A new class of non-C2-symmetric ligands for oxidative and redox-neutral palladium-catalyzed asymmetric allylic alkylations of 1,3-diketones

Article abstract of DOI:10.1021/jacs.5b00786

We report the discovery, synthesis, and application of a new class of non-C₂-symmetric phosphoramidite ligands derived from pyroglutamic acid for use in both oxidative and redox-neutral palladium-catalyzed asymmetric allylic alkylations of 1,3-diketones. The resulting chiral products are typically obtained in high yield with good to excellent levels of enantioselectivity.

Full text of DOI:10.1021/jacs.5b00786

Article
pubs.acs.org/JACS
A New Class of Non‑C2‑Symmetric Ligands for Oxidative and Redox-
Neutral Palladium-Catalyzed Asymmetric Allylic Alkylations of 1,3-
Diketones
Barry M. Trost,* Etienne J. Donckele, David A. Thaisrivongs, Maksim Osipov, and James T. Masters
Department of Chemistry, Stanford University, Stanford, California 94305-5080, United States
S
* Supporting Information
ABSTRACT: We report the discovery, synthesis, and application
of a new class of non-C2-symmetric phosphoramidite ligands
derived from pyroglutamic acid for use in both oxidative and redox-
neutral palladium-catalyzed asymmetric allylic alkylations of 1,3-
diketones. The resulting chiral products are typically obtained in
high yield with good to excellent levels of enantioselectivity.
breaking or -forming event typically occurs outside the
coordination sphere of the metal and therefore distal to the
chiral ligands.
INTRODUCTION
■
The ability to construct C−C bonds in an asymmetric fashion
continues to challenge synthetic chemists. Catalytic processes
are particularly effective to address this challenge because they
assemble molecules in the most atom-economical way
possible.1 Nevertheless, challenges persist even in such a
developed field, particularly in the area of asymmetric bond
formation. More precisely, the stereoselective construction of
C−C bonds, especially those of all-carbon quaternary stereo-
centers, remains a challenge that continues to receive broad
attention from the organic chemistry community.2
Within the class of asymmetric transformations that can be
utilized to generate C−C bonds, transition-metal-catalyzed
asymmetric allylic alkylation (AAA) reactions, principally those
employing palladium,3 copper,4 iridium,5 or molybdenum,6
have emerged as powerful tools for the synthesis of complex
molecules. The widespread use of AAA reactions speaks to not
only their ability to complete chemo-, regio-, diastereo-, and
enantioselective catalytic processes, but also the broad scope of
substrates that can be employed and the catalysts’ tolerance for
a variety of other functional groups.
There are three general characteristics of AAA reactions that
differentiate them from other asymmetric transition-metal-
catalyzed methods. First, in contrast to processes such as
asymmetric oxidations and reductions, AAA reactions exhibit
several mechanisms through which asymmetric induction can
be realized.7 These include the differentiation of enantiotopic
leaving groups or π-allyl termini and the preferential reaction of
one prochiral face of either the nucleophile or the electrophile.
Second, both reacting partners can simultaneously be prochiral,
allowing for the concurrent formation of multiple stereocenters
using a single catalytic species. Third, the asymmetric bond-
A common design element of widely applicable chiral
catalysts, including those employed in AAA reactions, is a C2-
symmetric scaffold around the metal.8 Conceptually, this style
of ligand is attractive because the inherent symmetry reduces
the number of potential catalyst−substrate arrangements,
simplifying the mechanistic rationale of the transformation
and minimizing the number of possible reaction pathways.
Generally, reducing the number of competing processes that
may lead to enantiomeric products can increase the catalyst’s
enantioselectivity.9 This basic strategy has been leveraged to
create countless metal complexes for asymmetric catalysis,
many of which are based on common ligands derived from
cinchona alkaloids,10 bis(oxazoline)s,11 salens,12 DIOPs,13
BINAPs,14 and BINOLs15 (Figure 1).
Our group was thus attracted to the advantages afforded by
C2 symmetry when we developed diphenylphosphinobenzoic
acid (DPPBA)-based ligands for palladium-catalyzed AAA
reactions (Figure 2). These modular ligands, derived from
derivatization of readily accessed chiral building blocks, have
been used over a broad variety of substrates and trans-
formations that is unmatched by any other family of chiral
catalysts in the field of palladium-catalyzed AAA reactions
(Figure 3).16
However, despite all the progress that has been made in
catalyst design, DPPBA ligands are not universally applicable.
Received: January 23, 2015
Accepted: January 28, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.5b00786
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. Common C2-symmetric ligands used for asymmetric
catalysis.
Figure 4. Comparison of oxidative and redox-neutral palladium-
catalyzed AAA reactions.
redox-neutral AAA reactions, but also would avert the
chemoselectivity issues inherent in having such groups in a
synthetic sequence.
In 2008, Shi and co-workers described the first catalytic
intramolecular allylic alkylation of both activated and
unactivated C−H bonds with β-dicarbonyl derivatives in the
presence of Pd(OAc)2, 1,2-bis(benzylsulfinyl) ethane,18 and
benzoquinone.19 Contemporaneous with that report, White
and Young demonstrated a similar intermolecular allylic C−H
alkylation of activated C−H bonds with methyl nitroacetate,20a
and they later extended that work to an analogous
intermolecular allylic C−H alkylation using unactivated C−H
bonds.20b Recently, Sharma and Hartwig reported a one-pot
process involving an initial Pd-catalyzed oxidation followed by
an in situ Ir-catalyzed allylic alkylation.21 Our group has also
been interested in performing palladium-catalyzed allylic C−H
alkylations, specifically of 1,4-dienes and 1,4-enynes.22 At the
time, however, there were no reports of either stoichiometric or
catalytic asymmetric allylic alkylation reactions that proceeded
via C−H activation. We recently provided a preliminary
communication of the first examples of such a method.23
These reactions construct all-carbon quaternary stereocenters
with high levels of enantiocontrol by utilizing prochiral
nucleophiles, a class of substrates upon which palladium AAA
catalysts often struggle to induce asymmetry.24 To date, there
has been no report wherein the same structural type of chiral
ligand can affect both oxidative and nonoxidative Pd-catalyzed
AAA reactions. Herein, we provide a full account of the design
and development of a new class of non-C2-symmetric chiral
ligands that enabled this advance. Also, we describe the first
direct comparison of the performance of the presumed π-
allylpalladium intermediate in the AAA reaction via oxidative
and redox-neutral reaction manifolds. This new ligand family
has provided an excellent approach to the difficult challenge of
AAA of 1,3-diketones.
Figure 2. C2-symmetric DPPBA ligands.
Figure 3. DPPBA ligands in palladium-catalyzed AAA reactions.
One prominent limitation is that for nearly all substrates that
undergo metal-catalyzed allylic alkylations, a leaving group is
required on the electrophile.17 Such functionality allows the
catalyst to convert one allylic substituent into another via a
redox-neutral event (Figure 4). One way to increase the
efficiency, chemoselectivity, and scope of these reactions would
be to eliminate this limitation by developing an oxidative
method to perform analogous AAA reactions with unfunction-
alized alkenes. The use of a relatively inert allylic hydrogen
atom as a “leaving group” not only would obviate the
installation of the allylic leaving groups necessary for traditional
RESULTS AND DISCUSSION
■
The first report of a nucleophilic addition to a stoichiometri-
cally prepared π-allylpalladium complex was made in 1965 by
Tsuji and co-workers, who demonstrated that sodium diethyl
malonate could be allylated with [(η3-C3H5)PdCl]2 in the
presence of DMSO.25 In 1973, we developed a sequence
wherein an overall allylic alkylation from olefins was achieved
by conversion of the olefin to a π-allylpalladium complex, which
B
DOI: 10.1021/jacs.5b00786
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
then was subjected to nucleophilic attack in the presence of
phosphine ligands.26 Despite nearly half a century of advances
in organometallic chemistry, only recently have there been
reports of methods that reduce this two-step procedure to a
single-step palladium-catalyzed process, and all have employed
sulfoxides as ligands (vide supra). In our early work, we found
that such conditions failed with more substituted allyl
substrates. While studying the alkylation of π-allylpalladium
species, our group discovered that phosphorus-based ligands
proved advantageous for promoting attack by stabilized
nucleophiles on such complexes, a discovery that has provided
the basis of Pd-AAA chemistry ever since.
Considering the success this family of ligands continues to
demonstrate in achieving chemo-, regio-, and stereocontrol in
palladium-catalyzed allylic substitution processes across a wide
array of contexts, we hypothesized that if phosphorus-based
ligands could be made to promote palladium-catalyzed allylic
C−H alkylations then there would be a similarly great
opportunity to dramatically expand both the scope and
selectivity of AAA reactions. At the outset of our research
program, however, there was a report from the White group
that plainly claimed that phosphorus ligands are not compatible
with the oxidative conditions necessary for C−H activation.27
These remarks, however, were made in the context of allylic
acetoxylation reactions (specifically, enantioselective allylic
acetoxylations of terminal alkenes using a heterobimetallic
catalyst system composed of both palladium(II) and
chromium(III) salts), and we immediately questioned whether
there might be meaningful differences between that process and
an analogous allylic alkylation reaction. More precisely, we
postulated that either the mechanism of allylic C−O bond
formation is dissimilar enough from allylic C−C bond
formation or the reaction conditions necessary to achieve
such transformations are distinct enough so that such a broad
statement is not generally applicable. Indeed, under the White
group’s optimized conditions, (E)-methyl 2-nitro-5-phenylpent-
4-enoate (4) is obtained in 62% yield as a 4:1 mixture of
regioisomers, favoring the linear product from the reaction of
methyl 2-nitroacetate (1) with allylbenzene (2) (eq 1). If 1,2-
racemization under the reaction conditions. Accordingly, we
began our exploratory work with 2-acetylcyclopentanone as the
nucleophile for alkylation with allylbenzene (2). The employ-
ment of 2,6-dimethylbenzoquinone (2.6-DMBQ) was neces-
sary to observe any desired reactivity, but the 1.5 equiv reported
by White and Young could be reduced to 1.0 equiv without any
observable decrease in catalytic efficiency. Importantly,
benzoquinone itself was not a competent oxidant for this
process, an observation that the White group has also made.28
The reaction temperature could also be reduced from 45 to 35
°C, which in some cases led to a concomitant increase in
enantioselectivity (vide infra); conducting the alkylation at
ambient temperature generally gave only trace amounts of the
desired product. Lowering the catalyst loading from 10 mol %
Pd(OAc)2 and 20 mol % PPh3 to 5.0 and 7.5 mol %,
respectively, did not result in a diminution in yield, but the
addition of 1.0 equiv of bases such as K2CO3 or NaOAc or
acids such as AcOH was significantly deleterious to the reaction
conversion. On the basis of these preliminary results, we turned
to the key challenge of our program: the design of a chiral
ligand class that would be compatible with the highly oxidizing
allylic C−H alkylation reaction conditions.
Palladium catalysts prepared using our DPPBA ligands are
not suitable for such a process because they undergo facile
oxidation to form stable palladium(II) adducts that are
catalytically inactive (Figure 5).29 A screen of many other
Figure 5. Catalytically inactive DPPBA Pd2+ complex.
known ligands for palladium provided no strong leads (for
examples, see the Supporting Information). However, at the
conception of this research program there was a vigorous and
sustained effort in our laboratories to discover new
phosphoramidite-based catalysts for palladium-catalyzed trime-
thylenemethane cycloadditions.30 Intrigued by the possibility
that such ligands might prove useful in this circumstance, we
discovered a new class of phosphoramidites that support
palladium-catalyzed asymmetric allylic C−H alkylation reac-
tions. These non-C2-symmetric ligands are modular in design
(i.e., each aryl ring and the biaryl backbone can be
independently changed) and are readily accessed from
pyroglutamic acid (Scheme 1). In a representative synthesis,
esterification of l-pyroglutamic acid (5) with 1-naphthalene
methanol using DCC in the presence of catalytic DMAP
followed by carbamate formation with Boc2O affords
pyrrolidine (6) in 63% overall yield. Chemoselective semi-
reduction of the lactam carbonyl with LiBHEt3 and then
treatment with methanol and catalytic p-TsOH provides the
corresponding aminal (7) in 37% yield. Copper(I) bromide
mediated addition of phenylmagnesium bromide in the
presence of BF3·OEt2 gives the desired trans-substituted
pyrrolidine (8) in 83% yield.31 Notably, in every such reaction
we have performed, the diastereocontrol of the cuprate addition
has been complete. Cleavage of the tert-butyl carboxyl group
with TFA liberates the unprotected pyrrolidine (9) in 74%
yield, which undergoes base-promoted coupling with a slight
excess of chlorophosphite (10) to afford representative ligand
L10 in 54% yield.
bis(phenylsulfinyl)ethane palladium acetate (3) is replaced with
Pd(OAc)2, 20 mol % PPh3 is added, and the reaction is
conducted without the addition of AcOH or DMSO, then (E)-
methyl 2-nitro-5-phenylpent-4-enoate (4) is formed in 47%
yield as a single regioisomer (eq 2).
We hypothesized that methyl 2-nitroacetate (1) would not
be an ideal nucleophile for our optimization studies partly
because we anticipated that distinguishing between the similarly
sized ester and nitro substituents of the activated methylene
would be challenging for a chiral catalyst and also because we
were concerned that any alkylation products might succumb to
C
DOI: 10.1021/jacs.5b00786
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 1. Synthesis of Chiral Phosphoramidite Ligands
Based on Pyroglutamic Acid
Although the enantioselective palladium-catalyzed allylic
alkylation of 1,3-diketone nucleophiles has been known since
a 1978 report from Kagan and co-workers,32 highly
enantioselective examples of such a transformation are rare.33
The most notable example of these is a 2003 report from
Kuwano, Uchida, and Ito, who demonstrated that a palladium
catalyst with BINAP as the chiral ligand is able to perform
unsymmetrically the nonoxidative enantioselective allylic
alkylations of various 1,3-diketones to give the corresponding
products in high yields with enantiomeric excesses that range
from 64 to 89%.34 This paucity is likely due to the steric and
electronic similarities of the two carbonyl groups that the
reacting center bears, between which many catalysts struggle to
discriminate.
Taking this class of nucleophiles as a significant challenge to
be met, we decided to evaluate our new set of phosphoramidite
ligands in oxidative palladium-catalyzed allylic C−H alkylations,
employing 2-acetyl-1-tetralone (11) as the nucleophile,
allylbenzene (2) as an electrophile, 2,6-DMBQ as an oxidant,
and Et3N as base in the presence of Pd(OAc)2 (5.0 mol %) and
phosphoramidite ligand (7.5 mol %) in THF at 60 °C for 24 h
(Figure 6). When the pyrrolidine subunit is substituted with a
phenyl ring (L11), allylic C−H alkylation product 12 was
obtained in 83% conversion and 72% ee. Substitution on this
aromatic ring is generally not well tolerated. A reaction with the
corresponding para-biphenyl ligand (L12) gave 12 in only 14%
conversion and 30% ee, and the meta-biphenyl analog (L13)
gave no desired product at all. The corresponding meta-
terphenyl ligand (L14) yielded 12 in 70% ee, but only at 10%
conversion. Further increasing the steric bulk of this substituent
is deleterious because the reaction with the 3,5-di-tert-
butylphenyl ligand (L15) gave no desired product. Neither
the 1-naphthyl (L16) nor 2-naphthyl (L17) ligand offered
improvement in terms of conversion or enantioselectivity, and
replacement of the aromatic ring with the saturated cyclo-
hexane (L18) provided 12 in only 29% conversion and 25% ee.
The effect of the ester substituent of L11 on both the
reactivity and enantioselectivity of the allylic C−H alkylation
Figure 6. Palladium-catalyzed enantioselective allylic C−H alkylations
with selected non-C2-symmetric phosphoramidite ligands. Reactions
were run for 24 h, and conversions to the product are indicated.
was next examined. Replacing the methyl substituent with
either a phenyl (L19) or a 2-naphthyl (L20) group gave
inactive catalysts, presumably because these sterically demand-
ing groups inhibit the C−H activation event. However,
substitution with a benzyl group (L21) yielded 12 in 43%
conversion and 85% ee. With a benzylic 1-naphthyl ester (L10),
the reaction delivered 12 in 54% conversion and 89% ee. The
reaction conversion increased to 100% with the corresponding
benzylic 2-naphthyl ligand (L22), and the desired product was
obtained in 55% ee. Increasing the steric bulk of the aryl group,
as in L23−L26, gave no significant improvement. Interestingly,
switching the methyl group to an adamantyl enhanced the
reactivity significantly and resulted in decent enantioselectivity.
Among the ligands we prepared and evaluated, phosphor-
amidite L10 gave the best combination of yield and
enantioselectivity, and conducting the reaction with 10 mol %
ligand enhanced its reproducibility. A solvent screen demon-
strated that THF was ideal for conducting the transformation,
in terms of both reactivity and enantioselectivity. Varying
concentration and stoichiometry did not provide a significant
benefit to the transformation. Ultimately, it was found that if
the ligand loading was increased to 10 mol %, the temperature
decreased to 50 °C, and the catalyst dosed in two portions at
the beginning of the reaction and after 3 h, then desired
product 12 could be isolated in 89% yield and 85% ee.
We then moved to evaluate the scope of allylarenes that will
undergo reaction with 2-acetyl-1-tetralone (11) (Figure 7). The
para-methyl benzoate electrophile (13) gave the corresponding
product in 83% yield and 69% ee, compared with 89% yield and
85% ee obtained with allylbenzene (2). Neither the para-
methoxy (14) nor para-nitrile (15) substrate reacts to afford
any alkylated product. In the former case, the electron-rich
D
DOI: 10.1021/jacs.5b00786
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
donor, a meta-methoxy is a predominantly σ-withdrawing
group. This result is consistent with the hypothesis that only
moderately electron-rich aromatic rings can be activated at the
benzylic position, and both neutral and strongly electron-
withdrawing aromatic rings are well-tolerated. The 3,5-difluoro
substrate (24) provided the allylated material in 59% yield and
65% ee, and the 2-naphthyl analog (25) performed well to
generate the desired product in 90% yield and 71% ee. With
this catalyst system, it appears that ortho substitution is not well
tolerated for electrophiles such as 26 and 27. Even when the
ortho substituent is only a methyl group (28), the desired
product was isolated in only 36% yield and the stereoselectivity
dropped precipitously. Surprisingly, despite the fact that
aldehydes, esters, and amides are well-tolerated, neither the
para-methyl ketone (29) nor a para-phenyl ketone (30)
substrate gave any product. It is noteworthy to mention that in
all cases examined, the linear regioisomer was obtained in >19:1
selectivity, demonstrating the remarkable regioselectivity of this
allylation process.
Substituted tetralones bearing both an electron-donating
group (6-methoxy) and an electron-withdrawing group (7-
nitro) were examined and delivered cinnamylated products (31
and 32) in comparable yields (82 and 91%, respectively) and
comparable enantioselectivities (78% ee) as illustrated in eq 3.
Figure 7. Enantioselective allylic C−H alkylation of 2-acetyl-1-
tetralones with various allylarene pro-electrophiles. Reactions were
run for 24 h, and isolated yields of the product are indicated.
Because ligand L10 performed so well in the oxidative Pd-
AAA, we were interested in seeing if this particular ligand is also
applicable in the traditional Pd-AAA and whether these two
approaches differ in selectivity−in other words, how identical
are the reactivities of the π-allylpalladium complexes generated
by these two different paradigms? To evaluate the phosphor-
amidite−palladium catalyst for the optimization of traditional
Pd-AAA of 1,3-diketones, 2-acetyl-1-tetralone (11) and
cinnamyl acetate (33) were chosen as model substrates.
Traditional Pd-AAA bidentate ligands L6−L9 (Table 1),
which have enjoyed incontestable success in the field, delivered
product 12 in high yield but with varying levels of
enantioselectivity (Table 1, entries 1−4). Using non-C2-
symmetric phosphoramidate ligand L10 gave dialkyl-1,3-
aromatic ring may sufficiently deactivate the benzylic protons
toward C−H cleavage by substantially decreasing their basicity.
In the latter case, it is possible that the electron-withdrawing
character of the nitrile is not at fault, but rather the ability of its
nitrogen lone pairs to coordinate palladium. Such engagement
likely disrupts both the reactivity and selectivity of the C−H
alkylation process. The para-methyl derivative (16), however,
performs well, giving the corresponding product in 61% yield
and 79% ee. This result suggests that a moderately electron-rich
aromatic ring is compatible with this allylic C−H alkylation, as
long as the electron-rich groups do not approach the donating
ability of a methoxy group. The analogous para-biphenyl
substrate (17) also performs well, giving the desired product in
70% yield and 67% ee, indicating that the reaction tolerates a
fair amount of steric bulk at the aromatic portion of the
electrophile. The para-fluoro (18) and para-trifluoromethyl
(19) substrates give the corresponding products in 75 and 76%
yield and 83 and 66% ee, respectively. Impressively, electro-
philes bearing either an aldehyde (20) or a dimethyl amide
(21) are well-tolerated when such substitution is at the para
position, giving the products in 84 and 85% yield and 71 and
74% ee, respectively. The successful alkylation of 20 under-
scores the remarkably mild reaction conditions and chemo-
selectivity of this allylic C−H alkylation protocol because no
competitive nucleophilic addition into the highly electrophilic
carbonyl is observed, even after prolonged reaction times.
The meta-methyl substrate (22) reacts well, providing the
desired product in 77% yield and 79% ee. Although the para-
methoxy substrate (14) failed to react, the meta-methoxy
substrate (23) gives the alkylated material in 63% yield and
79% ee. Unlike a para-methoxy substituent, which is a strong π-
a
Table 1. Ligand Screen for Palladium-Catalyzed AAA of 11
b
c
entry
ligand
base
yield (%)
ee (%)
1
2
3
4
5
(S,S)-L6
(S,S)-L7
(S,S)-L8
(S,S)-L9
L10
Et3N
Et3N
Et3N
Et3N
Et3N
Et3N
89
84
92
94
87
76
41
61
73
68
83
88
d
6
L10
a
All reactions were conducted at 0.2 M using 1.0 equiv of 11 and 33 in
b
c
0.2 mmol scale. Isolated yield. Enantiomeric excess determined by
chiral HPLC. Reaction run at 50 °C.
d
E
DOI: 10.1021/jacs.5b00786
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
diketone 12 in 86% yield and 83% ee after 12 h (Table 1, entry
5). Surprisingly, the enantiomeric excess showed a slight inverse
temperature effect. No formation of a regioisomeric product
was observed. An extensive screening of solvents showed DME
to be superior both in terms of reactivity and enantioselectivity
(Table 2). Furthermore, the choice of base had a significant
Table 4. Nucleophile Scope of Palladium-Catalyzed Enantio-
Selective Allylic Alkylations
a
a
Table 2. Solvent Screen for Palladium-Catalyzed AAA of 11
b
c
entry
solvent
yield (%)
ee (%)
1
2
3
4
5
THF
94
73
91
78
82
83
81
86
80
78
dioxane
DME
toluene
DCE
a
All reactions were conducted at 0.2 M using 1.0 equiv of 11 and 33 in
b
c
0.2 mmol scale. Isolated yield. Enantiomeric excess determined by
chiral HPLC.
a
Table 3. Base Screen for Palladium-Catalyzed AAA of 11
a
All reactions were conducted using 1.0 equiv of nucleophile and 1.0
equiv of electrophile in 0.1 mmol scale at 0.2 M in DME using 1 eq. of
Li2CO3, 2.5 mol % Pd2(dba)2(CHCl3) and 7.5 mol % L10.
nitro-1-tetralone analogously provided compound 32 in 95%
yield and 81% ee. Electron-donating nucleophiles gave decent
to excellent yields of products 34 and 35 (89 to 91%) with
comparable enantioselectivity (94 and 91%, respectively). The
reaction of a 2-propionyl substrate proceeded with slightly
lower enantioselectivity than that with the 2-acetyl substrate,
delivering product 36. We were pleased to observe similar
enantiodiscrimination for furanyl-fused product 37 and 2-acetyl
cyclohexanone product 38. A five-membered ring system
underwent alkylation with cinnamyl acetate as well, delivering
39 in 96% yield and 84% ee.
b
c
entry
base
Et3N
yield (%)
ee (%)
1
2
3
4
5
6
7
8
9
50 °C
room temperature
4 °C
−20 °C
4 °C
4 °C
4 °C
4 °C
4 °C
91
90
90
89
77
75
90
82
90
86
86
89
89
78
8
Et3N
Et3N
Et3N
DBU
Hunig’s base
Cs2CO3
Na2CO3
Li2CO3
81
90
92
a
All reactions were conducted at 0.2 M using 1.0 equiv of 11 and 33 in
0.2 mmol scale. Isolated yield. Enantiomeric excess determined by
Different electrophiles were also investigated to emphasize
the extension of the asymmetric allylation of unsymmetric 1,3-
diketones (Table 5). To test the scope of allyl electrophiles
b
c
chiral HPLC.
impact on the enantiomeric excess (Table 3). In particular,
employing Li2CO3 as the base proved optimal and delivered
desired product 12 in 90% yield and 92% ee; the advantageous
nature of Li+ is further noted by the use of cinnamyl t-
butylcarbonate as the substrate, as illustrated in eq 4. In this
Table 5. Electrophile Scope of Palladium-Catalyzed
Enantioselective Allylic Alkylations
a
case, the leaving group serves as the base in which the only
counterion is the Pd. In the absence of a counterion other than
the Pd, the enantiomeric excess turns out to be slightly lower
than that obtained when triethylamine was used as base.
With the optimized reaction conditions in hand (Table 3,
entry 9), a variety of 1,3-diketones were subjected to Pd-AAA
(Table 4). In general, good to excellent yields and
enantioselectivity were obtained. Acetyl-6-methoxy-1-tetralone
gave compound 31 in 78% yield and 90% ee, and 2-acetyl-7-
a
All reactions were conducted using 1.0 equiv of nucleophile and 1.0
equiv of electrophile in 0.1 mmol scale at 0.1 M in DME.
F
DOI: 10.1021/jacs.5b00786
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
under the optimized conditions, we synthesized a variety of
cinnamyl acetate derivatives via the Horner−Wadsworth−
Emmons reaction as a key step in a three-step−one-purification
protocol. Electron-donating substituents on the aromatic ring
of the cinnamyl group gave excellent yields (85−97%) with
high enantioselectivity in the range of 90%. Moderately
electron-rich substrates are tolerated because compound 16
was isolated in 80% yield and 90% ee and compound 25 was
obtained with 82% yield and 91% ee. Diverse electron-rich
aromatic systems such as thiophene (42) or furan (43)
derivatives are also tolerated.
Oxidative and traditional Pd-AAA could be utilized in this
transformation to afford the alkylated products in decent yields
and enantiopurity. In general, Pd-AAA via C−X activation gave
both a higher reactivity and a higher enantioselectivity (Table
6). The most significant exception is the alkylation with a 2-
achiral ligand under the oxidative process that competes with
the chiral phosphine ligand. The idea that quinones can serve as
ligands for Pd is well documented.36 We do note that for
traditional AAA we utilized a dba complex for our Pd(0)
source. Because dba can serve as a ligand during alkylation, we
performed the reaction described in Table 6 entry 5 under the
c
conditions described in Table 6 footnote , but we replaced
Pd2(dba)3·CHCl3 with CpPd(allyl), which has only the
phosphine as a ligand. The enantiomeric excess under these
revised conditions was the same, demonstrating that dba is not
involved in the catalytic cycle. Thus, the current results support
both a role for the cation associated with the nucleophilic anion
and the quinone being responsible for the differences in
enantiomeric excess between the oxidative and nonoxidative
processes involving the same π-allylpalladium core. An attempt
to see if the cation effect impacts the enantiomeric excess under
oxidative conditions failed because addition of lithium salts to
such reactions gave mainly decomposition and very low
conversion to allylated products.
Table 6. Comparison of Reactivity and Enantioselectivity
between Oxidative and Traditional Pd-AAA
CONCLUSIONS
■
The stability of the phosphorus toward the oxidant in these
reactions is curious. That quinones can oxidize trivalent
phosphorus compounds has been noted by Mukaiyama, who
employed them as the oxidant in a Mitsunobu reaction.35
Phosphoramidites, particularly, would be thought to be
susceptible to such oxidants. Presumably, the success of these
reactions derives from the coordination of the quinones as
ligands to the palladium. Such coordination is well-documented
and is the reason that quinones are particularly effective at
oxidizing low-valent Pd to higher-valent Pd.37 Further evidence
for such coordination does occur is illustrated by a control
experiment. When we performed the nonoxidative AAA in the
presence of the quinone, reaction stopped at 13% conversion.
This observation is consistent with the quinone coordinating
the Pd making it effectively a Pd(+2) complex which is not
catalytically active for the traditional AAA process.
The design of a novel class of non-C2-symmetric
pyroglutamic acid-derived phosphoramidite ligands has led us
to the development of the first examples of enantioselective
allyl C−H activation as well as the report of a traditional Pd-
AAA of 1,3-diketones, a most challenging pro-chiral nucleo-
phile. To our delight, the reaction, typically, provided chiral 2,2-
dialkyl-1,3-diketones in high yields (25 examples) in 65−89%
ee for the oxidative process and 81−94% ee for the
nonoxidative process. The broad substrate scope and mild
conditions make this procedure a valuable reaction for the
synthesis of unsymmetric 1,3-diketones. The enantioselectivity
of our nonoxidative are somewhat higher than the range of 64−
89% ee reported by Kuwano et al. Furthermore, they required
−60 °C; whereas, we operate at 4 °C. Remarkably, our
oxidative process exhibits a range similar to that of Kuwano et
al. even though our reaction temperature is 50 °C and theirs is
−60 °C. This study is the first to compare directly the effect of
method of generation of the π-allylpalladium intermediate on
ee. As Table 6 illustrates, higher enantiomeric excess is typically
obtained via the nonoxidative method compared to the
oxidative method. The source of this difference appears, most
likely, to derive from the change in the counterion for the
nucleophile (Table 3) combined with the quinone serving as an
achiral ligand that modifies the chiral environment around the
Pd. This striking difference suggests that the C−H activation in
the oxidative process is the rate-determining step. Further
a
b
Performed under conditions illustrated in Figure 6. Performed
c
under conditions illustrated in Table 4. Performed under conditions
illustrated in Figure 6, except using 2.5 mol % Pd2(dba)3·CHCl3, no
quinone and the appropriate cinnamyl acetate derivative.
napthalene substituent (Table 6, entry 6). A yield of 92% and
71% ee are observed with oxidative Pd-AAA, whereas a yield of
82% and an enantioselectivity of 91% were obtained with
traditional Pd-AAA. To some extent, the differences may stem
from a cation effect because an ion pair is the actual
nucleophile.
Support for such a conjecture derives from our studies on β-
ketoesters where the choice of cation had a large effect on
enantiomeric excess under otherwise identical conditions.35 To
examine this question, we studied traditional Pd-AAA under the
conditions of the oxidative process. Interestingly, the results
came midway between the optimized conditions for each. A
possible cause for the lowering of the enantiomeric excess for
the oxidative reaction could be the effect of the quinone as an
G
DOI: 10.1021/jacs.5b00786
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Chem. Rev. 2006, 106, 3561−3651. (d) Dagorne, S.; Bellemin-
studies to improve the selectivity and scope of the reaction as
well as the broader applicability of this new ligand type are
ongoing in our laboratory.
Laponnaz, S.; Maisse-Franco̧ is, A. Eur. J. Inorg. Chem. 2007, 913−925.
(e) Rasappan, R.; Lavetine, D.; Reiser, O. Coord. Chem. Rev. 2008, 252,
702−714. (f) Desimoni, G.; Faita, G.; Jørgensen, K. A. Chem. Rev.
2011, 111, PR284−PR437.
ASSOCIATED CONTENT
■
(12) (a) Katsuki, T. Synlett 2003, 281−297. (b) McGarrigle, E. M.;
Gilheany, D. G. Chem. Rev. 2005, 105, 1563−1602.
S
* Supporting Information
General remarks, experimental procedures, and H and 13C
1
(13) Tang, W.; Zhang, X. Chem. Rev. 2003, 103, 3029−3070.
NMR spectra. This material is available free of charge via the
Internet at http://pubs.acs.org.
(14) Borner, A., Ed. Phosphorus Ligands in Asymmetric Catalysis;
̈
Wiley-VCH: Weinheim, Germany, 2008.
(15) Zhou, Q.-L., Ed. Privileged Chiral Ligands and Catalysts; Wiley-
VCH: Weinheim, Germany, 2011.
AUTHOR INFORMATION
Corresponding Author
*bmtrost@stanford.edu
Notes
■
(16) (a) Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96, 395−
422. (b) Trost, B. M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921−
2944. (c) Trost, B. M. J. Org. Chem. 2004, 69, 5813−5837. (d) Lu, Z.;
Ma, S. Angew. Chem., Int. Ed. 2008, 47, 258−297. (e) Trost, B. M.;
Zhang, T.; Sieber, J. D. Chem. Sci. 2010, 1, 427−440. (f) Trost, B. M.
Org. Process Res. Dev. 2012, 16, 185−194.
(17) Palladium-catalyzed AAA reactions with allene electrophiles
proceed through allene hydropalladation rather than ionization of an
allylic leaving group. For examples, see (a) Trost, B. M.; Jakel, C.;
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the NSF (grant CHE-0948222) for their generous
support of our programs. D.A.T. acknowledges support from a
Stanford Graduate Fellowship (The William K. Bowes, Jr.
Foundation) and an Eli Lilly and Company Graduate
Fellowship. M.O. acknowledges support from a Stanford
Graduate Fellowship and a John Stauffer Memorial Fellowship.
̈
Plietker, B. J. Am. Chem. Soc. 2003, 125, 4438−4439. (b) Trost, B. M.;
Xie, J. J. Am. Chem. Soc. 2006, 128, 6044−6045.
(18) Chen, M. S.; White, M. C. J. Am. Chem. Soc. 2004, 126, 1346−
1347.
(19) Lin, S.; Song, C.-X.; Cai, G.-X.; Wang, W.-H.; Shi, Z.-J. J. Am.
Chem. Soc. 2008, 130, 12901−12903.
(20) (a) Young, A. J.; White, M. C. J. Am. Chem. Soc. 2008, 130,
14090−14091. (b) Young, A. J.; White, M. C. Angew. Chem., Int. Ed.
2011, 50, 6824−6827.
REFERENCES
■
(1) (a) Trost, B. M. Science 1991, 254, 1471−1477. (b) Trost, B. M.
Acc. Chem. Res. 2002, 35, 695−705.
(21) Sharma, A.; Hartwig, J. K. J. Am. Chem. Soc. 2013, 135, 17983−
17989.
(2) (a) Fuji, K. Chem. Rev. 1993, 93, 2037−2066. (b) Corey, E. J.;
Guzman-Perez, A. Angew. Chem., Int. Ed. 1998, 37, 388−401.
(c) Christoffers, J.; Mann, A. Angew. Chem., Int. Ed. 2001, 40,
4591−4597. (d) Denissova, I.; Barriault, L. Tetrahedron 2003, 59,
10105−10146. (e) Douglas, C. J.; Overman, L. E. Proc. Natl. Acad. Sci.
U.S.A. 2004, 101, 5363−5367. (f) Christoffers, J., Baro, A., Eds.
Quaternary Stereocenters: Challenges and Solutions for Organic Synthesis;
Wiley-VCH: Weinheim, Germany, 2005. (g) Trost, B. M.; Jiang, C.
Synthesis 2006, 369−396. (h) Wang, B.; Tu, Y. Q. Acc. Chem. Res.
2011, 44, 1207−1222. (i) Shimizu, M. Angew. Chem., Int. Ed. 2011, 50,
5998−6000.
(22) (a) Trost, B. M.; Hansmann, M. M.; Thaisrivongs, D. A. Angew.
Chem., Int. Ed. 2012, 51, 4950−4953. (b) Trost, B. M.; Thaisrivongs,
D. A.; Hansmann, M. M. Angew. Chem., Int. Ed. 2012, 51, 11522−
11526.
(23) Trost, B. M.; Thaisrivongs, D. A.; Donckele, E. J. Angew. Chem.,
Int. Ed. 2013, 52, 1523−1526.
(24) Trost, B. M.; Machacek, M. R.; Aponick, A. Acc. Chem. Res.
2006, 39, 747−760.
(25) Tsuji, J.; Takahashi, H.; Morikawa, M. Tetrahedron Lett. 1965, 6,
4387−4388.
(3) (a) Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96, 395−
422. (b) Trost, B. M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921−
2944. (c) Trost, B. M. J. Org. Chem. 2004, 69, 5813−5837. (d) Lu, Z.;
Ma, S. Angew. Chem., Int. Ed. 2008, 47, 258−297. (e) Trost, B. M.;
Zhang, T.; Sieber, J. D. Chem. Sci. 2010, 1, 427−440. (f) Trost, B. M.
Org. Process Res. Dev. 2012, 16, 185−194.
(26) (a) Trost, B. M.; Fullerton, T. J. J. Am. Chem. Soc. 1973, 95,
292−294. (b) Trost, B. M.; Weber, L.; Strege, P. E.; Fullerton, T. J.;
Dietsche, T. J. J. Am. Chem. Soc. 1978, 100, 3416−3426.
(27) Covell, D. J.; White, M. C. Angew. Chem., Int. Ed. 2008, 47,
6448−6451.
(28) The increased steric bulk of 2,6-dimethylbenzylquinone
compared with that of benzoquinone has been postulated by White
and co-workers to slow the irreversible conjugate addition of “soft”
nucleophiles into its electrophilic π-system. In our hands, when
benzoquinone was employed as the stoichiometric terminal oxidant,
the typically bright-yellow reaction solution turned black within
minutes of heating, and no desired product was observed in the crude
reaction mixture.
(4) (a) Yorimitsu, H.; Oshima, K. Angew. Chem., Int. Ed. 2005, 44,
4435−4439. (b) Alexakis, A.; Backvall, J. E.; Krause, N.; Pam
̀
ies, O.;
̈
Dieg
́
uez, M. Chem. Rev. 2008, 108, 2796−2823. (c) Falciola, C. A.;
Alexakis, A. Eur. J. Org. Chem. 2008, 22, 3765−3780.
(5) (a) Helmchen, G.; Dahnz, A.; Dubon, P.; Schelwies, M.;
̈
Wiehofen, R. Chem. Commun. 2007, 675−691. (b) Hartwig, J. F.;
Stanley, L. M. Acc. Chem. Res. 2010, 43, 1461−1475.
(6) Trost, B. M. Org. Process Res. Dev. 2012, 16, 185−194.
(7) (a) Jacobsen, E. N., Pfaltz, A., Yamamoto, H., Eds. Comprehensive
Asymmetric Synthesis; Springer: Berlin, 1999. (b) Ojimi, I., Ed. Catalytic
Asymmetric Synthesis, 3rd ed.; Wiley: Hoboken, NJ, 2010.
(8) (a) Yoon, T. P.; Jacobsen, E. N. Science 2003, 299, 1691−1693.
(b) Zhou, Q.-L., Ed. Privileged Chiral Ligands and Catalysts; Wiley-
VCH: Weinheim, Germany, 2011.
(29) (a) Trost, B. M.; Breit, B.; Organ, M. G. Tetrahedron Lett. 1994,
35, 5817−5820. (b) Swanson, R. A.; Patrick, B. O.; Ferguson, M. J.;
Daly, C. J. A. Inorg. Chim. Acta 2007, 360, 2455−2463. (c) Tsarev, V.
N.; Wolters, D.; Gais, H.-J. Chem.Eur. J. 2010, 16, 2904−2915.
(30) (a) Trost, B. M.; Stambuli, J. P.; Silverman, S. M.; Schworer, U.
̈
J. Am. Chem. Soc. 2006, 128, 13328−13329. (b) Trost, B. M.;
Silverman, S. M. J. Am. Chem. Soc. 2010, 132, 8238−8240. (c) Trost,
B. M.; Bringley, D. A.; Silverman, S. M. J. Am. Chem. Soc. 2011, 133,
7664−7667. (d) Trost, B. M.; Silverman, S. M.; Stambuli, J. P. J. Am.
Chem. Soc. 2011, 133, 19483−19497. (e) Trost, B. M.; Silverman, S.
M. J. Am. Chem. Soc. 2012, 134, 4941−4954. (f) Trost, B. M.; Bringley,
D. A.; Seng, P. S. Org. Lett. 2012, 14, 234−237. (g) Trost, B. M.; Lam,
T. M. J. Am. Chem. Soc. 2012, 134, 11319−11321. (h) Trost, B. M.;
Maruniak, A. Angew. Chem., Int. Ed. 2013, 52, 6262−6264.
(9) Whitesell, J. K. Chem. Rev. 1989, 89, 1581−1590.
́
(10) (a) Kacprzak, K.; Gawronski, J. Synthesis 2001, 961−998.
(b) Song, C. E., Ed. Cinchona Alkaloids in Synthesis and Catalysis:
Ligands, Immobilization and Organocatalysis; Wiley-VCH: Weinheim,
Germany, 2009.
(11) (a) Ghosh, A. K.; Mathivanan, P.; Cappiello, J. Tetrahedron:
Asymmetry 1998, 9, 1−45. (b) McManus, H. A.; Guiry, P. J. Chem. Rev.
2004, 104, 4151−4202. (c) Desimoni, G.; Faita, G.; Jørgensen, K. A.
H
DOI: 10.1021/jacs.5b00786
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(31) (a) The relative stereochemistry of these transformations was
assigned by analogy to reactions described by Wistrand and Skrinjar.
Under this set of reaction conditions, trans-substituted pyrrolidines are
always the major products; see Wistrand, L.-G.; Skrinjar, M.
Tetrahedron 1991, 47, 573−582. Similarly trans-selective cuprate
additions have been observed by many groups. For examples, see
(b) Collado, I.; Ezquerra, J.; Pedregal, C. J. Org. Chem. 1995, 60,
5011−5015. (c) Tong, Y.; Fobian, Y. M.; Wu, M.; Boyd, N. D.;
Moeller, K. D. J. Org. Chem. 2000, 65, 2484−2493. (d) Vaswani, R. G.;
Chamberlin, A. R. J. Org. Chem. 2008, 73, 1661−1681.
(32) Fiaud, J. C.; de Gournay, A. H.; Larcheveque, M.; Kagan, H. B. J.
Organomet. Chem. 1978, 154, 175−185.
(33) (a) Hayashi, T.; Kanehira, K.; Hagihara, T.; Kumada, M. J. Org.
Chem. 1988, 53, 113−120. (b) Sawamura, M.; Nagata, H.; Sakamoto,
H.; Ito, Y. J. Am. Chem. Soc. 1992, 114, 2586−2592. (c) Fukuda, Y.;
Kondo, K.; Aoyama, T. Tetrahedron Lett. 2007, 48, 3389−3391.
(34) Kuwano, R.; Uchida, K.; Ito, Y. Org. Lett. 2003, 5, 2177−2179.
(35) Trost, B. M.; Schaffner, B.; Osipov, M.; Wilton, D. A. Angew.
Chem., Int. Ed. 2011, 50, 3548−3551.
(36) Mukaiyama, T.; Kuroda, K.; Maruyama, Y. Heterocycles 2010, 80,
63−82.
(37) For examples, see (a) Baeckvall, J. E.; Nystrom, J. E.; Nordberg,
R. E. J. Am. Chem. Soc. 1985, 107, 3676−3686. (b) Anderson, P. G.;
Baeckvall, J. E. J. Am. Chem. Soc. 1992, 114, 8696−8698. (c) Boelle, M.
D. K.; Van Strijdonok, G. P. F.; De Vries, A. H. M.; Kamer, P. C. J.; De
Vries, J. G.; van Leeuwen, P. W. N. M. J. Am. Chem. Soc. 2002, 124,
1586−1587. (d) Lee, G. T.; Jiang, X.; Prasad, K.; Repre, O.; Blacklock,
T. J. Adv. Synth. Catal. 2005, 347, 1921−1924. (e) Young, A. J.; White,
M. C. J. Am. Chem. Soc. 2008, 130, 14090−14091. (f) Wang, D.H.;
Wasa, M.; Giri, R.; Yu, J.-Q. J. Am. Chem. Soc. 2008, 130, 7190−7191.
(g) Houlden, C.E.; Hutchby, M.; Bailey, C.D.; Ford, J.G.; Tyler,
S.N.G.; Gagne, M.R.; Lloyd-Jones, G.C.; Booker-Milbum, K.I. Angew.
Chem., Int. Ed. 2009, 48, 1830−1833. (h) Nishika, T.; Abelia, A.L.;
Huang, S.; Lipshutz, B.H. J. Am. Chem. Soc. 2010, 132, 4978−4979.
(i) Wang, D.H.; Mei, T.-S.; Yu, J.-Q. J. Am. Chem. Soc. 2011, 133,
18183−18193. (j) Trost, B.M.; Hansmann, M.M.; Thaisrivongs, D.A.
Angew. Chem., Int. Ed. 2012, 51, 4950−4953.
I
DOI: 10.1021/jacs.5b00786
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX