Synthesis and dynamic properties of cycloheptatrienyl(dipropyl)borane. Equilibrium with 7-dipropylborylnorcaradiene

Article abstract of DOI:10.1021/ja9724699

The synthesis of cycloheptatrienyl(dipropyl)borane (2a) was accomplished via the exchange reaction of trimethyl(cycloheptatrienyl)tin (6) and dipropylchloroborane. Compound 2a was found by NMR spectroscopy to equilibrate with its valence tautomer 7-exo-(d

Full text of DOI:10.1021/ja9724699

1034
J. Am. Chem. Soc. 1998, 120, 1034-1043
Synthesis and Dynamic Properties of
Cycloheptatrienyl(dipropyl)borane. Equilibrium with
7-Dipropylborylnorcaradiene
Ilya D. Gridnev,*,† Oleg L. Tok,† Natalya A. Gridneva,† Yuri N. Bubnov,† and
Peter R. Schreiner‡
Contribution from the A.N. NesmeyanoV Institute of Organoelement Compounds, VaViloVa 28,
117813 Moscow, Russia, and Institut fu¨r Organische Chemie der Georg-August-UniVersita¨t Go¨ttingen,
Tammannstrasse 2, D-37077 Go¨ttingen, Germany
ReceiVed July 21, 1997. ReVised Manuscript ReceiVed NoVember 5, 1997
Abstract: The synthesis of cycloheptatrienyl(dipropyl)borane (2a) was accomplished via the exchange reaction
of trimethyl(cycloheptatrienyl)tin (6) and dipropylchloroborane. Compound 2a was found by NMR spectroscopy
to equilibrate with its valence tautomer 7-exo-(dipropylboryl)norcaradiene (2b). The equilibrium between 2a
and 2b was studied in detail experimentally by variable temperature NMR and theoretically by ab initio
calculations of size-reduced molecular systems (2a and 2b, with methyl instead of n-propyl groups) at the
B3LYP/6-311+G*//B3LYP/6-31G* + ZPVE level. Experimentally determined thermodynamic parameters
of the equilibrium (∆H ) 2.4 ( 0.1 kcal mol-1; ∆S ) -5.5 ( 0.3 cal mol-1 K-1) and the activation barriers
at 228 K (∆Gq228(2af2b) ) 8.2 ( 0.1 kcal mol-1, ∆Gq228(2bf2a) ) 9.4 ( 0.1 kcal mol-1) are in reasonable
agreement with the computed results (∆H ) 2.0 kcal mol-1, ∆S ) -3.7 cal mol-1 K-1; ∆Gq228(2af2b) )
3.2 kcal mol-1 and ∆Gq228(2bf2a) ) 6.7 kcal mol-1). The computations also indicate that 7-endo-
(dimethylboryl)norcaradiene (2c) is 7.6 kcal mol-1 less stable than the exo-isomer 2b due to more favorable
overlap of the unoccupied boron 2p AO with the Walsh orbital of the three-membered ring moiety in 2b.
1
Line shape analyses together with 2D H EXSY data for the equilibrating system of 2a and 2b allowed the
detection of a [1,7] sigmatropic shift in 2a at temperatures above 293 K. This is confirmed by the computations
identifying the [1,7] B shift to have the lowest activation enthalpy (2a, ∆Hq ) 18.4 kcal mol-1). In the
unsymmetrical deuteriopyridine complex 9, the empty boron 2p AO interacts strongly with the nitrogen lone
pair. This reduces the stabilization of the exo-norcaradiene skeleton, yielding only the “pure” cycloheptatrienyl
form 9 in the NMR spectra. Both NMR data and the computations show that the rotation about the B-C
bond in 9 is hindered.
Introduction
(dipropyl)borane (1): a [1,3] B sigmatropic shift, a [1,7] B
sigmatropic shift accompanied by [1,2] migration of the iron
tricarbonyl group, and a [1,3] Fe haptotropic shift.3 The inherent
symmetry of 1 allowed us to exclude simultaneous [1,5] B
migrations. We concluded that [1,3] and [1,7] boron migrations
are allowed, whereas [1,5] B shifts are prohibited.3 However,
the question of whether the observed selectivity is only due to
the presence of the transition metal in 1 remained unanswered.
The synthesis of an uncoordinated cycloheptatrienylboron
derivative was therefore of particular interest.
Another motivation for synthesizing and studying boron
cycloheptatrienyls is to elucidate further on the long-standing
discussion over the tautomerism between cycloheptatrienes
(CHT) and norcaradienes (NCD).4 Despite Willsta¨tter’s for-
mulation of the problem as early as 1901,5 it has been tackled
experimentally only after NMR techniques developed into a
powerful tool for physical organic chemistry in the 1960s.4a,b
Examples were found where either norcaradienes6 or cyclohep-
tatrienes predominate,7 but also where rapid equilibria between
Cycloheptatrienes display a large variety of dynamic pro-
cesses which in most cases can be understood in terms of
qualitative molecular orbital theory. The sigmatropic1j migra-
tions of 7-η1 derivatives of cycloheptatriene afford typical
examples.1 The nature of a particular rearrangement depends
on the migrating atom: while [1,5] sigmatropic shifts are
observed for the proton1d-g and groups containing tin as the
migrating center,1a-c substituted Ru,1h Re,1i and S1j,k atoms only
rearrange in a [1,7] fashion. Polyhapto-metal complexes of
cycloheptatriene are also known to undergo haptotropic shifts.2
Recently we have found that three independent processes take
place in the iron tricarbonyl complex of cycloheptatrienyl-
† A.N. Nesmeyanov Institute of Organoelement Compounds.
‡ Georg-August-Universita¨t Go¨ttingen.
(1) (a) Larrabee, R. B. J. Am. Chem. Soc. 1971, 93, 1510. (b) Curtis, M.
D.; Fink, R. J. Organomet. Chem. 1972, 38, 299. (c) Mann, B. E.; Taylor,
B. F.; Taylor, N. A.; Wood, R. Ibid. 1978, 162, 137. (d) Spangler, C. W.
Chem. ReV. (Washington, D.C.) 1976, 76, 187. (e) Ashe, A. J., III J. Org.
Chem. 1972, 37, 2053. (f) Larrabee, R. B. J. Organomet. Chem. 1974, 74,
313. (g) Shono, T.; Maekawa, H.; Nozoe, T.; Kashimura, S. Tetrahedron
Lett. 1990, 31, 895. (h) Heinekey, D. M.; Graham, W. A. G. J. Am. Chem.
Soc. 1979, 101, 6115. (i) Heinekey, D. M.; Graham, W. A. G. Ibid. 1982,
104. (j) Dushenko, G. A.; Mikhailov, I. E.; Zschunke, A.; Hakam, N.;
Mu¨gge, C.; Minkin, V. I. MendeleeV Commun. 1995, 133. (k) Dushenko,
G. A.; Mikhailov, I. E.; Zschunke, A.; Hakam, N.; Mu¨gge, C.; Minkin, V.
I. MendeleeV Commun. 1997, 50.
(2) (a) Benn, R.; Cibura, K.; Hoffmann, P.; Jonas, K.; Rufinska, A.
Organometallics 1985, 4, 2214. (b) ShingMan, L. K. K.; Reuvers, J. G.
A.; Takats, J.; Deganello, G. Organometallics 1983, 2, 28. (c) Davies, E.
S.; Whitely, M. W. J. Organomet. Chem. 1996, 519, 216.
(3) Gridnev, I. D.; Tok, O. L.; Gurskii, M. E.; Bubnov, Yu. N. Chem.
Eur. J. 1996, 2, 1483.
S0002-7863(97)02469-4 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/23/1998

Synthesis and Properties of Cycloheptatrienyl(dipropyl)borane
J. Am. Chem. Soc., Vol. 120, No. 5, 1998 1035
these two forms exist.8 This stimulated extensive studies of
steric and electronic substituent effects on CHT-NCD equi-
libria.9 The first explanation of the observed stability of 7,7-
dicyanonorcaradiene proposed by Ingold,10 Viz. “quasi-conju-
gation of the cyclopropane ring with the cyano-group”, was
supported theoretically by Hoffmann11a and Gu¨nther,11b who
showed that NCD is stabilized by electron donation from the
cyclopropane Walsh orbital into an unoccupied π MO. It was
shown recently that CHT-NCD equilibria can be controlled
differentially through substitution.12 For instance, NCD can be
stabilized by introducing weak π donors in 3,5-positions.13
Nevertheless, it is still impossible to differentiate between the
steric and electronic effects operative in CHT-NCD equilibria.
The study of related new systems allowing these effects to be
analyzed more cleanly is desirable, especially in view of the
fundamental importance of this problem for the synthesis14 and
understanding of similar equilibria in larger molecules, such as
1,6-methano[10]annulenes,15 as well as fullerenes and fulle-
roids.16
The present work describes the synthesis of cycloheptatrienyl-
(dipropyl)borane (2) as well as its analysis by NMR and
electronic structure methods.
Results and Discussion
Synthesis of Cycloheptatrienyl(dipropyl)borane. Synthe-
sis of Trimethylcycloheptatrienyltin (6) and the Kinetics of
the [1,5] Sn Shift. Borylation of appropriate lithium or
potassium precursors for the synthesis of allylic boranes was
accomplished previously.3,17 Unfortunately, this approach failed
for the preparation of the target compound 2 since the corre-
sponding antiaromatic cycloheptatrienyl anion is highly un-
stable.18,19 Therefore, only covalent cycloheptatrienyl metals
could be considered as suitable synthetic precursors for the
desired borane. Among those, the tin compounds are most
readily accessible via reaction of tin-lithium derivatives with
tropylium salts. Thus, triphenylcyclo-heptatrienyltin (3) was
prepared by reacting Ph3SnLi with tropylium tetrafluoroborate
(Scheme 1).1a-c
(4) For reviews, see: (a) Maier, G. Angew. Chem. 1967, 79, 446. (b)
Vogel, E. Pure Appl. Chem. 1969, 20, 237. (c) le Noble, W. J. Highlights
of Organic Chemistry; Dekker: New York, 1974; p 402. (d) Liebman, J.
F. Chem. ReV. 1989, 89, 1225. (e) Okamura, W. H.; Delera, A. R. In
ComprehensiVe Organic Synthesis; Trost, B. M., Fleming, I., Eds.;
Pergamon: Oxford, 1991; Vol. 5, p 699.
(5) Willsta¨tter, R. Liebigs Ann. Chem. 1901, 317, 204.
(6) (a) Darms, R.; Threlfall, T.; Pesaro, M.; Eschenmoser, A. HelV. Chim.
Acta 1963, 46, 1893. (b) Vogel, E.; Wiedemann, W.; Kiefer, H.; Harrison,
V. F. Tetrahedron Lett. 1963, 673. (c) Ciganek, E. J. Am. Chem. Soc. 1967,
89, 1454. (d) Mukai, T.; Kubota, H.; Toda, T. Tetrahedron Lett. 1967, 3581.
(e) Scho¨nleber, D. Angew. Chem. 1969, 81, 83. (f) Jones, M., Jr. Ibid. 1969,
81, 87.
Scheme 1
(7) (a) Knox, L. H.; Velarde, E.; Cross, A. D. J. Am. Chem. Soc. 1963,
85, 5, 2533. (b) Knox, L. H.; Velarde, E.; Cross, A. D. ibid. 1965, 87,
3727. (c) Lambert, J. B.; Durham, L. J.; Lepoutere, P.; Roberts, J. D. Ibid.
1965, 87, 3896. (d) Vogel, E.; Maier, W.; Eimer, J. Tetrahedron Lett. 1966,
655. (e) Gale, D. M.; Middleton, W. J.; Krespan, C. G. J. Am. Chem. Soc.
1966, 88, 3617. (f) Bushweller, C. H.; Sharpe, M.; Weininger, S. J.
Tetrahedron Lett. 1970, 453.
Although the yields are low and many byproducts form,
stannane 3 is stable and can be purified by column chromatog-
raphy. However, the triphenyl derivative 3 is an unsuitable
starting compound for the synthesis of the boron analogue via
transmetalation, since phenyl radicals in organotin compounds
are more labile than alkyl groups,20 and the expected product
of a reaction between 3 and a dialkylboron halide is a
dialkylphenylboron species. Therefore, a trialkyltin derivative
of cycloheptatriene was needed; unsuccessful syntheses of
trimethyl- or triethylcycloheptatrienyltin are described in the
literature.1c In these experiments bicycloheptatrienyl (4) and
ditin compound 5 were the only products of the reaction of
R3SnLi (R ) Me,Et) with tropylium tetrafluoroborate (Scheme
2). Thus, we decided to try the less ionic tropylium bromide
for the synthesis of trimethylcycloheptatrienyltin (6).
(8) (a) Ciganek, E. J. Am. Chem. Soc. 1965, 87, 1149. (b) Go¨rlitz, M.;
Gu¨nter, H. Tetrahedron 1969, 4467. (c) Hall, G. E.; Roberts, J. D. J. Am.
Chem. Soc. 1971, 93, 2203. (d) Ciganek, E. Ibid. 1971, 93, 2207.
(9) (a) Daub, J.; Betz, W. Tetrahedron Lett. 1972, 3451. (b) Gu¨nther,
H.; Peters, W.; Wehner, R. Chem. Ber. 1973, 106, 3683. (c) Kla¨rner, F.-
G.; Tetrahedron Lett. 1974, 19. (d) Stohrer, W.-D.; Daub, J. Angew. Chem.
1974, 86, 54. (e) Wehner, R.; Gu¨nther, H. J. Am. Chem. Soc. 1975, 97,
923. (f) Staley, S. W.; Fox, M. A.; Cairncross, A. Ibid, 1977, 99, 4524. (g)
Balci, M.; Fischer, H.; Gu¨nther, H. Angew. Chem. 1980, 92, 316. (h)
Takeuchi, K.; Arima, M.; Okamoto, K. Tetrahedron Lett. 1981, 22, 3081.
(i) Takeuchi, K.; Fujimoto, H.; Okamoto, K. Ibid. 1981, 22, 4981. (j)
Takeuchi, K.; Kitagawa, T.; Toyama, T.; Okamoto, K. J. Chem. Soc., Chem.
Commun. 1982, 313. (k) Takeuchi, K.; Kitagawa, T.; Senzaki, Y.; Fulimoto,
H.; Okamoto, K. Chem. Lett. 1983, 69. (l) Takeuchi, K.; Kitagawa, T.;
Senzaki, Y.; Okamoto, K. Ibid. 1983, 73. (m) Takeuchi, K.; Senzaki, Y.;
Okamoto, K. J. Chem. Soc., Chem. Commun. 1984, 111. (n) Takeuchi, K.;
Fujimoto, H.; Kitagawa, T.; Fujii, H.; Okamoto, K. J. Chem. Soc., Perkin
Trans. 2 1984, 461. (o) Takeuchi, K.; Kitagawa, T.; Ueda, A.; Senzaki, Y.;
Okamo, K. Tetrahedron 1985, 41, 5455.
(10) Ingold, C. K. Structure and Mechanism in Organic Chemistry;
Cornell University Press: Ithaca, NY, 1969; p 882.
(11) (a) Hoffmann, R. Tetrahedron Lett. 1970, 2907. (b) Gu¨nther, H.
Ibid. 1970, 5173.
(12) (a) Saba, A. Tetrahedron Lett. 1990, 32, 4657. (b) Kohmoto, S.;
Funabashi, T.; Nakayama, N.; Nishio, T.; Iida, I.; Kishikawa, K.; Yamamoto,
M.; Yamada, K. J. Org. Chem. 1993, 58, 4764. (c) Oda, M.; Horiguchi,
H.; Nasaki, Y.; Sakamoto, Y.; Kuroda, S. Recl. TraV. Chim. Pays-Bas 1996,
115, 149.
(13) (a) Matsumoto, M.; Shiono, T.; Mutoh, H.; Amano, M.; Arimitsu,
S. J. Chem. Soc., Chem. Commun. 1995, 101. (b) Matsumoto, M.; Shiono,
T.; Kasuga, N. C. Tetrahedron Lett. 1995, 36, 8817.
(15) (a) Roth, W. R.; Kla¨rner, F. G.; Siepert, G.; Lennartz, H.-W. Chem.
Ber. 1992, 125, 217. (b) Dorn, H. C.; Yannoni, C. S.; Limbach, H.-H.;
Vogel, E. J. Phys. Chem. 1994, 98, 11628. (c) Laue, J.; Seitz, G. Liebigs
Ann. Chem. 1996, 773. (d) Mealli, C.; Ienco, A.; Hoyt, E. B.; Zoellner, R.
W. Chem. Eur. J. 1997, 3, 958.
(16) (a) Warner, P. M. J. Am. Chem. Soc. 1994, 116, 11059. (b) Li, Z.;
Shevlin, P. B. Ibid. 1997, 119, 1149.
(17) (a) Gurskii, M. E.; Gridnev, I. D.; Geiderikh, A. V.; Ignatenko, A.
V.; Bubnov, Y. N.; Mstislavky, V. I.; Ustynyuk, Y. A. Organometallics
1992, 11, 4056. (b) Gurskii, M. E.; Gridnev, I. D.; Il’ichev, Y. V.; Ignatenko,
A. V.; Bubnov, Y. N. Angew. Chem. 1992, 104, 762. (c) Gridnev, I. D.;
Gurskii, M. E.; Ignatenko, A. V.; Bubnov, Y. N.; Il’ichev, Y. V.
Organometallics 1993, 12, 2487. (d) Gurskii, M. E.; Gridnev, I. D.; Buevich,
A. V.; Bubnov, Y. N. Organometallics 1994, 13, 4658. (e) Gridnev, I. D.;
Gurskii, M. E.; Buevich, A. V.; Bubnov, Y. N. Russ. Chem. Bull. 1996,
107. (f) Gridnev, I. D.; Gurskii, M. E.; Bubnov, Y. N. Organometallics
1996, 15, 3696.
(18) Kuwajima, S.; Soos, Z. G. J. Am. Chem. Soc. 1987, 109, 107.
(19) (a) Staley, S. W.; Orvedal, A. W. J. Am. Chem. Soc. 1974, 96, 1618.
(b) Cunion, R. F.; Karney, W.; Wenthold, P. G.; Borden, W. T.; Lineberger,
W. C. Ibid. 1996, 118, 5074.
(14) (a) Banwell, M. G.; Onrust, R. Tetrahedron Lett. 1985, 26, 4543.
(b) Jeener, G.; Papadopoulos, M. J. Org. Chem. 1986, 51, 1, 585. (c)
Kohmoto, S.; Nakayama, N.; Takami, J.-i.; Kishikawa, K.; Yamamoto, M.;
Yamada, K. Tetrahedron Lett. 1996, 37, 7761. (d) Oda, M.; Horiguchi, H.;
Kajioka, T.; Kuroda, S. Recl. TraV. Chim. Pays-Bas 1996, 115, 151. (e)
Kitagawa, T.; Miyabo, A.; Fujii, H.; Okazaki, T.; Mori, T.; Matsudou, M.;
Sugie, T.; Takeuchi, K. J. Org. Chem. 1997, 62, 888.
(20) (a) Moedritzer, K. Organomet. Chem. ReV. 1966, 1, 179. (b) Hann,
N. S.; Mole, T. Prog. NMR 1969, 4, 91.

1036 J. Am. Chem. Soc., Vol. 120, No. 5, 1998
GridneV et al.
Scheme 2
Scheme 3
The reaction between tropylium bromide and lithium tri-
methyltin proved to be very sensitive to the reaction conditions.
As found in earlier reports, only 4 and 5 (R ) Me) were detected
in the reaction mixtures when a solution of Me3SnLi was added
to a suspension of tropylium bromide at either 0 or -78 °C.
Reversing the order of addition also was unsuccessful. Only
when tropylium bromide was added very quickly at once to a
solution of Me3SnLi at -78 °C, followed by immediate
treatment with Et2O-H2O-NaHCO3, could 6 be detected by
NMR (Scheme 3). In some experiments the yield for 6 was
35-40%; however, 6 is rather unstable, and the best preparative
yield after purification did not exceed 5%. The only satisfactory
way of purification was by column chromatography on silica
gel in hexane; attempts to distill 6 in vacuo led to complete
decomposition.
Figure 1. 13C NMR EXSY spectrum of trimethylcycloheptatrienyltin
(6) (100 MHz, CDCl3, 328 K): mixing time 0.08 s, initial delay 1 s,
number of scans 256, a matrix 512 × 256 was acquired by the TPPI
technique and zero-filled to the size 1024 × 1024.
Triphenylcycloheptatrienyltin (3) was the first fluxional
cycloheptatrienyl derivative.1a The nature of tin sigmatropic
migrations in this compound was the key question in the
discussions of the factors governing the reaction pathways in
cyclic polyunsaturated organometallic compounds.1f Two in-
dependent investigations on the dynamic behavior of 3 showed
that a [1,5] Sn shift occurs in 3, clearly indicating orbital control
of sigmatropic tin migrations.1b,c There is some discrepancy
in the activation parameters reported for the [1,5] Sn shift in 3,
but the values given in the most recent work seem more
reliable1c (∆Hq ) 13.8 ( 0.4 kcal mol-1, ∆Sq ) -5.6 ( 1.2
cal mol-1 K-1, ∆Gq ) 15.4 ( 0.1 kcal mol-1).
300
We have studied the fluxional behavior of trimethylcyclo-
heptatrienyltin (6). The 2D EXSY NMR experiments (e.g.,
Figure 1) carried out in the 298-338 K temperature interval
provided strong evidence for a [1,5] Sn shift in 6. Kinetic data
for the [1,5] Sn migration in 6 were deduced from 2D EXSY
13C NMR spectra. It is known that at small mixing times the
rate constant can be obtained from the slope of the linear
dependence of the volume of a cross-peak on the tempera-
ture.21,22 The result of such a treatment for four different
temperatures is shown in Figure 2. The activation parameters
found from these data by the Eyring approach are shown in
Scheme 4. A decrease of the rate of the [1,5] Sn shift in 6
compared to that in 3 is due to the known tendency of the
phenyl groups increasing the migratory aptitude of a tin
atom.20
Figure 2. Kinetic curves for the [1,5] tin shift in 6.
Scheme 4
deuteriochloroform (monitored by NMR). However, when
approximately half of 6 is consumed, side reactions (i.e., the
exchange of the radicals between 6 and trimethyltin chloride
(7), giving chlorodimethylcycloheptatrienyltin (8), and tetra-
methyltin (Scheme 5)) become more significant. Compound 8
Replacement of Tin with Boron in Trimethylcyclohep-
tatrienyltin (6). Preliminary experiments showed that trans-
metalation of 6 readily proceeds at room temperature in
was characterized by one- and two-dimensional H, 13C, and
1
(21) Ernst, R. R.; Bodenhausen, G.; Wokaun, A. Principles of Nuclear
Magnetic Resonance in One and Two Dimensions; Clarendon: Oxford,
119Sn NMR spectra. Interestingly, 2D EXSY experiments
showed that the [1,5] Sn shift in 8 is significantly faster than
that in 6. This observation is in agreement with the reported
1987.
(22) Perrin, C. L.; Dwyer, T. J. Chem. ReV. (Washington, D.C.) 1990,
90, 935.

Synthesis and Properties of Cycloheptatrienyl(dipropyl)borane
J. Am. Chem. Soc., Vol. 120, No. 5, 1998 1037
Figure 3. 1H NMR spectrum of 2 (400 MHz, CD2Cl2-CDCl3-CCl4,
298 K).
Scheme 5
Figure 4. Temperature dependence of the high-field region of the 13
NMR spectrum (100 MHz, CD2Cl2-CDCl3-CCl4) of compound 2.
C
Scheme 6
Scheme 7
acceleration of tin sigmatropic shifts when an alkyl group at
the tin atom is replaced by chlorine.23
Further experiments showed that the amount of 8 produced
in the reaction of 6 with Pr2BCl is reduced when an excess of
dipropylboron chloride is used. The synthesis of the pure
organoboron compound was accomplished by treating 6 with
an 8-fold excess of dipropylboron chloride without solvent for
30 min followed by evaporation of volatile compounds in vacuo,
treatment with an additional 8 equiv of Pr2BCl, and vacuum
distillation (Scheme 6).
Dynamic Equilibrium between Cycloheptatrienyl(dipro-
pyl)borane (2) and 7-(Dipropylboryl)norcaradiene (2b).
Analysis of the NMR Spectra of 2. Figure 3 displays the 1H
NMR spectrum of the transmetalation product 2 recorded at
room temperature. In contrast to a typical 7-monosubstituted
cycloheptatriene NMR spectrum, it contains a high-field triplet
at 0.16 ppm, a multiplet at 3.18 ppm, and two multiplets in the
olefinic regionsat 6.02 and 6.22 ppm. The proton at 0.16 ppm
is attached to a carbon atom (δ ) 23 ppm) adjacent to boron;
this follows from the CHCORR spectrum and the characteristic
broadening of the signal in the 13C NMR spectrum caused by
C-B coupling. Figure 4 displays the temperature dependence
of the 13C NMR spectra of 2. Particularly dramatic changes
are observed for the signal at δ ) 59.7 ppm at 313 K. When
the temperature is decreased to 273 K, the chemical shift of
this signal decreases and it becomes much broader. The signal
almost disappears at 228 K but reappears as a very broad signal
at 203 K; a relatively sharp resonance at δ ) 33.6 ppm is
observed at 153 K.
The temperature-dependent spectra correspond to an equi-
librium between cycloheptatrienylborane 2a (endo and exo
forms) [the eqilibrium between the exo and endo forms of 2a
is fast on the NMR time scale in the studied temperature range;
only at 153 K can some indications of this process be detected,
viz., the broadening of the signals of the propyl groups in the
13C NMR spectrum, the other signals remaining relatively
narrow] and its valence tautomer 7-(dipropylboryl)norcaradiene
1
(2b) (Scheme 7).9 However, the H and 13C NMR spectra
measured at the lowest temperature (153 K, Figure 4) display
only one set of signals easily assigned to the norcaradienyl 2b.
Attempts to detect the exchange cross-peaks between 2a and
(23) Naruta, Y.; Nishigaichi, Y.; Maruyama, K. Tetrahedron 1990, 45,
1067.

1038 J. Am. Chem. Soc., Vol. 120, No. 5, 1998
GridneV et al.
Figure 6. B3LYP/6-311+G*//B3LYP/6-31G* + ZPVE structures and
energies for the [1,n] boron shifts (n ) 1, 3, 7) in 7-(dimethylboryl)-
norcaradiene.
Figure 5. B3LYP/6-311+G*//B3LYP/6-31G* + ZPVE structures and
energies (∆G values at 228 K at B3LYP/6-31G* in parentheses, cf.
Table 1) for the cycloheptatrienyl(dimethyl)borane f 7-(dimethylbo-
ryl)norcaradiene rearrangement.
from minimization of steric repulsion between the alkyl groups
and the cyclohexadiene moiety. This is analogous to the
bisected and perpendicular geometries of the cyclopropylcarbinyl
cation.24 The relative energy difference (7.6 kcal mol-1
)
2b in 1H and 13C 2D EXSY experiments at 153 K were
unsuccessful, although we had observed the exchange cross-
peaks between two tautomers with a ratio of 24:1 previously.17f
Therefore, the equilibrium must be shifted even more toward
2b at lower temperatures.
between the two isomers 2b and 2c (note that both are minima),
however, is only about half of that found experimentally (NMR
at 253 K) for the (1,1-dimethylcyclopropyl)carbinyl cation (13.7
kcal mol-1),25 owing to the greater electron demand of the
positively charged carbon system.
The temperature dependence of the stabilities of 2a and 2b
(see NMR data above) is well reproduced by our computations.
While 2b is 0.5 kcal mol-1 more stable than 2a (endo) at 298
K, this difference increases to 0.7 kcal mol-1 at 228 K, mostly
due to entropic effects favoring 2b (Table 1). The nearly
exclusive formation of 2b is also supported by the computed
NMR data (Table 2). The observed and computed proton
chemical shifts agree remarkably well; the largest deviation of
only 0.3 ppm is found for the proton experiencing the ring
current of the π system (measured, -0.5 ppm; computed, -0.8
ppm). The computed 13C shifts are systematically somewhat
too high, but agree qualitatively quite well.
The thermodynamic parameters for the equilibrium between
2a and 2b were derived from six 13C NMR spectra at 263-313
K, i.e., in the temperature interval where the chemical shifts of
the averaged C(1,6) signals could be measured accurately. The
chemical shift of C(1,6) in 2a was estimated as being ap-
proximately that of the structurally similar cyclononatetraenyl-
(dipropyl)borane (135.8 ppm).17e Since the difference between
chemical shifts of C(1,6) in 2a and 2b (approximately 100 ppm)
is about 2 orders of magnitude larger than the possible error in
Only the exo isomer of the norcaradienyl form 2b was ob-
served in the NMR spectra. This assignment has been made
on the basis of δ(H7) which is -0.47 ppm at 153 K, owing to
its orientation over the shielding cone of the conjugated π sys-
tem; larger δ(H7) values (2.0-2.5 ppm) are characteristic for
exo H7 protons in the corresponding endo isomers8b,9b {which
could form from 2a (endo)}. Since 2c was not detected in the
NMR spectra, it must be thermodynamically less stable than
2b.
To elaborate further on these considerations, we carried out
density functional computations on the isomers of 2 and their
transition states for interconversion; computational details can
be found below. To reduce the size of the problem, we replaced
the propyl groups by methyl groups in our computational
studies; to remain consistent with the structure numbering, we
labeled the computed “reduced” systems in bold italics {e.g.,
2b refers to C7H7B(CH3)2}. All structures are depicted in
Figures 5 and 6; energies are listed in Table 1.
From the computed geometries of 2b vs 2c one can easily
deduce why the latter must be higher in energy. While the
empty p AO on boron is perfectly aligned with the Walsh orbital
of the cyclopropane moiety in 2b, this interaction is essentially
“switched off” in 2c due to a perpendicular orientation resulting
(24) Hehre, W. J. Acc. Chem. Res. 1975, 8, 369.
(25) Kabakoff, D. S.; Namanworth, E. J. Am. Chem. Soc. 1970, 92, 3234.

Synthesis and Properties of Cycloheptatrienyl(dipropyl)borane
J. Am. Chem. Soc., Vol. 120, No. 5, 1998 1039
Table 1. Energies and Thermodynamic Quantitites of C7H7B(CH3)2 Isomers Related to 2 at B3LYP
abs energy
(6-31G*),
-au
abs energy incl
thermal corrections
to ∆Ga(-au)
rel energya,b
(6-31G*),
abs energyc
(6-311+G*),
-au
rel energyc
(6-311+G*),
kcal mol-1
ZPVE,a
entropy,
eua
species
kcal mol-1
kcal mol-1
2a (exo)
375.59855
375.59877
375.60204
375.58993
375.58870
375.57380
375.57196
375.55786
375.56004
122.4
109.3
122.5
109.4
123.3
110.1
122.7
109.5
122.1
109.1
121.9
108.8
122.3
109.2
123.4
110.2
122.3
109.2
96.8
90.0
103.8
96.8
100.5
93.6
103.8
96.8
102.9
96.0
100.9
94.1
95.3
88.4
95.2
88.1
92.3
89.6
375.43857
375.44697
375.44093
375.45138
375.44177
375.45254
375.43202
375.44247
375.43150
375.44191
375.42616
375.43675
375.41134
375.4226
2.0
3.5
0.5
0.7
0.0
0.0
6.1
6.3
6.4
6.7
9.8
9.9
19.1
18.8
28.9
28.5
27.2
16.7
375.68335
375.68317
375.68359
375.67154
375.67208
375.66690
375.65382
375.63940
375.64210
0.2
2a (endo)
2b
0.3
0.0
2c
7.6
TS1
TS2
TS3
TS4
TS5
7.2
10.5
18.7
27.7
26.0
375.39564
375.40708
375.39849
375.41003
a First entry at 298.15 K; second entry at 228 K, unless noted otherwise. b Including energy corrections. c Uncorrected.
2.0 kcal mol-1 and ∆S ) -3.7 cal mol-1 K-1. Note, however,
that it is quite difficult to reliably recover such small relative
energy differences computationally. A much more elaborate
electron correlation treatment (e.g., the coupled cluster method)
and a larger basis set would be needed for absolute quantitative
agreement, but this is currently not feasible computationally.
The equilibrium constant at 153 K calculated from these data
equals 182; therefore, the failure to observe exchange cross-
peaks at this temperature is quite understandable.
Table 2. Selected Measured (298 K) and Computed
(GIAO-B3LYP/6-311+G**//B3LYP/6-31G*) NMR Chemical Shifts
(vs TMS and F3B‚OEt2) of 2b and Complex 9a
C-1
C-2
C-3
C-7
B
HC1 HC2 HC3 HC7
2bb,c
exptl 33.6 128.3 121.1 16.8 76.3 2.5 6.2 5.9 -0.5
theor 39.3 137.5 129.1 22.9 85.9 2.4 6.5 6.1 -0.8
9d
exptl 124.5 124.9 129.0 129.4 8.2e 5.1 6.1 6.5
theor 133.6 138.5 140.9 143.6 0.8 5.1 6.6 6.9
1.1
0.5
The rate constants for the rearrangement of 2a to 2b and the
reverse reaction at the coalescence temperature were derived
by the method of Anet and Basus26 for systems with unequally
populated sites. It has been shown that when the differences
in both the populations and the chemical shifts are large, the
use of eqs 1a and 1b for determining the rate constants at the
a The computed species were slightly reduced in size (n-propyl
groups replaced by methyl groups) due to computational limitations.
Thus, the boron shifts are slightly shifted downfield in the computed
chemical shifts. b For numbering see Scheme 9. c Averaged with slight
traces of 2a (endo); see the text. d For numbering see Figure 12. e At
303 K.
kA ) k(major f minor) ) 2πpBδν
kB ) k(minor f major) ) 2πpAδν
(1a)
(1b)
maximum broadening temperature does not lead to significant
errors. pA and pB are the populations of the major and the minor
components, respectively, and δν is the difference in the
chemical shifts in hertz.
In our case pA ) 0.93, pB ) 0.07 (calculated from the
thermodynamic parameters of the equilibrium found above), δν
) 10 220 Hz (for the signals C(1,6) in 2a and 2b), and the
temperature of maximum broadening is 228 K (Figure 8). This
gives kA(228) ) 4.5 × 103, kB(228) ) 6.0 × 104 s-1, ∆Gq
-
228
(2af2b) ) 8.2 ( 0.1 kcal mol-1, and ∆Gq228(2bf2a) ) 9.4
( 0.1 kcal mol-1 27
.
The corresponding transition states for electrocyclic ring
opening of 2a and 2b also were computed to calibrate theoretical
with experimental results (Figure 5). We computed ∆Gq
-
228
Figure 7. Dependence of ln K versus 1/T for the equilibrium between
2a and 2b.
{2a(exo) f TS1} ) 3.2 kcal mol-1 and ∆Gq228(2b f TS1) )
6.7 kcal mol-1 at B3LYP/6-31G*, in reasonable agreement with
experiment. Note that the corresponding barrier for electrocyclic
ring closure of 2a (endo) is significantly higher (∆Gq228 ) 9.2
the estimation of the corresponding chemical shift in 2a (about
(1 ppm), this treatment should not result in a significant error.
The thermodynamic equilibrium parameters, determined through
a linear correlation of ln K vs 1/T (Figure 7), are as follows:
(26) Sandstro¨m, J. Dynamic NMR Spectroscopy; Academic Press:
London, 1984.
(27) Friebolin, H. Ein- und zweidimensionale NMR-Spectroskopie, 2nd
ed.; VCH: Weinheim, 1992.
(28) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. (Washington,
D.C.) 1988, 88, 899.
∆H ) 2.4 ( 0.1 kcal mol-1; ∆S ) -5.5 ( 0.3 cal mol-1 K-1
.
These thermodynamic equilibrium parameters are in excellent
agreement with our computations (at 298 K) which give ∆H )

1040 J. Am. Chem. Soc., Vol. 120, No. 5, 1998
GridneV et al.
Figure 10. Structure of complex 9 at B3LYP/6-31G*.
Figure 8. Dependence of the line width of the various resonances in
the 13C NMR spectrum of 2 on temperature.
Figure 11. Temperature dependence of the olefinic part of the 1H NMR
spectrum of 9 (400 MHz, pyridine-d5).
Scheme 8
Figure 9. 1H NMR spectra of 2 (400 MHz, 298 K): (a) in CD2Cl2-
CDCl3-CCl4; (b) in pyridine-d5.
pyridine by 0.054-0.056 Å, complex 9 is rather “tight” with
some particularly short atom-atom distances. The pyridine
ortho-hydrogen pointing toward the cycloheptatrienyl moiety
has an optimized distance of only 2.472 Å to one of the carbons.
The other ortho-hydrogen also has a close contact with one of
the methyl group hydrogens (2.213 Å).
Upon raising the temperature, reversible dynamic effects were
observed in the 1H and 13C NMR spectra of 9 (e.g., Figure 11).
The 2D EXSY experiments suggest that the exchange between
the positions 1 and 6, 2 and 5, and 3 and 4 in the cyclohep-
tatrienyl moiety of 9 is observed. The activation barrier for
this process was derived from the 2D 1H EXSY spectrum of 9
at 313 K. Each pair of exchanging signals in 9 can be regarded
as the exchange between two equally populated sites, and the
rate constant can be calculated from the intensities of diagonal
and off-diagonal peaks by eq 2,21,22 where IAA and IBB are the
intensities of the diagonal peaks and IAB and IBA are the
intensities of the cross-peaks.
kcal mol-1) whereas the barrier for ring opening of 2c is small
(∆Gq ) 3.6 kcal mol-1).
228
Dynamic Behavior of the Deuteriopyridine Complex 9.
Since the increased thermodynamic stability of 2b vs 2a is due
to overlap of the empty boron 2p AO with the Walsh orbital of
the cyclopropane moiety (note that there is no evidence for
norcaradienyl forms in cycloheptatrienyl derivatives of silicon,
germanium, tin, rhodium, and ruthenium due to the absense of
empty orbitals of proper size and symmetry), we speculated that
filling the empty boron 2p AO with an electron pair donor would
lead to exclusive formation of the cycloheptatrienyl form.
Figure 9 illustrates impressively that this is indeed the case:
1
the H NMR spectrum of 2 changes dramatically when the
neutral solvent is changed to deuteriopyridine. The cyclohep-
tatrienyl moiety in the deuteriopyridine complex 9 (Scheme 8)
is evidenced by 2D correlation experiments. However, initially
somewhat unexpected is the lack of symmetry in the spectrum
of 9: all nuclei are magnetically inequivalent. This observation
is confirmed by the nonsymmetrical (C1 point group) minimum
energy structure for the optimized complex 9 (Figure 10).
Although the B-C bonds lengthen upon complexation with
IAA + IBB
IAB + IBA
1
r + 1
k ) ln
tm r - 1
r )
(2)

Synthesis and Properties of Cycloheptatrienyl(dipropyl)borane
J. Am. Chem. Soc., Vol. 120, No. 5, 1998 1041
Scheme 9
Application of eq 2 to the 1H 2D EXSY spectrum of 9
is raised from 228 to 293 K is due to a rate increase for the
equilibrium between 2a and 2b. However, significant broaden-
ing of this signal at higher temperatures can only be due to the
existence of another dynamic process, viz., sigmatropic migra-
tion of the dipropylboryl group in 2a. In the same temperature
interval (293-313 K) the width of the C(2,5) signal changes
from 4.0 to 7.5 Hz, whereas the widths of those for C(3,4) and
the other resonances in the spectrum remain constants4.0 Hz
(the signal of C(7) broadens significantly, but the same effect
is expected due to the quadrupolar coupling with boron).
Scheme 9 displays the analysis of the exchange patterns for
the three possible mechanisms of boron migration in 2a.
In each case one of the signals must display a relatively weak
line width broadening since one of the exchanges transforms it
into itself. For each type of boron migration this signal is
different, C(3,4) for [1,7] B shift, C (1,6) for [1,3] boron
migration, and C(2,5) for [1,5] boron migration. Hence, we
conclude that a [1,7] B shift must occur.
The same conclusion follows from the phase-sensitive 2D
1H EXSY spectrum recorded at 323 K (Figure 12). Since the
positions of the protons relative to the position of the dipropyl-
boryl group remain unchanged when 2a interchanges with 2b
(Scheme 9), the exchange cross-peaks between the averaged
signals directly indicate the mode of the sigmatropic boron
migration in 2a. The H(7) resonance gives a cross-peak with
the H(1,6) resonance, corresponding to a formal [1,2] migration,
i.e., a [1,7] B shift in 2a. The positions of other cross-peaks in
the 2D EXSY spectrum at 323 K are also in agreement with a
sigmatropic [1,7] B shift in cycloheptatrienyl(dipropyl)borane
(2a) (Scheme 9).
afforded the rate constant for the oserved dynamic process in
9: k313 ) 1.75 s-1 (∆Gq ) 18.0 ( 0.1 kcal mol-1). The
313
chemical shifts for 9 were computed (Table 2) and show good
qualitative agreement with the measured, highly temperature
dependent NMR spectrum. Again, the proton shifts agree
exceptionally well, while the computed 13C shifts are uniformly
somewhat too high.
The measured values of the 11B chemical shift at 303 and
313 K for complex 9 differ considerably from the calculated
value at 0 K. Moreover, δ(11B) increases noticeably from 8.2
at 303 K to 10.4 at 313 K. These findings indicate that complex
9 is probably somewhat less stable than other pyridine com-
plexes of triorganoboranes. A considerable amount of uncom-
plexed 2a equilibrates with 9 already at 303 K. Therefore, the
dynamic effects observed in the 1H and 13C NMR spectra of 9
are most likely due to a dissociation-association mechanism,
and the determined activation barrier corresponds probably to
the dissociation step.
Analysis of the Mechanism of Boron Sigmatropic Migra-
tions in 2. As noted above, cycloheptatrienylborane 2a cannot
be observed in its pure form because of equilibration with its
tautomer 2b. However, since at temperatures above 300 K this
equilibrium is fast on the NMR time scale and the content of
2a is about 30%, some conclusions about the mechanism and
approximate rates of sigmatropic boron migrations in 2a may
be drawn from an analysis of the equilibrium mixture. If a
sigmatropic boron migration occurs in 2, it should lead to mixing
of the chemical shifts in 2a, which would inevitably affect the
line shape of the averaged spectrum. We conclude that the
barrier for such migrations must be comparatively high (see
computations below), since no changes in the line shapes besides
the ones belonging to the equilibrium between 2a and 2b were
observed at temperatures below 293 K. Nevertheless, such
changes were detected at higher temperatures (Figure 8). The
line width of C(1,6) displays a minimum at 293 K and increases
by 10 Hz in the 293-313 K temperature regime. The
sharpening of the averaged C(1,6) signal when the temperature
The computed relative energies also identify the [1,7] B shift
as the most favorable shift mechanism (∆Hq ) 18.4 kcal mol-1
,
Table 1, Figure 6), followed by the [1,5] (∆Hq ) 25.7 kcal
mol-1) and the [1,3] B shift (∆Hq ) 27.4 kcal mol-1). The
[1,7] B transition structure TS3 can be rationalized as a favorable
Mo¨bius aromatic system with eight electrons in eight active
orbitals connected in a circle with one phase inversion. This
circle consists of the seven p orbitals on carbon and the boron

1042 J. Am. Chem. Soc., Vol. 120, No. 5, 1998
GridneV et al.
only by 1.7 kcal mol-1. Several factors can account for such a
difference, for instance the presence of a transition metal atom
in 1 or inevitable endo orientation of the dipropylboryl sub-
stituent in 1. Studies on the effects governing the selectivity
of the sigmatropic migrations of organometallic groups in
cyclononatetraene compounds are well underway in our labo-
ratories.
Experimental Section
All experiments were performed under dry argon atmosphere. 1H,
13C, 11B, and 119Sn NMR spectra were recorded using a Bruker AMX-
400 spectrometer. 1H and 13C 2D EXSY NMR spectra were acquired
using a NOESYTP pulse program, slightly modified in the case of 13
C
EXSY to allow the 1H decoupling during acquisition. Areas of cross-
peaks and diagonal peaks were obtained by volume integration of
appropriate voxels surrounding the peaks.
Trimethyl(cyclohepta-1,3,5-trien-7-yl)tin (6). To a 0.25 M solution
of Me3SnLi in THF (110 mL, 27.5 mmol) at -60 °C was added
tropylium bromide (1.28 g, 7.5 mmol) in one portion, and the solution
was then poured into water (200 mL). The organic layer was separated,
dried over Na2SO4, filtered, and concentrated in vacuo. The residue
was purified by column chromatography on silica gel with hexane as
the eluent, affording 0.36 g (5.1%) of 6 as a yellow oil. 1H NMR
(400 MHz, CDCl3, 298 K): δ ) 0.10 (s, 9H, CH3, 2J(H,119Sn) ) 51.2
Hz); 2.79 (t, 1H, H-4, 3J(H,H) ) 9.0 Hz, 2J(H,119Sn) ) 63.3 Hz); 5.40
(dd, 2H, H-1, 3J(H,H) ) 9.3 Hz, 3J(H,H) ) 9.0 Hz, 3J(H,119Sn) ) 38.0
Hz); 5.76 (ddd, 2H, H-2, 3J(H,H) ) 9.3 Hz, 3J(H,H) ) 3.5 Hz, 4J(H,H)
Figure 12. 1H NMR EXSY spectrum of compound 2 (400 MHz, CD2-
Cl2-CDCl3-CCl4, 323 K): mixing time 0.5 s, initial delay 1 s, number
of scans 256, a matrix 512 × 256 was aquired by the TPPI technique
and zero-filled to the size 1024 × 1024.
4
3
) 2.8 Hz, J(H,119Sn) ) 28.9 Hz); 6.20 (dd, 2H, H-3, J(H,H) ) 3.5
Hz, 4J ) 2.8 Hz, 5J(H,119Sn) ) 23.4 Hz). 13C NMR (100 MHz, CDCl3,
298 K): δ ) -6.32 (CH3, 1J(13C,119Sn) ) 305.5 Hz); 33.29 (C-4,
1J(13C,119Sn) ) 328.1 Hz); 124.59 (C-2, 4J(13C,119Sn) ) 31.2 Hz); 129.62
(C-1, 2J(13C,119Sn) ) 38.5 Hz); 133.26 (C-3). 119Sn NMR (149 MHz,
CDCl3, 298 K): δ ) 12.41.
p AO; the latter has a nodal plane (phase inversion). This leads
to electron delocalization which causes a reduction of the
positive charge (NBO) on boron (+0.6 in TS3) vs 2a (+1.0)
and bond length equalization from C1 to C7, but not between
these two (note the long C1-C7 bond, 1.673 Å, Figure 6). In
contrast, TS4 and TS5 are strongly alternating and considerably
higher in energy (Figure 6, Table 1).
NMR Experiments on the Exchange Reaction between 6 and
Dipropylchloroborane. Pr2BCl (0.1 g, 0.8 mmol) was added to a
solution of 0.15 g (0.6 mmol) of trimethylcycloheptatrienyltin (6) in
0.5 mL of CDCl3. NMR spectra taken immediately at room temperature
showed 50% conversion of 6 into 2 (see below). Further storage of
the sample at room temperature or heating to 343 K for several hours
led to complete conversion of 6; however, the amount of 2 in the
reaction mixture did not increase. Instead, equal amounts of chlo-
rodimethylcycloheptatrienyltin (8) and tetramethylstannane were de-
tected by 119Sn NMR. The structure of compound 8 was confirmed
by 2D chemical shift correlation experiments. 1H NMR (400 MHz,
Conclusions
The synthesis as well as the study of the dynamic behavior
of a boron cycloheptatrienyl (2) allows the analysis of several
important problems. Compound 2 is the first organometallic
derivative of cycloheptatriene in which the norcaradiene form
predominates. Many compounds with the organometallic group
in the 7-position are known to exist exclusively as cyclohep-
tatrienes. Ab initio computations show that the donor-acceptor
interaction between the boron 2p AO and the Walsh orbital of
the cyclopropane is most effective in the exo-norcaradiene form
2b. Our results clearly demonstrate that the boron 2p AO is
necessary for stabilizing the norcaradiene form. The deuteri-
opyridine complex 9 only exists in its cycloheptatrienyl form.
Note that in the iron tricarbonyl complex 1 the corresponding
norcaradiene is not detected3 owing to the effect of the iron
tricarbonyl forcing the dipropylboryl group into the endo
position.
The mechanism of boron migrations in 2a has also been
clarified. Both experimental and computational data indicate
that the [1,7] B shift is the fastest borotropic rearrangement in
2a. Since the same type of sigmatropic boron shift has
previously been found in the iron tricarbonyl complex 1, one
can conclude that a suprafacial [1,7] B shift is a facile process,
regardless of the presence of the iron tricarbonyl group. Orbital
control but more importantly charge distribution and geometric
factors are dominating. However, the data concerning two other
possible boron shifts in cycloheptatrienylboranes are not entirely
conclusive. While a relatively slow [1,3] B migration was
detected in 1 ([1,5] B shift was not observed), computational
data for 2a energetically favor [1,5] B over [1,3] B shifts, but
2
CDCl3, 298 K): δ ) 0.62 (s, 6H, CH3, J(H,119Sn) ) 57.2 Hz); 3.49
3
2
(t, 1H, H-4, J(H,H) ) 8.7 Hz, J(H,119Sn) ) 77.4 Hz); 5.84 (t, 2H,
H-1, J(H,H) ) 9.1 Hz, J(H,119Sn) ) 70.4 Hz); 6.13 (m, 2H, H-2);
3
3
6.56 (m, 2H, H-3). 13C NMR (100 MHz, CDCl3, 298 K): δ ) 3.9
(CH3, J(13C,119Sn) ) 256.5 Hz); 41.47 (C-4); 125.70 (C-2); 128.69
1
(C-1, J(13C,119Sn) ) 34.3 Hz); 134.52 (C-3).
2
Dipropyl(cyclohepta-1,3,5-trien-7-yl)borane (2a) in Equilibrium
with 7-(Dipropylboryl)norcaradiene (2b). Pr2BCl (6.2 g, 48 mmol)
was added in one portion to trimethylcycloheptatrienyltin (6) (1.5 g, 6
mmol). The reaction mixture was stirred at room temperature for 1 h,
and then the volatile products were removed in vacuo. An additional
portion of Pr2BCl (6.2 g, 48 mmol) was added to the reaction mixture,
and the above treatment was repeated. The residue was distilled in
vacuo, and a fraction collected at 45-55 °C (0.5 mmHg) afforded 0.3
g (27%) of 2 as a colorless extremely air-sensitive liquid. 1H NMR
(400 MHz, CCl4-CDCl3-CD2Cl2 (13:27:60), 298 K): δ ) 0.28 (t,
1H, H-4, 3J(H,H) ) 5.5 Hz); 0.96 (t, 6H, CH2CH2CH3, 3J(H,H) ) 7.3
Hz); 1.41 (t, 4H, CH2CH2CH3, 3J(H,H) ) 7.8 Hz); 1.58 (m, 4H,
3
CH2CH2CH3); 3.28 (m, 2H, H-1), 6.11 (dd, 2H, H-3, J(H,H) ) 6.5
Hz, 3J(H,H) ) 3.0 Hz); 6.31 (m, 2H, H-2). 13C NMR (100 MHz,
CCl4-CDCl3-CD2Cl2 (13:27:60), 298 K): δ ) 16.45 (Pr); 18.02 (Pr);
23.01 (br, C-4); 30.12 (br, Pr); 52.55 (br, C-1); 123.65 (C-3); 128.05
(C-2). 11B NMR (128 MHz, CCl4-CDCl3-CD2Cl2 (13:27:60), 298
K): δ ) 76.3. When 2 was dissolved in deuteriopyridine, quantitative
formation of complex 9 was observed. 1H NMR (400 MHz, CCl4-
CDCl3-CD2Cl2 (13:27:60), 298 K): δ ) 1.10 (t, 1H, H-4, 3J(H,H) )

Synthesis and Properties of Cycloheptatrienyl(dipropyl)borane
J. Am. Chem. Soc., Vol. 120, No. 5, 1998 1043
3
6.3 Hz); 0.6-1.4 (m, 14H, Pr); 5.08 (dd, 1H, H-1, J(H,H) ) 8.9 Hz,
vibrational frequencies at B3LYP/6-31G*. Single-point energies were
evaluated using a standard 6-311+G* basis set;37 final energies thus
refer to B3LYP/6-311+G*//B3LYP/6-31G* + ZPVE, unless noted
otherwise. Standard notation is used, i.e., “//” means energy computed
at//geometry.
3J(H,H) ) 6.4 Hz); 5.43 (dd, 1H, H-6, J(H,H) ) 9.1 Hz, J(H,H) )
3
3
3
3
6.1 Hz); 6.09 (dd, 1H, H-2, J(H,H) ) 9.2 Hz, J(H,H) ) 4.8 Hz);
6.20 (dd, 1H, H-5, 3J(H,H) ) 9.3 Hz, 3J(H,H) ) 5.2 Hz); 6.49 (m, 2H,
H-3,4). 13C NMR (100 MHz, CCl4-CDCl3-CD2Cl2 (13:27:60), 298
K): δ ) 16.91 (Pr); 17.77 (Pr); 28.76 (br, Pr); 34.51 (br, C-7); 124.45
(C-5); 124.51 (C-1); 124.90 (C-2); 125.96 (C-6); 129.04 (C-3); 129.36
(C-4). 11B NMR (128 MHz, pyridine-d5, 313 K): δ ) 10.4.
Computational Methods. Geometries of all stationary points were
optimized using analytical energy gradients of self-consistent field29
and density functional theory (DFT).30 The latter utilized Becke’s three-
parameter exchange-correlation functional31 including the nonlocal
gradient corrections described by Lee-Yang-Parr (LYP),32 as imple-
mented in the Gaussian 94 program package.33 All geometry optimiza-
tions were performed using the 6-31G* basis set.34 Previous studies
showed convincingly that DFT methods in conjunction with double-ú
basis sets are quite suitable for studying transition states of pericyclic
reactions35 and electron-deficient rearrangements.36
NMR chemical shifts were computed with the gauge-invariant atomic
orbital (GIAO) approach38 in conjunction with the B3LYP func-
tional.31,32 Since a flexible valence basis plus polarization functions is
needed for accurate chemical shifts, we utilized the 6-311+G** basis
for chemical shift calculations on the B3LYP/6-31G* geometries. The
absolute shifts (σ) of tetramethylsilane (TMS) and the etherate of boron
trifluoride (Et2O‚BF3) were used to compute the relative chemical shifts
δ ) σreference - σcompound
.
Acknowledgment. This work was financially supported by
the Russian Foundation for Basic Research (Project No. 97-
03-32714). An A. v. Humboldt Research Fellowship for I.D.G.
is gratefully acknowledged. P.R.S. thanks the Fonds der
Chemischen Industrie for financial support (Liebig-Fellowship),
Professor Armin de Meijere for his enouragement, and the
Gesellschaft fu¨r wissenschaftliche Datenverarbeitung (GWGD)
as well as the Regional Rechenzentrum Niedersachsen (RRZN
Hannover) for generous allotment of computer time.
Thermodynamic corrections (228 and 298 K) and zero-point
vibrational energy corrections (ZPVE) were derived from analytical
(29) Pulay, P. In Modern Theoretical Chemistry; Schaefer, H. F., Ed.;
Plenum: New York, 1977; Vol. 4, p 153.
(30) Parr, R. G.; Yang, W. Density Functional Theory of Atoms and
Molecules; Oxford University Press: New York, 1989.
(31) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
JA9724699
(32) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(33) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G.
A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski,
V. G.; Ortiz, J. V.; Foresman, J. B.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian 94, ReVision B.3; Gaussian,
Inc.: Pittsburgh, PA, 1995.
(36) For a review of applications of DFT to electron-deficient rearrange-
ments involving carbon see: Bettinger, H. F.; Schleyer, P. v. R.; Schreiner,
P. R.; Schaefer, H. F. In Modern Electronic Structure Theory and its
Applications to Organic Chemistry; Davidson, E. L., Ed.; World Scientific
Press: River Edge, NJ, in press.
(37) (a) Spitznagel, G. W.; Clark, T.; Chandrasekhar, J.; Schleyer, P. v.
R. J. Comput. Chem. 1982, 3, 363. (b) Clark, T.; Chandrasekhar, J.;
Spitznagel, G. W.; Schleyer, P. v. R. J. Comput. Chem. 1983, 4, 294.
(38) Wolinski, K.; Hinton, J. F.; Pulay, P. J. Am. Chem. Soc. 1990, 112,
8251.
(34) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213.
(35) Wiest, O.; Houk, K. N. Top. Curr. Chem. 1996, 183, 1.