Development of chiral bis-hydrazone ligands for the enantioselective cross-coupling reactions of aryldimethylsilanolates

Article abstract of DOI:10.1021/jo502388r

A palladium-catalyzed, enantioselective, aryl-aryl cross-coupling reaction using 1-naphthyldimethylsilanolates and chiral bis-hydrazone ligands has been developed. A family of glyoxal bis-hydrazone ligands containing various 2,5-diarylpyrrolidine groups was prepared to evaluate the influence of ligand structure on the rate and enantioselectivity of the cross-coupling. New synthetic routes to the 1-amino-2,5-diarylpyrrolidines were developed to enable the structure/reactivity-selectivity studies. Role reversal experiments of aryldimethylsilanolates and aryl bromides result in biaryl products with the same configuration and similar enantioselectivities implying that reductive elimination is the stereodetermining step. The origin of stereoselectivity is rationalized through computational modeling of diarylpalldium(II) complex which occurs through a conrotatory motion for the two aryl groups undergoing C-C bond formation.

Full text of DOI:10.1021/jo502388r

Article
pubs.acs.org/joc
Development of Chiral Bis-hydrazone Ligands for the
Enantioselective Cross-Coupling Reactions of
Aryldimethylsilanolates
,†
†
‡
‡
Scott E. Denmark,* Wen-Tau T. Chang, K. N. Houk, and Peng Liu
†
Roger Adams Laboratory, Department of Chemistry, University of Illinois, Urbana, Illinois 61801, United States
Department of Chemistry and Biochemistry, University of California, Los Angeles, California 90095, United States
‡
*
S Supporting Information
ABSTRACT: A palladium-catalyzed, enantioselective, aryl−
aryl cross-coupling reaction using 1-naphthyldimethylsilano-
lates and chiral bis-hydrazone ligands has been developed. A
family of glyoxal bis-hydrazone ligands containing various 2,5-
diarylpyrrolidine groups was prepared to evaluate the influence
of ligand structure on the rate and enantioselectivity of the
cross-coupling. New synthetic routes to the 1-amino-2,5-diaryl-
pyrrolidines were developed to enable the structure/reactivity−
selectivity studies. Role reversal experiments of aryldimethylsilanolates and aryl bromides result in biaryl products with the same
configuration and similar enantioselectivities implying that reductive elimination is the stereodetermining step. The origin of
stereoselectivity is rationalized through computational modeling of diarylpalldium(II) complex which occurs through a
conrotatory motion for the two aryl groups undergoing C−C bond formation.
1
. INTRODUCTION
Recent years have witnessed a growing interest in the design
of chiral ligands to facilitate catalytic enantioselective biaryl
coupling. The key developments and mechanistic studies of
catalytic enantioselective cross-coupling are briefly discussed in
the following section.
The importance of chiral biaryl compounds cannot be
overstated; this motif is the central feature of many ligands and
catalysts. For example, phosphoramidites derived from BINOL
have been applied to asymmetric conjugate addition and allylic
amination reactions (Figure 1). A large family of binaphthyl-
based ammonium salts has been employed for asymmetric phase-
transfer catalysis. The chiral biaryl structural motif is also
found in numerous natural products such as michellamine B,
steganacin, and vancomycin.
1
1
b
2
. BACKGROUND
1c
2.1. Evolution of Catalytic Asymmetric Aryl−Aryl
2
3
Coupling. The use of chiral ligands to control enantioselective
aryl−aryl coupling was first reported in 1975 using aryl Grignard
4
5
13
The importance of this structural motif has led to the develop-
reagents, aryl halides, and nickel catalysts. After more than a
decade, the enantioselectivity of the reaction was substantially
improved by tuning the reaction conditions and using mono-
dentate phosphine ligand 1 (Figure 2) with a ferrocene back-
2
ment of many methods for the preparation of chiral biaryls.
Some common methods that have been applied to the synthesis
of natural product include (1) oxidative coupling of aryl-
cyanocuprates using a chiral tether derived from tartaric
acid, (2) Ullmann coupling or S Ar reaction mediated by an
o-oxazoline derived from valinol, (3) dynamic kinetic resolution
of configurationally labile lactones, (4) oxidative homocoupling
of naphthol derivatives catalyzed by a diaza-cis-decalin copper(II)
complex, and (5) chiral Brønsted acid catalyzed [3,3]-sigmatropic
14
bone.
6
The first, catalytic, enantioselective Suzuki−Miyaura coupling,
N
7
15
was reported in 2000. The dimethylamino-substituted
8
ferrocenyl phosphine 2 is more efficient at stereocontrol than
methoxy analogue 1. The KenPhos system reported in the same
2c,9
16
year also features the dimethylamino functionality. Impor-
1
0
tantly, aryl substrates with polar functionalities such as phos-
phonate and nitro at the ortho position are demonstrated for the
first time. In addition, non-naphthyl-derived substrates such as
rearrangement of diarylhydrazines.
Palladium-catalyzed cross-coupling reactions are arguably the
most widely used method for the construction of aryl−aryl bond.
However, the corresponding enantioselective reaction has not
reached maturity as evident in their limited application in total
2
-substituted phenylboronic acids and halides became competent
coupling partners. The scope of this method has been sub-
17
2b,11
sequently expanded to 2-halobenzamides. Comparable or
improved selectivity for the preparation of biarylphosphonates
synthesis.
Notable examples include the formation of the
chiral biaryl units in (−)-steganone, vancomycin and rupens-
amines A and B which have been prepared by the coupling of
chiral tricarbonylchromium complex of aryl halides in high
Received: October 17, 2014
12
diastereoselectivity.
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Figure 1. Examples of ligands and natural products containing a chiral biaryl motif.
Figure 2. Representative ligands that have been used in enantioselective aryl−aryl coupling reactions.
has been reported with the use of monophosphine ligands 3
also been employed. Other important developments that have
emerged in this research area include a recyclable chiral catalyst
1
8
and 4.
In 2008, Fernan
hydrazone ligand 5 for the cross-coupling of arylboronic
21
22h
́
dez et al. reported C -symmetric bis-
on a polymer support, catalysis by palladium nanoparticles,
2
25
and C−H activation.
1
9
acids. Excellent enantioselectivities are achieved for a number
of biaryls when the reactions are conducted at 20 °C, albeit at the
expense of reaction time (7 days). The conversion can be
accelerated at 80 °C (<17 h) with some erosion in selectivity.
Noting the relatively low catalytic activity at room temperature
and the limited scope in this work, the same group designed a
2.2. Mechanistic Hypotheses of the Stereodetermining
Step and the Origin of Enantioselectivity. Despite the
preparative advances in asymmetric biaryl, insights required to
decipher the stereodetermining step (SDS) are usually not
available. As such, the knowledge behind the stereocontrol im-
posed by chiral ligands often remains speculative except in a few
17,18a,20
novel P/N-hybrid ligand 6 derived from C -symmetric 2,5-
cases.
Hayashi et al. employed the role reversal experi-
2
20
bis(isopropylamino)pyrrolidine.
ment of cross-coupling partners to probe the SDS. In the pre-
paration of 2-methyl-1,1′-binaphthalene a very different stereo-
chemical outcome is observed when the 2-methyl sub-
stituent arises from the Grignard reagent rather than from the
Catalytic systems that maintain high enantioselectivity for
both polar and nonpolar coupling partners are rare. One notable
success employs polymer-supported chiral imidazoindole
21
14
phosphine 7 (Figure 2). For example, the highly hindered
but nonpolar 2,2′-dimethylbinaphthalene is obtained in 95%
yield and 97:3 er. The less hindered but more polar 2-methyl-1-
bromide (Scheme 1). Under the assumption that the
diarylnickel(II) intermediate does not undergo racemization,
the SDS is thought to be transmetalation.
(
2-nitrophenyl)naphthalene is obtained in 96% yield and 96:4 er.
Catalytic, enantioselective cross-coupling reactions have been
In a Suzuki−Miyaura coupling using (R)-BINAP as the chiral
22c
ligand, transmetalation is also suggested to be the SDS.
1
8−22
23
24
accomplished for arylboron,
reagents. Although many systems involve phosphine-derived
chiral ligands, diene, oxazoline,
zone−phosphine hybrid ligands and helical polymers have
-zinc, and -indium
Altering the identity of the boron reagent from pinacol ester to
ethylene glycol ester leads to 2,2′-dimethylbinaphthalene with
the opposite configuration as the major product, although
the enantioselectivity in both reactions are low. It should be
22i
22g,25
19
bis-hydrazone, hydra-
20
22l
B
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 1
ability to isolate the pretransmetalation intermediate, it was of
interest to systematically examine the ligand effects on the
enantioselectivity of the coupling and correlate those trends with
structural variations. In addition to experimental interrogation,
we also planned a computational study to investigate the origin of
enantioselectivity.
3
. RESULTS
.1. Preparation of Bis-hydrazone Ligands. 3.1.1. Glyox-
3
al-Derived Bis-hydrazones. To date, only a limited number of
chiral glyoxal bis-hydrazones have been prepared (Figure 3).
Among these variants, ligand 5 featuring (2S,5S)-diphenylpyrro-
mentioned that the interpretation of SDS can be complicated by
the palladium/ligand ratio.
28
22a
The SDS for the coupling promoted by phosphine−hydrazone
lidine moiety has shown promise in enantioselective biaryl
hybrid ligand 6 was inferred from the configuration stability of
19
20
coupling. Despite the report of this ligand and its potential
oxidative addition complex (Scheme 2). The 2-methoxy-1-
naphthylpalladium bromide complex exists as a 7:1 mixture of
atropisomers. Treatment of this complex with 1-naphthylboronic
acid affords the coupled product in 80:20 er at room temperature
similar to that obtained from the catalytic reaction (83:17). The
stereochemical outcome of this coupling appears to be mainly
controlled by oxidative addition. In contrast, free C−Pd bond
rotation is observed for the less hindered 1-naphthylpalladium
bromide complex. These results imply that the stereodetermin-
ing step may be substrate dependent in this system.
28d
modularity almost a decade ago, no analogues with other 2,5-
disubstituted pyrrolidines have appeared. A possible explanation for
this lack of development is the challenge associated with the synthesis
of the requisite building blocks, an analysis of which follows.
The first route considered was that reported for the synthesis
of (2S,5S)-diphenylpyrrolidine bis-hydrazone 5 (Scheme 3,
28d
route a). The key step is the enantio- and diastereoselective
reduction of 1,4-diphenyl-1,4-butanedione by the Corey−Itsuno
29
protocol. The scope of the diketone amenable to this reduction
has not been extensively explored, and it is not applicable to the
2
9
2
-naphthyl substituted diketone. Although these issues may be
Scheme 2
30
addressed by the use of a chiral cobalt catalyst, the repro-
3
1
ducibility has been questioned.
Noting the inability to install a 2-naphthyl moiety at the 2- and
-positions of the pyrrolidine in a stereoselective manner, an
5
approach based on sequential allylic amination and ring-closing
31
metathesis (RCM) has been introduced (route b). Mild, albeit
specialized, conditions are needed for the reduction of 2,5-diaryl-
2
,5-dihydro-1H-pyrrole to avoid cleavage of allylic C−N bond. It
is unclear if this route can be generalized for other aromatic
variants as only one example is reported.
A detailed computational study on the origin of enantiose-
lectivity promoted by KenPhos (Figure 2) has been carried out
The third disconnection is based on the α-arylation of pro-
tected pyrrolidine, which consists of three transformations:
1
7
by Buchwald et al. These authors assumed that reductive elimi-
nation is the SDS and then calculated a number of diary-
lpalladium transition structures from the crystallographic coor-
dinates of the corresponding oxidative addition adduct. The most
stable transition structure correctly predicts the configuration of
the major biaryl product.
2
3
enantioselective deprotonation, transmetalation, and sp −sp
32
coupling (route c). This strategy has been applied to the
33
preparation of a number of chiral phosphoramidite ligands.
Although the most expedient approach to build up pyrrolidine
ring is through [3 + 2]-cycloaddition (route d and e), a reliable
and stereoselective method for the preparation of 2,5-diaryl
Despite the significant progress of catalytic, enantioselective,
biaryl coupling reactions in the past two decades, a general cross-
coupling process to access a range of chiral biaryls has yet to
be developed. In continuation of our interest in the use of
organosilanols and their derivatives in cross-coupling reactions,
we considered the use of arylsilicon reagents for this purpose.
The competency of an aryldimethylsilanolate unit to transfer a
variety of aromatic groups to palladium has been reported from
34
substituted analogues has not been developed.
This analysis suggests that α-arylation (route c) is the only
approach that has demonstrated generality for the synthesis of
33
nonracemic 2,5-diarylpyrrolidines. However, the critical bis-
hydrazone ligand precursor, 1-amino-2,5-diarylpyrrolidine 8
cannot be obtained directly; therefore, this intermediate needs
to be accessed through one of the three following methods:
26
35
these laboratories. In addition, the ability to isolate silanolate
complexes of palladium has facilitated our understanding of the
Hofmann-type rearrangement of the 1-carbamoylpyrrolidine,
36
direct amination of pyrrolidine, or reduction of 1-nitro-
2
7
37
catalytic cycle focusing on the transmetalation step. Given the
sopyrrolidine (Scheme 4).
Figure 3. Chiral glyoxal bis-hydrazone ligands.
C
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 3
Scheme 4
Because of the short synthesis by route (a) and the generality
of α-arylation by route (c), these two synthetic plans were
investigated for the preparation of chiral glyoxal bis-hydrazone
ligands with 2,5-diarylpyrrolidine substituents.
enantioselectivities were maintained and higher diastereoselec-
tivities were uniformly observed. Notably, electron-deficient diol
10d was obtained in greater than 99:1 enantiopurity and the
diastereoselectivity was improved from 63:37 to 79:21 (entry 4).
In general, the dr could be upgraded to at least 98:2 by recry-
stallization except in the case of diol 10b, in which the meso
isomer was more crystalline. Analysis of the mother liquor
showed a dr of 95:5.
3
.1.1.1. Ligand Preparation via Corey−Itsuno Reduction of
,4-Diketones. The investigation of route (a) began with
Corey−Itsuno reduction of 1,4-diaryl-1,4-butanedione 9
1
38
(
Table 1, method A). Guided by the protocol reported by
2
9
Steel et al., the preparation of chiral diols 10 with various
aromatic substituents was evaluated. Excellent enantioselectiv-
ities were observed for electron-rich and moderately electron-
deficient substrates, whereas the diastereoselectivities were
moderate (entries 1−3). The results for 3,5-bis(trifluoromethyl)-
phenyl variant 9d were less satisfactory in both categories (entry
With the 1,4-diols 10a, 10c ,and 10d available in high
diastereo- and enantiomeric purities, their activation for
subsequent reaction with hydrazine was studied (Scheme 5).
Electron-rich dimesylate 11a could not be obtained because of
polymerization. Partial decomposition and epimerization of
40
dimesylate 11b were observed, whereas electron-deficient
mesylate 11c was isolated in crystalline form and 11d was
obtained as an oil. Displacement of the latter two dimesylates
with hydrazine occurred smoothly to afford 1-aminopyrrolidines
8c and 8d. These compounds appeared to be oxygen sensitive
and unstable to storage at ambient temperature, and were used
shortly after their preparation. The condensation of 8c and 8d
4
). The best diastereoselectivity was obtained for the reduction
29
of diphenyl dione 9e (entry 5).
In view of the moderate diastereoselectivities, a modified
protocol that employs tin(II) chloride and sodium borohy-
dride was evaluated for the same set of diketones (Table 1,
3
9
method B). Despite the higher reaction temperature, excellent
D
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Table 1. Stereoselective Reduction of 1,4-Diaryl-1,4-butanediones 9
method A
method B
a
a
a
,
b
a
entry
diol
R
dr
er
dr
93:7, [99:1]
er
1
2
3
4
5
10a
10b
10c
10d
10e
4-MeOC H
78:22
80:20
84:16
63:37
91:9
>99:1
99:1
99:1
99:1
6
4
4-t-BuC H
90:10, [87:13]
89:11, [98:2]
79:21, [>99:1]
n/a, n/a
6
4
4-F CC H
4
>99:1
87:13
>99:1
>99:1
>99:1
n/a
3
6
c
c
c
c
3,5-(F C) C H
3
3
2
6
Ph
a
b
Diastereomeric ratios (chiral:meso) and enantiomeric ratios were determined by CSP-SFC. The values in brackets are ratios after trituration or
c
recrystallization. Determined through dibenzoate derivative.
Scheme 5
with glyoxal was carried out in two separate steps to provide the
electron-deficient bis-hydrazone ligands 13c and 13d. A slightly
elevated temperature was required to ensure condensation of
substrates were prepared by α-arylation of N-Boc-pyrrolidine
through a sequence of enantioselective deprotonation me-
diated by (−)-sparteine, transmetalation to zinc chloride, and
32,33
3
2
1
2d in a timely fashion. Malondialdehyde-derived ligand 15 was
also prepared using this route.
.1.1.2. Ligand Preparation via α-Arylation of N-Boc-
pyrrolidine. In view of the inability to reduce 1,4-bis(2-
palladium-catalyzed sp −sp coupling of aromatic bromides.
The Boc protecting group was easily removed from 16f using an
excess of trifluoroacetic acid. Electron-rich substrate 16a
decomposed under these conditions but could be deprotected
by treatment with a slight excess of iodotrimethylsilane at 0 °C to
3
2
9
naphthyl)-1,4-dione and the instability of electron-rich
dimesylates 11a and 11b, a new synthetic approach to a broader
range of bis-hydrazone ligands was needed. The preparation of
ligands with monoaryl substituted pyrrolidine was studied
initially to probe the accessibility of the 1-aminopyrrolidines
from the parent heterocycle (Scheme 4). The investigation
began with 4-methoxyphenyl- and 2-naphthyl-substituted
N-Boc-(R)-2-arylpyrrolidines 16a and 16f (Scheme 6). These
41
afford the deprotected 2-arylpyrrolidine 17a in 93% yield.
42
Nitrosation of 17a with nitrosonium tetrafluoroborate took
place in 1.5 h at 0 °C, but a higher temperature was required for
the 2-naphthyl analogue 17f. The enantiomeric purity of 1-
nitroso-2-arylpyrrolidine can be confirmed at this stage by CSP-
SFC analysis; 16a of 92:8 er was converted to 18a of 91:9 er. The
enantiomeric purity of 16a was improved to 96:4 by careful
E
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 6
Table 2. Preparation of N-Boc-2-arylpyrrolidine 16 and N-Boc-2,5-diarylpyrrolidine 22
a
b
c
a
d
entry
R
ref
yield (%)
er
ref
yield (%)
1
2
3
4
5
6
7
8
4-MeOC H
16a
16f
16g
16h
16i
71
61
56
61
54
96:4
96:4
96:4
96:4
95:5
95:5
96:4
96:4
22a
22f
22g
22h
22i
18
6
4
g
2-naphthyl
37
3,5-Me C H
3
21
18
2
6
3,5-Ph C H
3
2
6
2-tolyl
17
e
5-Me-2-thienyl
3,5-(F C) C H
16j
∼47
55
22j
13
f
f
16d
16k
22d
22k
trace
0
3
2
6
3
1-naphthyl
48
a
b
c
Reference number for compound. Isolated yield after column chromatography and Kugelrohr distillation if necessary. Determined by CSP-SFC.
d
e
Isolated yield after column chromatography and recrystallization if necessary. Cannot completely purify by chromatography and Kugelrohr
f
g
distillation. Determined through 3,5-dinitrobenzoate derivative. Contains about 6% of the corresponding pyrroline.
control of the reaction temperature of α-arylation during the
deprotonation step. The reduction of 1-nitrosopyrrolidines 18a
The second arylation was significantly more challenging, and
N-Boc-(2R,5R)-2,5-diarylpyrrolidines 22 were obtained in less
satisfactory yields (Table 2). A significant amount of the starting
material was not consumed. For example, 16a was recovered in
64% yield. The low yields were also attributed to the difficulty in
purification from unconsumed 16 or removal of a small amount
of the corresponding dehydropyrrolidine. Nevertheless, this pro-
cedure provided a range of Boc-protected diarylpyrrolidines 22
without a detectable amount of the meso-diastereomer. The
second α-arylation could not provide 1-naphthyl- and 3,5-
bis(trifluoromethyl)phenyl-substituted pyrrolidines 22d and
22k (entries 7 and 8).
and 18f by LiAlH in THF proceeded smoothly at 0 °C. The two-
4
step condensation protocol with glyoxal afforded the target bis-
hydrazone ligands 21a and 21f.
With the successful preparation of bis-hydrazones 21a and 21f,
the focus next shifted to analogues featuring C -symmetric 2,5-
2
diarylpyrrolidine moieties. The α-arylation protocol developed
32
33
by Campos et al. and subsequently modified by Trost et al.
was adopted in two consecutive steps from N-Boc-pyrrolidine to
provide N-Boc-(2R,5R)-diarylpyrrolidines 22 (Table 2).
In general, the N-Boc-(R)-2-arylpyrrolidines 16 were obtained
in acceptable yields considering that three transformations are
involved in each step. The enantioselectivity of this process is
high, and ratios of 95:5 to 96:4 were achieved reproducibly.
With a number of N-Boc pyrrolidines 22 in hand, the stage was
set to reveal the unprotected C -symmetric diarylpyrrolidines 23.
2
Standard reaction conditions using 20 equiv of TFA afforded 23f,
F
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Table 3. Preparation of 2,5-Diarylpyrrolidine 23 and 1-Nitroso-2,5-diarylpyrrolidine 24
a
b
a
b
entry
R
ref
temp (°C)
time (h)
yield (%)
ref
temp (°C)
time (h)
yield (%)
c
1
2
3
4
5
6
4-MeOC H
23a
23f
23g
23h
23i
23j
0
1
6
82
81
86
93
24a
24f
24g
24h
24i
24j
0
22
22
22
22
2
2
96
92
96
93
95
6
4
d
2-naphthyl
0
d
3,5-Me C H
3
0
2
20
12
12
0.25
2
6
d
3,5-Ph C H
3
0
6
2
6
c
f
2-tolyl
5-Me-2-thienyl
0
2
97
g
e
h
200
18
∼64
0
57
a
b
c
d
Reference number for compound. Yield of chromatographed product. Reaction condition A (TMSI) was used. Reaction condition B (TFA) was
used. Boc group was removed by heating at 200 °C under argon. Yield of crude product. Cannot completely purify by chromatography. Yield
e
f
g
h
after two steps.
Table 4. Preparation of 2,5-Diarylpyrrolidine-Based Bis-hydrazone Ligands 13 from 1-Nitroso- 2,5-diarylpyrrolidine 24
a
b
a
b
a
b
entry
R
ref
yield (%)
ref
yield (%)
ref
yield (%)
1
2
3
4
5
6
4-MeOC H
8a
8f
47
51
58
43
12a
12f
12g
12h
12i
12j
99
94
84
13a
13f
13g
13h
13i
13j
84
87
83
6
4
2-naphthyl
3,5-Me C H
3
8g
8h
8i
2
6
d
e
3,5-Ph C H
3
∼87
∼88
59
2
6
c
d
2-tolyl
5-Me-2-thienyl
78
43
f
8j
47
99
82
a
b
c
Reference number for compound. Yield of chromatographed product. Estimated yield; reduction was carried out by a Zn/HCl protocol; see
d
e
f
Scheme 7. A small amount of hexanes adsorbed. 8h partially decomposed and so 12h was not consumed. Reaction time was 21 h.
2
3g, and 23h in 81−93% yield (Table 3, entries 2−4). The
nitroso-2,5-di(2-naphthyl)pyrrolidine 24f at 0 °C, and extensive
decomposition occurred at 22 °C. Other reagents known to
reaction temperature and time required were dependent on the
bulk of the aromatic group. The TMSI-assisted deprotection was
applied to substrates that were incompatible with acid-promoted
deprotection (entries 1 and 5). The reaction was slower for
hindered 2-tolyl substituted 22i and was conducted for 2 h. For
44
effect N−N bond cleavage (titanium trichloride, zinc/acetic
acid, sodium borohydride, borane or LiEt BH) did not yield any
3
(diarylamino)pyrrolidine 8f.
The most promising results were obtained using DIBAL-H in
45
5-methyl-2-thienyl-substituted pyrrolidine 22j, condition A also
dichloromethane for 2 h. Longer reaction times led to a
decrease of desired product 8f and an increase in the amount of
byproducts 1,2-di(2-naphthyl)cyclobutane 25 and 2-ethenyl-
naphthalene 26. Doubling the amount of the reductant
marginally improved the conversion but complicated the workup
because of greater amount of gelatinous aluminate. The con-
version of 24f in THF was slower and gave 25 as the dominant
product. No N−N cleavage product 23f was observed in any
case.
failed to provide the desired 2,5-diarylpyrrolidine 23j cleanly. A
number of other deprotection conditions including HCl/Et O,
Ce(NH ) (NO ) , ZnBr , or BF Et O were also not fruitful.
Eventually, heat-promoted decarboalkoxylation (200 °C)
allowed the isolation of desired product, although further opti-
mization of the reaction time is required to improve the product
yield.
An excess amount of nitrosonium tetrafluoroborate (2 equiv)
was used in the nitrosation step (Table 3) because of the in-
creased steric demands from two flanking aromatic substituents
in 23 compared to one in 2-arylpyrrolidine 17. For the relatively
unhindered bis(4-methoxyphenyl)pyrrolidine 23a, nitrosation
proceeded at 0 °C in 2 h to provide 24a. In contrast, full
consumption for the 2-naphthyl substituted analogue 23f was
achieved at 22 °C. The bulkier substrates required extended
reaction times (12−20 h) (entries 3−5). In general, the desired
nitroso-2,5-diarylpyrrolidines 24 were obtained in excellent
yields.
2
43
4
2
3 6
2
3
2
Although a perfect solution was not found for the preparation
46
of (2,5-diarylamino)pyrrolidine 8, DIBAL-H reduction of 2,5-
diaryl-1-nitrosopyrrolidines delivered the most promising
results. Optimization of this route involved stopping the reaction
at 2 h to minimize decomposition of the product 8, and the
unconsumed nitrosopyrrolidine 24 can be recycled (Table 4).
For example, reduction of 24f with DIBAL-H at ambient
temperature in methylene chloride afforded a 51% yield of the
desired aminopyrrolidine 8f (Table 4, entry 2). The unconsumed
24f was recovered in 38% yield and was subjected to a second
round of reduction (52%). Other (diarylamino)pyrrolidines
were prepared successfully with the same procedur proce-
dure except for 2-tolyl analogue 8i. The two-step condensation
The reduction of nitroso group was particularly challenging.
Unlike the clean reduction of the monoaryl 1-nitrosopyrrolidines
1
8a and 18f by LiAlH (Scheme 6), no reaction was observed for
4
G
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 7
Scheme 8
proceeded well in general at room temperature, affording various
chiral bis-hydrazone ligands 13 featuring the diaryl-substituted
C2-symmetric pyrrolidine scaffold.
synthetic route for Maruoka’s phase-transfer catalysts, hydra-
48
zones 33a, 33b, and 34 were prepared. Double displacement of
the dibromide 30 with excess hydrazine provided aminoazepine
31. Stepwise condensation of 31 with glyoxal completed the
synthesis of 33. The exclusion of oxygen is important for the
second condensation to minimize decomposition of amino-
azepine 31. The reaction of 32a with picolinaldehyde yielded
hydrazone/pyridine hybrid ligand 34.
The reduction of 2,5-di(2-tolyl)-1-nitrosopyrrolidine 24i
could not be effected by either DIBALH or LiAlH . Fortunately,
4
the reduction with Zn/HCl provided the 22i as the major
product accompanied by over-reduced diarylpyrrolidine 23i
(
83:17) (Scheme 7). Because of the difficulty in separating these
amines, this mixture was subjected directly to the stepwise
condensation with glyoxal to furnish bis-hydrazone 13i.
3
.2. Structure−Activity and Structure−Selectivity
Relationship Studies. With a number of chiral bis-hydrazone
ligands in hand, the study of asymmetric biaryl coupling of
aryldimethylsilanolates was initiated. The optimal reaction
conditions for ligand survey was established using bis-hydrazone
In summary, synthetic route (a) (Scheme 3) allowed the pre-
paration of bis-hydrazone ligands 5, 13c, 13d, and 15 with
electron-deficient and -neutral aromatic substituents. The scope
was complemented and expanded to include electron-rich and
sterically hindered analogues 13a,f−i via synthetic route (c).
5
bearing two (2S,5S)-diphenylpyrrolidine units (Table 5). Good
conversion and enantioselectivity were observed at 90 °C in 1 h
entry 1). Lowering the reaction temperature improved the
selectivity at the expense of conversion, except at 70 °C
3
.1.1.3. (R,R)-Bis(α-methylbenzyl)amine-Derived Glyoxal
Bis-hydrazone Ligand. The preparation of acyclic bis-hydrazone
9 (Scheme 8) commenced with nitrosation of commercially
(
2
+
−
47
(entries 2−4). The loading of silanolate K 35a was decreased to
available (R)-bis((R)-1-phenylethyl)amine. Sodium−metal
reduction of nitrosamine 27 in ethanol at reflux afforded the
desired 1,1-bis((R)-1-phenylethyl)hydrazine and the over-
reduced product in 84:16 ratio. This mixture was subjected
to a substoichiometric amount of glyoxal to provide chiral bis-
hydrazone 29 in a one-step preparation.
1
6
.5 equiv without detrimental impact on the product yield (entry
). Doubling the ligand loading did not improve the
47
enantioselectivity (entry 8). Although the use of (MeCN) PdCl
2
2
as palladium source did not affect the selectivity, the reaction was
considerably slower (entry 9). The reaction in 1,4-dioxane was
less efficient than in toluene in all aspects (entries 5 and 10). The
optimal reaction conditions involved the use of 1.5 equiv of
3.1.1.4. Binaphthalene-Based Glyoxal Bis-hydrazone Li-
gands. Inspired by the structural feature of binaphthalene-based
phosphine ligands such as BINAP, a novel type of bis-hydrazone
ligand was proposed (Scheme 9). Adopting the reported
silanolate at 70 °C in toluene with 2.5 mol % of [allylPdCl] and
2
5 mol % of ligand 5 (entries 6 and 7).
H
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 9
+
−
Table 5. Reaction Optimization for the Cross-Coupling of 2-Methylnaphthylsilanolate (K 35a ) Using Bis-hydrazone Ligand 5
+
−
a
b
c
entry
K 35a (equiv)
temp (°C)
time (h)
37 (%)
38 (%)
er, 37 (R/S)
1
2
3
4
1.75
1.75
1.75
1.75
1.75
1.5
90
70
50
21
70
70
70
70
70
70
1
1
85
85
71
43
59
85
2.4
2.3
3.4
5.4
10
92:8
95:5
96:4
98:2
90:10
94:6
95:5
96:4
95:5
90:10
4
7 d
24
2
d
5
6
7
3.3
2.5
2.3
2.3
4.3
e
1.5
2
90
f
8
g
1.5
1
84
79
70
9
1.5
10
7.5
d
1
0
1.5
a
b
Yield of chromatographed product taking into account of 1,1′-binaphthalene. Percentage of 1,1′-binaphthalene in the chromatographed product
c
d
e
f
estimated by CSP-SFC. Ratio of (R)-37:(S)-37 determined by CSP-SFC. Reaction was conducted in 1,4-dioxane. 0.5 mmol scale. 10 mol % of
ligand. (MeCN) PdCl (5 mol %) was used as the palladium source.
g
2
2
For the purpose of establishing structure−activity and −
selectivity relationships (SAR and SSR) and further improving
the enantioselectivity of the cross coupling, pyrrolidine-based
bis-hydrazones 13 bearing various aromatic substituents were
studied using the optimized reaction conditions (Table 6).
Increasing the steric bulk of the aromatic group as in 13g and 13h
had a negative impact on the yield and the enantiomeric purity of
the coupled product (entries 2 and 3). In one extreme case, the
(cf. entries 4 and 3). The cross-coupling with the more hindered
electrophile 1-bromo-2-methylnaphthalene was possible,
although higher temperature was required (110 °C) to produced
even a modest amount of 40, but gratifyingly with enantio-
selectivity in this preliminary experiment (entry 11).
In an effort to probe the stereodetermining step, the donor/
acceptor role of the two coupling substrates was reversed. The
+
−
reaction was slower when 1-naphthyldimethylsilanolate K 35b was
used as the nucleophile and 1-bromo-2-methylnaphthalene was used
as the electrophile (entry 12). Interestingly, the major enantiomer
also possessed the (R)-configuration, and the enantiomeric ratios of
the products from these reactions are the same (cf. entries 1 and 12).
Further SAR studies revealed that the presence of substituents
at both the 2- and 5-positions on the pyrrolidine moiety was
critical for high enantioselectivity (Table 7, entries 3 and 4). Bis-
hydrazone 29, an analogue of 5 that lacks the C(3)−C(4) bond,
was ineffective at inducing selective coupling (entry 5). The
introduction of dimethylmethide linkage to the bis-hydrazone
2-tolyl-substituted ligand 13i yielded a near-racemic mixture of
37 (entry 5). The use of 2-naphthyl- (13f) or 4-methoxy- (13a)
substituted ligand restores the selectivity of the coupling process
compared to that of the parent ligand (5) (cf. entries 6, 7, and 1).
On the contrary, electron-deficient ligands 13c and 13d were less
selective, and the reaction was significantly slower for 13d
(
entries 8 and 9). The cross-coupling stalled when 5-methyl-
-thienyl-substituted hydrazone 13j was used (entry 10).
Replacing the palladium source with (MeCN) PdCl did not
2
2
2
affect the enantioselectivity but resulted in a sluggish reaction
I
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Table 6. SAR and SSR Studies of Bis-hydrazone Ligands for the Coupling of 2-Methylnaphthylsilanolate (K 35a )
Article
+
−
a
b
c
d
entry
ligand
config
aryl
time (h)
product
yield (%)
yield, 38 (%)
product, er (R/S)
e
1
2
3
5
S,S,S,S
Ph
3,5-Me C H
3
2
3
37
37
37
37
37
37
37
37
37
37
40
37
90
2.5
3.6
4.1
3.0
9.7
1.4
2.0
3.8
8.3
5.8
n/a
1.2
95:5
13g
13h
13h
13i
13f
13a
13c
13d
13j
5
R,R,R,R
R,R,R,R
R,R,R,R
R,R,R,R
R,R,R,R
R,R,R,R
S,S,S,S
75
73
12
56
86
81
81
52
45
8:92
2
6
3,5-Ph C H
3
3
13:87
12:88
44:56
5:95
2
6
f
4
3,5-Ph C H
3
2
6
3
5
6
7
8
9
2-tolyl
12
2
2-naphthyl
4-MeOC H
2
6:94
6
4
4-F CC H
4
2
90:10
71:29
20:80
3
6
S,S,S,S
3,5-(F C) C H
3
18
7
3
2
6
10
11
12
R,R,R,R
S,S,S,S
5-Me-thienyl
Ph
g
j
h
i
19
7
24
92:8
5
S,S,S,S
Ph
72
95:5
a
b
c
Configuration of the bis-hydrazone ligand. Yield of chromatographed product taking into account of 1,1′-binaphthalene. Percentage of 1,1′-
d
e
binaphthalene in the chromatographed product estimated by CSP-SFC. Enantiomeric ratio of product determined by CSP-SFC. 0.5 mmol scale.
f
g
+
−
h
(
MeCN) PdCl (5 mol %) was used as the palladium source. K 35a and 39 were used in the experiment. Yield of 2,2′-dimethyl-1,1′-
2
2
i
binaphthalene; reaction was conducted at 110 °C. The configuration was assigned based on the sense of induction for 2-methyl-1,1′-binaphthalene
j
+
−
(R)-37 and optical rotation. K 35b and 39 were used in the experiment.
ligand was also detrimental; this reaction stalled after 2 h and a
racemic product was obtained (entry 6). A novel type of bis-
hydrazone ligand featuring the dinaphthylazepine backbone was
also tested, which provided the biaryl product in 81:19 er (entry
palladium source improved the enantioselectivity, albeit at a
significant expense of conversion (cf. entries 8 and 1).
To probe the stereodetermining step, the donor/acceptor role
of the two coupling substrates was reversed (entries 9 and 10).
The reactions were considerably slower; both 1-naphthyldime-
7
). Intriguingly, the reaction employing 3- and 3′-phenyl-
substituted variant 33b favors the formation of the enantiomer
32:68 er), albeit with moderate yield after a prolonged reaction
+
−
thylsilanolate K 35b and 2-methoxy-1-bromonaphthalene were
(
not consumed within 12 h. The reactions were quenched after
2
4 h, at which point no more silanol was detected. The major
time (entry 8). The hydrazone−pyridine hybrid loses the ability
to provide stereocontrol (entry 9).
enantiomer had the same configuration as before (cf. entry 1
and 9, and entry 4 and 10), although the enantiomeric com-
position differed slightly.
To evaluate the effect of the substituent next to the silanolate
on the rate and selectivity of the coupling, the reaction of a
more electron-rich, but less sterically hindered nucleophile,
+
−
2
(
-methoxynaphthylsilanolate (K 35c ) was investigated
Table 8). Under standard conditions, the progress of the
reaction was indicated by the observable changes in the physical
4. DISCUSSION
4
.1. Effect of Ligand on the Rate and Selectivity of the
Cross-Coupling. 4.1.1. 2-Methylnaphthylsilanolate. During
the initial stage of the ligand survey, a high temperature (110 °C)
was found to be necessary for the coupling of 2-methylnaph-
+
−
state of the reaction mixture. Potassium silanolate K 35c , has
limited solubility in toluene even at elevated temperature
(70 °C). The gradual consumption of this nucleophile was manifest
+
−
thylsilanolate K 8a when chiral bidentate phosphine ligands
were used (see the Supporting Information for the initial ligand
survey). This is not surprising since partial ligand dissociation is
required to generate an empty coordination site for trans-
metalation. Consistent with this notion, the monodentate, MOP-
type ligands allowed the couplings to be conducted at 90 °C.
Good reactivities were also observed with diene and bis-
hydrazone ligands at this temperature because of facile ligand-
dissociation from palladium that originates from their weaker
by its dissolution and the darkening of the reaction mixture.
Using 5, biaryl product 41 was obtained in 76% yield (GC) and
+
−
7
9:21 er (entry 1). Although silanolate K 35c was fully con-
verted in 3 h (GC analysis), a small amount of 1-bromo-
naphthlene was not consumed. Further purification of the
product to remove siloxanes led to a slightly enhanced er (entry 2).
Under “ligandless” conditions, a significant amount of the
coupling product was also observed (56%) (entry 3). Whereas
the employment of 2-naphthyl substituted ligand 13f marginally
improved the enantioselectivity (entry 4), the reaction was
slower. To improve the enantiomeric purity of the coupling
product, bulkier ligands bearing 3,5-Me C H (13g), 3,5-
49
coordinating abilities compared to bisphosphines.
In the initial reaction optimization using bis-hydrazone ligand
5, toluene was found to be a superior solvent than 1,4-dioxane
with respect to reaction time, yield, and enantioselectivity
(Table 5, entries 2 and 5). This observation may be rationalized
by the ability of polar solvent to interact with palladium
intermediates. The dioxane molecule could potentially occupy a
2
6
3
Ph C H (13h), and 2-tolyl (13i) substituents were examined
2
6
3
(
entries 5−7). Disappointingly, the results were less satisfactory
in both yield and er. The use of (MeCN) PdCl as an alternative
2
2
J
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Table 7. SAR and SSR Studies of Hydrazone Ligands for the Coupling of 2-Methylnaphthylsilanolate (K 35a )
Article
+
−
a
b
c
entry
ligand
time (h)
37 (%)
38 (%)
er, 37 (R/S)
1
2
3
4
5
6
7
8
9
13a
13f
21a
21f
29
2
2
4
6
4
81
86
70
76
78
2.0
1.4
3.5
2.8
6.6
12.8
3.0
9.0
8.8
6:94
5:95
40:60
43:57
48:52
50:50
19:81
68:32
52:48
d
e
15
2
∼21
84
33a
33b
34
2
18
24
f
59
34
a
b
Yield of chromatographed product taking into account of 1,1′-binaphthalene. Percentage of 1,1′-binaphthalene in the chromatographed product
c
d
e
f
estimated by CSP-SFC. Ratio of (R)-37:(S)-37 determined by CSP-SFC. Reaction stalled. Contains a contaminant. 0.1 mmol scale.
+
−
Table 8. SAR and SSR Studies of Bis-hydrazone Ligands for the Coupling of 2-Methoxynaphthylsilanolate (K 35c )
R1
R2
yield (%)
a
er (S/R)
b
entry
ligand
ligand config
aryl
time (h)
1
2
3
4
5
6
7
5
S,S,S,S
S,S,S,S
Ph
Ph
3
3
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
H
H
76
79:21
84:16
c
5
H
60
none
13f
13g
13h
13i
5
3
H
56
53
61
38
66
41
R,R,R,R
R,R,R,R
R,R,R,R
R,R,R,R
S,S,S,S
2-naphthyl
3
H
16:84
31:69
26:74
44:56
88:12
84:16
18:82
3,5-Me C H
3
3
H
2
6
3,5-Ph C H
3
3
H
2
6
2-tolyl
Ph
3
H
d
8
3
H
c
9
5
S,S,S,S
Ph
24
24
OMe
OMe
60
c
1
0
13f
R,R,R,R
2-naphthyl
H
46
a
b
c
Determined by GC using biphenyl as the internal standard. Determined by CSP-SFC of chromatographed product. Isolated yield on a 0.25 mmol
d
scale. (MeCN) PdCl (5 mol %) was used as the palladium source.
2
2
coordination site competitively because of its high concentration.
As a consequence, transmetalation is slower, and the chelation of
the chiral hydrazone ligand is disrupted.
The choice of palladium source had a more dramatic impact
on the rate of the reaction. The cross-coupling employing
(MeCN) PdCl proceeded slowly possibly because of the
2
2
K
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Figure 4. Critical features of 2,5-diaryl-based bis-hydrazone ligands to facilitate high enantioselectivity in the catalytic, asymmetric biaryl coupling.
1
9
difficulty in double transmetalations from two aryldimethylsila-
nolates, which provides diarylpalladium complex necessary for
the formation of active Pd(0) species by reductive elimination
also provides lower selectivity. A five-membered palladacycle
formed from the chelation by the bis-hydrazone ligand is a much
superior catalyst than the six-membered analogue. The presence
of each chiral hydrazone unit is important as demonstrated by the
coupling using pyridine−hydrazone hybrid 34 (52:48 er) (Table
7, entry 9).
(
entry 9). In contrast, Pd(0) is generated readily from
[
allylPdCl] by nucleophilic attack on the allyl moiety by the
2
5
0
silanolate.
4
.1.2. 2-Methoxynaphthylsilanolate. The selectivity and
reactivity for the coupling of 2-methoxynaphthylsilanolate
K 35c ) mimic those of 2-methylnaphthylsilanolate
K 35a ), a bis-hydrazone ligand with a bulkier aromatic sub-
stituent than phenyl correlates with a lower enantioselectivity
and product yield. Nevertheless, further discussion can be made
by scrutinizing the data in Table 8.
The structure−activity relationship revealed that the use of
bulkier bis-hydrazone ligands generally correlates with longer
reaction times and lower product yields (Table 6). The increased
steric encumbrance likely raises the energy barriers for both
oxidative addition and transmetalation steps, thus resulting in a
more sluggish reaction.
+
−
(
(
+
−
Further SAR analyses have revealed the critical features of bis-
The results from entries 1−3 warrant comment. First, the
enantiomeric purity of the chromatographed product (entry 1) is
lower than that of the further purified product after removal of
residual siloxanes (entry 2). This situation arises by accidental
enantiomeric enrichment during purification. The solvent used
to wash away siloxanes from the chromatographed product was
analyzed, and the small amount of dissolved product showed a
substantially lower er (33:67).
hydrazone ligands (Figure 4). Aromatic substituents at both the
2
- and 5-positions of the pyrrolidine are necessary for high
enantioselectivity. The rigidity of the pyrrolidine ring is required
because no asymmetric induction was observed if the C(3)−
C(4) linkage is missing. The 2,6-diphenylpiperidine analogue
Scheme 10
Second, a significant amount of background reaction was
detected (entry 3). It is intriguing that coupling product was
produced in substantial amount (56%) in the absence of
hydrazone ligand. This phenomenon has been documented
50
previously from these laboratories. Presumably, the product
Scheme 11
L
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
generated from the reduction of palladium(II) is noninnocent
second equivalent of the silanolate for the transfer of aryl group.
The possibility of unproductive coordination of the methoxy
group to palladium over π-coordination prior to transmetalation
cannot be excluded (Scheme 11). The moderate yield of the
biaryl product can potentially be improved by simply increasing
the amount of silanolate K 35c to ensure full consumption of
the 1-bromonaphthalene in the cross-coupling reaction and to
compensate for unproductive siloxane formation.
(Scheme 10). A small portion of the silanolate reacts with the
allyl group of the allylpalladium chloride dimer, and the resulting
silyl allyl ether 42 can, in principle, stabilize palladium(0) to some
51
extent by acting as a π-ligand. Therefore, it should be noted that
52
+
−
this is not a true “ligandless” condition. To support the above
hypothesis, (MeCN) PdCl was used in place of [allylPdCl] as
2
2
2
the precatalyst to avoid formation of silyl allyl ether 42. Indeed,
an increase of enantiomeric ratio of the biaryl product from 21:79
to 12:88 was observed (Table 8, entries 1 and 8).
Considering the extent of background reaction, it is remark-
able that meaningful enantioenrichment can be observed when a
phenyl- or naphthyl-substituted ligand is present (entries 1
and 4). This observation implies that palladium associates to the
bis-hydrazone ligand much more strongly than to allyl silyl ether
As has been noted previously for the role-reversal experiments,
the reaction is significantly slower when the electron-donating
group is on the aromatic bromide rather than on the aromatic
silanolate (Table 8, cf. entries 1, 9, 4 and 10). This observation
can be rationalized by examining the electronic and steric
27b
influence of the substituent on the catalytic cycle.
Slower
oxidative addition is expected for a more electron-rich aryl
bromide such as 2-methoxy-1-bromonaphthalene compared to
4
2. However, such speculation may not be applied to the bulkier
53
bis-hydrazone analogues in which lower enantioselectivities or
near racemic biaryl products were obtained (entries 5−7). The
increased steric bulk may hinder the coordination of these
ligands to palladium effectively resulting in competitive back-
ground reaction.
1-bromonaphthalene. Additionally, the 2-methoxy substituent
could engender some steric hindrance toward the site of C−Br
bond breakage. These two factors can also manifest in the
transmetalation step. Since the palladium(II) center is less
electron-deficient but more sterically hindered, the transfer of
aromatic moiety from the dimethylsilyl unit would be less
efficient (Figure 5).
+
−
The coupling of 2-methoxynaphthylsilanolate K 35c , was
more sluggish than the less electron-rich 2-methylnaphthylsila-
+
−
nolate K 35a (Scheme 11). These results are somewhat sur-
prising given that a Hammett study revealed an electron-rich
nucleophile should accelerate the rate of the reaction by stabi-
27b
lizing a partial positive charge of a transition state. Addition-
ally, the steric hindrance of a 2-methyl substituent should further
retard the rate of transmetalation. This conundrum may be
reconciled by the difference in the solubility of the silanolates.
+
−
Whereas the 2-methyl-substituted silanolate K 35a can be
+
−
characterized in benzene-d , the 2-methoxy congener K 35c is
6
almost completely insoluble, and its characterization was
+
−
conducted in THF-d . At 70 °C, silanolate K 35a almost
8
Figure 5. Electronic and steric environments of the palladium(II) center
of the pretransmetalation intermediate.
completely dissolved in toluene after 10 min, whereas silanolate
+
−
K 35c did not. Consequently, the lower concentration of
silanolate in the solution may result in a slower reaction, because
the activated transmetalation is the major pathway in the catalytic
4.2. Stereodetermining Step. Preliminary mechanistic
insight of the stereodetermining step is gained by the donor−
acceptor reversal of the coupling partners (Scheme 12, eqs 1 and 2).
27a
cycle and it is concentration dependent. Contrariwise, the
minor, thermal transmetalation pathway does not depend on a
Scheme 12
M
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 13
Figure 6. Optimized structure of the most stable conformer of the diarylpalladium intermediate with chiral ligand (R,R,R,R)-5: (a) top view; (b) side
view; (c) side view with R and R′ groups omitted.
Interestingly, 2-methyl-1,1′-binaphthalene is obtained as the
R)-configured isomer from both reactions with the same
enantioselectivity (95:5 er). Further comparison with the results
from the Suzuki-coupling (eq 3) indicates the stereodetermin-
ing step is also independent of the element of the transmetalating
unit (Si vs B). This statement is also true for the coupling of
4.3. Origin of Enantioselectivity. To aid the understanding
of the origin of enantioselectivity and the effects of ligand in the
reductive elimination step, the possible isomers of the diaryl-
palladium(II) intermediates (C and C′, Scheme 13) and the
reductive elimination transition states were computed using
density functional theory (DFT). Reductive elimination from
palladium(II) complexes, an important step in many cross-
coupling reactions, has been studied thoroughly with computa-
(
19
2-methoxynaphthyl and naphthyl moieties (see the Supporting
Information). These observations strongly suggest a common
intermediate and the same elementary step in the catalytic cycle
is involved in the determination of product configuration.
Analysis of the proposed catalytic cycle (Scheme 13) reveals
that Pd(II) complexes C and C′ are the common intermediates
involved in the equations in Scheme 12; the aryl substituents on
the palladium have no memory of their origins, either from the
donor or the acceptor. In contrast, complexes A, A′ and B, B′ are
directly associated with the identity of aryl bromide and trans-
metalating agent, respectively. On the basis of these preliminary
findings, reductive elimination is hypothesized to be the stereo-
determining step.
5
4
tions. However, only a few computational studies on
17
asymmetric biaryl coupling have been reported. In the present
study, we investigated the origins of enantioselectivities with
chiral bis-hydrazone ligands and the effects of substituents on the
ligands. Calculations with (R,R,R,R)-5 were carried out first as the
representative bis-hydrazone ligand. Geometry optimizations
55
were performed with B3LYP and a mixed basis set of SDD for
Pd and 6-31G(d) for other atoms. Single-point calculations were
56
performed with M06 and a mixed basis set of SDD for Pd and
57
6-311+G(d,p) and the SMD solvation model in toluene. All
58
calculations were performed with Gaussian 09. This combination
N
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Figure 7. Numbers of potential arene−arene interaction for diarylpalladium complexes. Gibbs free energies of the isomers are with respect to the most
stable isomer (5)PdRR′-A.
Figure 8. Reductive elimination transition structures for diarylpalladium complex of phenyl substituted bis-hydrazone ligand (R,R,R,R)-5. Repulsions
between the two naphthyl moieties are indicated by the distances between two atoms, when such distance is less than 80% of the sum of van der Waals
radii. The H−H and H−C distances and the distances between the centroids of the B-rings are shown in black, green, and blue, respectively.
of methods has been widely tested in the studies of mecha-
(
(
R = 1-naphthyl; R′ = 2-methyl-1-naphthyl) is shown in Figure 6
nisms and selectivities in various transition metal-catalyzed reac-
see the Supporting Information for other higher energy
5
9,60
tions.
Computations employing a similar level of theory re-
isomers). For a clearer view of the spatial environment created
by the ligand, the two front naphthyl moieties have been tem-
porarily removed in Figure 6c. Because of the C -symmetry of the
ligand, the NW and SE quadrants are both occupied by phenyl
groups on the ligand. On the contrary, empty spaces can be seen
produced the experimental barrier of C−C reductive elimination
61
from a phenylpalladium(II) N-heterocyclic carbene complex.
2
The optimized geometry of the lowest energy conformer
of the diarylpalladium-bishydrazone complex (5)PdRR′
O
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Figure 9. Transition structures with 4-(trifluoromethyl)phenyl-substituted bis-hydrazone ligand (R,R,R,R)-13c.
in the SW and NE quadrants. These chiral pockets are expected
to dictate the orientation of the substrates. In the most stable
conformer of the diarylpalladium intermediate, the two empty
quadrants are occupied by the naphthyl rings, and the relatively
less hindered MeC(2) and HC(2) substituents are positioned in
the more congested quadrants. The two naphthyl groups are not
aligned but are tilted in a propeller sense. The naphthyl B-rings,
which are located on the unoccupied quadrants, are tilted toward
the ligand, while the MeC(2) and HC(2) groups on the filled
quadrants are tilted outward to avoid repulsions with the phenyl
groups on the ligand.
transition structure TS-A, which leads to the formation of biaryl
(S)-37. The clockwise motion of the aryl substituents tilts the
naphthyl B-rings toward the unoccupied NE and SW quadrants,
away from the C(2)-Me and C(2′)-H groups. In contrast,
transition structure TS-A′, which arises from counterclockwise
motion of the aryl substituents and would lead to the (R)-
product, has significantly higher energy. The counterclockwise
motion in TS-A′ rotates the naphthyl B-rings toward the C(2)-
Me and C(2′)-H groups, causing significant steric repulsion
between Me-C(2)/HC(8′) (H−H distance of 1.99 Å) and
HC(8)/HC(2′) (H−H distance of 1.86 Å). In TS-A′, the
geometry of palladium is considerably distorted from square
planar to minimize steric strain.
To facilitate discussion, a further-simplified picture is pro-
vided to depict the steric environment created by the ligand
(
red squares) and the position of the naphthyl group (letter B,
The second most stable transition structure TS-B has two
B-rings at the NW and SE quadrants. The minor enantiomer (R)-
product is expected from this transition state. The counter-
clockwise motion in TS-B rotates the naphthyl B-rings toward
the occupied quadrants of the ligand. To minimize steric inter-
action between substrates and phenyl groups of the ligand, the
palladium is again distorted from the perfect square-planar geo-
metry. Although relieved from the substrate−ligand repulsion,
the clockwise conrotatory motion implied by TS-B′ engenders
significant steric strain between naphthyl B-rings and C(2)-Me/
C(2′)-H groups observed similarly for TS-A′; the C−H distance
between C(2)/H−C(8′) (2.26 Å) and the H−H between
H−C(2′)/H−C(8) (1.86 Å) are both much shorter than the
sum of the van der Waals radii (2.9 and 2.4 Å for C−H and H−H,
respectively).
The two naphthyl moieties in transition structures TS-C,
TS-C′, TS-D, and TS-D′ have a syn relationship. These structures
all have unfavorable interaction between HC(8)/C(8′) or C(8)/
HC(8′). Therefore, energies higher than those for TS-A and TS-B
were found.
On the basis of the energy difference (2.9 kcal/mol) between
the most stable transition structures TS-A and TS-B leading to
the enantiomeric products, the predicted er (99:1) is comparable
to observed er (95:5) at 70 °C.
Figure 6b). The shaded red block represents filled quadrant and
the label B represents the location of naphthyl B-ring.
́
As was suggested by Fernandez et al., arene−arene interaction
may favor an alternative orientation of the two naphthyl groups
with respect to the ligand such that the B-rings are in the same
19
quadrants as the phenyl groups of the ligand. With this
consideration, three more conformers of the diarylpalladium
complexes can be formulated (Figure 7, B−D); one of which has
two potential arene−arene interactions (B). These complexes
are only about 1 kcal/mol less stable than the most stable isomer
(5)PdRR′-A.
Each of the four conformers of the diarylpalladium inter-
mediate may lead to either (S)- or (R)-coupling product via the
C−C bond forming transition state accompanied by either
clockwise or counterclockwise conrotatory motion of both
naphthyl groups about the aryl−Pd bond. Such conrotatory
motion is necessary to avoid head-on collision between C(2)/
(
8) and C(2′)/(8′) substituents. All eight possible reductive
elimination transition states were computed (Figure 8), four of
which simulate clockwise conrotatory reductive elimination (TS-
A, TS-B′, TS-C′, and TS-D), and the other four simulate
counterclockwise conrotatory reductive elimination (TS-A′,
TS-B, TS-C, TS-D′). In all transition states, the conformation
of the bis-hydrazone ligand remains relatively rigid, with minimal
changes from the intermediate complex (see Figure S1,
Supporting Information, for an overlay of the intermediate and
transition-state geometries). The relative energies of the
transition states are determined by steric repulsions between
the ligand and the aryl groups and between the C(2)/C(8) and
C(2′)/C(8′) substituents about the forming aryl−aryl bond.
Both types of steric repulsions are minimized in the most stable
The electron-deficient ligand 13c was similarly analyzed for
the two lowest energy transition structures, TS-E and TS-F
(Figure 9), which have the same conformation as TS-A and
TS-B, respectively. This exercise revealed a smaller energy
difference between the two transition structures (1.4 kcal/mol),
which may be attributed to the increased arene−arene dispersion
interaction between substrates (B-ring) and the more electron-
deficient aromatic ligand (4-trifluoromethylphenyl) in TS-F.
P
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 14
The distances between the centroids of the two π-systems are
were found to be generally less effective in the cross-coupling
reaction of aryldimethylsilanolate. Transmetalation to a tetra-
coordinate palladium is unfavorable based on prior calcu-
lations. Therefore, an opportunity exists for the equili-
bration between diarylpalladium A−D through tricoordinate
palladium after transmetalation and before reassociation of
hydrazone ligand.
3
.96 and 4.47 Å in TS-F, noticeably shorter than the
corresponding distances in the phenyl hydrazone-ligated TS-B
4.27 and 4.60 Å, respectively). See Figure S2 (Supporting
63
(
Information) for an overlay of the geometries of TS-A/TS-E and
TS-B/TS-F. These results are consistent with a stronger arene−
arene interaction between electron-rich and electron-deficient
62
π-system than between two electron-rich π-systems. In either
case, a longer distance is measured between the π-systems
located at the NW quadrant than at the SE because the 2-methyl
substituent is repelled by the adjacent naphthyl moiety. The com-
puted er (91:9) based on the energy difference between TS-E and
TS-F closely approximates the observed er (90:10) at 70 °C.
6
. CONCLUSIONS
A variety of diarylpyrrolidine-based chiral bis-hydrazone ligands
have been evaluated in the asymmetric cross-coupling reaction of
aryldimethylsilanolates. Ligands with electron-rich/neutral and
unhindered aromatic substituents on the 2,5-positions of the
pyrrolidine ring generally correlate with higher enantioselectiv-
ities and reactivities. Preliminary mechanistic studies indicate
that reductive elimination is likely the stereodetermining
step. The interpretation of the origin of enantioselectivity has
been facilitated by computational modeling. The lowest energy
reductive elimination transition state features a propeller align-
ment of the aryl substituents, which tilts the naphthyl B-rings
toward the unoccupied quadrants of the ligand. The arene−
arene interaction has been identified as the potential culprit of
lower selectivity observed for ligands with electron-deficient
aromatic substituent. This result implies that eliminating
the π-density on the ligand should enhance the enantio-
selectivity by raising the transition state energy toward the
minor enantiomer.
5. OUTLOOK AND FUTURE DEVELOPMENT
The modeling studies provided much insight into the origin and
magnitude of enantioselectivity. Nevertheless, a few issues re-
mained to be addressed. First, the conclusions from the calcu-
lations can be tested by using ligands such as 3,5-dimethyl- and
3
,5-bis(trifluoromethyl)-substituted bis-hydrazones. Second, the
interconversion barriers between diarylpalladium complexes
A−D need to be estimated. High energy barriers would suggest
that the reductive-elimination step does not determine the
enantioselectivity and the product composition is a consequence
of the ratio of A/B/C/D. Preliminary calculations at the PM6
semiempirical level indicates that interconversion is unlikely if
the hydrazone ligand is bound to palladium in bidentate mode.
However, interconversion may be possible if the ligand is bound
in a monodentate mode leaving palladium with an empty coor-
7
. EXPERIMENTAL SECTION
Stereoselective Reduction of 1,4-Diaryl-1,4-butane-
6
3
dination site. Circumstantial evidence indicates that conversion
diones 9 (Table 1). Asymmetric Reduction of 1,4-Diaryl-1,4-
butanedione Using Borane Dimethyl Sulfide in the Presence
of Methylborate Complex of (S)-(−)-α,α-Diphenyl-2-pyrroli-
between diarylpalladium complexes A−D is a possibility through
64
partial ligand dissociation. Importantly, the results from the
donor/acceptor reversal experiments support reductive elimi-
nation as the stereodetermining step implying that diastereo-
meric complexes A−D are in equilibrium.
19
dinemethanol General Procedure 1.
The empty coordination site required for the transmetalation
event necessitates the partial dissociation of the bidentate ligand
(
Scheme 14). The use of tri-tert-butylphosphine in the prepa-
26
rative cross-coupling reactions ensures a tricoordinate palla-
dium intermediate to facilitate transmetalation. Bidentate phos-
phine ligands, stronger chelating ligands than bis-hydrazones,
Q
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
To an oven-dried Schlenk reaction flask equipped with a stir
bar and a septum was added a methylborate complex of (S)-
(1R,4R)-1,4-Bis(4-(trifluoromethyl)phenyl)butane-1,4-diol
(10c) (Entry 3).
(
−)-α,α-diphenyl-2-pyrrolidinemethanol (0.2 equiv), THF,
and borane·dimethyl sulfide (2 equiv) under argon. A solution
of 1,4-diarylbutane-1,4-dione (1 equiv) in THF was added by a
syringe pump (0.5 mmol/h) at room temperature. After
complete addition of the diketone, stirring was continued for
Following general procedure 1, a mixture of catalyst (6.2 mg,
1−4 h. The reaction was quenched with MeOH, and the mixture
2
(
0 μmol, 0.2 equiv), THF (0.25 mL), borane·dimethyl sulfide
19 μL, 0.2 mmol, 2 equiv), and a solution of 1,4-bis(4-
trifluoromethylphenyl)butane-1,4-dione (37 mg, 0.1 mmol,
equiv) in THF (0.75 mL) were combined in a 5 mL Schlenk
reaction flask. After 4 h, the reaction was quenched with MeOH
2 mL), stirred for 5 min, and then concentrated. Purification by
was stirred until no bubbling was observed (5−10 min). The
volatiles were evaporated, and the crude product was purified by
column chromatography to provide 1,4-diarylbutane-1,4-diol.
The diastereomeric and enantiomeric purities were analyzed by
CSP-SFC analysis.
1
(
(
1R,4R)-1,4-Bis(4-methoxyphenyl)butane-1,4-diol (10a)
column chromatography (SiO , 1 × 8 cm, petroleum ether/
2
(
Entry 1).
EtOAc, gradient elution, 65/35, 50/50, then 0/100) afforded
1
3
2 mg (85%) of 10c as a white solid. Data for 10c: H NMR (500
MHz, CDCl ) 7.60 (d, J = 8.2 Hz, 4 H), 7.45 (d, J = 8.1 Hz, 4 H),
3
4
H
.82 (app br s, 2 H), 2.78 (d, J = 2.9 Hz, 2 H), 1.95−1.82 (m, 2 H,
C(1)); TLC R 0.30 (EtOAc/petroleum ether, 1/1) [silica gel,
Following general procedure 1, a mixture of catalyst (7.8 mg, 26
2
f
μmol, 0.17 equiv), THF (0.16 mL), borane·dimethyl sulfide
CAM]; SFC (S,S)-10c, t 9.2 min (below detection limit); (R,R)-
R
(
28.5 μL, 0.3 mmol, 2 equiv), and a solution of 1,4-bis(4-
10c, t 10.2 min (84%); meso-10c, 11.7 min (26%) (Chiralpak
R
methoxyphenyl)butane-1,4-dione (45 mg, 0.15 mmol, 1 equiv)
in THF (3 mL) were combined in a 5 mL Schlenk reaction flask.
Within 4 h, TLC showed complete consumption of the diketone.
The reaction was quenched with MeOH (1.5 mL) and stirred for
AD, 200 bar, 3 mg/mL, 5% MeOH in CO , 2 mL/min, 220 nm,
40 °C).
(1R,4R)-1,4-Bis(3,5-bis(trifluoromethyl)phenyl)butane-1,4-
diol (10d) (Entry 4).
2
5
min. Purification by column chromatography (SiO , 1 × 8 cm,
2
petroleum ether/EtOAc, gradient elution, 65/35, 50/50, then
/100) afforded 36 mg (82%) of 10a as a white solid. The
0
2
0
spectroscopic data matched those from the literature. Data for
1
1
0a: H NMR (500 MHz, CDCl ) 7.25 (d, J = 8.5 Hz, 4 H), 6.87
3
(
(
d, J = 8.5 Hz, 4 H), 4.73−4.64 (m, 2 H), 3.80 (s, 6 H), 1.90−1.70
m, 4 H); TLC R 0.13 (EtOAc/petroleum ether, 35/65) [silica
To an oven-dried, 5 mL, one-necked, round-bottom flask
equipped with a stir bar and an argon inlet adaptor with a
septum was added (S)-(−)-α,α-diphenyl-2-pyrrolidinemethanol
(8.7 mg, 34 μmol, 0.17 equiv). After two cycles of evacuation/
argon fill, THF (0.22 mL) and trimethyl borate (5.0 μL, 44 μmol,
f
gel, p-anisaldehyde]; SFC (S,S)-10a, t 20.5 min (0.4); meso-10a,
tR 25.8 min (22.1%); (R,R)-10a, 29.0 min (77.5%) (Chiralpak
AD, 200 bar, 2 mg/mL, 10% MeOH in CO , 2 mL/min, 220 nm,
R
2
4
0 °C).
1R,4R)-1,4-Bis(4-tert-butylphenyl)butane-1,4-diol (10b)
Entry 2).
(
0
.22 equiv) were added. The colorless solution was stirred at
room temperature for 1 h, and borane dimethyl sulfide (41 μL,
.42 mmol, 2 equiv) was added. A solution of 1,4-bis(3,5-
bis(trifluoromethyl)phenyl)butane-1,4-dione (102 mg,
(
0
0
0
.2 mmol, 1 equiv) in THF (1 mL) was added at a rate of
.5 mmol/h, and stirring was continued at 22 °C for 1 h (the
Following general procedure 1, a mixture of catalyst (6.2 mg,
0 μmol, 0.2 equiv), THF (0.25 mL), borane·dimethyl sulfide
19 μL, 0.2 mmol, 2 equiv), and a solution of 1,4-bis(4-tert-
butylphenyl)butane-1,4-dione (35 mg, 0.1 mmol, 1 equiv) in
THF (0.75 mL) were combined in a 5 mL Schlenk reaction flask.
After 1 h, reaction was quenched with MeOH (1.5 mL), stirred
for 5 min, and then concentrated. Purification by column
diketone was not consumed). The reaction was cooled to 0 °C
and quenched slowly with HCl (1 M, 1.5 mL). The aqueous
phase was extracted with EtOAc (5 mL × 3), dried over Na SO ,
filtered, and concentrated. Purification by column chromatog-
raphy (SiO , 3 × 10 cm, CH Cl , then hexanes/EtOAc, 1/1)
afforded 48 mg (47%) of 10d as a white solid. The diastereomeric
and enantiomeric purities were analyzed through dibenzoate
derivative. Data for 10d: H NMR (500 MHz, CD OD) 7.92 (s,
4
2
(
2
4
2
2
2
chromatography (SiO , 1 × 8.5 cm, petroleum ether/EtOAc,
2
1
gradient elution, 65/35) afforded 29 mg (83%) of 10b as a white
3
H, HC(4)), 7.82 (s, 2 H, HC(6)), 4.91−2.83 (m, 2 H, HC(2)),
solid. The spectroscopic data matched those from the
2
1
1
1
1
.87−1.78 (m, 4 H, H C(1)); TLC R 0.57 (hexanes/EtOAc,
literature. Data for 10b: H NMR (500 MHz, CDCl ) 7.36
2
f
3
/1) [silica gel, KMnO ]; SFC (S,S)-10d, t 4.5 min (9%);
4 R
(
2
d, J = 8.4 Hz, 4 H), 7.27 (d, J = 8.4 Hz, 4 H), 4.75−4.67 (m, 2 H),
.37 (br s, 2 H), 1.99−1.79 (m, 4 H), 1.31 (s, 18 H); TLC R 0.22
meso-10d, t 5.5 min (27%); (R,R)-10d, 6.6 min (64%) ((R,R)-
R
Whelk-O1, 200 bar, 2% MeOH in CO , 2 mL/min, 220 nm,
40 °C).
f
(
1
EtOAc/petroleum ether, 35/65) [silica gel, CAM]; SFC (S,S)-
0b, t 3.3 min (below detection limit); meso-10b, t 4.2 min
2
R
R
(
2
20%); (R,R)-10b, 5.6 min (80%) (Chiralpak OB, 200 bar,
mg/mL, 10% MeOH in CO , 2 mL/min, 220 nm, 40 °C).
Asymmetric Reduction of 1,4-Diaryl-1,4-butanedione
Using Sodium Borohydride and Tin(II) Chloride in the
2
R
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Presence of (S)-(−)-α,α-Diphenyl-2-pyrrolidinemethanol
(1R,4R)-1,4-Bis(4-tert-butylphenyl)butane-1,4-diol (10b)
(Entry 2).
20
(
Table 1). General Procedure 2.
Following general procedure 2, a mixture of NaBH (182 mg,
4
4
1
.8 mmol, 2.4 equiv), THF (20 mL), SnCl (455 mg, 2.4 mmol,
.2 equiv), (S)-(−)-α,α-diphenyl-2-pyrrolidinemethanol
2
(
101 mg, 0.4 mmol, 0.2 equiv), and a solution of 1,4-bis(4-tert-
To an oven-dried, one-necked, round-bottom flask equipped
with a stir bar, a short reflux-condenser, and an argon inlet
butylphenyl)butane-1,4-dione (703 mg, 2 mmol, 1 equiv) in
THF (15 mL) were combined in a 100 mL round-bottom flask
equipped with a short reflux condenser. After complete addi-
tion of the diketone (0.5 mmol/h) at mild reflux, stirring was
continued for 30 min. The reaction was cooled to 0 °C and
quenched with MeOH (30 mL). The mixture was filtered,
concentrated, and purified by column chromatography (SiO2,
adaptor with a septum was added NaBH (2.4 equiv). After the
4
flask was purged with argon, THF was added. Under a positive
argon pressure and with rapid stirring, the septum was removed
temporarily to allow the addition of SnCl (1.2 equiv), imme-
2
diately resulting in gas evolution and the formation of a gray
mixture. After the mixture was stirred at room temperature for
4
× 11 cm, petroleum ether/EtOAc, gradient elution, 65/35) to
1
h, (S)-(−)-α,α-diphenyl-2-pyrrolidinemethanol (0.2 equiv)
afford 651 mg (92%) of 10b as a white solid. Recrystallization
from hexanes/CH Cl (2.5/1) afforded 350 mg (49%) of 10b
was added. The mixture was heated at reflux for 0.5 h, and a solu-
tion of 1,4-diaryl-1,4-butanedione (1 equiv) in THF was added
by a syringe pump (0.5 mmol/h) under mild reflux. After com-
plete addition of diketone, the reaction was stirred for another
2
2
as a white solid. The spectroscopic data matched those from
21
1
the literature. Data for 10b: H NMR (500 MHz, CDCl ) 7.36
3
(
2
d, J = 8.4 Hz, 4 H), 7.27 (d, J = 8.4 Hz, 4 H), 4.75−4.67 (m, 2 H),
10 min. The reaction was cooled to room temperature and then
.37 (br s, 2 H), 1.99−1.79 (m, 4 H), 1.31 (s, 18 H); TLC R 0.22
f
quenched with MeOH at 0 °C. After bubbling had subsided, the
ice/water bath was removed, and stirring was continued at room
temperature for 10 min. The gray solid was removed by filtration
through a pad of Celite (2 cm deep, medium-porosity fritted
(
EtOAc/petroleum ether, 35/65) [silica gel, CAM]; SFC before
recrystallization: (S,S)-10b, tR 5.2 min (1%); meso-10b, tR
.1 min (10%); (R,R)-10b, 9.8 min (89%), after recrystallization:
S,S)-10b, t 5.2 min (below detection limit); meso-10b, t
7
(
R
R
funnel), eluted with Et O or EtOAc, and concentrated to give the
2
7
.1 min (13%); (R,R)-10b, 9.8 min (87%) (Chiralpak OB, 200
crude product. Purification by column chromatography, recrys-
tallization, or trituration provided 1,4-diarylbutane-1,4-diol.
The diastereomeric and enantiomeric purities were analyzed by
CSP-SFC analysis.
bar, 2 mg/mL, 7.5% MeOH in CO , 2 mL/min, 220 nm, 40 °C).
2
(
1R, 4R)-1,4-Bis(4-(trifluoromethyl)phenyl)butane-1,4-diol
(
10c).
(
1R,4R)-1,4-Bis(4-methoxyphenyl)butane-1,4-diol (10a)
(
Entry 1).
Following general procedure 2, a mixture of NaBH (188 mg,
4
5
.0 mmol, 2.4 equiv), THF (21 mL), SnCl (472 mg, 2.5 mmol,
2
1
.2 equiv), (S)-(−)-α,α-diphenyl-2-pyrrolidinemethanol
(
106 mg, 0.41 mmol, 0.2 equiv), and a solution of 1,4-bis(4-
methoxyphenyl)butane-1,4-dione (622 mg, 2.1 mmol, 1 equiv)
in THF (41 mL) were combined in a 250 mL round-bottom flask
equipped with a short reflux condenser. After complete addition
of the diketone (0.5 mmol/h) at mild reflux, stirring was con-
tinued for 35 min. The reaction was cooled to 0 °C and quenched
with MeOH (40 mL). The mixture was filtered and concentrated
To an oven-dried, 100 mL, one-necked, round-bottom flask
equipped with a stir bar, a short reflux-condenser, and an argon
inlet adaptor with a septum was added sodium borohydride
(
182 mg, 4.8 mmol, 2.4 equiv). After three cycles of evacuation/
argon fill, THF (18 mL) was added. Under a positive argon
pressure and with rapid stirring, the septum was removed
temporarily to allow the addition of tin(II) chloride (457 mg,
2.4 mmol, 1.2 equiv), immediately resulting in gas evolution and
the formation of gray mixture. Additional THF (2 mL) was
added to rinse all the reagents down the wall of the condenser.
After the mixture was stirred at room temperature for 1 h, (S)-
(−)-α,α-diphenyl-2-pyrrolidinemethanol (103 mg, 0.4 mmol,
0.2 equiv) was added. The mixture was heated to reflux for 0.5 h,
and a solution of 1,4-bis(4-(trifluoromethyl)phenyl)butane-1,4-
dione (749 mg, 2 mmol, 1 equiv) in THF (20 mL) was added by
a syringe pump (0.5 mmol/h) under mild reflux. The reaction
was stirred for another 0.5 h after complete addition of diketone.
The reaction was cooled to room temperature and then
to give a white solid. Trituration of this solid with CH Cl2
2
(
∼4 mL) followed by addition of hexanes (∼15 mL) and
filtration afforded 435 mg (69%) of 10a as a white solid. The
spectroscopic data matched those from the literature. Data for
1
20
1
0a: H NMR (500 MHz, CDCl ) 7.25 (d, J = 8.5 Hz, 4 H), 6.87
3
(
(
d, J = 8.5 Hz, 4 H), 4.73−4.64 (m, 2 H), 3.80 (s, 6 H), 1.90−1.70
m, 4 H); TLC R 0.13 (EtOAc/petroleum ether, 35/65) [silica
f
gel, p-anisaldehyde]; SFC Before trituration: (S,S)-10a, tR
0.5 min (below detection limit); meso-10a, t 25.8 min (7%);
2
R
(
2
R,R)-10a, 29.0 min (93%). After trituration: (S,S)-10a, tR
0.5 min (below detection limit); meso-10a, t 25.8 min (1%);
R
(R,R)-10a, 29.0 min (99%) (Chiralpak AD, 200 bar, 2 mg/mL,
1
0% MeOH in CO , 2 mL/min, 220 nm, 40 °C).
2
S
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
quenched with MeOH (30 mL) at 0 °C. After bubbling subsided
Celite (2 cm deep, 30 mL size, medium-porosity fritted funnel),
eluted with EtOAc (40 mL), concentrated. The white solid was
taken up with EtOAc (∼5 mL) and filtered through a pad of silica
(2 cm deep, 30 mL size, medium-porosity fritted funnel), eluted
with EtOAc (80 mL), concentrated to give 10d as a white
powder, 1.20 g (>99:1 er, 83:17 dr). Recrystallization twice from
hexanes/dichloromethane (4/1, 0.04 g/mL then 0.03 g/mL,
>99:1 er, 88:12 dr, 1.04 g), and once from hexanes/dichloro-
methane (3/1, 0.035 g/mL) afforded 490 mg (40%) of 10d as
white needles (>99:1 er, >99:1 dr). Note that the third
(
∼5 min), the ice/water bath was removed and stirred at room
temperature for 10 min. The gray solid was removed by filtration
through a pad of Celite (2 cm deep, 30 mL size, medium-porosity
fritted funnel), eluted with EtOAc (30 mL), concentrated to give
830 mg of 10c as a white solid (>99:1 er, 89:11 dr). Recrys-
tallization from dichloromethane afforded 560 mg (74%) of 10c
1
as colorless needles (>99:1 er, 98:2 dr). Data for 10c: H NMR
(
500 MHz, CD OD) 7.49 (d, J = 8.2 Hz, 4 H, HC(5)), 7.39 (d,
3
J = 8.2 Hz, 4 H, HC(4)), 4.62 (t, J = 4.5 Hz, 2 H, HC(2)), 1.80−
1
3
1
1
(
.69 (m, 2 H, H C(1)), 1.67−1.58 (m, 2 H, H C(1)); C NMR
recrystallization was not cooled to 0 °C. Data for 10d: H
2
2
126 MHz, CD OD) 151.1 (C(3)), 130.3 (q, J = 32.0 Hz, C(6)),
NMR (500 MHz, CD
OD) 7.92 (s, 4 H, HC(4)), 7.82 (s, 2 H,
3
3
1
26.1 (q, J = 3.9 Hz, C(5)), 125.8 (q, J = 270.6 Hz, C(7)), 74.0
HC(6)), 4.91−4.83 (m, 2 H, HC(2)), 1.87−1.78 (m, 4 H,
1
9
13
(C(2)), 36.2 (C(1)); F NMR (470 MHz, CD OD) −63.64; IR
(Nujol) 3236 (m, broad), 1335 (s), 1166 (m), 1114 (s), 1085
(w), 1069 (m), 1040 (w), 1016 (w), 946 (w), 904 (w), 842 (w);
H C(1)); C NMR (126 MHz, CD OD) 150.3 (C(3)), 132.6
2 3
3
(q, J = 33.2 Hz, C(5)), 127.4 (app d, J = 2.8 Hz, C(4)), 124.9 (q,
J = 271.6 Hz, C(7)), 121.8 (m, C(6)), 72.9 (C(2)), 36.1 (C(1));
19
MS (ESI) 517.0 (19, M + 2NaCl), 459.1 (28, M + NaCl), 401.1
M + Na, 32), 361.1 (100), 343.1 (29), 254.2 (20), 236.1 (39);
HRMS (ESI) calcd for C H O F Na 401.0952, found
F NMR (470 MHz, CD OD) −64.14; IR (Nujol) 3413 (w,
3
(
broad), 3182 (w, broad), 1623 (w), 1347 (m), 1279 (s), 1170
(s), 1125 (s), 1040 (w), 904 (m), 844 (w), 710 (m), 682 (m);
MS (ESI, no acid) 515.1 (M + H, 8), 497.1 (100), 479.1 (7);
1
8
16
2 6
4
01.0957; TLC R 0.16 (EtOAc/hexanes, 3/1) [silica gel,
f
2
4
KMnO ]; [α]D +19.0 (c = 0.1, CHCl ); SFC Before
HRMS (ESI, no acid) calcd for C20
515.0880, found 515.0881; TLC R 0.74 (EtOAc/hexanes,
+26.4 (c = 0.2, methanol);
HPLC before recrystallization, (S,S)-10d, t 5.6 min (below
detection limit); meso-10d, t 6.8 min (21%); (R,R)-10d, t
7.6 min (79%), after recrystallization, (S,S)-10d, t 5.6 min
(below detection limit); meso-10d, t 6.8 min (below detection
limit); (R,R)-10d, t 7.6 min (>99%) (Chiralpak AD-H,
200 bar, 1.5 mg/mL, 3.5% i-PrOH in hexanes, 1 mL/min,
H O F12 [M + H]
15 2
4
3
recrystallization: (S,S)-10c, t 9.2 min (below detection limit);
f
R
24
(
R,R)-10c, t 10.2 min (89%); meso-10c, 11.7 min (11%). After
1/1) [silica gel, KMnO ]; [α]
4 D
R
recrystallization: (S,S)-10c, t 9.2 min (below detection limit);
R
R
(
(
R,R)-10c, tR 10.2 min (98%); meso-10c, 11.7 min (2%)
R
R
Chiralpak AD, 200 bar, 3 mg/mL, 5% MeOH in CO , 2 mL/
R
2
min, 220 nm, 40 °C).
R
(
1R, 4R)-1,4-Bis(3,5-bis(trifluoromethyl)phenyl)butane-1,4-
R
diol (10d).
2
2 °C)
Preparation of (2S,5S)-Diarylpyrrolidine-Based Bis-
hydrazone Ligand by the Corey−Itsuno Reduction
Route (Scheme 5). (1R,4R)-1,4-Bis(methanesulfonyloxy)-
1,4-bis(4-trifluoromethylphenyl)butane (11c).
To an oven-dried, 250 mL, one-necked, round-bottom flask
equipped with a stir bar, a short reflux-condenser, and an argon
inlet adaptor with a septum was added sodium borohydride
(
217 mg, 5.64 mmol, 2.4 equiv). After three cycles of evacuation/
An oven-dried, 50 mL, three-necked, round-bottom flask
equipped with a stir bar, two septa, an argon inlet, and an
internal temperature probe was evacuated and backfilled with
argon (three cycles). Dichloromethane (2.6 mL) and meth-
anesulfonyl chloride (210 μL, 2.7 mmol, 2.6 equiv) were added,
and the flask was submerged to a −20 °C bath. A solution of
(1R,4R)-1,4-bis(4-(trifluoromethyl)phenyl)butane-1,4-diol 10c
(390 mg, 1.0 mmol, 1.0 equiv) in dichloromethane (2.6 mL) and
triethylamine (435 μL, 3.1 mmol, 3.0 equiv) was cannulated to
the methanesulfonyl chloride solution at a rate that the tem-
perature did not exceed −10 °C. The resulting heterogeneous
mixture was stirred at −20 °C for 1.5 h and then quenched with a
argon fill, THF (24 mL) was added. Under a positive argon
pressure and with rapid stirring, the septum was removed
temporarily to allow the addition of tin(II) chloride (546 mg,
2
.82 mmol, 1.2 equiv), immediately resulting in gas evolution and
the formation of gray mixture. After the mixture was stirred at
room temperature for 1 h, (S)-(−)-α,α-diphenyl-2-pyrrolidine-
methanol (120 mg, 0.47 mmol, 0.2 equiv) was added. The
mixture was heated to reflux for 0.5 h, and a solution of 1,4-
bis(3,5-bis(trifluoromethyl)phenyl)butane-1,4-dione (1.20 g,
2
.35 mmol, 1 equiv) in THF (18 mL) was added by a syringe
pump (0.5 mmol/h) under mild reflux. The reaction was stirred
for another 0.5 h after complete addition of diketone. The
reaction was cooled to room temperature and then quenched
with MeOH (40 mL) at 0 °C. After bubbling subsided (∼5 min),
ice/water bath was removed and stirred at room temperature for
saturated, aqueous solution of NH Cl (4 mL) with vigorous
4
stirring. The biphasic layers were poured into a separatory funnel
containing EtOAc (20 mL), and the organic layer was washed
sequentially with a 1:2:1 solution of H O−brine-saturated
2
1
0 min. The gray solid was removed by filtration through a pad of
NaHCO (7 mL × 4) and saturated NaHCO (7 mL × 2).
3
3
T
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
The organic layer was dried over Na SO , filtered, and con-
(2S,5S)-1-Amino-2,5-bis(4-trifluoromethylphenyl)-
pyrrolidine (8c).
2
4
centrated to the 7 mL mark. The solution was cooled to 0 °C, and
hexanes (10 mL) was added dropwise to the flask with occasional
swirling. After aging at 0 °C for 0.5 h, the colorless, fine crystals
were collected by filtration and washed with hexanes (5 mL) to
afford 352 mg (64%) of 11c. A second crop of the product was
obtained by concentrating the mother liquor to the 7 mL mark,
cooled to 0 °C for 5 min, filtered, and hexanes washed (5 mL) to
afford 95 mg (17%) of 11c as a white powder. The product
should not be dried under high vacuum because of its tendency
1
to decompose and polymerize. Data for 11c: H NMR (500
To a 10 mL Schlenk reaction flask equipped with a stir bar and a
septum was added (1R,4R)-1,4-bis(methanesulfonyloxy)-1,4-
bis(4-trifluoromethylphenyl)butane 11c (396 mg, 0.74 mmol,
MHz, C D ) 7.25 (d, J = 8.1 Hz, 4 H, HC(5)), 7.09 (d, J = 8.1 Hz,
6
6
4
H, HC(4)), 5.76 (d, J = 10.4 Hz, 2 H, HC(2)), 1.90 (s, 6 H,
1
3
HC(8)), 1.87−1.75 (m, 4 H, H C(1)); C NMR (126 MHz,
CDCl ) 142.2 (C(3)), 131.3 (q, J = 32.1 Hz, C(6)), 126.6
2
1
.0 equiv). The flask was purged with argon, and 2-propanol
3
(
(
1.5 mL) was added to give a suspension. Anhydrous hydrazine
0.47 mL, 15 mmol, 20 equiv) was added, and the flask was
(
C(4)), 126.1 (q, J = 3.9 Hz, C(5)), 123.7 (q, J = 271.4 Hz,
19
C(7)), 81.1 (C(2)), 38.9 (C(8)), 33.1 (C(1)); F NMR
470 MHz, CDCl ) −62.91; MS (ESI) 557.0 (10, M + Na),
submerged to a 40 °C oil bath. After being stirred at this tem-
perature for 14 h, the suspension disappeared. The reaction was
(
3
3
43.1 (100), 159.0 (18); HRMS (ESI) calcd for C H O -
20 20 6
cooled to room temperature, and taken up with Et O (30 mL).
2
F S Na [M + Na] 557.0503, found 557.0499, calcd for
6
2
The mixture was washed with saturated NaHCO (10 mL × 2)
3
C H O F S 535.0684, found 535.0692.
20
21
6 6 2
and then with brine (10 mL). The organic layer was dried over
(
1R,4R)-1,4-Bis(methanesulfonyloxy)-1,4-bis(3,5-bis-
Na SO , filtered, and concentrated to afford 249 mg (90%) of 8c
2
4
(
trifluoromethyl)phenyl)butane (11d).
1
as a white, waxy solid. Data for 8c: H NMR (500 MHz, CDCl )
3
7
.64 (d, J = 8.1 Hz, 4 H, HC(5)), 7.49 (d, J = 8.1 Hz, 4 H,
HC(4)), 4.19 (t, J = 6.5 Hz, 2 H, HC(1)), 2.64−2.33 (m, 4 H,
1
3
HC(2) and NH ), 2.12−1.94 (m, 2 H, HC(2)); C NMR
2
(
126 MHz, CDCl ) 145.5 (C(3)), 129.8 (q, J = 32.5 Hz, C(6)),
3
1
28.6 (C(4)), 125.4 (q, J = 3.8 Hz, C(5)), 124.1 (q, J = 272.5 Hz,
19
C(7)), 69.0 (C(1)), 30.6 (C(2)); F NMR (470 MHz, CDCl )
3
−
62.92; IR (CHCl film) 2954 (w), 2914 (w), 2815 (w), 1618
3
(
m), 1594 (w), 1470 (w), 1421 (m), 1325 (s) 1167 (s), 1124 (s),
067 (s), 1017 (m), 932 (m), 833 (m); MS (ESI) 375.1 (100,
M + H), 358.1 (6); HRMS (ESI) calcd for C H N F [M + H]
1
18
17
2 6
3
75.1296, found 375.1294; TLC R 0.14 (CH Cl /MeOH, 98/2)
f 2 2
[silica gel, UV, KMnO4].
(
2S,5S)- 1-Amino-2,5-bis(3,5-bis(trifluoromethyl)phenyl)-
An oven-dried, 25 mL, three-necked, round-bottom flask
equipped with a stir bar, two septa, an argon inlet, and an
internal temperature probe was evacuated and backfilled with
argon (three cycles). Dichloromethane (2.1 mL) and meth-
anesulfonyl chloride (160 μL, 2.1 mmol, 2.6 equiv) were added,
and the flask was submerged to a −20 °C bath. A solution of
pyrrolidine (8d).
(
1R,4R)-1,4-bis(3,5-bis(trifluoromethyl)phenyl)butane-1,4-diol
0d (400 mg, 0.77 mmol, 1.0 equiv) in dichloromethane
2.1 mL) and triethylamine (325 μL, 2.3 mmol, 3.0 equiv) was
1
(
cannulated to the methanesulfonyl chloride solution at a rate that
the temperature did not exceed −10 °C. The resulting
heterogeneous mixture was stirred at −20 °C for 1.5 h and
then quenched with a saturated, aqueous solution of NH Cl
To a 50 mL, round-bottom flask containing crude (1R,4R)-1,4-
bis(methanesulfonyloxy)-1,4-bis(3,5-bis(trifluoromethyl)-
phenyl)butane 11d obtained above was added 2-propanol
(7.7 mL) and swirled to give a homogeneous and colorless solu-
tion. A stir bar was added, and an argon inlet adaptor equipped
with a septum was attached. The flask was purged with argon, and
anhydrous hydrazine (0.48 mL, 15 mmol, 20 equiv) was added.
The flask was submerged to a 40 °C oil bath and stirred for 14 h.
The reaction was cooled to room temperature and taken up with
Et O (60 mL). The mixture was washed with saturated NaHCO
4
(
3.5 mL) with vigorous stirring. The biphasic layers were poured
into a separatory funnel containing EtOAc (20 mL), and the
organic layer was washed sequentially with a 1:2:1 solution of
H O−brine−saturated NaHCO (6 mL × 4) and saturated
2
3
NaHCO (6 mL × 2). The organic layer was dried over Na SO ,
3
2
4
filtered, and concentrated. The residue was freed from EtOAc by
repeated azeotrope with hexanes (10 mL × 4) under reduced
pressure. The pale yellow gum thus obtained was used without
further purification. The crude product was not dried under high
2
3
vacuum to avoid potential decomposition and polymerization.
(20 mL × 2), then with brine (20 mL). The organic layer was
1
Data for 11d: H NMR (400 MHz, CDCl ) 7.90 (s, 6 H, HC(4)
dried over Na SO , filtered, and concentrated. Purification by
3
2
4
and HC(6)), 6.00−5.92 (m, 2 H, HC(2)), 2.95 (s, 6 H, HC(8)),
column chromatography (SiO , 3.0 × 16 cm, hexanes/CH Cl,
2
2
1
9
2
.20−2.10 (m, 4 H, H C(1)); F NMR (376 MHz, CDCl )
1/1) afforded 313 mg (80%, two steps from 10d) of 8d as a
2
3
1
−63.33.
viscous, colorless oil. Data for 8d: H NMR (500 MHz, CDCl )
3
U
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
7
4
.84 (s, 2 H, HC(6)), 7.83 (s, 4 H, HC(4)), 4.27 (dd, J = 7.1,
(2S,5S)-(2,5-Bis(3,5-bis(trifluoromethyl)phenyl)-
pyrrolidine)-N-iminoacetaldehyde (12d).
.5 Hz, 2 H, HC(1)), 2.68−2.56 (m, 2 H, H C(2)), 2.55 (s, 2 H,
2
1
3
NH ), 2.10−2.01 (m, 2 H, H C(2)); C NMR (126 MHz,
2
2
CDCl ) 143.9 (C(3)), 131.9 (q, J = 33.3 Hz, C(5)), 128.3 (app
3
d, J = 4.0 Hz, C(4)), 123.3 (q, J = 272.9 Hz, C(7)), 121.7 (m,
1
9
C(6)), 68.8 (C(1)), 30.5 (C(2)); F NMR (470 MHz, CDCl )
3
−
63.19; IR (neat) 1322 (w), 2962 (m), 2926 (m), 2871 (w),
2
1
8
794 (w), 1810 (w), 1623 (m), 1465 (m), 1377 (s), 1353 (s),
321 (m), 1279 (s), 1129 (s), 1033 (w), 994 (w), 896 (s),
43 (s), 792 (w), 708 (s), 682 (s); MS (ESI) 511.1 (100);
HRMS (ESI) calcd for C H N F [M + H] 511.1044, found
2
0
15
2
12
To a 10 mL, one-necked, round-bottom flask equipped with a stir
bar was added an aqueous solution of glyoxal (40% w/w, 370 μL,
3.2 mmol, 20 equiv). A solution of N-amino-(2S,5S)-2,5-bis(3,5-
bis(trifluoromethyl)phenyl)pyrrolidine 8d (82 mg, 0.16 mmol,
5
11.1036; TLC R 0.16 (hexanes/CH Cl , 1/1) [silica gel,
f 2 2
PMA].
(
2S,5S)-(2,5-Bis(4-trifluoromethylphenyl)pyrrolidine)-N-
1
.0 equiv) in THF (2.1 mL) was added at 0 °C. The ice/water
iminoacetaldehyde (12c).
bath was removed, and the flask was capped with a glass stopper.
After the mixture was stirred at 22 °C for 1 h, THF was
evaporated under reduced pressure. The residue was taken up by
dichloromethane (10 mL) and basified with saturated NaHCO3
(
15 mL) in a separatory funnel. The organic layer was saved, and
the aqueous layer was extracted further with dichloromethane
10 mL × 2). The combined organic extract was dried over
Na SO , filtered, and concentrated to give a semisolid.
(
2
4
Purification by column chromatography (SiO , 1.5 × 12 cm,
2
hexanes/CH Cl , 1/1, with 1% Et N) afforded 82 mg (93%) of
2
2
3
To a 25 mL, one-necked, round-bottom flask equipped with a stir
bar was added an aqueous solution of glyoxal (40% w/w, 720 μL,
1
1
2d as a white solid. Data for 12d: H NMR (500 MHz, CDCl )
3
9
.20 (d, J = 7.4 Hz, 1 H, HC(9)), 7.87 (s, 2 H, HC(6)), 7.61 (s, 4
6
.3 mmol, 20 equiv). A solution of N-amino-(2S,5S)-2,5-bis(4-
trifluoromethylphenyl)pyrrolidine 8c (118 mg, 0.32 mmol,
.0 equiv) in THF (4.2 mL) was added at 0 °C. The ice/water
H, HC(4)), 6.41 (d, J = 7.4 Hz, 1 H, HC(8)), 5.34 (app br s, 2 H,
HC(1)), 2.75−2.61 (m, 2 H, HC(2)), 2.13−1.99 (m, 2 H,
1
HC(2)); 13C NMR (126 MHz, CDCl ) 190.1 (C(9)), 134.1
3
(
1
C(8)), 132.7 (app d, J = 32.3 Hz, C(5)), 126.2 (br s C(4)),
23.0 (q, J = 273.0 Hz, C(7)), 122.2 (C(4)), 31.4 (br s, C(2));
bath was removed, and the flask was capped with a glass stopper.
After the mixture was stirred at 22 °C for 1 h, THF was evapora-
ted under reduced pressure. The residue was taken up by
dichloromethane (10 mL) and basified with saturated NaHCO3
1
9
C(1) and C(3) were not observed due to hindered rotation; F
NMR (470 MHz, CDCl ) −63.33; IR (CDCl film) 3059 (w),
2
1
1130 (s), 1032 (w), 989 (w), 894 (s), 846 (m), 707 (m), 682 (s);
MS (ESI) 551.1 (100, M + H); HRMS (ESI) calcd for
C H N OF [M + H] 551.0993, found 551.0988; TLC R 0.14
3
3
989 (w), 2947 (w), 1885 (w), 1816 (w), 1681 (s), 1625 (m),
531 (s), 1468 (m), 1378 (s), 1340 (m), 1320 (m), 1279 (s),
(
15 mL) in a separatory funnel. The organic layer was saved, and
the aqueous layer was extracted further with dichloromethane
10 mL × 2). The combined organic extract was dried over
Na SO , filtered, and concentrated to give a very light yellow oil.
(
22 15
2
12
f
2
4
(
hexanes/CH Cl , 1/1 with 0.1% Et N) [silica gel, UV,
2 2 3
Purification by column chromatography (SiO , 1.5 × 16 cm,
2
KMnO4].
CH Cl with 1% Et N) afforded 111 mg (85%) of 12c as a white
2
2
3
N,N′-(Ethane-1,2-diylidene)bis((2S,5S)-(2,5-bis(4-(trifluoro-
1
foam. Data for 12c: H NMR (500 MHz, C D , 60 °C) 9.49 (d,
6
6
methyl)phenyl)pyrrolidin-1-amine)) (13c).
J = 7.5 Hz, 1 H, HC(9)), 7.39 (d, J = 8.0 Hz, 4 H, HC(5)), 6.71
d, J = 7.9 Hz, 4 H, HC(4)), 6.35 (app d, J = 7.4 Hz, 1 H, HC(8)),
(
4
1
.44 (app br s, 2 H, HC(1)), 1.85−1.75 (m, 2 H, HC(2)), 1.25−
1
3
.13 (m, 2 H, HC(2)); C NMR (126 MHz, C D , 60 °C) 188.9
6
6
(
C(9)), 144.9 (C(3), broad), 134.0 (C(8)), 130.2 (q, J =
2.2 Hz, C(6)), 126.6 (C(4)), 126.0 (d, J = 3.9 Hz, C(5)), 124.7
q, J = 272.4 Hz, C(7)), 65.4 (C(1), broad), 30.9 (C(2)); 19
NMR (470 MHz, CDCl ) −62.60. IR (CH Cl film) 3052 (w),
3
(
F
3
2
2
2
1
1
983 (w), 2940 (w), 2815 (w), 1923 (w), 1673 (s), 1620 (m),
529 (s), 1450 (w), 1417 (m), 1388 (s), 1324 (s), 1247 (m),
210 (w), 1128 (s), 1067 (s), 1016 (s), 896 (w), 837 (s), 765
(
(
w), 737 (w), 715 (w); MS (ESI) 415.1 (100, M + H); HRMS
ESI) calcd for C H N OF [M + H] 415.1245, found
2
0
17
2
6
4
15.1238; TLC R 0.37 (CH Cl with 1% Et N) [silica gel, UV,
f 2 2 3
To an oven-dried, 15 mL, one-necked, round-bottom flask
KMnO4].
equipped with a stir bar, an argon gas inlet, and a septum were
V
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
added (2S,5S)-(2,5-bis(4-(trifluoromethyl)phenyl)pyrrolidine)-
N-iminoacetaldehyde 12c (109 mg, 0.27 mmol, 1.0 equiv) and
Na SO (19 mg, 0.13 mmol, 0.5 equiv). A solution of N-amino-
(m), 1466 (w), 1377 (m), 1336 (m), 1319 (m), 1279 (s), 1171
(s), 1129 (s), 987 (w), 894 (m), 846 (m), 707 (m); MS (ESI)
1043.2 (M + H, 100); HRMS (ESI) calcd for C H N F [M +
2
4
42 27
4 24
(
2S,5S)-2,5-bis(4-(trifluoromethyl)phenyl)pyrrolidine 8c (110
H] 1043.1853, found 1043.1853; TLC R 0.34 (hexanes/CH Cl ,
f 2 2
2
4
mg, 0.29 mmol, 1.1.0 equiv) in dichloromethane (1.0 mL) was
added to give a yellow solution. After the flask was purged with
argon, the reaction was stirred at 22 °C for 24 h. The solution was
filtered and concentrated to give a lightly yellow solid. Purifi-
cation by column chromatography (SiO , 1.5 × 16 cm, hexanes/
EtOAc, 9/1, with 1% Et N) afforded 183 mg (90%) of 13c as a
4/1, with 0.5% Et N) [silica gel, UV, KMnO ]; [α] −196.7
3
4
D
(c = 0.15, chloroform).
Preparation of N,N′-(2,2-Dimethylpropane-1,3-
diylidene)bis((2R,5R)-2,5-diphenylpyrrolidin-1-amine)
65
(Scheme 8). 2,2-Dimethylmalonaldehyde.
2
3
1
white powder. Data for 13c: H NMR (500 MHz, CDCl ) 7.58
3
(
d, J = 8.0 Hz, 8 H, HC(5)), 7.22 (d, J = 8.0 Hz, 8 H, HC(4)),
6
2
.55 (s, 2 H, HC(8)), 5.06 (d, J = 7.1 Hz, 4 H, HC(1)), 2.53−
To an oven-dried, 100 mL, three-necked, round-bottom flask
equipped with a stir bar, an argon inlet adaptor, and two septa
was charged 2,2-dimethyl-1,3-propanediol (208 mg, 2.0 mmol,
13
.34 (m, 4 H, HC(2)), 1.80−1.65 (m, 4 H, HC(2)); C NMR
(
126 MHz, CDCl ) 146.9 (C(3)), 134.3 (C(8)), 129.3 (q, J =
3
3
1
2.2 Hz, C(6)), 126.5 (C(4)), 125.5 (app d, J = 3.7 Hz, C(5)),
24.2 (q, J = 271.9 Hz, C(7)), 64.4 (C(1)), 30.9 (C(2)). 19
1
.0 equiv). After three cycles of evacuation/argon fill, dichloro-
F
methane (20 mL) was added and the mixture stirred for 5 min to
NMR (470 MHz, CDCl ) −62.81; IR (CDCl film) 2984 (w),
2
1
3
3
obtain a homogeneous solution. Dess−Martin periodinane
947 (w), 2878 (w), 1618 (m), 1546 (m), 1448 (w), 1417 (m),
324 (s), 1224 (m), 1167 (s), 1123 (s), 1067 (s), 1016 (s), 870
(
2.54 g, 6 mmol, 3.0 equiv) was added in one portion with
rapid stirring. The cloudy solution was stirred at room tempera-
ture for 1 h. The insoluble particulates were removed by filtering
through a pad of Celite (1 cm deep, 15 mL size, medium-porosity
fritted funnel), eluted with dichloromethane (2.5 mL). To the
filtrate was added NaHCO −Na S O (1:1, 15 mL) to give a
(
(
w), 836 (s), 732 (m); MS (ESI) 771.2 (100, M + H); HRMS
ESI) calcd for C H N F [M + H] 771.2357, found 771.2354;
38
31
4 12
TLC R 0.45 (hexanes/EtOAc, 9/1 with 0.1% Et N) [silica gel,
f
3
2
4
UV, KMnO ]; [α] −291.3 (c = 0.15, chloroform).
4
D
3
2
2
3
N,N′-(Ethane-1,2-diylidene)bis((2S,5S)-(2,5-Bis(3,5-bis-
trifluoromethyl)phenyl)pyrrolidin-1-amine)) (13d).
murky mixture. After vigorous stirring for several minutes, both
aqueous and organic layers became transparent and fizzing
subsided. The organic layer was dried over Na SO and filtered to
(
2
4
afford a solution of the target dialdehyde and possibly some
polymeric species. The solution was used directly without con-
centration to minimize polymerization.
(
2R,5R)-((2,5-Diphenylpyrrolidin-1-yl)imino)-2,2-dimethyl-
propanal (14).
28d
A solution of N-amino-(R,R)-2,5-bisphenylpyrrolidine 8e
(
48 mg, 0.2 mmol) in dichloromethane (3 mL) was added to a
solution of 2,2-dimethylmalonaldehyde prepared by oxidation of
the corresponding diol in the scale described in the procedure
above. After the flask was purged with argon, the flask was capped
with a stopper. After being stirred at 22 °C for 2.5 h, the solution
was concentrated. Purification by column chromatography
To an oven-dried, one-piece, 5 mL round-bottom flask and a
reflux condenser equipped with a stir bar, an argon gas inlet and a
septum were added Na SO (10 mg, 0.07 mmol, 0.5 equiv) and a
solution of (2S,5S)-(2,5-bis(3,5-bis(trifluoromethyl)phenyl)-
pyrrolidine)-N-iminoacetaldehyde 12d (80 mg, 0.15 mmol,
2
4
(SiO , 1 × 20 cm, hexanes/CH Cl , 1/1, with 1% Et N) af-
2 2 2 3
1
.0 equiv) and N-amino-(2S,5S)-2,5-bis(4-trifluoromethylphenyl)-
forded 33 mg of 14 as a colorless, viscous oil. The desired product
is contaminated with some other aldehydes based on CHO
chemical shifts in NMR analysis, which may arise from the
condensation of aminopyrrolidine 8e with polymeric species of
pyrrolidine 8d (82 mg, 0.16 mmol, 1.1.0 equiv) in dichloro-
methane. The solvent was carefully removed under reduced
pressure and CH Cl (0.73 mL) was added to give a yellow
solution. After purging the flask with argon, the reaction was
stirred at reflux for 20 h. The solution was filtered and con-
2
2
1
2,2-dimethylmalonaldehyde. Data for 14: H NMR (500 MHz,
CDCl ) 9.22 (s, 1 H, HC(11)), 7.34 (t, J = 7.6 Hz, 4 H, HC(5)),
3
centrated. Purification by column chromatography (SiO , 1.5 ×
7.25 (t, J = 7.4 Hz, 2 H, HC(6)), 7.18 (d, J = 7.2 Hz, 4 H, HC(4)),
5.92 (s, 1 H, HC(7)), 4.99 (d, J = 6.7 Hz, 2 H, HC(1)), 2.56−
2.45 (m, 2 H, H C(2)), 1.87−1.79 (m, 2 H, H C(2)), 1.01 (s,
2
2
0 cm, hexanes/CH Cl , 4/1, with 0.5% Et N) afforded 138 mg
2 2 3
1
(
91%) of 13d as a white solid. Data for 13d: H NMR (500 MHz,
2
2
13
CDCl ) 7.79 (s, 4 H, HC(6)), 7.52 (s, 8 H, HC(4)), 6.50 (s, 2 H,
3 H, HC(9)), 0.98 (s, 3 H, HC(10)); C NMR (126 MHz,
3
HC(8)), 5.26 (dd, J = 7.0, 4.5 Hz, 4 H, HC(1)), 2.53−2.43 (m, 4
CDCl ) 202.8 (C(11)), 143.0 (C(3)), 134.2 (C(7)), 128.3
(C(5)), 126.7 (C(6)), 126.3 (C(4)), 65.1 (C(1)), 49.4 (C(8)),
3
1
3
H, H C(2)), 1.87−1.80 (m, 4 H, H C(2)); C NMR (126 MHz,
2
2
CDCl ) 144.7 (C(3)), 134.7 (C(8)), 132.1 (q, J = 33.4 Hz,
31.5 (C(2)), 20.5 (C(9)), 20.5 (C(10)); IR (CDCl film) 3080
3
3
C(5)), 126.2 (C(4)), 123.1 (q, J = 272.7 Hz, C(7)), 121.4
(w), 3061 (m), 3027 (m), 2971 (s), 2870 (m), 2808 (m), 2704
(w), 1947 (w), 1878 (w), 1802 (w), 1727 (s), 1601 (m), 1580
(m), 1493 (m), 1451 (s), 1391 (m), 1360 (m), 1320 (m), 1303
1
9
(
s, C(6)), 64.0 (C(1)), 30.9 (C(2)). F NMR (470 MHz,
CDCl ) −63.45; IR (CDCl film) 2989 (w), 1624 (w), 1552
3
3
W
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
(
(
(
m), 1268 (m), 1213 (m), 1174 (m), 1139 (m), 1074 (m), 1028
m), 999 (w), 980 (w), 955 (w), 909 (m), 862 (m), 795 (m), 751
s), 700 (s), 677 (w); MS (ESI) 321.2 (100, M + H), 293.2 (15);
added directly to the solution to avoid crystallization on the wall
of the flask.) The resulting light yellow solution was stirred at
−75 ± 1 °C for 3 h. A solution of ZnCl (1 M in Et O, 7.2 mL,
7.2 mmol, 0.72 equiv) was then added to the reaction dropwise
with rapid stirring, keeping the temperature below −69 °C. The
cloudy solution was stirred at −75 ± 1 °C for 0.5 h and then
warmed to 22 °C resulting in a heterogeneous mixture. After
being stirred at 22 °C for 0.5 h, the septum was secured by copper
wire and the flask was brought into a glovebox. Aryl bromide
2
2
HRMS (ESI) calcd for C H N O [M + H] 321.1967, found
3
KMnO4].
2
1
25
2
21.1963; TLC R 0.50 (CH Cl with 1% Et N) [silica gel, UV,
f 2 2 3
N,N′-(2,2-Dimethylpropane-1,3-diylidene)bis((2R,5R)-2,5-
diphenylpyrrolidin-1-amine) (15).
(
0
10 mmol, 1.0 equiv), t-Bu P-HBF (174 mg, 0.6 mmol,
.06 equiv), and Pd(OAc) (110 mg, 0.48 mmol, 0.048 equiv)
2
3 4
were added to the flask with rapid stirring. After the mixture was
stirred at 22 °C for 18 h under argon, the reaction was quenched
with an aqueous solution of NH OH (30%, 1 mL), and the
4
To a 15 mL, one-necked, round-bottom flask equipped with a stir
bar, an argon gas inlet, and a septum were added (2R,5R)-((2,5-
diphenylpyrrolidin-1-yl)imino)-2,2-dimethylpropanal 14
mixture was stirred at room temperature for 1 h. The resulting
slurry was filtered through Celite (2 cm deep, 30 mL, medium-
porosity fritted funnel) and eluted with 60 mL of TBME. The
filtrate was washed with 50 mL of 1 M HCl and then with 50 mL
of deionized water. The organic phase was dried over magnesium
sulfate, filtered, and concentrated to give a brown oil. Purification
by column chromatography provided N-Boc-(R)-2-arylpyrroli-
dine 16.
(
33 mg, 0.1 mmol, 1.0 equiv) and Na SO (7.5 mg, 0.05
2 4
mmol, 0.5 equiv). A solution of (2R,5R)-1amino-2,5-diphenyl-
pyrrolidine 8e (28 mg, 0.12 mmol, 1.2 equiv) in dichloromethane
(
1.5 mL) was added to give a colorless solution. After the flask
was purged with argon, the reaction was stirred at 22 °C for 12 h.
The solution was filtered and concentrated to give a colorless
(−)-Sparteine was recovered by washing its acidic aqueous
film. Purification by column chromatography (SiO , 1 × 30 cm,
solution (in 1 M HCl) with Et O (×6) to remove t-Bu P(O) and
2
3
2
hexanes/CH Cl , 1/1, with 0.2% Et N) afforded 47 mg (84%) of
then basified with an aqueous solution of NaOH (20%). The
aqueous phase was extracted with Et O ( × 4), and the combined
organic phases were dried over K CO , filtered, and concen-
trated. The orange oil was stirred overnight over CaH then
2
2
3
1
5 as a colorless, sticky gel (44% yield from condensation with
2
1
2,2-dimethylmalonaldehyde). Data for 15: H NMR (500 MHz,
2
3
CDCl ) 7.29 (t, J = 7.4 Hz, 8 H, HC(5)), 7.22 (t, J = 7.3 Hz, 4 H,
2
3
HC(6)), 7.11 (d, 8 H, HC(4)), 6.00 (s, 2 H, HC(7)), 4.80 (d, J =
distilled under vacuum (bp: 115−120 °C/0.3 mmHg) to afford
6
.6 Hz, 4 H, HC(1)), 2.50−2.38 (m, 4 H, H C(2)), 1.81−1.71
pure (−)-sparteine as a colorless oil.
2
13
(
m, 4 H, H C(2)), 0.74 (s, 6 H, H C(9)); C NMR (126 MHz,
Preparation of N-Boc-(R)-2-arylpyrrolidine. N-Boc-(R)-
2-(4-methoxyphenyl)pyrrolidine (16a).
2
3
CDCl ) 144.1 (C(3)), 141.4 (C(7)), 128.0 (C(5)), 126.5
3
(
C(4)), 126.3 (C(6)), 65.0 (C(1)), 40.7 (C(8)), 31.5 (C(2)),
2
2
1
4.8 (C(9)); IR (CDCl film) 3085 (m), 3062 (m), 3026 (s),
969 (s), 2871 (s), 1946 (w), 1874 (w), 1806 (w), 1727 (w),
602 (m), 1494 (s), 1453 (s), 1385 (m), 1359 (s), 1303 (s), 1285
3
(
(
(
m), 1214 (s), 1170 (s), 1126 (s), 1074 (m), 1052 (m), 1028
m), 981 (m), 950 (w), 910 (s), 868 (w), 799 (w), 749 (s), 648
m); MS (ESI) 541.3 (100, M + H), 291.2 (8); HRMS (ESI)
calcd for C H N [M + H] 541.3331, found 541.3333; TLC R
Following general procedure 3, (−)-sparteine (2.8 mL,
12 mmol, 1.2 equiv), N-Boc-pyrrolidine (2.1 mL, 12 mmol,
1.2 equiv) and TBME (30 mL), s-BuLi (1.74 M in cyclohexanes,
6.9 mL, 12 mmol, 1.2 equiv), ZnCl (1 M in Et O, 7.2 mL,
37
41
4
f
24
0
.66 (CH Cl with 1% Et N) [silica gel, UV, I ]; [α] −184.8
2
2
3
2
D
(c = 0.15, chloroform).
Preparation of (2R)-Aryl- and (2R,5R)-Diarylpyrroli-
2
2
dine-Based Bis-hydrazone Ligands by the α-Arylation
Route (Scheme 6 and Tables 2−4). General Procedure 3:
7.2 mmol, 0.72 equiv), 4-bromoanisole (1.25 mL, 10 mmol,
1.0 equiv), t-Bu P-HBF (174 mg, 0.6 mmol, 0.06 equiv), and
3
4
3
2,33
α-Arylation of N-Boc-pyrrolidine (Table 2).
Pd(OAc) (110 mg, 0.48 mmol, 0.048 equiv) were combined in a
2
1
00 mL Schlenk reaction flask. After the mixture was stirred at
2 °C for 18 h under argon, the reaction was quenched, filtered,
2
subjected to aqueous washes, and then concentrated. Purification
by column chromatography (SiO , 4.5 × 26 cm, hexanes/EtOAc,
2
gradient elution, 14/1, 9/1 then 7/1) afforded 1.61 g (71%) of
1
6a as a slightly tanned white solid (a rotameric mixture, 66:34,
−
20 °C). The spectroscopic data matched those from the
66
1
literature. Data for 16a: H NMR (500 MHz, CDCl , −20 °C)
3
major 7.07 (d, J = 8.6 Hz, 2 H, HC(9)), 6.82 (d, J = 8.6 Hz, 2 H,
HC(10)), 4.71 (dd, J = 7.7, 4.3 Hz, 1 H, HC(1)), 3.79 (s, 3 H,
H C(12)), 3.66−3.52 (m, 2 H, H C(4)), 2.32−2.24 (m, 1 H,
An oven-dried, 100 mL Schlenk reaction flask equipped with a
stir bar, a septum, and an internal temperature probe was
evacuated then filled with argon (three cycles). (−)-Sparteine
3
2
HC(2)), 1.92−1.74 (m, 3 H, HC(2) and H C(3)), 1.18 (s, 9 H,
2
(
2.8 mL, 12 mmol, 1.2 equiv), N-Boc-pyrrolidine (2.1 mL,
H C(7)), minor 7.10 (d, J = 8.5 Hz, 2 H, HC(9)), 6.83 (d, J =
3
1
2 mmol, 1.2 equiv), and TBME (30 mL) were added. The
8.5 Hz, 2 H, HC(10)), 4.89 (dd, J = 8.0, 2.9 Hz, 1 H, HC(1)),
3.77 (s, 3 H, H C(12)), 3.50−3.44 (m, 2 H, H C(4)), 2.25−2.19
colorless solution was cooled to −75 °C, and s-BuLi (1.74 M in
cyclohexanes, 6.9 mL, 12 mmol, 1.2 equiv) was added dropwise,
keeping the temperature below −69 °C. (Note: s-BuLi should be
3
2
(m, 1 H, HC(2)), 1.92−1.74 (m, 3 H, HC(2) and H C(3)), 1.44
2
13
(s, 9 H, H C(7)); C NMR (126 MHz, CDCl , −20 °C), major
3
3
X
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
57.9 (C(11)), 154.6 (C(5)), 137.1 (C(8)), 126.5 (C(9)), 113.2
Article
24
1
(hexanes/CH Cl , 1/4) [silicagel, KMnO ];[α] +68.9 (c = 0.5,
2 2 4 D
(C(10)), 79.0 (C(6)), 60.5 (C(1)), 55.2 (C(12)), 46.9 (C(4)),
acetone).
3
1
6.0 (C(2)), 28.1 (C(7)), 23.0 (C(3)), minor 158.0 (C(11)),
54.4 (C(5)), 136.1 (C(8)), 126.3 (C(9)), 113.6 (C(10)), 79.1
(R)-2-(3,5-Bis(trifluoromethyl)phenyl)pyrrolidine (17d).
C(6)), 59.9 (C(1)), 55.2 (C(12)), 47.2 (C(4)), 34.9 (C(2)),
2
2
1
8.4 (C(7)), 23.3 (C(3)); IR (CDCl film) 2975 (s), 2836 (m),
058 (w), 1882 (w), 1693 (s), 1681 (s), 1651 (m), 1613 (s),
586 (m), 1514 (s), 1454 (s), 1392 (s), 1246 (s), 1163 (s), 1113
3
(
8
1
s), 1079 (m), 1036 (s), 971 (w), 917 (m), 901 (m), 875 (m),
To an oven-dried, 10 mL, round-bottom flask equipped with a
stir bar, an argon inlet adaptor, and a septum was added N-Boc-
+
28 (s), 768 (m); MS (EI, 70 eV) 277.2 (M , 7), 220.2 (100),
76.2 (84), 148.1 (22); HRMS (ESI) calcd for C H NO [M +
1
6
24
3
(
0
R)-2-(3,5-bis(trifluoromethyl)phenyl)pyrrolidine 16d (61 mg,
.16 mmol, 1.0 equiv). After three cycles of evacuation and argon
H] 278.1756, found 278.1765; TLC R 0.16 (hexanes/EtOAc, 9/
1
f
2
4
) [silica gel, UV, KMnO ]; [α] +90.6 (c = 0.5, acetone); SFC
4 D
fill, dichloromethane (1.6 mL) was added. The colorless solution
was cooled to 0 °C, and TMSI (25 μL, 0.17 mmol, 1.05 equiv)
was added. The resulting light burgundy solution was stirred at
(S)-16a, t 3.8 min (4%); (R)-16a, t 6.7 min (96%) ((R,R)-
R
R
Whelk-O1, 200 bar, 5% MeOH in CO , 2.5 mL/min, 220 nm,
4
2
0 °C).
0
°C for 1.5 h and then quenched with an aqueous of saturated
N-Boc-(R)-2-(3,5-bis(trifluoromethyl)phenyl)pyrrolidine
NaHCO /saturated Na S O (1:1, 3.2 mL). After vigorous stir-
3
2
2
3
(16d).
ring for a few minutes, the organic phase was saved and the
aqueous phase was extracted with dichloromethane (1.5 mL × 3).
The combined organic extracts was dried over Na SO , filtered,
2
4
and concentrated. Purification by column chromatography (SiO2,
× 16 cm, CH Cl /MeOH, 98/2) afforded 40 mg (88%) of 17d
1
2
2
1
as a pale yellow liquid. Data for 17d: H NMR (500 MHz, CDCl )
3
7
.86 (s, 2 H, HC(6)), 7.73 (s, 1 H, HC(8)), 4.28 (t, J = 4.3 Hz, 1 H,
HC(1)), 3.22−3.17 (m, 1 H, HC(4)), 3.12−3.06 (m, 1 H,
Following general procedure 3, (−)-sparteine (2.4 mL,
HC(4)), 2.31−2.24 (m, 1 H, HC(2)), 1.98 (s, 1 H, NH), 1.96−
1
1
5
6
0 mmol, 1.2 equiv), N-Boc-pyrrolidine (1.8 mL, 10 mmol,
.2 equiv), TBME (26 mL), s-BuLi (1.74 M in cyclohexanes,
.8 mL, 10 mmol, 1.2 equiv), ZnCl (1 M in Et O, 6.1 mL,
1
3
1
(
.82 (m, 2 H, HC(3)), 1.67−1.59 (m, 1 H, HC(2)); C NMR
126 MHz, CDCl ) 148.4 (C(5)), 131.4 (q, J = 33.0 Hz, C(7)),
3
2
2
1
26.7 (C(6)), 123.5 (q, J = 272.6 Hz, C(9)), 120.6 (C(8)), 61.4
.1 mmol, 0.72 equiv), 1-bromo-3,5-bis(trifluoromethyl)-
19
(
C(1)), 47.0 (C(4)), 34.9 (C(2)), 25.5 (C(3)); F NMR (470
benzene (1.45 mL, 8.4 mmol, 1.0 equiv), t-Bu P-HBF (146 mg,
3
4
MHz, CDCl ) −63.21;IR(CDCl film) 2965 (w), 2871 (w), 1622
0
.5 mmol, 0.06 equiv), and Pd(OAc) (90 mg, 0.4 mmol,
3
3
2
(
8
w), 1464 (w), 1380 (m), 1352 (w), 1279 (s), 1171 (s), 1131 (s),
96 (m), 842 (m).
R)-(2-(3,5-Bis(trifluoromethyl)phenyl)pyrrolidin-1-yl)(3,5-
0.048 equiv) were combined in a 100 mL Schlenk reaction flask.
After the mixture was stirred at 21 °C for 18 h, the reaction was
quenched, filtered, subjected to aqueous washes, and then con-
(
dinitrophenyl)methanone (17d-DNB).
centrated. Purification by column chromatography (SiO , 4.5 ×
2
2
1
6
3 cm, CH Cl /hexanes, 4/1) and Kugelrohr distillation afforded
.73 g (54%) of 16d as a very pale yellow oil (a rotameric mixture,
4:36, −20 °C). This compound was derivatized to 3,5-dinitro-
2 2
phenylmethane for the assessment of enantiomeric purity (96:4
er). Data for 16d: bp 150 °C [0.1 mmHg, ABT]; H NMR
1
(
2
(
500 MHz, CDCl −20 °C) major 7.75 (s, 1 H, HC(11)), 7.62 (s,
3
,
H, HC(9)), 4.82 (dd, J = 7.6, 5.8 Hz, 1 H, HC(1)), 3.72−3.62
m, 2 H, HC(4)), 2.45−2.42 (m, 1 H, HC(2)), 1.93−1.87 (m, 2
H, H C(3)), 1.86−1.82 (m, 1 H, HC(2)), 1.14 (s, 9 H, H C(7)),
2
3
minor 7.72 (s, 1 H, HC(11)), 7.59 (s, 2 H, HC(9)), 5.01 (dd, J =
.1, 3.5 Hz, 1 H, HC(1)), 3.58−3.52 (m, 2 H, HC(4)), 2.40−2.33
m, 1 H, HC(2)), 1.93−1.87 (m, 2 H, H C(3)), 1.81−1.75 (m, 1
8
(
2
1
3
H, HC(2)), 1.44 (s, 9 H, H C(7)); C NMR (126 MHz, CDCl ,
To a 10 mL, one-necked round-bottom flask equipped with a stir
bar, an argon inlet adaptor, and a septum was added (R)-2-(3,5-
bis(trifluoromethyl)phenyl)pyrrolidine 17d (38 mg, 0.13 mmol,
1.0 equiv). After three quick cycles of evacuation and argon fill
(17d is volatile), THF (1.4 mL), 3,5-dinitrobenzoyl chloride
3
3
−
20 °C) major 154.1 (C(5)), 147.7 (C(8)), 131.3 (q, J = 33.4 Hz,
C(10)), 125.8 (C(9)), 123.2 (q, J = 272.6 Hz, C(12)), 120.5
C(11)), 79.9 (C(6)), 60.8 (C(1)), 47.2 (C(4)), 36.1 (C(2)),
(
2
1
7.9 (C(7)), 23.4 (C(3)), minor 154.5 (C(5)), 146.6 (C(8)),
31.4 (q, J = 33.2 Hz, C(10)), 125.4 (C(9)), 123.3 (q, J = 272.6
(35 mg, 0.15 mmol, 1.1.0 equiv) and Et N (190 μL, 1.34 mmol,
3
Hz, C(12)), 120.5 (C(11)), 80.1 (C(6)), 60.2 (C(1)), 47.4
10 equiv) were added to give a heterogeneous mixture. After the
mixture was stirred at 21 °C for 4.5 h, THF was removed under
1
9
(
C(4)), 34.8 (C(2)), 28.3 (C(7)), 23.5 (C(3)). F NMR (470
MHz, CDCl ) −63.31; IR (neat) 2978 (m), 2882 (m), 1699 (s),
reduced pressure. The crude product was taken up with Et O
(1.5 mL) and washed with 5% NaOH (3 mL). The organic phase
3
2
1
624 (w), 1479 (m), 1457 (m),1392 (s), 1278 (s), 1169 (s), 1133
(
(
(
s), 1027 (w), 974 (w), 928 (w), 898 (m), 846 (w), 776 (w), 707
m). MS (EI, 70 eV) 383.1 (M , 7), 328.1 (100), 308.1 (50), 282.1
48), 267.1 (49), 255.1 (48), 227.0 (34); HRMS (ESI) calcd for
was saved, and the aqueous phase was extracted with Et O
(1.5 mL × 3). The combined organic phases was dried over
2
+
Na SO , filtered, and concentrated to give a brown sticky oil.
2
4
C H NO F [M + H] 384.1398, found 384.1404; TLC R 0.26
Purification by column chromatography (SiO , 1 × 21 cm,
17
20
2
6
f
2
Y
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
CH Cl /hexanes, 9/1) afforded 61 mg (94%) of 17d-DNB as a
(C(7)), 23.4 (C(3)); IR (CDCl film) 3053 (m), 2973 (m),
2
2
3
white, foamy liquid. Spectroscopic data for major conformer is
2876 (m), 1693 (s), 1634 (w), 1602 (w), 1509 (w), 1478 (m),
1454 (m), 1392 (s), 1364 (s), 1255 (m), 1165 (s), 1109 (s), 1081
1
shown. Data for 17d-DNB: H NMR (500 MHz, CDCl ) 9.14
3
(
7
(
s, 1 H, HC(9)), 8.77 (s, 2 H, HC(7)), 7.80 (s, 1 H, HC(13)),
.77 (s, 2 H, HC(11)), 5.36 (t, J = 7.0 Hz, 1 H, HC(1)), 3.96
app q, J = 7.5 Hz, 1 H, HC(4)), 3.75−3.71 (m, 1 H, HC(4)),
.64−2.58 (m, 1 H, HC(2)), 2.18−2.00 (m, 3 H, HC(2) and
(
(
w), 974 (w), 912 (w), 896 (w), 856 (w), 817 (m), 746 (m). MS
EI, 70 eV) 297.1 (M , 20), 241.1 (80), 196.1 (100), 168.0 (34),
+
1
54.0 (23), 84.0 (23), 57.1 (75); HRMS (ESI) calcd for
2
13
C H NO [M + H] 298.1807, found 298.1801; TLC R 0.11
H C(3)); C NMR (126 MHz, CDCl ) 165.2 (C(5)), 148.5
19 24
2
f
2
3
24
(
C(Aryl)), 144.7 (C(Aryl)), 139.2 (C(Aryl)), 123.1 (q, J = 33.5
(CH
2
Cl
2
) [silica gel, UV, KMnO
]; [α]
4
D
+117.7 (c = 0.5,
Hz, C(12)), 127.7 (C(Aryl)), 126.2 (C(Aryl)), 123.2 (q, J =
acetone). SFC (S)-16f, t 3.8 min (4%); (R)-16f, t 11.6 min
R
R
2
72.8 Hz, C(14)), 122.6 (C(Aryl)), 120.4 (C(Aryl)), 61.9
(
96%) ((R,R)-Whelk-O1, 200 bar, 10% MeOH in CO , 2.5mL/min,
2
(
(
(
(
C(1)), 51.2 (C(4)), 34.6 (C(2)), 25.6 (C(3)); 19F NMR
220 nm, 40 °C).
470 MHz, CDCl ) −63.23; IR (CDCl film) 3104 (m), 2981
3
3
N-Boc-(R)-2-(3,5-dimethylphenyl)pyrrolidine (16g).
m), 2883 (m), 1818 (w), 1643 (s), 1591 (m), 1546 (s), 1467
s), 1409 (s), 1380 (s), 1344 (s), 1278 (s), 1181 (s), 1132 (s),
9
4
68 (m), 870 (m), 846 (m), 836 (m), 706 (s). MS (EI, 70 eV)
78.1 (M + H, 100); HRMS (ESI) calcd for C H N O F [M +
1
9
14
3
5 6
H] 478.0838, found 478.0835; TLC R 0.23 (CH Cl ) [silica gel,
f
2
2
UV]; SFC (S)-17d-DNB, t 7.6 min (4%); (R)-17d-DNB, t
R
R
1
2
8.5 min (96%) ((R,R)-Whelk-O1, 200 bar, 5% MeOH in CO2,
.5 mL/min, 220 nm, 40 °C).
N-Boc-(R)-2-(2-naphthyl)pyrrolidine (16f).
Following general procedure 3, (−)-sparteine (2.4 mL,
1
1
5
6
8
0
0 mmol, 1.2 equiv), N-Boc-pyrrolidine (1.8 mL, 10 mmol,
.2 equiv), TBME (26 mL), s-BuLi (1.74 M in cyclohexanes,
.8 mL, 10 mmol, 1.2 equiv), ZnCl (1 M in Et O, 6.1 mL,
2
2
.1 mmol, 0.72 equiv), 1-bromo-3,5-dimethylbenzene (1.15 mL,
.4 mmol, 1.0 equiv), t-Bu P-HBF4 (148 mg, 0.5 mmol,
3
.06 equiv), and Pd(OAc) (92 mg, 0.4 mmol, 0.048 equiv)
2
were combined in a 100 mL Schlenk reaction flask. After the
mixture was stirred at 22 °C for 18 h, the reaction was quenched,
filtered, subjected to aqueous washes, and then concentrated.
Following general procedure 3, (−)-sparteine (2.8 mL,
1
1
6
7
1
2 mmol, 1.2 equiv), N-Boc-pyrrolidine (2.1 mL, 12 mmol,
.2 equiv), TBME (30 mL), s-BuLi (1.74 M in cyclohexanes,
.9 mL, 12 mmol, 1.2 equiv), ZnCl (1 M in Et O, 7.2 mL,
Purification by column chromatography (SiO , 4.5 × 23 cm,
2
2
2
CH Cl /hexanes, 9/1) and Kugelrohr distillation afforded 1.30 g
2
2
.2 mmol, 0.72 equiv), 2-bromonaphthalene (2.07 g, 10 mmol,
.0 equiv), t-Bu P-HBF (174 mg, 0.6 mmol, 0.06 equiv), and
(
−
56%) of 16g as a colorless oil (a rotameric mixture, 66:34,
20 °C). Data for 16g: bp 175 °C [0.02 mmHg, ABT]; H NMR
3
4
1
Pd(OAc) (110 mg, 0.48 mmol, 0.048 equiv) were combined in a
2
1
00 mL Schlenk reaction flask. After being stirred at 22 °C for
(500 MHz, CDCl3, −20 °C) major 6.84 (s, 1 H, HC(11)), 6.76
(s, 2 H, HC(9)), 4.68 (dd, J = 7.7, 4.2 Hz, 1 H, HC(1)), 3.65−
16 h, the reaction was quenched, filtered, subjected to aqueous
washes, and then concentrated. Purification by column chro-
matography (SiO , 4.5 × 20 cm, CH Cl /hexanes, gradient
elution, 4/1, 9/1 then 1/0) afforded 1.82 g (61%) of 16f as a
white solid (a rotameric mixture, 70:30, −20 °C). The spectro-
scopic data matched those from the literature. Data for 16f: H
NMR (500 MHz, CDCl , −20 °C) major 7.85−7.77 (m, 3 H,
HC(Aryl)), 7.58 (s, 1 H, HC(8)), 7.50−7.40 (m, 2 H,
HC(Aryl)), 7.35−7.30 (m, 1 H, HC(16) or HC(10)), 4.95
3
.56 (m, 2 H, HC(4)), 2.31−2.20 (m, 1 H, HC(2)), 2.28 (s, 6 H,
2
2
2
HC(12)), 1.93−1.79 (m, 3 H, H C(3) and HC(2)), 1.17 (s, 9 H,
2
H C(7)), minor 6.84 (s, 1 H, HC(11)), 6.76 (s, 2 H, HC(9)),
3
24
1
4.88−4.84 (m, 1 H, HC(1)), 3.51−3.46 (m, 2 H, HC(4)), 2.31−
2
.20 (m, 1 H, HC(2)), 2.28 (s, 6 H, HC(12)), 1.93−1.72 (m, 3
3
1
3
H, H C(3) and HC(2)), 1.45 (s, 9 H, H C(7)); C NMR
2
3
(126 MHz, CDCl3, −20 °C), major 154.6 (C(5)), 144.8
(C(10)), 137.4 (C(8)), 127.9 (C(11)), 123.3 (C(9)), 79.3
(
dd, J = 7.9, 3.6 Hz, 1 H, HC(1)), 3.76−3.63 (m, 2 H, H C(4)),
2
2
.43−2.29 (m, 1 H, HC(2)), 1.97−1.83 (m, 3 H, HC(2) and
(
C(6)), 61.0 (C(1)), 46.9 (C(4)), 35.8 (C(2)), 28.0 (C(7)),
H C(3)), 1.13 (s, 9 H, H C(7)), minor 7.85−7.77 (m, 3 H,
2
3
2
1
3.1 (C(3)), 21.3 (C(12)), minor 154.4 (C(5)), 144.0 (C(10)),
37.7 (C(8)), 128.3 (C(11)), 122.9 (C(9)), 79.1 (C(6)), 60.5
HC(Aryl)), 7.59 (s, 1 H, HC(8)), 7.50−7.40 (m, 2 H,
HC(Aryl)), 7.35−7.30 (m, 1 H, HC(16) or HC(10)), 5.11 (d,
(
2
C(1)), 47.3 (C(4)), 34.9 (C(2)), 28.4 (C(7)), 23.4 (C(3)),
1.4 (C(12)). IR (neat) 3373 (w), 2972 (s), 1693 (s), 1681 (s),
J = 6.5 Hz, 1 H, HC(1)), 3.59−3.55 (m, 2 H, H C(4)), 2.43−
2
2
.29 (m, 1 H, HC(2)), 1.97−1.83 (m, 3 H, HC(2) and H C(3)),
.47 (s, 9 H, H C(7)); C NMR (126 MHz, CDCl , −20 °C)
2
13
1
1651 (m), 1605 (s), 1454 (s), 1392 (s), 1255 (s), 1163 (s), 1112
(s), 1036 (m), 1009 (m), 972 (m), 951 (m), 918 (m), 897 (m),
3
3
major 154.7 (C(5)), 142.2 (C(9)), 133.0 (C(12) or C(17)),
+
1
1
32.2 ((C(12) or C(17)), 127.9 (C(Aryl)), 127.5 (C(Aryl)),
27.5 (C(Aryl)), 126.0 (C(Aryl)), 125.3 (C(Aryl)), 124.1
8
7
1
2
43 (m), 774 (m), 730 (m), 704 (s). MS (EI, 70 eV) 275.2 (M ,
), 219.1 (100), 204.1 (24), 174.1 (99), 160.1 (56), 147.1 (28),
32.1 (74); HRMS (ESI) calcd for C H NO [M + H]
(
C(8)), 123.6 (C(8) or C(10)), 79.2 (C(6)), 61.2 (C(1)),
7.0 (C(4)), 35.7 (C(2)), 28.0 (C(7)), 23.0 (C(3)), minor 154.5
C(5)), 141.3 (C(9)), 133.2 (C(12) or C(17)), 132.4 (C(12) or
C(17)), 128.2 (C(Aryl)), 127.7 (C(Aryl)), 127.5 (C(Aryl)),
25.8 (C(Aryl)), 125.3 (C(Aryl)), 124.0 (C(8)), 123.5 (C(8) or
C(10)), 79.3 C(6)), 60.7 (C(1)), 47.4 (C(4)), 34.7 (C(2)), 28.4
17
26
2
4
(
76.1964, found 276.1971; TLC R 0.17 (CH Cl ) [silica gel,
UV, KMnO ]; [α]D +88.3 (c = 0.51, acetone); SFC (S)-16g, tR
f
2
2
24
4
1
3.1 min (4%); (R)-16g, t 5.6 min (96%) ((R,R)-Whelk-O1, 200
R
bar, 5% MeOH in CO , 2.5 mL/min, 220 nm, 40 °C).
2
Z
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
N-Boc-(R)-2-(5-phenylbiphenyl-3-yl)pyrrolidine (16h).
Pd(OAc) (132 mg, 0.58 mmol, 0.048 equiv) were combined in a
2
100 mL Schlenk reaction flask. After the mixture was stirred at
2 °C for 18 h, the reaction was quenched, filtered, subjected to
2
aqueous washes, and then concentrated. Purification by column
chromatography (SiO , 4.5 × 21 cm, hexanes/EtOAc, 19/1) and
2
Kugelrohr distillation afforded 1.69 g (54%) of 16i as a colorless
oil (a rotameric mixture, 66:34, −20 °C). N-Boc-pyrrolidine was
removed at 80−90 °C [0.02 mmHg, ABT]. The spectroscopic
32
data matched those from the literature. Data for 16i: bp 125 °C
Following general procedure 3, (−)-sparteine (2.8 mL,
1
[
0.02 mmHg, ABT]; H NMR (500 MHz, CDCl , −20 °C)
3
12 mmol, 1.2 equiv), N-Boc-pyrrolidine (2.1 mL, 12 mmol, 1.2
major 7.17−7.09 (m, 3 H, HC(10), HC(11) and HC(12)), 7.07
(
equiv), TBME (30 mL), s-BuLi (1.74 M in cyclohexanes, 6.9 mL,
m, 1 H, HC(13)), 4.94 (dd, J = 7.9, 4.5 Hz, 1 H, HC(1)), 3.71−
1
0
1
2 mmol, 1.2 equiv), ZnCl (1 M in Et O, 7.2 mL, 7.2 mmol,
.72 equiv), 1-bromo-3,5-diphenylbenzene (3.10 g, 10 mmol,
2 2
3.57 (m, 2 H, H C(4)), 2.32 (s, 3 H, H C(14)), 2.36−2.18 (m, 1
2
3
H, HC(2)), 1.98−1.82 (m, 2 H, H C(3)), 1.75−1.65 (m, 2 H,
2
.0 equiv), t-Bu P-HBF (175 mg, 0.6 mmol, 0.06 equiv), and
3
4
H C(3)), 1.13 (s, 9 H, H C(7)), minor 7.17−7.09 (m, 3 H,
2
3
Pd(OAc) (112 mg, 0.48 mmol, 0.048 equiv) were combined in a
2
HC(10), HC(11) and HC(12)), 7.03 (d, J = 7.1 Hz, 1 H,
HC(13)), 5.11 (dd, J = 8.2, 1.8 Hz, 1 H, HC(1)), 3.71−3.57 (m,
1
2
00 mL Schlenk reaction flask. After the mixture was stirred at
2 °C for 18 h, the reaction was quenched, filtered, subjected to
1
H, H C(4)), 3.50−3.45 (m, 1 H, H C(4)), 2.34 (s, 3 H,
2
2
aqueous washes, and then concentrated. Purification by column
chromatography (SiO , 4.5 × 26 cm, CH Cl /hexanes, gradient
H C(14)), 2.36−2.18 (m, 1 H, HC(2)), 1.98−1.82 (m, 2 H,
3
13
2
2
2
HC(3)), 1.75−1.65 (m, 2 H, HC(3)), 1.15 (s, 9 H, H C(7)); C
3
elution, 6/4, 7/3, 8/2 then 9/1) afforded 2.44 g (61%) of 16h as a
foamy, white solid (a rotameric mixture, 65:35, −20 °C). The
NMR (126 MHz, CDCl , −20 °C) major 154.3 (C(5)), 143.0
3
(
C(8)), 133.8 (C(9)), 129.8 (C(10)), 126.1 (C(11) or C(12)),
24
spectroscopic data matched those from the literature. Data for
1
25.8 (C(11) or C(12)), 124.2 (C(13)), 78.9 (C(6)), 57.7
1
1
7
2
1
3
6h: H NMR (500 MHz, CDCl , −20 °C) major 4.89 (dd, J =
3
(
C(1)), 46.9 (C(4)), 33.9 (C(2)), 27.9 (C(7)), 23.0 (C(3)),
.6, 4.6 Hz, 1 H, HC(1)), 3.75−3.60 (m, 2 H, H C(4)), 2.45−
2
1
1
9.4 (C(14)), minor 154.3 (C(5)), 141.5 (C(8)), 134.0 (C(9)),
30.4 (C(10)), 126.4 (C(11) or C(12)), 125.7 (C(11) or
.38 (m, 1 H, HC(2)), 2.04−1.86 (m, 3 H, HC(2) and H C(3)),
2
.20 (s, 9 H, H C(7)), minor 5.12 (d, J = 6.6, 1 H, HC(1)), 3.57−
3
C(12)), 123.7 (C(13)), 79.2 (C(6)), 57.8 (C(1)), 47.2 (C(4)),
.51 (m, 2 H, H C(4)), 2.38−2.30 (m, 1 H, HC(2)), 2.04−1.86
2
3
2
1
2.5 (C(2)), 28.4 (C(7)), 22.9 (C(3)), 19.3 (C(14)). IR (neat)
973 (s), 2874 (m), 1697 (s), 1605 (w), 1479 (m), 1455 (m),
392 (s), 1364 (s), 1275 (m), 1246 (m), 1161 (s), 1121 (s), 1102
(
m, 3 H, HC(2) and H C(3)), 1.51 (s, 9 H, H C(7)), aromatic
2
3
protons for major and minor rotameric isomers 7.68−7.62 (m, 5
1
3
H), 7.50−7.45 (m, 4 H), 7.41−7.36 (m, 4 H); C NMR (126
(
m), 1078 (m), 1051 (w), 1032 (w), 971 (w), 921 (m), 875 (m),
MHz, CDCl , −20 °C) major 154.5 (C(5)), 79.4 (C(6)), 61.4
+
3
7
71 (m), 752 (m), 725 (m); MS (EI, 70 eV) 261.2 (M , 3), 205.1
67), 190.1 (15), 160.1 (25), 144.1 (44), 133.1 (26), 114.1 (29),
1.1 (12), 70.1 (30), 57.1 (100); HRMS (ESI) calcd for
C H NO [M + H] 262.1807, found 262.1798; TLC R 0.15
(
C(1)), 47.1 (C(4)), 36.1 (C(2)), 28.1 (C(7)), 23.3 (C(3)),
(
9
minor 154.5 (C(5)), 79.3 C(6)), 60.5 (C(1)), 47.3 (C(4)), 34.9
C(2)), 28.4 (C(7)), 23.3 (C(3)), aromatic carbons for major
and minor rotameric isomers: 146.3, 144.8, 141.7, 141.5, 141.2,
(
16
24
2
f
24
(
hexanes/EtOAc, 19/1) [silica gel, KMnO ]; [α]D +70.9 (c =
4
1
1
1
1
1
7
2
40.9, 128.7, 128.6, 127.4, 127.3, 127.2, 127.1, 124.8, 124.4,
0
.5, acetone); SFC (S)-16i, t 7.1 min (5%); (R)-16i, t 10.1 min
R R
23.2, 123.0; IR (CDCl film) 3033 (w), 2973 (m), 2874 (w),
3
(
95%) ((R,R)-Whelk-O1, 200 bar, 2% MeOH in CO2,
693 (s), 1596 (m), 1577 (w), 1498 (m), 1477 (m), 1454 (m),
435 (m), 1392 (s), 1365 (s), 1249 (w), 1165 (s), 1116 (m),
081 (w), 1029 (w), 971 (w), 871 (w), 758 (s), 698 (s); MS (EI,
2
mL/min, 220 nm, 40 °C).
N-Boc-(R)-2-(5-methyl-2-thienyl)pyrrolidine (16j).
+
0 eV) 399.2 (M , 20), 343.1 (72), 298.1 (100), 270.1 (45),
56.1 (65), 241.1 (22), 194.1 (19), 165.1 (10), 70.1 (31), 57.1
(
84); HRMS (ESI) calcd for C H NO [M + H] 400.2277,
27 30 2
found 400.2283; TLC R 0.15 (CH Cl /hexanes, 9/1) [silica gel,
f
2
2
UV]; [α]D24 +83.1 (c = 0.5, acetone); SFC (S)-16h, t 18.9 min
R
(
4%); (R)-16h, t 21.0 min (96%) (Chiralpak-OD, 200 bar,
R
1
−10% MeOH gradient in CO (30 min), 2 mL/min, 220 nm,
An oven-dried, three-necked, round-bottom flask equipped with
a stir bar, a septum, an internal temperature probe, and a solid
addition bulb containing t-Bu P-HBF (273 mg, 0.94 mmol, 0.06
2
40 °C).
N-Boc-(R)-2-(2-tolyl)pyrrolidine (16i).
3
4
equiv) and Pd(OAc) (174 mg, 0.75 mmol, 0.048 equiv) was
2
carefully evacuated and then filled with argon (two cycles).
(
−)-Sparteine (4.4 mL, 18.8 mmol, 1.2 equiv), N-Boc-pyrro-
lidine (3.3 mL, 18.8 mmol, 1.2 equiv), and TBME (47 mL) were
added. The colorless solution was cooled to −75 °C, and s-BuLi
(
1.47 M in cyclohexanes, 12.9 mL, 18.8 mmol, 1.2 equiv) was
added dropwise, keeping the temperature below −69 °C. (Note:
s-BuLi should be added directly to the solution to avoid
crystallization on the wall of the flask.) The resulting light yellow
Following general procedure 3, (−)-sparteine (3.3 mL,
1
1
8
8
1
4.4 mmol, 1.2 equiv), N-Boc-pyrrolidine (2.6 mL, 14.4 mmol,
.2 equiv), TBME (36 mL), s-BuLi (1.74 M in cyclohexanes,
.3 mL, 12 mmol, 1.2 equiv), ZnCl (1 M in Et O, 8.7 mL,
solution was stirred at −76 for 3 h. A solution of ZnCl (1 M in
2
Et O, 11.3 mL, 11.3 mmol, 0.72 equiv) was then added to the
2
2
2
.7 mmol, 0.72 equiv), 2-bromotoluene (1.5 mL, 12 mmol,
.0 equiv), t-Bu P-HBF (210 mg, 0.72 mmol, 0.06 equiv), and
reaction dropwise with rapid stirring, keeping the temperature
below −69 °C. The cloudy solution was stirred at −76 °C for
3
4
AA
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
.5 h and then warmed to 22 °C resulting in a heterogeneous
Article
0
UV, KMnO ]. SFC (R)-16k, t 6.2 min (6%); (S)-16k, t 7.1 min
(94%) (Chiralpak OJ, 200 bar, 1 mg/mL, 1−10% MeOH in CO ,
4
R
R
mixture. After the solution was stirred at 21 °C for 0.5 h, 2-
bromo-5-methylthiophene (1.83 mL, 15.7 mmol, 1.0 equiv), t-
Bu P-HBF , and Pd(OAc) were added with rapid stirring
2
2 mL/min, 220 nm, 40 °C).
General Procedure 4: α-Arylation of N-Boc-(R)-2-arylpyrro-
3
4
2
3
2,33
(
exothermed from 21 to 29 °C). After the mixture was stirred at
lidine (Table 2).
21 °C for 18 h under argon, the reaction was quenched with an
aqueous solution of NH OH (30%, 1.6 mL), and the mixture was
4
stirred at room temperature for 1 h. The resulting slurry was
filtered through Celite (2 cm deep, 60 mL, medium-porosity
fritted funnel) and eluted with 80 mL of TBME. The filtrate was
washed with 60 mL of 1 M HCl and then with 60 mL of
deionized water. The organic phase was dried over magnesium
sulfate, filtered, and concentrated to give an orange oil. Puri-
fication by column chromatography (SiO , 6 × 30 cm, gradient
2
An oven-dried, 50 mL Schlenk reaction flask equipped with a stir
bar, a septum and an internal temperature probe was added
N-Boc-(R)-2-arylpyrrolidine 16 (4.4 mmol, 1.2 equiv). After
three cycles of evacuation/argon fill, (−)-sparteine (1.0 mL,
elution, hexanes/EtOAc, 19/1 then 9/1) and Kugelrohr
distillation afforded 1.99 g (∼40%) of 16j as a semisolid (a
rotameric mixture, 61:39, 21 °C) with an unknown side product.
N-Boc-pyrrolidine was removed at 80−90 °C [0.02 mmHg,
ABT]. The contaminated product was used for the second
4
.4 mmol, 1.2 equiv), TBME (11 mL) and toluene (2.2 mL) were
added. The light yellow solution was cooled to −75 °C, and
s-BuLi (1.74 M in cyclohexanes, 2.5 mL, 4.4 mmol, 1.2 equiv)
was added dropwise, keeping the temperature below −69 °C.
α-arylation without further purification. Data for 16j: bp 130 °C
1
[
(
0
0.025 mmHg, ABT]; H NMR (500 MHz, CDCl ) 6.68−6.54
3
m, 2 H, HC(7)), 5.14 (app br s, 0.39 H, HC(1)), 5.01 (app br s,
.61 H, HC(1)), 3.60−3.28 (m, 2 H, H C(2)), 2.45 (s, 3 H,
H C(11)), 2.42 (s, 3 H, H C(7)), 2.23 (app br s, 1 H, HC(2) or
(
Note: s-BuLi should be added directly to the solution to avoid
crystallization on the wall of the flask.) The intensively colored
solution was stirred at −75 ± 1 °C for 3 h. A solution of ZnCl
(1 M in Et O, 2.65 mL, 2.65 mmol, 0.72 equiv) was then added to
2
3
3
HC(3)), 2.10−1.97 (m, 2 H, HC(2) or HC(3)), 1.94−1.88 (m,
2
1
H, HC(2) or HC(3)), 1.48 (s, ∼2.5 H, H C(7)), 1.37 (s, ∼6.5
2
3
H, H C(7)); IR (CDCl film) 2975 (s), 2878 (m), 1694 (s),
the reaction dropwise with rapid stirring, keeping the temper-
ature below −69 °C. The yellow solution was stirred at −75 ±
1 °C for 0.5 h and then warmed to 22 °C resulting in a
heterogeneous mixture. After the mixture was stirred at 22 °C for
0.5 h, the septum was secured by copper wire and the flask was
brought into a glovebox. Aryl bromide (3.7 mmol, 1.0 equiv),
t-Bu P-HBF (64 mg, 0.22 mmol, 0.06 equiv), and Pd(OAc)
3
3
1
1
478 (m), 1453 (m), 1392 (s), 1365 (s), 1271 (m), 1255 (m),
225 (m), 1167 (s), 1108 (s), 1039 (w), 964 (w), 916 (w), 886
+
(m), 796 (m), 770 (m); MS (EI, 70 eV) 267.1 (M , 34), 211.0
(100), 196.0 (43), 166.0 (94), 152.0 (36), 139.0 (38), 124.0
(25), 111.0 (29), 97.0 (17); HRMS (EI, 70 eV) calcd for
C H NO S 267.12930, found 267.12882; TLC R 0.32
1
4
21
2
f
3
4
2
(
hexanes/EtOAc, 9/1) [silica gel, UV, KMnO ]; SFC (S)-16j,
(41 mg, 0.18 mmol, 0.048 equiv) were added to the flask with
rapid stirring. In a fume hood, the flask was submerged to a 60 °C
oil bath and stirred under argon. Within 2 h, the mixture turned
gray. After the mixture was stirred at 60 °C for a total of 14 h, the
4
t 5.4 min (6%); (R)-16j, t 7.7 min (94%) ((R,R)-Whelk-O1,
R
2
R
00 bar, 3% MeOH in CO , 2 mL/min, 220 nm, 40 °C).
2
N-Boc-(R)-2-(1-naphthyl)pyrrolidine (16k).
reaction was quenched with an aqueous solution of NH OH
4
(
30%, 0.37 mL), and stirred at room temperature for 0.5 h. The
resulting dark gray mixture was filtered through Celite (2 cm
deep, 30 mL, medium-porosity fritted funnel) and eluted with 50
mL of TBME. The filtrate was washed with 40 mL of 1 M HCl
and then with 40 mL of deionized water. The organic layer was
dried over magnesium sulfate, filtered, and concentrated to give
an orange oil. Purification by column chromatography provided
N-Boc-(2R,5R)-2,5-diarylpyrrolidine 22. (−)-Sparteine can be
recovered as described in experimental procedure I.
Following general procedure 3, (−)-sparteine (1.4 mL,
6
1
3
0
mmol, 1.2 equiv), N-Boc-pyrrolidine 1.02 g, 6 mmol,
.2 equiv), TBME (14 mL), s-BuLi (1.66 M in cyclohexanes,
.65 mL, 6 mmol, 1.2 equiv), ZnCl (1 M in Et O, 3 mL, 3 mmol,
2
2
Preparation of N-Boc-(2R,5R)-2,5-diarylpyrrolidine.
N-Boc-(2R,5R)-2,5-bis(4-methoxyphenyl)pyrrolidine (22a).
.6 equiv), 1-bromonaphthalene (0.64 mL, 5 mmol, 1.0 equiv),
t-Bu P-HBF (88 mg, 0.3 mmol, 0.06 equiv), and Pd(OAc)
3
4
2
(
55 mg, 0.24 mmol, 0.048 equiv) were combined in a 50 mL
Schlenk reaction flask. After the mixture was stirred at 20 °C for
0.5 h, the reaction was quenched, filtered, subjected to aqueous
2
washes, and then concentrated. Purification by column chroma-
tography (SiO , 3 × 26 cm, CH Cl /hexanes, 9/1) afforded
2
2
2
0
6
.71 g (48%) of 16k as a pale yellow solid (a rotameric mixture,
2:38). The spectroscopic data matched those from the
2
4
1
literature. Data for 16k: H NMR (400 MHz, CDCl ) 8.00
3
(
d, J = 8.1 Hz, 1 H), 7.89−7.83 (m, 1 H), 7.75−7.72 (m, 1 H),
.52−7.43 (m, 2 H), 7.41 (t, J = 7.7 Hz, 1 H), 7.26−7.24 (m, 1
H), 5.77−5.75 (m, 0.38 H), 5.61−5.60 (m, 0.62 H), 3.79−3.58
m, 2 H), 2.49−2.41 (m, 1 H), 1.92−1.81 (m, 3 H), 1.48 (s, 3 H),
.10 (s, 6 H); TLC R 0.33 (CH Cl /hexanes, 9/1) [silica gel,
7
(
1
Following general procedure 4, N-Boc-(R)-2-(4-
methoxyphenyl)pyrrolidine 16a (1.22 g, 4.4 mmol, 1.2 equiv),
f
2
2
AB
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
−)-sparteine (1.0 mL, 4.4 mmol, 1.2 equiv), TBME (11 mL)
Article
(
hindered rotation of Boc group. The spectroscopic data matched
those from the literature. Data for 22f: H NMR (500 MHz,
24
1
and toluene (2.2 mL), s-BuLi (1.74 M in cyclohexanes, 2.5 mL,
4
0
.4 mmol, 1.2 equiv), ZnCl (1 M in Et O, 2.65 mL, 2.65 mmol,
.72 equiv), 4-bromoanisole (0.46 mL, 3.7 mmol, 1.0 equiv),
CDCl , −20 °C) 7.90−7.82 (m, 6 H, HC(Aryl)), 7.71 (s, 1 H,
2
2
3
HC(6)), 7.69 (s, 1 H, HC(6)), 7.55−7.40 (m, 6 H, HC(Aryl)),
5.58 (d, J = 7.6 Hz, 1 H, HC(1)), 5.44 (d, J = 7.8 Hz, 1 H,
t-Bu P-HBF (64 mg, 0.22 mmol, 0.06 equiv), and Pd(OAc)
3
4
2
(
41 mg, 0.18 mmol, 0.048 equiv) were combined in a 50 mL
Schlenk reaction flask. After the mixture was stirred at 60 °C for
4 h under argon, the reaction was quenched, filtered, subjected
HC(1)), 2.61−2.50 (m, 2 H, H C(2)), 1.87−1.77 (m, 2 H,
2
13
H C(2)), 1.13 (s, 9 H, H C(5)); C NMR (126 MHz, CDCl ,
2
3
3
1
−20 °C) 154.2 (C(3)), 142.2 (C(7)), 141.0 (C(7)), 133.2
(C(10) or C(15)), 133.0 ((C(10) or C(15)), 132.5 (C(10) or
C(15)), 132.2 ((C(10) or C(15)), 128.4 (C(Aryl)), 128.0
(C(Aryl)), 127.8 (C(Aryl)), 127.6 (C(Aryl)), 127.6 (C(Aryl)),
127.5 (C(Aryl)), 126.1 (C(Aryl)), 125.9 (C(Aryl)), 125.4
(C(Aryl)), 125.4 (C(Aryl)), 124.2 (C(Aryl)), 124.0 (C(Aryl)),
123.2 (C(Aryl)), 79.6 (C(4)), 62.3 (C(1)), 61.7 (C(1)), 31.9
to aqueous washes, and then concentrated to give an orange oil.
Purification by column chromatography (SiO , 4.5 × 28 cm,
hexanes/EtOAc, gradient elution, 7/1 then 4/1) afforded 364 mg
2
(
26%) of 22a as a mixture of white solid and colorless oil.
Recrystallization from hexanes afforded 259 mg (18%) of 22a as
colorless, star-shaped crystals. Two sets of NMR signals for the
2
,5-diarylpyrrolidine moiety were observed due to hindered
(C(2)), 31.2 (C(2)), 28.0 (C(5)); IR (CDCl film) 3053 (m),
3
1
rotation of Boc group. Data for 22a: H NMR (500 MHz,
CDCl , −20 °C) 7.17 (d, J = 8.6 Hz, 2 H, HC(7)), 7.13 (d, J =
8
2975 (m), 2247 (w), 1694 (s), 1633 (m), 1601 (m), 1508 (m),
1477(m), 1454 (m), 1383 (s), 1320 (m), 1269 (m), 1255 (m),
1171 (m), 1128 (m), 1111 (m), 1049 (w), 1018 (w), 981 (w),
961 (w), 909 (m), 855 (m), 817 (m), 780 (m), 732 (s), 646 (m).
3
.6 Hz, 2 H, HC(7)), 6.88 (d, J = 8.7 Hz, 2 H, HC(8)), 6.86 (d,
J = 8.6 Hz, 2 H, HC(8)), 5.25 (d, J = 7.2 Hz, 1 H, HC(1)), 5.09
d, J = 7.1 Hz, 1 H, HC(1)), 3.82 (s, 3 H, HC(10)), 3.79 (s, 3 H,
HC(10)), 2.49−2.30 (m, 2 H, HC(2)), 1.72−1.63 (m, 2 H,
+
(
MS (EI, 70 eV) 423.2 (M , 4), 367.1 (6), 354.1 (9), 322.1 (6),
295.1 (11), 279.1 (7), 239.1 (13), 213.1 (16), 194.1 (33), 167.0
(24), 155.0 (60), 149.0 (36), 127.0 (47), 115.0 (10), 97.1 (12),
1
3
HC(2)), 1.14 (s, 9 H, H C(5)); C NMR (126 MHz, CDCl ,
3
3
−
20 °C) 158.1 (C(9)), 157.9 (C(9)), 154.0 (C(3)), 137.2
83.1 (14), 71.1 (19); HRMS (ESI) calcd for C29
H] 424.2277, found 424.2285; TLC R 0.26 (CH
+146.6 (c = 0.2, acetone).
N-Boc-(2R,5R)-2,5-bis(3,5-dimethylphenyl)pyrrolidine
(22g).
H
30NO
2
[M +
Cl /hexanes,
2 2
(
(
C(6)), 135.9 (C(6)), 126.2 (C(7)), 126.1 (C(7)), 113.7
C(8)), 113.2 (C(8)), 79.3 (C(4)), 61.5 (C(1)), 60.8 (C(1)),
f
24
9/1) [silica gel, UV, KMnO ]; [α]
4 D
5
5.2 (C(10)), 32.1 (C(2)), 31.5 (C(2)), 28.0 (C(5)); IR
(
(
(
(
CDCl film) 2974 (m), 2836 (m), 1693 (s), 1612 (m), 1586
m), 1513 (s), 1462 (m), 1294 (m), 1247 (s), 1208 (m), 1174
s), 1119 (s), 1036 (s), 977 (w), 901 (m), 829 (s), 810 (m), 779
3
+
m), 731 (m), 647 (w), 633 (w). MS (EI, 70 eV) 383.2 (M , 7),
3
1
26.1 (100), 282.1 (27), 255.1 (54), 220.1 (23), 193.1 (84),
76.1 (24), 148.1 (25), 134.1 (73), 121.1 (18), 57.1 (54); HRMS
(
ESI) calcd for C H NO [M + H] 384.2175, found 384.2175;
23 30 4
TLC R 0.38 (hexanes/EtOAc, 4/1) [silica gel, UV, KMnO ];
f
4
[
α]D24 +155.6 (c = 0.2, acetone).
N-Boc-(2R,5R)-2,5-bis(2-naphthyl)pyrrolidine (22f).
Following general procedure 4, N-Boc-(R)-2-(3,5-
dimethyphenyl)pyrrolidine 16g (1.28 g, 4.7 mmol, 1.2 equiv),
(
−)-sparteine (1.1 mL, 4.7 mmol, 1.2 equiv), TBME (12 mL)
and toluene (2.4 mL), s-BuLi (1.74 M in cyclohexanes, 2.7 mL,
4
0
1
.7 mmol, 1.2 equiv), ZnCl (1 M in Et O, 2.8 mL, 2.8 mmol,
2 2
.72 equiv), 1-bromo-3,5-dimethylbenzene (0.54 mL, 3.9 mmol,
.0 equiv), t-Bu P-HBF (69 mg, 0.23 mmol, 0.06 equiv), and
3
4
Pd(OAc) (45 mg, 0.19 mmol, 0.048 equiv) were combined in a
2
5
0 mL Schlenk reaction flask. After the mixture was stirred at
Following general procedure 4, N-Boc-(R)-2-(2-naphthyl)-
pyrrolidine 16f (1.43 g, 4.8 mmol, 1.2 equiv), (−)-sparteine
60 °C for 14 h under argon, the reaction was quenched, filtered,
subjected to aqueous washes, and then concentrated to give an
orange oil. Purification by column chromatography (SiO , 4.5 ×
23 cm, CH Cl /hexanes, gradient elution, 1/1 then 2/1)
(
(
1.1 mL, 4.8 mmol, 1.2 equiv), TBME (12 mL) and toluene
2
2.4 mL), s-BuLi (1.74 M in cyclohexanes, 2.75 mL, 4.8 mmol,
2
2
1
2
.2 equiv), ZnCl (1 M in Et O, 2.9 mL, 2.9 mmol, 0.72 equiv),
-bromonaphthalene (828 mg, 4.0 mmol, 1.0 equiv),
afforded 390 mg (27%) of 22g as a white solid. Recrystallization
from hexanes afforded 311 mg (21%) of 22g as a light pink cube.
Two sets of NMR signals for the 2,5-diarylpyrrolidine moiety
2
2
t-Bu P-HBF (71 mg, 0.23 mmol, 0.06 equiv), and Pd(OAc)
3
4
2
(
44 mg, 0.19 mmol, 0.048 equiv) were combined in a 50 mL
Schlenk reaction flask. After the mixture was stirred at 60 °C for
4 h under argon, the reaction was quenched, filtered, subjected
were observed due to hindered rotation of Boc group. Data for
1
22g: H NMR (500 MHz, CDCl ) 6.86 (br s, 2 H, HC(6)), 6.83
3
1
(br s, 2 H, HC(4)), 6.81 (br s, 2 H, HC(4)), 5.22 (d, J = 7.1, 1 H,
HC(1)), 5.05 (d, J = 7.1, 1 H, HC(1)), 2.45−2.41 (m, 2 H,
H C(2)), 2.31 (s, 12 H, H C(10)), 1.72−1.67 (m, 2 H, HC(2)),
to aqueous washes, and then concentrated to give an orange oil.
Purification by column chromatography (SiO , 5.5 × 22 cm,
CH Cl /hexanes, gradient elution, 7/3, 8/2 then 9/1) afforded
2
2
3
13
1.15 (s, 9 H, H C(5)); C NMR (126 MHz, CDCl ) 154.1
2
2
3
3
6
∼
19 mg (37%) of 22f as a white solid which is contaminated by
6% of the corresponding pyrroline. Two sets of NMR signals
for the 2,5-diarylpyrrolidine moiety were observed due to
(C(3)), 145.2 (C(6)), 144.0 (C(6)), 137.8 (C(8)), 137.5
(C(8)), 128.5 (C(9)), 128.0 (C(9)), 123.2 (C(7)), 123.0
(C(7)), 79.2 (C(4)), 62.3 (C(1)), 61.7 (C(1)), 32.2 (C(2)),
AC
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
1.6 (C(2)), 28.1 (C(5)), 21.5 (C(10)), 21.3 (C(10)); IR
Article
3
Following general procedure 4, N-Boc-(R)-2-(2-tolyl)-
pyrrolidine 16i (1.68 g, 6.4 mmol, 1.2 equiv), (−)-sparteine
(1.5 mL, 6.4 mmol, 1.2 equiv), TBME (16 mL) and toluene
(3.2 mL), s-BuLi (1.74 M in cyclohexanes, 3.7 mL, 6.4 mmol,
1.2 equiv), ZnCl (1 M in Et O, 3.9 mL, 3.9 mmol, 0.72 equiv),
(
(
(
CDCl film) 2974 (m), 2918 (m), 1699 (s), 1604 (m), 1476
m), 1455 (m), 1383 (s), 1268 (w), 1255 (w), 1172 (m), 1117
m), 1055 (w), 957 (w), 845 (m), 775 (w), 703 (m); MS (ESI)
3
380.3 (M + H, 28), 324.2 (100), 218.1 (6); HRMS (ESI) calcd
2
2
for C H NO [M + H] 380.2590, found 380.2594; TLC R 0.33
2-bromotoluene (0.64 mL, 5.4 mmol, 1.0 equiv), t-Bu P-HBF
25
34
2
f
3
4
2
4
(
hexanes/CH Cl , 9/1) [silica gel, UV]; [α] 142.0 (c = 0.2,
(95 mg, 0.32 mmol, 0.06 equiv) and Pd(OAc) (59 mg,
2
2
D
2
acetone).
0.26 mmol, 0.048 equiv) were combined in a 100 mL Schlenk
reaction flask. After the mixture was stirred 60 °C for 14 h under
argon, the reaction was quenched, filtered, subjected to aqueous
washes, and then concentrated to give an orange oil. Purification
N-Boc-(2R,5R)-2,5-bis(5-phenylbiphenyl-3-yl)pyrrolidine
22h).
(
by column chromatography (SiO , 4.5 × 30 cm, hexanes/EtOAc,
2
9
2/8) afforded 413 mg (22%) of 22i as a white solid. Recrys-
tallization from hexanes afforded 320 mg (17%) of 22i as a light
pink cube. Two sets of NMR signals for the 2,5-diarylpyrrolidine
moiety were observed due to hindered rotation of Boc group.
1
Following general procedure 4, N-Boc-(R)-2-(5-phenylbiphenyl-
Data for 22i: H NMR (500 MHz, CDCl ) 7.23−7.14 (m, 8 H,
3
3
-yl)pyrrolidine 16h (2.33 g, 5.8 mmol, 1.2 equiv), (−)-sparteine
HC(Aryl)), 5.50 (d, J = 8.0 Hz, 1 H, HC(1)), 5.37 (d, J = 8.2 Hz,
(
(
1.35 mL, 5.8 mmol, 1.2 equiv), TBME (15 mL) and toluene
3.0 mL), s-BuLi (1.74 M in cyclohexanes, 3.35 mL, 5.8 mmol,
1
H, HC(1)), 2.52−2.36 (m, 2 H, H C(2)), 2.42 (s, 3 H,
2
H C(12)), 2.39 (s, 3 H, H C(12)), 1.64 (dd, J = 12.1, 6.3 Hz,
2 H, H C(2)), 1.13 (s, 6 H, H C(5)), C NMR (126 MHz,
3
3
13
1
1
.2 equiv), ZnCl (1 M in Et O, 3.5 mL, 3.5 mmol, 0.72 equiv),
-bromo-3,5-diphenylbenzene (1.50 g, 4.9 mmol, 1.0 equiv),
2 2
2
3
CDCl ) 153.7 (C(3)), 143.0 (C(Aryl)), 141.5 (C(Aryl)), 134.2
3
t-Bu P-HBF (85 mg, 0.29 mmol, 0.06 equiv), and Pd(OAc)
3
4
2
(C(Aryl)), 133.7 (C(Aryl)), 130.8 (C(Aryl)), 130.1 (C(Aryl)),
(54 mg, 0.23 mmol, 0.048 equiv) were combined in a 50 mL
1
(
5
(
26.6 (C(Aryl)), 126.4 (C(Aryl)), 125.8 (C(Aryl)), 125.7
C(Aryl)), 124.3 (C(Aryl)), 123.7 (C(Aryl)), 79.2 (C(4)),
9.1 (C(1)), 58.9 (C(1)), 30.0 (C(2)), 29.5 (C(2)), 28.0
C(5)), 19.3 (C(12)), 19.3 (C(12)); IR (CDCl film) 3066 (w),
Schlenk reaction flask. After the mixture was stirred at 60 °C
for 14 h under argon, the reaction was quenched, filtered,
subjected to aqueous washes, and then concentrated to give
an orange oil. Purification by column chromatography (SiO2,
3
3017 (w), 2973 (m), 2871 (w), 1698 (m), 1604 (w), 1485 (w),
1461 (w), 1383 (m), 1285 (w), 1250 (w), 1180 (m), 1159 (m),
1128 (m), 1101 (w), 976 (w), 792 (w), 752 (w). MS (EI, 70 eV)
4.5 × 25 cm, hexanes/EtOAc, 14/1) afforded 633 mg of 22h and
contaminants. Further purification by column chromatography
+
(
SiO , 3.5 × 25 cm, hexanes/CH Cl , 1/1) afforded 540 mg
351.2 (M , 5), 295.2 (100), 223.2 (71), 164.1 (57), 119.1 (69);
2
2
2
(
18%) of 22h as a fine white powder. Two sets of NMR signals
HRMS (ESI) calcd for C H NO [M + H] 352.2277, found
23
30
2
for the 2,5-diarylpyrrolidine moiety were observed due to
hindered rotation of Boc group. The spectroscopic data matched
those from the literature. Data for 22h: H NMR (500 MHz,
352.2278; TLC R 0.33 (hexanes/EtOAc, 92/8) [silica gel, UV,
f
24
KMnO ]; [α]D +126.4 (c = 0.2, acetone).
4
24
1
N-Boc-(2R,5R)-2,5-bis(5-methyl-2-thienyl)pyrrolidine (22j).
CDCl ) 7.72−7.70 (m, 1 H, HC(Aryl)), 7.68−7.64 (m, 9 H,
3
HC(Aryl)), 7.51−7.44 (m, 12 H, HC(Aryl)), 7.42−7.36 (m, 4 H,
HC(Aryl)), 5.53 (d, J = 7.7 Hz, 1 H, HC(1)), 5.32 (d, J = 7.9 Hz,
1
1
H, HC(1)), 2.68−2.52 (m, 2 H, H C(2)), 1.87−1.77 (dd, J =
13
2
2.7, 5.9 Hz, 2 H, HC(2)), 1.21 (s, 9 H, H C(5)); C NMR (126
3
MHz, CDCl ) 154.1 (C(3)), 146.4 (C(6)), 144.8 (C(6)), 142.1
3
Following general procedure 4, N-Boc-(R)-2-(5-methyl-2-
thienyl)pyrrolidine 16j (1.94 g, 7.3 mmol, 1.2 equiv),
(
1
(
C(Aryl)), 141.9 (C(Aryl)), 141.5 (C(Aryl)), 141.1 (C(Aryl)),
28.8 (C(Aryl)), 128.7 (C(Aryl)), 127.5 (C(Aryl)), 127.4
C(Aryl)), 127.3 (C(Aryl)), 127.2 (C(Aryl)), 125.0 (C(Aryl)),
(
−)-sparteine (1.7 mL, 7.3 mmol, 1.2 equiv), TBME (18 mL)
and toluene (3.6 mL), s-BuLi (1.465 M in cyclohexanes, 5.0 mL,
1
6
24.6 (C(Aryl)), 123.2 (C(Aryl)), 123.1 (C(Aryl)), 79.7 (C(4)),
2.6 (C(1)), 61.8 (C(1)), 32.3 (C(2)), 31.9 (C(2)), 28.2
7
0
1
.3 mmol, 1.2 equiv), ZnCl (1 M in Et O, 4.35 mL, 4.35 mmol,
2 2
.72 equiv), 2-bromo-5-methylthiophene (0.59 mL, 6.1 mmol,
.0 equiv), t-Bu P-HBF (108 mg, 0.36 mmol, 0.06 equiv), and
(
(
(
(
(
C(5)); IR (CDCl film) 3032 (w), 2974 (m), 1695 (s), 1596
3
m), 1576 (m), 1497 (m), 1477 (w), 1454 (m), 1434 (m), 1385
s), 1365 (m), 1272 (w), 1159 (m), 1120 (m), 1075 (w), 1048
w), 1029 (w), 909 (m), 875 (m), 757 (s), 741 (s), 698 (s). MS
3
4
Pd(OAc) (68 mg, 0.29 mmol, 0.048 equiv) were combined in a
2
1
00 mL Schlenk reaction flask. After the mixture was stirred
+
60 °C for 18 h under argon, the reaction was quenched, filtered,
subjected to aqueous washes and then concentrated to give an
EI, 70 eV) 627.4 (M , 6), 570.3 (34), 526.3 (46), 499.3 (38),
446.2 (60), 315.2 (66), 271.2 (100), 256.1 (80), 241.1 (21), 56.1
orange oil. Purification by column chromatography (SiO , 4.5 ×
(
78); HRMS (ESI) calcd for C H NO [M + H] 628.3216,
2
4
5
42
2
3
0 cm, CH Cl /hexanes, 3/2) afforded 440 mg (20%) of 22j as a
found 628.3221; TLC R 0.12 (hexanes/CH Cl , 1/1) [silica gel,
2 2
f
2
2
UV]; [α]D24 +49.6 (c = 0.2, acetone).
tanned solid. Recrystallization from hexanes afforded 291 mg
13%) of 22j as a white fluffy needle. Two sets of NMR signals for
(
N-Boc-(2R,5R)-2,5-bis(2-tolyl)pyrrolidine (22i).
the 2,5-diarylpyrrolidine moiety were observed due to hindered
1
rotation of Boc group. Data for 22j: H NMR (500 MHz,
CDCl ) 6.68 (d, J = 3.2 Hz, 1 H, HC(7)), 6.61 (d, J = 3.2 Hz, 1 H,
3
HC(7)), 6.57−6.54 (m, 2 H, HC(8)), 5.32 (d, J = 6.9 Hz, 1 H,
HC(1)), 5.15 (d, J = 6.9 Hz, 1 H, HC(1)), 2.59−2.49 (m, 2 H,
H C(2)), 2.44 (s, 3 H, H C(7)), 2.42 (s, 3 H, H C(7)), 1.94−
2
3
3
1
3
1
.84 (m, 2 H, H C(2)), 1.26 (s, 9 H, H C(5)); C NMR
2
3
AD
dx.doi.org/10.1021/jo502388r | J. Org. Chem. XXXX, XXX, XXX−XXX