Fluorinated 9-borafluorenes vs. conventional perfluoroaryl boranes - Comparative Lewis acidity

Article abstract of DOI:10.1139/v05-240

Perfluorinated 9-phenyl-9-borafluorene, 1, is an antiaromatic analog of the well-known tris(pentafluorophenyl)borane. Spectroscopic, structural, and electrochemical studies have been performed on 1 and its Lewis base adducts with MeCN, THF, and PMe₃ with a view to assessing its comparative Lewis acid strength relative to B(C₆F₅)₃. For the sterically undemanding Lewis base MeCN, 1 and B(C₆F₅) ₃ have comparable LA strengths, while for more sterically prominent THF, 1 is clearly the stronger Lewis acid (LA) based on competition experiments. We conclude that steric factors, rather than antiaromaticity, are the most important determinants in the LA strength differences between 1 and B(C ₆F₅)₃.

Full text of DOI:10.1139/v05-240

2098
Fluorinated 9-borafluorenes vs. conventional
perfluoroaryl boranes — Comparative Lewis acidity
Preston A. Chase, Patricio E. Romero, Warren E. Piers, Masood Parvez, and
Brian O. Patrick
Abstract: Perfluorinated 9-phenyl-9-borafluorene, 1, is an antiaromatic analog of the well-known tris(pentafluoro-
phenyl)borane. Spectroscopic, structural, and electrochemical studies have been performed on 1 and its Lewis base ad-
ducts with MeCN, THF, and PMe3 with a view to assessing its comparative Lewis acid strength relative to B(C6F5)3.
For the sterically undemanding Lewis base MeCN, 1 and B(C6F5)3 have comparable LA strengths, while for more
sterically prominent THF, 1 is clearly the stronger Lewis acid (LA) based on competition experiments. We conclude
that steric factors, rather than antiaromaticity, are the most important determinants in the LA strength differences be-
tween 1 and B(C6F5)3.
Key words: boranes, Lewis acids, fluorinated compounds, heterocycles.
Résumé : Le 9-phényl-9-borafluorène perfluoré (1) est un analogue antiaromatique du tris(pentafluorophényl)borane
bien connu. Afin d’évaluer son acidité relative à celle du B(C6F5)3, on a effectué des études spectroscopiques, structu-
rales et électrochimiques sur le composé 1 et sur les adduits qui résultent de sa réaction avec les bases de Lewis
MeCN, THF et PMe3. Sur la base d’expériences de compétition, on a observé que dans le cas de la base de Lewis
MeCN qui n’impose pas de demandes stériques spéciales, les forces des acides de Lewis du composé 1 et du B(C6F5)3
sont comparables alors que, avec le THF qui impose des contraintes stériques plus grandes, l’acidité de Lewis du com-
posé 1 est nettement plus grande. On en conclut que les facteurs stériques, plutôt que l’antiaromaticité, sont plus im-
portants dans la détermination des différences des forces d’acidité de Lewis entre le composé 1 et le B(C6F5)3.
Mots clés : boranes, acides de Lewis, composés fluorés, hétérocycles.
[Traduit par la Rédaction] Chase et al. 2105
Introduction
plication (10). It is thus clear that the ability to modulate
Lewis acid strength is beneficial for the applications of this
family of boranes.
Highly fluorinated aryl boranes are an important class of
Lewis acids because of their high thermal stability, resis-
tance to hydrolysis, and Lewis acid strength (1). Although
they have been known for many years (2), their application
as activators for single-site olefin polymerization catalysts
(3) has sparked interest in defining other applications in ar-
eas as diverse as organic synthesis (4), weakly coordinating
anion development (5), and cationic polymerization initia-
tion (6). Since Lewis acid strength is related to the effective-
ness of the boranes in many of these applications (7), efforts
to modulate the Lewis acid properties of the parent tris-
(pentafluorophenyl)borane have utilized several strategies,
including the use of larger, more fluorinated aryl groups (8)
and the replacement of one C6F5 group with poorly π-donat-
ing but electronegative substitutents like -NC4H4 (9). In some
instances, weakening the Lewis acid strength by substitution
with nonfluorinated aryl groups is beneficial for a given ap-
One intriguing strategy for increasing the Lewis acid
strength of B(C6F5)3 without drastically elevating the steric
imposition of the aryl groups about boron is to incorporate
the boron center into an antiaromatic ring framework. Using
this strategy, we and Marks and co-workers (11) independ-
ently prepared the perfluorinated 9,10-diboraanthracene
Lewis acid shown in Chart 1. Marks found it to be an excep-
tionally strong Lewis acid (12), eclipsing B(C6F5)3 in com-
petition experiments with low-steric-demand Lewis bases
like acetonitrile. In fact, the second boron center is also
highly Lewis acidic in these compounds, often performing
independently from the first. Taking a cue from early work
on boroles by Eisch et al. (13), we developed a second fam-
ily of antiaromatic perfluoroaryl boranes, the 9-borafluorene
compounds 1 and 2 (14) (Chart 1), a closer structural analog
of the parent B(C6F5)3. A detailed study on the coordination
chemistry of compounds 1 and 2 with the unusual Lewis
base “Cp*AlI” showed that they exhibit chemical behavior
that is more borane-like than borole-like (15); that is, com-
Received 11 August 2005. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on 10 February 2006.
P.A. Chase, P.E. Romero, W.E. Piers,1 and M. Parvez.
Department of Chemistry, University of Calgary, 2500
University Drive N.W., Calgary, AB T2N 1N4, Canada.
B.O. Patrick. Department of Chemistry, University of British
Columbia, 2036 Main Mall, Vancouver, BC V6T 1Z1, Canada.
1
pounds 1 and 2 coordinate the aluminum Lewis base in an η
fashion, rather than undergoing two-electron reduction to
give higher hapticity π complexes. Here we show that these
reactivity patterns can be extended to more conventional
Lewis bases and discuss the relative LA strength of 1 rela-
tive to the parent borane B(C6F5)3. Structural, competition,
1Corresponding author (e-mail: wpiers@ucalgary.ca).
Can. J. Chem. 83: 2098–2105 (2005)
doi: 10.1139/V05-240
© 2005 NRC Canada

Chase et al.
2099
Chart 1.
gap causes the absorption to shift out of the visible range.
Similar behavior has been observed in unfluorinated 9-
borafluorenes, which have been employed as colourimetric
sensors for fluoride ions (17). Although labile (vide infra),
the equilibria strongly favor the adducts 1·LB and 2·LB; for
the stronger Lewis acid 1, adducts 1·LB do not relinquish
their Lewis bases even with gentle heating under vacuum.
All were formed as analytically pure, colorless solids that
could be fully characterized using standard methods.
C6F5
B
F
F
F
F
F
F
F
F
B
C6F5
F
F
F
F
F
B
F
F
F
B
1
F
F
F
F
Upon complexation, resonances in the H NMR spectra
F
F
F
for the coordinated bases in 1·LB shift downfield signifi-
cantly; for example, the resonance for the methyl group of
the MeCN base in 1·MeCN shifts to 2.66 ppm vs. 2.10 ppm
for free MeCN (18) as electron density is removed by the
Lewis acid. For 2·MeCN, the opposite trend is observed for
the B–Me signal, which shifts upfield by 0.7 ppm upon
complexation. In the 19F NMR spectra of 1·LB, similar ef-
fects to those observed in the adducts of B(C6F5)3 are exhib-
ited: namely, a contraction of the chemical shift difference
between the para- and meta-fluorines (∆δp,m) (19). In B(C6F5)3,
this value is approximately 18 ppm in noncoordinating sol-
vents, but upon treatment with acetonitrile a value of
8.4 ppm is found (20). The changes observed in 1·MeCN vs.
free 1 indicate that the ∆δp,m value in the -C6F5 ring goes
from 14.7 to 6.4 ppm upon complexation of MeCN. Upfield
shifts of varying severity are also observed in the fluorines
of the borafluorene framework and analogous patterns are
revealed in the THF and PMe3 adducts.
CH3
F
F
F
F
F
F
1
2
F
F
F
F
F
F
F
F
B
F
F
F
LB
R
[1]
B
F
R
+
LB
F
F
F
F
R = C6F5, 1
R = Me, 2
R = C6F5, 1•MeCN, 1•THF, 1•PMe3
R = Me, 2•MeCN, 2•THF, 2•PMe3
The 11B NMR spectra also show a distinct difference between
the tricoordinate, base-free Lewis acids and the tetracoordi-
nated adducts (21). In the case of the free acids, the 11B
NMR chemical shift for the synthesized 9-borafluorene de-
rivatives ranges from δ 56.0 to δ 67.3. In general, the range
in 11B chemical shift for tetrahedral boron occurs at much
higher field than related tricoordinate systems. The 11B
NMR chemical shifts for the adducts of 1·LB and 2·LB oc-
cur between δ 8.6 and δ –15.2 ppm. Both 1·PMe3 and
2·PMe3 exhibit coupling in the 11B NMR spectra (δ –14.2
and electrochemical studies suggest that steric factors domi-
nate in determining the relative Lewis acidity in these com-
pounds and that the antiaromaticity of 1 is not a significant
progenitor of its high Lewis acidity.
Results and discussion
The pentafluorophenyl (1) and methyl (2) 9-borafluorene
derivatives were prepared as previously reported (14) and re-
acted with the sterically differentiated but comparably basic
(16) Lewis bases acetonitrile (MeCN) and tetrahydrofuran
(THF) and the stronger base trimethylphosphine (PMe3) to
form 1:1 Lewis acid – Lewis base adducts (eq. [1]). In the
case of MeCN and THF, an excess of base was used, while
for PMe3 near stoichiometric amounts were employed. In all
cases, the compounds 1·LB and 2·LB were isolated in good
yields as solids simply by removal of the solvent in vacuo.
Immediately upon introduction of the Lewis base (LB), or-
ange solutions of 1 and lime green solutions of 2
decolorized, as coordination of the Lewis base disrupted the
strong π–π transition involving the low-lying boron-based
LUMO, which is responsible for the colors in 1 and 2. Lewis
base coordination fills this orbital, and the LUMO is now a
higher energy formally π* orbital; the increase in the energy
1
and –11.4, JB-P = 87 and 66 Hz, respectively).
The IR spectra of both 1·MeCN and 2·MeCN were re-
corded, and a diagnostic band for the CϵN stretching mode
was observed at 2361 and 2364 cm–1, respectively, similar to
that observed for the B(C6F5)3 adduct (νCϵN = 2367 cm–1).
The frequency of the CϵN stretching mode for the free base
is found at significantly lower wavenumbers (νCϵN
=
2253 cm–1), indicative of a weaker carbon–nitrogen bond
compared with that in the adducts. As the donor orbital of
MeCN has slight antibonding character with respect to the
CN linkage, complexation will remove some of the electron
density from this orbital, which will strengthen the CN bond
and result in the observed shift in νCϵN (20).
A crystal structure of 1·MeCN was obtained, and a view
with selected metrical data is presented in Fig. 1 (see also
Table 1).2 Compared to the structure of 1, there is elongation
2 Supplementary data for this article are available on the journal Web site (http://canjchem.nrc.ca) or may be purchased from the Depository
of Unpublished Data, Document Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0R6, Canada. DUD 4074. For more
information on obtaining material refer to http://cisti-icist.nrc-cnrc.gc.ca/irm/unpub_e.shtml. CCDC 280874, 1·MeCN and 280873,
[Cp*2Co]+[1·Cl]– contain the crystallographic data for this manuscript. These data can be obtained, free of charge, via
www.ccdc.cam.ac.uk/conts/retrieving.html (Or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ,
UK; fax +44 1223 336033; or deposit@ccdc.cam.ac.uk).
© 2005 NRC Canada

2100
Can. J. Chem. Vol. 83, 2005
Fig. 1. Crystalmaker model of the molecular structure of
1·MeCN with atom labeling scheme. Selected bond lengths (Å):
B(1)—N(1) 1.581(3), B(1)—C(1) 1.620(3), B(1)—C(12) 1.612,
B(1)—C(13) 1.617(3), N(1)—C(19) 1.133(3), C(1)—C(2)
1.366(3), C(1)—C(6) 1.413(3), C(2)—C(3) 1.384(3), C(3)—C(4)
1.364(3), C(4)—C(5) 1.384(3), C(5)—C(6) 1.389(3), C(6)—C(7)
1.481(3). Selected bond angles (°): C(1)-B(1)-N(1) 106.5(2),
C(1)-B(1)-C(12) 98.8(2), B(1)-N(1)-C(19) 171.0(2), C(1)-C(6)-
C(7)-C(12) 1.6(2).
Table 1. Summary of data collection and structure refinement
details for 1·MeCN and [Cp*2Co]+[1·Cl]–.
1·MeCN
[Cp*2Co]+[1·Cl]–
Formula
C26H9BF13N
C38H30F13BClCo
fw
593.15
838.83
Temperature (K)
Crystal size (mm3)
λ (Å)
Crystal system
Space group
a (Å)
198.2
170.15
0.028
0.710 69
Triclinic
P1(#2)
13.597(3)
13.678(3)
9.927 5(11)
90.641(15)
90.757(13)
103.348(16)
1796.0(6)
2
0.006 3
0.710 69
Triclinic
P1(#2)
8.967 5(9)
11.599 8(8)
12.026 4(8)
108.632(3)
102.337(3)
92.938(5)
1148.2(2)
2
b (Å)
c (Å)
α (°)
β (°)
γ (°)
V (Å3)
Z
Reflections
Unique reflections
No. of variables
R1
8855
3822
370
0.035
6647
6357
487
0.050
wR2
0.050
1.18
0.049
2.09
GoF
framework, namely the C=C bond of the five-membered
core ring. This result was also noted in the structures of
perfluorinated 9,10-diphenyl-9,10-diboranthracene (11) and
1,2-bis[bis(pentafluorophenyl)boryl] tetrafluorobenzene (11b,
23). This effect is also apparent in the adduct, 1·MeCN. The
bond lengths for C(1)—C(6) and C(7)—C(12) are 1.413(3)
and 1.415(3) Å, respectively, while the average of the re-
maining bonds is 1.376(3) Å. The difference in the bond
lengths is similar to those observed in the aforementioned
cases.
of the B—C bonds in both the C6F5 ring and the 9-
borafluorene fragment. In 1·MeCN, B(1)—C(1), B(1)—
C(12), and B(1)—C(13) are 1.620(3), 1.612(3), and 1.617(3)
Å, respectively (cf. 1.555(4) and 1.547(3) Å, respectively,
for the analogous bonds in 1). This is the result of the
rehybridization of the boron from sp2 to sp3 and the increase
in coordination number. The N(1)-B(1)-C(1) and N(1)-B(1)-
C(13) angles are 106.5(2)° and 110.2(2)°, respectively, in-
dicative of a distorted tetrahedral environment about boron.
The internal angle in the 9-borafluorene is even more com-
pressed than in the free acid; the C(1)-B(1)-C(12) angle is
98.8(2)°, compared with 103.5(2)°. The MeCN base has re-
mained linear (N(1)-C(19)-C(20) = 179.1(2)°) and is bonded
in an essentially linear orientation with the boron center. The
B(1)-N(1)-C(19) angle is 171.0(2)°, and the acetonitrile
methyl group is oriented slightly over the 9-borafluorene
ring. The B(1)—N(1) distance (1.581(3) Å) and the CϵN tri-
ple bond (N(1)—C(19) = 1.133(3) Å) are quite close to
those observed in the MeCN adduct of B(C6F5)3 reported by
Erker and co-workers (20) (for the equivalent bonds, the
B—N and C—N distances are 1.616(3) and 1.124(3) Å, re-
spectively). The same group has performed an accurate de-
termination of the crystal structure of free acetonitrile using
a Laser zone melting apparatus (22). From this, the bond
length of the CϵN triple bond (α phase, T = 208 K) was
found to be 1.141(2) Å. Thus, there is a measurable contrac-
tion of the CϵN bond upon complexation in both cases. This
correlates with the observed bond strengthening in the IR
spectra of 1·MeCN as the νCϵN band shifts to considerably
higher wavenumbers.
In our original communication, we noted that 1 exhibits
LA strength comparable with that of B(C6F5)3 by both the
Childs method (24) and that of Laszlo and Teston (25),
which uses the relative energies of the π* orbital of LA-
complexed crotonaldehyde to assess LA strength. Further-
more, in a competition experiment for the sterically unde-
manding MeCN, we found a slight preference for 1·MeCN
over MeCN·B(C6F5)3, the equilibrium constant being 1.3 at
25 °C. This preference is further emphasized when the
sterically more bulky base THF is used in this experiment.
Thus, when a 1:1:1 mixture of 1, B(C6F5)3, and THF in d8-
1
toluene was analyzed by H and 19F NMR spectroscopy, it
was found that the equilibrium lies completely towards
1·THF within the detection limits of NMR spectroscopy.
Remarkably, even though B(C6F5)3 is known to form a sta-
ble, isolable complex with THF, in a competitive situation, 1
dominates the competition for this Lewis base. That this is a
thermodynamic preference is illustrated by the fact that
identical results are obtained when equilibrium is ap-
proached from either side of the equation and by EXSY
NMR experiments in the MeCN system, which show that the
acetonitriles coordinated to both Lewis acids are exchanging
with each other on this timescale. Unfortunately, similar
competition experiments using PMe3 were precluded by the
In the case of the structure of base free 1, Marks and co-
workers (12) noted a small but statistically relevant elonga-
tion of one of the aryl C=C bonds in the 9-borafluorene
© 2005 NRC Canada

Chase et al.
2101
nearly total insolubility of the Me3P·B(C6F5)3 adduct in
common solvents (1).
ployed. Model experiments on the unfluorinated analogue of
1, 9-phenyl-9-borafluorene, show that, at scan rates of
500 mV s–1 (33), a reversible reduction and oxidation
process occurs with a reduction potential of –2.16 V vs.
ferrocene (Fc), a similar value to that found for BPh3. From
the available data, it is not possible to determine if the CV
of 9-phenyl-9-borafluorene is indicative of a one- or a two-
electron process. Under the same conditions, the CV of 1
revealed an irreversible process with a reduction potential
at –2.42 V vs. Fc. The very negative potential obtained for 1
in THF was puzzling since it implies that 1 is more difficult
to reduce than the unfluorinated analog and runs counter to
the expected trend evidenced in the MesnB(C6F5)3–n series.
On the other hand, the value for the reduction potential of 1
in THF is similar to those reported by Campbell (34) on the
electrochemistry of simple aromatic fluorinated compounds.
Although not directly comparable, since they are determined
in different solvents, pentafluorobenzene and decafluorobi-
phenyl showed a first reduction wave at –2.84 and –2.20 V,
respectively. These observations suggest the possibility that
the reduction process does not actually take place at the bo-
ron center, but at the fluorinated aromatic framework, i.e.,
we are observing the reduction of 1·THF rather than 1 under
these conditions. Campbell also detected the presence of
species arising from the rupture of aromatic C—F bonds,
which suggests that a similar decomposition pathway might
be operating in our case, i.e., via Ar—F scission. Alterna-
tively, the high negative potential for the onset of reduction
of 1 may be due to a large kinetic barrier to its reduction, as
was observed for B(C6F5)3 (31).
The rather marked dependence of the relative LA strength
on the steric properties of the Lewis base employed suggests
that the differences in Lewis acidity observed between 1 and
B(C6F5)3 are grounded in steric factors, rather than the fact
that 1 exhibits antiaromaticity. Indeed, a comparison be-
tween the bond lengths associated with the borafluorene
moiety in 1 vs. 1·MeCN shows that little change in the C—
C distances is observed upon complexation of the base. This
suggests that the π character of the ring framework is not
perturbed significantly, i.e., the system does not become
more aromatic (or less antiaromatic) as the base coordinates
the boron center. Accordingly, the Julg A values (26) (a mea-
sure of aromaticity based on metrical bond distances (27))
for both the five- and six-membered rings do not change sig-
nificantly in 1·MeCN vs. 1. In fact, while the A values for
the flanking six-membered rings remain essentially the
same, at 0.96 and 0.97 in 1 and 1·MeCN, respectively,3 the
A(5) value for 1·MeCN is actually less (0.89) than that for
uncomplexed 1 (0.92). The A(5) value in 1·MeCN is compa-
rable to that of 0.88 observed by West and co-workers (27b)
for the neutral 9-sila- and 9-germa-fluorenyl systems, which
also contain pyramidal atoms in the 9 position.
Given this discussion, we conclude that the reason 9-
borafluorene 1 is a Lewis acid of competitive or superior
strength to B(C6F5)3 (despite being less fluorinated) stems
from the greater steric accessibility of the boron center and
the lowered back strain encountered upon the pyramidali-
zation of the boron center, which occurs upon complexation
of a Lewis base. The flatness of the borafluorene ring
substitutent is the main reason for the relaxing of these steric
factors; further, the pyramidalization of boron in going from
1 to 1·LB likely relieves strain in the C-B-C moiety of the
five-membered ring. Although long a matter of debate, it has
been recently acknowledged that steric factors, particularly
the onset of back strain, dominate in determining the relative
Lewis acidity of the boron trihalides BX3 (28, 29).
Although antiaromaticity is likely not playing a strong
role in determining the relative Lewis acidity of 1, given the
propensity of boroles to undergo two-electron reduction to
form aromatic borollides (30), it seemed reasonable to pre-
sume that 1 could be reduced more readily than B(C6F5)3.
Norton and co-workers (31) have reported on the difficulties
associated with reduction of B(C6F5)3, but estimate an Ered
of –0.9 V vs. SCE based on measured Ered values for Mes3B,
Mes2B(C6F5), and MesB(C6F5)2 (Mes = 2,4,6-Me3C6H2)
(31c). While the electron-withdrawing pentafluorophenyl
groups make the boranes easier to reduce from a thermody-
namic perspective, the lack of steric protection afforded by
the C6F5 groups (as opposed to the Mes substitutent, for ex-
ample) renders the radical anion extremely reactive, even
with the surfaces of electrodes.
Since attempts to generate 12– via electrochemical means
failed, chemical reduction of 1 was attempted. Given that an
earlier study indicated that Al(I) reagents were not suffi-
ciently reducing (15), harsher reducing agents (C8K, Li, Na,
or K) were employed in a variety of solvents (Et2O, THF, to-
luene, benzene, bromobenzene, or α,α,α-trifluorotoluene).
Unfortunately, these invariably produced ill-defined mix-
tures. Not surprisingly, reactions involving alkali metals in
nondonor solvents proved extremely slow, and after work up,
the main products recovered were unreacted 1 and decompo-
sition by-products. Furthermore, attempts to stabilize the pu-
tative 12– by addition of crown ethers were also
unsuccessful; only coordination of one of the ether oxygen
atoms to the boron center was observed in the 19F NMR
spectrum. Given the slow kinetics associated with heteroge-
neous reductants, we sought a reducing agent that would al-
low the process to be carried out under homogeneous
conditions.
Norton and co-workers (31) have shown that Cp*2Co is an
effective reductant for B(C6F5)3. Reaction between 1 and
2 equiv. of Cp*2Co in benzene or toluene led to the precipi-
tation of an oily phase with recovery of 2 equiv. of unreacted
Cp*2Co from the supernatant. Analysis of the oily material
by 19F NMR spectroscopy showed an extremely complex
pattern arising from decomposition. However, this oily mate-
rial was diamagnetic in nature, as judged by its silent EPR
spectrum, even at low temperatures (ca. –50 °C). The pres-
ence of one remaining equivalent of reducing agent suggests
Cyclic voltammetry (CV) studies were performed on 1 in
a variety of solvents using tetrabutylammonium tetrakis-
(pentafluorophenyl)borate (32) as the supporting electrolyte.
Experiments in the noncoordinating solvents CH2Cl2 and
α,α,α-trifluorotoluene were unsuccessful, so THF was em-
3 These values are slightly depressed from the expected values near 1.00 because of the observed lengthening of the C—C bond that the six-
and five-membered rings share compared with the other C—C bonds in the flanking six-membered rings observed in these and other sys-
tems. Despite the slightly lower A value, these rings are essentially fully aromatic.
© 2005 NRC Canada

2102
Can. J. Chem. Vol. 83, 2005
Fig. 2. Crystalmaker model of the molecular structure of
[Cp*2Co]+[1·Cl]– with atom labeling scheme. Selected bond
lengths (Å): B(1)—Cl(1) 1.910(9), B(1)—C(13) 1.618(11),
B(1)—C(1) 1.617(10), C(1)—C(2) 1.381(10), C(2)—C(3)
1.373(10), C(3)—C(4) 1.350(11), C(4)—C(5) 1.397(11), C(5)—
C(6) 1.368(9), C(6)—C(7) 1.491(10). Selected bond angles (°):
C(1)-B(1)-C(12) 97.9(6), C(1)-B(1)-C(13) 113.1(7), C(12)-B(1)-
C(13) 114.9(6), C(1)-B(1)-Cl(1) 109.8(5), C(12)-B(1)-Cl(1)
106.9(6), C(13)-B(1)-Cl(1) 113.1(6).
Scheme 1.
[Cp*2Co]
F
F
F
F
F
F
B
F
F
B
F
F
CH2Cl2
Cl
C6F5
C6F5
+
Co
F
F
F
F
F
F
CH2Cl2
[Cp*2Co]
F
F
F
B
F
F
C6F5
F
F
F
anion rapidly abstracts Cl· from solvent to form
[Cp* Co]+[1·Cl]–. Alternatively, the CH2Cl2 solvent may
2
mediate this reaction by electron transfer from Cp*2Co to
the solvent, which then gives up a chloride ion to 1. Al-
though this option is not depicted in the scheme, our data do
not exclude this as a possibility.
Taken together, the observations garnered from the experi-
ments aimed at two-electron chemical and electrochemical
reduction of 1 suggest that this compound behaves more
similarly to its perfluoroaryl borane brethren than its non-
fluorinated 9-borafluorene analogs. While one-electron
reduction is fairly facile, the resulting radical anion is ex-
tremely reactive, abstracting atoms from solvent and (or) un-
dergoing dimerization processes.
In conclusion, we find the antiaromatic perfluorinated 9-
phenyl-9-borafluorene 1 to be of comparable or greater
Lewis acid strength than the noncyclic parent borane
B(C6F5)3, mainly because of the lesser back strain engen-
dered in adducts upon pyramidalization of the boron center,
owing to the flat C12F8B ring system. The antiaromaticity
associated with this function does not appear to play a sig-
nificant role in determining the Lewis acidity of the system.
The high Lewis acidity of 1 contributes to the difficulty in
carrying out two-electron reduction chemistry that is facile
in unfluorinated analogs.
that the initially formed radical anion is highly reactive and
that Cp*2Co is not sufficiently reducing to produce 12–. In
accord with this postulate, the reaction of Cp*2Co and 1 in
CH2Cl2 took place cleanly within seconds, as judged by
1
NMR spectroscopy. The H NMR spectrum showed only a
singlet at δ 1.70, corresponding to the expected hydrogen
resonances for the Cp* units of the diamagnetic Cp*2Co+
cation. 19F NMR spectroscopy clearly revealed a pattern of
seven different well-defined signals in the expected range.
The difference in chemical shift between the meta- and
para-fluorines of the C6F5 ring (∆δm,p 4.7 ppm) was indica-
tive of borate formation and a significant degree of charge
separation. The negative value obtained by 11B{1H} NMR
spectroscopy of δ –14 is also consistent with formation of a
borate center.
Single crystals of the product of this reaction were ob-
tained by slow diffusion of hexanes into a CH2Cl2 solution
of the compound, generated as described above. The molec-
ular structure of the salt [Cp* Co]+[1·Cl]– is shown in Fig. 2
2
and is in accord with all the spectral data. There are no un-
usual contacts in the structure, and the ions are well sepa-
rated with a boron—cobalt distance of 7.556(59) Å. The
B(1)—Cl(1) distance of 1.910(9) Å is in the expected range
Experimental section
for chloroborates, and the boron center in [Cp* Co]+[1·Cl]–
2
shows a distorted tetrahedral environment with a C(1)-B(1)-
C(12) angle of 97.9(6)°, also within the expected range by
comparison with related compounds.
General procedures
An Innovative Technologies or Vacuum Atmospheres
argon-filled glove box was employed for manipulation and
storage of oxygen and moisture-sensitive compounds. Reac-
tions were performed using a double manifold argon/vacuum
line and modified Schlenk line techniques. Matheson
Oxisorb-W gas purification cartridges were used to remove
residual oxygen and moisture in the argon stream. Glassware
was dried overnight (80–110 °C) and then assembled warm
and evacuated on the vacuum line or placed in the glovebox
antechamber while hot. Glassware consisted of swivel frit
apparatus and line connectors equipped with Kontes valves
and round-bottom flasks equipped with ground glass joints.
Unless otherwise noted, the introduction of solvent in all
The outcome can be rationalized in terms of chloride ab-
straction from the CH2Cl2 solvent, the sole possible source
of chloride in the reaction (Scheme 1). A potential direct ab-
straction of Cl– by 1 is negated, since 1 is an indefinitely sta-
ble species in CH2Cl2, as well as in most chlorinated
solvents. Thus, the first step may involve the direct reaction
between 1 and Cp*2Co. Based on the previous observations
for the reaction of 1 with Cp*2Co in less polar solvents,
where only 1 equiv. of Cp*2Co was found to react with 1, a
plausible mechanism might involve the intermediacy of a
highly reactive, monoanionic radical species. The radical
© 2005 NRC Canada

Chase et al.
2103
manipulations using the argon/vacuum line was via vacuum
transfer with condensation at –80 °C (acetone – dry ice
bath). Solvents were purchased from Aldrich and dried and
deoxygenated before use in the following procedures. Hex-
anes, toluene, and THF were dried and deoxygenated via
passage through a Grubbs/Dow purification system (35) and
were stored in evacuated bombs over titanocene (36) (hex-
anes and toluene) or sodium – benzophenone ketyl (THF).
Benzene was deoxygenated by bubbling argon through the
solvent and was dried over sodium and stored over the so-
dium – benzophenone ketyl. Diethyl ether and methylene
chloride were predried over LiAlH4 and CaH2, respectively,
and subsequently stored over sodium – benzophenone ketyl
and CaH2, respectively. All deuterated solvents for NMR ex-
periments were purchased from Cambridge Isotopes.
Deuterated benzene, toluene, and THF were dried over 4 Å
molecular sieves, stored over sodium – benzophenone ketyl,
and distilled prior to use. Deuterated methylene chloride and
bromobenzene were dried over 4 Å molecular sieves, stored
over CaH2, and distilled prior to use. After distillation, the
deuterated solvents were stored in the glove box, in glass
bombs equipped with Kontes valves. NMR spectra were ob-
tained on Bruker AC 200 MHz (1H), AMX 300 MHz (1H,
19F, 11B{1H}), DRX 400 MHz (1H, 13C{1H}), or Varian
200 MHz (11B, 11B{1H}, 31P{1H}) spectrometers. All NMR
spectral data are reported in ppm, and NMR spectra of 1H and
13C were internally referenced to the residual solvent peak. The
other nuclei were referenced externally with the following
standards: 11B (BF3·OEt2, δ 0), 19F (C6F6, δ –163.0), 31P
(H3PO4, δ 0). For all air- and (or) moisture-sensitive com-
pounds and reactions, NMR samples were prepared in the
glove box and the NMR tubes were capped with rubber
septa. Temperature calibration for the NMR spectra was
spectral samples were analyzed as KBr plates using a
NEXUS 470 FT-IR ESP spectrometer. Electron paramag-
netic spectra (EPR) were recorded on a Bruker EMX spec-
trometer at –50 °C, referencing externally against
diphenylpicrylhydrazyl (DPPH, g = 2.0036) as a standard.
C6F5BC12F8 (1), CH3BC12F8 (2), C6H5BC12H8, and
B(C6F5)3·MeCN were synthesized via literature procedures.
B(C6F5)3 was purchased from Boulder Scientific and puri-
fied by reaction with Me2SiH(Cl) in hexanes followed by
vacuum sublimation (120 °C, dynamic vacuum). Reagent
quantities of PMe3 and MeCN were dried using 4 Å molecu-
lar sieves and CaH2, respectively, then freshly distilled be-
fore use or vacuum transferred directly into reaction vessels.
All other reagents were purchased from Aldrich Chemicals
and used as received.
Preparation of C12F8BC6F5(CH3CN) (1·MeCN)
C12F8BC6F5 (0.081 g, 0.17 mmol) was placed in a round-
bottom flask, and the flask was evacuated and cooled to
–78 °C. Methylene chloride (10 mL) was condensed onto
the solid, and the solution was stirred. CH3CN (excess) was
then condensed onto the stirred solution. Immediately on ex-
posure to the CH3CN vapours, the green colour of the 9-
borafluorene discharged. The reaction was warmed to room
temperature and stirred for 30 min. The solvent was re-
moved in vacuo, and a white powder was obtained. Yield:
1
0.088 g, 99%. IR (KBr, cm–1) ν: 2361 (CϵN stretch). H
NMR (CD2Cl2) δ: 2.66 (s, 3H, MeCN). 11B{1H} NMR
(C6D6) δ: –8.7 (br). 19F NMR (CD2Cl2) δ: –134.6 (m, 4F,
overlapped 2F o-C6F5), –136.3 (m, 2F), –155.6 (m, 2F),
–156.9 (m, 2F), –157.7 (m, 1F, p-C6F5), –164.1 (m, 2F,
m-C6F5). Anal. calcd. for C16H3BF13N: C 46.64, H 0.59, N
2.72; found: C 46.25, H 0.85, N 2.71.
1
achieved by monitoring the H NMR spectrum of methanol
or ethylene glycol (37).
Preparation of C12F8BC6F5(THF) (1·THF)
Synthesis was identical to the method employed for
C12F8BC6F5(CH3CN) using C12F8BC6F5 (0.084 g,
Cyclic voltammetry studies were carried out in a three-
electrode cell using a PerkinElmer EG & G instrument
(Princeton Applied Research) potentiostat, model 283, con-
nected to a drybox and equipped with a PowerCVTM
PowerSuite cyclic voltammetry software. A platinum disk
electrode (1.6 mm diameter) and a platinum wire were used
as the working electrode and the counter electrode, respec-
tively. A silver wire was employed as a pseudoreference
electrode. All the electrodes were purchased from BAS, Inc.
All cyclic voltammetry experiments and handling of solids
were performed in an argon-filled drybox at room tempera-
ture. THF (5 mL, previously dried and degassed) was syringed
into an electrochemical cell containing [n-Bu4N][B(C6F5)4],
and the mixture was stirred until complete dissolution to
give a solution of concentration 0.1 mol L–1. A background
test was run, and in cases where traces of oxygen were de-
tected, the system was additionally degassed by blowing ar-
gon through the solution with the aid of a Pasteur pipette.
The samples (0.01 mol L–1) were then run at 50, 100, 200,
500, and 1000 mV s–1 using Cp2Fe (0.01 mol L–1) as an inter-
nal standard. Electrochemical grade [n-Bu4N][B(C6F5)4] was
synthesized according to literature procedures (32) and dried
overnight at 130 °C.
1
0.18 mmol). Yield: 0.097 g, 100%. H NMR (CD2Cl2) δ:
4.24 (m, 4H, THF), 2.15 (m, 4H, THF). 11B{1H} NMR
(C6D6) δ: 4.1. 19F NMR (CD2Cl2) δ: –126.2 (m, 2F, o-
C6F5), –132.6 (m, 2F), –133.7 (m, 2F), –134.5 (m, 2F),
–154.6 (m, 2F), –157.1 (m, 1F, p-C6F5), –163.7 (m, 2F, m-
C6F5). Anal. calcd. for C18H8BF13O: C 48.39, H 1.48; found:
C 48.19, H 1.27.
Preparation of C12F8BC6F5(PMe3) (1·PMe3)
Synthesis was identical to the method employed for
C12F8BC6F5(CH3CN) using C12F8BC6F5 (0.082 g,
0.17 mmol). Yield: 0.094 g, 99%. 1H NMR (CD2Cl2) δ: 1.31
2
(d, 9H, JH-P = 11.0 Hz). 11B{1H} NMR (C6D6) δ: –15.2 (d,
1JB-P = 87 Hz). 19F NMR (CD2Cl2) δ: –130.4 (m, 2F, o-
C6F5), –133.4 (m, 2F) –133.7 (m, 2F), –156.0 (m, 2F),
–156.6 (m, 2F), –157.8 (m, 1F, p-C6F5), –163.7 (m, 2F, m-
C6F5). 31P{1H} NMR (C6D6) δ: –14.2 (m). Anal. calcd. for
C17H9BF13P: C 45.85, H 1.65; found: C 45.57, H 1.60.
Typical conditions for competition experiments
1·THF (0.008 g, 15 µmol) and B(C6F5)3 (0.008 g,
16 µmol) were weighed into an NMR tube and d8-toluene
(~0.5 mL) was added. The signals in the NMR spectra were
identified by comparison to authentic samples. Equilibrium
Elemental analysis was performed by Dorothy Fox or
Roxanna Smith within the department using a Control
Equipment Corporation (CEC) 440 elemental analyzer. IR
© 2005 NRC Canada

2104
Can. J. Chem. Vol. 83, 2005
constants were evaluated by integration of the 19F NMR
spectra of all peaks that were baseline separated.
References
1. W.E. Piers. Adv. Organomet. Chem. 52, 1 (2005).
2. A.G. Massey, A.J. Park, and F.G.A. Stone. Proc. Chem. Soc.
212 (1963).
3. E.Y.-X. Chen and T.J. Marks. Chem. Rev. 100, 1391 (2000),
and refs. cited therein.
4. (a) K. Ishihara and H. Yamamoto. Eur. J. Org. Chem. 527
(1999); (b) H. Yamamoto (Editor). Lewis acid reagents. Ox-
ford University Press, Oxford. 1999.
Preparation of [Cp*2Co]+[1·Cl]–
Cp*2Co (35 mg, 0.11 mmol) was dissolved in ca. 3 mL of
CH2Cl2 in a vial and added dropwise to a solution of 1
(50 mg, 0.11 mmol), also dissolved in ca. 3 mL of CH2Cl2.
The brown solution was stirred for 20 min and then layered
with hexanes. After cooling overnight at –35 °C, brown
crystals of [Cp*2Co]+[1·Cl]– were obtained. Yield: 45 mg,
53%. 1H NMR (CD2Cl2) δ: 1.70 (s, 15H, C5(CH3)5).
13C{1H} NMR (C6D6) δ: 94.0 (s, C5(CH3)5), 7.1 (s,
C5(CH3)5). 11B{1H} NMR (C6D6) δ: –14.0. 19F{1H} NMR
(CD2Cl2) δ: –130.2 (2F, o-F), –136.5 (2F), –137.2 (2F),
–158.5 (2F), –159.9 (2F), –161.0 (1F, m-F), –165.7 (2F, p-
F). Anal. calcd. for C38H30F13BClCo: C 54.36, H 3.58;
found: C 54.21, H 3.35.
5. (a) S.J. Lancaster, A. Rodriguez, A. Lara-Sanchez, M.D.
Hannant, D.A. Walker, D.H. Hughes, and M. Bochmann.
Organometallics, 21, 451 (2002); (b) J. Zhou, S.J. Lancaster,
D.A. Walker, S. Beck, M. Thornton-Pett, and M. Bochmann. J.
Am. Chem. Soc. 123, 223 (2001); (c) R.E. Lapointe, G.R.
Roof, K.A. Abboud, and J. Klosin. J. Am. Chem. Soc. 122,
9560 (2000); (d) see ref. 1 and refs. cited therein.
6. (a) S.P. Lewis, N.J. Taylor, W.E. Piers, and S. Collins. J. Am.
Chem. Soc. 125, 14686 (2003); (b) S.P. Lewis, L.D.
Henderson, M. Parvez, B.D. Chandler, W.E. Piers, and S. Col-
lins. J. Am. Chem. Soc. 127, 46 (2005).
7. (a) Z. Xu, K. Vanka, and T. Ziegler. Organometallics, 23, 104
(2004); (b) K. Vanka, Z. Xu, and T. Ziegler. Can. J. Chem. 81,
1413 (2003); (c) Z. Xu, K. Vanka, T. Firman, M. Michalak,
and T. Ziegler. Organometallics, 21, 2444 (2002); (d) K.
Vanka, M.S.W. Chan, C.C. Pye, and T. Ziegler. Organo-
metallics, 19, 1841 (2000); (e) G. Lanza, I.L. Fragala, and T.J.
Marks. Organometallics, 21, 5595 (2002); ( f ) G. Lanza, I.L.
Fragala, and T.J. Marks. Organometallics, 20, 4006 (2001).
8. (a) Y.-X.E. Chen, M.V. Metz, L. Li, C.L. Stern, and T.J.
Marks. J. Am. Chem. Soc. 120, 6287 (1998); (b) L. Li, C.L.
Stern, and T.J. Marks. Organometallics, 19, 3332 (2000);
(c) L. Li and T.J. Marks. Organometallics, 17, 3996 (1998).
9. G. Kehr, R. Fröhlich, B. Wibbeling, and G. Erker. Chem. Eur.
J. 6, 258 (2000).
Preparation of C12F8BMe(CH3CN) (2·MeCN)
C12F8BMe (0.053 g, 0.16 mmol) was weighed into a
round-bottom flask, which was evacuated and cooled to
–78 °C. Methylene chloride (10 mL) was condensed onto
the solid, and the solution was stirred. CH3CN (~2 mL, ex-
cess) was then condensed onto the stirred solution. Immedi-
ately on exposure to the CH3CN vapors, the green colour of
the 9-borafluorene discharged. The reaction was warmed to
room temperature and stirred for an additional 30 min. The
solvent was removed in vacuo, and a white powder was ob-
tained. Yield: 0.058 g, 97%. IR (KBr, cm–1) ν: 2364 (CϵN
1
stretch). H NMR (CD2Cl2) δ: 2.42 (s, 3H, MeCN), 0.38 (s,
3H, BMe). 11B{1H} NMR (C6D6) δ: –5.1. 19F NMR
(CD2Cl2) δ: –135.5 (m, 2F), –138.4 (m, 2F), –157.6 (m, 2F),
–158.1 (m, 2F). Anal. calcd. for C15H6BF8N: C 49.87, H
1.67, N 3.86; found: C 49.54, H 1.37, N 3.57.
10. D.J. Morrison and W.E. Piers. Org. Lett. 5, 2857 (2003).
11. (a) M.V. Metz, D.J. Schwartz, C.L. Stern, P.N. Nickias, and
T.J. Marks. Angew. Chem. Int. Ed. 39, 1312 (2000); (b) V.C.
Williams, W.E. Piers, W. Clegg, M.R.J. Elsegood, S. Collins,
and T.B. Marder. J. Am. Chem. Soc. 121, 3244 (1999).
12. M.V. Metz, D.J. Schwartz, C.L. Stern, T.J. Marks, and P.N.
Nickias. Organometallics, 21, 4158 (2002).
13. J.J. Eisch, J.E. Galle, and S. Kozima. J. Am. Chem. Soc. 108,
379 (1986).
14. P.A. Chase, W.E. Piers, and B.O. Patrick. J. Am. Chem. Soc.
Preparation of C12F8BMe(THF) (2·THF)
Synthesis was identical to the method employed for
C12F8BMe(CH3CN) using C12F8BMe (0.057 g, 0.18 mmol).
1
Yield: 0.069 g, 99%. H NMR (CD2Cl2) δ: 4.04 (m, 4H,
THF), 2.04 (m, 4H, THF), 0.40 (s, 3H, BMe). 11B{1H}
NMR (C6D6) δ: 8.6. 19F NMR (CD2Cl2) δ: –135.0 (m, 2F),
–136.2 (m, 2F), –156.9 (m, 2F), –157.5 (m, 2F). Anal. calcd.
for C17H11BF8O: C 51.81, H 2.81; found: C 51.65, H 2.65.
122, 12911 (2000).
15. P.E. Romero, W.E. Piers, S.A. Decker, D. Chau, T.K. Woo, and
M. Parvez. Organometallics, 22, 1266 (2003).
16. They have similar proton affinities: W.G. Mallard (Editor).
NIST Chemistry WebBook [online]. National Institute of Stan-
dards and Technology, Gaithersburg, Md. Available from
http://webbook.nist.gov/chemistry/pa-ser.html [accessed 2 Au-
gust 2005]. 2005.
Preparation of C12F8BMe(PMe3) (2·PMe3)
Synthesis was identical to the method employed for
C12F8BMe(CH3CN) using C12F8BMe (0.053 g, 0.17 mmol).
1
Yield: 0.060 g, 92%. H NMR (CD2Cl2) δ: 1.20 (d, 9H,
3
2JH-P = 11.0 Hz, PMe3), 0.36 (d, 3H, JH-P = 19.0 Hz, BMe).
1
11B{1H} NMR (C6D6) δ: –14.0 (d, JB-P = 66 Hz). 19F
NMR (CD2Cl2) δ: –134.5 (m, 2F), –134.7 (m, 2F),
–158.1 (m, 2F), –158.3 (m, 2F). 31P{1H} NMR (C6D6) δ:
–11.4 (m). Anal. calcd. for C16H12BF8P: C 48.28, H 3.04;
found: C 48.20, H 2.47.
17. S. Yamaguchi, T. Shirasaka, S. Akiyama, and K. Tameo. J.
Am. Chem. Soc. 124, 8816 (2002).
18. H.E. Gottlieb, V. Kotlyar, and A. Nudelman. J. Org. Chem. 62,
7512 (1997).
19. A.D. Horton, J. de With, A.J. van der Linden, and H. van de
Weg. Organometallics, 15, 2672 (1996).
Acknowledgements
20. H. Jacobsen, H. Berke, S. Doring, G. Kehr, G. Erker, R.
Frohlich, and O. Meyer. Organometallics, 18, 1724 (1999).
21. R.G. Kidd. In NMR of the newly accessible nuclei. Vol. 2.
Edited by P. Laszlo. Academic, London. 1983. pp. 49–77.
22. (a) T. Brackenmeyer, G. Erker, R. Frohlich, J. Prigge, and U.
Funding for this work was provided by the Natural Sci-
ences and Engineering Research Council of Canada
(NSERC) through a Discovery grant to WEP and postgradu-
ate scholarships (PGSA&B) to PAC.
© 2005 NRC Canada

Chase et al.
Peuchert. Chem. Ber. 130, 899 (1997); (b) D. Brodalla, D.
2105
29. (a) V. Branchadell and V. Oliva. J. Mol. Struct.
(THEOCHEM), 236, 75 (1991); (b) V. Branchadell, A. Sbai,
and V. Oliva. J. Phys. Chem. 99, 6472 (1995); (c) P. Fowler.
Nature (London), 355, 586 (1992); (d) K. Morokuma. Acc.
Chem. Res. 10, 294 (1977); (e) V. Jonas, G. Frenking, and
M.T. Reetz. J. Am. Chem. Soc. 116, 8741 (1994).
30. (a) R.J. Wehmschulte, M.A. Khan, B. Twamley, and B.
Schiemenz. Organometallics, 20, 844 (2001); (b) R.J.
Wehmschulte, A.A. Diaz, and M.A. Khan. Organometallics,
22, 83 (2003).
Mootz, R. Boese, and W. Osswald. J. Appl. Crystallogr. 18,
316 (1985).
23. (a) V.C. Williams, C.Y. Dai, Z.M. Li, S. Collins, W.E. Piers,
W. Clegg, M.R.J. Elsegood, and T.B. Marder. Angew. Chem.
Int. Ed. 38, 3695 (1999); (b) V.C. Williams, G.J. Irvine, W.E.
Piers, Z.M. Li, S. Collins, W. Clegg, M.R.J. Elsegood, and
T.B. Marder. Organometallics, 19, 1619 (2000); (c) L.D.
Henderson, W.E. Piers, G.J. Irvine, and R. McDonald.
Organometallics, 21, 340 (2002).
24. (a) R.F. Childs, D.L. Mulholland, and A. Nixon. Can. J. Chem.
60, 801 (1982); (b) R.F. Childs, D.L. Mulholland, and A.
Nixon. Can. J. Chem. 60, 809 (1982).
31. (a) C.J. Harlan, T. Hascall, E. Fujita, and J.R. Norton. J. Am.
Chem. Soc. 121, 7274 (1999); (b) R.J. Kwaan, C.J. Harlan,
and J.R. Norton. Organometallics, 20, 3818 (2001); (c) S.A.
Cummings, M. Iimura, C.J. Harlan, R.J. Kwaan, I. Vu Trieu,
J.R. Norton, B.M. Bridgwater, F. Jäkle, A. Sundararaman, and
M. Tilset. Organometallics. In press.
25. P. Laszlo and M. Teston. J. Am. Chem. Soc. 112, 8750 (1990).
26. A. Julg and P. Francois. Theor. Chim. Acta, 8, 249 (1967).
27. The Julg A value is calculated according to the equation A = 1 –
(225/n)Σ(1 – ri/r)2, where n = No. of C—C bonds in the cyclic
system, ri = C—C bond length, r = average C—C bond length.
For recent examples where this analysis has been used in simi-
lar heterocycles, see (a) P.v.R. Schleyer, P.K. Freeman, H. Jiao,
and H.B. Goldfuss. Angew. Chem. Int. Ed. Engl. 34, 337
(1995); (b) Y. Liu, D. Ballweg, T. Muller, I.A. Guzei, R.W.
Clark, and R. West. J. Am. Chem. Soc. 124, 12174 (2002).
28. (a) P. Cassoux, R.L. Kuczkowski. and A. Serafini. Inorg.
Chem. 16, 3005 (1977); (b) K. Iigima, N. Adachi, and S.
Shibata. Bull. Chem. Soc. Jpn. 57, 3269 (1984); (c) A.C.
Legon and H.E. Warner. J. Chem. Soc. Chem. Commun. 1397
(1991); (d) A.G. Avent, P.B. Hitchcock, M.F. Lappert, D.-S.
Liu, G. Mignani, C. Richard, and E. Roche. J. Chem. Soc.
Chem. Commun. 855 (1995).
32. R.J. LeSuer and W.E. Geiger. Angew. Chem. Int. Ed. 39, 248
(2000).
33. D. Astruc. Electron transfer and radical processes in transition-
metal chemistry. VCH Publishers Inc., New York. 1995.
34. B.H. Campbell. Anal. Chem. 44, 659 (1972).
35. A.B. Pangborn, M.A. Giardello, R.H. Grubbs, R.K. Rosen, and
F.J. Timmers. Organometallics, 15, 1518 (1996).
36. R.H. Marvich and H.H. Brintzinger. J. Am. Chem. Soc. 93,
2046 (1971).
37. C. Ammann, P. Meier, and A.E. Merbach. J. Magn. Reson. 46,
319 (1982).
© 2005 NRC Canada