses platinum and palladium halides dis Saturated linear carboxylic acids and their esters may persed in molten quaternary ammonium or phospho be prepared by a large number of general procedures nium trihalostannate(II) and trihalogermanate(II) salts such as the oxidation of primary alcohols or aldehdyes, for the catalytic carboxylation of alpha-olefins, no pal the catalytic hydrolysis of nitriles, the reaction of ladium-containing catalysts are specifically named or Grignard reagents with carbon dioxide as well as by 35 described by Parshall, much less employed. several special procedures for specific acids including The sole “disclosure' can be found in Column 2, fermentation, the acetoacetic ester synthesis, the ma lines 21-32, more specifically, lines 22-25, ". (A) lonic ester synthesis and the Reformatsky reaction. at least 0.05 weight percent of a chloride, bromide or In recent years, with the availability of large quanti iodide-containing salt of a metal having an atomic num ties of alpha-olefins from wax cracking at relatively low 40 ber of 22–28, 40-46 or 72-79, and (B) . ' costs, alpha-olefins have been considered as starting The following reasons are evidenciary of why the materials for fatty acid production. Parshall patent does not constitute a bona fide disclo The carboxylation of olefins in the presence of metal sure of the use of the analogous palladium catalysts carbonyls or carbonyl precursors to produce carboxylic even to those skilled in the art: acids or esters is old in the literature, originally having 45 1. While it is known that palladium has an atomic been developed by Reppe and his coworkers. How number of 46 and Column 2, lines 21 to 32 states that ever, the nickel and iron salts or carbonyl precursors metal halides of 40-46 are useful as catalytic entities, suffer from some major drawbacks. These drawbacks this disclosure is sufficient to characterize the specific include the high toxicity of the carbonyl type reagents, components of active palladium containing catalysts in the production of a variety of undesirable side products 50 order to reject applicant's specifically claimed catalysts due to polymerization, isomerization and reduction of on Parshall. the olefin substrates and most importantly, the reaction 2. A further reason why this type of "tour de force' results in the production of large quantities of branched disclosure is a meaningless teaching to reject appli isomers in addition to the desired linear fatty acid prod cant's claims is that it is well documented in the patent uct. 55 literature that catalysis is a most unpredictable art and *This work is reviewed by C. W. Bird, Rev.62,283 (1962) does not lend itself to broad generic disclosures sup Recently, it has been shown that the production of ported by only a few isolated species, particularly when polymeric, isomeric and reduced products can be such broad disclosures can encompass many catalyst avoided through the use of alternative homogeneous species which are not available for testing or are even catalyst systems which are active under mild reaction 60 unknown. For instance, virtually every component of a conditions and which give good yields and selectivity to multi-component homogeneous platinum(II) or pal the desired linear fatty acids or esters. These catalyst ladium(II) catalyst system, such as the noble metal systems consist of ligand-stabilized platinum(II) and employed, the nature of the halide salt, the presence of palladium(II) halide complexes in combination with absence of stabilizing ligands, the structure of the mol Group IVB metal halides. They are exemplified by: 65 ten quaternary ammonium or phosphonium trihlaos PtCl(ASCCH)-SnCl2 (U.S. Pat. No. 3,819,669) tannate(II) or trihalogermanate(II) salts present can and each affect not only the degree of conversion, selectiv PdCl(P(CH)l-SnCl (U.S. Pat. No. 3,700,706). ity and yield but, in some cases, even the operability. 3,968,133 3 4. These statements are confirmed by the data in Tables 1 g. 10 (C2H5)4NSnCla-PodCl2P(p-CHO.CH4)32 and 2. h. 10 (CH)NSnCl-PdCl2P(CH3)2C6H5) In re Grant, 134U.S.P.O.248, among others. i. 5 (CH)NIGeCl-PdCl2P(C6H5)2 3. Many of the catalyst compositions disclosed gener j. 5 (CH)N(GeCla-PdCl2(AsOCH5)3)2 ally in U.S. Pat. No. 3,657,368 (See Column 2, lines 5 k. 10 (CH)N(SnCla-PodCl2(P(C6H5)3)2 21-32) and specifically (See Column 2, lines 38-45) 1. 10 (CH)NSnCla)-PdCl2(AsOCH5)3)2 for the carbonylation of olefins have been found inac Of the 12 catalyst compositions described supra, the tive by us for the carboxylation of a-olefins, as exempli palladium calalyst compositions (a) through (j) (inclu fied by 1-octene, to linear carboxylic acid esters, such sive) have been found to show superior performances as vinyl nonanoate under experimental conditions 10 compared to analogous platinum catalysts disclosed by where the preferred ligand-stabilized palladium(II) Parshall for the carboxylation of a-olefins. Compara catalysts, such as (C2H5)4NSnCl-PdCl2(P(C6H5)3)2 tive data are summarized in Tables 1 and 2 for the show good activity. This is confirmed by the data in typical synthesis of ethyl nonanoate from 1-octene Table 2. under standard experimental conditions. 4. When the closest platinum-containing catalyst 15 More generally it is anticipated that the palladium disclosed by Parshall for the carboxylation of alpha catalysts of this invention should consist of ligand olefins; stabilized palladium(II) halide complexes dispersed in 10 (CH)NSnCl)-PtCl(P(CH)) quaternary ammonium, phosphonium and arsonium is run side by side with the analogous palladium catalyst salts of trihalostannate(II) and trihalogermanate(II) utilized by applicant, viz.