Pd-PEPPSI-IPentCl: A highly effective catalyst for the selective cross-coupling of secondary organozinc reagents

Article abstract of DOI:10.1002/anie.201205747

No migration? No problem. A series of new N-heterocyclic carbene based Pd complexes has been created and evaluated in the Negishi cross-coupling of aryl and heteroaryl chlorides, bromides, and triflates with a variety of secondary alkylzinc reagents (see scheme). The direct elimination product is nearly exclusively formed; in most examples there is no migratory insertion at all. Copyright

Full text of DOI:10.1002/anie.201205747

.
Angewandte
Communications
DOI: 10.1002/anie.201205747
Homogeneous Catalysis
Pd-PEPPSI-IPentCl: A Highly Effective Catalyst for the Selective
Cross-Coupling of Secondary Organozinc Reagents**
Matthew Pompeo, Robert D. J. Froese, Niloufar Hadei, and Michael G. Organ*
Molecules with architectural complexity, especially those
possessing high alkyl content, have been avoided in the
pharmaceutical sector owing to the challenges associated with
their synthesis.[1] Consequently, medicinal chemistry efforts
have focused primarily on developing drug candidates that
are relatively flat (e.g., biaryls), and can be prepared readily
using cross-coupling methodology.[1] However, such com-
pounds are prone to promiscuous protein binding and side
effects.[2] Consequently, there is new interest in the pursuit of
molecules with increased alkyl composition resulting in more
three-dimensional topology to improve target specificity.[2]
The cross-coupling of alkyl nucleophiles can figure prom-
inently in this approach and the mechanism is detailed in
Scheme 1.[3,4]
primary organometallic compounds (e.g., Scheme 1, R1 = H)
reinsertion into the olefin favors reformation of the primary
metal-alkyl (i.e., 4) for steric and electronic reasons. The same
would be true if the starting organometallic compound were
secondary (e.g., R1 = CH3, R2 = H), however this migratory
insertion (MI) leads to the undesired isomer (8) rather than
the expected one (5) after RE. To avoid formation of this
undesired isomer, the barrier to RE, relative to BHE, must be
lower (see below). We envisioned that this could be accom-
plished by designing a catalyst that increases the steric bulk
around the Pd center while, at the same time, reduces the
electron density on it.[4a] This increased ligand bulk would
increase the strain in the TM intermediate (e.g., 4) and favor
RE, which would relieve the strain, and a more easily reduced
metal would drive the same step electronically. To examine
this, we have created a series of new Pd-PEPPSI precatalysts,
including Pd-PEPPSI-IPrCl (12), Pd-PEPPSI-IPrMe (13), Pd-
PEPPSI-IPrQuino (14), and Pd-PEPPSI-IPentCl (16), and
evaluated them in alkyl cross-coupling (Table 1).
To establish a baseline of inherent selectivity by minimiz-
ing substrate interactions, para-substituted derivatives were
first coupled (Table 1; entries 1–10). These results indeed
demonstrate that ligand sterics are important as the catalyst
featuring IPent[5] (15) showed much higher selectivity than
that with IPr[6] (11). There is a general trend that electron-
withdrawing groups (EWGs) on the OA partner improve
selectivity for the branched (normal) product (e.g., entry 1 vs.
6, and 2 vs. 7). When the substituent on 9 was positioned
closer to Br, and thus Pd in intermediate 4, immediate effects
on selectivity were observed. When Pd-PEPPSI-IPr (11) was
used, a meta group tipped the balance towards BHE, as the
linear product (8) was favored (e.g., entry 11). When there
was a substituent at the ortho position, the rearranged product
(8) now dominated with Pd-PEPPSI-IPr (11; entry 21), thus
indicating that the rate of RE, relative to BHE, has been
profoundly impacted. However, the reaction catalyzed by Pd-
PEPPSI-IPent (15) shows that ligand bulk helps to counteract
the negative effects of substrate hindrance as the branched
product remained dominant (e.g., entries 15 and 25), so
selectivity is primarily under catalyst control with 15.
When substituents were placed on the backbone of the IPr
NHC the preference for BHE versus RE, seen with 11 was
completely reversed (see catalysts 12, 13, and 14) for meta-
substituted coupling partners. For example, for catalysts 12,
13, and 14 (entries 12–14), the selectivity for the branched
product 5 was a sharp improvement on that seen for catalyst
11 (entry 11). An even more dramatic change was observed in
the selectivity obtained with ortho-substituted aryl halides,
e.g., entry 21 versus entries 22–24. When catalyst 16 having
a ligand with the bulkier N-phenyl NHC substituent and
Scheme 1. Cross-coupling mechanism with alkylzinc nucleophiles.
Problematically, the Pdll center in transmetallation (TM)
intermediate 4 can readily undergo b-hydride elimination
(BHE) resulting in the formation of olefin by-products. Also,
an electron-rich oxidative addition (OA) partner will slow
reductive elimination (RE), leading to more BHE. Energeti-
cally, the metal-alkyl complex (4 or 7) is favored over the
olefin-coordinated metal hydride (6; see below). With
[*] M. Pompeo, Dr. N. Hadei, Prof. M. G. Organ
Department of Chemistry, York University
4700 Keele Street, Toronto, ON, M3J 1P3 (Canada)
E-mail: organ@yorku.ca
Homepage: http://www.yorku.ca/organ/
Dr. R. D. J. Froese
The Dow Chemical Company, Midland, MI 48674 (USA)
[**] This work was supported by NSERC (Canada) and the Ontario
Research Fund (ORF, Ontario). PEPPSI=pyridine-enhanced pre-
catalyst preparation, stabilization, and initiation; IPentCl refers to an
N-heterocyclic carbene ligand.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201205747.
11354
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11354 –11357

Angewandte
Chemie
performed.[7] Figure 1 depicts the potential energy surface
(PES) for the selectivity associated with formation of the
Table 1: Control study coupling aryl bromides (9) to isopropylzinc
bromide (10) using Pd-PEPPSI precatalysts 11–16.[a,b]
normal (5) and rearranged products (8) for catalysts 11 and
Entry
Cat.
5/8
Prod.
Yield [%]
DDE°[c]
ꢀ
Ar Br
1
2
3
4
5
6
7
8
4-CO2CH3
4-CO2CH3
4-CO2CH3
4-CHO
4-CHO
4-OCH3
4-OCH3
4-OCH3
4-CN
4-CN
3-CN
3-CN
3-CN
3-CN
3-CN
3-CN
3-CHO
3-CHO
3-OCH3
3-OCH3
2-CN
11
15
16
15
16
11
15
16
15
16
11
12
13
14
15
16
15
16
15
16
11
12
13
14
15
16
15
16
5.0:1
40:1
56:1
39:1
59:1
2.5:1
33:1
46:1
27:1
59:1
1:1.4
15:1
15:1
13:1
11:1
61:1
22:1
48:1
34:1
39:1
1:6.6
4.3:1
5.2:1
8.5:1
2.4:1
28:1
1.9:1
23:1
a
a
a
b
b
c
c
c
d
d
e
e
e
e
e
e
f
99
1.3
98[d]
94
78[d]
92
89[d]
95[d]
79
1.2
9
92[d]
92
Figure 1. The DFT potential energy surface for RE versus BHE for the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
82[d]
81
ꢀ0.9
2.0
2.5
ꢀ
coupling of Ph X and isopropylzinc halide with precatalysts 11 (c)
and 12 (g). L=NHC ligand.
79
78
1.6
66[d]
81
12.[8] Figure 1 illustrates that from the intermediate species
(4), which was formed from an isopropyl zinc reagent, RE can
lead directly to the normal product (5). This pathway
competes with BHE to form an olefin-hydrido complex
(6a), which through reinsertion of the olefin into the hydride
with opposite regiochemistry (MI) leads to linear alkyl 7 that
can reductively eliminate to form isomer 8.
71[d]
84
f
g
g
h
h
h
h
h
h
i
57[d]
66
38
72
66
ꢀ1.8
1.0
1.6
2-CN
2-CN
2-CN
2-CN
2-CN
2-OCH3
2-OCH3
On looking at the shape of the PES, it can be seen that the
selectivity between normal (5) and rearranged (8) products is
decided at transition states (TSs) TS4!5 and TS4!6a,
directly leading from the intermediate 4. If BHE and
reinsertion occurs to form the linear intermediate (7), RE
of the linear chain (TS7!8) has a lower barrier and is
significantly faster than BHE. This feature on the PES differs
from the cationic PdII olefin polymerization process, wherein
BHE from primary or secondary alkyls is much faster than
70
1.1
80[d]
76
46[d]
54
i
[a] Ratio determined by 1H NMR spectroscopy after purification.
[b] Yields of the isolated mixtures of iPr and nPr products, averaged over
two runs. [c] DDE° is the DFTselectivity (see text). [d] These entries have
been reported previously, see Ref. [4g].
olefin insertion into the Pd alkyl bond.[9] This polymerization
ꢀ
surface leads to a Curtin–Hammett regime where the
analogous Pd-alkyl complexes 4 and 7 are in equilibrium
and the selectivity is defined by the relative height of the two
TSs for insertion into the respective metal alkyls.[9] In the case
of the coupling reactions in this work, TS4!6a was always
found to be slightly higher than TS6b!7 whereas the RE of
the primary alkyl group (TS7!8) always has a substantially
lower energy TS. Thus, the relative heights of TS4!5
(associated rate constant k5 leading to product 5) and TS4!
6a (k8 leading to product 8) define the selectivity for this
transformation.
chlorines on the NHC backbone (IPentCl) was used, selectiv-
ity for the non-rearranged ꢀnormalꢁ product was now near
optimal for the meta-substituted aryl halides (entry 16) and
still excellent (approximately 28:1) for the ortho case
(entry 26). When we examined the scope of this transforma-
tion using Pd-PEPPSI-IPentCl (16) to prepare compounds of
biological and medicinal interest, it became clear that this
catalyst shows very high reactivity, broad functional group
tolerance, and most importantly, virtually exclusive selectivity
for the desired, non-rearranged product (Scheme 2).
Based on an Eyring expression,[10] the relative rate
constants (i.e., selectivity) for the formation of 5 versus 8
can be expressed as: ln{k5/k8} = DDE°/RT, where DDE° =
[11]
DE°8ꢀDE°
.
In the example shown in Figure 1 for Pd-
5
PEPPSI-IPr (11), the normal product is favored over the
rearranged product by 1.4 kcalmolꢀ1 (a positive DDE°
implies a selectivity for the normal product). Upon introduc-
tion of the chlorides onto the NHC backbone, as in Pd-
PEPPSI-IPrCl (12), this energy difference increases signifi-
cantly to 4.1 kcalmolꢀ1. To get a direct comparison to
In an effort to understand the selectivity demonstrated in
this study, density functional theory (DFT) calculations were
Angew. Chem. Int. Ed. 2012, 51, 11354 –11357
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11355

.
Angewandte
Communications
agreement between theory and experiment is excellent
(Figure 2a). In the examples shown in Table 1, entries 11
and 21, the rearranged product was favored and in these cases,
the BHE TS is lower in energy than the RE TS, and the
greatest experimental selectivities toward the normal product
(entries 12 and 13) corresponded to the largest computed
DDE° (2.0–2.5 kcalmolꢀ1).
Figure 2. a) A comparison of the computed difference in energy
(DDE°) between RE and BHE (ordinate) versus the log of the
experimental selectivity (abscissa) for catalysts 13–16, shown in
Table 1. Each number on the graph corresponds to a table entry. b) A
comparison of the computed DDE° value between RE and BHE
(ordinate) versus the average of the two computed C-N-aryl bond
angles (see arrows on structure) of the NHC (abscissa) for a set of
related IPr-based NHC catalysts for the reaction shown in Table 1 with
bromobenzene. Each group on the graph corresponds to the R groups
on the NHC backbone.
As the calculations accurately predicted the selectivity
with different catalysts and aryl groups, we examined other
catalysts, including virtual ones, based on the IPr NHC core[8]
(Figure 2b). From an electronic perspective the data are
interesting in that we expected the ligand with hydrogen
substituents on the NHC backbone (i.e., in catalyst 11) to
provide intermediate results, yet it is at the extreme. Com-
paratively higher DDE° values of 2.2 and 2.7 kcalmolꢀ1 were
obtained for the ligands with electron-donating CH3 and
electron-withdrawing CF3 groups, respectively. The most
compelling trend is that as the groups on the NHC backbone
become bulkier, the DDE° value gets larger with very high
selectivity predicted for ligands with tBu and TMS sustituents
(Figure 2b). It would seem there is a strong steric component
in that a larger group on the backbone prevents isomer-
ization. This fact can be rationalized if one considers that the
selectivity is a function of the difference between the barriers
for RE, having a three-center TS (e.g., TS4!5), and BHE,
having a four-center TS (e.g., TS4!6a). Thus, if larger groups
Scheme 2. Scope study for C(sp2)–C(sp3) couplings using Pd-PEPPSI-
IPentCl. The ratio was determined by 1H NMR spectroscopy after
purification; isomers were inseparable. The yields of the isolated
products include that of the rearranged isomer, where relevant,
averaged over two runs. A ratio of >99:1 means that minor isomers
are not observed at all. n=normal product, r=rearranged product.
experiment, the selectivities for Pd-PEPPSI-IPr (11), Pd-
PEPPSI-IPrCl (12), Pd-PEPPSI-IPrMe (13), and Pd-PEPPSI-
IPrQuino (14) were computed for a variety of aryl groups
coupling to iPrZnX (Table 1, entries 1, 6, 11–14, 21–24);
11356 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11354 –11357

Angewandte
Chemie
[4] For cross-coupling with secondary alkyl nucleophiles, see: for
are on the NHC backbone, they act to push the aryl groups
towards the active site, thus destabilizing the four-center TS of
BHE to a greater extent than the three-center TS of RE, and
therefore increasing selectivity towards the non-rearranged
product.
Mg: a) K. Tamao, Y. Kiso, K. Sumitani, M. Kumada, J. Am.
Chem. Soc. 1972, 94, 1268 – 9263; for Si see: b) T. Hayashi, M.
Konishi, Y. Kobori, M. Kumada, T. Higuchi, K. Hirotsy, J. Am.
Chem. Soc. 1984, 106, 158 – 163; for B, see: c) S. D. Dreher, P. G.
Dormer, D. L. Sandrock, G. A. Molander, J. Am. Chem. Soc.
2008, 130, 9257 – 9259; d) A. van den Hoogenband, J. H. M.
Lange, J. W. Terpstra, M. Koch, G. M. Visser, M. Visser, T. J.
Korstanje, J. T. B. H. Jastrzebski, Tetrahedron Lett. 2008, 49,
4122 – 4124; for Zn, see: e) C. Han, S. L. Buchwald, J. Am.
Chem. Soc. 2009, 131, 7532 – 7533; f) T. Thaler, B. Haag, A.
Gavryushin, K. Schober, E. Hartmann, R. M. Gschwind, H.
Zipse, P. Knochel, Nat. Chem. 2010, 2, 125 – 130; g) S. C¸ alimsiz,
M. G. Organ, Chem. Commun. 2011, 47, 5181 – 5183; for in situ
Zn, see: h) C. Duplais, A. Krasovskiy, B. H. Lipshutz, Organo-
metallics 2011, 30, 6090 – 6097.
A series of Pd-PEPPSI complexes have been created with
various substitution patterns on the NHC backbone that have
demonstrated extraordinary selectivity in the Negishi cross-
coupling of secondary zinc nucleophiles. Whether the sub-
stituents render the metal center more electron rich (e.g., 13)
or poor (e.g., 12, 14), selectivity for the normal, non-
rearranged product is favored. Computational studies reveal
that the relative barrier difference between RE and BHE
correlate very well with observed selectivities. Further, it
would appear that the effect imparted by the NHC substitu-
ents is primarily steric in origin.[12] This analysis has led to the
creation of Pd-PEPPSI-IPentCl (16) that has shown unprece-
dented selectivity, leading to virtually one single (desired)
isomer for reactions of a wide selection of alkylzincs and
highly functionalized (hetero)aromatic halides.
[5] For a review of couplings with 15, see: C. Valente, S. C¸ alimsiz,
K. H. Hoi, D. Mallik, M. Sayah, M. G. Organ, Angew. Chem.
2012, 124, 3370 – 3388; Angew. Chem. Int. Ed. 2012, 51, 3314 –
3332.
[6] C. J. OꢁBrien, E. A. B. Kantchev, N. Hadei, C. Valente, G. A.
Chass, J. C. Nasielski, A. Lough, A. C. Hopkinson, M. G. Organ,
Chem. Eur. J. 2006, 12, 4743 – 4748.
[7] Calculations were performed at the B3LYP level with
LANL2TZ(F) on Pd and 6-31G* on the remaining atoms.
Further details are provided in the Supporting Information.
[8] In this work, IPent catalysts were not studied owing to the
complicated conformations of the eight ethyl groups of the N-
aryl moiety.
[9] a) L. H. Shultz, D. J. Tempel, M. Brookhart, J. Am. Chem. Soc.
2001, 123, 11539 – 11555; b) D. G. Musaev, M. Svensson, K.
Morokuma, S. Strçmberg, K. Zetterberg, P. E. M. Siegbahn,
Organometallics 1997, 16, 1933 – 1945.
Received: July 18, 2012
Revised: August 29, 2012
Published online: October 4, 2012
Keywords: homogeneous catalysis · Negishi reaction ·
.
palladium · PEPPSI
[1] F. Lovering, J. Bikker, C. Humblet, J. Med. Chem. 2009, 52,
6752 – 6756.
[2] a) S. Dandapani, L. A. Marcaurelle, Curr. Opin. Chem. Biol.
2010, 14, 362 – 370; b) A. Ganesan, Curr. Opin. Chem. Biol. 2008,
12, 306 – 317.
[10] H. Eyring, J. Chem. Phys. 1935, 3, 107 – 115.
[11] See the Supporting Information for derivation of relative rate
constants.
[12] Tolman electronic parameter (TEP) analysis failed to yield
consistent trends between selectivity and electronic ligand
effects. For complete TEP analysis, see the Supporting Informa-
tion.
[3] For a recent comprehensive review, see: R. Jana, T. P. Pathak,
M. S. Sigman, Chem. Rev. 2011, 111, 1417 – 1492.
Angew. Chem. Int. Ed. 2012, 51, 11354 –11357
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11357