-
Chemical Engineering Journal 179 (2012) 285 294
Contents lists available at SciVerse ScienceDirect
Chemical Engineering Journal
j ourna l ho mepage: www.elsev ier .c
Direct f poxygen
Van-Huy unla Department ob Institute of En
a r t i c l
Article history:Received 10 AReceived in reAccepted 2 No
Keywords:EpoxidationPropylene oxidPhotocatalysisMolecular
oxyMCM-41TiO2
opylee envd at iO2 aic epopress
and malyticwith
ssuretreamg the
2011 Elsevier B.V. All rights reserved.
1. Introduction
As the dicals demanspecicallytant in theas key inteinstant,
prointermediatindustry sinwas approxin 2009. Altby 5.6% in 2annual
growrecovers [4]
Indeed, studies per
CorresponE-mail add
epoxidation by using molecular oxygen under high temperature
1385-8947/$ doi:10.1016/j.evelopment of third world countries
continues, chem-d worldwide reaches an unprecedented level.
More
, the epoxidation of hydrocarbons becomes more impor- industrial
production of chemicals, which are usedrmediates for polymer and
organic syntheses. Forpylene oxide (PO), the second most important
chemicale, has become increasingly important to the chemicalce
1950s [13]. In 1991, the global PO consumptionimately 3.9 million
tons and reached to 6.0 million tonshough the global PO consumption
has been decreased009 due to economic crisis, it is predicted that
averageth of over 7% will be seen from 2009 to 2014 as demand.
there has been great experimental and theoreticalformed recently
focusing on gas-phase propylene
ding author. Tel.: +886 2 23631994; fax: +886 2 23623040.ress:
[email protected] (J.C.S. Wu).
and/or high pressure as shown in Eq. (1).
CH3 CH CH2 + O2 CH 3 CH CH 2
Othermocatalyst
high temp., high pressure12
(1)
In 2004, Jin et al. reported that PO selectivity of 60.3%
wasachieved over the 20%Ag4%MoO3/ZrO2 catalyst under 400 C,0.1 MPa
and GHSV = 7500 h1 [5]. In 2008, Suo et al. obtainedPO selectivity
of 17.9% with 0.9% propylene conversion at theinitial 10 min of
reaction over Au/SiO2 catalyst at 325 C, andGHSV = 6000 h1 [6]. The
other approaches on exploring the selec-tive catalytic epoxidation
of propylene via the mixture of hydrogenand oxygen are shown in Eq.
(2).
CH3 CH CH + H2 2 + O2 CH3 CH CH + H 2 2O
Othermocatalyst
high temp., high pressure (2)
In 2005, Cumaranatunge and Delgass enhanced Au content inAu/TS-1
and achieved 10% conversion, 76% selectivity by using agas mixture
at 473 K [7]. In 2007, the study from Oyama and co-workers achieved
a conversion of 1.4%, a PO selectivity of 99%,and a H2 efciency of
17% under 423 K and 0.1 MPa by usinggold supported on a
Ti-containing silicate mesoporous material[8]. The approaches
developed for the thermocatalysts by using
see front matter 2011 Elsevier B.V. All rights
reserved.cej.2011.11.003gas-phase photocatalytic epoxidation o by
photocatalysts
Nguyena, Hsiang-Yu Chana, Jeffrey C.S. Wua,, Hsf Chemical
Engineering, National Taiwan University, Taipei 10617,
Taiwanvironmental Engineering, National Chiao Tung University, Hsin
Chu 300, Taiwan
e i n f o
ugust 2011vised form 28 October 2011vember 2011
e
gen
a b s t r a c t
The photocatalytic epoxidation of prduce propylene oxide (PO)
from thachieve high PO selectivity and yielTiO2, V-Ti/MCM-41,
V2O5/SiO2, Au/Tphoto-epoxidation. The photocatalytC3H6:O2:N2 =
1:1:18 at atmospheric hyde, acetone, acetaldehyde, ethanolfavorable
photocatalysts for photocatformation rate of 114 mol g cat1 h1
after 4 h at 50 C and atmospheric pretivity of products was very
stable on sknowledge of species presented durinom/ locate /ce j
ropylene with molecular
ing Baib
ne by molecular oxygen is one of the best methods to
pro-ironmental viewpoints. The key is using photo-energy tomild
conditions. A series of photocatalysts including SiO2,nd TS-1, were
used to evaluate their performance in thexidation of propylene was
carried out in the gas mixture of
ure. In addition to PO, other products, such as
propionalde-ethanol, were detected. The results indicated that the
most
epoxidation were silicates supported group. The highest PO
selectivity of 47% was obtained over V-Ti/MCM-41 on stream
under UVA irradiation of 0.2 mW/cm2. In addition, the selec-. A
possible reaction mechanism was proposed based on the
photocatalytic reaction.
-
286 V.-H. Nguyen et al. / Chemical Engineering Journal 179
(2012) 285 294
hydrogenoxygen mixture or only molecular oxygen are very
cleanand have shown good activity performance. However, the
extraenergy lost in the reaction, low hydrogen efciency and safety
con-sideration that requires H2O2 production facilities to be
restrictedin size have
It is wellsource andneer work engineeringcatalysis, hafew
researclysts and thYoshida et metal oxideTiOx/SiO2 wyield and 1Xe
lamp irrproposed Vwhich has dand a PO sewith UV-C lronmental
require furtidation proprocess (20dation
procliquid-phasepoxidationronmentallconditions still needs t
It is expwith the aidapproach fothe main stdirect gas-pis to
explorvia an idealenergy to acEq. (3).
CH3 CH
The objephotocatalySiO2, TiO2, conditions. for photocayields of
POcatalytic epof species p
2. Experim
2.1. Prepara
Titaniumcial P25 (Dwas carriedHAuCl4 solpH 5.5 by uwas
irradia
stirring. The photocatalyst was ltered and washed before
dryingin an oven at 393 K and calcined in a ow of air at 773 K for
5 h.
Amorphous SiO2 was prepared from tetraethyl orthosilicate(TEOS,
Aldrich, 98%) by using the solgel method. The pH of a mix-
ntairoxim
for 2n a r.niumlizatnd tsenc
Flukaurathe ra5 (0
ed by cheK forere csopod bimetng te.01Tisis ped int of
van titane waf 2 M0.5 e sowas ae wae wan at sultinand te w
arac
lighectiv00). Bectrotrum
on Hiffra
strucred bline photohe chof vaecics, ASdiedotro
andon b2G Crepa made these approaches only a temporary solution.
known that photo-energy is the Earths ultimate energy
friendly to environment. Following the excellent pio-of
Fujishima and Honda in 1972 [9], scientic and
interests in TiO2, particularly in the eld of photo-ve grown
signicantly [1016]. However, there are onlyhes on photo-epoxidation
in the presence of photocata-eir performances are not advanced yet
[1723]. In 2000,al. investigated more than 50 types of
silica-supporteds for their photo-oxidation activities and found
thatas the most effective photocatalyst attaining 4.7% in PO9.2% in
PO selectivity under batch reaction with 200 Wadiation and at 318 5
K [18]. In 2006, Amano et al.2O5 (0.1%)/SiO2 as the best
photocatalyst until now,emonstrated a PO formation rate of 85 mol g
cat1 h1
lectivity of 37% under continuous reaction at 303 K andight of
240 < < 440 nm [19]. From economic and envi-viewpoints, most
of the current studies on epoxidationher improvements. Although
there were two new epox-cesses developed recently, a
cumenehydroperoxide03 in Japan) and a hydrogen peroxide (H2O2)
epoxi-ess (2008 in Belgium), they still based on multi-stagee
reactions and required hydrogen. The photocatalytic
of propylene is a promising process because it is envi-y
friendly, and the process can be carried out under mildin the
absence of hydrogen; however, its performanceo be improved.ected
that direct epoxidation process to produce PO
of light energy over photocatalysts will be a promisingr the
production of chemicals in the future. Therefore,rategy being
considered in this study is to develop ahase one-pass reaction
under mild conditions. Our aime the selective photocatalytic
epoxidation of propylene
oxidant, i.e., molecular oxygen. The key is using photo-hieve
high PO selectivity at mild conditions, as show in
CH2 + O2 CH3 CH CH2
Ophotocatalyst
mild conditions12
(3)
ctive of this study is to evaluate the direct gas-phasetic
epoxidation on a series of photocatalysts such asV-Ti/MCM-41,
V2O5/SiO2, Au/TiO2 and TS-1 at mildThe optimal reaction conditions
were also studiedtalytic epoxidation by comparing the selectivities
and
for different photocatalysts. The mechanism of photo-oxidation
was also proposed based on the knowledgeresent during the
photocatalytic reaction.
ental
tion of photocatalysts
dioxide (TiO2) powder used in this study was commer-egussa). The
photo-deposition of Au particles on TiO2
out by the method suggested in the literature [24].ution and
TiO2 were mixed together and adjusted tosing 0.1 M Na2CO3 before
illumination. The solution
ted with a 200 W mercuryarc lamp for 120 min with
ture coto appstirredcined ipowde
Titacrystal98%) athe prewater,monol[25]. T
V2OreportStreamat 353 ples w
MethesizecetyltrdirectiSiO2:0synthedissolvamounand/orfor
themixtur20 ml opH to 1ring, thwater mixturmixturan ovethe rewater
templa
2.2. Ch
Theby reCary 1ning el230 insformedX-ray
dtallinemeasucrystalX-ray mine tstatus The spmetricwas
stuSynchrK-edgemode DE-20were pning ethanol, TEOS, and deionized
water was adjustedately 2 by HCl (0.1 M) solution and the mixture
was
h. After that, it was dried in an oven at 393 K and cal-ow of
air at 973 K for 5 h to yield the amorphous SiO2
silicalite-1 (TS-1) was prepared by the hydrothermalion of gel
from tetraethylorthosilicate (TEOS, Aldrich,itanium (IV)
tetrabutoxide (TBOT, Alfa Aesar, 98%) ine of tetrapropylammonium
hydroxide (TPAOH, 20% ina) with a small amount of polyoxyethylene
sorbitane (Tween 20), which was reported by Khomane et al.tio of Si
to Ti for TS-1 was 19:1 in this study..5 wt%)/SiO2 was prepared
using the same method as
Amano and Tanaka [26]. First, silicon(IV) oxide (99.5%;mical)
was stirred with an aqueous solution of NH4VO3
2 h, followed by evaporation to dryness. Next, the sam-alcined
in dry air at 773 K for 5 h.rous V-Ti/MCM-41 molecular sieves were
syn-y simple hydrothermal treatment method usinghylammonium bromide
(CTAB) as the structure-mplate. The molar composition of the gel
mixture wasO2:0.01V2O5:0.2CTAB:0.89H2SO4:120H2O. In a
typicalrocedure, 21.2 g of sodium metasilicate monohydrate
100 ml DI water was combined with the appropriate metal
precursors like titanium oxysulfate hydrateadyl sulfate hydrate
(dissolved in 20 ml of 2 M H2SO4)ium and vanadium sources,
respectively. The resultings stirred vigorously for 30 min. Then,
approximately
H2SO4 was added to the above mixture to adjust thewith constant
stirring to form a uniform gel. After stir-lution containing 7.28 g
of CTAB dissolved in 25 ml of DIdded slowly into the above mixture
and the combineds stirred for three additional hours. The resulting
gels transferred into a Teon coated autoclave and kept in145 C for
36 h. After cooling to the room temperature,g solid was recovered
by ltration, washed with DIdried in an oven at 110 C for 8 h.
Finally the organicas removed by calcination at 550 C for 10 h.
terization
t absorption of photocatalysts was fully characterizede
diffusive UVvis spectroscopy (UVvis DRS, VarianaSO4 was used as a
standard reection reference. Scan-n microscopy (SEM) was carried
out on Nova Nano SEMent. Transmission electron microscopy (TEM) was
per-itachi H-7100 electron microscope instrument. Powderction (XRD,
Xray-M03XHF, Ultima IV) was used for crys-ture and calculation of
photocatalyst sizes. All peak datay XRD analysis were checked and
assigned to known
phases by comparing with those of JCPDS database. Theelectron
spectroscopy (XPS) was carried out to deter-emical composition of
the elements and the chemicalrious species on the Thermo Theta
Probe instrument.
surface area of photocatalyst was measured by Micro-AP 2000. The
atomic structure of V-Ti/MCM-41 and TS-1
by X-ray absorption spectroscopy (XAS) at the Nationaln
Radiation Research Center in Hsin Chu, Taiwan. For Ti
V K-edge spectra, they were analyzed in uorescenceeam-line 16A.
The analysis was performed using ARSlosed Cycle Cryostat with Lytle
detector. The samplesred as powder form, and dispersed on Kapton
tape. The
-
V.-H. Nguyen et al. / Chemical Engineering Journal 179 (2012)
285 294 287
Fig. 1. The scheme of powder packed-bed reactor.
energy step used was 0.2 eV, and the counting time per step was2
s around the absorption edge.
2.3. Direct gas-phase photocatalytic epoxidation of
propylene
All photocatalytic epoxidation experiments were
conductedcontinuously with the reactant gas mixture of C3H6:O2:N2 =
1:1:18(controlled by rotameters) at a gas hourly space velocity
(GHSV) of6000 h1 under atmospheric pressure. The photocatalyst
(0.02 g)was spread in a powder packed-bed reactor (0.55 cm3) with
aquartz window on top for transmission of light as shown inFig. 1.
The UV-A (wavelength 320500 nm) from Exfo S1500 (USA)
Fig. 2. XRD paAu/TiO2; V2O5
Fig. 3. Diffuse reectance UVvis spectroscopy of different
photocatalysts (a) SiO2,(b) V-Ti/MCM-41, (c) TS-1, (d) P25, (e)
V2O5(0.5 wt%)/SiO2, and (f) Au(3 wt%)/TiO2.
equipped with 200 W mercuryarc lamp was used as the irradi-ation
source and guided to reactor by an optical ber. The lightintensity
was measured at the window of the reactor by GOLDILUXRadiometertor
was placwas requiregas chromaame ioniz(TCD). A 6-wthe two colto
separatehyde (PA), aand ethanoarate O2 anpropylene cselectivity
o
C3H6 consu
C3H6 conve
=
rattterns of different photocatalysts (a) V-Ti-/MCM-41; TS-1 and
(b) P25;/SiO2; SiO2./Photometer (UV-A Probe/UV-C Probe). The
photoreac-ed on the hot plate in case high-temperature conditiond.
The feed gases were analyzed periodically using atograph (Young
Lin, YL6100 GC) equipped with both aation detector (FID) and a
thermal conductivity detectoray valve was used to switch the ow of
the sample into
umns of the GC. A molecular sieve-5A column was used propylene
and oxygenates, including PO, propionalde-cetone (AC), acetaldehyde
(AA) and alcohol (methanoll). The Porapak-N on the other hand was
used to sep-d CO2. The following (Eqs. (47)) are the denitions
ofonversion, consumption and adsorption rates, and thef products
for the reaction.
mption rate = C3H6feed rate C3H6out rate (4)
rsion rate
e of all products covered to C3 products (5)Fig. 4. SEM image of
titanium silicalite (TS-1).
-
288 V.-H. Nguyen et al. / Chemical Engineering Journal 179
(2012) 285 294
Fig. 5. TEM im
C3H6 adsorp
= C3H6 co
Product sele
= moles o
3. Results
3.1. Photoc
The X-rshown in Fof V-Ti/MCMpore struct24.3 and 4shown in thtive
peaks from monoc[28]. In Fig. sive XRD peP25. Au is acteristic Xand
77.55
structure.ages of photocatalysts: (a) Au(3 wt%)/TiO2; and (b)
V-Ti/MCM-41.
tion rate
nsumption rate C3H6 conversion rate (6)
ctivity
f the formation one C3 productmoles of all C3 products
100% (7)
and discussions
atalyst characterization
ay diffraction patterns of various photocatalysts areig. 2. As
shown in Fig. 2(a), the main diffraction peaks
-41 at 2 = 1.9 and 3.9 conrm the family of meso-ure. The
diffraction peaks of TS-1, at 2 = 23.2, 23.8,5, are consistent with
the MFI topology structures ase inset of Fig. 2(a) [27]. The
presence of single diffrac-at 2 = 24.3 in the XRD pattern indicates
a changelinic symmetry (S-1) to orthorhombic symmetry (TS-1)2(b),
the photocatalysts containing TiO2 showed inten-aks for anatase
phase and rutile phase as expected inhighly dispersed on the TiO2
which exhibits a char-RD pattern with ne structures at 2 = 44.38,
64.56
. No peak is shown for SiO2 due to their amorphous
Fig. 6. XPS spV-Ti/MCM-41
Fig. 3 shoferent phot200380 nminset of Figwidely usedTi4+ in a
tecenters. Moindicates thTS-1 [7]. Foobserved atof anatase ectra
of: (a) Si 2p: TS-1, V-Ti/MCM-41, V2O5/SiO2; (b) Ti 2p: TS-1, and
(c) V 2p: V2O5/SiO2, V-Ti/MCM-41.
ws the UVvis diffusive reectance spectroscopy of dif-ocatalysts.
There is a slight absorption edge at around
for amorphous SiO2 and V-Ti/MCM-41 as shown in the. 3(a)(b). The
spectrum of TS-1 as shown in Fig. 3(c) is
to conrm the electronic transitions between O2 andtrahedral
environment of isolated framework of Ti(IV)reover, the absence of
absorption band at 320400 nmat no anatase or extra framework of Ti
is observed inr TiO2 (P25) shown in Fig. 3(d), the absorption edge
is
around 400 nm, which is between the absorption edge(387 nm) and
rutile (418 nm). For V2O5 (0.5 wt%)/SiO2
-
V.-H. Nguyen et al. / Chemical Engineering Journal 179 (2012)
285 294 289
shown in Ffrom O2 toto be a tetris shown ilight absorpabsorption
surface plas
Fig. 4 shoamount of Fig. 5(a) anTi/MCM-41as shown inMCM-41
ca
Fig. 6(a)104.1 eV. TFig. 7. Summary of the Ti K-edge
characterization of (a) V-Ti/MCM-41 and (b) TS-1 by
ig. 3(e), a ligandmetal-charge-transfer (LMCT) band V5+ emerges
at around 200550 nm, which is assignedahedral VO4 monomer [19]. The
spectrum of Au/TiO2n Fig. 3(f). The Au on TiO2 signicantly affects
thetion as compared with bare TiO2, and shows a secondband in the
visible-light region at 600 nm, revealingmonic resonance [29].ws
the SEM micrograph of TS-1, indicating that a small
TiO2 aggregated on the SiO2 surface in some regions.d (b) shows
the TEM images of Au (3 wt%)/TiO2 and V-, respectively. The size of
Au particles is around 10 nm
Fig. 5(a). With obvious contrast, the nano-structure ofn be
observed in Fig. 5(b).
shows the binding energies of Si 2p at 103.2 eV andhe peak of
TS-1 at 104.1 eV belongs to silica oxide
compound.V2O5/SiO2oxides but hFig. 6(b) shconrmingon V-Ti/MCTi
state of VFig. 6(c) shTi/MCM-41The intensitloading of V
The Ti Kand TS-1 arcan be obta(XANES). Th X-ray near edge
absorption spectroscopy (XANES).
There is no distinguished shift for V-Ti/MCM-41 andat 103.2 eV,
which indicates that silicates are mainlyex-uorosilicate and other
anion are also included [30].ows the binding energies of Ti 2p for
TS-1 at 457.5 eV,
the presence of Ti4+ state. Due to the little amount of TiM-41,
its peak of Ti 2p could not be observed clearly. The-Ti/MCM-41 was
conrmed by XAS as shown below.
ows the binding energies of V 2p3/2 and 2p1/2 for V- and
V2O5/SiO2 at 517.3 eV and 520.9 eV, respectively.ies of peaks were
closed to background due to the small2O5 (0.5 wt%).-edge X-ray
absorption spectroscopy of V-Ti/MCM-41e shown in Fig. 7. Generally,
two kinds of informationined from the X-ray absorption near edge
spectroscopye shift of absorption edge indicates the oxidation
state
-
290 V.-H. Nguyen et al. / Chemical Engineering Journal 179
(2012) 285 294
Table 1Photocatalytic epoxidation of propylene.a
Entry Catalysts BET surfacearea(m2 g1)b
Temp.(C)
Lightintensity(mW/cm2)
C3H6 PO formationrate (mol g1 h1)
Selectivity (%)c
Adsorptionrate (molg1 h1)
Conv. rate(molg1 h1)
AA ROH PO PA AC CO2
1 P25 50.0 50 0.2 3444.8 619.5 7.6 ND ND 0.4 17.6 74.42 Au
(3%)/TiO2 37.9 50 0.1 2900.9 455.4 4.13 33.7 ND 1.0 5.4 21.7 38.23
SiO2 560.1 50 0.1 2113.6 4.5 0.21 77.3 ND 10.6 12.1 ND ND4 V2O5
(0.5%)/SiO2 50 0.2 1459.9 26.1 49.9 ND ND 50.1 ND ND
5 TS-1 368.9 25 0.2 59.6 25.4 1.31 31.6 ND 24.1 ND ND 44.350 0.2
3886.7 50.7 19.69 29.0 0.9 40.1 22.9 7.1 trace50 0.1 521.9 55.7
24.08 15.9 ND 44.7 32.3 7.0 trace70 0.1 n.s. 59.7 17.91 38.3 ND
31.0 24.6 6.1 trace50 0.05 n.s. 16.9 5.02 38.5 ND 33.2 16.6 12.0
trace
6 V-Ti/MCM-41 790 50 0.2 11972.5 243.5 114.18 21.3 ND 46.8 26.2
5.7 ND25 0.2 n.s. 163.1 54.52 41.4 2.3 32.0 17.8 6.5 ND75 0.2
2013.6 223.2 87.92 22.0 1.7 40.2 26.9 9.2 ND120 0.2 13624.9 207.5
81.96 11.8 ND 39.6 45.1 3.5 ND50 0.4 9323.8 283.0 92.80 24.3 1.4
33.2 22.6 18.5 ND50 0.1 12915.6 121.8 48.35 21.1 ND 40.1 29.8 9.0
ND
ND: not detected by GC.a Reaction conditions: photocatalyst 0.02
g; feed gas C3H6:O2:N2 = 1:1:18 vol% at a gas hourly space velocity
(GHSV) of 6000 h1. The data is the mean value obtained on
stream after 4 h.b BET surface area was measured by
Micromeritics, ASAP 2000.c PO: propylene oxide; PA:
propionaldehyde; AC: acetone; AA: acetaldehyde; ROH: ethanol and
methanol.
of elementcoordinatiothe averageshould be Tand (b),
rescoordinatioshown in Fthat of TS-1is close to gdistorted. Tto
propylen
irect
le 1 sene wckedpylenvity rch a
nol wramand . The pre-edge peak shows the information of localn
structure [31]. Compared with the Ti(III) and Ti(IV),
oxidation state of titanium in V-Ti/MCM-41 and TS-1i4+ from the
shifts of absorption edges shown in Fig. 7(a)pectively. From the
pre-edge peaks of Fig. 7(a), the localn structure of Ti4+ is
suggested to be tetrahedral. Asig. 7(b), the pre-edge peak of TS-1
fresh is higher than
reacted. The local coordination structure of TS-1 freshenuine
tetrahedral while that of TS-1 reacted is slightlyhis suggests that
TS-1 is distorted due to Ti connectinge or other oxygenates during
reaction.
3.2. D
Tabpropylous paof proselectiucts, sumethamost d(TiO2Fig. 8. The
effective light of intensity and temperature for photocatalytic
epoxidation ogas-phase photocatalytic epoxidation of propylene
ummarizes the results of photocatalytic epoxidation ofith oxygen
over a series of photocatalysts in a continu-
-bed reactor at mild conditions. The conversion ratese range
from 4.5 to 619.5 mol g cat1 h1 while POanges from 1% to 46.8%. In
addition to PO, other prod-s propionaldehyde, acetone,
acetaldehyde, ethanol andere detected and their selectivities were
calculated. The
atic differences between two supported photocatalystsSiO2) are
in the production of side products and thever TS-1: (a) PO
formation rate and (b) PO selectivity.
-
V.-H. Nguyen et al. / Chemical Engineering Journal 179 (2012)
285 294 291
Fig. 9. The time-dependent behavior of selectivity for the
formation of PO (), PA (), AC (), ROH () and MeCHO () on stream
over TS-1 under different light intensityconditions: (a) 0.1 mW/cm2
and (b) 0.2 mW/cm2.
Fig. 10. The effects of light intensity and temperature on the
photocatalytic epoxidation over V-Ti/MCM-41: (a) PO formation rate
(b) PO selectivity and (c) C3H6 conversionrate.
-
292 V.-H. Nguyen et al. / Chemical Engineering Journal 179
(2012) 285 294
Slec
tivity
(%)
0
10
20
30
40
50MeCHOPO PA AC
(a)_V- Ti/MCM -41_0.1mW_50 oC
Time on strea m (min)0 50 10 0 15 0 20 0 25 0
50
60MeCHOPO PA AC
(b)_V -Ti/MCM -41_0.2mW_50 oCSl
ectiv
ity (%
)
0
10
20
30
40MeCHOPO PA AC
(d)_V -Ti/MCM -41_0.2mW_120 oC
Fig. 11. The t0.1 mW/cm2, 5
conversion total oxidatC3H6 conve
The adsoduring phothe averagereaction. Thand chemicface area
anV-Ti/MCM4material.
Commerity. The C3HCO2 in the pis very impoAu particle of C3H6 to
CCO2 is prodlene, but alproducts. HTiO2.
The UV avery low C3Table 1. HoThe main r560.1 m2 g cin order
tocatalyst andTime on strea m (min)0 50 10 0 15 0 20 0 25 0
Slec
tivity
(%)
0
10
20
30
40
MeCHOPO PA AC ROH
(c)_V- Ti/MCM -41_0.4mW_50 oC
Slec
tivity
(%)
0
10
20
30
40
50Time on strea m (min)0 50 10 0 15 0 20 0 25 0 0
ime-dependent behavior of selectivity for the formation of PO
(), PA (), AC (), ROH0 C; (b) 0.2 mW/cm2, 50 C; (c) 0.4 mW/cm2, 50
C and (d) 0.2 mW/cm2, 120 C.
rate of C3H6. TiO2 support can convert more C3H6 toion while
SiO2 support leads to low oxidation at lowerrsion.rption of
propylene was observed for all photocatalyststoreaction after
switching on the UV light. Table 1 lists
adsorption rates based on the carbon balance in 4 he propylene
adsorption is the combination of physicalal adsorptions on the
supports, which depends on sur-d TiO2 content. The amount of
propylene adsorbed on1 was substantial because of the mesoporous
nature of
cial TiO2 (entry 1) shows high photo-oxidation activ-6
consumption rate is 619.5 mol g1 h1 with mostroducts, and no PO is
observed. The size of Au particlesrtant on epoxidation. This study
also conrms that the
with size higher than 5.0 nm prefers complete oxidationO2 as
shown in entry 2 [32]. A considerable amount of
uced not only from the direct photo-oxidation of propy-so from
the successive photo-oxidation of oxygenatedence, only a few PO can
be generated over Au/TiO2 and
bsorption of amorphous SiO2 is very low (Fig. 3), thusH6
conversion rate was obtained as listed in entry 3 ofwever, the
selectivity to form PO still can reach 10.6%.eason may be the high
specic surface area of SiO2,at1. Therefore, SiO2 can harvest more
photo-energy
produce sufcient electronhole pairs on the photo- carry out
partial oxidation successfully. The dispersed
vanadium absorption Table 1, POC3H6 convephous SiO2
The moand V-Ti/Mentry 5 and
0.05 C3H
6 con
v. &
PO
form
atio
n ra
te (
mol
.g-1
.h-1
)
0
50
100
150
200
250
300
Fig. 12. The lifor V-Ti/MCMTime on strea m (min) 50 10 0 15 0 20
0 25 0
() and MeCHO () over V-Ti/MCM-41 at different conditions:
(a)
oxide on silica had signicantly improved the lightat around
200550 nm (Fig. 3). As listed in entry 4 of
still cannot be produced over V2O5/SiO2 although thersion rate
increased 5 times as compared with amor-.st favorable
photocatalytic epoxidation was on TS-1CM-41 under various reaction
conditions as shown in
6 of Table 1, respectively. These two photocatalysts
Light i ntensit y (mW/cm2)0.10 0.15 0.20 0.25 0.30 0.35 0.40
0.45
C3H
6 ads
orpt
ion
rate
(m
ol.g
-1.h
-1)
9000
10000
11000
12000
13000
14000
Conv. ra te PO ra te Adsorp tion ra te
ght effect on PO formation, C3H6 conversion and C3H6 adsorption
rate-41.
-
V.-H. Nguyen et al. / Chemical Engineering Journal 179 (2012)
285 294 293
C
C
CH3H
O*O
C
C
CH3H*
O
C
C
CH3
O
ucts o
prepared frble for the poptimal con114.18 moV-Ti/MCM-
The phoferent lightoptimum cowhich can Further invtivity are
shachieved uFig. 9(a). Thof the Ti spfore, stableepoxidationlight
intenssupplied, malyst. Thus mresulting in
V-Ti/MCepoxidationthe light anfor V-Ti/MCimpacts on condition
inin Fig. 10(a)than TS-1 (Fenhances thever, oversuand formati
Comparetime-depenepoxidationble selectiv0.10.4 mWties at the is
that V-TiTS-1, hencePO is still thslight increat 50 C (Figtially
increathat of PO r
Fig. 12 tion, C3H6 cThe adsorpto its porouwith increathe C3H6
c
er, fu, the m P
opos
intrible f speativeataly
ataly
-phan as
e spes. He
by sho
+ h+
he epts. Foct wducts
CH C
C
CH3H
H H
O*
H H
C
C
CH3H
H H
O*
Fig. 13. The formation mechanism of PO and oxygenate prod
om SiO2 supported materials are all active and sta-roduction of
PO, despite having signicantly differentditions. For instance, the
highest PO formation rate ofl g cat1 h1 with selectivity of 46.8%
was observed for41at 50 C under light intensity of 0.2
mW/cm2.tocatalytic epoxidation over TS-1 was studied at dif-
intensities and temperatures as shown in Fig. 8. Thenditions of
0.10.15 mW/cm2 and 4050 C were foundachieve the highest PO
formation rate and selectivity.estigations on time-dependent
behavior of PO selec-own in Fig. 9. The stable selectivities of
products are
nder the light intensity of 0.1 mW/cm2 as shown ine XANES (Fig.
7(b)) revealed that the oxidation statesecies are the same before
and after reaction. There-
active sites can be maintained during photocatalytic. However,
the selectivities are unstable with higherity as shown in Fig.
9(b). When higher light intensity isore elections and holes are
generated on the photocat-ore active intermediate oxygenates may be
generated
unstable selectivities.M-41 exhibited excellent selectivity for
the photo-
of propylene with molecular oxygen. Fig. 10 showsd temperature
effects on photocatalytic epoxidationM-41. Light intensity and
temperature have signicantphotocatalytic epoxidation and there
exists an optimum
the range of 0.20.3 mW/cm2 and 4060 C as revealed. V-Ti/MCM-41
has much lower UVvis absorption edgeig. 3). Hence, increase in
light intensity can signicantlye C3H6 conversion rate as shown in
Fig. 10(c). How-pply of light would eventually reduce the
selectivityon rate of PO as shown in Fig. 10(a) and (b).
to TS-1, V-Ti/MCM-41 also shows a quite similardent behavior in
PO selectivity for the photocatalytic. However, as shown in Fig.
11, V-Ti/MCM-41 holds sta-ities on stream within a wide range of
light intensity,/cm2, while TS-1 can only maintain stable
selectivi-light irradiation of 0.1 mW/cm2. The possible
reason/MCM-41 has the lower UVvis light absorption than
HowevHenceoptimu
3.3. Pr
TheA possedge oof negphotoc
photoc
Gaselectro
O2(g) +The
procesformedhole as
O2(ads)
In tproduccan reaall pro
CH3 more photo-energy is required than TS-1. Althoughe dominant
product with increasing light intensity, a
ase in acetone and alcohol concentration is observed. 11(a)(c)).
Increasing temperature to 120 C substan-ses the amount of
propionaldehyde produced whileemains at the same level (Fig.
11(d)).shows the inuence of light intensity on PO forma-onversion
and C3H6 adsorption rate over V-Ti/MCM-41.tion rate of C3H6 is
signicant on V-Ti/MCM-41 dues structure. The adsorption rate
decreases noticeablyse in light intensity. Increasing light
intensity enhancesonversion rate thus also increases PO formation
rate.
Carter ananism of proxypropenyangle of H-aand the reaate
intermeabstractionintermediapionaldehywe suggestthe pore reHH
CC
CH3
H
H
H
*
HH
O
CC
CH3
HH
H
ver V-Ti/MCM-41 photocatalyst.
rther increasing light intensity inhibits PO selectivity.PO
formation rate starts to decline after reaching anO formation rate
at 0.2 mW/cm2.
ed mechanism of photocatalytic epoxidation
insic epoxidation mechanism is still not quite clear.reaction
mechanism is proposed based on the knowl-cies present during the
photocatalytic reaction. A pair
electron (e) and positive hole (h+) is formed in thest with
light irradiation as in Eq. (8).
sthve + h+ (8)
se O2 is adsorbed and becomes O2 when react with anshown in Eq.
(9).
O2(ads) (9)
cies O2(ads) is stable and not directly active in thisnce, the
active species, surface atomic oxyradicals, arethe reaction of the
preceding species with the positivewn in Eq. (10).
2O(ads) (10)
oxide reaction, there are several pathways to create ther
example, the above surface atomic oxyradicals O(ads)
ith propylene to yield 3 kinds of products (8090% for) as listed
in Eq. (11).
O* CH2 +
CH3 CH CH 2
O
CH3 CH 2 CH
O
CH3 C CH 3
O
(11)d Goddard [33] proposed the selective oxidation mech-opylene
to epoxide via H-atom abstraction by formingl intermediate. The
product selectivities depend on thetom abstraction from the
propylene attacked by O(ads)
ction conditions. Fig. 13 displays the mechanism to cre-diate
species of PO and another product via H-atom
in a limited space, e.g. inside pore. The oxypropenylte of PO is
favorable inside the pore while that of pro-de cannot be formed due
to space limitation. Therefore,ed that the selectivity of PO could
be increased due tostriction of MCM-41.
-
294 V.-H. Nguyen et al. / Chemical Engineering Journal 179
(2012) 285 294
4. Conclusions
A direct photo epoxidation process to produce PO from propy-lene
by O2 with the aid of light energy over photocatalysts is
apromising technology for the production of chemicals in the
nearfuture. Compared to other thermal processes of propylene
epox-idation, no hydrogen is required in the photocatalytic
process.Furthermore, sunlight may be applied in the
photo-epoxidationin the future. In this study, under mild
conditions, Ti-V/MCM-41and TS-1 promote the photocatalytic
epoxidation of propylene withmolecular oxygen at steady state in a
ow reactor system. The high-est PO formation rate and PO
selectivity achieved on stream after 4 hover Ti-V/MCM-41 is 114 mol
g cat1 h1 and 47%, respectively,at 50 C and atmospheric pressure.
In summary, V-Ti/MCM-41 isthe most favorable photocatalyst for the
photocatalytic epoxidationin this study, and the selectivity of PO
could be further enhancedpossibly by solving the pore restriction
of reaction intermediate.
Acknowledgment
We gratefully acknowledge the National Science Council ofTaiwan
for nancial supporting this research under contractnumbers,NSC
99-2923-E-002-002-MY2 and NSC 99-2221-E-009-037-MY3.
References
[1] K. Weissermel, H.-J. Arpe, C.R. Lindley, Industrial Organic
Chemistry, 4th illus-trated ed., Wiley-VCH, 2003, pp. 265276.
[2] T.A. Nijhuis, M. Makkee, J.A. Moulijn, B.M. Weckhuysen,
Industrial and Engi-neering Chemistry Research 45 (2006)
34473459.
[3] D. Kahlich, U. Wiechern, J. Lindner, Propylene Oxide, in:
Ullmanns Encyclopediaof Industrial Chemistry, John Wiley & Sons
Inc, 2009, pp. 2135521379.
[4] WP Coverage of Propylene Oxide, 2010.[5] G. Jin, G. Lu, Y.
Guo, Y. Guo, J. Wang, X. Liu, Catalysis Today 9395 (2004)
173182.[6] Z. Suo, M
184190.[7] L. Cumara
[8] J. Lu, X. Zhang, J.J. Bravo-Surez, K.K. Bando, T. Fujitani,
S.T. Oyama, Journal ofCatalysis 250 (2007) 350359.
[9] A. Fujishima, K. Honda, Nature 238 (1972) 3738.[10] Y.-H.
Yu, Y.-T. Pan, Y.-T. Wu, J. Lasek, J.C.S. Wu, Catalysis Today 174
(2011)
141147.[11] S. Cao, K.L. Yeung, P.-L. Yue, Applied Catalysis B:
Environmental 76 (2007)
6472.[12] J. Wu, C.-W. Huang, Frontiers of Chemical Engineering
in China 4 (2010)
120126.[13] A.J. Maira, W.N. Lau, C.Y. Lee, P.L. Yue, C.K. Chan,
K.L. Yeung, Chemical Engineer-
ing Science 58 (2003) 959962.[14] T.-V. Nguyen, J.C.S. Wu,
Applied Catalysis A: General 335 (2008) 112120.[15] T.-V. Nguyen,
J.C.S. Wu, Solar Energy Materials and Solar Cells 92 (2008)
864872.[16] K.L. Yeung, W.K. Leung, N. Yao, S. Cao, Catalysis
Today 143 (2009) 218224.[17] C. Murata, H. Yoshida, J. Kumagai, T.
Hattori, Journal of Physical Chemistry B
107 (2003) 43644373.[18] H. Yoshida, C. Murata, T. Hattori,
Journal of Catalysis 194 (2000) 364372.[19] F. Amano, T. Yamaguchi,
T. Tanaka, Journal of Physical Chemistry B 110 (2006)
281288.[20] V.-H. Nguyen, J.C.S. Wu, Photocatalytic epoxidation
of propylene with molec-
ular oxygen by photocatalysts, in: 28th Taiwan Symposium of
Catalysis andReaction Engineering, Taipei, Taiwan, 2010, Abtract
book, p. 48.
[21] H. Yoshida, C. Murata, T. Hattori, Chemical Communication
(1999) 15511552.[22] H. Yoshida, T. Tanaka, M. Yamamoto, T.
Funabiki, S. Yoshida, Chemical Commu-
nication (1996) 21252126.[23] H. Kanai, M. Shono, K. Hamada, S.
Imamura, Journal of Molecular Catalysis A:
Chemistry 172 (2001) 2531.[24] Y.-F. Yang, P. Sangeetha, Y.-W.
Chen, International Journal of Hydrogen Energy
34 (2009) 89128920.[25] R.B. Khomane, B.D. Kulkarni, A.
Paraskar, S.R. Sainkar, Materials Chemistry and
Physics 76 (2002) 99103.[26] F. Amano, T. Tanaka, Catalysis
Communications 6 (2005) 269273.[27] B.F. Mentzen, F. Lefebvre,
Comptes Rendus de lAcadmie des Sciences Series
IIC Chemistry 3 (2000) 843847.[28] L. Wang, Y. Zhou, Z. Mi,
Journal of Chemical Technology and Biotechnology 82
(2007) 414420.[29] J.-J. Chen, J.C.S. Wu, P.C. Wu, D.P. Tsai,
Journal of Physical Chemistry C 115 (2010)
210216.[30] J.F. Moulder, J.E. Chastain, Handbook of X-ray
Photoelectron Spectroscopy: A
Reference Book of Standard Spectra for Identication and
Interpretation of XPSData, Perkin-Elmer Corporation, Physical
Electronics Division, Eden Prairie, MN,1992, pp. 5657.
[31] T. Maschmeyer, F. Rey, G. Sankar, J.M. Thomas, Nature 378
(1995) 159162.. Oyamalysis. Carte. Jin, J. Lu, Z. Wei, C. Li,
Journal of Natural Gas Chemistry 17 (2008)
natunge, W.N. Delgass, Journal of Catalysis 232 (2005) 3842.
[32] S.TCat
[33] E.Aa, Mechanisms in Homogeneous and Heterogeneous
Epoxidation, 1st ed., Elsevier, Amsterdam, Boston, 2008, Chapter
11, pp. 315338.r, W.A. Goddard, Journal of Catalysis 112 (1988)
8092.
Direct gas-phase photocatalytic epoxidation of propylene with
molecular oxygen by photocatalysts1 Introduction2 Experimental2.1
Preparation of photocatalysts2.2 Characterization2.3 Direct
gas-phase photocatalytic epoxidation of propylene
3 Results and discussions3.1 Photocatalyst characterization3.2
Direct gas-phase photocatalytic epoxidation of propylene3.3
Proposed mechanism of photocatalytic epoxidation
4 ConclusionsAcknowledgmentReferences