Efficient, regioselective palladium-catalyzed tandem heck-lsomerization reaction of aryl bromides and non-allylic benzyl alcohols

Article abstract of DOI:10.1021/ol900036y

An efficient and mild method to couple aryl bromides and activated non-allylic alcohols in a Heck reaction with tandem isomerization to selectively afford high yields of 1,5-diarylalkan-1-ones has been developed. Mechanistic insight was gained through NMR studies of products derived from deuterium-labeled intermediates.

Full text of DOI:10.1021/ol900036y

ORGANIC
LETTERS
2009
Vol. 11, No. 5
1183-1185
Efficient, Regioselective
Palladium-Catalyzed Tandem
Heck-Isomerization Reaction of Aryl
Bromides and Non-Allylic Benzyl
Alcohols
Matthew L. Crawley,*,† Kristin M. Phipps,† Igor Goljer,‡ John F. Mehlmann,†
Joseph T. Lundquist,† John W. Ullrich,† Cuijian Yang,† and Paige E. Mahaney†
DiscoVery Medicinal Chemistry and DiscoVery Analytical Chemistry, Chemical
Sciences, Wyeth Research, 500 Arcola Road, CollegeVille, PennsylVania 19426
crawlem@wyeth.com
Received January 8, 2009
ABSTRACT
An efficient and mild method to couple aryl bromides and activated non-allylic alcohols in a Heck reaction with tandem isomerization to
selectively afford high yields of 1,5-diarylalkan-1-ones has been developed. Mechanistic insight was gained through NMR studies of products
derived from deuterium-labeled intermediates.
Development of palladium-catalyzed coupling reactions has
allowed access to diverse arrays of complex, biologically
important molecules that would have otherwise been difficult
to synthesize.1 In this class of transformations, the Heck
reaction has become a widely utilized method for the
formation of carbon-carbon bonds.2 When used in tandem
with other reactions, such as subsequent palladium-catalyzed
transformations, rearragements, or isomerizations, the power
of this approach is magnified.3
During our investigations of nonsteroidal agonists of the
farnesoid X receptor (FXR), we attempted to employ a Heck
reaction to couple our functionalized heteroaryl core scaffold
appended with γ-hydroxy alcohol 1 to generate an intermedi-
ate styrene derivative 3 (eq 1). Much to our surprise, the
only product isolated, albeit in low yield, was ketone 4. While
tandem Heck-isomerization reactions are well documented
with allylic alcohols4 and homoallylic alcohols5 with aryl
iodides, the few reports with longer chain lengths or aryl
(3) Recent represenative reviews and examples: (a) Muzart, J. Tetra-
hedron 2005, 61, 4179. (b) Alberico, D.; Rudolph, A.; Lautens, M. J. Org.
Chem. 2007, 72, 775. (c) Cheung, W. S.; Patch, R. L.; Player, M. R. J.
Org. Chem. 2005, 70, 3741. (d) Artman, G. D.; Weinreb, S. M. Org. Lett.
2003, 9, 1523. (e) Pinho, P.; Minnaard, A. J.; Feringa, B. L. Org. Lett.
2003, 5, 259.
† Discovery Medicinal Chemistry.
‡ Discovery Analytical Chemistry.
(1) For selected reviews, see: (a) Tietz, L. F.; Ila, H.; Bell, H. P. Chem.
ReV. 2004, 104, 3453. (b) Trost, B. M.; Crawley, M. L. Chem. ReV. 2003,
103, 2921. (c) Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457.
(2) For selected reviews, see: (a) Beletskaya, I. P.; Cheprakov, A. V.
Chem. ReV. 2000, 100, 3909. (b) Dounay, A. B.; Overman, L. E. Chem.
ReV. 2003, 103, 2945.
(4) Classic examples: (a) Melpoder, J. B.; Heck, R. F. J. Org. Chem.
1976, 41, 265. (b) Chalk, A. J.; Magennis, S. A. J. Org. Chem. 1976, 41,
273.
10.1021/ol900036y CCC: $40.75
Published on Web 02/11/2009
2009 American Chemical Society

bromides require either forcing conditions, long time periods,
or unusual ligands and in all cases give mixtures of
regioisomeric and intermediate products, including cyclized
adducts.6 In fact, some research groups have deliberately
taken advantage of the cyclized “side products” to synthesize
a variety of tetrahydrofurans from γ-hydroxy terminal
alkenes.7 Additionally, under mild conditions reactions of
homoallyl alcohols have been exclusive for aryl iodides in
the presence aryl bromides, as is the case with 2-iodobro-
mobenzene and pent-4-en-2-ol.5b
1). When the reaction was run for only half the time (1 h)
under these conditions, conversion was still complete and
the isolated yield improved to 44% (Table 1, entry 2).
An examination of reaction progress at different temper-
atures and times (Table 1, entries 2-6) revealed that an
improved yield (72%) of product was possible at a lower
temperarture (50 °C). However, at room temperature the
reaction did not reach full conversion even after 24 h had
elapsed. Changing the base from potassium carbonate to
triethylamine or cesium carbonate resulted in lower conver-
sion and higher formation of side products (Table 1, entries
7 and 8). Notably, the use of sodium bicarbonate as a base
allowed complete conversion to product in 3 h and high yield
of product in only 0.5 h at 50 °C (Table 1, entries 9 and
10). Further optimization of conditions was not achieved by
variation of the solvent (Table 1, entries 12-14). As
expected, the reaction did not proceed in the absence of
lithium chloride (Table 1, entry 11), confirming its role as a
ligand.
A small array of electron-rich and electron-poor fragments
were coupled under the optimized conditions (Table 2). A
To better understand the scope and limitations of the
process we observed and whether it was specific to our
unusual system, we attempted to optimize the reaction
conditions on the simple γ-hydroxy alkene derivative 1a
(Table 1). Starting with the same conditions as applied to
Table 2. Pd-Cataylzed Tandem Heck-Isomerization Reactionsa
Table 1. Optimization of Reaction Conditionsa
entry
R1
4-Me
H
4-OMe
4-CO2Me
H
H
H
R2
n
product
yield %
1
2
3
4
5
6
7
8
9
H
H
H
H
2
2
2
2
2
2
2
1
1
1
4a
4b
4c
4d
4e
4f
4g
4h
4i
87
82
71
59
75
88
67
21c
23c
45
ArBr time temp
baseb
(3 equiv)
yield %
(convn)c
4-OMe
3-CN
4-CO2Et
H
entry equiv
(h)
(°C)
solvent
1
2
3
4
5
6
7
8
1.2
2.5
3.0
3.0
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
3.0
2.5
2
1
1
3
5
24
3
3
3
0.5
0.5
3
24
3
100
100
50
50
50
25
50
50
50
50
50
50
50
50
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
TEA
Cs2CO3
NaHCO3
NaHCO3 DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
34 (100)
44 (100)
(90)
H
4-OMe
4-CO2Me
H
H
(>95)
10
4j
72 (100)
(76)
a All reactions were run with alkene as the limiting reagent on a 1.00
mmol scale under N2 atmosphere. b All bases and solvents used were
anhydrous. c Significant amounts of side products formed.
(88)d
71 (95)
(>95)
9
10
11
12
13
14
87 (100)
range of substituents gave moderate to high yields of targets
under optimized conditions when n ) 2 (Table 2, entries
1-7). Notably, there was not any obvious trend or difference
in reactivity between the systems of varied electronics.
However, surprisingly, when smaller homoallyl substrates
1h-1j (n ) 1) were utilized, the yields fell significantly and
the reaction produced traditional isomeric side products
(Table 2, entries 8-10).
NaHCO3
NaHCO3
NaHCO3
NaHCO3
DMF
(0)e
CH3CN (35)d
CH3CN (40)d
THF
(71)
a All reactions were run with alkene 1a as the limiting reagent on a
1.00 mmol scale under N2 atmosphere. b All bases and solvents used were
anhydrous unless other specified. c The conversion was determined by
LCMS monitoring using UV detection at 220 nM. d Significant amounts of
side products formed. e LiCl was not added to the reaction.
The rapid and efficient nature of the transformation when n
) 2 coupled with the low yield and generation of isomers when
n ) 1 caused us to wonder if a mechanism other than the
traditional palladium-catalyzed olefin migration8 was needed
our hydroxy alkene fragment, compound 1a was successfully
reacted with bromobenzene to give ketone 4a (Table 1, entry
1184
Org. Lett., Vol. 11, No. 5, 2009

to rationalize the results when n ) 2. One idea considered was
whether or not some type of direct C-H abstraction or other
process terminated the Heck addition, possibly stabilized
through a six-membered palladacycle, facilitating the transposi-
tion of the olefin to the ketone (eq 2).
Table 3. Probe of the Tandem Heck-Isomerization Mechanism
entry reagent
conditions
convn (%)
To test this hypothesis, deuterium-labeled substrate 1b was
synthesized and subjected to reaction conditions (eq 3).9
Migration of the deuterium was observed yielding a 2:1
mixture of isomers of 4b. If a direct C-H activation was
the active pathway, the deteurium label should have remained
unaffected. This result is consistent with formation of a
traditional Heck adduct containing an olefin, followed by
reversible hydropalladation-dehydropalladation as the mech-
anism of migration. This would eventually lead to enol
formation (which is a thermodynamic sink as the enol
tautomerizes to and is trapped as the ketone).
5 mol % Pd(OAc)2, 1.5 equiv LiCl,
NaHCO3, DMF, 50 °C, 0.5 h
5 mol % Pd(OAc)2, 1.5 equiv LiCl,
NaHCO3, DMF, 50 °C, 0.5 h
5 mol % Pd(OAc)2, 1.5 equiv LiCl,
NaHCO3, DMF, 50 °C, 24 h
5 mol % Pd(OAc)2, 1.5 equiv LiCl,
NaHCO3, DMF, 50 °C, 24 h
10 mol % Et3N, 10 mol % Pd/C
(5 mol % Pd), toluene, 80 °C
10 mol % Et3N, 10 mol % Pd/C
(5 mol % Pd), toluene, 80 °C
5 mol % Pd(OAc)2, 5 mol % PhBr,
1.5 equiv LiCl, NaHCO3, DMF, 50 °C, 1 h
5 mol % Pd(OAc)2, 1.5 equiv LiCl,
NaHCO3, Bu4NBr, DMF, 50 °C, 1 h,
1
2
3
4
5
6
7
8
3b
5b
3b
5b
3b
5b
5b
5b
0
0
<5
<5
<5
20c
∼5
<5
a All bases and solvents used were anhydrous unless otherwise specified,
and reactions were run on 0.500 mmol scale under N2 atmosphere. b The
conversion to 4b was determined by LCMS monitoring using UV detection
at 220 nM. c A complex mixture of several adducts and starting material
formed.
conditions did cause 5b to isomerize to some extent, albeit
with a complex mixture of side products. When a 1:1 ratio
of aryl bromide versus palladium catalyst was added, no
appreciable isomerization occurred (Table 3, entry 7).
Addition of tetrabutylammonium bromide to the reaction also
had no effect (Table 3, entry 8). The combination of these
results suggest that although clearly olefin migration may
be part of the mechanism, there must be additional nonclas-
sical factors in the reaction at work to facilitate this relatively
facile transformation.
In conclusion, we have developed mild and efficient
conditions that led in high yields to functionalized 1,5-
diarylalkanone products. Notably, this protocol avoids the
use of phosphine ligands and the formation of regioisomeric
and cyclized products that are common during this type of
transformation. We were able to apply this route to synthesize
several arrays of otherwise difficult to access medicinal
chemistry targets, the results of which will be reported in
due course.
Although the deteurium labeling result was consistent with
a stepwise isomerization mechanism, additional information
was needed to support this result. When adducts 3b and 5b,10
presumably initial adducts formed from the Heck reaction,
were subjected to the optimized reaction conditions, less than
5% migration product (4b) was observed after 30 min (Table
3, entries 1 and 2). Prolonged reaction time did not
significantly increase migration and resulted in the formation
of multiple products (Table 3, entries 3 and 4). This result
was not unexpected as the requisite palladium hydride would
likely not be generated without the initial aryl bromide
present. Attempts to apply modified conditions, whereby a
palladium hydride-iminium complex is generated in situ,11
still did not produce isomerization of 3b to 4b. These
(5) (a) Gangjee, A.; Yu, J.; Kisliuk, R. L.; Haile, W. H.; Sobrero, G.;
McGuire, J. J. J. Med. Chem. 2003, 46, 591. (b) Qadir, M.; Priestley, R. E.;
Rising, T. W. D. F.; Gelbrich, T.; Coles, S. J.; Hursthouse, M. B.; Sheldrake,
P. W.; Whittall, N.; Hii, K. K. Tetrahedron Lett. 2003, 44, 3675. (c) Ohno,
H.; Okumura, M.; Maeda, S.-I.; Iwasaki, H.; Wakayama, R.; Tanaka, T. J.
Org. Chem. 2003, 68, 7722. (d) Dyker, G.; Grundt, P. HelV. Chim. Acta
1999, 82, 588. (e) Dyker, G.; Markwitz, H. Synthesis 1998, 1750. (f) Dyker,
G.; Grundt, P.; Markwitz, H.; Henkel, G. J. Org. Chem. 1998, 63, 6043.
(6) (a) Larock, R. C.; Leung, W.-Y.; Stolz-Dunn, S. Tetrahedron Lett.
1989, 30, 6629. (b) Berthiol, F.; Doucet, H.; Santelli, M. Synthesis 2005,
20, 3589.
Supporting Information Available: Experimental details;
1H and 13C NMR, HRMS, and HPLC results for selected
products. This material is available free of charge via the
Internet at http://pubs.acs.org.
OL900036Y
(7) Hay, M. B.; Hardin, A. R.; Wolfe, J. P. J. Org. Chem. 2005, 70,
3099.
(10) Adduct 5b was prepared by the procedure of Kimura, M.; Matsuo,
S.; Shibata, K.; Tamaru, Y. Angew. Chem., Int. Ed. 1999, 38, 3386.
(11) Recently reported conditions to isomerize allyl alcohols to the
corresponding carbonyl compounds: Coquerel, Y.; Bremond, P.; Rodriguez,
J. J. Organomet. Chem. 2007, 4805.
(8) Migration by sequential hydropalladation-dehydropalladation.
(9) 1b-(D) was synthesized with >95% incorporation of deuterium in
four steps from benzaldehyde (see Supporting Information).
Org. Lett., Vol. 11, No. 5, 2009
1185