Asymmetric cross-coupling of aryl triflates to the benzylic position of benzylamines

Article abstract of DOI:10.1002/anie.201201874

Ligand in a haystack: The first catalytic asymmetric cross-coupling of benzyllithiums α to tertiary amines using [Cr(CO)₃] activation of benzylic C sp 3-H bonds is described. The stabilized organolithium undergoes Pd-catalyzed coupling with ary

Full text of DOI:10.1002/anie.201201874

.
Angewandte
Communications
DOI: 10.1002/anie.201201874
Synthetic Methods
Asymmetric Cross-Coupling of Aryl Triflates to the Benzylic Position
of Benzylamines**
Genette I. McGrew, Corneliu Stanciu, Jiadi Zhang, Patrick J. Carroll, Spencer D. Dreher,* and
Patrick J. Walsh*
The development of new transition-metal-catalyzed cross-
coupling methods has been a focus of intense research.[1]
More recently, direct arylations are emerging as a more
the palladium-catalyzed cycle, arene exchange between F and
free arene G liberates the product and regenerates D. To
simplify the extremely challenging development of such
a dual catalyst system, we separated it into three phases. In
the first, we focused solely on the palladium-catalyzed left-
hand cycle by selecting an arene-activating moiety, {Cr(CO)3},
which would not undergo arene exchange (i.e., F does not
react with G). This enabled the initial proof-of-concept cross-
coupling, which is summarized in Scheme 2.[3] In the current
efficient method to C C bond formations.[2] Certain types of
ꢀ
ꢀ
C H bonds, however, have proven difficult to arylate, and
represent a particular challenge for asymmetric processes. We
envisioned a novel and potentially very powerful dual catalyst
cycle for the enantioselective functionalization of very weakly
ꢀ
acidic benzylic C H groups (Scheme 1). The left-hand cycle
Scheme 2. Tricarbonylchromium-assisted synthesis of achiral and race-
mic polyarylmethanes.
phase of this project, we introduce the palladium-catalyzed
enantioselective version of the reaction in Scheme 2, a reac-
tion which proceeds by an unusual dynamic kinetic resolution
(DKR). Future work (Phase 3) will focus on closing the right-
hand cycle.[4]
Scheme 1. Dual catalyst cycle for the asymmetric benzylic functionali-
zation process.
Development of an enantioselective version of the
coupling reaction to afford compounds like 2 in Scheme 2
was perceived to be particularly challenging because of some
unique features of our proposed catalytic cycle. First, the
diarylmethane-based products in Scheme 2 are more acidic
than the starting materials, and might be expected to lead to
racemization of the enantioenriched products. Second, our
proposed mechanism, described in more detail below
(Scheme 3), involves achieving enantioselectivity through
a novel DKR.[5] This mechanism requires that one of the
reversibly formed lithiated planar-chiral Cr adducts, either 1-
Li or 1-Li’, undergoes transmetallation with the enantioen-
riched palladium catalyst much faster than the other.
Together, these characteristics require identification of
a chiral ligand/metal complex which is exquisitely tuned to
promote the chemistry under mild reaction conditions.
Mechanistically, initial deprotonation of [(h6-benzylami-
ne)Cr(CO)3] by LiN(SiMe3)2 was anticipated to be rapid
based on the chemistry in Scheme 2.[3a] Because of the ability
of {Cr(CO)3} to delocalize negative charge,[6] the lithiated
intermediates 1-Li and 1-Li’ (Scheme 3) were expected to be
planar chiral and configurationally stable owing to the partial
double bond character between the ipso and benzylic carbon
atoms.[7] We hypothesized that rapid and reversible deproto-
involves the palladium-catalyzed cross-coupling through
oxidative addition to Pd0, transmetallation with M’ of the
organometallic intermediate E to generate C and reductive
elimination to close the cycle. The right-hand cycle is initiated
with reversible deprotonation of the activated benzylic CH
groups of the h6-arene complex to generate E. Emerging from
[*] G. I. McGrew, Dr. C. Stanciu, J. Zhang, Dr. P. J. Carroll,
Prof. P. J. Walsh
UPenn/Merck High Throughput Experimentation Laboratories
University of Pennsylvania Department of Chemistry
231 South 34th St. Philadelphia, PA 19104 (USA)
E-mail: pwalsh@sas.upenn.edu
Homepage: http://titanium.chem.upenn.edu/walsh/
Dr. S. D. Dreher
Department of Process Research, Merck and Co. Inc.
126 East Lincoln Avenue, Rahway, NJ 07065 (USA)
[**] We thank the NSF for a GOALI Award to UPenn and Merck (CHE-
0848460). P.J.W. is grateful to the NSF (CHE-1152488) for partial
support of this work.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201201874.
11510
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11510 –11513

Angewandte
Chemie
run in parallel in 250 mL HPLC vials on a 2 mmol scale (0.5 mg
substrate). Although several ligands showed high conversion
to product 2a (by HPLC), only certain ferrocenyl phosphine
ligands in the Taniaphos and Mandyphos family exhibited
high ee values (ꢁ 80%). The Cy-Mandyphos-based (Cy = cy-
clohexyl) palladium catalyst exhibited the highest enantiose-
lectivity, although conversion was low relative to Taniaphos.
Interestingly, these selective catalysts are phosphine ligands
which contain a tertiary amine. We hypothesize that the
amino group is important in coordinating to planar-chiral 1-Li
or 1-Li’ during the transmetallation step (Scheme 3). A
similar hypothesis was made by Kumada and co-workers in
the asymmetric cross-coupling of sBuMgBr with aryl bro-
mides, although the reactions are quite different.[11]
Unfortunately, attempts to increase conversion while
maintaining high enantioselectivity proved difficult. Con-
ducting the reaction at higher concentrations or with addi-
tional equivalents of base accelerated product decomposition.
Lowering the temperature to increase enantioselectivity
resulted in very low conversions. Furthermore, the product
ee value eroded as the reaction progressed. To explore the
possibility of product racemization, the benzylmorpholino
complex 2b was recrystallized to 94% ee. It was then heated
at 408C with 4 equivalents of LiN(SiMe3)2 for 16 hours and
quenched with D2O. No detectable deuterium incorporation
(1H NMR) and negligible erosion of the ee value (< 1%)
were observed. Substrate 1b was then used for further
investigation.
Scheme 3. Proposed asymmetric arylation of tricarbonylchromium-sta-
bilized benzyllithiums showing the reversible deprotonation and inter-
conversion of lithiated enantiomers, enantiodetermining transmetalla-
tion to palladium, and reductive elimination to form 2.
nation of 1 by LiN(SiMe3)2 would provide a necessary
pathway to equilibrate the enantiomers 1-Li and 1-Li’. In
this proposed scenario, the enantiodetermining step would
then be transmetallation of the enantiomers 1-Li and 1-Li’
with the enantioenriched PdII center. Given the high nucleo-
philicity of these organolithiums,[8] transmetallation is
expected to be irreversible. At the same time, we were
concerned that the high reactivity of 1-Li and 1-Li’ would
render selective transmetallation of 1-Li over 1-Li’ difficult.
Considering the significant demands on the catalyst in this
unusual DKR, it is quite likely that the number of promising,
highly enantioselective ligands will be very small. Hedging
our bets, we first examined the privileged bis(phosphine) 2,2’-
bis(diphenylphosphino)-1,1’-binaphthyl (binap). Under the
successful reaction conditions presented in Scheme 2, very
low conversion and enantioselectivity were observed. Thus,
our strategy for ligand discovery relied on low-barrier high-
throughput experimentation (HTE), in which a large library
of enantioenriched ligands could be screened at once on the
microscale.[9]
We next examined the effects of solvent composition and
additives. Use of toluene as a cosolvent (40%) led to an
increase in enantioselectivity (up to 81% ee) and reduced
product decomposition, although yields of the isolated
product remained below 50%.
Amines and LiX salts are known to influence aggregation
and reactivity of organolithium species,[12] and were examined
next (Table 1, entries 3–8). Additionally, we hypothesized
that amines could potentially influence the transmetallation
The asymmetric coupling of the N,N’-dimethylbenzyl
amine complex [(h6-Me2NCH2Ph)Cr(CO)3] 1a[10] with 4-
bromotoluene using LiN(SiMe3)2 was examined in an HTE
screen with a large collection (> 190, see the Supporting
Information for a complete list) of enantioenriched mono-
and bidentate phosphine ligands (Scheme 4). Reactions were
Table 1: Initial optimization of conditions in the asymmetric coupling
reaction of 1b with 4-bromotoluene.
Additives
(equiv)
Cosolvent
(%)[a]
T [8C]
t [h]
Conv. [%][b]
ee [%]
1
2
3
4
5
6
7
8
–
–
–
50
40
40
18
18
18
18
18
9
9
9
12
40
40
48
42
95
62
56
58
41
38
60
35[c]
55
81
82
75
91
86
77
90
tol (40)
tol (40)
tol (40)
tol (40)
tol (40)
tol (40)
tol (40)
iPr2NEt (2)
TEEDA (2)
TMEDA (2)
TMEDA (1)
PMDTA (1)
PMDTA (2)
Scheme 4. Initial screen of chiral ligands for the asymmetric cross-
coupling of 1a with 4-bromotoluene to yield 2a. cod=cyclo-1,5-
octadiene, THF=tetrahydrofuran.
[a] Reactions were run at 0.08–0.1m [1b]. [b] The conversion was
determined by SFC. [c] Yield of isolated product. TEEDA=N,N,N’,N’-
tetraethylethylenediamine
Angew. Chem. Int. Ed. 2012, 51, 11510 –11513
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
11511

.
Angewandte
Communications
Table 3: Scope of the cross-coupling of tertiary benzylamine and R’X
and, therefore, enantioselectivity, by binding to organolithium
intermediates. Indeed, polydentate tertiary amines acceler-
ated the coupling, thus allowing reactions to proceed at 188C
(entries 4–8). Furthermore, 2 equivalents of N,N,N’,N’’,N’’-
pentamethyldiethylenetriamine (PMDTA) or N,N,N’,N’-tet-
ramethylethylenediamine (TMEDA) led to an increase in
enantioselectivities up to 91% (entries 5 and 8). Lithium
chloride, bromide, and iodide additives[13] had a detrimental
effect on enantioselectivity, presumably impacting the trans-
metallation (see the Supporting Information for details).
Thus, generation of a LiBr by-product during the reaction
with aryl bromides may be responsible for the erosion of the
ee value with increasing conversion as outlined earlier for
reactions using 1a in the absence of TMEDA or PMDTA. We
have previously demonstrated that the negative impacts of
LiCl on enantioselective reactions can be counteracted by
sequestering the salt with polydentate amines.[14]
using optimized reaction conditions.[a]
NR2
R’X
Product T [8C][b] t [h] Yield ee
[%][c] [%]
1
2b
2c
2d
RT
8
76
74
72
92
90
87
2[c]
3[c]
0!RT 24
0!RT 18
4[c]
2e
0!RT 20
0!RT 16
72
90
To circumvent the detrimental effects of LiBr altogether,
other ArX derivatives were examined (Table 2, entries 1–3).
4-Tolyl triflate resulted in much faster and cleaner formation
5
2 f
80
59
89
84
6[d]
2g
RT
60
7[d]
2h
RT
48
36
62
Table 2: Effect of electrophile leaving group in the asymmetric coupling
reaction of 4-Tol-X with 1b to yield 2b.
8
2i
2j
2k
4
0!RT 24
8–10 60
66
72
69
80
87
85
88
53
X
Additives (equiv)[a]
T [8C] t [h] Yield [%][b] ee [%]
9[d]
10
11
1
2
3
4
5
6
Cl
TMEDA (2)
35
50
18
18
24
24
7
–
–
–
–
OTs PMDTA (1)
OTf PMDTA (2)
OTf PMDTA (1)
OTf PMDTA (1)
80
80
73
18[c]
23[c]
76
81
87
89
79
63
92
0!RT 24
0!RT 18
12
5–10 48
OTf
–
–
18
18
12
12
8
7[d] OTf
[a] Reaction conditions: 1 (1 equiv; 0.08–0.1m [1b]), LiN(SiMe3)2
8
OTf PMDTA (1), PhCl (2%) 18
(4 equiv), R’X (1.5–2 equiv), [{PdCl(allyl)}2] (2–2.5 mol%), Cy-Mandy-
phos (12 mol%), tol/THF (40:60), PhCl (2 vol%), PMDTA (1 equiv).
[b] RT is typically 18–198C. [c] Reported yields and ee values are the
average for two runs. [d] For reactions using ArBr or benzyl piperidine,
PhCl was not used. TIPS=triisopropylsilyl.
[a] Unless otherwise stated, reactions were run at 0.08–0.1m. [1b] in
40:60 toluene/THF as the solvent. [b] Yields of isolated products unless
otherwise specified. [c] The conversion was determined by SFC.
[d] Reaction in THF without toluene. Tf=trifluoromethanesulfonyl.
of the product with only a slightly lower ee value at room
temperature (entry 3). By decreasing the amount of PMDTA
(entries 4–7) and employing 2% chlorobenzene as a solvent
additive, an increase in the product ee value was observed: up
to 92% with a much-improved yield (76%; entry 8). Note
that chlorobenzene is unreactive under these reaction con-
ditions, and its role in the reaction is currently under
investigation.
Using the reaction conditions as given for entry 8 in
Table 2, the scope of the asymmetric coupling reaction was
next examined. Electron-rich aryl triflates were good sub-
strates (Table 3, entries 1–6). Unfortunately, electron-poor
substrates were less compatible with these reaction condi-
tions. For example, 3-CF3C6H4OTf triflate formed benzyne
under the basic reactions conditions (see the Supporting
Information) and 3-pyridyl triflate gave no product. Aryl
triflates with ortho-methyl or ortho-methoxy substituents
exhibited less than 5% conversion. In contrast, the ee value
was relatively insensitive to the nature of amine substitution.
Reactions using piperazino, piperidino, and N,N-dimethyl-
amino derivatives (entries 8–10) behaved similarly to the
morpholino substrate (entry 5). Cyclohexenyl triflate
(entry 11) exhibited good reactivity but moderate enantiose-
lectivity (53%). Given that transmetallation to [L*PdR-
(OTf)] is enantiodetermining, it is not surprising that chang-
ing from R = Ph to R = cyclohexenyl has a significant impact
on the product ee value.
Enantioenriched chromium-coordinated diarylmethyl-
amines did not suffer a significant decrease in ee value upon
demetallation. A solution of the purified complex may be
exposed to sunlight and air and then filtered to obtain
diarylmethylamine product. Alternatively, the concentrated
crude arylation reaction mixture can be directly decomplexed
and then purified by chromatography. The crude reaction
mixtures for 2b–e furnished the chromium-free diarylmethyl-
amines 5b–e in 63–91% yields upon isolation with 85–91% ee
(Scheme 5).
The absolute configuration of 2b using the (+)-enantio-
mer of Cy-Mandyphos was obtained by single-crystal X-ray
diffraction and found to be R (see the Supporting Informa-
tion).[15]
11512 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11510 –11513

Angewandte
Chemie
[3] a) G. I. McGrew, J. Temaismithi, P. J. Carroll, P. J.
Walsh, Angew. Chem. 2010, 122, 5673 – 5676; Angew.
Chem. Int. Ed. 2010, 49, 5541 – 5544; b) Previous
attemps at benzylic coupling of [(h6-arene)Cr(CO)3]
complexes used irreversible generation of benzylic zinc
species and provided very low yields of desired prod-
ucts. See: V. N. Kalinin, I. A. Cherepanov, S. K. Moi-
seev, J. Organomet. Chem. 1997, 536 – 537, 437 – 455;
c) J. Zhang, C. Stanciu, B. Wang, M. M. Hussain, C.-S.
Da, P. J. Carroll, S. D. Dreher, P. J. Walsh, J. Am. Chem.
Soc. 2011, 133, 20552 – 20560.
Scheme 5. One-pot arylation/decomplexation allows access to enantioenriched
diarylmethylamines.
[4] For a related approach to the synthesis of racemic
triarylmethanes, see:J. Zhang, A. Bellomo, A. D.
Creamer, S. D. Dreher, P. J. Walsh, J. Am. Chem. Soc.
2012, 134, 13765 – 13772.
[5] P. J. Walsh, M. C. Kozlowski, Fundamentals of Asym-
metric Catalysis, University Science Books, Sausalito, 2008.
[6] G. Jaouen, S. Top, M. J. McGlinchey, J. Organomet. Chem. 1980,
195, C9 – C12.
[7] a) C. A. Merlic, J. C. Walsh, D. J. Tantillo, K. N. Houk, J. Am.
Chem. Soc. 1999, 121, 3596 – 3606; b) A. Pfletschinger, T. K.
Dargel, J. W. Bats, H.-G. Schmalz, W. Koch, Chem. Eur. J. 1999,
5, 537 – 545; c) C. A. Merlic, M. M. Miller, B. N. Hietbrink, K. N.
Houk, J. Am. Chem. Soc. 2001, 123, 4904 – 4918; d) C. A. Merlic,
B. N. Hietbrink, K. N. Houk, J. Org. Chem. 2001, 66, 6738 – 6744.
[8] M. Uemura in Org. React., Vol. 67 (Ed.: L. E. Overman), Wiley,
Hoboken, 2006, pp. 217 – 659.
In summary, a novel dynamic kinetic resolution involving
palladium-catalyzed cross-coupling of aryl triflates with [(h6-
benzylamine)Cr(CO)3] has been presented. Diastereoselec-
tive transmetallation of one enantiomer of a rapidly equili-
brating planar-chiral secondary benzyllithium species is likely
the enantioselectivity-determining step in this process.
It is noteworthy that enantioenriched diarylmethylamines
are core structures present in several current pharmaceut-
icals[16] and our approach to them represents a novel dis-
connection. Although this chemistry does not yet constitute
a practical synthesis of these valuable compounds, this
strategy and the conceptual understanding gained herein is
critically important for future development of the dual
catalyst enantioselective functionalization of weakly acidic
[9] S. D. Dreher, P. G. Dormer, D. L. Sandrock, G. A. Molander, J.
Am. Chem. Soc. 2008, 130, 9257 – 9259.
[10] J. P. Djukic, A. Maisse, M. Pfeffer, A. deCian, J. Fischer,
Organometallics 1997, 16, 657 – 667.
ꢀ
[11] T. Hayashi, M. Fukushima, M. Konishi, M. Kumada, Tetrahedron
Lett. 1980, 21, 79 – 82.
C H bonds.
[12] V. H. Gessner, C. Dꢀschlein, C. Strohmann, Chem. Eur. J. 2009,
15, 3320 – 3334.
Received: March 8, 2012
Revised: August 11, 2012
[13] a) E. Hevia, R. E. Mulvey, Angew. Chem. 2011, 123, 6576 – 6578;
Angew. Chem. Int. Ed. 2011, 50, 6448 – 6450; b) Y. Ma, A. C.
Hoepker, L. Gupta, M. F. Faggin, D. B. Collum, J. Am. Chem.
Soc. 2010, 132, 15610 – 15623; c) A. Krasovskiy, V. Krasovskaya,
P. Knochel, Angew. Chem. 2006, 118, 3024 – 3027; Angew. Chem.
Int. Ed. 2006, 45, 2958 – 2961.
[14] a) J. G. Kim, P. J. Walsh, Angew. Chem. 2006, 118, 4281 – 4284;
Angew. Chem. Int. Ed. 2006, 45, 4175 – 4178; b) L. Salvi, S.-J.
Jeon, E. L. Fisher, P. J. Carroll, P. J. Walsh, J. Am. Chem. Soc.
2007, 129, 16119 – 16125; c) L. Salvi, J. G. Kim, P. J. Walsh, J. Am.
Chem. Soc. 2009, 131, 12483 – 12493.
[15] CCDC 869404 (2b) contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif.
[16] J. J. Li, D. S. Johnson, D. R. Sliskovic, B. Roth, Contemporary
Drug Synthesis, Wiley-Interscience, New York, 2004.
Published online: October 4, 2012
ꢀ
Keywords: asymmetric catalysis · C C coupling · chromium ·
.
palladium · synthetic methods
[1] a) R. Jana, T. P. Pathak, M. S. Sigman, Chem. Rev. 2011, 111,
1417 – 1492; b) C. S. Yeung, V. M. Dong, Chem. Rev. 2011, 111,
1215 – 1292; c) B. M. Rosen, K. W. Quasdorf, D. A. Wilson, N.
Zhang, A.-M. Resmerita, N. K. Garg, V. Percec, Chem. Rev.
2011, 111, 1346 – 1416; d) C. M. So, F. Y. Kwong, Chem. Soc. Rev.
2011, 40, 4963 – 4972.
[2] a) O. Daugulis, H.-Q. Do, D. Shabashov, Acc. Chem. Res. 2009,
42, 1074 – 1086; b) A. A. Kulkarni, O. Daugulis, Synthesis 2009,
4087 – 4109; c) L. Ackermann, Chem. Commun. 2010, 46, 4866 –
4877; d) T. W. Lyons, M. S. Sanford, Chem. Rev. 2010, 110, 1147 –
1169; e) I. A. I. Mkhalid, J. H. Barnard, T. B. Marder, J. M.
Murphy, J. F. Hartwig, Chem. Rev. 2010, 110, 890.
Angew. Chem. Int. Ed. 2012, 51, 11510 –11513
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11513