Communications
doi.org/10.1002/cssc.202002459
ChemSusChem
water-splitting based synergistic strategy highlighted by
broad reaction scope, excellent functional group tolerance,
high yield and selectivity, and scalability is therefore a
promising alternative to conventional N-alkylation processes.
First, the visible-light-response semiconductors such as
PCN, KPCN, CdS, etc. with sufficient reduction potential for
water reduction were chosen as the photoredox catalysts
(the preparation procedures and corresponding character-
ization details are provided in the Supplemental Experimen-
tal section and Figures S1–S8) (Figure 1e). As shown in
Figure 1f, under visible light illumination, both CdS and KPCN
enabled high H2 evolution activities in the presence of
sacrificial reagents. Low H2 was produced over ZnS (~3.4 eV)
and TiO2 (~3.2 eV) due to their limited absorption of visible
light. With these initial results, photoredox centers coopera-
tive catalytic water-splitting coupled with methanol oxida-
tion for selective N-methylation of nitrobenzene is evaluated
(Figure 1g). The reaction was performed under visible light
(420 nm LED) with 0.4 mmol of nitrobenzene, methanol and
H2O (1.5 ml/1.5 ml) and 0.3 mmol of AlCl3 as the acidic
additive over Pd (1 wt%)/semiconductors. Among a variety of
semiconductors, KPCN was screened out as the most active
material to effectively drive the cascade di-alkylation of
nitrobenzene to produce N,N-dimethyl-4-methylaniline (Fig-
ure 1h). CH3CN has been demonstrated as the best solvent
compared with DMF and ethylacetate (EA) (more details can
be found in Table S1). The different loading of Pd cocatalyst
(1 wt%, 3 wt%, 5 wt%) has little impact on the reactivity and
selectivity. Interestingly, the control experiment without
water showed a remarkably decreased yield from 94% to
19%, attesting to the significant role of water in this N-
methylation reaction. Time-dependent experiments were
carried out with the optimized conditions. As shown in
Figure 1i, the p-nitrotoluene was exhausted gradually, along
with the formation of a mono-methylation product at the
initial period (0–12 h), which could undergo another meth-
ylation step by prolonging the reaction time to produce N,N-
dimethyl 4-methylaniline with excellent yields.
To evaluate the generality of our proposal, various nitro
compounds and alcohols were tested under optimized
conditions as shown in Scheme 1. Initially, the di-alkylation of
different substituted nitroarenes with methanol was ex-
plored. Reactions of nitroarenes bearing electron-donating
groups (pÀ Me, pÀ OMe) or electron-withdrawing groups
(pÀ CN, pÀ F, pÀ Cl) proceeded smoothly with methanol to
afford the desired N,N-dimethylated anilines in good to
excellent yields (61–94% yields) (2aa-2ak). In particular,
sterically hindered ortho-substituted substrates were also
active (2ag-2ai), giving the corresponding products with
excellent yields. Besides, substrates with a series of function-
alities such as aryl fluoride (2ab), chloride (2ac, 2af, 2ah),
nitrile (2ag), carbonyl (2aj) and sulfhydryl (2ak) substituents
were well-tolerated. This protocol was also applicable for the
synthesis of heteroaromatic dimethylanilines (2al-2am),
which are important substructures in bioactive molecules,
agrochemicals and advanced materials. For example, 4-
dimethylaminopyridine (DMAP, 2am), a useful nucleophilic
catalyst for a variety of reactions was obtained in 86% yield.
Derivatives of tocopherol (product 2an), L-menthol (product
2ao), indomethacin (product 2ap), adamantane (product 2aq)
and mefenamate (product 2ar) underwent reductive di-meth-
ylation in good to excellent yields. To extend the applicability
of this methodology, N,N-di-methylation of a variety of
biologically active molecules with nitro moieties was inves-
tigated. For example, di-methylation of nimesulide (a non-
steroidal anti-inflammatory drug, 2as)[44] and flutamide (2at)
proceeded smoothly in good to excellent yields (78% and
90% yields) without affecting the amide and sulfamine
functionalities. Gram scale synthesis of 2at proved the good
practical utility of our protocol. A rhodamine derivative
(2 au), which is widely used as a fluorescent probe, was
successfully obtained in good yield. Next, we explored the
possibility of applying different alkanols as the alkylation
partner. To our delight, this strategy could be well-extended
by using ethanol as the alkylation reagent, affording the di-
ethylated anilines (2ba–2bd) in good yields. Interestingly,
when applying this strategy to substituted benzyl alcohols,
mono-alkylated products were obtained in acceptable yields
with high selectivity (2ca–2cd). Finally, aliphatic nitro com-
pounds were also tested, but unfortunately these substrates
cannot function in this system.
On the basis of these obtained results, we are interested
in exploring the applicability of mono-methylation of nitro-
arenes. To this end, stepwise N-methylation of flutamide was
investigated in greater detail (Scheme 2A). By controlling the
reaction time, flutamide-NH2 (3aa), flutamide-NHMe (3ab)
and flutamide-N(Me)2 (2at) were successfully obtained, which
was very helpful to understand the reaction process as well
as the mechanism. With these stepwise results, the generality
of a mono-alkylation protocol was studied. As shown in
Scheme 2B, several nitroarenes were subjected to the opti-
mized conditions with carefully controlled reaction time,
which smoothly afforded the NÀ Me and NÀ Et anilines in 45–
81% yields.
N-alkylation is an important tool to regulate the biological
and pharmaceutical activities of life science molecules,
especially, N-alkylation reactions are of significance in the
synthesis of existing pharmaceuticals that belong to the 200
top selling drugs. Since our photocatalytic protocol is also
applicable to N-alkylation of amines from the stepwise
results, a variety of important bioactive or pharmaceutical-
related secondary amines were examined by utilizing this
mild strategy. First, late-stage functionalization of pharma-
ceutical amines was evaluated (Scheme 2C). Methylation of
vortioxetine and paroxetine, ethylation of paroxetine and
atomoxetine and benzylation of fluoxetine (3ca-3ce) all
proceeded smoothly in good to excellent yields without
affecting the core structures of the pharmaceutical mole-
cules. Acyclic amines, piperazine and piperidine rings all
worked well. Furthermore, our process could enable access
to several important pharmaceutical amines in a single step
with high efficiency and selectivity. For example, Benadryl
(3cf), an antihistamine, was prepared in 88 % yield in a
straightforward manner. Loxapine (3cg), an antipsychotic
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
ChemSusChem 2020, 13, 1–9
3
© 2020 Wiley-VCH GmbH
��
These are not the final page numbers!