On the catalytic hydrodefluorination of fluoroaromatics using nickel complexes: The true role of the phosphine

Article abstract of DOI:10.1021/ja412268y

Homogeneous catalytic hydrodefluorination (HDF) of fluoroaromatics under thermal conditions was achieved using nickel(0) compounds of the type [(dippe)Ni(??²-C₆F_6-nH_n)] where n = 0-2, as the catalytic precursors. These complexes were prepared in situ by reacting the compound [(dippe)Ni(?-H)]₂ with the respective fluoroaromatic substrate. HDF seems to occur homogeneously, as tested by mercury drop experiments, producing the hydrodefluorinated products. However, despite previous findings by other groups, we found that these HDF reactions were actually the result of direct reaction of the alkylphosphine with the fluoroaromatic substrate. This metal- and silane-free system is the first reported example of a phosphine being able to hydrodefluorinate on its own.

Full text of DOI:10.1021/ja412268y

Article
pubs.acs.org/JACS
On the Catalytic Hydrodefluorination of Fluoroaromatics Using
Nickel Complexes: The True Role of the Phosphine
Alma Arevalo, Adrian Tlahuext-Aca, Marcos Flores-Alamo, and Juventino J. García*
́
Facultad de Química, Universidad Nacional Auton
́ ́
oma de Mexico, Circuito Interior, Ciudad Universitaria, Mexico City 04510,
Mexico
́
S
* Supporting Information
ABSTRACT: Homogeneous catalytic hydrodefluorination
(HDF) of fluoroaromatics under thermal conditions was
achieved using nickel(0) compounds of the type [(dippe)Ni-
(η2-C6F6‑nHn)] where n = 0−2, as the catalytic precursors.
These complexes were prepared in situ by reacting the com-
pound [(dippe)Ni(μ-H)]2 with the respective fluoroaromatic
substrate. HDF seems to occur homogeneously, as tested by mercury drop experiments, producing the hydrodefluorinated
products. However, despite previous findings by other groups, we found that these HDF reactions were actually the result of
direct reaction of the alkylphosphine with the fluoroaromatic substrate. This metal- and silane-free system is the first reported
example of a phosphine being able to hydrodefluorinate on its own.
Other examples of homogeneous catalytic HDF have appeared.
In 2005, Holland14 reported the reaction of perfluorinated
aromatic compounds with silanes using 20 mol% of Fe(II)
complexes as catalyst precursors. Another catalytic example
using Ru complexes was reported by Whittlesey,15 employing
1 mol% of a ruthenium catalyst using Et3SiH as hydrogen
source. A recent report shows catalytic HDF using [(NHC)-
AuH]16 in the presence of DMAP and silanes. Recent examples
of catalytic reactions of fluoroarenes with nickel compounds
include the works reported by Cao17 and Radius.18
In the case of heterofluoroaromatics, one of the first exam-
ples of oxidative addition of the C−F bond in pentafluoropyr-
idine (PFP) was reported by Perutz,6b using Ni(COD)2 in the
presence of triethylphosphine or with [Ni(PEt3)4], resulting in
the formation of trans-[Ni(PEt3)2(C5F4N)F]. Other stoichio-
metric reactions with transition metals and PFP have been
found to occur at the 4-position for palladium or platinum,6b
rhodium,20 and gold,21 and at the 2-position dominantly for
nickel,19 [Rh(SiPh3)(PMe3)3].22 Examples of catalytic HDF
of PFP are relatively scarce, with relevant examples using
rhodium,20 gold,21 palladium,23 and copper.24 Herein we report
our findings on the search for a catalytic HDF system initially
based on nickel compounds and the surprising results of the
study where we found that PEt3 was the actual HDF agent.
INTRODUCTION
■
Organofluorine chemistry has grown enormously in significance
and scope in the past three decades. Fluorine is unique in many
ways for its ability to replace hydrogen in a variety of organic com-
pounds without important distortion of the geometry. However,
naturally occurring organofluorines are rare.1 These fluoro com-
pounds have a wide range of uses in industry.2 Fluorine sub-
stituents increase the hydrophobicity of an organic substrate,
which can increase the rate of transport of the drug to the active
site and also enhance resistance to metabolic oxidation, which
means less toxic byproducts and increased safety of a drug.3
The synthesis of polyfluoroaromatic molecules usually
requires several steps and the use of harsh conditions (200−
600 °C). Yields range from moderate to good (50−90%), and
the reactions occur via electrophilic or nucleophilic fluorina-
tions4 conducted on activated enol-like functionalities, ethers,
or allyl silanes.5 An alternative approach for the synthesis of
partially fluorinated products is the cleavage of C−F bonds of
perfluorinated compounds mediated by transition metals,6
involving an oxidative addition of the C−F bond, followed by
fluorine elimination with the use of silanes or hydrogen (eq 1),
in a reaction known as hydrodefluorination (HDF).6b
LnM
H2
RF‐F ⎯⎯⎯→ RF‐MLn−2‐F ⎯⎯⎯⎯⎯→⎯ RF‐H + LnM
(1)
RESULTS AND DISCUSSION
“−HF”
■
Reaction of 1 with Hexafluorobenzene (C6F6). The
reaction of a blood red solution of [(dippe)Ni(μ-H)]2 (1)
with C6F6 in THF results in an immediate color change to
yellow-brown (eq 2). The molecular structure of 2 has been
Perutz reported in 1989 the coordination of fluoroaromatics
to a rhodium center.7 He found that [CpRh(PMe3)(C2H4)]
forms an isolable η2-complex with C6F6. Additional examples of
C−F bond oxidative addition with transition metal compounds
include those using late transition metals, such as rhodium8 and
nickel,9−11 and early transition metals, such as zirconium.12
Milstein reported the first catalytic version of that activation
leading to HDF by transition metals in 1994,13 using rhodium.
−H2
1 + C6F ⎯⎯⎯→ 2
(2)
6
Received: December 6, 2013
Published: January 16, 2014
© 2014 American Chemical Society
4634
dx.doi.org/10.1021/ja412268y | J. Am. Chem. Soc. 2014, 136, 4634−4639

Journal of the American Chemical Society
Article
very same reaction conditions, the C−F oxidative addition
product, complex 5, was isolated and fully characterized (vide
infra).
Reaction of 1 with Pentafluorobenzene (C6F5H). To
shed more light on the reactivity with fluoroaromatics, the
reaction of complex 1 with C6F5H was investigated. Complex 4
was obtained in analogy to complex 2 (vide supra) and closely
related nickel compounds bearing trialkyl monophosphines.26
The 31P{1H} NMR spectrum at room temperature for
complex 4 displays a very broad signal centered at 71.8 ppm,
indicative of a fluxional behavior, which corresponds to the
product shown in Figure 2. The corresponding 1H NMR
Figure 1. ORTEP drawing (50% probability) for complex 2. Selected
bond distances (Å): Ni1−P1 2.1648(8), Ni1−P2 2.1613(8), Ni1−C1
1.916(3), Ni1−C2 1.941(3), C2−C1 1.440(4), C1−C6 1.448(4),
C2−C3 1.467(4), C6−C5 1.323(4). Selected bond angles (deg): P1−
Ni1−P2 90.38(3), C2−Ni1−C1 43.84(12), C1−C2−C3 119.9(2),
C3−C4−C5 117.8(2), C4−C5−C6 122.0(3).
Figure 2. Complex 4.
spectrum features an upfield-shifted broad multiplet at δ 3.0 for
the aromatic proton, and the signals due to the alkyl
substituents at δ 1.35 (m, CH3), 1.92 (d, −CH2−), and δ 2.3
(−CH−). The 19F NMR spectrum shows five signals located at
δ −132.1, −147.7, −153.2, −160.9, and −184.5, indicative of
five different fluorine atoms and of a rather unsymmetrical
coordination of the fluoro-aromatic moiety. Upon gentle
warming of a sample of complex 4 to 80 °C for 4 h, compound
5 was obtained (Figure 3).
determined by X-ray crystallography (130 K), and the
corresponding ORTEP diagram is depicted in Figure 1.
The complex displays a C2 axis of symmetry, which lies in the
plane of the C6F6 ligand and passes through the midpoints of
the bonds C2−C3 and C5−C6. In 2, two [(dippe)Ni] moieties
are antifacially coordinated to adjacent CC bonds of a
bridging C6F6 ligand. The Ni1−C1 and Ni2−C4 bond lengths
are the same, 1.916(3) Å, and are slightly shorter than the Ni1−
C2 and Ni2−C3 bond lengths, which are 1.941(3) and
1.940(3) Å, respectively. The arene ring shows significant
elongation of the coordinated double bonds from typical
aromatic values, for instance, C1−C2 and C3−C4 distances of
1.440(4) and 1.434(4) Å, respectively. The fluorine substit-
uents show a considerable nonplanarity. The remaining C−C
bond lengths of the C6F6 moiety are also significantly per-
turbed: the C5−C6 distance of 1.323(4) Å is shorter than
expected for a C−C bond in C6F6 (1.394 Å),25 whereas C2−C3
(1.467(4) Å), C4−C5 (1.453(4) Å), and C6−C1 (1.448(4) Å)
are closer to single bonds. These distances are evidence of
electron density donation/backbonding between the two
[(dippe)Ni] fragments and the fluoroaromatic ring; therefore,
the formally sp2 carbons are more hybridized to sp3-like. The
Ni−C bond lengths of 1.928(3) Å (mean) are as expected for
Ni(0)−η2-fluoroarene complexes.9,26
Figure 3. C−F oxidative addition product, complex 5.
The room-temperature 31P{1H} NMR spectrum for 5 shows
the signals expected for the structure depicted in Figure 3: two
2
doublet of doublets located at δ 70.4 (2JP−F = 30 Hz, JP−P
=
10 Hz) assigned to the phosphorus trans to C, and the other at
δ 65.6 (2JP−F = 34 Hz, 2JP−P = 10 Hz) for the phosphorus trans
to F. The 19F NMR spectrum shows four multiplets located at
δ −115.6, −141.2, −162.6, and −164.4 for the four non-
The room-temperature 31P{1H} NMR spectrum of 2 features
a resolved multiplet centered at 70.7 ppm (3JP−F = 12 Hz) due
equivalent fluorine atoms at the aromatic ring and the Ni−F
1
2
to coupling to the fluorine atoms in the C6F6 moiety. The H
resonance assigned at δ −345.4 as a multiplet with JP−F
=
NMR spectrum exhibits three proton signals assigned to the
dippe (δ 1.1, m, CH3; δ 1.67, d, CH2; δ 2.1, m, CH). The 19F
NMR spectrum shows a very broad signal at δ −172 due to the
fluxionality of the C6F6 moiety, as reported for closely related
nickel complexes.26
34 Hz. The selectivity of the [Ni(0)] fragment at the ortho-
position in pentafluorobenzene is atypical, since nucleophilic
substitution in C6F5H occurs at the para-position with respect
to the H atom. In a few cases meta-substititution occurs.1
Single crystals suitable for an X-ray determination of 5
were obtained by slow evaporation of a THF/hexane solution.
Figure 4 displays the corresponding ORTEP drawing. This
structure allowed us to confirm the C−F activation of C6F5H at
the ortho-position with respect to the original hydrogen atom.
The values of the angles around the nickel atom are indicative
of a distorted square-planar geometry, probably due to the
chelating phosphine. The Ni(1)−C(1) bond distance in 5 is
0.099 Å longer that in the analogue, trans-(PEt3)2NiF(C6F5H),
Upon warming of a THF solution in a sealed NMR tube to
80 °C for 4 h, the characteristic multiplet detected in the
31P{1H} and 19F NMR spectra for complex 2 diminished in
intensity, and several new signals were observed in the reaction
mixture. Unfortunately, none of these signals can be unambig-
uously assigned to the expected oxidative addition product
2
[(dippe)Ni(C6F5)F] (3) on the basis of JP−P in a Ni(II) com-
plex.27 However, when pentafluorobenzene was used under the
4635
dx.doi.org/10.1021/ja412268y | J. Am. Chem. Soc. 2014, 136, 4634−4639

Journal of the American Chemical Society
Article
Table 1. HDF Products of C6F6 Using Phosphines as
Ancillary Ligands and Triethylsilane as the Hydrogen Source
1 (5%)
2C6F + 12Et3SiH + 6 phosphine ⎯⎯⎯⎯⎯→⎯ HDF product
6
(3)
a
entry
1
phosphine
HDF product (yield)
triphenylphosphine
C6F6 (28%), C6F5H (62%),
1,2,4,5-C6F4H2 (10%)
2
3
diisopropylphosphinoethane C6F5H (2%), 1,2,4,5-C6F4H2 (98%)
triethylphosphine 1,2,4,5-C6F4H2 (100%)
a
Determined by GC-MS. Reaction conditions: 5% 1, dioxane, T =
120 °C, t = 3 d.
Table 2. HDF Products of C6F6 Using Phosphines as
Ancillary Ligands
1 (5%)
2C6F + 6 phosphine ⎯⎯⎯⎯⎯→⎯ HDF products
(4)
6
a
entry
phosphine
HDF products (yield)
1
2
dippe
PEt3
C6F5H (2%), 1,2,4,5-C6F4H2 (98%)
1,2,4,5-C6F4H2 (100%)
Figure 4. ORTEP drawing (50% probability) for complex 5. Selected
bond lengths (Å): Ni(1)−C(1) 1.957(8), Ni(1)−F(5) 1.850(4),
Ni(1)−P(1) 2.139(2), Ni(1)−P(2) 2.191(2). Selected bond angles
(deg): P(2)−Ni(1)−P(1) 88.14(7), P(1)−Ni(1)−C(1) 93.8(2),
C(1)−Ni(1)−F(5) 88.7(3), F(5)−Ni(1)−P(2) 89.36(15), P(1)−
Ni(1)−F(5) 177.49(17).
a
Determined by GC-MS. Reaction conditions: 5% 1, dioxane, T =
120 °C, t = 3 d.
Table 3. Catalytic HDF of Fluoroarenes by [Ni(dippe)H]2 +
a
reported by Johnson and co-workers,26 whereas the Ni(1)−F(5)
bond distance is 0.066 Å shorter.
PEt3
With the aim to apply the observed reactivity of the above-
discussed nickel complexes into the catalytic HDF reactions, a
variety of reaction conditions were assessed. Among them, the
use of Et3SiH was considered due to the characteristic of being
a potential hydride source and taking advantage of the possible
formation of a strong Si−F bond.6d When using a 1:1:6 ratio of
Ni:C6F6:Et3SiH at 120 °C for 3 d, the conversion toward HDF
products was about 90%. Under these conditions, the selectivity
was low, since a mixture of fluoroaromatic products was de-
tected by GC-MS: C6F6 (6%), C6F5H (10%), C6F4H2 (50%),
and C6F3H3 (34%). Using a load of 5 mol% of 1, the
fluoroaromatics were C6F6 (70%), C6F5H (22%), and C6F4H2
(8%). Along with the low selectivity, the presence of black
metallic nickel residues was also observed. Therefore, to
suppress nickel metal formation due to decomposition and to
improve yields and selectivity, we decided to use an extra
amount of phosphine, considered initially as an ancillary ligand
to avoid metal deposition. Relevant results are summarized in
eq 3 and Table 1.
It is noteworthy that the alkylphosphines improved yields
and selectivity. In the above results, the presence of minute
amounts of Et3SiF, detected by GC-MS, was constantly lower
than expected. Therefore, we hypothesized that just the nickel−
phosphine mixture could be the actual hydrodefluorinating
system. Consequently, we performed experiments to test such a
possibility. Key results are summarized in eq 4 and Table 2.
Again, good yields and selectivity were observed, but the
source of hydrogen was still unclear. Therefore, the very same
reaction from Table 2, entry 2, was done using deuterated
dioxane as solvent to find out if the solvent was the hydrogen
source of this reaction. The result was exactly the same as in
entry 2; i.e., the HDF product obtained was not deuterated.
Consequently, the hydrogen in the product is likely to arise
only from the phosphine.
a
Reaction conditions: 5 mol% cat. 1, dioxane, 6 equiv PEt3, T =
120 °C, t = 72 h.
was then assessed. In particular, PEt3 was very effective for this
purpose, as can be seen in the results included in Table 3.
The results in Table 3 indicate a high yield and selectivity
toward HDF in the presence of a very basic phosphine such as
PEt3. In sharp contrast, the use of triphenylphosphine gave
poor HDF results, <1%, and dippe gave again low yields, <50%.
A mercury drop test using PEt3 as in Table 2, entry 2, showed a
high catalytic activity.
Other fluoroaromatic substrates with fewer fluorines in the
aromatic ring, such as 1,2-, 1,3-, and 1,4-difluorobenzene, did
not react in the presence or absence of nickel catalysts. These
results are also in agreement with previous reports.11b
A careful analysis of the intermediates involved in the reac-
tions in Table 3 was carried out. The 31P{1H} NMR spectrum
indicated the presence of fluorophosphorous derivatives formed
under the reaction conditions. We propose the possible for-
mation of fluorophosphines of the type R2PF (see Scheme 1).
The use of phosphine as a sacrificial agent for the HDF of a
variety of fluoroarenes in the presence of catalytic amounts of 1
4636
dx.doi.org/10.1021/ja412268y | J. Am. Chem. Soc. 2014, 136, 4634−4639

Journal of the American Chemical Society
Article
Scheme 1. Phosphine-Based Mechanistic Proposal for
Table 4. HDF Products Using PEt3 Only
Hydrodefluorination
fluoroarene + 3PEt3 → HDF product + 8 + 9
(6)
a
entry
fluoroarene
HDF product (yield)
1
2
3
4
C6F6
1,2,4,5-C6F4H2 (100%)
1,2,4,5-C6F4H2 (100%)
2,3,5,6-C5HF4N (100%)
1,2,4-C6F3H3 (100%)
C6F5H
C5F5N
1,2,3,4-C6F4H2
a
Determined by GC-MS and multinuclear NMR. Reaction conditions:
dioxane, T = 120 °C, t = 3 d.
conversion to yield HDF of all substrates, as indicated in eq 6
and Table 4.
These results rule out the required participation of any
nickel-based organometallic intermediates leading to HDF.
Although there are several reports30 in which the importance
of the metal−phosphine complexes (metallophosphoranes) in
C−F bond activation is highlighted, including a recent review in
this field,31 to the best of our knowledge there are no reactions
reported where the sole action of an alkylphosphine is able to
selectively defluorinate a polyfluoroarene. Closely related
reactivity of organofluorophosphonium salts along with the
use of a silane has been very recently reported by Stephan
and co-workers for HDF of fluoroalkanes.32a Also related are
Ozerov’s findings of fluoroalkyl HDF with a silylium−carborane
catalyst.32b Of note, compound 8 is the major organo-
phosphorous product containing fluorine (>93%).
It is well known that such fluorophosphines are unstable toward
their expected disproportionation reaction,28 as shown in eq 5.
A follow-up to the reaction in Table 4, entry 1, using a sealed
NMR tube monitoring by multinuclear NMR spectroscopy
revealed that the reaction was actually complete in 3 h, with the
immediate formation of 8 instead of Et2PF as the HDF pro-
gressed. Again, only small amounts of 9 were observed during
the course of the reaction. The relative amount of 9 barely
changed after 3 h. Also for entry 1, the GC-MS trace allowed
detection of a weak and broad peak consistent with the mass of
a “Et2-P-F2” fragment (see SI). To shed additional light on this
process, the reaction under the same conditions but using PMe3
as the phosphine showed no progress after 3 h.
In the current case, compounds 6 and 7 were not directly
detected, but some products of further reactivity or exchange
were found (vide infra). The 31P{1H} NMR spectrum identified
the formation of the difluorophosphorane 8 as the main com-
ponent in the mixture of final products (Figure 5), displaying a
Considering the results in Table 4, and the evidence found in
this study (vide supra), a tentative mechanistic proposal is
depicted in Scheme 1.
The initial step in Scheme 1 is a nucleophilic attack by the
phosphine (a), followed by F-migration (b). From there, a β-
hydride elimination is proposed (c), considering the fact that
the reaction does not depend on silane or on the participation
of the solvent as a hydrogen source. Next, a hydride nucleo-
philic addition (d), followed by a final elimination of fluo-
rophosphine Et2PF, yields the hydrodefluorinated product (e).
Such a fluorophosphine would be expected to decompose to
products 8 and 9 via the well-known disproportion depicted
above in eq 5, or alternatively following a closely related reac-
tion between the fluorophosphine and PEt3 depicted in eq 7.
Figure 5. Phospho-fluorinated product, 8.
1
signature signal at δ −11.6 (t, JP−F = 587 Hz). The 19F NMR
spectrum showed a doublet of septuplets at δ −40.0 (dsept,
3
JP−F = 588 Hz, JF−H = 10.5 Hz) due to the coupling of both
axial fluorine atoms with the CH2 of the Et moiety, each
displaying a septuplet signal (see SI). These NMR data are in
agreement with those reported for 8.29 Small amounts of a
compound labeled as 9 can be seen, with the characteristic
signals at δ −40.5 (d, JP−F = 620 Hz) in the 19F NMR and at
1
δ −16.3 (t, 1JP−F = 635 Hz) in the 31P{1H} NMR. These signals
allow us to suggest the presence of a “Et2-P-F2” derivative that
could not be fully characterized due to the low abundance.
Considering the presence of phospho-fluorinated products as
the final fate of fluorine, control experiments in the absence of
the nickel catalyst and silane revealed, to our surprise, very high
This proposal is inspired by the known reactivity of chloro-
phosphines and alkylphosphines to yield 8 and a phosphorus
polymer material.33 Ethylene was not detected, probably due to
evaporation from the reaction mixture at the high temperatures
4637
dx.doi.org/10.1021/ja412268y | J. Am. Chem. Soc. 2014, 136, 4634−4639

Journal of the American Chemical Society
Article
solvent, the amount of solution was reduced to 1 mL and then trans-
ferred into an NMR tube with a J-Young valve to permit multinuclear
following by NMR spectroscopy. After up to 3 d, the reaction mixture
was analyzed by both GC-MS and multinuclear NMR spectroscopy,
cooling the reaction mixture to room temperature and then taking a
sample from the original reaction mixture by separating the volatiles
under a vacuum. The addition of hexane at −30 °C allowed crys-
tallization of nickel-containing complexes 2 and 5. Compound 2 can
be crystallized shortly after reaction at room temperature and complex
5 after heating at 80 °C for 4 h. Complexes were further dried for 4 h
under high vacuum.
employed, or possibly polymerization. The lack of reactivity of
PMe3 supports the β-elimination proposal.
As presented earlier, many metal-mediated and catalytic
HDF reactions have been studied, some of them postulated to
be phosphine-assisted C−F bond activation at iridium34 and
platinum35 along with a P−C bond cleavage and P−F bond for-
mation, with retention of the oxidation state at the P atom and
also with the formation of metallophosphoranes.31 Here we
found that HDF occurs by the sole action of the phosphine
acting as fluorine acceptor and reducing agent, with oxidation
of the P atom and with regioselectivity.
Complex 2. Yield 90%. Anal. Calcd for 2, C34H64F6Ni2P4: C, 49.31;
1
H, 7.79; F, 13.76. Found: C, 49.1; H, 7.69; F, 13.5. H NMR (22 °C,
300 MHz, THF-d8): δ 1.1 (m, CH3); 1.67 (d, CH2); 2.1 (m, CH). 19
F
CONCLUSIONS
NMR (22 °C, 282 MHz, THF-d8): δ −172 (m, br). 31P{1H} NMR
(22 °C, 121 MHz, THF-d8): δ 70.7 (3JP−F = 12 Hz). Crystals of 2
suitable for X-ray diffraction studies were obtained by cooling the
THF-d8 solution to −30 °C, layering with hexane, and isolating the
crystals by decantation.
■
We demonstrated the formation of η2-CC complexes
between fluoroaromatics and a Ni(0) fragment, along with
the oxidative addition of the metal into the C−F bond of the
fluoroarene. Also reported are the catalytic HDF reactions of
polyfluoroarenes using a nickel catalyst and an added alkyl
phosphine, to yield the hydrodefluorination product with high
selectivity toward partially hydrodefluorinated products and
excellent yields. We show for the first time that the same
reactivity toward HDF was obtained without the use of a metal
catalyst or a silane. Thus, the current results point out that
many of the HDF reactions may not only be phosphine-
mediated but indeed possibly metal- and silane-free.
1
Complex 4. H NMR (22 °C, 300 MHz, THF-d8): δ 1.35 (m,
CH3); 1.92 (d, −CH2−); 2.3 (m, CH). 19F NMR (22 °C, 282 MHz,
THF-d8): δ −132.1 (m); −147.7 (m); −153.2 (m); −160.9 (m);
−184.5 (m). 31P{1H} (22 °C, 121 MHz, THF-d8): δ 71.8 (m, br).
Complex 5. Yield 88%. Anal. Calcd for 5, C20H33F5NiP2: C, 49.11;
1
H, 6.8; F, 19.42. Found: C, 48.9; H, 6.77; F, 19.2. H NMR (22 °C,
300 MHz, THF-d8): δ 1.33 (m, CH3); 1.90 (d, −CH2−); 2.31 (m,
CH). 19F NMR (22 °C, 282 MHz, THF-d8): δ −115.6 (m); −141.2
2
(m); −162.6 (m); −164.4 (m); −345.4 (m, JP2−F = 34 Hz). 31P{1H}
NMR (22 °C, 121 MHz, THF-d8): δ 70.4 (dd, JP−P = 10 Hz, 2JP−F
=
2
2
30 Hz); 65.6 (dd, JP−P = 10 Hz, JP−F = 34 Hz).
EXPERIMENTAL SECTION
■
Signature Signals for Et3PF2 (8). 19F NMR (22 °C, 282 MHz,
dioxane-d8): δ −40.0 (dsept, JP−F = 588 Hz, 3JF−H = 10.5 Hz). 31P{1H}
NMR (22 °C, 121 MHz, dioxane-d8): δ −11.6 (t, 1JP−F = 587 Hz). All
data are in agreement with the literature values.28
General Considerations. Unless otherwise noted, all manipu-
lations were performed under an argon atmosphere in an MBraun
glovebox (<1 ppm H2O and O2) or using standard Schlenk tech-
niques. Hexafluorobenzene, pentafluorobenzene, 1,2,3,4-tetrafluoro-
benzene, pentafluoropyridine, and 1,2-, 1,3-, and 1,4-difluorobenzene
were purchased from Aldrich; all substances were reagent grade, stored
over 3 Å molecular sieves, and manipulated in a drybox. All other
solvents used were dried and distilled from sodium/benzophenone
ketyl. Deuterated solvents for NMR experiments were purchased from
Aldrich and stored over 3 Å molecular sieves in the glovebox. The
nickel(I) complex (1)36 was prepared from Super-Hydride and
[(dippe)NiCl2]37 suspended in hexane, similarly to the literature
1
Signature Signals for Et2F2PH (9). H NMR (22 °C, 300 MHz,
dioxane-d8): δ 3.9 (m, P−H). 19F NMR (22 °C, 282 MHz, dioxane-
1
d8): δ −40.5 (d, JP−F = 620 Hz). 31P{1H} NMR (22 °C, 121 MHz,
1
dioxane-d8): δ −16.3 (t, JP−F = 635 Hz).
2. HDF Catalytic Reactions. A procedure similar to that previously
described was followed, but instead using 5 mol% of 1 (0.018 g,
0.002 mmol) in 1,4-dioxane with fluoroaromatic substrate (0.0210 g,
0.112 mmol in the case of C6F6; 0.0781 g, 0.672 mmol for Et3SiH; and
0.0399 g, 0.336 mmol for PEt3). The mixture was heated to 120 °C for
72 h. After this time, the reaction mixture was vacuum transferred for
the separation of volatiles. The volatile fraction and the residue were
analyzed by GC-MS and NMR.
procedure.38 All compounds were verified by the corresponding H,
1
19F, and 31P{1H} NMR spectra. All catalytic experiments were loaded
in a Schlenk flask with J-Young valves inside the glovebox. All of the
HDF products were distilled from the flask by static vacuum on a
vacuum line (P < 10−3 mmHg) prior to their quantification and char-
acterization. All NMR spectra of complexes and products were
recorded at room temperature on a 300 MHz Varian Unity spec-
trometer. NMR determinations for air-sensitive samples were collected
3. HDF Reactions without Catalyst. A procedure similar to that
previously described was followed, but using only the fluoroarene of
choice and PEt3 in a molar ratio of 1:3, respectively (for instance,
0.0210 g, 0.112 mmol of C6F6; 0.0399 g, 0.336 mmol of PEt3).
X-ray Structure Determination. Crystals of 2 or 5 mounted on a
glass fiber were studied with an Oxford Diffraction Gemini “A”
diffractometer with a CCD area detector, using a sealed tube X-ray
radiation source with λMo Kα = 0.71073 Å for 2 and λCu Kα = 1.54184 Å
for 5, with a monochromator of graphite at 130 K. Unit cell constants
were determined with a set of 15/3 narrow frame/runs (1° in ω)
scans. Data sets consisted of 207 and 1091 frames of intensity data
collected with a frame width of 1° in ω for 2 and 5, respectively, and a
crystal-to-detector distance of 55.00 mm. The double pass method of
scanning was used to exclude any noise. The collected frames were
integrated by using an orientation matrix determined from the narrow
frame scans. CrysAlisPro and CrysAlis RED software packages39a were
used for data collection and data integration. Analysis of the integrated
data did not reveal any decay. Final cell constants were determined by
a global refinement of 7225 (θ < 25.68°) and 1501 (θ < 68.10°) reflec-
tions for 2 and 5, respectively. Collected data were corrected for
absorbance by using analytical numeric absorption correction39b using
a multifaceted crystal model based on expressions upon the Laue
symmetry using equivalent reflections.
1
using sealed J-Young NMR tubes. The H, 19F, and 31P{1H} NMR
spectra of the hydrodefluorinated compounds were obtained from
1
reaction solutions in CDCl3 unless otherwise stated. H NMR chem-
ical shifts (δ, ppm) are reported relative to the residual proton reso-
nance in the deuterated solvent. 19F and 31P{H} spectra are reported
relative to external 10% trifluoroacetic acid and 85% H3PO4, respec-
tively. GC-MS determinations were performed using an Agilent 5975C
instrument equipped with a 30 m DB-5MS capillary (0.32 mm i.d.)
column.
The following general procedures were used for synthetic and
catalytic work.
1. Stoichiometric Reaction of 1 with Fluoroaromatics. To a
solution of [(dippe)NiH]2 (1) (0.0365 g, 0.056 mmol) in THF or
dioxane (5 mL) was added the corresponding fluoroaromatic substrate
(0.0210 g, 0.112 mmol in the case of C6F6; 0.0133 g, 0.112 mmol in
the case of PEt3). An immediate color change from deep purple to
yellow-brownish and release of gas were observed. The flask was
vented into the drybox and then heated to the desired temperature,
typically in an oil bath at 120 °C. Using THF-d8 or dioxane-d8 as
4638
dx.doi.org/10.1021/ja412268y | J. Am. Chem. Soc. 2014, 136, 4634−4639

Journal of the American Chemical Society
Article
Structure solution and refinement were carried out with the
programs SHELXS9739c and SHELXL97;39c for molecular graphics,
ORTEP-3 for Windows39d was used; and the software used to prepare
material for publication was WinGX.39e
(20) Barun, T.; Noveski, D.; Ahijado, M.; Wehmeier, F. Dalton Trans.
2007, 34, 3820−3825.
(21) Zhan, J.-H.; Lv, H.; Yu, Y.; Zhang, J.-L. Adv. Synth. Catal. 2012,
354, 1529−1541.
Full-matrix least-squares refinement was carried out by minimizing
(22) Lindup, R. J.; Marder, T. B.; Perutz, R. N.; Whitwood, A. C.
Chem. Commun. 2007, 3664−3666.
2
(Fo − Fc2)2. All non-hydrogen atoms were refined anisotropically. H
atoms attached to C atoms were placed in geometrically idealized
positions and refined as riding on their parent atoms, with CH =
0.95−1.00 Å and Uiso(H) = 1.2Ueq(C) or 1.5Ueq(C) for aromatic,
methylene, methyne, and methyl groups.
(23) Breyer, D.; Braun, T.; Penner, A. Dalton Trans. 2010, 39, 7513−
7520.
(24) Lv, H.; Cai, Y.-B.; Zhang, J.-L. Angew. Chem., Int. Ed. 2013, 52,
1−6.
(25) Almenningen, A.; Bastiansen, O.; Seip, R.; Seip, H. M. Acta
Chem. Scand. 1964, 18, 2115−2124.
ASSOCIATED CONTENT
* Supporting Information
■
(26) Johnson, S. A.; Taylor, E. T.; Cruise, S. J. Organometallics 2009,
28, 3842−3855.
S
Figures with selected multinuclear NMR spectra, GC-MS, and
crystallographic data for complex 2 and 5. This material is
available free of charge via the Internet at http://pubs.acs.org.
(27) (a) Garcia, J. J.; Brunkan, N. M.; Jones, W. D. J. Am. Chem. Soc.
2002, 124, 9547−9555. (b) Garcia, J. J.; Arevalo, A.; Brunkan, N. M.;
Jones, W. D. Organometallics 2004, 23, 3997−4002. (c) Arevalo, A.;
Ovando-Segovia, S.; Flores-Alamo, M.; Garcia, J. J. Organometallics
2013, 32, 2939−2943.
AUTHOR INFORMATION
Corresponding Author
juvent@unam.mx
■
(28) Fey, N.; Garland, M.; Hopewell, J. P.; McMullin, C. L.;
Mastroianni, S.; Orpen, A. G.; Pringle, P. G. Angew. Chem., Int. Ed.
2012, 51, 118−122.
Notes
(29) (a) Bartsch, R.; Stelzer, O.; Schmutzler, R. J. Fluorine Chem.
1982, 20, 85−88. (b) Doxsee, K. M.; Hanawalt, E. M.; Weakley, T. J.
R. Inorg. Chem. 1992, 31, 4420−4421.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank PAPIIT-DGAPA-UNAM (IN-210613) and CON-
ACYT (0178265) for their financial support of this work.
(30) Erhardt, S.; Macgregor, S. A. J. Am. Chem. Soc. 2008, 130,
15490−15498.
■
(31) Goodman, J.; Macgregor, S. A. Coord. Chem. Rev. 2010, 254,
1295−1306.
(32) (a) Caputo, C. B.; Hounjet, L. J.; Dobrovetsky, R.; Stephan, D.
W. Science 2013, 341, 1374−1377. (b) Douvris, C.; Ozeroz, O. V.
Science 2008, 321, 1188−1190.
REFERENCES
■
(1) Chambers, R. D. Fluorine in Organic Chemistry; Blackwell
Publishing Ltd.: Oxford, UK, 2004.
(33) Spangenberg, S. F.; Sisler, H. H. Inorg. Chem. 1969, 8, 1006−
(2) Ritter, S. K. Chem. Eng. News 2012, 90 (Feb 27), 10.
(3) Amii, H.; Uneyama, K. Chem. Rev. 2009, 109, 2119−2183.
(4) Hennecke, U. Angew. Chem., Int. Ed. 2012, 51, 4532−4534.
(5) Gladysz, J. A. Science 1994, 266, 55−56.
1010.
(34) (a) Blum, O.; Frollow, F.; Milstein, D. J. Chem. Soc., Chem.
Commun. 1991, 258−259. (b) Erhardt, S.; Macgregor, S. A. J. Am.
Chem. Soc. 2008, 130, 15490−15498.
(6) (a) Grushin, V. V. Acc. Chem. Res. 2010, 43, 160−171. (b) Clot,
E.; Eisenstein, O.; Jasim, N.; McGregor, S. A.; McGrady, J. E.; Perutz,
R. N. Acc. Chem. Res. 2011, 44, 333−348. (c) Braun, T.; Wehmeier, F.
(35) (a) Jasim, N. A.; Perutz, R. N.; Whitwood, A. C.; Braun, T.;
Izundu, J.; Neumann, B.; Rothfeld, S.; Stammler, H.-G. Organometallics
2004, 23, 6140−6149. (b) Nova, A.; Erhardt, S.; Jasim, N. A.; Perutz,
R. N.; Macgregor, S. A.; McGrady, J. E.; Whitwood, A. C. J. Am. Chem.
Soc. 2008, 130, 15499−15511.
Eur. J. Inorg. Chem. 2011, 613−625. (d) Scott, V. J.; Çelenligil-Çetin,
R.; Ozerov, O. V. J. Am. Chem. Soc. 2005, 127, 2852−2853.
(7) Belt, S. T.; Duckett, S. B.; Helliwell, M.; Perutz, R. N. J. Chem.
Soc., Chem. Commun. 1989, 928−930.
(36) A quintet at δ −9.81 (2JH‑P = 24 Hz) in the 1H NMR spectrum
is characteristic of nickel−bisphosphine hydride dimers.
(8) Jones, W. D.; Partridge, M. G.; Perutz, R. N. J. Chem. Soc., Chem.
Commun. 1991, 264−266.
(37) Scott, F.; Kruger, C.; Betz, P. J. Organomet. Chem. 1990, 387,
̈
113−121.
(9) Fahey, D. R.; Mahan, J. E. J. Am. Chem. Soc. 1977, 99, 2501−
(38) Jonas, K.; Wilke, G. Angew. Chem., Int. Ed. 1970, 9, 312−313.
(39) (a) Oxford Diffraction CrysAlis CCD and CrysAlis RED; Oxford
Diffraction Ltd.: Abingdon, UK, 2010. (b) Clark, C.; Ried, J. S. Acta
Crystallogr., Sect. A: Found. Crystallogr. 1995, 51, 887. (c) Sheldrick, G.
M.SHELXS97 and SHELXL97; University of Gottingen, Germany,
2008. (d) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565. (e) Farrugia,
L. J. J. Appl. Crystallogr. 1999, 32, 837.
2508.
(10) Bach, I.; Porschke, K. R.; Goddard, R.; Kopiske, C.; Kruger, C.;
̈
̈
Rufinska, A.; Seevogel, K. Organometallics 1996, 15, 4959−4966.
(11) (a) Cronin, L.; Higgit, C. L.; Karch, R. N.; Perutz, R. N.
Organometallics 1997, 16, 4920−4928. (b) Johnson, S. A.; Huff, C. F.;
Mustafa, F.; Saliba, M. J. Am. Chem. Soc. 2008, 130, 17278−17280.
(12) Jones, W. D. Dalton Trans. 2003, 3991−3995.
(13) (a) Aizenberg, M.; Milstein, D. Science 1994, 265, 359−361.
(b) Aizenberg, M.; Milstein, D. J. Am. Chem. Soc. 1995, 117, 8674−
8675.
(14) Vela, J.; Smith, J. M.; Yu, Y.; Ketterer, N. A.; Flaschenriem, C. J.;
Lachicotte, R. J.; Holland, P. L. J. Am. Chem. Soc. 2005, 127, 7857−
7870.
(15) Reade, S. P.; Mahon, M. F.; Whittlesey, M. K. J. Am. Chem. Soc.
2009, 131, 1847−1861.
(16) Hongbin, L.; Zhan, J.-H.; Cai, Y.-B.; Yi, Y.; Bingwu, W.; Zhang,
J.-L. J. Am. Chem. Soc. 2012, 134, 16216−16227.
(17) (a) Wu, J.-J.; Cao, S. Chem. Cat. Chem. 2011, 3, 1582−1586.
(b) Zhao, W.; Wu, J.; Cao, S. Adv. Synth. Catal. 2012, 354, 574−578.
(18) Fischer, P.; Gotz, K.; Radius, U. Organometallics 2012, 31,
̈
1374−1383.
(19) Archibald, S. J.; Braun, T.; Gaunt, J. F.; Hobson, J. E.; Perutz, R.
N. J. Chem. Soc., Dalton. Trans. 2000, 2013−2018.
4639
dx.doi.org/10.1021/ja412268y | J. Am. Chem. Soc. 2014, 136, 4634−4639