Organic Letters
Letter
Our investigation into this photoredox-catalyzed protocol
began by studying the model reaction of nitrone 1a with N-
phenylglycine 2a. After various photocatalysts, bases, solvents,
and light sources were screened (Table S1), the optimized
reaction conditions were established as follows: By using 7 W
blue LEDs as the light source, the cross-coupling was carried
out in the presence of Ir[dF(CF3)ppy]2(dtbbpy)(PF6) (1.0
mol%) and Li2CO3 (20 mol%) in DMF at room temperature
for 24 h to give the desired β-amino hydroxylamine 3a-1 in
83% yield. Control experiments showed that no product was
formed in the absence of a photocatalyst, light source, or argon
protection, which back up a photoredox catalytic oxidation
mechanism. Furthermore, β-amino hydroxylamine 3a-1 was
obtained in only 30% yield without a base additive. This
indicates that the deprotonation of α-amino acid significantly
increase the concentration of α-aminoalkyl radicals, and thus a
modified condition (Protocol 1) was adopted to give cross-
coupling product 3j-1 in 85% yield. Compared with acyclic
nitrones, cyclic nitrones exhibited higher reactivity under the
general reaction conditions, and the desired products were
formed in excellent yields (3q−s).
We also explored the scope of α-amino acids. As shown in
Scheme 3, decarboxylative cross-coupling of N-arylglycines
Scheme 3. Photoredox Decarboxylative Cross-Coupling of
a b
,
N-Substituted Glycines with Nitrones
After identifying the optimized reaction conditions, we
examined the scope of nitrones and found that the reaction
tolerates a wide array of nitrone substrates. As for acyclic
nitrones, various C-substituents (1a−l) and N-substituents
(1m−p) on the nitrones were examined using N-phenylglycine
2a as the reaction partner, and the desired β-amino
hydroxylamines were obtained in moderate to good yields
(Scheme 2). When electron-rich nitrone 1j was used as an
electrophilic radical acceptor, more Li2CO3 with water was
needed to promote the deprotonation of α-amino acids and
Scheme 2. Photoredox Decarboxylative Cross-Coupling of
a b
,
N-Phenylglycine with Nitrones
a
General method: Nitrone (0.30 mmol), α-amino acid (0.90 mmol),
Ir[dF(CF3)ppy]2(dtbbpy)(PF6) (1.0 mol%), Li2CO3 (20 mol%),
b
DMF (1.5 mL), 7 W blue LEDs, room temperature, 24 h. Isolated
c
yield. Protocol 2: Nitrone (0.40 mmol), α-amino acid (0.40 mmol),
Ir[dF(CF3)ppy]2(dtbbpy)(PF6) (1.0 mol%), CsF (100 mol%), DMF
d
e
(2.0 mL), 32 W blue LEDs, room temperature, 36 h. Protocol 1. dr
1
values were determined by H NMR analysis of crude products.
with nitrone 1a gave the desired β-amino hydroxylamines in
moderate to good yields. As for para-substituted N-
arylglycines, electron-donating methoxy gave a lower yield
compared with electron-withdrawing halides and cyano
substituents (3a-5 vs 3a-2−4). Ortho-, meta-, or para-Cl-
substituted N-arylglycines gave almost identical and highest
yields (3a-3, 3a-6, and 3a-8). N-Arylglycine 2i with a
polysubstituted aryl group gave a moderate yield of 65%
(3a-9). For C-aryl nitrone 1g, the reaction outcome was
irrelevant to N-arylglycines, as products 3g-2−4 were obtained
in nearly identical yields. Furthermore, the reaction of N-bis-
substituted glycines proceeded smoothly under the general
conditions to provide desired cross-coupling products in
moderate yields (3a-10−a-12). It is worth mentioning that
the decarboxylation of N-Boc glycine proceeded under harsher
conditions (Protocol 2) to give the desired product 3a-13 in
61% yield. In addition, the reaction could also be applied to α-
branched N-phenyl α-amino acids in moderate to good yields
when the less sterically hindered nitrone 1j was used as the
a
General method: Nitrone (0.30 mmol), α-amino acid (0.90 mmol),
Ir[dF(CF3)ppy]2(dtbbpy)(PF6) (1.0 mol%), Li2CO3 (20 mol%),
DMF (1.5 mL), 7 W blue LEDs, room temperature, 24 h. Isolated
yield. Protocol 1: Nitrone (0.30 mmol), α-amino acid (0.90 mmol),
Ir[dF(CF3)ppy]2(dtbbpy)(PF6) (1.0 mol%) and Li2CO3 (100 mol%),
DMF (1.2 mL) and H2O (0.30 mL), 7 W blue LEDs, room
temperature, 24 h.
b
c
877
Org. Lett. 2021, 23, 876−880