Table 2 Photodehalogenation of aryl fluorides
was satisfactory even in the irradiation of aryl halide 12, (U =
r
0
.22 at 254 nm) although it was not completely consumed
a
H
3
PO
2
(0.5 M)
i-PrOH
ArH (% yield) t /h
after 20 h. In a couple of cases, when using H
3
PO as the
2
b
c
d
c
d
ArF (0.05 M)
ArH (% yield)
t /h
reductant, an acetone/water 5 : 1 mixture was tested as an
alternative reaction medium. As a result, photoreduction of
e
1
00
15
100
34
16
16
6
2
and 10 occurred in a shorter irradiation time and in a
comparable yield to those carried out in MeCN/water 5 : 1
(
Table 1).
e
f
f
9
8
8
8 (81); 43 (70) 30, 28 22 (45)
The extension to fluorides was successful; indeed the reduction
yields were in most cases quantitative and higher than those
observed with the corresponding aryl chlorides. In particular,
7
16
15
100
97
4
-fluorophenol (14) was easily converted to phenol by irradiation
either in an acetonitrile solution of hypophosphorous acid or in
neat i-PrOH. 4-Fluoroanisole (15) was not completely consumed
after 30 h irradiation but, at least in an acetonitrile solution
g
g,h
2 , 71
of H
3
PO , the reduction yield was good (98%). In this case,
2
a
b
c
substituting MeCN with acetone was deleterious since both
the consumption of 15 and the yield of formation of anisole
were decreased (Table 2). The photoreduction by H PO was
See Table 1. See Table 1. GC Yields based on consumed aryl halides.
In parentheses and italic the consumption of the aryl halide when
different from 100%. Irradiation at 254 nm. Lamp flux: ~6 ¥ 10
Einstein min cm . MeCN as the solvent. Reaction carried out in
d
-6
3
2
-1
-2
e
f
also effective with fluoroanilines 16 and 17 with a decrease
of the yield when using a lower amount of H PO (0.2 M).
g
h
2 2 3 2
H O/Me CO 1 : 5. MeOH as the solvent. H PO 0.2 M.
3
2
However, with these compounds irradiation in i-PrOH was by
far the best choice, with a yield of 100% and 97%, respec-
tively.
In detail, 4-chlorophenol (1), generally used as a model
compound for chlorinated pollutants, was found to give phenol
in 86% or 100% yield after 15 h irradiation in i-PrOH or with
The above data demonstrate a straightforward and
H
3
PO
2
, respectively. Methyl substituted chlorophenols 2 and 3
environment-friendly procedure for the clean hydrodehalogena-
tion of electron-rich aryl chlorides and fluorides via photo-
heterolysis and reduction of the thus generated triplet phenyl
cation. Homolytic hydrogen abstraction from the solvent by
this cation was supported by the formation of variable amounts
of pinacol (from the dimerization of the dimethylketyl radical
intermediate) in the experiments carried out in i-PrOH. It is
noteworthy that the reduction is equally effective with both
were likewise efficiently photoreduced despite the sterical hinder-
ing in the latter case. However, the yield of 3,5-dimethylphenol
from 3 dropped from 86 to 58% when the amount of H
3
PO was
2
decreased to 0.2 M. As for isomeric chloroanisoles, irradiation
in a methanol solution of hypophosphorous acid gave the best
yields in the reduction of the para isomer (4, 80%), whereas
isomers 5 and 6 were completely consumed only in i-PrOH
and reduced in >90% yield. 3,5-Dimethyl-4-chloroanisole (7)
was likewise transformed in 3,5-dimethylanisole in a good yield
independently of the conditions used. Photoreduction, however,
was less efficient (47% yield) when the reaction was carried
out in i-PrOH/water 3 : 1. It is noteworthy that increasing
the concentration of 7 to 0.1 M increased the yield of the
H
3
PO
i-PrOH (involving a C–H bond), as well as, though somewhat
less efficiently, with Et SiH (involving an Si–H bond). The
2
(where it involves abstraction from a P–H bond) and
3
smooth reaction in i-PrOH is due to the combined enhanced
photocleavage yield in this polar solvent and to its good
H-donating properties. The present protocol is advantageous
with respect to thermal methods in that it avoids the use of
neurotoxic organotin hydrides or aggressive metal hydrides as
well as the use of expensive and labile metal catalysts. Moreover,
anisole up to 86%. As for bis-ethers, H
the best solvent for the H PO mediated photoreduction of both
-chloro[1,3]benzodioxole (8) and chlorosesamol (9) in yields
similar to those obtained in neat i-PrOH. In the reduction of
-chloro-N,N-dimethylaniline (10) the irradiation time required
2
O/MeCN 1 : 5 was
3
2
5
6,20
contrary to some metal-catalyzed reductions
the present
4
reaction took place at room temperature. A possible drawback of
the method is the large consumption of energy for the UV lamps.
The reactions in Tables 1 and 2 are referred to a small amount
of aryl halide but the reaction can be scaled up to 15 mmol
by using a 150 ml photochemical reactor with an internal
lamp rather than a test tube with an external illumination.
Furthermore, experiments in progress show that irradiation by
solar light is also effective at least for derivatives absorbing in
the UVA, such as anilines. Moreover, at least for aryl chlorides,
an eco-friendly solvent such as acetone can be used in place
of acetonitrile. At any rate, the simple experimental procedure
does not require sophisticated operation, e.g. is not affected
by moisture. Indeed, water, rarely used in reductions because
of the chemical incompatibility with most reducing agents, is
a convenient cosolvent in the reaction with hypophosphorous
acid.
was shorter (4–6 h). i-PrOH was the best reducing agent and
in this solvent N,N-dimethylaniline was formed quantitatively
even using 0.1 M of 10. The preparative value of the reaction
was demonstrated in the photoreduction of 10 (15 mmol)
where N,N-dimethylaniline was isolated in 77% yield after
2
h irradiation in an immersion-well apparatus followed by
simple evaporation of the solvent and bulb to bulb distillation
see ESI).† This protocol was less efficient with S- or Si-
bonded substituents, as shown in the case of 4-chlorothioanisole
11) and 1-chloro-4-(trimethylsilyl)benzene (12), which gave the
(
(
corresponding dechlorinated derivatives in a medium yield and
in some cases reached only a partial conversion. 1-n-Butyl-4-
chlorobenzene, however, was quantitatively hydrodehalogenated
in i-PrOH. As for the experiments in neat i-PrOH the quantum
19
yield of reaction of compound 10 was high U
r
= 0.90, and
9
44 | Green Chem., 2009, 11, 942–945
This journal is © The Royal Society of Chemistry 2009
Dichiarante, Valentina
