<%BANNER%>

Silica-Titania Composites for Water Treatment


PAGE 1

SILICA-TITANIA COMPOSITES FOR WATER TREATMENT By DANIELLE JULIA LONDEREE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2002

PAGE 2

Copyright 2002 by Danielle Julia Londeree

PAGE 3

ACKNOWLEDGMENTS First of all, there is no way I could have accomplished this research without the help of my Lord and Savior. He has not only given me this opportunity, but also gave me peace and patience during the times of frustration. Second, without the help of my husband who kept me calm when my computer crashed several times, this work would not be done. He also stayed up with me many nights while I was writing and constantly gave me love and support. I would like to thank Dr. David Mazyck for giving me the opportunity to work for him and learn from him. He has challenged me in many ways and I feel that I have learned more in the past two years of graduate school than the four years of undergrad. I thank him for patience and for purchasing the spectrophotometer. I thank Dr. Powers for allowing me to do an independent study with him. His excitement for research is contagious and motivational. I would also like to thank Dr. Paul Chadik and Dr. Chang-Yu Wu for giving advice during the research meetings and helping to provide answers for my many questions. I thank the students in my research group and for those who have helped me get good data, specifically Matt Tennant (my mentor), Ameena Khan, Jennifer Hobbs, Christina Termaath, Jon Powell, Noworat Coowanitwong, and Julee MacKenzie. iii

PAGE 4

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES..........................................................................................................vii ABSTRACT.......................................................................................................................ix CHAPTER 1 INTRODUCTION...........................................................................................................1 2 LITERATURE REVIEW................................................................................................4 2.1 Heterogeneous Photocatalysis..................................................................................4 2.2 Titanium Dioxide......................................................................................................8 2.3 Reactor Design........................................................................................................11 2.4 Catalyst Supports....................................................................................................13 2.5 Silica-Titania Composites.......................................................................................16 2.5.1 Sol-Gel Chemistry.........................................................................................18 2.5.2 Surface Chemistry and Adsorption Characteristics......................................22 3 MATERIALS AND METHODS..................................................................................25 3.1 Silica-Titania Composites.......................................................................................25 3.2 Batch Performance Studies.....................................................................................26 3.3 Column Performance Studies.................................................................................28 4 RESULTS AND DISCUSSION...................................................................................30 4.1 Dye Comparison.....................................................................................................30 4.2 Optimization of Gels...............................................................................................34 4.2.1 Titania Loading.............................................................................................34 4.2.2 Curing Temperature......................................................................................40 4.2.3 Pore Size.......................................................................................................43 4.3 Column Studies.......................................................................................................44 iv

PAGE 5

5 SUMMARY AND CONCLUSIONS............................................................................48 APPENDIX A COMMON PREPARATION METHODS OF MIXED AND SUPPORTED OXIDES AS DISCUSSED BY GAO AND WACHS (1999)......................................50 A.1 Mixed Oxides.........................................................................................................50 A.1.1 Sol-Gel Hydrolysis.......................................................................................50 A.1.2 Coprecipitation.............................................................................................50 A.2 Supported Oxides...................................................................................................51 A.2.1 Impregnation................................................................................................51 A.2.2 Chemical Vapor Deposition ........................................................................51 B DYE STRUCTURES....................................................................................................52 B.1 Methylene Blue......................................................................................................52 B.2 Malachite Green.....................................................................................................52 B.3 Crystal Violet.........................................................................................................53 B.4 Reactive Red..........................................................................................................53 C TEM PICTURES...........................................................................................................54 LIST OF REFERENCES...................................................................................................55 BIOGRAPHICAL SKETCH.............................................................................................62 v

PAGE 6

LIST OF TABLES Table page 2-1 Oxidation power for various species relative to chlorine........................................6 2-2 Characteristics of Degussa P25 Titanium Dioxide..................................................9 4-1 Peak absorbance wavelengths................................................................................30 4-2 Exhaustion (C/Co = 1) times (minutes) for column studies..................................46 vi

PAGE 7

LIST OF FIGURES Figure page 2-1 Schematic of oxidation and reduction occurring on the semiconductor surface.....5 2-2 Example of hydroxyl radical attack a) on tetrachloroethylene by self-addition b) on 1,1,1-trichloroethane by hydrogen abstraction (notice chlorine shift in the radical).....................................................................................................................8 2-3 Diagram of anatase and rutile crystal structures....................................................10 2-4 Simplified diagram of a catalyst support...............................................................13 2-5 Linkages of SiO2 tetrahedras.................................................................................17 2-6 Structural changes during the sol-gel process........................................................20 2-7 Silanol groups on the silica surface.......................................................................23 2-8 Representation of dehydroxylation........................................................................23 3-1 Column setup.........................................................................................................29 4-1 Photolysis of dyes as a function of UV flux for 2 hours of exposure....................31 4-2 Photolysis of MB and MG (2 hours UV exposure at 0.45 mW/cm2) as a function of initial dye concentration....................................................................................32 4-3 Photolysis of RR (2 hours UV exposure) as a function of initial dye concentration and flux..................................................................................................................32 4-4 Change in absorbance of RR at 538 nm as a function of pH.................................33 4-5 Spectrophotometer scan of 10 mg/L RR as a function of pH, before and after UV exposure.................................................................................................................34 4-6 Destruction of RR (pH 7.6) after 2 hours UV exposure versus TiO2 loading as a function of flux......................................................................................................35 4-7 Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm2) versus TiO2 loading as a function of pH..........................................................................36 vii

PAGE 8

4-8 Destruction of RR (10 mg/L) after 1 hour UV with TiO2 slurry at various pHs..37 4-9 Effects of bicarbonate at pH 7.6 versus TiO2 loading...........................................38 4-10 BET surface area and pore size versus titania loading..........................................39 4-11 Adsorption of CV (10 mg/L) versus TiO2 loading after 24 hours of mixing........39 4-12 Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm2) versus curing temperature as a function of pH..................................................................40 4-13 Effect of curing temperature on surface area.........................................................41 4-14 XRD analysis of various temperature-cured silica-titania composites..................42 4-15 Effect of curing temperature on adsorption of CV on gels with and without TiO2.......................................................................................................................43 4-16 Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm2) versus pore size.................................................................................................................44 4-17 Adsorption of CV on gels (6% TiO2) versus pore size.........................................44 4-18 Column exhaustion curve for 140 pellet (12% TiO2)........................................45 4-19 Column exhaustion curve for 30 pellets (12% TiO2)........................................46 4-20 Effect of regeneration time on 140 column runs................................................47 viii

PAGE 9

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering SILICA-TITANIA COMPOSITES FOR WATER TREATMENT By Danielle Julia Londeree December 2002 Chair: David Mazyck Major Department: Environmental Engineering Sciences Heterogeneous photocatalysis can be used for mineralizing organic dyes found in the effluent of textile dyeing operations. Incorporating the catalyst TiO 2 into an adsorbent material, such as silica, has many advantages over using a TiO 2 slurry for water purification. The goal of this research was twofold: 1) to produce a silica-titania composite using a sol-gel hydrolysis method that dopes the catalyst into the silica matrix during gelation and 2) to optimize the titania loading, curing temperature, and pore size of the material based upon maximizing its destruction and adsorption ability for textile dyes. The optimal titania loading found for reactive red dye was 30 wt% TiO 2 The optimal pore size and curing temperature found were 140 and 180 o C, respectively. These composites were also made into small cylindrical pellets and tested in a flow-through column to be used in a regenerative system. It was found that diffusivity was very important to efficiently regenerate the column using photocatalysis. ix

PAGE 10

CHAPTER 1 INTRODUCTION The textile industrys dyeing operations are of primary environmental concern, due to the variety of toxic chemicals (mostly solvents, surfactants, and acids) used in the dyeing of garments. According to the Environmental Protection Agency [EPA] (1997), chemicals such as methanol, methyl ethyl ketone, trichloroethylene, trimethylbenzene, and dyes such as basic green 4 and dispersive yellow 3 are released into the environment, either in the air phase or water phase. These contaminants can create large amounts of BOD (biochemical oxygen demand) and COD (chemical oxygen demand), as well as cause eutrophication and aquatic toxicity in effluent streams and natural water bodies. Color itself is increasingly being regulated, as there have been links to aquatic toxicity and lower DO (dissolved oxygen) values in receiving streams (EPA, 1996). There are several chemical and physical methods of removal available for pre-release treatment, such as adsorption by activated carbon, biodegradation, and ozonation. Yet, dyes can be resistant to biodegradation and ozonation due to the large number of aromatic rings found in modern dyes (Sheng and Chi, 1993). Therefore, advanced oxidation processes are being considered as an emerging technology to handle large volumes of textile wastewater (Sauer et al., 2002; Wu et al., 1999). Of these processes, heterogeneous photocatalysis can be used for mineralization (complete conversion to CO 2 and H 2 O) of several classes of organic compounds (including dyes), instead of transferring the pollutants from one phase to another (associated with carbon use). 1

PAGE 11

2 Dyes can also be used as a surrogate for more toxic organic compounds in order to measure the photocatalytic ability of a system. For instance, the structure of methylene blue is similar to chloropromazine, an antipsychotic drug, and methylene blue destruction could be used to estimate the destruction of this drug. In this research, dyes were used to represent hard to degrade aromatic organic compounds in order to measure the performance of the silica-titania composites for water treatment. In addition, NASA, with its construction of the International Space Station and plans for a Mars mission, requires a water recovery system that will minimize chemical additions and energy and provide water at a potable level. The water source to be treated will come mostly from grey water, including urine, humidity condensate, and wash water. NASA Document CTSD-ADV-245 also lists numerous organic compounds (many aromatic) that may be present in water in space flight (1998). The dyes used in this study are surrogates for those compounds that may be present in space flight for NASA. There were two main objectives of this research. First, to develop a silica-titania composite by utilizing a sol-gel hydrolysis method and doping a highly efficient TiO 2 catalyst into the silica matrix during gelation. These composites could also be created into any shape and size, such as small cylindrical pellets to be used for a packed column. Second, the titania loading, curing temperature, and pore size of the composites were to be optimized based upon maximizing its destruction and adsorption ability for textile dyes. The result would be a combined adsorbent/photocatalyst that efficiently adsorbs contaminants into its pores for oxidation where the hydroxyl radicals would predominantly exist. It was hypothesized that a packed column of these composites made

PAGE 12

3 into small pellets could create a regenerable system that could be used over and over again.

PAGE 13

CHAPTER 2 LITERATURE REVIEW 2.1 Heterogeneous Photocatalysis Advanced oxidation processes (AOPs) are considered to be an alternative water treatment technology for removing harmful compounds and microbes from public water supply, comparable with traditional practices such as activated carbon adsorption, air stripping, and chlorine disinfection. AOPs can not only disinfect the water from virulent microbes, but can also oxidize toxic heavy metals, organic pesticides and solvents, chlorinated compounds, and inorganic chemicals from the water without producing a waste stream unsuitable for disposal or consumption. For example, these processes can be used for chemical spill cleanup, treatment of industrial effluents, and wastewater treatment (Turchi and Ollis, 1990). AOPs involve the generation of highly reactive oxidative species produced by several pathways, including reactions with H 2 O 2 and O 3 with or without ultraviolet (UV) irradiation and heterogeneous photocatalysis. Heterogeneous photocatalysis entails the use of a solid photocatalyst (usually a semiconductor) in contact with either a liquid or gas, while homogeneous photocatalysis uses a catalyst of the same phase as the contaminated media. While UV use or an oxidant alone may yield partial degradation of a compound, the combined use of UV with an oxidant (with or without a photocatalyst) has been shown to yield complete mineralization of organic contaminants to carbon dioxide (Ollis et al., 1991). 4

PAGE 14

5 The reactants in heterogeneous photocatalysis in water are naturally occurring species (such as oxygen, water molecules, and hydroxide ions (OH )) as compared to chemically supplied reactants (such as hydrogen peroxide and ozone) required for other AOPs. When the surface of the photocatalyst absorbs a specific amount of energy, an electron from the valence band jumps to the conduction band, thereby leaving a positive hole in the valence band. These electrons (e ) and holes (h + ) may either recombine, releasing heat, or migrate to the surface of the catalyst. The reactive oxidative species are then generated from reactions with the electrons or holes on the surface (see Figure 2-1). Figure 2-1. Schematic of oxidation and reduction occurring on the semiconductor surface. The primary oxidizing species in water is the hydroxyl radical (*OH) (Turchi and Ollis, 1990). It is the most reactive of the oxidizing agents (see Table 2-1) and can be generated via several pathways: H 2 O + h + *OH + H + (1) OH + h + *OH (2)

PAGE 15

6 O 2 + 2H + + 2e H 2 O 2 (3a) H 2 O 2 + e *OH + OH (3b) H 2 O 2 hv 2*OH (3c) O 2 + e O 2 (4a) O 2 + H 2 O 2 *OH + OH + O 2 (4b) As can be seen from equations 1 4b, molecular oxygen plays a dual role in heterogeneous photocatalysis by 1) creating an oxidative species, the superoxide radical (O 2 ), and 2) acting as an electron acceptor to prevent the electrons and holes from recombining. These functions are very important in establishing an efficient photocatalytic system, for it has been shown that with an increase in oxygen concentration, there is a subsequent increase in the reaction rate of contaminant degradation (Gerischer and Heller, 1991). Similarly, hydrogen peroxide plays a parallel role. Table 2-1. Oxidation power for various species relative to chlorine. Species Relative Oxidation Power Hydroxyl radical 2.06 Singlet oxygen radical 1.78 Hydrogen peroxide 1.31 Perhydroxyl radical 1.25 Chlorine dioxide 1.15 Chlorine 1.00 Source: Elizardo, 1991. Adsorption of compounds to the surface of the catalyst is also important in an efficient photocatalytic system. For instance, competition with other compounds can be

PAGE 16

7 deleterious to the adsorption of specific contaminants on the catalyst surface. Chen et al. (1997) showed that the effect of inhibition due to competitive adsorption will hinder the degradation rate, especially in the presence of bicarbonate and phosphate. In addition, carbonate ions will also scavenge the hydroxyl radicals, thereby decreasing the destruction rate (Chen et al., 1997). Another factor affecting adsorption is pH and the effect of pH is highly compound specific. For example, in the photocatalysis of textile dyes, a lower pH enhanced adsorption of Orange II (containing an anionic sulfonate group) to the catalyst surface, thereby increasing the rate of mineralization (Wu et al., 1999), while an alkaline pH increased adsorption and subsequent degradation of methylene blue (a cationic dye) (Houas et al., 2001). Sauer et al. (2000) also found similar relationships between adsorption and destruction of reactive dyes. The photocatalytic mineralization of an organic compound begins with either reactions with hydroxyl radicals (or other less significant oxidizing species) or by direct oxidation with the holes on the surface of the catalyst. A hydroxyl radical attacks an organic compound by either removing an available hydrogen atom to form water or adding itself to any unsaturated carbon bonds (Grabner et al., 1991; Halmann, 1996; Hoffman et al., 1995; Hoigne, 1990; Mao et al., 1991, 1992, 1993). Mao et al. (1991, 1992, 1993) found that the rate of oxidation of organics correlated with the C-H bond strengths, proving that H atom abstraction by *OH may be a rate-limiting step. Consequently, from these primary reactions, an organic radical is formed (see Figure 2-2). The subsequent radical transformations and radical to radical interactions lead to an

PAGE 17

8 array of intermediate products, eventually resulting in CO 2 and H 2 O. For chlorinated compounds, HCl is also produced. Figure 2-2. Example of hydroxyl radical attack a) on tetrachloroethylene by self-addition b) on 1,1,1-trichloroethane by hydrogen abstraction (notice chlorine shift in the radical) (adapted from Mao et al., 1991 and 1992). There are several literature reviews available on the mineralization of particular classes of organic chemicals (Hoffman et al., 1995; Matthews, 1988; Ollis et al., 1991; Serpone, 1995). Blake (1995) and Halmann (1996) have compiled extensive surveys of this research and have also included sections on typical reactor designs and methods for catalyst improvement. Research has also been focused on the determination of the intermediate degradation pathways, for Mao et al. (1991 and 1992) has shown that the conditions of a system (pH, oxygen concentration, competition with other compounds) can affect the pathway a single compound can take towards complete mineralization. 2.2 Titanium Dioxide Metal oxides are the most popular catalysts used for potable water and air treatment; of these, TiO 2 and ZnO have been the most researched and are considered the most efficient. Of these two, TiO 2 tends to be favored due to its stability in extreme conditions (low or high pH), insolubility, heat resistance, non-toxicity, and low cost. TiO 2

PAGE 18

9 is also photostable as compared to ZnO which can undergo photocorrosion, decreasing the effective lifetime of the catalyst (Okamoto et al., 1985). Titanium is the ninth most abundant element in the earths crust and occurs in nature only in combination with other elements, such as with oxygen to form TiO 2 (US Geological Survey, 2002). Naturally occurring forms of titanium dioxide are usually combined with iron (FeTiO 3 or FeO-TiO 2 ) and contain other impurities. Therefore, the TiO 2 (titania) used as a photocatalyst is manufactured from synthetic means. Of the commercially available titanium dioxides, Degussas P25 seems to be the most researched for its photocatalytic ability in water (Halmann, 1996). Table 2-2 shows a few specific characteristics of P25. The product is produced from high-temperature flame hydrolysis of TiCl 4 in the presence of oxygen and hydrogen and is then steam treated to remove HCl, resulting in a 70:30 anatase to rutile phase TiO 2 (Degussa Technical Bulletin, 1990). The phase of TiO 2 is important since the anatase crystal form is considered the more photoreactive of the two (Tanaka et al., 1991). Figure 2-3 shows the crystal structures of the two phases. A difference in reactivity of the phases is attributed to 1) the more positive conduction band of the rutile phase, hindering molecular oxygen to act as an electron acceptor, and 2) the difference in surface properties between the two phases (Tanaka et al., 1991). Table 2-2. Characteristics of Degussa P25 Titanium Dioxide BET Surface Area 50 m 2 /g Average Primary Particle Size 21 nm Band Gap: anatase, rutile 3.29 eV, 3.05 eV Point of Zero Charge pH 6.0 Source: adapted from Degussa Technical Bulletin (1990)

PAGE 19

10 Figure 2-3. Diagram of anatase and rutile crystal structures (Millennium Chemical, 2002). Crystal morphology, as well as a variety of other factors, can influence the activity of a catalyst. Surface area is known to affect the adsorption ability of the catalyst, as well as the available contact area for oxidation and reduction reactions. It would make sense that a catalyst with a large surface area would have an increased rate of reaction, but Tanaka et al. (1991) and Mills et al. (1994) showed that the destruction efficiencies of several different titanium dioxide samples were independent of surface area (within the range of 2.75 m 2 /g to 177 m 2 /g). Therefore, variables besides surface area may have a more influential role in the activity of a catalyst. These variables are crystal morphology, as well as crystallite size, prevention of electron-hole recombinations, zeta potential, and band gap energy (Okamoto et al., 1985; Palmisano and Sclafani, 1997; Suri et al., 1993; Tanaka et al, 1991; Zhang et al., 2001). Tanaka et al. (1991) showed that the photocatalytic ability of a TiO 2 catalyst is foremost affected by the anatase content and secondarily affected by the size of the

PAGE 20

11 crystal. For example, a photocatalyst containing larger crystals is more efficient than one with smaller crystals, even when both contain equal percentages of anatase. This could be due to the larger migration distance of the holes and electrons to the surface of the catalyst, thereby decreasing the possibility of recombination (Tanaka et al., 1991). Modifications in the catalyst surface have also been investigated for prevention of electron-hole recombination. The addition of platinum and other transition metals have been successfully arrayed on the titanium dioxide surface (Abrahams et al., 1985; Okamoto et al., 1985; Martin et al., 1994; Suri et al., 1993). These metal additions have an optimum at low weight percentages (less than 5%), above which the metal actually hinders the photocatalytic ability. The isoelectric point or point of zero charge (PZC) represents the pH at which an immersed solid oxide would have zero net charge, resulting in electrically equivalent concentrations of positive and negative complexes on the surface. At a pH above the PZC, interactions with cationic electron donors and acceptors will be favored; while anionic electron donors and acceptors will be favored at pH below the PZC. The band gap is equal to the amount of energy required to activate the surface of the catalyst. A band gap of 3.29eV corresponds to a wavelength of 378 nm. Therefore, wavelengths below 378 nm (such as those emitted by UV light) would have the required energy to activate the anatase phase of TiO 2 2.3 Reactor Design A problem in the practical application of heterogeneous photocatalysis of environmental pollutants is the design of a reactor that will maximize photocatalytic efficiency while utilizing the least amount of energy. The simplest reactor configuration uses an aqueous suspension or slurry of the photocatalyst. Most photocatalysis studies

PAGE 21

12 use a slurry system in preliminary testing to establish the feasibility of pollutant mineralization or microbe inactivation. For these systems, an optimum weight of TiO 2 per volume of solution (wt/vol) was found that yielded the most degradation for specific types of compounds at a specific light intensity. For instance, Goswami et al. (1993) found 0.1% (wt/vol) TiO 2 to be the optimum loading for most hydrocarbons and Block and Goswami (1995) found 0.01% (wt/vol) as the best for microbial destruction. Although slurry systems are considered to be the most photocatalytically efficient, there is a dilemma to be addressed in the design of the reactor. Due to the extremely small size of the particles (between 0.1 30 nm, depending on the manufacturing source), there is a difficulty in recovering the catalyst from the purified water. This problem has directed research to investigate the use of catalyst supports. There are many research groups that have examined the immobilization of photocatalysts on nonporous glass surfaces as a way of providing a backbone of support for the particles, while also allowing the penetration of light to activate the catalyst surface. Successful immobilization has been done on glass beads (Jackson et al., 1991; Serpone et al., 1986; Zhang et al., 1994) and on the inside surface of glass tubes (Al-Ekabi and Serpone, 1988; Matthews, 1988). The oxidation rates were usually lower with the immobilized catalysts than with free suspensions. The determinate factor was assumed to be the result of mass transfer limitations, for it was observed that as the flow rate increased, resulting in more efficient mixing, the oxidation rate likewise increased. While these designs may have certain advantages, it is hypothesized that an efficient support for a catalyst would be a material with adsorption capability that would bring the contaminants into close contact with the catalyst surface (see Figure 2-4) where

PAGE 22

13 the hydroxyl radicals predominantly exist (Turchi and Ollis, 1990). Indeed, there has been a reported increase in destruction efficiency when the reactants are brought into close contact with the surface of activated TiO 2 particles. This increase could be due to the greater concentration of the substrate around the catalyst surface due to adsorption interactions, for it has been proven that the highest efficiency of photocatalytic degradation of organic pollutants is observed at relatively high concentrations of reagents (Emeline et al., 2000). Figure 2-4. Simplified diagram of a catalyst support. 2.4 Catalyst Supports Adsorbents researched as possible catalyst supports include activated carbon, silica gel, mordenite, alumina, zeolite, and aerogels (Takeda et al., 1995; Torimoto et al., 1996; Yoneyama and Torimoto, 2000). The highest decomposition rate for propyzamide (herbicide) was obtained with supports having medium adsorption strengths, such as silica and mordenite (Torimoto et al., 1996). With supports of too low (zeolite) and too high (activated carbon) adsorption strengths, no significant effect on the rate of destruction was found.

PAGE 23

14 Yet, research has focused on the benefits of using activated carbon (AC) as a catalyst support versus other supports. Uchida et al. (1993) and Lu et al. (1999) found that while the oxidation rates of a specific compound were lower with the TiO 2 loaded AC, the complete mineralization rates were greater than with a plain slurry of TiO 2 or any other catalyst loaded support. They concluded that the lower oxidation rate can be attributed to the carbon blocking some of the photons of light from reaching the catalyst surface and the higher mineralization rate can be attributed to the high adsorptive ability of the carbon to adsorb the intermediate compounds that are being formed. An additional benefit of TiO 2 coated carbon is its potential for in situ AC regeneration. Many water utilities use granular activated carbon (GAC) for the removal of organic compounds. However, after the absorbent is spent, removal of the carbon from the system is required for reactivation, disposal in a hazardous waste landfill, or incineration. While on-site thermal reactivation is an option for some utilities, it is not the most economical option for many. Therefore, research has been looking at optimizing both TiO 2 -coated carbons adsorption capability and its regeneration efficiency for the removal and subsequent oxidation of organic pollutants adsorbed on the spent carbon surface (Crittenden et al., 1993; Khan et al., 2002; Sheintuch and Matatov-Meytal, 1999). While activated carbon has some benefits as a catalyst support, its use requires a reactor that efficiently exposes the catalyst surface to the photons of light. Since this may require mixing or fluidizing the particles, the inherent attrition can cause the catalyst to detach itself from the carbon (Lu et al., 1999). An additional problem is that the surface chemistry of a carbon may hinder effective coating (Khan et al, 2002). The various methods of activating carbon can create different functional groups on the surface, thus

PAGE 24

15 affecting the properties a specific carbon will display. This can greatly affect the immobilization of a catalyst to the carbon surface. Silica gels, on the other hand, have many advantages over activated carbon as a catalyst support. The transparency of silica allows the penetration of photons to the catalyst surface. This is extremely beneficial and allows for a fixed-bed reactor design that can be highly efficient in relation to input energy. Silica also has high mechanical strength, thermal stability, and can be synthetically formed into any shape, such as cylindrical pellets (Yamazaki et al., 2001) or fibers (Brinker and Scherer, 1990). In addition, silica is an adsorbent and is commonly used in chromatography columns for adsorption and the resulting separation of compounds in an aqueous sample. Furthermore, TiO 2 and SiO 2 can be chemically combined, enabling the formation of highly efficient photocatalysts. These TiO 2 SiO 2 photocatalysts allow the placement of the catalyst on both external surfaces and internal surfaces within the porous silica matrix where pollutants are adsorbed. Anderson and Bard (1995, 1997) found a strong synergy between silica and titania as a combined oxide. For substances that are easily adsorbed to the silica surface, a higher decomposition rate was found with TiO 2 SiO 2 composites versus a plain TiO 2 slurry. The presence of an adsorbent was considered to promote efficiency by increasing the concentration of the substrate near the TiO 2 sites relative to the solution concentration. But for substrates that are not readily adsorbed to the surface of silica, the degradation rates were lower than compared to a slurry of pure TiO 2 However, when these rates were normalized to the TiO 2 content of the material, the TiO 2 -SiO 2 material

PAGE 25

16 demonstrated a more efficient use of the TiO 2 sites than for TiO 2 alone. The highest initial degradation rate was found with a 30/70 wt% TiO 2 -SiO 2 composite. Jung and Park (2000) also found 30 wt% titania (with gels prepared in a similar process to Anderson and Bard) to be the optimum in their studies for the mineralization of trichloroethylene. In addition, they concluded that the high porosity and large pore size of the silica facilitated the mass transfer of reactants, resulting in a higher rate of degradation than a plain slurry of Degussa P25. Chun et al. (2001) used a different preparation method than Jung and Park (2000) and Anderson and Bard (1995) and still found 30 wt% TiO 2 to be an optimum. Chun et al. additionally showed, using two organic compounds with differing characteristics, that there is a strong correlation between adsorption of the compound on the mixed oxide surface and the destruction rate of that compound. In conclusion, silica-titania composites can be efficiently utilized in heterogeneous photocatalysis systems for the adsorption and subsequent destruction of unwanted compounds in an aqueous solution. In addition, a dosage of 30 wt% (equal to 12% on a weight per volume of silica precursor) was also found to be an optimum loading in this research (see Chapter 4) for the silica-titanium composites created using a sol-gel doping method (see Chapter 3). 2.5 Silica-Titania Composites Silica is the most abundant oxide on the earth, yet despite this abundance, silica is predominantly made by synthetic means for its use in technological applications. Since synthetic silicas have a higher surface area than naturally occurring forms of silica, the synthetic silicas provide the best adsorption and catalyst support structure for heterogeneous photocatalysis. Although silica has a simple chemical formula (SiO 2 ), it can exist in a variety of forms, each with its own structural characteristics, as well as

PAGE 26

17 chemical and physical properties. In general, SiO 2 is a SiO 4 tetrahedra, where each silica atom is bonded to four oxygen atoms and each oxygen atom is bound to two silica atoms. Figure 2-5. Linkages of SiO 2 tetrahedras (Hench and West, 1990). There has been much research on the different preparation methods for creating a solid SiO 2 -TiO 2 material with photocatalytic ability. The characteristics (e.g., pore size, surface charge, mechanical strength, and adsorption sites) of the final product are dependent on the synthesis conditions and the type of interaction between TiO 2 and SiO 2 There are two forms of interaction: physical forces of attraction (such as Van der Waals forces) and chemical bonding (creation of a Ti O Si bond). The physically supported TiO 2 on SiO 2 preparation methods have been the least researched, while the chemically bonded TiO 2 on SiO 2 methods (also called mixed oxides) have been thoroughly investigated (Gao and Wachs, 1999). The most widely used methods of preparation (for details on these individual methods, see Appendix A) for creating mixed oxides are sol-gel hydrolysis and coprecipitation. Chemical vapor deposition, precipitation, and impregnation are the methods commonly used for creating supported oxides, yet there is evidence of Ti-O-Si bonds attaching the titania to the silica surface (Chun et al., 2001). A problem with supported oxides is the inconsistency in the coating from one batch to the next and the

PAGE 27

18 nonuniformity of the coating due to the sensitivity of the procedure to its conditions (Hanprasopwattana et al., 1996; Lei et al., 1999). Another problem with any of these methods is the assurance of the formation of an anatase phase of TiO 2 Even though X-ray diffraction (XRD) can be used to determine the phase of TiO 2 the question remains whether the preparation method has formed the most efficient semiconductor particle, as compared to commercially available photocatalysts. For example, Liu and Cheng (1995) produced a TiO 2 -SiO 2 mixed oxide using coprecipitation and found that the SiO 2 inhibits the growth of TiO 2 crystals, resulting in amorphous TiO 2 within the silica matrix. A method of titania incorporation in silica has been devised to overcome both of these potential problems. This method uses sol-gel hydrolysis to create the silica matrix, then during gelation the solution is doped with the commercially available highly efficient Degussa P25 TiO 2 The silica network may either form around the titania particle and/or form a bond with the titania to secure its position in the matrix. 2.5.1 Sol-Gel Chemistry A sol is a colloidal suspension of particles, while the term gel refers to the semi-rigid material formed when the colloidal particles link together in a liquid to form a network. Sol-gel precursors consist of metals or metalloids surrounded by various ligands, of which the most widely used are metal alkoxides (Mauritz, 2002). Aluminates, titanates, borates, and silicates are just some of the alkoxides that can be used to create a sol-gel. A gel that contains SiO 2 networks is given the term silica gel. Hydrolysis and condensation. Gels derived from alkoxides form through hydrolysis reactions followed by condensation reactions (Equations 5-7). Generally speaking, hydrolysis replaces the alkoxide group with a hydrogen ion while condensation produces siloxane bonding (Si-O-Si) with the products of the reaction being water or

PAGE 28

19 alcohol. As the number of siloxane bonds increases, the individual molecules join together to create a silica network. See Equations 5-7 (Mauritz, 2000). Gelation. After condensation reactions begin, gelation can occur (see Figure 2-6 for structure changes). Gelation is considered the growth period for the colloidal particles, which grow by polymerization or aggregation, depending on the conditions. Eventually, as the last links are made, the sols viscosity increases to the point where the sol becomes a semi-rigid gel that consists of a network of pores filled with liquid. Aging. Aging occurs after gelation. Here, terminal silanols re-orient themselves and react with each other to form additional network linkages. This results in shrinkage of the gel and forces the removal of some liquid from the pores. As the gel stiffens, Ostwald ripening occurs. This is the phenomenon whereby necks begin to form between primary particles causing the filling of small pores (see Figure 2-6 part c).

PAGE 29

20 Figure 2-6. Structural changes during the sol-gel process (adapted from Hench and West, 1990) Drying. The drying step involves the removal of the solvent from within the pores of the gel. This is very important in the determination of pore size. There are three types of gels characterized by the solvent removal process: xerogels, aerogels, and cryogels. As defined by Legrand (1998), a xerogel is obtained by evaporation of the liquid component at ambient pressure and temperatures below the critical temperature of the liquid; an aerogel is obtained from the evaporation of the liquid component above its critical temperature; and a cryogel is obtained from the freezing and subsequent sublimation of the liquid component. Aerogels have higher surface areas than xerogels and do not have a material volume loss due to shrinkage. Yet, because supercritically drying gels can be cumbrous and expensive, research has focused on making xerogels with similar

PAGE 30

21 characteristics to aerogels. As far as water treatment is concerned, xerogels would be preferred over aerogels due to the higher mechanical strength of the xerogel. Curing. Calcination or curing is the final heat treatment step in producing a mechanically strong gel. This thermal treatment also affects several characteristics of the silica gel. With increased heat treatment, the BET (Brunauer, Emmett, and Teller equation) surface area will decrease due to sintering (Papirer, 2000) and the dehydration (around 180-200 o C) and subsequent dehydroxylation (above 200 o C) of the surface of the silica, resulting in the formation of siloxane bridges (Holysz, 1998). The final product resulting from all these steps (hydrolysis, condensation, gelation, aging, drying, curing) is a gel containing a relatively monodisperse pore size and displaying specific characteristics associated with the conditions it experienced during processing. Effect of catalyst. The two alkoxides most often used are tetramethoxysilane (TMOS) and tetraethoxysilane (TEOS). TMOS has the advantage of rapid hydrolysis under a variety of conditions, but the toxic methanol produced can be hazardous to the eyes and lungs. Conversely, TEOS is hydrolysis rate limited (Brinker and Scherer, 1990), but produces less toxic ethanol during the reaction. To increase the hydrolysis reaction rate of TEOS, an acid or base can be added as a catalyst. However, the nature and concentration of the catalyst will affect the characteristics of the gel. For example, the addition of hydrofluoric acid will rapidly increase the rate of gelation as well as increase the pore size (Powers, 1998). Other important parameters to consider in the production of a gel are H 2 O to Si molar ratio (called the R factor) and the aging and drying schedules. A high R ratio will promote more rapid hydrolysis, for the concentration of water affects the rate of

PAGE 31

22 hydrolysis. The drying time and temperature also affect the pore size of the gel as well as the shape of the gel. For example, the longer the time to dry, the more Ostwald ripening occurs and the less likely the gel will crack, thereby producing a larger pore size than a gel of the same composition dried in short period of time (less than 48 hours). In summary, there are many parameters in the production of a gel that will greatly affect the final products characteristics. It is these differences in silicas produced by various preparation methods that make the comparison of data found in the literature a very difficult process. 2.5.2 Surface Chemistry and Adsorption Characteristics The surface of silica consists of two types of functional groups: silanol groups (Si O H) and siloxane groups (Si O Si). The silanol groups are the locale of activity for any process taking place on the surface, while siloxane sites are considered nonreactive (Unger, 1979). Porous amorphous silica contains three types of silanols on its surface: isolated, geminal, and vicinal (or associated). Figure 2-7 shows the distinction between these groups. The unequal distribution of the silanols in the sililca matrix, resulting from irregular packing of macromolecules as well as incomplete condensation, results in a heterogeneous surface (non-uniformity in the dispersion of silanol groups) for synthesized silica (El Shafei, 2000). The various silanols can have different adsorption activity and current knowledge indicates that the isolated silanols are the more reactive species (Nawrocki, 1997; El Shafei, 2000). With increasing temperature of heat treatment, the silica surface becomes hydrophobic due to the condensation of surface hydroxyls (dehydroxylation) and the formation of siloxane bridges (see Figure 2-8). This increases the ratio of isolated sites versus other silanol groups on the surface and therefore increases adsorption.

PAGE 32

23 Figure 2-7. Silanol groups on the silica surface (Legrand, 1998). At temperatures up to 450 o C, the removal of the hydroxyl groups does not result in visible changes in structure, density, or specific surface area of the silica because the rehydroxylation makes it possible to return to the initial state (Davydov, 2000). Rehydroxylation of the silica surface by exposure to water vapor is reversible up to 600 o C. At temperatures above this, the rehydroxylation becomes a slow process or remains incomplete (Unger et al., 2000). Also, there is a decrease in surface area as a result of sintering at temperatures above 500 o C (Persello, 2000). Figure 2-8. Representation of dehydroxylation (Burneau and Gallas, 1998). Pore size can be affected by the heat treatment of the silica, but is also a function of the manufacturing process. The effect of pore size on adsorption is dependent on the compounds characteristics, such as molecule size, shape, and functional groups (Dabrowski, 2001; Goworek et al., 1997; Goworek et al., 1999). In general, the smaller the pore, the larger the surface area and the larger the capacity for adsorption of

PAGE 33

24 contaminants. Yet, Goworek et al. (1997) found that some compounds may be hindered by a narrow pore size due to steric forces. Therefore, the adsorption ability for that particular compound would be higher in a silica with larger pores. Adsorption on silica surfaces is primarily due to ion interactions and hydrogen bonding with the silanol groups. Therefore, the adsorption kinetics in relation to these silanols is greatly dependent on the pH. Silicas point of zero charge ranges between 2 and 3 (Papirer, 2000; Persello, 2000). At a pH less than the PZC, the surface has a positive charge due to the formation of Si-OH 2 + At a pH above the PZC, the surface of the silica has a negative charge due to the deprotonation of the silanol group resulting in Si-O As the pH approaches 7, the Si-O sites become significant (the pKa for SiOH = SiO + H + is 7) (Cox, 1993). These Si-O sites readily react with cations in solution and an increase in pH would greatly increase the adsorption of these compounds (Kinniburgh and Jackson, 1981). For nonionic compounds in solution, adsorption can occur at the silanol groups by hydrogen bonding or by Van der Waals forces. For anionic compounds, there may be some adsorption at a pH above the PZC of silica, but it would be occurring at the TiO 2 surface because it is positively charged at pH < 6.0. One factor to consider when adjusting pH in order to achieve the highest adsorption onto silica-titania composites is the dissolution of silica at a pH greater than 9 (Iler, 1979). In order to create a long-lasting photocatalytic composite, the dissolution of silica would need to be avoided.

PAGE 34

CHAPTER 3 MATERIALS AND METHODS 3.1 Silica-Titania Composites The silica-titania composites were made by a sol-gel method (Powers, 1998) using nitric acid and hydrofluoric acid as catalysts to increase the hydrolysis and condensation rates, thereby decreasing the gelation time. The basic formula used to create gels with a pore size of roughly 140 is as follows: 25 mL water, 50 mL ethanol, 35 mL TEOS (tetraethylorthsilicate), 4 mL nitric acid (1N), and 4 mL HF (3%). TEOS was used in place of TMOS (tetramethylorthosilicate) because it produces less toxic ethanol instead of highly toxic methanol that would be created by reactions with TMOS. Varying the volume of hydrofluoric acid allowed the manipulation of the pore size. Decreasing the volume to 1 mL HF resulted in 30 pores, while increasing it to 8 mL resulted in 320 pores. The chemicals (reagent grade from Fisher) were added individually, in no particular order, to a polymethylpentene container. A magnetic stir plate provided sufficient mixing. The solution was allowed to mix until gelation occurred (2 hours for 140 gel). During this time, a known mass of Degussa P25 was added to the batch and the percentage of titania recorded is given as a percent by volume of silica precursor. After gelation, the lid was put on, preventing any air exposure in order to protect against premature evaporation. Next, it was aged at room temperature for two days, then at 65 o C for two days. After aging, the gel was removed from the container, rinsed with deionized water to remove any residual acid or ethanol, and placed in a Teflon container for the 25

PAGE 35

26 next series of heat treatments. A small hole in the lid of the container allowed slow and uniform drying of the gel. It was then placed in an oven and the temperature was ramped from room temperature to 103 o C (2 o /min) and kept constant for 18 hours, resulting in the vaporization of liquid solution within the silica network. Next, the temperature was ramped to 180 o C (2 o /min) to remove any physically adsorbed water. It was kept constant for 6 hours and then was slowly decreased back to room temperature over a 90 minute period. These gels were powdered and sieved through 325 mesh to be used for batch performance studies. The titania-doped pellets were made in a similar fashion, except the solution (including the P25) was pipeted into polystyrene 96-well assay plates before gelation. The volume added to each well was approximately 0.3 mL. The plates were then covered with lids and wrapped in foil to prevent premature evaporation. The same aging and drying schedule as previously discussed was used, but before the drying step, the pellets were removed from their plates and placed in a Teflon container. The resultant size of an individual cylindrical pellet after drying was approximately 5 mm in length with a diameter of 3 mm. These pellets were used in column performance studies. 3.2 Batch Performance Studies The batch performance studies were done to determine an optimal titania loading, curing temperature, and pore size that is efficient for both adsorption and destruction of contaminants. The powdered gels were compared based upon destruction (removal of color) of reactive red dye (RR) from Sigma-Aldrich and adsorption of crystal violet dye (CV) from Fisher. All dye solutions were made with nanopure water. The destruction studies were performed as follows: 30 mg of gel was added to 100 mL of a 10 mg/L RR solution buffered with 200 mg/L of sodium bicarbonate in a 125 mL Erlenmeyer flask.

PAGE 36

27 Then pH adjustments were made with sodium hydroxide or nitric acid. The flask opening was then covered with parafilm to maintain a closed system with a limited amount of dissolved oxygen (DO). In a system designed for NASA, oxygen would be a limiting factor due to the problem of continually supplying oxygen for water treatment for long duration space flight. The flasks were placed within a box containing four magnetic mixers relatively equidistant from a single 4-Watt UV lamp. The UV exposure time was two hours and the percent destruction was calculated by comparing the initial and final absorbance readings using a Hach DR/4000U (Loveland, Colorado) Spectrophotometer at RRs peak absorption wavelength of 538 nm. The spectrophotometer gave a linear response across the concentration range for each dye. The UV intensity (mW/ cm 2 ) was varied by changing the position of the mixers in regard to the lamp and by the use of additional lamps. UV intensity (wavelengths between 320 to 380 nm) was measured with a VWR Ultraviolet Light Meter (Atlanta, GA). Each mixer was placed where the desired UV intensity was achieved at the mixers center. The adsorption studies were performed as follows: 30 mg of gel was added to 100 mL of 10 mg/L CV solution with no bicarbonate added. The studies were conducted at a pH of 4, using nitric acid and sodium hydroxide for the pH adjustments. The flasks were covered with parafilm and placed in a box with no penetration of light and allowed to mix for 24 hours. The percent adsorbed was calculated from the absorbance reading at a wavelength of 590 nm. The BET (Brunauer, Emmett, and Teller equation) surface area and pore volume analysis was performed on a Quantachrome NOVA 1200 Gas Sorption Analyzer (Boynton Beach, FL). The powdered samples were outgassed at 110 o C for

PAGE 37

28 approximately 24 hours and analyzed using nitrogen adsorption isotherms. The average pore size was calculated by dividing the pore volume by the surface area. Pore size distribution curves were also attained and coincided with the calculated pore size. XRD (X-ray diffraction) analysis was performed on cured gels to determine the phase of titania, anatase or rutile. These gels were cured at temperatures up to 800 o C as an additional curing step after the regular drying schedule. The analysis was performed on a APD 3720 from Philips Analytical (Almelo, Netherlands). The reflections of the two crystalline TiO 2 phases, anatase (2 = 25.3) and rutile (2 = 27.4), were taken from Grieken et al. (2002). 3.3 Column Performance Studies The purpose of these studies was to show the performance of the pellets in a regenerative system, which places equal significance on the adsorption of a contaminant as well as the feasibility of the contaminant to be mineralized. A cylindrical column (6 inches in length, 0.5 inch diameter) with a porous frit was filled with 10 mL of pellets. The column setup is shown in Figure 3-1. A CV dye solution of 2 mg/L was pumped through the column until the effluent concentration equaled the influent concentration. Then the CV pump was turned off and the deionized water and UV light was turned on (flux = 1.1 mW/cm 2 ) for a set time (e.g., one hour). During this regeneration time, the water was pumped through the column to provide the reactants (O 2 H 2 O) required for photocatalysis. After the regeneration time, the CV pump was turned back on. This cycle was continued several times for separate packed columns of 140 gels and 30 pellets, both loaded with 12% TiO 2 The flowrate for both the deionized water and CV dye was 8 mL/min.

PAGE 38

29 Figure 3-1. Column setup.

PAGE 39

CHAPTER 4 RESULTS AND DISCUSSION 4.1 Dye Comparison In order to accurately measure the photocatalytic ability of the silica-titania (SiO 2 TiO 2 ) composites, it was important to choose a dye that displayed little or no photolytic behavior, i.e. change in absorbance as a response to UV light exposure alone. The following four dyes were examined in order to determine the most suitable dye for the studies herein: malachite green (MG), methylene blue (MB), crystal violet (CV), and reactive red (RR). See Appendix B for structures of these dyes. Using a spectrophotometer, each dye was scanned at a range of 290 nm to 800 nm in order to find each dyes peak absorbance wavelength. A summary of the peak wavelengths is shown in Table 4-1. The absorbance measured at this peak wavelength was used for comparing initial and final readings. Table 4-1. Peak absorbance wavelengths. D y e Peak Wavelen g th MG 619 n m MB 656 n m CV 590 n m R R 538 n m A 100 mL solution of each dye (10 mg/L) was prepared in a 125 mL Erlenmeyer flask and exposed to three different intensities of UV light for two hours. The findings are presented in Figure 4-1. Each dye experienced a change in absorbance of less than 5% at each flux, except for crystal violet, which experienced a 19% change at a flux of 1.1 30

PAGE 40

31 mW/cm 2 Based upon this undesirable change, crystal violet was excluded for future destruction studies. Subsequently, the remaining dyes were evaluated for photolysis at various concentrations (0.5 mg/L to 10 mg/L). At 0.5 mg/L, there was a 98% and 20% change between the initial absorbance readings for MB and MG respectively, compared to their final readings after two hours of UV exposure (0.45 mW/cm 2 ) (Figure 4-2). Large discrepancies were also observed at 1 mg/L and 2 mg/L. Therefore, MB and MG were eliminated as viable candidates to investigate the photocatalytic ability of the composites. For RR (see Figure 4-3), the largest change in absorbance was only 13% at a flux of 1.1 mW/cm 2 the highest flux achievable. Thus, RR was chosen for the photocatalysis studies. The amount of photolysis was not subtracted from the results of the destruction studies. 0510152025RRCVMBMGDyePercent Change in Absorbance 0.45 mW/cm^2 0.75 mW/cm^2 1.1 mW/cm^2 Figure 4-1. Photolysis of dyes as a function of UV flux for 2 hours of exposure.

PAGE 41

32 00.20.40.60.811.21.41.61.820246810Concentration (mg/L)Absorbance Read at Peak Wavelength MB Initial Absorbance MB Final Absorbance MG Initial Absorbance MG Final Absorbance 45% for MG and 20% for MB 98% for MB and 20% for MG 12 Figure 4-2. Photolysis of MB and MG (2 hours UV exposure at 0.45 mW/cm 2 ) as a function of initial dye concentration. 00.050.10.150.20.250.3012345678910111Concentration (mg/L)Absorbance Initial Absorbance Final Absorbance (0.45 flux) Final Absorbance (1.1 flux) 2 Figure 4-3. Photolysis of RR (2 hours UV exposure) as a function of initial dye concentration and flux.

PAGE 42

33 Besides photolysis, pH can also affect a dyes peak absorption wavelength. Therefore, RR solutions of 10 mg/L were prepared at different pHs and compared before and after UV exposure. The change in absorbance (at 538 nm) as a function of pH is shown in Figure 4-4. At a pH above 8, there was a decrease in absorbance along with a slight visible change of color from a dark pink to a red. Figure 4-5 displays a section of the spectrophotometer scan and shows the changes that occur at an alkaline pH. At a pH of 7, the maximum peak wavelength is 538 nm. At a pH of 9.5, the highest peak occurs at both 514 nm, as well as 538 nm. After two hours of UV exposure, the max peak returns back to 538 nm. Also noticeable at pH 9.5 is a slight downshift in the curve after UV exposure. Therefore, the pH was kept below 8 for comparison studies of gels to ensure accurate measurements of destruction ability. 00.050.10.150.20.250.30246810pHAbsorbance Abs. (initial) Abs. (final) 12 Figure 4-4. Change in absorbance of RR at 538 nm as a function of pH.

PAGE 43

34 00.050.10.150.20.250.3400450500550600Wavelength (nm)Absorbance Absorbance pH 7 Absorbance pH 9.5 Absorbance ph 9.5 after UV Figure 4-5. Spectrophotometer scan of 10 mg/L RR as a function of pH, before and after UV exposure. 4.2 Optimization of Gels 4.2.1 Titania Loading In order to determine the optimal titania loading, 14 gels (140 ) of varied TiO 2 loadings (from 0.77% to 40%) were created. The desired optimal loading would be found as the composite having the highest rate of destruction for RR. These gels were performance tested at three UV intensities for the destruction of RR (10 mg/L) and the results are displayed in Figure 4-6. There is a definite plateau where the percent destruction remains constant, even with the addition of more TiO 2 It appears that this plateau occurs at lower TiO 2 percentages with higher UV intensities. For 1.1 mW/cm 2 the plateau occurs at 5% TiO 2 and it occurs at 10% for 0.75 mW/cm 2 At the lowest UV intensity, the destruction levels off at 12%. These results were triplicated and the error bars are included in the figure.

PAGE 44

35 10203040506070800510152025303540Percent TiO2Percent Destruction 1.1mW/cm^2 0.75mW/cm^2 0.45mW/cm^2 Figure 4-6. Destruction of RR (pH 7.6) after 2 hours UV exposure versus TiO 2 loading as a function of flux. The plateau can be explained by the TiO 2 is agglomerating in solution during gel preparation, which was visually observed with TiO 2 loadings greater than 8%, and therefore is decreasing the effective surface area of titania available for excitation and/or oxidation and reduction reactions on its surface. The 12% optimal loading (wt/vol precursor) at the lowest energy is approximately equal to 30 wt%, which was found to be an optimum for destruction of organic compounds with other methods of producing silica-titania composites (Jung and Park, 2000; Chun et al., 2001; Anderson and Bard, 1995). The structure of RR includes sodium sulfonate groups that would make the compound anionic when the sodium disassociates as the compound dissolves. Therefore, the effect of pH on adsorption and destruction of RR was also investigated. It was observed that at a pH range of 3 to 10, there was no adsorption of RR dye to the silica

PAGE 45

36 titania gels. Thus, the color disappearance of the dye can be directly attributed to photocatalysis with a minimal amount due to photolysis. The effect of pH on the destruction of RR is shown in Figure 4-7. It was observed that acidic conditions enhanced the photocatalytic destruction of RR and it was initially hypothesized that this was due to the TiO 2 having a positive charge at pH 3, therefore, the amount of RR adsorption that would occur on pure TiO 2 at low pH would increase the destruction rate. 0102030405060708090100012345678910111213Percent TiO2Percent Destruction pH 3, bicarb pH 4, no bicarb pH 7.6, bicarb pH 7.6, no bicarb Figure 4-7. Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm 2 ) versus TiO 2 loading as a function of pH. Yet, when a slurry of titania was tested for destruction and adsorption at a range of pHs, the results were contrary to what was hypothesized (Figure 4-8). While the adsorption onto the titania surface may have slightly increased at a lower pH, the level of total removal (destruction and adsorption) did not change. It is then presumable that at a pH greater than 6, the silica surface becomes more concentrated with SiO groups, thus

PAGE 46

37 repelling RR. As the pH approaches 3, the silica has less of a negative charge and would allow the penetration of RR into the pores. Hence, acidic conditions would increase the ability of RR to get within the vicinity of the titania. 0102030405060708090100012345678pHPercent Removal0102030405060708090100Percent Adsorption Destruction Adsorption Figure 4-8. Destruction of RR (10 mg/L) after 1 hour UV with TiO 2 slurry at various pHs. Grey water contains many sulfates, phosphates, and salts that may inhibit the destruction of unwanted compounds (Chen et al., 1997). The addition of sodium bicarbonate allows one to see the performance of the silica-titania composites in a realistic feed stream. In Figure 4-7, it is also interesting to notice the affect of bicarbonate on the destruction efficiency. There is little increase in destruction (with no bicarbonate) at a pH of 7.6 within a range of 0.77% to 6% TiO 2 At a larger TiO 2 percentage, the difference becomes very noticeable. However, Figure 4-9 shows that the increase in destruction without bicarbonate present is only seen within a certain range of TiO 2 loading (8% to 12%).

PAGE 47

38 01020304050607080901000510152025303540Percent TiO2Percent Destruction no bicarbonate with bicarbonate Figure 4-9. Effects of bicarbonate at pH 7.6 versus TiO 2 loading. It was also thought that surface area may play a role in the leveling-off seen in destruction of RR. As can be seen in Figure 4-10, there is no real trend between BET surface area and titania loading, as well as between pore volume and titania loading. Notice that this surface area is normalized on a volume basis. This trend is different for silica-titania composites prepared by other methods, for which there is a drastic change (up to a loss of 200 m 2 /g) in surface area with an increase in percentage of titania (Greiken et al., 2002). In conclusion, there is no correlation between surface area or pore volume and the destruction of RR. For adsorption studies, CV dye was used because of its cationic nature resulting in a high affinity for the silica-titania composites. These studies were conducted at a low pH (worst case scenario for adsorption of a cation contaminant onto a silica surface). Figure 4-11 compares the amount of adsorption versus titania loading. There is an increase in

PAGE 48

39 adsorption with additional titania in the gel. It is interesting to note, however, that adsorption, similar to the destruction studies, reached a plateau at 12% loading. Therefore, based upon the destruction studies and adsorption studies, an optimal titania loading of 12% TiO 2 was chosen. 010203040506070051015202530354045Percent TiO2BET SA (m2/g/cc)00.10.20.30.40.5Pore Volume (cc/g/cc) BET SA/vol Pore Volume Figure 4-10. BET surface area and pore size versus titania loading. 05101520253035051015202530354045Percent TiO2Percent Adsorption Figure 4-11. Adsorption of CV (10 mg/L) versus TiO 2 loading after 24 hours of mixing.

PAGE 49

40 4.2.2 Curing Temperature The effect of curing temperature was investigated on gels doped with 6% TiO 2 and those with no TiO 2 (all made with the 140 initial formula). The literature showed that there can be an increase in adsorption with gels produced at higher curing temperatures (up to 400 o C) due to increasing the ratio of isolated silanols on the surface versus other types of silanols. Yet, the curing temperature may have detrimental effects on the photocatalytic ability of the composites. In looking at the destruction of RR (10 mg/L), Figure 4-12 shows that there is a 30% difference between the 180 o C gel versus the 500 o C gel at a pH of 7.8. This difference is less pronounced at pH 4. At 800 o C, there is a marked decrease in destruction, regardless of pH. This can be explained by both a decrease in surface area of the gels at 800 o C (see Figure 4-13) from an average of 275 m 2 /g to 150 m 2 /g for the titania loaded gel and a phase change of titania. 01020304050607080901000100200300400500600700800Curing Temp (oC)Percent Destruction pH 4 pH 7.8 Figure 4-12. Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm 2 ) versus curing temperature as a function of pH.

PAGE 50

41 0501001502002503003504000200400600800Curing Temp (oC)BET SA (m2/g) 6% TiO2 0% TiO2 Figure 4-13. Effect of curing temperature on surface area. The anatase phase of TiO 2 begins transforming into rutile phase at a temperature of 600 o C and becomes complete at 1000 o C (Zhang et al., 2001). It was thought that the shift in the phase of titania at 800 o C may have also affected its photocatalytic ability, but XRD analysis (Figure 4-14) revealed only a slight shift from the anatase phase (2 = 25.3 o ) to the rutile phase (2 = 27.4 o ) at 800 o C which may or may not be enough to have caused a drastic decrease in destruction. It may also be that the phase change is mostly occurring on the surface of the titania where the reactions take place. In other words, the large anatase peak is due to internal anatase crystal formation and the rutile formation is occurring foremost on the surface where it is greatly affecting photocatalysis.

PAGE 51

42 Figure 4-14. XRD analysis of various temperature-cured silica-titania composites. In looking at adsorption of CV at various curing temperatures, Figure 4-15 shows 800 o C with the highest adsorption capacity even though it has the lowest surface area. This may be explained by the cleaner surface (decrease of silanol sights) at this temperature, resulting in the gel displaying a more hydrophobic behavior (Holysz, 1998). This would decrease the competition between water and CV, allowing CV to efficiently interact with the isolated silanols. However, the clean surface can slowly become rehydroxylated when it is in contact with water (Davydov, 2000; Unger et al., 2000). The adsorption ability would then most likely follow the same trend as gels that were treated at lower curing temperatures. Therefore, based upon this and the destruction studies, the optimal curing temperature chosen was 180 o C.

PAGE 52

43 05101520250100200300400500600700800900Curing Temp (oC)Percent Adsorbed 6% TiO2 0% TiO2 Figure 4-15. Effect of curing temperature on adsorption of CV on gels with and without TiO 2 4.2.3 Pore Size Pore size was also investigated for its effects on destruction ability, as well as adsorption ability. Figure 4-16 shows that a large pore size (>140 ) increases the destruction rates for RR. This is consistent with what is found in the literature. Jung and Park (2000) concluded that high porosity and large pore size facilitate the mass transfer of reactants, such as oxygen and reaction intermediates. Thus, the increase in surface area by small pores is not always effective for high photoactivity. The opposite is true for adsorption, because a smaller pore size results in more surface area available for adsorption. Figure 4-17 shows the largest adsorption with a pore size of 30 Adsorption decreases 73% with larger pores (> 30 ), and in this range there is no significant difference in adsorption from 60 versus 140

PAGE 53

44 0102030405060708090100050100150200250300350Pore Size (A)Percent Destruction pH = 4 pH = 8 Figure 4-16. Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm 2 ) versus pore size. 051015202530354045503060140320Pore Size (A)Percent Adsorbed Figure 4-17. Adsorption of CV on gels (6% TiO 2 ) versus pore size. 4.3 Column Studies Since the optimal pore size for adsorption is contrary to the optimal pore size for destruction, it was decided to test two pore sizes in a flow-through regenerative column. These studies utilized 12% optimal titania loaded pellets of 30 and 140 pore sizes. A larger pore size was not analyzed for these studies due to the frailty of the gel at pore

PAGE 54

45 sizes larger than 200 The duplicated column exhaustion curves for each pore size are shown in Figures 4-18 and 4-19. The duplicated runs were done with pellets made from a separate batch. Table 4-2 summarizes the exhaustion results and shows a longer time to breakthrough of the initial concentration for the column of 30 gels. Yet, after the 30 pellets reached exhaustion and the system was regenerated by photocatalysis, its level of adsorption ability decreased after each cycle. This was also visually observed. After one hour of UV exposure the 30 pellets were still slightly purple, while the 140 pellets had regenerated and returned to its initial white state. 00.10.20.30.40.50.60.70.80.91050100150200250300350400450Time (minutes)C/Co Run 1 Run 2 Figure 4-18. Column exhaustion curve for 140 pellet (12% TiO 2 ).

PAGE 55

46 00.10.20.30.40.50.60.70.80.910100200300400500Time (minutes)C/Co Run1 Run2 Figure 4-19. Column exhaustion curve for 30 pellets (12% TiO 2 ). Table 4-2. Exhaustion (C/Co = 1) times (minutes) for column studies. Pellet Pore Size 140 30 Run 1 Exhaustion Times 36 157 26 48 50 64 27 Run 2 Exhaustion Times 28 132 35 128 58 98 65 Average Exhaustion Time 40.6 104.5 Both sets of pellets were examined during a column run after exhaustion of the column and it was observed that the 30 pellets were still white on the inside, revealing that the total surface available for adsorption was not utilized. The 140 pellets were purple throughout their interior, revealing that the available space for adsorption had been exhausted. Thus, the small pore size of the 30 gels inhibits the transit of CV solution through the porous matrix. Therefore, the 140 sized gels exhibit the greatest efficiency in utilizing its surface area for adsorption and its photocatalytic ability to regenerate.

PAGE 56

47 Thus, the 140 pore size was chosen as the optimal pore size for the regenerative system. In order to determine if a shorter regeneration time was achievable while still keeping the same trend as previously observed, the effect of regeneration time with UV light was examined for the 140 column. The time was decreased from 1 hour to 10 minutes and Figure 4-20 shows the comparison of this breakthrough curve with the 1 hour curves. The average breakthrough time was 36 minutes for the 10 minute run versus 46 for the 60 minutes run. Yet, the same trend seems to be exhibited by all 140 breakthrough curves. This increase in breakthrough time with each cycle shows the benefit of using the 140 pellets, for this trend was not observed with the 30 pellets. 00.10.20.30.40.50.60.70.80.91050100150200250300Time (minutes)C/Co Run1 Run2 10 minute cycle Figure 4-20. Effect of regeneration time on 140 column runs.

PAGE 57

CHAPTER 5 SUMMARY AND CONCLUSIONS Silica-titania composites were made using a sol-gel method that allows the doping of titania during gelation. The gelation rate was increased by use of acid catalysts and allowed the composites to be made into pellets with sufficient dispersion of TiO 2 within the silica matrix. Varying the concentration of hydrofluoric acid allowed the manipulation of pore size. Three variables (TiO 2 loading, curing temperature, and pore size) were investigated to determine the best silica-titania composites for a packed column system. These variables were optimized based upon destruction and adsorption studies and the gels were characterized by surface area and XRD analysis. Reactive red dye (RR) was chosen for its non-photolytic behavior in the analysis of photocatalytic ability for the destruction studies and crystal violet dye was chosen because of its cationic nature for the adsorption studies. An optimal loading of 12% TiO 2 was chosen based upon the results of studies involving the destruction of RR and adsorption of CV. At higher loadings, it was observed that the titania is agglomerating during mixing, limiting the effective surface area available for reactions on its surface. Also, there was no trend with BET surface area on a volume basis versus titania loading. In addition, the adsorption studies revealed a plateau in uptake beginning at 12% TiO 2 A curing temperature of 800 o C showed the largest effects (a decrease of over 80% destruction) on photocatalytic ability. At this temperature, the surface area decreased and 48

PAGE 58

49 the phase change of anatase to rutile is evident. This resulted in a dramatic decrease in destruction of RR. Yet, this temperature showed the highest level of adsorption. It was concluded that the clean silica surface and more hydrophobic behavior the silica surface displays at this temperature increased adsorption. This increase in adsorption would decrease over time due to rehydroxylation. The 180 o C gel performed the best for destruction and was chosen as the optimal curing temperature. Concerning pore size, the highest destruction rates were obtained with large pore sizes (> 140 ). The opposite was true for adsorption, which was greater with small pore sizes (30 ). Column studies were done in order to determine the best pore size for a regenerative system. The smaller pore size had the longer time to exhaustion, but did not efficiently regenerate. The small pores inhibited the ability of water or the contaminant to move through the matrix, i.e. the path of least resistance was around the pellet versus through the pellet. On the other hand, the 140 gels achieved complete regeneration after one hour of UV exposure and efficiently utilized their surface area for adsorption. In conclusion, the optimal pore size chosen for the regenerative flow-through column was 140

PAGE 59

APPENDIX A COMMON PREPARATION METHODS OF MIXED AND SUPPORTED OXIDES AS DISCUSSED BY GAO AND WACHS (1999). A.1 Mixed Oxides A.1.1 Sol-Gel Hydrolysis Sol-gel hydrolysis involves the acid hydrolysis and condensation of chemically mixed Tiand Sialkoxides. In summary, a Ti atom acts as a substitute for a Si atom in the silica network. It is the most widely used method due to its capability in controlling the textural and surface properties of the mixed oxides, but the synthesis conditions can greatly affect the homogeneity of the final product. There is a recently developed two-stage hydrolysis procedure that solves this problem and results in the highest homogeneity. Yet, these mixed oxides can only be obtained at low TiO 2 content, less than 15 wt%. At higher Ti content, due to the larger size of the Ti atom versus Si, atom, there is distortion of the SiO 2 network. This increases the OH content on the silica surface and affects the crystal phase of the TiO 2 A.1.2 Coprecipitation (Not Used In This Study) Coprecipitation involves the simultaneous precipitation of SiO 2 and TiO 2 (at high pH) that also includes the formation of linkages with each other. This method also has similar problems to sol-gel hydrolysis. 50

PAGE 60

51 A.2 Supported Oxides A.2.1 Impregnation (Not Used In This Study) Impregnation involves the coating of silica with Ti by mixing with a titanium precursor, such as TiCl 4 or a Ti-alkoxide. The following equation summarizes the process: Si OH + Ti +4 Si O Ti The homogeneity of the supported oxide is dependent on the concentration of surface silanols, pretreatment temperature, and molecular size of precursor. A maximum dispersion of TiO 2 was found to be roughly 4.0 Ti atom/nm 2 Consistent creation and determination of the titania phase for catalytic purposes is a major problem with this method. A.2.2 Chemical Vapor Deposition (not used in this study) The TiO 2 is deposited onto the porous silica surface from a gaseous metal precursor. Consistent coating is difficult to attain from batch to batch and another problem is the reduction in surface area after coating.

PAGE 61

APPENDIX B DYE STRUCTURES B.1 Methylene Blue MW: 319.86 Empirical Formula: C 16 H 18 ClN 3 S B.2 Malachite Green MW: 365 Empirical Formula: C 23 H 25 N 2 Cl 52

PAGE 62

53 B.3 Crystal Violet MW: 407.996 Empirical Formula: C 25 H 30 N 3 Cl B.4 Reactive Red MW: 615.34 Empirical Formula: C 19 H 12 Cl 2 N 6 O 7 S 2

PAGE 63

APPENDIX C TEM AND SEM PICTURES TEM of 4% TiO 2 gel at 20,000x 54

PAGE 64

LIST OF REFERENCES Abrahams, J., Davidson, S. and Morrison, C., Optimization of the photocatalytic properties of titanium dioxide, Journal of Photochemistry, Vol. 29, pp. 353-361, 1985. Al-Ekabi, H. and Serpone, N., Kinetic studies in heterogeneous photocatalysis: photocatalytic degradation of chlorinated phenols in aerated aqueous solutions over TiO 2 supported on a glass matrix, Journal of Physical Chemistry, Vol. 92, pp. 5726-5731, 1988. Anderson, C. and Bard, A., An improved photocatalyst of TiO 2 /SiO 2 prepared by a sol-gel synthesis Journal of Physical Chemistry, Vol. 99, pp. 9882 9885, 1995. Anderson, C. and Bard, A., Improved photocatalytic activity and characterization of mixed TiO 2 /SiO 2 and TiO 2 /Al 2 O 3 materials, Journal of Physical Chemistry B, Vol. 101, pp. 2611-2616, 1997. Blake, D., Bibliography of work on photocatalytic removal of hazardous compounds from water and air, Update Number 1, NREL/TP-473-20300, Golden, CO: National Renewable Energy Laboratory, 1995. Block, S. and Goswami, D., Proceedings of ASME Solar Energy Conference, Kohala Coast, Hawaii, pp. 431-438, 1995. Brinker, C. and Scherer, G., Sol-Gel Science: The Physics and Chemistry of Sol-Gel Processing, Academic Press Inc., San Diego, 1990. Burneau, A. and Gallas, J., Hydroxyl groups on silica surfaces, in: The Surface Properties of Silicas, Legrand, P. (Ed), John Wiley & Sons, Inc., New York, pp. 147-234, 1998. Chen, H., Zahraa, O. and Bouchy, M., Inhibition of the adsorption and photocatalytic degradation of an organic contaminant in an aqueous suspension of TiO 2 by inorganic ions, Journal of Physical Chemistry A: Chemistry, Vol. 108, pp. 37-44, 1997. Chun, H., Yizhong, W. and Hongxiao, T., Preparation and characterization of surface bond-conjugated TiO 2 /SiO 2 and photocatalysis for azo dyes, Applied Catalysis B: Environmental, Vol. 30, pp. 277-285, 2001. 55

PAGE 65

56 Cox, G., The influence of silica structure on reversed-phase retention, Journal of Chromatography, Vol. 656, pp. 353-367, 1993. Crittenden, J., Notthakun, S., Hand, D. and Perram, D., United States Patent 5,182,030, January 26, 1993. Dabrowski, A., Adsorption from theory to practice, Advances in Colloid and Interface Science, Vol. 93, pp. 135-224, 2001. Davydov, V., Adsorption on silica surfaces, in: Adsorption on Silica Surfaces, Papirer, E. (Ed), John Wiley and Sons Ltd., New York, 2000. Degussa Technical Bulletin No. 56, Highly Dispersed Metallic Oxides Produced by the AEROSIL Process, 1990. Elizardo, K., Fighting pollution with hydrogen peroxide, Pollution Engineering, pp. 106-109, September, 1991. El Shafei, G., Silica surface chemical properties, in: Adsorption on Silica Surfaces, Papirer, E. (Ed), John Wiley and Sons Ltd., New York, pp. 35-62, 2000. Emeline, A., Ryabchuk, V. and Serpone, N., Factors affecting the efficiency of a photocatalyzed process in aqueous metal-oxide dispersions: prospect of distinguishing between two kinetic models, Journal of Photochemistry and Photobiology A : Chemistry, Vol. 133, pp. 89-97, 2000. Environmental Protection Agency, Best management practices for pollution prevention in the textiles industry, EPA/625/R-96/004, September, 1996. Environmental Protection Agency, Sector notebook project profile of the textile industry part 2, EPA/310-R-97-009, September, 1997. Gao, X. and Wachs, I., Titania-silica as catalysts: molecular structural characteristics and physico-chemical properties, Catalysis Today, Vol. 51, pp. 233-254, 1999. Gerischer, H. and Heller, A., The role of oxygen in photooxidation of organic molecules on semiconductor particles, Journal of Physical Chemistry, Vol. 95, pp. 5261-5267, 1991. Goswami, D., Klausner, J., Wyness, P., Martin, A., Mathur, G., Schanze, K., Turchi, C. and Marchand, E., Solar photocatalytic treatment of groundwater at Tyndall AFB : field test results, University of Florida, LTFNM/SEECL-9302, 1993. Goworek, J., Nieradka, A. and Dabrowski, A., Adsorption from ternary mixtures on silica gel of different mesoporosity, Fluid Phase Equilibria, Vol. 136, pp. 333-343, 1997.

PAGE 66

57 Goworek, J., Derylo-Marczewska, A. and Borowka, A., Adsorption from the liquid phase on silica gels of various structural heterogeneity, Langmuir, Vol. 15, pp. 6103-6106, 1999. Grabner, G., Li, G., Quint, R., Quint, R. and Getoff, N., Pulse laser-induced oxidation of phenol in acid aqueous TiO 2 , Journal of the Chemical Society Faraday Transactions, Vol. 87, pp. 1097-1101, 1991. Grieken, R., Aguado, J., Lopez-Munoz, M. and Marugan, J., Synthesis of size-controlled silica-supported TiO 2 photocatalysts, Journal of Photochemistry and Photobiology A: Chemistry, Vol. 148, pp. 315-322, 2002. Halmann, M., Photodegradation of Water Pollutants, CRC Press, Boca Raton, 1996. Hanprasopwattana, A., Srinivason, S., Sault, A. and Datye, A., Titania coatings on monodisperse silica spheres: characterization using 2-propanol dehydration and TEM, Langmuir, Vol. 12, pp. 3173-3179, 1996. Hench, L. and West, J., The sol-gel process, Chemical Review, Vol. 90, pp. 33-72, 1990. Hoffman, M., Martin, S., Choi, W. and Bahnemann, D., Environmental application of semiconductor photocatalysis, Chemical Reviews, Vol. 95, pp. 69-96, 1995. Hoigne, J., Formulation and calibration of environmental reaction kinetics; oxidations by aqueous photooxidants as an example, in: Aquatic Chemical Kinetics, Stumm, W. (Ed.), John Wiley & Sons, Inc., New York, pp. 43-70, 1990. Holysz, L., The effect of thermal treatment of silica gel on its surface free energy components, Colloids and Surfaces A: Physicochemical and Engineering Aspects, Vol. 134, pp. 321-329, 1998. Houas, A., Lachheb, H., Ksibi, M., Elaloui, E., Guillard, C. and Hermann, J., Photocatalytic degradation pathway of methylene blue in water, Applied Catalysis B: Environmental, Vol. 31, pp. 145-157, 2001. Iler, R., The Chemistry of Silica, John Wiley & Sons, New York, 1979. Jackson, N., Wang, C., Luo, Z., Schwitzgebel, J., Ekert, J., Brock, J. and Heller, A., Attachment of TiO 2 powders to hollow glass microbeads: activity of the TiO 2 -coated beads in the photoassisted oxidation of ethanol to acetaldehyde, Journal of the Electrochemical Society, Vol. 138, pp. 3660-3664, 1991. Jung, K. and Park, S., Enhanced photoactivity of silica-embedded titania particles prepares by sol-gel process for the decomposition of trichloroethylene, Applied Catalysis B: Environmental, Vol. 25, pp. 249-256, 2000.

PAGE 67

58 Khan, A., Mazyck, D. and Wu, C., TiO 2 coated activated carbon: a regenerative technology for water recovery, 021CES-73, International Conference on Environmental Systems, San Antonio, 2002. Kinniburgh, D. and Jackson, M., Cation adsorption by hydrous metal oxides and clay, in: Adsorption of Inorganics at Solid-Liquid Interfaces, Anderson, M and Rubin, A (Eds.), Ann Arbor Publishers, Ann Arbor, pp. 200-230, 1981. Legrand, P., On the silica edge, in: The Surface Properties of Silicas, Legrand, P. (Ed), John Wiley & Sons, Inc., New York, pp. 130-200, 1998. Lei, L., Chu, H., Hu, X. and Yue, P., Preparation of heterogeneous photocatalysts (TiO 2 /alumina) by metallo-organic chemical vapor deposition, Industrial Engineering Chemical Research, Vol. 38, pp. 3381-3385, 1999. Liu, T. and Cheng, T., Effects of SiO 2 on the catalytic properties of TiO 2 for the incineration of chloroform, Catalysis Today, Vol. 26, pp. 71-77, 1995. Lu, M., Chen, J. and Chang, K., Effect of adsorbents coated with titanium dioxide on the photocatalytic degradation of propoxur, Chemosphere, Vol. 38, pp. 617-627, 1999. Mao, Y., Schoneich, C. and Asmus, K., Identification of organic acids and other intermediates in oxidation in oxidation degradation of chlorinated ethanes on TiO 2 surfaces en route to mineralization: a combined photocatalytic and radiation chemical study, Journal of Physical Chemistry, Vol. 95, pp. 10080 10089, 1991. Mao, Y., Schoneich, C. and Asmus, K., Influence of TiO 2 surface on 1,2-chlorine shift in -chlorine substituted radicals as studied by radiation chemistry and photocatalysis, Journal of Physical Chemistry, Vol. 96, pp. 8522-8529, 1992. Mao, Y., Schoneich, C. and Asmus, K., Radical mediated degradation mechanisms of halogenated organic compounds as studied by photocatalysis at TiO 2 and by radiation chemistry, in: Photocatalytic Purification and Treatment of Water and Air, Ollis, D.F. and Al-Ekabi, H. (Eds.), Elsevier Science Publishers, Amsterdam, pp. 49-66, 1993. Martin, S., Morrison, C. and Hoffman, M., Photochemical mechanism of size-quantized vanadium-doped TiO 2 particles, Journal of Physical Chemistry, Vol. 98, pp. 13695-13704, 1994. Matthews, R., Kinetics of photocatalytic oxidation of organic solutes over titanium dioxide, Journal of Catalysis, Vol. 111, pp. 264-272, 1988. Mauritz, K., Sol-gel chemistry, http:// www.solgel.com August 4, 2002.

PAGE 68

59 Millennium Chemical, Titanium dioxide crystal types, www.millenniumchem.com November 1, 2002. Mills, A. and Sawunyama, P., Photocatalytic degradation of 4-chlorophenol mediated by TiO 2 : a comparative study of the activity of laboratory made and commercial TiO 2 samples, Journal of Photochemistry and Photobiology A: Chemistry, Vol. 84, pp. 305-309, 1994. Nawrocki, J., The silanol group and its role in liquid chromatography, Journal of Chromatography A, Vol. 779, pp. 29-71, 1997. NASA, Requirements definition and design consideration, CTSD-ADV-245, 1998. Okamoto, K., Yamamoto, Y., Tanaka, H., Tanaka, M. and Itaya, A., Heterogeneous photocatalytic decomposition of phenol over TiO 2 powder, Bulletin of the Chemical Society of Japan, Vol. 58, pp. 2015-2022, 1985. Ollis, D., Pelizzetti, E. and Serpone, N., Destruction of water contaminants, Environmental Science and Technology, Vol. 25, pp. 1523-1529, 1991. Palmisano, L. and Sclafani, A., Thermodynamics and kinetics for heterogeneous photocatalytic processes, in: Heterogeneous Photocatalysis, Schiavello, M. (Ed), John Wiley & Sons Ltd, New York, pp. 109-132, 1997. Papirer, E., Adsorption on Silica Surfaces, John Wiley and Sons Ltd., New York, 2000. Persello, J., Surface and interface structure of silicas, in: Adsorption on Silica Surfaces, Papirer, E. (Ed), John Wiley and Sons Ltd., New York, pp. 297-342, 2000. Powers, K., The development and characterization of sol gel substrates for chemical and optical applications, Ph.D. Dissertation, University of Florida, 1998. Sauer, T., Neto, G., Jose, H. and Moreira, R., Kinetics of photocatalytic degradation of reactive dyes in a TiO 2 slurry reactor, Journal of Photochemistry and Photobiology A: Chemistry, Vol. 149, pp. 47-154, 2002. Serpone, N., Brief introductory remarks on heterogeneous photocatalysis, Solar Energy Materials and Solar Cells, Vol. 38, pp. 369-379, 1995. Serpone, N., Borgarello, E., Harris, R., Cahill, P., Borgarello, M. and Pelizzetti, E., Photocatalysis over TiO 2 supported on a glass substrate, Solar Energy Materials, Vol. 14, pp. 121-127, 1986. Sheintuch, M. and Matatov-Meytal, Y., Comparison of catalytic processes with other regeneration methods of activated carbon, Catalysis Today, Vol. 53, pp. 73-80, 1999.

PAGE 69

60 Sheng, H. and Chi, M., Treatment of textile waste effluent by ozonation and chemical coagulation, Water Resources, Vol. 27, pp. 1743-1748, 1993. Suri, R., Liu, J., Hand, D., Crittenden, J., Perram, D. and Mullins, M., Heterogeneous photocatalytic oxidation of hazardous organic contaminants in water, Water Environment Resources, Vol. 65, pp. 665 673, 1993. Takeda, N., Torimoto, T., Sampath, S., Kuwabata, S. and Yoneyama, H., Effect of inert supports for titanium dioxide loading on enhancement of photodecomposition rate of gaseous propionaldehyde, Journal of Physical Chemistry, Vol. 99, pp. 9986-9991, 1995. Tanaka, K., Capule, M. and Hisanaga, T., Effect of crystallinity of TiO 2 on its photocatalytic action, Chemical Physics Letters, Vol. 187, pp. 73-76, 1991. Torimoto, T., Ito, S., Kuwabata, S. and Yoneyama, H., Effects of adsorbents used as supports for titanium dioxide loading photocatalytic degradation of propyzamide, Environmental Science and Technology, Vol. 30, pp. 1275-1281, 1996. Turchi, C. and Ollis, D., Photocatalytic degradation of organic water contaminant: mechanisms involving hydroxyl radical attack, Journal of Catalysis, Vol. 122, pp. 178-192, 1990. Uchida, H., Itoh, S. and Yoneyama, H., Photocatalytic decomposition of propyzamide using TiO 2 supported on activated carbon, Chemistry Letters, pp. 1995-1998, 1993. Unger, K., Porous Silica, Elsevier, 1979. Unger, K.,Kumar, D., Ehwald, V. and Grossman, F., Adsorption on silcia surfaces from solution and its impact on chromatographic separation techniques, in: Adsorption on Silica Surfaces, Papirer, E. (Ed), John Wiley and Sons Ltd., New York, pp. 565-594, 2000. US Geological Survey Minerals Information, Titanium statistical compendium, http://minerals.usgs.gov/minerals/pubs/commodity/titanium/stat/ August 1, 2002. Voinov, M. and Augustynski, J., Introduction to the physics of semiconductor photocatalysts, in: Heterogeneous Photocatalysis, Schiavello, M. (Ed.), John Wiley & Sons Ltd., New York, pp. 1-34, 1997. Wu, T., Liu, G., Zhao, J., Hidaka, H. and Serpone, N., Evidence for H 2 O 2 generation during the TiO 2 assisted photodegradation of dyes in aqueous dispersions under visible light illumination, Journal of Physical Chemistry B, Vol. 103, pp. 4862-4867, 1999.

PAGE 70

61 Yamazaki, S., Matsunaga, S. and Hori, K., Photocatalytic degradation of trichloroethylene in water using TiO 2 pellets, Water Resources, Vol. 35, pp. 1022-1028, 2001. Yoneyama, H. and Torimoto, T., Titanium dioxide/adsorbent hybrid photocatalysts for photodestruction of organic substances of dilute concentrations, Catalysis Today, Vol. 58, pp. 133-140, 2000. Zhang, Y., Crittenden, J., Hand, D. and Perram, D., Fixed-bed photocatalysts for solar decontamination of water, Environmental Science and Technology, Vol. 28, pp. 435-442, 1994. Zhang, Y., Weidenkaff, A. and Reller, A., Mesoporous structure and phase transition of nanocrystalline TiO 2 , Materials Letters, Vol. 3429, 2001.

PAGE 71

BIOGRAPHICAL SKETCH Danielle Julia Eller Londeree was born in June of 1979 in Deerfield Beach, Florida. Her dad graduated from UF with an agricultural engineering degree and took over the family business of manufacturing large-scale water pumps. Her brothers also graduated from UF and followed in his footsteps, both receiving engineering degrees and working for the family business. Their encouragement is what made Danielle want to be an engineer, and following high school graduation from Zion Lutheran Christian School in 1997, she pursued an engineering degree from UF. After getting married to Donald Londeree in August of 2001, she received her B.S. in Environmental Engineering in December of 2001. Then her thirst for knowledge and her desire to contribute to the family business led her to pursue graduate school, focusing on water quality. In the future, she hopes to use the knowledge that she has obtained in graduate school to increase the scope of the family business and assist in solving the worlds water problems, on both a quantity and quality level. 62


Permanent Link: http://ufdc.ufl.edu/UFE0000564/00001

Material Information

Title: Silica-Titania Composites for Water Treatment
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000564:00001

Permanent Link: http://ufdc.ufl.edu/UFE0000564/00001

Material Information

Title: Silica-Titania Composites for Water Treatment
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000564:00001


This item has the following downloads:


Full Text












SILICA-TITANIA COMPOSITES FOR WATER TREATMENT


By

DANIELLE JULIA LONDEREE















A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ENGINEERING

UNIVERSITY OF FLORIDA


2002




























Copyright 2002

by

Danielle Julia Londeree















ACKNOWLEDGMENTS

First of all, there is no way I could have accomplished this research without the

help of my Lord and Savior. He has not only given me this opportunity, but also gave me

peace and patience during the times of frustration. Second, without the help of my

husband who kept me calm when my computer crashed several times, this work would

not be done. He also stayed up with me many nights while I was writing and constantly

gave me love and support.

I would like to thank Dr. David Mazyck for giving me the opportunity to work for

him and learn from him. He has challenged me in many ways and I feel that I have

learned more in the past two years of graduate school than the four years of undergrad. I

thank him for patience and for purchasing the spectrophotometer. I thank Dr. Powers for

allowing me to do an independent study with him. His excitement for research is

contagious and motivational. I would also like to thank Dr. Paul Chadik and Dr. Chang-

Yu Wu for giving advice during the research meetings and helping to provide answers for

my many questions.

I thank the students in my research group and for those who have helped me get

good data, specifically Matt Tennant (my mentor), Ameena Khan, Jennifer Hobbs,

Christina Termaath, Jon Powell, Noworat Coowanitwong, and Julee MacKenzie.
















TABLE OF CONTENTS
page

A C K N O W L E D G M E N T S ......... .................................................................................... iii

LIST OF TABLES .............................. ......... .... ..... .. .... ....... ....... vi

L IST O F F IG U R E S .... ...... ................................................ .. .. ..... .............. vii

ABSTRACT ........ .............. ............. ...... .......... .......... ix

CHAPTER

1 INTRODUCTION .................. .................................... ............ ..................

2 LITERA TURE REVIEW ............................................................... ............4

2.1 H eterogeneous Photocatalysis ........................... ....... .................................. 4
2.2 Titanium D dioxide .................. ................................ ....... .. .......... .. 8
2 .3 R actor D design ..................................................... 11
2.4 C analyst Supports ....................................................... .... ... ........ .. 13
2.5 Silica-Titania Composites ......... ................... ......... .. .... .............. 16
2.5.1 Sol-Gel Chem istry......................... ...... .. ....... ........ ....... 18
2.5.2 Surface Chemistry and Adsorption Characteristics .................................... 22

3 M ATERIALS AND M ETHODS ............................................................................25

3.1 Silica-Titania Com posites ......... .. ................. ................. ... ....................... 25
3.2 Batch Performance Studies .......... ........ ................. ... .............. 26
3.3 Colum n Perform ance Studies ........................................... .......................... 28

4 RESULTS AND DISCUSSION ............................................................................30

4.1 D ye Com prison ......... ..................... ... ........... ......... ........... 30
4.2 Optimization of Gels ........... ...... .. ..... .... .. .............. .......... 34
4.2.1 Titania Loading .................................... ........ ... .. .......... .. 34
4.2.2 Curing Temperature ........... .... ............... .................... 40
4.2.3 Pore Size ........ ....................... ...................... ........... 43
4.3 Column Studies ........................................................................ .... ......... .................. 44









5 SUM M ARY AN D CON CLU SION S ...........................................................................48

APPENDIX

A COMMON PREPARATION METHODS OF MIXED AND SUPPORTED
OXIDES AS DISCUSSED BY GAO AND WACHS (1999). ....................................50

A 1 M ixed O xides ..................................................... 50
A 1.1 Sol-G el H ydrolysis.................................................. ......................... 50
A 1.2 C precipitation ............................................................ .. ........ ..... 50
A .2 Supported Oxides ............. .................. ......... .. ................ 51
A .2.1 Im pregnation .......... .. ........ ........................ ...... ............ ..... 51
A.2.2 Chemical Vapor Deposition ......................................................... 51

B D Y E ST R U C T U R E S .......................................................................... ....................52

B 1 M ethylene B lue ....................................... ............ ........... ....... ....... 52
B .2 M alachite G reen .................. ........................................ ............ .. 52
B .3 Crystal Violet ......................................... 53
B .4 R active R ed ........................................ 53

C T E M P IC T U R E S ..................................................................................................... 54

LIST OF REFERENCES ................................................................. ...........55

BIO GR A PH ICA L SK ETCH ........................................................................... 62
















LIST OF TABLES

Table p

2-1 Oxidation power for various species relative to chlorine. ......................................6

2-2 Characteristics of Degussa P25 Titanium Dioxide............................. ...............

4-1 Peak absorbance w avelengths ................................................................... ....... 30

4-2 Exhaustion (C/Co = 1) times (minutes) for column studies. ................................46
















LIST OF FIGURES


Figure page

2-1 Schematic of oxidation and reduction occurring on the semiconductor surface. ....5

2-2 Example of hydroxyl radical attack a) on tetrachloroethylene by self-addition b)
on 1,1,1-trichloroethane by hydrogen abstraction (notice chlorine shift in the
ra d ic al) .............................................................................. 8

2-3 Diagram of anatase and rutile crystal structures..............................................10

2-4 Simplified diagram of a catalyst support. .................................... ...............13

2-5 Linkages of SiO 2 tetrahedras........................................................................ .... 17

2-6 Structural changes during the sol-gel process............................................ 20

2-7 Silanol groups on the silica surface ............................................ ............... 23

2-8 Representation of dehydroxylation............................................................23

3-1 C olum n setup. .......................................................................29

4-1 Photolysis of dyes as a function of UV flux for 2 hours of exposure .................31

4-2 Photolysis of MB and MG (2 hours UV exposure at 0.45 mW/cm2) as a function
of initial dye concentration. ....................................................................... ......... 32

4-3 Photolysis of RR (2 hours UV exposure) as a function of initial dye concentration
an d flu x ......................................................... ................ 3 2

4-4 Change in absorbance of RR at 538 nm as a function of pH .............................33

4-5 Spectrophotometer scan of 10 mg/L RR as a function of pH, before and after UV
exposure. ............................................................................34

4-6 Destruction ofRR (pH 7.6) after 2 hours UV exposure versus TiO2 loading as a
function of flux. ........... .. ............... .... .......... ................. ...... .....35

4-7 Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm2) versus
TiO2 loading as a function of pH ................. ....... ....................................... 36









4-8 Destruction of RR (10 mg/L) after 1 hour UV with TiO2 slurry at various pHs ..37

4-9 Effects of bicarbonate at pH 7.6 versus TiO2 loading ............... .....................38

4-10 BET surface area and pore size versus titania loading. .......................................39

4-11 Adsorption of CV (10 mg/L) versus TiO2 loading after 24 hours of mixing........39

4-12 Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm2) versus
curing temperature as a function of pH.......... ............. .......................40

4-13 Effect of curing temperature on surface area....................................... ........... 41

4-14 XRD analysis of various temperature-cured silica-titania composites.................42

4-15 Effect of curing temperature on adsorption of CV on gels with and without
T iO 2 ................................................................................ 4 3

4-16 Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm2) versus
pore size. ...........................................................................44

4-17 Adsorption of CV on gels (6% TiO2) versus pore size. ......................................44

4-18 Column exhaustion curve for 140 A pellet (12% TiO2).....................................45

4-19 Column exhaustion curve for 30 A pellets (12% Ti2). .....................................46

4-20 Effect of regeneration time on 140 A column runs..............................................47















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Engineering

SILICA-TITANIA COMPOSITES FOR WATER TREATMENT
By

Danielle Julia Londeree

December 2002

Chair: David Mazyck
Major Department: Environmental Engineering Sciences

Heterogeneous photocatalysis can be used for mineralizing organic dyes found in

the effluent of textile dyeing operations. Incorporating the catalyst TiO2 into an adsorbent

material, such as silica, has many advantages over using a TiO2 slurry for water

purification. The goal of this research was twofold: 1) to produce a silica-titania

composite using a sol-gel hydrolysis method that dopes the catalyst into the silica matrix

during gelation and 2) to optimize the titania loading, curing temperature, and pore size

of the material based upon maximizing its destruction and adsorption ability for textile

dyes. The optimal titania loading found for reactive red dye was 30 wt% TiO2. The

optimal pore size and curing temperature found were 140 A and 180C, respectively.

These composites were also made into small cylindrical pellets and tested in a flow-

through column to be used in a regenerative system. It was found that diffusivity was

very important to efficiently regenerate the column using photocatalysis.














CHAPTER 1
INTRODUCTION

The textile industry's dyeing operations are of primary environmental concern,

due to the variety of toxic chemicals (mostly solvents, surfactants, and acids) used in the

dyeing of garments. According to the Environmental Protection Agency [EPA] (1997),

chemicals such as methanol, methyl ethyl ketone, trichloroethylene, trimethylbenzene,

and dyes such as basic green 4 and dispersive yellow 3 are released into the environment,

either in the air phase or water phase. These contaminants can create large amounts of

BOD (biochemical oxygen demand) and COD (chemical oxygen demand), as well as

cause eutrophication and aquatic toxicity in effluent streams and natural water bodies.

Color itself is increasingly being regulated, as there have been links to aquatic toxicity

and lower DO (dissolved oxygen) values in receiving streams (EPA, 1996).

There are several chemical and physical methods of removal available for pre-

release treatment, such as adsorption by activated carbon, biodegradation, and ozonation.

Yet, dyes can be resistant to biodegradation and ozonation due to the large number of

aromatic rings found in modern dyes (Sheng and Chi, 1993). Therefore, advanced

oxidation processes are being considered as an emerging technology to handle large

volumes of textile wastewater (Sauer et al., 2002; Wu et al., 1999). Of these processes,

heterogeneous photocatalysis can be used for mineralization (complete conversion to CO2

and H20) of several classes of organic compounds (including dyes), instead of

transferring the pollutants from one phase to another (associated with carbon use).









Dyes can also be used as a surrogate for more toxic organic compounds in order

to measure the photocatalytic ability of a system. For instance, the structure of methylene

blue is similar to chloropromazine, an antipsychotic drug, and methylene blue destruction

could be used to estimate the destruction of this drug. In this research, dyes were used to

represent hard to degrade aromatic organic compounds in order to measure the

performance of the silica-titania composites for water treatment.

In addition, NASA, with its construction of the International Space Station and

plans for a Mars mission, requires a water recovery system that will minimize chemical

additions and energy and provide water at a potable level. The water source to be treated

will come mostly from grey water, including urine, humidity condensate, and wash water.

NASA Document CTSD-ADV-245 also lists numerous organic compounds (many

aromatic) that may be present in water in space flight (1998). The dyes used in this study

are surrogates for those compounds that may be present in space flight for NASA.

There were two main objectives of this research. First, to develop a silica-titania

composite by utilizing a sol-gel hydrolysis method and doping a highly efficient TiO2

catalyst into the silica matrix during gelation. These composites could also be created into

any shape and size, such as small cylindrical pellets to be used for a packed column.

Second, the titania loading, curing temperature, and pore size of the composites were to

be optimized based upon maximizing its destruction and adsorption ability for textile

dyes. The result would be a combined adsorbent/photocatalyst that efficiently adsorbs

contaminants into its pores for oxidation where the hydroxyl radicals would

predominantly exist. It was hypothesized that a packed column of these composites made










into small pellets could create a regenerable system that could be used over and over

again.














CHAPTER 2
LITERATURE REVIEW

2.1 Heterogeneous Photocatalysis

Advanced oxidation processes (AOPs) are considered to be an alternative water

treatment technology for removing harmful compounds and microbes from public water

supply, comparable with traditional practices such as activated carbon adsorption, air

stripping, and chlorine disinfection. AOPs can not only disinfect the water from virulent

microbes, but can also oxidize toxic heavy metals, organic pesticides and solvents,

chlorinated compounds, and inorganic chemicals from the water without producing a

waste stream unsuitable for disposal or consumption. For example, these processes can be

used for chemical spill cleanup, treatment of industrial effluents, and wastewater

treatment (Turchi and Ollis, 1990).

AOPs involve the generation of highly reactive oxidative species produced by

several pathways, including reactions with H202 and 03 with or without ultraviolet (UV)

irradiation and heterogeneous photocatalysis. Heterogeneous photocatalysis entails the

use of a solid photocatalyst (usually a semiconductor) in contact with either a liquid or

gas, while homogeneous photocatalysis uses a catalyst of the same phase as the

contaminated media. While UV use or an oxidant alone may yield partial degradation of

a compound, the combined use of UV with an oxidant (with or without a photocatalyst)

has been shown to yield complete mineralization of organic contaminants to carbon

dioxide (Ollis et al., 1991).









The reactants in heterogeneous photocatalysis in water are naturally occurring

species (such as oxygen, water molecules, and hydroxide ions (OH-)) as compared to

chemically supplied reactants (such as hydrogen peroxide and ozone) required for other

AOPs. When the surface of the photocatalyst absorbs a specific amount of energy, an

electron from the valence band jumps to the conduction band, thereby leaving a positive

"hole" in the valence band. These electrons (e-) and holes (h ) may either recombine,

releasing heat, or migrate to the surface of the catalyst. The reactive oxidative species are

then generated from reactions with the electrons or holes on the surface (see Figure 2-1).




Acceptor-
Reduction





S}COxidation B
hv Release of heat





Do onorr
Donor+


Figure 2-1. Schematic of oxidation and reduction occurring on the semiconductor
surface.

The primary oxidizing species in water is the hydroxyl radical (*OH) (Turchi and

Ollis, 1990). It is the most reactive of the oxidizing agents (see Table 2-1) and can be

generated via several pathways:

H20 + h+ *OH + H+ (1)

OH- + h *OH (2)









02 + 2H+ + 2e- H202 (3a)

H202 + e *OH + OH- (3b)

H202 hL- 2*OH (3c)

02 + e 02- (4a)

02- + H202 *OH + OH- + 02 (4b)

As can be seen from equations 1 4b, molecular oxygen plays a dual role in

heterogeneous photocatalysis by 1) creating an oxidative species, the superoxide radical

(02-), and 2) acting as an electron acceptor to prevent the electrons and holes from

recombining. These functions are very important in establishing an efficient

photocatalytic system, for it has been shown that with an increase in oxygen

concentration, there is a subsequent increase in the reaction rate of contaminant

degradation (Gerischer and Heller, 1991). Similarly, hydrogen peroxide plays a parallel

role.

Table 2-1. Oxidation power for various species relative to chlorine.
Species Relative Oxidation Power

Hydroxyl radical 2.06

Singlet oxygen radical 1.78

Hydrogen peroxide 1.31

Perhydroxyl radical 1.25

Chlorine dioxide 1.15

Chlorine 1.00

Source: Elizardo, 1991.
Adsorption of compounds to the surface of the catalyst is also important in an

efficient photocatalytic system. For instance, competition with other compounds can be









deleterious to the adsorption of specific contaminants on the catalyst surface. Chen et al.

(1997) showed that the effect of inhibition due to competitive adsorption will hinder the

degradation rate, especially in the presence of bicarbonate and phosphate. In addition,

carbonate ions will also scavenge the hydroxyl radicals, thereby decreasing the

destruction rate (Chen et al., 1997).

Another factor affecting adsorption is pH and the effect of pH is highly compound

specific. For example, in the photocatalysis of textile dyes, a lower pH enhanced

adsorption of Orange II (containing an anionic sulfonate group) to the catalyst surface,

thereby increasing the rate of mineralization (Wu et al., 1999), while an alkaline pH

increased adsorption and subsequent degradation of methylene blue (a cationic dye)

(Houas et al., 2001). Sauer et al. (2000) also found similar relationships between

adsorption and destruction of reactive dyes.

The photocatalytic mineralization of an organic compound begins with either

reactions with hydroxyl radicals (or other less significant oxidizing species) or by direct

oxidation with the holes on the surface of the catalyst. A hydroxyl radical "attacks" an

organic compound by either removing an available hydrogen atom to form water or

adding itself to any unsaturated carbon bonds (Grabner et al., 1991; Halmann, 1996;

Hoffman et al., 1995; Hoigne, 1990; Mao et al., 1991, 1992, 1993). Mao et al. (1991,

1992, 1993) found that the rate of oxidation of organic correlated with the C-H bond

strengths, proving that H atom abstraction by *OH may be a rate-limiting step.

Consequently, from these primary reactions, an organic radical is formed (see Figure 2-

2). The subsequent radical transformations and radical to radical interactions lead to an









array of intermediate products, eventually resulting in CO2 and H20. For chlorinated

compounds, HC1 is also produced.


Cl ci Ci CI
C Q= + *OH C C -OH
a) CI CI Cl

H Cl H Cl
I I I I
H-C-C-CI + *OH H-C-C. + H20
b) H CI ci Cl




Figure 2-2. Example of hydroxyl radical attack a) on tetrachloroethylene by self-addition
b) on 1,1,1-trichloroethane by hydrogen abstraction (notice chlorine shift in
the radical) (adapted from Mao et al., 1991 and 1992).

There are several literature reviews available on the mineralization of particular

classes of organic chemicals (Hoffman et al., 1995; Matthews, 1988; Ollis et al., 1991;

Serpone, 1995). Blake (1995) and Halmann (1996) have compiled extensive surveys of

this research and have also included sections on typical reactor designs and methods for

catalyst improvement. Research has also been focused on the determination of the

intermediate degradation pathways, for Mao et al. (1991 and 1992) has shown that the

conditions of a system (pH, oxygen concentration, competition with other compounds)

can affect the pathway a single compound can take towards complete mineralization.

2.2 Titanium Dioxide

Metal oxides are the most popular catalysts used for potable water and air

treatment; of these, TiO2 and ZnO have been the most researched and are considered the

most efficient. Of these two, TiO2 tends to be favored due to its stability in extreme

conditions (low or high pH), insolubility, heat resistance, non-toxicity, and low cost. TiO2









is also photostable as compared to ZnO which can undergo photocorrosion, decreasing

the effective lifetime of the catalyst (Okamoto et al., 1985).

Titanium is the ninth most abundant element in the earth's crust and occurs in

nature only in combination with other elements, such as with oxygen to form TiO2 (US

Geological Survey, 2002). Naturally occurring forms of titanium dioxide are usually

combined with iron (FeTiO3 or FeO-TiO2) and contain other impurities. Therefore, the

TiO2 (titania) used as a photocatalyst is manufactured from synthetic means.

Of the commercially available titanium dioxides, Degussa's P25 seems to be the

most researched for its photocatalytic ability in water (Halmann, 1996). Table 2-2 shows

a few specific characteristics of P25. The product is produced from high-temperature

flame hydrolysis of TiC14 in the presence of oxygen and hydrogen and is then steam

treated to remove HC1, resulting in a 70:30 anatase to rutile phase TiO2 (Degussa

Technical Bulletin, 1990). The phase of TiO2 is important since the anatase crystal form

is considered the more photoreactive of the two (Tanaka et al., 1991). Figure 2-3 shows

the crystal structures of the two phases. A difference in reactivity of the phases is

attributed to 1) the more positive conduction band of the rutile phase, hindering

molecular oxygen to act as an electron acceptor, and 2) the difference in surface

properties between the two phases (Tanaka et al., 1991).

Table 2-2. Characteristics of Degussa P25 Titanium Dioxide
BET Surface Area 50 m2/g
Average Primary Particle Size 21 nm
Band Gap: anatase, rutile 3.29 eV, 3.05 eV
Point of Zero Charge pH 6.0
Source: adapted from Degussa Technical Bulletin (1990)













iP 1


A 11 skf ;^ !













Crystal morphology, as well as a variety of other factors, can influence the
k Y T4.5929A -9

I 4 -- C OXYGEN



Anatast Ru ile
Figure 2-3. Diagram of( anatase and rutle crystal structures (Millennium Chemical,
2002).

Crystal morphology, as well as a variety of other factors, can influence the

activity of a catalyst. Surface area is known to affect the adsorption ability of the catalyst,

as well as the available contact area for oxidation and reduction reactions. It would make

sense that a catalyst with a large surface area would have an increased rate of reaction,

but Tanaka et al. (1991) and Mills et al. (1994) showed that the destruction efficiencies of

several different titanium dioxide samples were independent of surface area (within the

range of 2.75 m2/g to 177 m2/g). Therefore, variables besides surface area may have a

more influential role in the activity of a catalyst. These variables are crystal morphology,

as well as crystallite size, prevention of electron-hole recombinations, zeta potential, and

band gap energy (Okamoto et al., 1985; Palmisano and Sclafani, 1997; Suri et al., 1993;

Tanaka et al, 1991; Zhang et al., 2001).

Tanaka et al. (1991) showed that the photocatalytic ability of a TiO2 catalyst is

foremost affected by the anatase content and secondarily affected by the size of the









crystal. For example, a photocatalyst containing larger crystals is more efficient than one

with smaller crystals, even when both contain equal percentages of anatase. This could be

due to the larger migration distance of the holes and electrons to the surface of the

catalyst, thereby decreasing the possibility of recombination (Tanaka et al., 1991).

Modifications in the catalyst surface have also been investigated for prevention of

electron-hole recombination. The addition of platinum and other transition metals have

been successfully arrayed on the titanium dioxide surface (Abrahams et al., 1985;

Okamoto et al., 1985; Martin et al., 1994; Suri et al., 1993). These metal additions have

an optimum at low weight percentages (less than 5%), above which the metal actually

hinders the photocatalytic ability.

The isoelectric point or point of zero charge (PZC) represents the pH at which an

immersed solid oxide would have zero net charge, resulting in electrically equivalent

concentrations of positive and negative complexes on the surface. At a pH above the

PZC, interactions with cationic electron donors and acceptors will be favored; while

anionic electron donors and acceptors will be favored at pH below the PZC.

The band gap is equal to the amount of energy required to activate the surface of

the catalyst. A band gap of 3.29eV corresponds to a wavelength of 378 nm. Therefore,

wavelengths below 378 nm (such as those emitted by UV light) would have the required

energy to activate the anatase phase of TiO2.

2.3 Reactor Design

A problem in the practical application of heterogeneous photocatalysis of

environmental pollutants is the design of a reactor that will maximize photocatalytic

efficiency while utilizing the least amount of energy. The simplest reactor configuration

uses an aqueous suspension or slurry of the photocatalyst. Most photocatalysis studies









use a slurry system in preliminary testing to establish the feasibility of pollutant

mineralization or microbe inactivation. For these systems, an optimum weight of TiO2

per volume of solution (wt/vol) was found that yielded the most degradation for specific

types of compounds at a specific light intensity. For instance, Goswami et al. (1993)

found 0.1% (wt/vol) TiO2 to be the optimum loading for most hydrocarbons and Block

and Goswami (1995) found 0.01% (wt/vol) as the best for microbial destruction.

Although slurry systems are considered to be the most photocatalytically efficient,

there is a dilemma to be addressed in the design of the reactor. Due to the extremely

small size of the particles (between 0.1 30 nm, depending on the manufacturing

source), there is a difficulty in recovering the catalyst from the purified water. This

problem has directed research to investigate the use of catalyst supports.

There are many research groups that have examined the immobilization of

photocatalysts on nonporous glass surfaces as a way of providing a backbone of support

for the particles, while also allowing the penetration of light to activate the catalyst

surface. Successful immobilization has been done on glass beads (Jackson et al., 1991;

Serpone et al., 1986; Zhang et al., 1994) and on the inside surface of glass tubes (Al-

Ekabi and Serpone, 1988; Matthews, 1988). The oxidation rates were usually lower with

the immobilized catalysts than with free suspensions. The determinate factor was

assumed to be the result of mass transfer limitations, for it was observed that as the flow

rate increased, resulting in more efficient mixing, the oxidation rate likewise increased.

While these designs may have certain advantages, it is hypothesized that an

efficient support for a catalyst would be a material with adsorption capability that would

bring the contaminants into close contact with the catalyst surface (see Figure 2-4) where









the hydroxyl radicals predominantly exist (Turchi and Ollis, 1990). Indeed, there has

been a reported increase in destruction efficiency when the reactants are brought into

close contact with the surface of activated TiO2 particles. This increase could be due to

the greater concentration of the substrate around the catalyst surface due to adsorption

interactions, for it has been proven that the highest efficiency of photocatalytic

degradation of organic pollutants is observed at relatively high concentrations of reagents

(Emeline et al., 2000).


H20


OH' .


Figure 2-4. Simplified diagram of a catalyst support.

2.4 Catalyst Supports

Adsorbents researched as possible catalyst supports include activated carbon,

silica gel, mordenite, alumina, zeolite, and aerogels (Takeda et al., 1995; Torimoto et al.,

1996; Yoneyama and Torimoto, 2000). The highest decomposition rate for propyzamide

(herbicide) was obtained with supports having medium adsorption strengths, such as

silica and mordenite (Torimoto et al., 1996). With supports of too low zeolitee) and too

high (activated carbon) adsorption strengths, no significant effect on the rate of

destruction was found.


0 :ate^^


AdsorhenlL









Yet, research has focused on the benefits of using activated carbon (AC) as a

catalyst support versus other supports. Uchida et al. (1993) and Lu et al. (1999) found

that while the oxidation rates of a specific compound were lower with the TiO2 loaded

AC, the complete mineralization rates were greater than with a plain slurry of TiO2 or

any other catalyst loaded support. They concluded that the lower oxidation rate can be

attributed to the carbon blocking some of the photons of light from reaching the catalyst

surface and the higher mineralization rate can be attributed to the high adsorptive ability

of the carbon to adsorb the intermediate compounds that are being formed.

An additional benefit of TiO2 coated carbon is its potential for in situ AC

regeneration. Many water utilities use granular activated carbon (GAC) for the removal

of organic compounds. However, after the absorbent is spent, removal of the carbon from

the system is required for reactivation, disposal in a hazardous waste landfill, or

incineration. While on-site thermal reactivation is an option for some utilities, it is not the

most economical option for many. Therefore, research has been looking at optimizing

both TiO2-coated carbon's adsorption capability and its regeneration efficiency for the

removal and subsequent oxidation of organic pollutants adsorbed on the spent carbon

surface (Crittenden et al., 1993; Khan et al., 2002; Sheintuch and Matatov-Meytal, 1999).

While activated carbon has some benefits as a catalyst support, its use requires a

reactor that efficiently exposes the catalyst surface to the photons of light. Since this may

require mixing or fluidizing the particles, the inherent attrition can cause the catalyst to

detach itself from the carbon (Lu et al., 1999). An additional problem is that the surface

chemistry of a carbon may hinder effective coating (Khan et al, 2002). The various

methods of activating carbon can create different functional groups on the surface, thus









affecting the properties a specific carbon will display. This can greatly affect the

immobilization of a catalyst to the carbon surface.

Silica gels, on the other hand, have many advantages over activated carbon as a

catalyst support. The transparency of silica allows the penetration of photons to the

catalyst surface. This is extremely beneficial and allows for a fixed-bed reactor design

that can be highly efficient in relation to input energy. Silica also has high mechanical

strength, thermal stability, and can be synthetically formed into any shape, such as

cylindrical pellets (Yamazaki et al., 2001) or fibers (Brinker and Scherer, 1990). In

addition, silica is an adsorbent and is commonly used in chromatography columns for

adsorption and the resulting separation of compounds in an aqueous sample.

Furthermore, TiO2 and SiO2 can be chemically combined, enabling the formation of

highly efficient photocatalysts. These TiO2 SiO2 photocatalysts allow the placement of

the catalyst on both external surfaces and internal surfaces within the porous silica matrix

where pollutants are adsorbed.

Anderson and Bard (1995, 1997) found a strong synergy between silica and

titania as a combined oxide. For substances that are easily adsorbed to the silica surface, a

higher decomposition rate was found with TiO2 SiO2 composites versus a plain TiO2

slurry. The presence of an adsorbent was considered to promote efficiency by increasing

the concentration of the substrate near the TiO2 sites relative to the solution

concentration. But for substrates that are not readily adsorbed to the surface of silica, the

degradation rates were lower than compared to a slurry of pure TiO2. However, when

these rates were normalized to the TiO2 content of the material, the TiO2-SiO2 material









demonstrated a more efficient use of the TiO2 sites than for TiO2 alone. The highest

initial degradation rate was found with a 30/70 wt% TiO2-SiO2 composite.

Jung and Park (2000) also found 30 wt% titania (with gels prepared in a similar

process to Anderson and Bard) to be the optimum in their studies for the mineralization

of trichloroethylene. In addition, they concluded that the high porosity and large pore size

of the silica facilitated the mass transfer of reactants, resulting in a higher rate of

degradation than a plain slurry of Degussa P25.

Chun et al. (2001) used a different preparation method than Jung and Park (2000)

and Anderson and Bard (1995) and still found 30 wt% TiO2 to be an optimum. Chun et

al. additionally showed, using two organic compounds with differing characteristics, that

there is a strong correlation between adsorption of the compound on the mixed oxide

surface and the destruction rate of that compound. In conclusion, silica-titania composites

can be efficiently utilized in heterogeneous photocatalysis systems for the adsorption and

subsequent destruction of unwanted compounds in an aqueous solution. In addition, a

dosage of 30 wt% (equal to 12% on a weight per volume of silica precursor) was also

found to be an optimum loading in this research (see Chapter 4) for the silica-titanium

composites created using a sol-gel doping method (see Chapter 3).

2.5 Silica-Titania Composites

Silica is the most abundant oxide on the earth, yet despite this abundance, silica is

predominantly made by synthetic means for its use in technological applications. Since

synthetic silicas have a higher surface area than naturally occurring forms of silica, the

synthetic silicas provide the best adsorption and catalyst support structure for

heterogeneous photocatalysis. Although silica has a simple chemical formula (SiO2), it

can exist in a variety of forms, each with its own structural characteristics, as well as









chemical and physical properties. In general, SiO2 is a SiO4 tetrahedra, where each silica

atom is bonded to four oxygen atoms and each oxygen atom is bound to two silica atoms.

HO OH
HO-Si-OH HO-Si-OH
I I
HO 0 OH
I I 1



HO-S--OH HO-Si-OH



Figure 2-5. Linkages of SiO2 tetrahedras (Hench and West, 1990).

There has been much research on the different preparation methods for creating a

solid SiO2-TiO2 material with photocatalytic ability. The characteristics (e.g., pore size,

surface charge, mechanical strength, and adsorption sites) of the final product are

dependent on the synthesis conditions and the type of interaction between TiO2 and SiO2.

There are two forms of interaction: physical forces of attraction (such as Van der Waals

forces) and chemical bonding (creation of a Ti O Si bond). The physically supported

TiO2 on SiO2 preparation methods have been the least researched, while the chemically

bonded TiO2 on SiO2 methods (also called mixed oxides) have been thoroughly

investigated (Gao and Wachs, 1999).

The most widely used methods of preparation (for details on these individual

methods, see Appendix A) for creating mixed oxides are sol-gel hydrolysis and

coprecipitation. Chemical vapor deposition, precipitation, and impregnation are the

methods commonly used for creating supported oxides, yet there is evidence of Ti-O-Si

bonds attaching the titania to the silica surface (Chun et al., 2001). A problem with

supported oxides is the inconsistency in the coating from one batch to the next and the









nonuniformity of the coating due to the sensitivity of the procedure to its conditions

(Hanprasopwattana et al., 1996; Lei et al., 1999). Another problem with any of these

methods is the assurance of the formation of an anatase phase of TiO2. Even though X-

ray diffraction (XRD) can be used to determine the phase of TiO2, the question remains

whether the preparation method has formed the most efficient semiconductor particle, as

compared to commercially available photocatalysts. For example, Liu and Cheng (1995)

produced a TiO2-SiO2 mixed oxide using coprecipitation and found that the SiO2 inhibits

the growth of TiO2 crystals, resulting in amorphous TiO2 within the silica matrix.

A method of titania incorporation in silica has been devised to overcome both of

these potential problems. This method uses sol-gel hydrolysis to create the silica matrix,

then during gelation the solution is doped with the commercially available highly

efficient Degussa P25 TiO2. The silica network may either form around the titania

particle and/or form a bond with the titania to secure its position in the matrix.

2.5.1 Sol-Gel Chemistry

A sol is a colloidal suspension of particles, while the term gel refers to the semi-

rigid material formed when the colloidal particles link together in a liquid to form a

network. Sol-gel precursors consist of metals or metalloids surrounded by various

ligands, of which the most widely used are metal alkoxides (Mauritz, 2002). Aluminates,

titanates, borates, and silicates are just some of the alkoxides that can be used to create a

sol-gel. A gel that contains SiO2 networks is given the term silica gel.

Hydrolysis and condensation. Gels derived from alkoxides form through

hydrolysis reactions followed by condensation reactions (Equations 5-7). Generally

speaking, hydrolysis replaces the alkoxide group with a hydrogen ion while condensation

produces siloxane bonding (Si-O-Si) with the products of the reaction being water or






19


alcohol. As the number of siloxane bonds increases, the individual molecules join

together to create a silica network. See Equations 5-7 (Mauritz, 2000).


I Hydrolysis
-Si-OR + HOH -Si-OH + ROH (5)
SReesterfication


Water
ICondensation 6)
-Si-OH + -Si-OH i -Si- + HOH
Hydrolysis |


Alcohol
Condensation
-Si-OH + -Si-OR --Si-O-Si- + ROH (7)
SAlcoholysis


Gelation. After condensation reactions begin, gelation can occur (see Figure 2-6

for structure changes). Gelation is considered the growth period for the colloidal

particles, which grow by polymerization or aggregation, depending on the conditions.

Eventually, as the last links are made, the sol's viscosity increases to the point where the

sol becomes a semi-rigid gel that consists of a network of pores filled with liquid.

Aging. Aging occurs after gelation. Here, terminal silanols re-orient themselves

and react with each other to form additional network linkages. This results in shrinkage

of the gel and forces the removal of some liquid from the pores. As the gel stiffens,

"Ostwald ripening" occurs. This is the phenomenon whereby necks begin to form

between primary particles causing the filling of small pores (see Figure 2-6 part c).










A B
A UQUID B LIOUDQUD NECKING





0 EGE
SQL GEL AGED GEL

NO 0 PORE





DRIED GEL PARTIALLY
O ENSIFIED
GEL

GLASS PORE




Figure 2-6. Structural changes during the sol-gel process (adapted from Hench and West,
1990)

Drying. The drying step involves the removal of the solvent from within the pores

of the gel. This is very important in the determination of pore size. There are three types

of gels characterized by the solvent removal process: xerogels, aerogels, and cryogels. As

defined by Legrand (1998), a xerogel is obtained by evaporation of the liquid component

at ambient pressure and temperatures below the critical temperature of the liquid; an

aerogel is obtained from the evaporation of the liquid component above its critical

temperature; and a cryogel is obtained from the freezing and subsequent sublimation of

the liquid component. Aerogels have higher surface areas than xerogels and do not have a

material volume loss due to shrinkage. Yet, because supercritically drying gels can be

cumbrous and expensive, research has focused on making xerogels with similar









characteristics to aerogels. As far as water treatment is concerned, xerogels would be

preferred over aerogels due to the higher mechanical strength of the xerogel.

Curing. Calcination or curing is the final heat treatment step in producing a

mechanically strong gel. This thermal treatment also affects several characteristics of the

silica gel. With increased heat treatment, the BET (Brunauer, Emmett, and Teller

equation) surface area will decrease due to sintering (Papirer, 2000) and the dehydration

(around 180-200 C) and subsequent dehydroxylation (above 2000 C) of the surface of

the silica, resulting in the formation of siloxane bridges (Holysz, 1998). The final product

resulting from all these steps hydrolysiss, condensation, gelation, aging, drying, curing) is

a gel containing a relatively monodisperse pore size and displaying specific

characteristics associated with the conditions it experienced during processing.

Effect of catalyst. The two alkoxides most often used are tetramethoxysilane

(TMOS) and tetraethoxysilane (TEOS). TMOS has the advantage of rapid hydrolysis

under a variety of conditions, but the toxic methanol produced can be hazardous to the

eyes and lungs. Conversely, TEOS is hydrolysis rate limited (Brinker and Scherer, 1990),

but produces less toxic ethanol during the reaction. To increase the hydrolysis reaction

rate of TEOS, an acid or base can be added as a catalyst. However, the nature and

concentration of the catalyst will affect the characteristics of the gel. For example, the

addition of hydrofluoric acid will rapidly increase the rate of gelation as well as increase

the pore size (Powers, 1998).

Other important parameters to consider in the production of a gel are H20 to Si

molar ratio (called the "R" factor) and the aging and drying schedules. A high R ratio will

promote more rapid hydrolysis, for the concentration of water affects the rate of









hydrolysis. The drying time and temperature also affect the pore size of the gel as well as

the shape of the gel. For example, the longer the time to dry, the more Ostwald ripening

occurs and the less likely the gel will crack, thereby producing a larger pore size than a

gel of the same composition dried in short period of time (less than 48 hours). In

summary, there are many parameters in the production of a gel that will greatly affect the

final product's characteristics. It is these differences in silicas produced by various

preparation methods that make the comparison of data found in the literature a very

difficult process.

2.5.2 Surface Chemistry and Adsorption Characteristics

The surface of silica consists of two types of functional groups: silanol groups (Si

- O H) and siloxane groups (Si O Si). The silanol groups are the locale of activity

for any process taking place on the surface, while siloxane sites are considered

nonreactive (Unger, 1979). Porous amorphous silica contains three types of silanols on its

surface: isolated, geminal, and vicinal (or associated). Figure 2-7 shows the distinction

between these groups. The unequal distribution of the silanols in the sililca matrix,

resulting from irregular packing of macromolecules as well as incomplete condensation,

results in a heterogeneous surface (non-uniformity in the dispersion of silanol groups) for

synthesized silica (El Shafei, 2000).

The various silanols can have different adsorption activity and current knowledge

indicates that the isolated silanols are the more reactive species (Nawrocki, 1997; El

Shafei, 2000). With increasing temperature of heat treatment, the silica surface becomes

hydrophobic due to the condensation of surface hydroxyls (dehydroxylation) and the

formation of siloxane bridges (see Figure 2-8). This increases the ratio of isolated sites

versus other silanol groups on the surface and therefore increases adsorption.














geminal isolated single associated silanols
H H H




hydrogen bounded water molecules


Figure 2-7. Silanol groups on the silica surface (Legrand, 1998).

At temperatures up to 450C, the removal of the hydroxyl groups does not result

in visible changes in structure, density, or specific surface area of the silica because the

rehydroxylation makes it possible to return to the initial state (Davydov, 2000).

Rehydroxylation of the silica surface by exposure to water vapor is reversible up to

600C. At temperatures above this, the rehydroxylation becomes a slow process or

remains incomplete (Unger et al., 2000). Also, there is a decrease in surface area as a

result of sintering at temperatures above 500C (Persello, 2000).


H- Oi H- -...-H
Si Si


Si Si + HO


Figure 2-8. Representation of dehydroxylation (Burneau and Gallas, 1998).

Pore size can be affected by the heat treatment of the silica, but is also a function

of the manufacturing process. The effect of pore size on adsorption is dependent on the

compound's characteristics, such as molecule size, shape, and functional groups

(Dabrowski, 2001; Goworek et al., 1997; Goworek et al., 1999). In general, the smaller

the pore, the larger the surface area and the larger the capacity for adsorption of









contaminants. Yet, Goworek et al. (1997) found that some compounds may be hindered

by a narrow pore size due to steric forces. Therefore, the adsorption ability for that

particular compound would be higher in a silica with larger pores.

Adsorption on silica surfaces is primarily due to ion interactions and hydrogen

bonding with the silanol groups. Therefore, the adsorption kinetics in relation to these

silanols is greatly dependent on the pH. Silica's point of zero charge ranges between 2

and 3 (Papirer, 2000; Persello, 2000). At a pH less than the PZC, the surface has a

positive charge due to the formation of Si-OH2 At a pH above the PZC, the surface of

the silica has a negative charge due to the deprotonation of the silanol group resulting in

Si-O-. As the pH approaches 7, the Si-O- sites become significant (the pKa for SiOH =

SiO- + H is 7) (Cox, 1993). These Si-O- sites readily react with cations in solution and

an increase in pH would greatly increase the adsorption of these compounds (Kinniburgh

and Jackson, 1981). For nonionic compounds in solution, adsorption can occur at the

silanol groups by hydrogen bonding or by Van der Waals forces. For anionic compounds,

there may be some adsorption at a pH above the PZC of silica, but it would be occurring

at the TiO2 surface because it is positively charged at pH < 6.0. One factor to consider

when adjusting pH in order to achieve the highest adsorption onto silica-titania

composites is the dissolution of silica at a pH greater than 9 (Iler, 1979). In order to

create a long-lasting photocatalytic composite, the dissolution of silica would need to be

avoided.














CHAPTER 3
MATERIALS AND METHODS

3.1 Silica-Titania Composites

The silica-titania composites were made by a sol-gel method (Powers, 1998)

using nitric acid and hydrofluoric acid as catalysts to increase the hydrolysis and

condensation rates, thereby decreasing the gelation time. The basic formula used to create

gels with a pore size of roughly 140 D is as follows: 25 mL water, 50 mL ethanol, 35 mL

TEOS (tetraethylorthsilicate), 4 mL nitric acid (IN), and 4 mL HF (3%). TEOS was used

in place of TMOS (tetramethylorthosilicate) because it produces less toxic ethanol

instead of highly toxic methanol that would be created by reactions with TMOS. Varying

the volume of hydrofluoric acid allowed the manipulation of the pore size. Decreasing the

volume to 1 mL HF resulted in 30 A pores, while increasing it to 8 mL resulted in 320 A

pores.

The chemicals (reagent grade from Fisher) were added individually, in no

particular order, to a polymethylpentene container. A magnetic stir plate provided

sufficient mixing. The solution was allowed to mix until gelation occurred (2 hours for

140 A gel). During this time, a known mass of Degussa P25 was added to the batch and

the percentage of titania recorded is given as a percent by volume of silica precursor.

After gelation, the lid was put on, preventing any air exposure in order to protect against

premature evaporation. Next, it was aged at room temperature for two days, then at 650 C

for two days. After aging, the gel was removed from the container, rinsed with deionized

water to remove any residual acid or ethanol, and placed in a Teflon container for the









next series of heat treatments. A small hole in the lid of the container allowed slow and

uniform drying of the gel. It was then placed in an oven and the temperature was ramped

from room temperature to 1030 C (2/min) and kept constant for 18 hours, resulting in the

vaporization of liquid solution within the silica network. Next, the temperature was

ramped to 1800 C (2/min) to remove any physically adsorbed water. It was kept constant

for 6 hours and then was slowly decreased back to room temperature over a 90 minute

period. These gels were powdered and sieved through 325 mesh to be used for batch

performance studies.

The titania-doped pellets were made in a similar fashion, except the solution

(including the P25) was pipeted into polystyrene 96-well assay plates before gelation.

The volume added to each well was approximately 0.3 mL. The plates were then covered

with lids and wrapped in foil to prevent premature evaporation. The same aging and

drying schedule as previously discussed was used, but before the drying step, the pellets

were removed from their plates and placed in a Teflon container. The resultant size of an

individual cylindrical pellet after drying was approximately 5 mm in length with a

diameter of 3 mm. These pellets were used in column performance studies.

3.2 Batch Performance Studies

The batch performance studies were done to determine an optimal titania loading,

curing temperature, and pore size that is efficient for both adsorption and destruction of

contaminants. The powdered gels were compared based upon destruction (removal of

color) of reactive red dye (RR) from Sigma-Aldrich and adsorption of crystal violet dye

(CV) from Fisher. All dye solutions were made with nanopure water. The destruction

studies were performed as follows: 30 mg of gel was added to 100 mL of a 10 mg/L RR

solution buffered with 200 mg/L of sodium bicarbonate in a 125 mL Erlenmeyer flask.









Then pH adjustments were made with sodium hydroxide or nitric acid. The flask opening

was then covered with parafilm to maintain a closed system with a limited amount of

dissolved oxygen (DO). In a system designed for NASA, oxygen would be a limiting

factor due to the problem of continually supplying oxygen for water treatment for long

duration space flight. The flasks were placed within a box containing four magnetic

mixers relatively equidistant from a single 4-Watt UV lamp. The UV exposure time was

two hours and the percent destruction was calculated by comparing the initial and final

absorbance readings using a Hach DR/4000U (Loveland, Colorado) Spectrophotometer at

RR's peak absorption wavelength of 538 nm. The spectrophotometer gave a linear

response across the concentration range for each dye. The UV intensity (mW/ cm2) was

varied by changing the position of the mixers in regard to the lamp and by the use of

additional lamps. UV intensity (wavelengths between 320 to 380 nm) was measured with

a VWR Ultraviolet Light Meter (Atlanta, GA). Each mixer was placed where the desired

UV intensity was achieved at the mixer's center.

The adsorption studies were performed as follows: 30 mg of gel was added to 100

mL of 10 mg/L CV solution with no bicarbonate added. The studies were conducted at a

pH of 4, using nitric acid and sodium hydroxide for the pH adjustments. The flasks were

covered with parafilm and placed in a box with no penetration of light and allowed to mix

for 24 hours. The percent adsorbed was calculated from the absorbance reading at a

wavelength of 590 nm.

The BET (Brunauer, Emmett, and Teller equation) surface area and pore volume

analysis was performed on a Quantachrome NOVA 1200 Gas Sorption Analyzer

(Boynton Beach, FL). The powdered samples were outgassed at 110 C for









approximately 24 hours and analyzed using nitrogen adsorption isotherms. The average

pore size was calculated by dividing the pore volume by the surface area. Pore size

distribution curves were also attained and coincided with the calculated pore size.

XRD (X-ray diffraction) analysis was performed on cured gels to determine the

phase of titania, anatase or rutile. These gels were cured at temperatures up to 8000 C as

an additional curing step after the regular drying schedule. The analysis was performed

on a APD 3720 from Philips Analytical (Almelo, Netherlands). The reflections of the two

crystalline TiO2 phases, anatase (20 = 25.3) and rutile (20 = 27.4), were taken from

Grieken et al. (2002).

3.3 Column Performance Studies

The purpose of these studies was to show the performance of the pellets in a

regenerative system, which places equal significance on the adsorption of a contaminant

as well as the feasibility of the contaminant to be mineralized. A cylindrical column (6

inches in length, 0.5 inch diameter) with a porous frit was filled with 10 mL of pellets.

The column setup is shown in Figure 3-1. A CV dye solution of 2 mg/L was pumped

through the column until the effluent concentration equaled the influent concentration.

Then the CV pump was turned off and the deionized water and UV light was turned on

(flux = 1.1 mW/cm2) for a set time (e.g., one hour). During this regeneration time, the

water was pumped through the column to provide the reactants (02, H20) required for

photocatalysis. After the regeneration time, the CV pump was turned back on. This cycle

was continued several times for separate packed columns of 140 A gels and 30 A pellets,

both loaded with 12% TiO2. The flowrate for both the deionized water and CV dye was 8

mL/min.






































UV Light


Effluent


Front View


- u





I I


Side View


Figure 3-1. Column setup.


Reflective
- Cover















CHAPTER 4
RESULTS AND DISCUSSION

4.1 Dye Comparison

In order to accurately measure the photocatalytic ability of the silica-titania (Si02

- TiO2) composites, it was important to choose a dye that displayed little or no photolytic

behavior, i.e. change in absorbance as a response to UV light exposure alone. The

following four dyes were examined in order to determine the most suitable dye for the

studies herein: malachite green (MG), methylene blue (MB), crystal violet (CV), and

reactive red (RR). See Appendix B for structures of these dyes.

Using a spectrophotometer, each dye was scanned at a range of 290 nm to 800 nm

in order to find each dye's peak absorbance wavelength. A summary of the peak

wavelengths is shown in Table 4-1. The absorbance measured at this peak wavelength

was used for comparing initial and final readings.

Table 4-1. Peak absorbance wavelengths.
Dye Peak Wavelength
MG 619 nm
MB 656 nm
CV 590 nm
RR 538 nm


A 100 mL solution of each dye (10 mg/L) was prepared in a 125 mL Erlenmeyer

flask and exposed to three different intensities of UV light for two hours. The findings are

presented in Figure 4-1. Each dye experienced a change in absorbance of less than 5% at

each flux, except for crystal violet, which experienced a 19% change at a flux of 1.1










mW/cm2. Based upon this undesirable change, crystal violet was excluded for future

destruction studies.

Subsequently, the remaining dyes were evaluated for photolysis at various

concentrations (0.5 mg/L to 10 mg/L). At 0.5 mg/L, there was a 98% and 20% change

between the initial absorbance readings for MB and MG respectively, compared to their

final readings after two hours of UV exposure (0.45 mW/cm2) (Figure 4-2). Large

discrepancies were also observed at 1 mg/L and 2 mg/L. Therefore, MB and MG were

eliminated as viable candidates to investigate the photocatalytic ability of the composites.

For RR (see Figure 4-3), the largest change in absorbance was only 13% at a flux of 1.1

mW/cm2, the highest flux achievable. Thus, RR was chosen for the photocatalysis

studies. The amount of photolysis was not subtracted from the results of the destruction

studies.



25

c 0.45 mW/cm^2
S20 -- 0.75 mW/cm^2
0
o1
] 1.1 mW/cm^2



I 10
)



5



RR CV MB MG
Dye

Figure 4-1. Photolysis of dyes as a function of UV flux for 2 hours of exposure.











- MB Initial Absorbance
* MG Initial Absorbance


---.--- MB Final Absorbance
MG Final Absorbance


2
1.8-
W 1.6
a-
'S 1.4 ---
1.2

ns = 1

S 0.8 98% for MB
S0. and 20% for
$ E 0.6 -----
0 0.4 M,
45% for MG
< ^B Y and 20% for
0
0 2 4 6 8 10
Concentration (mg/L)

Figure 4-2. Photolysis of MB and MG (2 hours UV exposure at 0.45 mW/cm2) as a
function of initial dye concentration.


-o- Initial Absorbance
- Final Absorbance (1.1 flux)


---- Final Absorbance (0.45 flux)


0 1 2 3 4 5 6 7 8
Concentration (mg/L)


9 10 11 12


Figure 4-3. Photolysis of RR (2 hours UV exposure) as a function of initial dye
concentration and flux.


0.3

0.25

o 0.2

S0.15
0
.0 0.1

0.05

0









Besides photolysis, pH can also affect a dye's peak absorption wavelength.

Therefore, RR solutions of 10 mg/L were prepared at different pHs and compared before

and after UV exposure. The change in absorbance (at 538 nm) as a function of pH is

shown in Figure 4-4. At a pH above 8, there was a decrease in absorbance along with a

slight visible change of color from a dark pink to a red. Figure 4-5 displays a section of

the spectrophotometer scan and shows the changes that occur at an alkaline pH. At a pH

of 7, the maximum peak wavelength is 538 nm. At a pH of 9.5, the highest peak occurs at

both 514 nm, as well as 538 nm. After two hours of UV exposure, the max peak returns

back to 538 nm. Also noticeable at pH 9.5 is a slight downshift in the curve after UV

exposure. Therefore, the pH was kept below 8 for comparison studies of gels to ensure

accurate measurements of destruction ability.




0.3

0.25

o 0.2

S0.15
0
0.1
S0.1 -- Abs. (initial)

0.05 -- Abs. (final)

0
0 2 4 pH 6 8 10 12
Figure 4-4. Change in absorbance of RR at 538 nm as a function of pH.
Figure 4-4. Change in absorbance of RR at 538 nm as a function ofpH.










Absorbance pH 7 ------.Absorbance pH 9.5
-- Absorbance ph 9.5 after UV

0.3

0.25 ,

o 0.2

0.15
0
0 0.1

0.05 -Z


400 450 500 550 600
Wavelength (nm)

Figure 4-5. Spectrophotometer scan of 10 mg/L RR as a function of pH, before and after
UV exposure.

4.2 Optimization of Gels

4.2.1 Titania Loading

In order to determine the optimal titania loading, 14 gels (140 A) of varied TiO2

loadings (from 0.77% to 40%) were created. The desired optimal loading would be found

as the composite having the highest rate of destruction for RR. These gels were

performance tested at three UV intensities for the destruction of RR (10 mg/L) and the

results are displayed in Figure 4-6. There is a definite plateau where the percent

destruction remains constant, even with the addition of more TiO2. It appears that this

plateau occurs at lower TiO2 percentages with higher UV intensities. For 1.1 mW/cm2,

the plateau occurs at 5% TiO2 and it occurs at 10% for 0.75 mW/cm2. At the lowest UV

intensity, the destruction levels off at 12%. These results were triplicated and the error

bars are included in the figure.












80 -

70-
0 1.1
S60 mW/cm^2
2 A 0.75
S50 mW/cm^2

40 0.45
W mW/cm^2
30

20

10
0 5 10 15 20 25 30 35 40
Percent TiO2

Figure 4-6. Destruction ofRR (pH 7.6) after 2 hours UV exposure versus TiO2 loading as
a function of flux.

The plateau can be explained by the TiO2 is agglomerating in solution during gel

preparation, which was visually observed with TiO2 loadings greater than 8%, and

therefore is decreasing the effective surface area of titania available for excitation and/or

oxidation and reduction reactions on its surface. The 12% optimal loading (wt/vol

precursor) at the lowest energy is approximately equal to 30 wt%, which was found to be

an optimum for destruction of organic compounds with other methods of producing

silica-titania composites (Jung and Park, 2000; Chun et al., 2001; Anderson and Bard,

1995).

The structure of RR includes sodium sulfonate groups that would make the

compound anionic when the sodium disassociates as the compound dissolves. Therefore,

the effect of pH on adsorption and destruction of RR was also investigated. It was

observed that at a pH range of 3 to 10, there was no adsorption of RR dye to the silica-










titania gels. Thus, the color disappearance of the dye can be directly attributed to

photocatalysis with a minimal amount due to photolysis.

The effect of pH on the destruction of RR is shown in Figure 4-7. It was observed

that acidic conditions enhanced the photocatalytic destruction of RR and it was initially

hypothesized that this was due to the TiO2 having a positive charge at pH 3, therefore, the

amount of RR adsorption that would occur on pure TiO2 at low pH would increase the

destruction rate.


pH 3, bicarb m pH 4, no bicarb
A pH 7.6, bicarb pH 7.6, no bicarb
100
90-
C 80
70

60
0 50
40
2 30
a. 20

0
0 --------------------------
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Percent TiO2

Figure 4-7. Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm2)
versus TiO2 loading as a function of pH.

Yet, when a slurry of titania was tested for destruction and adsorption at a range

of pHs, the results were contrary to what was hypothesized (Figure 4-8). While the

adsorption onto the titania surface may have slightly increased at a lower pH, the level of

total removal (destruction and adsorption) did not change. It is then presumable that at a

pH greater than 6, the silica surface becomes more concentrated with SiO- groups, thus










repelling RR. As the pH approaches 3, the silica has less of a negative charge and would

allow the penetration of RR into the pores. Hence, acidic conditions would increase the

ability of RR to get within the vicinity of the titania.



Destruction U Adsorption
100 100
90 90
80 80
70 70 c
0 60 60
E
0 50 50 C.
S 40 40 0
-%
30 30 "
S20 20
10 10
0 0
0 1 2 3 H 4 5 6 7 8
pH

Figure 4-8. Destruction of RR (10 mg/L) after 1 hour UV with TiO2 slurry at various
pHs.

Grey water contains many sulfates, phosphates, and salts that may inhibit the

destruction of unwanted compounds (Chen et al., 1997). The addition of sodium

bicarbonate allows one to see the performance of the silica-titania composites in a

realistic feed stream. In Figure 4-7, it is also interesting to notice the affect of bicarbonate

on the destruction efficiency. There is little increase in destruction (with no bicarbonate)

at a pH of 7.6 within a range of 0.77% to 6% TiO2. At a larger TiO2 percentage, the

difference becomes very noticeable. However, Figure 4-9 shows that the increase in

destruction without bicarbonate present is only seen within a certain range of TiO2

loading (8% to 12%).











no bicarbonate Awith bicarbonate
100
90
0 80
0 70
S 0 -
5 30 -
40 A

20
10 -
0
0 5 10 15 20 25 30 35 40
Percent Ti02

Figure 4-9. Effects of bicarbonate at pH 7.6 versus TiO2 loading.

It was also thought that surface area may play a role in the leveling-off seen in

destruction of RR. As can be seen in Figure 4-10, there is no real trend between BET

surface area and titania loading, as well as between pore volume and titania loading.

Notice that this surface area is normalized on a volume basis. This trend is different for

silica-titania composites prepared by other methods, for which there is a drastic change

(up to a loss of 200 m2/g) in surface area with an increase in percentage of titania

(Greiken et al., 2002). In conclusion, there is no correlation between surface area or pore

volume and the destruction of RR.

For adsorption studies, CV dye was used because of its cationic nature resulting in

a high affinity for the silica-titania composites. These studies were conducted at a low pH

(worst case scenario for adsorption of a cation contaminant onto a silica surface). Figure

4-11 compares the amount of adsorption versus titania loading. There is an increase in










adsorption with additional titania in the gel. It is interesting to note, however, that

adsorption, similar to the destruction studies, reached a plateau at 12% loading.

Therefore, based upon the destruction studies and adsorption studies, an optimal titania

loading of 12% TiO2 was chosen.


BET SA/vol A Pore Volume
0.5
70
-0
60 0.4
o e
50 *
0.3 O
E 40 *

30 AAAA A 0.2a
AA A A A A 0
j 20 A
m 0.1
10 -

0 0
0 5 10 15 20 25 30 35 40 45
Percent TiO2

Figure 4-10. BET surface area and pore size versus titania loading.

35

30-.

0




r
25 -
2 10M

0 a
0 -------------------------
0 5 10 15 20 25 30 35 40 45
Percent TiO2

Figure 4-11. Adsorption of CV (10 mg/L) versus TiO2 loading after 24 hours of mixing.









4.2.2 Curing Temperature

The effect of curing temperature was investigated on gels doped with 6% TiO2

and those with no TiO2 (all made with the 140 A initial formula). The literature showed

that there can be an increase in adsorption with gels produced at higher curing

temperatures (up to 4000C) due to increasing the ratio of isolated silanols on the surface

versus other types of silanols. Yet, the curing temperature may have detrimental effects

on the photocatalytic ability of the composites. In looking at the destruction of RR (10

mg/L), Figure 4-12 shows that there is a 30% difference between the 180C gel versus the

500C gel at a pH of 7.8. This difference is less pronounced at pH 4. At 800C, there is a

marked decrease in destruction, regardless of pH. This can be explained by both a

decrease in surface area of the gels at 800C (see Figure 4-13) from an average of 275

m2/g to 150 m2/g for the titania loaded gel and a phase change of titania.


*pH 4 m pH 7.8
100
90
S80o
S70-
& 60
W 50-
0 ,40
2 30-
20
10 -
0 -------------------------
0 100 200 300 400 500 600 700 800
Curing Temp (oC)

Figure 4-12. Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm2)
versus curing temperature as a function of pH.










*6% 102 0% 102
400

350

300



0)
S150--
W
100-

50


0 200 400 600 800
Curing Temp (OC)

Figure 4-13. Effect of curing temperature on surface area.

The anatase phase of TiO2 begins transforming into rutile phase at a temperature

of 600C and becomes complete at 1000C (Zhang et al., 2001). It was thought that the

shift in the phase of titania at 800C may have also affected its photocatalytic ability, but

XRD analysis (Figure 4-14) revealed only a slight shift from the anatase phase (29 =

25.3) to the rutile phase (29 = 27.4) at 800C which may or may not be enough to have

caused a drastic decrease in destruction. It may also be that the phase change is mostly

occurring on the surface of the titania where the reactions take place. In other words, the

large anatase peak is due to internal anatase crystal formation and the rutile formation is

occurring foremost on the surface where it is greatly affecting photocatalysis.










(counts]-

2800 C


156B


lees-


see- ISOC


ze ......... *2" ......... 4* ..*.. *.... ** *'* ** ... ... ** ^.[' 3 e
20 22 24 26 28 [201 38

Figure 4-14. XRD analysis of various temperature-cured silica-titania composites.

In looking at adsorption of CV at various curing temperatures, Figure 4-15 shows

800C with the highest adsorption capacity even though it has the lowest surface area.

This may be explained by the cleaner surface (decrease of silanol sights) at this

temperature, resulting in the gel displaying a more hydrophobic behavior (Holysz, 1998).

This would decrease the competition between water and CV, allowing CV to efficiently

interact with the isolated silanols. However, the clean surface can slowly become

rehydroxylated when it is in contact with water (Davydov, 2000; Unger et al., 2000). The

adsorption ability would then most likely follow the same trend as gels that were treated

at lower curing temperatures. Therefore, based upon this and the destruction studies, the

optimal curing temperature chosen was 180C.










6% Ti02 0% Ti02
25


S20

0
, 15


- 10

S5


0


U







U U


0 100 200 300 400 500 600 700 800 900
Curing Temp (oC)

Figure 4-15. Effect of curing temperature on adsorption of CV on gels with and without
TiO2.

4.2.3 Pore Size

Pore size was also investigated for its effects on destruction ability, as well as

adsorption ability. Figure 4-16 shows that a large pore size (>140 A) increases the

destruction rates for RR. This is consistent with what is found in the literature. Jung and

Park (2000) concluded that high porosity and large pore size facilitate the mass transfer

of reactants, such as oxygen and reaction intermediates. Thus, the increase in surface area

by small pores is not always effective for high photoactivity. The opposite is true for

adsorption, because a smaller pore size results in more surface area available for

adsorption. Figure 4-17 shows the largest adsorption with a pore size of 30 A. Adsorption

decreases 73% with larger pores (> 30 A), and in this range there is no significant

difference in adsorption from 60 A versus 140 A.












0 pH=4 *pH= 8
100
90
S80 -
70
70
5O
2 60
50 -
4-
r 40


0.. 20
10
0
Q 20 ---------------------------------


0 50 100 150 200 250 300 3E
Pore Size (A)

Figure 4-16. Destruction of RR (10 mg/L) after 2 hours UV exposure (0.45 mW/cm2)
versus pore size.


50
45
40
.M 35
o 30
V 25
20

10
15
0

30 60 140 320

Pore Size (A)

Figure 4-17. Adsorption of CV on gels (6% TiO2) versus pore size.

4.3 Column Studies

Since the optimal pore size for adsorption is contrary to the optimal pore size for

destruction, it was decided to test two pore sizes in a flow-through regenerative column.

These studies utilized 12% optimal titania loaded pellets of 30 A and 140 A pore sizes. A

larger pore size was not analyzed for these studies due to the frailty of the gel at pore










sizes larger than 200 A. The duplicated column exhaustion curves for each pore size are

shown in Figures 4-18 and 4-19. The duplicated runs were done with pellets made from a

separate batch.

Table 4-2 summarizes the exhaustion results and shows a longer time to

breakthrough of the initial concentration for the column of 30 A gels. Yet, after the 30 A

pellets reached exhaustion and the system was regenerated by photocatalysis, its level of

adsorption ability decreased after each cycle. This was also visually observed. After one

hour of UV exposure the 30 A pellets were still slightly purple, while the 140 A

pellets had regenerated and returned to its initial white state.


-- Run 1 ---*--- Run 2

0.9
0.8
0.7----
0.6
0
0.5---
00.4.
0.3 -
0.2
0.1 ,
0 $- ------__, -.----

0 50 100 150 200 250 300 350 400 451
Time (minutes)

Figure 4-18. Column exhaustion curve for 140 A pellet (12% TiO2).











--- --- Run1


-- Run2


0 100 200 300 400
Time (minutes)

Figure 4-19. Column exhaustion curve for 30 A pellets (12% TiO2).

Table 4-2. Exhaustion (C/Co = 1) times (minutes) for column studies.
Pellet Pore Size 140 A 30 A
Run 1 Exhaustion Times 36 157
26 48
50 64
27
Run 2 Exhaustion Times 28 132
35 128
58 98
65
Average Exhaustion Time 40.6 104.5


500


Both sets of pellets were examined during a column run after exhaustion of the

column and it was observed that the 30 A pellets were still white on the inside, revealing

that the total surface available for adsorption was not utilized. The 140 A pellets were

purple throughout their interior, revealing that the available space for adsorption had been

exhausted. Thus, the small pore size of the 30 A gels inhibits the transit of CV solution

through the porous matrix. Therefore, the 140 A sized gels exhibit the greatest efficiency

in utilizing its surface area for adsorption and its photocatalytic ability to regenerate.










Thus, the 140 A pore size was chosen as the optimal pore size for the regenerative

system.

In order to determine if a shorter regeneration time was achievable while still

keeping the same trend as previously observed, the effect of regeneration time with UV

light was examined for the 140 A column. The time was decreased from 1 hour to 10

minutes and Figure 4-20 shows the comparison of this breakthrough curve with the 1

hour curves. The average breakthrough time was 36 minutes for the 10 minute run versus

46 for the 60 minutes run. Yet, the same trend seems to be exhibited by all 140 A

breakthrough curves. This increase in breakthrough time with each cycle shows the

benefit of using the 140 A pellets, for this trend was not observed with the 30A pellets.


--Run1 ...*... Run2 10 minute cycle

0.9
0.8
0.7 ,,------
0.6 _
p0.5
0.4
0.3
0.2 -
0.1
0 ,-- --------" > -,-v -* -.. -.-. -
0 50 100 150 200 250 30
Time (minutes)

Figure 4-20. Effect of regeneration time on 140 A column runs.














CHAPTER 5
SUMMARY AND CONCLUSIONS

Silica-titania composites were made using a sol-gel method that allows the doping

of titania during gelation. The gelation rate was increased by use of acid catalysts and

allowed the composites to be made into pellets with sufficient dispersion of TiO2 within

the silica matrix. Varying the concentration of hydrofluoric acid allowed the

manipulation of pore size.

Three variables (TiO2 loading, curing temperature, and pore size) were

investigated to determine the best silica-titania composites for a packed column system.

These variables were optimized based upon destruction and adsorption studies and the

gels were characterized by surface area and XRD analysis. Reactive red dye (RR) was

chosen for its non-photolytic behavior in the analysis of photocatalytic ability for the

destruction studies and crystal violet dye was chosen because of its cationic nature for the

adsorption studies.

An optimal loading of 12% TiO2 was chosen based upon the results of studies

involving the destruction of RR and adsorption of CV. At higher loadings, it was

observed that the titania is agglomerating during mixing, limiting the effective surface

area available for reactions on its surface. Also, there was no trend with BET surface area

on a volume basis versus titania loading. In addition, the adsorption studies revealed a

plateau in uptake beginning at 12% TiO2.

A curing temperature of 800C showed the largest effects (a decrease of over 80%

destruction) on photocatalytic ability. At this temperature, the surface area decreased and









the phase change of anatase to rutile is evident. This resulted in a dramatic decrease in

destruction of RR. Yet, this temperature showed the highest level of adsorption. It was

concluded that the clean silica surface and more hydrophobic behavior the silica surface

displays at this temperature increased adsorption. This increase in adsorption would

decrease over time due to rehydroxylation. The 180C gel performed the best for

destruction and was chosen as the optimal curing temperature.

Concerning pore size, the highest destruction rates were obtained with large pore

sizes (> 140 A). The opposite was true for adsorption, which was greater with small pore

sizes (30 A). Column studies were done in order to determine the best pore size for a

regenerative system. The smaller pore size had the longer time to exhaustion, but did not

efficiently regenerate. The small pores inhibited the ability of water or the contaminant to

move through the matrix, i.e. the path of least resistance was around the pellet versus

through the pellet. On the other hand, the 140 A gels achieved complete regeneration

after one hour of UV exposure and efficiently utilized their surface area for adsorption. In

conclusion, the optimal pore size chosen for the regenerative flow-through column was

140 A.














APPENDIX A
COMMON PREPARATION METHODS OF MIXED AND SUPPORTED OXIDES AS
DISCUSSED BY GAO AND WACHS (1999).

A.1 Mixed Oxides

A.1.1 Sol-Gel Hydrolysis

Sol-gel hydrolysis involves the acid hydrolysis and condensation of chemically

mixed Ti- and Si- alkoxides. In summary, a Ti atom acts as a substitute for a Si atom in

the silica network. It is the most widely used method due to its capability in controlling

the textural and surface properties of the mixed oxides, but the synthesis conditions can

greatly affect the homogeneity of the final product. There is a recently developed two-

stage hydrolysis procedure that solves this problem and results in the highest

homogeneity. Yet, these mixed oxides can only be obtained at low TiO2 content, less than

15 wt%. At higher Ti content, due to the larger size of the Ti atom versus Si, atom, there

is distortion of the SiO2 network. This increases the OH content on the silica surface and

affects the crystal phase of the TiO2.

A.1.2 Coprecipitation (Not Used In This Study)

Coprecipitation involves the simultaneous precipitation of Si02 and TiO2 (at high

pH) that also includes the formation of linkages with each other. This method also has

similar problems to sol-gel hydrolysis.









A.2 Supported Oxides

A.2.1 Impregnation (Not Used In This Study)

Impregnation involves the coating of silica with Ti by mixing with a titanium

precursor, such as TiC14 or a Ti-alkoxide. The following equation summarizes the

process:

Si OH + Ti+4 Si-O-Ti

The homogeneity of the supported oxide is dependent on the concentration of surface

silanols, pretreatment temperature, and molecular size of precursor. A maximum

dispersion of TiO2 was found to be roughly 4.0 Ti atom/nm2. Consistent creation and

determination of the titania phase for catalytic purposes is a major problem with this

method.

A.2.2 Chemical Vapor Deposition (not used in this study)

The TiO2 is deposited onto the porous silica surface from a gaseous metal

precursor. Consistent coating is difficult to attain from batch to batch and another

problem is the reduction in surface area after coating.












APPENDIX B
DYE STRUCTURES
B.1 Methylene Blue
MW: 319.86

Empirical Formula: C16H18ClN3S






(H3C)2N Jc Sxr% N(CH3)2
C1



B.2 Malachite Green


MW: 365


Empirical Formula: C23H25N2C1


N(CH3)2


N(CH3)2








B.3 Crystal Violet
MW: 407.996

Empirical Formula: C25H30N3C1


N(CH3)2


(C H3)2 N


N(CH3)2


B.4 Reactive Red

MW: 615.34

Empirical Formula: C19H12C12N607S2


NaOS SONa




NH OH

NIN

Cl N CI















APPENDIX C
TEM AND SEM PICTURES


TEM of 4% TiO2 gel at 20,000x















LIST OF REFERENCES


Abrahams, J., Davidson, S. and Morrison, C., "Optimization of the photocatalytic
properties of titanium dioxide," Journal of Photochemistry, Vol. 29, pp. 353-361,
1985.

Al-Ekabi, H. and Serpone, N., "Kinetic studies in heterogeneous photocatalysis:
photocatalytic degradation of chlorinated phenols in aerated aqueous solutions
over TiO2 supported on a glass matrix," Journal of Physical Chemistry, Vol. 92,
pp. 5726-5731, 1988.

Anderson, C. and Bard, A., "An improved photocatalyst of TiO2/SiO2 prepared by a sol-
gel synthesis" Journal of Physical Chemistry, Vol. 99, pp. 9882 9885, 1995.

Anderson, C. and Bard, A., "Improved photocatalytic activity and characterization of
mixed TiO2/SiO2 and TiO2/A1203 materials," Journal of Physical Chemistry B,
Vol. 101, pp. 2611-2616, 1997.

Blake, D., "Bibliography of work on photocatalytic removal of hazardous compounds
from water and air," Update Number 1, NREL/TP-473-20300, Golden, CO:
National Renewable Energy Laboratory, 1995.

Block, S. and Goswami, D., "Proceedings of ASME Solar Energy Conference," Kohala
Coast, Hawaii, pp. 431-438, 1995.

Brinker, C. and Scherer, G., Sol-Gel Science: The Physics and Chemistry of Sol-Gel
Processing, Academic Press Inc., San Diego, 1990.

Burneau, A. and Gallas, J., "Hydroxyl groups on silica surfaces," in: The Surface
Properties of Silicas, Legrand, P. (Ed), John Wiley & Sons, Inc., New York, pp.
147-234, 1998.

Chen, H., Zahraa, O. and Bouchy, M., "Inhibition of the adsorption and photocatalytic
degradation of an organic contaminant in an aqueous suspension of TiO2 by
inorganic ions," Journal of Physical Chemistry A: Chemistry, Vol. 108, pp. 37-44,
1997.

Chun, H., Yizhong, W. and Hongxiao, T., "Preparation and characterization of surface
bond-conjugated TiO2/SiO2 and photocatalysis for azo dyes," Applied Catalysis
B: Environmental, Vol. 30, pp. 277-285, 2001.









Cox, G., "The influence of silica structure on reversed-phase retention," Journal of
Chromatography, Vol. 656, pp. 353-367, 1993.

Crittenden, J., Notthakun, S., Hand, D. and Perram, D., United States Patent 5,182,030,
January 26, 1993.

Dabrowski, A., "Adsorption from theory to practice," Advances in Colloid and
Interface Science, Vol. 93, pp. 135-224, 2001.

Davydov, V., "Adsorption on silica surfaces," in: Adsorption on Silica Surfaces, Papirer,
E. (Ed), John Wiley and Sons Ltd., New York, 2000.

Degussa Technical Bulletin No. 56, "Highly Dispersed Metallic Oxides Produced by the
AEROSIL Process," 1990.

Elizardo, K., "Fighting pollution with hydrogen peroxide," Pollution Engineering, pp.
106-109, September, 1991.

El Shafei, G., "Silica surface chemical properties," in: Adsorption on Silica Surfaces,
Papirer, E. (Ed), John Wiley and Sons Ltd., New York, pp. 35-62, 2000.

Emeline, A., Ryabchuk, V. and Serpone, N., "Factors affecting the efficiency of a
photocatalyzed process in aqueous metal-oxide dispersions: prospect of
distinguishing between two kinetic models," Journal of Photochemistry and
Photobiology A : Chemistry, Vol. 133, pp. 89-97, 2000.

Environmental Protection Agency, "Best management practices for pollution prevention
in the textiles industry," EPA/625/R-96/004, September, 1996.

Environmental Protection Agency, "Sector notebook project profile of the textile
industry part 2," EPA/310-R-97-009, September, 1997.

Gao, X. and Wachs, I., "Titania-silica as catalysts: molecular structural characteristics
and physico-chemical properties," Catalysis Today, Vol. 51, pp. 233-254, 1999.

Gerischer, H. and Heller, A., "The role of oxygen in photooxidation of organic molecules
on semiconductor particles," Journal of Physical Chemistry, Vol. 95, pp. 5261-
5267, 1991.

Goswami, D., Klausner, J., Wyness, P., Martin, A., Mathur, G., Schanze, K., Turchi, C.
and Marchand, E., "Solar photocatalytic treatment of groundwater at Tyndall
AFB : field test results," University of Florida, LTFNM/SEECL-9302, 1993.

Goworek, J., Nieradka, A. and Dabrowski, A., "Adsorption from ternary mixtures on
silica gel of different mesoporosity," Fluid Phase Equilibria, Vol. 136, pp. 333-
343, 1997.









Goworek, J., Derylo-Marczewska, A. and Borowka, A., "Adsorption from the liquid
phase on silica gels of various structural heterogeneity," Langmuir, Vol. 15, pp.
6103-6106, 1999.

Grabner, G., Li, G., Quint, R., Quint, R. and Getoff, N., "Pulse laser-induced oxidation of
phenol in acid aqueous TiO2," Journal of the Chemical Society Faraday
Transactions, Vol. 87, pp. 1097-1101, 1991.

Grieken, R., Aguado, J., Lopez-Munoz, M. and Marugan, J., "Synthesis of size-controlled
silica-supported TiO2 photocatalysts," Journal of Photochemistry and
Photobiology A: Chemistry, Vol. 148, pp. 315-322, 2002.

Halmann, M., Photodegradation of Water Pollutants, CRC Press, Boca Raton, 1996.

Hanprasopwattana, A., Srinivason, S., Sault, A. and Datye, A., "Titania coatings on
monodisperse silica spheres: characterization using 2-propanol dehydration and
TEM," Langmuir, Vol. 12, pp. 3173-3179, 1996.

Hench, L. and West, J.," The sol-gel process," Chemical Review, Vol. 90, pp. 33-72,
1990.

Hoffman, M., Martin, S., Choi, W. and Bahnemann, D., "Environmental application of
semiconductor photocatalysis," Chemical Reviews, Vol. 95, pp. 69-96, 1995.

Hoigne, J., "Formulation and calibration of environmental reaction kinetics; oxidations
by aqueous photooxidants as an example," in: Aquatic Chemical Kinetics, Stumm,
W. (Ed.), John Wiley & Sons, Inc., New York, pp. 43-70, 1990.

Holysz, L., "The effect of thermal treatment of silica gel on its surface free energy
components," Colloids and Surfaces A: Physicochemical and Engineering
Aspects, Vol. 134, pp. 321-329, 1998.

Houas, A., Lachheb, H., Ksibi, M., Elaloui, E., Guillard, C. and Hermann, J.,
"Photocatalytic degradation pathway of methylene blue in water," Applied
Catalysis B: Environmental, Vol. 31, pp. 145-157, 2001.

Iler, R., The Chemistry of Silica, John Wiley & Sons, New York, 1979.

Jackson, N., Wang, C., Luo, Z., Schwitzgebel, J., Ekert, J., Brock, J. and Heller, A.,
"Attachment of TiO2 powders to hollow glass microbeads: activity of the TiO2-
coated beads in the photoassisted oxidation of ethanol to acetaldehyde," Journal
of the Electrochemical Society, Vol. 138, pp. 3660-3664, 1991.

Jung, K. and Park, S., "Enhanced photoactivity of silica-embedded titania particles
prepares by sol-gel process for the decomposition of trichloroethylene," Applied
Catalysis B: Environmental, Vol. 25, pp. 249-256, 2000.









Khan, A., Mazyck, D. and Wu, C., "TiO2 coated activated carbon: a regenerative
technology for water recovery," 021CES-73, International Conference on
Environmental Systems, San Antonio, 2002.

Kinniburgh, D. and Jackson, M., "Cation adsorption by hydrous metal oxides and clay,"
in: Adsorption oflnorganics at Solid-Liquid Interfaces, Anderson, M and Rubin,
A (Eds.), Ann Arbor Publishers, Ann Arbor, pp. 200-230, 1981.

Legrand, P., "On the silica edge," in: The Surface Properties of Silicas, Legrand, P. (Ed),
John Wiley & Sons, Inc., New York, pp. 130-200, 1998.

Lei, L., Chu, H., Hu, X. and Yue, P., "Preparation of heterogeneous photocatalysts
(TiO2/alumina) by metallo-organic chemical vapor deposition," Industrial
Engineering Chemical Research, Vol. 38, pp. 3381-3385, 1999.

Liu, T. and Cheng, T., "Effects of SiO2 on the catalytic properties of TiO2 for the
incineration of chloroform," Catalysis Today, Vol. 26, pp. 71-77, 1995.

Lu, M., Chen, J. and Chang, K., "Effect of adsorbents coated with titanium dioxide on the
photocatalytic degradation of propoxur," Chemosphere, Vol. 38, pp. 617-627,
1999.

Mao, Y., Schoneich, C. and Asmus, K., "Identification of organic acids and other
intermediates in oxidation in oxidation degradation of chlorinated ethanes on TiO2
surfaces en route to mineralization: a combined photocatalytic and radiation
chemical study," Journal of Physical Chemistry, Vol. 95, pp. 10080 10089,
1991.

Mao, Y., Schoneich, C. and Asmus, K., "Influence of TiO2 surface on 1,2-chlorine shift
in 0-chlorine substituted radicals as studied by radiation chemistry and
photocatalysis," Journal of Physical Chemistry, Vol. 96, pp. 8522-8529, 1992.

Mao, Y., Schoneich, C. and Asmus, K., "Radical mediated degradation mechanisms of
halogenated organic compounds as studied by photocatalysis at TiO2 and by
radiation chemistry," in: Photocatalytic Purification and Treatment of Water and
Air, Ollis, D.F. and Al-Ekabi, H. (Eds.), Elsevier Science Publishers, Amsterdam,
pp. 49-66, 1993.

Martin, S., Morrison, C. and Hoffman, M., "Photochemical mechanism of size-quantized
vanadium-doped TiO2 particles," Journal of Physical Chemistry, Vol. 98, pp.
13695-13704, 1994.

Matthews, R., "Kinetics of photocatalytic oxidation of organic solutes over titanium
dioxide," Journal of Catalysis, Vol. 111, pp. 264-272, 1988.

Mauritz, K., "Sol-gel chemistry," http://www.solgel.com, August 4, 2002.









Millennium Chemical, "Titanium dioxide crystal types," www.millenniumchem.com,
November 1, 2002.

Mills, A. and Sawunyama, P., "Photocatalytic degradation of 4-chlorophenol mediated by
TiO2: a comparative study of the activity of laboratory made and commercial
TiO2 samples," Journal of Photochemistry and Photobiology A: Chemistry, Vol.
84, pp. 305-309, 1994.

Nawrocki, J., "The silanol group and its role in liquid chromatography," Journal of
Chromatography A, Vol. 779, pp. 29-71, 1997.

NASA, "Requirements definition and design consideration," CTSD-ADV-245, 1998.

Okamoto, K., Yamamoto, Y., Tanaka, H., Tanaka, M. and Itaya, A., "Heterogeneous
photocatalytic decomposition of phenol over TiO2 powder," Bulletin of the
Chemical Society of Japan, Vol. 58, pp. 2015-2022, 1985.

Ollis, D., Pelizzetti, E. and Serpone, N., "Destruction of water contaminants,"
Environmental Science and Technology, Vol. 25, pp. 1523-1529, 1991.

Palmisano, L. and Sclafani, A., "Thermodynamics and kinetics for heterogeneous
photocatalytic processes," in: Heterogeneous Photocatalysis, Schiavello, M. (Ed),
John Wiley & Sons Ltd, New York, pp. 109-132, 1997.

Papirer, E., Adsorption on Silica Surfaces, John Wiley and Sons Ltd., New York, 2000.

Persello, J., "Surface and interface structure of silicas," in: Adsorption on Silica Surfaces,
Papirer, E. (Ed), John Wiley and Sons Ltd., New York, pp. 297-342, 2000.

Powers, K., The development and characterization of sol gel substrates for chemical
and optical applications," Ph.D. Dissertation, University of Florida, 1998.

Sauer, T., Neto, G., Jose, H. and Moreira, R., "Kinetics of photocatalytic degradation of
reactive dyes in a TiO2 slurry reactor," Journal of Photochemistry and
Photobiology A: Chemistry, Vol. 149, pp. 47-154, 2002.

Serpone, N., "Brief introductory remarks on heterogeneous photocatalysis," Solar Energy
Materials and Solar Cells, Vol. 38, pp. 369-379, 1995.

Serpone, N., Borgarello, E., Harris, R., Cahill, P., Borgarello, M. and Pelizzetti, E.,
"Photocatalysis over TiO2 supported on a glass substrate," Solar Energy
Materials, Vol. 14, pp. 121-127, 1986.

Sheintuch, M. and Matatov-Meytal, Y., "Comparison of catalytic processes with other
regeneration methods of activated carbon," Catalysis Today, Vol. 53, pp. 73-80,
1999.









Sheng, H. and Chi, M., "Treatment of textile waste effluent by ozonation and chemical
coagulation," Water Resources, Vol. 27, pp. 1743-1748, 1993.

Suri, R., Liu, J., Hand, D., Crittenden, J., Perram, D. and Mullins, M., "Heterogeneous
photocatalytic oxidation of hazardous organic contaminants in water," Water
Environment Resources, Vol. 65, pp. 665 673, 1993.

Takeda, N., Torimoto, T., Sampath, S., Kuwabata, S. and Yoneyama, H., "Effect of inert
supports for titanium dioxide loading on enhancement of photodecomposition rate
of gaseous propionaldehyde," Journal of Physical Chemistry, Vol. 99, pp. 9986-
9991, 1995.

Tanaka, K., Capule, M. and Hisanaga, T., "Effect of crystallinity of TiO2 on its
photocatalytic action," Chemical Physics Letters, Vol. 187, pp. 73-76, 1991.

Torimoto, T., Ito, S., Kuwabata, S. and Yoneyama, H., "Effects of adsorbents used as
supports for titanium dioxide loading photocatalytic degradation of
propyzamide," Environmental Science and Technology, Vol. 30, pp. 1275-1281,
1996.

Turchi, C. and Ollis, D., "Photocatalytic degradation of organic water contaminant:
mechanisms involving hydroxyl radical attack," Journal of Catalysis, Vol. 122,
pp. 178-192, 1990.

Uchida, H., Itoh, S. and Yoneyama, H., Photocatalytic decomposition of propyzamide
using TiO2 supported on activated carbon," Chemistry Letters, pp. 1995-1998,
1993.

Unger, K., Porous Silica, Elsevier, 1979.

Unger, K.,Kumar, D., Ehwald, V. and Grossman, F., "Adsorption on silcia surfaces from
solution and its impact on chromatographic separation techniques," in: Adsorption
on Silica Surfaces, Papirer, E. (Ed), John Wiley and Sons Ltd., New York, pp.
565-594, 2000.

US Geological Survey Minerals Information, "Titanium statistical compendium,"
http://minerals.usgs.gov/minerals/pubs/commodity/titanium/stat/, August 1, 2002.

Voinov, M. and Augustynski, J., "Introduction to the physics of semiconductor
photocatalysts," in: Heterogeneous Photocatalysis, Schiavello, M. (Ed.), John
Wiley & Sons Ltd., New York, pp. 1-34, 1997.

Wu, T., Liu, G., Zhao, J., Hidaka, H. and Serpone, N., "Evidence for H202 generation
during the TiO2 assisted photodegradation of dyes in aqueous dispersions under
visible light illumination," Journal of Physical Chemistry B, Vol. 103, pp. 4862-
4867, 1999.






61


Yamazaki, S., Matsunaga, S. and Hori, K., "Photocatalytic degradation of
trichloroethylene in water using TiO2 pellets," Water Resources, Vol. 35, pp.
1022-1028, 2001.

Yoneyama, H. and Torimoto, T., "Titanium dioxide/adsorbent hybrid photocatalysts for
photodestruction of organic substances of dilute concentrations," Catalysis Today,
Vol. 58, pp. 133-140, 2000.

Zhang, Y., Crittenden, J., Hand, D. and Perram, D., "Fixed-bed photocatalysts for solar
decontamination of water," Environmental Science and Technology, Vol. 28, pp.
435-442, 1994.

Zhang, Y., Weidenkaff, A. and Reller, A., "Mesoporous structure and phase transition of
nanocrystalline TiO2," Materials Letters, Vol. 3429, 2001.















BIOGRAPHICAL SKETCH

Danielle Julia Eller Londeree was born in June of 1979 in Deerfield Beach,

Florida. Her dad graduated from UF with an agricultural engineering degree and took

over the family business of manufacturing large-scale water pumps. Her brothers also

graduated from UF and followed in his footsteps, both receiving engineering degrees and

working for the family business. Their encouragement is what made Danielle want to be

an engineer, and following high school graduation from Zion Lutheran Christian School

in 1997, she pursued an engineering degree from UF. After getting married to Donald

Londeree in August of 2001, she received her B.S. in Environmental Engineering in

December of 2001. Then her thirst for knowledge and her desire to contribute to the

family business led her to pursue graduate school, focusing on water quality. In the

future, she hopes to use the knowledge that she has obtained in graduate school to

increase the scope of the family business and assist in solving the world's water

problems, on both a quantity and quality level.