1,4-Shifts in a dinuclear Ni(I) biarylyl complex: A mechanistic study of C-H bond activation by monovalent nickel

Article abstract of DOI:10.1021/ja067112w

The known aryne complex (PEt₃)₂Ni(η²- C₆H₂-4,5-F₂) (1a) reacts with a catalytic amount of Br₂-Ni(PEt₃)₂ over 1% Na/Hg to afford the dinuclear Ni(I) biarylyl

Full text of DOI:10.1021/ja067112w

Published on Web 01/09/2007
1,4-Shifts in a Dinuclear Ni(I) Biarylyl Complex: A Mechanistic
Study of C-H Bond Activation by Monovalent Nickel
Alana L. Keen, Meghan Doster, and Samuel A. Johnson*
Contribution from the Department of Chemistry & Biochemistry, UniVersity of Windsor,
Windsor, Ontario, Canada, N9B 3P4
Received October 4, 2006; E-mail: sjohnson@uwindsor.ca
Abstract: The known aryne complex (PEt3)2Ni(η2-C6H2-4,5-F2) (1a) reacts with a catalytic amount of Br2-
Ni(PEt3)2 over 1% Na/Hg to afford the dinuclear Ni(I) biarylyl complex [(PEt3)2Ni]2(µ-η1:η1-3,4-F2C6H2-3′,4′-
F2C6H2) (2a), which results from a combination of C-C bond formation and C-H bond rearrangement.
The dinuclear benzyne [(PEt3)2Ni]2(µ-η2:η2-C6H2-4,5-F2) (3) was obtained by the reaction of 1a with a
stoichiometric amount of Br2Ni(PEt3)2 over excess 1% Na/Hg, and 3 was found to catalyze the conversion
of 1a to 2a. The reaction of 1a with B(C6F5)3 produced the trinuclear complex (PEt3)3Ni3(µ3:η1:η1:η2-4,5-
F2C6H2)(µ3:η1:η1:η2-4,5-F2C6H2-4′,5′-F2C6H2) (6). The addition of PEt3 to 6 produced 1 equiv of 1a and 1
equiv of [(PEt3)2Ni]2(µ-η1:η1-4,5-F2C6H2-4′,5′-F2C6H2) (7a). Both 6 and 7a were identified as intermediates
in the conversion of 1a to 2a. The analogue [(PEt3)(PMe3)Ni]2(µ-η1:η1-4,5-F2C6H2-4′,5′-F2C6H2) (7b) was
prepared by the addition of PMe3 to 6 and was structurally characterized. NMR spectroscopic evidence
identified the additional asymmetric biarylyl [(PEt3)2Ni]2(µ-η1:η1-4,5-F2C6H2-3′,4′-F2C6H2) (8a) during the
conversion of 1a to 2a. The initial observation of 2 equiv of 8a for every equivalent of 2a produced from
solutions of 7a suggests that 8a and 2a are formed from a common intermediate. A crossover labeling
experiment shows that the C-H bond rearrangement steps in the conversion of 1a to 2a occur with the
intermolecular scrambling of hydrogen and deuterium labels. The evidence collected suggests that Ni(I)
complexes are capable of activating aromatic C-H bonds.
Introduction
We recently communicated the isomerization of the aryne
complex (PEt3)2Ni(η2-C6H2-4,5-F2) (1a) to the dinuclear Ni(I)
complex [(PEt3)2Ni]2[µ-η1:η1-(C6H2-3,4-F2)]2 (2a) when treated
with Br2Ni(PEt3)2 over 1% Na/Hg, as shown in Scheme 1.7 Two
possible mechanistic pathways were proposed, one in which
C-H bond activation precedes C-C bond formation, and a
second where C-C bond formation occurs first, and the resultant
biarylyl complex undergoes isomerization by C-H bond
activation. These two mechanistic manifolds are shown in
Scheme 1 as mechanisms A and B. In each reaction step,
multiple elementary processes could be involved.
Both stoichiometric and catalytic transformations involving
C-H activation are of current interest, due to the synthetic utility
that the facile functionalization of unreactive C-H bonds would
provide.1 The relatively low cost of nickel complexes, compared
to their more expensive second and third row transition metal
counterparts, provides an impetus for their utilization in catalytic
or stoichiometric transformations that proceed by C-H bond
activation.2 Unfortunately, calculations have demonstrated that,
although the Ni(PEt3)2 fragment is capable of C-F bond
activation,3 the activation of aromatic C-H bonds by Ni(PEt3)2
is not thermodynamically favorable.4 Although a few examples
of nickel complexes that perform C-H bond activation have
been reported,5,6 many of these involve C-H bonds rendered
more reactive by adjacent substituents.
The involvement of d9 Ni(I) intermediates in C-H bond
activation, which is implied in mechanism B, would render this
C-H bond activation unique compared to those that rely on
second and third row transition metal complexes with d6 and
d8 electronic configurations. Only a few related organometallic
dinuclear Ni(I) complexes are known,8,9 although their reported
(1) (a) Labinger, J. A.; Bercaw, J. E. Nature 2002, 417, 507-514. (b) Arndtsen,
B. A.; Bergman, R. G.; Mobley, T. A.; Peterson, T. H. Acc. Chem. Res.
1995, 28, 154-162. (c) Crabtree, R. H. J. Chem. Soc., Dalton Trans. 2001,
2437-2450. (d) Crabtree, R. H. J. Organomet. Chem. 2004, 689, 4083-
4091. (e) Lersch, M.; Tilset, M. Chem. ReV. 2005, 105, 2471-2526. (f)
Shilov, A. E.; Shul’pin, G. B. Chem. ReV. 1997, 97, 2879-2932. (g)
Kakiuchi, F.; Murai, S. Acc. Chem. Res. 2002, 35, 826-834. (h) Godula,
K.; Sames, D. Science 2006, 312, 67-72.
(2) Jolly, P. W.; Wilke, G. The Organic Chemistry of Nickel; Organometallic
Chemistry Series; Academic Press: New York, 1974; Vol. 1.
(3) (a) Braun, T.; Foxon, S. P.; Perutz, R. N.; Walton, P. H. Angew. Chem.,
Int. Ed. 1999, 38, 3326-3329. (b) Braun, T.; Cronin, L.; Higgitt, C. L.;
McGrady, J. E.; Perutz, R. N.; Reinhold, M. New J. Chem. 2001, 25, 19-
21. (c) Bach, I.; Poerschke, K.-R.; Goddard, R.; Kopiske, C.; Krueger, C.;
Rufinska, A.; Seevogel, K. Organometallics 1996, 15, 4959-4966.
(4) Reinhold, M.; McGrady, J. E.; Perutz, R. N. J. Am. Chem. Soc. 2004, 126,
5268-5276.
(5) (a) Nakao, Y.; Kanyiva, K. S.; Oda, S.; Hiyama, T. J. Am. Chem. Soc.
2006, 128, 8146-8147. (b) Kleiman, J. P.; Dubeck, M. J. Am. Chem. Soc.
1963, 85, 1544-1545. (c) Tsuda, T.; Kiyoi, T.; Saegusa, T. J. Org. Chem.
1990, 55, 2554-2558. (d) Brunkan, N. M.; Brestensky, D. M.; Jones, W.
D. J. Am. Chem. Soc. 2004, 126, 3627-3641. (e) Clement, N. D.; Cavell,
K. J. Angew. Chem., Int. Ed. 2004, 43, 3845-3847. (f) Clement, N. D.;
Cavell, K. J.; Jones, C.; Elsevier, C. J. Angew. Chem., Int. Ed. 2004, 43,
1277-1279. (g) Ogoshi, S.; Ueta, M.; Oka, M.-A.; Kurosawa, H. Chem.
Commun. 2004, 2732-2733. (h) Brunkan, N. M.; Brestensky, D. M.; Jones,
W. D. J. Am. Chem. Soc. 2004, 126, 3627-3641. (i) van der Boom, M.
E.; Liou, S.-Y.; Shimon, L. J. W.; Ben-David, Y.; Milstein, D. Inorg. Chim.
Acta 2004, 357, 4015-4023.
(6) Vicic, D. A.; Jones, W. D. J. Am. Chem. Soc. 1999, 121, 7606-7617.
(7) Keen, A. L.; Johnson, S. A. J. Am. Chem. Soc. 2006, 128, 1806-1807.
9
810
J. AM. CHEM. SOC. 2007, 129, 810-819
10.1021/ja067112w CCC: $37.00 © 2007 American Chemical Society

1,4-Shifts in a Dinuclear Ni(I) Biarylyl Complex
A R T I C L E S
Scheme 1
reactivities include reversible C-C bond activation.10 The
activation of C-H bonds at room temperature by Ni(I)
complexes could provide a new economically viable methodol-
ogy for syntheses involving C-H bond activation. The study
provides full experimental details for the conversion of 1a to
2a and a mechanistic study that provides insight into the nature
of the Ni-mediated C-H bond activation step.
Figure 1. Solid-state molecular structure of 2a as determined by X-ray
crystallography. Hydrogen atoms and two ethyl substituents, associated with
P(2) and P(3), are omitted for clarity. Selected bond lengths (Å), angles
(deg), and dihedral angles (deg): Ni(1)-Ni(2), 2.3710(5); Ni(1)-C(1),
1.965(3); Ni(2)-C(1), 2.308(3); Ni(1)-P(1), 2.2091(8); Ni(1)-P(2), 2.2232-
(8); P(1)-Ni(1)-Ni(2), 157.72(3); P(2)-Ni(1)-Ni(2), 103.25(2); C(1)-
Ni(1)-Ni(2), 63.50(7); C(1)-Ni(1)-P(1), 94.98(8); C(1)-C(6)-C(7),
112.7(2); C(1)-C(6)-C(7)-C(8), 30.8(6).
Results and Discussion
The alkali metal reduction of (PEt3)2NiBr(2-Br-4,5-F2C6H2)
in hydrocarbon solvents to form the aryne complex 1a is a
known reaction;11 however, in our hands, no reduction took
place within 24 h when the starting material, (PEt3)2NiBr(2-
Br-4,5-F2C6H2), was rigorously purified. The addition of
catalysts such as BrNi(PEt3)3 and Br2Ni(PEt3)2 allows this
reduction to proceed with Na/Hg in 4 h to produce 1a; however,
when Br2Ni(PEt3)2 was utilized the reaction continues, and the
conversion of 1a to the dinuclear Ni(I) complex 2a, shown in
Scheme 1, was observed after 1-2 days. Complex 2a can also
be prepared using isolated 1a in the presence of Na/Hg and
catalytic Br2Ni(PEt3)2.
An ORTEP depiction of the solid-state molecular structure
of 2a is shown in Figure 1. The structure demonstrates that 2a
is a dinuclear Ni(I) complex, with a Ni-Ni bond distance of
2.3710(5) Å and approximate C2 symmetry. The biarylyl
fragment bridges two Ni(PEt3)2 fragments in a µ-η1:η1 manner,
with a C(1)-C(6)-C(7)-C(8) torsion angle of 30.8(6)°. This
is an unusual bonding mode for a biarylyl ligand bridging a
metal-metal bond.12,13 Complex 2a displays Ni(1)-C(1) and
Ni(2)-C(8) bond lengths of 1.965(3) and 1.952(3) Å, as well
as longer Ni(1)-C(8) and Ni(2)-C(1) distances of 2.317(3) and
2.308(3) Å. Neither Ni atom lies in a plane with either of the
aryl rings, possibly because these longer Ni-C distances are
weak bonding interactions, not simply short contacts. Excluding
the Ni(1)-C(8) and N(2)-C(1) interactions, the geometries at
the Ni centers are distorted square planar. The donors associated
with P(1) and P(4) are approximately trans to the Ni-Ni bond,
whereas the donors associated with P(2) and P(3) are ap-
proximately trans to the shorter Ni-C bonds. The longer Ni-
(1)-C(8) and Ni(2)-C(1) interactions are directed toward the
Ni centers at an approximately 45° angle to the planes described
by the other four attached atoms.
The 213 K 31P{1H} NMR spectrum contains two 31P
environments, consistent with the C2 symmetry observed in
the solid state. One of these is a virtual triplet with a JPP
value of 5.4 Hz. To account for the strong coupling
between chemically equivalent phosphorus nuclei that is
mandatory for virtual coupling to occur, a Ni-Ni bond must
be present. The second 31P environment can be modeled as a
virtual triplet of doublets of doublets, due to coupling to 31P
as well as both 19F environments (JPP ) 5.4 Hz, JPF ) 7.1 Hz,
JPF ) 11.3 Hz).
As the sample is warmed, the 1H and 31P NMR signals
associated with the PEt3 groups begin to broaden. The barrier
1
to this process is estimated to be 23 kJ/mol from the H and
31P{1H} NMR data prior to coalescence; however, at near 323
K the sample decomposes rapidly. Thus, although it is apparent
that the phosphorus environments can exchange rapidly at higher
temperatures, the mechanism of this rearrangement is not clear.
Synthesis of a µ-η2:η2 Aryne Complex. The slow conver-
sion of 1a to 2a allowed for the observation of several long-
lived intermediates by 1H, 31P{1H}, and 19F NMR spectroscopy,
and the reaction of a pentane solution of 1a with 1 equiv of
Br2Ni(PEt3)2 and excess Na/Hg, as shown in eq 1, allowed for
the isolation of orange crystals of one of these species, [(PEt3)2-
Ni]2(µ-η2:η2-C6H2F2) (3). Complex 3 exhibits a virtual triplet
in the aromatic region of the 1H NMR spectrum, similar to that
of 1a. The 19F NMR spectrum contains a single environment at
δ -74.0, which is a virtual triplet of pentets due to coupling to
(8) Eisch, J. J.; Piotrowski, A. M.; Han, K. I.; Kruger, C.; Tsay, Y. H.
Organometallics 1985, 4, 224-231.
(9) Diercks, R.; Stamp, L.; Kopf, J.; Heindirk, T. D. Angew. Chem. 1984, 96,
891-895.
(10) Ramakrishna, T. V. V.; Sharp, P. R. Organometallics 2004, 23, 3079-
3081.
(11) Bennett, M. A.; Wenger, E. Organometallics 1995, 14, 1267-1277.
(12) (a) Perthuisot, C.; Edelbach, B. L.; Zubris, D. L.; Jones, W. D. Organo-
metallics 1997, 16, 2016-2023. (b) Leong, W. K.; Chen, G. Organome-
tallics 2001, 20, 2280-2287. (c) Leong, W. K.; Chen, G. Organometallics
2001, 20, 5771-5773. (d) Chehata, A.; Oviedo, A.; Arevalo, A.; Bernes,
S.; Garcia, J. J. Organometallics 2003, 22, 1585-1587. (e) Yeh, W.-Y.;
Hsu, S. C. N.; Peng, S.-M.; Lee, G.-H. Organometallics 1998, 17, 2477-
2483. (f) Mueller, J.; Haensch, C.; Pickardt, J. J. Organomet. Chem. 1983,
259, C21-C25.
5
the aromatic protons and a 2.3 Hz JPF to four chemically
equivalent 31P environments.
The solid-state structure of 3 was determined by X-ray
crystallography and is shown in Figure 2. The aryne fragment
bridges two Ni(PEt2)2 fragments in a µ-η2:η2 bonding mode.
(13) Bennett, M. A.; Griffiths, K. D.; Okano, T.; Parthasarathi, V.; Robertson,
G. B. J. Am. Chem. Soc. 1990, 112, 7047-7048.
9
J. AM. CHEM. SOC. VOL. 129, NO. 4, 2007 811

A R T I C L E S
Keen et al.
Scheme 2
catalyze this conversion; thus, Hg does not appear to be involved
directly in either the C-C bond forming or C-H activation
steps.16 Using 19F NMR spectroscopy, we could observe a
second intermediate during these reactions. Similar to complex
3, this intermediate also appears early in the reaction but has
two fluorine environments in the 19F NMR spectrum. Both
mechanisms from Scheme 1 have intermediates that should
exhibit two 19F environments, and thus attempts were made to
synthesize these complexes or related analogues to determine
which pathway operates.
Figure 2. Solid-state molecular structure of 3 as determined by X-ray
crystallography. Hydrogen atoms are omitted for clarity. Selected bond
lengths (Å), angles (deg), and dihedral angles (deg): Ni(1)‚‚‚Ni(2), 2.7242-
(5); Ni(1)-C(1), 1.916(2); Ni(1)-C(2), 1.962(2); Ni(2)-C(2), 1.920(2);
Ni(2)-C(1), 1.963(2); Ni(1)-P(1), 2.1664(7); Ni(1)-P(2), 2.1572(7); C(1)-
C(2), 1.390(3); C(2)-C(3), 1.427(3); C(3)-C(4), 1.353(3); C(4)-C(5),
1.405(4); C(5)-C(6), 1.356(4); C(6)-C(1), 1.420(3); P(2)-Ni(1)-P(1),
109.80(3).
This bonding mode has been observed before with nickel alkyne
complexes14 but is unique for aryne complexes. More typically,
aryne moieties bridging two metal centers adopt a µ-η1:η1
bonding mode, which is best described as a 1,2-disubstituted
phenylene.15 In complex 3, the C(1)-C(2) distance of 1.390(3)
Å is typical for an aromatic C-C bond, but the remainder of
the aryne ring shows bond length alternations due to backbond-
ing to the aromatic π-system. The Ni(1)-Ni(2) distance of
2.7242(3) Å in 3 is 0.3532(6) Å longer than the Ni-Ni bond
in 2a, which indicates that no significant Ni-Ni bond is present
in this complex. The geometries around the nickel centers are
slightly distorted from planar, with the Ni(PEt3)2 groups twisted
slightly with respect to each other, as can be seen from the P(1)-
Ni(1)-Ni(2)-P(3) torsion angle of -159.5(3)°. As a result, the
aryne moiety does not bridge perfectly symmetrically, and the
Ni(1)-C(1) and Ni(2)-C(2) distances are slightly shorter than
the Ni(1)-C(2) and Ni(2)-C(1) distances. Related arene
π-adducts have been observed as intermediates for C-F bond
cleavage,3 but no C-F bond activation occurs in these com-
plexes.
Synthesis of a 3-Fluoroaryne Complex. There is no efficient
synthetic route to the speculative aryne complex intermediate
of mechanism A in Scheme 1, (PEt3)2Ni(η2-3,4-F2C6H2);
however, the question of whether the location of the fluoro
substituents on the aryne fragment can promote the dinuclear
coupling of aryne complexes, as suggested in the second step
of mechanism A of Scheme 1, can be addressed by synthesizing
a related aryne complex bearing a fluoro-substituent in the
3-position. The oxidative addition of 1-fluoro-2,3-diodoben-
zene17 to (COD)Ni(PEt3)2 was regioselective and produced a
single isomer, (PEt3)2NiI(3-F-2-I-C6H3) (4) as shown in Scheme
2. The connectivity was determined by X-ray crystallography,
and an ORTEP depiction of this structure is provided in the
Supporting Information. The reduction of this product over Na/
Hg was rapid and produced the aryne complex (PEt3)2Ni(η2-
3-F-C6H3) (5). Unfortunately, this product is a low-melting oil;
similar behavior has been noted for the closely related
complex (PEt3)2Ni(η2-C6H4).11 The 31P{1H} NMR spectrum of
the pure product displays two signals; however, with the
slightest impurity of PEt3 a single resonance is observed.
Attempts to promote the dinuclear coupling of this aryne
complex failed. The complex is thermally stable for days in
refluxing toluene, and thus it seems unlikely that the intermediate
in mechanism B of Scheme 1 would spontaneously couple in
preference to 1a.
Pure solutions of complex 3 do not convert cleanly into 2a,
but the addition of small amounts of 3 to solutions of 1a does
Coupling of Arynes. It is plausible that the Ni(PEt3)2
fragment could behave as a Lewis acid and abstract PEt3 from
1a to form an unstable (PEt3)Ni(C6H2-4,5-F2) moiety, which
could then dimerize and undergo C-C bond formation to form
a precursor to 2a. In an attempt to mimic this reaction sequence
with a reagent that would react strictly as a Lewis acid, we added
1 equiv of B(C6F5)3 to a pentane solution of 1a, as shown in eq
2. This reaction produced the white precipitate Et3P‚B(C6F5)3
and a brown solution. Cooling the brown pentane solution to
(14) Day, V. W.; Abdel-Meguid, S. S.; Dabestani, S.; Thomas, M. G.; Pretzer,
W. R.; Muetterties, E. L. J. Am. Chem. Soc. 1976, 98, 8289-8291.
(15) (a) Retboll, M.; Edwards, A. J.; Rae, A. D.; Willis, A. C.; Bennett, M. A.;
Wenger, E. J. Am. Chem. Soc. 2002, 124, 8348-8360. (b) Bennett, M. J.;
Graham, W. A. G.; Stewart, R. P., Jr.; Tuggle, R. M. Inorg. Chem. 1973,
12, 2944-2949. (c) Rausch, M. D.; Gastinger, R. G.; Gardner, S. A.; Brown,
R. K.; Wood, J. S. J. Am. Chem. Soc. 1977, 99, 7870-7876. (d) Grushin,
V. V.; Vymenits, A. B.; Yanovskii, A. I.; Struchkov, Y. T.; Vol’pin, M. E.
Organometallics 1991, 10, 48-49. (e) Bennett, M. A.; Schwemlein, H. P.
Angew. Chem. 1989, 101, 1349-1373. (f) Jones, W. M.; Klosin, J. AdV.
Organomet. Chem. 1998, 42, 147-221. (g) Bennett, M. A.; Wenger, E.
Chem. Ber. 1997, 130, 1029-1042. (h) Cullen, W. R.; Rettig, S. J.; Zhang,
H. Organometallics 1993, 12, 1964-1968.
(16) Begum, R. A.; Sharp, P. R. Organometallics 2005, 24, 2670-2678.
(17) Rausis, T.; Schlosser, M. Eur. J. Org. Chem. 2002, 19, 3351-3358.
9
812 J. AM. CHEM. SOC. VOL. 129, NO. 4, 2007

1,4-Shifts in a Dinuclear Ni(I) Biarylyl Complex
A R T I C L E S
Figure 3. Solid-state molecular structure of 6 as determined by X-ray
crystallography. Ethyl substituents on the phosphine ligands and hydrogen
atoms are omitted for clarity. Selected bond lengths (Å), angles (deg), and
dihedral angles (deg): Ni(1)-Ni(2), 2.3798(6); Ni(1)-C(1), 1.962(3); Ni-
(2)-C(1), 1.974(3); Ni(1)-P(1), 2.1685(1); P(1)-Ni(1)-Ni(2), 161.16-
(3); P(2)-Ni(2)-Ni(1), 132.88(3); C(1)-Ni(1)-Ni(2), 53.03(9); C(1)-
Ni(1)-P(1), 145.79(9); C(1)-C(6)-C(7), 115.5(3). Dihedral angles (deg)
C(1)-C(6)-C(7)-C(8), 30.9(6).
Figure 4. Solid-state molecular structure of 7b as determined by X-ray
crystallography. Hydrogen atoms are omitted for clarity. Selected bond
lengths (Å): Ni(1)-C(1), 1.959(3) and 2.346(3); Ni(1)-P(1), 2.1823(10);
Ni(1)-P(2), 2.1928(10); Ni(1)-Ni(1), 2.3079(8).
shown in Scheme 1. The addition of PEt3 to the trinuclear
complex 6 provided the dinuclear complex [(PEt3)2Ni]2(3,4-
F2C6H2-3′,4′-F2C6H2) (7a), along with 1 equiv of 1a, as shown
1
in Scheme 3. As monitored by H, 31P{1H}, and 19F NMR
-40 °C provided [(PEt3)Ni]3(µ3-4,5-F2C6H2)(µ3-4,5-F2C6H2-
4′,5′-F2C6H2) (6) as dark brown crystals.
spectroscopy, the reaction of 6 with PEt3 is relatively slow.
Complex 7a exhibits two 31P resonances at δ 8.1 and -6.6,
similar to 2a, as well as two 19F environments, and two
1
aromatic H environments. A comparison of the NMR param-
eters of this complex with the intermediate observed almost
immediately after the addition of a source of Ni(PEt3)2 to 1a
allowed us to assign this intermediate as 7a, which confirms
that mechanism B in Scheme 1 operates in the conversion of
1a to 2a.
Unfortunately, compound 7a is too thermally unstable to be
isolated as a pure compound, and the equivalent of 1a produced
from the reaction that forms 7a cannot be separated by
recrystallization. In the absence of excess phosphine, 7a is
thermally unstable and converts to 2a over the course of
∼48 h at room temperature. This occurs without the addition
of any catalyst, though trace amounts of Ni(PEt3)4 were difficult
to remove.
An ORTEP representation of the solid-state molecular
structure of trinuclear complex 6 is shown in Figure 3. The
complex is composed of one uncoupled 4,5-F2C6H2-aryne unit
and a biarylyl unit with the fluorine substituents in the positions
expected in the absence of C-H bond activation. Open
trinuclear nickel clusters are not common,18 although a related
compound has been reported as a minor byproduct from
the reduction of NiCl(2-ClC6H4)(PiPr3)2 with 1% sodium
amalgam.13
When relatively large catalyst loadings of (PEt3)2NiBr2 over
Na/Hg are used in the conversion of 1a to 2a, the presence of
complex 6 is ascertained by 19F NMR spectroscopy, and under
these conditions 6 is found to slowly convert to 2a. Density
functional theory (DFT) calculations on model complexes where
the PEt3 ligands are replaced by PMe3 donors predict that the
formation of the biarylyl intermediate from Scheme 1, [(PEt3)2-
Ni]2(3,4-F2C6H2-3′,4′-F2C6H2), from 1a should be both sub-
stantially exothermic (∆H ) -33 kcal/mol) and thermody-
namically favorable, unlike in some similar systems, where C-C
coupling is nearly thermoneutral.10 One plausible role of the
catalytic Ni(PEt3)2 moiety in the conversion of 1a to 2a is as a
Lewis acid capable of abstracting a PEt3 ligand, which provides
a low activation energy pathway for the coupling of two aryne
complexes.
The addition of PMe3 in lieu of PEt3 to a pentane solution of
6 provided an isolable analogue to 7a, as shown in Scheme 3.
The aryne complex byproduct, (PMe3)(PEt3)Ni(η2-4,5-F2C6H2)
(1b), is considerably less soluble than the dinuclear product,
[(PEt3)(PMe3)Ni]2(3,4-F2C6H2-3′,4′-F2C6H2) (7b), and precipi-
tates from pentane as a yellow solid. The resultant brown
pentane solution can then be concentrated and cooled to
-40 °C to provide X-ray quality crystals of 7b. The 19F NMR
1
spectrum and aromatic region of the H NMR spectrum of 7b
are nearly identical to that of 7a, and the 31P{1H} NMR
spectrum exhibits two phosphorus environments.
The solid-state molecular structure of 7b is shown in Figure
4 and has crystallographically imposed C2 symmetry. The
complex is similar to 2a, in that a biarylyl ligand is bridged by
two Ni(I) moieties. Unlike 2a, the fluorine substituents are in
the 4,5 and 4′,5′ positions of the biarylyl fragment. The bulkier
PEt3 groups occupy the sites trans to the Ni-Ni bond, whereas
the smaller PMe3 donors are adjacent to each other and trans to
the Ni(1)-C(1) bonds. The Ni-Ni bond in 7b is 0.0631(9) Å
shorter than the Ni-Ni bond in 2, likely due to the lesser steric
bulk of PMe3 relative to that of PEt3.
Synthesis of Biarylyl Intermediates. Complex 6 can be used
as a precursor to the intermediate postulated for mechanism B
(18) Pasynkiewicz, S.; Pietrzykowski, A.; Kryza-Niemiec, B.; Zachara, J. J.
Organomet. Chem. 1998, 566, 217-224.
9
J. AM. CHEM. SOC. VOL. 129, NO. 4, 2007 813

A R T I C L E S
Scheme 3
Keen et al.
Scheme 4
Similar to those of 7a, solutions of 7b are thermally unstable
but convert much slower, over the course of 4 days, to an
analogue of 2a, [(PEt3)(PMe3)Ni]2(3,4-F2C6H2-3′,4′-F2C6H2)
(2b). Complex 2b has 19F and 1H NMR spectra that are nearly
identical to those of 2a. The 31P{1H} NMR spectrum has a
virtual triplet at δ -0.52, which is assigned as the PEt3 donor,
and a multiplet at δ -20.2, which can be ascribed to the PMe3
donor.
These alternate routes to intermediate 7a and its analogue
7b allowed for the identification of another intermediate in the
conversion of 1a to 2a. For both the thermal conversion of 7a
to 2a and the conversion of 7b to 2b, intermediates are observed
that exhibit four 19F resonances of equal intensity, a pair of
which have chemical shifts similar to those of 7a, and a pair of
which have chemical shifts similar to those of 2a. During the
isomerization of 7b, it was possible to characterize this
asymmetric intermediate by a combination of 19F and 31P{1H}
NMR spectroscopy. The 31P{1H} NMR spectra exhibit four 31P-
{1H} resonances. Two 31P resonances at δ 0.4 and -1.0 are
associated with the PEt3 moieties and are coupled to each other
mixtures with complexes of similar solubility and thus cannot
be isolated.
Comparison of the concentrations of 8a and 2a, as determined
by peak integrals in the 19F NMR spectrum, demonstrated that
at the start of the isomerization of 7a the asymmetric species
8a and the final product 2a are produced in a 2:1 ratio,
respectively. This implies that a common intermediate exists
that forms 8a twice as rapidly as it forms 2a. The immediate
appearance of 2a also reveals that although 8a can be converted
to 2a, it is not directly on the reaction pathway from 7a. A
similar result is seen for the isomerization of 7b.
Any postulated mechanism for the conversion of 7a,b to 2a,b
must explain this initial 2:1 ratio of 8a,b to 2a,b, which
eliminates many possible reaction pathways. The isomerization
of 7a to 2a requires the migration of the hydrogen substituents
in the ortho (6 and 6′) positions of the biarylyl ligand. It can be
speculated that the Ni-Ni bond reversibly homolytically cleaves
to give a binuclear Ni(I) complex with ortho-C-H agostic
interactions that result in a four-coordinate environment about
the Ni(I) centers. These agostic interactions could aid in the
activation of the ortho-C-H bonds, although there is precedent
for stable three-coordinate Ni(I) complexes.19 The increased rate
of rearrangement observed for 7a compared to that of 7b can
be explained by the increased steric interactions of the adjacent
PEt3 group on opposing metal centers decreasing the barrier to
3
2
with a JPP value of 97 Hz. Both resonances also exhibit JPP
values of 15.0 and 15.3 Hz, respectively, due to coupling with
cis-disposed PMe3 groups, and the resonance at δ -1.0 has an
4
additional poorly resolved JPF value of approximately 5 Hz.
The 31P resonances associated with the PMe3 moieties are a
multiplet at δ -19.5 and a doublet at δ -20.3 with 2JPP ) 15.0
Hz. This intermediate is assigned as the asymmetric complex
[(PEt3)(PMe3)Ni]2(µ-η1:η1-3,4-F2C6H2-4′,5′-F2C6H2) (8b), and
the analogous intermediate observed in the conversion of 7a to
2a can be assigned as [(PEt3)2Ni]2(µ-η1:η1-3,4-F2C6H2-4′,5′-
F2C6H2) (8a). These intermediates are always present as
(19) (a) Bradley, D. C.; Hursthouse, M. B.; Smallwood, R. J.; Welch, A. J.
Chem. Commun. 1972, 872. (b) Eaborn, C.; Hill, M. S.; Hitchcock, P. B.;
Smith, J. D. Chem. Commun. 2000, 691-692. (c) Kitiachvili, K. D.;
Mindiola, D. J.; Hillhouse, G. L. J. Am. Chem. Soc. 2004, 126, 10554-
10555. (d) Bai, G.; Wei, P.; Das, A. K.; Stephan, D. W. J. Chem. Soc.,
Dalton Trans. 2006, 1141-1146. (e) Bai, G.; Wei, P.; Stephan, D. W.
Organometallics 2005, 24, 5901-5908. (f) Melenkivitz, R.; Mindiola, D.
J.; Hillhouse, G. L. J. Am. Chem. Soc. 2002, 124, 3846-3847.
9
814 J. AM. CHEM. SOC. VOL. 129, NO. 4, 2007

1,4-Shifts in a Dinuclear Ni(I) Biarylyl Complex
A R T I C L E S
Ni-Ni bond homolysis. A plausible common intermediate for
complexes 8a and 2a could be obtained from Ni-Ni bond
homolysis, followed by a 1,4-shift20-23 of the Nia(PEt3)2 moiety,
as shown in Scheme 4. Such an intermediate could convert to
8a by an additional 1,4-shift that exchanged Nia and Ha or via
a 1,4-shift that exchanged the locations of Nib and Hb. The
conversion to 2a could only occur via the postulated common
intermediate undergoing a 1,4-shift that exchanged the locations
of Nib and Ha. The only other possible 1,4-shift, which
exchanges Nia and Hb, simply regenerates 7a. If the enthalpic
barriers for these rates are approximately equal, the two
permutations by which 8a can be formed from the intermediate
would result in the value of the rate constant k2 being twice the
value of k3 shown in Scheme 4, despite approximately equal
activation energies. That the enthalpic barriers should be similar
is reasonable, since this is likely a relatively high-energy
intermediate, and the conversions to 2a, 7a, and 8a should all
be exothermic and have early transition states.
Figure 5. Solid-state molecular structure of 10 as determined by X-ray
crystallography. Hydrogen atoms are omitted for clarity. Selected bond
lengths (Å), angles (deg), and dihedral angles (deg): Ni(1)-C(1), 1.971-
(5); Ni(1)-P(1), 2.236(7); C(1)-Ni(1)-P(1), 96.99(6).
Initially, complex 2a forms rapidly from 7a at room tem-
perature, and appreciable conversions to 2a and 8a are observed
after 15 min. The rate at which 2a is produced slows as 7a is
consumed. The 2 equiv of 8a formed then take much longer to
convert to 2a. It is known that the presence of proximal electron-
withdrawing groups stabilize late transition metal-carbon
bonds,24 and DFT calculations show that the model complex
for 2a, [(PMe3)2Ni]2(µ-η1:η1-3,4-F2C6H2-3′,4′-F2C6H2), is pre-
dicted to be 10.8 kcal/mol more stable than the model complex
for 7a, [(PMe3)2Ni]2(µ-η1:η1-4,5-F2C6H2-4′,5′-F2C6H2). DFT
calculations therefore predict that complex 8a should be
approximately 5.4 kcal/mol more stable than 7a. If one presumes
that the activation energies for the reactions associated with the
rate constants k1, k2, and k3 are approximately equal, as
evidenced from the observation of 2 equiv of 8a for every
equivalent of 2a produced early in the reaction, then the rate
constant k-2 should be approximately 2 to 3 orders of magnitude
smaller than k-1, due to an approximately 5 kcal/mol larger
activation energy for this reaction. This result is consistent with
the rapid formation of 2a and 8a from 7a, as well as the
observation that reaction times of approximately 48 h are needed
for the complete conversion of 8a to 2a. These results also
predict that the rate constant k-3 is so small that this reverse
reaction occurs to a negligible degree under the experimental
conditions used.
Figure 6. Experimental mass spectrum for the mixture of isotopomers of
12 shown as white bars, alongside the predicted mass spectrum for an
intermolecular reaction pathway with a 1:4:6:4:1 ratio of nondeuterated
through to tetradeuterated species shown as black bars. The best fit was
obtained using a model with 52% H and 48% D.
environments, two 19F environments, a single 31P environment,
and JFH values consistent with no change in the substitution
pattern of the biarylyl ligand. Complex 9a could not be isolated
because it is thermally unstable and decomposes in the absence
of PEt3 to a variety of unidentified compounds, none of which
are complex 2a. The observation of Ni(PEt3)4 by 31P{1H} NMR
spectroscopy as a byproduct in the conversion of 7a to 9a in
excess PEt3 leads us to believe that this compound is the
mononuclear species (PEt3)2Ni(3,4-F2C6H2-3′,4′-F2C6H2), de-
rived from the loss of Ni(PEt3)2 from 7a as Ni(PEt3)4. The
related mononuclear biarylyl complex (PEt3)2Ni(η2-C6H4C6H4)
is known to be thermally unstable in the absence of excess
phosphine.8 DFT calculations revealed that the reaction of the
model compound (PMe3)2Ni(µ-η1:η1-4,5-F2C6H2-4′,5′-F2C6H2)
with 2 equiv of PMe3 to provide Ni(PMe3)4 and (PMe3)2Ni(η2-
4,5-F2C6H2-4′,5′-F2C6H2) should occur with only a slightly
exothermic ∆H of -7.6 kcal/mol and a slightly unfavorable
∆G°298 of 4.8 kcal/mol. The small value of ∆G°298 indicates
that the conversion of 7a to 9a could be thermodynamically
viable in the presence of excess phosphine. A similar calculation
with an analogue of 2a, (PMe3)2Ni(η2-3,4-F2C6H2-3′,4′-F2C6H2),
demonstrated that this complex is 2 kcal/mol more stable with
respect to this disproportionation reaction. Although we could
not isolate and crystallize 9a, the DFT calculations on the model
complex predict a structure distorted between square planar and
tetrahedral, which is an unusual geometry for a nickel(II)
complex bearing strong-field ligands. For comparison, the solid-
Evidence against Mononuclear Biarylyls as Intermediates.
Alternate pathways to C-H bond activation that do not require
the involvement of Ni(I) in the C-H bond activation steps were
also considered. In the presence of excess phosphine, 7a converts
cleanly into a new product, 9a, which displays two 1H
(20) (a) Masselot, D.; Charmant, J. P. H.; Gallagher, T. J. Am. Chem. Soc. 2006,
128, 694-695. (b) Ma, S.; Gu, Z. Angew. Chem., Int. Ed. 2005, 44, 7512-
7517. (c) Singh, A.; Sharp, P. R. J. Am. Chem. Soc. 2006, 128, 5998-
5999. (d) Liu, Z.; Zhang, X.; Larock, R. C. J. Am. Chem. Soc. 2005, 127,
15716-15717. (e) Karig, G.; Moon, M.-T.; Thasana, N.; Gallagher, T. Org.
Lett. 2002, 4, 3115-3118.
(21) (a) Campeau, L.-C.; Parisien, M.; Jean, A.; Fagnou, K. J. Am. Chem. Soc.
2006, 128, 581-590. (b) Garcia-Cuadrado, D.; Braga, A. A. C.; Maseras,
F.; Echavarren, A. M. J. Am. Chem. Soc. 2006, 128, 1066-1067. (c) Bour,
C.; Suffert, J. Org. Lett. 2005, 7, 653-656.
(22) (a) Miura, T.; Sasaki, T.; Nakazawa, H.; Murakami, M. J. Am. Chem. Soc.
2005, 127, 1390-1391. (b) Oguma, K.; Miura, M.; Satoh, T.; Nomura, M.
J. Am. Chem. Soc. 2000, 122, 10464-10465. (c) Yamabe, H.; Mizuno,
A.; Kusama, H.; Iwasawa, N. J. Am. Chem. Soc. 2005, 127, 3248-3249.
(23) (a) Davies, D. L.; Donald, S. M. A.; Macgregor, S. A. J. Am. Chem. Soc.
2005, 127, 13754-13755. (b) Mota, A. J.; Dedieu, A.; Bour, C.; Suffert,
J. J. Am. Chem. Soc. 2005, 127, 7171-7182.
(24) Hughes, R. P. AdV. Organomet. Chem. 1990, 31, 183-267.
9
J. AM. CHEM. SOC. VOL. 129, NO. 4, 2007 815

A R T I C L E S
Scheme 5
Keen et al.
Scheme 6
state molecular structure of the known compound (PEt3)2Ni-
(η2-C6H4C6H4) (10)8 was determined by X-ray crystallography
and is shown in Figure 5.
cleavage of the Ni-Ni bond to provide phosphine complexes
of Ni(0) and a mononuclear Ni(II) biarylyl is unlikely under
the conditions used to isomerize 1a to 2a.
The geometry at the nickel center in 10 is intermediate
between square planar and tetrahedral, as evidenced by the angle
of 45.7° between the planes defined by P(1)-Ni(1)-P(1) and
C(2)-Ni(1)-C(2). It is unlikely that strain in the biaryl chelate
is the cause of this distortion; the C(1)-Ni(1)-C(1a) angle of
82.88(12)° is not particularly small compared to other Ni(II)
complexes, and the C1-C2-C(2a) angle of 112.29(11)° is
nearly identical to the corresponding angle in the structure of
2a. Also, many complexes similar to 10, where the chelating
bis(diisopropylphosphino)ethane6 or bis(dicyclohexylphosphi-
no)ethane25 ligands are used in lieu of triethylphosphine, are
only slighted distorted from a planar geometry. Related examples
where slightly distorted geometries occur at Ni(II) centers have
been reported.26 It seems that solely steric interactions between
the triethylphosphine ligands and the hydrogen substitutuents
of the biphenyldiyl in the 3,3′ positions cause the distortions
evident in (PEt3)2Ni(η2-C6H4C6H4). This distortion may partly
account for the relative apparent thermodynamic stability of the
dinuclear Ni(I) complexes 2a,b with respect to disproportion-
ation into Ni(0) phosphine complexes and Ni(II) biarylyl
complexes. The fact that 9a was never identified as an
intermediate in the isomerization of 1a to 2a and the failure to
observe the conversion of 9a to 2a both provide evidence against
the participation of mononuclear 9a in this isomerization. These
observations are consistent with the fact that the addition of
PEt3 inhibits the isomerization of 1a. Attempts to generate a
dinuclear Ni(I) complex from 10 by the addition of Br2Ni(PEt3)2
and excess Na/Hg as a source of the Ni(PEt3)2 moiety failed;
the Lewis acidic nature of the Ni(PEt3)2 moiety catalyzed the
decomposition of 10 via the known C-C bond coupling
pathway.8 This provides evidence that the reversible intercon-
version of the dinuclear Ni(I) biarlylyl 7a via heterolytic
Deuterium Labeling Studies. A deuterium labeling crossover
experiment was performed to determine if the isomerization of
7a to 2a was intramolecular or involved intermolecular transfer
of hydrogen. Multiple treatments of 1,2-dibromo-4,5-difluo-
robenzene with LiNiPr2 followed by quenching with CH3OD
provided 1,2-dibromo-3,6-dideutero-4,5-difluorobenzene (11-
d2). This precursor could then be used to produce the dideu-
terated aryne 1a-d2, as shown in Scheme 5.
A solution consisting of approximately equal amounts of 1
and 1-d2 in C6D6 was reacted with a catalytic amount of Br2-
Ni(PEt3)2 over Na/Hg. When the reaction was complete, as
monitored by 19F NMR spectroscopy, the sample was reacted
with Br2 to replace the Ni centers with bromine atoms and
filtered through a plug of alumina to produce 2,2′-Br2-3,3′,4,4′-
F4-C12HnD4-n (12-dn), where n ) 0 to 4 as shown in
Scheme 6.
The mass spectrum of the resultant biaryl isotopomers, 12-
dn, is shown in Figure 6, alongside the spectrum predicted for
an intermolecular scrambling of H/D labels. For an equimolar
mixture of 1a and 1a-d2, a mechanism that occurs with
intermolecular exchange of the hydrogen/deuterium atom of the
activated C-H/C-D bond should result in a 1:4:6:4:1 ratio of
isotopomers that contain 0-4 deuterium atoms. The data do
not fit the modeled spectrum for intramolecular transfer, which
should produce a 1:0:2:0:1 ratio of isotopomers. They are also
poorly fit by the modeled spectrum for a 1:2:2:2:1 ratio of
isotopomers predicted if one of the two hydrogen/deuterium
atoms was transferred intramolecularly and the other intermo-
lecularly. A similar experiment performed in C6D6 with 1a
produced only 2,2′-Br2-3,3′,4,4′-F4-C12H4, which excludes the
involvement of the solvent in these reactions.
By following the course of the reaction by 19F NMR, it was
possible to determine that there is not a large kinetic isotope
effect in the rate of activation of the C-H bonds over the C-D
bonds. The fluorine atoms ortho to deuterium substituents on
(25) Bennett, M. A.; Kopp, M. R.; Wenger, E.; Willis, A. C. J. Organomet.
Chem. 2003, 667, 8-15.
(26) Jones, G. D.; Anderson, T. J.; Chang, N.; Brandon, R. J.; Ong, G. L.; Vicic,
D. A. Organometallics 2004, 23, 3071-3074.
9
816 J. AM. CHEM. SOC. VOL. 129, NO. 4, 2007

1,4-Shifts in a Dinuclear Ni(I) Biarylyl Complex
A R T I C L E S
Table 1. X-ray Crystallographic Data and Structure Refinement Parameters for Complexes 4, 6, 7b, and 10
4
6
7b
10
empirical formula
formula weight
T (K)
C18H33F12NiP2
642.89
173(2)
C36H51F6NiP3
866.81
223(2)
C30H52F4Ni2P4
730.02
173(2)
C24H38NiP2
447.19
173(2)
crystal system
space group
a, Å
b, Å
c, Å
monoclinic
Pn
8.9858(15)
12.415(2)
10.8491(18)
90
95.168(2)
90
1205.4(3)
2
monoclinic
P2(1)/n
12.420(3)
14.719(3)
21.426(5)
90
95.076(4)
90
3901.8(15)
4
monoclinic
C2/c
19.576(2)
28.862(3)
13.9109(15)
90
117.1080(10)
90
6996.3(13)
8
tetragonal
I4(1)/a
11.0586(19)
11.0586(19)
38.711(13)
90
90
90
4734(2)
8
0.961
R, deg
â, deg
γ, deg
V, Å3
Z
µ (mm-1
)
3.506
13221
1.609
37187
1.299
39037
total reflns
26292
No. unique; Rint
residuals: R1, wR2
5356; 0.029
0.045, 0.098
8938; 0.026
0.054, 0.136
7946; 0.055
0.073, 0.124
2827; 0.052
0.046, 0.092
the aromatic rings all exhibit an easily observable isotopic
chemical shift of approximately -0.3 ppm compared to their
hydrogen-substituted analogues, and thus the ratio of the
integrals of these peaks can be used to estimate if rings bearing
C-H bonds are activated faster than the rings bearing C-D
bonds. Throughout the course of the reaction, the ratio of peak
integrals for F atoms ortho to H versus F atoms ortho to D is
approximately 1 for the reactants, intermediates, and products,
although this technique is not accurate enough to determine the
very small inverse kinetic isotope effects that are sometimes
observed for C-H bond activations.27
The intermolecular scrambling observed implies that the rate
constants shown in Scheme 4 represent processes with at least
two elementary steps and that nickel hydride complexes may
exist as intermediates in these reactions, whether the mechanism
involves the oxidative addition of C-H bonds or electrophilic
C-H bond activation. Ni(0) phosphine complexes could
potentially act as bases to form Ni(II) hydrides. All attempts to
identify peaks in the 1H NMR spectra that could be assigned as
Ni hydrides during the conversion of 1a to 2a failed; species
such as the dinuclear bis(phosphine) Ni(I) hydrides28 or cationic
mononuclear Ni(II) hydrides29 should exhibit distinctive shifts
and couplings to phosphorus nuclei but would be expected to
be present in very low concentrations. Mechanisms involving
electrophilic activation that produce acidic intermediates are
unlikely, because any strongly acidic protons in these systems
should be reduced by sodium when these reactions are per-
formed using 1a and catalytic Br2Ni(PEt3)2 over Na/Hg.
as a Lewis acid that can abstract PEt3 from 1a. The proposed
mechanism of C-H bond rearrangement in these biarylyl
complexes utilizes 1,4-shifts, which have been observed in
biphenyl and related aromatic complexes of Pt(II) and Pd(II),20
some of which require high temperatures. Facile 1,4- and 1,5-
shifts have also been reported in a variety of systems that contain
Pd(II)21 and Rh(I).22 None of these examples utilize first row
metals, such as Ni, and they also all differ in that they involve
d8 metal centers. A deuterium labeling crossover experiment
demonstrates the intermolecular scrambling of hydrogen and
deuterium labels, which indicates a true C-H bond-breaking
step must occur; the low barrier to this rearrangement cannot
be due to a concerted mechanism. The observation of 2 equiv
of intermediate 8a for every equivalent of 2a early in the reaction
implies that 8a is not directly on the reaction pathway and
implies that both complexes share a common intermediate. The
ability of Ni(I) complexes to undergo these rearrangements at
room temperature holds promise for increasing the scope of
these C-H bond activations to a variety of aromatic systems.
Experimental Section
General Procedures. Unless otherwise stated, all manipulations were
performed under an inert atmosphere of nitrogen using either standard
Schlenk techniques or an MBraun glove box. Dry, oxygen-free solvents
were employed throughout. Anhydrous pentane was purchased from
Aldrich, sparged with dinitrogen, and passed through activated alumina
under a positive pressure of nitrogen gas and further deoxygenated using
Ridox catalyst columns.30 Deuterated benzene was dried by heating at
reflux with sodium/potassium alloy in a sealed vessel under partial
pressure, then trap-to-trap distilled, and freeze-pump-thaw degassed
three times. Deuterated toluene was purified in an analogous manner
by being heated at reflux over Na. NMR spectra were recorded on
Bruker AMX (300 MHz) or Bruker AMX (500 MHz) spectrometer.
All chemical shifts are reported in parts per million (ppm), and all
coupling constants are in hertz (Hz). For 19F{1H} NMR spectra,
trifluoroacetic acid was used as the external reference at 0.00 ppm. 1H
NMR spectra were referenced to residual protons (C6D5H, 7.15; C7D7H,
δ 2.09) with respect to tetramethylsilane at δ 0.00. 31P{1H} NMR
spectra were referenced to external 85% H3PO4 at δ 0.0. 13C{1H} spectra
were referenced relative to solvent resonances (C6D6, δ 128.0; C7D8,
δ 20.4). UV/visible spectra were obtained on a Varian Carey 50
spectrophotometer. Elemental analyses were performed by the Centre
Conclusions
Both the identification of intermediates and the relative rate
at which they are formed and consumed have allowed for a
detailed understanding of the C-C bond forming and C-H bond
activation steps that isomerize the aryne complex 1a into the
biarylyl complex 2b. This study demonstrates that the initial
step involves C-C bond formation, depicted as mechanism B
in Scheme 1, and is due to the behavior of the Ni(PEt3)2 moiety
(27) (a) Jones, W. D. Acc. Chem. Res. 2003, 36, 140-146. (b) Churchill, D.
G.; Janak, K. E.; Wittenberg, J. S.; Parkin, G. J. Am. Chem. Soc. 2003,
125, 1403-1420.
(28) (a) Bach, I.; Goddard, R.; Kopiske, C.; Seevogel, K.; Poerschke, K.-R.
Organometallics 1999, 18, 10-20. (b) Jonas, K.; Wilke, G. Angew. Chem.,
Int. Ed. 1970, 9, 312-313. (c) Vicic, D. A.; Jones, W. D. J. Am. Chem.
Soc. 1997, 119, 10855-10856.
(30) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.; Timmers,
(29) Schunn, R. A. Inorg. Chem. 1976, 15, 208-212.
F. J. Organometallics 1996, 15, 1518-1520.
9
J. AM. CHEM. SOC. VOL. 129, NO. 4, 2007 817

A R T I C L E S
Keen et al.
for Catalysis and Materials Research, Windsor, Ontario, Canada. The
compounds (Et3P)2NiBr(2-Br-4,5-F2-C6H2),11 Ni(PEt3)4,31 and 1-F-2,3-
I2C-6H317 were prepared by literature procedures. The complex Br2Ni-
(PEt3)2 was prepared by the addition of PEt3 to commercially available
(DME)NiBr2 followed by recrystallization from pentane. The com-
pounds PEt3, B(C6F5)3, Na, and Hg were purchased from Aldrich. The
compound B(C6F5)3 was sublimed prior to use.
6.7 Hz, aromatic 6-H), 7.18 (ddd, 1H, JHF ) 3.3 Hz, JHH ) 6.7 Hz,
JHH ) 7.4 Hz, aromatic-5-H), 7.41 (dd, 1H, JHF ) 6.3 Hz, JHH ) 7.4
Hz, aromatic-4-H). 31P{1H} NMR (C6D6, 298 K, 202.47 MHz): δ 29.7
(br s), 27.1 (br s). 19F NMR (C6D6, 298 K, 282.48 MHz): δ -28.4
(ddd, JHF ) 6.3 Hz, JHF ) 3.3 Hz, JHF ) 3.3 Hz). 13C{1H} NMR (C6D6,
298 K, 125.27 MHz): δ 9.0 and 9.3 (br, PCH2Me), 20.0 and 20.7 (br,
PCH2), 113.0 (d, J ) 32.5 Hz), 118.0 (s), 129.4 (d, J ) 97 Hz), 130.2
1
Synthesis of (PEt3)4Ni2(µ-η1:η1-3,3′,4,4′-F4-C12H4) (2a). To a stirred
solution of NiBr(2-Br-4,5-F2C6H2)(PEt3)2 (0.5 g, 0.88 mmol) in 20 mL
of pentane over excess 1% Na/Hg amalgam (40 g) was added a solution
of Br2Ni(PEt3)2 (0.1 g, 0.22 mmol) in 15 mL of pentane. The solution
was then stirred slowly for 20 h. The solution was then filtered through
Celite, and the volume was reduced to ca. 20 mL. The solution was
cooled to -40 °C, upon which dark brown crystals of (PEt3)4Ni2-
(s), 153.9 (br d, J ) 10 Hz), 163.1 (d, JCF ) 254.8 Hz).
Synthesis of (PEt3)3Ni3(µ3-4,5-F2C6H2)(µ3-4,5-F2C6H2-4′,5′-F2C6H2)
(6). To a stirred solution of Ni(η2-C6H2F2)(PEt3)2 (0.65 g, 0.0016 mol)
in 25 mL of pentane was added dropwise a solution of tris-
(pentafluorophenyl)borane (B(C6F5)3) (0.82 g, 0.0016 mol) in 15 mL
of pentane. The resulting brown solution was stirred for 15 min and
then filtered to remove the white precipitate. The solvent was then
reduced to half volume, and the product was recrystallized at -40 °C
to yield (PEt3)3Ni3(η2-C6H2F2)(η2-C6H2F2-C6H2F2) as dark brown
crystals in a 31% yield. 1H NMR (C6D6, 298 K, 500 MHz): δ 0.66 (br
m, 27 H, CH3), 1.00(br m, 18 H, CH2), 6.97(dd, 2 H, JHF ) 12.3 Hz,
JHF ) 7.2 Hz), 7.34 (vt, 2 H, JHF ) 9.5 Hz), 7.79(dd, 2 H, JHF ) 10.6
Hz, JHF ) 8.8 Hz). 31P{1H} NMR (C6D6, 298 K, 202.47 MHz): δ
13.2 (br, 2P), 5.1 (br, 1P). 13C{1H} NMR (C6D6, 298 K, 75.50 MHz):
1
(C6H2F2)2 were obtained (0.22 g, 60% yield). H NMR (C6D6, 20 °C,
500 MHz): δ 1.05 (br, 36H, CH3), 1.90 (br, 24H, CH2), 6.55 (m, 4H).
1H NMR (C6D6, -60 °C, 500 MHz, aromatic region): δ 6.44 (ddd,
2H, 3JHH ) 8.0, 4JHF ) 4.0, 5JHF ) 0.7 Hz), 6.64 (ddd, 3JHH ) 8.0 Hz,
4
4JH(a)F(b) ) 7.3 Hz, JH(a)F(a) ) 10.2 Hz). 31P{1H} NMR (C6D6, 20 °C,
202.47 MHz): δ 8.8 (br s), -3.8 (br s). 31P{1H} NMR (C6D6, -60 °C,
202.47 MHz): δ -3.8 (vt, JPP ) 5.4 Hz), 8.8 (vtdd, JPP ) 5.4 Hz, JFP
) 7.1, 11.3 Hz). 19F NMR (C6D6, 20 °C, 282.48 MHz): δ -34.3 (dddd,
4JH(a)F(a) ) 10.2 Hz, 4JH(b)F(a) ) 4.0 Hz, 3JFF ) 35.6 Hz, JFP ) 6.0 Hz),
δ 108.25 (d, 2JCF ) 12.6), 111.8 (d, 2JCF ) 18.9 Hz), 131.0 (d, 2JCF
12.6 Hz), 135.8 (s, ipso C), 146.1 (dd, JCF ) 251.5 Hz, JCF ) 18.9
)
1
2
3
1
2
1
-71.1 (ddd, JFP ) 11.5 Hz, JH(a)F(b) ) 7.3 Hz, JFF ) 35.6 Hz). 13C-
Hz), 148.4 (dd, JCF ) 213.8, JCF ) 12.6), 149.1 (dd, JCF ) 232.6,
{1H} NMR (C6D6, 298 K, 75.50 MHz): δ 8.2 and 9.5 (br, CH3), 17.1
and 21.5 (br, CH2), 109.9 (d, 2JCF ) 19.8 Hz), 114.23 (s, ipso-C), 150.5
2JCF ) 12.6), 151.4 (s, ipso C), 168.1 (s, ipso C). 19F NMR (C6D6, 298
3
K, 282.48 MHz): δ -69.1 (ddd, JFF ) 19.9 Hz, JHF ) 12.3 Hz, JHF
1
2
2
(dd, JCF ) 244.3 Hz, JCF ) 22.0 Hz), 151.5 (d, JCF ) 17.6 Hz),
155.7 (dd, 1JCF ) 226.1 Hz, 2JCF ) 14.7 Hz), 155.9 (s, ipso-C). Anal.
Calcd: C, 53.11; H, 7.92. Found: C, 52.67; H, 7.81. UV/vis: ꢀ )
72 000 L mol-1 cm-1. λmax ) 428 nm.
) 8.8 Hz), -66.5 (vt, JHF ) 9.5 Hz), -61.6 (ddd, 3JFF ) 19.9 Hz, JHF
) 10.6 Hz, JHF ) 7.2 Hz). Anal. Calcd for C39H77F6Ni3P3: C, 49.88;
H, 5.93. Found: C, 49.56; H, 5.91.
Synthesis of [(PEt3)2Ni]2(µ-η2:η2-4,5-F2C6H2-4′,5′-F2C6H2) (7a).
To a solution of 6 (0.12 g, 0.154 mmol) in 20 mL of pentane was
added PEt3 (0.06 mL, 0.406 mmol, 2.6 equiv) and stirred for 5 min.
The mixture was then reduced to ca. 10 mL and placed in the freezer
Synthesis of [(PEt3)2Ni]2(µ-η2:η2-C6H2F2) (3). To a stirred solution
of Ni(η2-C6H2F2)(PEt3)2 (80 mg, 0.197 mmol) in 15 mL of pentane
over excess 1% Na/Hg amalgam (50 g) was added dropwise, over 1 h,
a dilute solution of Br2Ni(PEt3)2 (89 mg, 0.197 mmol) in 25 mL of
pentane. The solution was then stirred for 20 min before being filtered
through Celite, and the solvent was reduced to ca. 20 mL. The solution
was then cooled to -40 °C, and pale orange crystals of Ni2(µ-η2:η2-
1
at -40 °C. The products separated as a dark oily product. H, 31P-
{1H}, and 19F NMR data were used to identify (PEt3)4Ni2(η2-C6H2-
3,4-F2)2 (1a) and a second product assigned as [(PEt3)2Ni]2(4,5-F2C6H2-
1
4′,5′-F2C6H2) (7a). H NMR (C6D6, 298 K, 300 MHz): δ 1.09 (br,
1
C6H2F2)(PEt3)2 were obtained (60 mg, 43% yield). H NMR (C6D6,
36H, CH3), 1.86 (br, 24H, CH2), 6.58 (dd, JHF ) 7.6, 11.9 Hz, aromatic),
6.77 (dd, JHF ) 9.4, 11.2 Hz, aromatic). 31P{1H} NMR (C6D6, 20 °C,
202.47 MHz): δ 8.1 (br), -6.6 (br). 19F NMR (C6D6, 298 K, 282.48
20 °C, 300 MHz): δ 0.91 (m, 18H, CH3), 1.42 (m, 12H, CH2), 7.33 (t,
3
2H, JHF ) 7.3 Hz). 31P{1H} NMR (C6D6, 20 °C, 121.54 MHz): δ
14.8 (s). 19F NMR (C6D6, 20 °C, 282.46 MHz): δ -74.0 (m). Anal.
Calcd: C, 51.32; H, 8.90. Found: C, 51.68; H, 8.70.
MHz): δ -72.7 (m, JHF ) 11.9, 9.4 Hz, JFF ) 20.0 Hz), -73.0 JHF
11.2, 7.6 Hz, JFF ) 20.0 Hz).
)
Synthesis of [(PEt3)(PMe3)Ni]2(µ-η1:η1-4,5-F2C6H2-4′,5′-F2C6H2)
(7b). To a solution of 6 (0.111 g, 0.142 mmol) in 15 mL of pentane
was added PMe3 (0.044 mL, 0.43 mmol). The solution was stirred for
5 min and then cooled to -40 °C. The first compound to precipitate
out of solution was Ni(η2-C6H2F2)(PMe3)(PEt3), 1b, as an orange solid.
This was filtered off, and the remaining solution was concentrated and
cooled to -40 °C. Brown crystals of (PEt3)2(PMe3)2Ni2(η2-C6H2-3,4-
Synthesis of NiI(3-F-2-I-C6H3)(PEt3)2 (4). A suspension of Ni-
(COD)2 (0.55 g, 0.002 mol) in 25 mL of pentane was cooled to 0 °C.
A solution of PEt3 (0.59 g, 0.005 mol) was added dropwise to the
solution and stirred for 5 min. A solution of 2,3-diiodofluorobenzene
(1.02 g, 0.003 mol) in 25 mL of pentane was then added. The mixture
was stirred for 10 min at 0 °C and then for 4 h at room temperature.
The solvent was then removed under vacuum, and the residue was
extracted with pentane and filtered through Celite. The solution was
then placed in the freezer at -40 °C. Dark yellow crystals of NiI(2-
1
F2)2 were isolated (0.03 g, 32% yield). H NMR (C6D6, 298 K, 300
MHz): δ 1.03 (m, 18H, CH3, PEt3), 1.15 (d, 18H, CH3 of PMe3), 1.38
(m, 12H, PCH2), 6.57 (dd, J ) 7.1 Hz, J ) 11.6 Hz, 2H, aromatic),
6.70 (dd, 2H). 31P{1H} NMR (C6D6, 298 K, 202.47 MHz): δ -0.3
(vt, J ) 10.3 Hz, Ni-PEt3), -21.2 (vt, J ) 10.3 Hz, Ni-PMe3). 19F
NMR (C6D6, 298 K, 282.48 MHz): δ -73.0 (m), -73.6 (m). Anal.
Calcd for C30H52F4Ni2P4: C, 49.36; H, 7.18. Found: C, 49.56; H, 7.32.
Complex 9a. To a solution of 6 (0.14 g, 0.179 mmol) in 15 mL of
pentane was added PEt3 (0.20 mL g, 1.35 mmol) and stirred. The
mixture was then reduced to ca. 10 mL and placed in the freezer at
-40 °C. The products separated as a pale brown oil. 1H, 31P{1H}, and
19F NMR data were used to identify excess PEt3, Ni(PEt3)4, (PEt3)4-
Ni2(η2-C6H2-3,4-F2)2 (1), and a product assigned tentatively as [(PEt3)2-
Ni](4,5-F2C6H2-4′,5′-F2C6H2) (9a); complex 9a is thermally unstable
in the absence of Pet3 and could not be isolated pure. 1H NMR (C6D6,
298 K, 300 MHz): δ 6.79 (dd, 2H, JHF ) 8.6 Hz, JHF ) 10.7 Hz,
aromatic H), 6.81 (dd, 2H, JHF ) 7.5 Hz, JHF ) 11.8 Hz, aromatic H).
1
I-3-FC6H3)(PEt3)2 were isolated (0.27 g, 30% yield). H NMR (C6D6,
298 K, 300 MHz): δ 0.96 (m, 18H), 1.52 (m, 12H), 6.44 (dd, 1H, JHF
) 8.8 Hz, JHH ) 7.6 Hz), 6.65 (m, 2H). 31P{1H} NMR (C6D6, 298 K,
202.47 MHz): δ 9.9 (s). 19F NMR (C6D6, 298 K, 282.48 MHz): δ
-13.2 (d, JHF ) 8.8 Hz). Anal. Calcd for C18H33I2FP2Ni: C, 33.63; H,
5.17. Found: C, 33.78; H, 5.24.
Synthesis of Ni(η2-3-F-C6H3)(PEt3)2 (5). A solution of NiI(2-I-3-
FC6H3)(PEt3)2 (0.10 g, 0.26 mmol) in 25 mL of pentane was stirred
over excess 1% Na/Hg (40 g) for 1 h. The solution was then filtered
through Celite. The solvent was then removed under vacuum, leaving
1
a dark yellow oil that freezes below 0 °C (0.037 g, 62% yield). H
NMR (C6D6, 298 K, 300 MHz): δ 0.97 (overlapping br m, 18H, CH3),
1.46 and 1.56 (br m, 12H, CH2), 6.93 (dd, 1H, JHF ) 3.3 Hz, JHH
)
(31) Cundy, C. S. J. Organomet. Chem. 1974, 69, 305-310.
9
818 J. AM. CHEM. SOC. VOL. 129, NO. 4, 2007

1,4-Shifts in a Dinuclear Ni(I) Biarylyl Complex
A R T I C L E S
3
3
31P{1H} NMR (C6D6, 298 K, 202.47 MHz): δ 3.6 (br). 19F NMR (C6D6,
isomer; JFF ) 22.0 Hz, JHF ) 9.3 Hz, 4-H-5-D isomer), -55.6
(overlapping m, 3JFF ) 22.0 Hz, 4,5-D2 and 4-D-5-H isomer). GC/MS
(mass, intensity): 382, 6.99; 383, 34.15; 384, 60.97; 385, 94.70; 386,
99.97; 387, 81.06; 388, 53.07; 389, 25.79, 390, 7.40; 391, 0.786.
An analogous procedure using only 1a produced 2,2′-dibromo-
3,3′,4,4′-tetrafluorobiphenyl. 19F NMR (CDCl3, 300 K, 282.48 MHz):
298 K, 282.48 MHz): δ -67.7 (ddd, JFF ) 20.3 Hz, JHF ) 11.8 Hz,
4
4JHF ) 8.6 Hz), -65.4 (dddd, JFF ) 20.3 Hz, JHF ) 10.7 Hz, JHF
)
7.5 Hz, JPF ) 2.3 Hz).
Synthesis of 1,2-Dibromo-3,6-dideutero-4,5-difluorobenzene
(11-d2). To a stirred solution of 1,2-dibromo-4,5-difluorobenzene (18.4
mmol, 5 g) in 25 mL of 5% THF in Et2O at -78 °C was added dropwise
a suspension of lithiumdiisopropylamide (18.4 mmol, 1.97 g, 1 equiv)
in 20 mL of 5% THF in Et2O. The solution was stirred for 20 min.
DOCH3 (18.4 mmol, 0.608 g, 1 equiv) was added to the mixture
dropwise. The solution was then allowed to slowly warm to room
temperature over 1.5 h, at which point a sample was taken to monitor
progress. This process was repeated for a total of six additions of LDA
and DOCH3. The solvent was removed under vacuum, and the residue
was extracted with four portions of ca. 50 mL of pentane. The pentane
was removed under vacuum, which afforded a yellow oil. The oil was
cooled to 0 °C to provide colorless crystals of 1,2-Br2-4,5-F2C6D2 (1.0
g, 20% yield). 19F NMR (C6D6, 300 K, 282.48 MHz): δ -58.5 (virtual
pentet, J ) 1.3 Hz). Anal. Calcd for C6F2D2: C, 26.31; D, 1.47.
Found: C, 26.55; D, 1.53.
3
4
5
δ -48.2 (ddd, JFF ) 21.9 Hz, JHF ) 7.4 Hz, JHF ) 2 Hz), -55.5
(dd, JFF ) 21.9 Hz, JHF ) 9.4 Hz, JHF ) 5.1 Hz).
3
3
4
X-ray Crystallography. The X-ray structures were obtained at low
temperature, with the crystal covered in Paratone and placed rapidly
into the cold N2 stream of the Kryo-Flex low-temperature device. The
data were collected using the SMART32 software on a Bruker APEX
CCD diffractometer using a graphite monochromator with Mo KR
radiation (λ ) 0.71073 Å). A hemisphere of data was collected using
a counting time of 10-30 s per frame. Details of crystal data, data
collection, and structure refinement are listed in Table 1. Data reductions
were performed using the SAINT33 software, and the data were
corrected for absorption using SADABS.34 The structures were solved
by direct methods using SIR9735 and refined by full-matrix least-squares
on F2 with anisotropic displacement parameters for the non-hydrogen
atoms using SHELXL-9736 and the WinGX37 software package, and
thermal ellipsoid plots were produced using ORTEP32.38 Complex 6
features disorder of some of the ethyl substituents on the PEt3 donors;
the carbon atoms were treated as 2-fold disordered and refined
isotropically. The hydrogen atoms associated with the disordered
fragments were omitted from the model.
Synthesis of NiBr(2-Br-4,5-F2C6D2)(PEt3)2. A stirred suspension
of Ni(COD)2 (2.6 mmol, 0.715 g) in 20 mL of pentane was cooled to
0 °C, and PEt3 (5.1 mmol, 0.604 g, 2 equiv) was added followed by a
solution of 1,2-dibromo-3,6-dideutero-4,5-difluorobenzene (2.6 mmol,
0.7 g) in ca. 15 mL of pentane. The mixture was stirred for 10 min at
0 °C and then for 4 h at room temperature. The solvent was removed
under vacuum. The residue was extracted with 50 mL of pentane and
filtered through Celite. The solvent was then reduced to 25 mL and
cooled to -40 °C to provide NiBr(2-Br-4,5-F2C6D2)(PEt3)2 as a yellow
Calculations. Ab initio DFT calculations were performed using the
hybrid functional B3LYP39 method with the Gaussian 03 package.40
The basis functions used were the 6-311+G(d) set, provided in the
Gaussian 03 program.
1
solid from pentane at (0.752 g, 51% yield). H NMR (C6D6, 300 K,
300 MHz): δ 0.91 (m, 18H, CH3), 1.32 (m, 12H, CH2). 31P{1H} NMR
Acknowledgment. Acknowledgment is made to the National
Sciences and Engineering Council (NSERC) of Canada and the
Ontario Research and Development Challenge Fund (ORDCF)
for their financial support, and to Paul R. Sharp for his helpful
comments and for providing a preprint of a relevant publication.
(C6D6, 300 K, 121.54 MHz): δ 10.6 (s). 19F NMR (C6D6, 300 K, 282.48
MHz): δ -67.2 (dt, 3JFF ) 20.4 Hz, JFD ) 1.6 Hz), -64.7 (dt, 3JFF
)
20.4 Hz, JFD ) 1.6 Hz). Anal. Calcd for C18H30D2Br2F2P2Ni: C, 38.00;
D, 6.02. Found: C, 38.04; D, 5.99.
Synthesis of (PEt3)2Ni(η2-C6D2-4,5-F2), (1a-d2). The synthesis of
1a-d2 was performed in an analogous manner to the known literature
procedure for the preparation of 1a. 31P{1H} NMR (C6D6, 298 K,
202.47 MHz): δ 29.5 (br). 19F NMR (C6D6, 300 K, 282.48 MHz): δ
-60.3 (br).
Supporting Information Available: Full details of ref 40;
crystallographic information in CIF format for 4, 6, 7b, and
10; ORTEP depiction of the solid-state molecular structure of
4, as determined by X-ray crystallography; optimized coordi-
nates and energies for DFT calculations. This material is
available free of charge via the Internet at http://pubs.acs.org.
Synthesis of [(PEt3)2Ni]2(µ-η1:η1-3,4-F2C6D2-3′,4′-F2C6D2), (2a-
d4). To a solution of 1a-d2 (0.04 g, 0.098 mmol) in 2 mL of pentane
was added (PEt3)2NiBr2 (0.004 g, 0.0088 mmol, 0.09 equiv) and 0.5 g
of 1% Na/Hg. The solution was stirred for 24 h. No aromatic proton
resonances associated with the biarylyl complex were observed in the
1H NMR spectrum. 19F NMR (C6D6, 300 K, 282.48 MHz): δ -34.3
JA067112W
(32) SMART, Molecular analysis research tool; Bruker AXS: Madison, WI, 2001.
(33) SAINTPlus, Data reduction and correction program; Bruker AXS: Madison,
WI, 2001.
(34) SADABS, An empirical absorption correction program; Bruker AXS:
Madison, WI, 2001.
(35) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.; Giacovazzo,
C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna, R. J. Appl.
Crystallogr. 1999, 32, 115-119.
(36) Sheldrick, G. M. SHELXL-97; University of Go¨ttingen: Go¨ttingen,
Germany, 1997.
(37) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837-838.
(38) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.
(39) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(40) Frisch, M. J.; et al. Gaussian 03, revision C.01; Gaussian, Inc.: Wallingford,
CT, 2004.
3
3
(br d, JFF ) 33.8 Hz), -71.4 (br d, JFF ) 33.8 Hz).
Deuterium Labeling Crossover Study. Approximately equal masses
of 1a and 1a-d2 (∼10 mg) were dissolved in C6D6. The solution was
then stirred with 1 mg of Br2Ni(PEt3)2 and 0.5 g of 1% Na/Hg, and the
reaction was monitored by 19F NMR spectroscopy. When the conversion
to 2a-dn (n ) 1-4) was complete, the solution was quenched by the
direct addition of Br2, and the solution was passed through a plug of
acidic alumina and analyzed by 19F NMR and GC/MS. 19F NMR
(CDCl3, 300 K, 282.48 MHz): δ -48.2 (overlapping multiplets), -55.3
(overlapping m, 3JFF ) 22.0 Hz, 3JHF ) 9.3 Hz, 4JHF ) 5.0 Hz, 4,5-H2
9
J. AM. CHEM. SOC. VOL. 129, NO. 4, 2007 819