High Pressure1

8/2/2019 High Pressure1

1/5

ELSEVIER Journal of Photochemistry and Photobiology A: Che mistry 109 ( 1997) 59-63

Photocatalytic reduction of CO2 using TiO, powders in liquid CO,mediumSatoshiKaneco **, Hidekazu Kurimoto a, Kiyohisa Ohta a,Takayuki Mizuno , Akira Saji b

* Department of Chemistry for Materials, FaculQ of Engineering. Mie University, T& Mie S/4. Japan Electric Power R&D Center, Chubu Electric Power Co. Ltd. , 20-l Ki tasekiyama, Odaka-cho, Midori -ku, Nu,yow 459, Japan

Received 28 January 1997; accepted 12 March 1997

AbstractThe photocatalyt ic reduct ion of CO1 was invest igated using TiOZ powders in l iquid CO, medium . The TiOZ powders with l iquid CO,

medium were i l luminated in a stainless steel vessel. After reducing CO? pressure to the ordinary state, purified water was added to the vesselcontaining the TiOZ powders without air contaminat ion. No gaseous reduct ion products were observed and formic acid was exclusivelyobtained in the aqueous so lution. It see ms that formic acid is produced through the protonation of the reaction intermediate s on the TiOlpowders with purif ied water. 0 1997 Elsevier Science S.A.Ke,words: Photocatalyt ic reduct ion: Carbon dioxide; Titanium dioxide; Liquid CO, medium

1. IntroductionPhotocata lytic reduction of CO, using photosensitive

semiconductor powders has been widely studied in aqueoussolutions [ l-201. So far, the studies have been performedexclus ively at ordinary temperature and pressure. In theseconditions, however, the concentration of CO, in water isvery smal l because of its low solubilit y [ 21,221 and, further-more, photocatalytic reduction of CO2 is competitive with Hzformation via water reduction. Therefore, selective reductionof CO2 is one of the major problems for the photocatalyticreduction of CO2 in aqueous solutions. Increase in CO* pres-sure is one of the measures for increasing the concentrationof CO* and improving the CO, reduction selectiv ity [ 23.241.At the ultimate, liquid CO, can be used as a medium. LiquidCO,, however, is non-polar and non-electroconductive. andhence the addition of supporting salts to liquid CO, needs forelectrochemical or photoelectrochemical reduction [ 25-271.On the other hand, these properties should be of no conse-quence in the photocatalytic reduction of CO2 using thesemiconductor powders. But, little information on thephotoreduction of CO, using the semiconductor powders inliquid CO? medium has been presented.

* Co rresponding author. Tel: + 81 592 3 I 942 7; fax : + 8 1 592 3 1 9442,9471 or 9427; e-mail: [email protected]. jp1010.6030 /97/$17.00 % 1997 Elsevier Science S.A. All rights reservedPUSlOlO-6030(97)00107-X

Recently, in many studies for the photocatalytic reductionof CO,, TiO, has been shown to be one of the effectivecatalysts for the reduction by illumination with light [ 5-201.Therefore, it has been found that the photocatalytic effect ofTiO, on reduction of CO1 is of great interest.

In this work. we present a photocatalytic reduction of CO,using TiO, powders in only liquid CO2 medium. Moreover,the mechanism of the photoreduction of CO2 in liquid CO,medium is discussed. based on the characterizations of theTiO, powders by electron spin resonance (ESR) after theend of illumination.

2. Experimental2. I . Procedures

The photocatalytic reduction was carried out in a stain lesssteel vessel (inner volume, 57.5 ml) with a window throughwhich T iO, powders in liquid CO2 medium was illuminated,as shown in Fig. 1. A commercially available pressure glassdevice with 21 mm diameter aperture (KLINPORT KPT -ClQ, Nihon Klingage CO. Ltd.) was f i t ted in the window,because the window glass transmitted almost all the light raysof wavelength above 340 nm and energy of 3.6 eV, which is


2/5

60

Gas adepvlicegWater inletc;l I Pressure gauge

S&g barFig. 1. Stainless steel vessel for photocatalytic reduction of CO, under highco2 pressure.

higher than the bandgap energy of TiO, (3.0 eV) The innersurface was completely covered by Teflon to preventcontam-ination from the stainless steel and its catalytic effect on CO,reduction. The vessel was placed in a water bath at 293 K.TiO, (Wako Junyaku, anatase; specific surface, 8.7 mg -; diameter, 230 nm; purity. 99.9%) was pretreated byboiling in 1 M ni tric acid and then thoroughly rinsed withdistilled-deionized water prior to use. The TiOZ powders (50mg) were placed in the stainless steel vessel. After deoxy-genation by flowing highly purity CO, (99.9999%) throughthe vesse l for 30 min. the vessel was closed tigh tly and thenthe CO> pressure was increased to 6.5 MPa. The TiOz pow-ders were continuously dispersed in liquid CO2 medium bya magnetic stirrer during illumination with a Xe lamp (990W, Ushio Electronics Co.) through the window. The lightintensity was 0.96 kW m-. After the end of illumination(30 h), CO, pressure was reduced to the ordinary state. Tothe vessel containing the TiOz powders was added degassed-purified water (5 ml) without air contamination to protonatethe reaction intermediates on the TiO, powders. Gaseousreduction products were sampled through a sampling valveand analyzed by FID and TCD gas chromatography. Theanalysis of gaseous products was performed immediatelyafter illumination and the addition of purified water, respec-tively. The liquid sample was analyzed by high performanceliquid chromatography with a UV detector.2.2. ESR analysis

After illumination, CO, pressure was reduced to ambientstate and the TiO, powders were transferred to an ESR quartzglass tube in an atmosphere of CO2 without air contamination.Then, ESR spectra of the TiO, powders were measured at 77K with a JES-TE200 (JEOL, X-band) spectrometer with 100kHz held modulation and a low temperature accessory. Theg values were measured relative to a DPPH sample (2.0036).

3. Results and discussionIn the beginning, we tried to photocata lytically reduce CO,

on TiO, powders in liquid CO1 medium without any protonsource, because little water can be disso lved into liquid CO21281. After illumination . no reduction products were identi-fied in vaporized CO,. However. since any reduced productscould be adsorbed on the TiO- powders illuminated, the TiO zpowders were washed with a smal l amount of purified waterwithout air contamination after pressure reduction to the ordi-nary state. Consequently, it was found that gaseous reductionproducts were not obtained and formic acid was exc lusivelyproduced in the water. Th is suggested that reaction interme-diates on the surface of TiO-, might react with the water toform formic acid. Hence, we made it our procedures to reducephotocatalytically CO, in liquid CO, medium through twosteps, i.e. illumination of the TiO, powders and then theaddition of purified water. When the same procedures wereperformed in highly purity nitrogen medium without CO2 asa reference. no reduction products were detected. Also, whenthe dark reduction of CO2 was performed in liquid CO?medium, CO? was not reduced. From the resul ts, it was con-cluded that formic acid was produced by the photocatalyticreduction of CO,.

The infuence of temperature and pressure on the yield offormic acid was studied for 30 h illuminations. Accordingly.illumination of the TiO, powders (SO mg) was carried out atvarious temperatures (273. 278, 283, 288, 293 and 298 K)and pressures (5.0,6.0.6.5.7.0 and 8.0 MPa) However. theyield of formic acid was almost constant and was not affectedby temperature and pressure. The density of liqu id CO, cannot almost change in these conditions [ 291. Thus, the resultsmay indicate that the interaction of CO2 with the surfacespecies of the TiO, powders or the catalytic activity of TiO-.can not be effected by temperature and pressure. Conse-quently. a ll subsequent illuminations were made with liqu idCO? medium of 293 K and 6.5 MPa, because of the repro-ducibility for the production of formic acid.

In order to select optimal illumination time, the effect ofillumination time on the yield of formic acid was investigated.The results are shown in Fig. 2. The yield of formic acidincreased sharply with illumination time until 5 h. After thetime, the production yield was suppressed gradually until 30h. During this period, it was observed that the dispersed TiO,particles began to coagulate and the state of dispersionbecame worse. Therefore, the suppression of the productionyield after 5 h might be attributed to deterioration of photo-catalytic activity due to coagulation of the TiO, particles. Theyield of formic acid reached a maximum value of 8.4 pmol(g-cat) for 30 h illuminations. This value was several timeslarger than those obtained by Ishitani et al. [ 141 and Hiranoet al. [ 151. Over 30 h, the yield of formic acid turned froman increase into a decrease. The reason for this turn is not


3/5

I approximately 0.4 m2. In excess of the area, the productionof formic acid plateaued. These phenomena may be due thatthe amount of TiO, in liquid CO, medium was too large forthe light to reach every particle.

Illumination time , hFig. 2. Effect of illumin ation time on the yield of formic acid. TiOZ, SO mg;temperature, 293 K; pressure, 6.5 MPa.

clear. It may be attributable to deterioration of photocatalyticactiv ity due to coagulation, diminishment of the adsorptionpower of the particles, saturation of the adsorption sites onthe surface of TiO, with intermediate products, and anincrease in the recombination of the intermediate productswith positive ho les on the surface of TiO,.3.2. E ffect of the sutjke area of TiO z

If the reaction intermediates from CO2 are accumulated onthe surface of TiO,, the yield of formic acid should increaselinearly with the total surface area of TiO,. In order to dem-onstrate the speculation, the effect of the total surface area ofTiO, on the yield of formic acid was evaluated. The resultsare shown in Fig. 3. Before measurements, it was confirmedby the BET method that the total surface area of Ti02 usedwas almost directly proportional to the amount of Ti02 up toabout 1 .O m*. The production of formic acid almost linear lyincreased with increasing the total sur face area of Ti02 up to

4- /d I igr= 2.015 72-

.;,,

0 t I0 0.3 0.6 O.!Total surface area o f Ti02 , m2

Fig. 3. Effect of total surface area of TiO, on the yield of formic acid.Illumina tion time, 30 h; temperature, 293 K; pressure, 6.5 MPa.

Fig. 4. ESR signals measured at 77 K with the TiO, powders after illumi -nation in l iquid CO, medium. TiOZ, 50 mg; i llumination time, 30 h; tem-perature, 293 K: pressure, 6.5 MPa.

1 gl= 1.9810 Amount of TiO, , mg50 50G

S. Kaneco et al . /Joumol ofPhotochemistry and Photobiology A: Chemistry 109 (1997) 59-63 61

From the data in the present study, we tried to estimate thedensity of the intermediate products on the surface of TiO,,assuming that all of the intermediate products were situatedon the surface of TiO, and protonated to formic acid withpurified water. The density was approximately 5.8 X lOImolecules of the intermediate products per 1 mm. The valuewas about a few tenths smaller than the density of the CO,monolayer [ 2 11. Therefore, the density was sparse and theintermediate products were considered to be situated on thespecific surface sites of TiOZ.3.3. ESR characterization

In order to cla rify a reaction scheme for the photoreductionof CO? on the TiO z powders in liquid CO, medium, westudied the ESR characterizations of the Ti02 powders. Fig. 4shows the ESR signals measured at 77 K with the Ti02 cat-alys t after the end of illumination. The ESR signals consistedtwo different radica l species. From the literature reported[ l&18-20,30-331, the signal with a g, value of 1.981 ca nbe attributed to the characteris tic photogenerated Ti3+ ions.Henglein et al. have recently proposed that in the two-electronreduction .CO; radicals formed in the uptake of the firstelectron were strongly adsorbed by ZnS particles [4].Yamanaka and coworkers descr ibed that . CO; radicals weredetected near a g value of 2.00 by ESR measurements[ 34.351. Therefore, another signal with a g, value of 2.015may probably consist of . CO; radicals. When illuminationof the TiO, powders was performed in high-purity nitrogenmedium without CO2 as a blank, this signal could not beobserved.


4/5

62 S. Kaneco Yt 01. /Journal qf Photochemistry and Photobiology A: Chemi.ct~ 109 (1997) X-6.3From the results in the present study and the literature [ 16-

20,30-33,3641], the formation process of formic acid byillumination of the Ti02 powders in liquid CO, medium andthe protonation with purified water is suggested to be asfollows:Illumination;

TiO, e-(Ti+) +hf[ (Ti4+ -(y-) 2 (Ti+-O-)*]

co2hv-3 CO; (ads)(TV+PO-)*

Protonation;El- +


5/5

S. Ku~eco et al . /Journal of Photochemistv and Photobiology A: Chemistry 109 (1997) 5943 631341 H. Kohno, C. Yamanaka, M. Ikeya, S. Ikeda, Y. Horino, Nucl. lnstrum.

Meth. Phys. Res. B91 (1994) 366.[ 351 M. Hirai, M. Ikeya, Y. Tsukamoto, C. Yamanaka, Jpn. J. Appl. Phys.

33 (1994) 1453.[36] M. Anpo, Res. Chem. Intermed. 11 (1989) 67.[ 371 J. Kiwi, J. Phys. Chem. 90 ( 1986) 1493.[38] G.T. Brown, J.R. Darwent, J. Phys. Chem. 88 ( 1984) 495 5.

[39] A. Wahl. M. Ulmann, A. Carroy, B. Jermann, M. Dolata. P.Kedzierzawski, C. Chatelain, A. Monnier, J. Augustynski, J.Electroanal. Chem. 396 ( 1995) 41.

[ 40 I A. Wahl, M. Ulmann, A. Carroy, J. Augustynski, J. Chem. Sot. Chem.Commun. ( 1994) 2271.

[41] G.R. Barnwends, S. Tsubota, T. Nakamura, M. Harata, J. Photochem.Photobio l. A: Chem. 89 (1995) 177.

Documents

High Pressure1