Oxidative C-arylation of free (NH)-heterocycles via direct (sp3) C-H bond functionalization

Article abstract of DOI:10.1021/ja045402b

The development of a new chemical transformation, namely oxidative C-arylation of saturated (NH)-heterocycles, is described. This reaction combines dehydrogenation and arylation in one process, leading to cross-coupling of (NH)-heterocycles and haloarenes. Typical reaction conditions involve heating the reaction partners in anhydrous dioxane at 120-150 °C in the presence of RhCl(CO)[P(Fur)₃]₂ as the catalyst and Cs₂CO₃ as the base. Addition of tert-butylethylene as the hydrogen acceptor increases the chemical yield by diminishing the dehalogenation pathway. This method demonstrated a good substrate scope, allowing for cross-coupling of a variety of (NH)-heterocycles (e.g., pyrrolidine, piperidine, piperazine, morpholine) and halo(hetero)arenes to afford valuable heterocyclic products in one step. The preliminary mechanistic studies provided some insight regarding the key events in the proposed catalytic cycle, including β-hydride elimination of an amido rhodium complex and carbometalation of the resulting imine. A large kinetic isotope effect [KIE (k_C-H/k_C-D) = 4.3] suggests that one or both β-hydride elimination steps are rate determining. The central role for the phosphine ligand was established in controlling the partitioning between the oxidative C-arylation and N-arylation pathways. Copyright

Full text of DOI:10.1021/ja045402b

Published on Web 09/23/2004
Oxidative C-Arylation of Free (NH)-Heterocycles via Direct (sp3) C-H Bond
Functionalization
Bengu¨ Sezen and Dalibor Sames*
Department of Chemistry, Columbia UniVersity, 3000 Broadway, New York, New York 10027
Received July 30, 2004; E-mail: sames@chem.columbia.edu
This paper was retracted on March 1, 2006 (J. Am. Chem. Soc. 2006, 128, 3102).
Guided by the concept of direct and systematic elaboration of
heterocyclic compounds via C-H bond functionalization,1 we have
recently developed new methods of C-arylation of (NH)-heteroare-
nes such as indoles, pyrazoles, and imidazoles.1,2 In this context,
we became interested in the possibility of C-arylation of saturated
(NH)-heterocycles with haloarene donors. To achieve this goal
Figure 1. C-H bond versus N-H bond arylation.
would require disfavoring known N-arylation pathways to the
benefit of (sp3) C-H bond functionalization (Figure 1). Although
this proposition may seem unlikely, it has been demonstrated that
ruthenium fragment [RuHCl(PiPr3)2], generated in situ, undergoes
selective insertion into C-H bonds at the R-position of free
pyrrolidine.3 We followed this lead, however, with no success,
highlighting the challenge of translating the knowledge acquired
by C-H activation (metalation) studies into direct C-H bond
arylation processes. An alternative strategy would involve N-H
bond metalation, followed by â-hydride elimination and arylation
of the resulting imine intermediate. This latter scenario forms a
mechanistic rationale for a new synthetic method described herein
for oxidative C-arylation of free (NH)-heterocycles.
Scheme 1. Lead Identification
Scheme 2. Ligand Degradation and Dehalogenation
To explore the feasibility of this cross-coupling reaction,
pyrrolidine and iodobenzene were selected as the parent substrates.
A systematic study was undertaken to investigate three key
variables, namely the metal source, solvent, and base. This exercise
provided us with an exciting lead as well as indispensable insight
into the system in question (Scheme 1). Remarkably, rhodium(I)
complexes 5 and 6 catalyzed the formation of 2-phenylpyrroline
(1) as the main product in 45% yield. We note that only rhodium
complexes out of a broad array of examined transition metals were
capable of affording appreciable amounts of C-arylation products.
Compound 2 was also formed as the minor product, together with
a small amount N-phenylpyrrolidine.4 Upon acidic treatment of the
crude reaction mixture, additional product 4 was detected; this
compound is a result of 2′-arylation of product 1.
Scheme 3. Optimized Conditionsa
Interestingly, the cross-coupling between pyrrolidine and 4-io-
doanisole afforded the desired product 7 in 31% yield, as well as
2-phenylpyrroline 1 and anisole (Scheme 2). This result revealed
two major side reactions largely responsible for the low efficiency
of this process. The formation of compound 1 reflects carbon-
phosphorus bond cleavage in the triphenylphosphine ligand and
subsequent phenyl group migration, while the presence of a
significant amount of anisole reveals facile dehalogenation.
This insight guided our optimization studies, which began by
addressing the ligand stability and phenyl group scrambling. A panel
of ligands containing a plethora of phosphines and imidazolyl
carbenes was examined (Supporting Information). A clear parallel
between rhodium and palladium chemistry was seen in this context,
as ligands known to suppress â-hydride elimination, such as bulky
monodentate phosphines or bidentate phoshines, promoted N-
arylation to give 3.5 In contrast, triphenylphosphine favored
oxidative C-arylation, albeit with modest overall efficiency. In our
search for a robust alternative to Ph3P, we identified tri-(2-furyl)-
a P(Fur)3 ) tri-(2-furyl)phosphine, TBE ) tert-butylethylene. Condi-
tions: pyrrolidine (1.2 equiv), PhX (1 equiv), RhCl(CO)[P(Fur)3]2 (5 mol
%), TBE (2 equiv), Cs2CO3 (1.2 equiv), dioxane, 120 °C, 10 h. 10-15%
dehalogenation occurred. Compound 4 was also formed, <4%. The given
yields are the averages of three runs with a deviation of 2-3%. Purification
of reagents and anhydrous conditions are required (see Supporting Informa-
tion for details).
phosphine [(Fur)3P] as a promising candidate.5c,6 Indeed, the
corresponding rhodium complex RhCl(CO)[P(Fur)3]2 (8) proved
to be a superior catalyst, leading to a 2-fold increase in the efficiency
of the coupling reaction, while no transfer of the furyl group was
observed and the extent of dehalogenation was reduced. The latter
process was further abated (to approximately 10-15%) by addition
of tert-butylethylene (TBE) as the hydrogen acceptor (Scheme 3).
Under these optimized conditions, pyrrolidine and iodobenzene
can be coupled in one step to furnish product 1 in very good yield
9
13244
J. AM. CHEM. SOC. 2004, 126, 13244-13246
10.1021/ja045402b CCC: $27.50 © 2004 American Chemical Society

C O M M U N I C A T I O N S
Scheme 4. C-Arylation of Chiral Substratesa
Table 1. Reaction Scope: Halo(hetero)arene Donorsa
a X ) I for compounds 7, 9, 10; X ) Br for 11, 12. Conditions:
pyrrolidine (1.2 equiv), (Het)Ar-X (1 equiv), RhCl(CO)[P(Fur)3]2 (5 mol
%), TBE (2 equiv), Cs2CO3 (1.2 equiv), dioxane, 120 °C, 12-14 h. 10-
15% of dehalogenation occurred for compounds 7, 9. 25-30% of dehalo-
genation occurred for compounds 10-12.
a Conditions: (a) heterocycle (1.2 equiv), PhI (1 equiv), RhCl(CO)[P-
(Fur)3]2 (5 mol %), TBE (2 equiv), Cs2CO3 (1.2 equiv), dioxane, 120 °C,
15 h; (b) heterocycle (1.2 equiv), (Het)Ar-X (1 equiv), RhCl(CO)[P(Fur)3]2
(5 mol %), TBE (2 equiv), Cs2CO3 (1.2 equiv), dioxane, 150 °C, 16 h.
Table 2. (NH)-Heterocycle Substrate Scopea
and racemization occurred, affording dehydroproline 19 and race-
mized starting material (Scheme 4).
In contrast, â-substituted substrates worked well, showing
exclusive arylation at the less hindered R-methylene group. Thus,
unknown compounds 21 and 23 were prepared from protected
3-hydroxypyrrolidine 20 and ethyl nipecotate 22 in 59% and 52%
yield, respectively. Although the isolated yields were modest, the
direct method described herein, even in the present state of
optimization, competes favorably with traditional multistep ap-
proaches. This is particularly true in the case of more complex
compounds such as 23.9,10
The analysis of this system revealed that relatively low efficiency
may be ascribed to three key processes. First, nonproductive
reduction of haloarene donors (dehalogenation) is always present,
to the extent of 10-15% of the halide (and up to 30% with
haloheteroarene donors, Table 1).5a,11 Second, heterocyclic rings
larger than pyrrolidine require a higher reaction temperature, which
in turn increases other haloarene degradation pathways. Third,
catalyst decomposition represents the most serious limiting factor,
particularly at higher reaction temperatures.
The limits of this method can be illustrated in the condensation
of a mismatched reaction couple, such as ethyl nipecotate and
2-bromothiophene, which afforded a low yield of product 24 (11%
yield, Scheme 4). In this case, an unreactive acceptor required a
high reaction temperature, which at the same time caused fast
decomposition of an unstable heteroarene donor.
a Conditions: heterocycle (1.2 equiv), PhI (1 equiv), RhCl(CO)[P(Fur)3]2
(5 mol %), TBE (2 equiv), Cs2CO3 (1.2 equiv), dioxane. Reaction was
performed at 120 °C for 1. Reactions were performed at 150 °C for 13-18.
(78%), together with a small amount of amidine 2 and compound
4 (<5%). The oxidative C-arylation of pyrrolidine with iodobenzene
represents the parent example of a new chemical transformation
which unites dehydrogenation and arylation in one process.
Subsequently, we investigated the scope of this methodology
starting with haloarene donors. Both electron-donating and electron-
withdrawing substituents (in the 4-position) were tolerated, although
the yields were lower compared to the parent substrate (Table 1). We
were particularly interested in the prospect of (NH)-heterocycle-
heteroarene coupling. To our delight, compounds 10-12 were
prepared from pyrrolidine and the corresponding iodo- or bromo-
heteroarenes in one step. Lower efficiency may be ascribed to lower
stability of the haloheteroarenes in comparison to that of haloben-
zenes (Table 1).
In the next step, we evaluated the scope of (NH)-heterocyclic
substrates in terms of the size and character of the ring (Table 2).
Aware of the lower reactivity of piperidine in comparison with
pyrrolidine,7 we were pleased with 65% yield of product 13. Seven-
and eight-membered rings were also arylated, in >50% yield.
Furthermore, heterocycles of higher complexity, such as morpholine
and piperazine, provided the corresponding products in good yields,
demonstrating promising functional group compatibility of this
method.8
Last, preliminary mechanistic studies were carried out to establish
a crude framework of the catalytic cycle. The parent reaction
between pyrrolidine and iodobenzene was used for these investiga-
1
tions. When the reaction mixture was monitored by H and 31P
NMR at 80 °C, two major products were identified as complexes
25 and 30, and their structures were confirmed by independent
synthesis. Complex 25 showed identical behavior to the catalyst
precursor 8 in terms of both kinetic profile and chemical yield of
the catalytic reaction. Our observations showed that oxidative
addition of iodobenzene to rhodium(I) complex 8 was facile and
thus may represent the first, and probably reversible, step of the
cycle (Figure 2, cycle A). This may be followed by a fast metalation
of the N-H bond [KIE (kNH/kND) ) 1.0], proceeding, for instance,
via Lewis-acid-promoted deprotonation of pyrrolidine to furnish
rhodium(III) amido complex 26. The subsequent events involve
â-hydride elimination and formation of imine rhodium hydride 27,
followed by sequential carbometalation and a second â-hydride
elimination to yield 2-phenyl-1-pyrroline, coordinated to the
Finally, oxidative C-arylation of chiral substrates was examined.
As expected, R-substituted pyrrolidines exemplified by proline
methyl ester did not undergo arylation; instead, dehydrogenation
9
J. AM. CHEM. SOC. VOL. 126, NO. 41, 2004 13245

C O M M U N I C A T I O N S
2-aryltetrahydropyridines can be prepared in one synthetic step from
the corresponding saturated (NH)-heterocycles and haloarenes. To
further expand the substrate scope of this method, a deeper
mechanistic insight needs to be gained, particularly with respect to
the catalyst stability.
Acknowledgment. This work was supported by the NIGMS,
GlaxoSmithKline, and Merck Research Laboratories. D.S. is a
recipient of the AstraZeneca Excellence in Chemistry Award. B.S.
is a recipient of The Arun Guthikonda Memorial Fund graduate
fellowship and the Bristol-Myers Squibb graduate fellowship. We
thank Dr. J. B. Schwarz for editorial assistance, and Vitas Votier
Chmelar for intellectual support.
Supporting Information Available: Experimental procedures,
spectral data for all products, and kinetics of stoichiometric and catalytic
experiments with complexes 8, 25, and 30. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
Figure 2. Proposed mechanistic rationale. Cycle A, fast and more
productive cycle. Cycle B, slow and less productive cycle. Compounds 8,
25, and 30 were prepared and characterized.
(1) Sezen, B.; Sames, D. J. Am. Chem. Soc. 2003, 125, 10580-10585.
(2) Sezen, B.; Sames, D. J. Am. Chem. Soc. 2003, 125, 5274-5275.
(3) Ferrando-Miguel, G.; Coalter, J. N., III; Gerard, H.; Huffman, J. C.;
Eisenstein, O.; Caulton, K. G. New J. Chem. 2002, 26, 687-700.
(4) Pd(0)-catalyzed formation of 2 from pyrrolidine was reported: Murahashi,
S.-I.; Yoshimura, N.; Tsumiyama, T.; Kojima, T. J. Am. Chem. Soc. 1983,
105, 5002-5011.
(5) (a) Hartwig, J. F.; Richards, S.; Baranano, D.; Paul, F. J. Am. Chem. Soc.
1996, 118, 3626-3633. (b) Wagaw, S.; Rennels, R. A.; Buchwald, S. L.
J. Am. Chem. Soc. 1997, 119, 8451-8458. (c) Hamann, B. C.; Hartwig,
J. F. J. Am. Chem. Soc. 1998, 120, 3694-3703. (d) Strieter, E. R.;
Blackmond, D. G.; Buchwald, S. L. J. Am. Chem. Soc. 2003, 125, 13978-
13980.
(6) Andersen, N. G.; Keay, B. A. Chem. ReV. 2001, 101, 997-1030.
(7) Piperidine did not undergo R-metalation in Caulton’s studies; see ref 3.
(8) Note that no arylation was observed at the R-methylene positions adjacent
to the protected amine group in piperazine or oxygen in morpholine.
(9) Examples of traditional methods for preparation of R-substituted cyclic
imines: (a) Giovannini, A.; Savoia, D.; Umani-Ronchi, A. J. Org. Chem.
1989, 54, 228-234. (b) Dieter, R. K.; Li, S. J. Org. Chem. 1997, 62,
7726-7735.
(10) For examples of direct C-H to C-C transformations at the R-methylene
positions of N-protected pyrrolidines or tertiary amines, see: (a) Chatani,
N.; Asaumi, T.; Yorimitsu, S.; Ikeda, T.; Kakiuchi, F.; Murai, S. J. Am.
Chem. Soc. 2001, 123, 10935-10941. (b) Davies, H. W. L. J. Mol. Catal.
A 2002, 189, 125-135. (c) Murahashi, S.-I.; Komiya, N.; Terai, H.; Nakae,
T. J. Am. Chem. Soc. 2003, 125, 15312-15313. (d) DeBoef, B.; Pastine,
S. J.; Sames, D. J. Am. Chem. Soc. 2004, 126, 6556-6557.
(11) Lewis, J. C.; Wiedemann, S. H.; Bergman, R. G.; Ellman, J. A. Org.
Lett. 2004, 6, 35-38.
rhodium metal (cf. 29).12 Replacement of the organic product by
the phosphine ligand at the metal center would complete the cycle.
The key elementary reactions proposed in this cycle have firm
precedent, including â-hydride elimination of amido complexes13
and carbometalation of imines12a,14 with rhodium or closely related
metals. A large kinetic isotope effect [KIE (kC-H/kC-D) ) 4.3]
suggests that one or both â-hydride elimination steps are rate
determining. The phosphine ligand plays a key role in controlling
the partitioning between the oxidative arylation and N-arylation
pathways. Reductive elimination of benzene from complex 27
represents a plausible pathway for the competing dehalogenation
of arene donors.
An alternative cycle may begin with conversion of 8 to amido
complex 30 by the displacement of the chloride with pyrrolidine
(Figure 2, cycle B). This species may subsequently be converted
to complex 27 in two steps involving â-hydride elimination and
oxidative addition, in either order. Examination of complex 30 in
detail revealed that it could function as the catalyst, however, with
slower rates and significantly lower efficiency in comparison to
complex 8 and 25. The experiments involving both catalytic and
stoichiometric amounts of metal complexes 8, 25, and 30 (see
Supporting Information) support the notion that both catalytic cycles
may be operative, wherein cycle B represents the slower and less
productive one.
(12) An alternative mechanistic route from 27 to 29 may proceed via insertion
of the metal into the C-H bond of the imine, followed by reductive
elimination. (a) Ishiyama, T.; Hartwig, J. F. J. Am. Chem. Soc. 2000, 122,
12043-12044. (b) Krug, C.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 124,
1674-1679.
(13) (a) Diamond, S. E.; Mares, F. J. Organomet. Chem. 1977, 142, C55-
C57. (b) Hartwig, J. F. J. Am. Chem. Soc. 1996, 118, 7010-7011.
(14) (a) Krug, C.; Hartwig, J. F. J. Am. Chem. Soc. 2004, 126, 2694-2695.
(b) Oi, S.; Moro, M.; Fukuhara, H.; Kawanishi, T.; Inoue, Y. Tetrahedron
Lett. 1999, 40, 9259-9262. (c) Hayashi, T.; Ishigedani, M. J. Am. Chem.
Soc. 2000, 122, 976-977.
In summary, oxidative C-arylation of (NH)-heterocycles repre-
sents a new chemical transformation which unites two reactions,
namely dehydrogenation and arylation, into one process. Thus,
valuable heterocyclic compounds such as 2-arylpyrrolines and
JA045402B
9
13246 J. AM. CHEM. SOC. VOL. 126, NO. 41, 2004