Communications
vided the expected ring-opened product 3e (entry 5) along
The HOMO of free anion C shows high electron density
on the boron atom delocalized into the NHC ring
(Figure 1c,d). Just as the radical D is loosely analogous to a
benzyl radical, the anion C is loosely analogous to a benzyl
anion. As such, the anion C is structurally different from both
the phosphine boryl anion A[16] and NHC boryl anion B.
In summary, we have generated an unsubstituted NHC
boryl anion in situ by reduction of a readily available boryl
iodide with LDBB. The anion can be trapped with a diverse
range of electrophiles to give an assortment of new types of
NHC boranes that would be difficult or impossible to access
with existing methods. These compounds form a very small
family of tricoordinate boryl anions. Its unusual features in
reactions with electrophiles (for example, para addition to
benzonitrile and substitution of adamantyl iodide) warrant
additional mechanistic study, as radical mechanisms might be
possible. The availability of the new classes of NHC borane
products (acyl boranes, a-boryl alcohols, etc.) opens the door
to the study of their properties and chemistry.
with the double adduct 5 (30%). Compound 5 presumably
arises from deprotonation of 3e on the imidazolyl ring.[12]
Accordingly, the imidazolylidene protons on NHC borane
complexes must be weakly acidic; a property that merits
additional study.
Interestingly, benzonitrile did not give the standard
1,2-addition product expected from highly reactive anions.
The para-substituted product 3 f (entry 6) was isolated
instead. This product resembles those observed by Imamoto
in reactions of anion A.[6]
Alkyl halides delivered the corresponding B-alkyl deriv-
atives 3g–l (entries 7–13) in useful yields whereas hexafluoro-
benzene provided the addition/elimination product 3n
(entry 15). Again, the observations were unusual, with less
reactive halides like butyl chloride, isopropyl iodide, and
adamantyl iodide (entries 10, 12–13) generally providing
better yields than more reactive ones like butyl iodide and
crotyl chloride (entries 8–9). Interestingly, the reaction with
dichloromethane produced the methylated product 3m, not
the chloromethylated product (entry 14).
Received: July 9, 2010
Published online: September 17, 2010
Most of the products in Table 1 are members of new
classes of substituted NHC boranes that would be difficult to
make by the current method of hydroboration and subse-
quent complexation. The free acyl boranes needed to make
3a and 3b, for example, are rare and reactive types of
molecules.[13] The unsaturated boranes needed to make 3g,
3h, and 3j would doubtless hydroborate themselves. The a-
boryl alcohols 3c and 3d and the isopropyl borane 3k are all
formally products of hydroboration with regioselectivity
opposite to a classical hydroboration reaction. And products
3e and 3 f are formally analogous to hydroboration products
of an enol and a benzyne, respectively.
To complement the experiments, we conducted DFT
calculations on the starting iodide 2 and the intermediate
radical D and free anion C.[14] Figure 1 shows diagrams of the
lowest unoccupied molecular orbital (LUMO) of iodide 2
(Figure 1a), the spin density radical on D (Figure 1b), the
highest occupied molecular orbital (HOMO) of anion C
(Figure 1c), and the electrostatic potential mapped onto the
electronic density on C (Figure 1d).
Keywords: anions · boron · carbenes · nucleophilic addition ·
nucleophilic substitution
.
[1] a) Y. Wang, B. Quillian, P. Wei, C. S. Wannere, Y. Xie, R. B.
King, H. F. Schaefer, P. v. R. Schleyer, G. H. Robinson, J. Am.
Wei, Y. Xie, C. S. Wannere, R. B. King, H. F. Schaefer, P. v. R.
3299; c) T. K. Wood, W. E. Piers, B. A. Keay, M. Parvez, Angew.
[2] a) S.-H. Ueng, A. Solovyev, X. Yuan, S. J. Geib, L. Fensterbank,
E. Lacꢀte, M. Malacria, M. Newcomb, J. C. Walton, D. P. Curran,
Makhlouf Brahmi, L. Fensterbank, E. Lacꢀte, M. Malacria, Q.
Parvez, Chem. Commun. 2009, 5147 – 5149.
4253; b) D. McArthur, PhD Thesis, University of Bristol, 2009.
[4] Divalent boryl anions with six valence electrons and one
negative charge are likewise rare. M. Yamashita, K. Nozaki,
The calculated LUMO of 2 is mostly situated on the NHC
À
fragment (Figure 1a), not the s* orbital of the B I bond. The
calculations suggest that the radical anion resulting from
injection of an electron into the LUMO of 2 does not lead to a
[5] a) M. Yamashita, Angew. Chem. 2010, 122, 2524 – 2526; Angew.
115; d) H. Braunschweig, M. Burzler, R. D. Dewhurst, K.
Amine-boryl anions have also been used. See: T. D. Parsons,
[7] H. Braunschweig, C.-W. Chiu, K. Radacki, T. Kupfer, Angew.
Chem. 2010, 122, 2085 – 2088; Angew. Chem. Int. Ed. 2010, 49,
2041 – 2044.
À
stable radical anion. Instead, B I bond cleavage occurs
simultaneously and boryl radical D is formed directly. This
can also be inferred by the reduction wave width and
variation of peak potential with scan rate.[15]
The calculated singly occupied molecular orbital (SOMO)
of D is similar to that of other NHC boryl radicals
(Figure 1b),[2,8d] with a spin density partially on the boron
atom and partially delocalized throughout the NHC ring. The
radical D resists electrochemical reduction but is chemically
reduced by LDBB to C. The calculated electron affinity for
this reduction to the boryl anion is + 42.1 kcalmolÀ1. This may
explain why one needs a powerful reducing agent to generate
4.
[8] a) S.-H. Ueng, M. Makhlouf Brahmi, ꢂ. Derat, L. Fensterbank,
Ueng, C. Robert, M. D.-E. Murr, D. P. Curran, M. Malacria, L.
9168
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 9166 –9169