Journal of Photochemistry and Photobiology A: Chemistry 163 (2004) 503–508

Photo-catalytic reduction of carbon dioxide to methane

using TiO2 as suspension in water
G.R. Dey a,∗ , A.D. Belapurkar b , K. Kishore a
a Radiation Chemistry and Chemical Dynamics Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400 085, India
b Applied Chemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400 085, India

Received 11 August 2003; received in revised form 1 December 2003; accepted 27 January 2004

Abstract
Photo-catalytic reduction of CO2 to methane was studied in CO2 -saturated aqueous solution in presence of TiO2 photo-catalyst (0.1%,
w/v) as a suspension using 350 nm light. CO2 methanation rate was very much enhanced in the presence of 2-propanol as a hole scavenger.
In addition to CO2 reduction, photo-catalytic reduction of methanol in N2 -purged system was also tried but no methane was generated in
presence of TiO2 without 2-propanol. The yield of methane was quite low even in presence of 2-propanol in this system. In the methanol
photolysis CO2 was found to be the main product with a yield of 60 × 10−8 mol to 200 × 10−8 mol. This shows that the generation of CH4
from CO2 does not proceed via methanol as an intermediate under these conditions. In the aerated system also, methane was produced
during photolysis of TiO2 suspension in presence of 2-propanol and its yield was comparable to that in CO2 -saturated system. This suggests
that surface-adsorbed as well as in situ generated CO2 are equally responsible for methane formation through photo-reduction in presence
of TiO2 . In the aerated system, 45 ␮mol CO2 was produced by photo-degradation of 2-propanol. In O2 -saturated system, the methane yield
was lower as compared to that in aerated system whereas CO2 yield was higher. Overall, the yield of methane was quite high if extrapolated
to per gram of TiO2 .
© 2004 Elsevier B.V. All rights reserved.
Keywords: CO2 -reduction; Methane; TiO2 ; Methanol; Hole scavenger

1. Introduction The use of sunlight to convert potentially harmful organic

chemicals to harmless mineral products or useful organic
Use of solar energy in solving environmental problems has chemicals at accelerated rates by the use of a comparatively
become very important [1–7]. Insecticides, herbicides and cheap and non-toxic photo-catalyst is an attractive alterna-
pesticides used in farming with subsequent run-off to catch- tive and is aimed at restoration of the environment. Various
ments areas lead to organic contamination of soil and water. photo-catalysts such as TiO2 , ZnO, WO3 , CdS, ZnS, SnO2
Other common examples of pollution are accidental leakages and Fe3 O4 have been tried independently or in the mixed or
and spills and continuous discharge of waste by-products in metal-doped forms for this purpose [4,5]. The initial step in
effluent streams from essential industries. This water even- TiO2 photo-catalysed oxidation is believed to be the forma-
tually contaminates ground and surface water, which is the tion of both hydroxyl radicals and super oxide radical anions
primary source of human contact with toxic chemicals. In [2]:
order to tackle this problem, extensive researches are being hν
carried out to develop advanced physicochemical method- → e− + h +
TiO2 − (1)
ologies for the elimination of hazardous chemicals from
air, soil and water. Purification procedures based on photo- h+ + H2 O → OH + H+ (2)
chemical reactions are potentially useful for this purpose. In presence of air,
Photochemical means to destroy organic pollutants through
oxidation as well as to convert them to useful material by e− + O2 → O2 − (3)
means of reduction are the best ways to tackle this problem.
Thus two species, OH (E◦ = 2.72 V) [8] a strong oxidant
and O2 − (E◦ = −0.33 V) [9] a weak reductant, are gener-
∗ Corresponding author. Fax: +91-22-25505151. ated from the above photo-catalysed oxidation. In the ab-
E-mail address: grdey@magnum.barc.ernet.in (G.R. Dey). sence of oxygen, OH radical, a strong oxidant, and e− , a

1010-6030/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochem.2004.01.022
504 G.R. Dey et al. / Journal of Photochemistry and Photobiology A: Chemistry 163 (2004) 503–508

strong reductant, are the reactive species generated on pho- The results show methane can be produced under these
tolysis of TiO2 in aqueous medium [10–12]. In the presence experimental conditions with good yield.
of an OH or an h+ scavenger, it is possible to use TiO2 for
photo-reduction processes.
Conversion of CO2 into methane or other organic sub- 2. Experimental
stances is of importance not only for the development of
alternative fuel sources, but also for the prevention of rise TiO2 (anatase, −325 mesh) was purchased from Aldrich
in temperature due to green house effect. Photo-catalytic re- and used as received. A quartz cell of 16.5 ml capacity hav-
duction of CO2 using photosensitive semiconductor powders ing a gas purging inlet and a gas sampling port was used for
has been studied in aqueous solutions [13], liquid CO2 [14] photo-irradiation with 350 nm light using Rayonet photore-
and high pressure CO2 [15]. Electrochemical reduction of actor having photon flux 4 × 1015 photons cm−2 s−1 . The
CO2 has also been extensively studied using various kinds amount of solution used for photolysis experiments was
of metals as cathode, producing CO or formic acid as the 5 ml. The ambient gas after photolysis of sample was anal-
main product depending on the kind of cathode materials ysed by gas chromatography (GC) using a poropak Q-S col-
[16,17] and the route for the electrochemical reduction reac- umn of 2 m length and diameter 1/8 in. Helium was used
tion has been established. Kaneco et al. [15] have reported as a carrier gas. Five hundred microlitres of gaseous sam-
that CO2 reduction does not take place in the absence of ple was injected into the GC for the analysis of CH4 and
electron donors. They tried 2-propanol as a hole scavenger CO2 . Thermal conductivity detector (TCD) and flame ion-
during photo-catalytic reduction of high pressure CO2 us- ization detector (FID) were used for analyzing the gaseous
ing TiO2 suspensions in water and reported the formation samples. Gases such as CO2 , N2 and O2 used in the exper-
of methane. In the work on photo-reduction of liquid CO2 , iments were of the best quality available from Indian oxy-
no gaseous products were observed; only formic acid was gen. The purged/saturated systems were photo-irradiated at
reported as a reduction product [14]. Methanol, formalde- normal atmospheric pressure and at room temperature. The
hyde and traces of formic acid in liquid phase and CO in the estimated values of methane yield have an error of ±10%.
gas phase were reported as the products for photo-catalysed
reduction of CO2 in the TiO2 suspension containing copper
powder [18]. Low yield of methane was observed after a long 3. Results
illumination time when KHCO3 was used in the system.
We have studied the photo-reduction of CO2 , and The yield of methane obtained from the photolysis of
methanol in the presence of TiO2 semiconductor in aqueous different solutions containing various concentrations of
medium at room temperature under normal pressure condi- 2-propanol in the presence of 0.1% (w/v) TiO2 is given
tions and observed the formation of methane in presence in Table 1. The amount of solution used for photolysis
of 2-propanol. The main objective of this work is to elu- experiments was 5 ml and it contained 5 mg of TiO2 . In
cidate the mechanism of CO2 methanation in the presence the CO2 -saturated system, the yield of methane observed
of 2-propanol as a hole scavenger at ambient conditions after 5 h photolysis was 5.4 × 10−8 mol and was found to
and also to compare the results with deoxygenated and increase with the time of photolysis as shown in Fig. 1.
oxygenated systems where methane was also produced. It may be mentioned here that this yield of methane was

Table 1
The yield of methane obtained from photolysis of 5 ml solutions containing 0.1% TiO2 (w/v) in the presence and absence of 2-propanol
Systems Photolysis time (h) Yield of CH4 and CO2 (×108 mol−1 ) from a solution containing 2-propanol

0M 0.05 M 0.5 M

CH4 CO2 CH4 CO2 CH4 CO2

CO2 -saturated 0 0 – 0 – 0 –
5 2.2 – 2.1 – 5.4 –
10 12.2 – 11.7 – 29.7 –
N2 -saturated 0 0 0 0 0 0 0
5 0 0 0 0 2.2 20
10 0 0 2.3 20 11.3 50
Aerated 0 0 0 0 0 0 0
5 0 0 12.4 2930 22.1 3150
10 0 0 11.3 4350 31.5 4650
O2 -saturated 0 0 0 0 0 0 0
5 0 0 10.2 3740 11.3 4000
10 0 0 13.2 6500 15.3 5870
 G.R. Dey et al. / Journal of Photochemistry and Photobiology A: Chemistry 163 (2004) 503–508 505

60 50

40
Methane Yield (x10 mol )
yield of methane (x10 mol )

-1
-1

40
30
8
8

20

20

10

0
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
0 4 8 12 16 20
Photolysis time/hr [TiO2] in % w/v

Fig. 1. Plot of yield of methane vs. photolysis time obtained for CO2 - Fig. 2. The yield of methane obtained from various concentrations of
saturated 2-propanol solutions containing 0.1% TiO2 . TiO2 in CO2 -saturated system containing 0.5 M 2-propanol after 10 h
photolysis.

obtained by using 5 mg of TiO2 in 5 ml of solution and

extrapolated methane per gram of TiO2 will be 200 times tion of TiO2 as suspension. The results obtained are shown
this value. In the N2 -purged system, methane yield was in Fig. 2. It was found that at ≥0.05% TiO2 concentration,
quite low even after 10 h of photolysis whereas in the aer- the methane yield reached an optimum value. Earlier, most
ated system the yield of methane was higher as compared of the researchers have used 0.5% or higher concentration
to that in N2 -purged solution and is comparable to that in of the semiconductor [13,21,22]. Using a high concentration
CO2 -saturated system. Methane yield was also found to of TiO2 makes the solution more opaque and the fraction
increase with increase in concentration of 2-propanol in the of light absorbed by the system may actually decrease with
presence of the photo-catalyst under identical conditions. increasing concentration of TiO2 leading to a lower yield
It can be seen that the yield of methane more than dou- of CH4 . In our experiment we have used only 0.1% TiO2
bled when 2-propanol concentration was increased from (w/v) and it has given much higher values of CH4 yield
0.05 to 0.5 M. No methane was obtained on photolysis from CO2 as compared to those reported earlier [15]. The
of a CO2 -saturated solution containing 0.5 M 2-propanol yield of methane without the hole scavenger [15,19] has not
in the absence of TiO2 photo-catalyst. Hence, methane been measured earlier but under this condition a measurable
formation is not due to any thermal decomposition or amount of methane was found.
photolytic degradation of 2-propanol. It is mainly due to Methanol, formic acid and formaldehyde are well known
the photo-catalytic reduction of carbon dioxide over TiO2 . products formed during CO2 reduction process [13,21–25].
2-Propanol acts as a positive hole scavenger [15,19] and Methanol is an intermediate for methane formation from
reacts with the positive hole ion (an oxidizing species) pro- CO2 as per the mechanism (CO2 –HCOO− /HCOOH–
ducing a reducing radical [(CH3 )2 C• OH]. The hole (h+ ) CH3 OH–CH4 ) as reported elsewhere [18,23]. If it is so
and OH radical have similar properties. The rate of reaction under the present system, it should be possible to produce
of 2-propanol with the hydroxyl radical has been reported methane by two-electron reduction of methanol also. Hence,
[20] to be 1 × 109 dm3 mol−1 s−1 in aqueous system. The we tried to reduce methanol by the photo-catalytic reduc-
low yield of methane found in N2 -purged system has been tion route in the presence of TiO2 . The results obtained
explained earlier [13,21,22] as due to the reduction of dis- are listed in Table 2. As seen from this table, no methane
solved CO2 which is present in water even in a deaerated was produced in the absence of 2-propanol. Even in the
system. CO2 produced from degradation of 2-propanol can presence of 2-propanol the yield was low and almost con-
also contribute to this photo-reduction. stant for different concentrations of methanol. It is known
To evaluate the optimum concentration of TiO2 for that 2-propanol is a better hole scavenger than methanol
photo-reduction of CO2 , photolysis experiments were [15]. This may be the possible reason why no methane was
carried out in CO2 -saturated systems containing 0.5 M produced in the presence of methanol alone. These results
2-propanol in aqueous solution and varying the concentra- probably indicate that methanol may not be an intermediate
506 G.R. Dey et al. / Journal of Photochemistry and Photobiology A: Chemistry 163 (2004) 503–508

Table 2
The yield of methane with respect to photolysis time from N2 -purged system containing various concentrations of methanol and 2-propanol in presence
of 0.1% TiO2
Systems containing Photolysis time (h) Photolysis yield (×108 mol−1 )

Methanol (M) 2-Propanol (M) [CH4 ] [CO2 ]

0 0 0 0 0
0.5 0 10 0 63.2
0.005 0.5 10 7.7 76.5
0.05 0.5 10 6.3 135.0
0.5 0.5 10 7.7 202.5

in the reduction of CO2 under this condition. Whatever methane depends on both the concentrations of oxygen and
yield of methane was obtained in this system may be com- the hole scavenger (2-propanol). This can be explained in
ing from the photo-reduction of dissolved CO2 or CO2 better way by the following proposed mechanism:
produced during photolysis of methanol and 2-propanol.
h+ + 2-propanol → organic radical + H+ (4)

4. Discussion organic radical + O2 → CO2 (mineralisation reaction)

(5)
It is clear from Table 1 and Fig. 2 that the amount of −
semiconductor catalyst and the ambient are very important CO2 + e → CH4 (methanation) (6)
for the reduction of CO2 . If we compare the yield of methane e− + O2 → O2 − (7)
in CO2 , N2 and aerated systems we find that a higher yield
of methane was obtained in both aerated and CO2 -saturated In the O2 -saturated system, where oxygen concentration
systems whereas in N2 -purged system the yield was lower. (1.2 × 10−3 mol dm−3 ) is five times higher as compared to
On photo-irradiation, hole (h+ ) and e− are generated as that in the aerated system, part of e− react with O2 rather
explained in reaction (1). The presence of 0.5 M 2-propanol than CO2 , thus reducing the yield of CH4 . In the aerated
is enough to scavenge h+ an oxidizing species, to produce a system although CO2 yield is lower than that in O2 -saturated
reducing species (reaction 4). In N2 -purged system, absence system, more e− are available for photo-reduction of CO2 .
of oxygen does not allow 2-propanol to mineralize to CO2 Further, if it can be assumed that the surface adsorption of
to a great extent as indicated by the small amount of CO2 CO2 is higher than that of O2 , then the surface-adsorbed
produced. The low yield of methane generated in this system CO2 has better chance of getting reduced to methane. In this
may be due to the photo-reduction of a small amount of CO2 system, it is difficult to differentiate surface-adsorbed CO2
produced by photolysis of 2-propanol. It is worth noting that from the in situ generated CO2 , possibly at the surface of
without 2-propanol as a hole scavenger methane was not the photo-catalyst.
produced in the N2 -purged system. This agrees well with Acetone [15], formed due to oxidation of 2-propanol was
the views reported earlier [15,19] about the requirement of observed in our experiments also. We also tried to esti-
a hole scavenger. mate propane as a photolysis product of 2-propanol under
The time profile obtained for the yield of methane these experimental conditions but its yield was very low
as shown in Fig. 1 indicates that the methane yield in both in aerated and O2 -saturated systems. Thus, methane
CO2 -saturated system increases with the photolysis time af- may not be formed directly on photolysis of 2-propanol, but
ter a small induction period. With increasing light exposure may be a product of photo-reduction of CO2 produced from
time, more and more photo-electrons are generated result- 2-propanol. Photolysis of 2-propanol system without TiO2
ing in increased reduction yield. The high yield of methane semiconductor did not yield any methane.
in the aerated system was unexpected. However, this can The yields of methane given in Tables 1 and 2 and Figs. 1
be explained as due to greater mineralization of 2-propanol and 2 were obtained from 5 ml samples containing 0.005 g
to CO2 by photo-oxidation in the presence of O2 . The min- of TiO2 as suspension. These values can also be extrapo-
eralization of the alcohols has also been discussed earlier lated to yield per gram of TiO2 per hour as in other reports
[26,27] and the process enhances with the modification of available in the literature. For example 29.7 × 10−8 mol of
photo-catalysts. Thus, CO2 generated in situ, on the surface methane obtained in the CO2 -saturated system after 10 h
of the semiconductor has an important role in the formation photolysis (see Table 1) is equivalent to 6 ␮mol h−1 g−1 af-
of methane under this condition. This is also supported by ter extrapolation. Hence, the yield of methane per gram of
the results obtained in the case of an oxygenated system TiO2 per hour of photolysis is quite high in comparison to
where the concentration of oxygen is expected to be five that reported even for high pressure CO2 system [15]. This
times higher and the methane yield was lower. The yield of may be due to the fact that when one employs 0.5% or more
 G.R. Dey et al. / Journal of Photochemistry and Photobiology A: Chemistry 163 (2004) 503–508 507

TiO2 (w/v), less light is absorbed by the system due to the semiconductor. In the oxygen-saturated system, the yield of
opacity leading to low methane yield as observed earlier methane was lower due to competition for e− by both O2
[15]. and CO2 .
Polychromic light sources, such as Xe arc lamps, medium
pressure mercury lamps and blacklight bulbs were used
in most of the photolysis experiments in the presence Acknowledgements
of the semiconductor catalyst. In such systems quantum
efficiency or photonic efficiency cannot be calculated The authors are thankful to the referees for their useful
because of scattering/reflection of the incident light by suggestions. The authors also thank to Dr. D.B. Naik for
TiO2 semiconductor particles [28]. Hence, in our stud- useful discussions, Dr. T. Mukherjee, Head RC & CDD and
ies we do not report the quantum yield of the reduction Dr. N.M. Gupta, Head ApCD for their encouragement.
products.
CO2 has a solubility of about 90 cm3 per 100 ml of water.
In aqueous solution, it exists in many forms [29,30]. First, References
it simply dissolves,
[1] R. Memming, Photoelectrochemical solar energy conversion,
CO2 (gas) → CO2 (aq) (8) Top. Curr. Chem. 143 (1988) 79.
[2] D.W. Bahnemann, in: E. Pelizzetti, M. Schiavello (Eds.),
Then, equilibrium is established between the dissolved CO2 Photochemical Conversion and Storage of Solar Energy, Kluwer
Academic Publishers, Dordrecht, 1991, p. 251.
and H2 CO3 , carbonic acid
[3] G. Mills, M.R. Hoffmann, Environ. Sci. Technol. 27 (1993) 1681.
CO2 (aq) + H2 O(l)
H2 CO3 (aq) (9) [4] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem.
Rev. 95 (1995) 69.
Only about 1% of the dissolved CO2 exists as H2 CO3 a [5] B.C. Faust, M.R. Hoffmann, D.W. Bahnemann, J. Phys. Chem. 93
(1989) 6371.
weak acid, which dissociates in two steps, [6] Y. Mao, C. Schoneich, K.-D. Asmus, J. Phys. Chem. 95 (1991)
H2 CO3
H+ + HCO3 − (10) 10080.
[7] K. Tanaka, T. Hisanaga, A.P. Rivera, in: H. Al-Ekabi, D.F. Ollis
HCO3 −
H+ + CO3 − (11) (Eds.), Photocatalytic Purification and Treatment of Water and Air,
Elsevier, Amsterdam, 1993, p.169.
H2 CO3 (aq) + CO3 2− (aq)
2HCO3 − (aq) (12) [8] J. Rabani, M.S. Metheson, J. Am. Chem. Soc. 80 (1964) 3175;
Y. Tabata, in: Pulse Radiolysis, CRC Press, Boca Raton, Ann
Arbar-Botson, 1991, p. 399.
Our experimental results especially in the aerated system
[9] B.H.J. Bieski, D.E. Cabelli, R.L. Arudi, A.B. Ross, J. Phys. Chem.
indicate that the methanation of CO2 is occurring on the Ref. Data 14 (1985) 1041.
surface of semiconductor catalyst but not through methanol [10] M. Anpo, Res. Chem. Intermed. 11 (1989) 67.
pathway. Also, HCO3 − , CO3 2− , though present in the [11] C.D. Jaeger, A.J. Bard, J. Phys. Chem. 83 (1979) 3146.
CO2 -saturated system, do not take part in the formation of [12] J.R. Harbour, M.L. Hair, J. Phys. Chem. 83 (1979) 652.
[13] T. Inoue, A. Fujishima, S. Konishi, K. Honda, Nature 277 (1979)
methane under the experimental conditions as shown by ex-
637.
periments conducted with these systems. The above results [14] S. Kaneco, H. Kurimoto, K. Ohta, T. Mizuno, A. Saji, J. Photochem.
agree with the mechanism reported [15] for the methanation Photobiol., A: Chem. 109 (1997) 59, and references therein.
process for high pressure CO2 where the authors explained [15] S. Kaneco, Y. Shimizu, K. Ohta, T. Mizuno, J. Photochem. Photobiol.,
the methane formation through C• radical generation mech- A: Chem. 115 (1998) 223.
[16] H. Noda, S. Ikeda, Y. Oda, K. Imai, M. Maeda, K. Ito, Bull. Chem.
anism and also with the mechanism suggested by Anpo Soc. Jpn. 63 (1990) 2459.
et al. [31] where the authors used different types of active ti- [17] R.G. Russel, N. Novac, S. Srinivasan, M. Steinberg, J. Electrochem.
tanium oxide to reduce CO2 with H2 O by photo-irradiation. Soc. 124 (1977) 1329.
[18] K. Hirano, K. Inoue, T. Yatsu, J. Photochem. Photobiol., A: Chem.
64 (1992) 255.
[19] W. Choi, M.R. Hoffmann, Environ. Sci. Technol. 31 (1997) 89.
5. Conclusions [20] G.V. Buxton, C.L. Greenstock, W.P. Helman, A.B. Ross, J. Phys.
Chem. Ref. Data 17 (1988) 513–886.
Photo-catalytic reduction of CO2 using TiO2 suspension [21] O. Ishitani, C. Inoue, Y. Suzuki, T. Ibusuki, J. Photochem. Photobiol.,
in aqueous solutions containing 2-propanol as a hole scav- A: Chem. 72 (1993) 269.
[22] T. Mizuno, K. Adhachi, K. Ohta, A. Saji, J. Photochem. Photobiol.,
enger leads to the formation of methane. CO2 reduction is A: Chem. 98 (1996) 87.
not possible without light or without a positive hole scav- [23] K. Tennakone, A.H. Jayatissa, S. Punchihewa, J. Photochem.
enger such as 2-propanol. CO2 reduction on TiO2 was found Photobiol., A: Chem. 49 (1989) 369.
to be very less in the absence of a hole scavenger. Methane [24] I.-H. Tseng, W.-C. Chang, J.C.S. Wu, Appl. Catal. B: Environ. 37
was also produced on photolysis of aerated 2-propanol so- (2002) 37.
[25] S. Kuwabata, H. Uchida, A. Ogawa, S. Hirao, H. Yoneyama, J.
lutions containing TiO2 . The methane formed in this system Chem. Soc. Chem. Commun. (1995) 829.
could be explained as due to the production of CO2 (pro- [26] J. Chen, D.F. Ollis, W.H. Rulkens, H. Bruning, Water Res. 33 (1999)
duced from 2-propanol mineralization) on the surface of the 1173.
508 G.R. Dey et al. / Journal of Photochemistry and Photobiology A: Chemistry 163 (2004) 503–508

[27] M. Harada, M. Honda, H. Yamashita, M. Anpo, Res. Chem. Intermed. [30] F.A. Cotton, G. Wilkinson, in: Advanced Inorganic Chemistry: A
25 (1999) 757. Comprehensive Test, third ed., Wiley Eastern Limited, New Delhi,
[28] A. Mills, S. Le Hunte, J. Photochem. Photobiol., A: Chem. 108 1984, p. 296.
(1997) 1–35. [31] M. Anpo, H. Yamashita, Y. Ichihashi, S. Ehara, J. Electroanal. Chem.
[29] R.C. Weast, CRC Handbook of Chemistry and Physics, Boca Rotan, 396 (1995) 21.
1979–1980, p. B-85.