Multimetallic Ni- and Pd-Catalyzed Cross-Electrophile Coupling to Form Highly Substituted 1,3-Dienes

Article abstract of DOI:10.1021/jacs.7b13601

The synthesis of highly substituted 1,3-dienes from the coupling of vinyl bromides with vinyl triflates is reported for the first time. The coupling is catalyzed by a combination of (5,5′-bis(trifluoromethyl)-2,2′-bipyridine)NiBr₂ and (1,3-bis(diphenylphosphino)propane)PdCl₂ in the presence of a zinc reductant. This method affords tetra- and penta-substituted 1,3-dienes that would otherwise be difficult to access and tolerates electron-rich and -poor substituents, heterocycles, an aryl bromide, and a pinacol boronate ester. Mechanistically, the reaction appears to proceed by an unusual zinc-mediated transfer of a vinyl group between the nickel and palladium centers.

Full text of DOI:10.1021/jacs.7b13601

Communication
pubs.acs.org/JACS
Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
Multimetallic Ni- and Pd-Catalyzed Cross-Electrophile Coupling To
Form Highly Substituted 1,3-Dienes
Astrid M. Olivares† and Daniel J. Weix*,†,‡
†University of Rochester, Rochester, New York 14627-0216, United States
‡University of Wisconsin−Madison, Madison, Wisconsin 53706, United States
S
* Supporting Information
Scheme 1. Cross-Electrophile 1,3-Diene Synthesis
ABSTRACT: The synthesis of highly substituted 1,3-
dienes from the coupling of vinyl bromides with vinyl
triflates is reported for the first time. The coupling is
catalyzed by a combination of (5,5′-bis(trifluoromethyl)-
2,2′-bipyridine)NiBr2 and (1,3-bis(diphenylphosphino)-
propane)PdCl2 in the presence of a zinc reductant. This
method affords tetra- and penta-substituted 1,3-dienes that
would otherwise be difficult to access and tolerates
electron-rich and -poor substituents, heterocycles, an aryl
bromide, and a pinacol boronate ester. Mechanistically, the
reaction appears to proceed by an unusual zinc-mediated
transfer of a vinyl group between the nickel and palladium
centers.
he 1,3-diene is an important synthetic target because it
Tparticipates in an increasing variety of useful trans-
formations, including hydrofunctionalizations, ring-forming
reactions, and polymerizations. The diene also occurs in many
natural products with interesting biological activity.1 A natural
disconnection for the 1,3-diene motif is the formation of the
central single bond. For linear and less-substituted branched
dienes, Mizoroki-Heck,2 ene-yne metathesis,3 C−H vinylation,4
and cross-coupling approaches5 have been successful. For highly
substituted and branched dienes with multiple rings, the most-
used approach is cross-coupling of vinylmetal reagents with
vinyl halides.1,6 Though these approaches have proven useful,
relatively few highly substituted dienes have been synthesized
and often require synthesis of an organometallic reagent.
An alternative strategy to synthesize highly substituted 1,3-
dienes would be the cross-electrophile coupling of two cyclic
vinyl electrophiles, but no general approach of this type has been
reported. We recently reported a cross-Ullman strategy toward
biaryls that relied on complementary reactivity of nickel7 and
palladium toward aryl bromides and aryl triflates, but the
selectivity was poor for vinyl bromides with vinyl triflates and
the scope was limited (Scheme 1A).8,9 We hypothesized that
this was due to the fact that the nickel catalyst was poorly
selective for vinyl bromides over vinyl triflates,10 but it was not
clear if the nickel and palladium catalyst system could be applied
to these more reactive substrates. Furthermore, the mechanism
of the reaction was poorly understood, limiting development.
Overcoming these challenges would be useful in synthesis: the
stoichiometric version of this coupling was recently used at a late
stage in an approach to Batrachotoxin.11 Herein, we report a
new nickel catalyst that allows for the selective cross-coupling of
vinyl bromides with vinyl triflates without an excess of either
coupling partner (Scheme 1B) as well as insight into the dual
role of zinc in the process.
We began our studies by testing the coupling of vinyl bromide
1a with vinyl triflate 2a under catalysis by nickel and palladium
with a variety of ligands and examined the distribution of
products between the cross-coupled product 3a and the
homocoupled vinyl bromide 5 and the homocoupled vinyl
triflate 4. These studies, summarized in Table 1, showed that the
nitrogen ligand has a large influence on selectivity whereas the
phosphorus ligand does not (see Supporting Information, Chart
S2).12 Tridentate and hindered bidentate ligands were selective
for homodimer formation (L10−L12), but a variety of bidentate
bipyridine and phenanthroline ligands provided promising
results. Among these ligands, bipyridine L6, containing
electron-withdrawing trifluoromethyl groups, provided the
best balance of selectivity and reactivity.13 Bipyrimidines L7
and L8 were selective, but reactions did not proceed to
completion. More electron-rich ligands (L3 and L4) were both
poorly selective and slow. These results are in stark contrast to
our results on biaryl synthesis,8 where the phosphine was crucial
and the nitrogen ligand was relatively unimportant, demonstrat-
ing the adaptability of this new system.
Received: December 22, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.7b13601
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
a
a
Table 1. Nitrogen Ligand Effect on Diene Synthesis
Table 2. Optimization of Reaction Conditions
b
entry
change in conditions
3a (%)
3a:4 (3a:5)
1
2
3
4
5
6
KF (1 equiv), 5 mol % catalysts
5 mol % catalysts
none
95
94
19 (32)
18 (31)
19 (32)
16 (19)
7.0 (8.4)
5.3 (4.2)
19 (32)
3.1 (3.5)
NA
c
95/89
94
84
79
97
59
0
40 °C
60 °C
80 °C
1.25 mol % catalysts
no (dppp)PdCl2
no L6
d
7
8
e
9
10
no dppp
83
0
7.5 (6.4)
NA
b
b
b
entry
1
L
3a (%)
4 (%)
5 (%)
3a:4 (3a:5)
e
11
no Ni
e
L1
70
37
28
18
53
82
62
26
9
31
9
16
16
0
2.3 (4.4)
4.1 (2.3)
1.8 (NA)
1.2 (NA)
2.0 (13)
5.9 (9.1)
6.9 (21)
4.3 (26)
4.5 (NA)
0.4 (NA)
0.6 (0.9)
0.6 (1.1)
12
no Zn
0
NA
c
2
L2
13
14
under N2 on bench
under air on bench
94
76
25 (47)
c
3
L3
16
15
26
14
9
5.2 (31)
b
c
a
4
L4
0
Reactions were run on 0.5 mmol scale in 2 mL of solvent. Yields are
corrected GC yields vs dodecane as internal standard. Balance of yield
was dimeric products unless otherwise noted. Isolated yield after
purification. Reaction complete after 24 h. Starting materials were
c
5
L5
5
d
c
6
L6
9
c
d
e
7
L7
3
c
not consumed.
8
L8
6
1
c
9
L9
2
0
c
10
L10
L11
L12
8
20
73
60
0
bromide (3k). Although substrates with little steric or electronic
differentiation provided lower selectivity (3p), even small
differences in sterics or electronics could allow for reasonable
selectivity (3x). Finally, the yield of 3v was the same on bench at
a 5.0 mmol scale as in the glovebox at a 1.0 mmol scale (see
Supporting Information).
Most of the dienes synthesized in Table 3 would be difficult or
impossible to form by Heck or metathesis approaches because of
their substitution patterns. The difficulty in accessing these types
of dienes is reflected by the fact that only one example (3u) in
Table 3 has been previously reported. Compared to other cross-
coupling approaches,6 the direct use of two different electro-
philes can save steps and improve functional-group compati-
bility. For example, dienes like 3x (Table 3) have been
previously synthesized by a Stille-coupling strategy,16 but our
approach avoids the need to synthesize a vinyltin reagent. These
dienes can participate in a variety of pericyclic reactions to form
steroid analogs (e.g., 3z in eq 1 and Supporting Information).
11
12
43
38
50
34
a
Reactions were run on 0.5 mmol scale in 2 mL of solvent. Byproduct
4 corresponds to the dimer of 2a; byproduct 5 corresponds to the
dimer of 1a. dppp = 1,3-bis(diphenylphosphino)propane. Yields are
corrected GC yields vs dodecane as internal standard. Both starting
b
c
d
materials remained after 24 h. Reaction complete in 3 h.
Examining the other reaction parameters also led to further
improvements in conditions (Table 2). Higher temperatures led
to lower selectivity (entries 3−6), but lower catalyst loading did
not alter the yield dramatically (entries 2, 3 and 7). Finally,
reactions set up on the bench under N2 (entry 13), or even
under air (entry 14), provided good yields of product.
Control reactions showed that the addition of KF did not
have an effect on yield or selectivity (entries 1−2). This result
again is in sharp contrast to our previous biaryl study,8 where
potassium fluoride was crucial to high selectivity, presumably by
improving the selectivity of the palladium catalyst for aryl triflate
over aryl bromide.14 This suggests different factors are crucial to
selectivity in this system. Though no product was formed in the
absence of L6, Ni, or Zn, significant product was formed in the
absence of dppp (entry 10) or even (dppp)PdCl2 (entry 8). The
lower selectivities in these reactions suggest that both nickel and
palladium are essential for high cross-product selectivity, but not
for product formation.
These conditions proved to be general for a variety of
cycloalkenyl as well as several acyclic alkenyl electrophiles (3m,
3o, 3t, 3w) (Table 3). Thiophene (3v) and benzothiophene
(3j) could also be selectively coupled with vinyl triflates, but
these conditions were unselective for the coupling of aryl
bromides with aryl triflates.8,15 A variety of functional groups
were tolerated, including protected nitrogen (3e), enone (3h,
3u), boronic acid ester (3l), nitrile (3s), and even an aryl
A key mechanistic question for this multimetallic reaction is
how vinyl groups are transferred between metal centers. Though
direct nickel−palladium transmetalation has been proposed,8
little evidence has been reported. In order to better understand
the roles of the two catalysts in these reactions, we decided to
study reactions catalyzed by nickel alone as well as reactions
catalyzed by a combination of nickel and palladium. Monitoring
reaction progress under both conditions provided an important
B
DOI: 10.1021/jacs.7b13601
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
a
a
Table 3. Scope of Cross-Coupling Reaction
Scheme 2. Evidence for a Vinylzinc Intermediate
a
See Supporting Information for additional information. The
intermediates were quantitated by GC, corrected.
concentration of 6 and 7 that built up during the first 30 min
of reactions was small, it exceeded the nickel loading ([7]/[Ni]
> 1) and decreased with time (Scheme 2C and additional
experiments in the Supporting Information). This suggests that
both 6 and 7 are derived from quenching of the same vinylzinc
intermediate and not from quenching of a vinylnickel
intermediate.18
These results are most consistent19 with zinc-mediated
transfer of the vinyl group of 1a from nickel to palladium20
via a transient vinylzinc intermediate21,22 and subsequent
palladium-catalyzed C−C bond formation.1 Though the use of
organozinc intermediates in cross-coupling is well-known, and
in situ organozinc synthesis and coupling has been reported,23,24
the zinc mediated transfer of groups between two different
catalysts has not been exploited in small molecule synthesis. In
polymer chemistry, however, this strategy has been used in chain
shuttling copolymerization, where a zinc-co-catalyst transfers
growing polymer chains between two different catalysts.25 This
important concept in polymer chemistry could also have wide
application in small molecule synthesis. Further mechanistic
studies are ongoing.
In conclusion, we have developed a general reductive
approach to highly substituted 1,3-dienes that tolerates a wide
array of functionality. This approach demonstrates that the
nickel and palladium system can be tuned to selectively couple
two highly reactive electrophiles. Furthermore, the zinc-
mediated transfer of organic moieties between two catalysts
has not been extensively explored and could provide a new
general mechanism for cross-coupling.
a
Reactions were run on 1.0 mmol scale in 4 mL of DMF. Yields of
b
isolated, purified material. See Supporting Information. BINAP was
used instead of dppp. Yield of 3e was adjusted to account for 8.6 wt %
of inseparable tert-butyl 3,6-dihydropyridine-1(2H)-carboxylate. Yield
c
d
of 3i was adjusted to account for 4.7 wt % of inseparable [1,1′-
e
f
bi(cyclohexane)]-1,1′diene. 1.5 equiv of vinyl bromide used. 2-
Chlorocyclohex-2-en-1-one was used.
observation: aliquots quenched with acid contained 1,2-
dihydronaphthalene (6, Scheme 2A) and aliquots quenched
with I2 contained 1-iodo-1,2-dihydronaphthalene (7, Scheme
2B). Both 6 and 7 could be formed from reaction of either a
17
vinylnickel or vinylzinc reagent. However, although the
C
DOI: 10.1021/jacs.7b13601
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
because further experiments demonstrated that L19 and L14 worked
equally well (see Supporting Information, Chart S4).
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.7b13601.
■
S
(13) Our ranking of the relative donicity of a subset of these ligands:
L6 < L7 < L1 < L4 < L3. We ranked the ligands using an adaptation of
Girolami’s DFT method: Flener Lovitt, C.; Frenking, G.; Girolami, G.
S. Organometallics 2012, 31, 4122 The absolute values and range of
values for the calculated νCO are very close to those calculated by
Girolami for dppe derivatives substituted with aryl and alkyl
substituents.
(14) (a) Littke, A. F.; Dai, C.; Fu, G. C. J. Am. Chem. Soc. 2000, 122,
4020. (b) Proutiere, F.; Schoenebeck, F. Angew. Chem., Int. Ed. 2011,
50, 8192.
Optimization data, mechanistic experiments, detailed
experimental procedures, and spectral data (PDF)
AUTHOR INFORMATION
Corresponding Author
*dweix@wisc.edu
■
(15) These conditions give poor cross-selectivity for biaryl syntheses
in reference 8. Applying the conditions from Table 3 to the coupling of
bromobenzene with 4-anisyl triflate resulted in a 47% isolated yield of
cross-product, 4-methoxy-1,1′-biphenyl. The remaining mass balance
consisted of homodimers of the starting materials.
ORCID
Daniel J. Weix: 0000-0002-9552-3378
Notes
The authors declare no competing financial interest.
(16) (a) Sunnemann, H. W.; Hofmeister, A.; Magull, J.; Banwell, M.
̈
G.; de Meijere, A. Org. Lett. 2007, 9, 517. (b) Sunnemann, H. W.;
ACKNOWLEDGMENTS
̈
■
Hofmeister, A.; Magull, J.; de Meijere, A. Chem. - Eur. J. 2007, 13, 3739.
(17) (a) Krasovskiy, A.; Knochel, P. Synthesis 2006, 2006, 890. (b) Jin,
L.; Xin, J.; Huang, Z.; He, J.; Lei, A. J. Am. Chem. Soc. 2010, 132, 9607.
(18) We cannot rule out the possibility that a small portion of the 6
and 7 observed are derived from reaction of vinylnickel(II)
intermediates with acid and I2. In Scheme 2C, this could account for
a maximum [7]/Ni of 2. We observed only trace 1-iodocyclohexene,
suggesting that a vinylzinc species derived from 2a is not important.
(19) The possibility that 6 could be converted to product by a Heck
reaction was ruled out by studying the reactivity of 6 with 2a under
catalytic conditions. See Supporting Information, page S35.
(20) A small amount of direct nickel-palladium transmetalation
cannot be ruled out. At this time, it is not clear if zinc mediated transfer
occurs in reactions with aryl bromides. The vinylzinc is not formed by
direct insertion, see Table 1, entry 11.
The authors gratefully acknowledge funding from the NIH
NIGMS (GM097243 to D.J.W.) and the NSF (DGE-1419118
to A.M.O.). D.J.W. is a Camille Dreyfus Teacher-Scholar.
Additional funding from Novartis, Pfizer, and Boehringer
Ingelheim is also gratefully acknowledged. Dylan E. Parsons
(Univ. of Rochester) is acknowledged for assisting in synthesis
and for scientific discussions. We thank Prof. Frank Huo and Dr.
Sharma SRK Chaitanya Yamijala (Univ. of Rochester) for
assistance with the calculations.
REFERENCES
■
(1) (a) Mehta, G.; Rao, H. S. P. Synthesis of Conjugated Dienes and
Polyenes. In The Chemistry of Dienes and Polyenes; Rappoport, Z., Ed.;
Wiley: New York, 1997; Vol. 1, pp 359. (b) De Paolis, M.; Chataigner,
I.; Maddaluno, J. Top. Curr. Chem. 2012, 327, 87. (c) 1,3-Dienes; Rawal,
V. H., Kozmin, S. A., Eds.; Science of Synthesis Vol. 46; Georg Thieme
Verlag KG: Stuttgart, 2009. (d) Negishi, E.-I.; Huang, Z.; Wang, G.;
Mohan, S.; Wang, C.; Hattori, H. Acc. Chem. Res. 2008, 41, 1474.
(e) Yang, X.-H.; Dong, V. M. J. Am. Chem. Soc. 2017, 139, 1774.
(2) (a) Hansen, A. L.; Skrydstrup, T. Org. Lett. 2005, 7, 5585.
(b) Hansen, A. L.; Ebran, J. P.; Ahlquist, M.; Norrby, P. O.; Skrydstrup,
T. Angew. Chem., Int. Ed. 2006, 45, 3349. (c) Zhou, P.; Jiang, H.;
Huang, L.; Li, X. Chem. Commun. 2011, 47, 1003. (d) Zheng, C.; Wang,
D.; Stahl, S. S. J. Am. Chem. Soc. 2012, 134, 16496.
(21) Nickel-catalyzed electrosynthesis of arylzinc reagents: Sibille, S.;
Ratovelomanana, V.; Perichon, J. J. Chem. Soc., Chem. Commun. 1992,
283.
(22) A related cobalt-catalyzed method that uses zinc powder: Fillon,
H.; Gosmini, C.; Perichon, J. J. Am. Chem. Soc. 2003, 125, 3867.
(23) Coupling of in-situ-formed alkylzinc with aryl halides:
Krasovskiy, A.; Duplais, C.; Lipshutz, B. J. Am. Chem. Soc. 2009, 131,
15592.
(24) See also related iron-catalyzed work: Czaplik, W. M.; Mayer, M.;
Jacobi von Wangelin, A. Angew. Chem., Int. Ed. 2009, 48, 607.
(25) (a) Arriola, D.; Carnahan, E.; Hustad, P.; Kuhlman, R.; Wenzel,
T. Science 2006, 2006, 786. (b) Hustad, P. Science 2009, 325, 704.
(3) Diver, S. T.; Giessert, A. J. Chem. Rev. 2004, 104, 1317.
(4) Hu, X.-H.; Zhang, J.; Yang, X.-F.; Xu, Y. H.; Loh, T.-P. J. Am.
Chem. Soc. 2015, 137, 3169.
(5) Traditional cross-coupling approaches as well as newer variants,
such as decarboxylative coupling, have been reported. (a) Wang, G.;
Mohan, S.; Negishi, E.-I. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 11344.
(b) Yamashita, M.; Hirano, K.; Satoh, T.; Miura, M. Org. Lett. 2010, 12,
592.
(6) Keck, D.; Muller, T.; Brase, S. Synlett 2006, 2006, 3457.
̈
(7) Aryl triflates were not reactive with nickel-bipyridine catalysts and
zinc as the reducing agent: Everson, D. A.; Jones, B. A.; Weix, D. J. J.
Am. Chem. Soc. 2012, 134, 6146.
(8) Ackerman, L. K. G.; Lovell, M. W.; Weix, D. J. Nature 2015, 524,
454.
(9) Beng reported a related method catalyzed by cobalt that had
similar limitations and required slow addition of one coupling partner.
Beng, T. K.; Sincavage, K.; Silaire, A. W. V.; Alwali, A.; Bassler, D. P.;
Spence, L. E.; Beale, O. Org. Biomol. Chem. 2015, 13, 5349.
(10) Vinyl bromides and vinyl triflates are both reactive under nickel
catalysis: (a) Zhao, Y.; Weix, D. J. J. Am. Chem. Soc. 2015, 137, 3237.
(b) Johnson, K. A.; Biswas, S.; Weix, D. J. Chem. - Eur. J. 2016, 22, 7399.
(11) Sakata, K.; Wang, Y.; Urabe, D.; Inoue, M. Org. Lett. 2018, 20,
130.
(12) Several phosphine ligands proved to be effective, including L13
and L14. We chose to keep L19 as the phosphine ligand for this system
D
DOI: 10.1021/jacs.7b13601
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX