Photoreductive defluorination of hexafluorobenzene on metal-doped ZnS photocatalysts under visible light irradiation

Article abstract of DOI:10.1246/cl.2005.1056

Metal-(Cu, Ni, or Pb) doped ZnS photocatalysts have the ability to photoreductively defluorinate hexafluorobenzene under visible light irradiation; the defluorination mechanism on Pb-, or Cu-doped ZnS is different from that on Ni-doped ZnS. The role of Pb and Cu dopants in ZnS is not only to extend the photocatalytic response toward visible light but also to act as redox reaction sites in the photoreductive defluorination. Copyright

Full text of DOI:10.1246/cl.2005.1056

1
056
Chemistry Letters Vol.34, No.7 (2005)
Photoreductive Defluorination of Hexafluorobenzene on Metal-doped ZnS Photocatalysts
under Visible Light Irradiation
ꢀ
y
y;yy
and Ryoichi Nakagaki
Shigeru Kohtani, Yoshihiko Ohama, Yu Ohno, Issei Tsuji, Akihiko Kudo,
Graduate School of Natural Science and Technology, Kanazawa University, Kakuma-machi, Kanazawa 920-1192
Department of Applied Chemistry, Faculty of Science, Science University of Tokyo, 1-3 Kagurazaka, Shinjuku-ku, Tokyo 162-8601
y
yy
Core Research for Evolutional Science and Technology, Japan Science and Technology Agency, Kawaguchi, Saitama 332-0012
(Received April 19, 2005; CL-050539)
ꢁ
3
Metal-(Cu, Ni, or Pb) doped ZnS photocatalysts have the
ability to photoreductively defluorinate hexafluorobenzene
under visible light irradiation; the defluorination mechanism
on Pb-, or Cu-doped ZnS is different from that on Ni-doped
ZnS. The role of Pb and Cu dopants in ZnS is not only to extend
the photocatalytic response toward visible light but also to act as
redox reaction sites in the photoreductive defluorination.
photocatalytic performance. F6 (25 mmol dm ), triethylamine
ꢁ3
ꢁ3
(1 mol dm ), and photocatalyst powders (1 g dm ) contained
3
in acetonitrile (25 cm ) were placed in a cylindrical quartz vessel
and degassed with N2 gas flow. Irradiation experiments were
carried out using a 300-W xenon lamp equipped with an appro-
priate cut-off filter with stirring in a water bath (280 K).
Figure 1 shows the decrease in F6 and evolution of F5 ob-
tained with the three metal-doped and nondoped ZnS photocata-
lysts under visible light irradiation. When the metal-doped ZnS
was used, F5 was produced with decreasing F6 in all cases. The
three F6 degradation reactions follow first-order rate kinetics,
Perfluorinated organic compounds are well known to be
thermally and chemically stable, because of their strong C–F
bonds and the small size of the fluorine atom. Photochemical re-
actions have attracted much attention with respect to inducing
defluorination or decomposition for such perfluorinated com-
ꢁ
1
with nearly the same rate constant of 0.1 day . On the contrary,
in a nondoped ZnS sample, the decrease rate for F6 was much
slower (indicated by crosses in Figure 1) and negligible evolu-
tion of F5 was observed. It is demonstrated that Ni, Cu, and
Pb metals doped in ZnS are responsible for the defluorination
reaction observed upon photoexcitation with visible light.
Decrease in the photocatalytic activity for Cu-ZnS was brought
about when dried Cu-ZnS powders were used. On the other
hand, efficient degradation of F6 as well as formation of F5 were
observed in a nondoped ZnS solution under UV irradiation; F6
was completely decomposed within 2 days, and ca. 50% of F6
was converted into F5.
1
–3
pounds under mild conditions. Photoredox reactions on semi-
conductor photocatalysts can also lead to such defluorination re-
actions. In particular, zinc sulfide (ZnS) has an advantage in pho-
toreductive defluorination because of its high conduction band
4
(
CB) levels. Recently, Yin et al. have reported the photode-
fluorination of hexafluorobenzene (F6) using ZnS and CdS nano-
5
crystallites, under UV and visible light irradiation, respectively.
Concurrently, one of us (A. K.) has developed Cu-doped ZnS
(
(
6
7
Cu-ZnS), Ni-doped ZnS (Ni-ZnS), and Pb, Cl-codoped ZnS
8
Pb,Cl-ZnS) photocatalysts in order to extend the photocatalytic
In GC-MS analyses, F5, as well as a considerable amount of
diethylamine, were observed from all of the metal-doped ZnS
suspensions. Furthermore, a trace amount of acetaldehyde was
detected from the head space gas above the solutions in the
vessels for all samples. Diethylamine and acetaldehyde are
produced in the oxidative dealkylation of triethylamine induced
by hole oxidation at the surface of ZnS, as reported in the liter-
response of ZnS toward the visible region (Diffuse reflectance
spectra are given in Electronic Supporting Information). The
new absorption bands that appeared in the 350–540 nm region
have been assigned to transitions from the metal impurity levels
(
Cu 3d, Ni 3d, and Pb 6s) formed in the bandgap to the CB of
ZnS. Although the impurity levels of the three metal-doped
ZnS are different, the CB levels are common and identical to that
of nondoped ZnS. Therefore, the photoreductive defluorination
of F6 is expected to occur upon photoexcitation with visible light.
In this letter, we report the photoreductive defluorination of
F6 using Cu-ZnS, Ni-ZnS, and Pb,Cl-ZnS in triethylamine (a
reducing reagent), dissolved in acetonitrile, under visible light ir-
radiation. Products have been identified by gas chromatography-
mass spectrometry (GC-MS). In order to clarify the hydrogen
source for the pentafluorobenzene (F5) produced by the defluori-
nation of F6, deuterated reagents such as triethyl-d15-amine
100
80
60
F
H
F
F
F
F
F
F
F
F
4
2
0
0
0
F
F
F6
F5
(
N(C2D5)3, 99.7 atom % D) and D2O (>99:75 atom % D) were
added to the solutions and reaction mixtures were analyzed by
GC-MS.
0
1
2
3
4
5
6
7
Irradiation time / day
Ni (0.1 atom %)-ZnS and Pb (1.4 atom %), Cl (2.5 atom %)-
ZnS photocatalysts were prepared by the coprecipitation
Figure 1. Degradation of F6 (filled symbols) and formation of F5
(open symbols) with Cu-ZnS (triangles), Pb,Cl-ZnS (circles), and
Ni-ZnS (squares) under visible light irradiation (>400 nm, 1.6 W
7
,8
method and then heat-treated as reported previously. Cu (4.3
atom %)-ZnS was also synthesized by the coprecipitation
ꢁ2
cm ) at 280 K. The solid decay curve is a simulated one, with a
ꢁ
6
1
method. The Cu-ZnS was used immediately after preparation
without drying, because this condition was favorable to the
first-order rate constant of 0.1 day . The cross plots indicate the
decrease in F6 with nondoped ZnS.
Copyright Ó 2005 The Chemical Society of Japan

Chemistry Letters Vol.34, No.7 (2005)
1057
ature.5,9 In the mechanism, water contained in solution takes part
in the dealkylation of triethylamine.
a relatively small contribution to the defluorination of F6. On
the contrary, for Pb,Cl-ZnS, a considerable amount of the hydro-
gen of F5 was derived from not only triethylamine but also water
contained in acetonitrile. This result is supported by the fact that
the ratio of C6F5H was diminished in anhydrous acetonitrile con-
taining deuterated N(C2D5)3 (solution (c)). Thus, a small amount
of water in acetonitrile contributes greatly to the defluorination
of F6 for Pb,Cl-ZnS, compared (b) to (c). This is not due to
A trace amount of 1,2,4,5-tetrafluorobenzene (1,2,4,5-F4)
was also detected. However, further defluorination of 1,2,4,5-
F4 was not observed in our study. Yin et al. have reported that
nanocrystalline ZnS has the ability to induce photoreductive de-
fluorination of F5, F4, and trifluorobenzenes to afford several F4
isomers, trifluorobenzene, and difluorobenzene. This discrepan-
cy may be caused by the difference in the CB levels between
metal-doped ZnS and nanocrystalline ZnS. The CB edges of
the three metal-doped ZnS would be located at the same level
as that of the ZnS single crystal (ca.ꢁ2:4 V vs SCE in acetoni-
ꢁ
the presence of codoped Cl ion in Pb,Cl-ZnS powders, because
ꢁ
Pb-doped ZnS (not including Cl ) also gave a similar molar ratio
in solution (a), although the defluorination rate was slower than
that for Pb,Cl-codoped ZnS. The rate reduction may arise from
enhancement of electron–hole recombination at defect sites,
which would be formed by the doping of large Pb cations in
4
ꢁ
10
trile). Therefore, F6 (F6/F6 :ꢁ2:11 V vs SCE in DMF) is
easily reduced by CB electrons generated in the metal-doped
ꢁ
10
ꢁ
ZnS, while F5 (F5/F5 :ꢁ2:35 V vs SCE) and 1,2,4,5-F4
the ZnS lattice. The codoped Cl anions may be useful for the
relaxation of the distortion induced by the Pb doping, as reported
ꢁ
10
(
F4/F4 :ꢁ2:40 V vs SCE) may be reduced only to a slight
8
extent. On the other hand, the CB level of nanocrystalline ZnS
is located at a more negative potential than that of bulk ZnS be-
previously. For Cu-ZnS, the molar ratio of F5 in solution (b)
shows a similar tendency to that of Pb,Cl-ZnS, while the result
in solution (a) seemingly resemble that of Ni-ZnS or nondoped
ZnS. However, the small C6F5D ratio in solution (a) is probably
underestimated because a relatively large amount of normal H2O
from the wet Cu-ZnS powders can contribute to the production
of C6F5H.
5
cause of the quantum size effect.
There are two candidates for the hydrogen source of F5: one
is triethylamine, and the other is water contained in acetonitrile.
Here, we prepared the following complementary solutions and
examined the molar ratios of C6F5H and C6F5D by use of GC-
MS analysis: (a) a normal N(C2H5)3 and a small amount of
In conclusion, the mechanism of photoreductive defluorina-
tion of F6 on Pb,Cl-ZnS and Cu-ZnS is different from that of
nondoped ZnS. Pb and Cu cations on the surface of metal-doped
ZnS should act as redox reaction sites, where electrons and holes
generated upon visible light irradiation may be concentrated and
react with F6, triethylamine and/or water. Thus, the role of Pb
and Cu in ZnS is not only to extend the ZnS photocatalytic re-
sponse toward visible light but also to act as the redox reaction
sites in the photoreductive defluorination. On the other hand, the
result for Ni-ZnS in Figure 2 is very similar to that of nondoped
ZnS. This means that Ni cations on the surface of ZnS would not
be the main redox reaction sites because of the very small
amounts of Ni doping (0.1%). The defluorination reaction on
Ni-ZnS would therefore occur on the pure ZnS surface, as in
nondoped ZnS.
ꢁ3
D2O (0.46 mol dm ) contained in anhydrous acetonitrile
H2O < 30 ppm), and (b) deuterated N(C2D5)3 contained in
(
HPLC grade acetonitrile (H2O < 500 ppm). Molar ratios were
obtained from peak areas observed at specific charges (m=z) of
168 and 169, corresponding to the molecular ions of C6F5H
and C6F5D, respectively (Representative data for Pb,Cl-ZnS
are given in Electronic Supporting Information).
Figure 2 shows molar ratios after a 1-day irradiation. The
indicated ratios remained nearly unchanged during several days
for all samples. In the case of nondoped ZnS (UV irradiation)
and Ni-ZnS, the hydrogen of F5 came predominantly from
triethylamine. Therefore, water contained in acetonitrile made
a
d
non-doped ZnS
b
The present work was supported by a Grant-in Aid for
Scientific Research on Priority Areas, ‘‘Photo-functional
Interfaces’’ (Area 417, No. 17029023), from MEXT of Japan.
a
e
Ni (0.1%)-ZnS
b
a
e
Cu (4.3%)-ZnS
References
1
b
a) J. Burdeniuc and R. H. Crabtree, J. Am. Chem. Soc., 118, 2525 (1996).
b) J. Burdeniuc, P. E. M. Siegbahn, and R. H. Crabtree, New J. Chem., 22,
503 (1998).
N. A. Kaprinidis and N. J. Turro, Tetrahedron Lett., 37, 2373 (1996).
a) H. Hori, Y. Takano, K. Koike, K. Takeuchi, and H. Einaga, Environ. Sci.
Technol., 37, 418 (2003). b) H. Hori, Y. Takano, K. Koike, S. Kutsuna, H.
Einaga, and T. Ibusuki, Appl. Catal., B, 46, 333 (2003). c) H. Hori, E.
Hayakawa, K. Koike, H. Einaga, and T. Ibusuki, J. Mol. Catal. A: Chem.,
a
b
c
a
C F H
6 5
C F D
6
5
2
3
Pb(1.4%),Cl(2.5%)-
e
ZnS
e
Pb(1.4%)-ZnS
2
11, 35 (2004). d) H. Hori, E. Hayakawa, H. Einaga, S. Kutsuna, K. Koike, T.
0
%
25%
50%
75%
100%
Ibusuki, H. Kitagawa, and R. Arakawa, Environ. Sci. Technol., 38, 6118 (2004).
H. Kisch, ‘‘Semiconductor Photocatalysis for Organic Synthesis,’’ in ‘‘Advances
in Photochemistry,’’ ed. by D. C. Neckers, G. Bunau, and W. S. Jenks,
Wiley and Sons, New York (2001), Vol. 26, pp 93–143.
H. Yin, Y. Wada, T. Kitamura, and S. Yanagida, Environ. Sci. Technol., 35, 227
(2001).
A. Kudo and M. Sekizawa, Catal. Lett., 58, 241 (1999).
A. Kudo and M. Sekizawa, Chem. Commun., 2000, 1371.
I. Tsuji and A. Kudo, J. Photochem. Photobiol., A, 156, 249 (2003).
M. Kanemoto, H. Hosokawa, Y. Wada, K. Murakoshi, S. Yanagida, T. Sakata,
H. Mori, M. Ishikawa, and H. Kobayashi, J. Chem. Soc., Faraday Trans., 92,
2401 (1996).
4
5
Figure 2. Molar ratios of C6F5H and C6F5D in nondoped and metal-
doped ZnS suspensions after a 1-day irradiation (280 K). All indicated
1
3
values were corrected by considering natural abundance of C (1.1%)
ꢁ3
a
in C6F5H or C6F5D. Symbols: normal N(C2H5)3 (1 mol dm ) and
D2O (0.46 mol dm ) contained in 25 cm3 of anhydrous acetonitrile
ꢁ
b
3
6
7
8
9
ꢁ3
(
1
H2O < 30 ppm); deuterated N(C2D5)3 (1 mol dm ) contained in
0 cm of HPLC grade acetonitrile (H2O < 500 ppm); deuterated
N(C2D5)3 (1 mol dm ) contained in 10 cm of anhydrous acetonitrile
H2O < 30 ppm); dUV and visible light irradiation (>300 nm,
3
c
ꢁ
3
3
(
1
ꢁ2
e
ꢁ2
.9 W cm ); visible light irradiation (>400 nm, 1.6 W cm ).
10 R. O. Loutfy and R. O. Loutfy, Can. J. Chem., 54, 1454 (1976).
Published on the web (Advance View) June 25, 2005; DOI 10.1246/cl.2005.1056