Environmental Monitoring and Assessment (2006) 115: 395–403 DOI: 10.1007/s10661-006-7236-y
c Springer 2006
PATHWAYS AND KINETICS ON PHOTOCATALYTIC DESTRUCTION OF AQUEOUS PHENOL LIANFENG ZHANG1,∗ , TATSUO KANKI2 , NORIAKI SANO2 and ATSUSHI TOYODA3 1
University of Waterloo, Faculty of Engineering, Department of Chemical Engineering, Waterloo, Ontario, Canada; 2 University of Hygo, Graduate School of Mechanical System Engineering, Himeji, Hyogo, Japan; 3 Envisys Co. Ltd., R&D Division, Himeji, Hyogo, Japan (∗ author for correspondence, e-mail:
[email protected])
(Received 15 January 2005; accepted 10 May 2005)
Abstract. In this study, the TiO2 photocatalytic decomposition process of aqueous phenol was investigated. The intermediate products generated in the elementary reaction steps in the mineralization process were experimentally identified as hydroquinone, catechol and hydroxyhydroquinone. The concentration variations of these intermediate products with time passage were traced by high performance liquid chromatograph. The pathways of the decomposition process were given. Based on Langmuir isothermal theory and Langmuir-Hinshelwood mechanism, the multi-compounds competition kinetic model was established. In this model, the observed time-dependent concentrations of phenol and the intermediate products were simulated. Keywords: photocatalyst, TiO2 , kinetics, phenol, pathway, Langmuir-Hinishelwood
Notation Ci : molar concentration xi : number of carbons ki : reaction rate constant K i : adsorption equilibrium coefficient ai j : fraction related to elementary reaction t : time
S UBSCRIPTS 1: phenol 2: catechol, abbreviated as CC 3: hydroquinone, abbreviated as HQ 4: hydroxyhydroquinone, abbreviated as HHQ 5: organic acids, abbreviated as OA 6: total organic carbons, TOC
396
L. ZHANG ET AL.
1. Introduction The photocatalytic purification of environment has been extensively investigated. Many semiconductors have photocatalytic function (Dom´enech, 1993; Lepore et al., 1993; Faust et al., 1989; Bahnemann et al., 1987; Frank and Bard, 1977), but most researches focused on titanium dioxide due to its high efficiency, stability, and non-toxicity (Nogueira and Jardim, 1996; Okamoto et al., 1985; Hoffmann et al., 1995). The UV-illuminated semiconductor can generate positive holes and negative electrons, which react with oxygen and H+ , OH− in water to form ·OH and ·OOH radicals. Organic compounds can be destructed by such radicals. Small molecule organic compounds can be directly oxidized to carbon dioxide. Large molecule organic compounds have to degrade to small molecules, and then are further oxidized to carbon dioxide. The intermediate products compete for adsorption site, radicals with parent compounds. And in some cases, it will be very meaningful only for toxic compounds to become non-toxic (Wu et al., 2004; Choi et al., 2000). It’s very important to investigate the degradation pathways of large molecule organic compounds. Many kinetic models of photcatalytic decomposition of organic compounds were reported, most of them are based on Lamuir-Hinshelwood model (L-H model), i.e., L-H model or transformed L-H model. For instance, the first order relation is accepted in many reports, actually it is the special case of L-H model: the product of adsorption coefficient and the concentration of reactant is greatly smaller than 1; Puma and co-worker reported various models for different photocatalytic reactors, in which the reaction kinetics is similar to L-H model, or can be regarded as the special cases of L-H model (Puma, 2003; Puma and Yue, 2003). Organic compounds with aromatic ring might act as toxic pollutants to environment (Bhatkande et al., 2004; Liu et al., 2001). To photocatalytically decompose these compounds, the cleavage of aromatic ring is a key intermediate step. Hydroxyl group addition plays an important role for this cleavage (Wang and Ku, 2003; Sivalingam et al., 2004). To further investigate the decomposition mechanism of such aromatic compounds, it’s a base work to understand the degradation pathways of phenol. Several studies have been conducted on the degradation of phenol and substituted phenol (Sobczy´nski et al., 2004; Sivalingam et al., 2004; Akbal and Onar, 2003; Okamoto et al., 1985). However, the degradation pathways are still being disputed. This work thus aims at exploring the intrinsic decomposition pathways, and establishing the kinetic model for photocatalytic decomposition process with multiintermediates to kinetically explain these pathways.
2. Experimental The laboratory scale rotating-drum reactor, which originally designed for purifying water by solar light (Zhang et al., 2001), was used in this work. The photocatalyst
DECOMPOSITION MECHANISM OF PHENOL
397
is commercial titanium dioxide (TP-2, Fuji Titanium Ltd.); purity: 99.98%, mean diameter: 0.15 μm, specific area: 23.2 m2 /g, crystal type: anatase. It was coated on the outer surface of a glass drum, which is 5 cm in diameter and 18 cm in length. Both ends of the drum were sealed with metallic caps with rubber packing. About 40% of the drum was immersed in water to be treated and the other part was exposed in the atmosphere. The drum was rotated, and the water was viscid enough to be drawn up continuously onto the drum surface and formed thin water film. The rotation speed of the drum was set appropriately at 25 rpm. The TiO2 photocatalyst surface carrying water film was illuminated by two 6W mercury ultraviolet lamps (GL-6, 253.7 nm, Toshiba Ltd.). We used aqueous solution of 50.0 mg/dm3 phenol as the water to be treated. A series of batch experiments were conducted in different running times, 5, 10, 30, 45, 65, 75 minutes respectively. 25 mL aqueous solution was fed in the vessel. The concentrations of phenol and its intermediate products were analyzed by high performance liquid chromatograph (HPLC). The Column is Develosil Packed Column (Nomura Chemical), the liquid pump unit is LC-10Advp (Shimadzu), and the detector is UV-VIS detector (SPD-10Avp/10Avvp, Shimadzu). The eluent stream is composed of 0.1%H3 PO4 and CNCH3 at the ratio of 98:2. The analysis condition is 1.0 ml/min of liquid flow rate, 20 ◦ C, and the wavelength for the detector is 277 nm. The total organic carbons were measured by TOC meter (TOC-5000, Shimadzu Ltd.).
3. Results and Discussion In the decomposition process of aqueous phenol, the color variation could be observed: initially the solution was transparent without color, then changed to red, and finally to no color again. This indicates that phenol was mineralized in a series of reaction steps through intermediate products. Figure 1 shows an illustration of the HPLC charts during the decomposition process of phenol, and the HPLC charts of pure phenol, pyrogallol (PG), catechol (CC), hydroquinone (HQ), and hydroxyhydroquinone (HHQ). In the figure, we can find that the retention times of intermediate products match those of pure HQ, CC, and HHQ. In phenol molecule, hydroxyl group disturbs the evenness of π -electron density on aromatic ring and increases the electron density at the ortho and para positions, i.e., the hydroxyl group is ortho, para-directing group. Accordingly, CC, HQ, HHQ, and PG are theoretically possible intermediate products in the photocatalytic decomposition of phenol. Thus both theoretical and experimental analyses proved the existence of CC, HQ and HHQ. This identification is also consistent with Sivalingam and co-woker’s report (Sivalingam et al., 2004). Okamoto and co-workers thought that intermediate products also include pyrogallol (Okamoto et al., 1985), which is in agreement with the analysis of molecular structure. However, the experimental results denied the existence of pyrogallol (see Figure 1).
398
L. ZHANG ET AL.
Figure 1. (a) Variation of HPLC charts in the decomposition process of phenol; (b) The HPLC charts of pure phenol, PG, CC, HQ, and HHQ.
Figure 2 shows the variation of concentrations of phenol, intermediate products and corresponding TOC in the process of decomposition of phenol. The figure indicates that the intermediate products HQ and CC are generated immediately after phenol starts to be oxidized and HHQ is generated from the precursor intermediate products. This is consistent with Sivalingam’s research report (Sivalingam et al., 2004). The TOC is found to still remain appreciably high value even when the intermediate products HQ, CC and HHQ are completely oxidized to disappear. This implies that the further oxidized intermediate products exist. These intermediate products are thought to be organic acid (OA) without aromatic ring (Okamoto, 1985). According to these facts, we can write the reaction pathways
DECOMPOSITION MECHANISM OF PHENOL
399
Figure 2. Time course of phenol decomposition, photocatalyst: TiO2 , reaction liquid: 25 ml, rotation speed: 5 rpm, UV light: 30 W/m2 .
Figure 3. Decomposition pathways of aqueous phenol.
for photocatalytic decomposition of phenol as shown in Figure 3. Not only can the aromatic ring with three hydroxyl groups be broken (Sobczy´nski et al., 2004; Sivalingam et al., 2004; Okamoto,1985), but also those with one and two hydroxyl groups can be directly broken. On the surface of TiO2 whose crystal system is tetragonal system, the coordination sphere of titanium atom is incomplete at the (110) plane, (110) × (100) and (010)×(100) edges, and at several corners of the crystal. In aqueous circumstances, such titanium atom can bond to oxygen atoms of water molecules. Hydroxyl groups
400
L. ZHANG ET AL.
on the TiO2 surface undergo the following acid-base equilibriums (Kormann et al., 1988) + ≡ TiOH+ 2 ↔≡ TiOH + H −
+
≡ TiOH ↔≡ TiO + H
pK1 pK2
(1) (2)
The pH of zero point of charge, pHzpc , is given by pHzpc = (pK1 + pK2 )/2
(3)
The cationic electron donors will be favored by heterogeneous photocatalyst at high pH, pH>pHzpc , while anionic electron donors will be favored at low pH, pH
