EPR studies of the generation, structure, and reactivity of n-heterocyclic carbene borane radicals

Article abstract of DOI:10.1021/ja909502q

N-Heterocyclic carbene boranes (NHC-boranes) are a new "clean" class of reagents suitable for reductive radical chain transformations. Their structures are well suited for their reactivity to be tuned by inclusion of different NHC ring units and by approp

Full text of DOI:10.1021/ja909502q

Published on Web 01/27/2010
EPR Studies of the Generation, Structure, and Reactivity of
N-Heterocyclic Carbene Borane Radicals
John C. Walton,*,† Malika Makhlouf Brahmi,‡ Louis Fensterbank,‡
Emmanuel Lacoˆte,*,‡ Max Malacria,‡ Qianli Chu,§ Shau-Hua Ueng,§
Andrey Solovyev,§ and Dennis P. Curran*,§
School of Chemistry, EaStChem, UniVersity of St. Andrews, St. Andrews, Fife KY16 9ST, United
Kingdom, UPMC UniVersity Paris 06, Institut parisien de chimie mole´culaire (UMR 7201),
C. 229, 4 place Jussieu, 75005 Paris, France, Department of Chemistry, UniVersity of
Pittsburgh, Pittsburgh, PennsylVania 15260
Received November 9, 2009; E-mail: jcw@st-andrews.ac.uk; emmanuel.lacote@upmc.fr; curran@pitt.edu
Abstract: N-Heterocyclic carbene boranes (NHC-boranes) are a new “clean” class of reagents suitable
for reductive radical chain transformations. Their structures are well suited for their reactivity to be tuned
by inclusion of different NHC ring units and by appropriate placement of diverse substituents. EPR spectra
were obtained for the boron-centered radicals generated on removal of one of the BH3 hydrogen atoms.
This spectroscopic data, coupled with DFT computations, demonstrated that the NHC-BH2• radicals are
planar π-delocalized species. tert-Butoxyl radicals abstracted hydrogen atoms from NHC-boranes more
than 3 orders of magnitude faster than did C-centered radicals, although the rate decreased markedly for
sterically shielded NHC-BH3 centers. Combinations of two NHC-boryl radicals afforded 1,2-bis-
NHC-diboranes at rates which also depended strongly on steric shielding. The termination rate increased
to the diffusion-controlled limit for sterically unhindered NHC-boryls. Bromine atoms were rapidly transferred
to imidazole-based NHC-boryl radicals from alkyl, allyl, and benzyl bromides. Chlorine-atom abstraction
was, however, much less efficient and only observed for sterically unhindered NHC-boryls reacting with
allylic and benzylic chlorides. For an NHC-borane containing a bulky thexyl substituent at boron, the tertiary
H atom of the thexyl group was selectively removed. The resulting ꢀ-boron-containing alkyl radical rapidly
underwent ꢀ scission of the B-C bond with production of an NHC-boryl radical and an alkene.
cyclization reactions.6 Although organoboranes play an impor-
tant role in delivering carbon radicals in many radical-mediated
Introduction
Group 14 metal hydrides, especially tin1 and silicon hydrides,2
have proved to be very useful reagents for reductive chain
transformations of a variety of organic compounds. Efforts to
enlarge the scope of such reductions, and decrease their
environmental impact,3 have steered research toward analogous
reactions of Group 13 metal hydrides. For example, it has been
shown that HInCl2 is a mild, chemoselective reagent for reducing
aromatic aldehydes and halides,4 as well as organic azides,5 and
that HGaCl2 is effective for mediating reductive radical-based
processes,7 boranes did not appear to be promising as H-atom
donors in radical reductions because of their strong boron-
hydrogen bonds.8 Computations at the CBS-4 level provided
large BDEs of 105.5 and 100.3 kcal mol-1 for H2B-H and
H3BBH2-H, respectively.9
It has been known for some time, however, that the H atoms
of boranes ligated to amines and phosphines are more labile.10
A wide variety of amine- and phosphine-boryl radicals have
been generated by H-atom abstraction from the corresponding
ligated boranes and take part in rearrangement, abstraction, and
addition reactions.11 The B-H bonds of alkyl amine-boranes
are still strong. Computations indicate DHo(B-H) values of
103 and 101 kcal mol-1 (MP3/6-31G**) for H3NBH3 and
MeNH2BH3, respectively,10d and 101.1 kcal mol-1 (CBS-4) for
† University of St. Andrews.
‡ UPMC.
§ University of Pittsburgh.
(1) (a) In Radicals in Organic Synthesis; Renaud, P., Sibi, M. P., Eds.;
Wiley-VCH: Weinheim, 2001. (b) In Radicals in Synthesis; Gansa¨uer,
A., Ed.; Springer-Verlag: Berlin, 2006; Vols. 263 and 264.
(2) Chatgilialoglu, C. Organosilanes in Radical Chemistry; Wiley: Chich-
ester, 2004.
(3) (a) Baguley, P. A.; Walton, J. C. Angew. Chem., Int. Ed. 1998, 37,
3072–3082. (b) Studer, A.; Amrein, S. Synthesis 2002, 835–849. (c)
Darmency, V.; Renaud, P. Top. Curr. Chem. 2006, 263, 71–106.
(4) (a) Inoue, K.; Sawada, A.; Shibata, I.; Baba, A. Tetrahedron Lett.
2001, 42, 4661–4663. (b) Inoue, K.; Sawada, A.; Shibata, I.; Baba,
A. J. Am. Chem. Soc. 2002, 124, 906–907. (c) Takami, K.; Yorimitsu,
H.; Oshima, K. Org. Lett. 2002, 4, 2993–2995. (d) Takami, K.;
Mikami, S.; Yorimitsu, H.; Shinokubo, H.; Oshima, K. Tetrahedron
2003, 59, 6627–6635.
(6) (a) Takami, K.; Mikami, S.; Yorimitsu, H.; Shinokubo, H.; Oshima,
K. J. Org. Chem. 2003, 68, 6627–6631. (b) Mikami, S.; Fujita, K.;
Nakamura, T.; Yorimitsu, H.; Shinokubo, H.; Matsubara, S.; Oshima,
K. Org. Lett. 2001, 3, 1853–1855.
(7) Darmency, V.; Renaud, P. In Radicals in Synthesis I; Gansauer, A.,
Ed.; Springer-Verlag: Berlin, 2006; Vol. 263, pp71-106.
(8) Ryu, I.; Uehara, S.; Hirao, H.; Fukuyama, T. Org. Lett. 2008, 10,
1005–1008.
(9) Rablen, P. R. J. Am. Chem. Soc. 1997, 119, 8350–8360. See
also: Barreto, P. R. P.; Vilela, A. F. A.; Gargano, R. Int. J. Quantum
Chem. 2005, 103, 656–684.
(5) Benati, L.; Bencivenni, G.; Leardini, R.; Nanni, D.; Minozzi, M.;
Spagnolo, P.; Scialpi, R.; Zanardi, G. Org. Lett. 2006, 8, 2499–2502.
9
2350 J. AM. CHEM. SOC. 2010, 132, 2350–2358
10.1021/ja909502q  2010 American Chemical Society

Reactivity of NHC-Boryl Radicals
A R T I C L E S
Me3NBH3.9 In a recent paper,12 however, H-atom abstraction
from N-heteroaryl-boranes was found to be very fast and the
corresponding DHo(B-H) values were much smaller and in the
range 67-77 kcal mol-1. Applications of amine-boranes to
reductive chain transformations have so far been rather limited.
A theoretical study of a series of donor-acceptor com-
plexes of boron revealed that their B-H BDEs were greatly
reduced by complexation of the borane moiety with Lewis
bases.9 This finding led some of us to postulate that NHC’s
would be very effective partners of boron for this purpose
and that the resulting NHC-boranes could be serviceable
radical chain reducing agents. This prediction was recently
confirmed in that secondary xanthates were indeed reductively
deoxygenated by the stable NHC-boranes 1,3-bis(2,6-di-
isopropylphenyl)imidazol-2-ylidene borane, 1a, and 2-phenyl-
1,2,4-triazol-3-ylidene borane, 7 (Barton-McCombie reac-
tion).13 Furthermore, we obtained good evidence that the
mechanism involved chains propagated by NHC-boryl
radicals and suggested that the NHC-BH2-H BDE was as
low as 88 kcal mol-1. A preliminary EPR spectrum of the
radical derived from 1a provided key evidence for the
postulated mechanism.14
These findings signaled that NHC-boranes could be devel-
oped as a unique class of reagents capable of mediating a range
of radical processes, possibly including some not accessible to
tin and silicon hydrides. The modifiable structures of the NHC’s
should enable the reactivity and selectivity of the NHC-boranes
to be readily tuned for particular purposes. Accordingly, we
studied the reactivity of a representative set of NHC-boranes
mainly by EPR spectroscopy. We found that H atoms were
selectively removed by tert-butoxyl radicals in a very rapid
process, giving previously unknown NHC-boryl radicals. EPR
spectra and DFT calculations provided important structural
information about this new class of radicals. The termination
and atom-transfer reactions of these boron-centered species were
Figure 1. Structures of the N-heterocyclic carbene boranes studied. The
corresponding NHC-BH2• radicals are designated 1a•, etc.
strongly influenced by steric shielding from substituents pendant
from the N atoms of the NHC rings.
Results and Discussion
Generation and EPR Spectroscopic Characterization of
NHC-Boryl Radicals. The set of NHC-boranes shown in
Figure 1 was chosen to acquaint us with the characteristic
reactions of these reagents and to probe the effect of key
structural features on their reactivity. A preliminary analysis of
the EPR spectrum of the radical 1a• derived from 1a has already
been communicated.14 Compounds 1a and 2-6 are imidazole-
based NHC-boranes with 6 π electrons in the NHC moiety.
This series contains both alkyl and aryl substituents on the N
atoms, with consequent variation in the steric shielding around
the BH3 group. In addition, compound 4 has a fused aryl ring
in place of the imidazole CdC bond, while 6 has a saturated
ring fused to this bond. Compound 1b was chosen as an example
of an imidazolidin-2-ylidene, lacking the 5-center 6-electron π
delocalization as a stabilizing factor. Likewise, tetrahydropyr-
rolotriazol-3-ylidene 7 was selected to exemplify the reactivity
of triazole-based NHC-boranes. The effect of alkyl substituents
on the B atom was investigated by means of NHC-thexyl-
borane 8.
(10) (a) Baban, J. A.; Marti, V. P. J.; Roberts, B. P. J. Chem. Soc., Perkin
Trans. 2 1985, 1723–1733. (b) Paul, V.; Roberts, B. P. J. Chem. Soc.,
Perkin Trans. 2 1988, 1183–1193. (c) Baban, J. A.; Roberts, B. P.
J. Chem. Soc., Perkin Trans. 2 1988, 1195–1200. (d) Kirwan, N.;
Roberts, B. P. J. Chem. Soc., Perkin Trans. 2 1989, 539–550. (e) Dang,
H. S.; Roberts, B. P. Tetrahedron Lett. 1992, 33, 6169–6172. (f) Dang,
H. S.; Diart, V.; Roberts, B. P.; Tocher, D. A. J. Chem. Soc., Perkin
Trans. 2 1994, 1039–1045. (g) Roberts, B. P.; Steel, A. J. J. Chem.
Soc., Perkin Trans. 2 1994, 2411–2422. (h) Lucarini, M.; Pedulli, G. F.;
Valgimigli, L. J. Org. Chem. 1996, 61, 1161–1164. (i) Barton,
D. H. R.; Jacob, M. Tetrahedron Lett. 1998, 39, 1331–1334. (j) Sheeler,
B.; Ingold, K. U. J. Chem. Soc., Perkin Trans. 2 2001, 480–486.
(11) Reviews of amine and phosphine boranes: (a) Carboni, B.; Monnier,
L. Tetrahedron 1999, 55, 1197–1248. (b) Carboni, B.; Carreaux, F.
In Science of Synthesis, Organometallics: Boron Compounds; Kauf-
mann, D. E., Matteson, D. S., Eds.; Verlag: Stuttgart, 2004; Vol. 6,
pp 455-484. (c) Gaumont, A. C.; Carboni, B. In Science of Synthesis,
Organometallics: Boron Compounds; Kaufmann, D. E., Matteson,
D. S., Eds.; Verlag: Stuttgart, 2004; Vol. 6, pp 485-512.
(12) (a) Laleve´e, J.; Blanchard, N.; Chany, A.-C.; Tehfe, M.-A.; Allonas,
X.; Fouassier, J.-P. J. Phys. Org. Chem. 2009, 22, 986–993. (b) Thefe,
M. A.; Makhlouf Brahmi, M.; Fouassier, J.-P.; Curran, D. P.; Malacria,
M.; Fensterbank, F.; Lacoˆte, E.; Laleve´e, J. Submitted for publication.
(13) Ueng, S.-H.; Makhlouf Brahmi, M.; Derat, E.; Fensterbank, L.; Lacoˆte,
E.; Malacria, M.; Curran, D. P. J. Am. Chem. Soc. 2008, 130, 10082–
10083.
All these NHC-boranes are air- and moisture-stable white
solids that are readily prepared by deprotonation of the corre-
sponding imidazolium or triazolium salt and reaction of the in
situ generated NHC with borane. Compounds 1a and 7 are
known,13 whereas the other complexes are new. Procedures for
the preparations and characterization data for all the new
complexes are in the Supporting Information.
(14) Ueng, S.-H.; Solovyev, A.; Yuan, X.; Geib, S. J.; Fensterbank, L.;
Lacoˆte, E.; Malacria, M.; Newcomb, M.; Walton, J. C.; Curran, D. P.
J. Am. Chem. Soc. 2009, 131, 11256–11262.
9
J. AM. CHEM. SOC. VOL. 132, NO. 7, 2010 2351

A R T I C L E S
Walton et al.
Table 1. Experimental and Computed hfs for NHC-Boryl and
with all the magnetic nuclei of the BH2 group, the N-CH3
groups and aromatic H atoms should be significant (Table 1),
thus giving rise to a maximum of 3780 resonance lines. In view
of the complex fine structure, it is not surprising that most of
the spectral lines of this NHC-boryl radical and of the radicals
derived from 5 and 7 are not visible above the noise level. The
DFT-computed parameters for 4• and 7• are included in Table
1. Hydrogen-atom abstraction from imidazolidin-2-ylidene 1b,
lacking the 5-center 6-electron π delocalization, also took place
efficiently, and the hfs of the radical were only slightly smaller
in magnitude than those of 1a•.
It was important to establish if NHC-BH2• radicals could
be generated in sufficient quantity for EPR spectroscopic
detection when t-BuO• was replaced by a C-centered radical.
We prepared a solution of 1a and dilauroyl peroxide as the
source of undecyl radicals. However, only the undecyl radical
[a(2HR) ) 21.8, a(2Hꢀ) ) 27.5 G] was observed by EPR
spectroscopy on UV photolysis in the temperature range
300-340 K. Even with the sterically much less hindered
NHC-borane 3, no NHC-BH2• radicals were detected but only
undecyl radicals. It is evident that H-atom abstraction from
NHC-BH3 by t-BuO• is much faster than abstraction by n-alkyl
radicals. This accords well with the measured rate constant [4
× 104 M-1 s-1 at 301 K] for H-atom abstraction from 1a by a
sec-alkyl radical.14
The EPR spectra of amine-boryl radicals such as Me3N-
BH2• (and phosphine-boryl radicals) showed them to be σ-type
radicals, pyramidal at boron.10d,18 Comparison with the hfs of
the radicals from 1a, 2, and 3 shows NHC-BH2• radicals have
much smaller a(11B) and larger a(N) than Me3N-BH2• (Table
1). This indicates that the NHC-BH2• radicals are roughly
planar at boron and that the unpaired electron is delocalized
into the NHC ring. The structures of the NHC-boryl radicals
and their SOMOs were computed by DFT employing the
B3LYP functional.19 The structures of model A [N,N′-diphenyl-
imidazol-2-ylidene-BH2•, (9•)] and benzimidazole-derived radi-
cal 4•, optimized with the 6-31G(d) basis set, and their
associated SOMOs are shown in Figure 3.
Related Radicalsa
T/K or DFT
NHC-BH2• radical basis setb
a(HR)G
a(11B)G a(2N) G a(2H)G
a(other)G
1a ·
300
EPR-ii
270
EPR-iii -10.0
245 12.0
11.4
-10.1
11.4
7.3 4.0
5.1 2.3
7.6 4.0
5.3 3.1
6.7 4.1
5.2 2.3
2.8 2.0
1.0
-1.0
1.1
-1.2
1.2 2.5(6H)
c
1a ·
2 ·
d
e
model A (9 · )
3 ·
3 ·
4 ·
EPR-iii -10.2
EPR-iii
-2.0 2.8(6H)
-7.8
-2.5, -1.2,
3.3(6H)
1.2(4H)
1b ·
model B
7 ·
NHC-Mes2B•
Me3N-BH2•
295
10.4
-7.5
6.4 4.3
2.7 2.2
f
EPR-iii
EPR-iii
293
3.8(4H)
a(N) ) 1.4
g
-8.6,-8.4 4.0 4.3, 1.0
h
7.9 3.0
51.3 1.4(1N)
0.7 2.8(6H)
i
280
9.6
1.4(9H)
a All NHC-BH2• g factors ) 2.0028 ( 0.0005. Note that only the
magnitudes and not the signs of hfs can be derived from the EPR
spectra. b DFT computations with the B3LYP functional: geometries
optimized with the 6-31G(d) basis set. c Computations with the EPR-iii
basis set failed. d N,N′-Diphenylimidazol-2-ylidene-BH2•. e hfs of
Ph-ring
H atoms all <
(0.2 G. f N,N′-Diphenyl-4,5-dihydroimidazol-
2-ylidene-BH2•. g a(1H) ) 4.3, 2.7, -1.8, and -1.1. a(2Ho) ) -2.0.
a(2Hm) ) 0.7. a(Hp) ) -1.7 G. h 1,3-Dimethylimidazol-2-ylidene-
B(•)Mes2; data from Matsumoto et al.17 i Data from Paul et al.18
Di-tert-butyl peroxide (DTBP) is a convenient photochemical
source of tert-butoxyl radicals, which are very efficient abstrac-
tors of H atoms and are EPR “silent”. Deaerated solutions of
individual NHC-boranes and DTBP in tert-butylbenzene
solvent were irradiated in the resonant cavity of a 9 GHz EPR
spectrometer in the expectation that the initial t-BuO• radicals
would abstract hydrogen from the BH3 groups of the substrates,
thus generating the NHC-boryl radicals. Strong, well-resolved
spectra of the corresponding NHC-BH2• radicals were obtained
from 1a and 1b, while 2 and 3 gave somewhat weaker signals.
All the spectra were satisfactorily simulated using the Bruker
SimFonia and NIEHS WinSim software,15 and the resulting
hyperfine splittings (hfs) are listed in Table 1. The actual and
simulated spectra obtained from 2 are shown in Figure 2, and
the other spectra are in the Supporting Information.
The whole imidazole-BH2 unit in 9• and the benzim-
idazole-BH2 unit in 4• were found to be planar, and interest-
ingly, the two Ph rings in 9• were orthogonal to the NHC moiety
and hence out of conjugation. This is in good accord with the
absence of resolvable hfs from the aromatic H atoms in the EPR
spectra of the radicals 1a• and 2•. The C-B bond lengths of
1.498 and 1.500 Å in 9• and 4•, respectively, were shorter than
the C-B bonds of the corresponding NHC-BH3 precursors by
about 0.1 Å. Concomitantly, the C-N bonds were longer than
those of the precursors by about 0.03 Å. These structural changes
are analogous to those found for C-centered radicals in
comparison with their hydrocarbon precursors.20 The barrier to
internal rotation about the C-BH2 bond in the radical 3• was
computed to be 19.5 kcal mol-1, confirming the partial double-
bond character of this bond. The SOMO in 9• is an essentially
pure π system extending over the BH2 group and the imidazole
ring but, as expected from the EPR data, not to the Ph rings.
Similarly, in 4•, the SOMO extends to the whole benzimidazole
moiety. The NHC-boryl radicals are quite unlike the bent
σ-amine-boryl radicals but quite like the π-delocalized benzyl
The hyperfine splittings are entirely appropriate for NHC-boryl
radicals, and comparisons with the DFT-computed values for
individual NHC-BH2• radicals (Table 1) provide further support
for the radicals’ identities.16 Furthermore, good correspondence
between our hfs and those of the 1,3-dimethylimidazol-2-
ylidene-BMes2 radical, obtained by Matsumoto et al. by
reduction of the corresponding cation,17 was also observed
(Table 1). Thus, t-BuO• radicals efficiently abstracted H atoms
from a variety of NHC-boranes to generate NHC-BH2•
radicals.
ΝΗC-ΒΗ3 + t-Βuꢀ• f ΝΗC-ΒΗ2• + t-ΒuꢀΗ
(1)
With NHC-borane 4, an extremely weak EPR spectrum
showing only a few central lines was observed. The DFT
computations indicated that interaction of the unpaired electron
(15) Duling, D. R. J. Magn. Reson. B 1994, 104, 105–110. The minor
contributions to the EPR spectra from the 10B isotopomers were
omitted.
(16) The DFT-computed 11B hfs depended strongly on the basis set varying
from 13.6 (6-31G(d) basis set) to 4.9 G (6-311G(d) basis set) (See
Table S1 in the Supporting Information).
(19) Frisch, M. J.; et al. Gaussian 03, Revision A.1; Gaussian, Inc.:
Pittsburgh, PA, 2003.
(17) Matsumoto, T.; Gabba¨ı, F. P. Organometallics 2009, 28, 4252–4253.
(18) Paul, V.; Roberts, B. P.; Robinson, C. A. S. J. Chem. Res. (S) 1988,
264–265.
(20) See, for example: (a) Pacansky, J.; Dupuis, M. J. Chem. Phys. 1979,
71, 2095–2098. (b) Pacansky, J.; Schubert, W. J. Chem. Phys. 1982,
76, 1459–1466.
9
2352 J. AM. CHEM. SOC. VOL. 132, NO. 7, 2010

Reactivity of NHC-Boryl Radicals
A R T I C L E S
Figure 2. EPR (9.5 GHz) spectrum of the NHC-BH2• radical 2•. (Top) First-derivative experimental spectrum at 270 K in t-BuPh solution. (Bottom)
Computer simulation with parameters noted in Table 1.
Figure 4. Allyl-type radicals derived from 6a,b.
Similar spectra were obtained throughout the temperature
range 250-300 K. At first glance this suggests the tert-butoxyl
radicals abstract the H atoms more rapidly from the allylic
positions than from the BH3 group. However, the great
persistence of 10a and 10b could lead to masking of any
NHC-BH2• radicals. Radicals 10a,b are strongly resonance
stabilized but owe their persistence to steric shielding. Sterically
shielded neutral boronium radicals, based on the 9-bora-9,10-
dihydroanthracene scaffold with substituted 2,2′-bipyridyl ligands,
had significant spin density on boron and were also found to
be persistent.21
Figure 3. DFT-computed structures and SOMOs of NHC-BH2• radicals.
radical. Similar delocalized electronic structures are expected
for imidazolylidene, triazolylidene, and thiazolylidene NHC-boryl
species. Consequently, NHC-boryl radicals should possess
significant resonance delocalization, and therefore, they will be
thermodynamically stabilized. This will be a major factor in
lowering the transition-state energy of H-atom transfer reactions
of NHC-boranes.
When a solution of NHC-borane 6a and DTBP was
photolyzed in the EPR cavity, a strong spectrum of a very
persistent radical was obtained. The spectrum did not decay in
21 h in the dark, although it degraded in a few minutes if UV
photolysis was continued. Analysis and simulation gave the
following results: a(2H) ) 0.91, a(2H) ) 2.74, a(1H) ) 9.66,
a(2H) ) 9.75, a(1N) ) 10.50 G at 300 K. A similar persistent
spectrum having a(2H) ) 0.94, a(2H) ) 2.74, a(1H) ) 10.55,
a(2H) ) 9.68, a(1N) ) 10.41, a(1N) ) 0.95 G at 300 K was
obtained from the related cyclohexyl-substituted NHC-borane
6b. It is evident that these are not NHC-boryl radicals but rather
allyl-type radicals 10a,b derived by abstraction of the allylic H
atoms of the cyclohexane rings (Figure 4).
Combination Reactions of NHC-Boryl Radicals. The EPR
spectrum of the radical from NHC-borane 1a did not disappear
instantly when the UV light was shuttered, suggesting it had
some persistence. To learn more about this, samples of 1a and
DTBP in t-BuPh were prepared and degassed and the EPR
magnetic field was fixed at the maximum intensity of the radical
signal. The decay of the signal was then monitored as a function
of time on shuttering the UV photolysis beam. Several separate
samples were examined, and a typical curve is shown in Figure
5 for the decay at 300 K. The initial concentration of the
NHC-BH2• radical was determined from the average of the
double integrals of the full spectra before and after photolysis.
The decay could not be satisfactorily fitted by a first-order plot
(see Supporting Information), but good fit was obtained for
second-order decay having 2kt ) 9.0 × 106 M-1 s-1 (Figure 5).
The decay of this radical was monitored at three lower
temperatures, and the best-fit 2kt values were plotted in
(21) Wood, T. K.; Piers, W. E.; Keay, B. A.; Parvez, M. Chem. Commun.
2009, 5147–5149.
9
J. AM. CHEM. SOC. VOL. 132, NO. 7, 2010 2353

A R T I C L E S
Walton et al.
Scheme 1. Dimerization of NHC-Boryl Radicals
Figure 5. Decay of the EPR signal of radical 1a• as a function of time at
300 K.
Table 2. Combination Rate Constants for NHC-Boryl Radicalsa
steric energy,c
kcal mol-1
radical
T, K
[NHC-BH2•]i, Mb
2kt, M-1 s-1
d
1a•
1b•
2•
300
300
290
300
5.1 × 10-7
3.1 × 10-7
7.3 × 10-7
9.0 × 106
260.7 [83]
281.4 [95]
249.6 [62]
74.5 [17]
8.1 × 106
50 × 106
3•
∼5000 × 106
a In t-BuPh solution. b Initial concentration of NHC-BH2• radical.
c Computed by the MM2 method for the B3LYP/6-31G(d) optimum
structures: values for MM2-optimized structures in brackets. d Arrhenius
) 10.7, Et ) 1.94 kcal mol-1]24 and for t-Bu• radical termination
in n-heptane at 294 K 2kt ) 8 × 109 [log At/M-1 s-1 ) 11.7, Et
) 2.4 kcal mol-1].23a The sterically unhindered radical 3•
probably dimerizes at about this rate, but radical 1a• dimerizes
nearly 3 orders of magnitude more slowly and there is a much
higher energy barrier for formation of the B-B bond.
parameters: log(At/M-1 s-1) ) 11.9 ( 0.8, Et ) 6.9 ( 0.4 kcal mol-1
.
Arrhenius form (see Supporting Information). Linearity was
good (R2 ) 0.991), yielding Et ) 6.9 ( 0.4 kcal mol-1 and
log(At/M-1 s-1) ) 11.9 ( 0.8. The EPR spectral intensities were
poorer for the radicals 1b• and 2•; however, second-order decay
rate constants were obtained in the same way, at a single
temperature, for each radical. The individual resonance lines
of the EPR spectrum of radical 3• were too weak for satisfactory
decay curves to be obtained. However, it was apparent that the
signal from this radical disappeared the instant the UV light
was shuttered. This demonstrated that termination was extremely
rapid and comparable to that of small C-centered radicals. It is
probable therefore that termination is diffusion controlled or
nearly so. The decay rate constants are collected in Table 2,
including a typical diffusion rate constant for the radical 3•.
The decay curves established that terminations were second
order in the NHC-BH2• radical concentrations. Since dispro-
portionations are not possible for these radicals, their decay
processes must be dimerizations leading to the formation of 1,2-
bis-NHC-diborane derivatives such as 11 and 12, Scheme 1.
There are literature precedents for the stability of such com-
pounds.22 In one experiment with 1a at 300 K, photolysis was
prolonged for 30 min so that a significant amount of dimer built
up. Subsequent heating of this solution in the dark at 350 K
did not lead to any detectable NHC-boryl radical. These
observations suggest the dimerization process is not significantly
reversible in the accessible temperature range. The absence of
curvature in the Arrhenius plot was consistent with this
conclusion.
The DFT-computed structure for 1a• and the semiempirical
AM1 structure for 11 (Figure 6) give an idea of the substantial
steric shielding of the BH2 group by the 2,6-diisopropylphenyl
substituents. The “steric energies”, computed by the empirical
MM2 method, for radical structures optimized at the B3LYP/
6-31G(d) level are shown in Table 2 for comparison. These
SEs indicate that shielding is rather substantial for the radicals
from 1a, 1b, and 2. The rough correlation between the SE values
and the log 2kt data suggests that steric shielding is indeed the
main factor slowing the combination reactions of the
NHC-BH2• radicals. The radical 1b• lacks the imidazole ring
conjugation and hence the 5-center 6-electron π delocalization
possessed by the other species. Steric shielding of the radical
center is about the same as in the radical 1a•. The fact that the
2kt values for the radicals derived from 1a and 1b are rather
similar (Table 2) shows it is steric shielding rather than
electronic stabilization that controls the dimerization process.
Hydrogen Abstraction from NHC-Boranes. The success of
the EPR experiments showed that H-atom abstraction from the
NHC-boranes by tert-butoxyl radicals was very fast. To
measure the rate of this process, a substrate of comparable
reactivity toward t-BuO• was sought for competition experi-
ments. After some trials, we found that UV irradiation of
Most small C-centered radicals dimerize (and/or dispropor-
tionate) extremely rapidly at the diffusion-controlled limit.23 For
example, the measured 2kt for benzyl radical dimerization at
300 K in cyclohexane was 4.1 × 109 M-1 s-1 [log At/M-1 s-1
(22) Wang, Y.; Quillian, B.; Wei, P.; Wannere, C. S.; Xie, Y.; King, R. B.;
Schaefer, H. F.; Schleyer, P. V.; Robinson, G. H. J. Am. Chem. Soc.
2007, 129, 12412–12413.
(23) (a) Schuh, H.-H.; Fischer, H. HelV. Chim. Acta 1978, 61, 2130–2164.
(b) Schuh, H.-H.; Fischer, H. Int. J. Chem. Kinet. 1976, 8, 341–356.
Figure 6. Structures of radical 1a• and dimer 11.
9
2354 J. AM. CHEM. SOC. VOL. 132, NO. 7, 2010

Reactivity of NHC-Boryl Radicals
A R T I C L E S
Table 3. Rate Constants for H-Atom Abstraction from
NHC-Boranes and Related Moleculesa
BDE, [NHC-BH2•][i-PrOH]/
substrate kcal mol-1 [Me2C(•)OH][NHC-BH3] kH, M-1 s-1
radical
ref
b
1a
88
88
4 × 104
R2CH•
14
t-BuO• 1a
82
456
1440
e1.4 × 108 this work
e8.2 × 108 this work
t-BuO•
t-BuO•
t-BuO•
RCH2•
t-BuO• Bu3SnH
RCH2• Bu3SnH
2
3
2.6 × 109
5.4 × 107
2 × 105
this work
25a, 27
28
29
30
c
c
76
76
78
78
1,4-CHD
1,4-CHD
2.2 × 109
2.4 × 106
a Solution data at 295 K. b 1-Cyclobutyldodecyl radical.
c 1,4-Cyclohexadiene.
Nonetheless, these upper limits may not be very different from
the actual rate constants because rapid cross-termination with
t-BuO• radicals may predominate. In parallel work on the use
of NHC-boranes as co-initiators in radical polymerizations,
Laleve´e and co-workers used time-resolved laser flash photolysis
experiments to measure kH ) 9.5 × 107 M-1 s-1 for the reaction
of t-BuO• with 1a at 300 K.12b This compares favorably with
the upper limit measured by EPR experiments, kH e 1.4 × 108
M-1 s-1 (Table 3, second entry). For radical 3•, derived from
the unhindered NHC-borane 3, the termination rate constants
of the two radicals will be equal to within the experimental
error so the derived kH will be reliable.
Table 3 shows the kH values for the t-BuO• radical increase
in magnitude for the NHC-BH3 substrates from 1a to 2 to 3,
and this is in line with the decrease in steric shielding of the
BH3 group along this series. The magnitude of kH for t-BuO•
with 3 is comparable to that of the same radical with Bu3SnH.
These reactions are extremely fast, and H-atom abstraction
occurs on almost every encounter of the reactants in solution.
For abstraction from the NHC-borane 1a, the increased
magnitude of kH for the t-BuO• radical, as compared with the
C-centered radical, is particularly striking. The data for other
substrates show that kH values are always greater for t-BuO•
but by smaller factors. The BDE for Bu3SnH is 10 kcal mol-1
less than that of 1a. It is not surprising therefore that kH(1a) for
a C-centered radical is nearly 2 orders of magnitude slower than
kH(Bu3SnH) for a C-centered radical (Table 3). However, the
fact that kH for t-BuO• with 3 is comparable to that of the same
radical with Bu3SnH suggests the H abstraction by t-BuO• is
not simply controlled by the thermochemistry but that other
factors must play a part.
Figure 7. EPR spectrum showing 1a• and Me2C(•)OH from photoreaction
of 1a and propan-2-ol with DTBP. (Top) Experimental spectrum in t-BuPh
at 295 K. (Bottom) Computer simulation.
solutions containing an NHC-borane and propan-2-ol together
with DTBP gave rise to EPR spectra showing both the
NHC-boryl radical and the 2-hydroxypropan-2-yl radical. Stock
solutions of known concentrations of NHC-boranes 1a, 2, and
3 together with DTBP were prepared. To each was added known
amounts of propan-2-ol, and the EPR spectra were recorded
during photolysis. Figure 7 shows the spectra of the mixture of
1a• and Me2C(•)OH obtained from the reaction with 1a and
propan-2-ol. The ratio of the concentrations of the two radicals
was obtained from spectral simulations. For each NHC-borane,
spectra were obtained with three different concentrations of
propan-2-ol and the results were averaged.
kΗ
ΝΗCdΒΗ3 + t-Βuꢀ•
98 ΝΗC-ΒΗ2• + t-ΒuꢀΗ
kS
•
Μe2CΗꢀΗ + t-Βuꢀ• 9
Η8 Μe2CꢀΗ + t-ΒuꢀΗ
It can easily be shown that
kΗ/ksΗ ) [ΝΗC-ΒΗ2•][i-ΡrꢀΗ]/[Μe2C(•)ꢀΗ][ΝΗC-ΒΗ3]
(2)
provided the termination rates of the two radicals are the same.
The average values of the concentration ratios are listed in Table
3 along with data for some related H-atom abstractions. The
rate constant for H-atom abstraction from Me2CHOH is known
to be 1.8 × 106 M-1 s-1 at 295 K,25 and hence, kH values can
be derived for the three NHC-boranes. However, as shown
above, the 2kt values for the sterically congested NHC-BH2•
radicals from 1a and 2 are considerably smaller than 2kt for the
Me2C(•)OH radical, which is known to be ca. 4 × 109 M-1 s-1
at 300 K in hydrocarbon solvents.26 The measured ratios will,
therefore, lead only to upper limits for the corresponding kH
values.
Halogen-Atom Abstraction by NHC-Boryl Radicals. Group
14 metal-centered radicals readily abstract halogen atoms from
alkyl halides. We therefore investigated the propensity of
NHC-boryl radicals to abstract halogen atoms by generating
them in the presence of various organic halides and then
scrutinizing the EPR spectra. Known quantities of various alkyl
halides (usually 10 µL) were added to aliquots (0.2 mL) from
a stock solution of 3 (ca. 0.1 mM) and DTBP in tert-
butylbenzene. During UV photolyses at 300 K the EPR spectra
shown in Figure 8 were obtained.
(24) Lehni, M.; Schuh, H.; Fischer, H. Int. J. Chem. Kinet. 1979, 11, 705–
713.
Good quality spectra of the corresponding n- and tert-butyl
radicals were observed during UV photolyses with added
(25) (a) Paul, H.; Small, R. D.; Scaiano, J. C. J. Am. Chem. Soc. 1978,
100, 4520–4527. (b) Malatesta, V.; Scaiano, J. C. J. Org. Chem. 1982,
47, 1455–1459.
(26) Griller, D. Radical Reaction Rates in Liquids. In Landolt-Bo¨rnstein:
Numerical Data and Functional Relationships in Science and Technol-
ogy; Fischer, H., Ed.; Springer-Verlag: Berlin, 1984; Vol. 13a, pp 5-
130.
(29) Scaiano, J. C. J. Am. Chem. Soc. 1980, 102, 5399–5400.
(30) Chatgilialoglu, C.; Ingold, K. U.; Scaiano, J. C. J. Am. Chem. Soc.
1981, 103, 7739–7742.
(31) Most of the EPR spectra were recorded with 2.0 mW power and 1.0
Gpp modulation intensity. However, for detection of the easily saturated
benzyl radical 1.0 mW power and 0.2 Gpp modulation intensity was
needed.
(27) Effio, A.; Griller, D.; Ingold, K. U.; Scaiano, J. C.; Sheng, S. J. J. Am.
Chem. Soc. 1980, 102, 6063–6068.
(28) Newcomb, M.; Park, S. U. J. Am. Chem. Soc. 1986, 108, 4132–4134.
9
J. AM. CHEM. SOC. VOL. 132, NO. 7, 2010 2355

A R T I C L E S
Walton et al.
Table 4. Relative EPR Signal Intensities in Reactions of
NHC-BH2• Radicals with 1-Bromobutane and
2-Bromo-2-methylpropanea
t-Bu•
relative EPR intensity
n-Bu•
relative EPR intensity
precursor NHC-BH3
1a
2
3
4
5
1b
7
2.2
1.6
1.0
0.3
0.4
0.4
0.2
0.6
1.6
3.4
n.d.
n.d.
e0.1
n.d.
a Determined from the double integrals of the EPR spectra relative to
the t-Bu• intensity with 3; n.d. ) not determined.
Figure 8. EPR spectra obtained from UV photolyses of solutions of 3 and
DTBP with various alkyl halides at 300 K: (a) with added n-BuBr showing
part of the n-Bu• radical spectrum, (b) with added t-BuBr showing part of
the t-Bu• radical spectrum, (c) with added PhCH2Cl showing the
NHC-BH2• radical 3•, (d) same sample as c but with low power and smaller
modulation intensity,31 showing the benzyl radical, (e) with added allyl
chloride showing the allyl radical.
1-bromobutane and 2-bromo-2-methylpropane, respectively
(Figure 8a and 8b), showing that boron-centered radical 3•
readily abstracted primary and tertiary bromine atoms at room
temperature. With added Me3CCl, however, only radical 3• was
detectable, indicating that chlorine-atom abstraction was too
slow for spectroscopic observation. With PhCH2Cl as substrate,
we initially observed only the NHC-boryl radical 3• (Figure
8c), but on reducing the microwave power and modulation
intensity,31 a new spectrum appeared with hfs identical to those
of the benzyl radical. In fact, a mixture of 3• and PhCH2•
actually resulted. Thus, radical 3• did abstract Cl atoms from
benzyl chloride to give the resonance-stabilized benzyl radical,
although the process was comparatively slow. Similarly, a weak
spectrum of the allyl radical was observed with allyl chloride
as substrate (Figure 8e).
Similar experiments were carried out with NHC-boranes 1a,
2-5, and 7, and good spectra of alkyl radicals were observed
in every case with an alkyl bromide as a reaction partner, even
for precursors 4, 5, and 7 where the initial NHC-BH2• radicals
had not been detected. With 1b as reactant the spectra showed
a mixture of the NHC-BH2• radical with minor amounts of
the tert-butyl and a trace of the n-butyl radical with tert- and
n-alkyl bromide coreactants, respectively. Chlorine-atom ab-
straction was only detected, however, with the sterically
unhindered radical 3•.
Figure 9. EPR spectra on treatment of 8 with tert-butoxyl radicals. (a)
Radical 1a• at 230 K in t-BuPh solution. (b) Spectrum at 155 K in
cyclopropane solution. (c) Computer simulation of radical 14.
slower cross-combination of the t-Bu• radical with the sterically
congested NHC-boryl radical 1a•. The imidazole-based re-
agents with 6 π-electron delocalization were more efficient than
the others.
Bromine-atom abstraction was least efficient with the im-
idazolidin-2-ylidene-based radical 1b• and the triazole-based
radical 7•. The imidazolidine-containing radical lacks the quasi-
aromatic structure but has similar steric shielding to 1a• and a
similar termination rate. Thus, the much slower bromine-atom
transfer rate is due to electronic effects. The smaller hfs on boron
and on the R-H atoms (Table 1) indicate lower unpaired spin
on boron for the imidazolidine-containing radical 1b•, and this
is consistent with a slower rate of bromine-atom abstraction.
Experimental hfs were not obtained for the triazole-based radical
7•, but the computed hfs (Table 1) predict lower unpaired spin
on boron for this system too. It is clear that NHC-borane 3
should be a useful reagent for reductive transformations of a
wide range of alkyl bromides but will probably only reduce
allyl, benzyl, and similarly activated chlorides.
NHC - ΒΗ2• + RΒr f ΝΗC-ΒΗ2Βr + R•
R ) primary-alkyl, tert-alkyl, benzyl, allyl
The relative intensities of the released alkyl radicals are recorded
in Table 4. The initial reactant concentrations and reaction
conditions were essentially the same for all these experiments,
so the relative intensities are related to the overall efficiency of
the abstraction. However, the relative intensities will depend
on the rates of termination reactions as well as on the rates of
bromine-atom abstraction. Column 3 of Table 4 shows that the
n-Bu• radical intensity increased as the steric shielding of the
NHC-boryl radical center decreased on going from 1a to 2 to
3 as might be expected. Surprisingly, however, the reverse was
the case for the t-Bu• radical. Possibly this is a consequence of
Elimination Reaction of the NHC Thexyl-Borane 8. The
thexyl-borane 8 was examined as an example of an NHC-
borane with alkyl substitution at boron. UV photolysis of a
solution containing 8 and DTBP in t-BuPh gave rise to a good
spectrum of an NHC-boryl radical (Figure 9a). Surprisingly,
this was not the NHC-BH(•)thexyl radical 13. Inspection and
simulation showed this to be radical 1a• identical to that obtained
from 1a. To check if the boron-carbon bond simply underwent
homolytic scission, a solution of 8 in t-BuPh, without added
9
2356 J. AM. CHEM. SOC. VOL. 132, NO. 7, 2010

Reactivity of NHC-Boryl Radicals
Scheme 2. Radical Reaction of 8
A R T I C L E S
) 27.0 G at 180 K].33 Additional evidence supporting the
ꢀ-scission process of Scheme 2 was obtained by GC-MS
analysis of the products from a 2 h UV photolysis of a solution
of 8 in DTBP at rt. The main products were shown from their
mass spectra to be Me2C-CMe2 and 1a together with products
derived from the DTBP (ethane, acetone, tert-butanol, etc.).
The EPR spectra were too weak for radical concentration
measurements, but we estimated by visual inspection and
simulation that radicals 14 and 1a• were of equal concentration
at about Tmid ) 180 K. The activation energies Er of many
unimolecular reactions correlate with EPR-derived Tmid values
according to34
Εr/kcalmol-1 ) 0.044Τmid + 0.22
(3)
Strictly, this linear correlation is only expected to hold for
processes in which the rearranged and unrearranged radicals
terminate at the diffusion-controlled limit. In the case of radical
14 the ejected radical 1a• terminates more slowly than this, so
only a lower limit for Er will be obtained. Use of eq 3 leads to
Er g ca. 8 kcal mol-1, and for a normal pre-exponential factor
of 1013 s-1, the rate constant for the ꢀ scission of 14 at 300 K
will be eca. 107 s-1. We can conclude that ꢀ scission of 14 is
a fast process, and this is consistent with the resonance
stabilization of the released NHC-boryl radical 1a• and the
comparative stability of the tetrasubstituted alkene. It also
corresponds well with the idea35 that NHC complexation
weakens boron-carbon bonds as well as boron-hydrogen
bonds. ꢀ Scission was also reported for the only other known
ꢀ-boron-containing radical 15 (and one analog),33 and of course,
ꢀ scissions are rapid for LnMCH2CH2• radicals, where M )
second and subsequent row element.
DTBP, was heated up to 360 K in the EPR cavity. No significant
signals were observed. Similarly, on heating up to 360 K, with
simultaneous UV irradiation, no EPR signals were observed.
Direct homolysis of the B-C bond was not therefore detectable
in the accessible temperature range.
Conclusions
The most likely explanation seemed to be that the tert-butoxyl
radicals were abstracting the tertiary H atom from the branched
thexyl group in preference to the BH3 group. This is not too
surprising given the additional steric shielding of the boron
center compared to 1a. The resulting C-centered radical 14
subsequently underwent ꢀ scission to release the NHC-BH2•
radical 1a• and 2,3-dimethylbut-2-ene (Scheme 2). If this were
correct, we reasoned that the unrearranged C-centered radical
14 might be spectroscopically observable at lower temperatures.
Accordingly, a sample of 8 (3.3 mg) with DTBP (ca. 0.02 mL)
in cyclopropane (ca. 0.3 mL) was prepared on a vacuum line
and then UV irradiated at several temperatures in the EPR
cavity. At 180 K the spectrum showed a mixture of radical 1a•
and a new radical, and at 155 K (Figure 9b) this new radical
predominated.
NHC-boryls are a new class of boron-centered radicals, and
an understanding of their structures and reactions is beginning
to emerge from these EPR and complementary laser flash
photolysis studies,12b both supplemented by calculations.
NHC-boryl radicals can readily be generated by selective
H-atom abstraction by t-BuO• radicals from “clean” precursor
NHC-boranes with long shelf lives. A good range of NHC-
ring structures and substituent patterns are available, enabling
the electronic and steric properties to be tuned. Imidazole-based
NHC-boranes having 4,5-alkyl substituents were an exception
in that allylic H-atom abstraction competed, thus generating
long-lived allyl-type radicals rather than the desired NHC-
boryls. The EPR hfs indicated that the NHC-BH2• radicals had
planar structures in which the unpaired electron was contained
in a π-orbital delocalized round the whole system. They differed
markedly therefore from amine- and phosphine-boryls, the
only other well-investigated class of boron-centered radicals,
which are σ-type species pyramidal at boron.
The new radical was satisfactorily simulated with the fol-
lowing EPR parameters: a(6H) ) 20.8 G, a(11B) ) 16.0 G
(Figure 9c).32 These parameters are about what would be
expected for the C-centered radical 14. For comparison, the EPR
spectrum of the 2-(tetramethyldioxaborolane)ethyl radical 15,
which also contains a ꢀ-boron atom, was found to have fairly
similar hfs [a(HR) ) 22.1, a(11Bꢀ) ) 17.7, a(2Hꢀ) ) 22.1, a(2Hꢀ)
tert-Butoxyl radicals abstracted hydrogen atoms from
NHC-boranes more than 3 orders of magnitude faster than
C-centered radicals. The rate constant at 300 K for the
unhindered molecule 3 was comparable to that for H-atom
abstraction from Bu3SnH by tert-butoxyl, even though the BDE
(32) The simulation assumed isotropic behavior. However, at the low
temperature of 155 K, restricted internal rotation in radical 14 is very
probable because of the agglomeration of large groups. The nonideal
agreement between some of the experimental line intensities with the
simulated ones can probably be attributed to the selective line
broadening resulting from this anisotropic motion.
(33) Walton, J. C.; McCarroll, A. J.; Chen, Q.; Carboni, B.; Nziengui, R.
J. Am. Chem. Soc. 2000, 122, 5455–5463.
(34) Walton, J. C. J. Chem. Soc., Perkin Trans. 2 1998, 603–605.
(35) Monot, J.; Makhlouf Brahmi, M.; Ueng, S.-H.; Robert, C.; Desage-El
Murr, M.; Curran, D. P.; Malacria, M.; Fensterbank, L.; Lacoˆte, E.
Org. Lett. 2009, 11, 4914–4917.
9
J. AM. CHEM. SOC. VOL. 132, NO. 7, 2010 2357

A R T I C L E S
Walton et al.
vacuum line by distilling in the cyclopropane, degassing with three
freeze-pump-thaw cycles, and finally flame sealing the tubes.
Most of the EPR spectra were recorded with 2.0 mW power, a 1.0
Gpp modulation intensity, and a gain of ca. 106. In all cases where
spectra were obtained, hfs were assigned with the aid of computer
simulations using the Bruker SimFonia and NIEHS Winsim2002
software packages. For kinetic measurements, precursor samples
were used mainly in ‘single shot’ experiments; that is, new samples
were prepared for each temperature and each concentration to
minimize sample depletion effects. In a few cases second shot data
were included. EPR signals were double integrated using the Bruker
WinEPR software, and radical concentrations were calculated by
reference to the double integral of the signal from a known
concentration of the stable radical 2,2-diphenyl-1-picrylhydrazyl
(DPPH) [1 × 10-3 M in PhMe], run under identical conditions, as
described previously.36
of the B-H bond is considerably greater. For preparative
reductions with the new reagents, DTBP/UV and/or tert-butyl
hyponitrite/heat will probably be the best initiator systems. The
rate of H-atom abstraction was very sensitive to steric shielding
around the BH3 center, so the rate could easily be tuned. In the
absence of an internal reaction partner, the NHC-boryl radicals
dimerized to 1,2-bis-NHC-diboranes. The rate of this process
was also strongly dependent on steric shielding but increased
to the diffusion-controlled limit for sterically unhindered
NHC-boryls.
Imidazole-based NHC-boryl radicals readily abstracted
bromine atoms from alkyl, allyl, and benzyl bromides. Chlorine-
atom abstraction was, however, much less efficient and only
observed for the sterically unhindered radical 3• reacting with
allylic and benzylic chlorides. The radical 1b•, lacking the quasi-
aromatic structure, and the triazole-based radical 7• were poorer
bromine-atom acceptors, probably as a consequence of the lower
unpaired spin on boron for these two species. An interesting
elimination process was discovered for the NHC-borane 8
containing a bulky thexyl substituent at boron. The tertiary H
atom of the thexyl group was selectively removed, yielding an
unusual ꢀ-boron-containing alkyl radical. The latter rapidly
underwent ꢀ scission of its B-C bond with production of an
NHC-boryl radical and an alkene. In this respect,
NHC-BH2-alkyl radicals resembled LnM-CH2CH2• radicals
where M is an element from the second or subsequent row of
the Periodic Table. The ease of this elimination suggests that
addition reactions of NHC-boryl radicals with simple alkenes
will be reversible at room temperature, so that hydroborations
with NHC-boranes will only be achievable under special
conditions.
Acknowledgment. This work was supported by grants from
the US National Science Foundation (CHE-0645998 to D.P.C.),
Agence Nationale de la Recherche (ANR, BLAN0309 Radicaux
Verts, and 08-CEXC-011-01, Borane), l′e´tat et la re´gion Ile-de-
France (Chaire Blaise Pascal to D.P.C.), UPMC, CNRS, and IUF
(M.M., L.F.). Technical assistance (MS, elemental analyses) was
generously offered by FR 2769. A.S. thanks the University of
Pittsburgh for a Graduate Excellence Fellowship. J.C.W. thanks
EaStChem for financial support. We dedicate this paper to Professor
Athelstan L. J. Beckwith for his 80th birthday.
Supporting Information Available: Experimental details of
the preparations of imidazolium and triazolium salts and the
corresponding boranes; NMR spectra of all new compounds;
sample EPR spectra of NHC-boryl and related radicals;
experimental and computed hfs of NHC-boryl radicals; com-
puted torsion barrier in radical 3•; kinetic data for NHC-boryl
radical decays; details of photochemical reaction of
NHC-BH2-thexyl 8 with DTBP; complete ref 19; DFT-
optimized geometries and energies for the NHC-boryl radicals.
This material is available free of charge via the Internet at http://
pubs.acs.org.
Experimental Section
EPR Spectroscopy. EPR spectra were obtained with a Bruker
EMX 10/12 spectrometer fitted with a rectangular ER4122 SP
resonant cavity and operating at 9.5 GHz with 100 kHz modulation.
Usually stock solutions of the NHC-borane (2-15 mg) and DTBP
(ca. 0.1 mL) in tert-butylbenzene (1.0 mL) were prepared and
sonicated if necessary. An aliquot (0.2 mL), to which any additional
reactant had been added, was placed in a 4 mm o.d. quartz tube,
deaerated by bubbling nitrogen for 20 min, and photolyzed in the
resonant cavity by unfiltered light from a 500 W super pressure
mercury arc lamp. Solutions in cyclopropane were prepared on a
JA909502Q
(36) Maillard, B.; Walton, J. C. J. Chem. Soc., Perkin Trans. 2 1985, 443–
450.
9
2358 J. AM. CHEM. SOC. VOL. 132, NO. 7, 2010