2961 Experimental Section Nmr single resonance spectra were obtained on Varian HA 100 (1H) and A56-50 ('OF) spectrometers. lH(lgF), 'OF("), and l@F( 1gF) double-resonance experiments were performed on a Varian HA-60 instrument operating at 56.4 and 60.0 MHz. Octafluorocyclobutane, CClsF, CF2C1CF2C1,and CeHsCFa were used as lock signals for the lgFexperiments. The necessary irradiating field for the heteronuclear decoupling experiments was provided by a NMR Specialties Model SD-60B heteronuclear spin decoupler. All of the benzotrifluorides 5 were obtained from Columbia Organic Chemicals. a,a-Difluorotoluene was prepared from benzaldehyde and SF4.l6 a,a,a-Difluorochlorotoluene. In a three-necked, 250-ml flask equipped with a thermometer, gas inlet tube, and reflux condensor was placed 30 ml of apdifluorotoluene and 50 mg of azobisisobutyronitrile. The flask was then heated to 70" while chlorine gas was bubbled through the reaction mixture. The reaction was followed by lgF nmr and by testing the effluent gas for HCI and was found to be over in about 90 min. The product was then distilled: bp 140", 61" (50 mm); nmr (CCL) multiplet 6 7.5, 19F C$
49.36.
2,3,4,5,6-Pentafluorobenzyl Fluoride. The apparatus consisted of a 250-ml, three-necked flask equipped with an addition funnel, a motor-driven stirrer, and an exit tube leading to a Dry Ice trap connected to a vacuum pump. The apparatus was flushed with dry nitrogen and then 48 g (0.2 mol) of finely powdered HgFz was placed in the flask and 10 g (0.04 mol) pentafluorobenzyl bromide (Penninsular Chem Research) was placed in the funnel. After evacuation of the flask the bromide was then added dropwise to the rapidly stirred mercuric fluoride over a period of 15 min. The product which collected in the Dry Ice trap was then filtered from a few grams of NaF, yield 1.8 g ( 4 5 x ) . The nmr spectrum is given in Table IV. Decafluorobenzhydrol was obtained from Imperial Smelters Ltd. Tris(pentafluoropheny1)carbinol was a gift from Professor R. Filler. 1 2 (15) W. R. Hasek, W. C. Smith, and V. A . Englehardt, J . Amer. Chem. Soc., 82, 543 (1960).
a,&-Difluoroethylbenzenewas prepared from H F and phenylacetylene according to Matsuda, et al.16 A procedural change suggested in ref 27 of this paper was followed and resulted in a yield of 24% rather than the 18 reported. a,a-Dichloroethylbenzene. A solution of 50 g of phenylacetylene in 100 ml of methylene chloride was cooled to - 4 0 " and saturated with anhydrous HCl for 4 hr. After removal of the solvent and excess HCI on a rotary evaporator, the product was vacuum distilled. After collection of a small portion of a-chlorostyrene the product was collected and then redistilled: bp 71" (4 mm), 50" (1 mm); nmr (CC14) multiplet 7.68 (2), multiplet 7.28 (3), 2.50 (3), yield 75 %. The a-chlorostyrene fraction was redistilled: bp 98" (46 mm); nmr (CC14) multiplet 7.55 (2), multiplet 7.25 (3), doublet 5.67 (l), doublet 5.44 ( l ) , yield 10%. The title compound should be stored at 0". a,a-Dibromoethylbenzene was prepared from anhydrous HBr and phenylacetylene using a procedure analogous to the one used for preparing apdichloroethylbenzene, yields: a,a-dibromoethylbenzene, bp 76" (0.15 mm); nmr (CC14)multiplet 7.70 (2), multiplet 7.25 (3), 3.94 (3), 25%; a-bromostyrene, bp 61" (0.5 mm); nmr (CCld) multiplet 7.52 (2), multiplet 7.25 (3), doublet 6.03 ( l ) , doublet 5.71 ( l ) , 60%. The title compound should be stored at 0". 22-Dihalopropanes. The fluorine compound was prepared according to the literature16 from acetone and SF4. The chlorine compound was obtained from J. T. Baker and the bromine compound from K and K Laboratories. Dihalodiphenylmethanes. The fluorine compound was prepared from SF4 and benzophenone.15 The chlorine compound was obtained from Frinton Laboratories and distilled before use. Acknowledgment. Support of this research by the National Science Foundation and the Petroleum Research Fund, administered by the American Chemical Society is greatly appreciated. Professor R. Filler is thanked for the sample of tris(pentafluoropheny1)carbinol. (16) K. Matsuda, J. A . Dedlak, J. S. Noland, and E. C. Glocker, J . Org. Chem., 27,4018 (1962).
The Nature of the Carbonium Ion. I. The T-Route Norbornyl Cation from a Thiocyanate-Isothiocyanate Isomerization Langley A. Spurlock and
Walter G . Cox
Contribution from the Department of Chemistry, Temple University of the Commonwealth System of Higher Education, Philadelphia, Pennsylvania 19122. Received October 10, 1968 Abstract: 2-(A3-Cyclopenteny1)ethylthiocyanate was isomerized to a mixture of exo-2-norbornyl thiocyanate, ex0-2norbornyl isothiocyanate, and 2-(A3-~yclopentenyl)ethylisothiocyanate in a variety of solvents. No endo-norbornyl products were detected. The rate of reaction and product composition were directly governed by the solvent employed. The necessity of the double bond for isomerization was established by the failure of the saturated 2-cyclopentylethyl thiocyanate to isomerize under these conditions, Rate measurements confirmed the isomerization as a first-order process and activation parameters were calculated. The catalytic effects of potassium perchlorate, potassium thiocyanate, and boron trifluoride were studied and the results were used in deducing the nature of the norbornyl cations which serve as intermediates for isomerization.
Pthermal their
rior investigations have given evidence t h a t the rearrangements of alkyl thiocyanates to isomeric isothiocyanates can proceed by several mechanistic pathways. It can be seen that the choice of isomerization mode is dependent on the structure of the alkyl moiety, the nature of the solvent employed as an isomerizing medium, and on catalysts. In the cases of most allylic thiocyanates, rearrangement occurs by
way o f a six-membered cyclic transition state involving little to no charge separation.lP2 These reactions are therefore relatively insensitive to solvent and catalyst e f f e c t ~ . ~ Nonallylic ,~~ thiocyanates present a (1) 0. Billeter, Helu. Chim. Acta, 8, 337 (1925). (2) 0. Mumm and H. Richter, Ber., 73,843 (1940). (3) P. A. S. Smith and D . W. Emerson, J . Am. Chem. Soc., 82,3076 (1960).
Spurlock, Cox 1 Thiocyanate-Isothiocyanate Isomerization
2962
somewhat more complicated picture. Their conversion to isothiocyanates is strongly solvent influenced, and is often catalyzed by the Lewis base thiocyanate ion4 as well as by Lewis acids.j The over-all aspect of these isomerizations is one of a kinetically favored thiocyanate with a relatively weak carbon-sulfur bond proceeding Diu bimolecular displacement or unimolecular dissociation-recombination to the thermodynamically favored isothiocyanate. Intensive investigation of the benzhydryl thiocyanates by Fava, Iliceto, and coworkers4has revealed that they undergo ionization with formation of ion pairs followed by recombination to give the corresponding isothiocyanates. It was estimated that only a small fraction of the ion pairs in this system, which can produce a relatively stable intermediate carbonium ion, proceed to degrees of dissociation past the intimate pair stage. Our own observation that cyclopropylcarbinyl thiocyanate, in contrast t o most derivatives of this skeleton, gives primarily unrearranged product6 would seem to bear out the notion of closely associated ionic intermediates. T o further explore this possibility we wished to examine thiocyanate isomerizations in which participation by a remote double bond supplies a nucleophilic driving force for the ionization. Principal analogies for this process may be found in reports? that the solvolyses of 2-(A~-cyclopentenyl)ethyl arylsulfonates (1) proceed at rates substantially faster than their saturated analogs giving nearly exclusively cyclized product. Explanations for these observations,
limiting structure corresponding to a starting material from which the ion may be generated and, in theory, corresponding to a possible product derived from this ion. That the product derived from 2c is virtually never obtained from solvolyses of any of the starting materials casts doubt on the representation of the fully formed ion as having this limiting structure. Correspondingly, many authors have adopted the bridged notation eliminating this form. The desirability of thiocyanate ionizations for possible clarification of this point is indicated by the predominant intimate-pair status of intermediates, rendering the thiocyanate ion an effective carbonium ion "trap." The choice of 2(A3-cyclopenteny1)ethyl thiocyanate (3) was thus dictated by the wealth of information already available for ionic dissociations involving this carbon skeleton and the ease of synthesis from the known esters 1.
Results The p-toluenesulfonate la, precursor of %(A3cyclopenteny1)ethyl thiocyanate (3), was prepared from 2-(A3-cyclopentenyl)ethanol.7 Treatment of l a with potassium thiocyanate in acetone or sulfolane gave a 96:4 ratio of 3 to isothiocyanate 4 (Scheme I, eq 1). Separation of this mixture by chromatography on silica gel or by preferential reaction of 4 with alkyl amines gave pure 3. 2-(A3-Cyclopenteny1)ethyl isothiocyanate (4) was also prepared by the reaction of carbon disulfide, ethyl chloroformate, and potassium hydroxide with the corresponding amine (eq 2). The isothiocyanate 4, Scheme I
gs8 -
1
xscs
2 Li4lH
subsequently elaborated,s have assumed a large degree of n-electron participation in the transition states, and mainly the intermediacy of species resembling the bridged ion 2. This bridged representation can also be described in valence bond notation (2a-c) with each (4) (a) A. Iliceto, A . Fava, and U. Mazzucato, Tetrahedron Letters, 27 (1960); (b) A. Iliceto, A . Fava, U. Mazzucato, and P. Radici, Gazz. Chim. I t a / . , 99, 919 (1960); (c) A. Iliceto, A . Fava, U. Mazzucato, and 0. Rosseto, J . A m . Chem. Soc., 83, 2729 (1961); (d) A. Fava, A. Iliceto, A. Ceccon, and P. Koch, ibid., 87, 1045 (1965); (e) A. Fava, A. Iliceto, and S . Bresadola, ibid., 87, 4791 (1965). ( 5 ) A. Smits and H. Vixseboxe, Verslag Koninkl. Akad. Wetenschap., 46 (1913); Chem. Abstr., 8, 649 (1914); J. Gillis, Rec. Trau. Chim., 39,330 (1920); E.Schmidt, W. Striewsky, M. Secfelder, and F. Hitzler, Ann., 568, 192 (1950). (6) L.A. Spurlock and P. E. Newallis, Tetrahedron Letters, 303 (1966). (7) R. G. Lawton, J . A m . Chem. Soc., 83, 2399 (1961); P. D. Bartlett and S . Bank, ibid., 83,2591 (1961).
(8) (a) P. D. Bartlett, S. Bank, R. J. Crawford, and G. H. Schmid, ibid., 87, 1288 (1965); (b) K. Humski, S. Borcic, and D. Sunko, Croat. Chem. Acta, 37,3 (1965); (c) C. C. Lee and L. I
