Titanosilicate molecular sieve for size-screening photocatalytic conversion

Article abstract of DOI:10.1021/ja042707h

Titanosilicate molecular sieves, when activated by ultraviolet light irradiation in water in the presence of molecular oxygen, catalyze a conversion of molecules having a size close to the pore of the catalysts but are inactive for molecules having much larger or smaller size. This unprecedented size-screening photocatalytic activity is triggered by a combination of H₂O-induced shortened lifetime of active species (charge-transfer excited state of tetrahedrally coordinated titanium oxide) and restricted diffusion of a molecule inside the pore. This catalytic property demonstrates a potential utility of the catalyst for selective transformation of molecules that is associated with a size reduction of molecules, so-labeled molecular shave transformation. Copyright

Full text of DOI:10.1021/ja042707h

Published on Web 05/17/2005
Titanosilicate Molecular Sieve for Size-Screening Photocatalytic Conversion
Yasuhiro Shiraishi,* Naoya Saito, and Takayuki Hirai
Research Center for Solar Energy Chemistry, and DiVision of Chemical Engineering, Graduate School of
Engineering Science, Osaka UniVersity, Toyonaka 560-8531, Japan
Received December 3, 2004; E-mail: shiraish@cheng.es.osaka-u.ac.jp
Development of efficient and highly selective methods for
photochemical organic synthesis is one of the biggest challenges
1
in chemistry. Photocatalytic version of organic synthesis on a
heterogeneous catalyst is particularly attractive for practical ap-
plications because of the ease of catalyst recovery. Much investiga-
tion has so far been carried out on semiconductor (bulk) TiO
this lacks sufficient product selectivity. Practical use of bulk TiO
2
, but
2
2
is limited mainly to a photocatalytic water treatment (decomposition
3
2 2
of hazardous chemicals into CO and H O).
Titanosilicate molecular sieve is a silicious zeolite with titanium
oxide [Ti-O ; x (coordination number of Ti atom with oxygen
x
atom) ) 4-6)] species covalently linked into a porous SiO
2
framework.4 This material has already been identified as an
5
oxidation catalyst for organic substrates and as an adsorbent for
6
gaseous ones. Application of the material as a photocatalyst has
7
Figure 1. Maximum length, molecular width, and minimum width of (A)
,3,5-trihydroxybenzene (20; EMW, 0.6132 nm) and (B) 1,2,4-trihydroxy-
benzene (10; 0.5762 nm). (C) Schematic representation of photoexcitation-
deactivation process of the isolated Ti-O4 species.
recently received much attention. In this case, substrate reactivity
1
and product selectivity depend strongly on the nature of the Ti-
O
x
species. On the catalyst with low Ti content, the Ti-O
exist in a tetrahedral coordination (Ti-O species) and are highly
dispersed in the SiO matrixes. These “isolated” Ti-O species,
therefore, lack bulk semiconducting character and display photo-
catalytic activities quite different from those of bulk TiO , which
consists of octahedrally coordinated (aggregated) Ti-O species.
On the Ti-O , some substrates are transformed selectively, for
example, NO into N and O , CO with H O into CH and CH
into CH
x
species
4
both containing 0.4 mol % Ti () Ti/(Ti+Si) ratio).14 Both catalysts
exhibit a narrow absorption band at λ < 290 nm,14 which is assigned
to the LMCT (ligand-to-metal charge transfer) band from lattice
2
4
2
2-
4+ 15
oxygen (O
consist of predominantly the isolated Ti-O
Measuring catalytic conversions of 25 kinds of phenol derivatives
1-25),14 in water with O
under photoirradiation (λ > 280 nm;
.5 h), revealed that the activity of TS-1 and TS-2 is rather lower
L
) to the titanium ion (Ti ), suggesting that both
6
4
species.
4
2
2
2
2
4
3
-
4
,
(
0
2
OH, CO with CH
and propene with O
3
OH into methyl formate, CO
2
with H
2
8
2
into propene oxide. These unique activities
2
than that of bulk TiO (anatase); conversions of 17 kinds of these
of the titanosilicate molecular sieves are, however, inspired simply
by the Ti-O active species; structural advantages of the pore
substrates are <3%, whose values are less than 20% of those
obtained on bulk TiO . However, some substrates, such as 20 (1,3,5-
4
2
system on the catalyst are left unexploited.
Here, we report unprecedented “size-screening” photocatalytic
activity of titanosilicate molecular sieves with the Ti-O species,
4
which promotes a selective conversion of molecules having a size
trihydroxybenzene), demonstrate specifically high conversion (>10%)
both on TS-1 and on TS-2, implying that some factor affects the
catalytic activity. Considering the large internal surface area of the
14
catalysts, substrate conversion likely occurs inside the pore. We
close to the pore of the catalyst. This catalytic property is triggered
therefore hypothesized that substrate size may affect the catalytic
activity. To evaluate whether a substrate can enter the pore or not
and how it is oriented inside the pore, the substrate size must be
determined. Molecules inside the pore align their long axis parallel
by a combination of the Ti-O
on the catalysts, when photoactivated in water with molecular
oxygen (O ). Size-selective conversion of molecules is promoted
by several catalytic systems based on zeolites (including photo-
4
active species and the pore system
2
5,9
16
17
to the pore wall and rotate about the axis. By MO calculations,
10
catalytic systems ), which allow a preferential conversion of “either
small or large molecules”. The present photocatalytic system
promotes a selective conversion of “appropriate” size molecules,
while being inactive for too small or too large molecules. Some
maximum length of substrate parallel to the benzene plane was
determined, and its perpendicular length was defined as molecular
width (Figure 1A). For asymmetric substrates (Figure 1B), whose
minimum width is smaller than the molecular width, we termed an
average value of the molecular width and the minimum width as
11
catalytic systems show such selectivity, while none of the
photocatalytic system does such. We highlight here a successful
application of this unusual activity to a selective transformation of
molecules that is associated with a size reduction of molecules,
so-labeled “molecular shave” transformation.
18
effective molecular width (EMW). Relation between EMW and
conversion of substrates on TS-1 and TS-2 is summarized in Figure
2A and B (open symbol). On both catalysts, conversion of substrates
of <0.55 nm EMW is nearly zero, but obviously higher for
substrates of 0.6-0.65 nm EMW, which is 10-18 and 12-21%
larger than the average pore diameter of TS-1 (0.55 nm) and TS-2
(0.535 nm), respectively. However, substrates of >0.68 nm EMW
We used two representative microporous titanosilicate, such as
titanium silicalite-1 [TS-1; MFI structure; pore dimensions, 0.54
×
0.56 nm (sinusoid channel) and 0.52 × 0.58 nm (straight
1
2
1
3
channel)] and titanium silicalite-2 (TS-2; MEL; 0.53 × 0.54 nm),
show almost zero conversion on both catalysts. On bulk TiO
2
8304
9
J. AM. CHEM. SOC. 2005, 127, 8304-8306
10.1021/ja042707h CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
Figure 2. (Open symbol) Relation between EMW and photoconversion
(
0.5 h) of substrates (1-25) on (A) TS-1, (B) TS-2, (C) bulk TiO2, and
(
D) Ti-SiO2. Substrates are (in the EMW order): 1, hydroquinone; 2, benzyl
alcohol; 3, phenol; 4, 4-chlorophenol; 5, p-cresol; 6, 4-chlororesorcinol; 7,
-chlorophenol; 8, 3-chlorocatechol; 9, resorcinol; 10, 1,2,4-trihydroxy-
2
benzene; 11, 3-chlorophenol; 12, 1,2,3-trihydroxybenzene; 13, 2,5-dichlo-
rophenol; 14, 4-chlorocatechol; 15, catechol; 16, m-cresol; 17, 2,6-
dichlorophenol; 18, 3,5-dichlorophenol; 19, 2,4-dichlorophenol; 20, 1,3,5-
trihydroxybenzene; 21, 2-chlorohydroquinone; 22, 5-chlororesorcinol; 23,
3
,4-dichlorophenol; 24, 2,6-bis(hydroxymethyl)-p-cresol; 25, 2,4,6-trichlo-
3+
rophenol. (Red symbol) Relation between EMW and decrease in the Ti
intensity on (A) TS-1 and (D) Ti-SiO2 in the presence of substrates, which
3
+
were obtained by double integration of the Ti signal on ESR spectra.
3
+
Figure 3. ESR spectra of (A) TS-1 and (B) Ti-SiO2 obtained at 77 K
with photoirradiation. The spectra were measured (i) in vacuo, (ii) with
H2O (56 mmol/g of catalyst), (iii) with H2O (56 mmol/g of catalyst)
containing 3 (0.1 mmol/g of catalyst), and (iv) with H2O (56 mmol/g of
catalyst) containing 20 (0.1 mmol/g of catalyst).
The decrease in the Ti intensity obtained with only H2O is set as 0%.
Substrates used are 3, 5, 7, 10, 20, 21, and 23. The dotted lines denote
average pore diameter of the catalysts.
(Figure 2C), such EMW-dependent profile is not observed. These
impressive profiles for TS-1 and TS-2 imply a potential size-
screening activity of the catalysts promoting selective conversion
of appropriate size molecules.
H
2
O (Figure 3ii). Addition of H
2
O containing 20, of 11% larger
sample also did
EMW than that of the TS-1 pore, to the Ti-SiO
not show any signal change (Figure 3iv, B), but the addition to the
TS-1 sample led to apparent signal decrease (Figure 3iv, A).
However, addition of H
that of the TS-1 pore, still did not show any signal change for both
catalysts. The Ti3 signal was doubly integrated to obtain Ti
intensity, and the decrease in the intensity in the presence of
substrates was plotted against EMW (Figure 2; red symbol), where
the decrease in the presence of only H
the Ti3 intensity profile on TS-1 (Figure 2A) agrees well with
that for the substrate conversion (open symbol). For Ti-SiO
(Figure 2D), almost no decrease in the Ti intensity is observed
in the presence of any substrates, whose profile still agrees with
the conversion profile. These indicate that the substrate conversion
is closely related to the degree of the interaction between [Ti
O
The above findings led us to assume the onset mechanism for
the size-screening activity of TS-1 and TS-2, as follows. The excited
state [Ti3 -O
by H O, such that their lifetime is shortened significantly. A slim
molecule, of rather smaller EMW than that of the pore, diffuses
smoothly inside the pore and is scarcely trapped by the short-lived
[Ti -O
close to that of the pore is restricted by the pore wall, such that the
“interlocked” molecule is trapped easily by [Ti
in high conversion. Zeolite framework is distorted in solution, and
the pore structure changes elastically. The molecules of ca. 10-
20% larger EMW than that of the pore may, therefore, be
interlocked more easily and allowed for the specifically high
conversion (Figure 2A and B). However, a fat molecule of >20%
larger EMW cannot enter the pore, thus showing zero conversion.
To confirm the above, when reaction was carried out in acetonitrile,
2
On heterogeneous photocatalyst, substrate adsorption onto the
catalyst surface is an important factor for the catalytic activity.19
However, any relation between the amount of the substrates
adsorbed and the substrate conversion is not observed for both TS-1
2
O containing 24, of 25% larger EMW than
+
3+
14
and TS-2. Oxidation potential, HOMO level, and hydrophobicity
log P) of substrates also do not affect the catalytic activity.14 The
size-screening catalytic activity of TS-1 and TS-2 may be affected
by the isolated Ti-O species and/or the pore system on the
catalysts. To check the former effect, when two catalysts, such as
TiO -impregnated on a Ti-free silicalite-1 (TiO
/S-1; 5 mol % Ti)20
and TS-1 of high Ti content (TS-1high; 5 mol % Ti) both containing
Ti-O species, were employed, no EMW dependence was ob-
served.14 When the latter effect was checked with a nonporous silica
with Ti-O species, prepared by a sol-gel method (Ti-SiO ; Ti
.4 mol %),8 almost zero conversion was obtained for all of the
(
2
O is set as 0%. Surprisingly,
+
4
2
3+
2
2
6
3+
-
-
4
2
L
]* and substrates.
e,21
0
substrates (Figure 2D). These indicate that the size-screening activity
of TS-1 and TS-2 is triggered by a combination of the isolated
+
-
L
]* species on the catalysts are deactivated strongly
2
2
Ti-O
4
species and the pore of the catalysts.
2
In this photocatalytic system, the Ti-O
4
species [Ti4+-O
L
2-]
on the catalysts are photoexcited to form charge-transfer excited
state [Ti3 -O
+
-
]* (Figure 1C), which acts as the active species.
21
3+
-
L
L
]*. On the contrary, diffusion of a molecule of a size
UV irradiation of TS-1 and Ti-SiO
2
catalysts in vacuo at 77 K
gave rise to ESR signals assigned to Ti3 and O
+
-
(Figure 3i, A
O (56 mmol/g of catalyst) to both
+
-
L
3
L
-O ]*, resulting
and B). Upon addition of H
2
samples, the Ti3 signals were decreased (Figure 3ii) because of
+
24
the deactivation of [Ti3 -O
O addition (0.5 mol/g of catalyst) led to a complete disappearance
of the signal (not shown) on both catalysts. The present photore-
action system employs enormous quantity of H O (56 mol/g of
]* is deactivated strongly. Upon
O (56 mmol/g of catalyst) containing 3 (0.1 mmol/g
+
-
]* by H
22,23
Further
L
2
O (Figure 1C).
H
2
2
catalyst), indicating that [Ti3 -O
+
-
L
3+
-
25
addition of H
2
L
which scarcely deactivates [Ti -O ]*, no EMW dependence
of catalyst) of rather smaller EMW than that of the TS-1 pore
was observed.14 However, the dependence became apparent with
Figure 3iii), further decrease in the Ti3 signal was scarcely
+
an increase in the H
of [Ti3 -O
O quantity. This indicates that the deactivation
14
(
2
+
-]* by H
observed for both catalysts, as compared to that obtained with only
L
2
O is essential for the size-screening activity,
J. AM. CHEM. SOC.
9
VOL. 127, NO. 23, 2005 8305

C O M M U N I C A T I O N S
Table 1. Photocatalytic Transformation of 1,2,4-Trisubstituted
Chlorophenols Promoted on Various Catalysts
Acknowledgment. This work is supported by the Grant-in-Aids
for Scientific Research (No. 15360430) and on Priority Areas
a
“
Fundamental Science and Technology of Photofunctional Inter-
faces (417)” (No. 15033244) from the Ministry of Education,
Culture, Sports, Science and Technology, Japan (MEXT).
Supporting Information Available: Materials and methods, dis-
cussion, tables, and figures. This material is available free of charge
via the Internet at http://pubs.acs.org.
References
(
1) Coyle, J. D.; Carless, H. A. C. Photochemistry in Organic Synthesis; Royal
Society of Chemistry: London, 1986.
(
2) Fox, M. A.; Dulay, M. T. Chem. ReV. 1993, 93, 341-357.
3) Fujishima, A.; Rao, T. N.; Tryk, D. A. J. Photochem. Photobiol., C:
Photochem. ReV. 2000, 1, 1-21.
(
(
4) Centi, G.; Trifiro, F. New DeVelopments in SelectiVe Oxidation; Elsevier:
Amsterdam, 1990.
a
Conditions: substrates, 20 µmol; catalyst, 10 mg; photoirradiation time,
(5) Arends, I. W. C. E.; Sheldon, R. A.; Wallau, M.; Schuchardt, U. Angew.
Chem., Int. Ed. Engl. 1997, 36, 1144-1163.
2
h; H2O, 10 mL; temperature, 313 K. b EMW: 10 (0.5762 nm); 19 (0.6051
nm); 21 (0.6149 nm); 23 (0.6431 nm).
(6) Kuznicki, S. M.; Bell, V. A.; Nair, S.; Hillhouse, H. W.; Jacubinas, R.
M.; Braunbarth, C. M.; Toby, B. H.; Tsapatsis, M. Nature 2001, 412,
720-724.
and the activity is triggered by a combination of the short-lived
(7) Anpo, M. Photofunctional Zeolites; Nova Science Publishers Inc.: New
3
+
-
York, 2000.
[
Ti -O
L
]* and the restricted diffusion of a molecule inside the
(Figure 2D) is, therefore,
(
8) (a) Zhang, S. G.; Ichihashi, Y.; Yamashita, H.; Tatsumi, T.; Anpo, M.
Chem. Lett. 1996, 895-896. (b) Yamashita, H.; Fujii, Y.; Ichihashi, Y.;
Zhang, S. G.; Ikeue, K.; Park, D. R.; Koyano, K.; Tatsumi, T.; Anpo, M.
Catal. Today 1998, 45, 221-227. (c) Yeom, Y. H.; Frei, H. J. Phys. Chem.
A 2001, 105, 5334-5339. (d) Yamagata, S.; Nishijo, M.; Murao, N.; Ohta,
S.; Mizoguchi, I. Zeolites 1995, 15, 490-493. (e) Yoshida, H.; Murata,
C.; Hattori, T. Chem. Commun. 1999, 1551-1152.
pore. Zero conversion on Ti-SiO
attributed to the lack of the pore that regulates the motion of
substrates.26
The size-screening catalytic activity, inspired by titanosilicate
molecular sieves, is highlighted by their application to a selective
transformation of 1,2,4-trisubstituted harmful chlorophenols (Table
). Photoirradiation of bulk TiO with 21 (run 1) afforded the
2
corresponding trihydroxylated benzene derivative 10, which is
nontoxic and industrially valuable, as an initial product via a
hydroxyl radical addition.14 The yield of 10 is, however, signifi-
cantly low (1%) because the sequential decomposition of 10 occurs
on TiO
than the TS-1 and TS-2 pores, respectively, thus allowing effective
photoconversion of 21 on both catalysts (>65%) (runs 2 and 3).
On these catalysts, the substitution of -Cl by -OH also occurs,
via reaction of 21 trapped on the O
formed by a reduction of O
2
(9) (a) Corbin, D. R.; Herron, N. J. Mol. Catal. 1994, 86, 343-369. (b)
Tatsumi, T.; Nakamura, M.; Negishi, S.; Tominaga, H. Chem. Commun.
1990, 476-477.
1
(
10) (a) Ramamurthy, V.; Eaton, D. F.; Casper, J. V. Acc. Chem. Res. 1992,
2
5, 299-307. (b) Ramamurthy, V.; Shailaja, J.; Kaanumalle, L. S.; Sunoj,
R. B.; Chandrasekhar, J. Chem. Commun. 2003, 1987-1999. (c) Corma,
A.; Garcia, H. Chem. Commun. 2004, 1443-1459.
(11) Degnan, T. F., Jr. J. Catal. 2003, 216, 32-46.
(
(
12) Clerici, M. G.; Bellussi, G.; Romano, U. J. Catal. 1991, 129, 159-167.
13) Reddy, J. S.; Kumar, R. J. Catal. 1991, 130, 440-446.
14,27
2
.
The EMW of 21 (0.6149 nm) is 12 and 15% larger
(14) See Supporting Information.
(
15) Bordiga, S.; Coluccia, S.; Lamberti, C.; Marchese, L. Zecchina, A.;
Boscherini, F.; Buffa, F.; Genoni, F.; Leofanti, G.; Petrini, G.; Vlaic, G.
J. Phys. Chem. 1994, 98, 4125-4132.
(
(
16) Cox, S. D.; Gier, T. E.; Stucky, G. D. Chem. Mater. 1990, 2, 609-619.
17) Calculations were performed with the MNDO-PM3 method within the
WinMOPAC version 3.0 software (Fujitsu Inc.). Molecular thickness of
the substrates is <0.3 nm, which is rather smaller than the pore diameter
of the catalysts, such that the thickness is not considered in this study.
18) J a¨ ger, R. Chem.sEur. J. 2004, 10, 247-256.
-
L
site with a superoxide anion
on the Ti3+ site. Astoundingly,
14
2
excellent high selectivity of 10 was obtained on TS-1 (85%) and
TS-2 (> 99%). Since the -OH is less bulky than -Cl, the EMW
of 10 (0.5762 nm) is 6% smaller than that of 21 and is only 5 and
(
(19) Anderson, C.; Bard, A. J. J. Phys. Chem. B 1997, 101, 2611-2616.
(
20) Hsien, Y.-H.; Chang, C.-F.; Chen, Y.-H.; Cheng, S. Appl. Catal. B 2001,
1, 241-249.
(21) Murata, C.; Yoshida, H.; Kumagai, J.; Hattori, T. J. Phys. Chem. B 2003,
3
8
% larger than the TS-1 and TS-2 pores, respectively, thus allowing
107, 4364-4373.
relatively smooth diffusion of 10 inside the pores. Sequential
reaction of 10 on the catalysts, therefore, scarcely occurs, thus
affording 10 in high selectivity. These catalysts also promoted
selective transformation of dichlorophenols (19 and 23) into 10 in
high selectivity (>80%) (runs 8, 9, 11, and 12), where in both cases
intermediately formed 21 is sequentially transformed into 10. The
above findings demonstrate that the titanosilicate molecular sieves
shave” a molecule, interlocked inside the pore, into a smoothly
(
22) Ikeue, K.; Yamashita, H.; Anpo, M.; Takewaki, T. J. Phys. Chem. B 2001,
105, 8350-8355.
3+
(
23) Double integration of the Ti signal on ESR spectra (Figure 3ii) indicates
3
+
that decrease in the Ti intensity in the presence of H
2
O (56 mmol/g of
catalyst) is 12% (TS-1) and 82% (Ti-SiO ), based on the intensity
2
2
obtained in vacuo (Figure 3i). Surface hydrophilicity of Ti-SiO is
significantly higher than that of TS-1 (Supporting Information Table S1;
14
3+
monolayer adsorption capacity of H O). The strong Ti deactivation on
2
3+
-
L
2 2
Ti-SiO is, therefore, due to the enhanced access of H O to [Ti -O ]*.
(
(
(
24) van Koningsveld, H.; Jansen, J. C. Microporous Mater. 1996, 6, 159-
167. The straight pore dimension of silicalite-1 (0.52 × 0.58 nm) changes
to 0.46 × 0.64 nm (9.4% expansion) in the presence of naphthalene.
“
diffusive slim molecule. In general, harmful chlorophenols are
treated (detoxificated) by chemical28 and photochemical
3
+
25) Relative Ti intensity on TS-1, obtained by double integration of ESR
spectra, is 1 (in vacuo), 0.99 [with acetonitrile (56 mmol/g of catalyst)],
3,27
total
2
and 0.88 [with H O (56 mmol/g of catalyst)].
destruction processes. The present molecular shave catalytic system
enables the detoxification of chlorophenols and the synthesis of
valuable phenol derivatives all at once, proving that this “green”
photocatalyst may contribute to the development of an economically
and environmentally friendly chemical process.
In conclusion, we have found the size-screening photocatalytic
activity of titanosilicate molecular sieves. Various titanosilicate
molecular sieves, of different pore sizes and structures, have so far
been synthesized using a structure directing agent. The catalytic
system based on titanosilicate molecular sieves has the potential
to become a very powerful tool for photocatalytic organic synthesis.
26) A question arising is why the amount of substrates adsorbed on the catalyst
surface does not affect the substrate conversion on TS-1 and TS-2
(Supporting Information Table S2), although the substrate adsorption may
occur mainly inside the pore of high surface area and the strongly adsorbed
3+
-
substrate is likely to be trapped easily by [Ti -O ]*. The substrate
L
adsorption may occur mainly on the silicious surface of the catalysts
3
+
-
L
because of very low Ti content. The substrate trapping by [Ti -O ]*
is, therefore, not related to the degree of the substrate adsorption.
(27) D’Oliveira, J.-C.; Minero, C.; Pelizzetti, E.; Pichat, P. J. Photochem.
Photobiol., A: Chem. 1993, 72, 261-267.
(
28) Gupta, S. S.; Stadler, M.; Noser, C. A.; Ghosh, A.; Steinhoff, B.; Lenoir,
D.; Horwitz, C. P.; Schramn, K.-W.; Collins, T. J. Science 2002, 296,
326-328.
JA042707H
8306 J. AM. CHEM. SOC.
9
VOL. 127, NO. 23, 2005