RKCL4233 HETEROGENEOUS OXIDATION OF ...

Jointly published by Akadémiai Kiadó, Budapest and Kluwer Academic Publishers, Dordrecht

React.Kinet.Catal.Lett. Vol. 78, No. 2, 341-348 (2003)

RKCL4233 HETEROGENEOUS OXIDATION OF ETHANOL OVER Cu(OH)2/α-Fe2O3 UNDER VISIBLE LIGHT Ekaterina V. Kuznetsovaa, Evgueni N. Savinova, Alexandre V. Vorontsova and Panagiotis Smirniotisb a

Boreskov Institute of Catalysis, Novosibirsk 630090, Russia, [email protected] b Chemical Engineering Department, University of Cincinnati, Cincinnati, OH 45221-0171, USA Received October 28, 2002 Accepted November 7, 2002

Abstract A Cu(OH)2/α-Fe2O3 photocatalyst is shown to be active in the gas phase oxidation of ethanol under visible light. The calculated initial quantum efficiency of the ethanol photooxidation is 0.1–1%. However, ethanol is oxidized only into acetic acid, which deactivates the catalyst. Keywords: Photocatalysis, copper hydroxides, multielectron reactions, organic oxidation

INTRODUCTION Photocatalysts for deep oxidation with relatively high activity under the action of near ultraviolet (UV) and visible light attract considerable interest. Most photocatalytic studies are devoted to properties of titanium dioxide owing to its high activity, chemical stability under UV light, and to its environmental harmlessness. However, pure TiO2 does not absorb visible light. Therefore, under the action of sun light, photocatalytic processes with the use of TiO2 cannot be intensive, while the traditional sources of UV light are relatively expensive. In an effort to use visible light in photocatalysis, various methods for the production of active photocatalysts have been proposed in the literature, for example: 0133-1736/2003/US$ 20.00. © Akadémiai Kiadó, Budapest. All rights reserved.

342

KUZNETSOVA et al.: PHOTOCATALYSIS

I. Extension of the response of TiO2 to visible light by its modifications [1–6], e.g.: O TiO2 photosensitization using appropriate dyes [1–3]. On the one hand, colored organic compounds may play the role of TiO2 sensitizers in the degradation of other substances, provided that dyes are constantly regenerated. On the other hand, the colored organic compounds themselves can be oxidized over TiO2 under the visible light [1–3]. O TiO2 reduction by hydrogen plasma treatment [4] that results in an activation of TiO2 to the oxidative removal of NO under visible light. O Cr ion implantation into TiO2 films [5]. In this case, Cr ions are found to be located in the Ti sites of the TiO2 lattice and to form a Ti-O-Cr-O-Ti structure. The shift of the absorption band toward visible region depends on the amount of the Cr ions implanted. Decomposition of NO into N2, O2 and N2O proceeds over such a catalyst under visible light (λ > 450 nm) at D O

Preparation of TiO2-xNx films [6] that show a photocatalytic activity in the decomposition of methylene blue and in the oxidation of acetaldehyde under visible light. The substitutional doping of N for O has been proven to be effective in the band-gap narrowing and in inducing a photocatalytic activity under visible light due to mixing the N 2p and O 2p states in the valence bands [6].

The photocatalysts cited above did show some activity under the action of visible light, but in all cases, their actual quantum efficiency was not determined and probably was very small. II. Another way to prepare a catalyst active under the visible light is the use of inexpensive iron oxide, which absorbs readily the radiation with λ > 570 nm. A number of studies are devoted to photochemistry of different forms of iron oxides (α-Fe2O3, γ-Fe2O3, δ-FeOOH, β-FeOOH, γ-FeOOH, α-FeOOH) [7-14]. It has been shown in [7,8] that Fe2O3 particles of 100–200 nm size can photolyze water under visible light. For very small particles (less than 100 nm), the quantum size effect results in shifting the band gap towards the UV, whereas for the particles larger than 200 nm, the H2 generation is hampered by the e––h+ recombination [7,8]. Colloidal solutions of iron oxides show a photocatalytic activity under visible light in the oxidation of strong reductants or complexing agents such as sulfite [9–11], iodide [12], oxalate [9] and salicylic acid [13]. The maximum reported quantum yields in the photooxidation of SO32–, C2O42–, and I– over αFe2O3 were 30%, 12% and 80%, respectively. However, iron oxide is not active in the oxidation of cyanides and chlorine-containing organic substances, such as


343

chloroform, chloroacetic acid, and other organic compounds [14]. This may be due to the fact that the process of one-electron transfer from α-Fe2O3 to the O2 PROHFXOH ZLWK WKH IRUPDWLRQ RI H2.– is thermodynamically unfavorable. While the potential of the conduction band for bulk α-Fe2O3 is 0.3 V at jG = 2 [14], the electrode potential for the reduction of oxygen (e– + O2 + H+ = HO2) is -0.15 V vs normal hydrogen electrode (here and later vs NHE) [14]. The electrode potential for the two-electron reduction of oxygen with the formation RI H22– (2O2 + 2e– + 2H+ = H2O2) is 0.682 V vs NHE [15]. In the present work, we attempted to prepare a photocatalyst which is active under the action of visible light (λ > 400 nm) and consists of α-Fe2O3 loaded with Cu(OH)2. Copper hydroxide can act as a catalyst for multielectron processes [16]. So we made an effort at O2 activation by a more favorable reaction of two-electron transfer yielding a peroxy-anion.

EXPERIMENTAL Preparation of Cu(OH)2/α-Fe2O3. The catalyst was prepared via the following technique: 1 M NaOH was added to 250 mL of 0.04 M Fe(NO)3·6H2O solution until the pH became 10.8. The precipitate obtained was rinsed in distilled water until the pH became 9.3. Then 10 mL of 1 M HCl and 1 mL 0.1 M KH2PO4 were added to the mixture. The final volume of 500 mL was obtained by adding the required amount of distilled water. The mixture was held at 100oC for 24 hours. The precipitate obtained was rinsed in distilled water. Then 20 mL of 0.05 M Cu(NO3)2 was added to 70 mL of the precipitate suspension, 1 M NaOH was added until the pH became 10.0. The catalyst obtained was rinsed in distilled water and dried at room temperature. Experimental conditions. 10 mg catalyst powder with an illuminated area of 3.14 cm2 was used for experiments in a 434 mL reactor, at room temperature and 100% moisture. Instruments. A Xenon lamp LOS-2 (Russia) and a Mercury lamp DRSH1000 (Russia) with a cutoff filter (λ > 400 nm) were used for catalyst illumination. Light intensity in the interval of 400–600 nm, which was measured by LM-2 Carl-Zeiss-Jena photometer (GDR), equaled 30 mW/cm2 and 34 mW/cm2 for the Xenon and Mercury lamps, respectively. Ethanol, acetaldehyde, benzene, and CO2 were recorded with a gas chromatographs Tsvet-500 (Russia) and LHM 8MD (Russia). To analyze the reaction products, which were adsorbed on the catalyst surface, the powders of the catalyst before and after the reaction were suspended in 2 mL water acidified with H2SO4 to pH = 4, then the organic reaction products were extracted in 2 mL diethyl ether. The GC-MS analysis of the extracts was performed using a Saturn-2000 (USA).

344


The maximal content of carbonates on the catalyst surface was determined from the quantity of gaseous CO2 liberated from the catalyst during the adsorption of the acetic acid vapors. The adsorption was carried out at the following conditions: catalyst powder 10 mg, reactor volume 434 mL, the initial acetic acid concentration in the gaseous phase 605 ppm. The adsorption reached a steady state in 2 hours. The X-ray fluorescence analysis with a SPRUT-001 (Russia) by Mrs. I.L. Kraevskaya in the analytical laboratory of the Boreskov Institute of Catalysis showed that the Cu(OH)2/α-Fe2O3 catalyst contained 20 wt.% Cu. Quantum efficiency calculation. In general, quantum efficiency ϕ is OLJKW − GDUN , where Wlight and Wdark are the determined as follows: ϕ = Ö reaction rates under the light and in the darkness, respectively (mol/s); Φ is the quantum flux at the current wavelength (moles of quanta/s). Φ = , ⋅ 6 ⋅ λ , K ⋅ F ⋅ 1,

:

:

where I is the light intensity (W/cm2); S is the illuminated catalyst area (cm2); λ is the current wavelength of light (m); h is Planck's constant; c is the speed of light in vacuum; NA is Avogadro's number. In our case, the initial reaction rate under visible light was used for the calculation of the quantum efficiency, and the initial reaction rate in the dark equaled zero. In the calculations, we used the average wavelength λav. = 500 nm. Since illumination was carried out with the light in the range of 400–600 nm, the calculated value of quantum efficiency may be smaller than the actual value. RESULTS AND DISCUSSION The dependence of ethanol concentration in the reactor loaded with the Cu(OH)2/α-Fe2O3 catalyst on the time of illumination with visible light is presented in Fig. 1. The rate of decrease of ethanol concentration is maximal at the moment when the irradiation is just turned on. Until about 250 min of illumination, the ethanol concentration remained practically constant. In the control experiments, without illumination or in the absence of oxygen in the reactor, the ethanol concentration did not change (Fig. 1). Traces of acetaldehyde were found in the gas phase only under the illumination in the presence of O2. Therefore, ethanol is oxidized in the system only under the action of visible light. Illumination with the mercury or xenon lamp for 200 min resulted in the photooxidation of 2x10-6 and 5x10-6 moles of ethanol, respectively. The ethanol conversions in these cases were 14 and 35%, respectively. The quantum efficiency determined from the initial reaction rate was 0.1% and 1%, when mercury or xenon illumination was used, respectively.


345

The latter value of the quantum efficiency was higher probably due to catalyst heating by the infrared irradiation from the xenon lamp.

Light

C2H5OH concentration, ppm

1400

1200

1000

800

600

400 0

50

100

150

200

250

300

350

400

Time, min Fig. 1. Time dependence of the ethanol concentration in the reactor with Cu(OH)2/α-Fe2O3 ZLWKRXW LUUDGLDWLRQ O DQG XQGHU WKH YLVLEOH OLJKW RI ;HQRQ (▼) or Mercury (●) lamps. ■ – the same under the visible light of Mercury lamp but in the absence of oxygen (O2 is removed by blowing the reactor with He)

Only a negligible decrease in the ethanol concentration was observed during the illumination of α-Fe2O3 alone or Cu(OH)2 alone with the visible light of the mercury lamp. The ethanol conversions in these cases were ca. 4 and 6%, respectively. The changes in the CO2 concentration in the gas volume of the reactor during ethanol photooxidation over the Cu(OH)2/α-Fe2O3 catalyst are shown in Fig. 2. After 200 min of illumination with the mercury or xenon lamp, the amounts of CO2 liberated from 10 mg of the catalyst were 0.2x10-6 and 1x10-6 mol, respectively. However, we think that the observed CO2 liberation is not connected directly with ethanol oxidation. We believe that CO2 was liberated in our case due to desorption of CO2, which is chemisorbed on the catalyst surface probably in the form of copper carbonates. CO2 dissolved in NaOH or other aqueous solutions, probably produced CO2 chemisorbed on the hydroxylated catalyst during its preparation. Further, the desorption of chemisorbed CO2 can

346


result from the binding of acetic acid produced in ethanol oxidation to the catalyst surface. Indeed, the maximal amount of CO2 liberated from 10 mg of the catalyst was 2.2x10-6 mol in excess of the acetic acid vapors and without illumination. The GC-MS analysis confirmed the presence of acetic acid on the catalyst surface after the reaction (see EXPERIMENTAL). So, during illumination of the Cu(OH)2/α-Fe2O3 catalyst, ethanol was oxidized only to acetic acid and the observed deactivation of the catalyst can be accounted for by strong binding of acetic acid to the surface. It has been shown that the oxidative photoactivity of the catalyst depends on the amount of carbonates. For example, the sample with the minimal content of carbonates (10 mg of the catalyst contained only 0.6x10-6 PRO RI KH2) showed no photoactivity under visible light. In all cases, the copper content in the catalysts was 3x10-5 mol.

/LJKW

&2 FRQFHQWUDWLRQ SSP

120

100

80

60

40

20

0 0

50

100

150

200

250

Fig. 2. &KDQJHV LQ WKH KH2 concentration with time over the Cu(OH)2/α-Fe2O3 catalyst under the visible light of Xenon (▼) and Mercury (●) lamps

In addition, the following experiment with the Cu(OH)2/α-Fe2O3 catalyst was carried out: 0.074 g of the catalyst powder was placed in 0.74 mL of 0.005 VROXWLRQ RI +2PtCl6 at jG = 12 (1 M NaOH was added to the H2PtCl6 solution to avoid noncatalytic reduction of Pt4+ upon exposure to light). This suspension was illuminated with visible light under continuous stirring for 5 hours. Two control experiments were carried out as well: 1) the catalyst

.


347

suspension with H2PtCl6 prepared as cited above was stirred without the illumination, and 2) 0.005 M solution of H2PtCl6 at pH = 12 was illuminated with visible light without a catalyst. The X-ray fluorescence analysis showed that the catalyst was platinized with the formation of 0.08% Pt on its surface under exposure to visible light. On the contrary, the sample from the first control experiment did not contain Pt. In the second control experiment, no Pt(OH)2 was produced in the H2PtCl6 solution, as well. Thus, these data confirmed again that the obtained catalyst showed photoactivity under the visible light. The Cu(OH)2/α-Fe2O3 catalyst was active in the benzene oxidation as well. After the reaction under the visible light, a certain amount of phenol was found on the catalyst surface using the GC-MS analysis of extracts. CONCLUSION In contrast to pure α-Fe2O3 and Cu(OH)2, the Cu(OH)2/α-Fe2O3 catalyst showed a photoactivity in the oxidation of organic compounds such as ethanol and benzene. Ethanol oxidation proceeds only to the formation of acetic acid, which deactivates the catalyst. Copper hydroxide most probably enables or facilitates electron transfer from α-Fe2O3 to the oxygen molecule. This makes the catalyst active under the action of visible light. Acknowledgment. We gratefully acknowledge financial support of the NATO Science for Peace Program (Project Reference Number: sfp 974209) and of the Russian Foundation of Basic Research (Project “Scientific schools” # 00-1597446). We are also grateful to Drs. G.L. Elizarova and I.L. Kraevskaya for the assistance provided.

REFERENCES 1. 2. 3. 4.

5.

6.

S. Ferrere, B.A. Gregg: J. Am. Chem. Soc., 120, 843 (1998). K.T. Ranjit, I. Willner, S. Bossmann, A. Braun: J. Phys. Chem. B, 102, 9397 (1998). P.V. Kamat, K.Vinodgopal: Photocatalytic Purification and Treatment of Water and Air. Eds. D.F. Ollis, H. Al-Ekabi, 1993, p. 83. K. Takeuchi, I. Nakamura, O. Matsumoto, S. Sugihara, M. Andoh, T. Ihara: 5th International Conference on TiO2 Photocatalytic Purification and Treatment of Water and Air. London, Ontario, Canada, June 26-30, 2000, p. 11. M. Anpo, M. Takeuchi, H. Yamashita, T. Hirao, N. Itoh, N. Iwamoto: 5th International Conference on TiO2 Photocatalytic Purification and Treatment of Water and Air. London, Ontario, Canada, June 26-30, 2000, p. 9. T. Morikawa, R. Asahi, T. Ohwaki, K. Aoki, Y. Taga: 5th International Conference on Semiconductor Photochemistry. Glasgow (U.K.), July 23-25, 2001, p. 53.

348


7.

S. Chatterjee, S. Sarkar, S.N. Bhattacharyya: J. Photochem. Photobiol. A: Chem., 72, 183 (1993). J. Kiwi, M. Gratzel: J. Chem. Soc., Faraday Trans., 83, 1101 (1987). J.K. Leland, A.J. Bard: J. Phys. Chem., 91, 5076 (1987). B.C. Faust, M.R. Hoffmann, D.W. Bahnemann: J. Phys. Chem., 93, 6371 (1989). S.N. Frank, A.J. Bard: J. Phys. Chem., 81, 1484 (1977). J. Moser, M. Gratzel: Helvetica Chimica Acta, 65, 1436 (1982). P. Bonamali, Sh. Maheshwar: J. Chem. Technol. Biotechnol., 73, 269 (1998). C. Kormann, D.W. Bahnemann: J. Photochem. Photobiol. A: Chem., 48, 161 (1989). Brief Handbook of Physical-Chemical Values. Eds A.A. Ravdel, A.M. Ponomareva, Chemistry, Leningrad 1983, p. 144. G.L. Elizarova, L.G. Matvienko, O.L. Ogorodnikova, V.N. Parmon: Kinet. Catal., 41, 366 (2000).

8. 9. 10. 11. 12. 13. 14. 15. 16.