Organic Letters
Letter
a
Scheme 1. Cross Coupling of Aziridines
Table 1. Optimization of Reaction Conditions
entry
standard reaction conditions
yield (%)
1
2
none
70
0
without Ru(bpy)3Cl2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Ru(bpy)3(PF6)2 instead of Ru(bpy)3Cl2
[Ir(ppy)2(dtbbpy)]PF6 instead of Ru(bpy)3Cl2
NaI/PPh3 instead of Ru(bpy)3Cl2
Eosin Y instead of Ru(bpy)3Cl2
without blue LEDs light (in dark)
white light instead of blue LEDs light
without NiBr2·DME
NiBr2 instead of NiBr2·DME
NiCl2 instead of NiBr2·DME
without L1
L2 instead of L1
L3 instead of L1
L4 instead of L1
without Et3N
Na2HPO4 or Na2CO3 instead of Et3N
MeCONMe2 instead of MeCONMe2/MeCN
MeCN instead of MeCONMe2/MeCN
none
63
50
57
15
0
<5
trace
10
<5
trace
30
40
trace
trace
trace
43
cooperative catalysis, affording β-phenethylamine derivatives
via the C3−N bond cleavage (Scheme 1A-c).6b However, the
utilities of these protocols all are limited because of the
requirement of prefunctionalized reaction partners (e.g.,
organometallics, aryl iodides). Although the transition-metal-
catalyzed aryl C(sp2)−H coupling of aziridines with function-
alized arenes (e.g., 2-arylpyridines, aromatic acids) has been
developed in recent years,7 the directing groups are pivotal to
enable the success of the reaction.
We envisioned that direct use of electrophilic arenes, such as
pyridin-1-ium salts,8 to replace the common aryl halide
derivatives, would establish new cross-electrophile coupling
methods to allow transformations of aziridines. Herein, we
report a new visible light photoredox/nickel-cocatalyzed cross-
electrophile coupling of aziridines with pyridin-1-ium salts for
producing β-(1,4-dihydropyridin-4-yl)-ethylamines. By means
of the photoredox reductive quenching cycle strategy, a variety
of aziridines undergo the C2−N bond cleavage and
dearomatization alkylation cascades with pyridin-1-ium salts
to enable the formation of C(sp3)−C(sp3) bonds, which has a
different regioselectivity from the precedent work reported by
the Doyle group. Notably, the resulting products, especially β-
aryl ethylamines, are important structural units in natural
products and pharmaceuticals,9 as well as consist of the
unsaturated 1,4-dihydropyridin-4-yl group and the highly
reactive amino group which provide the potential to build
and modify pharmaceutically important molecule derivatives.8
We initially conducted the ring opening reaction using 2-(p-
tolyl)-1-tosylaziridine 1a and 1-cyclohexyl-2,4,6- triphenyl-
pyridin-1-ium tetrafluoroborate 2a as the model substrates
for optimization of reaction conditions (Table 1). We found
that the reaction of aziridine 1a and pyridin-1-ium 2a afforded
the desired product 3aa in a satisfactory yield (70%) in the
presence of Ru(bpy)3Cl2 (1 mol %), NiBr2·DME (5 mol %),
2,6-di(1H-pyrazol-1-yl)pyridine L1 (10 mol %) and Et3N (2
equiv) in anhydrous MeCONMe2/MeCN (1:4; 2 mL) at
room temperature for 12 h (entry 1). Further screening
showed that the cooperative catalytic system, including
Ru(bpy)3Cl2, NiBr2·DME, 2,6-di(1H-pyrazol-1-yl)pyridine L1
and Et3N (2 equiv), were crucial for the success of this reaction
because omission of each one resulted in no reaction (entries
2, 7, 9, 12, and 16). Other photocatalysts, such as
Ru(bpy)3PF6, [Ir(ppy)2(dtbbpy)]PF6, NaI/PPh3 and Eosin
Y, showed reactivity, but they all were less efficient than
Ru(bpy)3Cl2 (entries 2−6). The reaction is sensitive to blue
light as white light had no effect (entry 8). Using NiBr2 and
NiCl2 instead of NiBr2·DME resulted in lower yields (entries
55
62
b
20
a
Standard reaction conditions: 1a (0.2 mmol), 2a (1.2 equiv),
Ru(bpy)3Cl2 (1 mol %), NiBr2·DME (5 mol %), L1 (10 mol %), Et3N
(2 equiv), anhydrous MeCONMe2/MeCN (1:4; 2 mL), room
temperature, and 12 h. 1a (1 mmol) and 24 h.
b
10−11). Substituted bipyridine ligands L2 and L3 are highly
reactive (entries 12−13), but bipyridine L4 as ligand was inert
(entry 15). Notably, other inorganic bases, including Na2HPO4
and Na2CO3, completely inhibited the reaction, suggesting that
Et3N may serve as a single electron transfer reagent to effect
the reaction. Using MeCONMe2 or MeCN as the medium led
to less efficiency than a mixed MeCONMe2/MeCN solvent
(entries 18−19). Gratifyingly, the reaction was consistent with
the loading of 1a up to 1 mmol, giving 3aa in good yield (entry
20).
With the optimized reaction conditions in hand, we set out
to explore the substrate scope of this ring opening protocol
using various aziridines 1 and pyridin-1-ium tetrafluoroborates
2 (Scheme 2). Initially, we evaluated the substitution effect on
2 position of 1-tosylaziridine 1 in the presence of pyridin-1-ium
salt 1a, Ru(bpy)3Cl2, NiBr2·DME, L1 and NEt3 (2 equiv)
under blue LEDs light (3ba−3wa). The results showed that an
aryl group presented at 2 position was crucial for making the
reaction successful (3ba−3ua, 3wa−3ya) since an alkyl group
led to no reactivity (3va). Moreover, a wide range of
t
substituents, including bulking Bu, MeO, Ph, Br, Cl, F, CF3,
and CO2Et, on the 2-aryl ring were well tolerated, and the
electronic and steric hindrance effects both affected the
reaction (reactive order: electron-donating group > electron-
withdrawing, para > meta > ortho; 3ba−3ma). The use of
aziridine 1c bearing a electron-donating p-MeO group afforded
3ca in 72% yield, whereas aziridines 1h−i possessing an
electron-withdrawing p-CF3 or CO2Et group resulted in the
3697
Org. Lett. 2021, 23, 3696−3700