Bleaching of lignin solution by a photocatalyzed reaction on

Bleaching of lignin solution by a photocatalyzed reaction on semiconductor photocatalysts. Hideyuki Ohnishi, Michio Matsumura, Hiroshi Tsubomura, and ...
1 downloads 0 Views 750KB Size
719

I n d . E n g . C h e m . Res. 1989,28, 719-724 and Water. Makromol. Chem. 1960, 42, 52-67. Patterson, D. Free-Volume and Polymer Solubility. A Qualitative View. Macromolecules 1969,2, 672-677. Perry, R. H., Chilton, C. H., Eds. Chemical Engineers Handbook, 5th ed.; McGraw-Hill: New York, 1973. Sayegh, S, G.; Vera, J. H. Lattice Model Expressions for the Combinatorial Entropy of Liquid Mixtures: a Critical Discussion. Chem. Eng. J. 1980, 19, 1-10. Semlyen, J. A. Ring-Chain Equilibria and the Conformations of Polymer Chains. Adu. Polym. Sei. 1976, 21, 41-75.

Tiegs, D.; Gmehling, J.; Rasmussen, P.; Fredenslund, A. VaporLiquid Equilibria by UNIFAC Group Contribution. 4. Revision and Extension. Znd. Eng. Chem. Res. 1987, 26, 159-161. Truesdell, C. On a Function which Occurs in the Theory of the Structure of Polymers. Ann. Math. 1945,46,144-157. Van Krevelen, D. W.; Hoftijzer, P. J. Properties of Polymers, 2nd ed.; Elsevier: Amsterdam, 1976.

Received for review June 1, 1988 Accepted January 27, 1989

Bleaching of Lignin Solution by a Photocatalyzed Reaction on Semiconductor Photocatalysts Hideyuki Ohnishi,? Michio Matsumura,?Hiroshi Tsubomura,*stand Makoto Iwasakif Laboratory for Chemical Conversion of Solar Energy and Department of Chemistry, Faculty of Engineering Science, Osaka University, Toyonaka, Osaka 560, Japan, and Central Research Laboratory, Oji Paper Company, Ltd., Shinonome, Koto-ku, Tokyo 135, J a p a n

Lignin is oxidatively decomposed by a semiconductor-photocatalyzed reaction in aqueous solutions containing oxygen and semiconductor powder. Under continued illumination, the solution becomes colorless and the chemical oxygen demand (COD) value decreases, generating carbon dioxide and a small amount of carbon monoxide. TiOz, ZnO, and CdS have high photocatalytic activities in neutral solutions, and TiOz, ZnO, and W 0 3 also show high activities in alkaline solutions. Oxygen plays the role of accepting electrons from the conduction band of semiconductor particles and promotes the reaction. T h e catalytic efficiency is improved by loading TiOz with noble metals, such as P t , Ag, and Au, or by heat treatment under a hydrogen atmosphere. The mechanism of the photocatalytic reactions was discussed based on the experimental results, including those obtained from the electrochemical measurements of the semiconductor electrodes. Lignin occupies about one-quarter of dry wood and is the most important component of wood, next to cellulose. The utilization of lignin as a raw material for organic chemicals has been extensively studied, but so far without notable success. In pulp and paper mills, lignin is separated from cellulose, and most of the extracted lignin is utilized only as fuel. Lignin and its derivatives also cause the dark brown coloration of the pulp factory effluent. Its chemical or biological degradation is demanded for environmental reasons. New methods for the decolorization of lignin solutions have been developed, such as electrolysis (Oehr, 1978; Nassar et al., 1983) and ultrafiltration (Lundahl and Mansson, 1980). It has also been reported that lignin can be decomposed effectively by photochemical reactions under ultraviolet light in aqueous solutions containing hydrogen peroxide or ozone (Coburn et al., 1984). The semiconductor photocatalysts developed for the past 10 years are powerful tools for decomposing many organic materials (Hashimoto et al., 1984; Izumi et al., 1980) as well as harmful inorganic materials such as cyanides (Frank and Bard, 1977), sulfides (Borgarello et al., 1983), and sulfites (Matsumura et al., 1985; Buhler et al., 1984). The reactions are initiated by electrons and holes photogenerated in semiconductor particles (Pelizzetti and Serpone, 1986). The wavelength region available for the reactions is determined by the band gaps of the semiconductors and is usually wider than those for pure organic photochemical reactions or reactions employing hydrogen peroxide and chloride. This makes the reaction more efficient and even makes it possible to use solar energy in some cases. ?Osaka University. Oji Paper Company, Ltd.

In this paper, the decomposition of lignin by the semiconductor-photocatalyzed reactions is reported and the mechanism is discussed.

Experimental Section Preparation of the Photocatalysts. ZnO and TiOz were purchased from Kanto Chemical Co., Ltd., and W03, In203,Fe203, and CdS were purchased from Furuuchi Chemical Co., Ltd. Their particle sizes were about 0.2,0.1, 4, 1, 4, and 1 pm, respectively. They were used as photocatalysts without any further treatments unless otherwise stated. Some of the experiments were carried out using semiconductor powder loaded with noble metals by the photodeposition technique (Reiche et al., 1979) in solutions containing H2PtC1,, HAuC14,and AgN03, respectively, at the range of concentrations from 0.15 to 0.3 M, together with 0.4 M formaldehyde as the hole acceptor. The amount of the semiconductor powder in the solution was set so as to have 1.5 wt '70 metal loaded on the semiconductor. For the photodeposition of metals on semiconductor particles, the semiconductor powder was suspended in the solution, which was deaerated by bubbling with nitrogen for 30 min and subsequently irradiated for 1 h with a 500-W super-pressure Hg lamp (Wacom, BMD500D). Then the suspension was centrifuged, and the metal-loaded semiconductor powder was filtered, rinsed with water, and dried in an evacuated desiccator. Photocatalytic Reactions. Alkaline lignin (Tokyo Chemical Industry Co., Ltd.) was used as the reactant without further purification. By employing the gel-filtration chromatography, it was found that the molecular weight distribution of the lignin had a maximum at about 6000. In this work, the reactions were carried out in quartz vessels, containing a 10-mL aqueous solution of Alkaline

0888-5885/89/2628-0719$01.50/0 0 1989 American Chemical Society

720 Ind. Eng. Chem. Res., Vol. 28, No. 6, 1989

Lignin (0.1 g/L) and 50-mg photocatalyst unless otherwise stated. The solution had a pH between 6.8 and 7.0. The photocatalyst was ultrasonically suspended in the solution, and the solution was illuminated with light from a 500-W super-pressure mercury lamp through a water filter to remove the infrared portion. During illumination, the solution was magnetically stirred at a speed of about 350 rpm and bubbled with oxygen at a rate of about 10 mL/min. In some cases, the solution was bubbled with nitrogen or oxygen diluted with nitrogen. For the study of the photocatalytic reactions in alkaline solutions, 0.1 M sodium hydroxide solutions were used. The quantum efficiencies of the photocatalytic reactions were determined from the experiments performed under light monochromatized by a Japan-Jarrell-Ash monochromator, JE-25. The light intensity was measured with a thermopile (Eppley Laboratories). All other chemicals used were reagent grade and were used as obtained. Analysis of the Products. After illumination for a desired period, the suspension was centrifuged to separate the photocatalyst, and the absorption spectrum of the solution was measured in a quartz cell, 1-cm path length, with a Shimadzu UV-300 spectrophotometer. The change of the molecular weight distribution of lignin was measured with a high-performance liquid chromatography system (TOSOH, CCDP, and UV-800) equipped with a gel-filtration column, G300PWa. A mixed solution of 20% acetonitrile and a phosphate buffer solution (pH 6.8) was used as the eluting solvent. The gaseous products were analyzed by use of a Shimadzu GC-4A gas chromatograph equipped with a column packed with active carbon or molecular sieves, 13X. The COD (chemical oxygen demand) value was determined by the permanganate titration method according to JIS KO102 (Japanese Industrial Standards Committee, 1974). Electrochemical Measurements. The Ti02-sintered disks, 1-cm diameter, were prepared by compressing Ti02 powder at 15 kg cm-, and heating at 1300 "C for 6 h under a nitrogen atmosphere. The electrochemical properties of the Ti0,-sintered electrodes were measured in a threeelectrode cell consisting of a Ti02electrode, a Pt counter electrode, and an Ag/AgCl reference electrode. The potential of the TiO, electrode was controlled with a POtentiostat (Nikkokeisoku, NPGFZ-2501A). Metal loading of the TiO, electrodes was carried out by the photodeposition method in a solution similar to those used for the photodeposition of metals on the semiconductor powder.

Results An aqueous solution of lignin has an absorption peak a t about 210 nm and a shoulder at 275 nm, as shown by the top curve in Figure 1. The color of lignin arises from the absorption tailing to the long-wavelength region. When the lignin solution with suspended semiconductor powder was bubbled with oxygen and illuminated, the color faded gradually. As an example, the absorption decrease with the TiOz photocatalyst is shown in Figure 1. The absorbance of lignin decreased monotonically throughout this wavelength region, without other changes in the spectrum, and disappeared completely after 2 h. This suggests that not only the portion of the unsaturated chains included in the lignin molecule but also the aromatic rings themselves are decomposed by the photocatalytic reaction. The decolorization did not occur in the absence of the photocatalyst even by prolonged illumination. The rate of decolorization increased with the amount of the photocatalyst added to solution and approached a limiting

illuminctior

A

20C

30C

400

Wavelengtl?

/

Period /

min

503

nm

Figure 1. Absorption spectra of a 10-mL aqueous solution (pH 6.9) of lignin (0.1 g/L) containing 50 mg of TiOz photocatalyst, the illumination periods for curves from top to bottom being shown in the figure.

value when the amount of photocatalyst became large enough to absorb the incident light completely. No decolorization occurred in the dark in the suspension of semiconductor or metal powder. The photocatalytic decolorization was investigated in aqueous solutions of lignin (10 mL) with various semiconductors (50 mg). The amounts of the photocatalysts were large enough to absorb the incident light completely and saturate the reaction rate. the photocatalytic activity in the neutral lignin solution was the highest for ZnO and went down in the order ZnO > TiO, > CdS >> In203 > WO, Fe203,as seen in Figure 2a. Their activities were enhanced by making the solutions alkaline in all cases except for CdS and ZnO (Figure 2b). In the case of CdS, the activity was lowered by about 10 times in alkaline solutions. The rate of the decolorization decreased when the partial pressure of the oxygen bubbled into the solution was lowered by diluting it with nitrogen, as shown in Figure 3. Practically no decolorization occurred in the absence of oxygen. Even in the absence of oxygen, however, decolorization was observed when silver ions were present in the solution, as shown in Figure 4. In this case, the decolorization proceeded very fast for the initial few minutes and stopped in about 10 min. The semiconductor powder irradiated in the presence of silver ions turned black from the deposition of silver on semiconductor particles. The decay of the decolorization rate during illumination is explained by taking account of the lowering of the photocatalytic activity caused by too much metal loading (Uchihara et al., 1989b). The rate of bleaching lignin solutions bubbled with oxygen was enhanced by loading the semiconductor powder with noble metals. The change of the absorbance of the lignin solution with bare and noble metal (1.5 wt % ) loaded TiO, powder is shown in Figure 5. The effect becomes stronger in the order Au, Ag, and Pt. For the case of Ti02, heat treatment at 550 "C under a hydrogen stream also enhanced its activity. For the heat-treated photocatalyst, the rate of decolorization changed little with noble metals loading. The gaseous products formed during illumination were analyzed by use of a closed reaction vessel containing a 30-mL solution of lignin, 100 mg of the photocatalyst, and an oxygen atmosphere. The gas in the vessel was sampled periodically with a microsyringe through a septum attached to the vessel and analyzed with a gas chromatograph. The

-

Ind. Eng. Chem. Res., Vol. 28, No. 6, 1989 721

!

I

0

0,s

1.0 I ,5 Illumination Time / h

0 1 0' 2.0

I

I

I

I

I

1

2

3

4

5

Concentration o f Silver Yitrate /

Figure 4. Absorbance of the same lignin solution as that for Figure 3 after illumination for 2 min without oxygen bubbling in the presence of ZnO and TiOz photocatalysts against the concentration of silver nitrate added to the solution.

0,25

0,20

-E 2

0,15

-7

Y 0

E 0.10 0

f v)

Q

0.05

0

0.5

0

1.0

1.5

2.0

Illumlnatlon Time / h

Figure 2. Change of the absorbance at 465 nm of the (a, top) neutral and (b, bottom) alkaline (0.1 M sodium hydroxide) solutions of lignin (0.1 g/L) by the photocatalytic reaction in the presence of various kinds of semiconductor powder (50 mg) against illumination time.

IO

0

20 Illumination Time / m l n

30

40

Figure 5. Absorbance of lignin solutions (the same as those for Figure 3) illuminated in the presence of TiOz powder and noble metal (1.5 wt %) loaded TiOz powder against the illumination time.

- 400

E

- 300 1 - 200

-= D

0 0

OJ

I

I

I

I

I

-

- 100

-

m-

10 _1

0

0.2 0.0 0.6 0.8 Partial Pressure o f Oxysen (Po2 / Po2 + P N 2 )

IO

Figure 3. Absorbance of the lignin solution after illumination for 30 min in the presence of ZnO photocatalyst and after illumination for 60 min in the presence of TiOz photocatalyst against the partial pressure of oxygen bubbled into the solutions. In this experiment, 100 mg of photocatalysts was added to the 30-mL lignin solutions (0.1 g/L, pH 6.9). "0

products obtained by the Ti02-photocatalyzed reaction were mainly carbon dioxide and a small amount of carbon monoxide, as shown in Figure 6b. When Pt-loaded TiOz was used as the photocatalyst, the evolution of carbon monoxide was negligible. Figure 6a shows the changes of the absorbance of the lignin solution and the COD value by the photocatalytic reaction measured under the same experimental condition

2

4 6 lllminatlon Time / h

8

Figure 6. (a) Change of absorbance and that of COD value of the lignin solution by illumination in the presence of TiOz photocatalyst and (b) gaseous products generated by illumination. In these measurements 100 mg of TiOz powder was added to 30-mL lignin solutions (0.5 g/L).

as above. By comparing these changes, it is seen that the rates of the evolution of carbon dioxide and carbon mon-

722 Ind. Eng. Chem. Res., Vol. 28, No. 6, 1989 Molecular io I

102 1

103 I

Weigh1

XI'

135

I I I I I , ~ : I C ~ I O -ZP'IK I

10' I

!r

?+

electroce

~

Figure 8. i-V curves of the Pt and TiOz electrodes measured in the solution of 0.5 M sodium sulfate buffered at pH 6.9 by addition of potassium dihydrogen phosphate and sodium hydroxide (solid lines) and those measured in the solution of 0.5 M sodium sulfate and 0.1 M sodium hydroxide (pH 12.7, broken lines). The solutions were bubbled with oxygen during measurements. L

-1.0 11.5

10.0

8.5

Elution

7,:

5.5

L O

Vglume / m t

I

-0.5 I

I

/ V

'4s.

AJ / AgCI

0

05

I

I

Figure 7. Elution curves of lignin solutions illuminated in the presence of TiOz photocatalyst measured with a gel-filtration chromatograph, the illumination periods being shown in the figure. Each of the lignin solutions has a volume of 10 mL and contains 0.5 g/L of lignin and 0.5 g of TiOz.

oxide agree well with the rates of decrease of the absorbance and the COD value. The total amount of carbon dioxide and carbon monoxide generated during the illumination for 8 h was in good agreement with the amount predicted from the decrease of the COD value. When ZnO was used as a photocatalyst, the COD value did not drop to zero even after illumination over 8 h, where the absorbance at the wavelength of 465 nm diminished almost completely. This suggests that a small amount of hydrocarbons remained unreacted in the decolored solution, their identification being unsuccessful. The quantum efficiencies of the generation of carbon dioxide and carbon monoxide from the lignin solution were measured under monochromatic irradiation (A 366 nm, photon flux 2.15 X 1015cm-2 s-l). The efficiencies for the generation of carbon dioxide and carbon monoxide by the Ti02photocatalyst in the neutral solution were 0.0285 and 0.0019 molecule/photon, respectively. The quantum efficiency of the generation of carbon dioxide by the platinum-loaded Ti02photocatalyst was nearly twice that of the nonloaded Ti02 photocatalyst. The change of the molecular weight distribution of lignin during illumination in the presence of the TiOz photocatalyst was studied with a gel-filtration chromatograph. The maximum of the curve decayed and shifted to the direction of smaller molecular weight as the reaction proceeded, and no peaks due to the units containing one or several benzene rings were observed (Figure 7 ) . This suggests that each of the lignin molecules is decomposed sequentially at the point of contact with the photocatalyst. The Ti02photocatalyst was very stable, and no evidence for its decomposition was observed by atomic absorption analysis of the test solutions. On the other hand, the ZnO and CdS photocatalysts were found to be decomposed a little when irradiated in the presence of oxygen. Finally, we studied the electrochemical properties of TiOpand metal-loaded Ti02 electrodes in order to clarify the mechanism of the photocatalysts. The reduction current of oxygen on the Ti02electrode appears at potentials more negative than on a Pt electrode in both

I

I I '

200t-

Figure 9. i-V curves of the TiOz and metal-loaded TiOz electrodes measured in the solution of 0.5 M sodium sulfate buffered at pH 6.9. The solution was bubbled with oxygen.

neutral and alkaline solutions, as shown in Figure 8. The current-potential curve of the Pt electrode shifts toward the negative at a rate of about -55 mV/pH, while that of the TiOz electrode shifts a t a smaller rate. The current-potential curves of the Ti02 and Pt-, Ag-, and Au-loaded Ti02 electrodes are shown in Figure 9. The onset potentials for the cathodic current to reduce oxygen shift to the positive in the order of bare to Au-, Ag-, Ptloaded electrodes. Discussion It is well-known that semiconductor-photocatalyzed reactions are initiated by either electrons or holes photogenerated in the conduction band (CB) or the valence band (VB) of the semiconductor particles, eq 1 (Griitzel, 1983; Pelizzetti and Serpone, 1986). The electrons in the CB are transferred to an electron-accepting substance (A) in the solution and reduce it to A- as expressed by eq 2, while holes (h) in the VB oxidize electron-donating substances (D) in the solution as expressed by eq 3. The electrons and holes can also recombine, producing either heat or luminescence (eq 4). The whole reaction scheme can therefore be expressed by the following series of equations: photon(semiconductor) e(CB) + h(VB) (1) e(CB) + A A(2) h(VB) + D D+ (3) e(CB) + h(VB) heat or luminescence (4)

--

Ind. Eng. Chem. Res., Vol. 28, No. 6, 1989 723 n- t y p e Semi conauctor

Solution A

Figure 10. Band structure of an n-type semiconductor particle causing photocatalytic reactions.

For the case of n-type semiconductors in the dark, the electron-transfer equilibrium will be attained between the semiconductor particle and the solution in such a way that a positive space-charge layer is formed in the semiconductor and the band bending will be produced, as shown in Figure 10. The presence of such a band bending is favorable for the reaction of h(VB), because the hole will be attracted toward the surface by the electric force and reacts with the donor a t the surface (Uchihara et al., 1989b). For the electrons in the CB, this means that they should surmount a barrier to reach the surface and react with the electron acceptors in the solutions. However, it is empirically known that electrons in the CB can surmount such a barrier fairly efficiently, if the barrier is of the order of 0.3 eV or less (Matsumura et al., 1989). In the previous sections, we have shown that lignin is oxidatively decomposed by the photocatalytic reactions. Three possible reaction paths can be considered: (1) Lignin is directly oxidized by h(VB). (2) Water or hydroxide anion (OH-) is oxidized by h(VB), and it in turn oxidize lignin. (3) Oxygen is reduced by e(CB) and the superoxide anion formed will initiate the oxidation of lignin. The third postulate may be discarded taking as fact that the reaction proceeds without oxygen if only silver ions are present in the solution. It is most probable that oxygen or a silver ion plays the role of removing e(CB) efficiently and maintaining the band bending mentioned above in the working stage of the photocatalyst. Without these electron acceptors, the bands are soon flattened, most of the photogenerated electrons and holes will be lost by their recombination (eq 4), and the efficiency of the reactions becomes very low (Uchihara et al., 1989a). The differences in the bleaching efficiency for various semiconductor particles as shown in Figure 2 can be explained by taking account of the size of the band gap and the position of the conduction band in the semiconductor (Sculfort and Gautron, 1984). Generally speaking, semiconductors having large band gaps have strong photocatalytic activities. In the present case, Ti02and ZnO have band gaps larger than 3 eV and show strong activities. CdS, having a smaller band gap, shows less activity. The remaining semiconductors, In203,W03, and Fe203,have band gaps nearly as large as that of CdS but show very weak activities. Since their CBs are much lower than that of CdS, e(CB) in these semiconductors cannot move into the electron acceptor in the solution rapdily. It should be noted that the activity of the photocatalysts is also affected by the particle size, crystallinity, and concentration of impurities included in the catalysts.

Next, let us discuss the effect of pH on the reactivity. As Figure 2 indicates, the bleaching by the oxide photocatalysts is faster if the pH of the solution is high, except for the case of ZnO. It is well-known that the energy levels of oxide semiconductors move upward, i.e., toward the negative, by -59 mV/pH. This means that the reducing capability of electrons in the CB becomes stronger in alkaline solutions than in neutral solutions. However, since the thermodynamic potential for the reduction of oxygen into water also shifts to the negative at the same rate, there should be no merit for the reactivity by change of pH from the thermodynamic standpoint. In order to discuss the pH effect, we have investigated the properties of platinum and semiconductor electrodes on the reduction of oxygen. A platinum electrode has a high catalytic activity for oxygen reduction, and its onset potential for the oxygen-reducing current shows a -55 mV/pH dependence (Figure 8), in nearly good agreement with the thermodynamic reasonings. On the other hand, the oxygen-reducing current for the TiOz electrode in a neutral solution starts at a potential much more negative than that for a platinum electrode, indicating that it has a high activation energy for the oxygen reduction. The change of the onset potential for TiOz with the pH is much smaller than that for a Pt electrode, as shown in Figure 8. This means that the oxygen reduction in TiOz becomes easier in high pH solutions, thus indicating that bleaching with Ti02 is more efficient in alkaline solutions. The similar increase of efficiencies in other oxide semiconductor photocatalysts can be explained by the same reason. The slight lowering of the activity of ZnO photocatalyst in alkaline solutions is considered to be related to the pHdependent reactivity of h(CB) in ZnO (Lee et al., 1984); however, the exact reason is still not clear. For the case of CdS, the energy levels do not change with pH (Watanabe et al., 1974). It is thought, therefore, that the reduction of oxygen becomes more difficult in solutions of higher pH. This explains well that only for the case of CdS is bleaching slower by 1order of magnitude in alkaline solutions than in neutral solutions. The photocatalytic activity of TiOz is markedly increased by loading with noble metals, as shown in Figure 5. This effect is the highest for Pt, followed by Ag and Au. The onset potential for the oxygen reduction current in metal-loaded Ti02 electrodes also shifts to the positive in just the same order, Le., Pt, Ag, and Au, as seen in Figure 9. These results show clearly that the photocatalytic activity is improved by the electrocatalytic effect of metals to accelerate the oxygen reduction. The bleaching rate for the Ti02photocatalyst was also improved by heat-treating T i 0 2 powder in a hydrogen stream. It is well-known that donor levels are formed in high density by the partial reduction of Ti02 with such a treatment, resulting in an increase of the electron density in the conduction band. This is thought to increase the rate of electron transfer across the interface. It is interesting to note that the TiOz photocatalyst heated in hydrogen showed no effect of improvement by the metal loading. This can be explained by assuming that the rate of the oxygen reduction in a partially reduced TiOz.powder is already fast enough compared with the oxidation processes on the same semiconductor particle. Conclusion It has been demonstrated that lignin dissolved in aqueous solutions is decomposed by the semiconductor photocatalysis efficiently in the presence of oxygen. The

Ind. Eng. Chem. Res. 1989,28, 724-728

724

results suggest a promising method for treating the effluent from pulp and paper mills. Acknowledgment The authors are grateful to Professor S. Yanagida of the Faculty of Engineering, Osaka University, for helpful discussions and for the use of the gel-filtration chromatograph. Registry No. Fe203, 1309-37-1; W03, 12036-22-5; In203, 1312-43-2; CdS, 1306-23-6; Ti02, 13463-67-7; ZnO, 1314-13-2; Pt, 7440-06-4; Ag, 7440-22-4; Au, 7440-57-5; COZ, 124-38-9; CO, 630-08-0; lignin, 9005-53-2.

Literature Cited Borgarello, E.; Erbs, W.; Gratzel, M.; Pelizzetti, E. Visible Ligh Induced Cleavage of Hydrogen Sulfide and Photosynthesis of Thiosulfate in CdS Dispersions. Nouv. J. Chim. 1983, 7, 195-198. Buhler, N.; Meier, K.; Reber, J.-F. Photochemical Hydrogen Production with Cadmium Sulfide Suspensions. J. Phys. Chem. 1984, 88, 3261-3268. Coburn, L. A,; Lockwood, M. P.; Englende, A. J., Jr.; Collins, T. Kraft Bleach Plant Effluent Treatment by Ultraviolet/Oxidation Process. Tappi Proceedings of the 1984 Environmental Conference, Savannah, GA 1984; pp 277-286. Frank, S. N.; Bard, A. J. Heterogeneous Photocatalytic Oxidation of Cyanide Ion in Aqueous Solutions at TiOp Powder. J . Am. Chem. SOC.1977, 99, 303-304. Gratzel, M. Energy Resources through Photochemistry and Catalysis; Academic Press: New York, 1983. Hashimoto, K.; Kawai, T.; Sakata, T. Photocatalytic Reactions of Hydrocarbons and Fossile Fuels with Water. Hydrogen Production and Oxidation. J. Phys. Chem. 1984, 88, 4083-4088. Izumi, I.; Dunn, W. W.; Wilbourn, K. 0.;Fan, F.-R. F.; Bard, A. J. Heterogeneous Photocatalytic Oxidation of Hydrocarbons on Platinized TiOz Powders. J. Phys. Chem. 1980, 84, 3207-3210. Japanese Industrial Standards Committee Testing Methods for Industrial Waste Water. JIS K 0102-1974, 1974; pp 26-28.

Lee, J.; Kato, T.; Fujishima, A.; Honda, K. Photoelectrochemical Oxidation of Alcohols on Polycrystalline Zinc Oxide. Bull. Chem. SOC. Jpn. 1984, 57, 1179-1183. Lundahl, H.; Mansson, I. Ultrsfiltration for Removing Color from Bleach Plant Effluent. Tappi 1980, 63, 97-101. Matsumura, M.; Furukawa, S.; Saho, Y.; Tsubomura, H. Cadmium Sulfide Photocatalyzed Hydrogen Production from Aqueous Solutions of Sulfite: Effect of Crystal Structure and Preparation Method of the Catalyst. J . Phys. Chem. 1985, 89, 1327-1329. Matsumura, M.; Uchihara, T.; Hanafusa, K.; Tsubomura, H. Interfacial Band Structure of Platinum-Loaded CdS Powder and Its correlation with the Photocatalytic Activity. J . Electrochem. SOC. 1989, in press. Nassar, M. M.; Fadali, 0. A.; Sedahmed, G. H. Decolorization of Pulp Mill Bleaching Effluents by Electrochemical Oxidation. Pulp Paper Can. 1983,84, T275-278. Oehr, K. Electrochemical Decolorization of Kraft Mill Effluents. J . Water Poll. Control Fed. 1978, 50, 286-289. Pelizzetti, E., Serpone, N., Eds. Homogeneous and Heterogeneous Photocatalysis; Reidel: Dordrecht, 1986. Reiche, H.; Dunn, W. W.; Bard, A. J. Heterogeneous Photocatalytic and Photosynthetic Deposition of Copper on TiOz and WOg Powders. J . Phys. Chem. 1979,83, 2248-2251. Sculfort, J.-L.; Gautron, J. The Role of the Anion Electronegativity in Semiconductor-Electrolyte and Semiconductor-Metal Junctions. J . Chem. Phys. 1984,80,3767-3773. Uchihara, T.; Matsumura, M.; Tsubomura, H. Effect of Dissolved Electron Acceptors and Platinum Loading on the Luminescence of CdS Powder in Aqueous Solutions. J . Phys. Chem. 1989a, in press. Uchihara, T.; Matsumura, M.; Yamamoto, A.; Tsubomura, H. Effect of Platinum-Loading on the Photocatalytic Activity and Luminescence of Cadmium Sulfide Powder. J.Phys. Chem. 1989b, submitted for publication. Watanabe, T.; Fujishima, A.; Honda, K. Potential Variation at the Semiconductor-Electrolyte Interface through a Change in pH of the Solution. Chem. Lett. 1974, 897-900.

Received for review May 25, 1988 Accepted February 15, 1989

Promotion of Iron-Group Catalysts by a Calcium Salt in Hydrogasification of Coal Chars T e t s u y a Haga* and Yoshiyuki Nishiyama Chemical Research Institute of Non-Aqueous Solutions, Tohoku University, Katahira, Sendai 980, J a p a n

+

The effects of Fe Ca and Ni + Ca dual catalysts on hydrogasification of coal char were studied between 1 and 30 atm and between 750 and 900 "C. Promotions of Fe- and Ni-catalyzed reactions by added Ca were observed in the reactions of all coal chars used. The degree of promotion increased with increases in the amount of added Ca, temperature, and pressure. T h e initial surface area of coal char was found to greatly affect the activities of dual catalysts, probably due to different dispersions of catalyst particles a t the stage of impregnation. The rate of Ni-catalyzed gasification increased with time during the isothermal mode, and the added Ca accelerated the rate increase. T h e results are discussed in comparison with our previous study using model carbons. It is known that the iron-group metals catalyze coal gasifications in steam, carbon dioxide, and hydrogen, and the alkali and alkaline-earth salts catalyze the gasifications in oxygen-containing gases (e.g., McKee (1981) and Pullen (1984)). Interestingly, several researchers have recently reported on the enhancement of catalyst activity by mixing different catalytic components (Rao et al., 1980; Suzuki et al., 1984; Carrazza et al., 1985; Huttinger and Minges, 1985; McKee et al., 1985) or by adding a foreign component *To whom all correspondence should be addressed. 0888-5885/89/2628-0724$01.50/0

to a catalyst, such as the addition of La203 (Inui et al., 1979,1985) or ZrOz (Ryabtchenko et al., 1987) to the Fegroup catalysts. We have previously reported that hydrogasification of carbons (pitch coke and activated carbon) catalyzed by the Fe-group metals is greatly promoted by the added Ca, which is not a catalyst for hydrogasification (Haga and Nishiyama, 1983a,b, 1987). It has recently been found that Fe and Ca also catalyze cooperatively in steam gasification of carbons (Haga et al., 1989). The above studies on mixed catalyst systems, involving our earlier works, were mostly carried out utilizing simpler

0 1989 American Chemical Society