J. Org. Chem. 1993,58, 3575-3577
3871
Large Formation Constant of a Transient 1:l
dl-Da-Trishomocubylidene-Da-trishomocubaneBr2 Charge-Transfer Complex: General Implications for the Mechanism of Electrophilic Bromination of Olefins G. Bellucci,'*+R. Bianchinij C. Chiappe,' V. R. Gadgil,@and A. P. Marchando Dipartimento di Chimica Bioorganica, via Bonanno 33, Pisa 56126, Italy, and Department of Chemistry, University of North Texas, Denton, Texas 76203-5068 Received Febrwry 3,1993
Olefin-Brz charge-transfer complexes (CTC)have been shown1*2to be the precursors of the bromonium-bromide or polybromide (n= 3, 5I3v4ion pair intermediatea involved in the ionic bromination of olefins (eq 1). It has been
kioh
JBr
so
r e c ~ g n i z e dthat ~ * ~this preequilibrium tends to reduce the information content of the k o w for bromination, unless structural effeds on Kr,k-1, and k2 (or ~ S O H )are much smaller than those on kl. The following recent data show, however, that alkyl substitution on the double bond as well as substituent steric effeds have a very large influence on Kt: in 1,2-dichloroethane at 25 "C Kf= 0.47 M-' for cyclohexene,29.71 M-l for tetraisobutylethylene? and 2.89 x 102 M-l for adamantylideneadamantane (11.' The comparison with the last compound was, however, subjected to criticism,' because 1 is an atypical olefin in that its bromination does not lead to the normal addition product but stops at the stage of bromonium ion formation.' In this paper we are reporting on the determination of Kr of a 1:l CTC between Br2 and dl-03-trishomocubylidene-D3-trishomocubane(2), a tetrasubstituted %age" olefin similar to 1,which, however, does react with Br2 to give the expected 1,Qdibromide 3. The results confiim a large increase in Kfwith increasing substitution on the double bond. Equal volumes of a 3.2 X 1O-a M solution of Br2 and of 3.2 X 1O-a to 1.5 X 1P2M solutions of 28 in 1,2dichloroethane were mixed at 25 "C in a stopped-flow apparatus, and the reactions were monitored in a wavelength range between 250 and 350 nm. Initial absorbances,
Br 1
P
8
which were measured immediatelyafter the reactante had been mixed, were significantly higher thanthose expected on the basis of analytically determinedBr2 concentrations. These measured absorbances were found to increase with increasing concentrations of substrate 2 at constant Br2 concentration. Figure 1shows these initial absorbances (A,) measured at 25 OC and 300 nm, where the Br2 absorption is near to its minimum (tw = 13M-l cm-I), as a function of the concentration of 2. The curve is consistent with a progressive complexation of Br2 by the olefin to give a CTC, saturation being attained at a concentration of 2 above ca. 6 X 10-9 M. The data of Figure 1were fittad to the Scott equation for a 1:lcompleses,9 ignoring the small contribution to 4 by free Br2. A satisfactory fit was obtained, giving Kt = 7 X 102 M-' and tw = 3.4 X 10s M-l cm-l for the 1:l CTC. These estimates were used as starting parameters in a nonlinear leaet-squares fitting procedure' including the entire set of A,-concentration data obtained in the 260350-nm range and taking into account the free Br2 contribution to A,. This yielded a final value of Kfof 7.68 (0.85) X 102 M-' at 25 "C and the molar extinction coefficients of the 1:l CTC at all examined wavelengths. This Kr is only 2.6 times larger than that found for 1.The root-mean-square residual of the fit was 0.048,and the correlation coefficients between the Kfand any of the Q never exceeded hO.8. The obtainment of a very satisfactory fit to an equation derived for 1:l complexes excluded any significant contribution to the CTC band by speciesof 1:2stoichiometrylike a bromonium-tribromide, formed by Brz-aeeistad ionization of the CTC, whose Br3counteranion should absorbin the same region. In contrast to the 1-Br2 system, where the impossibility to proceed beyond the formation of the bromonium-polybromide species allowstheir equilibrium accumulation, in the case of 2 the collapse to dibromide3 prevents this accumulation in relevant amounts. The computed best fit W spectrum of the 2-Br2 CTC is shown in Figure 2. The absorption maximum, A,, = 267(1)nm(em=9.1(O.7) X 10sM-lcm-l),isneartothose observed for the analogoustetraisobutylethylene.Br2(260 nmP and adamantylideneadamantaneBr2CTC (272nm).' The CT band decayed fast according to the integrated second-order rate law of eq 2, with a k,w, measured at several wavelength, of 3.0 (0.5)X 106M-ls-l. The dibromo
Dipartimento di Chimica Bioorganica. t Present address: Dipartimentodi S c i e m Chimiche, vide A. Doria 6, Catania 96125, Italy. I University of North Texas. adduct 3 was the final product.1° This h o w cannot simply (1) Gamier, F.; Donnay, R. H.; Dubois, J. E. Chem. Commun. 1971, be the rate constant for the dissociation of the CTC to a 829. bromonium-tribromide ion pair because such dissociation ( 2 ) Bellucci,G.;Bianchini,R.; Ambrosetti, R. J. Am. Chem. SOC. 1986, 107.2464. is too fast for the stopped flow time scale in the case of i3)Bellucci,G.;Bianchini,R.;Chiappe,C.;Ambroeetti,R. J. Am. Chem. 1 and is expected to be as fast for 2. The how is rather SOC. 1989,111,119. (4) Bellucci, G.;Bianchini, R.; Chiappe, C.; Marioni, F.; Ambroeetti, related to the product-forming step, which is retarded, R.; Brown, R. S.; Slebocka-Tilk,H.J. Am. Chem. SOC. 1989, I l l , 2640. although to a much lower extent than in the case of 1, by (5)Ruasse, M.-F. Acc. Chem. Res. 1990,23, 87. (6)Brown,R.S.;Slebocka-Tilk,H.;Bennet,A.;Bellucci,G.;Bianchini, (9) Scott, R. L. Recl. Trcrv. Chim, Pays-Rae 1966, 76, 787. R.; Ambrosetti, R. J. Am. Chem. Soc. 1990,112,6310. (10) Thin is at variance with the behavior of meso-D&ris.homocuby(7) Reference 5, footnote 17b. lidene-Da-trishomocubane,giving a rearrangedspiro compound;see: ref (8) Marchand,A. P.;Reddy, G. M.; Deshpande, M. N.;Watson,W. H.; 8. Lee, 0. S.; oeawa, E.Bull. Korean. Chem. Soc. 1992, I S , 69. Nagl, A.; Lee, 0.S.;Oeawa, E. J. Am. Chem. SOC. 1990,112,3621. f
~
Q 1993 American
Chemical Society
3576 J. Org. Chem., Vol. 58, No. 13, 1993
Notes
the Kf increases at least by a factor of lo3on passing from a 1,2-disubstituted to a tetrasubstituted olefin. These data provide therefore a definite c o n f i t i o n for the previously inferred large dependence of Kt of olefinBr2 CTC's on substituent effects.4 This dependence is likely to be s i m i i in other solvents because solvent effects on Kf are T0.6 modest.2tBJ2 V It has been assumed that electronic effects of substiti0.4 uents on kl are much higher than those on Kf because of the much larger charge development in the TS for the ionization step than in CTC formation.6 Since in crowded olefins steric effects influence howmuch more than Kf,13 0.2 a rigorous check of this assumption would require a knowledge of both Kf and h o w for uncrowded olefins in 0.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . the same solvent. Unfortunately, this is not possible at 0.00 2.00 4.00 6.00 8.00 present because in solvents like 1,2-dichloroethane tri[og x i ~ 3 , ~ and tetrasubstituted uncrowded olefins react so fast that Figure 1. Initial absorbances (Ao)at 300 nm of mixtures of 1.6 neither h o w nor Kf are accessible, even using fast kinetic X lo-' to 7.5 X 1PM dl-Dg-trishomocubylidene-D~-trishomocu- techniques like the stopped flow. Kinetic data for monobane and 1.6 X lo-' M Brz in 1,2-dichloroethaueat 25 O C against to tetraalkyl-substituted ethylenes have been reported in the olefin concentrations. Freon112,14but, having been obtained by poorly perfor10 mant techniques, they could be not reliable. Kinetic constants have been accurately measured for ethylene to tetramethylethylene in methanol,ls but again the relative %.loa Kfare not experimentally accessible in this solvent because of the too high bromination rates. However, kinetic data for the bromination of ethylene- and methyl-substituted ethylenes show that very similar rate increases for the same changes in substitution occur in solventsas different as methanol or ethanol and acetic acid.lsJ6 Thus, the conversion of ethylene to its 1,l-dimethyl and to its cis51 and trans-l,2-dimethyl derivative causes respective rate enhancements in methanol of 8 X lp-,5.1 X 103-and 2.8 X 103-fold;lS likewise, the conversion of propene to trimethylethylene gives a rate increase of 3.2 X 103-fold 0 1 in methanol and 3.8 X 103-fold in acetic acid,16 while 250.00 350.00 1. In.) changing from 1,l-dimethyl-,cis-l,2-dimethyl-, and trans1,2-dimethyl-substitutedto tetramethyl-substituted ethFigure 2. Computed best fit spectrum of the 1:l dl-Daylene produces respective rate enhancementsof 2.9 X lo2-, trishomocubylidene-D~-trishomocubaneBr~ CTC in 1,2-dichlo6.1 X lo2-and 1.1 X 103-fold in methanol and 4.6 X 102-, roethane. 5.7 X lo2-and 7.5 X 102-fold in acetic acid.ls Similar rate increases of 2.6 X lo2and 4.1 X 102have been respectively steric effects on nucleophilic attack by the counteranion found on passing from cis-2-penteneand trans-2-pentene at the bromonium carbons. to tetramethylethylene in ethanol.ls The results obtained in the present study for bromiAll these data suggest that the reactivity ratios of these nation of 2 suggest that the C=C double bond in "cage" olefins are scarcely affected by the solvent, and therefore olefms such as 1 and 2 is more accessible for Br2 they should likely be of the same magnitude in 1,2complexation than in other tetraalkylated olefins which dichloroethane, too. On this basis, and on account of the contain branched alkyl groups (e.g., tetraisobutylethylmodest solvent effects on Kf, it can be inferred that an ene).B The change in hybridization concomitant with increase by two in number of alkyl substituents on the bromonium ion formation in 1and 2 causes the two 'cage" double bond increases both the Kf and the h o w roughly moietiea in each substrate to approach one another. The by a factor of lo3. This point indicates that substituent proximity of the "cage" moieties in the resulting bromoeffects are not much more influential on h o w than on Kt. nium ions in each case effectively impedes subsequent A possible rationalization of the lower accelerating effect backside attack by the counteranion. This deleterious by alkyl substituents on the brominationrates, relative to steric effect is more pronounced in the case of bromine what could be expected for an AdECl mechanism on the addition to 1than to 2." The Kf of 7.68 X lo2M-l found basis of their effecta on Kf, could be found in a reversible for the 2.Br2 CTC can therefore be taken as a lowest ionization of CTC's to bromonium-bromide ion pairs. This limiting evaluation for unhinderedtetrasubstituted olefii. reversibility,which has been unambiguouslydemonstrated The comparison of this value with that found for the cyclohexeneBr2 CTC indicates that in 1,2-dichloroethane "O
1
1
I/
,/
!n
\\
(11)A reviewer has suggested that the lack of dibromide formation from 1 may be due to reversibilityof the produd-forming step,the covalent dibromidebeing unstable toward ionizationunder the reactionconditions. Although this hypothesis cannot be checked experimentally because of the unavailabilityof this dibromide, it seem to us to be less probable, since the s i m i i l y substituted dibromide 8 is stable.
(12) Ruesee, M.-F.; Motallebi, S. J. Phys. Org. Chem. 1991,4, 527. (13) In the limiting case of adamantylideneadamantaneKrie 289 M-l, whereas k o w IS mr0.4 (14) Gamier, F.; Dubia, J.-E. Bull. SOC.Chim. I+. 1968, 3797. (15) Ruaese, M.-F.;Argile, A.; Bienvenue-Mtz, E.; Dubois, J.-E. J . Org. Chem. 1979,44,2758. (16) Ruaese, M.-F.; Zhang, B.-L. J. Org. Chem. 1984, 49, 3207.
J. Org. Chem., Vol. 58, No. 13,1993 8677
Notas
by different approaches,17 should practially result in a decreased ionization rate and therefore in a decrease in the koM for bromination. Experimental Section dl-D~-Triahomocubylidene-D~-triahomocubane (2), was pre1,2-Dichloroethane (Fluka, pared and purified as >99.6%) and bromine (C. Erba RPE >99.6%) were treated as previously described.l8 1,2-Dichloroethane solutions of 2 and Brz were prepared by weighingthe reagenta in accuratelycalibratedvolumetric flasks. The Br2 solutionswere stored in the dark, checked spectrophotometricallyla for their concentrations, and discarded when absorptionsaround the Br2 W minimum (280-360 nm)higher than expected were found. &ual volumes of solutions of 2 and of Bra, prethermoatated at 25 OC,were mixed in a Durrum Model D-110 stopped flow (17)Bellucci, G.;Bianchini, R.; Chiappe, C.; Brown, R. S.;SlebockaTilk, H.J. Am. Chem. SOC.1991,113,8012 and references cited therein. (18)Bellucci, G.; Berti, G.; Bianchini,R.; Ingroseo, G.; Ambrosetti,R. J. Am. Chem. Soc. 1980,102, 7480.
apparatus equipped with a 2-cm observation cell and coupled to a data acquisitionsystem,* Absorbanceswererecordedat several wavelengths between 260 and 360 nm. For the determination of Kf,the initial absorbancee (A,) at 300 nm were f i t fit to the Scott equationein order to obtain a preliminary evaluation. The compositeA, data for all rune were then fit by the already described nonlinear least-squarea procedure'~~ that yielded both Kfand the spectrum at the measured wavelengths for the CTC. The kineticconstants ( k , ~have ) been obtainedby fitting the A ( t )valuesrecorded at each wavelength to eq 2, wing for em the values obtainedin the abovefitting (weFigure 2). Reproducible ~ valuesof k~ were obtainedat eachwavelength. Thek , quoted in the text is an average value.
Acknowledgment. This work was supported in part by granta from Consiglio Nazionale delle Ricerche (CNR, Roma) and Minister0 dell'universita' e d e b Ricerca Scientifica e Tecnologica (MURST, Roma). A.P.M. thanks the Office of Naval Research (Contract No. N0001492-5-1362)and the Robert A. Welch Foundation (Grant No. B-963)for financial support of this study.
