Organic Letters
Letter
benign protocol was also demonstrated by the successful
functionalization of natural products and drug-based complex
molecules. (Seven examples are shown in Scheme 2.)
Initially, N-Me-quinoxalin-2(1H)-ones 1a and 4-benzyl-1,4-
dihydropyridine-3,5-dicarboxylate 2a (Bn-DHP) were selected
as model substrates to test the feasibility of our photochemical
reaction design. Irradiating a solution of 1a and 2a in the
presence of 2.0 equiv of BI-OAc in CH3CN with a 24 W blue
LED for 3 h afforded the desired C3-benzylation product 3aa
in 52% isolated yield (Table 1, entry 1). Notably, hypervalent
(e.g., methyl) and electron-withdrawing (e.g., fluoro, chloro,)
groups on the benzene scaffold in 1 can afford the
corresponding alkylation products in generally good yields
(3ba−ea). Moreover, N-protecting group in quinoxalin-
2(1H)-ones containing reactive functional groups, including
alkenes, alkynes, alcohols, halides, ketones, and esters,
underwent the desired radical alkylation with high efficiency
(products 3fa−la, 73−98% yields). The unprotected quinox-
alin-2(1H)-one 1m is also a viable reaction partner for this
transformation. Note that the high efficiency can almost be
maintained when running the reaction on a 3 mmol scale
(3ma, 73%).
a
Table 1. Reaction Optimization
We then evaluated the scope of the 4-alkyl-1,4-dihydropyr-
idine component in this photoredox-catalyst-free protocol. As
revealed in Scheme 2, the protocol was amenable to
dihydropyridines bearing benzyl groups with either electron-
withdrawing (−F, −Cl) or electron-donating (−Me) sub-
stituents at the para positions on the phenyl ring, leading to the
desired alkylated products 3ab−ad in 75−91% yield.
Furthermore, R-DHPs containing α-Me-branched benzyl
groups were also proven to be amenable substrates (3ae and
3af, 83 and 70% yields, respectively). Nonbenzylic primary and
secondary alkyl groups, such as R-DHPs derived from 3-
phenylpropanal (2h), pentanal (2i), 3-methylbutanal (2j),
isobutyraldehyde (2k), 2-ethylhexanal (2l), and 2,6-dimethyl-
hept-5-enal (2m), were also successfully converted to the
corresponding products in good yields (3ah−am, 33−98%
yield). Note that the C3−H methylation product 3ag could
not be obtained by using the developed method, likely due to
the low stability of the formed methyl radical species. Good
tolerability was observed when a secondary cyclic alkyl group
was introduced (3ak−am, 82−97% yield). It is worth noting
that a pyranose-derived DHP could be involved to achieve a
new pyranose derivative 3an in 76% yield. In addition, the
tertiary alkyl group also worked well to give the product 3ar in
51% yield.
b
entry
deviation from standard conditions
yield (%)
1
2
3
4
5
6
7
8
none
52
41
PIDA instead of BI-OAc
PIFA instead of BI-OAc
BI-OH instead of BI-OAc
Et2O instead of CH3CN
acetone instead of CH3CN
H2O instead of CH3CN
THF instead of CH3CN
EA instead of CH3CN
1,4-dioxane instead of CH3CN
trace
n.r.
trace
trace
trace
45
62
65
9
10
To further highlight the potential application of this
photoredox-catalyst-free radical alkylation protocol, we have
employed some natural isolates and drug-derived complex
molecules in the reaction substrates. As the result reveals in
Scheme 2, quinoxalin-2(1H)-ones derived from natural
products, such as o-vanillin, vanillin, vanillylacetone, and
piperonylic acid, and pharmaceutically compounds, including
ibuprofen, aspirin, and vitamin E, all functioned as competent
substrates, affording the desired C3-benzylation products
3na−ta in good to excellent yields (63−92%).
With the aim of showing the utility and practicality of this
radical alkylation reactions, some useful synthetic trans-
formations of final products were conducted, as shown in
Scheme 3. The treatment of N-allyl-derived quinoxalin-2(1H)-
one product 3ga with m-chloroperoxybenzoic acid (m-CPBA)
afforded epoxidation product 4 in 59% yield (Scheme 3A).13a
Moreover, a copper-catalyzed click reaction could easily
transfer the alkyne moiety in 3ha into an important triazole
skeleton in 78% yield (Scheme 3B).13b Parvaquone is always
employed for the treatment of theileriosis.13c To our delight,
the strategy could be successfully extended to the one-step
synthesis of parvaquone in good yield from lawsone (Scheme
3C).
iodine radical acceptors had a significant influence on the
reaction efficiency. The replacement of BI-OAc with other
commonly used hypervalent iodine reagents, such as PIDA
(phenyliodine(III) diacetate), PIFA ([bis(trifluoroacetoxy)-
iodo]benzene), and BI−OH (hydroxybenziodoxoles), resulted
in a dramatic decrease in the reaction efficiency (Table 1,
entries 2−4). When PIFA was used, only a trace amount of 3aa
was observed, which might be due to the formation of some
unknown byproducts. To further increase the reaction yield,
we then turned our attention to identify the influence of the
reaction media. Only a trace amount of 3aa was observed when
the reaction was performed in diethyl ether, acetone, or water
(Table 1, entries 5−7). Other reaction solvents, including ethyl
acetate, THF, and 1,4-dioxane, gave a comparable yield (Table
1, entries 8−10). After extensive screening, we were gratified to
find that CHCl3 gave an optimal 94% yield (Table 1, entry 11).
(For more details of the optimization, see the SI.)
By using the optimized reaction conditions shown in Table
1, entry 11, we then investigated the generality of this
photochemical radical alkylation process (Scheme 2). We first
evaluated the substrate scope of quinoxalin-2(1H)-ones 1. It
was found that this radical alkylation reaction tolerates a wide
range of substituents on the aromatic ring of quinoxalin-
2(1H)-ones. The incorporation of both electron-donating
To probe the plausible reaction mechanism, some
preliminary mechanistic experiments were conducted, as
shown in Figure 1. It was found that the reaction was
completely shut down when 2.0 equiv of radical scavenger
C
Org. Lett. XXXX, XXX, XXX−XXX
He, Xiang-Kui
