A well-defined (POCOP)Rh catalyst for the coupling of aryl halides with thiols

Article abstract of DOI:10.1021/ja505576g

This article describes a well-defined pincer-Rh catalyst for C-S cross-coupling reactions. (POCOP)Rh(H)(Cl) serves as an active precatalyst for the coupling of aryl chlorides and bromides with aryl and alkyl thiols under reasonable conditions (3% mol cat., 110 °C, 2-24 h, >90% yield). For select substrates, >90% yields were obtained with catalyst loading as low as 0.1%. Key mechanistic intermediates have been isolated and fully characterized, including (POCOP)Rh(Ph)(SPh) (6a) and (POCOP)Rh(SPh₂) (6b). The aryl/bis(phosphinite) (POCOP)Rh system has been shown to favor aryl thiolate reductive elimination at elevated temperatures and in some cases at room temperature, compared with the analogous diarylamido/bis(phosphine) (PNP)Rh pincer system. Concerted reductive elimination has been studied with 6a directly and in the presence of aryl bromide and aryl chloride traps. This investigation demonstrates a clear rate dependence on aryl chloride concentration during catalysis, a dependence that is absent when using aryl bromides. The rate of catalysis is dramatically reduced or brought to zero for ortho-tolyl halides, which can be traced to slower C-S coupling and slower carbon-halogen oxidative addition for ortho-substituted aryls. The influence of the sterics in the thiol component is less straightforward. The S-H oxidative addition product (POCOP)Rh(H)(SPh) (16) has been fully characterized and its reactivity has been examined, resulting in the isolation of the sodium-thiolate adduct (POCOP)Rh(NaSPh) (19). The solid-state structure of 19 shows Na interactions not only with sulfur, but also with a neighboring Rh and the chelating aryl carbon of the pincer framework. The reactivity of 16 and 19 indicates that these potential side products should not hinder catalysis.

Full text of DOI:10.1021/ja505576g

Article
pubs.acs.org/JACS
A Well-Defined (POCOP)Rh Catalyst for the Coupling of Aryl Halides
with Thiols
Samuel D. Timpa, Christopher J. Pell, and Oleg V. Ozerov*
Department of Chemistry, Texas A&M University, College Station, Texas 77842, United States
S
* Supporting Information
ABSTRACT: This article describes a well-defined pincer-Rh
catalyst for C−S cross-coupling reactions. (POCOP)Rh(H)-
(Cl) serves as an active precatalyst for the coupling of aryl
chlorides and bromides with aryl and alkyl thiols under
reasonable conditions (3% mol cat., 110 °C, 2−24 h, >90%
yield). For select substrates, >90% yields were obtained with
catalyst loading as low as 0.1%. Key mechanistic intermediates
have been isolated and fully characterized, including
(POCOP)Rh(Ph)(SPh) (6a) and (POCOP)Rh(SPh2) (6b).
The aryl/bis(phosphinite) (POCOP)Rh system has been
shown to favor aryl thiolate reductive elimination at elevated
temperatures and in some cases at room temperature,
compared with the analogous diarylamido/bis(phosphine) (PNP)Rh pincer system. Concerted reductive elimination has
been studied with 6a directly and in the presence of aryl bromide and aryl chloride traps. This investigation demonstrates a clear
rate dependence on aryl chloride concentration during catalysis, a dependence that is absent when using aryl bromides. The rate
of catalysis is dramatically reduced or brought to zero for ortho-tolyl halides, which can be traced to slower C−S coupling and
slower carbon−halogen oxidative addition for ortho-substituted aryls. The influence of the sterics in the thiol component is less
straightforward. The S−H oxidative addition product (POCOP)Rh(H)(SPh) (16) has been fully characterized and its reactivity
has been examined, resulting in the isolation of the sodium-thiolate adduct (POCOP)Rh(NaSPh) (19). The solid-state structure
of 19 shows Na interactions not only with sulfur, but also with a neighboring Rh and the chelating aryl carbon of the pincer
framework. The reactivity of 16 and 19 indicates that these potential side products should not hinder catalysis.
bromides. It still required a relatively high temperature for
operation (>100 °C) and the cost of the CyPF-tBu ligand is
quite high (hundreds of dollars per g).
The wide-scale success in using Pd to effectively perform
coupling reactions owes to its capacity to undergo two-electron
oxidative addition (OA) and reductive elimination (RE) events,
alternating between Pd0 and PdII oxidation states. The
mechanism of CyPF-tBu/Pd-catalyzed C−S cross-coupling
reactions was extensively examined by Hartwig and co-
workers.12 Scheme 1 shows the three steps that in broad
INTRODUCTION
■
The prevalence of aryl sulfides in biologically and pharmaceuti-
cally active compounds has sparked an increased interest
toward improving methodologies to form these groups.1 Aryl-
sulfur bonds can be formed by the direct reaction of thiolates
with aryl halides under harsh conditions (>200 °C).2 The
development of transition metal catalysts for C−S coupling of
aryl halides with thiols is complicated by the poisoning of a
number of catalysts by thiols, as noted by Hartwig.3 The Cu-
catalyzed Ullman reaction4 allowed progress in selectivity and
more reasonable reaction conditions. More recently, other
transition metals, including Ni5 and Co,6 have also been used to
catalyze the coupling of aryl halides with thiols to form diaryl
thioethers. However, they are characterized by modest turnover
numbers and difficulty in engaging aryl chlorides.
Scheme 1. General Pd0/PdII and RhI/RhIII Cycles for the
Coupling of Aryl Halides with Thiols To Form Sulfides
Pd catalysts are ubiquitous in carbon−carbon and carbon−
heteroatom coupling,7,8 and they have also been used for C−S
coupling.3,9−11 The utility of Pd catalysis with aryl chlorides
hinges on the use on bulky designer NHC ligands10 or “fourth-
generation” bidentate phosphines.3 Arguably the most
successful system was developed by Hartwig et al., with the
Pd center supported by the Josiphos-type “CyPF-tBu”
ligand.3,11 The CyPF-tBu/Pd system was capable of impressive
102−104 turnovers over 5−50 h with aryl chlorides and
Received: June 3, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja505576g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
strokes are typical7,8 for Pd-catalyzed aryl halide coupling
reactions in general: OA of aryl halide to Pd0, transmetalation
at PdII, and RE of the coupled product from PdII to regenerate
Pd0.
fused about Rh, resulting in a different P−Rh−P pincer bite
angle. Using Fryzuk’s notation for pincer ligands,22 4 can be
described as a {[5,6]-PCP} ligand in contrast to {[5,5]-PCP}
for 1 and 3 and {[5,5]-PNP} for 2.
We have recently reported13,14 on the ability of Rh
complexes supported by PNP and POCOP pincer15 ligands
to mimic the chemistry typical for LnPd-based catalysts. This
mimicry is somewhat suprpising given that the RhI/RhIII couple
(d8/d6) and Pd0/PdII (d10/d8) possess different numbers of d-
electrons and different geometries about the metal center. Our
investigations have shown that (pincer)RhI fragments easily
undergo OA with aryl halides to give (pincer)RhIII(Ar)(Hal)
complexes, and clean carbon−carbon RE from (pincer)-
RhIII(Ar)(Ar′).13 Besides stoichiometric reactions, we were
able to demonstrate the catalytic proficiency of (POCOP)Rh in
Kumada coupling of aryl iodides,14a and modest activity in C−
N coupling of aryl bromides and chlorides.14b Non-pincer
complexes of Rh have also been used for select aryl halide
coupling reactions.16,17 They likely make use of the RhI/RhIII
cycle, as well, but mechanistic information about these
processes is scarce, with the exception of the Bergman−Ellman
arylation of heterocycles.17 Previous reports of aryl−sulfur
coupling with Rh involved thiolation of electron deficient aryl
fluorides18 or reaction of aryl fluorides with a R2S2/R′3P
combination of reagents,19 but not the coupling of unactivated
aryl chlorides with thiols.
Each compound 1−4 was tested as a catalyst under the same
conditions (3% loading, 2 h, 110 °C) for the coupling of p-
MeC6H4SH or c-C6H11SH with p-BrC6H4CF3 using NaOtBu as
a base (Scheme 2). Dehydrochlorination of (pincer)Rh(H)(Cl)
with NaOtBu is expected13,14 to provide access to the
catalytically relevant three-coordinate (pincer)Rh intermedi-
ate.23 The use of the CF3-substituted aryl halide allowed for the
reactions to be conveniently monitored by 19F NMR
spectroscopy.
The reactions using the bulkiest complex 3 produced only a
near-stoichiometric amount of the expected C−S coupling
products, but the use of complexes that contained smaller PiPr2
side arms (1, 2, and 4) clearly led to catalytic diorganosulfide
production. 1 displayed the greatest activity yielding 97% of the
C−S coupled products with both thiol substrates. The
diarylamido-based 2 worked well with c-C6H11SH producing
95% of the coupled product, but gave only 20% conversion for
the reaction with p-MeC6H4SH. The naphthalenediol-based 4
worked almost as well as 1 for both coupling test reactions. We
selected 1 for more in-depth studies.
Scope. The results of a survey of the catalytic activity of 1 in
the coupling of aryl chlorides and bromides with alkyl or aryl
thiols are shown in Figure 1. Several general observations can
Against this backdrop, we sought to explore whether a well-
defined (pincer)Rh catalyst can be effective for C−S coupling
according to the envisioned RhI/RhIII mechanism shown in
Scheme 1. This inquiry has been successful, and here we
describe the efficacy of the (POCOP)Rh pincer system in
catalyzing the coupling of aryl chlorides and bromides with aryl
and alkyl thiols, and report preliminary mechanistic findings.
RESULTS AND DISCUSSION
■
Catalyst Screening. We set out to evaluate (POCOP)Rh-
(H)(Cl) (1), (PNP)Rh(H)(Cl) (2), (POCOPtBu)Rh(H)(Cl)
(3), and (NaptPOCOP)Rh(H)(Cl) (4) as a representative
group of precatalysts (Scheme 2). The synthesis of 1−4 and
their respective ligands has been described elsewhere.14,20,21
Complexes 1−4 are similar in that the Rh center in each is
supported by a monoanionic pincer ligand with opposing
phosphine/phosphinite arms. Complex 2 provides for a
different central donor atom: amido in 2 vs aryl in others.
Complex 3 possesses much greater steric bulk in the PtBu2
“arms”. Complex 4 presents a five- and a six-membered ring
Scheme 2. Catalyst Screening for the Coupling of p-
BrC6H4CF3 with c-C6H11SH and p-MeC6H4SH
Figure 1. Scope of (POCOP)Rh(H)(Cl) (1) catalyzed coupling of
aryl halides with aryl and alkyl thiols. NMR yields are given; isolated
yields in parentheses.
B
dx.doi.org/10.1021/ja505576g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
be made that dovetail related trends in Pd catalysis.7 Aryl
bromides were more reactive than aryl chlorides across the
board. More electron-poor aryl halides also reacted faster. This
can be exemplified by the ca. 50% yield of a/b starting from p-
ClC6H4F vs >90% yield of c/d starting from p-ClC6H4CF3 after
the same period of time. Primary and secondary (c-C6H11SH)
alkyl thiols reacted faster than aromatic thiols (e.g., l/v vs a/b),
as they do with Pd catalysts.11 Most reactions between alkyl
thiols and para-substituted aryl bromides were complete after 3
h at 110 °C, with NMR yields of 96% or greater. The analogous
reactions with aryl chlorides were slower to reach completion,
20 −48 h, but product yields were still high (88−98%, except
x). The formation of q with the more electron rich p-
MeC6H4Cl was also successful, yielding 93% after 36 h.
However, the use of p-MeOC6H4Cl gave only 50% of x after 36
h. The ester-substituted aryl bromides (m, n) gave diminished
yields of the corresponding thioethers (the balance being the
hydrodehalogenation products). The −CONEt2 (o) and
−NMe2 (g) substituted aryl bromides worked well, however.
c-C6H11SH gave a modest 30% yield of the coupling product
when paired with o-MeC6H4Br and none with o-MeC6H4Cl (r).
The attempted coupling of o-MeC6H4Cl or o-MeC6H4Br was
entirely unsuccessful with an unencumbered aryl thiol (p-
MeC6H4SH, h). However, ortho-substituted diaryl sulfide
products were synthesized in moderate to good yields via the
coupling of ortho-substituted aryl thiols with p-ClC6H4F or p-
BrC6H4F (i−k). As the steric bulk was increased from i to j and
k, the reaction times required for high conversion also
increased to 40 and 80 h, respectively. Increased steric bulk
of the reagents appears to be inhibitive of the catalysis, but it
affects the thiol to a different degree than the aryl halide
coupling partner (vide infra). We did not observe catalytic
formation of thioethers in reactions with p-fluorophenyl
tosylate (a, l). p-Fluorophenyl triflate reacted only sluggishly
with cyclohexanethiol (l) and not at all with thiophenol (a).
Many of the reactions also produced a minor quantity (≤5%,
except esters m and n, which gave 24% and 17%, respectively)
of the corresponding Ar−H (Tables S4 and S5 in the
Supporting Information). Its origin remains unclear but we
have been able to establish that traces of moisture are not
responsible. We tested this by examining formation of PhF in
the synthesis of b and v with ArBr using reagents either simply
degassed or rigorously dried by distillation from CaH2.
However, all reactions produced an equal amount of ca. 3%
PhF. The formation of Ar−H is not detrimental to product
isolation when aryl halides of relatively low molecular weight
are used.
Scheme 3. Synthesis of Complexes 5−9
Treatment of 5-Br with an additional equivalent of NaOtBu
in the presence of HSPh resulted in immediate formation of
(POCOP)Rh(Ph)(SPh) (6a) (Scheme 3). Inspection of the 1H
and 31P{1H} NMR spectra of a C6D6 solution of 6a after 1 wk
at ambient temperature indicated approximately 3% conversion
to (POCOP)Rh(SPh2) (6b). 6b was synthesized directly from
1 by treatment with NaOtBu in the presence of SPh2 at ambient
temperature (Scheme 3). Independent thermolysis of 6a or 6b
at 65 °C produced an equilibrium mixture of 6b:6a in a 48:1
ratio.
The alkylthiolate analogues, (POCOP)Rh(Ph)(SnC5H11)
(7a) and (POCOP)Rh(C6H4F)(StBu) (9a), were synthesized
using the same method as for 6a (Scheme 3). 7a could only be
observed in solution in a mixture and underwent 60%
conversion to the reductive coupling product (POCOP)Rh-
(SPh(nC5H11)) (7b) after 15 min at room temperature, while
9a produced no observable quantity of its reductive coupling
product (POCOP)Rh(S(C6H4F)tBu) (9b) after 24 h at room
temperature. However, thermolysis of 9a (9 h at 65 °C)
resulted in quantitative conversion to 9b.
Molecular structures of 6a and 6b in the solid state were
determined by single-crystal X-ray diffractometry (Figure 2). X-
ray-quality crystals of 6a were obtained from a saturated
pentane solution at −35 °C. The solid-state structure presented
a distorted square pyramidal coordination environment about
the d6 RhIII metal center.24 The geometry about Rh is quite
similar to the structure of (POCOP)Rh(Ph)(I) we reported
previously.14 The phenyl group in 6a minimizes the steric clash
with the iPr substituents on the phosphorus atoms by adopting
a conformation where the plane of the phenyl ring is
approximately perpendicular to the P−Rh−P vector. X-ray-
quality crystals of 6b were grown from a saturated toluene
solution layered with pentane at −35 °C. 6b adopts a distorted
square planar geometry about Rh, as would be expected for a
four-coordinate RhI complex.
In order to examine the longevity of the catalyst, we
conducted a series of coupling reactions with decreased catalyst
loadings (Tables S6 and S7 in the Supporting Information)
using p-MeC6H4SH and p-F3CC6H4X (X = Br or Cl) as
substrates in one set of experiments, and c-C6H11SH and p-
FC6H4X (X = Br or Cl) in another. We were able to observe
TON of up to 2500 with ArBr and 350 with ArCl but at low
conversions. The lowest practical catalyst loading (>90%
conversion in <3 d) is probably 0.1% for aryl bromides and
1% for aryl chlorides.
Synthesis of Intermediates. We previously described the
synthesis of various (POCOP)Rh(Ar)(X) (X = Cl, Br, I)
analogues.14,16 (POCOP)Rh(Ph)(Br) (5-Br) was formed by
treatment of 1 with NaOtBu in the presence of PhBr (Scheme
3). The reaction occurred upon mixing at room temperature as
indicated by a distinct color change from orange to red.
C
dx.doi.org/10.1021/ja505576g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(12b → 10), and OA (10 → 11-X). Scheme 5 illustrates these
three steps and highlights the rate constants relevant to the
Scheme 5. Elementary Steps of the Catalytic Cycle with
Notations for Rate Constants
following discussion. The OA step is clearly irreversible in all
cases, and we need only concern ourselves with the forward
reaction (k3). On the other hand, the reductive coupling step
(k1/k−1) and the product dissociation step (k2/k−2) deserved
closer attention.
Figure 2. ORTEP drawings (50% probability ellipsoids) of (POCOP)-
Rh(Ph)(SPh) (6a) (left) and (POCOP)Rh(SPh2) (6b) (right)
showing selected atom labeling. Hydrogen atoms are omitted for
clarity. Selected bond distances (Å) and angles (deg) for 6a: Rh1−S1,
2.366(1); S1−C1, 1.774(5); Rh1−C7, 2.3852(8), Rh1−S1−C1,
114.7(2); C13−Rh1−S1, 154.6(1); C13−Rh−C7, 87.7(1); C7−
Rh1−S1, 116.2(1). Selected bond distances (Å) and angles (deg)
for 6b: Rh1−S1, 2.3266(6); S1−C1, 1.786(2); S1−C7, 1.797(2);
Rh1−S1−C1, 116.93(8); Rh1−S1−C7, 118.00(7); C13−Rh1−S1,
166.90(6); C1−S1−C7, 99.8(1).
As was mentioned above, the equilibria between 6a and 6b or
7a and 7b strongly favor the C−S coupled RhI products. This
means that the k1 ≫ k−1 and that the oxidative cleavage
reaction (corresponding to k−1) is not relevant to consid-
erations of the rate-limiting step. Interestingly, we previously
described an equilibrium observed for the (PNP)Rh analogues
of 6a and 6b with a 1:1 ratio of (PNP)Rh(Ph)(SPh) and
(PNP)Rh(SPh2).26 The increased preference for the RhI isomer
may be attributed to the less electron rich nature of the
(POCOP)Rh system vs (PNP)Rh;27 which likely contributes to
the enhanced catalytic performance with 1 compared to 2.
Next, we explored the influence of the aryl halide substrate
(Br vs Cl) on the reaction rates. In general (Figure 1), the use
of ArBr led to faster catalysis than the use of analogous ArCl. In
a specific comparison, we examined the apparent rate of the
coupling of p-FC6H4X (X = Br, Cl) with c-C6H11SH using
either 1.1 or 10 equiv of p-FC6H4X (Table 1). The apparent
Mechanism. The proposed mechanism for catalytic C−S
coupling with 1 is shown in Scheme 4. NaOtBu serves to
Scheme 4. Proposed Catalytic Cycle for the Coupling of Aryl
Halides with Thiols Using 1 as the Catalyst
Table 1. Effects of Aryl Halide and Thioether
Concentrations
yield (%)
X
equiv Ar−X
equiv thioether
1 h
3 h
Br
Br
Br
Cl
Cl
Cl
1.1
1.1
10
0
1
0
0
1
0
80
88
81
17
6
97
96
97
40
17
90
1.1
1.1
10
dehydrochlorinate 1 and produce the unobserved, unsaturated
(POCOP)RhI fragment (10). We have previously demon-
strated the importance of three-coordinate (pincer)RhI frag-
ments in concerted OA reactions.13,14 OA of an aryl halide to
10 would then produce (POCOP)Rh(Ar)(X) (11-X). This
species would undergo transmetalation with NaSR25 to produce
(POCOP)Rh(Ar)(SR) (12a). The aryl/thiolato complex 12a
can undergo reductive coupling to form the thioether adduct
(POCOP)Rh(ArSR) (12b), followed by thioether dissociation
to re-form the unsaturated (POCOP)RhI fragment.
Isolation of compounds 5/6 allowed for closer study of each
of the elementary reactions making up the proposed catalytic
cycle. We were especially interested in elucidating parameters
affecting the overall rate of catalysis. The transmetalation step
can be confidently dismissed from this consideration given that
it is rapid and irreversible even at room temperature. Thus, we
focused on the remaining three steps with potential to be rate-
limiting: reductive coupling (12a → 12b), product dissociation
50
rate of catalysis was faster with p-FC6H4Br than with p-
FC6H4Cl. It was the same with 1.1 and 10 equiv of p-FC6H4Br,
while the analogous reactions with FC6H4Cl exhibited
enhanced conversion with 10 equiv of FC6H4Cl.28 In addition,
we found (Table 1) that excess thioether inhibits the coupling
of cyclohexanethiol with p-FC6H4Cl but not with p-FC6H4Br.
In a related pair of experiments, we examined the thermolysis
of 6a in the presence of PhBr or PhCl. In either reaction, 40%
consumption of 6a was observed after 4 h at 65 °C. But,
whereas in the reaction with PhBr, the only product detected
was 5-Br in 40% yield, the reaction with PhCl at the same point
in time produced 30% 6b and 10% (POCOP)Rh(Ph)(Cl) (5-
Cl). Additional thermolysis at 100 °C resulted in complete
conversion to the respective (POCOP)Rh(Ph)(X) products
(5-Br or 5-Cl) in both reactions. Related observations were
D
dx.doi.org/10.1021/ja505576g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
made in analyzing catalytic mixtures at intermediate conversion
for the apparent resting state of the catalyst. In the case of
PhCl/HSPh coupling we detected both 6a and 6b, while in the
PhBr/HSPh coupling, only 6a was detected.
Scheme 7 contains a series of reactions shedding light on the
influence of the ortho-substituent in the aryl halide. Compound
Scheme 7. Reactions Arising from o-Tolyl Halides
These results indicate several things about the kinetics and
mechanism for the coupling of simple aryl halides with PhSH.
The rate of reductive coupling is independent of the identity of
ArX, which is consistent with concerted, monomolecular C−S
reductive coupling. The reductive coupling (k1) is the rate-
limiting step in catalysis involving ArBr. Dissociation of ArSAr′
is much faster (k2 ≫ k1) and k−1 is of no concern given that k1
≫ k−1. Both the non-observation of 8 and the lack of apparent
dependence on [ArBr] in catalysis suggest that once generated,
the (POCOP)Rh intermediate is always trapped by ArBr to
irreversibly form the OA product (POCOP)Rh(Ar)(Br) and
the reverse trapping of (POCOP)Rh by the thioether product
is not competitive (k3[ArBr] ≫ k−2[ArSAr′] in the relevant
concentration ranges). In contrast, the trapping of (POCOP)
Rh with the thioether product is competitive with ArCl:
k3[ArCl] is comparable to k−2[ArSAr′], and so the rate of
catalysis is sensitive to [PhCl].
Changes in the steric bulk of the substrates have the potential
to affect various steps along the catalytic cycle. Bulkier aryl
halides or organothiols would result in sterically more imposing
thioether products, which may accelerate catalysis by virtue of
increasing k2 and decreasing k−2 (Scheme 5). This may be
especially in play in the catalytic synthesis of StBu-derived s
(Figure 1), one the bulkiest thoethers in our study. The
catalysis proceeded quite well compared to other thiols (cf. a, l,
t), considering that the rate of reductive coupling (9a → 9b)
was rather slow. On the other hand, increased steric bulk can
adversely affect k1 and k3. Goldman, Krogh-Jespersen, et al.
previously showed that, for Ar−X (X = Me, Ph, CHCH2,
CCH) RE from five-coordinate (PCP)Ir(Ar)(X), the aryl group
must be oriented “face-on” toward X in the transition state.29
The steric influence of the phosphines in the ground state of
(PCP)Ir(Ar)(X) favors the “side-on” orientation of the aryl
group with respect to X, and the necessary rotation of the aryl
can be a considerable component of the activation barrier. This
situation can be contrasted with the typical (R3P)nPd(Ar)(X)
(n = 1, 2) intermediates, where the steric bulk of the
phosphines favors “face-on” orientation of the aryl with respect
to X in the ground state and encourages faster RE.30 The
(POCOP)Rh(Ar)(X) system is sterically and electronically
analogous to Goldman’s (PCP)Ir(Ar)(X) system. From this
vantage point (Scheme 6), the rotation of the aryl group in
(POCOP)Rh(Ar)(SR) should be expected to be greatly
inhibited by the ortho-substituent. Ortho-substitution also has
the potential to inhibit OA of the aryl halide to (POCOP)Rh
since it presumably proceeds via a transition state with similar
spatial requirements.
13-Br was synthesized in the manner analogous to 5-Br. It was
found to exist as a mixture of two rotamers. Variable-
temperature NMR observations in the 20−110 °C range
showed no change in the width of the signals of 13-Br and no
change in the ratio of rotamers, indicating slow rotation on the
NMR time scale even at 110 °C. In contrast, examination of
1
NMR spectra of 5-Br showed coalescence of the Rh−Ph H
NMR resonances above 100 °C, consistent with a faster
rotation about the Rh−Caryl bond in the absence of the ortho-
substituent.
Transmetalation gave access to o- and p-tolyl isomers 14a
and 15a. Their thermolyses demonstrated that C−S reductive
coupling is indeed slower for the o-tolyl: 33% conversion to the
reductively coupled product 14b was recorded for 14a after 15
min at 80 °C (4 h to 95%), while the conversion of 15a to 15b
was complete31 by that point (12% conversion observed in
several min at room temperature during preparation). Oxidative
addition was examined using reactions of o-MeC6H4Br and
PhBr with 8. With PhBr, 40% conversion was observed after 10
min at room temperature (5 h at room temperature to
complete conversion), while for o-MeC6H4Br, reaching 40%
conversion required 1 h at 65 °C (20 h at 65 °C to complete
conversion). Looking at the OA comparison, we found that the
reaction of 6b proceeded much faster with PhBr than with o-
MeC6H4Br (Scheme 7). Thus, while RE and OA reactions
stemming from o-tolyl halides are in principle accessible, they
are both much slower. Whether that alone accounts for the very
poor reactivity of o-tolyl halides in (POCOP)Rh catalysis is not
fully clear. The reduced rate of the reactions in the desired cycle
Scheme 6. Schematic Illustration of the Rotation of the Aryl
Group Necessary for Attaining the Transition-State
Geometry for C−S Reductive Coupling
E
dx.doi.org/10.1021/ja505576g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
potentially opens the door to catalyst-killing side reactions,
although we have not observed any direct evidence of them.
The Hartwig group carried out extensive studies on the
influence of steric and electronic parameters on the rate of C−S
RE from square-planar Pd(II) compounds.32 In contrast to our
findings here, they found that introduction of a single o-methyl
group into the Pd-bound aryl slightly accelerates RE from
(DPPE)Pd(Ar)(StBu), although further buildup of steric
pressure in the aryl is deactivating.
Scheme 8. Formation of 16 and 19
Steric bulk would only impact the rate of catalysis if the RLS
is accelerated or decelerated. The observed effect of the sterics
of the thiolate group on the rate of catalysis and RE is less
straightforward. It is clear from the relative stability of 9a vs that
of 7a that increased steric bulk on the thiolate ligand ultimately
increases the barrier for reductive coupling. However, the
catalytic reaction to form l from the secondary alkyl thiol (c-
C6H11SH) and p-FC6H4Cl was faster (98%, 24 h) than the
reaction to form v from the primary alkyl thiol (n-C5H11SH)
and p-FC6H4Cl (94%, 48 h). Examination of the catalytic
coupling reactions with ortho-substituted aryl thiols (Figure 1,
i−k) shows an eventual decrease in the productivity of catalysis
as the steric bulk of the substituent is increased from i to j and
k. Compared to i, the reaction time increased by a factor of 2
for the formation of j and increased by a factor of 4 for k,
consistent with an increased barrier for reductive coupling with
bulky thiolate groups. Interestingly, the rate of formation of i is
similar or even greater to that of the para-substituted b and
non-substituted a. Ortho-substitution in the aryl thiol does not
appear to have as dramatic of an inhibiting effect as ortho-
substitution in the aryl halide, likely because it is farther
removed from the Rh center and has less impact on the ease of
achieving the orientation necessary for the C−S bond-forming
transition state. At the same time, ortho-substitution in the
diaryl thioether may have a positive effect on k2 (Scheme 5). All
of these results indicate that there is a delicate balance between
the sterics of the thiolate group and its effect on the rate of
catalysis. The 1998 Hartwig study30 of C−S RE from Pd(II)
found that increasing steric bulk in alkylthiolates was beneficial,
while increasing steric bulk in arylthiolates was detrimental to
the rate of RE.
1
and four distinct methyl resonances were visible in the H
NMR spectrum, consistent with the expected Cs-symmetric
structure for 16. As the solution reached −80 °C, a new set of
resonances appeared, corresponding to an additional Cs-
symmetric compound. The new resonances included a new
hydride resonance at −16.9 ppm, consistent with a hydride that
is not trans to an open coordination site. These data led us to
tentatively assign the new complex as [(POCOP)Rh(H)-
(SPh)]2 (18), the dimer of 16. A rapid equilibrium between 16
and 18 would not explain the C2v symmetry observed for 16 at
room temperature. Moreover, the observed hydride chemical
shift at −23.00 ppm at room temperature is consistent with a
hydride trans to an empty site and is similar to the chemical
shift of the hydride in 16 observed at −80 °C. Thus, the
dimeric 18 is not present in solutions of 16 at room
temperature in a significant amount.
Treatment of 16 with 1 equiv NaOtBu resulted in an
immediate color change to yellow-brown (Scheme 8). Analysis
of the reaction by 1H NMR spectroscopy showed conversion to
a new compound accompanied by the disappearance of the
hydride resonance and the appearance of HOtBu. The identity
of the new compound was determined to be (POCOP)Rh-
(NaSPh) (19) by X-ray crystallography (Figure 3), and is
consistent with the formal deprotonation of the coordinated
thiol by NaOtBu. The extended solid-state structure of 16
shows a chain structure where sodium has close interactions in
one (POCOP)Rh(SPh) unit with sulfur (Na−S, 2.751(2) Å),
and in another with the rhodium center (Na···Rh, 2.8958(13)
Å) and the Rh-bound carbon (Na···C, 2.881(3) Å).33
16 and 19 both convert to the product-forming 6a under the
conditions of catalysis (Scheme 8). 16 converted to 6a when
treated with NaOtBu (to yield 19) and then PhBr after 1 h at
room temperature (cf. 6b requiring 5 h for complete reaction
with PhBr). Complex 6b showed no reaction with NaSPh at
room temperature or after thermolysis at 110 °C. These
reactions illustrate that neither 16 nor 19 should form in an
appreciable concentration under the conditions of catalysis
(and we have not observed them in analyzing catalytic mixtures,
vide supra). These findings pertain to the reactions involving
NaSPh, PhBr, and SPh2, but it is possible that other
combinations of substituents may favor sodium thiolate
complexes to a greater degree.
Additional Reactivity. Hartwig and co-workers previously
identified LnPd(H)(SR) as the catalyst resting state for C−S
coupling reactions catalyzed by LnPd(Ar)(X).12 We were
interested to see if the (POCOP)Rh system would react with
thiols in a similar fashion, as well as the potential for a hydrido/
thiolato complex of this nature to impede catalysis. Treatment
of 6b with HSPh resulted in formation of (POCOP)Rh(H)-
1
(SPh) (16) upon mixing (Scheme 8). The H NMR spectrum
of 16 displayed a hydride resonance at −23.00 ppm and slightly
broadened POCOP signals. The spectrum displayed 1 methine
resonance and 2 resonances for the methyl protons of the
isopropyl groups, which is the pattern for a C2v-symmetric
complex. However, 16 should be Cs-symmetric (by analogy
with 1, for example) and should display two distinct methine
resonances and four different resonances for the methyl
protons. The apparent C2v symmetry could be the result of
fast reversible S−H reductive coupling on the metal, which
would create a rapid exchange between two degenerate forms
of 16 and the unobserved C2v-symmetric thiol adduct
(POCOP)Rh(HSPh) (17). 19 was examined by NMR
spectroscopy in the −80 °C to +20 °C range (Figure S15−
S17 in the Supporting Information). As a solution of 16 in d8-
toluene was cooled to −60 °C, two distinct methine resonances
CONCLUSION
■
In summary, we have presented a well-defined (POCOP)Rh
system for the catalytic coupling of aryl bromides and chlorides
with aryl and alkyl thiols. Our preliminary mechanistic studies
F
dx.doi.org/10.1021/ja505576g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Foundation (grant A-1717 to O.V.O.). We also thank Ms.
Linda Redd for editorial assistance.
REFERENCES
■
(1) (a) Dvorak, C. A.; Schmitz, W. D.; Poon, D. J.; Pryde, D. C.;
Lawson, J. P.; Amos, R. A.; Meyers, A. I. Angew. Chem., Int. Ed. 2000,
39, 1664. (b) Thomas, G. L.; Spandl, R. J.; Glansdorp, F. G.; Welch,
M.; Bender, A.; Cockfield, J.; Lindsay, J. A.; Bryant, C.; Brown, D. F. J.;
Loiseleur, O.; Rudyk, H.; Ladlow, M.; Spring, D. R. Angew. Chem., Int.
Ed. 2008, 47, 2808. (c) Kemperman, G. J.; Zhu, J.; Klunder, A. J. H.;
Zwaneburg, B. Eur. J. Org. Chem. 2001, 1817. (d) Le Grand, B.;
Pignier, C.; Letienne, R.; Cuisiat, F.; Rolland, F.; Mad, A.; Vacher, B. J.
Med. Chem. 2008, 51, 3856. (e) Liu, G.; Link, J. T.; Pei, Z.; Reilly, E.
B.; Leitza, S.; Nguyen, B.; Marsh, K. C.; Okasinski, G. F.; von Geldern,
T. W.; Ormes, M.; Fowler, K.; Gallatin, M. J. Med. Chem. 2000, 43,
4025.
(2) (a) Ley, S. V.; Thomas, A. W. Angew. Chem., Int. Ed. 2003, 42,
5400−5449. (b) Cristau, H. J.; Chabaud, B.; Chene, A.; Christol, H.
Synthesis 1981, 892.
Figure 3. ORTEP drawings (50% probability ellipsoids) of (POCOP)-
Rh(H)(Ph)(SPh) (16) (top left) and (POCOP)Rh(NaSPh) (19)
(top right). Hydrogen atoms are omitted for clarity. Selected bond
distances (Å) and angles (deg) for 16: Rh1−S1, 2.323(3); S1−C1,
1.78(1); Rh1−S1−C1, 113.8(8), C7−Rh1−S1, 157.7(3). Selected
bond distances (Å) and angles (deg) for 16: Rh1−S1, 2.3852(8); S1−
C1, 1.790(3); S1−Na1, 2.751(2); Rh1−S1−C1, 112.60(8); Rh1−S1−
Na1, 132.70(4); Na1−S1−C1, 87.59(8); C7−Rh1−S1, 174.45(7)(3).
The extended crystal packing of 19 (bottom) showing selected atom
labeling. Hydrogen atoms and isopropyl methyl groups are omitted for
clarity. The extended crystal packing structure of 19 shows sodium
interactions with sulfur (Na−S, 2.751(2) Å), neighboring rhodium
centers (Na···Rh, 2.8958(13) Å), and the chelating aryl carbon of the
pincer framework (Na···C, 2.881(3) Å).
(3) Hartwig, J. F. Acc. Chem. Res. 2008, 41, 1534.
(4) (a) Kondo, T.; Mitsudo, T.-A. Chem. Rev. 2000, 100, 3205.
(b) Bates, C. G.; Gujadhur, R. K.; Venkataraman, D. Org. Lett. 2002, 4,
2803. (c) Kwong, F. Y.; Buchwald, S. L. Org. Lett. 2002, 4, 3517.
(5) (a) Taniguchi, N. J. Org. Chem. 2004, 69, 6904. (b) Millois, C.;
Diaz, P. Org. Lett. 2000, 2, 1705. (c) Zhang, Y.; Ngeow, K. C.; Ying, J.
Y. Org. Lett. 2007, 9, 3495. (d) Percec, V.; Bae, J.-Y.; Hill, D. H. J. Org.
Chem. 1995, 60, 6895.
(6) Wong, Y.-C.; Jayanth, T. T.; Cheng, C.-H. Org. Lett. 2006, 8,
5613.
(7) Hartwig, J. F. Organotransition Metal Chemistry: From Bonding to
Catalysis; University Science Books: Sausalito, CA, 2009; Ch. 19.
(8) Negishi, E. I., Ed. Handbook of Organopalladium Chemistry for
Organic Synthesis; Wiley-Interscience: New York, 2002.
(9) (a) Murata, M.; Buchwald, S. L. Tetrahedron 2004, 60, 7397.
(b) Li, G. Y. Angew. Chem., Int. Ed. 2001, 40, 1513. (c) Schopfer, U.;
Schlapbach, A. Tetrahedron 2001, 57, 3069. (d) Zheng, N.;
McWilliams, J. C.; Fleitz, F. J.; Armstrong, J. D., III; Volante, R. P.
J. Org. Chem. 1998, 63, 9606.
(10) Bastug, G.; Nolan, S. P. J. Org. Chem. 2013, 78, 9303.
(11) (a) Fernandez-Rodriguez, M. A.; Shen, Q.; Hartwig, J. F.
Chem.Eur. J. 2006, 12, 7782. (b) Fernandez-Rodriguez, M. A.; Shen,
Q.; Hartwig, J. F. J. Am. Chem. Soc. 2006, 128, 2180.
showed increased favorability toward aryl−sulfur RE with
(POCOP)Rh relative to the analogous (PNP)Rh system. They
also exposed the apparent dependence of the rate of catalysis
on aryl chloride but not aryl bromide concentration. The
analysis of steric effects in the aryl halide and in the thiol
component reveals a rather complex picture, where the
influence of the sterics depends on the nature of the rate-
limiting step for various substrate couples. Reactions of PhSH
and PhSNa with the catalytically active species were examined
and deemed not to be detrimental under catalytic conditions.
These results support the proposed RhI/RhIII mechanism for
aryl halide thiol cross-coupling reactions with Rh.
(12) Alvaro, E.; Hartwig, J. F. J. Am. Chem. Soc. 2009, 131, 7858.
(13) (a) Gatard, S.; Celenligil-Cetin, R.; Guo, C.; Foxman, B. M.;
Ozerov, O. V. J. Am. Chem. Soc. 2006, 128, 2808. (b) Puri, M.; Gatard,
S.; Smith, D. A.; Ozerov, O. V. Organometallics 2011, 30, 2472.
(14) (a) Timpa, S. D.; Fafard, C. M.; Herbert, D. E.; Ozerov, O. V.
Dalton Trans. 2011, 40, 5426. (b) Timpa, S. D.; Pell, C. J.; Zhou, J.;
Ozerov, O. V. Organometallics 2014, DOI: 10.1021/om500577w.
(15) For general information on pincer complex chemistry, see:
(a) van der Boom, M. E.; Milstein, D. Chem. Rev. 2003, 103, 1759. (b)
The Chemistry of Pincer Compounds; Morales-Morales, D., Jensen, C.
M., Eds.; Elsevier: Amsterdam, 2007.
(16) (a) Grushin, V. V.; Alper, H. Organometallics 1991, 10, 1620.
(b) Ishiyama, T.; Hartwig, J. J. Am. Chem. Soc. 2000, 122, 12043.
(c) Murata, M.; Ishikura, M.; Nagata, M.; Watanabe, S.; Masuda, Y.
Org. Lett. 2002, 4, 1843. (d) Bedford, R. B.; Limmert, M. E. J. Org.
Chem. 2003, 68, 8669. (e) Ueura, K.; Satoh, T.; Miura, M. Org. Lett.
2005, 7, 2229. (f) Wang, X.; Lane, B. S.; Sames, D. J. Am. Chem. Soc.
2005, 127, 4996. (g) Yanagisawa, S.; Sudo, T.; Noyori, R.; Itami, K. J.
Am. Chem. Soc. 2006, 128, 11748. (h) Wu, J.; Zhang, L.; Gao, K. Eur. J.
Inorg. Chem. 2006, 5260. (i) Shintani, R.; Yamagami, T.; Hayashi, T.
Org. Lett. 2006, 8, 4799. (j) Ito, J.-I.; Miyakawa, T.; Nishiyama, H.
Organometallics 2008, 27, 3312. (k) Yamanoi, Y.; Nishihara, H. J. Org.
Chem. 2008, 73, 6671. (l) Zhang, L.; Wu, J. Adv. Synth. Catal. 2008,
350, 2409. (m) Kim, M.; Chang, S. Org. Lett. 2010, 12, 1640.
(17) Lewis, J. C.; Bergman, R. G.; Ellman, J. A. Acc. Chem. Res. 2008,
41, 1013.
ASSOCIATED CONTENT
* Supporting Information
■
S
Details of experimental procedures and characterization;
crystallographic information in the form of CIF files. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
ozerov@chem.tamu.edu
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We are grateful for the support of this work by the US National
Science Foundation (grant CHE-1300299), and the Welch
■
G
dx.doi.org/10.1021/ja505576g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(18) Arisawa, M.; Suzuki, T.; Ishikawa, T.; Yamaguchi, M. J. Am.
Chem. Soc. 2008, 130, 12214.
(19) Ajiki, K.; Hirano, M.; Tanaka, K. Org. Lett. 2005, 7, 4193.
(20) Pell, C. J.; Ozerov, O. V. ACS Catal. 2014, 4, 3470.
(21) (a) Ozerov, O. V.; Guo, C.; Papkov, V. A.; Foxman, B. M. J. Am.
Chem. Soc. 2004, 126, 4792. (b) Salem, H.; Shimon, L. J. W.; Leitus,
G.; Weiner, L.; Milstein, D. Organometallics 2008, 27, 2293.
(22) Fryzuk, M. D.; MacNeil, P. A. J. Am. Chem. Soc. 1981, 103, 3592.
(23) Treatment of 1 with 1 equiv of NaOtBu in C6D6 produced what
appeared to be a C2v-symmetric Rh(I) complex of the POCOP ligand,
but we have not pursued its isolation and characterization. It is likely a
solvent adduct, or a rapidly exchanging mixture of solvent or ligand
CH adducts and/or C−H oxidative addition products. For related
pseudo-three-coordinate Rh complexes of PBP and PNP ligands, see:
(a) Hasegawa, M.; Segawa, Y.; Yamashita, M.; Nozaki, K. Angew.
Chem., Int. Ed. 2012, 51, 6956. (b) Verat, A. Y.; Pink, M.; Fan, H.;
Tomaszewski, J.; Caulton, K. G. Organometallics 2008, 27, 166.
(c) Zhu, Y.; Smith, D. A.; Herbert, D. E.; Gatard, S.; Ozerov, O. V.
Chem. Commun. 2012, 48, 218.
(24) For discussions of structural preferences of five-coordinate d6
complexes, see: (a) Lam, W. H.; Shimada, S.; Batsanov, A. S.; Lin, Z.;
Marder, T. B.; Cowan, J. A.; Howard, J. A. K.; Mason, S. A.; McIntyre,
G. J. Organometallics 2003, 22, 4557. (b) Rachidi, I. E.-I.; Eisenstein,
O.; Jean, Y. New J. Chem. 1990, 14, 671. (c) Riehl, J.-F.; Jean, Y.;
Eisenstein, O.; Pelissier, M. Organometallics 1992, 11, 729. (d) Olivan,
M.; Eisenstein, O.; Caulton, K. G. Organometallics 1997, 16, 2227.
(25) NaOtBu is a sufficiently strong base to completely deprotonate
aryl or alkyl thiols in situ.
(26) Gatard, S.; Guo, C.; Foxman, B. M.; Ozerov, O. V.
Organometallics 2007, 26, 6066.
(27) νCO for (POCOP)Rh(CO) is 1962 cm−1 (ref 27a), vs 1945
cm−1 for (PNP)Rh(CO) (ref 27b): (a) Salem, H.; Ben-David, Y.;
Shimon, L. J. W.; Milstein, D. Organometallics 2006, 25, 2292.
(b) Weng, W.; Guo, C.; C
O. V. Chem. Commun. 2006, 197.
̧elenligil-Çetin, R.; Foxman, B. M.; Ozerov,
(28) The reaction with 10 equiv of ArCl also produced an increased
concentration of C6H5F, approximately 10%.
(29) Ghosh, R.; Emge, T. J.; Krogh-Jespersen, K.; Goldman, A. S. J.
Am. Chem. Soc. 2008, 130, 11317.
(30) Mann, G.; Shelby, Q.; Roy, A. H.; Hartwig, J. F. Organometallics
2003, 22, 2775.
(31) 15b and 15a apparently form an equilibrium mixture in a ca.
95:5 ratio.
(32) (a) Mann, G.; Baranano, D.; Hartwig, J. F.; Rheingold, A. L.;
Guzei, I. A. J. Am. Chem. Soc. 1998, 120, 9205. (b) Hartwig, J. F. Acc.
Chem. Res. 1998, 31, 852.
(33) (a) Jeigel, E.; Bock, H.; Krenzel, V.; Sievert, M. Acta Crystallogr.
2001, C57, 154. (b) Englich, U.; Chandwick, S.; Ruhlandt-Senge, K.
Inorg. Chem. 1998, 37, 283.
H
dx.doi.org/10.1021/ja505576g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX