Organic Letters
Letter
a
Achieving atom-economic reactions is an important pursuit
of sustainable development.8 If the CO2 released from
decarboxylation could be sequestered as a C1 building
block9 for subsequent bond formation, the whole process
could be carbon-neutral and thus contribute to atom-economy.
In continuation of our effort in CO2 conversion11 and inspired
by the above-mentioned developments in decarboxylation,
herein we report our progress toward the above goal.
Specifically, we were curious about whether CO2 released
from decarboxylation could serve as an electrophile to react
with the carbanion prior to protonation. If successful, α-
aminomethyl carboxylation of styrenes would occur instead of
simple hydroalkylation (Scheme 1c). It is important to note
that such a process could provide straightforward access to
those useful α,α-disubstituted γ-amino acids12 and their
derivatives, such as γ-lactams (Scheme 1d). During the
preparation of our manuscript, Yu and co-workers reported a
work based on the similar concept, in which CO2 released from
decarboxylation of α-amino acids and peptides could be reused
for carboxylation (Scheme 1b).10
Table 1. Evaluation of Conditions
b
entry
PC
Solvent
3ab yield (%)
1
2
3
4
Ir[(dF(CF3)ppy)]2(dtbbpy)PF6
4CzIPN
Ru(bpy)3Cl2·6H2O
Ru(bpy)3·2PF6
DMSO
DMSO
DMSO
DMSO
DMSO
DMF
CH3CN
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
36
<5
22
40
49
16
NR
74
68
0
57
68
64
70
27
c
5
Ru(bpy)3·2PF6
6
7
Ru(bpy)3·2PF6
Ru(bpy)3·2PF6
Ru(bpy)3·2PF6
Ru(bpy)3·2PF6
d
8
9
de
,
df
,
10
11
12
13
14
15
Ru(bpy)3·2PF6
d
d
d
d
d
Ru(bpy)3Cl2·6H2O
Ir[(dF(CF3)ppy)]2(dtbbpy)·PF6
Ir[(2.4-2Fppy)]2(dtbbpy)·PF6
Ir[(2.4-2Fppy)]2(bpy)·PF6
4CzIPN
A proposed mechanism of this design is depicted in Scheme
2. With photocatalysis, decarboxylation takes places to
a
Reaction conditions: ethene-1,1-diyldibenzene 1a (0.2 mmol),
sodium 2-(methyl(phenyl)amino)acetate 2b (0.3 mmol), catalyst (2
Scheme 2. Design Plan
mol %), under N2, solvent (2 mL), 48 h, in a sealed Schlenk tube,
b
irradiated by a 2 × 3 W blue LEDs, unless otherwise noted. Isolated
c
d
yield. Sodium N-methyl-N-phenylglycinate 2b (0.6 mmol), CO2 (1
e
f
atm) was used in place of N2, Catalyst (1 mol %). Without catalyst
or blue LEDs. NR = no reaction.
Ru(bpy)3·2PF6 proved to be the best catalyst (3ab, 40%, entry
4). When the amount of 2b was increased to 3 equiv, the yield
of 3ab could be improved to 49% (entry 5). A survey of
solvents indicated that DMSO worked best (entries 4, 6, and
7). In this reaction, it is clear that the CO2 released from the
decarboxylation step was used as the carbon source of the
newly formed carboxylic acid group. Subsequent considerable
efforts were devoted to further optimization. Unfortunately,
higher efficiency was not achieved. Nevertheless, it serves as an
important proof of concept for CO2 reutilization. The low
yield might be attributed to the low concentration of CO2
released from 2b. Therefore, in order to promote the reaction
to complete carboxylation, external CO2 (1 atm) was used to
replace N2, and the yield of 3ab increased to 74% (entry 8).
Lowering the catalyst loading to 1 mol % resulted in slightly
diminished efficiency (68%, entry 9). Furthermore, control
experiments confirmed that both photocatalyst and blue light
are essential for this reaction (entry 10). Finally, other catalysts
were compared again with the best conditions, but they were
all inferior (entries 11−15).
With the optimized conditions (entry 8, Table 1), the
substrate scope of styrenes was investigated. As shown in
Scheme 3, these reactions proceeded smoothly under the
standard conditions regardless of the electronic nature of the
substituents in the para-position of the phenyl ring (3ab−3ib).
Furthermore, steric hindrance has no effect on the reaction
efficiency. Those substrates bearing an m- (1i) or o-methyl
(1j) group could afford both of the corresponding products
3jb and 3kb in 70% and 67% yields, respectively. Multi-
substituted styrenes 1k and 1l were also good substrates,
generating the desired products in good yields (3 lb, 75%;
3mb, 67%). To our delight, this reaction also proceeded
smoothly with styrenes bearing 2-naphthyl (1n), 2-furyl (1o),
and 2-thienyl (1p) substituents, forming the corresponding
generate aminomethyl radical B, which then adds to the olefin
double bond to form carbon radical D. Single-electron transfer
from the reduced form of photocatalyst to this radical
generates carbanion E. Next, sequestration of the released
CO2 with this anion followed by protonation leads to γ-amino
acid 3. Additionally, when the amino moiety containing a free
N−H bond (R2 = H), intramolecular lactamization of
intermediate 3 could generate diverse γ-lactams 5.
At the outset, ethene-1,1-diyldibenzene 1a and N-methyl-N-
phenylglycine 2a were employed to test our hypothesis (see
Table S1 for details). Unfortunately, initial examination of the
reaction under N2 led to trace desired product 3a. Instead,
hydroalkylation to form 6 as the main byproduct, even with
various photocatalysts (Table S1, entries 1−4). When 2 equiv
of K2CO3 was added, the yield of 3ab was dramatically
increased to 42%; other inorganic bases gave comparable
results, while organic bases such as TBD and DBU only
afforded 6 as the main product (Table S1, entries 5−9).
When sodium 2-(methyl(phenyl)amino)acetate 2b was used
instead of 2a, the desired product 3ab was isolated in 36%
yield in the presence of 2 mol % of Ir[(dF(CF3)-
ppy)]2(dtbbpy)PF6 as the catalyst (Table 1, entry 1). To
further improve the reaction efficiency, other catalysts were
screened. 4CzIPN showed low catalytic activity (entry 2).
Ru(bpy)3Cl2·6H2O resulted in only 22% yield (entry 3), but
2896
Org. Lett. 2021, 23, 2895−2899