- Email: [email protected]
A comparative study on the photocatalytic degradation of humic substances of various origins
DESALINATION Desalination 176 (2005) 167-176
ELSEVIER
www.elsevier.com/locate/desal
A comparative study on the photocatalytic degradation of humic substances of various origins Ceyda Senem Uyguner, Miray Bekbolet* Institute of Environmental Sciences, Bogazici University, 34342, Bebek, Istanbul, Turkey TeL +90 (212) 359-7012; Fax: +90 (212) 257-5033; email: [email protected] Received 26 October 2004; accepted 5 November 2004
Abstract The photocatalytic removal of model humic and fulvic acids of different origins (terrestrial and aquatic) was investigated using TiO2 Degussa P-25 as the photocatalyst. The results were presented comparatively based on pseudofirst-order and the Langmuir-Hinshelwood kinetic models. The reaction rates calculated for the UV-vis parameters follow the general trend: IHSS soil humic acid > Aldrich humic acid > IHSS humic acid > Roth humic acid > IHSS fulvic acid. However, the rate constants show differences depending on the UV-vis absorption parameters. The evaluation of the Langmuir-Hinshelwood kinetic model reveals that, in all cases, Roth hurnic acid has the lowest Langmuir-Hinshelwood rate, rate constant and adsorption coefficient, and the ordering of the rates follows the same trend with that of pseudo-first order. Furthermore, data related to the molecular size distribution profiles of raw and treated humic acids suggest that the photocatalytic degradation occurs irrespective of the molecular size fractions. Keywords: Humic acid; Fulvic acid; Photocatalysis; Langmuir-Hinshelwood kinetics; Molecular size fractionation
1. Introduction Humic substances are the predominant type of natural organic matter present in ground and surface waters. On the basis o f their solubility in water and as a function o f p H , humic substances are generally classified into humic acids (HA),
fulvic acids (FA) and humin. HA are comprised o f high-molecular-weight organic substances that are soluble in alkaline media and insoluble in acidic media, whereas FA comprise moderate molecular weight organic substances of nonspecific composition that are soluble at all pH values.
*Corresponding author. Presented at the Seminar in Environmental Science and Technology: Evaluation of Alternative Water Treatment Systems for Obtaining Safe Water. Organized by the University of Salerno with support of NATO Science Programme. September 27, 2004, Fiseiano (SA), Italy. 0011-9164/05/$- See front matter © 2005 Elsevier B.V. All rights reserved doi: 10.1016/j.desal.2004.11.006
168
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
Humic substances are known to be a complex class ofbiogenic polyelectrolytes that can interact with organic and inorganic substances in the environment. Imparting undesirable color and acting as precursors for undesirable trihalomethane formation during the chlorination process, their presence in drinking water supplies causes problems, the consequences of which may lead to the deterioration of human health [1,2]. Hence, their removal from water supplies has received specific attention. During the last decade, TiO2-mediated heterogeneous photocatalysis has emerged as an alternative advanced oxidation technology for the removal ofhumic substances from drinking water supplies. Since 1996, Bekbolet and coworkers have been studying the photocatalytic degradation of HA in detail, focusing on operational parameters that affect the process efficiency [3-6]. Complementary to previous research and in parallel to the recent results of a study where UV-vis spectroscopic properties of humic substances have been investigated, this study presents data on the photocatalytic removal of model HA and FA of different origins using TiO2 Degussa P-25 as the photocatalyst [7]. The UVvis properties ofhumic substances are utilized to assess the degradation kinetics that is presented comparatively based on pseudo-first-order and Langmuir-Hinshelwood kinetic models [5,8]. Aqueous solutions of HA are generally polydisperse, with size ranges differing according to their origins, thereby strongly affecting treatment efficiencies. The oxidative behavior of humic substances towards photocatalytic degradation could also be evaluated by use of molecular size distribution profiles.
2. Experimental HA and FA of different origins (terrestrial and aquatic) were used in bench-scale photocatalytic oxidation experiments. The Suwannee River
fulvic acid (IHSS FA) and Suwannee River humic acid (IHSS HA) standard materials, isolated from the Suwannee River, Georgia, as well as soil humic acid (IHSS SHA) (Lot no. 1S 102H) as a terrestrial source were purchased from the International Humic Substance Society. Commercial HA samples were supplied by Aldrich (humic acid salt) (AHA) and Roth (humic acid) (RHA). The term "humic substances" is used throughout the text to define humic and fulvic acids. TiO 2 Degussa P-25 was used as the photocatalyst. Photocatalytic oxidation of humic substances (50 mg L-l) was carried out at pH 6.5+0.5 according to a previously outlined procedure using 0.25 mg mL -l of TiO 2 Degussa P-25 [5]. After each run, the absorption of the supernatant was determined using a Shimadzu UV160A double-beam spectrophotometer at 436 nm (Color436, m-l), 400 nm (Color400, m-l), 365 nm (UV365, m-l), 300 nm (UV300, m-l), 280 nm (UV2so, m -1) and 254 nm (UV254, m -l) for the evaluation of the kinetics of the photocatalytic degradation ofhumic substances. The absorbance values measured at 254 nm (UV254) and 280 nm (UV2s0) are known to explain aromaticity removal whereas, 436 nm is used to monitor decolorization. Furthermore, specific UV absorbance (SUVA254, m -l mg -1 L) was used to represent total organic carbon (TOC) normalized aromatic moieties. TOC (mg L-l) measurements of humic substances were performed on a Shimadzu TOCV CSH TOC analyzer. Calibration of the instrument was done using potassium hydrogen phthalate in the concentration range of 5-25 mg L -1. Molecular size fractionation of the raw and photocatalytically treated humic substances (50% degradation with respect to Color436) was performed by ultrafiltration (Amicon) through a sequence of membranes of decreasing pore sizes in the range of 100,000 to 1,000 Dalton. At the beginning of each run, the samples were filtered through 0.45 #m (approximately 450 kDa) Millipore membrane filters.
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
3. Results and discussion 3.1. Sample characterization
Light absorption in the UV-vis region is a typical property ofhumic substances. A variety of absorption wavelengths and ratios has been proposed for the spectroscopic characterization of humic substances [9-14]. Amongst them, UV254 is interchangeably measured with TOC as a surrogate parameter to indicate the natural organic matter content in natural waters [15]. The UV absorptivity at 280 nm is known to represent total aromaticity for phenolic arenes, benzoic acids, aniline derivatives, polyenes and polycyclic aromatic hydrocarbons with two or more rings that are considered as common structural subunits in humic matter [ 10,11 ]. Furthermore, absorption values at 400 and 436 nm are used to measure color removal ofhumic substances. In accordance with the reported diverse nature of UV-vis data, the specified UV-vis parameters are recorded for the assessment of the photocatalytic degradation kinetics. Depending on the concentration of the working solutions, UV-vis and TOC parameters of the humic substances display certain values as given in Tables 1 and 2. Humic acids of terrestrial origin exhibited Color436 values in the range of 18.8-38, whereas HA of aquatic origin, IHSS HA, had a relatively lower content of colorforming moieties. Obviously, IHSS FA displayed quite low color characteristics. The UV365 parameter is found to be irrespective of the source of the samples as expressed in a decreasing order of RHA > IHSS SHA > IHSS HA > AHA > IHSS FA. On the other hand, RHA has remarkably high UV280 and UV254values with respect to the other samples emphasizing strong aromatic character. The order of UV-absorbing moieties of the samples is followed by a decreasing trend as RHA > IHSS HA > IHSS SHA > AHA > IHSS FA, reflecting non-source dependency. The TOC contents of the humic substances are expressed in
169
a decreasing order of IHSS FA> RHA > IHSS HA > AHA > IHSS SHA. 3.2. Kinetics o f photocatalytic degradation
The kinetics of the photocatalytic degradation of humic substances is known to obey pseudofirst order as expressed by the equation [5,8,16, 17]: R = - dA/dt = k A
(1)
where R is the pseudo-first-order rate (m- Jmin-~), A the specified UV-vis parameters of humic substances (m-l), t the irradiation time (rain) and k the pseudo-first-order reaction rate constant (min-'). The evaluation of the kinetic data revealed the following model parameters for C010r436, UV365, UV2so and UVz54 as presented in Table 1. The results of the relevant data for Color400 and UV300 are only discussed for simplicity purposes. The color-forming moieties of humic substances are usually expressed with C010r436and/or Color400 values; hence, the removal of similar moieties might be considered. A comparison of the rate constants of Color436 and Color4oo for the photocatalytic degradation of HA in a diverse solution matrix indicates that the difference is <10% [4]. Accordingly, in this study the rate constants calculated for C010r436 and Color400 differ from each other by 4 to 7%; hence the C010r436 parameter was preferred. Recent results indicate the importance ofabsorbance ratios such as E254/E436, E28o/E36s and their respective changes during the photocatalytic degradation of HA [ 14]. Therefore, the removal ofUV36 s forming moieties was also analyzed in this study. The degradation rates calculated for all the indicated parameters follow the general trend: IHSS SHA > AHA> IHSS HA > RHA > IHSS FA. However, the rate constants show differences depending on the UV-vis parameters employed. The ordering is similar for Color436 and UV36sand
170
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
Table 1 Pseudo-first-order kinetic parameters of photocatalytic degradation of model compounds (R 2 >0.95); Ao, initial value of specified UV-vis parameters (m -~)
Color436 IHSS FA IHSS HA IHSS SHA AHA RHA
A o, m -1
k, min -t
6/2, min
R, m -~ rain -1
7.22× 10 -3 8.30× 10 3 1.35x 10 -2 1.60× 10 2
0.0368 0.132 0.379 0.301 0.115
5.1 15.9 28.1 18.8 38
3.04x 10 -3
96 84 51 43 228
19.4 44.3 48.4 39.2 72.9
6.46× 10 -3 7.03 x 10 -3 1.24x 10 -2 1.49x 10 -2 2.69x 10 -3
107 99 56 47 258
0.125 0.311 0.600 0.584 0.196
68.4 120.2 99.3 92.1 159.4
4.02× 10 -3 5.45x10 -3 1.10x 10 -2 1.09x 10 -2 2.13×10 -3
172 127 63 64 325
0.275 0.655 1.092 1.004 0.339
91.1 148.1 114.9 108.1 186.3
3.76x 10 -3 4.90x 10 -3 1.07× 10 -2 1.03x10 2 2.01 ×10 -3
184 141 65 67 345
0.342 0.726 1.229 1.113 0.374
UV365 IHSS FA IHSS HA IHSS SHA AHA RHA
UV28o IHSS FA IHSS HA IHSS SHA AHA RHA
UV2s4 IHSS FA IHSS HA IHSS SHA AHA RHA
Table 2 Pseudo-first-order kinetic parameters o f photocatalytic degradation of model compounds in terms of TOC (R2>0.95) TOC
TOCo, mg L-i
k, min-l
tl/2, min
R, mg L-l min-1
IHSS FA IHSS HA IHSS SHA AHA RHA
20.59 16.16 15.48 15.5 19.53
1.07× 10 -2 2.65x 10 -3 2.10x 10 -2 9.75×10 -3 3.02 x 10 -3
65 261 33 71 230
0.220 0.0501 0.339 0.151 0.0590
d e c r e a s e s in t h e f o l l o w i n g o r d e r , A H A > I H S S SHA> IHSS HA> IHSS FA> RHA. IHSS SHA a n d A H A e x h i b i t q u i t e s i m i l a r UVzs 0 a n d UV254
degradation rate constants that could be cons i d e r e d as d i f f e r i n g i n s i g n i f i c a n t l y . T h e l o w e s t d e g r a d a t i o n r a t e c o n s t a n t is o b t a i n e d for R H A
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
followed by IHSS FA and IHSS HA for both of the UV absorbing moieties (UV2s0 and UV254)that signify benzene such as polycyclic aromatic hydrocarbons with two or more rings. On the other hand, UV300 removal rates are found to be relatively lower than UV365 but exhibit higher removal rates than UVzso. In accordance with the low photocatalytic degradation rate constants, quite high half-life values for all of the parameters are expected. These values are presented in Table 1. FA are generally characterized by their low molecular weight and a smaller number of total and aromatic carbons than their HA counterparts. Unexpectedly, their respective removal rates in terms of the specified UV-vis parameters are found to be relatively lower than the removal rates of HA that have longer chain fatty acid products and a higher hydrophobicity than FA. The structural and compositional characteristics of the samples could possibly affect the photocatalytic degradation pathway leading to various degradation products comprising composite UVvis properties. Considering that UV254 is generally used as a surrogate parameter for TOC, the rate constants calculated for TOC removal might be expected to follow the same trend observed for UV254. However, UV254 rate constants for IHSS FA, IHSS SHA, and RHA are relatively low compared to the TOC rate constants (Tables 1 and 2). Contrary to that, lower values for the rate constants of TOC were obtained for IHSS HA and AHA with respect to the rate constants calculated for UV254 removal. For IHSS FA, the UV254rate constant is one-third of the rate constant calculated for TOC. This indicates that low-molecular-weight and fewer UV-absorbing compounds are formed. Con-equently, the half-life values for UVz54 removal are ordered as IHSS SHA > AHA > IHSS HA > IHSS FA > RHA, and the TOC halflife values follow the decreasing trend as IHSS SHA > IHSS FA > AHA > RHA > IHSS HA (Tables 1 and 2).
171
Prior to photocatalysis, initial adsorption levels were determined as 6, 7, 15, 30, and 47% TOC removal for IHSS FA, IHSS HA, RHA, AHA and IHSS SHA, respectively. As verified by the discrepancy between TOC values, the initial adsorption oflHSS SHA onto TiO 2 is reflected as high removal rates in terms of the specified UVvis parameters and TOC. However, for the other samples, removal rates were not found to be in accordance with the initial adsorption trends. The reason for this anomaly might be explained by the conformational changes present in the adsorbed state of HA and FA samples on TiO 2. In view of the generally presented optimum loading of 1.0 mg mL -j of TiO2, photocatalytic experiments were also carried out under higher photocatalyst conditions [3]. Color436degradation rate constants are found to be 4.30×10 .2, 4.19× 10-2, 2.58x10 -2, 1.52x10 -2, and 8.31×10 -3 for IHSS SHA, AHA, IHSS FA, IHSS HA and RHA, respectively. Following the same decreasing trend, considerably lower UV254rate constants are calculated as 3.25×10 -2, 2.49×10 -2, 1.22×I0 -2, 1.08× 10 -2 and 6.22× 10 -3 for the photocatalytic degradation of the humic substances. The most prominent effect was observed for IHSS FA with a four-fold increase of the Color436removal rate. On the other hand, the effect of increasing the photocatalyst loading by fourfold affected the removal rate of IHSS HA by a two-fold increase that is comparatively less than the other samples investigated (Table 1). A similar trend was observed for the UV254removal rates. For IHSS FA and IHSS SHA, increasing TiO z loading has the effect of increasing UV254 removal rates by approximately threefold. It could be stated that increasing the photocatalyst loading obviously results in an enhancement of the removal efficieneies. However, to assess a precise UV-vis parameter evaluation for molecular size distribution data after photocatalytic treatment, the use of 0.25 mg mL-1 photocatalyst loading was preferred.
172
c.s. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
Considering that photocatalysis occurs at the surface, the concentration ofsubstrate adsorbed to the surface directly affects the overall rate of adsorption. A Langmuir-Hinshelwood model is widely used to describe the kinetics of semiconductor photocatalysis [18,19]. The rate o f degradation o f humic substances in terms o f C010r436, UV254 and other UV-vis parameters is expressed by the following empirical type relationship:
= kL. KL. A/(1 +
A)
(2)
where RLrt is the rate o f the reaction (m -~ min-J), kLrt the reaction rate constant (min-1), KLH the adsorption coefficient o f the reactant onto TiO2 under irradiation conditions (m) and A the initial value of the specified UV-vis parameters of humic substances (m-1).
A new series of photocatalytic experiments was designed to assess the Langmuir-Hinshelwood kinetics utilizing 20-50 mg L-1 humic substances in the presence o f 0.25 mg mL -1 TiOz Degussa P-25. The data are evaluated in terms of the above-given Langmuir-Hinshelwood model, and the related parameters are given in Table 3. For all of the HA, the L-H rate constants increase in the order offolor436, UV365, UV2~0and UV254, revealing that UV absorbing centers may be removed more selectively than color-forming moieties. However, within the studied sample pool, the ordering o f the rate constants follows the general trend IHSS SHA > AHA > IHSS HA > RHA for all of the parameters. Contrary to the rate constants, the adsorption coefficients display a decreasing trend o f Color436, UV365,UV280and UV2s4 for all o f the samples.
Table 3 Langmuir-Hinshelwood kinetic parameters of humic substances (rate was calculated for 50 mg L -~ of each humic substance)
IHSS HA
Color436
UV365
UV280
UV254
kEn, min-t KEn, m
0.0903 1.47 93 0.0866
0.206 0.337 118 0.193
0.435 0.134 150 0.410
0.469 0.0838 176 0.434
0.244 0.398 65 0.224
0.399 0.345 66 0.376
0.714 0.120 78 0.659
0.805 0.105 80 0.743
0.188 0.584 56 0.172
0.376 0.360 57 0.351
0.627 0.122 83 0.576
0.697 0.109 87 0.642
0.0698 0.217 318 0.0623
0.115 0.0936 381 0.100
0.195 0.0395 499 0.168
0.214 0.0332 533 0.184
ttl2, min m -~ min-I IHSS SHA kLn, min-j KEn, m tt/2, min RLn,m -I min-t AHA kLn, min-i KLH,m fia, min RLr~,m-l min-1 RHA ken, min -l KL~,m tt/2, min RLn, m- i min- 1 RLn ,
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
A comparison within the samples results in different orders such that; Color436 and UV280 exhibit similar trend as, IHSS HA> AHA> IHSS SHA> RHA and UV365 and UV254 are ordered as AHA> IHSS SHA> IHSS HA> R_HA. The effect of the model parameters on the overall rate equation is expressed by the L-H rates following the same trend of the L-H rate constants. Considerably higher L-H degradation rates are achieved for UV254 than C010r436 for all of the samples. Considering the related half-life values, the following orders could be presented for Color436 and UV365 as: AHA < IHSS SHA< IHSS HA < RHA; and for UVaso and UV254 as: IHSS SHA < AHA < IHSS HA < RHA. The ordering of the L-H rates is similar to that of first-order rates; however, the calculated L-H rates for the HA are approximately half the value of the pseudo-first-order rates for all of the studied parameters. The discrepancy observed in the kinetic modeling of the photocatalytic degradation of the humic substances could not be clearly explained - - neither by the properties of the parameters nor by the character of the samples.
3.3. Evaluation of molecular size distribution
profiles The effect of the photocatalytic degradation pathway could be visualized by the structural changes that could be resolved in terms of molecular size distribution profiles. To achieve this goal, a new series of photocatalytic degradation experiments was performed to prepare samples that were partially oxidized as 50% with respect to Color436 to follow the possible changes in molecular size distribution. Molecular size distribution characteristics of raw and photocatalytically treated HA samples are presented through the SUVA254 values 0 f each fraction provided that the SUVA254 parameter comprises the counterbalancing effect of UV254 and TOC (Figs. 1 and 2). It is known that the molecular size distribution profiles of the humic substances exhibit a general decreasing trend from high molecular size fractions to the lower molecular size fractions irrespective of the origin and source of the organic matter. Depending on the explanation that the specific UV absorbance (SUVA254)parameter can be used to describe the composition of humic
8.00 7,00 "7
173
6.00 5.00 4.00 3.00 2.00 1.00 0.00
450-100 kDa 100-30kDa 30-10kDa 10-1 kDa Molecular size distribution
<1 kDa
Fig. 1. SUVA254values for raw humic acid samples with respect to molecular size distribution.
174
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
8.00 7.00 "7
6.00 5.00
"7
4.00 3.00 2.00 1.00
0.00 450-100 kDa
100-30kDa
30-10kDa
10-1 kDa
<1 kDa
Molecular size distribution
Fig. 2. SUVA254values for photocatalytically treated humic acid samples with respect to molecular size distribution.
material in terms of hydrophobicity and hydrophilicity, SUVA254>4 indicates mainly hydrophobic and especially aromatic moieties while a SUVA254 <4 represents a hydrophilic organic fraction [20]. The hydrophobic character predominates for humic samples up to a size fraction of 10 kDa, whereas a comparatively hydrophilic character could be assessed for smaller molecular size fractions as 10-1 kDa and even smaller than 1 kDa fraction (Fig. 1). All of the studied HA samples exhibited close similarities in terms of SUVA254 up to the 30 kDa fraction. The medium size fractions of the 3010 kDa region of aquatic originated IHSS HA displayed a comparatively higher SUVA254value than the others that were of terrestrial origin. IHSS SHA has a quite low SUVA254 value whereas IHSS HA of aquatic origin still revealed a higher SUVA254 than the others for the 101 kDa fraction. On the other hand, the AHA sample acquired a comparatively high SUVA254 for fractions <1 kDa. As a general trend for all of the HA investigated, SUVAz54 values decrease after photocatalytic treatment indicating that the degradation efficiency is irrespective of molecular size fraction (Fig. 2). The primary effect ofphotocatalytic
treatment was observed in the 450-100 kDa fraction for all the HA except RHA. The probable resistance displayed by the 450--100 kDa fraction of RHA to photocatalytic degradation could also be explained by the formation of high-molecularweight fractions (450-100 kDa) via the combination of the photoproducts originating from lower molecular size fractions. A slight change is observed for SUVA254 of IHSS HA irrespective of the size fraction up to 10 kDa. IHSS HA and RHA retained a hydrophobic character up to the 10 kDa size fraction whereas AHA displayed more hydrophilic character for size fractions smaller than 30 kDa. IHSS SHA expressed a quite different photocatalytic reactivity towards the formation of lower molecular size fractions (<100 kDa) that could be considered as possessing more hydrophilic properties since SUVA254 values are found to be between 2-3. For AHA and IHSS SHA samples increased SUVA254 values are noticed for the smallest size fractions. The size fraction < 1 kDa for IHSS SHA exhibits a two-fold increase in SUVA254 relative to that of the 10 kDa fraction. Oxidation reactions taking place during photocatalysis may lead to a shift as an increase in lower molecular size fractions that is apparently observed in the case of
C.S. Uyguner, M. Bekbolet ~Desalination 176 (2005) 167-176
IHSS SHA and AHA. This could be explained by relatively high pseudo-first-order and L-H UV254 degradation rates as well as the TOC removal rates of the IHSS SHA and AHA (Tables 1 and 2). It is widely known that during photocatalytic degradation the main oxidative pathway is achieved by the nonselective attack o f . O H radicals. The different reactivities observed for each size fraction could be attributed to their structural and conformational diversity. The elucidation of the complex chemistry involved in these photocatalytic pathways will likely only be achieved through the application of sophisticated analytical tools.
4. Conclusions On the basis of their diverse chemical and physical properties such as molecular weight, molecular size, elemental composition and source of origin, substantial differences are observed in photocatalytic removal efficiencies ofhumic and fulvic acids. The pseudo-first-order rate constants calculated for Color436 of each of the humic substances show a decreasing trend as follows: AHA > IHSS SHA > IHSS FA > IHSS HA > tLHA. On the other hand, the order of the rate constants of aromaticity removal (UV254) is as follows: IHSS SHA > AHA > IHSS HA > IHSS FA > RHA. The rate constants calculated for TOC removal exhibit a different trend: IHSS SHA > IHSS FA > AHA > RHA > IHSS HA. Higher removal rates have been achieved with UV254 values compared to that of Color436values. All of the studied HA exhibit higher pseudo-firstorder removal rates compared to that for FA. The ordering of the L-H rates is similar to that of first-order rates; however, the calculated L-H rates for the HA are approximately half the value of the pseudo-first-order rates for all of the studied UV vis parameters. Comparison of the molecular size distribution profiles of the photocatalytically treated HA with
175
those of raw untreated samples reveals the confirmation of the formation of lower molecular size fractions exhibiting low SUVA254values.
Acknowledgements The support provided by the Research Fund of Bogazici University, Project No. 03S107 and NATO Grant EST.CLG.980506 is gratefully acknowledged. The valuable contribution by Patrick Dunlop and Marc Anderson is appreciated.
References [1] J.J. Rook, Water Treat. Exam., 23(2) (1974) 234243. [2] P.C. Singer, Water Sci. Technol., 40(9) (1999) 2530. [3] M. Bekbolet, J. Env. Sci. Health, A31(4) (1996) 845-858. [4] M. Bekbolet, Z. Boyacioglu and B. Ozkaraova, Water Sci. Technol., 38 (1998) 155-162. [5] M. Bekbolet, A.S. Suphandag and C.S. Uyguner, J. Photochem. Photobiol. A: Chem., 148 (2002) 121128. [6] C.S. Uyguner and M. Bekbolet, Intemat. J. Photoenergy, 6(2) (2004) 73-80. [7] C.S. Uyguner and M. Bekbolet, Desalination, 176 (2005) 47-55. [8] M. Bekbolet and I. Balcioglu, Water Sci. Technol., 34(9) (1996) 73-80. [9] H. De Haan, T. De Boer, H.A. Kramer and J. Voerman, Water Res., 16(6) (1982) 1047-1050. [10] S.J. Traina, J. Novak and N.E. Srneck, J. Environ. Quality, 19(1)(1990) 151-153. [11] Y.P. Chin, G. Aiken and E. O'Loughlin, Environ. Sci. Technol., 28(11) (1994) 1853-1858. [12] J. Peuravuori and K. Pihlaja, Anal. Chimica Acta, 337 (1997) 133-149. [13] J. Chen, B. Gu, E.J. LeBoeuf, H. Pan and S. Dai, Chemosphere, 48 (2002) 59-68. [14] C.S. Uyguner and M. Bekbolet, submitted for publication. [15] I.N. Najm, N.L. Patania, J.G. Jacangelo and S.W. Krasner, J. AWWA, 86(6) (1994) 98-106.
176
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
[16] J. Wiszniowski, D. Robert, J. Surmaez-Gorska, K. Miksch and J.V. Weber, Intemat. J. Photoenergy, 5 (2003) 69-74. [17] R. A1-Rasheed and D.J. Cardin, Appl. Catal. A: General, 8501 (2003) 1-10.
[18] R.W. Matthews, J. Catal., 111(2) (1988) 264-272. [19] C. Turchi and D.F. Ollis, J. Catal., 119 (1989) 483496. [20] J.K. Edzwald, W.C. Becket and K.L. Wattier, J. AWWA, 77(4) (1985) 122-132.
ELSEVIER
www.elsevier.com/locate/desal
A comparative study on the photocatalytic degradation of humic substances of various origins Ceyda Senem Uyguner, Miray Bekbolet* Institute of Environmental Sciences, Bogazici University, 34342, Bebek, Istanbul, Turkey TeL +90 (212) 359-7012; Fax: +90 (212) 257-5033; email: [email protected] Received 26 October 2004; accepted 5 November 2004
Abstract The photocatalytic removal of model humic and fulvic acids of different origins (terrestrial and aquatic) was investigated using TiO2 Degussa P-25 as the photocatalyst. The results were presented comparatively based on pseudofirst-order and the Langmuir-Hinshelwood kinetic models. The reaction rates calculated for the UV-vis parameters follow the general trend: IHSS soil humic acid > Aldrich humic acid > IHSS humic acid > Roth humic acid > IHSS fulvic acid. However, the rate constants show differences depending on the UV-vis absorption parameters. The evaluation of the Langmuir-Hinshelwood kinetic model reveals that, in all cases, Roth hurnic acid has the lowest Langmuir-Hinshelwood rate, rate constant and adsorption coefficient, and the ordering of the rates follows the same trend with that of pseudo-first order. Furthermore, data related to the molecular size distribution profiles of raw and treated humic acids suggest that the photocatalytic degradation occurs irrespective of the molecular size fractions. Keywords: Humic acid; Fulvic acid; Photocatalysis; Langmuir-Hinshelwood kinetics; Molecular size fractionation
1. Introduction Humic substances are the predominant type of natural organic matter present in ground and surface waters. On the basis o f their solubility in water and as a function o f p H , humic substances are generally classified into humic acids (HA),
fulvic acids (FA) and humin. HA are comprised o f high-molecular-weight organic substances that are soluble in alkaline media and insoluble in acidic media, whereas FA comprise moderate molecular weight organic substances of nonspecific composition that are soluble at all pH values.
*Corresponding author. Presented at the Seminar in Environmental Science and Technology: Evaluation of Alternative Water Treatment Systems for Obtaining Safe Water. Organized by the University of Salerno with support of NATO Science Programme. September 27, 2004, Fiseiano (SA), Italy. 0011-9164/05/$- See front matter © 2005 Elsevier B.V. All rights reserved doi: 10.1016/j.desal.2004.11.006
168
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
Humic substances are known to be a complex class ofbiogenic polyelectrolytes that can interact with organic and inorganic substances in the environment. Imparting undesirable color and acting as precursors for undesirable trihalomethane formation during the chlorination process, their presence in drinking water supplies causes problems, the consequences of which may lead to the deterioration of human health [1,2]. Hence, their removal from water supplies has received specific attention. During the last decade, TiO2-mediated heterogeneous photocatalysis has emerged as an alternative advanced oxidation technology for the removal ofhumic substances from drinking water supplies. Since 1996, Bekbolet and coworkers have been studying the photocatalytic degradation of HA in detail, focusing on operational parameters that affect the process efficiency [3-6]. Complementary to previous research and in parallel to the recent results of a study where UV-vis spectroscopic properties of humic substances have been investigated, this study presents data on the photocatalytic removal of model HA and FA of different origins using TiO2 Degussa P-25 as the photocatalyst [7]. The UVvis properties ofhumic substances are utilized to assess the degradation kinetics that is presented comparatively based on pseudo-first-order and Langmuir-Hinshelwood kinetic models [5,8]. Aqueous solutions of HA are generally polydisperse, with size ranges differing according to their origins, thereby strongly affecting treatment efficiencies. The oxidative behavior of humic substances towards photocatalytic degradation could also be evaluated by use of molecular size distribution profiles.
2. Experimental HA and FA of different origins (terrestrial and aquatic) were used in bench-scale photocatalytic oxidation experiments. The Suwannee River
fulvic acid (IHSS FA) and Suwannee River humic acid (IHSS HA) standard materials, isolated from the Suwannee River, Georgia, as well as soil humic acid (IHSS SHA) (Lot no. 1S 102H) as a terrestrial source were purchased from the International Humic Substance Society. Commercial HA samples were supplied by Aldrich (humic acid salt) (AHA) and Roth (humic acid) (RHA). The term "humic substances" is used throughout the text to define humic and fulvic acids. TiO 2 Degussa P-25 was used as the photocatalyst. Photocatalytic oxidation of humic substances (50 mg L-l) was carried out at pH 6.5+0.5 according to a previously outlined procedure using 0.25 mg mL -l of TiO 2 Degussa P-25 [5]. After each run, the absorption of the supernatant was determined using a Shimadzu UV160A double-beam spectrophotometer at 436 nm (Color436, m-l), 400 nm (Color400, m-l), 365 nm (UV365, m-l), 300 nm (UV300, m-l), 280 nm (UV2so, m -1) and 254 nm (UV254, m -l) for the evaluation of the kinetics of the photocatalytic degradation ofhumic substances. The absorbance values measured at 254 nm (UV254) and 280 nm (UV2s0) are known to explain aromaticity removal whereas, 436 nm is used to monitor decolorization. Furthermore, specific UV absorbance (SUVA254, m -l mg -1 L) was used to represent total organic carbon (TOC) normalized aromatic moieties. TOC (mg L-l) measurements of humic substances were performed on a Shimadzu TOCV CSH TOC analyzer. Calibration of the instrument was done using potassium hydrogen phthalate in the concentration range of 5-25 mg L -1. Molecular size fractionation of the raw and photocatalytically treated humic substances (50% degradation with respect to Color436) was performed by ultrafiltration (Amicon) through a sequence of membranes of decreasing pore sizes in the range of 100,000 to 1,000 Dalton. At the beginning of each run, the samples were filtered through 0.45 #m (approximately 450 kDa) Millipore membrane filters.
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
3. Results and discussion 3.1. Sample characterization
Light absorption in the UV-vis region is a typical property ofhumic substances. A variety of absorption wavelengths and ratios has been proposed for the spectroscopic characterization of humic substances [9-14]. Amongst them, UV254 is interchangeably measured with TOC as a surrogate parameter to indicate the natural organic matter content in natural waters [15]. The UV absorptivity at 280 nm is known to represent total aromaticity for phenolic arenes, benzoic acids, aniline derivatives, polyenes and polycyclic aromatic hydrocarbons with two or more rings that are considered as common structural subunits in humic matter [ 10,11 ]. Furthermore, absorption values at 400 and 436 nm are used to measure color removal ofhumic substances. In accordance with the reported diverse nature of UV-vis data, the specified UV-vis parameters are recorded for the assessment of the photocatalytic degradation kinetics. Depending on the concentration of the working solutions, UV-vis and TOC parameters of the humic substances display certain values as given in Tables 1 and 2. Humic acids of terrestrial origin exhibited Color436 values in the range of 18.8-38, whereas HA of aquatic origin, IHSS HA, had a relatively lower content of colorforming moieties. Obviously, IHSS FA displayed quite low color characteristics. The UV365 parameter is found to be irrespective of the source of the samples as expressed in a decreasing order of RHA > IHSS SHA > IHSS HA > AHA > IHSS FA. On the other hand, RHA has remarkably high UV280 and UV254values with respect to the other samples emphasizing strong aromatic character. The order of UV-absorbing moieties of the samples is followed by a decreasing trend as RHA > IHSS HA > IHSS SHA > AHA > IHSS FA, reflecting non-source dependency. The TOC contents of the humic substances are expressed in
169
a decreasing order of IHSS FA> RHA > IHSS HA > AHA > IHSS SHA. 3.2. Kinetics o f photocatalytic degradation
The kinetics of the photocatalytic degradation of humic substances is known to obey pseudofirst order as expressed by the equation [5,8,16, 17]: R = - dA/dt = k A
(1)
where R is the pseudo-first-order rate (m- Jmin-~), A the specified UV-vis parameters of humic substances (m-l), t the irradiation time (rain) and k the pseudo-first-order reaction rate constant (min-'). The evaluation of the kinetic data revealed the following model parameters for C010r436, UV365, UV2so and UVz54 as presented in Table 1. The results of the relevant data for Color400 and UV300 are only discussed for simplicity purposes. The color-forming moieties of humic substances are usually expressed with C010r436and/or Color400 values; hence, the removal of similar moieties might be considered. A comparison of the rate constants of Color436 and Color4oo for the photocatalytic degradation of HA in a diverse solution matrix indicates that the difference is <10% [4]. Accordingly, in this study the rate constants calculated for C010r436 and Color400 differ from each other by 4 to 7%; hence the C010r436 parameter was preferred. Recent results indicate the importance ofabsorbance ratios such as E254/E436, E28o/E36s and their respective changes during the photocatalytic degradation of HA [ 14]. Therefore, the removal ofUV36 s forming moieties was also analyzed in this study. The degradation rates calculated for all the indicated parameters follow the general trend: IHSS SHA > AHA> IHSS HA > RHA > IHSS FA. However, the rate constants show differences depending on the UV-vis parameters employed. The ordering is similar for Color436 and UV36sand
170
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
Table 1 Pseudo-first-order kinetic parameters of photocatalytic degradation of model compounds (R 2 >0.95); Ao, initial value of specified UV-vis parameters (m -~)
Color436 IHSS FA IHSS HA IHSS SHA AHA RHA
A o, m -1
k, min -t
6/2, min
R, m -~ rain -1
7.22× 10 -3 8.30× 10 3 1.35x 10 -2 1.60× 10 2
0.0368 0.132 0.379 0.301 0.115
5.1 15.9 28.1 18.8 38
3.04x 10 -3
96 84 51 43 228
19.4 44.3 48.4 39.2 72.9
6.46× 10 -3 7.03 x 10 -3 1.24x 10 -2 1.49x 10 -2 2.69x 10 -3
107 99 56 47 258
0.125 0.311 0.600 0.584 0.196
68.4 120.2 99.3 92.1 159.4
4.02× 10 -3 5.45x10 -3 1.10x 10 -2 1.09x 10 -2 2.13×10 -3
172 127 63 64 325
0.275 0.655 1.092 1.004 0.339
91.1 148.1 114.9 108.1 186.3
3.76x 10 -3 4.90x 10 -3 1.07× 10 -2 1.03x10 2 2.01 ×10 -3
184 141 65 67 345
0.342 0.726 1.229 1.113 0.374
UV365 IHSS FA IHSS HA IHSS SHA AHA RHA
UV28o IHSS FA IHSS HA IHSS SHA AHA RHA
UV2s4 IHSS FA IHSS HA IHSS SHA AHA RHA
Table 2 Pseudo-first-order kinetic parameters o f photocatalytic degradation of model compounds in terms of TOC (R2>0.95) TOC
TOCo, mg L-i
k, min-l
tl/2, min
R, mg L-l min-1
IHSS FA IHSS HA IHSS SHA AHA RHA
20.59 16.16 15.48 15.5 19.53
1.07× 10 -2 2.65x 10 -3 2.10x 10 -2 9.75×10 -3 3.02 x 10 -3
65 261 33 71 230
0.220 0.0501 0.339 0.151 0.0590
d e c r e a s e s in t h e f o l l o w i n g o r d e r , A H A > I H S S SHA> IHSS HA> IHSS FA> RHA. IHSS SHA a n d A H A e x h i b i t q u i t e s i m i l a r UVzs 0 a n d UV254
degradation rate constants that could be cons i d e r e d as d i f f e r i n g i n s i g n i f i c a n t l y . T h e l o w e s t d e g r a d a t i o n r a t e c o n s t a n t is o b t a i n e d for R H A
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
followed by IHSS FA and IHSS HA for both of the UV absorbing moieties (UV2s0 and UV254)that signify benzene such as polycyclic aromatic hydrocarbons with two or more rings. On the other hand, UV300 removal rates are found to be relatively lower than UV365 but exhibit higher removal rates than UVzso. In accordance with the low photocatalytic degradation rate constants, quite high half-life values for all of the parameters are expected. These values are presented in Table 1. FA are generally characterized by their low molecular weight and a smaller number of total and aromatic carbons than their HA counterparts. Unexpectedly, their respective removal rates in terms of the specified UV-vis parameters are found to be relatively lower than the removal rates of HA that have longer chain fatty acid products and a higher hydrophobicity than FA. The structural and compositional characteristics of the samples could possibly affect the photocatalytic degradation pathway leading to various degradation products comprising composite UVvis properties. Considering that UV254 is generally used as a surrogate parameter for TOC, the rate constants calculated for TOC removal might be expected to follow the same trend observed for UV254. However, UV254 rate constants for IHSS FA, IHSS SHA, and RHA are relatively low compared to the TOC rate constants (Tables 1 and 2). Contrary to that, lower values for the rate constants of TOC were obtained for IHSS HA and AHA with respect to the rate constants calculated for UV254 removal. For IHSS FA, the UV254rate constant is one-third of the rate constant calculated for TOC. This indicates that low-molecular-weight and fewer UV-absorbing compounds are formed. Con-equently, the half-life values for UVz54 removal are ordered as IHSS SHA > AHA > IHSS HA > IHSS FA > RHA, and the TOC halflife values follow the decreasing trend as IHSS SHA > IHSS FA > AHA > RHA > IHSS HA (Tables 1 and 2).
171
Prior to photocatalysis, initial adsorption levels were determined as 6, 7, 15, 30, and 47% TOC removal for IHSS FA, IHSS HA, RHA, AHA and IHSS SHA, respectively. As verified by the discrepancy between TOC values, the initial adsorption oflHSS SHA onto TiO 2 is reflected as high removal rates in terms of the specified UVvis parameters and TOC. However, for the other samples, removal rates were not found to be in accordance with the initial adsorption trends. The reason for this anomaly might be explained by the conformational changes present in the adsorbed state of HA and FA samples on TiO 2. In view of the generally presented optimum loading of 1.0 mg mL -j of TiO2, photocatalytic experiments were also carried out under higher photocatalyst conditions [3]. Color436degradation rate constants are found to be 4.30×10 .2, 4.19× 10-2, 2.58x10 -2, 1.52x10 -2, and 8.31×10 -3 for IHSS SHA, AHA, IHSS FA, IHSS HA and RHA, respectively. Following the same decreasing trend, considerably lower UV254rate constants are calculated as 3.25×10 -2, 2.49×10 -2, 1.22×I0 -2, 1.08× 10 -2 and 6.22× 10 -3 for the photocatalytic degradation of the humic substances. The most prominent effect was observed for IHSS FA with a four-fold increase of the Color436removal rate. On the other hand, the effect of increasing the photocatalyst loading by fourfold affected the removal rate of IHSS HA by a two-fold increase that is comparatively less than the other samples investigated (Table 1). A similar trend was observed for the UV254removal rates. For IHSS FA and IHSS SHA, increasing TiO z loading has the effect of increasing UV254 removal rates by approximately threefold. It could be stated that increasing the photocatalyst loading obviously results in an enhancement of the removal efficieneies. However, to assess a precise UV-vis parameter evaluation for molecular size distribution data after photocatalytic treatment, the use of 0.25 mg mL-1 photocatalyst loading was preferred.
172
c.s. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
Considering that photocatalysis occurs at the surface, the concentration ofsubstrate adsorbed to the surface directly affects the overall rate of adsorption. A Langmuir-Hinshelwood model is widely used to describe the kinetics of semiconductor photocatalysis [18,19]. The rate o f degradation o f humic substances in terms o f C010r436, UV254 and other UV-vis parameters is expressed by the following empirical type relationship:
= kL. KL. A/(1 +
A)
(2)
where RLrt is the rate o f the reaction (m -~ min-J), kLrt the reaction rate constant (min-1), KLH the adsorption coefficient o f the reactant onto TiO2 under irradiation conditions (m) and A the initial value of the specified UV-vis parameters of humic substances (m-1).
A new series of photocatalytic experiments was designed to assess the Langmuir-Hinshelwood kinetics utilizing 20-50 mg L-1 humic substances in the presence o f 0.25 mg mL -1 TiOz Degussa P-25. The data are evaluated in terms of the above-given Langmuir-Hinshelwood model, and the related parameters are given in Table 3. For all of the HA, the L-H rate constants increase in the order offolor436, UV365, UV2~0and UV254, revealing that UV absorbing centers may be removed more selectively than color-forming moieties. However, within the studied sample pool, the ordering o f the rate constants follows the general trend IHSS SHA > AHA > IHSS HA > RHA for all of the parameters. Contrary to the rate constants, the adsorption coefficients display a decreasing trend o f Color436, UV365,UV280and UV2s4 for all o f the samples.
Table 3 Langmuir-Hinshelwood kinetic parameters of humic substances (rate was calculated for 50 mg L -~ of each humic substance)
IHSS HA
Color436
UV365
UV280
UV254
kEn, min-t KEn, m
0.0903 1.47 93 0.0866
0.206 0.337 118 0.193
0.435 0.134 150 0.410
0.469 0.0838 176 0.434
0.244 0.398 65 0.224
0.399 0.345 66 0.376
0.714 0.120 78 0.659
0.805 0.105 80 0.743
0.188 0.584 56 0.172
0.376 0.360 57 0.351
0.627 0.122 83 0.576
0.697 0.109 87 0.642
0.0698 0.217 318 0.0623
0.115 0.0936 381 0.100
0.195 0.0395 499 0.168
0.214 0.0332 533 0.184
ttl2, min m -~ min-I IHSS SHA kLn, min-j KEn, m tt/2, min RLn,m -I min-t AHA kLn, min-i KLH,m fia, min RLr~,m-l min-1 RHA ken, min -l KL~,m tt/2, min RLn, m- i min- 1 RLn ,
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
A comparison within the samples results in different orders such that; Color436 and UV280 exhibit similar trend as, IHSS HA> AHA> IHSS SHA> RHA and UV365 and UV254 are ordered as AHA> IHSS SHA> IHSS HA> R_HA. The effect of the model parameters on the overall rate equation is expressed by the L-H rates following the same trend of the L-H rate constants. Considerably higher L-H degradation rates are achieved for UV254 than C010r436 for all of the samples. Considering the related half-life values, the following orders could be presented for Color436 and UV365 as: AHA < IHSS SHA< IHSS HA < RHA; and for UVaso and UV254 as: IHSS SHA < AHA < IHSS HA < RHA. The ordering of the L-H rates is similar to that of first-order rates; however, the calculated L-H rates for the HA are approximately half the value of the pseudo-first-order rates for all of the studied parameters. The discrepancy observed in the kinetic modeling of the photocatalytic degradation of the humic substances could not be clearly explained - - neither by the properties of the parameters nor by the character of the samples.
3.3. Evaluation of molecular size distribution
profiles The effect of the photocatalytic degradation pathway could be visualized by the structural changes that could be resolved in terms of molecular size distribution profiles. To achieve this goal, a new series of photocatalytic degradation experiments was performed to prepare samples that were partially oxidized as 50% with respect to Color436 to follow the possible changes in molecular size distribution. Molecular size distribution characteristics of raw and photocatalytically treated HA samples are presented through the SUVA254 values 0 f each fraction provided that the SUVA254 parameter comprises the counterbalancing effect of UV254 and TOC (Figs. 1 and 2). It is known that the molecular size distribution profiles of the humic substances exhibit a general decreasing trend from high molecular size fractions to the lower molecular size fractions irrespective of the origin and source of the organic matter. Depending on the explanation that the specific UV absorbance (SUVA254)parameter can be used to describe the composition of humic
8.00 7,00 "7
173
6.00 5.00 4.00 3.00 2.00 1.00 0.00
450-100 kDa 100-30kDa 30-10kDa 10-1 kDa Molecular size distribution
<1 kDa
Fig. 1. SUVA254values for raw humic acid samples with respect to molecular size distribution.
174
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
8.00 7.00 "7
6.00 5.00
"7
4.00 3.00 2.00 1.00
0.00 450-100 kDa
100-30kDa
30-10kDa
10-1 kDa
<1 kDa
Molecular size distribution
Fig. 2. SUVA254values for photocatalytically treated humic acid samples with respect to molecular size distribution.
material in terms of hydrophobicity and hydrophilicity, SUVA254>4 indicates mainly hydrophobic and especially aromatic moieties while a SUVA254 <4 represents a hydrophilic organic fraction [20]. The hydrophobic character predominates for humic samples up to a size fraction of 10 kDa, whereas a comparatively hydrophilic character could be assessed for smaller molecular size fractions as 10-1 kDa and even smaller than 1 kDa fraction (Fig. 1). All of the studied HA samples exhibited close similarities in terms of SUVA254 up to the 30 kDa fraction. The medium size fractions of the 3010 kDa region of aquatic originated IHSS HA displayed a comparatively higher SUVA254value than the others that were of terrestrial origin. IHSS SHA has a quite low SUVA254 value whereas IHSS HA of aquatic origin still revealed a higher SUVA254 than the others for the 101 kDa fraction. On the other hand, the AHA sample acquired a comparatively high SUVA254 for fractions <1 kDa. As a general trend for all of the HA investigated, SUVAz54 values decrease after photocatalytic treatment indicating that the degradation efficiency is irrespective of molecular size fraction (Fig. 2). The primary effect ofphotocatalytic
treatment was observed in the 450-100 kDa fraction for all the HA except RHA. The probable resistance displayed by the 450--100 kDa fraction of RHA to photocatalytic degradation could also be explained by the formation of high-molecularweight fractions (450-100 kDa) via the combination of the photoproducts originating from lower molecular size fractions. A slight change is observed for SUVA254 of IHSS HA irrespective of the size fraction up to 10 kDa. IHSS HA and RHA retained a hydrophobic character up to the 10 kDa size fraction whereas AHA displayed more hydrophilic character for size fractions smaller than 30 kDa. IHSS SHA expressed a quite different photocatalytic reactivity towards the formation of lower molecular size fractions (<100 kDa) that could be considered as possessing more hydrophilic properties since SUVA254 values are found to be between 2-3. For AHA and IHSS SHA samples increased SUVA254 values are noticed for the smallest size fractions. The size fraction < 1 kDa for IHSS SHA exhibits a two-fold increase in SUVA254 relative to that of the 10 kDa fraction. Oxidation reactions taking place during photocatalysis may lead to a shift as an increase in lower molecular size fractions that is apparently observed in the case of
C.S. Uyguner, M. Bekbolet ~Desalination 176 (2005) 167-176
IHSS SHA and AHA. This could be explained by relatively high pseudo-first-order and L-H UV254 degradation rates as well as the TOC removal rates of the IHSS SHA and AHA (Tables 1 and 2). It is widely known that during photocatalytic degradation the main oxidative pathway is achieved by the nonselective attack o f . O H radicals. The different reactivities observed for each size fraction could be attributed to their structural and conformational diversity. The elucidation of the complex chemistry involved in these photocatalytic pathways will likely only be achieved through the application of sophisticated analytical tools.
4. Conclusions On the basis of their diverse chemical and physical properties such as molecular weight, molecular size, elemental composition and source of origin, substantial differences are observed in photocatalytic removal efficiencies ofhumic and fulvic acids. The pseudo-first-order rate constants calculated for Color436 of each of the humic substances show a decreasing trend as follows: AHA > IHSS SHA > IHSS FA > IHSS HA > tLHA. On the other hand, the order of the rate constants of aromaticity removal (UV254) is as follows: IHSS SHA > AHA > IHSS HA > IHSS FA > RHA. The rate constants calculated for TOC removal exhibit a different trend: IHSS SHA > IHSS FA > AHA > RHA > IHSS HA. Higher removal rates have been achieved with UV254 values compared to that of Color436values. All of the studied HA exhibit higher pseudo-firstorder removal rates compared to that for FA. The ordering of the L-H rates is similar to that of first-order rates; however, the calculated L-H rates for the HA are approximately half the value of the pseudo-first-order rates for all of the studied UV vis parameters. Comparison of the molecular size distribution profiles of the photocatalytically treated HA with
175
those of raw untreated samples reveals the confirmation of the formation of lower molecular size fractions exhibiting low SUVA254values.
Acknowledgements The support provided by the Research Fund of Bogazici University, Project No. 03S107 and NATO Grant EST.CLG.980506 is gratefully acknowledged. The valuable contribution by Patrick Dunlop and Marc Anderson is appreciated.
References [1] J.J. Rook, Water Treat. Exam., 23(2) (1974) 234243. [2] P.C. Singer, Water Sci. Technol., 40(9) (1999) 2530. [3] M. Bekbolet, J. Env. Sci. Health, A31(4) (1996) 845-858. [4] M. Bekbolet, Z. Boyacioglu and B. Ozkaraova, Water Sci. Technol., 38 (1998) 155-162. [5] M. Bekbolet, A.S. Suphandag and C.S. Uyguner, J. Photochem. Photobiol. A: Chem., 148 (2002) 121128. [6] C.S. Uyguner and M. Bekbolet, Intemat. J. Photoenergy, 6(2) (2004) 73-80. [7] C.S. Uyguner and M. Bekbolet, Desalination, 176 (2005) 47-55. [8] M. Bekbolet and I. Balcioglu, Water Sci. Technol., 34(9) (1996) 73-80. [9] H. De Haan, T. De Boer, H.A. Kramer and J. Voerman, Water Res., 16(6) (1982) 1047-1050. [10] S.J. Traina, J. Novak and N.E. Srneck, J. Environ. Quality, 19(1)(1990) 151-153. [11] Y.P. Chin, G. Aiken and E. O'Loughlin, Environ. Sci. Technol., 28(11) (1994) 1853-1858. [12] J. Peuravuori and K. Pihlaja, Anal. Chimica Acta, 337 (1997) 133-149. [13] J. Chen, B. Gu, E.J. LeBoeuf, H. Pan and S. Dai, Chemosphere, 48 (2002) 59-68. [14] C.S. Uyguner and M. Bekbolet, submitted for publication. [15] I.N. Najm, N.L. Patania, J.G. Jacangelo and S.W. Krasner, J. AWWA, 86(6) (1994) 98-106.
176
C.S. Uyguner, M. Bekbolet / Desalination 176 (2005) 167-176
[16] J. Wiszniowski, D. Robert, J. Surmaez-Gorska, K. Miksch and J.V. Weber, Intemat. J. Photoenergy, 5 (2003) 69-74. [17] R. A1-Rasheed and D.J. Cardin, Appl. Catal. A: General, 8501 (2003) 1-10.
[18] R.W. Matthews, J. Catal., 111(2) (1988) 264-272. [19] C. Turchi and D.F. Ollis, J. Catal., 119 (1989) 483496. [20] J.K. Edzwald, W.C. Becket and K.L. Wattier, J. AWWA, 77(4) (1985) 122-132.