Evidence for single-electron pathways in the reaction between palladium(II) dialkyl complexes and alkyl bromides under thermal and photoinduced conditions

Article abstract of DOI:10.1021/acs.organomet.6b00893

Palladium(II) dialkyl complexes have previously been studied for their formation of alkanes through reductive elimination. More recently, these complexes, especially L₂Pd(CH₂TMS)₂ derived from Pd(COD)(CH₂TMS)₂, have found general use as palladium(0) precursors for stoichiometric formation of oxidative addition complexes through a two-electron reductive elimination/oxidative addition sequence. Herein, we report evidence for an alternative pathway, proceeding through single-electron elementary steps, when DPEPhosPd(CH₂TMS)₂ is treated with an α-bromo-α,α-difluoroacetamide. This new pathway does not take place through a palladium(0) intermediate, neither does it afford the expected oxidative addition complexes. Instead, stoichiometric amounts of carbon-centered alkyl radicals are formed, which can be trapped in high yields either by TEMPO or by an arene, leading to α-aryl-α,α-difluoroacetamides. The same overall transformation takes place under both thermal conditions (70 °C) and irradiation with a household light bulb (at 30 °C). It is also demonstrated that DPEPhosPdMe₂, made in situ from Pd(TMEDA)Me₂, displays a similar initial reactivity. Finally, electronically and structurally different alkyl bromides were evaluated as reaction partners.

Full text of DOI:10.1021/acs.organomet.6b00893

Article
pubs.acs.org/Organometallics
Evidence for Single-Electron Pathways in the Reaction between
Palladium(II) Dialkyl Complexes and Alkyl Bromides under Thermal
and Photoinduced Conditions
Thomas L. Andersen,† Søren Kramer,*,† Jacob Overgaard,‡ and Troels Skrydstrup*,†
†Carbon Dioxide Activation Center (CADIAC), Department of Chemistry and the Interdisciplinary Nanoscience Center (iNANO),
Aarhus University, Gustav Wieds Vej 14, 8000 Aarhus C, Denmark
‡Center for Materials Crystallography (CMC), Department of Chemistry and the Interdisciplinary Nanoscience Center (iNANO),
Aarhus University, Langelandsgade 140, 8000 Aarhus C, Denmark
S
* Supporting Information
ABSTRACT: Palladium(II) dialkyl complexes have previously
been studied for their formation of alkanes through reductive
elimination. More recently, these complexes, especially L2Pd-
(CH2TMS)2 derived from Pd(COD)(CH2TMS)2, have found
general use as palladium(0) precursors for stoichiometric
formation of oxidative addition complexes through a two-
electron reductive elimination/oxidative addition sequence.
Herein, we report evidence for an alternative pathway,
proceeding through single-electron elementary steps, when DPEPhosPd(CH2TMS)2 is treated with an α-bromo-α,α-
difluoroacetamide. This new pathway does not take place through a palladium(0) intermediate, neither does it afford the
expected oxidative addition complexes. Instead, stoichiometric amounts of carbon-centered alkyl radicals are formed, which can
be trapped in high yields either by TEMPO or by an arene, leading to α-aryl-α,α-difluoroacetamides. The same overall
transformation takes place under both thermal conditions (70 °C) and irradiation with a household light bulb (at 30 °C). It is
also demonstrated that DPEPhosPdMe2, made in situ from Pd(TMEDA)Me2, displays a similar initial reactivity. Finally,
electronically and structurally different alkyl bromides were evaluated as reaction partners.
INTRODUCTION
■
Palladium catalysis is one of the most widely used synthetic
tools in both the chemical industry and academia for the
construction of carbon−carbon and carbon−heteroatom
bonds.1 A typical catalytic cycle is initiated by oxidative
addition of a palladium(0) species into a carbon−halogen bond.
Unfortunately, many palladium(0) catalysts are not air-stable,
and instead it is common practice to employ palladium(II)
precatalysts. These precatalysts are usually air-stable but can be
readily transformed into a catalytically active palladium(0)
species under the reaction conditions.2 The activation from
palladium(II) to palladium(0) as well as the remainder of the
catalytic cycle is usually represented as proceeding via two-
electron processes (Figure 1).
The palladium(II) complex Pd(COD)(CH2TMS)2 (1) is
Figure 1. General catalytic cycle for palladium-catalyzed cross-
couplings. Ligands have been omitted for the sake of simplicity. M
typically used as a precursor for in situ formation of L2Pd(0)-
type complexes.3,4 Combining this complex with mono- or
bidentate L-type (dative) ligands has been exploited both for
catalyst generation and for the formation of oxidative addition
complexes at room temperature.5 The activation pathway is
believed to proceed via rapid substitution of COD for the
added ligand followed by reductive elimination of (CH2TMS)2
to generate L2Pd(0) (eq 1). Frequent isolation of oxidative
addition complexes (e.g., eq 2) in combination with a few
reports on the observation of (CH2TMS)2 indicates a
= metal; X = halide.
mechanism involving these two-electron elementary steps.5,6
A few examples of ligand exchange of a single CH2TMS group
on Pd(COD)(CH2TMS)2 or L2Pd(CH2TMS)2 without
reductive elimination have also been reported; however, no
Received: December 2, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.organomet.6b00893
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
phosphine ligands, such as DPEPhos, led to rapid and clean
displacement of the COD ligand as expected, forming complex
3 (eq 3).12 However, when mixing bromodifluoroacetamide 4,
Pd(COD)(CH2TMS)2, and DPEPhos in benzene, we did not
obtain the desired oxidative addition complex (Scheme 2a).
Instead, aryldifluoroacetamide 2a was obtained along with the
palladium(II) monobromide complex 5. Repeating the reaction
with two equivalents of the bromide led to the palladium(II)
dibromide complex 6 (Scheme 2b). The structures of the
comlexes 3, 5, and 6 were identified by X-ray analysis (Figure
2). The presence of further equivalents of the bromodifluor-
oacetamide remained untouched even after prolonged reaction
time (Scheme 2c).13 These results show that the thermally
induced reaction is a stepwise process whereby the first
CH2TMS ligand is replaced at a much higher rate than the
second and that the consumption of the bromodifluoroaceta-
mide 4 corresponds to the conversion of palladium complex.
Due to the formation of aryldifluoroacetamide 2a under the
thermal conditions, it seems likely that the reaction goes
through a palladium−carbon bond homolysis or a related
radical pathway.14,15 Hence, we were curious if the same
transformation could be performed under milder conditions
when irradiated by light. Indeed, when the reaction in Scheme
2b was repeated under irradiation of a household 26 W CFL,
the same transformation proceeded at 30 °C (Scheme 2d). In
contrast to the thermal reaction, only trace amounts of the
intermediate monobromide 5 were observed when following
the reaction by 31P NMR spectroscopy, demonstrating that the
second substitution is now faster than the first. A possible
explanation for this divergence can be made when comparing
the UV−vis absorption spectra of the starting palladium
complex, DPEPhosPd(CH2TMS)2, with the intermediate
complex 5 (Figure 3). While DPEPhosPd(CH2TMS)2 absorbs
only in the UV region, the intermediate has additional
absorption shifted into the visible region, which is where the
26 W CFL is emitting predominantly. This shift should favor
the excitation and hence further reaction of DPEPhosPdBr-
(CH2TMS) over DPEPhosPd(CH2TMS)2. In contrast, the
thermally induced reactions were performed in the dark, and
the change in absorption will play no role in this case.
comments on the mechanistic path for this exchange were
mentioned.7 In fact, none of the reports utilizing Pd(COD)-
(CH2TMS)2 contain evidence for or suggest the alternative
possibility of a single-electron pathway involving free radicals.
Herein, we report clear evidence for an alternative pathway
to the commonly accepted reductive elimination/oxidative
addition sequence for an L2Pd(II)(CH2TMS)2 complex. The
unexpected reactivity is established through the first study of
reactions between alkyl bromides and an L2Pd(II) (CH2TMS)2
complex. The alternative pathway proceeds through single-
electron elementary steps and involves the formation of free
radicals. The implication of similar modes of reactivity for a
palladium(II) dimethyl complex as well as different alkyl
bromides is also presented.
RESULTS AND DISCUSSION
■
Aryldifluoroamides 2 have recently received significant
attention due to their important pharmacological properties.8
Several reports on the synthesis of aryldifluoroamides have
appeared including examples starting from bromodifluoroace-
tamides (Scheme 1a).9 In certain cases using palladium(0) or
Scheme 1. (a) Synthesis of α-Aryl-α,α-difluoroacetamides;
(b) Our Synthesis of α-(Hetero)aryl α,α-Difluoro-β-
ketoamides Using a Carbonylative Protocol
In order to examine the influence of light irradiation on the
reaction progress, we performed light on/off cycles (Figure 4).
It was observed that the conversion rapidly ceases when the
irradiation is switched off. Although this does not exclude a
radical chain mechanism, it clearly demonstrates the pivotal role
of the light irradiation throughout the reaction.16
Next, we performed relative reactivity experiments to
examine the influence on the reaction rate of electronic effects
on the arene reaction partner (Scheme 3). Both an arene
bearing an electron-withdrawing substituent (CN) and an arene
bearing an electron-donating substituent (OMe) react faster
than benzene. These results are consistent with the addition of
a free radical to the arene; however, they are inconsistent with
pathways involving electrophilic aromatic substitution and
photoredox catalysts with these gem-difluorobromides, the
generation of radical intermediates, which can be trapped by
unsaturated carbon−carbon moieties, has been invoked.10
Recently, we and others reported the use of bromodifluor-
oacetamides and -acetates for the synthesis of α-(hetero)aryl
α,α-difluoro-β-ketoamides and α,α-difluoro-β-ketoesters using a
palladium-catalyzed carbonylative protocol (Scheme 1b).11
During our preliminary studies on the mechanism of this
reaction, we attempted to prepare oxidative addition complexes
starting from Pd(COD)(CH2TMS)2 as a palladium(0)
precursor. Mixing Pd(COD)(CH2TMS)2 with bidentate
B
DOI: 10.1021/acs.organomet.6b00893
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 2. Reactions of DPEPhosPd(CH2TMS)2 with Varying Equivalents of Bromodifluoroacetamide 4 under both Thermal
a
(Dark) and Light-Irradiation Conditions
a
Yields of 2a were determined by HPLC analysis with the aid of a calibrated internal standard, and the findings are in accordance with intensities in
19F NMR. Yield of 5 is based on H NMR analysis with the aid of an internal standard. Yields of 6 are isolated yields relative to the amount of
1
palladium complex added. In all cases, the remainder of the mass balance is hydro-debrominated starting material. The same reaction outcomes,
albeit with slightly inferior yields, were observed when starting from Pd(COD)(CH2TMS)2 and DPEPhos instead of the premade complex.
Figure 3. UV−vis absorption spectra for DPEPhosPd(CH2TMS)2 (3)
and DPEPhosPd(CH2TMS)Br (5), both recorded in benzene.
While arenes are not typically used as radical traps, 2,2,6,6-
tetramethylpiperidin-1-oxyl (TEMPO) is usually the reagent of
choice for this purpose. In order to trap potential radical
intermediates, some of the reactions were accordingly repeated
in the presence of TEMPO (Scheme 4). Both the thermal and
the photoinduced reaction led to the TEMPO adduct 7 being
formed as the major product. These observations further
support the intermediacy of a free alkyl radical under both sets
of conditions. Interestingly, under the thermal conditions, the
presence of TEMPO led to an acceleration of the second step,
as the intermediate monobromide 5 only started to appear once
the majority of TEMPO had been consumed (Scheme 4c).18
Neither heating at 70 °C in the dark nor irradiation of the
bromodifluoroacetamide 4 at 35 °C in the presence of TEMPO
afforded more than a trace (<1%) of the TEMPO adduct 7,
thus indicating that unassisted homolysis of the bromide 4 is
unlikely to be the first step under either set of conditions.20 The
presence of the palladium complex is pivotal for the reactivity of
the bromodifluoroacetamide. Further control experiments
Figure 2. X-ray crystal structures of DPEPhosPd(CH2TMS)2 (top),
DPEPhosPd(CH2TMS)Br (middle), and DPEPhosPdBr2 (bottom)
(ellipsoids are shown at 50% probability, and hydrogens are omitted
for clarity).
electrophilic palladation. The observed regioselectivities are
also in accordance with those previously reported for the
addition of electrophilic radicals to arenes.17 Importantly, the
ratios in the competition experiments are very similar for the
thermal and the light-induced reactions, supporting the
suggestion that they follow a similar mechanism.
C
DOI: 10.1021/acs.organomet.6b00893
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 4. Reactions Performed in the Presence of TEMPO
under (a) Light Irradiation Conditions and (b) Thermal
Conditions; (c) Reaction under Thermal Conditions in the
Presence of TEMPO Monitored Over Time19
Figure 4. Influence of light irradiation on reaction progress by light
on/off cycles. Yields were determined by GC-FID analysis with the aid
of a calibrated internal standard.
Scheme 3. Intermolecular Competition Experiments
between Benzene and an Electron-Poor Arene and an
a
Electron-Rich Arene, Respectively
to be the case for both the thermally and the light-induced
reactions; however, the relative rates depend on the applied
conditions. Under thermal conditions, the slowest step is found
during the substitution of the second CH2TMS, while the
slowest step for the light-irradiated reaction occurs during
substitution of the first CH2TMS. A mechanistic proposal for
the substitution of the first CH2TMS for a bromide under light
irradiation is shown in Scheme 5a. The reaction starts with
excitation of DPEPhosPd(CH2TMS)2 (3).22,23 The excited
complex 3* can now follow one of two pathways. In the first
pathway (step B), homolysis of the Pd−C bond occurs,
affording a CH2TMS radical and the Pd(I) complex 8. This
highly reactive intermediate can rapidly abstract a bromide from
the bromodifluoroacetamide 4, leading to Pd(II) complex 5
and a carbon-centered alkyl radical (step C). Addition of this
radical to benzene generates an aryl radical, which, upon
hydrogen abstraction, can rearomatize, producing 2a (vide
infra).24 Alternatively, the excited complex 3* can directly
abstract the bromide, leading to Pd(III) intermediate 9 and the
carbon-centered alkyl radical (step D).25−27 Loss of a CH2TMS
radical from 9 leads to the monobromide complex 5 (step E).
Since no photoexcitation can occur under the thermal
conditions, it is possible that heating provides sufficient energy
for the starting complex 3 to undergo either homolysis or
bromide abstraction (Scheme 5b). The subsequent steps are
likely highly related to the light-induced pathways.
a
The reported product ratios were determined by 19F NMR.
Regioselectivities observed for monosubstituted substrates (o:m:p):
b
c
d
e
(4.7:1.0:4.5). (4.5:1.0:4.2). (3.1:2.0:1.0). (2.9:2.3:1.0).
showed that the palladium complex is stable, in both the
presence and absence of the bromodifluoroacetamide 4, in
solution at 35 °C in the dark for at least 24 h. However, in the
absence of bromodifluoroacetamide 4, it decomposes within 6
h under irradiation of a 26 W CFL, and decomposition is also
observed within 2 h in the dark at 70 °C. Consistent with the
lack of reaction at 35 °C in the dark, the oxidation/reduction
potentials measured by cyclic voltammetry also indicated that it
is unlikely that bromodifluoroacetamide 4 acts as an outer-
sphere single-electron oxidant for the oxidation of Pd(II) to
Pd(III) with concurrent formation of the radical anion of 4
(irreversible oxidation of DPEPhosPd(CH2TMS)2 observed at
Ep/2 = +1.23 V vs Ag/AgI in DMF; irreversible reduction of
bromide 4 observed at Ep/2 = −1.25 V vs Ag/AgI in DMF).21
On the basis of the results presented so far, it seems plausible
that the elementary steps for the substitution of the first and the
second CH2TMS are the same. Furthermore, this also appears
When following the light-irradiated reactions by 31P NMR
spectroscopy, very little of the intermediate 5 is observed.
Instead, complexes 3 and 6 are major species, suggesting that
one of the steps A (excitation of 3), B (loss of the alkyl radical
from 3*), or D (bromide abstraction from RF−Br by 3*) is
overall rate-limiting (Figure 5e). Under thermal conditions with
D
DOI: 10.1021/acs.organomet.6b00893
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 5. Mechanistic Proposal for the Formation of
DPEPhosPd(CH2TMS)Br under Light Irradiation and
Thermal Conditions; Similar Steps are Presumed for the
Formation DPEPhosPdBr2 from 5
Figure 5. 31P NMR of the different complexes and reactions in
Scheme 2: (a) DPEPhosPd(CH2TMS)2. (b) DPEPhosPdBr-
(CH2TMS) observed in the crude reaction mixture with 1.0 equiv
of [Pd] under thermal conditions (Scheme 2a). (c) DPEPhosPdBr2.
(d) Reaction at partial conversion under thermal conditions (Scheme
2b). (e) Reaction at partial conversion under light irradiation (Scheme
2d).
(CH2TMS)2 under these conditions was further confirmed by
1H NMR spectroscopy of the reaction mixture, when the
reaction was repeated in C6D6. Instead of the dimer, only a
peak at 0 ppm was observed, indicating the formation of Me4Si.
Overall, these results are consistent with the mechanistic
proposals with the addition that the predominant fate of the
liberated CH2TMS radical is hydrogen abstraction (Scheme
5c).
In the presence of a base such as K2CO3, we observed that
the conversion of the palladium complex and bromodifluor-
oacetamide 4 did not correlate. Repeating the reaction with a
large excess of bromodifluoroacetamide relative to palladium,
we observed a 21% yield of 2a with only 2.5 mol %
Pd(COD)(CH2TMS)2 and 8 mol % DPEPhos (Scheme 6).
one equivalent of the bromodifluoroacetamide 4, only
complexes 3 and 5 are observed, while complexes 3, 5, and 6
are observed during the reaction with two equivalents of the
bromodifluoroacetamide (Figure 5d). Accordingly, the overall
rate-limiting step under thermal conditions takes place during
the substitution of the second CH2TMS group.
The selective substitution of one CH2TMS ligand on
L2Pd(CH2TMS)2 complexes has been demonstrated previously
with carboxylic acids and a protonated phosphine; however, no
comments on the mechanism were mentioned except for the
observation of the formation of Me4Si in a single example.7a
Next, we set out to investigate the fate of the CH2TMS ligands
under our reaction conditions. At the end of the reactions in
Scheme 2a−c, the dimerized ligand, (CH2TMS)2, was observed
by GC-FID in 41%, 31%, and 27% calibrated yields,
respectively.28 This product could originate from reductive
elimination; however, taking the nature and the yield of the
obtained palladium(II) complexes 5 and 6 into account, it is
difficult to imagine a feasible mechanism via palladium(0)
intermediates formed by the reductive elimination. It seems
more likely that the CH2TMS dimer arises from radical
dimerization of CH2TMS radicals, either free or through attack
of a free CH2TMS radical on a metal-bound CH2TMS.29
Importantly, under the photoinduced reaction conditions
(Scheme 2d) no (CH2TMS)2 was observed, but the palladium-
(II) complex 6 is still formed in a high yield. The absence of
Scheme 6. Reactions with Substoichiometric Amounts of
Palladium Complex Performed in the Presence of Base
a
a
Yields were determined by GC-FID analysis with the aid of a
calibrated internal standard.
However, at the end of this reaction, the only observed peak in
the 31P NMR spectrum corresponds to the oxidized phosphine,
making the intermediacy of a Pd(0) species likely. It appears
that in the presence of K2CO3 a new species is formed, which
can consume superstoichiometric amounts of 4.30,31 Other
bases, such as Cs2CO3 and LiOtBu provided similar results.
To examine the generality of the observations made with
DPEPhosPd(CH2TMS)2, we performed preliminary studies on
the related dialkyl palladium complex, Pd(TMEDA)Me2 (10).
E
DOI: 10.1021/acs.organomet.6b00893
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Such a comparison will also be valuable in terms of evaluating
the necessity of β-silicon atoms on the alkyl substituents as a
prerequisite for the enclosed reactivity.32,33 Thermolysis of 10
has been reported to lead to the formation of a mixture of
ethane and methane, the latter suggested deriving from a
unimolecular α-elimination.34a Also, Pd(TMEDA)Me2 has
been reported to undergo two-electron oxidative addition to
R−X-type electrophiles, mostly Me−X, providing Pd(IV)
intermediates, which collapse with concomitant and quantita-
tive release of ethane.34 Sanford et al. reported on the
mechanism for the formation of ethane from (t-Bu2bpy)PdMe2
in the presence of the one-electron oxidant [Cp2Fe]PF6.35 It
was discovered that following the initial formation of Pd(III), a
disproportion to Pd(II)/Pd(IV) takes place followed by
reductive elimination of ethane from the Pd(IV) species to
form Pd(II).36,37 However, later studies indicate that this high-
valent pathway is only operating when diamine ligands are
employed and that bis-phosphine ligands instead lead to
oxidation-induced Pd−C bond homolysis, producing methane
along with a substantial amount of ethane.38 Photodecompo-
sition of a bis-phosphine-ligated complex, (dppe)PdMe2, has
also been studied briefly.39 Interestingly, while photolysis of
(dppe)PdMeCl was found to proceed through radical path-
ways, a nonradical pathway was indicated for photodecompo-
sition of (dppe)PdMe2, although the presence of several
simultaneously operating mechanisms complicated the con-
clusions.
with equal integration and with a common P,P coupling
constant of 30.4 Hz. Isolation of this phosphine species as a
colorless solid was possible and 1H NMR analysis of the
obtained compound was consistent with the palladium(II)
monobromide complex (11).40 The result suggests that a
stepwise substitution on palladium via a radical pathway similar
to the one for DPEPhosPd(CH2TMS)2 could be operating
under thermal conditions. In order to examine if the
substitution of the second methyl group was feasible, the
reactions were repeated in the presence of excess bromodi-
fluoroacetamide 4 similar to the reactions in Scheme 2b and d.
Under these conditions, high yields of fluoroalkylated benzene
were obtained as expected. However, DPEPhosPdMeBr (11)
was still observed as the major phosphine species by 31P NMR
with only trace amounts of DPEPhosPdBr2.41 Possibly, the
presence of TMEDA, which can serve as a base similar to
K2CO3 (Scheme 6), can account for the discrepancy between
the yield of 2a and formation of DPEPhosPdMeBr. Overall,
these findings suggest that DPEPhosPdMe2 could follow a
similar single-electron pathway as described for DPEPhosPd-
(CH2TMS)2 under both thermal and photochemical con-
ditions, consuming one equivalent of bromodifluoroacetamide
4, affording DPEPhosPdMeBr (11). However, unlike the
related monobromide 5, this complex does not react further
under our reaction conditions.
The influence of the phosphine ligand on the pathway taken
by dialkyl palladium complexes 1 and 10 in their reaction with
bromodifluoroacetamide 4 was also assessed (Scheme 8). For
When we subjected Pd(TMEDA)Me2 to thermal conditions
corresponding to Scheme 2a in the presence of one equivalent
of DPEPhos, aryldifluoroacetamide 2a was again obtained in a
near-quantitative yield (Scheme 7). Examining the crude
reaction mixture by 31P{H} NMR spectroscopy revealed a
single set of doublet signals resonating at δ = 28.8 and 6.8 ppm
Scheme 8. Reactions of Pd(TMEDA)Me2 in the Absence of
Added Phosphine Ligand under Both Thermal (Dark) and
a
Light-Irradiation Conditions
Scheme 7. Reactions of Pd(TMEDA)Me2 in the Presence of
DPEPhos with Varying Equivalents of
Bromodifluoroacetamide 4 under Both Thermal (Dark) and
a
Light-Irradiation Conditions
a
Yields and conversions were determined by HPLC analysis with the
aid of a calibrated internal standard. Conversion refers to conversion of
bromodifluoroacetamide 4. Palladium-black was observed in all the
reactions.
both complexes, essentially none of the aryldifluoroacetamide
2a could be detected under thermal conditions in the absence
of DPEPhos. In contrast, under the photoinduced reaction
conditions, 2a is observed for both complexes albeit in low
yields. In all the reactions the conversion of bromodifluor-
oacetamide 4 was found to be significantly higher than the
formation of 2a due to undetermined side reactions. In the
reactions with Pd(COD)(CH2TMS)2 (1), formation of the
dimer, (CH2TMS)2, occurred in a 58% yield for the thermal
and a 14% yield for the photoinduced reaction.42 On the basis
of these results, the presence of the phosphine ligand seems
crucial for selectively accessing the new reactivity.
a
Unless otherwise specified, yields were determined by HPLC analysis
with the aid of a calibrated internal standard, and the findings are in
accordance with intensities in 19F NMR.
F
DOI: 10.1021/acs.organomet.6b00893
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Having examined the influence of the β-silicon groups, we
turned our attention to the alkyl bromide reaction partner.
Interestingly, for reactions with DPEPhosPd(CH2TMS)2 (3),
the bromodifluoroacetate (12a), perfluoroalkyl bromide (12c),
and benzyl bromide (12d) all led to clean formation of
palladium(II) dibromide complex 6 (Scheme 9). For
CONCLUSIONS
■
In summary, we report evidence for an alternative pathway to
the commonly proposed two-electron reductive elimination for
the often used L2Pd(CH2TMS)2 complexes, when ligated to
the bidentate ligand DPEPhos and in the presence of an α-
bromo-α,α-difluoroacetamide. The same overall transformation
takes place under both thermal and photoinduced conditions,
leading to either DPEPhosPd(CH2TMS)Br or DPEPhosPdBr2
instead of DPEPhosPd(0) or a Pd(II) oxidative addition
species derived thereof. In spite of the different conditions, the
mechanistic studies indicate that the processes are taking place
by similar pathways albeit with different overall rate-
determining steps. During the reaction, stoichiometric amounts
of a carbon-centered alkyl radical are formed, which can be
trapped in high yields either by the arene solvent, leading to
aryldifluoroacetamides, or by TEMPO. A similar reactivity was
also observed for the replacement of the first methyl group for
DPEPhosPdMe2, suggesting that this type of reactivity could be
a general reaction pathway for palladium dialkyl complexes.
Finally, this novel reactivity was also observed across a series of
alkyl bromides, which can form stabilized radicals, thus
indicating the discovered pathway may be of a more general
nature. This was further corroborated by preliminary
investigations with other alkyl halides and phosphine ligands.
Scheme 9. Examination of Formation of 5 and 6 from Other
Alkyl Bromides (12a−g) and of the Reactivity of an Alkyl
a
Chloride and an Alkyl Iodide
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.organo-
met.6b00893.
■
a
The reported complex formation is based on observation in the 31P
S
NMR spectrum of the crude reaction mixture.
bromoacetate 12b, the major peaks observed by 31P NMR
spectroscopy corresponded to a mixture of palladium(II)
monobromide 5 and palladium(II) dibromide 6. Importantly,
the fate of the alkyl moieties of 12a−d was consistent with the
formation of free alkyl radicals: For 12a, addition to benzene
producing ethyl phenyldifluoroacetate was observed exclusively
by 19F NMR spectroscopy. For 12b, addition to benzene
producing ethyl phenylacetate was observed in small amounts
by 1H NMR spectroscopy. Exclusive formation of Ph(CF2)7CF3
Experimental procedures and compound characterization
data (PDF)
Crystallographic data (CIF)
AUTHOR INFORMATION
Corresponding Authors
*E-mail (S. Kramer): kramer@chem.au.dk.
*E-mail (T. Skrydstrup): ts@chem.au.dk.
ORCID
Jacob Overgaard: 0000-0001-6492-7962
Troels Skrydstrup: 0000-0001-8090-5050
Author Contributions
All authors have given approval to the final version of the
manuscript.
■
was observed by H and 19F NMR spectroscopy for 12c.
Finally, for 12d, significant formation of 1,2-diphenylethane was
observed by H NMR spectroscopy. A series of unfunctional-
ized alkyl bromides (12e−g) did not lead to any formation of 5
or 6. These results indicate that a functional group stabilizing
the generated carbon-centered radical is necessary for accessing
the disclosed reaction pathway; however, once this requirement
is met, different alkyl bromide substrates successfully undergo
the new transformation.
1
1
Notes
The authors declare no competing financial interest.
A preliminary investigation of other halide-containing
substrates as well as other phosphine ligands was also
undertaken. An α-chloro-α,α-difluoroacetamide was recovered
quantitatively when subjected to the reaction conditions
represented in Scheme 9.21 In contrast, the iodide analogue
of 12c reacted smoothly, leading to clean formation of
DPEPhosPdI2 and Ph(CF2)7CF3, both consistent with the
disclosed reactivity. No conversion of (PPh3)2Pd(CH2TMS)2
was observed in the presence of the alkyl bromide 4 under the
conditions in Scheme 9. However, under the same conditions,
XantPhosPd(CH2TMS)2 led to 20% conversion to Xant-
PhosPdBr2 along with a corresponding 40% yield of 2a relative
to the palladium complex. Accordingly, the new reactivity is
also accessible with XantPhosPd(CH2TMS)2, albeit to a lesser
extent.
ACKNOWLEDGMENTS
■
We are deeply appreciative of generous financial support from
the Danish National Research Foundation (grant nos. DNRF93
and DNRF118) and Aarhus University for generous financial
support of this work. We thank Magnus Rønne (Aarhus
University) for assistance with cyclic voltammetry.
REFERENCES
■
(1) (a) Handbook of Organopalladium Chemistry for Organic Synthesis;
Negishi, E.-I., de Meijere, A., Eds.; Wiley: New York, 2002.
(b) Organotransition Metal Chemistry − From Bonding to Catalysis;
Hartwig, J. F., Ed.; University Science Books: Sausalito, CA, 2010.
(c) Johansson Seechurn, C. C. C.; Kitching, M. O.; Colacot, T. J.;
Snieckus, V. Angew. Chem., Int. Ed. 2012, 51, 5062−5085.
G
DOI: 10.1021/acs.organomet.6b00893
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(2) Amatore, C.; Carre,
1995, 14, 1818−1826.
(3) For the first report on Pd(COD)(CH2TMS)2, see: Pan, Y.;
Young, G. B. J. Organomet. Chem. 1999, 577, 257−264.
(4) For seminal reports on L2Pd(CH2TMS)2, see: (a) Wozniak, B.;
Ruddick, J. D.; Wilkinson, G. J. Chem. Soc. A 1971, 0, 3116−3120.
(b) Tooze, R.; Chiu, K. W.; Wilkinson, G. Polyhedron 1984, 3, 1025−
1028.
(5) (a) Fors, B. P.; Watson, D. A.; Biscoe, M. R.; Buchwald, S. L. J.
Am. Chem. Soc. 2008, 130, 13552−13554. (b) Watson, D. A.; Su, M.;
Teverovskiy, G.; Zhang, Y.; Garcia-Fortanet, J.; Kinzel, T.; Buchwald,
S. L. Science 2009, 325, 1661−1664. (c) Sergeev, A. G.; Schultz, T.;
Torborg, C.; Spannenberg, A.; Neumann, H.; Beller, M. Angew. Chem.,
Int. Ed. 2009, 48, 7595−7599. (d) McAtee, J. R.; Martin, S. E. S.;
Ahneman, D. T.; Johnson, K. A.; Watson, D. A. Angew. Chem., Int. Ed.
2012, 51, 3663−3667. (e) Larini, P.; Kefalidis, C. E.; Jazzar, R.;
Renaudat, A.; Clot, E.; Baudoin, O. Chem. - Eur. J. 2012, 18, 1932−
1944. (f) Ros, A.; Estepa, B.; Ramirez-Lopez, P.; Alvarez, E.;
Fernandez, R.; Lassaletta, J. M. J. Am. Chem. Soc. 2013, 135,
15730−15733. (g) Lee, H. G.; Milner, P. J.; Buchwald, S. L. Org. Lett.
2013, 15, 5602−5605. (h) Andersen, T. L.; Friis, S. D.; Audrain, H.;
Nordeman, P.; Antoni, G.; Skrydstrup, T. J. Am. Chem. Soc. 2015, 137,
1548−1555. (i) Milner, P. J.; Yang, Y.; Buchwald, S. L. Organometallics
2015, 34, 4775−4775.
́
E.; Jutand, A.; M’Barki, M. A. Organometallics
(b) Zhao, H.-Y.; Feng, Z.; Luo, Z.; Zhang, X. Angew. Chem., Int. Ed.
2016, 55, 10401−10405.
(12) The isolated complex, DPEPhosPd(CH2TMS)2, can be stored
as a solid for at least two weeks in air at room temperature without any
signs of decomposition.
(13) No change in the reaction mixture was observed after an
additional day of heating.
(14) For an early example of an oxidative addition palladium complex
formed via a radical pathway, see: Becker, Y.; Stille, J. K. J. Am. Chem.
Soc. 1978, 100, 838−844.
(15) For a review on palladium radical reactions, see: Liu, Q.; Dong,
X.; Li, J.; Xiao, J.; Dong, Y.; Liu, H. ACS Catal. 2015, 5, 6111−6137.
(16) Cismesia, M. A.; Yoon, T. P. Chem. Sci. 2015, 6, 5426−5434.
(17) (a) Wang, L.; Wei, X.-J.; Jia, W.-L.; Zhong, J.-J.; Wu, L.-Z.; Liu,
Q. Org. Lett. 2014, 16, 5842−5845. (b) Shi, G.; Shao, C.; Pan, S.; Yu,
J.; Zhang, Y. Org. Lett. 2015, 17, 38−41.
(18) TEMPO can accordingly not be considered an innocent
bystander for these reactions.
(19) For (a), 15% hydro-debrominated starting material was
observed. For (c), reaction monitored over time by 31P NMR was
performed with 0.33 equiv of DPEPhosPd(CH2TMS)2, 0.40 equiv of
TEMPO, and 1.0 equiv of bromodifluoroacetamide 4.
(20)
(6) For reductive elimination of TMSCH2CN from L2Pd(CH2TMS)
CN complexes, see: Marcone, J. E.; Moloy, K. G. J. Am. Chem. Soc.
1998, 120, 8527−8528.
(7) (a) Vela, J.; Lief, G. R.; Shen, Z.; Jordan, R. F. Organometallics
2007, 26, 6624−6635. (b) Cai, Y.; Shi, Y. Dalton Trans. 2013, 42,
5232−5236.
(8) (a) Dubowchik, G. M.; Vrudhula, V. M.; Dasgupta, B.; Ditta, J.;
Chen, T.; Sheriff, S.; Sipman, K.; Witmer, M.; Tredup, J.; Vyas, D. M.;
Verdoorn, T. A.; Bollini, S.; Vinitsky, A. Org. Lett. 2001, 3, 3987−3990.
(b) Ward, S. E.; Harries, M.; Aldegheri, L.; Austin, N. E.; Ballantine, S.;
Ballini, E.; Bradley, D. M.; Bax, B. D.; Clarke, B. P.; Harris, A. J.;
Harrison, S. A.; Melarange, R. A.; Mookherjee, C.; Mosley, J.; Dal
Negro, G.; Oliosi, B.; Smith, K. J.; Thewlis, K. M.; Woollard, P. M.;
Yusaf, S. P. J. Med. Chem. 2010, 53, 78−94.
(9) For synthesis of aryldifluoroamides, esters, and ketones, see:
(a) Ohtsuka, Y.; Yamakawa, T. Tetrahedron 2011, 67, 2323−2331.
(b) Fujikawa, K.; Fujioka, Y.; Kobayashi, A.; Amii, H. Org. Lett. 2011,
13, 5560−5563. (c) Guo, C.; Wang, R.-W.; Qing, F.-L. J. Fluorine
Chem. 2012, 143, 135−142. (d) Xiao, Y. L.; Guo, W. H.; He, G. Z.;
Pan, Q.; Zhang, X. Angew. Chem., Int. Ed. 2014, 53, 9909−9913.
(e) Ge, S.; Chaladaj, W.; Hartwig, J. F. J. Am. Chem. Soc. 2014, 136,
4149−4152. (f) Ge, S.; Arlow, S. I.; Mormino, M. G.; Hartwig, J. F. J.
Am. Chem. Soc. 2014, 136, 14401−14404. (g) Jung, J.; Kim, E.; You,
Y.; Cho, E. J. Adv. Synth. Catal. 2014, 356, 2741−2748. (h) Wang, L.;
Wei, X.-J.; Jia, W.-L.; Zhong, J.-J.; Wu, L.-Z.; Liu, Q. Org. Lett. 2014,
16, 5842−5845. (i) Shi, S.-L.; Buchwald, S. L. Angew. Chem., Int. Ed.
2015, 54, 1646−1650. (j) Arlow, S. I.; Hartwig, J. F. Angew. Chem., Int.
Ed. 2016, 55, 4567−4572. (k) Tarui, A.; Shinohara, S.; Sato, K.;
Omote, M.; Ando, A. Org. Lett. 2016, 18, 1128−1131.
(10) For the generation of alkyl radicals from gem-bromodifluor-
oacetamides and acetates, see: (a) Nguyen, J. D.; Tucker, J. W.;
Konieczynska, M. D.; Stephenson, C. R. J. J. Am. Chem. Soc. 2011, 133,
4160−4163. (b) Yu, C.; Iqbal, N.; Park, S.; Cho, E. J. Chem. Commun.
2014, 50, 12884−12887. (c) Su, Y. M.; Hou, Y.; Yin, F.; Xu, Y. M.; Li,
Y.; Zheng, X. Q.; Wang, X. S. Org. Lett. 2014, 16, 2958−2961.
(d) Feng, Z.; Min, Q.-Q.; Zhao, H.-Y.; Gu, J.-W.; Zhang, X. Angew.
Chem., Int. Ed. 2015, 54, 1270−1274. (e) Prieto, A.; Melot, R.;
Bouyssi, D.; Monteiro, N. Angew. Chem., Int. Ed. 2016, 55, 1885−1889.
(f) Li, Z.; Garcia-Dominguez, A.; Nevado, C. J. Am. Chem. Soc. 2015,
136, 11610−11613. (g) Gu, Z.; Zhang, H.; Xu, P.; Cheng, Y.; Zhu, C.
Adv. Synth. Catal. 2015, 357, 3057−3063. (h) Fu, W.; Zhu, M.; Zou,
G.; Xu, C.; Wang, Z.; Ji, B. J. Org. Chem. 2015, 80, 4766−4770 Also
see refs 9a, f, h, and 11..
(21) See Supporting Information for additional details.
(22) For examples of excitation of palladium(II) complexes, see:
(a) Ozawa, F.; Yamamoto, A.; Ikariya, T.; Grubbs, R. H. Organo-
metallics 1982, 1, 1481−1485. (b) Caspar, J. V. J. Am. Chem. Soc. 1985,
107, 6718−6719. (c) Wakatsuki, Y.; Yamazaki, H.; Grutsch, P. A.;
Santhanam, M.; Kutal, C. J. Am. Chem. Soc. 1985, 107, 8153−8159.
(d) To, W.-P.; Liu, Y.; Lau, T.-C.; Che, C.-M. Chem. - Eur. J. 2013, 19,
5654−5664.
(23) For a review on alkyl radical formation from alkyl halides in the
presence of palladium(0) complexes under light irradiation, see:
Sumino, S.; Fusano, A.; Fukuyama, T.; Ryu, I. Acc. Chem. Res. 2014,
47, 1563−1574.
(24) The major species responsible for the aryl hydrogen abstraction
is likely CH2TMS radicals. See discussion of the fate of the CH2TMS
ligand later in the article.
(25) It is also possible that step D takes place through the formation
of a palladium-alkyl bromide exciplex.
(26) For examples of bromide abstraction by palladium(II)
complexes, see: (a) Tsuji, J.; Sato, K.; Nagashima, H. Tetrahedron
1985, 41, 393−397. (b) Dneprovskii, A. S.; Ermoshkin, A. A.;
Kasatochkin, A. N.; Boyarskii, V. P. Russ. J. Org. Chem. 2003, 39, 933−
946. (c) Khusnutdinova, J. R.; Rath, N. P.; Mirica, L. M. Angew. Chem.,
Int. Ed. 2011, 50, 5532−5536.
(27) For examples of halide abstraction by palladium(0) complexes
for the formation of alkyl radicals, see: (a) Chen, Q.-Y.; Yang, Z.-Y.;
Zhao, C.-X.; Qiu, Z.-M. J. Chem. Soc., Perkin Trans. 1 1988, 563−567.
(b) Curran, D. P.; Chang, C.-T. Tetrahedron Lett. 1990, 31, 933−936.
(c) Ishiyama, T.; Abe, S.; Miyaura, N.; Suzuki, A. Chem. Lett. 1992, 21,
691−694. (d) Stadtmuller, H.; Vaupel, A.; Tucker, C. E.; Studemann,
T.; Knochel, P. Chem. - Eur. J. 1996, 2, 1204−1220. (e) Manolikakes,
G.; Knochel, P. Angew. Chem., Int. Ed. 2009, 48, 205−209. (f) Bloome,
K. S.; Alexanian, E. J. J. Am. Chem. Soc. 2010, 132, 12823−12825.
(g) Bloome, K. S.; McMahen, R. L.; Alexanian, E. J. J. Am. Chem. Soc.
2011, 133, 20146−20148. (h) Feng, Z.; Xiao, Y. L.; Zhang, B.; Zhang,
X. Angew. Chem., Int. Ed. 2014, 53, 1669−1673. (i) Besset, T.; Poisson,
T.; Pannecoucke, X. Chem. - Eur. J. 2014, 20, 16830−16845.
(11) (a) Andersen, T. L.; Frederiksen, M. W.; Domino, K.;
Skrydstrup, T. Angew. Chem., Int. Ed. 2016, 55, 10396−10400.
H
DOI: 10.1021/acs.organomet.6b00893
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(j) Parasram, M.; Chuentragool, P.; Sarkar, D.; Gevorgyan, V. J. Am.
Chem. Soc. 2016, 138, 6340−6343. (k) Sargent, B. T.; Alexanian, E. J. J.
Am. Chem. Soc. 2016, 138, 7520−7523. (l) Natte, K.; Jagadeesh, R. V.;
He, L.; Rabeah, J.; Chen, J.; Taeschler, C.; Ellinger, S.; Zaragoza, F.;
(41) Under these conditions and in contrast to the reactions with
DPEPhosPd(CH2TMS)2 several other species, which are unknown at
this stage, were observed in the 31P NMR spectrum of the crude
reaction mixture.
Neumann, H.; Bruckner, A.; Beller, M. Angew. Chem., Int. Ed. 2016, 55,
(42) The yield of dimer is relative to the amount of palladium added.
̈
2782−2786.
(28) Relative to the amount of palladium complex added. The
addition of additional equivalents of 4 further decreased the yield of
the dimer; however, it could not be completely suppressed under
thermal conditions.
(29) Seligson, A. L.; Trogler, W. C. J. Am. Chem. Soc. 1992, 114,
7085−7089.
(30) Accordingly, the mechanism in the presence of K2CO3 differs
from the stoichiometric reactions. Possibilities include Pd(0) catalysis
(ref 27) and a long radical chain pathway where DPE-
PhosPd(CH2TMS)2 serves as initiator. The reaction in Scheme 6
was completely inhibited by the presence of 20 mol % TEMPO or 20
mol % 1,4-dinitrobenzene. Two control reactions were performed: (1)
Under the conditions in Scheme 6, but in the absence of palladium
complex 3, no reaction occurs. (2) In the absence of 4, the palladium
complex decomposes in a highly unselective fashion under the
conditions in Scheme 6. The bis-phosphine oxide of DPEPhos was
independently synthesized by oxidation of DPEPhos with hydrogen
peroxide.
(31) For rate acceleration of palladium-catalyzed reactions by UV
light, see: (a) Imperato, G.; Konig, B. ChemSusChem 2008, 1, 993−
̈
996. (b) Fredricks, M. A.; Kohler, K. Catal. Lett. 2009, 133, 23−26.
̈
(c) Fredricks, M. A.; Drees, M.; Kohler, K. ChemCatChem 2010, 2,
̈
1467−1476.
(32) For a study on reductive elimination from L2PdMe2 complexes,
see: Gillie, A.; Stille, J. K. J. Am. Chem. Soc. 1980, 102, 4933−4941.
(33) For reports on reductive elimination from Pd(TMEDA)Me2,
see: (a) Calvin, G.; Coates, G. E. J. Chem. Soc. 1960, 2008−2016.
(b) Krause, J.; Cestaric, G.; Haack, K. J.; Seevogel, K.; Storm, W.;
Porschke, K. R. J. Am. Chem. Soc. 1999, 121, 9807−9823. (c) Reid, S.
M.; Mague, J. T.; Fink, M. J. J. Am. Chem. Soc. 2001, 123, 4081−4082.
(34) (a) Degraaf, W.; Boersma, J.; Smeets, W. J. J.; Spek, A. L.;
Vankoten, G. Organometallics 1989, 8, 2907−2917. (b) Degraaf, W.;
Boersma, J.; Vankoten, G. Organometallics 1990, 9, 1479−1484.
(35) Lanci, M. P.; Remy, M. S.; Kaminsky, W.; Mayer, J. M.; Sanford,
M. S. J. Am. Chem. Soc. 2009, 131, 15618−15620.
(36) Although our reactions are not performed in the presence of a
ferrocenium oxidant, a related mechanism via Pd(III) species 9 can be
hypothesized in the cases where (CH2TMS)2 is observed (see below).
Due to the limited amount of (CH2TMS)2 formed, this cannot be the
major pathway when two or more equivalents of 4 are used. Under the
photoinduced conditions, the absence of (CH2TMS)2 excludes this
pathway.
(37) For related studies of monomethyl palladium(III) complexes
made from monomethyl palladium(II) precursors and one-electron
oxidants, see: (a) Khusnutdinova, J. R.; Rath, N. P.; Mirica, L. M. J.
Am. Chem. Soc. 2010, 132, 7303−7305. (b) Lotz, D. M.; Remy, M. S.;
Lao, D. B.; Ariafard, A.; Yates, B. F.; Canty, A. J.; Mayer, J. M.;
Sanford, M. S. J. Am. Chem. Soc. 2014, 136, 8237−8242.
(38) Lanci, M. P.; McKeown, B. A.; Lao, D. B.; Spettel, K. E.; Mayer,
J. M. Polyhedron 2014, 84, 44−50.
(39) Van Leeuwen, P. W. N. M.; Roobeek, C. F.; Huis, R. J.
Organomet. Chem. 1977, 142, 233−241.
(40) Crystals of sufficient quality for X-ray analysis have not been
obtained. However, the formation of complex 11 is in line with the
reactivity displayed by complex 3 and correlates well with the
concomitant formation of 2a.
I
DOI: 10.1021/acs.organomet.6b00893
Organometallics XXXX, XXX, XXX−XXX