Palladium(0)-catalyzed synthesis of chiral ene-allenes using alkenyl trifluoroborates

Article abstract of DOI:10.1021/jo052201x

Enantioenriched allenes serve as chiral transfer reagents, making them attractive synthetic targets. Herein, the synthesis of enantioenriched allenes utilizing a Pd(0)-catalyzed cross-coupling reaction of propargylic carbonates and phosphates with alkenyl trifluoroborates is reported. Di-, tri-, and tetrasubstituted allenes were synthesized in moderate to high optical yields. Several racemic allenes possessing various functional groups were also synthesized.

Full text of DOI:10.1021/jo052201x

Palladium(0)-Catalyzed Synthesis of Chiral Ene-allenes Using
Alkenyl Trifluoroborates
Gary A. Molander,* Erin M. Sommers, and Sharon R. Baker
Roy and Diana Vagelos Laboratories, Department of Chemistry, UniVersity of PennsylVania,
Philadelphia, PennsylVania 19104-6323
gmolandr@sas.upenn.edu
ReceiVed October 21, 2005
Enantioenriched allenes serve as chiral transfer reagents, making them attractive synthetic targets. Herein,
the synthesis of enantioenriched allenes utilizing a Pd(0)-catalyzed cross-coupling reaction of propargylic
carbonates and phosphates with alkenyl trifluoroborates is reported. Di-, tri-, and tetrasubstituted allenes
were synthesized in moderate to high optical yields. Several racemic allenes possessing various functional
groups were also synthesized.
Introduction
synthesized either by reduction of the conjugated alkynones6
or from the asymmetric addition of alkynylmetallics to alde-
hydes,7 organocuprate methods suffer from some major draw-
backs, including racemization of the product allenes under the
reaction conditions.8 Attempts have been made to remedy this
problem (e.g., utilizing mixed organocuprates, additives, and
different leaving groups), but a general approach to the synthesis
of optically active allenes remains elusive.
Allenes are becoming increasingly important as synthetic
targets, both in natural products and in other biologically active
compounds.1 Moreover, chiral allenes with appropriate substitu-
tion are uniquely powerful intermediates for organic synthesis,
translating axial chirality to central chirality.2 For these reasons,
our laboratory seeks to develop methods for the efficient,
stereoselective synthesis of enantiomerically enriched allenes.
Previously, the synthesis of chiral nonracemic allenes by cationic
chromium(III) catalysis has been reported.3 However, this
method possesses several drawbacks. Not only must substrates
remain unfunctionalized, but a bulky tert-butyldimethylsilanoxy
group serves as the leaving group, generating wasteful byprod-
ucts. In an effort to avoid these complications, a milder, more
atom economical synthesis of enantioenriched allenes was
desired.
(4) A representative sample. (a) Rona, R.; Crabbe´, P. J. Am. Chem. Soc.
1968, 90, 4733-4734. (b) Buchwald, S. L.; Grubbs, R. H. J. Am. Chem.
Soc. 1983, 105, 5490-5491. (c) Varghese, J. P.; Knochel, P.; Marek, I.
Org. Lett. 2000, 2, 2849-2852.
(5) (a) Rona, R.; Crabbe´, P. J. Am. Chem. Soc. 1969, 91, 3289-3293.
(b) Myers, A. G.; Zheng, B. J. Am. Chem. Soc. 1996, 118, 4492-4493. (c)
Schultz-Fademrecht, C.; Wibbeling, B.; Fro¨hlich, R.; Hoppe, D. Org. Lett.
2001, 3, 1221-1224. (d) Shimizu, M.; Kurahashi, R.; Kitagawa, H.; Hiyama,
T. Org. Lett. 2003, 5, 225-227.
(6) (a) Nishizawa, M.; Yamada, M.; Noyori, R. Tetrahedron Lett. 1981,
22, 247-250. (b) Ramachandran, P. V.; Teodorovic, A. V.; Rangaishenvi,
M. V.; Brown, H. C. J. Org. Chem. 1992, 57, 2379-2386. (c) Parker, K.
A.; Ledeboer, M. W. J. Org. Chem. 1996, 61, 3214-3217. (d) Matsumura,
K.; Hashiguchi, S.; Ikariya, T.; Noyori, R. J. Am. Chem. Soc. 1997, 119,
8738-8739.
(7) (a) Boyall, D.; Frantz, D. E.; Carreira, E. M. Org. Lett. 2002, 4, 2605-
2606. (b) El-Sayed, E.; Anand, N. K.; Carreira, E. M. Org. Lett. 2001, 3,
3017-3020. (c) Boyall, D.; Lo´pez, F.; Sasaki, H.; Frantz, D. E.; Carreira,
E. M. Org. Lett. 2000, 2, 4233-4236. (d) Frantz, D. E.; Fa¨ssler, R.; Carreira,
E. M. J. Am. Chem. Soc. 2000, 122, 1806-1807. (e) Frantz, D. E.; Fa¨ssler,
R.; Tomooka, C. S.; Carreira, E. M. Acc. Chem. Res. 2000, 33, 373-381.
(f) Maezaki, N.; Kojima, N.; Tominaga, H.; Yanai, M.; Tanaka, T. Org.
Lett. 2003, 5, 1411-1414. (g) Pu, L. Tetrahedron 2003, 59, 9873-9886.
(8) Claesson, A.; Olsson, L.-I. J. Chem. Soc., Chem. Commun. 1979,
524-525.
Although a myriad of allene syntheses exist,4 few are achieved
in high ee.5 SN2′-type reactions on optically active propargylic
derivatives involving organocuprates are the most common and
convenient methods for generating enantioenriched allenes.5
Although chiral nonracemic propargylic alcohols are easily
(1) (a) Hoffman-Ro¨der, A.; Krause, M. Angew. Chem., Int. Ed. 2004,
43, 1196-1216.
(2) A few examples: (a) Marshall, J. A.; Bourbeau, M. P. Org. Lett.
2002, 4, 3931-3934. (b) Ma, S.; Wu, S. Chem. Commun. 2001, 441-442.
(c) Marshall, J. A.; Wolf, M. A. J. Org. Chem. 1996, 61, 3238-3239.
(3) Molander, G. A.; Sommers, E. M. Tetrahedron Lett. 2005, 46, 2345-
2349.
10.1021/jo052201x CCC: $33.50 © 2006 American Chemical Society
Published on Web 01/24/2006
J. Org. Chem. 2006, 71, 1563-1568
1563

Molander et al.
TABLE 1. Optimization of Reaction Conditions
Easy access to enantioenriched substrates, mild reaction
conditions, and atom economy make cross-coupling routes to
form allenes extraordinarily attractive.9 Despite these favorable
attributes, few general methods have been developed. Arylations
using organozinc reagents have been reported with varying
success.10 Depending on the substrate, anti selectivity ranges
from 4:1 to complete stereoselectivity, and sterically unbiased
internal and terminal propargylic substrates consistently yield
allenes with less than 90% ee. The analogous Suzuki-Miyaura
coupling reaction with boronic acids yielded an allene with
complete transfer of chirality in the sole case reported.11
Likewise, one example has been reported where a 4-aryl-2,3-
allenol was formed with complete retention of stereochemistry
when a propargylic oxirane undergoes the cross coupling
reaction with an aryl boronic acid.12 Although other aryl-,11
alkynyl-,13 alkenyl-,14 and alkylzincs10a as well as analogous
organoboron reagents have been used as coupling partners in
racemic reactions, the scope of the reactions has not been
extensively explored. Thus, the opportunity to refine this
transformation exists. Because only one example of a cross-
coupling type reaction with an alkenyl organoboron has been
reported,14 alkenyl trifluoroborates as coupling partners served
as a place to begin the studies described herein.
time,
h
% GC
yield
entry
1
catalyst, ligand
base
PdCl2(dppf)•CH2Cl2 (10 mol %) Cs2CO3
2
starting
material
14a
41a
93
66
100
61
2
3
4
5
6
7
8
9
10
Pd(OAc)2, 2 PPh3 (10 mol %)
Pd(PPh3)4 (10 mol %)
Pd(PPh3)4 (10 mol %)
Pd(PPh3)4 (10 mol %)
Pd(PPh3)4 (10 mol %)
Pd(PPh3)4 (10 mol %)
Pd(PPh3)4 (10 mol %)
Pd(PPh3)4 (1 mol %)
Cs2CO3
Cs2CO3
Cs2CO3
K2CO3
NaHCO3
i-Pr2NH
none
2
2
2
7
2
4
24
3
0
90
58
NaHCO3
NaHCO3
Pd(PPh3)4 (0.5 mol %)
4.5
a Isolated yields.
on complete consumption of the starting material. A comparison
of different Pd catalysts proved that our original choice of
Pd(PPh3)4 was indeed correct (Table 1, entry 3). The reaction
does not proceed in the absence of a Pd(0) catalyst. After
determining Pd(PPh3)4 to be the appropriate catalyst, several
bases were screened. Entries 4-8 (Table 1) depict the results
of changing the base. Inexpensive NaHCO3 proved to be the
most efficient base. Using NaHCO3, the catalyst loading was
lowered. Entries 9 and 10 (Table 1) display these results. A
10-fold decrease in catalyst loading from 10 to 1 mol % created
only a 10% decrease in yield (entries 6 and 9), although the
reaction time was longer with less catalyst. However, lowering
the catalyst loading to only 0.5 mol % on this reaction scale
not only lengthened the reaction time but also decreased the
yield (entry 10). Therefore, 1 mol % was chosen as the optimum
catalyst loading on the reaction scale employed. It should be
noted that up to 10 mol % was used in subsequent reactions to
achieve the desired reaction efficiency and high yields. The final
reaction parameter explored was the solvent. EtOH, DMSO,
and THF without water were all tried. EtOH led to deprotection
of 1 to yield only the free alcohol and no allene product. No
reaction was observed in DMSO or THF without water. Despite
the necessity of water, the water content could be lowered to
25:1 (THF/H2O) without observing any decrease in yield or
reaction rate.
Potassium organotrifluoroborate salts have many attractive
properties compared to other organoboron compounds.15 They
are monomeric, air-stable solids that generate nontoxic byprod-
ucts. This paper details the synthesis of allenes by Pd(0)-
catalyzed cross-coupling of alkenyl trifluoroborates and pro-
pargylic carbonates. The functional group scope of the reaction
is also presented. Additionally, the stereoselectivity of the
reaction with propargylic phosphates is probed.
Results and Discussion
Optimization of Reaction Conditions for Propargylic
Carbonates and Exploration of Substrate Scope. Initially,
using Pd(PPh3)4 as a catalyst and Cs2CO3 as a base in THF/
H2O16 yielded 41% of the desired racemic ene-allene 3 from 1
(eq 1). Encouraged by this initial result, the reaction was
One more aspect of the reaction optimization bears mention.
The palladium byproducts present appeared to polymerize/
degrade the allene product. Very low isolated yields were
obtained initially. To eliminate this problem, methods were
investigated to remove the offending palladium from solution
as quickly as possible upon completion of the reaction. Addition
of adsorbent carbon black (Darco) followed by a period of
stirring exhibited a marked increase in isolated yield (e.g., 41-
optimized for palladium source, catalyst loading, base, solvent,
and isolation conditions (Table 1). The yields given are
quantitative GC yields using an internal standard and are based
(9) (a) Jeffery-Luong, T.; Linstrumelle, G. Tetrahedron Lett. 1980, 21,
5019-5020. (b) Ma, S. Eur. J. Org. Chem. 2004, 1175-1183. (c)
Ogasawara, M.; Ikeda, H.; Nagano, T.; Hayashi, T. Org. Lett. 2001, 3,
2615-2617. (d) Lee, K.; Seomoon, D.; Lee, P. Angew. Chem., Int. Ed.
2002, 41, 3901-3903. (e) Ogasawara, M.; Ueyama, K.; Nagano, T.;
Mizuhata, Y.; Hayashi, T. Org. Lett. 2003, 5, 217-219.
(11) Yoshida, M.; Gotou, T.; Ihara, M. Tetrahedron Lett. 2004, 45, 5573-
5575.
(12) Yoshida, M.; Ueda, H.; Ihara, M. Tetrahedron Lett. 2005, 46, 6705-
6708.
(10) (a) Elsevier: C. J.; Stehouwer, P. M.; Westmijze, H.; Vermeer, P.
J. Org. Chem. 1983, 48, 1103-1105. (b) Dixneuf, P. H.; Guyot, T.; Ness,
M. D.; Roberts, S. M. Chem. Commun. 1997, 2083-2084. (c) Konno, T.;
Tanikawa, M.; Ishihara, T.; Yamanaka, H. Chem. Lett. 2000, 1360-1361.
(d) Elsevier, C. J.; Kleijn, H.; Boersma, J.; Vermeer, P. Organometallics
1986, 5, 716-720. (e) Elsevier, C. J.; Mooiweer, H. H.; Kleijn, H.; Vermeer,
P. Tetrahedron Lett. 1984, 25, 5571-5572.
(13) (a) Gueugnot, S.; Linstrumelle, G. Tetrahedron Lett. 1993, 34,
3853-3856. (b) Mandai, T.; Nakata, T.; Murayama, H.; Yamaoki, H.;
Ogawa, M.; Kawada, M.; Tsuji, J. Tetrahedron Lett. 1990, 31, 7179-7180.
(14) Moriya, T.; Miyaura, N.; Suzuki, A. Synlett 1994, 149-151.
(15) Molander, G. A.; Figueroa, R. Aldrichim. Acta 2005, 38, 49-56.
(16) Molander, G. A.; Yun, C, -S.; Ribagorda, M.; Biolatto, B. J. Org.
Chem. 2003, 68, 5534-5539.
1564 J. Org. Chem., Vol. 71, No. 4, 2006

Palladium(0)-Catalyzed Synthesis of Chiral Ene-allenes
TABLE 2. Exploration of Substrate Scope
entry
substrate
R1
R2
P
R3
C8H17
product yield (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
4
1
6
9
Me
H
H
Me
Me
Me
H
Me
Me
H
Ph, 2
Ph, 2
H, 7
Ph, 2
H, 7
H, 7
H, 7
H, 7
H, 7
H, 7
H, 7
CO2Me
CO2Me
P(O)(OEt)2
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
P(O)(OEt)2
CO2Me
CO2Me
CO2Me
CO2Me
5, 84
3, 70
8, 56
H
(CH2)3CN
(CH2)3CN
(CH2)3CN
(CH2)4OH
(CH2)3Cl
(CH2)4N(Bn)2
(CH2)3CHO
(CH2)4OTBS
(CH2)4OTBS
H
10, 76
11, 71
13, 63
15, 47
17, 53
19, 19
21, 74
21, 44
25, 69
27, 73
30, 60
32, 17a
9
12
14
16
18
20
22
23
26
28
31
H
Me
Me
H
(CH2)3Cl, 24
H, 7
C8H17, 29
C8H17, 29
(CH2)4SPh
(CH2)3SO2Ph
TMS
H
a Product 32; R3 ) H.
SCHEME 1. Mechanism of Pd(0)-Catalyzed Cross-Coupling
Reaction To Form Allenes
70% yield). It became standard practice to open the reactions
to air and quench with Darco for all subsequent reactions. Except
for the initial results (eq 1 and Table 1, entries 2 and 3), the
yields reported herein reflect this workup protocol.
After optimizing the reaction parameters, the functional group
tolerance was explored. The results are presented in Table 2.
Many functional groups were compatible with the reaction
conditions. For example, nitrile, naked and protected alcohol,
amine, chloride, and sulfur-containing substrates tolerated the
reaction conditions, producing racemic ene-allenes in moderate
to good yields. Aldehyde 18 was the exception, producing only
19% of the allene. The basic conditions necessary for the
reaction may have led to aldol products or polymerization,
lowering the yield. The moderate yields of the remaining
substrates can be explained by the above-mentioned workup
protocol. The Darco necessary to achieve reasonable yields may
also sequester the product. Excessive solvent washes of the
Darco will release the problematic palladium. As discussed
below, phosphates were thermally unstable, highlighted by the
results in entries 3 and 11. Unlike their carbonate counterparts
(entries 5 and 10, respectively), the phosphates suffered from
lower yields. To synthesize an allene capped with a trimethyl-
silyl group, 31 was subjected to the reaction conditions. Instead
of isolating the desired TMS-substituted allene, a small amount
(17%) of deprotected product was observed. The alkynyl-TMS
might be readily deprotected under the basic reaction conditions,
or the fluorine-containing trifluoroborate byproducts may have
removed TMS from the alkyne or allene. More robust silyl
protecting groups were not explored.
Carbonate (R)-1 was subjected to the optimized reaction
conditions (eq 2). Disappointingly, the product was nearly
racemic.
At this juncture, the literature was searched for examples of
palladium-catalyzed racemization of allenes. To the best of our
knowledge, palladium has been shown to racemize allenes by
three mechanisms. In the first mechanism, allenes can be
racemized through the η1-allenylpalladium(II) to η1-propargyl-
palladium(II) interconversion (Scheme 2).17 When the Pd
Probing the Stereospecificity of the Reaction: Optimiza-
tion and Application. Based on previous literature reports of
similar transformations, the substitution step was anticipated to
proceed with inversion to generate allenyl-Pd I (Scheme 1),
while the transmetalation/reductive elimination steps would
occur with retention of configuration to yield an allene product
III with overall anti substitution.2b Therefore, an enantioenriched
substrate should lead to an enantioenriched allene product.
(17) Mikami, K.; Yoshida, A. Angew. Chem., Int. Ed. Engl. 1997, 36,
858-860.
J. Org. Chem, Vol. 71, No. 4, 2006 1565

Molander et al.
SCHEME 2. Possible Mechanism for Racemization of
Allenes via Propargylic/Allenyl Pd Intermediate
completely inhibited the reaction. Water was introduced using
hydrated bases: KF‚2H2O, CsOH‚H2O, and Na2CO3‚H2O in
THF led to the formation of racemic allenes. Mimicking
Ogoshi’s work, excess PPh3 was added to the reaction. Except
for noticeably slowing the reaction, only a moderate rise in ee
was observed when the substrate was fully consumed. Other
metal catalysts (Ni, Rh, and Cu) were employed with no success.
In a final bid to raise the ee, different leaving groups were
examined. The acetate leaving group led to racemic product at
a slower reaction rate than the propargylic carbonate. The slower
reaction can be explained by the fact that the acetate anion acts
as a ligand for palladium, retarding the rate-determining
transmetalation step and thereby decreasing the reaction rate.20
When an enantioenriched phosphate was subjected to the
reaction conditions (eq 3), not only was there a rise in ee, but
the reaction could be conducted at room temperature instead of
at reflux. At this juncture, the catalyst/ligand combinations were
screened again using a propargylic phosphate. An array of reac-
tions using PdCl2, Pd(PhCN)2Cl2, Pd(MeCN)2Cl2, Pd(OAc)2,
Pd2dba3, and [Pd(allyl)Cl]2 as catalysts and Ph3As, (2-furyl)3P,
dppe, dppb, and dppf as ligands were conducted. Each catalyst/
ligand combination yielded products that were racemic.
Pd(PPh3)4 remained the most viable catalyst for this transforma-
tion. Again different metals, bases, solvents, and additives were
investigated to raise the ee; however, no additional headway
was made. Under the countless reaction conditions tried, the ee
of 34 was initially high; however, (S)-33 failed to be fully
consumed before racemization occurred.
SCHEME 3. Racemization of Allenes with Pd(II) and LiBr
via a Bromopalladation/Elimination Sequence
SCHEME 4. Racemization of Allenyl-Pd Intermediate via
a Bimetallic Complex
occupies the propargylic position, free rotation can occur,
resulting in racemization. This mechanism proved beneficial for
Mikami and co-workers in their report of a dynamic kinetic
protonation of intermediate allenylmetals.17
By the second mechanism (Scheme 3), enantioenriched
allenes are racemized by Pd(OAc)2-catalyzed bromopalladation/
LiBr elimination.18
These empirical observations prompted an investigation into
the nature of the racemization. We proposed that there was an
initial rapid formation of allene possessing high ee followed
by a prolonged period in which little to no additional product
was formed and racemization of the product occurred. This idea
was confirmed by examining the conversion versus the race-
mization of the product at several temperatures (Table 3).
As predicted, the ee decreased over time while the production
of product tapered off. At room temperature the racemization
was rapid, occurring within 2 min, while the reaction reached
completion after 1 h. Lowering the temperature was beneficial
to both the conversion and ee, but a similar trend was observed.
Because of a difficult analytical separation of the enantiomers
of 34, another trifluoroborate was examined to corroborate the
trend. Table 4 indicates the results of using vinyl trifluoroborate
7. Unlike the reaction with 2, the reaction was complete within
30 s, but the same loss of ee was observed over time.
In the third case, Ogoshi and co-workers postulated that an
enantioenriched η1-allenylpalladium(II) species will racemize
over time due to the generation of a configurationally labile
µ-η3-allenyl/propargyldipalladium complex under normal cata-
lytic conditions (Scheme 4). Racemization was accelerated in
the presence of oxygen and decreased when as little as 10 mol
% of PPh3 was added to the reaction.19
The reaction conditions were manipulated to maximize the
enantioselectivity based on these three mechanisms. Unfortu-
nately, success was limited. Palladium catalysts on solid supports
were used to generate single-site catalysis and to circumvent
the formation of potential bimetallic intermediates (Scheme 4).
Reactions using Pd(PPh3)4 on polystyrene, FibreCat 1029, and
FibreCat 1031 exhibited a decreased reaction rate or no reaction,
and any product isolated was racemic. Yoshida suggests that
water assists formation of the bimetallic species (Scheme 4).12
To eliminate water, other alcoholic solvents, such as tert-amyl
and sec-butyl alcohol were used. Unfortunately, these solvents
These results confirmed the hypothesis that the newly formed
ene-allenes sequester the palladium, perhaps via initial coordina-
tion shown in Figure 1, and undergo racemization, effectively
shutting down the desired coupling reaction. Alkenes with
electron withdrawing substituents are better ligands for pal-
ladium based on back-bonding into the empty π* orbital.21
(20) Hegedus, L. S. Transition Metals in the Synthesis of Complex
Organic Molecules; University Science Books: Sausalito, CA, 1999; p 256.
(21) Spessard, G. O.; Miessler, G. L. Organometallic Chemistry; Prentice
Hall: Upper Saddle River, NJ, 1997; p 49.
(18) Horva´th, A.; Ba¨ckvall, J. Chem. Commun. 2004, 964-965.
(19) Ogoshi, S.; Nishida, T.; Shinagawa, T.; Kurosawa, H. J. Am. Chem.
Soc. 2001, 123, 7164-7165.
1566 J. Org. Chem., Vol. 71, No. 4, 2006

Palladium(0)-Catalyzed Synthesis of Chiral Ene-allenes
TABLE 3. Time vs Conversion and % ee Using 2
TABLE 5. Synthesis of Enantioenriched Ene-allenes
entry
T, °C
time, min
conversion,a %
ee, %
catalyst, time, product,
mol % min yield (%) % ee
1
2
3
4
5
6
7
8
9
rt
rt
rt
rt
rt
0.5
2.5
5
12
60
5
22
5
60
15
54
73
80
88
99
80
93
92
95
90
67
0
0
0
0
85
76
80
59
74
entry
trifluoroborate
1
2
3
4
5
R1, R2 ) H, 7
2
1.75 35, 80
94
76
32
87
91
R1 ) Me, R2 ) H, 36
R1 ) H, R2 ) Ph, 2
R1 ) H, R2 ) (CH2)3CN, 38
38
2
10
10
5
10
37, 52
34, 42
39, 45a
39, 53a
2
20
20
0
0
-15
-15
-40
a Yield based on recovered starting material.
10
a Conversion based on quantitative GC analysis of starting material
consumption.
To expand the scope of the asymmetric reaction, several
classes of alkenyl trifluoroborates were subjected to the reaction
conditions. The results are displayed in Table 5 and reflect the
optimized result for each trifluoroborate. (S)-33 reacted with
vinyl trifluoroborate (7) to produce (R)-35 in 80% yield and
94% ee (Table 5, entry 1). The less substituted trifluoroborates
produced allenes possessing high ee (Table 5, entries 1 and 4).
Conversely, the more substituted potassium 1-methyl-1-vinyl
trifluoroborate (36) resulted in a poorer yield of 37, with a
moderate loss of ee. Ironically, our test trifluoroborate 2 led to
5 in the poorest yield and ee (Table 5, entry 3).23 In the case of
the 1,2-disubstituted trifluoroborate 38, the reaction time was
extended to increase the conversion. This extra time negatively
impacted the ee of the reaction (Table 5, entry 4), but loss of
ee was alleviated to some degree by decreasing the catalyst
loading (Table 5, entry 5).
TABLE 4. Time vs Conversion and % ee Using 7
entry
time, min
conversion,a %
ee, %
1
2
3
4
5
0.5
1
3
4
5
> 99
> 99
> 99
> 99
> 99
98
92
78
76
50
a Conversion based on quantitative GC analysis of starting material
consumption.
In addition to different classes of alkenyl trifluoroborates,
substrates possessing different substitution patterns were sub-
jected to the reaction conditions. Equations 4 and 5 illustrate
the formation of enantioenriched tri- and tetrasubstituted allenes.
In the case of 40, the reaction must be heated to reflux for
reasonable conversion to 41. However, phosphates were ther-
mally unstable at higher temperatures, resulting in a lower yield.
Additionally, a high reaction temperature negatively impacted
the ee. In contrast, the tetrasubstituted allene 43 was formed in
moderate yield with high retention of ee. Although 42 was more
sterically encumbered than 40, the phosphate was poised to be
a much better leaving group because of its benzylic, propargylic,
and tertiary character. Therefore, the reaction was performed
at room temperature. Because the synthesis of chiral nonracemic
tertiary propargylic alcohols has recently been reported,24 this
method is a convenient way to generate tetrasubstituted allenes
in high ee.
As a last area of interest, other classes of trifluoroborates,
such as aryl, alkynyl, or alkyl were subjected to the reaction
conditions optimized for alkenyl trifluoroborates. Unfortunately,
only an aryl trifluoroborate coupled to generate an allene under
these conditions (eq 6). Interestingly, aryl allenes appear to be
more resistant to racemization. After 40 min at reflux, 45 was
isolated in 80% ee. This result correlates with Yoshida’s
successful coupling of an aryl boronic acid and a carbonate at
FIGURE 1. Possible coordination of Pd(0) with ene-allenes 34 and
35.
Because the phenyl group is more electronegative than a
hydrogen substituent,22 allene 34 was a better ligand for Pd(0)
than 35, explaining the slower turnover and more rapid
racemization of 34. Additionally, the phenyl substituent may
further coordinate to the metal center, increasing its affinity.
Initially, high ee’s suggest that the first and third racemization
mechanisms (Schemes 2 and 4) are not operating. If racemiza-
tion occurred at the allenyl-Pd intermediate, any product formed
would be racemic because racemization must necessarily occur
before the transmetalation/reductive elimination step. Therefore,
the Pd(II) racemization mechanism remains a possibility with
NaOP(O)(OEt)2 or NaHCO3 playing the role of LiBr; however,
no experimental evidence for this proposal has been gathered.
Furthermore, this mechanism seems unlikely due to the weakly
nucleophilic character of the phosphate and bicarbonate anions
compared to bromide in the reported example.
(23) The apparent discrepancy between Table 2, entry 2, and Table 5,
entry 3, can be attributed to the different isolation procedures. Table 2 results
are based on GC and HPLC analysis of aliquots from the same reaction.
Table 5 results were generated from isolated material.
(22) Carey, F. A.; Sundberg, R. J. AdVanced Organic Chemistry, Part
A: Structure and Mechanism, 4th ed.; Kluwer Academic/Plenum Publish-
ers: New York, 2000; p 17.
(24) Jiang, B.; Chen, Z.; Tang, X. Org. Lett. 2002, 4, 3451-3453.
J. Org. Chem, Vol. 71, No. 4, 2006 1567

Molander et al.
Experimental Section
Cross-Coupling Reaction To Form Enantioenriched Al-
lenes: Method A. The alkenyl trifluoroborate, Pd(PPh3)4, and
NaHCO3 were combined in a round-bottom flask. The flask was
evacuated and refilled with nitrogen, and THF (3 mL) and H2O
(200 µL) were added. In a separate flask, the propargyl phosphate
was degassed by evacuation or by the freeze/pump/thaw method
and refilled with nitrogen. The substrate was then dissolved in THF
(1 mL). The substrate solution was transferred via cannula into the
catalyst solution. The flask containing the substrate was washed
with THF (1 mL), which was also transferred to the reaction
mixture. The reaction time was calculated beginning with the first
drop of substrate transferred. At the completion of the reaction,
Darco was added and the flask was left uncapped. After being stirred
for 30 min to 1 h, the mixture was filtered through a short plug of
Celite to remove the Darco. The plug was washed with Et2O. The
solution was transferred to a separatory funnel where the organic
layer was washed three times with brine (1 mL each time). The
ethereal layer was dried with MgSO4, filtered, and concentrated.
The crude product was then purified via column chromatography
with appropriate conditions for each allene product (see the
Supporting Information).
100 °C.12 Alkynyl and alkyl trifluoroborate coupling partners
did not yield any coupled product upon workup. In the case of
the alkyl coupling, almost complete decomposition of the
starting material was observed after 2 h at reflux.
Cross-Coupling Reaction To Form Racemic Allenes from
Carbonates/Phosphates: Method B. The alkenyl trifluoroborate,
Pd(PPh3)4, and NaHCO3 were combined in a round-bottom flask
with a sidearm. The flask was fitted with a reflux condenser and
was evacuated and refilled with nitrogen, and THF (3 mL) and
H2O (200 µL) were added. In a separate flask, the propargyl
carbonate/phosphate was degassed by evacuation or by the freeze/
pump/thaw method and refilled with nitrogen. The substrate was
then dissolved in THF (1 mL). The substrate solution was
transferred via cannula into the catalyst solution. The flask
containing the substrate was washed with THF (1 mL), which was
also transferred to the reaction mixture. The reaction was then
heated to reflux. At the completion of the reaction, Darco was added
and the flask was left uncapped. After being stirred for 30 min to
1 h, the mixture was filtered through a short plug of Celite to remove
the Darco. The plug was washed with Et2O. The solution was
transferred to a separatory funnel where the organic layer was
washed three times with brine (1 mL each time). The ethereal layer
was dried with MgSO4, filtered, and concentrated. The crude
product was then purified via column chromatography under the
appropriate conditions for each allene product (see the Supporting
Information).
Conclusions
Ene-allenes were synthesized in low to high enantiomeric
excess depending on the trifluoroborate coupling partner and
product formed. To the best of our knowledge, this report marks
the first time that ene-allenes have been synthesized asymmetri-
cally. Based on experimental observation of stunted conversion
and rampant loss of ee over time, we propose that racemization
occurs after the allene is formed by an unknown mechanism.
Rapid racemization of electron-poor ene-allene products (i.e.,
alkenes bearing a phenyl substituent) further supports this
proposal because these compounds serve as better ligands for
palladium than ene-allenes bearing hydrogen or methylene
substituents. In addition to synthesizing enantioenriched di-, tri-,
and tetrasubstituted allenes, functional group tolerance was
exhaustively examined. Despite the poor yield of an aldehyde
and the loss of a TMS protecting group, all other cases (nitrile,
alcohol, silyl ether, amine, thioether, and sulfone) provided
moderate to good yields of allenes. Although alkynyl and alkyl
trifluoroborates failed to couple under the reaction conditions,
a chiral nonracemic aryl allene was synthesized in 83% optical
yield. Furthermore, a reproducible workup procedure to remove
the rogue palladium was devised. The continued development
of methods for the synthesis of enantioenriched allenes is
ongoing in our laboratory.
Acknowledgment. We thank the National Institutes of
Health (GM35249), Merck Research Laboratories, and Amgen
for generous support of this work. We thank Johnson-Matthey
for their generous contribution of catalysts.
Supporting Information Available: Experimental details and
structural data for all new compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
JO052201X
1568 J. Org. Chem., Vol. 71, No. 4, 2006