Environ. Sci. Technol. 2002, 36, 5342-5347
Photocatalytic Atrazine Degradation by Synthetic Minerals, Atmospheric Aerosols, and Soil Particles MARION LACKHOFF AND REINHARD NIESSNER* Institute of Hydrochemistry, Technical University of Munich, Marchioninistrasse 17, D-81377 Munich, Germany
In this work, the photocatalytic atrazine degradation by seven synthetic minerals and five environmental particle samples was examined to investigate a possible contribution of photocatalysis to the abiotic degradation of atrazine in the environment. Particle suspensions containing 500 ng/L atrazine were irradiated with a sun simulator, and the atrazine degradation was monitored by enzyme-linked immunosorbent assay (ELISA). Atrazine detection by ELISA proved to be an useful analytical tool because of low crossreactivity of atrazine metabolites and high sensitivity with detection limits in the lower nanograms per liter range. The atrazine degradation followed first-order kinetics, and the obtained rate coefficients were compared with the rate of direct photolysis. Known photocatalysts, such as TiO2 and ZnO, showed the expected fast photocatalytic degradation (k ) 27-327 × 10-3 min-1) of atrazine. The degradation rates detected upon irradiation of titanium-, zinc-, or iron-containing minerals were orders of magnitudes lower (k ) 0.15-0.70 × 10-3 min-1) but still significantly faster than direct photolysis without particles (k ) 0.10 × 10-3 min-1). With environmental particle samples (soot, fly ash, sand, road dust, and volcanic ash), however, no significant photocatalytic activity was observed (k ) 0.070.16 × 10-3 min-1). The atrazine degradation rates were in the range of direct photolysis. Thus photocatalysis by aerosol or soil particles appears not to enhance abiotic atrazine degradation in the environment.
Introduction In the past decades, photocatalytic degradation of harmful pollutants by irradiated particles has received considerable attention. Many persistent organic substances are rapidly degraded in aqueous suspension of TiO2 particles irradiated with simulated sunlight (1, 2). Most investigations so far were carried out with aqueous suspensions of TiO2 (2, 3), but other mineral particles such as ZnO or Fe2O3 also showed photocatalytic activity (4-7). Taking into account the wide range of mineral particles in the environment (soils or atmospheric particles such as dust, fly ash, or volcanic ash), it can be assumed that photocatalytic reactions induced by solar radiation may occur in the natural environment. Since most photocatalytic reactions are initiated by photon absorption of UV radiation, an enhancement of UV radiation in the earth’s atmosphere will enhance possible photocatalytic reactions. It is therefore of special interest to investigate the photocatalytic potential of atmospheric aerosol and soil * Corresponding author phone: 49 89 7095 7980; fax: 49 89 7095 7999; e-mail:
[email protected]. 5342
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 24, 2002
particles. Only a few experimental studies on photocatalytic reactions using real atmospheric aerosols or soil samples have been described in the literature by now. In 1997, Isidorov et al. (8) reported the degradation of halocarbons and monoterpenes by different synthetic and natural photocatalyst (TiO2, ZnO, Fe2O3, desert sand, volcanic ash, sea salt, and chalk). Titanium and iron oxides are suspected to be responsible for the photodegradation of fuel additives such as ethanol and methyl tert-butyl ether on irradiated fly ash (9). However, it is still unclear whether environmental particles exhibit sufficient photocatalytic activity to degrade persistent organic substances (e.g., the herbicide atrazine). Atrazine [2-chloro-4-(ethylamino)-6-(isopropylamino)-striazine], a representative substance out of the class of triazine herbicides, is widely used for weed control in the cultivation of corn and other crops. Although it is banned or regulated in some countries (e.g., banned in Germany in 1991), atrazine is still found in many environmental samples. Concentrations in surface water are usually in the range of 100 ng/L or less (10), but levels above 1 µg/L have also been reported (11). Because of volatilization losses from fields (12) and atmospheric transport, atrazine is ubiquitously distributed in the environment (ref 13 and references therein). In rain and fogwater, atrazine concentrations up to 70 µg/L have been measured (14, 15). Atrazine is characterized by a lifetime of days up to years in different environmental compartments (10, 15, 16). However, it is rapidly degraded in aqueous suspension of TiO2 irradiated by UV radiation (17-23). Goolsby et al. (15) suggested a photocatalytic degradation of atrazine in the atmosphere because of the high deethylatrazine to atrazine ratio found in rainfall. In this study, induced atrazine degradation was studied on model minerals and real particle samples. The model minerals can be divided into known photocatalysts (TiO2, Fe2O3, ZnO, ZnS) and mixed oxides (FeTiO3, SrTiO3) representing naturally occurring minerals containing Ti and/or Fe. All investigated minerals are characterized by band gaps between 2.2 and 3.9 eV, which correspond to maximum absorption wavelengths in the visible or UV range of the sun spectrum (24, 25). The real samples were chosen to cover a range of typical natural and anthropogenic environmental particles (sand, soot, dust, fly ash, volcanic ash). The photocatalytic activity of the particles was determined by measuring the degradation rates of atrazine in comparison to direct photolysis without particles.
Experimental Section Materials and Reagents. The following commercially available metal oxides were used as photocatalyst: TiO2 (P 25, Degussa, Frankfurt am Main, Germany); TiO2 (anatase, Chempur, Karlsruhe, Germany); ZnO (Merck, Darmstadt, Germany); Fe2O3 (Sigma-Aldrich, Steinheim, Germany); ZnS, FeTiO3, and SrTiO3 (Alfa Aesar, Karlsruhe, Germany). Real samples were collected from soot (mixed soot separated from a vehicle test stand of MAN AG, Nu¨rnberg); fly ash (sample from the electrostatic filter of a corundum furnace; Lonza AG, Waldshut-Tiengen, Germany); sand (soil sample of Agades, Republic of Niger); road dust (Arizona test dust, Powder Technology Inc., Burnsville, AZ); and volcanic ash (soil sample from the mount Vesuvius, Italy). Atrazine was purchased from Riedel-de Hae¨n (Seelze, Germany); hydrogen peroxide, nitric acid, and hydrofluoric acid (all “suprapure”) were from Merck (Darmstadt, Germany); and Aerosil 200 was from Degussa (Frankfurt am Main, Germany). The urban particulate matter (Standard Reference Material 1648) for total-reflection X-ray fluorescence analysis (TXRF) was 10.1021/es025590a CCC: $22.00
2002 American Chemical Society Published on Web 11/07/2002
FIGURE 1. Schematic cross section of a reaction cell. The dimensions are given in mm. The whole system consists of a base plate (220 × 250 mm) with eight reaction cells (each 71.5 × 21.5 × 5 mm). received from NIST (Gaithersburg, MD), and the germanium standard and boric acid were from Merck (Darmstadt, Germany). Irradiation was carried out with a sun simulator (SOL 500, Dr. Ho¨nle AG, Planegg, Germany), which is equipped with a metal halide high-pressure (model HQI from Osram) lamp. The aqueous suspensions were filled in a home-built array of irradiation chambers consisting of a brass plate with an internal cooling loop to control and adjust the temperature. Eight reaction cells (Airr ) 1540 mm2, Vmax ) 7.7 mL) are sealed by a PTEF plate. A brass cap covers the reaction chamber with glass slides. The illuminance that is penetrating the reaction chambers through the glass slides is similar to the sun spectrum (λ ) 295 up to 3000 nm) and amounts to 115 klux. Variations of the illuminance between the different reaction chambers are less than 10%. The setup of a reaction cell is outlined in Figure 1. During irradiation, the whole system was shaken using a self-acting shaker (Titamax 100, Heidolph, Schwabach, Germany). Flat-bottom polystyrene 96-well microtiter plates were purchased from Greiner (Nu ¨ rtingen, Germany). A washer (Columbus, SLT, Crailsheim, Germany), a shaker (Easyshaker EAS 2/4, SLT, Gro¨dig, Austria), and a reader (340ATTC, SLT, Crailsheim, Germany) for microtiter plates were used. Goat anti-mouse IgG was obtained from ICN Biomedicals (Eschwege, Germany); TMB (3,3′,5,5′-tetramethylbenzidine) and H2O2 (30%) were from Merck (Darmstadt, Germany). The monoclonal antibody 4A54 can be obtained from Connex (D-82152 Martinsried, Germany). The microwave oven used to digest real samples was a Miele Supratronic M 750 with 30-mL PTFE vessels (both Berghof, Eningen, Germany). The digested samples were analyzed by TXRF (Extra IIA, Atomika Instruments, Oberschleissheim, Germany). Degradation Experiments. A suspension of 5 g/L catalyst and 500 ng/L atrazine in distilled water or in phosphatebuffered saline (PBS; 10 mmol/L potassium dihydrogen phosphate, 70 mmol/L dipotassium hydrogen phosphate, and 145 mmol/L sodium chloride, pH 7.6) was suspended for 15 min in an ultrasonic bath and subsequently saturated with oxygen for 30 min. Three milliliters of the well-stirred suspension was filled in each reaction cell of the irradiation chamber. The temperature was held at 20 °C during the whole experiment. The samples were shaken for a further 20 min in the dark to obtain a temperature and adsorption equilibrium. The overall sample preparation took around 90 min. Since dark adsorption of atrazine on TiO2 was reported to equilibrate within 30 min (23), this time was suggested to be
sufficient to attain an adsorption equilibrium. Before starting the irradiation, a first sample was drawn by transferring the whole sample (3 mL) in a plastic cup, which was stored in the dark until centrifugation. During the irradiation, seven further samples (3 mL) were taken from the well-shaken suspension after different time intervals. The pH value of each sample was measured, and 1 mL from each sample was centrifuged for 5 min at 25850g and 4 °C. The supernatant liquid was stored at 4 °C until analysis. If the pH value exceeded 8, the solution was diluted 1:5 (v/v) in PBS to obtain a suitable pH value for ELISA determination. Atrazine Analysis. Atrazine was detected by a direct, competitive ELISA, which is described elsewhere (26, 27). Microtiter plates were coated 14 h with 200 µL/well of goat anti-mouse IgG diluted 1:3000 in coating buffer (15 mmol/L sodium carbonate, 35 mmol/L sodium hydrogencarbonate, and 3 mmol/L sodium azide, pH 9.9). After three times rinsing with washing solution (1 mmol/L potassium dihydrogen phosphate, 7 mmol/L dipotassium hydrogen phosphate, 15 mmol/L sodium chloride containing 0.02 mmol/L potassium sorbate and 0.05 % v/v Tween 20), the plates were incubated for 3 h with 200 µL/well atrazine antibody (4A54, cell culture supernatant) diluted 1:10 000 in PBS (see Degradation Experiments). Then the plates were washed again three times with washing solution and filled successively with 50 µL/ well PBS, 100 µL/well standard or sample, and 50 µL/well tracer solution (peroxidase-labeled atrazine diluted 1:50 000 in PBS containing 25 mg/L 3,3′,5,5′-tetramethylbenzidine to protect peroxidase; 28). After 25-min incubation, the plates were rinsed three times with washing solution, and then 150 µL/well substrate solution (200 mmol/L potassium dihydrogen citrate and 0.7 mmol/L potassium sorbate containing 0.5 µmol/L tetramethylbenzidine and 12 µmol/L H2O2) was added. After a development time of 10-20 min, the reaction was stopped by adding 100 µL of sulfuric acid (5% v/v), and the absorbance of each well was measured at 450 nm. The values of the standards and the samples were determined by calculating the median of the data (n ) 4). The standard curves were received by fitting the obtained medians with the following four-parameter function:
y)
(A - D) +D x B 1+ C
(1)
[ ( )]
where x is the concentration of atrazine (µg/L), y is the absorbance at 450 nm, A is the maximum absorbance, D is the minimum absorbance, C is the midpoint (IC50 value) (µg/L), and B is the slope parameter. The detection limit was calculated from the blank absorbance value plus a 3-fold standard deviation (3s - definition, see ref 26 for further details). The cross-reactivity was determined by relating the midpoint of the atrazine calibration curve with the midpoint of the atrazine metabolite calibration curve according to
CR(metabolite) )
C(atrazine) C(metabolite)
× 100%
(2)
where CR is the cross-reactivity (%), C(atrazine) is the midpoint of atrazine calibration curve (µg/L), and C(metabolite) is the midpoint of metabolite calibration curve (µg/L) measured on the same microtiter plate as the atrazine standard curve. Analysis of the Elemental Composition of the Particles. Prior to use, the microwave vessels were rinsed carefully with detergent solution and with distilled water. Then they were filled with 10 mL of 1 M nitric acid and treated in the microwave oven under similar conditions as those used for sample digestion. After being cooled, the vessels were flushed again with distilled water. To digest the samples, 5 mg of accurately weighted real sample or reference material (urban VOL. 36, NO. 24, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5343
FIGURE 2. Atrazine calibration curve. The error bars represent the standard deviation of four single measurements. particulate matter) were filled into the cleaned PTFE vessels along with concentrated HNO3 (2 mL), H2O2 (4 mL; 30%), and concentrated HF (0.3 mL). The vessels were closed, placed in the microwave system, and digested using the following procedure modified after ref 29: Step 1: The temperature was ramped to 130 °C (1 °C/s) with the application of 210 W power and held for 8 min at 130 °C, then heated to 140 °C (1 °C/s) at 425 W, followed by a dwell time of 4 min, afterwards warmed to 150 °C (1 °C/s) at 595 W power, held by this temperature 6 min, and finally heated to 175 °C (1 °C/s) at 765 W and held for 15 min. The vessels were removed from the oven and cooled at room temperature for 1 h. Step 2: The vessels were opened, and 2 mL of H2O2 was added. Then the digestion procedure of step 1 was repeated. Step 3: Three milliliters of saturated boric acid in aqueous solution was added to each vessel. The vessels were heated to 130 °C (1 °C/s) at 765 W power with a dwell time of 15 min. After being cooled, the solution was diluted to a final volume of 50 mL. Ten microliters of the sample and 10 µL of germanium standard (1 mg/L) were applied to a quartz slide and analyzed by TXRF. All samples, standards, and blanks were digested twice, and each digest was analyzed in triplicate for a total of six measurements per sample. The digestion method was tested by determining the recovery of the urban particulate matter standard.
Results and Discussion Characterization and Cross-Reactivity of the Atrazine Detection by ELISA. An atrazine calibration curve obtained by ELSIA is shown in Figure 2. The dynamic range comprises about 3 orders of magnitude. The ELISA method is very sensitive with detection limits in the lower nanograms per liter range; in this case, the detection limit amounts to 0.5 ng/L. Only a sample volume of 100 µL is needed. To assess the photocatalytic degradation of atrazine by ELISA, it has to be assured that there is no signal interference of the degradation products with the atrazine signal. In Figure 3, the mechanism of the photocatalytic degradation of atrazine by irradiated TiO2 is shown (19). The cross-reactivities of the most important atrazine metabolites were determined and listed in Table 1. Obviously, the cross-reactivity by the atrazine metabolites is low. The highest cross-reactivity is found by deethylatrazine (3.0%) and deisopropylatrazine (0.73%); the other values are far below 1%. In particular, the fully dealkylated intermediates exhibit very low cross-reactivity. It can be concluded that the atrazine metabolites do not affect the atrazine signal significantly. Furthermore, a good agreement of the obtained cross-reactivities with known literature values (27) is notable. The cross-reactivity of two metabolites (2-amino-4-chloro5344
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 24, 2002
FIGURE 3. Mechanism of the photocatalytic degradation of atrazine by irradiated TiO2 (19).
TABLE 1. Cross-Reactivity of Atrazine Metabolites in Comparison to Literature Valuesa analyte atrazine (1) deethylatrazine (3) deisopropylatrazine (4) dealkylated atrazine (6) ammelide (7) ammeline (5) cyanuric acid (8) hydroxyatrazine (2)
cross-reactivity ( s (%) 100 3.0 ( 0.5 0.73 ( 0.02 0.019 ( 0.001 0.001b ( 0.0003 0.0001b ( 0.00005 0.0001b ( 0.00005
lit. (27) (%) 100 3.4 0.74 0.014
0.000025b
a
The margin of error s is given by the standard derivation of three separate measurements except for extrapolated cross-reactivities, where the margin of error was estimated. b Extrapolated because of solubility problems.
6-hydroxy-1,3,5-triazine and 2-chloro-4,6-dihydroxy-1,3,5triazine) was not determined since they are not commercially available. Because of the low cross-reactivity of 2,4-diamino6-chloro-1,3,5-triazine (substance 6 in Figure 3 with 0.02% cross-reactivity), it can be safely assumed that these metabolites will not influence the atrazine signal either. Degradation Experiments. Figure 4 shows the atrazine concentration by irradiation of three particle systems as function of the irradiation time. In case of TiO2 and ZnO, the degradation followed a firstorder rate equation, which is confirmed by the evidence of a straight line relationship of logarithmic atrazine concentration versus irradiation time. The rate coefficient is given
FIGURE 4. Degradation of atrazine by irradiation of TiO2 (pH 6.1), ZnO (pH 8.0), and soot (pH 3.8) in dependency of the irradiation time; [particle] ) 5 g/L, [atrazine]0 ) 500 ng/L. by the slope of the exponential fit. Appearance of first-order kinetics suggests that the reaction kinetics can be described by the Langmuir-Hinshelwood model (1):
kKC dC ) dt 1 + KC
(3)
where C is the concentration of atrazine, t is the irradiation time, k is the reaction rate coefficient, and K is the adsorption equilibrium constant. Integration of eq 3 results in
()
ln
C0 + K(C0 - C) ) kKt C
(4)
where C0 is the atrazine concentration at t ) 0. For small atrazine concentrations, eq 4 can be simplified to
()
ln
C0 ) kKt ) ktott C
(5)
describing a pseudo-first-order process with the pseudofirst-order rate coefficient, ktot. The suitability of the Langmuir-Hinshelwood model was confirmed exemplarily with TiO2 by varying initial atrazine concentrations from 0.1 to 5 µg/L. The linearity (r 2 ) 0.985) of the double reciprocal plot of initial atrazine concentration versus initial rate (data not shown) indicates that the Langmuir-Hinshelwood model is valid. The reaction rate constant (k) and the adsorption equilibrium constant (K) were found to be 0.416 µg L-1 min-1 and 1.88 L µg-1, respectively. Atrazine concentrations by irradiation of soot (Figure 4) showed a slight but statistically significant decrease with irradiation time. The degradation was too slow to determine the reaction kinetics exactly within the observed irradiation period of 7-30 h. However, longer irradiation times can result in systematic errors due to sedimentation and adsorption of the photocatalyst to the reaction chamber walls as observed in preliminary experiments. Fitting the values with a monoexponential model yielded correlation coefficients just as good as these obtained by a linear fit. The obtained rate coefficients for atrazine degradation were compared with the rate of direct photolysis to check a possible influence of photocatalysis to the abiotic degradation of atrazine in the environment. However, the rate of direct atrazine photolysis may be affected by different pH values. The pH dependence of atrazine photolysis was previously reported by Comber (16), who described an enhancement of photolysis in distilled water at pH 4 in comparison to pH 7 by a factor between 1.1 and 3 depending up the irradiation source. We also found a faster photolysis in acidic solution. In distilled water (pH 5.5), a photolysis rate of 1.3 ( 0.1 ×
FIGURE 5. First-order rate coefficients of atrazine degradation by irradiation of different semiconductor, mineral, and real particulate samples (pH 7.5-8.0, phosphate-buffered saline, [particle] ) 5 g/L, [atrazine]0 ) 500 ng/L). The dashed line represents the value of the photolysis rate plus single standard derivation (0.13 × 10-3 min-1). Below this value, particles were considered to be photocatalytically inactive. 10-3 min-1 was obtained, whereas in buffered solution (pH 7.6) photolysis was 1 order of magnitude slower. Whether the enhancement of the photolysis rate is due to the pH variation or caused by different ionic strengths remains unclear up to now. Thus, to ensure constant conditions, degradation experiments were carried out in buffered suspension. In addition, the experiments were repeated in unbuffered suspension to obtain environmentally realistic conditions. The atrazine degradation rate coefficients observed upon irradiation of synthetic mineral particles and real particle samples at buffered pH (pH 7.5-8.0) are shown in Figure 5. The errors bars represent the scatter of the data points recorded in a single experiment. Irradiation of known photocatalysts such as TiO2 and ZnO caused fast photocatalytic decomposition of atrazine. The reaction rates obtained by TiO2 degradation were nearly the same for P 25 (anatase:rutile ratio 70:30; 3) and pure anatase. Atrazine degradation by irradiated ZnO was around 10 times slower, confirming previously reported results (18). The degradation rates detected by irradiation of tausonite (SrTiO3), ZnS, and iron oxide containing minerals were orders of magnitudes smaller but significantly faster than direct photolysis without particles. The photocatalytic activity of real samples concerning atrazine degradation was low, the reaction rate was in the range of direct photolysis. Atrazine reduction by using irradiated Niger sand colloids was 40% faster than by direct photolysis. Irradiation of soot particles, Arizona test dust, fly ash, and volcanic ash did not enhance the degradation of atrazine, the rate coefficients were within the experimental error of direct photolysis. Furthermore, atrazine degradation by these particles was proved to be statistically insignificant in buffered suspension. Instead, a slight increase of atrazine lifetime was observed in the experiments with fly ash. The decrease of atrazine degradation by these particles will be discussed later. The large error bars calculated in some real sample experiments, for example, soot, can be attributed to the natural heterogeneity of real particle samples. Nevertheless, the observed rate coefficients were reproducible to within (25% as confirmed by repeating Niger sand degradation experiments exemplarily three times (k values: 0.158, 0.119, and 0.151 × 10-3 min-1). Since semiconducting metal oxides are photocatalytically active phases, the amount of metals, especially Ti, Zn, and Fe, in real samples seems to determine their photocatalytic activity. Therefore, the content of Ti, Zn, and Fe of the investigated real samples was measured by TXRF after microwave digestion (as outlined in Table 2). VOL. 36, NO. 24, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5345
TABLE 2. Ti, Zn, and Fe Content of the Investigated Real Samples in wt % and Recovery of Urban Particulate Matter Standard (NIST Standard Reference Material 1648) in % Determined by TXRF after Microwave Digestiona titanium recovery in relation to NIST data (%) Niger sand (wt %) soot (wt %) Arizona test dust (wt %) volcanic ash (wt %) fly ash (wt %) b
zinc
iron
107.2
94.9
96.7
0.25 ( 0.02 0.09 ( 0.05 0.39 ( 0.01 0.97 ( 0.02 0.42 ( 0.01
ndb 0.23 ( 0.01 nd nd nd
1.30 ( 0.03 1.12 ( 0.27 2.85 ( 0.05 7.46 ( 0.03 3.78 ( 0.05
a The errors represent the standard derivation of six measurements. nd, not detectable.
In all real samples, the content of Ti and Zn was below 1 wt %, and titanium amounted about 0.09-0.97 wt %. These values comprised titanium contents of typical environmental samples of 0.16-0.7 wt % found in fly ash, desert sand, and soils (8, 30, 31). Volcanic ash is an example for a real sample with high titanium content (0.97 wt %). Zinc concentrations were below the detection limit of 0.06 wt % except for soot (0.23 wt %). Most abundant of the three metals was iron with contents of 1.1-7.5%. However, the amount of iron seems to be of little importance because of the low photocatalytic activity of Fe2O3 and iron-containing compounds concerning atrazine degradation (see Figure 5). It can be assumed that the Ti and Zn content of the samples was not high enough to lead to a significant photocatalytic degradation of such a persistent substance like atrazine. Even the comparatively high amount of titanium in volcanic ash did not lead to significant atrazine decomposition. Moreover, irradiation of volcanic ash yielded an atrazine degradation that was slower than by the most other real samples. The conclusion can be drawn that other particles characteristics such as size, morphology, or surface constitution were responsible for the slight variations in photocatalytic activity of real samples. Atrazine degradation by irradiated real samples, especially by fly ash and volcanic ash, was slower than direct photolysis. It can be concluded that processes take place that inhibit or mask direct photolysis and possible photocatalysis. This could be attributed to radiation absorption or scattering of the particles, leading to a smaller penetration depth of light in the reaction chamber (32, 33). To minimize these effects, a very small path length of 1.9 mm was chosen in all experiments. Furthermore, the influence of light scattering was estimated by addition of photocatalytically inactive light scattering particles (Aerosil 200 containing >99.8% SiO2) characterized by a UV-Vis spectrum close to those of most of the real samples. No deceleration of direct photolysis was obtained. Thus, an experimental setup was realized where light scattering does not affect the rate of atrazine degradation. It can be concluded that light scattering within this small path length is not responsible for the decrease of atrazine degradation by real particle samples. Another potential reason for the reduced atrazine degradation by irradiated fly ash or volcanic ash is a modification of the adsorption equilibrium during the experiment. Dark adsorption on the studied photocatalysts can be estimated as
