Organic Letters
Letter
Other reactions of NHC-boranes that are induced by visible
light and a metal photocatalyst include 1,4-hydroborations of
electron-poor aromatic rings,6,7 homolytic substitutions of
fluorine atoms in various settings,6,8,9 arylboration and
hydroboration of fluorinated alkenes,8 and hydroboration of
imines.10 In addition to a metal catalyst, all these reactions
require other additives including bases and thiols.
We next studied the scope with respect to the alkene
compounds, and Scheme 3 shows results of reactions of
various β-aryl-α,β-unsaturated esters, lactones, amides, nitriles,
and ketones. A variety of β-phenyl-α,β-unsaturated esters
smoothly underwent regioselective hydroboration to afford α-
boryl-β-phenyl esters, regardless of the electronic nature of the
substituents on the aromatic ring (see 3b−3k). β-Heteroaryl-
α,β-unsaturated esters with pyridine, thiophene, or quinoline
rings were also viable substrates, delivering the corresponding
α-addition products in moderate yields (3l−3n). Hydro-
boration of two cumarins worked well to give a-boryl lactones
3o and 3p.
In contrast, α,β-unsaturated esters bearing an additional
ester or cyano group at α-position led exclusively to β-addition
products in excellent yield (3q−3t). When α,β-unsaturated
esters containing an additional methyl group at α-position
were subjected to this protocol, the β-addition pathway
continued to predominate, giving stereoisomeric β-addition
products 3u-β and 3u-β′ in 62% combined yield with an 8%
yield of α-addition product 3u-α. The change in regioselec-
tivity is probably caused by synergistic steric and electronic
effects of the added substituent. The steric effect disfavors α-
addition. When a second electron-withdrawing group is added
at the α-position, the nucleophilicity of the NHC-boryl radical
takes over and β-selectivity is favorable.
̂
́
Predating all this recent work, Lacote and Lalevee reported
in 2012 the first examples of generation of NHC-boryl radicals
by visible light.11 They described soft polymerization of
acylates with NHC-boranes and acridine orange. Since then,
organic photocatalysts (dyes) have seen little use in small
molecule reactions of NHC-boranes. The hydroboration
reactions of Zhu succeeded with an iridium photocatalyst
but failed with eosin Y and 4-CzIPN (see Table S1 for
structure).5 Wu reported a few successful defluoroborylation
reactions with 4-CzIPN, but conditions with Ir photocatalysts
were much preferred.6
We have discovered that visible-light irradiation of
unsaturated carbonyl compounds or nitriles, NHC-boranes,
and a small amount of photocatalyst 2,4,5,6-tetrakis(diphenyl-
amino)isophthalonitrile (4-DPAIPN) regioselectively delivers
either α- or β-NHC-boryl functionalized compounds, depend-
ing on the substrate structure (Scheme 2). The starting
Scheme 2. Strategy for the Synthesis of Boryl
Functionalized Molecules
All the reactions in the scope study were conducted on 0.2
mmol scale. To test the scalability, we conducted a reaction
with 1.00 g (5.8 mmol) of 1a and 2 (674 mg, 1.1 equiv) under
the standard conditions. This provided 1.26 g of 3a, and the
yield (78%) is comparable to the small-scale experiment (83%,
Next, we examined alternatives to esters as activating groups.
Pleasingly, representative α,β-unsaturated amides gave α-boryl
amides in good yields (3v−3z). Notably, several β-aryl-α,β-
unsaturated nitriles and a ketone also underwent the α-
additions affording the α-NHC-boryl products (3aa−3af).
However, cinnamic acid and cinnamaldehyde were inert to the
hydroboration process and did not provide the boryl carbonyl
compounds.
materials are structurally diverse and readily accessible, and the
scope is broad. The conditions are mild and metal-free, and no
base, cocatalyst, or other additive is needed.
The β-aryl group in these substrates stabilizes the adduct
radical by resonance, thereby promoting α-addition. In
contrast, β-addition might be favored due to the strong
nucleophilicity of NHC-boryl radical if this group is absent. To
test this notion, we next studied the reactions to β-alkyl-α,β-
unsaturated carbonyl compounds (Scheme 4). Indeed, a wide
range of β-alkyl-α,β-unsaturated esters and lactones were
suitable substrates and underwent β-addition, giving solely β-
boryl esters and lactones in generally good yields (3ag−3an).
Increasing the steric effect on the β-carbon, such as ethyl 3-
methylcrotonate, failed to give the desired hydroboration
product. Interestingly, ethyl acrylate, which is easily poly-
merized by NHC-boryl radicals,11 was hydroborated to provide
the β-NHC-boryl ester 3ao in good yield. Reactions of β-alkyl-
α,β-unsaturated amides were also highly β-selective, albeit with
moderate yields (3ap−3ar). Finally, a β-alkyl-α,β-unsaturated
nitrile also afforded the β-boryl nitrile in good yield (3as).
Scheme 5 summarizes experiments aimed at providing
mechanistic information. In radical-blocking experiments,
reactions of 1a and 2 were conducted under standard
conditions in the presence of 2,2,6,6-tetramethyl-1-piperidinyl
oxyl (TEMPO) as a radical scavenger. When 4 equiv of
TEMPO were used, the yield of α-boryl ester 3a was decreased
to 20%. When 6 equiv of TEMPO were used, 3a was not
We chose ethyl cinnamate 1a and 1,3-dimethylimidazoyl-2-
ylidine borane 212 as pilot reaction partners, and Table S1 (see
assortment of preliminary reactions. In the first experiment,
irradiation of a room temperature acetonitrile solution of 1a, 2
(2 equiv), and the photocatalyst 3-DPAFIPN13 (10 mol %)
with a 30 W blue light-emitting diode (LED) for 12 h provided
α-boryl ester 3a in 56% isolated yield (entry 1). No product
was produced in the absence of the light or photocatalyst
(entries 2 and 3). Next, we reduced the amount of borane and
found that 1.1 equiv gave about the same yield as 2 equiv
(entries 4 and 5). Among the various photocatalysts tested (4-
DPAIPN, 4-CzIPN, 4-CzPN, 5-CzBN, eosin Y, and DCN,
entries 6−11), 4-DPAIPN gave the best yield (99% NMR,
82% isolated, entry 6).14 Decreasing the catalyst loading to 5
mol % resulted in a moderate decrease in yield (94% NMR,
79% isolated, entry 12), whereas increasing the amount of
catalyst gave about the same yield (entry 13). Finally,
decreasing the reaction time from 12 to 8 h did not change
the yield, while a further decrease to 5 h gave a moderate
reduction (compare entries 6, 14, and 15).
Based on these results, we selected the conditions of entry
14 (10% mol 4-DPAIPN, 8 h, rt) as standard for the ensuing
study of scope.
4354
Org. Lett. 2021, 23, 4353−4357