Catalysts, boron fluoride

As an example, an NMR spectrum of a 1,3-dioxolane-/3-propiolactone copolymer, obtained by using a boron-fluoride catalyst, is shown in Fig. 1101. The 1,3-dioxolane (DOL) homopolymer spectrum contains two singlet peaks of area 1 2 numbered 1 and 5, whereas the spectrum of the 0-propiolactone (PL) homopolymer contains two triplet peaks of area 1 1 numbered 2 and 6. Variation of initial feed ratios disclosed that peaks 1,3 and S are associated with the DOL units and that... 

The addition of boron fluoride to a nonreacting mixture of isobutylene and titanium tetrachloride at -80° resulted in a rapid polymerization of isobutylene. In other words, as has already been mentioned, the liquid phase polymerization of isobutylene with boron fluoride catalyst apparently does not require the presence of water. 

The 68 geometric isomers of the dimethyldecalins, comprising 60 dl pairs and 8 meso isomers, were previously unknown, except for one. Schneider (77) prepared hydrogen fluoride-boron fluoride catalyst. Good yields were obtained if the temperature was kept below about —40°, so that the solid isomer continuously crystallized out. 

Acetic anhydride adds to acetaldehyde in the presence of dilute acid to form ethyUdene diacetate [542-10-9], boron fluoride also catalyzes the reaction (78). Ethyfldene diacetate decomposes to the anhydride and aldehyde at temperatures of 220—268°C and initial pressures of 14.6—21.3 kPa (110—160 mm Hg) (79), or upon heating to 150°C in the presence of a zinc chloride catalyst (80). Acetone (qv) [67-64-1] has been prepared in 90% yield by heating an aqueous solution of acetaldehyde to 410°C in the presence of a catalyst (81). Active methylene groups condense acetaldehyde. The reaction of isobutfyene/715-11-7] and aqueous solutions of acetaldehyde in the presence of 1—2% sulfuric acid yields alkyl-y -dioxanes 2,4,4,6-tetramethyl-y -dioxane [5182-37-6] is produced in yields up to 90% (82). 

Boron trifluoride catalyst may be recovered by distillation, chemical reactions, or a combination of these methods. Ammonia or amines are frequently added to the spent catalyst to form stable coordination compounds that can be separated from the reaction products. Subsequent treatment with sulfuric acid releases boron trifluoride. An organic compound may be added that forms an adduct more stable than that formed by the desired product and boron trifluoride. In another procedure, a fluoride is added to the reaction products to precipitate the boron trifluoride which is then released by heating. Selective solvents may also be employed in recovery procedures (see Catalysts,regeneration). 

A. V. Topchiev, S. V. Zavgorodnii, and Y. M. Paushkiu, Boron Fluoride and Its Compounds as Catalysts in Organic Chemistry, Pergamon Press, New York, 1959. 

Aromatic Aldehydes. Carbon monoxide reacts with aromatic hydrocarbons or aryl haHdes to yield aromatic aldehydes (see Aldehydes). The reaction of equation 24 proceeds with yields of 89% when carried out at 273 K and 0.4 MPa (4 atm) using a boron trifluoride—hydrogen fluoride catalyst (72), whereas conversion of aryl haHdes to aldehydes in 84% yield by reaction with CO + H2 requires conditions of 423 K and 7 MPa (70 atm) with a homogeneous palladium catalyst (73) and also produces HCl. 

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). 

In a similar reaction of ketene with diethyl phosphite in the presence of boron fluoride etherate as catalyst, diethyl 1-acetoxyvinylphosphonate is formed [113] see Eq. (69) ... 

A ubiquitous co-catalyst is water. This can be effective in extremely small quantities, as was first shown by Evans and Meadows [18] for the polymerisation of isobutene by boron fluoride at low temperatures, although they could give no quantitative estimate of the amount of water required to co-catalyse this reaction. Later [11, 13] it was shown that in methylene dichloride solution at temperatures below about -60° a few micromoles of water are sufficient to polymerise completely some decimoles of isobutene in the presence of millimolar quantities of titanium tetrachloride. With stannic chloride at -78° the maximum reaction rate is obtained with quantities of water equivalent to that of stannic chloride [31]. As far as aluminium chloride is concerned, there is no rigorous proof that it does require a co-catalyst in order to polymerise isobutene. However, the need for a co-catalyst in isomerisations and alkylations catalysed by aluminium bromide (which is more active than the chloride) has been proved [34-37], so that there is little doubt that even the polymerisations carried out by Kennedy and Thomas with aluminium chloride (see Section 5, iii, (a)) under fairly rigorous conditions depended critically on the presence of a co-catalyst - though whether this was water, or hydrogen chloride, or some other substance, cannot be decided at present. 

The only solid acidic catalyst which has given high polymers at an appreciable rate at low temperatures, and which has been studied in some detail, is that described by Wichterle [41, 42]. This was prepared as follows A 10 per cent solution in hexane of aluminium tri-(s- or t-butoxide) was saturated with boron fluoride at room temperature, and excess boron fluoride was removed from the precipitate by pumping off about half the hexane. Two moles of boron fluoride were absorbed per atom of aluminium, and butene oligomers equivalent to two-thirds of the alkoxy groups were found in the solution the resulting solid had hardly any catalytic activity. When titanium tetrachloride was added to the suspension in hexane, an extremely active catalyst was formed but the supernatant liquid phase had no catalytic activity. The DP of the polymers formed by the catalyst prepared from the s-butoxide was much lower than that of polymers formed with a catalyst prepared from the r-butoxidc. 

Many other compounds have been shown to act as co-catalysts in various systems, and their activity is interpreted by analogous reactions [30-33]. However, the confidence with which one previously generalised this simple picture has been shaken by some extremely important papers from Eastham s group [34], These authors have studied the isomerization of cis- and Zraws-but-2-ene and of but-l-ene and the polymerization of propene and of the butenes by boron fluoride with either methanol or acetic acid as cocatalyst. Their complicated kinetic results indicate that more than one complex may be involved in the reaction mechanism, and the authors have discussed the implications of their findings in some detail. 

It might be thought that the question whether a particular alkyl halide is a co-catalyst for a particular monomer-catalyst combination could be settled easily, by adding some of the compound in question to a non-reacting mixture of monomer and catalyst. This approach has been used [36, 44], but it must be carried out in a polar solvent which is itself not a co-catalyst, or only a very weak one. The ideal solvent for this kind of work remains to be found it may be that S02 or even CS2 (which behaves like a polar solvent) will provide the answer. If one wants to use alkyl chloride solvents without being troubled by the possibility of solvent co-catalysis, boron fluoride should be used as catalyst, since the ion BF3C1 is not formed under the conditions generally used for polymerisations. 

The first study with an oxygen compound which was sufficiently rigorous to provide evidence on the question of co-catalysis was that of Farthing and Reynolds [61]. They showed that 3,3-bischloromethyl oxetan could be polymerised in methyl chloride solution by boron fluoride only in the presence of water. Tater, Rose [62] obtained kinetic evidence for the need for a co-catalyst in the system oxetan—boron fluoride—methyl chloride, and he interpreted the low reaction rate when no water was added as due to residual water he also showed that water and a hydroxyl-terminated polymer could act as co-catalysts. 

With ethylene oxide the situation is complicated. Worsfold and Eastham [63] showed that under anhydrous conditions in ethylene dichloride solution boron fluoride reacts slowly with the monomer to form a co-catalyst and that after the induction period during which this occurs the polymerization proceeds smoothly and that the induction period can be eliminated if water, equivalent to the boron fluoride, is added to the system. 

The same considerations apply to the explanation which Kern and Jaacks [65] put forward to account for their observation that the polymerization of trioxane by boron fluoride in methylene dichloride not only appears not to require a co-catalyst, but is retarded by water. They write the initiation reaction as... 

The initiation reaction in the polymerization of vinyl ethers by BF3R20 (R20 = various dialkyl ethers and tetrahydrofuran) was shown by Eley to involve an alkyl ion from the dialkyl ether, which therefore acts as a (necessary) co-catalyst [35, 67]. This initiation by an alkyl ion from a BF3-ether complex means that the alkyl vinyl ethers are so much more basic than the mono-olefins, that they can abstract alkylium ions from the boron fluoride etherate. This difference in basicity is also illustrated by the observations that triethoxonium fluoroborate, Et30+BF4", will not polymerise isobutene [68] but polymerises w-butyl vinyl ether instantaneously [69]. It was also shown [67] that in an extremely dry system boron fluoride will not catalyse the polymerization of alkyl vinyl ethers in hydrocarbons thus, an earlier suggestion that an alkyl vinyl ether might act as its own co-catalyst [30] was shown to be invalid, at least under these conditions. 

According to the Keele theory the incorporation of styrene would be represented by Reaction (E). Since Yamashita s experiments were done with the catalyst boron fluoride etherate, whose mode of action is still quite obscure, we denote the activating group derived from the catalyst by Y (which may be Et). 

The metal halide catalysts include aluminum chloride, aluminum bromide, ferric chloride, zinc chloride, stannic chloride, titanium tetrachloride and other halides of the group known as the Friedel-Crafts catalysts. Boron fluoride, a nonmetal halide, has an activity similar to that of aluminum chloride. 

Aluminum chloride, boron fluoride and certain other Friedel-Crafts catalysts catalyze the polymerization of isobutylene, at temperatures below about —70° recent work has indicated that the presence of a promoter such as water is usually necessary (see Section V). A rubberlike polymer is obtained. 

With halide catalysts of the Friedel-Crafts type (e.g., aluminum chloride or boron fluoride) in the presence of hydrogen halide the formation of the carbonium ion results in the addition of the proton from the promoter to the pi electrons ... 

The polymerization of olefins in the presence of halides such as aluminum chloride and boron fluoride but in the absence of hydrogen halide promoter may also be described in terms of the complex carbonium ion formed by addition of the metal halide (without hydrogen chloride or hydrogen fluoride) to the olefin (cf. p. 28). These carbonium ions are apparently more stable than those of the purely hydrocarbon type the reaction resulting in their formation is less readily reversed than is that of the addition of a proton to an olefin (Whitmore, 18). Polymerization in the presence of such a complex catalyst, may be indicated as follows (cf. Hunter and Yohe, 17) ... 

The low temperature polymerization of isobutylene (that is, polymerization at temperatures below about —70°) in the presence of Friedel-Crafts catalysts (particularly boron fluoride, aluminum chloride, and titanium tetrachloride, has been studied quite intensely. The reaction is commercially important because it yields a high molecular weight... 

The catalytic activity of certain of the Friedel-Crafts catalysts was shown to decrease over a very wide range in the series boron fluoride, aluminum bromide, titanium tetrachloride, titanium tetrabromide, boron chloride, boron bromide and stannic chloride (Fairbrother and Seymour, mentioned in Plesch al., 83). When boron fluoride is added to isobutylene at dry ice temperatures, the olefin is converted to a solid polymer within a very few seconds. The time required for complete polymerization with aluminum bromide hardly extends to a few minutes while reaction times of hours are required with titanium chloride and periods of days with stannic chloride. 

Unlike boron fluoride, titanium tetrachloride does not catalyze the liquid phase polymerization of isobutylene under anhydrous conditions (Plesch et al., 83). The addition of titanium tetrachloride to a solution of the olefin in hexane at —80° failed to cause any reaction. Instantaneous polymerization occurred when moist air was added. Oxygen, nitrogen, carbon dioxide, and hydrogen chloride had no promoting effect. Ammonia and sulfur dioxide combined with the catalyst if these were added in small quantity only, subsequent addition of moist air permitted the polymerization to occur. Ethyl alcohol and ethyl ether, on the other hand, prevented the polymerization even on subsequent addition of moist air. They may be regarded as true poisons. 

That the liquid phase polymerization of isobutylene in the presence of aluminum chloride or boron fluoride occurs even in the absence of a third component may be seen in the results of another worker (Houtman, 88). With solid aluminum chloride polymerization occurred at room temperature when isobutylene vapors were passed over the catalyst and at —80° when the isobutylene was in the liquid state. In both cases the reaction was incomplete probably because the catalyst was covered with a layer of polymer. The reaction at — 80° was not however as explosive in character as when moisture was present. When boron fluoride was added to isobutylene at about —80°, the polymerization started immediately in the gas phase above the liquid (owing to a somewhat higher temperature) and sometimes continued for a considerable time before reaction was noticeable in the liquid phase. There was no difference in behavior between the experiments in which the catalyst was led through or over the liquid. It was pointed out, however, that these results are not necessarily in contradiction with those of Polanyi and Evans and co-workers and that it is possible that the presence of moisture is neces-... 

---