Other Catalysts

Other complexes also react with propagating radicals by catalytic chain transfer. These include certain chromium,molybdenumand iron complexes. To date the complexes described appear substantially less active than the cobaloximes and are more prone to side reactions. 

Catalytic chain transfer has now been applied under a wide range of reaction conditions (solution, bulk, emulsion, suspension) and solvents (methanol, butan-2-one, water). The selection of the particular complex, the initiator, the solvent and the reaction conditions can be critical. For example  

Catalysts suitable specifically for reduction of carbon-oxygen bonds are based on oxides of copper, zinc and chromium Adkins catalysts). The so-called copper chromite (which is not necessarily a stoichiometric compound) is prepared by thermal decomposition of ammonium chromate and copper nitrate [50]. Its activity and stability is improved if barium nitrate is added before the thermal decomposition [57]. Similarly prepared zinc chromite is suitable for reductions of unsaturated acids and esters to unsaturated alcohols [52]. These catalysts are used specifically for reduction of carbonyl- and carboxyl-containing compounds to alcohols. Aldehydes and ketones are reduced at 150-200° and 100-150 atm, whereas esters and acids require temperatures up to 300° and pressures up to 350 atm. Because such conditions require special equipment and because all reductions achievable with copper chromite catalysts can be accomplished by hydrides and complex hydrides the use of Adkins catalyst in the laboratory is very limited. 

The same is true of rhenium catalysts rhenium heptoxide [42], rhenium heptasulfide [5i] and rhenimn heptaselenide [54] all require temperatures of 100-300° and pressures of 100-300 atm. Rhenium heptasulfide is not sensitive to sulfur, and is more active than molybdenum and cobalt sulfides in hydrogenating oxygen-containing functions [55,55]. 

Other Catalysts.—Copper(ii) alkoxides can be carbonylated, giving dialkyl carbonates, by insertion of carbon monoxide into the copper-alkoxide-oxygen bond. The first well-characterized copper(i)-carbonyl compounds are [CuCl(CO)(en)] and [(en)Cu(/i-CO)aCu(en)]Cl2. The effects of pH on rates of hydration of acetylene have been compared for the three catalysts CugSOi-CuSO, HgSOi, and PdS04-Fe2(S04)3. Alkyl and aryl halides can be carbonylated in the presence of ethanol, antimony pentachloride, and liquid sulphur dioxide  

In addition to the binary catalysts from transition metal compounds and metal alkyls there 2ire an increasing number which are clearly of the same general type but which have very different structures. Several of these are crystalline in character, and have been subjected to an activation process which gives rise to lattice defects and catalytic activity. Thus, nickel and cobalt chlorides, which untreated are not catalysts, lose chlorine on irradiation and become active for the polymerization of butadiene to high cis 1,4-polymer [59]. Titanium dichloride, likewise not a catalyst, is transformed into an active catalyst (the activity of which is proportional to the Ti content) for the polymerization of ethylene [60]. In these the active sites evidently react with monomer to form organo-transition metal compounds which coordinate further monomer and initiate polymerization. 

Soluble single species catalysts are also known, such as the bis(7r-allyl nickel halides) [7]. These can be prepared separately or in situ by reacting bis-allyl nickel (which is an ologomerization catalyst for butadiene but does not give high molecular weight polymer) with an equimolar quantity of nickel halide, and thus bears some resemblance to the catalysts from titanium subhalides and alkyl titanium halides. It is of interest to note that the active species is the monomeric form of the initiator as TT-complex with butadiene (XII) [61]. 

4- and trans 1,4-structure are obtained. In the presence of triphenyl-phosphite or alcohols a different complex is produced which produces high trans 1,4-poly butadiene. 

Combinations of TT-allyl nickel chloride and bis(Tr-allyl) nickel produce macrocyclic oligomers from butadiene containing from four to above eight monomer units [138] in the absence of the chloride dimers and trimers only are produced. This is of interest in that termination occurs by ring closure rather than by hydrogen transfer. 

Tris-allyl chromium gives poly 1,2-butadiene [62] and the active entity involves two chromium atoms [63], Propagation presumably involves monomer addition to the allyl grouping without undue disturbance of the electronic distribution around the metal but there is no information on the manner of monomer coordination or of the activation process which results in polymerization. 

Only a few reports of the use of other metal catalytic systems for hydrogen isotope exchange labeling have appeared studies have generally been done with deuterium oxide and their potential for use with tritiated water of significant specific activity appears very limited. 

Nafion , a perfluoroalkyl polymer resin containing sulfonic acid groups, has been compared with trifluoromethanesulfonic acid for its regioselectivity in the labeling of model quinolines, pyridines, anilines, toluene and chlorobenzene the reaction mixtures contained substrate, resin and a small volume of tritiated water. Reactions were conducted at 90-180 °C for up to several days and gave product specific activities only a fraction of those of the tritiated water batches. Generally, both catalysts label positions subject to electrophilic attack, but Nafion behaved as a weaker acid than its homogeneous counterpart. 

Based on studies to date, there are few advantages in using acid- or base-functionalized resins compared with conventional homogeneous acid or base catalysts. 

One report describes a study of HNaY zeolite for the catalysis of tritium exchange into simple aromatic compounds. In analogy with the use of the water-sensitive EtAlCl2 (Section 3.1.2), the anhydrous active centers of the thermally activated zeolite were exposed to small quantities of tritiated water, and subsequent heating to 175 °C with substrate induced the exchange-labeling of the latter. Success was limited to aromatic compounds without bulky substituents or electron-withdrawing substituents, and the pattern of label incorporation was that of electrophilic substitution. 

Neutral alumina was used ° to mediate the exchange-labeling of the sensitive compound by treatment with carrier-free H20 followed by mixing with a benzene solution of the substrate for 35 min at room temperature. The recovered [ H]55 was labeled (by analogy with a deuterated sample prepared the same way) exclusively at the indicated site at a specific activity of 7 Ci/mmol. 

Iron hydrocarbonyl has been repeatedly reported to be an active hydro-formylation catalyst [165, 166]. Some papers state that it is active at a lower pressure than cobalt [23, 165, 166, 168-170]. However, in a recent paper it was shown that iron hydrocarbonyl is only 1 10 times as active as Co carbonyl [824, 980]. Other authors recommend adding iron carbonyl as a promoter to Co catalysts [171, 979]. 

The electrochemical reduction of aqueous bicarbonate to formic acid, 

Ogura et a/.153 reduced C02 to methanol using the so-called Everitt s salt (K2Fe2+[Fe2+(CN)6])-modified electrode by a somewhat complicated but interesting route in the presence of a metal complex, such as Fe(II), Co(II), and Ni(II) complexes of 1-nitroso-2-naphthyl-3,6-disulfonic acid, and additional methanol  

Electrocatalytic reduction of carbon dioxide to Q-C3 hydrocarbons with less than 0.2% electrochemical yield was reported154 at pH 7 in the presence of pyrocatechol, TiCl3, and Na2Mo04 at -1.55 V versus SCE. 

As described above, many reports published to date indicate that metal complexes are promising catalysts for C02 fixation. The catalytic activity is considered basically to be due to a C02-catalyst complex formation. Thus, the complexes have to provide a binding site for C02, and this can be realized for some catalysts by losing a ligand on reduction of the catalyst at the electrode. Also, the C02 molecule is not linear but is rather a bent structure155,156 in the activated state of the C02-catalyst complexes. Theoretical calculations of C02-catalyst bonding157 and general ideas about activation of C02 by metal complexes have been summarized in several recent articles.158,159 

In addition, catalysts for C02 reduction based on nonmetallic compounds have also been reported. Taniguchi et al100 reported 

In his pioneering paper, Kurz (1963) considered examples of catalysis by protons, water, hydroxide ion, and, to a limited extent, general acids and 

Complexes of a variety of metals, including Pd, Pt, Cr, Mn, and Ir, were screened under about 1500 atm of H2/CO and are reported not to produce 

Selectivities to ethanol are relatively high for certain cobalt and ruthenium catalyst systems. In both metal systems, most of the ethanol observed is 

Despite numerous screening studies, the literature contains little evidence that homogeneous catalyst systems based on metals other than Co, Rh, or Ru have significant activity for catalytic CO reduction. As seen for the known active catalytic systems, however, properties of solvents and additives or promoters can have enormous effects on catalytic activities. Solvents and additives can serve many roles in these catalytic systems. One important function of promoters in the Rh and Ru systems appears to be that of stabilizing metal oxidation states involved in catalytic chemistry. Other 

I am indebted to many colleagues at Union Carbide Corporation, some of whose names appear in the references. Their research has contributed much to the development of this area of catalysis, and I have enjoyed their encouragement, help, and example. Of course, any errors of fact or interpretation in this review are the responsibility of the author. 

Storch, H. H., Golumbic, M. S., and Anderson, R. B., The Fischer-Tropsch and Related Syntheses. Wiley, New York, 1951. 

Oxygen exchange between [Cr(NH3)s(OH2)] + and solvent water is catalysed by carbon dioxide. The rate law and appropriate rate constants have been determined. The mechanism involves oxygen-atom scrambling through formation of an intermediate carbonate complex. -  

---

Other catalysts which may be used in the Friedel - Crafts alkylation reaction include ferric chloride, antimony pentachloride, zirconium tetrachloride, boron trifluoride, zinc chloride and hydrogen fluoride but these are generally not so effective in academic laboratories. The alkylating agents include alkyl halides, alcohols and olefines. 

Different types of other coal liquefaction processes have been also developed to convert coals to liqnid hydrocarbon fnels. These include high-temperature solvent extraction processes in which no catalyst is added. The solvent is usually a hydroaromatic hydrogen donor, whereas molecnlar hydrogen is added as a secondary source of hydrogen. Similar but catalytic liquefaction processes use zinc chloride and other catalysts, usually under forceful conditions (375-425°C, 100-200 atm). In our own research, superacidic HF-BFo-induced hydroliquefaction of coals, which involves depolymerization-ionic hydrogenation, was found to be highly effective at relatively modest temperatnres (150-170°C). 

As a catalyst for ester and amide formation from acyl chlorides or anhydrides, 4-(di-methylamino)pyridine has been recommended (DMAP G. Hdfle, 1978). In the presence of this agent highly hindered hydroxyl groups, e.g. of steroids and carbohydrates, are acylated under mild conditions, which is difficult to achieve with other catalysts. 

Dehydrogenation of isopropyl alcohol accounts for most of the acetone production not obtained from cumene. The vapor is passed over a brass, copper, or other catalyst at 400—500°C, and a yield of about 95% is achieved (1.09 unit weight of alcohol per unit of acetone) (13). 

Gyclooctatetraene (GOT). Tetramerization of acetylene to cyclooctatetraene [629-20-9], CgHg, although interesting, does not seem to have been used commercially. Nickel salts serve as catalysts. Other catalysts give ben2ene. The mechanism of this cyclotetramerhation has been studied (4). 

Catalysis is usually accompHshed through the use of tertiary amines such as triethylenediamine. Other catalysts such as 2,4,6-/m(/V,/V-dimethylaminomethyl)phenol are used in the presence of high levels of cmde MDI to promote trimerization of the isocyanate and thus form isocyanurate ring stmctures. These groups are more thermally stable than the urethane stmcture and hence are desirable for improved flammabiUty resistance (236). Some urethane content is desirable for improved physical properties such as abrasion resistance. 

A typical catalyst bed is very shallow (10 to 50 mm) (76,77). In some plants the catalyst is contained in numerous small parallel reactors in others, catalyst-bed diameters up to 1.7 and 2.0 m (77,80) and capacities of up to 135,000 t/yr per reactor are reported (78). The silver catalyst has a useful life of three to eight months and can be recovered. It is easily poisoned by traces of transition group metals and by sulfur. 

Friedel-Crafts alkylation using alkenes has important industrial appHcations. The ethylation of benzene with ethylene to ethylbenzene used in the manufacture of styrene, is one of the largest scale industrial processes. The reaction is done under the catalysis of AlCl in the presence of a proton source, ie, H2O, HCl, etc, although other catalysts have also gained significance. 

Synthetic water-spHtting membranes that contain the biochemical and other catalysts necessary to form hydrogen also are under development. 

Using only the phenyhnagnesium chloride without the MnCI catalyst results ia a mixture of products. This mixture iacludes the alcohol(s) resulting from the diaddition of the Grignard reagent to the carbonyl groups. Other catalysts, such as Fe(III) and Ni(II), have also been used to achieve similar results... 

Hydrocarbon resins based on CPD are used heavily in the adhesive and road marking industries derivatives of these resins are used in the production of printing inks. These resins may be produced catalyticaHy using typical carbocationic polymerization techniques, but the large majority of these resins are synthesized under thermal polymerization conditions. The rate constants for the Diels-Alder based dimerization of CPD to DCPD are weU known (49). The abiHty to polymerize without Lewis acid catalysis reduces the amount of aluminous water or other catalyst effluents/emissions that must be addressed from an environmental standpoint. Both thermal and catalyticaHy polymerized DCPD/CPD-based resins contain a high degree of unsaturation. Therefore, many of these resins are hydrogenated for certain appHcations. 

Commercially, polymeric MDI is trimerized duting the manufacture of rigid foam to provide improved thermal stabiUty and flammabiUty performance. Numerous catalysts are known to promote the reaction. Tertiary amines and alkaU salts of carboxyUc acids are among the most effective. The common step ia all catalyzed trimerizations is the activatioa of the C=N double boad of the isocyanate group. The example (18) highlights the alkoxide assisted formation of the cycHc dimer and the importance of the subsequent iatermediates. Similar oligomerization steps have beea described previously for other catalysts (61). 

Mesityl oxide can also be produced by the direct condensation of acetone at higher temperatures. This reaction can be operated ia the vapor phase over 2iac oxide (182), or 2iac oxide—2irconium oxide (183), or ia the Hquid phase over cation-exchange resia (184) or 2irconium phosphate (185). Other catalysts are known (186). 

The interest in 2,6-dialkylnaphthalenes such as dimethyl, diethyl, diisopropyl, dihexyl, etc, is shown by the increasing number of patents relevant to thek preparation, isomeri2ation, and separation (109—111). No method for selectively preparing 2,6-dialkylnaphthalene has to date been discovered. The efforts that have been made have each failed, not only with conventional Fridel-Crafts catalysts (106,112,113), but also with other catalysts such as... 

Organochromium Catalysts. Several commercially important catalysts utilize organ ochromium compounds. Some of them are prepared by supporting bis(triphenylsilyl)chromate on siUca or siUca-alumina in a hydrocarbon slurry followed by a treatment with alkyl aluminum compounds (41). Other catalysts are based on bis(cyclopentadienyl)chromium deposited on siUca (42). The reactions between the hydroxyl groups in siUca and the chromium compounds leave various chromium species chemically linked to the siUca surface. The productivity of supported organochromium catalysts is also high, around 8—10 kg PE/g catalyst (800—1000 kg PE/g Cr). 

A wide variety of chromium oxide and Ziegler catalysts was developed for this process (61,62). Chromium-based catalysts produce HDPE with a relatively broad MWD other catalysts provide HDPE resins with low molecular weights (high melt indexes) and resins with a narrower MWD (63,64). 

Dow catalysts have a high capabihty to copolymetize linear a-olefias with ethylene. As a result, when these catalysts are used in solution-type polymerisation reactions, they also copolymerise ethylene with polymer molecules containing vinyl double bonds at their ends. This autocopolymerisation reaction is able to produce LLDPE molecules with long-chain branches that exhibit some beneficial processing properties (1,2,38,39). Distinct from other catalyst systems, Dow catalysts can also copolymerise ethylene with styrene and hindered olefins (40). 

Another synthesis of pyrogaHol is hydrolysis of cyclohexane-l,2,3-trione-l,3-dioxime derived from cyclohexanone and sodium nitrite (16). The dehydrogenation of cyclohexane-1,2,3-triol over platinum-group metal catalysts has been reported (17) (see Platinum-GROUP metals). Other catalysts, such as nickel, rhenium, and silver, have also been claimed for this reaction (18). 

Epoxy Resins. Polysulftdes may also be cured by reaction with epoxy resins (qv) according to the reaction in equation 2. Amines or other catalysts are used and often primary or secondary amine resins are cured together with the polysulfide. 

A continuous process has been described (14) which can produce either the amide or the nitrile by adjusting the reaction conditions. Boric acid has been used as a catalyst in the amidation of fatty acid (15). Other catalysts employed include alumina (16), titanium, and 2inc alkoxides (17). The difficulty of complete reaction during synthesis has been explained by the formation of RCOOH NH RCOO , a stable intermediate acid ammonium salt (18). 

Other catalysts that can be used are boron trifluoride (5), copper—chromium oxides (6), phosphoric acid (7), and siUca-alurnina (8). Under similar conditions, ethanol yields /V-ethylaniline [103-69-5] and /V,/V-diethylaniline [91-66-7] (9,10). 

Diphenylamine can also be produced by passing the vapors of aniline over a catalyst such as alumina, or alumina impregnated with ammonium fluoride (17). The reaction is carried out at 480°C and about 700 kPa (7 atm). Conversion per pass, expressed as parts diphenylamine per 100 parts of reactor effluent, is low (18—22%), and the unconverted aniline must be recycled. Other catalysts disclosed for the vapor-phase process are alumina modified with boron trifluoride (18), and alumina activated with boric acid or boric anhydride (19). 

In past years, metals in dilute sulfuric acid were used to produce the nascent hydrogen reductant (42). Today, the reducing agent is hydrogen in the presence of a catalyst. Nickel, preferably Raney nickel (34), chromium or molybdenum promoted nickel (43), or supported precious metals such as platinum or palladium (35,44) on activated carbon, or the oxides of these metals (36,45), are used as catalysts. Other catalysts have been suggested such as molybdenum and platinum sulfide (46,47), or a platinum—nithenium mixture (48). 

Use of mercuric catalysts has created a serious pollution problem thereby limiting the manufacture of such acids. Other catalysts such as palladium or mthenium have been proposed (17). Nitration of anthraquinone has been studied intensively in an effort to obtain 1-nitroanthraquinone [82-34-8] suitable for the manufacture of 1-aminoanthraquinone [82-45-1]. However, the nitration proceeds so rapidly that a mixture of mono- and dinitroanthraquinone is produced. It has not been possible, economically, to separate from this mixture 1-nitroanthraquinone in a yield and purity suitable for the manufacture of 1-aminoanthraquinone. Chlorination of anthraquinone cannot be used to manufacture 1-chloroanthraquinone [82-44-0] since polychlorinated products are formed readily. Consequentiy, 1-chloroanthraquinone is manufactured by reaction of anthraquinone-l-sulfonic acid [82-49-5] with sodium chlorate and hydrochloric acid (18). 

Acylation. To achieve acylation of thiophenes, acid anhydrides with phosphoric acid, iodine, or other catalysts have been widely used. Acid chlorides with AlCl, SnCl, ZnCl2, and BF also give 2-thienylketones. AH reactions give between 0.5 and 2.0% of the 3-isomer. There has been much striving to find catalyst systems that minimize the 3-isomer content attempting to meet to customer specifications. The standard procedure for formylation is via the Vil smeier-H a ack reaction, using phosphoms o xycbl o ri de / /V, / V- dim e tb yl fo rm a m i de (POCl /DMF) or /V-m ethyl form an i1 i de. 

If the production of vinyl chloride could be reduced to a single step, such as dkect chlorine substitution for hydrogen in ethylene or oxychlorination/cracking of ethylene to vinyl chloride, a major improvement over the traditional balanced process would be realized. The Hterature is filled with a variety of catalysts and processes for single-step manufacture of vinyl chloride (136—138). None has been commercialized because of the high temperatures, corrosive environments, and insufficient reaction selectivities so far encountered. Substitution of lower cost ethane or methane for ethylene in the manufacture of vinyl chloride has also been investigated. The Lummus-Transcat process (139), for instance, proposes a molten oxychlorination catalyst at 450—500°C to react ethane with chlorine to make vinyl chloride dkecfly. However, ethane conversion and selectivity to vinyl chloride are too low (30% and less than 40%, respectively) to make this process competitive. Numerous other catalysts and processes have been patented as weU, but none has been commercialized owing to problems with temperature, corrosion, and/or product selectivity (140—144). Because of the potential payback, however, this is a very active area of research. 

Conversion of the nitrile to the amide has been achieved by both chemical and biological means. Several patents have described the use of modified Raney nickel catalysts ia this appHcation (25,26). Also, alkaH metal perborates have demonstrated their utiHty (27). Typically, the hydrolysis is conducted ia the presence of sodium hydroxide (28—31). Owiag to the fact that the rate of hydrolysis of the nitrile to the amide is fast as compared to the hydrolysis of the amide to the acid, good yields of the amide are obtained. Other catalysts such as magnesium oxide (32), ammonia (28,29,33), and manganese dioxide (34) have also been employed. 

Although supported Pd catalysts have been the most extensively studied for butadiene hydrogenation, a number of other catalysts have also been the object of research studies. Some examples are Pd film catalysts, molybdenum sulfide, metal catalysts containing Fe, Co, Ni, Ru, Rh, Os, Ir, Pt, Cu, MgO, HCo(CN) on supports, and LaCoC Perovskite. There are many others (79—85). Studies on the weU-characteri2ed Mo(II) monomer and Mo(II) dimer on siUca carrier catalysts have shown wide variations not only in catalyst performance, but also of activation energies (86). 

Dry reduced nickel catalyst protected by fat is the most common catalyst for the hydrogenation of fatty acids. The composition of this type of catalyst is about 25% nickel, 25% inert carrier, and 50% soHd fat. Manufacturers of this catalyst include Calsicat (Mallinckrodt), Harshaw (Engelhard), United Catalysts (Sud Chemie), and Unichema. Other catalysts that stiH have some place in fatty acid hydrogenation are so-called wet reduced nickel catalysts (formate catalysts), Raney nickel catalysts, and precious metal catalysts, primarily palladium on carbon. The spent nickel catalysts are usually sent to a broker who seUs them for recovery of nickel value. Spent palladium catalysts are usually returned to the catalyst suppHer for credit of palladium value. 

Lubrication AND lubricants). Optimal results are obtained at 130 5°C at a pressure of 1.5—2.0 MPa (15—20 bars) using 0.2 wt % nickel catalyst. Other catalysts and processing parameters may be used to produce unique derivatives. Simple double-bond hydrogenation at 140°C in the presence of Raney nickel catalyst produces glyceryl tris(12-hydroxystearate) [139-44-6], having a melting point of 86°C (46,47). 

An aimual review of the worldwide catalyst industry identifies current technical and business trends within the catalyst industry and fists virtually aU industrial supported (and other) catalysts by manufacturers designations (3). Included are the applications for the catalysts, the composition, ie, active agents and support materials, and some physical properties. 

An acidic solvent is recommended for use with palladium. Other catalysts that have been used for this reduction include copper chromite and any of the three Raney catalysts, cobalt, iron, or nickel. 

Gas Phase. The gas-phase methanol hydrochlorination process is used more in Europe and Japan than in the United States, though there is a considerable body of Hterature available. The process is typicaHy carried out as foHows vaporized methanol and hydrogen chloride, mixed in equimolar proportions, are preheated to 180—200°C. Reaction occurs on passage through a converter packed with 1.68—2.38 mm (8—12 mesh) alumina gel at ca 350°C. The product gas is cooled, water-scmbbed, and Hquefied. Conversions of over 95% of the methanol are commonly obtained. Garnma-alurnina has been used as a catalyst at 295—340°C to obtain 97.8% yields of methyl chloride (25). Other catalysts may be used, eg, cuprous or zinc chloride on active alumina, carbon, sHica, or pumice (26—30) sHica—aluminas (31,32) zeoHtes (33) attapulgus clay (34) or carbon (35,36). Space velocities of up to 300 h , with volumes of gas at STP per hour per volume catalyst space, are employed. 

Dichloroethane is produced by the vapor- (28) or Hquid-phase chlorination of ethylene. Most Hquid-phase processes use small amounts of ferric chloride as the catalyst. Other catalysts claimed in the patent Hterature include aluminum chloride, antimony pentachloride, and cupric chloride and an ammonium, alkaU, or alkaline-earth tetrachloroferrate (29). The chlorination is carried out at 40—50°C with 5% air or other free-radical inhibitors (30) added to prevent substitution chlorination of the product. Selectivities under these conditions are nearly stoichiometric to the desired product. The exothermic heat of reaction vapori2es the 1,2-dichloroethane product, which is purified by distillation. 

Only trace amounts of side-chain chlorinated products are formed with suitably active catalysts. It is usually desirable to remove reactive chlorides prior to fractionation in order to niinimi2e the risk of equipment corrosion. The separation of o- and -chlorotoluenes by fractionation requires a high efficiency, isomer-separation column. The small amount of y -chlorotoluene formed in the chlorination cannot be separated by fractionation and remains in the -isomer fraction. The toluene feed should be essentially free of paraffinic impurities that may produce high boiling residues that foul heat-transfer surfaces. Trace water contamination has no effect on product composition. Steel can be used as constmction material for catalyst systems containing iron. However, glass-lined equipment is usually preferred and must be used with other catalyst systems. 

Strong dehydrating agents such as phosphorous pentoxide or sulfur trioxide convert chlorosulfuric acid to its anhydride, pyrosulfuryl chloride [7791-27-7] S20 Cl2. Analogous trisulfuryl compounds have been identified in mixtures with sulfur trioxide (3,19). When boiled in the presence of mercury salts or other catalysts, chlorosulfuric acid decomposes quantitatively to sulfuryl chloride and sulfuric acid. The reverse reaction has been claimed as a preparative method (20), but it appears to proceed only under special conditions. Noncatalytic decomposition at temperatures at and above the boiling point also generates sulfuryl chloride, chlorine, sulfur dioxide, and other compounds. 

Several other catalyst systems have been suggested, including boron fluoride and both crystalline and noncrystalline siUcas and alurninosihcates. Although no commercial faciUty exists, the concept of using a crystalline siUca or alurninosihcate catalyst in an integral reaction and distillation apparatus has been proposed (9). 

C-21 dicarboxyhc acids are produced by Westvaco Corporation in Charleston, South Carolina in multimillion kg quantities. The process involves reaction of tall oil fatty acids (TOFA) (containing about 50% oleic acid and 50% hnoleic acid) with acryhc acid [79-10-7] and iodine at 220—250°C for about 2 hours (90). A yield of C-21 as high as 42% was reported. The function of the iodine is apparendy to conjugate the double bond in linoleic acid, after which the acryhc acid adds via a Diels-Alder type reaction to form the cycHc reaction product. Other catalysts have been described and include clay (91), palladium, and sulfur dioxide (92). After the reaction is complete, the unreacted oleic acid is removed by distillation, and the cmde C-21 diacid can be further purified by thin film distillation or molecular distillation. 

---