A Mechanistic Model for Mercury Capture with In ... - YONSEI University

TECHNICAL PAPER

ISSN 1047-3289 J. Air & Waste Manage. Assoc. 54:149 –156 Copyright 2004 Air & Waste Management Association

A Mechanistic Model for Mercury Capture with In Situ–Generated Titania Particles: Role of Water Vapor Sylian Rodrı´guez Aerosol and Air Quality Research Laboratory, Environmental Engineering and Science Division, University of Cincinnati, Cincinnati, Ohio Catherine Almquist Department of Paper Science and Engineering, Miami University, Oxford, Ohio Tai Gyu Lee Department of Chemical Engineering, Yonsei University, Seoul, South Korea Masami Furuuchi Department of Chemical Engineering, Kanazawa University, Kanazawa, Japan Elizabeth Hedrick National Exposure Research Laboratory, U.S. Environmental Protection Agency, Cincinnati, Ohio Pratim Biswas Environmental Engineering Science, Washington University, St. Louis, Missouri

ABSTRACT A mechanistic model to predict the capture of gas-phase mercury (Hg) species using in situ– generated titania nanosize particles activated by UV irradiation is developed. The model is an extension of a recently reported model for photochemical reactions by Almquist and Biswas that accounts for the rates of electron-hole pair generation, the adsorption of the compound to be oxidized, and the adsorption of water vapor. The role of water vapor in the removal efficiency of Hg was investigated to evaluate the rates of Hg oxidation at different water vapor concentrations. As the water vapor concentration is increased, more hydroxy radical species are generated on the surface of the titania particle, increasing the number of active sites for the photooxidation and capture of Hg.

IMPLICATIONS Hg is a toxic pollutant whose emissions will be regulated from coal-fired boilers. Low-cost, inorganic titanium dioxide– based sorbents are effective at trapping the Hg in combustion exhausts. Detailed phenomenological models that are developed to predict Hg capture rates will be useful at designing and scaling up processes for use in largerscale coal combustion systems.

Volume 54 February 2004

At very high water vapor concentrations, competitive adsorption is expected to be important and reduce the number of sites available for photooxidation of Hg. The predictions of the developed phenomenological model agreed well with the measured Hg oxidation rates in this study and with the data on oxidation of organic compounds reported in the literature. INTRODUCTION A review mandated by the U.S. Congress in the Clean Air Act (CAA) Amendments of 1990 identified a number of hazardous air pollutants of concern to human health and the environment. Toxic metal emissions were stipulated to be controlled by the use of maximum achievable control technologies (MACT), particularly from combustion systems, which represent one of the largest sources of trace metals. There is a considerable interest in controlling mercury (Hg) emissions because of the risk to public health. It is a persistent toxin that even in small concentrations and short exposures causes serious damage to the nervous, gastrointestinal, and respiratory systems.1–7 In addition, once emitted, it remains airborne for long periods of time (residence time of 0.5–2 yr) before depositing into the soil or water.8 –11 In the ecosystem, it tends to bioaccumulate in the food chain, causing vegetation and Journal of the Air & Waste Management Association 149

Rodrı´guez et al. fish contamination.2– 4,6 Furthermore, physical and chemical transformations of Hg occur in the combustion flue gases, and these processes have been investigated extensively for understanding the transport and the fate of Hg released into the atmosphere.12–16 In some instances, toxic species such as methylmercury can be formed through complicated pathways from a fraction of the Hg emitted from these combustion sources.17–20 Hg itself could be present in different chemical forms in coal combustion flue gases, and reports in the literature are widely varied on the mix between elemental and oxidized states. While the exact chemistry and pathways have not been established, it is conjectured that other species in the coal matrix and quench rates play an important role.21 The speciation of Hg is important in the choice and effectiveness of the Hg capture methodology. Several Hg emission control methodologies have been proposed. Fixed or fluidized beds containing granular activated carbon have been shown to have high capture efficiencies, especially when the carbon (C) is coated/doped with sulfur (S) or iodine (I) functional groups.22–29 Activated C, however, has several disadvantages that limit its applicability in real combustor environments. Its low operable temperature range, the need of regeneration, and its slow adsorption rate are some reasons. Furthermore, its high cost also may limit its usefulness. Other sorbent injection methods have been demonstrated to be effective in capture of toxic metal species by chemisorption.30 –33 Biswas and coworkers34 –36 have developed an in situ– generated sorbent agglomerate that has been demonstrated to be very effective in the capture of trace metals. The agglomerate formation is controlled to obtain an active, highsurface-area sorbent that can effectively capture heavy metal species.35 While silicon (Si, SiO2)- and calcium (Ca, CaO)-based sorbents are effective for capture of several heavy metals, they are not effective for the capture of Hg.37 However, a titanium (Ti, TiO2)-based nanostructured sorbent agglomerate has been demonstrated to be very effective in the capture of Hg in combustor exhausts, when irradiated with ultraviolet (UV) light.30,37 An interesting observation is that the corona present in an electrostatic precipitator may be effective at promoting the TiO2-Hg reaction and the firm binding of it to the sorbent surface;36 thus, external sources of UV irradiation may not be necessary. The method has been shown to be effective in the capture of Hg in laboratory-scale coal combustors.38 The kinetics of the reactions are adequately rapid that the technique could be effectively used in full-scale coal combustors.39 To effectively design a low-cost TiO2 nanostructured sorbent agglomerate for Hg capture, it is important to develop a firmer understanding of the mechanistic steps of the process. In this paper, a mechanistic model for the 150 Journal of the Air & Waste Management Association

capture of Hg on TiO2 particles is developed. The role of water vapor on the capture rates of Hg also is established. MODEL DESCRIPTION Several authors1,40 – 42 have proposed mechanistic kinetic models for the photocatalytic oxidation of organic species over a range of conditions in aqueous media. A similar approach was used to derive a model to predict the photooxidation rates of Hg by UV-irradiated titania particles in the presence of water vapor (Figure 1). A complete sequence of elementary steps representing the photocatalytic oxidation reactions of Hg is summarized in Table 1. Upon irradiation with UV light, the electrons in the valence band are excited and jump to the conduction band, creating electronhole pairs within the catalyst particles. The pairs migrate to the surface of the particle, where they are involved in oxidation-reduction reactions. While electrons are trapped on the particle surface (by oxygen [O2], for example), holes react with adsorbed surface water molecules to create the reactive hydroxyl (OH⫺) radicals. The OH⫺ radicals are strong oxidants and oxidize the Hg on the surface. Electronhole recombination also can occur on the surface and in the volume, resulting in decreased photoactivity. Based on mechanistic steps in gas-phase photocatalysis of Hg on titania, the following assumptions were made in this model to develop a generalized kinetic equation: (1) The model does not consider the reaction: e⫺ ⫹ H⫹ 3 1/2 Hr2. This reaction would compete with the electron transfer reaction with adsorbed O2. In gas-phase systems in which O2 is at least 20% by volume of the total gas (as in air), the electrontransfer reaction with O2 would dominate over that with the hydrogen (H⫹) ion, which would be present in much lower concentrations than O2 in the reactor. In addition, the concentration of H⫹ in the system would be, in part, dependent upon the concentration of water vapor in the system. Therefore, the effects of H⫹ in the system on the

Figure 1. Mechanistics description of Hg capture by UV-irradiated titania and detail on a single TiO2 spherical particle.30 Volume 54 February 2004

Rodrı´guez et al. [h⫹], and OH⫺ radical [䡠OH] on the surface of the titania particles, the following expressions can be obtained:

Table 1. Photocatalytic reaction scheme for Hg. TiO2 ⫹ hv 3 e⫺ ⫹ h⫹ e⫺ ⫹ h⫹ 3 heat e⫺ ⫹ O2,ads 3 O2⫺

Excitation Recombination Trapping

h⫹ ⫹ H2O,ads 3 䡠OH ⫹ H⫹ h⫹ ⫹ OHads⫺ 3 䡠OH h⫹ ⫹ Hg,ads 3 䡠Hg ⫹ H⫹ 䡠OH ⫹ Hg,ads 3 HgOH HgOH 3 HgO ⫹ H⫹

Hydroxyl Attack

䡠OH ⫹ Yi,ads 3 Yj TiO2 ⫹ O2 7 O2,ads TiO2 ⫹ Hg 7 Hg,ads TiO2 ⫹ Yi 7 Yi,ads TiO2 ⫹ H2O 3 H2Oads

Adsorption

1 2

d关e ⫺ 兴 ⬇ 0 ⬇ k 1 fC TiO 2关I兴 ⫺ k2关e ⫺ 兴关h ⫹ 兴 ⫺ k3关e ⫺ 兴CO2,ads dt

3 4 5 6 7

k 1 fC TiO 2关I兴 ke⫺h⫺generation 关e 兴 ⫽ ⫽ k2关h ⫹ 兴 ⫹ k3KO2CO2 k2关h ⫹ 兴 ⫹ kO2

8 9 10 11 12 13

d关h ⫹ 兴 ⬇ 0 ⬇ k 1 fC TiO 2关I兴 ⫺ k2关e ⫺ 兴关h ⫹ 兴 ⫺ k4关h ⫹ 兴CH2O,ads dt k 1 fC TiO 2关I兴 ke⫺h_generation 关h ⫹ 兴 ⫽ ⫽ k2关e ⫺ 兴 ⫹ k4KH2OCH2O k2关e ⫺ 兴 ⫹ k4KH2OCH2O

observed rate of Hg capture by titania could not be differentiated from the effects of water vapor, and so neglecting this reaction should not pose a significant error in the mechanistic model, which highlights the effects of water vapor on the photocatalytic capture of Hg by titania. (2) OH⫺ radicals and water are the main oxidants of Hg, because they are the most abundant adsorbates and they generally act as reactive intermediates between photoexcited semiconductors and oxidizable organics in photocatalytic mechanisms. The oxidation potential for reactions on the surface of TiO2 becomes faster when H2O or OH⫺ are present because there is a necessity for hydroxylation of the photocatalyst surface for Hg degradation to occur. (3) The only source of OH⫺ radicals comes from the reaction of photo-generated holes with water vapor present in the system or OH⫺ groups on the titania particle surface. (4) The concentration of the OH⫺ radical is constant during the reaction, as the photo-generated holes and electrons also are constant throughout the experiments. Therefore, a photostationary state is reached on the surface of the titania particles. The reaction rate at which Hg is being photooxidized can be represented by dC Hg ⫽ ⫺k7关䡠OH兴KHgCHg dt

(1)

where CHg is the concentration of vapor-phase Hg in the system, [䡠OH] is the concentration of OH⫺ radicals on the surface of the titania particles, KHg is the adsorption coefficient of Hg on titania, and k7 is the intrinsic reaction rate constant for the OH⫺ radical and Hg. Taking into account the photostationary state assumption of the concentrations of the electron [e⫺], hole Volume 54 February 2004

(2)

⫺

d关䡠OH兴 ⬇ 0 ⬇ k4关h ⫹ 兴CH2O,ads ⫺ k7关䡠OH兴CHg,ads dt ⫺

冘

k⬘i 关䡠OH兴KYiCYi关䡠OH兴 ⫽

k4KH2OCH2O关h ⫹ 兴 k7KHgCHg ⫹

冘

(3)

(4)

k⬙i KYiCYi

where CHg is the concentration of Hg, CTiO2 is the titania concentration in the gas phase, [I] is the light intensity, and f the fraction of titania that is activated to form electronhole pairs on its surface. CH2O is the concentration of water vapor, CO2 is the concentration of O2, [e⫺] and [h⫹] are the electron and hole concentrations on the surface of the titania particles, and [OH] is the OH⫺ radical concentration on the surface of the titania particles. Equilibrium adsorption coefficients are referred to as KHg, KH2O, KO2, and KYi; likewise, ki’s are the reaction rate constants for reactions steps shown previously in the photocatalytic mechanism. ke-h_generation is the rate of electron-hole generation at the surface of the titania particles, which is assumed to be the same in the mechanism process, because the light source, the test system, the titania source, and the titania concentration are not changed during the experiments. In addition, the concentration of O2 adsorbed on the titania particles remains constant, leading to a rate of charge transfer from the titania to the adsorbed O2, which can be represented by kO2[e⫺], where kO2 ⫽ k3KO2CO2. By substituting the expression obtained for [e⫺] in eq 2 into the expression obtained for [h⫹] in eq 3 and simplifying, a quadratic expression is obtained for the concentration of photo-generated holes on the titania particles:

关h ⫹ 兴 2 ⫹

k O2 ⫹ ke⫺h_generationkO2 关h 兴 ⫺ ⬇0 k2 k2k4KH2OCH2O

(5)

which gives the following solution for [h⫹] Journal of the Air & Waste Management Association 151

Rodrı´guez et al.

再

关h ⫹ 兴 ⫽ ⫺1 ⫹

冑

冎

4k2ke⫺h⫺generation kO2 k4KH2OCH2OkO2 2k2

1⫹

(6)

A similar form of the equation was derived in a recent model1 for the photo-generated holes on the surface of the titania particles, including electron-hole recombination. The following equation is obtained when eq 6 is substituted into eq 4, and subsequently into eq 1:

dCHg ⫽ ⫺ dt

再 冑 冘

k7k4KH2OCH2OkO2KHgCHg ⫺1 ⫹ 2k2 k7KHgCHg ⫹

4k 2 k e⫺h⫺generation k4KH2OCH2OkO2

1⫹

k⬙jKYjCYj

冎

(7)

This equation can be reduced to the form

dCHg ⫺ ⫽ dt

再 冑

bCHgCH2O ⫺1 ⫹

1⫹

a C H 2O

1 ⫹ cCHg

冎

(8)

where a, b, and c are the model parameters for vapor-phase systems and remain constant in a given experiment:

a⫽

b⫽

4k 2 k e⫺h_generation k4KH2OkO2 k 7 k 4 K H 2O kO2KHg 2k2

c⫽

冘

冘

k⬙j KYjCYj

k 7 K Hg k⬙j KYjCYj

Parameter a depends on the characteristics of the reaction test system (geometry and volume), the source and concentration of titania, and light intensity (and therefore efficiencies of electron-hole formation and recombination). Parameters b and c are influenced by the photocatalytic reaction mechanisms of electron-hole excitation; generation and recombination; the equilibrium constants and adsorption coefficients for Hg, O2, and water vapor; and charge transfer rates. In this study, the main goal was to investigate the role of water vapor in the specific case of photooxidation of Hg with in situ– generated titania nanosize particles in aerosol systems. Initial photooxidation rates of Hg were experimentally measured at different initial water vapor concentrations, maintaining constant titania precursor and Hg feed rates. Parameters values for a, b, and c were therefore determined by fitting this data to the model, using a nonlinear regression analysis. Data from the literature on oxidation of organic substrates also was used to check whether the model accurately predicted the trends with varying water vapor concentrations. EXPERIMENTAL METHODS Apparatus and Materials The capture of Hg by in situ– generated titania nanosized agglomerates in the presence of water vapor was studied in a flow reactor. Figure 2 shows the experimental system for the present study. The setup consists of TiO2 particle generators, Hg and water vapor feeders, a tubular furnace for precursor oxidation, a test column of borosilicate glass tube, a filter holder for particle collection, a UV source, a

Figure 2. Hg capture system with in situ– generated titania. 152 Journal of the Air & Waste Management Association

Volume 54 February 2004

Rodrı´guez et al. thermohygrometer, and an online Hg detector. The alumina reactor tube was 91.44 cm long with an inner diameter of 2.54 cm. Compressed air was used as the carrier gas and was passed through a HEPA filter [75– 62 Fourier transform-infrared (FT-IR) purge gas generator, Balston Filter Products] to assure it was particle-free. Hg vapor was introduced into the system by passing particle-free air at a precisely controlled flow rate above liquid Hg contained in a gas-washing bottle. The sorbent precursor, titanium (IV) isopropoxide (TTIP), 97% (Ti[OCH(CH3)2]4), was injected into the system at a constant flow rate, by bubbling argon (Ar) gas (prepurified, 99.99%, Wright Brothers) through the precursor solution contained in a bubbler (Midget, 30 mL, Ace Glass). The bubbler was placed in a water bath (N. 5160 wide-neck flask, 500 mL, Pyrex; TM 106 heating mantle, 500 mL, Glas-Col), so the precursor was vaporized at a constant temperature of 75 °C, controlled by a power controller. The tubing before and after the TTIP bubbler was wrapped by a heating tape to prevent any condensation inside the tube. An additional inlet was connected to the reactor to introduce water vapor before the photochemical reactor so as to vary its concentration. This was done by bubbling particle-free clean air through water contained in a bubbler (Midget, 30 mL, Ace Glass). The humidity and the temperature of the system were measured between the photochemical cell and the filter by a thermohygrometer (Cole Palmer, Cat. No. 37950 –10). The furnace was kept at a temperature of ⬃1000 °C with a residence time of ⬃3 sec. A photochemical reaction cell, a borosilicate glass tube, was placed under the UV lamp at the exit of the furnace reaction tube. The UV lamp (type XX-40, Spectronics, 80 W) was 120 cm long and the light intensity was 1850 ␮W/cm2 at 354 nm, at a distance of 25 cm. A glass fiber filter (N. 61663, Gelman Science) was used downstream to collect particles. The stream was then sent to the online Hg analyzer for real-time monitoring of elemental Hg concentration in the gas phase. The online analyzer (Shimadzu UV1201S Spectrometer) was calibrated using a series of impingers that captured the gas-phase Hg that was subsequently measured by cold vapor atomic absorption (CVAA).21,30 The calibration curve was used to obtain the Hg concentrations and to ensure that there was no interference in the measurements over the range of water vapor concentrations used in the experiment.43 Procedures and Measurement The experimental conditions are summarized in Table 2. The total flow rate in the reactor was maintained at 1 L/m in all experiments. Hg vapor was entrained in a 120 mL/ min airstream at room temperature. Sufficient time was allowed for both the system and the online analyzer Volume 54 February 2004

Table 2. List of experimental conditions. Flow Rate Total TiO2 Hg

1 Lpm 200 cc/min (125 ␮g/min) 120 cc/min (⬃2 ␮g/min) Temperature ⬃1000 °C ⬃75–80 °C ⬃20–25 °C

Furnace TiO2 Hg Residence Time Furnace reactor Photochemical cell

3 sec 70 sec

readings to stabilize. Inlet mercury (Hg0) concentrations in the system were measured to be in the range of 1.5–2.5 ␮g/m3, typical of exhaust gas concentrations in some coal combustors. To understand the effect of water vapor concentration on capture efficiency, different initial humidities were used by varying the initial water vapor concentration (by varying the flow from 0 to 30 mL/min in the water impinger), while the flow rate through the TiO2 precursor bubbler was kept constant at 200 mL/min. The precursor feed rate was determined by weighing the particles collected on the filter. The Hg feed rate was determined by the online Hg analyzer before feeding titania and water vapor into the system. Because of the nature of Hg and the various reports in the literature on the difficulty in performing reliable experiments, extra precautions were exercised.30 Before every experiment, the reactor was scrubbed clean at a very high temperature (1200 °C) and blank measurements were carried out to ensure that no residual Hg remained in the system. Each measurement for an experiment was performed in triplicate, and the data were averaged. The standard deviations of the three measurements were typically less than 1%. RESULTS AND DISCUSSION At a fixed inlet concentration of elemental Hg, the reactivity of the titania was measured as a function of water vapor concentration in the reactor. The reactivity of the titania was assessed by comparing the inlet elemental Hg concentration to the outlet elemental Hg concentration after the system stabilized (⬃ 3 min after a change in experimental conditions was made). The experiments were performed under varying water vapor concentrations, and the results are plotted in Figure 3. The concentration of Hg decreased over the entire range of water vapor levels, indicating that competitive adsorption had not been initiated as yet. The increasing water vapor Journal of the Air & Waste Management Association 153

Rodrı´guez et al. Table 3. Comparison of model parameters for the photooxidation of mercury and other organics. Model Parameters

Study This work (Hg) Toluene42 Formaldehyde42

Figure 3. Ratio of outlet to inlet concentration of Hg for different humidity levels in the photochemical reactor.

a [ppmv]

b [minⴚ1 ppmvⴚ1]

c [ppmvⴚ1]

Correlation Coefficient

2.74 ⫻ 106

5.65 ⫻ 10⫺5 (⫾ 4.5%) 2.14 ⫻ 10⫺6 (⫾ 9.8%) 3.59 ⫻ 10⫺6 (⫾ 6.9%)

3.23 (⫾ 6.5%) 6.86 ⫻ 10⫺1 (⫾ 4.3%) 5.97 ⫻ 10⫺1 (⫾ 3.3%)

0.9

(⫾ 6.9%) 9.63 ⫻ 106 (⫾ 6.3%) 3.06 ⫻ 106 (⫾ 14.7%)

0.94 0.93

concentration resulted in enhanced production of OH radicals, thus resulting in increased Hg oxidation rates. The photooxidation rate for Hg was approximated by taking the difference in the inlet and outlet concentrations and dividing by the residence time (volume of the photocatalytic reaction cell/volumetric flow rate of reaction gases), and the results are plotted in Figure 4 for different water vapor concentrations. In the range of water vapor concentrations, the oxidation rate is increasing, indicating that the additional OH⫺ radicals produced result in an increased rate. At low water vapor concentrations, competitive adsorption does not inhibit the Hg from being adsorbed and limit the rate. On examining eq 8, at very low water vapor concentrations, one would expect a square root dependence of the oxidation rate on the water vapor concentration. Reasonable agreement is obtained between measured rates and predicted values (shown by the solid line in Figure 4). The model parameters are listed in Table 3. At the higher water vapor concentrations, the agreement is not as good, probably

because of the limited range of the data (one obtains rather high values of the parameter a. Experiments were not conducted at very high water vapor concentrations in this study. Data from other studies44,45 support these general trends. The model also was used to compare with the photocatalytic oxidation of other organic substrates (toluene and formaldehyde) on irradiated titania surfaces measured by other researchers,44,45 for humidity levels up to 12,000 ppmv (Figure 5). The model parameters obtained are listed in Table 3, and the percentage variation in the constants is less than 15%, with correlation coefficients on the order of 0.9. For the selected range of values, the constant a is rather large. Other researchers44 have proposed a bimolecular type Langmuir-Hinshelwood (L-H) expression for the substrate and water vapor and have fit this form of the equation to obtain the parameter values. The expression described in this paper was derived from a mechanistic description (eq 8), and the expression reduces to an L-H type expression for the substrate but retains a complicated form for the water vapor. It should be noted that in the limit of low water vapor concentrations,

Figure 4. Photooxidation rate of Hg in the photochemical reactor as a function of water vapor concentration.

Figure 5. Photooxidation rate of toluene and formaldehyde42,43 compared with model predictions at different water vapor concentrations.

154 Journal of the Air & Waste Management Association

Volume 54 February 2004

Rodrı´guez et al. the L-H type expression tends to a first-order dependence on the water vapor concentration, whereas the equation derived in this paper indicates a square-root dependence. At high water vapor concentrations, the L-H predicts a constant value for the rate of pollutant oxidation that is independent of the water vapor concentration. Using our model (eq 8), a constant rate of pollutant oxidation that is independent of water vapor concentration also is obtained. Starting with the model in eq 8, it is noted that at high water vapor concentrations, or more specifically, when CH2O ⬎⬎a, the term [⫺1 ⫹ (1 ⫹ a/CH2O)0.5] becomes approximately equal to 1/2 (a/CH2O). Substituting this approximation into eq 8 yields the expression in eq 9 for CH2O ⬎⬎a. Thus, under the limiting conditions of water vapor, the following limiting rates are obtained: dCHg bCHg 冑aCH2O ⫽ dt 1 ⫹ cCHg

for CH2O Ⰶ a

bCHga dCHg ⫽ ⫺ dt 共1 ⫹ cCHg兲2

for CH2O Ⰷ a

⫺

(9)

Thus, the value of a is an important parameter value that establishes the region where the rate would be increasing with increasing water vapor concentration or would reach its maximum value and be independent of the water vapor concentration. Figure 6 is a comparison of the predicted photooxidation rates using the model equations and the experimentally determined photooxidation rates for the entire data set for the three different substrates. There is

reasonable agreement over the entire range of values for the three different substrates, indicating that the model qualitatively explains the trends in the measured data. Clearly, the range of water vapor concentrations selected for the experimental study has been limited to what is encountered in typical coal combustion exhausts, and the data indicate that the rate increases with water vapor concentration, thus indicating a paucity of OH⫺ radicals, or that it is a rate-controlling factor. In this range, water vapor is not hindering the adsorption of Hg; however, at higher water vapor concentrations, competitive adsorption potentially would result in a reduction in the rate. CONCLUSIONS A mechanistic model was developed for predicting the oxidation rates of Hg on titania particle surfaces, and the expressions were used to evaluate the role of water vapor. The solution to the model equations agreed well with experimental data for Hg capture rates and other organic substrates reported in the literature. Under typical combustion conditions, the rate of Hg capture by the in situ– generated TiO2 sorbent particles increases with increasing water vapor concentration. At low water vapor concentrations, the rate was proportional to the square root of the water vapor concentration, and at higher values, it is expected that the rate would reach a constant value independent of the water vapor concentration. At very high water vapor concentrations, competitive adsorption also may play a role and inhibit the capture rates of Hg, though this is not a concern in practical combustion systems.

Figure 6. Comparison of model predictions to experimental data of the photooxidation rates of Hg with UV-irradiated titania (this work), and toluene and formaldehyde.42,43 Volume 54 February 2004

Journal of the Air & Waste Management Association 155

Rodrı´guez et al. REFERENCES 1. Almquist, C.; Biswas, P. A Mechanistic Approach to Modeling the Effect of Dissolved Oxygen in Photooxidation Reactions on Titanium Dioxide in Aqueous Systems; Chem. Eng. Sci. 2001, 56, 3421-3430. 2. Hall, B.; Schager, P.; Lindqvist, O. Chemical Reactions of Mercury in Combustion Flue Gases; Water, Air, Soil Pollut. 1991, 56, 3. 3. Aizpu´ n, B.; Fernandez, M.L.; Blanco, E.; SanzMedel, A. Speciation of Inorganic Mercury (II) and Methylmercury by Vesicle-Mediated HighPerformance Liquid Chromatography Coupled to Cold Vapour Atomic Absorption Spectrometry; J. Anal. Atom. Spectrom. 1994, 9, 1279-1284. 4. Seigneur, C.; Wrobel, J.; Constantinou, E. A Chemical Kinetic Mechanism for Atmospheric Inorganic Mercury; Environ. Sci. Technol., 1994, 28 (9), 1589-1597. 5. Grandjean, P.; Weihe, P.; White, R.F.; Debes, F. Cognitive Performance of Children Prenatally Exposed to “Safe” Levels of Methylmercury; Env. Res. Sec. A 1998, 77, 165-172. 6. Mercury Study Report to Congress; EPA-452/R-97– 008; U.S. Environmental Protection Agency, U.S. Government Printing Office: Washington, DC, 1997; Vol 6. 7. Persistent Bioaccumulative Toxic Chemicals. Proposed Rule. CFR 40, Part 372, January 1999. 8. Lindqvist, O.; Rodhe, H. Atmospheric Mercury: A Review; Tellus 1985, 37B, 136-159. 9. Lindqvist, O. Fluxes of Mercury in the Swedish Environment: Contributions from Waste Incineration; Waste Manage. Res. 1986, 4, 35-44. 10. Slemr, F.; Seiler, W.J.; Schuster, G. Distribution, Speciation and Budget of Atmospheric Mercury; J. Atmos. Chem. 1985, 3, 407-434. 11. Hall, B. The Gas Phase Oxidation of Elemental Mercury by Ozone; Water, Air, Soil Pollut. 1995, 80, 301-315. 12. Wu, C.Y.; Biswas, P. An Equilibrium Analysis to Determine the Speciation of Metals in an Incinerator; Combust. Flame, 1993, 93, 31. 13. Laudal, D.L.; Galbreath, K.C.; Heidt, M.K. A State-of-the-Art Review of the Flue Gas Speciation Methods; EPRI Report N. TR-107080; EPRI: Palo Alto, CA, 1996. 14. Carpi, A. Mercury from Combustion Sources: A Review of the Chemical Species Emitted and Their Transport in the Atmosphere; Water, Air, Soil Pollut. 1997, 98, 241-254. 15. Galbreath, K.C.; Zygarlicke, C.J. Mercury Transformations in Coal Combustion Flue Gas; Fuel Process. Technol. 2000, 65– 66, 289-310. 16. Gibbs, W.H.; Clarke, F.; Mehta, A.K. The Fate of Coal Mercury during Combustion; Fuel Process. Technol. 2000, 65– 66, 365-377. 17. Keeler, G.; Glinsorn, G.; Pirrone, N. Particulate Mercury in the Atmosphere: Its Significance, Transport, Transformation and Sources; Water, Air, Soil Pollut. 1995, 80, 159. 18. Fitzgerald, W.F.; Engstrom, D.R.; Mason, R.P.; Nater, E.A. The Case for Atmospheric Mecury Contamination in Remote Areas; Environ. Sci Technol. 1998, 32, 1. 19. Hanisch, C. Where Is Mercury Deposition Coming From?; Environ. Sci. Technol. 1998; 32, 176A. 20. Lindberg, S.E.; Stratton, W.J. Atmospheric Mercury Speciation: Concentrations and Behavior of Reactive Gaseous Mercury in Ambient Air; Environ. Sci. Technol. 1998; 32, 49. 21. Hedrick, E.; Lee, T.G.; Biswas, P. The Development of Iodine Based Impinger Solutions for the Efficient Capture of Hg-0 Using Direct Injection Nebulization—Inductively Coupled Plasma Mass Spectrometry Analysis; Environ. Sci. Technol. 2001, 35 (18), 3764-3773. 22. Sinha, R.K.; Walker, P.L. Removal of Mercury by Sulfurized Carbons; Carbon 1972, 10, 754-756. 23. Otani, Y.; Kanaoka, C.; Usul, C.; Matsui, S.; Emi, H. Adsorption of Mercury Vapor on Particles; Environ. Sci. Technol. 1986, 20, 735. 24. Krishnan, S.V.; Gullett, B.K.; Jozewicz, W. Sorption of Elemental Mercury by Activated Carbons; Environ. Sci. Technol. 1994, 28, 1506. 25. Livengood, C.D.; Huang, H.S.; Mendelsohn, M.H.; Wu, J.M. Mercury Capture in Bench-Scale Absorbers. In Proceedings of PETC’s 10th Annual Coal Precapture, Utilization and Environmental Control Contractors Conference, Pittsburgh, PA, 1994. 26. Morency, J.R. Control of Mercury in Fossil Fuel-Fired Power Generation. In Proceedings of PETC’s 10th Annual Coal Precapture, Utilization and Environmental Control Contractors Conference, Pittsburgh, PA, 1994. 27. Karatza, D.; Lancia, A.; Musmarra, D.; Pepe, F.; Volpicelli, G. Removal of Mercury Chloride from Flue Gas by Sulfur Impregnated Activated Carbon; Hazard. Waste Hazard. Mat. 1996, 13, 95. 28. Korpiel, J.A.; Vidic, R.D. Effect of Sulfur Impregnation Method on Activated Carbon Uptake of Gas-Phase Mercury; Environ. Sci. Technol. 1997, 31, 2319. 29. Vidic, R.D.; Chang, M.T. Kinetics of Vapor Phase Mercury Uptake by Virgin and Sulfur-Impregnated Activated Carbons; J. Air & Waste Manage. Assoc. 1998, 48, 247.

156 Journal of the Air & Waste Management Association

30. Wu, C.Y.; Lee, T.G.; Tyree, G.; Arar, E.; Biswas, P. Capture of Mercury in Combustion Systems by In Situ Generated Titania Particles with UV Irradiation; Environ. Eng. Sci. 1998, 15, 137. 31. Uberoi, M.; Shadman, F. Simultaneous Condensation and Reaction of Metal Compound Vapors in Porous Solids; Ind. Eng. Chem. Res. 1991, 30, 624. 32. Ho, T.C.; Chen, C.; Hopper, J.R.; Oberacker, D.A. Metal Capture during Fluidized Bed Incineration of Wastes Contaminated with Lead Chloride; Combust. Sci. Tech. 1992, 85, 101-116. 33. Gullett, B.K.; Ragnunathan, K. Reduction of Coal-Based Metal Emissions by Furnace Sorbent Injection; Energy Fuels 1994, 8, 1068. 34. Owens, T.M.; Biswas, P. Vapor Phase Sorbent Precursors for Toxic Metal Emissions Control from Combustors; Ind. Eng. Chem. Res. 1996, 35, 792. 35. Biswas, P.; Zachariah, M.R. In Situ Immobilization of Lead Species in Combustion Environments by Injection of Gas Phase Silica Sorbent Precursors; Environ. Eng. Sci. 1997, 31 (9), 2455-2463. 36. Biswas, P.; Wu, C.Y. Control of Toxic Metal Emission from Combustors Using Sorbents: A Review; J. Air & Waste Manage. Assoc. 1998, 48, 113-127. 37. Lee, T.G.; Hedrick, E.; Biswas, P. Comparison of Hg-0 Capture Efficiencies of Three In Situ Generated Sorbents; AIChE J. 2001, 47 (4), 954-961. 38. Zhuang, Y.; Biswas, P. Submicrometer Particle Formation and Control in a Bench Scale Pulverized Coal Combustor; Energy Fuels 2001, 15 (3), 510-516. 39. Lee, T.G.; Hedrick, E.; Biswas, P. Kinetics of TiO2-Hg Reactions; Ind. Eng. Chem. Res., in press. 40. Turchi, C.S.; Ollis, D.F. Photocatalytic Degradation of Organic Water Contaminants: Mechanisms Involving Hydroxyl Radical Attack; J. Catal. 1990, 122, 178-192. 41. Gerischer, H. Conditions for an Efficient Photocatalytic Activity of TiO2 Particles; Photocatalytic Purification and Treatment of Water and Air; Elsevier: Amsterdam, 1993, p 1. 42. Halmann, M.M. Photodegradation of Water Pollutants; CRC Press: Boca Raton, FL, 1996. 43. Rodriguez-Lattuada, S.J. In Situ Generated Sorbents for Mercury Capture in Combustor Exhausts: Role of Other Particles and Water Vapor. M.S. Thesis, Aerosol and Air Quality Research Laboratory, University of Cincinnati, Ohio, 2001. 44. Obee, T.N. Photooxidation of Sub-Parts-Per-Million Toluene and Formaldehyde Levels on Titania Using a Glass-Plate Reactor; Environ. Sci. Technol. 1996, 30, 3578-3584. 45. Obee, T.N.; Brown, R.T. TiO2 Photocatalysis for Indoor Air Applications: Effects of Humidity and Trace Contaminant Levels on the Oxidation Rates of Formaldehyde, Toluene and 1,3-Butadiene; Environ. Sci. Technol. 1995, 29, 1223-1231.

About the Authors Sylian Rodrı´guez is a graduate student at the University of Cincinnati. She obtained her M.S. from the Aerosol and Air Quality Research Laboratory. Catherine Almquist is an assistant professor at Miami University, Oxford, OH. Tai Gyu Lee is an assistant professor of chemical engineering at Yonsei University in Seoul, South Korea. Masami Furuuchi is at Kanazawa University in Kanazawa, Japan. Elizabeth Hedrick is at the National Exposure Research Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH. Pratim Biswas is the Stifel and Quinette Jens Professor and director of the Environmental Engineering Science Program at Washington University, St. Louis, MO. Address correspondence to: Pratim Biswas, Washington University, 1 Brookings Drive, Campus Box 1180, St. Louis, MO 63130: fax: (314) 935-5464; e-mail: [email protected].

Volume 54 February 2004