of 21
Title:
Ni
-
Catalyzed Enantioselective Desymmetrization: Development of Divergent Acyl and
1
Decarbonylative Cross
-
Coupling Reactions
2
Authors:
Á
ngel
D.
Hernández
-
Mejías
,
1
Alexander M. Shimozono,
1
Avijit Hazra,
2
Sven Richter,
1
3
Zhengjia
Tong,
1
Neil F. Langille,
3
Kyle Quasdorf,
4
Andrew
T.
Parsons,
3
Matthew S. Sigman,
*
2
Sarah E.
4
Reisman*
1
5
6
Affiliation:
7
1.
The Warren and Katharine Schlinger Laboratory for Chemistry and Chemical Engineering
8
Division of Chemistry and Chemical
Engineering, California Institute of Technology Pasadena,
9
California 91125, USA
10
2.
Department of Chemistry, University of Utah, Salt Lake City, Utah 84112, USA
11
3.
Drug Substance Technologies
Synthetics, Amgen, Inc. 360 Binney St, Cambridge, MA 02142.
12
4.
Drug Substance Technologies
Synthetics, Amgen, Inc. One Amgen Center Drive, Thousand
13
Oaks, CA 91320
14
15
Abstract:
16
Ni
-
catalyzed
asymmetric
reductive cross
-
coupling
reactions
provide rapid and modular access to
17
enantioenriched
building blocks
from simple electrophile precursors.
Reductive coupling reactions
18
that can diverge
through
a common organometallic intermediate to two distinct families of
19
enantioenriched products are particularly versatile but underdeveloped.
Here, we describe
the
20
development of a bis(oxazoline) ligand
that
enables
the desym
metrization of
meso
-
anhydrides
.
21
W
hen secondary benzylic electrophiles are employed, doubly stereoselective acyl cross
-
coupling
22
proceeds to give ketone products with
catalyst
control over three new
ly
formed stereogenic centers.
23
Alternatively, use of primary alkyl halides in the presence of an additional
halogen
atom transfer
24
catalyst
results in decarbonylative alkylation to give enantioenriched
b
-
alkyl acids
.
Analysis of
25
reaction rates for a range of both catalysts and substrates supports the notion that tuning the
26
different electrophile activation steps with the two catalysts is required for enhanced reaction
27
https://doi.org/10.26434/chemrxiv-2024-3r6z4
ORCID:
https://orcid.org/0000-0001-8244-9300
Content not peer-reviewed by ChemRxiv.
License:
CC BY-NC-ND 4.0
performance.
These studies illustrate how reaction design can diverge a common Ni
-
acyl
28
intermediate to either acyl or decarbonylative coupling
products and
highlight how dual ligand
29
systems can be used to engage unactivated alkyl halides in Ni
-
catalyzed asymmetric reductive
30
coupling.
31
32
Main:
33
Ni
-
catalyzed reductive
cross
-
coupling
(RCC)
reactions have emerged as powerful methods
to
34
construct C
C bonds
using
C(sp
3
)
electrophiles
(Figure 1a)
.
1
,
2
These reactions can cross
-
couple two
35
organic electrophiles by using a stoichiometric reductant to supply electrons
for catalyst turnover
. To
36
achieve high cross
-
selectivity, the Ni catalyst must oxidatively add each electrophile in sequence,
3
or the
37
reaction conditions must provide mechanistically distinct modes of activating each coupling partner
.
4
The
38
mechanisms and
substrate
scope of these reactions are dictated by the supporting ligands and the reduct
ant
39
system
s
, with 2,2
-
bipyridine
(bpy) ligands
emerging as a broadly employed and versatile ligand
40
class
.
1
,
4
,
5
,
6
,
7
,
8
Although less well
-
developed,
the
use of
chiral
supporting ligands such as bis(oxazolines)
41
(BOX) or bi(oxazolines) (BiOX)
enables enantioselective RCC with control over newly
forged
C(sp
3
)
42
center
s
(Figure 1a)
.
9
,
10
,
11
,
12
However,
the
C(sp
3
) electrophile
scope of these asymmetric RCC reactions
is
43
often
restricted to
activated alkyl coupling partners
11
,
12
that
can generate
stabilized radical intermediates
44
(Figure 1a)
.
The development of enantioselective RCCs
employing
new electrophile pairs
would expand
45
the
field
with potential applications in drug development and natural product synthesis.
46
An
appealing
class of
electrophiles for
Ni
-
catalyzed
RCCs
are
anhydrides,
c
arboxylic acid
47
derivatives
that are typically accessed
from feedstock chemicals
.
We
targeted
cyclic
-
fused
meso
-
48
anhydrides
as
versatile
substrates for Ni
-
catalyzed
enantioselective desymmetrization reactions
for the
49
following reasons
. First,
although limited to redox
-
neutral cross
-
coupling
with organometallic reagents
,
50
https://doi.org/10.26434/chemrxiv-2024-3r6z4
ORCID:
https://orcid.org/0000-0001-8244-9300
Content not peer-reviewed by ChemRxiv.
License:
CC BY-NC-ND 4.0
prior work from Rovis
and Doyle
suggest
ed
that chiral Ni catalysts can desymmetrize
meso
-
anhydrides
51
through
Ni
-
catalyzed
C
C bond formation
(Figure 1b)
.
13
,
14
,
15
Second, we hypothesized that
, depending on
52
the reaction conditions,
it could be possible to diverge the intermediate acyl
-
Ni
II
oxidative addition
53
complex
(Figure 1c,
I
)
through two different reaction pathways
:
acyl
coupling to give ketoacid
54
products,
9
,
16
or
decarbonylative
coupling
17
,
18
,
19
to give alkyl acid products.
Given that both reactions
55
would proceed through the same intermediate
after desymmetrization
,
we
reasoned
that
a single chiral
56
catalyst
could provide
access
to
two families of
enantioenriched
products.
Third, if the
meso
-
anhydride
57
was cross
-
coupled with a secondary alkyl electrophile, a doubly stereoselective cross
-
coupling
could set
58
multiple stereogenic centers during a single transformation.
20
,
21
,
22
And fourth,
the
resulting carboxylic acid
59
product
s of either the acyl or decarbonylative coupling
c
ould
be submitted to
further diversification
60
through decarboxylative
functionalizati
o
n
.
The synthetic versatility
of sequential cross
-
coupling of
meso
-
61
anhydride
s
was previously demonstrated by Baran and coworkers
(Figure 1
b
)
;
23
h
owever,
in this case
,
a
62
cinchona alkaloid
-
mediated desymmetriz
ing esterification
was performed in step
separate
before
the
63
cross
-
coupling reactions
.
In the proposed Ni
-
catalyzed asymmetric RCC, the desymmetrization step
64
would proceed in concert with one of the C
C bond forming steps.
65
https://doi.org/10.26434/chemrxiv-2024-3r6z4
ORCID:
https://orcid.org/0000-0001-8244-9300
Content not peer-reviewed by ChemRxiv.
License:
CC BY-NC-ND 4.0
66
Figure 1. Ni
-
catalyzed reductive cross
-
coupling reactions and context for current study.
67
Here
, we
report the discovery of a new chiral ligand that
supports
a
highly
selective
68
desymmetrization of
meso
-
anhydrides
(Figure 1
c
)
.
With secondary
benzylic
chloride
electrophiles,
meso
-
69
anhydrides
undergo
catalyst
-
controlled
doubly stereoselective
acyl
cross
-
coupling
to
give ketoacid
70
products
containing
up to
three stereogenic centers.
We then
demonstrate
that
a dual
ligand
system
enables
71
the coupling of
meso
-
anhydrides with unactivated C(sp
3
)
halides
, a traditionally challenging coupling
72
partner for enantioselective RCCs,
to give
b
-
alkyl acids
in a
Ni
-
catalyzed
decarbonylative
C(sp
3
)
C(sp
3
)
73
coupling
process.
These studies illustrate how reaction design can be used to selectively diverge between
74
two productive asymmetric cross
-
coupling pathways and lays the foundation for expanding the scope of
75
Ni
-
catalyzed asymmetric RCCs
by using
dual ligand systems.
76
Our studies began by investigating the Ni
-
catalyzed
RCC of commercially available anhydrides
77
1a
and
1b
with various alkyl electrophiles. Two promising experimental findings, shown in Figure 2,
78
a.
Ni-catalyzed reductive C(sp
2
)–C(sp
3
) coupling
+
(±)
Y
C(sp
3
)
electrophile
C(sp
2
)
electrophile
N
N
Ni
I
X
N
N
Ni
0
multiple redox states accessible:
can engage broad range of C(sp
3
)
electrophiles
N
N
Ni
I
Ar
Non-Asymmetric:
bipyridines (bpy)
b.
Prior work
achiral
Ni
O
O
O
CO
2
H
O
R
Ni
redox neutral
singly stereoselective redox neutral cross-coupling
13-15
O
O
O
CO
2
H
OMe
O
desymmetrize
cross-
couple
multistep desymmetrization/decarboxylative cross-coupling
23
N
O
O
N
Ni
I
X
limited redox states accessible:
restricted to activated C(sp)
3
electrophiles
Asymmetric:
chiral
Ni
X
achiral
Ni
reductant
+
(±)
Y
X
reductant
chiral
Ni
(±)
racemic product
broad
C(sp
3
) electrophile
scope
enantioenriched product
limited
C(sp
3
) electrophile
scope
O
O
O
chiral
Ni
O
[Ni]
O
O
CO
2
H
O
high selectivity:
desymmetrization
X
–CO
O
[Ni]
O
CO
2
H
acyl coupling
Ni-acyl (
I
)
Ni-alkyl (
II
)
decarbonylative
coupling
X
doubly
stereoselective
cross-coupling
ketoacid
alkyl acid
c. This work:
Ni-catalyzed asymmetric RCC of
meso-
anhydrides
racemic products
enantioenriched products
N
O
Ni
I
N
O
X
bioxazoline
(
BiOX)
bis(oxazoline)
(BOX)
https://doi.org/10.26434/chemrxiv-2024-3r6z4
ORCID:
https://orcid.org/0000-0001-8244-9300
Content not peer-reviewed by ChemRxiv.
License:
CC BY-NC-ND 4.0
guided further reaction development. First, when phenyl
-
substituted BOX ligand
L
1
was used, anhydride
79
1a
and benzylic chloride
2a
could be cross
-
coupled to give
3aa
in 30% yield, 6:1 dr, and 92% ee
.
9
We
80
note that addition of BF
3
·OEt
2
(
50 mol %
) was critical
for productive reactivity, as
omission of this
81
additive failed to
provide
product
3aa
.
Second, although no decarbonylative coupling was observed when
82
benzylic chloride
2a
was used, when primary alkyl bromide
4a
and cyclobut
a
ne
-
fused anhydride
1b
were
83
subjected to a stoichiometric mixture of Ni(cod)
2
and
L
2
, a mixture of
acyl coupling product
5
ba
and
84
decarbonylative coupling product
6ba
w
as
formed (
43% yield
5ba
, 95% ee
and 17% yield
6
ba
, 29% ee,
85
respectively
). Given these promising results, we undertook reaction optimization campaigns to try to
86
improve the yield and selectivity of these divergent reaction outcomes
.
87
https://doi.org/10.26434/chemrxiv-2024-3r6z4
ORCID:
https://orcid.org/0000-0001-8244-9300
Content not peer-reviewed by ChemRxiv.
License:
CC BY-NC-ND 4.0
88
89
Figure 2.
Initial discovery
of
doubly stereoselective acyl
RCC
and decarbonylative RCC.
Reactions
90
were performed on 0.2 mmol scale. Yields were determined by
1
H NMR analysis of the crude reaction
91
mixture using an internal standard. The dr and ee of
3aa
was determined using liquid
92
chromatography/supercritical fluid chromatography (LC/SFC) with a chiral stationary phase.
93
We
first targeted the optimization of the
doubly
-
stereoselective
RCC to form ketoacid
3aa
(Figure
94
2c
).
In preliminary experiments, t
racking the mass balance of th
e
reaction revealed that
unreacted
95
anhydride remained, even at longer reaction times, but
that benzylic chloride
2a
was
fully
converted to
96
the corresponding homodimer butane
-
2,3
-
diyldibenzene (
7
, 1:1 dr).
Thus, an extensive evaluation of
97
Br
OPh
4a
(2.0 equiv)
O
O
O
H
H
O
N
N
O
Ph
Ph
Ph
Ni
0
H
H
O
Ni
·
L2
O
O
DMA/THF
23 °C
b.
Decarbonylative RCC with unactivated electrophiles: preliminary results using stoichiometric
L2
·
Ni
0
a.
Doubly stereoselective acyl RCC: preliminary results
O
O
O
H
H
+
Ph
Cl
Me
NiCl
2
·dme (15 mol %)
L1
(15 mol %)
BF
3
·OEt
2
(50 mol %)
Mn (3.0 equiv)
DMA (1.0 equiv)
THF, 14 h, 23 °C
O
Ph
Me
CO
2
H
H
H
1a
(1.0 equiv)
2a
(2 equiv)
3aa
30% yield
6:1 dr
,
92% ee
Me
Ph
Me
Ph
(1:1 dr)
O
N
N
O
Ph
Ph
Me
Me
L1
1b
L2
·Ni
0
5ba
43% yield
95% ee
CO
2
H
H
H
O
OPh
acyl coupling
+
acyl coupling
CO
2
H
H
H
6ba
17% yield
29% ee
OPh
decarbonylative
coupling
+
N
O
O
O
H
H
+
Ph
Cl
Me
NiCl
2
·dme (15 mol %)
L7
(15 mol %)
BF
3
·OEt
2
(50 mol %)
Mn (3.0 equiv)
DMA (1.0 equiv)
THF, 14 h, 23 °C
O
Ph
Me
CO
2
H
H
H
1a
(1.0 equiv)
2a
(2 equiv)
3aa
84% yield
>20:1 dr
,
>99% ee
O
N
N
O
Ph
Ph
Me
Me
O
N
N
O
Ph
Ph
Ph
N
F
3
C
O
N
N
O
Ph
Ph
O
N
N
O
Ph
Ph
Ph
N
O
N
N
O
Ph
Ph
Ph
N
O
N
N
O
Ph
Ph
Ph
N
CH
3
CF
3
L1
30% yield
6:1 dr
92% ee
L3
47% yield
10:1 dr
>99% ee
L4
50% yield
>20:1 dr
>99% ee
L5
54% yield
>20:1 dr
>99% ee
L7
84% yield
>20:1 dr
>99% ee
L6
53% yield
>20:1 dr
>99% ee
increasing dr
acyl coupling
increasing yield
c.
Ni-catalyzed doubly stereoselective acyl RCC: reaction optimization
d.
Bis(oxazoline) ligand evolution
L1
,
2.0 equiv
2a
none
1.0 equiv
2a
5 mol % Ni, 10 mol %
L7
10 mol % Ni, 15 mol %
L7
15 mol % Ni, 20 mol %
L7
No BF
3
·OEt
2
No DMA
No Ni/
L7
No Ni/
L7
, No BF
3
·OEt
2
1
2
3
4
5
6
7
8
9
10
entry
change from optimal
yield
3aa
recovery
1a
homodimer
7
30
47
84
6
6
42
44
(99% ee)
58
27
5
74
12
5
40
28
25
11
26
56
15
94
91
50
Me
Ph
Me
Ph
7
(1:1 dr)
O
N
N
O
Ph
Ph
R
1
R
2
5
42
34
47
general
BOX structure
(92% ee)
https://doi.org/10.26434/chemrxiv-2024-3r6z4
ORCID:
https://orcid.org/0000-0001-8244-9300
Content not peer-reviewed by ChemRxiv.
License:
CC BY-NC-ND 4.0
chiral BOX ligands was performed with the objective of improving the enantio
-
and
diastereoselectivity
98
and increasing the yield (Figure
2d
). Similar to the alkylation of acyl chlorides
,
9
,
15
phenyl substitution on
99
the oxazolines was found to be necessary. However, the substituents (R
1
and R
2
) at the methylene linker
100
of the BOX ligand substantially influenced both the selectivity and the yield. Replacing the gem
-
dimethyl
101
(
L
1
)
with gem
-
dibenzyl substitution (
L3
) increased the dr from 6:1 to 10:1. The diastereoselectivity was
102
further enhanced to >20:1 by differentiating the R
1
and R
2
substituents (
e.g.
,
L4
). With this type of ligand,
103
uniformly high dr and ee were observed; however, the yield remained modest due to competing formation
104
of homodimer
7
(see Supplementary Materials for
additional ligand data)
. Ultimately, we discovered that
105
the use of
L7
, in which
R
2
= (6
-
CF
3
-
pyridyl)methyl
,
improved
the conversion of
anhydride
1a
, giving
106
ketoacid
3aa
in 84% yield, >20:1 dr, and 99% ee
under the optimal conditions
(Figure
2c
, entry 2). This
107
remarkably high selectivity results from two independent, catalyst
-
controlled steps
anhydride
108
desymmetrization and C
C bond formation
in which the ee of the major diastereomer is amplified
109
following the Horeau principle
.
24
,
25
,
26
110
We noted that as the yield of
3aa
increases across the ligand series
L4
L7
,
the amount of
111
homodimer
7
decreases (see
SI
), suggesting that the
formation of
3aa
was in direct competition with this
112
non
-
productive radical
-
radical dimerization process and that
L7
partitions the reaction pathway toward
113
acyl coupling.
Despite the improved reactivity with
L7
,
use of
1
.0
equiv of
2a
decreased the
yield of
3aa
114
(Figure
2c
, entry 3).
When the
catalyst loading was reduced or
L
7
was used in excess, the yield of
3aa
115
was lower
(entries 4
6). When BF
3
·OEt
2
was
omitted from the reaction,
3aa
was
formed
in
40% yield
116
with 28% recovery of
1a
and
25
% yield of homodimer
7
. These data
are
consistent with the
hypothesis
117
that BF
3
·OEt
2
serves as a
Lewis acid
to
increase the
reactivity
of
anhydride
1a
. Lower conversion of
1a
118
and lower yield of
3aa
was
also observed in the absence of
dimethylacetamide (
DMA
)
(entry 8); however,
119
cosolvent levels of DMA perform
ed
poorly. Finally, control studies confirm
ed
that Ni
was
required for
120
https://doi.org/10.26434/chemrxiv-2024-3r6z4
ORCID:
https://orcid.org/0000-0001-8244-9300
Content not peer-reviewed by ChemRxiv.
License:
CC BY-NC-ND 4.0
product formation and that low levels of homodimer
7
were
formed when both BF
3
·OEt
2
and Mn
0
were
121
present (entry 9), indicating that multiple species
could
contribute to the
conversion
of
2a
.
122
Having
identified conditions for a highly doubly stereoselective acyl cross
-
coupling
, we turned
123
our attention to optimizing the
decarbonylative
RCC of primary alkyl bromide
4a
.
Although cross
-
124
coupling products
5ba
and
6ba
were formed under the
stoichiometric
conditions shown in Figure 2
b
,
125
under the standard acyl coupling
conditions,
neither product was o
bserved
(see
Supplementary Materials
).
126
At this stage, we considered r
ecent mechanistic studies
of
the
related
asymmetric RCC between alkenyl
127
bromides and benzylic chlorides, which also employed
a
chiral
BOX
·Ni complex as the catalyst
.
27
Based
128
on this
prior study
,
we p
ostulated
a mechanism for the acyl RCC that
involves two
interdependent
cycles
129
of the same catalyst: a cross
-
coupling cycle that mediates the C
C bond
-
forming step, and a halide atom
130
transfer (XAT) cycle, which serves to generate a carbon
-
centered radical from the C(sp
3
) electrophile
131
(Figure
3
a
).
For benzylic chloride
2a
, it is proposed that
L7
·Ni
I
X
accessible by Mn
0
reduction of
132
L
7
·Ni
II
X
2
serves as the XAT catalyst.
In contrast, we hypothesized
that
L7
·Ni
I
X was not an effective
133
XAT
catalyst
for activation
of alkyl bromide
4a
, accounting for why no RCC products were formed under
134
catalytic conditions.
We note that in
the
stoichiometric conditions using Ni(
cod
)
2
(Figure 2
b
)
,
L
2
·Ni
0
135
could potentially activate the alkyl bromid
e
4a
or generate alternative XAT catalysts through
136
comproportionation
. Based on these observations
and recent studies by Weix demonstrating that some
137
RCCs of alkyl bromides and chlorides benefit from the
addition of a
secondary
tridentate
ligand
,
28
we
138
envisioned that
a dual ligand system
could
induce
reactivity with
bromide
4a
.
With
this design
,
the BOX
139
ligand
would
support the
cross
-
coupling cycle
and enable asymmetric desymmetrization
of the anhydride
,
140
and
a
tridentate ligand
would
support the XAT cycle
(Figure
3
a)
. This
dual ligand system
would allow
141
independent ligand tuning of
the C(sp
2
) and C(sp
3
) electrophile activatio
n
steps
.
142
https://doi.org/10.26434/chemrxiv-2024-3r6z4
ORCID:
https://orcid.org/0000-0001-8244-9300
Content not peer-reviewed by ChemRxiv.
License:
CC BY-NC-ND 4.0
143
Figure
3
.
Proposed divergent mechanism and
optimization of the decarbonylative alkylation of
144
anhydride 1b.
Reactions were performed on 0.2 mmol scale. Yields were determined by
1
H NMR
145
analysis of the crude reaction mixture using an internal standard. The ee of
6ba
was determined using
146
liquid chromatography/supercritical fluid chromatography (LC/SFC) with a chiral stationary phase.
147
decarbonylative
cross-coupling
cycle
Ni
II
N
N
*
O
O
O
H
H
Ni
II
N
N
*
O
O
H
H
Ni
III
N
N
*
O
O
H
H
Ni
I
X
N
N
*
CO
2
H
H
H
Ni
III
N
N
*
O
O
O
H
H
Ni
I
X
N
N
*
CO
2
H
O
H
H
acyl
cross-coupling
cycle
X
Ni
II
L
X
X
Ni
I
L
ketoacid
alkylacid
X
XAT cycle
1/2 M
0
1/2 MX
2
1
oxidative addition/
reduction/
decarbonylation
oxidative addition/
reduction
1
N
N
N
Ni
I
X
tridentate ligand:
activation of simple C(sp
3
)
electrophiles (X = Br, I)
X
R
R
X
R
slow radical
generation
fast radical
generation
N
O
O
N
Ni
I
X
BOX ligand:
limited to activated C(sp)
3
electrophiles
Cl
R
Ph
+
Br
OPh
CO
2
H
H
H
OPh
NiCl
2
·DME (22.5 mol %)
L7
(20 mol %),
L9
(2.5 mol %)
AlCl
3
(0.5 equiv)
Mn (3.0 equiv)
N
2
balloon
THF, 23 °C, 14 h
N
N
N
4a
(2.0 equiv)
decarbonylative
coupling
O
O
O
H
H
1b
(1.0 equiv)
6ba
86% yield
84% ee
N
N
N
N
N
N
O
O
t
Bu
t
Bu
N
N
N
N
N
c.
Survey of XAT ligands for the reaction to form
6ba
L9
86% yield
84% ee
L8
30% yield
84% ee
L10
41% yield
72% ee
L11
37% yield
76% ee
none
no
L7
no
L9
12.5 mol % Ni, 10 mol %
L7
and 2.5 mol %
L9
25 mol % Ni, 20 mol %
L7
and 5 mol %
L9
BF
3
·OEt
2
instead of AlCl
3
with DMA (1.0 equiv)
with ICH
2
CH
2
OPh (
4a-I
)
instead of
4a
No N
2
balloon
1
2
3
4
5
6
7
8
9
entry
change from optimal
yield
6ba
recovery
4a
86 (84% ee)
84
82
56
0%
6ba
, 0% recovery
4a
0%
6ba
53
28
63
25
67
37
54
62
34
0% recovery
4a-I
66
55
(74% ee)
(84% ee)
(84% ee)
(84% ee)
NiCl
2
·DME (17.5 mol %)
L7
(15 mol%),
L9
(2.5 mol %)
DMA (1 equiv)
BF
3
·OEt
2
(0.5 equiv)
Mn (3 equiv)
THF (0.4M), 23 °C, 14 h
O
Me
H
CO
2
H
H
O
Me
H
CO
2
H
H
53% yield
3aa
>20:1 dr, 99% ee
42% yield
7
55% yield
3ba
10:1 dr, 99% ee
41% yield
7
Cl
+
Me
Ph
O
O
O
2a
(2.0 equiv)
d.
Addition of
L9
does not induce decarbonylative coupling with
2a
t
Bu
t
Bu
t
Bu
a
. Postulated catalytic cycle for divergent acyl and decarbonylative coupling
b.
Optimization of decarbonylative cross-coupling
–CO
+CO
https://doi.org/10.26434/chemrxiv-2024-3r6z4
ORCID:
https://orcid.org/0000-0001-8244-9300
Content not peer-reviewed by ChemRxiv.
License:
CC BY-NC-ND 4.0
Thus, we began investigating
dual ligand systems that could
enable the cross
-
coupling of
meso
-
148
anhydrides with bromide
4a
(Figure
3
)
. A
literature review
suggested that tridentate ligands such as
149
terpyridines might be
good candidates
, as they
were previously
invoked for
XAT
in Ni
-
catalyzed cross
-
150
coupling
.
29
We initiated studies
using anhydride
1b
with the optimal
acyl coupling
conditions
,
but
used
a
151
combination
of
L7
a
nd
either 2,2′:6′,2′′
-
terpyridine (terpy,
L
8
)
or 4,4′,4′′
-
tri
-
tert
-
butyl
-
2,2′:6′,2′′
-
152
terpyridine (
t
Bu
-
terpy,
L
9
)
.
The
combined
ligand
loading
was equal to the
loading of NiCl
2
·DME
.
153
Consistent with our hypothesis, reactions containing
the
both
ligands provided
cross
-
coupl
ing
product,
154
favoring the
decarbonylative alkylation
6
b
a
in this case
(Figure
3
b)
.
Presumably, the overall slower rate
155
of primary alkyl
radical
formation resulted in CO deinsertion prior to radical capture (Figure
3
a).
This
156
enantioselective C(sp
3
)
C(sp
3
) couplin
g
demonstrated
that dual ligand systems can be used to
enable
157
classically challenging transformations such as asymmetric
C(sp
3
)
C(sp
3
)
coupling
.
The
cis
-
158
stereochemistry of the
cyclobutane
was confirmed by single crystal X
-
ray diffraction of
6b
d
(
vide infra
),
159
indicative
of
CO deinsertion
via a stereoretentive process.
160
Subsequent studies focused
on
optimizing the yield of cyclobutane
6ba
.
The u
se of terpyridine
161
ligands
L
8
and
L
9
furnished similar yields
of
6
b
a
(see
SI
for initial hit conditions)
; however,
L
9
was
162
selected for further studies due to better overall conversion of
4a
.
Ultimately, the optimal conditions
163
employed 22.5 mol% NiCl
2
·dme, 20 mol%
L7
, and 2.5 mol%
L
9
, with Mn
0
as the reductant, AlCl
3
as the
164
Lewis acid, and THF as solvent
(Figure
3
b
)
.
T
o mitigate catalyst inhibition by CO, the vials were attached
165
to
an argon
-
filled balloon to increase the effective headspace.
Under these conditions,
cyclobutane
6ba
166
was formed in 86% yield and 8
4
% ee.
Use of lower [Ni]
0
(Figure
3
c
, entry 4) or a higher ratio of
L
9
to
167
L7
(entry 5) resulted in lower yields of
6ba
.
In contrast to the acyl coupling (Figure
2c
),
use of
BF
3
·OEt
2
168
as the Lewis acid (entry 6) or addition of DMA (entry 7) resulted in lower yields of
6ba
.
We
also observed
169
lower yields of
6ba
when the corresponding alkyl iodide (
4a
-
I
)
wa
s used (entry 8)
.
The corresponding
170
https://doi.org/10.26434/chemrxiv-2024-3r6z4
ORCID:
https://orcid.org/0000-0001-8244-9300
Content not peer-reviewed by ChemRxiv.
License:
CC BY-NC-ND 4.0