185 nm photoreduction of CO2 to methane by water. Influence of the presence of a basic catalyst

Article abstract of DOI:10.1021/ja304930t

Photolysis of CO₂ at 185 nm using a Hg lamp in the presence of H₂ or H₂O leads to the formation of CH₄ (after an induction period) and a lesser amount of CO. Using H₂ gas as reducing agent, up to 40% conversion was achieved with almost complete selectivity to methane and energy consumption of 1.55 Gcal/mol. When CO ₂ irradiation is carried out in the gas phase in the presence of H₂O, both reagents can undergo photolysis, and H₂ generation from water is a competing process accompanying the formation of CO and CH₄ (CO₂ conversion 0.67% at 14 h). For the irradiation of CO₂ using H₂O as reducing agent, basic solids (K⁺-exchanged Y zeolite or hydrotalcites) can increase CO ₂ conversion up to 2.21% with total selectivity toward CH₄ and an energy consumption of 4.5 Gcal/mol. This result opens a door in the quest for efficient photocatalysts that could further increase the apparent quantum yield of the process, particularly at longer wavelengths.

Full text of DOI:10.1021/ja304930t

Article
pubs.acs.org/JACS
1
85 nm Photoreduction of CO to Methane by Water. Influence of
2
the Presence of a Basic Catalyst
Francesc Sastre, Avelino Corma,* and Hermenegildo García*
Instituto Universitario de Tecnología Química CSIC-UPV, Universidad Politec
Valencia, Spain
́
nica de Valencia, Av. De los Naranjos s/n, 46022
ABSTRACT: Photolysis of CO at 185 nm using a Hg lamp
2
in the presence of H or H O leads to the formation of CH
2
2
4
(
after an induction period) and a lesser amount of CO. Using
2
H gas as reducing agent, up to 40% conversion was achieved
with almost complete selectivity to methane and energy
consumption of 1.55 Gcal/mol. When CO irradiation is
2
carried out in the gas phase in the presence of H O, both
2
reagents can undergo photolysis, and H generation from
2
water is a competing process accompanying the formation of
+
CO and CH (CO conversion 0.67% at 14 h). For the irradiation of CO using H O as reducing agent, basic solids (K -
4
2
2
2
exchanged Y zeolite or hydrotalcites) can increase CO conversion up to 2.21% with total selectivity toward CH and an energy
2
4
consumption of 4.5 Gcal/mol. This result opens a door in the quest for efficient photocatalysts that could further increase the
apparent quantum yield of the process, particularly at longer wavelengths.
INTRODUCTION
chromophores can be selectively excited by irradiation at
wavelengths longer than 250 nm. Deep UV photochemistry can
be extremely useful and widely applicable for the activation of
simple molecules, particularly H O, CO , and CH in the gas
■
Activation of carbon dioxide is one of the present challenges in
1
−3
chemistry.
One current line of research that is related with
2
2
4
renewable energies is the replacement of fossil fuels by
phase or on solid surfaces. In this context, it is worth
commenting that there are commercially available simple light
sources for deep UV irradiation based on the emission of
mercury vapor (185 nm) or deuterium gas (165 nm). Metal
halides, and particularly calcium fluoride, are convenient
optically transparent materials in this region for chambers
and cells.
sustainable energy resources that do not exert a negative
4
−6
influence on the environment.
In this context, one
possibility is the photochemical reduction of CO to form
2
methanol, methane, and CO among other possible prod-
4
,6,7
ucts.
Most of the current approaches to achieve this
transformation are based on the use of UV or visible light in
8
,9
combination with a photocatalyst.
Herein, we have taken advantage of direct 185 nm irradiation
The present report deals with an innovative methodology in
which photochemical irradiation is performed in the “deep” UV
region (radiation wavelength shorter than 200 nm). Photons in
the deep UV zone can be directly absorbed by H O or CO and
of CO and H O to develop a simple and efficient methodology
2
2
for the photochemical reduction of CO to methane with an
2
energy consumption in the range of large industrial processes
such as steam re-forming or water−gas shift. Although the
strategy reported here cannot serve for direct solar light
conversion, it has in common the use of light and photo-
2
2
have enough energy to effect the monophotonic direct
photolysis of these two highly stable molecules. Deep UV
photolysis of CO to split into CO and oxygen atoms has been
2
reported in the literature to occur with high quantum
catalysts to effect CO reduction by hydrogen or water. In the
2
1
0−12
efficiency.
In addition, it is known that the photolysis of
context of solar fuels production, our method, requiring the
prior conversion of sunlight into electricity, can be competitive
with direct visible light photocatalytic conversion due to the
higher efficiency inherent to monophotonic processes based on
deep UV compared to multistep photon absorption mecha-
nisms.
CO and H O with deep UV can form alcohols and aldehydes
2
13
among other compounds.
The reaction of CO and H O promoted by light mimics in a
2
2
certain way natural photosynthesis in which these two
molecules are the substrates for the formation of carbohydrates
1
4−16
and oxygen.
We have also recently shown that, similar to
the direct photolysis of water in the gas phase, deep UV
photons can also effect surface photochemistry via O−H bond
cleavage of surface hydroxyl groups in solids that are otherwise
RESULTS AND DISCUSSION
■1
85 nm Reduction of CO by H . H and N gases do not
2
2
2
2
19
1
7,18
absorb at 185 nm, while CO absorbs from 165 to 205 nm.
2
photochemically inert to UV and visible radiation.
UV
photochemistry has been generally ignored as a means to
promote organic transformation due to the fact that most
solvents and gases absorb in this region, while some organic
Received: May 21, 2012
Published: August 6, 2012
©
2012 American Chemical Society
14137
dx.doi.org/10.1021/ja304930t | J. Am. Chem. Soc. 2012, 134, 14137−14141

Journal of the American Chemical Society
Article
When a mixture of CO and H diluted in N is photolyzed at
was one of the first processes studied in this wavelength
range, and that it leads to homolytic OH bond cleavage with a
2
2
2
20
185 nm, the highly selective formation of methane with very
minor amounts of CO is observed. The presence of methanol
or other products is not detected at any time during the
reaction. Figure 1 shows the temporal profile of CO2
quantum efficiency approaching unity (eq 2), we envisioned the
hν(<200 nm)
•
•
1
H O ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ →⎯ H + OH → → H +
O2
2
2
2
(2)
possibility of coupling direct H O splitting and H generation
2
2
with CO reduction. To address this possibility, we performed
2
1
85 nm irradiation of CO diluted in moisture-saturated N at
2
2
room temperature; the molar ratio of the CO and H O feed
2
2
was 1:20. Under these conditions, a gradual generation of
hydrogen was observed, indicating that eq 2 is operating. It
should be commented that although deep UV photons can
effect homolytic water splitting as indicated in eq 2 to generate
hydrogen atoms and hydroxyl radicals, controls by colorimetric
titration have not been able to detect the formation of
measurable amounts of hydroxyl peroxide.
The concentration of hydrogen coming from water increases
steadily with time. In addition of H , formation of CH after
2
4
some induction period is also observed. Figures 2 and 3 show
Figure 1. Temporal profile of the CO (■) conversion and CH (▲)
2
4
and CO (●) yields. Photolysis conditions: 185 nm at 65 °C; mixture
of H (44%), CO (9.5%), and N (45%).
2
2
2
conversion, as well as the yields of methane and CO vs the
irradiation time. As can been seen in this figure, under our
irradiation conditions, conversion of CO increases almost
2
linearly with the irradiation time in the time interval up to 3 h.
High conversions, about 40%, were obtained at 80 h under our
photolysis conditions. These conversions are without precedent
6
for a photochemical CO2 activation. Interestingly, the
formation of CO develops faster than that of methane (see
inset in Figure 1) but reaches a quasi stationary low level that
reaches a steady state at longer irradiation times. In contrast,
the formation of methane exhibits an induction period, but after
this initial time its concentration grows over time, parallel to
Figure 2. Temporal profile of the CO conversion (■, left vertical
2
axis) and H
production upon irradiation of CO and H O with deep UV light.
(▲
, right vertical axis) and CH (●, right vertical axis)
2
4
2
2
Photolysis conditions: mixture of CO (9.5%) and N (90.5%) with
the CO conversion.
2
2
2
saturated water; no catalyst; Hg lamp as irradiation source, 185 nm; 60
C.
In order to gain information on the reaction stoichiometry
and particularly on the products arising from oxygen removed
°
from CO , we performed an analogous experiment using
2
1
3
18
the temporal evolution of CO
2
conversion and formation of H
and H O (1:20) is
2
isotopically labeled C O and followed the reaction by mass
2
18
and CH over time when a mixture of CO
4
2
2
spectrometry. Although the presence of lesser amounts of O
2
photolyzed in the gas phase in the absence or in the presence of
Zn/Ti LDH solid. Neither methanol nor CO was detected
under any conditions during the reaction time. The absence of
was detected, the predominant oxygen-containing product
13
18
2
accompanying CH was H O. Accordingly, eq 1 summarizes
4
the stoichiometry of the photochemical CO splitting.
2
CO previously observed in the photolysis of CO in the
2
hν(<200 nm)
presence of H indicates that reduction of CO in the photolysis
2
CO + 4H ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ →⎯ CH + 2H O
(1)
2
2
4
2
where H O is present has to be considerably faster than its
2
formation.
Most of the present study has been carried out using a
mercury lamp that emits at two wavelengths, namely 185 nm
with lower intensity and 254 nm with higher intensity (see
Experimental Section for the lamp emission). Control experi-
ments using quasi monochromatic 254 nm light from a low
vapor pressure mercury lamp indicate that the photons causing
the chemical transformation are those at 185 nm and not those
at 254 nm.
To determine the stoichiometry of the CO
O and the possible evolution of oxygen in the photolysis, we
2
2
reduction by
H
1
3
18
studied the reaction using isotopically labeled C O
followed the reaction by mass spectrometry. Under these
conditions, in which there is an excess of H O with respect to
and
2
2
13
18
18
13 18
C O
, we observed a scrambling of O atoms from C O
2
2
16
16 18
18
to H O as well as the formation of O , O O, and O in
2
2
2
relative proportions of 100, 0.1, and 0.08%. These results
indicate that a fast equilibrium through the formation of
CO2 Reduction at 185 nm Promoted by H O.
2
1
8
Considering that the direct photolysis of H O with deep UV
carbonic acid with oxygen scrambling between C O and
2
2
1
4138
dx.doi.org/10.1021/ja304930t | J. Am. Chem. Soc. 2012, 134, 14137−14141

Journal of the American Chemical Society
Article
H O that also absorbs in the deep UV, two alternative pathways
2
could be available, depending on which substrate is photolyzed,
either CO or H O. The fact that CO conversions using water
2
2
2
as reducing agent are lower than those using H while the
2
product distribution is the same suggests that water photolysis
is the exclusive hydrogen source and the hydroxyl radical and
other intermediates derived from water photolysis are not
involved in the formation of CH4.
Influence of the Presence of a Basic Solid. We recently
reported that deep UV photolysis for methane activation can be
1
7,18
carried out more efficiently on a solid surface.
Thus, we
have shown that deep UV photolysis of silanol groups on the
surface of silicates can serve to activate methane toward the
17,18
formation of C1 oxygenates.
In the present case, the
situation is somewhat different since, in contrast to methane
■
, left vertical
activation, direct photolysis at 185 nm of CO can take place in
2
Figure 3. Temporal profile of the CO conversion (
2
axis) and H (▲, right vertical axis) and CH (●, right vertical axis)
production upon irradiation of CO and H O with deep UV light.
the gas phase regardless the presence or not of any solid
surface. However the presence of a surface could lead to an
increase on the photochemical yield by adsorbing and confining
2
4
2
2
Photolysis conditions: mixture of CO (9.5%) and N (90.5%) with
2
2
saturated water; Zn/Ti LDH, 0.5 g; Hg lamp as irradiation source, 185
CO and the reaction on a 2D or 3D space. In addition, the
2
nm; 60 °C.
presence of a solid can open new pathways for the reaction.
Aimed at addressing the influence of a solid surface on the
H216O takes place before the photochemical reaction (eq 3),
and depending on the initial proportion of O / O, a
statistical distribution of isotopically labeled molecular oxygen
evolves.
185 nm photolysis of CO we screened a series of materials
2
1
6
18
including [Ti]Beta, Beta(OH), and HY zeolites and TiO2,
CeO , and MgO metal oxides. In all these cases, we observed
2
that the conversion of CO2 either using H2 or H O as
2
reductant, decreased considerably, thus showing the negative
influence of the solid on the photoreaction. We also tested the
performance of a basic large pore zeolite (KY) and two
hydrotalcites (Zn/Ti LDH and Zn/Ce LDH) as promoters of
the photochemical CO2 reduction by water.
The case of TiO2 is interesting considering that the
experiments were carried out with a light source that contains
also 254 nm (see Figure 5 in the Experimental Section for the
It is noticeable that the CO conversion in the presence of an
2
excess of H O is about 15-fold smaller than for the irradiation
2
in H . This is due to the fact that H does not absorb 185 nm
2
2
20
light, while H O has a strong absorption at this wavelength.
2
Moreover, being in large excess, H O prevents the photolysis of
2
emission spectrum of the lamp) and that TiO has been
2
CO . To increase CO conversion, we performed photolysis
2
2
reported to have some photocatalytic activity for CO
2
with a much smaller H O excess. Upon photolysis of a CO /
2
2
21−24
reduction.
Thus, in principle the possibility of a synergy
H O 1:1 mixture, up to 420 μmol of H , 60 μmol of CO, and
2
2
between the direct deep UV photolysis (185 nm) and
photocatalytic CO reduction promoted by TiO (254 nm)
1
5 μmol of methane was achieved with an energy consumption
of 20 Gcal/mol.
2
2
could be expected. The fact that the conversion decreased in
the presence of TiO clearly indicates that the contribution of
Although CH is a common final product of deep UV
4
photolysis using H or H O as reducing agent, we notice that,
2
2
2
at least in part, the reaction mechanism must be different for
the two substrates. The difference must arise from the fact that
the direct gas phase photolysis of CO
2
is much larger than any
possible photocatalytic conversion occurring on the TiO
2
H does not absorb 185 nm photons, and, therefore, in this
surface. This prevalence of the deep UV pathway can be
simply understood considering that the accepted mechanism
for the photocatalytic TiO reduction of CO involves several
2
case, the CO conversion must come exclusively from the CO2
2
irradiation, probably leading to CO. However, in the case of
2
2
Table 1. CO Conversion Mass Balance, Selectivity and Energy consumption for the 185 nm Photolysis of CO in the Presence
2
2
a
of Basic Solids
selectivity
catalyst
CO conversion (%)
molar balance
CO
CH4
energy consumption (Gcal/mol)
μmol H /(g ·h)
μmol CH /(g ·h)
2
2
cat
4
2.1
3.1
4.3
3.3
cat
d
d
none
0.67
0.85
1.14
0.88
2.21
1.75
0.23
96.8
98.6
99.9
99.9
95.3
99.9
98.9
0
0
0
0
0
0
0
100
100
100
100
100
100
100
14.9
11.7
8.7
9.9
KY
63.2
33.1
41.4
Zn/Ce LDH
Zn/Ti LDH
11.2
4.5
b
Zn/Ti LDH
740
77
c
Zn/Ti LDH
5.4
31.2
14
3.77
0.63
MgO
42.8
a
Photolysis conditions: feed, water-saturated mixture of 90.5% N and 9.5% CO ; irradiation time, 14 h; solid, 500 mg unless indicated otherwise;
2
2
2
2
b
c
d
catalyst distributed in a 4 × 13 cm shallow bed and illuminated with 86 W/m . 100 mg of catalyst. 50 mg of catalyst. μmol of CH /h.
4
1
4139
dx.doi.org/10.1021/ja304930t | J. Am. Chem. Soc. 2012, 134, 14137−14141

Journal of the American Chemical Society
Article
Figure 4. Left: 185 photoreactor composed by In Hg lamp (1), synthetic quart windows (2), gas chamber (3). Right: Upper view showing the
shallow bed of the material placed in the photoreactor.
steps requiring light absorption and is a totally different process
than the direct bond splitting occurring in deep UV.
amounts of CO and if this were the only operating factor, i.e.
2
no photocatalysis is taking place, all of them should behave
similarly promoting the reaction This is apparently no the case
and the composition of the of the LDH and the nature of the
It is known that CO can be adsorbed on basic solids, this
2
being the key point for developing efficient CO adsorb-
2
2
5−27
ents.
In the case of microporous zeolites it has been shown
that basicity can be introduced by exchanging Na by other
solid play a role in the promotional effect of CO reduction.
2
+
This influence of the metals present in the LDH and,
particularly the fact that Zn-containing LDH have been found
to be photocatalysts for water splitting, suggest that also in this
case there is some contribution of photocatalysis when certain
LDHs solid are present in the photolysis. Therefore it seems
that in order to develop materials for the photocatalytic CO2
reduction, solids that combine CO2 absorption capacity
together with photoresponse at the appropriate wavelength
appear as promising materials. Considering the diversity of
metal combinations that can be present in their composition,
hydrotalcites appear as highly promising materials for this
photocatalytic reduction.
2
8,29
alkali metal cation with larger ionic radius.
Zeolite Y is a
microporous crystalline aluminosilicate having quasi spherical
cavities (about 1.3 nm diameter) tetrahedrally interconnected
30
by windows of about 0.72 nm. LDHs are layered materials
that contain a divalent metal cation together with a proportion
up to 30% of a tri- or tetravalent cation that is isomorphically
replacing some of the positions of the divalent cation in the
3
1,32
layer.
The presence of the tri- or tetravalent cation
introduces some positive charges in the framework that
requires the presence of charge compensating anions in the
3
1,32
intergallery space.
Carbonate is one of the most common
ions and even if other ion is initially present in the solid, it can
subsequently be replaced by carbonates due to the high affinity
CONCLUSIONS
■
3
1,32
for these dinegative anions.
LDHs have also been used as
In the present study we have shown that 185 nm photolysis of
basic catalysts to promote condensations and nucleophilic
CO in the presence of either hydrogen or water leads to the
3
3,34
2
substitutions, among other reactions.
formation of CH . When using water as reagent there is a
4
In this context the positive influence of the presence of KY
possibility to increase the yield of the process by performing the
reaction on the surface of solids with adequate basic strength.
The energy consumption per mole of converted CO varies
depending the conditions, but when water is used as reductant
it is only about 4.5 Gcal/mol of CO converted, a value in the
zeolite or LDHs as catalysts on the photoredution of CO can
2
be rationalized considering that due to their basicity these solids
2
are good CO adsorbents. Thus, due to the adsorption of CO2
2
on the surface of these solids, the 185 nm photoreduction of
2
CO by H O would occur preferentially on the solid surface,
2
2
same range as large-scale industrial processes such as steam re-
where it should be a high concentration of these two molecules
and the species derived from there.
In support of this proposal experiments was carried out using
17,18
forming.
Our results open the way to further implementa-
tion of solids as catalysts that can promote more efficiently the
photoreduction of CO by water.
2
H2 gas instead of H O as reducing agent. Under these
2
conditions and considering that H adsorption on the solid
2
EXPERIMENTAL SECTION
■
surface at 60 °C should be negligible in contrast to the strong
Materials. The Zn−Ti LDH was prepared by coprecipitation of
adsorption of CO , it was observed that the presence of any of
2
zinc and titanium salts from homogeneous solution. A solution of zinc
these three solids plays an adverse effect reducing CO2
conversion with respect to the conversions in the gas phase
photolysis. Therefore, comparison of the conversions in the
−2
−3
chloride (6 × 10 mol) and titanium tetraetoxide (9.5 × 10 mol)
was mixed with a 1 M urea solution under magnetic stirring at 100 °C,
for 48 h under inert atmosphere. After filtration and washing several
times in distilled water at pH 7, the product was dried at 60 °C for 13
h.
presence of solids using H or H O as reducing agent clearly
2
2
revealed notable differences in the reaction mechanism and
opens the door for implementing the use of solid surfaces as
Zn/Ce LDH was prepared also by the same method for
photocatalyst to enhance CO conversion in the deep UV
comparison. TiO P25 was a commercial sample from Degussa.
2
2
Basic zeolites were prepared starting from NaY (PQ, CBV100) by
submitting the solids to ion exchange using 1 M aqueous solution of
KAcO in a liquid/solid ratio of 10. The suspensions were magnetically
stirred at 80 °C for 4 h and then the solids were filtered, washed with
bidistilled water and dried at 200 °C under ambient atmosphere.
Description of the Photoreactor. Figure 4 provides a schematic
of the photoreactor used in this study. Photolyses in the presence of
photolysis.
Concerning the possibility that Zn/Ti or Zn/Ce LDHs were
acting merely as CO absorbent or if in addition to their
adsorption capacity these Zn-containing LDHs promote CO2
reduction also photocatalytically, absorbing light and injecting
electrons into CO , it is worth commenting the difference in
behavior of these Zn LDHs with other basic LDHs or MgO. In
principle all LDHs as well as MgO should adsorb significant
2
2
solids were carried out by spreading uniformly a bed of powdered
2
samples on the photoreactor covering a surface of 4 × 13 cm . CO gas
2
1
4140
dx.doi.org/10.1021/ja304930t | J. Am. Chem. Soc. 2012, 134, 14137−14141

Journal of the American Chemical Society
Article
was introduced into the reactor diluted with hydrogen or nitrogen
saturated with water. 185 nm Irradiation of the solid acting as
photocatalyst was carried out using three Hg lamps emitting at 185 nm
REFERENCES
■
(
1) Anpo, M.; Yamashita, H.; Ichihashi, Y.; Fujii, Y.; Honda, M. J.
Phys. Chem. B 1997, 101, 2632−2636.
(4 W) and 254 nm (8 W) (see Figure 5). Due to the heating by the
(
(
(
(
2) Sayama, K.; Arakawa, H. J. Phys. Chem. 1993, 97, 531−533.
action of the light the temperature of the reaction chamber reached a
stationary value of 60−65 °C.
3) Song, C. S. Catal. Today 2006, 115, 2−32.
4) Centi, G.; Perathoner, S. ChemSusChem 2010, 3, 195−208.
5) Hammarstrom, L.; Hammes-Schiffer, S. Acc. Chem. Res. 2009, 42,
1
(
2
(
3
(
859−1860.
6) Roy, S. C.; Varghese, O. K.; Paulose, M.; Grimes, C. A. ACS Nano
010, 4, 1259−1278.
7) Serpone, N.; Lawless, D.; Terzian, R. Sol. Energy 1992, 49, 221−
4.
8) Hwanga, J.-S.; Changa, J.-S.; Parkb, S.-E.; Ikeuec, K.; Anpo, M.
Top. Catal. 2005, 35, 311−319.
9) Mori, K.; Yamashita, H.; Anpo, M. RSC Adv. 2012, 2, 3165−
172.
(
3
(
(
(
10) Mahan, B. H. J. Chem. Phys. 1960, 33, 959−965.
11) Warneck, P. Discuss. Faraday Trans. 1964, 37, 57−65.
12) Reeves, R. R., Jr.; Harteck, P.; Thompson, B. A.; Waldron, R. W.
J. Phys. Chem. 1966, 70, 163−1640.
(
(
(
(
13) Bar-Nun, A.; Hartman, H. Origin of Life 1978, 9, 93−101.
14) Bard, A. J.; Fox, M. A. Acc. Chem. Res. 1995, 28, 141−145.
15) Gratzel, M. Acc. Chem. Res. 1981, 14, 376−384.
16) Gust, D.; Moore, T. A.; Moore, A. L. Acc. Chem. Res. 2001, 34,
4
0−48.
17) Sastre, F.; Fornes, V.; Corma, A.; Garcia, H. J. Am. Chem. Soc.
011, 133, 17257−17261.
18) Sastre, F.; Fornes, V.; Corma, A.; Garcia, H. Chem.Eur. J.
2012, 18, 1820−1825.
19) Parkinson, W. H.; Rufus1, J.; Yoshino, K. Chem. Phys. 2003, 293,
51−256.
20) Getoff, N.; Schenk, G. O. Photochem. Photobiol. 1968, 8, 167−
178.
(
2
(
Figure 5. Emission spectrum for the mercury lamp.
(
2
(
The course of the reaction was followed by analyzing the gas phase
and the organic compounds adsorbed on the solid, when a
photocatalyst was present. Reproducibility of the data was checked
by performing independent experiments in quadruplicate, whereby
consistent results with significantless variations were obtained.
The gas products were analyzed using a Rapid Refinery Gas
Analyser from Bruker that consists in a three channel GC. The first
(21) Dey, G. R. J. Nat. Gas Chem. 2007, 16, 217−226.
(22) Fujishima, A.; Zhang, X.; Tryk, D. A. Surf. Sci. Rep. 2008, 63,
515−582.
(23) Ilie, M.; Cojocaru, B.; Parvulescu, V. I.; Garcia, H. Int. J.
Hydrogen Energy XXXX, 36, 15509−15518.
(24) Koci, K.; Obalova,
́ ́ ́
́
L.; Matejova, L.; Placha, D.; Lacny, Z.;
channel analyses H using a micro packet HayeSep Q and Molsieve 5A
Jirkovsky, J.; Solcova, O. Appl. Catal. B-Environ. 2009, 89, 494−502.
́
́
2
column with Ar as carrier gas and TC detector. The second channel
analyses CO, CO , N , and O with a combination of micropacket
(25) Gascon, J.; Aktay, U.; Hernandez-Alonso, M. D.; van Klink, G.
P. M.; Kapteijn, F. J. Catal. 2009, 261, 75−87.
2
2
2
Haysep Q, H−N and Molsieve 13X columns using He as carrier gas
and a TC detector. The third channel analyses hydrocarbons from C
(26) Lavalley, J. C. Catal. Today 1996, 27, 377−401.
(27) Plaza, M. G.; Pevida, C.; Arenillas, A.; Rubiera, F.; Pis, J. J. Fuel
2007, 86, 2204−2212.
1
to C in an Al O plot column with He as carrier gas and a FI detector.
5
2
3
The mass balance of each experiment was determined by adding the
moles of the products in the gas phase to the moles of CO converted
on the solid. Quantification of the moles of the products in the gas
(
28) Barthomeuf, D. Catal. Rev.Sci. Eng. 1996, 38, 521−612.
2
(29) Barthomeuf, D.; Coudurier, G.; Vedrine, J. C. Mater. Chem.
Phys. 1988, 18, 553−575.
phase was carried out considering that N remains constant during the
2
(30) Meier, W. M.; Olson, D. H.; Baerlocher, C. Zeolites 1996, 17, 1−
experiment and using this gas as standard. The response factor of the
products with respect to N2 was determined by independent
calibrations.
2
29.
(31) Cavani, F.; Trifiro, F.; Vaccari, A. Catal. Today 1991, 11, 173−
3
01.
Controls to determine the possibility of H O2 formation were
2
(32) Miyata, S. Clays and Clay Minerals 1983, 31, 305−311.
(33) Climent, M. J.; Corma, A.; Iborra, S.; Primo, J. J. Catal. 1995,
performed by photolyzing for 60 min moisture saturated N and then
injecting 50 mL of the gas into an aqueous solution of K (TiO)-
2
2
151, 60−66.
(
C O ) (Aldrich) in H SO /HNO as colorimetric titrator. After 5
2 4 2 2 4 3
(34) Sels, B. F.; De Vos, D. E.; Jacobs, P. A. Catal. Rev.Sci. Eng.
min the solution was monitored at 420 nm for the formation of the
peroxytitanyl according to the standard method to titrate H O .
2
001, 43, 443−488.
3
5
2
2
(35) Navalon, S.; Martin, R.; Alvaro, M.; Garcia, H. ChemSusChem
011, 4, 650−657.
2
AUTHOR INFORMATION
Corresponding Author
acorma@itq.upv.es; hgarcia@qim.upv.es
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
Financial support by the Spanish Ministry of Economy and
Competitivity (CTQ-2009-11586) is gratefully acknowledged.
■
1
4141
dx.doi.org/10.1021/ja304930t | J. Am. Chem. Soc. 2012, 134, 14137−14141