4
E. Hasegawa et al. / Tetrahedron xxx (2016) 1e8
halobenzenes 4, promoted by 1 and t-BuOK in DMSO, were ex-
plored (Table 4). As expected, the corresponding dehalogenated
product, dodecyloxybenzene 5 is produced in these processes al-
though the yields are significantly lower than those for reactions of
good yields (entries 1 and 4). The reactions proceed more slowly
when carried out in CH Cl . Moreover, either the presence of DBU or
2
2
prolonged irradiation time leads to increases in conversions of the
starting materials (entries 2, 3, 5, and 6). In addition, the meth-
anesulfonyl indole 6c is less reactive than the aryl analogues yet,
even in this case, the presence of DBU enhances the efficiency of the
process that gives 7b in reasonably good yield (entries 7 and 8).
Further studies showed that photo-reduction reaction of N,N-
diphenyl-N-tosyl amine (8) using 1a as the photo-reductant gen-
erates the deprotected amine 9 (Table 6). LED irradiation of a mix-
ture of 8 and 1a for 2 h (same condition as in entry 1, Table 5) does
not induce the expected desulfonylation reaction (entry 1). Notably,
when DBU is present in this mixture, irradiation for the same time
period leads to formation of 9 (entry 2). In the absence of DBU,
a longer irradiation time (24 h) is required to promote formation of
21
the corresponding allyloxy-substituted substrates 2 (see above).
In reaction of iodide 4a, photo-reduction takes place in the ab-
sence of t-BuOK, but the presence of this base accelerates the re-
action (entries 1 and 2). On the other hand, irradiation of a solution
of 4a not containing 1a does not lead to formation of 5 (entry 3), in
a similar manner to that observed for 2a (entry 2 in Table 2). The
yield of 5 in the photoreaction of chloride 4b, promoted by irradi-
ation of 1a in the presence of t-BuOK is extremely low (14%) even
though 69% conversion of 4b occurs (entry 4). In contrast, the use of
1
b and shorter-wavelength light to promote photoreaction of 4b,
although requiring longer times, does lead to an improved yield of 5
entry 6).
2
3
(
9 in good yield (entry 3).
Table 4
a
Table 6
Photoreaction of 1-dodecloxy-2-halobenzenes 4 with 1 and t-BuOK
LED-promoted photoreaction of N,N-diphenyl-N-tosyl amine 8 with 1aa
Conv of 8 (%)b
Yield of 9 (%)b
Entry
4
1
t-BuOK
4a 1a Added >390
4a 1a >390
4a Added >390
4b 1a Added >390
4b 1a >390
4b 1b Added >340
l
(nm) Time (h) Conv of 4 (%)b Yield of 5 (%)b
Entry
DBU
Time (h)
1
2
3
4
5
6
3
3
3
100
62
8
31
26
0
1
2
3
d
2
2
24
0
31
86
0
27
75
d
Added
d
d
24
24
24
69
20
71
14
Trace
44
a
8
(0.20 mmol), 1a (1.2 equiv vs 8), DBU (1.2 equiv vs 8), DMF (2.0 mL); 7.3 W
d
white LED.
b
Determined by using 1H NMR.
a
4
(0.20 mmol), 1 (2.0 equiv vs 4), t-BuOK (2.0 equiv vs 4), solvent (2.0 mL);
5
00 W Xe lamp with glass filter L-42 (>390 nm) or UV-37 (>340 nm).
b
Determined by 1H NMR.
Recently, reductive decyanation reactions of malononitrile de-
rivatives were described independently by Curran group and
2
4e26
Murphy group.
The process developed by Curran et al. involves
In order to explore the generality of photo-promoted reductions
a free radical protocol using N-heterocyclic carbene borane and di-
t-butyl peroxide, while a ‘super electron donor’ is used to promote
the reductive cleavage reactions devised by Murphy et al. Both
byHOAr-DMBIH, other reducible substrates wereutilized. This effort
began by probing LED-promoted photo-reductive desulfonation
2
2
reactions of N-sulfonyl indoles 6 (Table 5). The results show that
e2 h white light LED irradiation of DMF solutions of 1a and 6a or 6b
ꢁ
reaction conditions require the use of temperatures above 100
C
1
and reaction times longer than 15 h. Because of its mild nature, we
anticipated that the visible light-promoted protocol at room tem-
perature might be an applicable decyanation protocol. To probe this
proposal, studies of the decyanation reaction of dibenzyl malono-
nitrile 10 were conducted (Table 7). Initially, we found that re-
actions of 10 using 1a and DBU or t-BuOK, and LED or HgeXe lamp
irradiation were unsuccessful (not shown in Table 7). However, 1b
in the presence of bases does promote the process that produces
mono-nitrile 11. Irradiation of a DMSO solution of 10, 1b and DBU
leads to desulfonylation forming the respective indoles 7a and 7b in
Table 5
a
LED-promoted photoreaction of N-sulfonyl-indoles 6 with 1a
using HgeXe lamp (l>340 nm) promotes this process although the
yield of 11 is low (entry 1). Using t-BuOK as a base and increase in
its quantity lead to the completion of the reaction to produce 11 in
reasonably good yield (entry 2). Finally, it was found that visible
Conv of 6 (%)b
Yield of 7 (%)b
light irradiation using HgeXe (l>390 nm) as well as LED promoted
Entry
6
DBU
Solvent
DMF
Time (h)
the reaction to give the modest yields of 11 while a longer irradi-
ation time was needed when LED was employed (entries 3 and 4).
In the final phase of this investigation, the visible light protocol
using LED was applied to reductive cleavage reactions of epoxy
ketones to form hydroxyl ketones. This process serves as an effec-
tive model to evaluate properties of amine donors in PET reac-
1
2
3
4
5
6a
6a
6a
6b
6b
6b
6c
6c
d
2
2
2
1
2
6
2
2
100
26
45
100
95
100
38
77
13
25
82
79
91
23
74
d
CH
CH
2
Cl
Cl
2
2
Added
d
2
DMF
d
CH
CH
2
Cl
Cl
2
2
c
6
d
2
7
8
d
DMF
DMF
12a,c,d,f,h,i,m
tions.
Optimization of the reaction was explored using
Added
100
the benzalacetophenone (chalcone) epoxides 12aed as well as
benzalacetone epoxide 12e (Table 8). The results show that the
percent conversion of the 6 h photoreaction of 12a in DMF to form
13a increases when larger quantities of the photo-reductants 1a are
a
6
(0.20 mmol), 1a (1.2 equiv vs 6), DBU (1.2 equiv vs 6), solvent (2.0 mL); 7.3 W
white LED.
b
c
Determined by using 1H NMR.
b (0.40 mmol), CH Cl (4.0 mL).
6
2
2