1 ENHANCEMENT OF TITANIUM DIOXIDE PHOTOCATALYSIS WITH POLYHYDROXY FULLERENES By VIJAY B. KRISHNA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007
2 2007 Vijay B Krishna
3 To my parents, B.S. Krishna and B.K. Saroja , for their unwavering en couragement, support and motivation to pursue my dreams
4 ACKNOWLEDGMENTS I would like to express my sincere gratitude to my advisor Dr. Brij Moudgil and my cochair Dr. Ben Koopman for their valuable gui dance, intellectual stimulation and support throughout this study. Their virtue s of critical assessment and ra ising the bar for achievement helped me expand my potential. I would like to thank Dr. Wolfgang Sigmund for his invaluable suggestions and fruitful discussi ons. I would also like to expre ss my sincere gratitude to Dr. Seymour Block and Dr. Sam Farrah, who introduced me to the exciting field of microbiology. I would like to thank my committee members Dr. Chris Batich, Dr. Hassan El-Shall and Dr. Laurie Gower for their time, guidance and cons tructive comments. I would also like to thank National Science Foundation (NSF Grant EEC-94-02989) , Particle Engineering Research Center (PERC) and its industrial part ners for financial support. Appreciation is also extended to PERC staff members Dr. Vic Jackson, Dr. Kevin Powers, Gill Brubaker, Gary Scheiffele, Greg Norton, Jo -Anne Standridge and Donna Jackson for their assistance throughout the course of this study. I am also thankful to Dr. Kerry Siebein, Lynda Schneider, Chuck Garretson and members of Wate r Reclamation Facility for their help. I gratefully acknowledge my colleagues Scott Brown, Madhavan Esayanur, Suresh Yeruva, Kyoung-Ho Bu, Yunmi Kim, Amit Vohra, Jue Zhao, Steve Tedeschi, Parvesh Sharma, Rhye Hamey, Georgios Pyrgiotaki s, Maria Palazuelos, Anna Fu ller, Ivan Vakarelski, Marco Verwjis, Dauntel Specht, Smithi Pumprueg, Sajit Daosukho, Dushyant Shekawat, Marie Kissinger, Witcha Imaram, Gaut am Kini, Ashutosh Agrawal, and Amit Singh for their valuable insights, intellectual discussions , friendship and moral support. Finally, I would like acknowledge my parents, who encouraged me to pursue my dreams.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES................................................................................................................ .........8 ABSTRACT....................................................................................................................... ............10 CHAPTERS 1 INTRODUCTION AND BACKGROUND...........................................................................12 General Overview of Inactivation Techniques.......................................................................13 Photocatalysis with TiO2........................................................................................................16 Modification of Electronic Prope rties of Titanium dioxide...................................................18 Microbial Inactivation by Photocatalysis...............................................................................20 Structure of Bacterial Endospores and Vegetative Bacterial Cells.................................20 Mechanism of Photocatalytic Inactivation......................................................................21 Carbon Nanotubes and Fullerenes..........................................................................................22 2 MATERIALS AND METHODS...........................................................................................30 Synthesis of TiO2 Coated Multi-wall Carbon Nanotubes.......................................................30 Culturing and Purification of Bacillus Spores........................................................................31 Culturing of Escherichia coli ..................................................................................................31 Experimental Apparatus.........................................................................................................31 Photocatalytic Inactivation of Bacillus Spores.......................................................................31 Photocatalytic Inactivation of Escherichia coli ......................................................................32 Optimization of PHF Concentration w ith Dye Degradation Experiments.............................33 Adsorption Experiments.........................................................................................................34 HR-TEM Imaging of PHF-TiO2 Nanocomposite...................................................................36 Electron Paramagnetic Resonance Spectrosc opy for Hydroxyl Radical Determination........36 Synthesis of Polyhydroxy Fullerenes.....................................................................................37 Dye Degradation Experiments with Synthesized PHF...........................................................38 Characterization of PHF........................................................................................................ .38 Mass Spectroscopy..........................................................................................................38 Gaussian Modeling..........................................................................................................38 FTIR and XPS.................................................................................................................39 3 RESULTS AND DISCUSSION.............................................................................................43 Enhancement of TiO2 Photocatalysis with Carbon Nanotubes..............................................43 Optimization of Photocat alyst Concentration.................................................................44 Photocatalytic Inactivation of Bacterial Endospores.......................................................44
6 B. cereus spores........................................................................................................45 B. subtilis spores.......................................................................................................46 Photocatalytic Inactivation of Vegetative Bacterial Cells...............................................46 Effect of Size and Aspect Ratio of Photocatalyst............................................................48 Approach-1: Mutants with and without surface appendages...................................49 Approach-2: Size reduction of TiO2 coated MWNT...............................................49 Enhancement of TiO2 Photocatalysis with Polyhydroxy Fullerenes......................................50 Preliminary Dye Degradation Experi ments with Self-assembled PHF-TiO2 Nanocompo sites...........................................................................................................51 Optimization of PHF-TiO2 Nanocomposites for Photocatalysis.....................................52 Enhancement of E. coli Inactivation with PHF-TiO2 Nanocomposite............................54 Adsorption of PHF on TiO2 Nanoparticles.....................................................................54 Detection of Hydroxyl Radicals with EPR......................................................................56 Influence of Composition of Polyhydroxy Fulle renes on Photocatalytic Enhancement........59 Photocatalytic Dye Degradati on with Fresh and Aged PHF...........................................59 Mass Spectroscopy for Fresh and Aged PHF..................................................................60 FTIR Analysis of Fresh and Aged PHF..........................................................................60 Gaussian simulation.................................................................................................61 Literature values.......................................................................................................62 XPS Analysis for Fresh and Aged PHF..........................................................................62 TGA Analysis for Fresh and Aged PHF..........................................................................63 Effect of Impure Functional Groups................................................................................64 4 CONCLUSIONS AND SUGGESTIONS FOR FUTURE RESEARCH...............................97 Conclusions.................................................................................................................... .........97 Suggestions for Future Research..........................................................................................101 APPENDIX CALCULATION OF SURFACE COVERAGE................................................105 LIST OF REFERENCES.............................................................................................................108 BIOGRAPHICAL SKETCH.......................................................................................................123
7 LIST OF TABLES Table page 3-1 Zeta potential valu es of photocatalysts and E. coli ............................................................94 3-2 Peak assignments of FTIR peaks for fresh and aged PHF based on results from Gaussian simulation of C60(OH)24 and literature...............................................................95 3-3 Elemental composition of fresh and aged PHF obtained with XPS analysis....................96 3-4 Peak position and elemental composition of fresh and aged PHF obtained with XPS analysis....................................................................................................................... ........96
8 LIST OF FIGURES Figures page 1-1 Comparison of current disinfection techniques.................................................................25 1-2 Steps in photocatalytic ge neration of reactive species.......................................................26 1-3 Complex shell structur e of bacterial endospores...............................................................27 1-4 Cell-wall structure of non-motile E. coli ...........................................................................28 1-5 Photocatalytic inactivation of a bacterium.........................................................................29 2-1 High-resolution TEM image of TiO2 coated on multi-wall nanotube...............................40 2-2 Experimental setup for photocatalysis experiments..........................................................41 2-3 High resolution TEM images of PHF coated on TiO2 nanoparticles.................................42 3-1 D value estimation for photocatalytic inactivation of E. coli with Degussa P25..............66 3-2 Optimum concentration of Degussa P25 for photocatalytic inactivation of E. coli ..........67 3-3 Photocatalytic inactivation of B. cereus spores.................................................................68 3-4 Photocatalytic inactivation of B. subtilis spores................................................................69 3-5 Photocatalytic inactivation of E. coli .................................................................................70 3-6 Interaction of TiO2 coated MWNT and Degussa P25 with surface appendages of E. coli ............................................................................................................................... .......71 3-7 Two different approaches undertaken to test the hypothesis th at surface appendages sterically hinder contact of high-aspect ratio photocatalyst with bacterial cell-wall.........72 3-8 Photocatalytic inactivation of S. aureus mutants with and without surface appendages..................................................................................................................... ....73 3-9 First-order degradation kinetics of Procion Red MX-5B upon UVA irradiation with PHF, TiO2 and a mixture of TiO2 and PHF.......................................................................74 3-10 Normalized pseudo-first-or der rate coefficient for dye degradation as a function of the ratio of added polyhydroxy fullerenes (PHF) to TiO2.................................................75 3-11 Photocatalytic inactivation of E. coli plotted as a function of survival ratio vs. time.......76 3-12 D values for E. coli inactivation with Degussa P25 alone and a mixture of Degussa P25 and PHF.................................................................................................................... ..77
9 3-13 Absorption spectrum of polyhydroxy fullerenes...............................................................78 3-14 Calibration curves for polyhydroxy fullerene s (PHF) at three different pH values..........79 3-15 Adsorption density of polyhydroxy fullere nes (PHF) on titanium dioxide and shift in zeta potential of titanium dioxide nan oparticles with adsorption of polyhydroxy fullerenes at different pH...................................................................................................80 3-16 Zeta potential shifts as a function of adsorption density....................................................81 3-17 Electron paramagnetic resonance spect ra obtained upon UVA irradiation of DMPO and TiO2 alone and TiO2+PHF..........................................................................................82 3-18 Effect of PHF on generation of hydr oxyl radicals by UVA irradiation of titanium dioxide........................................................................................................................ ........83 3-19 Hypothetical photocatalytic reac tions occurring upon UV irradiation..............................84 3-20 First-order degradation kinetics of Procion Red MX-5B upon UVA irradiation with TiO2 alone, TiO2 + aged PHF and TiO2 + fresh PHF........................................................85 3-21 APCI-MS of fresh and aged Polyhydroxy Fullerenes.......................................................86 3-22 FTIR spectrum of fr esh Polyhydroxy Fullerenes...............................................................87 3-23 FTIR spectrum of ag ed Polyhydroxy Fullerenes...............................................................88 3-24 Gaussian simulation of C60(OH)24.....................................................................................89 3-25 Gaussian simulation of vibrational spectrum for C60(OH)24..............................................90 3-26 Experimental C1s XPS spectrum (top cu rve) of fresh Polyhydroxy Fullerenes with fitted curves representing three different oxidation states of carbon.................................91 3-27 Experimental C1s XPS spectrum (top curve) of aged Polyhydroxy Fullerenes with fitted curves representing three different oxidation states of carbon.................................92 3-28 TGA spectra for fresh and aged Polyhydr oxy Fullerenes. Numbers refer to different stages of weight loss..........................................................................................................93 A-1 Estimated surface coverage and observed enhancement as a function of dosed ratios of PHF to TiO2.................................................................................................................107
10 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ENHANCEMENT OF TITANIUM DIOXIDE PHOTOCATALYSIS WITH POLYHYDROXY FULLERENES By Vijay B. Krishna May 2007 Chair: Brij Moudgil Cochair: Ben Koopman Major: Materials Science and Engineering Semiconductor photocatalysts, particularly TiO2, are attracting extensive research for destruction of environmentally hazardous chemicals (e.g., organic pollutants, greenhouse gases) and hazardous bioparticulates (e.g., bacterial en dospores, emerging pathogens) because they can achieve complete mineralization without generation of toxic bypr oducts. Several attempts have been made to improve the quantum efficiency of TiO2 by conjugating it with conductors such as metals and organic molecules for scavenging th e photo-generated electro ns. Another class of materials well known for their electron accepting pr operties is carbon nanotubes and fullerenes. TiO2 (anatase polymorph) was coated on mu lti-wall carbon nanotubes by sol-gel coating and the resulting nanocomposites were found to in activate bacteria l endospores two times faster than Degussa P25 (gold standard), but were ineffective against Escherichia coli . This was attributed to their high aspect ratio, which prev ented contact with the fi mbriae covered cell-wall of E. coli . Water-soluble and non-toxic polyhydroxy fullerene s (PHF) were employed as alternate to the TiO2 coated MWNT. Adsorption of PHF molecules onto TiO2 by electrostatic interaction was demonstrated. PHF-TiO2 nanocomposites enhanced the phot ocatalytic activity of TiO2 for
11 dye degradation and E. coli inactivation. Surface coverage of TiO2 nanoparticles by PHF molecules determined the extent of enhancem ent, with an optimum at 2% surface coverage. The rate of photocatalytic dye degradation by the TiO2-PHF nanocomposite was 2.6 times the rate found with TiO2 alone. The hypothesis that scavenging of photo-ge nerated electrons a nd therefore higher generation of hydroxyl radicals is the mechanism for the obser ved enhancement was validated. The concentration of hydroxyl ra dicals generated by PHF-TiO2 nanocomposite was up to 60% greater than the concentration obtained with TiO2 alone as determined with EPR. Influence of functional groups of PHF on its electron scavenging abi lity and stability was determined. Fresh and aged forms of PHF we re characterized by MS, FTIR, XPS and TGA. Higher concentrations of impure groups were de trimental to stability and electron scavenging ability of PHF. A ratio of impure groups to hydroxyl groups of 0.27 was associated with successful enhancement by PHF, whereas a ratio of 1.66 was associated with no enhancement. Guidelines for effective formulation of PHF-TiO2 nanocomposites were developed.
12 CHAPTER 1 INTRODUCTION AND BACKGROUND Semiconductor photocatalysts are attracting extensive research for destruction of environmentally hazardous chemicals (e.g., gree nhouse gases, organic pollutants) and hazardous bioparticulates (e.g., bacterial endospores and em erging pathogens) because they can achieve complete mineralization (complete oxidation of bacterial cell to carbon dioxide and water) without generation of to xic byproducts (Hoffmann et al. , 1995; Fujishima et al. , 1999; Block, 2001; Atrih and Foster, 2002) . Titanium dioxide (TiO2) has been commercially applied as a selfcleaning coating on buildings a nd glass materials, especially in Japan, South Korea and Singapore. However, the potential for such appli cations is limited by the low quantum efficiency (10%) of TiO2 photocatalysis (Hoffmann et al. , 1995). Several attempts have been made to improve the efficiency by conjugating titanium dioxi de with metals or organic molecules. Metals such as silver, gold and platinum are either deposited on titanium dioxide particles by reduction of their salts and electron beam evaporation, or co-synthesized w ith titanium dioxide precursors (Vamathevan et al. , 2002; Arabatzis et al. , 2003a; Subramanian et al. , 2003a; Sun et al. , 2003; Sreethawong and Yoshikawa, 2005). Doping of meta ls has also been achieved with ion beam implantation. Organic compounds, which can conduc t electrons, are covalently conjugated to titanium dioxide particles for scavenging of photo-ge nerated electrons (Rajh et al. , 1999; Gratzel, 2001; Paunesku et al. , 2003; Brune et al. , 2004). All of the above mentioned approaches for enhancing photocatalysis require complex conjugation chemistry and therefore additional unit operations in the synthesis processes. Furthe rmore, contradictory resu lts have been reported in the literature for enhan cement with metals. The obser ved enhancement is usually demonstrated for organic dye degradation and se ldom for bacterial inactivation. Certain metals are also undesirable where health effects and ultim ate disposal are concerned. Therefore there is
13 a need for environmentally benign materials th at demonstrate consistent enhancement of photocatalysis for inactiva tion of microorganisms as well as degradation of organic pollutants. This chapter provides a general overview of available inactiv ation techniques and identifies the need for a new inactivation methodology based on photocatalysis. The current understanding of titanium dioxide photocatalysis and efforts for improving its quantum efficiency is discussed. Next a general picture of mechanism of photocatalytic inactivation of microorganisms is provided. Finally, the pot ential of carbon nanotube s and fullerenes as enhancers for titanium dioxide photocatalysis is provided. General Overview of Inactivation Techniques Microorganisms are ubiquitous in nature a nd are essential compone nts of our ecosystem. They are the only organisms known to exist in ex treme environments, such as at temperature as high as 150C (NSF), at alkaline and acidic pH or in saline environments (Perry et al. , 2002). Microorganisms are present in depths of ocean floor and can survive extreme conditions in outer space, which led to the theory that earth biota was originally colonized from space (Mileikowsky et al. , 2000; Horneck et al. , 2001; Mastrapa et al. , 2001). Pathogenic microorganisms are major cause of c oncern for the well being of society. Total global spending on healthcare exceeds $ 4.1 tril lion (WHO, 2007). The most common pathogen, influenza virus, accounts for $12 billion every ye ar in the U.S. for medical expenses and productivity losses (Solvay, 2007). An average loss of 6.5 working da ys is associated with the common cold (Keech et al. , 1998). The recent outbreak of bird flu virus (H5N1), an example of emerging pathogen, has cost $10 billion globally (USAID). Another area of concern is emergence of antib iotic resistant strains of microorganisms such as methicillin resistant Staphylococcus aureus , which are major contributors to healthcare-
14 associated infection. An estimated 2 million nosoc omial infections are reported annually, which results in 90,000 deaths and $4.5 bi llion in healthcar e costs (Haley et al. , 1985; CDC, 2006a). Other non-pathogenic microorganisms present es pecially in indoor environment are major contributors to allergy and respir atory problems, which are the l eading cause of school and work absence. The Center for Disease Control and Pr evention (CDC) Summary He alth Statistics for U.S. children state that nine million U.S. childre n under the age of 18 have been diagnosed with asthma. An estimated $16 billion is spent (medi cal expenses + productivity loss) in U.S. alone annually due to asth ma (LungUSA, 2005). Apart from the natural threats from pathogens present in the environment, a new cause for concern is biological warfare. The anthrax attack in the U.S. resulted in 5 deaths and 22 infections (CDC, 2006b). The U.S. budget for homela nd security related progr ams for fiscal year 2007 is $5.2 billion, indicating the se riousness of this threat. As suggested above, there is a need for de velopment of better mi crobial inactivating agents. The current global market for disinfectant s is $3.5 billion and is expected to rise to $6 billion by 2010 (Kaiser, 2006). Th e prevailing technology for inactiv ation of microorganisms can be classified into three categoriesâ€”heat, radiatio n and chemical based inac tivating agents (Fig.11). These inactivation technique s along with their shortcomings are presented below. Heat based inactivation: Thermal inactivation of microorganisms is the most commonly employed disinfection technique. In fact, heat is the oldest known method for disinfecting drinking water and is still widely used. Heat can kill most of the common disease causing waterborne bacteria and thus is a common method for tr eating water. However, there are certain microorganisms, such as bacterial endospor es, which are not easily affected by heat.
15 Furthermore, thermal techniques cannot be applie d for inactivation of all microorganisms present on surfaces (Block, 2001). Radiation based inactivation: Ultraviolet (UV) and Gamma ray radiation are the two major types of radiation utili zed for inactivation. UV radiation is commonly used for water treatment and has also been applied for inactiv ation of microorganisms in air and on surfaces. Gamma radiation is employed for sterilization of food. The major disadvant age associated with this technique is cost. Radiation at high intensity is required to ensure complete inactivation of microorganisms, which needs higher energy c onsumption (Block, 2001) . Inactivation with Gamma radiation requires separate building units as they can be hazardous for human exposure. Although no special enclosure is necessary for UV radiation, they are harmful to humans. Chemical inactivating agents: The chemical disinfectants are in widespread use and they can be further classified in to mild and strong ch emical agents. Mild chem ical inactivating agents such as those based on alcohols, aldehydes and surfactants are commonly employed for cleaning surfaces in hygiene products for washing hands. Strong chemical agents such as those based on chlorine and peroxygen compounds find applications in water treatment. Chlorine bleach is commonly employed for treating water in swi mming pools. Strong chemical agents are not recommended for treating surfaces as they can easil y oxidize them. Chemical agents are also not applicable for inactivation of microorganisms present in air a nd food. Gaseous chemical agents such as ozone, ethylene oxide and chlorine dioxide are employed as sterilants for treating medical devices. However, their cost and safety i ssues with handling restri ct their application. A major disadvantage of chemical agents is th at they generate t oxic byproducts, including mutagens and carcinogens (Block, 2001).
16 Photocatalysis is a relatively new technol ogy which overcomes the disadvantages of the above mentioned inactivating agents. Photo catalyst based technology can be applied for inactivation of microorganisms in air and wa ter and on surfaces. Semiconductors such as titanium dioxide, zinc oxide, cadmium sulfide, vanadium oxide and cerium oxide are employed as photocatalysts (Liu an d Yang, 2003; Hernandez-Alonso et al. , 2004; Karunakaran and Senthilvelan, 2005). Titanium dioxide , in particular anatase crysta l structure, has an advantage over other semiconductor photocatalysts as it has higher quantum efficiency and in addition is low cost and is generally rec ognized as being non-toxic (Zeng et al. , 2005; Wang et al. , 2006). Photocatalysis with TiO2 In 1972, Fujishima and Honda first demonstrated electrochemical photolysis of water with titanium dioxide (Fujishima and Honda, 1972). Sin ce then extensive research has been conducted on titanium dioxide photocatalysis with elucidation of its mechanism. Titanium dioxide is a semiconductor, which natu rally exists in three different crystalline formsâ€”anatase, brookite and rutile . Rutile is the most stable of three polymorphs. The calculated structure energy for rutile is more stable than anatase by 3.8 kJ mole-1 and brookite by 24.8 kJ mole-1 (Post and Burnham, 1986). Anatase and brookite are known to exist at room temperature as metastable states. In fact due to very sma ll difference between structure energy of anatase and rutile, the former can be considered as stable. Anatase structure has attracted substantial re search for photocatalysis even though it has a higher band gap of 3.2 eV compared to 3.11 eV a nd 3.0 eV for brookite and rutile, respectively (Li et al. , 2007). The higher band gap also reduces th e recombination probability and therefore anatase exhibits better photocatal ytic activity than rutile (Yan et al. , 2005; Li et al. , 2007).
17 Titanium dioxide (TiO2) has a band gap of 3.2 eV. When ultraviolet (UV) light of wavelength less than 380 nm irradiates TiO2 particles, the photon energy is sufficient for excitation of electrons from valence band to co nduction band leaving a hole behind. Various reactions occurring during photocatalys is are summarized below, where eCB represents electrons in conduction band; h+ VB represents holes in valence band; D represents electron donors and A represents electron acceptors. Electron-hole pair generation TiO2 + h eCB + h+ VB (10-15 s) (1-1) Charge localization eCB + Ti4+ Ti3+ (trapped electron) (10-10 s) (1-2) h+ VB + O2O(trapped hole) (10-10 s) (1-3) Electron-hole recombination eCB + h+ VB phonons (10-9 s) (1-4) Surface reactions TiO2 (h+) + D TiO2 + Dâ€¢ (10-7 s) (1-5) Specific examples of above reaction TiO2 (h+) + -OH TiO2 + â€¢OH (1-6) TiO2 (h+) + H2O TiO2 + â€¢OH + H+ (1-7) TiO2 (e-) + A TiO2 + A- â€¢ (10-3 s) (1-8) Specific examples of above reaction TiO2 (e-) + O2 TiO2 + O2 - â€¢ (1-9) TiO2 (e-) + H2O2 TiO2 + OH+ â€¢OH (1-10)
18 The generation of electro n-hole pairs (Eq. 1-1) is fast, occurring in 10-15 seconds (Hoffmann et al. , 1995). The photo-generated el ectrons and holes are eith er localized in shallow traps (Eq. 1-2 and 1-3) or recombine (Eq. 1-4) . Charge localization is much slower (10-10 seconds) than generation of electron-hole pairs (Berger et al. , 2005). The localized electrons and holes can recombine or take part in redox reactio ns, as shown in Figure 1-2. More than 90% of the electron-hole pairs recombine at a timescale of 10-9 seconds (Hoffmann et al. , 1995; Berger et al. , 2005). The recombination en ergy is dissipated as phonons. The remaining electron-hole pairs migrate to the surface, wh ere they react with adsorbed electron acceptors and donors (Eq. 1-5 to 1-10). The migration and interfacial charge transfer processes are the slowest of all the steps, with an overall timescale in the range of 10-8-3 seconds (Hoffmann et al. , 1995). Recombination, which is faster than charge migr ation and interfacial charge transfer, is the biggest single factor limiting the quant um efficiency of photocatalysis. Modification of Electronic Properties of Titanium dioxide Electronic property modification is usually done by creating an additional band (metastable state) near the con duction band, where the electron can remain for a longer time (Hoffmann et al. , 1995). The photon-generated electron-hole pairs recombin e at a faster rate (109 seconds) than the generati on of reactive species (10-8 to 10-3 seconds) (Hoffmann et al. , 1995). Decreasing the recombination rate, by trappi ng the photon-generated electrons or holes, will increase the production of reactive species and hence the overall efficiency of photocatalysis. Several attempts have been made to separate the photo-generated electro ns and holes to reduce recombination. Titanium dioxide photocatalysts ha ve been conjugated with electron scavenging agents such as metals or organic molecules(Hoffmann et al. , 1995; Keleher et al. , 2002; Wang et al. , 2002; Arabatzis et al. , 2003a; Arabatzis et al. , 2003b; Hu et al. , 2003c). Metals such as silver, gold and platinum are generally prefe rred due to their high conductivity and corrosion
19 resistance. They are either deposited on titanium dioxide particles by reduction of their salts and electron beam evaporation, or co-synthesized with titanium dioxide precursors (Vamathevan et al. , 2002; Arabatzis et al. , 2003a; Subramanian et al. , 2003a; Sun et al. , 2003; Sreethawong and Yoshikawa, 2005). Doping of metals has also b een achieved with ion beam implantation (Wang et al. , 2002). However, contradictor y results have been reported with metal as enhancers. Experimental investigations with goldâ€“titanium dioxide nanocompos ite particles showed that the response of photocatalyst is extended to the visi ble region with a significant decrease in their photocatalytic performance under ul traviolet radiation (Arabatzis et al. , 2003a). Silver doped titanium dioxide nanocomposite exhibited increase in degradation rate fo r sucrose, however no enhancement was observed for degradati on of salicyclic acid (Vamathevan et al. , 2002). Organic compounds, which can conduct electrons , are covalently conjugated to titanium dioxide particles for scavenging of photo-generated electrons with applications in cluding solar cells and visible light photocatalysis(Rajh et al. , 1999; Gratzel, 2001; Paunesku et al. , 2003; Brune et al. , 2004). Another class of conductors used recently is the carbon nanotube . Titanium dioxide coated multi-wall carbon nanotubes (MWNT) have been employed to increase the photocatalytic degradation of organic polluta nts and inactivation of microorganisms (Lee, 2004; Pumprueg, 2004; Krishna et al. , 2005; Lee et al. , 2005; Yu et al. , 2005a; Pyrgiotakis, 2006). It was hypothesized that the photo-gene rated electrons are scavenge d by the MWNT. Use of coadsorbents such as silica, alumina, zeolites a nd activated carbon have al so been explored for increasing the efficacy of photocatalysis (Matos et al. , 2001).
20 Microbial Inactivation by Photocatalysis The extent of inactivation was observed to be inversely proportional to the thickness and complexity of the cell wall. Therefore it is important to understand the structure of microorganisms before examining the mechanism of inactivation. Structure of Bacterial Endospores and Vegetative Bacterial Cells Endospores of bacteria are the most resistan t microorganisms agains t all disinfection and sterilization tec hniques (Nicholson et al. , 2000; Block, 2001; Atrih and Foster, 2002). Their high degree of resistance is governed by a unique spore structure, as shown in Figure 1-3. Each level of spore structure provides resist ance to different di sinfectants. The small acid soluble proteins (SASP) in the core protect the DNA from UV radi ation, whereas cortex pr ovides heat resistance and the spore coat layers protects the spore fr om chemical attack (Driks, 1999; Henriques and Moran, 2000; Nicholson et al. , 2000; Riesenman and Nichol son, 2000; Block, 2001; Atrih and Foster, 2002). Endospores of Bacillus cereus and Bacillus anthracis have an extra outermost layer called exosporium, wh ich is not present in Bacillus subtilis . Filaments are also present in case of B. cereus spores (Mizuki et al. , 1998). In contrast to bacterial endospores, th e vegetative bacterial cells such as Escherichia coli have simple cell-wall stru cture (Fig. 1-4) (Perry et al. , 2002). The cytoplasmic membrane is surrounded by a thin layer of peptidoglycan (2 nm) with periplasmi c space in between. The peptidoglycan layer is surrounded by an outer pr otein membrane (8 nm) again with periplasmic space in between. E. coli , which is a Gram negative bacterium, also possesses surface appendages. Fimbriae are long filamentous polymeric protein structures an chored inside plasma membrane (Klemm, 1994; Mol and Oudega, 1996). Fimbriae can be as long as 2 microns and can vary in diameter from 2 to 10 nm. They can be either thick and rigi d or thin and flexible,
21 however their exact structure in natural surrounding is not known. The total cell-wall thickness is approximately 20 nm. Mechanism of Photocatalytic Inactivation A better understanding of microbial inactiva tion has evolved since the first proposed mechanism by Matsunaga and co-workers (Matsunaga et al. , 1985) . They proposed that the cell death was caused by decrease in respiratory activity due to photo catalytic oxidation of intracellular coenzyme A. Saito et al. proposed that the cell death occurs due to photocatalytic disruption of cell membrane, evident from leakage of intracellular K+ ions (Saito et al. , 1992c). Leakage of Ca+ ions has also been observ ed with cancer cells. Sunada et al. found that endotoxin, an integral component of the outer me mbrane was degraded by photocatalytic action of TiO2, which leads to membrane damage (Sunada et al. , 1998). Maness et al. (Maness et al. , 1999) showed that TiO2 photocatalytic reaction causes the lipid peroxidation reaction, which results in disruption of normal ac tivities associated with intact cell membrane, such as respiration (Jacoby et al. , 1998). The loss of membrane structure was proposed to be the cause of cell death. Lu et al. also showed that cell death was caused by the decomposition of th e cell-wall and cellmembrane resulting in leakag e of intracellular components, as show n in Figure 1-5 (Lu et al. , 2003). Since the cell death is caused by photocatalytic degradation of cellwall, the inactivation time is proportional to the complexity and density of cell wall structure (Kuhn et al. , 2003). Endospores, with their complex and dense shell st ructure, have the longest inactivation time (in hours compared to minutes for simple bacteria ). The photocatalytic in activation time can be decreased by increasing the generation of reac tive species, which can be achieved by delaying the recombination process.
22 Carbon Nanotubes and Fullerenes Fullerenes and carbon nanotubes are allotrope s of carbon with unique properties. The existence of fullerenes was first proposed by Osawa in 1970 (Osawa, 1970). Although Rohlfing and coworkers were first to spectroscopically determine the existence of fullerenes (Rohlfing et al. , 1984), Kroto et al. , were the first to synthesize and purify fullerenes, which they named Buckminsterfullerenes (Kroto et al. , 1985). The presence of carbon nanotubes was first observed by Iijima in 1991 (Iijima, 1991). Since then carb on nanotubes and fullerenes have attracted substantial research on their physical, el ectronic and chemi cal properties. Carbon nanotubes, especially single-wall nanotubes, are known to possess metallic as well as semiconducting properties depending on the angle of folding graphene sheets (Saito et al. , 1992a; Saito et al. , 1992b). Multi-wall carbon nanot ubes (MWNT) are generally semiconducting, with their outermost layer sometime s dictating the electroni c properties. In fact MWNT are speculated to be the first above r oom temperature superconductor (Zhao and Wang, 2001). The unique electronic propertie s of carbon nanotubes have b een researched for electronic applications, as reviewed by Anantram a nd Leonard (Anantram and Leonard, 2006). Carbon nanotubes have been employed as electron accepto r for bio and nanobio-sensor applications (Guiseppi-Elie et al. , 2002; Chen et al. , 2003; Goswami et al. , 2004). Kymakis and Amaratunga exploited the electron scavengi ng ability of singlewall carbon nanotubes in photovoltaic cells (Kymakis and Amaratunga, 2003). MWNT have al so been employed for enhancing titanium dioxide photocatalysis (Lee, 2004; Pumprueg, 2004; Krishna et al. , 2005; Lee et al. , 2005; Yu et al. , 2005b; Pyrgiotakis, 2006). Fullerenes, which are spherical version of na notubes, have also been exploited for their unique electronic, optical and magne tic properties (Guldi, 2000b; Gust et al. , 2000; Makarova, 2001). Fullerenes, specifically C60, which have truncated icosahed ra structure, possess three-
23 dimensional symmetry. Fullerenes have also be en reported as three dimensional electron acceptors (Guldi, 2000a). Fullerenes were experiment ally shown to accept up to six electrons per molecule (Xie et al. , 1992). Fullerenes can accumulate and discharge electrons, thus acting as electron relays (Guldi and Prat o, 2000). Electrochemical sensor s based on fullerenes and carbon nanotubes have been reviewed by Sh erigara and coworkers (Sherigara et al. , 2003). Kamat and coworkers have demonstrated the transfer of phot o-generated electrons from titanium dioxide to fullerenes with ethanol/benzene mixtur e as solvent (Kamat et al., 1994). Fullerenes are not soluble in water and are al so reported to be toxi c (Oberdorster, 2004), limiting their potential for many applications. The water-solubility of fullerenes is improved by coupling the fullerene cage with hydrophilic mol ecules (Taylor, 1999). Fu nctionalized fullerenes such as 6,6-phenyl-C60-butyric acid (PCBM) has been successfully employed as an electron acceptor in solar cells (McNeill et al. , 2004). Self-assembled nanocomposites of functionalized water-soluble fullerenes with CdTe nanoparticle s have been prepared, in which fullerenes scavenge electrons from CdTe nanoparticles (Guldi et al. , 2004). These functi onalized fullerenes usually possess a few hydrophilic chains, and thus are somewhat water soluble. Fullerenes have also been functionalized with 12 hydroxyl groups per molecule. The advantage of hydroxylated fullerenes is that they are reported to be non-toxic and have been researched for antioxidant applications. Po lyhydroxy fullerenes were reported to reduce oxidative stress on cells by scave nging reactive oxygen species (Chen et al. , 2004; Djordjevic et al. , 2004b). Furthermore, they are patented as therapeutics (US5994410) and are used in cosmetics (www.vc60.com ). As mentioned earlier, carbon nanotubes a nd fullerenes are known for their unique electronic properties including electr on accepting ability (Guldi, 2000a; Chen et al. , 2003;
24 Sherigara et al. , 2003; Guldi et al. , 2005). Their potential, howeve r, is restricted by limited solubility in water. Chemical functionaliza tion of nanotubes and fullerenes can promote their dispersal in aqueous system and thus increase their range of applica tion. However, loss in electron scavenging ability of fullerenes upon hydroxylation has been reported (Mohan et al. , 1997), questioning their applicab ility for enhancing photocatal ysis. Although carbon nanotubes have been employed for enhancement of titanium dioxide photocatalysis (Lee, 2004; Pumprueg, 2004; Krishna et al. , 2005; Lee et al. , 2005; Pyrgiotakis, 2006), no previous attempts have been made to harness the electron-scavenging ability of fullerenes as a means of enhancing photocatalysis. Accordingly, the goal of the present research was to develop photocatalytic nanocomposites based on electron accepting properti es of fullerenes. The specific objectives were fourfold: (1) verify if carbon nanot ubes could enhance the activity of TiO2 against microorganisms, (2) determine if hydr oxylated fullerenes could enhance TiO2 photocatalysis and avoid steric hindrance encountered with TiO2-carbon nanotube nanocomposites, (3) confirm mechanism of photocatalytic enhancement with PHF, and (4) evaluate effect of PHF composition on its performance. The results obtained from accomplishing the four objectives will be further analyzed to develop a set of design guidelines for enhancing TiO2 photocatalysis for app lications of societal significance, such as coatings, pollution cont rol, healthcare and control of confined environments.
25 Figure 1-1. Comparison of curre nt disinfection techniques. Surfaces Food Toxic by products Chemical agents (Mild) (Strong) Toxic by products High intensity reqd., expensive Some microbes are resistant Disadv. Air Water Photocatalysis Radiation Heat Systems Surfaces Food Toxic by products Chemical agents (Mild) (Strong) Toxic by products High intensity reqd., expensive Disadv. Air Water Photocatalysis Radiation Heat Systems Slow Surfaces Food Toxic by products Chemical agents (Mild) (Strong) Toxic by products High intensity reqd., expensive Some microbes are resistant Disadv. Air Water Photocatalysis Radiation Heat Systems Surfaces Food Toxic by products Chemical agents (Mild) (Strong) Toxic by products High intensity reqd., expensive Disadv. Air Water Photocatalysis Radiation Heat Systems Slow Surfaces Food Toxic by products Chemical agents (Mild) (Strong) Toxic by products High intensity reqd., expensive Some microbes are resistant Disadv. Air Water Photocatalysis Radiation Heat Systems Surfaces Food Toxic by products Chemical agents (Mild) (Strong) Toxic by products High intensity reqd., expensive Disadv. Air Water Photocatalysis Radiation Heat Systems Slow
26 Figure 1-2. Steps in photocatalytic generation of reactive species. A.A D.DTiO2h Conduction Band Valence Band eh+ 10-15 s10-3 s 10-7 s10-7 s A.A D.DTiO2h Conduction Band Valence Band eh+ 10-15 s10-3 s 10-7 s10-7 sA â€“e-acceptors D â€“e-donors .OH -OH .-O2 O2 A.A D.DTiO2h Conduction Band Valence Band eh+ 10-15 s10-3 s 10-7 s10-7 s A.A D.DTiO2h Conduction Band Valence Band eh+ 10-15 s10-3 s 10-7 s10-7 sA â€“e-acceptors D â€“e-donors .OH -OH .-O2 O2
27 Figure 1-3. Complex shell stru cture of bacterial endospores. Exosporium Outer Spore CoatInner Spore Coat Cortex Core DNA SASP Germ Cell-wall Exosporium Outer Spore CoatInner Spore Coat Cortex Core DNA SASP Germ Cell-wall (200 nm) (20 nm) (40 nm) Exosporium Outer Spore CoatInner Spore Coat Cortex Core DNA SASP Germ Cell-wall Exosporium Outer Spore CoatInner Spore Coat Cortex Core DNA SASP Germ Cell-wall (200 nm) (20 nm) (40 nm)
28 Figure 1-4. Cell-wall structure of non-motile E. coli . Plasma membrane(7 nm)Periplasmic space Peptidoglycan layer (2 nm)Outer membrane(8 nm)Fimbriae (0.2.0 m length; 2â€“10 nm thickness) Cytoplasm Membrane proteins Phospholipids Porin Lipopolysaccharides Plasma membrane(7 nm)Periplasmic space Peptidoglycan layer (2 nm)Outer membrane(8 nm)Fimbriae (0.2.0 m length; 2â€“10 nm thickness) Cytoplasm Membrane proteins Phospholipids Porin Lipopolysaccharides
29 Figure 1-5. Photocatalytic in activation of a bacterium. Bacterium TiO2nanoparticle UV Light Bacterium Leakage of intracellular constituents LysedCellCell membrane disruption Bacterium TiO2nanoparticle UV Light Bacterium Leakage of intracellular constituents LysedCellCell membrane disruption
30 CHAPTER 2 MATERIALS AND METHODS Synthesis of TiO2 Coated Multi-wall Carbon Nanotubes Commercially available arc-discharged multiwall carbon nanotubes (Alfa Aesar, 3 nm outer diameter, 0.5 m) were used as base material . Multi-wall carbon nanotubes (MWNT) have unique electronic propertie s that can trap photon-genera ted electrons. Although electron transfer from titanium dioxide nanoparticle to uncoated carbon nanotubes is feasible, covalent linkage increases the efficiency of electron transfer and hence photocatalysis. Covalent linkage of multi-wall carbon nanotubes with titanium dioxi de is also desirable for strong and stable coating. Since carbon nanotubes have smooth walls w ithout any defects, it is necessary to add functional groups on the surface, which can act as reaction sites for sol-gel coating of titanium dioxide. The nanotubes were functionaliz ed by chemical oxidation method (Tsang et al. , 1994). Functionalization was carried out by dispersing 100 mg of MWNT in HNO3, using 30 minute sonication, followed by refluxing at 140 C for 10 hours. The functiona lization process involves oxidation of surface to yield â€“COOH, >C=O, â€“O H groups on the surface of MWNT, as also determined by FTIR (Hiura et al. , 1995; Ebbesen et al. , 1996). After functionalization step, MWNTs were washed several times with deionised water until no residual acids in the wash water were detected. In order to coat f unctionalized surface with titanium dioxide, 20 L of Titanium(III) sulfate (99.9+%) solution was added to MWNTs, dispersed in deionized water, with constant stirring for 1 hour. The TiO2 coated MWNTs were then washed and dried at 60 C for 2 days. The crystallization of titanium di oxide coating was done by heat treatment at 500 C. Figure 2-1 shows high resolution transmission elect ron micrograph of anatase coated multi-wall nanotubes (Lee, 2004; Krishna et al. , 2005; Lee et al. , 2005).
31 Culturing and Purification of Bacillus Spores Endospores of Bacillus cereus ATTC 2 was used as a surrogate of Bacillus anthracis . The spores were cultured according to ASTM E2111 protocol. Presence of vegetative cells can induce artifacts in the experiments, due to th eir higher susceptibility towards photocatalytic inactivation. Spores were harv ested and purified using the ly sozyme treatment (Harwood and Cutting, 1990). The lysozyme treatment was foll owed by heat shock treatment (80C, 10 min) for complete killing of vegetative cells. Finally spores were suspended in sterile deinoized water and refrigerated at 4C until further use. Culturing of Escherichia coli E. coli C3000 was cultured for 24 hours at 37C a nd harvested to give a stock solution according to the protocol of Rincon and Pulgarin (Rincon and Pulgarin, 2003). E. coli was stored in deionized water at 4C for a maximum of 24 hours. Experimental Apparatus Dye degradation experiments were performed in side an ultraviolet (UV) chamber with 16 solar UV lamps (Southern New England Ultra Vi olet Company, Branfield, CT). Cold air was passed over the reaction mixtures to limit heating. The surfaces of the reaction mixtures were positioned 100 mm below the lamps. A multiplate magnetic stirrer was used to provide agitation. Photocatalytic Inactivation of Bacillus Spores The photocatalytic inactivation efficiency co mparison of anatase coated multi-wall carbon nanotubes (MWNT) and commercial titanium dioxi de nanopowders (Degussa P25) was done on surface area basis. The surface area of anatase coated MWNT and Degussa P25 were found to be 172 m2 g-1 and 50 m2 g-1 respectively, by BET analysis (Pumprueg, 2004; Krishna et al. , 2005;
32 Lee et al. , 2005). 0.01% of commercial tit anium dioxide (Degussa P25) was used for disinfection experiments, as this is shown to be the optimum concentration by Block et al . (1997). The different reaction slurries pr epared are listed below. Control â€“ 20ml of water + 10ml of spore suspension Commercial titanium dioxide (Degussa P25) â€“ 3mg of commercial TiO2 P25 in 20ml of water + 10ml of spore suspension TiO2 coated multi-wall carbon nanotubes â€“ 0.8mg of TiO2 coated nanotubes in 20ml of water + 10ml of spore suspension. Functionalized multi-wall carbon nanotubes â€“ 0.8 mg of functionalized nanotubes in 20ml of water + 10ml of spore suspension. The suspension was transferred to a sterile Pe tri dish containing a st erile magnetic stirring bar. The Petri dish containing th e suspension was then exposed to ultraviolet (UV) light with sixteen 350 nm UV lamps. A cooling fan was used to avoid increase in temperature of the suspension. The experimental set up is shown in Figure 2-2. The UV lamps were stabilized for 30 min prior to the start of the experiment to obt ain constant intensity. Sa mples were collected at regular intervals of time, includ ing a sample at the start of the experiment. The samples were diluted appropriately and plated to determine the survival concentration. The results were analyzed by plotting the survival ratio against UV irradiation time. D values are determined as the time required for one log reduction in exponential portion of the curve and LD90 is the time required for 90% reduction of population (Block, 2001). Photocatalytic Inactivation of Escherichia coli The photocatalytic inactivation experiment s were conducted with Degussa P25. The photocatalyst suspension was prepared at a concentration of 30 mg L-1 and sonicated (Misonix Sonicator 3000, Farmingdale, NY) for one hour at 165 W. Subsequently 0.03 mg/L of PHF was added to 20 mL of the photo catalyst suspension followed by addition of 10 mL of stock E. coli
33 suspension. The initial concentration of E. coli in the reaction mixture was 3.67 cfu mL-1. The reaction mixture was stirred us ing a magnetic stirrer, with i rradiation by UVA at intensity of 86 W m-2. Three 0.333 mL samples were collected ev ery 15 minutes for one hour. The collected samples were then serially diluted with phospha te buffered saline. Appr opriate dilutions were plated on tryptic soy agar and incubated at 37 C for 20 hours prior to counting. Semilog graphs of survival ratio (concentration of E. coli at any given time normalized with initial concentration) vs. time were plotted to determine the first-order inactivation rate coeffici ents and D-value (time required for one log reduction in linear region of the graph). Optimization of PHF Concentration with Dye Degradation Experiments Titanium dioxide was dispersed in deionized water as described previously to give a concentration of 30 or 100 mg L-1 in the reaction mixture. I mmediately after sonication, PHF was added, followed closely by addition of dye. Different dosed weight ratios of PHF to titanium dioxide (0, 0.0002, 0.001, 0.003, 0.005, 0.05 and 0.093) were tested. The initial dye concentration in the reaction mixture was 3 mg L-1. The pH of the reaction mixture was approximately 6. The reaction mixture was poured into a 35 mm Petri dish and mixed for 10 minutes in the dark using a magnetic stirrer. Two 0.5 mL samples were collected and pipetted into a cuvette. The UV lamps were then turned on for the duration of the experiment. Samples were collected every 15 minutes for one hour. Becau se of the need to process multiple samples, the titanium dioxide was separated by sediment ation in the dark over a 10-day period. Reaction mixture containing a representativ e PHF concentration (150 g L-1) and no titanium dioxide was carried through the test procedur e as a control. The supernat ants were analyzed for dye concentration by UV-Vis spectroscopy at peak heights at 512 and 538 nm. The pseudo-firstorder rate coefficients for photoc atalytic degradation of Procion red dye were determined from
34 semi-log plots of normalized absorbance (absorban ce of dye solution at any given time divided by the dye absorbance at time zero) versus time. The statistical significance of adsorption and dye degradation re sults were analyzed by the Tukey HSD and Tukey-Kramer tests at = 0.01(Sokal and Rohlf, 1997; Montgomery, 2005). Adsorption Experiments Polyhydroxy fullerenes (C60(OH)n, n=18) were purchased from BuckyUSA (Houston, TX). Titanium dioxide (anatase polymorph, 5 nm particle size) was obtained from Alfa Aesar (Ward Hill, MA). All other chemicals were acquired from Fisher Scientific (Hampton, NH). The triazine monoazo organic dye, Procion Red MX-5B, was used in dye degradation studies. Stock solution of the organic dye was formul ated at a concentration of 100 mg L-1 with sterile deionized water. A stock solu tion of polyhydroxy fullerenes was prepared by dissolving 1.4 mg of C60(OH)n in 10 mL of sterile deionized water. The concentration of PHF was measured by UV absorbance (Perkin-Elmer Lambda 800, Wellesley, MA). In order to construct calibrati on curves, PHF solutions were prepared in supernatants of titanium dioxide suspensions to be consistent with the methodology of adsorption experiments and to verify the efficacy of centrif ugation step. Each tita nium dioxide suspension (50, 100, or 200 mg L-1) was prepared in deionized water. The pH of the suspension was adjusted with HCl or NaOH solution so that a separate cali bration curve could be prepared for each target pH value (3.3, 6.1, 9.9) used in the adsorption experiments. The suspension was sonicated at 165 mW in a water bath sonicator (Misonix Soni cator 3000, Farmingdale, NY) for one hour and the pH was measured again. Titanium dioxide was separated from the suspension by centrifugation at 15,000g for 15 minutes. Supernatant was coll ected and centrifuged again for 25 minutes. The final supernatant was found to be free of tit anium dioxide particles based on results from inductively coupled plasma (Perkin-Elmer Plas ma 3200, Wellesley, MA). PHF was added to the
35 supernatants to give desired fi nal concentrations. Calibration curv es (PHF absorbance at 254 nm vs. concentration) were constructe d at the selected pH values. Re sults of experiments carried out to determine if the concentration of titanium dioxide had an effect on the calibration curves indicated that initial con centration of titanium dioxide in the range of 50 mg L-1 had no effect (data not shown). Prior to an adsorption experiment; a su spension of titanium dioxide (100 mg L-1) was prepared as described previously at a selected pH. Immediatel y after sonication, PHF was added to the suspension to give a final concentration of 10 mg L-1 titanium dioxide and the suspension was mixed on an orbital shaker at 200 rev/min and 35 C for 12 hr. Zeta potential and pH of the suspensions were measured before addition of PHF and after completion of the agitation period. Zeta potential measurements were performed with Brookhaven ZetaPlus (Holtsville, NY). To separate unadsorbed PHF from the suspensi on, the mixture was centrifuged as described previously. The concentration of PHF in the supernatant was determined by UV spectroscopy ( = 254 nm). The quantity of adsorbed PHF was th en calculated as the difference between the dosed quantity and the amount remaining in th e supernatant. The adsorption density was calculated as the number of molecules of PHF adso rbed per unit surface ar ea of titanium dioxide nanoparticles. The surface area of titanium dioxide nanoparticles was determined by BET (Quantachrome Autosorb 1C-MS, Boynton Beach, FL). Surface coverage of titanium dioxide by PHF was calculated using a PHF diameter of 1.3 nm(Ozawa, 1995), whic h gives a projected surface area of 1.33 nm2 per molecule. The extent of adso rption was estimated from adsorption results at neutral pH. Control experiments were conduc ted without titanium dioxide.
36 HR-TEM Imaging of PHF-TiO2 Nanocomposite Samples for high resolution transmission el ectron microscopy (HR-TEM) were prepared by sonicating 30 mg L-1 of TiO2 suspension for one hour, followed by ten fold dilution of the suspension. The diluted suspension was again s onicated for 30 minutes and PHF was added to the suspension to give 3 mg L-1 of TiO2 and 0.3 mg L-1 of PHF in the final suspension. A 10 L drop of suspension was pipetted on top of TEM grid (Lacey carbon grids, EMS, Hatfield, PA) and allowed to dry overnight. The samples were imaged with field emission HR-TEM (JEOL 2010F) at 15 kV accelerati ng voltage. The PHF-TiO2 ratio was chosen to be 100 times higher than that employed for photocatalysis experiment s to increase the probabi lity of detection of PHF on TiO2 surface. HR-TEM images of PHF coated on TiO2 are presented in Figure 2-3. TiO2 is in the form of agglomerates of 20 nm size. A single crystal size of 5 nm is apparent from other images (not shown). PHF is present as clus ters of 2 nm with apparent lattice diffraction. Electron Paramagnetic Resonance Spectrosc opy for Hydroxyl Radical Determination EPR experiments were carried out to investig ate the radical generati on capability of PHFTiO2 nanocomposite. EPR spectra were recorded at room temperature using a commercial Bruker Elexsys E580 spectrometer employing Br ukerâ€™s high-Q cavity, ER 4123SHQE, and using quartz capillaries of approximately 1x2 mm (IDx OD). Spectral parameters were typically 100 kHz modulation frequency, 1 G modulation amplitude, 2 mW microwave power, 9.87 GHz microwave frequency, 20.48 ms time cons tant and 81.92 conversion time/point. A photocatalyst-spin trap mixture was ma de up by combining 190 L of photocatalyst suspension containing 30 mg L-1 TiO2 and 0.03 mg/L of PHF with 10 L of 5,5-dimethyl-1pyrroline N-oxide (DMPO) (Alexis) and 6 quartz cap illaries were filled with the mixture. EPR spectra were obtained before a nd after a predetermined time of exposure of each tube to UVA lamps (13 W m-2). To quantify the concentration of hydr oxyl radicals generated, the integrated
37 peak area of the DMPO-OH spectra was compared with the in tegrated peak area of a spectrum using 10 M 4-hydroxy-2,2,6,6-tetramethylpiperidi nyloxy radical (HTEMPO) (Fluka) at each UVA irradiation time (Uchino et al., 2002). Analysis and double in tegration of the spectra was performed with Grace, version 5.1.20, a public doma in 2D-plotting and data analysis package released under the GNU public license [http ://plasma-gate.weizmann.ac.il/Grace/]. EPR experiments were also carried out to in vestigate the radical ge neration capability of PHF alone. Procedure was same as described above except that TiO2 was omitted from the reaction mixture. Control experi ments with neither PHF nor TiO2 present were performed in the dark and UVA irradiation. Synthesis of Polyhydroxy Fullerenes Polyhydroxy fullerenes (PHF) were synthesized via alkali route simila r to Li et al. A solution of fullerenes was prep ared by adding 80 mg of C60 (95%, BuckyUSA, Houston TX) to 60 mL of benzene (HPLC grade, Fisher). A mixt ure of 2 mL of NaOH solution (1 g/mL) and 0.3 mL of tetra butyl ammonium hydroxide (40% solu tion) was prepared in a separate Erlenmeyer flask. The fullerene solution was added to the al kali-surfactant solution under vigorous stirring. After 30 minutes, the stirring was stopped and the mixture was allowed to phase separate. The top clear phase was decanted and remaining sl urry was stirred with additional 12 mL of deionized water for 24 hours. The mixture was then filtered through What man 40 filter paper and the filtrate was concentrated to 5 mL in a vac uum oven at 60C. The resultant slurry was washed four times with 50 mL of methanol by alternate centrifugation (5000 g, 10 min) and resuspension. After the final wash, PHF were su spended in 20 mL of methanol and dried under vacuum at 60C. The mass of PHF obtained was 120 mg.
38 Dye Degradation Experiments with Synthesized PHF Dye degradation experiments were conducted wi th anatase (5 nm, Alfa-Aesar) titanium dioxide as the photocatalyst. A photocatalyst su spension was prepared by sonicating 30 mg L-1 of anatase for 1 hour. PHF was added to the susp ension to give a final concentration of 0.03 mg L-1. Procion Red MX5B was then added to the photocatalyst suspensi on to give a final concentration of 3 mg L-1. The reaction mixture was transferre d to a Petri dish with a magnetic stirrer and placed 100 mm below a bank of 16 UVA lamps (Southern New England Ultra Violet Company, Branfield, CT). The mixture was stirred in the dark for 10 minutes and then exposed to UVA at 86 W m-2 intensity. Immediately prior to turn ing on the lights, two 1.5 mL samples were pipetted into a plastic vial. Subsequent samp les were collected at 15 min intervals for one hour. Each collected sample was centrifuge d twice at 10,000 g for 15 min. and the final supernatant was transferred to a plastic (PMME) cuvette. The UV-Vis spectrum was then obtained, with absorbance at 512 nm and 538 nm us ed for data analysis. The log of normalized sample absorbance was plotted vs. irradiation ti me and the slopes measured to obtain pseudofirst order degradation coefficients. Characterization of PHF Mass Spectroscopy PHF samples were analyzed with atmospheri c pressure chemical ionization (APCI) mass spectroscopy (ThermoFinnigan, San Jose, CA). Gaussian Modeling Structure optimization and vi bration frequency prediction for PHF was performed with Gaussian software. To simplify the computati on, PHF was assumed to have 24 hydroxyl groups based on information from the supplier. Beckestyle 3-Parameter Density Functional Theory
39 (using the Lee-Yang-Parr correlation function) 6-31 G* basis set was employed. B3LYP is a hybrid of density functional and Ha rtree-Fock theory (Foresman, 1996). FTIR and XPS Both fresh and aged form of PHF was char acterized by Diffuse Reflectance Infrared Fourier Transform (DRIFT) and X-ray Photoele ctron Spectroscopy (XPS). DRIFT experiments were carried out with Thermo Electron Magna 760 with potassium bromide as background. XPS experiments were performed with Kratos Analyt ical Surface Analyzer XSAM 800 in survey and multiplex mode. The C1s spectrum was subjected to peak fitting analysis with Grams 7.01 software (Thermo Fisher Scientific, Waltham, MA) to determine the various oxidation states of carbon.
40 Figure 2-1. High-resolution TEM image of TiO2 coated on multi-wall nanotube. Anatase coated MWNT MWNT TiO2 Anatase coated MWNT MWNT TiO2
41 Figure 2-2. Experimental setup fo r photocatalysis experiments. UV Lamps Magnetic stirrer Spore/E. coli suspension in Petri dish Fan Fan UV Lamps Magnetic stirrer Spore/E. coli suspension in Petri dish Fan Fan
42 Figure 2-3. High resolution TEM images of PHF coated on TiO2 nanoparticles. 5 nm 5 nmTiO2agglomerate PHF PHF Lattice Diffraction 5 nm 5 nmTiO2agglomerate PHF PHF Lattice Diffraction clusters 5 nm 5 nmTiO2agglomerate PHF PHF Lattice Diffraction 5 nm 5 nmTiO2agglomerate PHF PHF Lattice Diffraction clusters
43 CHAPTER 3 RESULTS AND DISCUSSION The present work explores ut ilization of electron accepting character of carbon nanotubes and fullerenes for scavenging photo-generated elect rons from titanium dioxide for accelerating photocatalysis. The photocatalytic activity of carbon nanotubes coated with titanium dioxide was studied initially. This was followed by asse ssment of self-assembled nanocomposites of polyhydroxy fullerenes and titanium dioxide. Phot ocatalytic activity and mechanisms of interaction and enhancement were characterized . Finally, the influence of the composition of polyhydroxy fullerenes on electron sc avenging ability was evaluated. Enhancement of TiO2 Photocatalysis with Carbon Nanotubes Carbon nanotubes are known for their unique electr onic properties. They have been employed as electrodes for sensing applications (Popov, 2004; Guldi et al. , 2005; Anantram and Leonard, 2006). Guldi et al. has also developed photoelectrochemi cal devices based on electron accepting ability of carbon nanotubes (Guldi et al. , 2005). This ability was exploited in the present study to scavenge photo-generated electrons from titanium dioxide. Photocatalytic nanocomposites were prepared by coating titanium dioxide on func tionalized multi-wall carbon nanotubes (Lee, 2004; Lee et al. , 2005; Pyrgiotakis, 2006). Th e photocatalytic activity of nanocomposite was tested against the best commercially available phot ocatalystâ€”Degussa P25, on equal surface area basis, since contact area between the photocatal yst particle and microbi al surface governs the inactivation process. The concentration of Degus sa P25 was first optimized for inactivation of vegetative bacterial cells. The phot ocatalysts were then tested for their efficacy to inactivate bacterial endospores, which are the most resistant life forms on the planet. This was compared to performance against vegetative bacterial cells. Fi nally, size and aspect ra tio of photocatalyst was proposed as an important factor in th e photocatalytic inactivation process.
44 Optimization of Photocatalyst Concentration The optimum concentration of Degussa P25 was determined by inactivating E. coli with different concentrations of photocatalyst. The performance of the photocatalyst was characterized in terms of D value, which is th e time required for one log reduction in viable cell concentration. A typical example for D valu e estimation is presented in Figure 3-1. Photocatalytic inactivation of E. coli was carried out with Degussa P25 at 10, 50, 100, 1000 and 5000 mg L-1 concentrations. The D values for different concentrations of Degussa P25 are plotted in Figure 3-2. Initially D value decreases with increase in concentration of photocatalyst as expected, with the minimum of 4 min observed at the concentra tion of 100 mg L-1. Further increase in concentration of Degussa P25 appears to exhibit a trend towards higher D values. Increase in D value at higher concentrat ions of photocatalyst can be attributed to multilayer photocatalyst coating of bacteria, wh ich blocks the UVA irra diation activating the photocatalyst layer in contact with the bacter ial surface. The shielding phenomenon has been previously reported by Block et al. (1997). The optimum concentr ation of Degussa P25 at 100 mg L-1, which is similar to reported values (Onoda et al. , 1988; Block et al. , 1997), was chosen for subsequent experiments involving photocatalytic inactivation of bacter ial endospores as well as vegetative bacterial cells. Photocatalytic Inactivation of Bacterial Endospores Bacterial endospores are the most resistant microorganisms against all disinfection techniques. Bacillus cereus and Bacillus subtilis were employed as surrogates for Bacillus anthracis , which causes anthrax. Bacillus cereus is very close to Bacillus anthracis in genetic makeup (Ivanova et al. , 2003; Read et al. , 2003) and has thus become popular as a surrogate. Bacillus subtilis spores have previously been employed as a surrogate for B. anthracis spores
45 (Nicholson and Galeano, 2003). TiO2 coated MWNT nanocomposites were compared with Degussa P25 on the basis of equal surface areas. The concentration of Degussa P25 was 100 mg L-1, as determined by optimization experiments. The BET surface area of TiO2 coated MWNT was three times that of the Degussa P25 and therefore the mass concentration of 33 mg L-1 of TiO2 coated MWNT was employed to give equal surface area. B. cereus spores The D values for inactivation of B. cereus spores are presented in Figure 3-3. The B. cereus spores are susceptible to UVA irradiation with a D value of 170 min (Control). Photocatalytic inactivation with Degussa P 25, at an optimum con centration of 100 mg L-1, has a D value similar to UVA inactivation. The time requi red for spore inactivation is significantly longer as compared to that for a bacteria, such as E. coli , which has a D value of ~ 5 minutes (Lu et al. , 2003). The longer time for photo catalytic inactivation of spor es can be explained on the basis of its complex and thick shell structure (~ 2 00 nm), which needs to be degraded for inactivation. It has been reporte d that increase in complexity and density of cell wall structure increases the photocatalytic inactivation time (Zheng et al. , 2000; Kuhn et al. , 2003). Photocatalytic inactivation experiments using TiO2 coated multi-wall carbon nanotubes yielded D value of 72 minutes for inactivation, which is ha lf that of best commercial titanium dioxideâ€” Degussa P25 (Lee, 2004; Pumprueg, 2004; Krishna et al. , 2005; Lee et al. , 2005). Experiments conducted with functionalized and uncoated MW NT yielded a D value of 240 min, which is higher than UVA inactivation. The functionali zed MWNT can shield spores from UVA irradiation thereby increasing the time required fo r photolytic inactivation and thus D value. The results indicate that anatase coated multi-wa ll carbon nanotubes are able to increase the photocatalysis efficiency, which is hypothesized to be due to delay in recombination process thereby increasing genera tion of reactive species.
46 B. subtilis spores A single set of inactivation expe riments were conducted with B. subtilis spores. The D values obtained for photocatalytic inactivation experiments are presented in Figure 3-4. The D value for UVA inactivation was 240 min. This is higher than that obtained for B. cereus spores. Photocatalytic inactivation with Degussa P25 gave a D value of 90 min, which suggests that B. subtilis spores are susceptible to pho tocatalysis with P25, whereas B. cereus spores are not. The higher resistance of B. cereus spores compared to B. subtilis spores toward s photocatalytic inactivation could be hypothesized to re flect the presence of exosporium. TiO2 coated MWNT exhibited a D value of 56 min, indicating that these photocatalytic nanocomposites inactivate B. subtilis spores faster than Degussa P25. Photocatalytic Inactivation of Vegetative Bacterial Cells Based on the encouraging results obtained fr om photocatalytic inactivation of bacterial endospores with TiO2 coated MWNT, experiments were carr ied out on bacterial cells, which are generally much more susceptible to disinfectants than endospores. Escherichia coli , which are common enteric bacteria, were selected as a model microorganism for inactivation studies. E. coli are rod shaped Gram negative bacteria that have a cell membrane comprising an inner membrane, a middle peptidoglycan layer a nd an outer cell wall. The cell-wall of E. coli is about 10 times thinner than that of bacterial endospores . Thus the photocatalyt ic inactivation time is expected to be shorter. The D value for photocatalytic inactivation of E. coli with commercial titanium dioxide nanoparticles (Degussa P25) was found to be 4 min (Fig. 3-5), wh ich is in agreement with the literature (Lu et al. , 2003; Rincon and Pulgarin, 2003). This illustrates the greater susceptibility of E. coli compared to endospores, consideri ng that Degussa P25 had no effect on B. cereus
47 endospores and acted 20 times more slowly against B. subtilis spores. In contrast, no statistical difference was observed between inactivation of E. coli with TiO2 coated MWNT and the control containing no photocatalyst. These results suggests that TiO2 coated MWNT were not active against E. coli . Failure of TiO2 coated MWNT to inactivate E. col i can be rationalized on the basis of mechanism of photocatalytic inactivation. The fi rst crucial step involved in the process of photocatalytic inactivation of a bacterium, as discussed in Ch apter 2, is the contact of the photocatalyst with cell-w all of the bacterium. It is hypothe sized that limited or no contact of TiO2 coated MWNT with cell-wall of E. coli is the cause for lack of photocatalytic inactivation. Insufficient contact of TiO2 coated MWNT with cell-wall can be due to either electrostatic repulsion or steric repulsion. Electrostatic interactions were investig ated by measuring zeta potential for TiO2 coated MWNT, Degussa P25 and E. coli . The zeta potential values ar e presented in Table 3-1. TiO2 coated MWNT has higher negative zeta poten tial than Degussa P25 and thus can be electrostatically repelled from E. coli surface. One approach to identify the role of electrostatic interactions is to conduct photo catalysis experiments with TiO2 coated MWNT at a lower pH, where the zeta potential of th e photocatalyst has a smaller negative value. However, susceptibility of E. coli to inactivating agents at lower pH can introduce an artifact in this approach. Alternatively, photocatalysis experi ments can be conducted with Degussa P25 at higher pH where the zeta potential of the phot ocatalyst has a higher negative value. Accordingly, photocatalysis experiments were ca rried out with Degussa P25 at pH 7.5. The zeta potential values of Degussa P25 and E. coli at pH 7.5 are -25 mV and -40 mV, respectively. The large negative values of zeta pote ntial suggests that favorable el ectrostatic repulsion exists and
48 therefore Degussa P25 should not contribute towards photo catalytic inactivation of E. coli . However, the D value increased from 4 min at pH 6 to 6 min at pH 7.5. A marginal increase in D value suggests that electrostatic interactions may not be the do minant factor in limiting the contact of TiO2 coated MWNT with cell-wall of E. coli . Effect of Size and Aspect Ratio of Photocatalyst Given that electrostatic interac tions were shown not to be re sponsible for lack of contact, steric repulsion would seem to be the most lik ely cause of insufficient contact between TiO2 coated MWNT and bacterial cell-wall. The cau se of steric repulsion could be the surface appendages of E. coli , such as fimbriae and flagella. Fimb riae are a characteristic feature of Gram negative bacteria, such as E. coli and are usually 1 microns in length and 2 nm in diameter (Mol and Oudega, 1996). Base d on the TEM image (Klemm, 1994) of E. coli , the average distance between two fimbriae can be esti mated to be 100 nm. Therefore the size of the photocatalyst (or radius of gyrati on in case of high-aspect ratio pa rticles) must be less than 100 nm for particles to penetrate fimbri ae and contact with cell-wall of E. coli . The dynamic nature of fimbriae, due to thermal vibrations, can fu rther reduce maximum size limit of particles for penetration. TiO2 coated MWNT are high-aspect ratio phot ocatalysts with 25 nm diameter and 2 m length. Since fimbriae are known to hinde r interaction of macr ophages (10 m longest dimension) with E. coli cells, it is plausible that fimbriae can also sterically hinder contact between the TiO2 coated MWNT and cell-wall, since MW NT have a length of 2 m. Steric hindrance of TiO2 coated MWNT by fimbriae is depicted in Figure 3-6. In contrast, Degussa P25 has a primary particle size of 30 nm and ther efore can easily penetrate between fimbriae and adsorb on surface of E. coli , as shown in Figure 3-6. Two different approaches were undertaken, as illustrated in Figur e 3-7, to test the hypothesis that surface appendages st erically hinder contact of high-aspect ratio photocatalyst
49 with bacterial cell-wall. In the first approach, high-aspect rati o photocatalyst were tested for inactivation of mutants with a nd without surface appendages. A ccording to the hypothesis, the D value for inactivation of mutant with surface appendages should be higher than that for the mutant without surface appendages. In second approach, TiO2 coated MWNT were ground to decrease the aspect ratio, which should result in more effective penetration between fimbriae, thereby increasing contact between photocatalys t and the bacterial cellwall and, concomitantly, bacterial inactivation (l owering of D value). Approach-1: Mutants with a nd without surface appendages Staphylococcus aureus was used as a model for bact eria with and without surface appendages. E. coli could not be used as model as no mutants are known that are completely devoid of surface appendages. Furthermore genetic engineering approaches to producing mutants completely without fimbriae have been unsuccessful. The strains of S. aureus with surface appendages used in this study were CP5 and Sm ith Diffuse. The mutant strains without surface appendages were CPand Smith Compact. The results with CP5 and CP(Fig. 3-8) indicate that these two strains are highly su sceptible to UVA radiation. In f act an increase in D value was observed in the presence of photocatalyst, which may be due to the UVA shielding by photocatalyst particles. In a different experime nt, the Smith diffuse and Smith Compact strains were also found to be highly susceptible (D value = 4 min) to UVA radiation. No further experiments were conducted for this approach due to high sensitivity of these strains to UVA irradiation. Approach-2: Size reduction of TiO2 coated MWNT As mentioned earlier, TiO2 coated MWNT photocatalyst are 2 m long and 20 nm in diameter. Stirred media mill was employed for size reduction of this photocatalyst with zirconia beads as grinding media. Grinding of TiO2 for 2 hours leads to phase transformation
50 from anatase (photocatalytically ac tive) to rutile or br ookite (photocatalytically inactive) (Criado and Real, 1983). Therefore, the grinding time in the present study was limite d to four hours. The particle size of ground TiO2 coat ed MWNT was experimentally determined to be 500 nm, which is larger than the estimated si ze of 100 nm required for suffici ent adsorption of photocatalyst on to cell-wall of E. coli . The Photocatalysis experi ments with size-reduced TiO2 coated MWNT particles did not inactivate E. coli . Failure of photocatalytic inac tivation can be attributed to either insufficient reduction of size or lo ss in photocatalytic ac tivity. Grinding of TiO2 coated MWNT can also result in delamination of TiO2 coating from the surface of carbon nanotubes, which reduces the photocatalytic activity of the nanocomposite. No further experiments were performed with size-reduction approach due to inherent flaws in the methodology, as mentioned above. The electron accepting ability of carbon na notubes was successfully exploited for enhancing TiO2 photocatalysis. However, the photocat alytic nanocomposites were unable to inactivate E. coli due to their large size. Fullerenes, which are a spherical version of carbon nanotubes, have electron accepting properties similar to carbon nanotubes (Sherigara et al. , 2003). Another advantage of fullerenes is their small molecular size of 1 nm, which overcomes the size constraint for inactivation of microorganisms with surface appendages. Enhancement of TiO2 Photocatalysis with Polyhydroxy Fullerenes Fullerenes (C60) are known for their unique elect ronic properties (Guldi, 2000b; Gust et al. , 2000; Makarova, 2001). Kamat and coworkers have demonstrated the transfer of photogenerated electrons from titanium dioxide to full erenes with ethanol/benzene mixture as solvent (Kamat et al., 1994). Fullerenes are not soluble in water and toxic, limiting their use in aqueous media for enhancing photocatalysis . The water-solubility of fu llerenes is improved by coupling hydroxyl groups to the molecules, creating the possibi lity of utilizing thei r electronic properties
51 in aqueous systems. Hydroxylated fullerenes are reported to be nontoxic and have been proposed for therapeutic applications. However, addition of hydroxyl groups to the fullerene structure modifies the electr onic properties of the fulle renes. In one study, polyhydroxy fullerenes synthesized via acid rout e could not scavenge electrons fr om electron rich amines such as diazabicyclooctane (DABCO) and hexamine (Mohan et al., 1997). Furthermore, Brownian motion of water-soluble fullerenes could impede their effective contact with titanium dioxide surface for electron transfer process. A nanocomposite of PHF and TiO2 can be synthesized either by coating a cluster of fullerene molecules with TiO2 or by coating TiO2 with fullerenes. In the latter case, the negatively charged PHF could pot entially adsorb to the TiO2 near its isoelect ric point (pH 5), in effect representing self-assembly of photocatal ytic nanocomposite. This approach is attractive because it does not require chemical bonding of PHF to TiO2, and may also be more practical. Accordingly, experiments with self-a ssembled nanocomposites of PHF and TiO2 were conducted to test the ability of PHF to enha nce the photocatalytic activity of TiO2. Preliminary Dye Degradation Experime nts with Self-assembled PHF-TiO2 Nanocomposites Preliminary experiments were conducted to ve rify if PHF can enhance the photocatalytic activity of TiO2 for degradation of an organic polluta nt. Procion Red MX 5B, which is an aromatic triazine monoazo compound, served as a model organic pollutant as it is relatively resistant to oxidation. Anatase (99%) was chos en as the photocatalyst for dye degradation experiments. Measured trends in dye absorbance versus time, representing PHF alone, titanium dioxide alone, and a mixture of PHF and titanium dioxide under solar UVA irradiation are presented in Figure 3-9. The photocatalytic dye degradation r eaction follows first-order reaction kinetics and has been characterized with pseudo-f irst-order rate coefficients (So et al. , 2002; Hu et al. , 2003b;
52 Hu et al. , 2003a; Yu et al. , 2005a). Pseudo-first-order relations hips were fitted to the data by least squares linear regression of log absorbance vs. time. The fitted relationships, in which the exponents give the photocatalysis ra te coefficients, are shown adjacent to the respective curves. The rate of degradation w ith self-assembled PHF-TiO2 nanocomposite, as indicated by the slope of the fit, was statistically more significant than the rate with TiO2 alone (alpha=0.01). The pseudo-first order rate coeffici ent of nanocomposite (0.0119.0001 min-1) was 1.6 times higher than the rate coefficient with TiO2 alone (0.0073.001 min-1). It is noteworthy that the rate of dye degradation with PHF alone (6 x 10-5 min-1) was two orders of magnitude less than obtained with TiO2 alone. Previous studies have reported that PHF can generate supe roxide radicals with UVA and visible light (Kamat et al. , 2000; Vileno et al. , 2004; Pickering and Wiesner, 2005). However, the rate of generation of superoxide ra dicals is dependent on the pH; the rate at pH 7 (9 x 10-5 min-1) is 100 times slower than at pH 5 (Pic kering and Wiesner, 2005). In the present study, at pH 6, the generation of superoxide radicals may not be significant and therefore no dye degradation was observed. Optimization of PHF-TiO2 Nanocomposites for Photocatalysis Polyhydroxy fullerenes were successfully employe d to enhance the photocatalytic activity of TiO2. Preliminary dye degradation experiments with a self-assembled nanocomposite showed that amount of PHF required to enhance the photocatalytic activity was three orders of magnitude less than TiO2 concentration. Assuming that electr on scavenging is the mechanism of observed enhancement, the magnitude of enhancement is dependent on the surface coverage of titanium dioxide nanoparticles with PHF. Ther efore, an optimum concentration of PHF is expected which will result in maximum enhancement for degradation of dye. Further dye degradation experi ments were carried out to optimize PHF concentration. The photocatalysis rate coefficients for experiments performed with di fferent dosed weight ratios of
53 PHF to titanium dioxide are shown in Figure 3-10. Ratios in the range of 0.001 to 0.003 gave the highest rate coefficients, whic h were 1.64.68 times higher than the rate coefficient for titanium dioxide alone. Further increase in the ratio caused a decrease of dye degradation rate, with finally no dye degradation observed at a PHF to titanium dioxide ratio of 0.09. The presence of optimum dosed weight ratio of PHF to titanium dioxide can be explained on the basis of surface coverage of titanium dioxide nanoparticles by PHF. At the ratio of 0.0002, the surface coverage of titanium dioxide nanopa rticles is very low (< 1%) and therefore few photo-generated electrons are scavenged by PHF. Further increase in PHF concentration leads to greater surface covera ge and therefore the number of scavenged photo-generated electron also increases, resulti ng in higher photocatalysis efficien cy (lower recombination). This phenomenon is observed at PHF to titani um dioxide ratios of 0.001.003 (2 %). Further increase in the ratio can lead to complete monolayer coverage of titanium dioxide surface exposed to light, resulting in no apparent enhancement. The reduction in enhancement can be due to blocking of active sites for ra dical generation and shielding of TiO2 from UVA irradiation. The calculated surface coverage (232%) at which phot ocatalysis is almost completely suppressed represents multi layer coating of PHF on TiO2. Fullerenes are well known for their optical limiting abilityâ€”higher absorption of light in excited (triplet) stateâ€” so that a few layers of PHF should be sufficient to block the UVA radiation en tirely [ref]. These calculations assumed that PHF molecules were in non-aggregated state, si nce aggregation has been predicted to occur only at much higher concentra tions (Jeng et al., 1999). After successful application of PHF to enhan ce photocatalytic activity of anatase for dye degradation, experiments were cond ucted to test the ability of PH F to enhance the activity of the best commercially available photocatal yst, Degussa P25. The optimum PHF/TiO2 ratio obtained
54 with dye degradation studies was em ployed for accelerating inactivation of E. col i with Degussa P25. Enhancement of E. coli Inactivation with PHF-TiO2 Nanocomposite Kinetics of the photocatalytic inactivation of E. coli are presented in Figure 3-11. Inactivation rate in the absen ce of P25 (i.e., control and PHF alone) was essentially zero. The inactivation rate coefficient for P25 + PHF (0.177 0.022 min-1) is 1.9 times faster than the inactivation rate for P25 alone (0.094 0.027 min-1). Figure 3-12 shows that the D value (time required for 1 log reduction of E. coli ) with P25 + PHF is 7.1 min, which is significantly less ( = 0.05) than the time of 11.7 mi n required with P25 alone. Complete destruction (7 log10 reduction) of E. coli is achieved in 40 minutes with P25 + PHF, compared to 60 minutes with P25 alone. These results are noteworthy, sin ce P25 is the best commercially available photocatalyst. These results indicate that the PHF is capable of enha ncing the action of both anatase and Degussa P25 (mixture of anatase an d rutile) and may also enhance photocatalytic activity of materials such as zinc ox ide, vanadium oxide and cerium oxide. In order to gain a better understanding of the interac tions between PHF and TiO2, the mechanisms of interaction both in terms of self-assembly and photocat alysis need to be examined. Thus the adsorption of PHF was firs t studied, followed by analysis of free hydroxyl formation. Adsorption of PHF on TiO2 Nanoparticles Adsorption experiments were carried out to investigate self-assembly of PHF on TiO2. Predetermined concentrations of PHF were added to TiO2 suspensions and the amount of adsorbed PHF was determined by depletion met hod. The concentration of PHF remaining in the solution was ascertained with UV-Vis spectrosco py. The UV-Vis absorption spectrum of PHF in
55 the range of 200 nm wavelengths is shown in Figure 3-13. The absorbance at 254 nm was selected for calibration, as orga nic carbon analysis is usually performed at this wavelength (Deflandre and Gagne, 2001; Potter, 2005). The wa velength of 254 nm minimizes the influence of interferences while maximizing sensitivity. Examples of calibration curves at three different pH values are shown in Figure 3-14. The experimental data indicate that PHF molecu les adsorb to titanium dioxide particles. Adsorption density of PHF on titanium dioxide na noparticles increased with lowering of pH (Fig. 3-15). Correspondingly, the sh ift in zeta potential (differe nce between zeta potential of nanocomposites and titanium dioxide alone) beca me more negative with lowering of pH. The zeta potential shifts we re highly correlated ( = 0.01) with adsorption density (Fig. 3-16). PHF molecules have affinity towards titaniu m dioxide, as indicated by adsorption density. The isoelectric point (i.e., the pH at which the zeta potential is zero) of titanium dioxide is in the pH range 5 (Kosmulski, 2002; Miyauchi et al. , 2004; Liufu et al. , 2005). Upon adsorption of PHF molecules at pH 3.3, the ze ta potential of titanium dioxide changed from +25 mV to -30 mV indicating that PHF molecules are negatively charged. The trends of adsorption density and shifts in zeta potential values with varying pH are consistent with th e PHF molecules having a negative charge (Mohan et al., 1997) based on the following reasoning. At acidic pH values, the surfaces of titanium dioxide na noparticles have a net positive charge and larger amount of PHF molecules are electrostatically attracted towa rds the titanium dioxide resulting in higher adsorption density. Since some of the positive sites on titanium dioxide surface are covered by PHF molecules, the net charge displayed by th e surface decreases leadi ng to lowering of zeta potential and hence a pronounced sh ift in zeta potentia l towards negative values. Above IEP of TiO2, the PHF molecules are electrostatically repe lled from the negatively charged titanium
56 dioxide surfaces and hence the adsorption density is low. At very low adsorption, the surface charge remains essentially unaltered and theref ore the shift in zeta potential is negligible. Electrostatic forces govern the adsorption of PHF molecules onto surface of titanium dioxide nanoparticles. Detection of Hydroxyl Radicals with EPR One approach to validate the hypothesis that PHF enhances TiO2 photocatalysis by electron scavenging is by measuring the concentr ation of hydroxyl radicals. The concentration of hydroxyl radicals can be determined by either fl uorescence techniques or by detecting spin trapped radicals with electron pa ramagnetic resonance spectroscopy (EPR). The former approach comprises of organic molecules that become fl uorescent when attacked by hydroxyl radicals. However, superoxide radicals, which are produced duri ng photocatalysis, can also react with the organic molecules thereby contribut ing to the fluorescent signal. In contrast, the latter approach utilizes spin-trap agents, which are not susceptible to attack by superoxide radicals. Accordingly, EPR experiments were conducted to test the el ectron scavenging hypothesis with 5,5-dimethyl-1pyrroline N-oxide (DMPO) as spin-trap for hydroxyl radicals. The EPR spectra obtained after UVA irradiation of DMPO and TiO2 alone or TiO2+PHF (Fig. 3-17) were identical in term s of peak locations and the char acteristic 1:2:2:1 relative peak magnitudes associated with trapped hydroxyl radi cals in the form of DMPO-OH (Uchino et al., 2002). Presence of PHF in the reaction mixtures had no eff ect on peak locations, but substantially increased peak magnitudes. This shows that addition of PHF increases the concentration of DMPO-OH, indicating higher rates of fr ee hydroxyl radical generation.
57 Plots of the DMPO-OH concentrations achieved after UVA irradiation of various reaction mixtures is presented in Figur e 3-18. No production of DMPO-OH was observed in the dark control. A concentration of 1.2 M DMPOOH was formed in the UVA control over 60 min, corresponding to a yield of 0.33 M cm2 J-1. This is comparable to the irradiated control yield of 0.5 M cm2 J-1 obtained by Uchino et al. (Uchino et al., 2002). Production of DMPO-OH by irradiation of PHF alone was similar to the UVA control, indicating th at PHF alone does not generate more hydroxyl radicals than UVA alone. Substantially higher DMPO-OH concentrations were obtained when TiO2 was present in the reaction mixture, reflecting the contribution of photocatalysis. The rate of DMPO-OH production in the presence of both TiO2 and PHF was in the range of 20% highe r than the rate obtained with TiO2 alone. The concentration of DMPO-OH generated exhibits a maximum as a function of UVA irradiation time followed by a gradual decrease. Similar observations ha ve been reported by others for ceria (HernandezAlonso et al., 2004) and TiO2 photocatalysts (Dvoranova et al., 2002). The gradual decrease after reaching a maximum may be attr ibuted to degradation of trapped radicals, accompanied by a drop-off in the rate of DMPO-OH production. DMPO-OH is known to be unstable, with a half-life of 20 min (Grela et al., 1996). One possible mechanism of degradation is multiple additions of hydr oxyl radicals on DMPO (Dvoranova et al., 2002). However, Hernandez-Alonso et al. (Hernandez-Alons o et al., 2004) showed experimental results that conflict with the multiple addition mechan ism, and argued that direct oxidation of DMPOOH radicals by photo-generated holes could better explain their results. A slow-down in the production rate of DMPO-OH could occur due to consumption of oxygen (Brezova et al., 1991), resulting in accumulation of electrons thus promoting elec tron-hole recombination. Direct
58 oxidation of DMPO-OH radicals by photo-generated holes w ould lead to a slow-down in free hydroxyl generation from holes and thus decreased DMPO-OH production. Contrary to Mohan and coworkersâ€™ conclusi on that electron accepting ability of fullerenes is lost upon hydroxylation (Mohan et al., 1997), th e observed enhancement in dye degradation, E. coli inactivation and hydroxyl radical generation indicates that PHF molecules retain their electron scavenging ability. The enhancement is speculated to proceed in the following way. The photo-generated electrons and holes in titanium dioxide particles take pa rt in redox reactions at the surface or recombine, as depicted in Figure 3-19a. The r ecombination process has a faster kinetics and controls the efficien cy of photocatalysis (Hoffmann et al., 1995). In the presence of adsorbed PHF molecules, photo-generated elec trons are scavenged by PHF molecules (Fig. 319b), decreasing the number available for reco mbination. The scavenged electrons ultimately enter reduction reactions, assuming that PHF act as an electron relay similar to fullerenes as suggested by Guldi and Prato (Guldi, 2000b). The discrepancy in electron scavenging ability of PHF between our study and the literature can be attributed to the mo lecular composition. The PHF empl oyed in the present study was synthesized via alkali route, wh ereas Mohan et al. produced PHF via acid synthesis route. The methodology employed for synthesis can dictate the composition of functional groups present on the fullerene cage. The estimated number of hydroxyl groups for PHF synthesized via alkali route is usually higher (24) than those synthesized via aci d route (18). PHF typically contains impure groups such as hemiketal, epoxi de and carbonyl. As discussed below, increasing the concentration of certain functional groups can either enhance or result in loss of electron scavenging ability.
59 Influence of Composition of Polyhydroxy Fu llerenes on Photocatalytic Enhancement Recent literature has emphasized the need for extensiv e characterization of PHF as the type and concentration of functional groups pres ent on fullerene cage can significantly influence its properties (Xing et al. , 2004; Rodriguez-Zavala and Guir ado-Lopez, 2006). PHF usually contains impure groups such as hemiketal, epoxi de and carbonyl. The functionality of PHF can also influence its electron scavenging ability and, hence, its role in photocatalysis. Increasing the concentration of certain functional groups can either enhance or resu lt in loss of electron scavenging ability. Accordingly, ex periments were carried out with two different PHF molecules for enhancement of titanium dioxide photocatal ysis. A freshly synthesized batch of PHF molecules, which was able to show enhancement, was termed â€œfresh PHF,â€ whereas a batch of PHF molecules aged for 18 months and showi ng no enhancement were termed â€œaged PHF.â€ Both types of PHF were extensively characterize d to determine the func tional groups present and correlate composition with el ectron scavenging ability. Photocatalytic Dye Degradation with Fresh and Aged PHF Photocatalytic degradation of Procion Red dye with TiO2 alone, TiO2 + fresh PHF and TiO2 + aged PHF are presented in Fi gure 3-20. The ratio of PHF to TiO2 was set at the optimum value of 0.001 as previously determined. The activ ity of PHF was determined by its ability to enhance photocatalytic degradation rate. The pse udo-first order rate coefficient with fresh PHF (0.0128.0029 min-1) was 2.6 times higher than the rate coefficient without PHF (0.0048.0005 min-1). The enhancement (2.6) with the synthesized batch of fullerenes is higher than previously reported by Krishna et al . (2006). Presence of aged PHF gave a photocatalytic rate that was not significantly different than TiO2 alone.
60 Mass Spectroscopy for Fresh and Aged PHF Mass spectroscopy was employed to investigate th e stability of fullere ne cage of fresh and aged PHF. The portions of the MS spectra in the range of 700 m/z for fresh and aged PHF are presented in Figure 321. The base fullerene (C60) peak at 720.6 m/z is present at 100 relative intensity in both fresh and aged PHF, indi cating that fullerene cage is intact in both PHF samples. The presence of peaks in the range expected for PHF ions (approximately 1600 m/z based on XPS data) would assist in estimating th e molecular weight of the PHF. However, no peaks were observed in this range (data not shown). The absence of peaks representing PHF ions has also been reported by Xing et al. (2004), Chen et al. (2001) and Chiang et al. (1993). FTIR Analysis of Fresh and Aged PHF FTIR spectroscopy has been employed in litera ture to identify various functional groups present on fullerene cage (Chiang et al. , 1993; Li et al. , 1993; Husebo et al. , 2004; Xing et al. , 2004). The FTIR-DRIFT spectra for fresh and aged forms of PHF are presented in Figures 3-22 and 3-23, respectively. (Note that th e wavenumbers in this plot are increasing from right to left as is commonly portrayed in the literature.) The fresh PHF exhibits five peaks at 3300, 1591, 1450, 1062, 881 and 471 wavenumbers along w ith shoulders at 1661, 1357 and 1165 wavenumbers. The aged form of PHF ha s peaks at 3300, 1595, 1408 and 1074 wavenumbers with a shoulder at 1680 cm-1. The vibrational modes of PHF were assigned to FTIR peaks of fresh and aged PHF based on the information obtained from Gaussian simu lation and the literature (Table 3-2).
61 Gaussian simulation Semi-empirical computation (PM3) has been employed in the literature for structure optimization of various hydroxylat ed fullerenes and possible stable isomers are reported (Slanina et al. , 1996; Guirado-Lopez and Ri ncon, 2006; Rodriguez-Zavala and Guirado-Lopez, 2006) . However no reports are present on theoretical pr ediction of vibrational peaks for hydroxylated fullerenes. The purpose of utilizing hybrid quantum chemical computation in the present study was to identify the vibration modes responsible for experimental FTIR peaks. For ease of simulation PHF was assumed to have 24 hydroxyl groups and no carbonyl, hemiketal and epoxide functionalities. The structure optimizat ion and vibration spectrum generation were performed with hybrid quantum-chemical basi s set (B3LYP 6-31G*) (Foresman, 1996). The optimized structure (Fig. 3-24) is similar to the reported C60(OH)24 structure obtained by a semiempirical quantum-chemical optimization (PM3) (Slanina et al. , 1996). The hydroxyl groups are present as intramoleculary hydrogen bonded island s on fullerene cage similar to the results obtained in literature (GuiradoLopez and Rincon, 2006; Rodrigue z-Zavala and Guirado-Lopez, 2006). The average Câ€“O and Oâ€“H bond length are 1.43 and 0.98 , respectively, and the average Câ€“Oâ€“H bond angle is 107. These results ar e similar to the values obtained with PM3 optimization for C60(OH)26 by Guirado-Lopez and Rincon (2006). Additionally, weak intramolecular hydrogen bonding was observed with average bond length of 1.78 , which is similar to the hydrogen bond length in water. The vibration spectrum generated for C60(OH)24 with B3LYP 6-31 G* basis set is presented in Figure 3-25, with wavenumbers in creasing from left to right on x axis. The simulated FTIR peak exhibits four major peaks. The broad peak at 3450 cm-1 originates from Oâ€“ H stretching. The peak at 1450 cm-1, 1070 cm-1 and 370 cm-1 represents Câ€“Oâ€“H bending, Câ€“O stretching and Oâ€“H rocking vibrations, respectively.
62 In reality, PHF molecules synt hesized via alkali route have hemiketal and epoxide groups along with hydroxyl groups as revealed by FTIR and XPS analysis in the present study and reports in the literature (Husebo et al. , 2004; Xing et al. , 2004). Rodriguez-Zavala and GuiradoLopez (2006) have performed theoretical studi es on hydroxylated fullerenes with different numbers of epoxide groups. They found that presence of epoxide groups on hydroxylated fullerenes can have significant effects on their structure and el ectronic and optical properties. Further theoretical and experiment al research is required to eluc idate the effect of different functionalities on physical , electronic and chemical properties of PHF. Literature values The peak at 1591 was not observed in the simu lation and was ascribed to C=C vibrations as reported in the literature (Chiang et al. , 1993; Li et al. , 1993; Xing et al. , 2004). The shoulders at 1658 cm-1, 1357 cm-1 and 1165 cm-1 were attributed to hemiketal (Xing et al. , 2004), epoxides and esters (Schraff et al. , 2005)respectively. XPS Analysis for Fresh and Aged PHF Polyhydroxy fullerene (PHF) molecules have hemiketal, epoxide and sometimes carbonyl functional groups in addition to hydroxyl groups. Th erefore extensive characterization of PHF as well as determination of its empirical formula is necessary to compare th e properties with those reported in the literature. Unfortunately, there is a lack of consistency in techniques employed for determining the empirical molecular formul a of PHF. The three common methodologies employed are elemental analysis (Li et al. , 1993; Chen et al. , 2001; Xing et al. , 2004; Vileno et al. , 2006), thermo gravimetric analysis (Goswami et al. , 2004; Alves et al. , 2006) and XPS (Husebo et al. , 2004). However, as Husebo et al . reported, the presence of residue in TGA and elemental analysis can influence the empirical formula. The pres ent study therefore utilizes XPS data, similar to Husebo et al ., to determine the empirical form ulas for fresh and aged PHF.
63 XPS analysis indicated presence of C, O a nd Na for both fresh and aged forms of PHF. The relative atomic concentrations determined for each element are provided in Table 3-3. The C1s region was further analyzed to reveal three oxidation states of carbon as shown in Figures 326 and 3-27, which is in agreem ent with the lit erature (Chiang et al. , 1993; Xing et al. , 2004). Since no carbonyl peak was presen t in FTIR spectra, the highest oxidation state was assumed to be due to hemiketal structure as revealed from FTIR. Curve fitting analysis was performed for the C1s spectrum of fresh and aged PHF to cal culate the relative con centration of each carbon oxidation state (Table 3-4). The relative concen tration of mono-oxygenated carbon is three times higher in fresh PHF than aged form. The relative atomic concentrations of C, O a nd Na and along with con centrations of monoand di-oxygenated states of carbon were empl oyed to deduce the composition of PHF according to Husebo et al. (2004). The molecular formula for fresh PHF was calculated as C60O8(OH)28Na10 and for aged PHF as C60O16(OH)10Na24. As seen from the empirical formulas, the number of functional groups added to fullerene cage in this study is within the range (24) of reported values (Xing et al. , 2004; Vileno et al. , 2006) however, fresh PHF has a higher number of functional groups per molecule than aged PHF. TGA Analysis for Fresh and Aged PHF Thermo gravimetric analyses (TGA) of fres h and aged PHF are presented in Figure 3-28. The weight loss generally occurs in 4 st ages, similar to that reported by Chiang et al. (1993). The first stage in weight loss is attributed to desorption of physica lly adsorbed water and accounted for 16% of the weight of both fresh and aged PHF. The second stage is attributed to desorption of hydroxyl functional groups and was 20% and 21% of the weight of fresh and aged PHF respectively. The weight loss in stage 3 is a ttributed to desorption of hemiketal functional groups and accounted for 11% of weight of fres h PHF and 34% of the weight of aged PHF. A
64 small weight loss (6%) in stage 4 was observed for fresh PHF and is attributed to desorption of carbonyl or epoxide groups. No weight loss in stage 4 was obs erved in agede PHF suggesting absence of carbonyl or epoxide groups. Finally aged PHF exhibited thermal degradation of structure resulting in further 19% weight loss (sta ge 5). No loss in stage 5 was observed for fresh PHF. The total weight loss up to a temperature of 1000 C was 54% for fresh PHF, which is close to the 52% weight loss reported by Husebo et al . (2004). TGA shows that the ratio of the weights of hemiketal to hydroxyl groups is higher for aged PHF. The same trend in terms of number of hemiketal to hydroxyl groups is al so observed in XPS results. Effect of Impure Functional Groups Stability: As mentioned earlier, functional groups other than hydroxyls can influence the properties of PHF. Xing et al . (2004) has synthesized PHF with different concentrations of functional groups by controlling the concentra tion of alkali. The various PHF were then characterized extensively to de termine the effect of presen ce of impure groups (hemiketal, epoxide and carbonyl) on stability of fullerene cage. Laser induced dissociation and XPS analyses indicated that the stability of th e fullerene molecule decreases with higher concentrations of impure groups. Xing et al. de fine a parameter R (ratio of impure groups to hydroxyl groups as determined by XPS analysis) and suggest that PHF molecules with R less than 0.2 and number of hydroxyl groups less than 36 are mostly stable. In the present study, the values of R (as determined by XPS analyses) for fresh and aged PHF are 0.27 and 1.66 respectively, suggesting lower stability of aged PHF. The lower stability of aged PHF is confirmed by thermal degradation observed at temperatures greater th an 800 C (Fig. 3-28). Electron scavenging ability: Theoretical simulation by Ro driguez-Zavala and GuiradoLopez (2006) suggest that presence of impure groups (hemiketal and epoxide) can significantly influence the electronic prope rty of PHF molecules. In ab sence of impure groups, the PHF
65 molecules were able to take up to 6 electrons wi thout rupture of carbon st ructure. It should be noted that addition of six elec trons resulted in cleavage of a single hydroxyl bond, suggesting further addition of electrons can promote dehydr oxylation of fullerene cage. In presence of impure groups, addition of electrons did not result in desorpti on of hydroxyl moieties, instead leading to surface reconstruction where Câ€“ C bonds containing an adsorbed oxygen were replaced by Câ€“Oâ€“C bonds. Based on these theoretical studies and experiment al investigations by Xing et al . (2004), the relative concentr ation of impure groups can be proposed to influence the electron scavenging ability of PHF molecules. PHF molecules with lower ratios of impure groups to hydroxyl groups (and with fewer th an 36 hydroxyl groups) are needed for efficient scavenging of electrons leading to greater enhancement in titanium dioxide photocatalysis. These observations indicate that more e fficient design of PHF is possible .
66 Figure 3-1. D value estimation for photocatalytic inactivation of E. coli with Degussa P25. y = -0.1647x + 0.0004 -7 -6 -5 -4 -3 -2 -1 0 0102030405060Time (min)Log (Survival Ratio) S t D value = t S Rate Coefficient = t S y = -0.1647x + 0.0004 -7 -6 -5 -4 -3 -2 -1 0 0102030405060Time (min)Log (Survival Ratio) S t D value = t S D value = t S Rate Coefficient = t S
67 Figure 3-2. Optimum concentration of Degu ssa P25 for photocatalytic inactivation of E. coli . Error bars are .0 SD; some of the error bars are too small to be visible. 0 10 20 30 40 50 60 70 80 10100100010000Titanium dioxide concentration (mg/L)D Value (min)
68 Figure 3-3. Photocatal ytic inactivation of B. cereus spores. 0 50 100 150 200 250 300 ControlFunct. MWNTDegussa P25TiO2 coated MWNTD Value (min) TiO2coated MWNT 0 50 100 150 200 250 300 ControlFunct. MWNTDegussa P25TiO2 coated MWNTD Value (min) TiO2coated MWNT
69 Figure 3-4. Photocatal ytic inactivation of B. subtilis spores. 0 50 100 150 200 250 ControlDegussa P25TiO2 coated MWNTD Value (min) TiO2coated MWNT 0 50 100 150 200 250 ControlDegussa P25TiO2 coated MWNTD Value (min) TiO2coated MWNT
70 Figure 3-5. Photocatal ytic inactivation of E. coli . 0 20 40 60 80 100 ControlDegussa P25TiO2 coated MWNTD Value (min) TiO2coated MWNT 0 20 40 60 80 100 ControlDegussa P25TiO2 coated MWNTD Value (min) TiO2coated MWNT
71 Figure 3-6. Interaction of TiO2 coated MWNT and Degussa P25 with surface appendages of E. coli . TiO2coated nanotubes (Dia= 25 nm) 2 â€“10 micronsFimbriae Degussa P25 (Dia= 30 nm) TiO2coated nanotubes (Dia= 25 nm) 2 â€“10 micronsFimbriae Degussa P25 (Dia= 30 nm)
72 Figure 3-7. Two different approaches undert aken to test the hypothesis that surface a ppendages sterically hinder contact of hi ghaspect ratio photocatalyst w ith bacterial cell-wall. High aspect ratio photocatalysts cannot inactivate microbes with surface appendagesExperiments with low aspect ratio photocatalysts on microbes with surface appendages. Experiments with high aspect ratio photocatalysts on microbes with & without surface appendagesExperiments with mutant strains of bact eria with and without surface appendagesHypothesis Fullerenes as e-acceptors Grinding of TiO2coated multi-wall carbon nanotubes UV was sufficient to inactivate S.aureus mutants They did not inactivate E. coli High aspect ratio photocatalysts cannot inactivate microbes with surface appendagesExperiments with low aspect ratio photocatalysts on microbes with surface appendages. Experiments with high aspect ratio photocatalysts on microbes with & without surface appendagesExperiments with mutant strains of bact eria with and without surface appendagesHypothesis Fullerenes as e-acceptors Fullerenes as e-acceptors Grinding of TiO2coated multi-wall carbon nanotubes Grinding of TiO2coated multi-wall carbon nanotubes UV was sufficient to inactivate S.aureus mutants UV was sufficient to inactivate S.aureus mutants They did not inactivate E. coli They did not inactivate E. coli
73 Figure 3-8. Photocatal ytic inactivation of S. aureus mutants with and without surface a ppendages. The strain CP5 has surface appendages and CPis mutant strain without appendages. 0 2 4 6 8 10 ControlDegussa P25TiO2 coated MWNTD Value (min) CP5 CPTiO2coated MWNT 0 2 4 6 8 10 ControlDegussa P25TiO2 coated MWNTD Value (min) CP5 CPTiO2coated MWNT
74 Figure 3-9. First-order degradati on kinetics of Procion Red MX-5B upon UVA irradiation with PHF, TiO2 and a mixture of TiO2 and PHF. Error bars are .0 SD; some of the error bars are too sma ll to be visible. y = -6E-05x 0.0003 y = -0.0073x 0.0398 y = -0.0119x 0.0174 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 015304560Time (min)log (C/Co)TiO2+ PHF TiO2PHF y = -6E-05x 0.0003 y = -0.0073x 0.0398 y = -0.0119x 0.0174 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 015304560Time (min)log (C/Co)TiO2+ PHF TiO2PHF
75 Figure 3-10. Normalized pseudo-first-order rate coefficient for dye degradation as a function of the ratio of added polyhydrox y fullerenes (PHF) to TiO2. Values above 1.0 indicate enhancement of photocat alysis relative to that achieved with TiO2 alone. Means of the normalized pseudo-firs t order rate coefficient at PHF rati os of 0, 0.001 and 0.093 were significantly different from each other at = 0.01, as indicated with the different letters a bove the respective data points. The error bars at point b are too small to be seen. b a 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 -0.00100.0010.0020.0030.0040.005PHF/ TiO2 Ratio (Initial wt. basis)Normalized Pseudo-FirstOrder Rate Coeffecient 0 c 0.1 TiO2 TiO2 TiO2 TiO2 TiO2 b a 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 -0.00100.0010.0020.0030.0040.005PHF/ TiO2 Ratio (Initial wt. basis)Normalized Pseudo-FirstOrder Rate Coeffecient 0 c 0.1 TiO2 TiO2 TiO2 TiO2 TiO2 TiO2 TiO2 TiO2 TiO2 TiO2 TiO2 TiO2
76 Figure 3-11. Photocatal ytic inactivation of E. coli plotted as a function of surv ival ratio vs. time. Error bars are .0 SD; error bars for PHF and control are too small to be visibl e. The rate with P25+PHF is significantly greater than the rate with P25 alone at = 0.05. -8 -7 -6 -5 -4 -3 -2 -1 0 0102030405060Time (min)Log (Survival Ratio)P25+PHF P25 Control PHF -8 -7 -6 -5 -4 -3 -2 -1 0 0102030405060Time (min)Log (Survival Ratio)P25+PHF P25 Control PHF
77 Figure 3-12. D values for E. coli inactivation with Degussa P25 alone and a mixture of Degussa P25 a nd PHF. Error bars are .0 SD. 0 3 6 9 12 15D value (min)P25 + PHF P25 0 3 6 9 12 15D value (min)P25 + PHF P25
78 Figure 3-13. Absorption spectrum of polyhydroxy fullerenes. 0 0.1 0.2 0.3 0.4 200300400500600700800Wavelength (nm)Absorbance
79 Figure 3-14. Calibration curves for polyhydroxy fulle renes (PHF) at three different pH values. pH 3.3 pH 6.1 pH 9.9R2 = 0.995 R2 = 1 R2 = 10 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0246810 Concentration of PHF (mg/l)Absorbance at =254 nm
80 Figure 3-15. Adsorption density of polyhydroxy fullerenes (PHF) on titanium dioxide and shift in zeta potential of titanium di oxide nanoparticles with adsorption of polyhydroxy fullerenes at di fferent pH. Means labeled with same letter are not significantly different at =0.01. a b b-70 -60 -50 -40 -30 -20 -10 0 Zeta Potential (mV) a b b0 5E+16 1E+17 1.5E+17 2E+17 2.5E+17 024681012pHAdsorption Density (# of molecules of PHF/ m2) A B B a b b-70 -60 -50 -40 -30 -20 -10 0 Zeta Potential (mV) a b b0 5E+16 1E+17 1.5E+17 2E+17 2.5E+17 024681012pHAdsorption Density (# of molecules of PHF/ m2) A B B
81 Figure 3-16. Zeta potential shifts as a function of adsorption density. y = -3E-16x + 1.2002 R2 = 0.9621-70 -60 -50 -40 -30 -20 -10 0 -5E+1605E+161E+171.5E+172E+172.5E+17Adsorption Density (# of molecules of PHF/ m2) Zeta Potential (mV) 0
82 Figure 3-17. Electron paramagnetic resonance spect ra obtained upon UVA irradi ation of DMPO and TiO2 alone and TiO2+PHF. -1200 -900 -600 -300 0 300 600 900 1200 3470350035303560Magnetic Field (Gauss)EPR Intensity (arb. units) TiO2 TiO2 + PHF TiO2 TiO2+ PHF -1200 -900 -600 -300 0 300 600 900 1200 3470350035303560Magnetic Field (Gauss)EPR Intensity (arb. units) TiO2 TiO2 + PHF TiO2 TiO2+ PHF
83 Figure 3-18. Effect of PHF on ge neration of hydroxyl radicals by UVA irradiation of titanium dioxide. 0 1 2 3 4 0102030405060Time (min)DMPO-OHTiO2TiO2+PHF(M)UVA Control PHF Dark Control 0 1 2 3 4 0102030405060Time (min)DMPO-OHTiO2TiO2+PHF(M)UVA Control PHF Dark Control
84 Figure 3-19. Hypothetical photocatalytic reactions occurring upon UV irradiation. a) with TiO2 alone and b) in presence of adsorbed PHF molecules. The time-scale for photocatalytic processes ar e from Hoffman et al. (1995). Note: A and A represents acceptor molecules befo re and after reduction (for e.g., O2 and O2). D and D represents donor molecules before and after oxidati on respectively (for e.g., OH and OH). A.A D.DTiO2h Conduction Band Valence Band eh+ 10-15 sec10-3 sec 10-7 sec10-7 sec A.A D.DTiO2h Conduction Band Valence Band eh+ 10-15 sec10-3 sec 10-7 sec10-7 sec A.A D.DTiO2h Conduction Band Valence Band eh+ PHF A.A A.A D.DTiO2h Conduction Band Valence Band eh+ PHF A.A a) b)
85 Figure 3-20. First-order degrad ation kinetics of Procion Red MX-5B upon UVA irradiation with TiO2 alone, TiO2 + aged PHF and TiO2 + fresh PHF. Error bars are .0 SD; some of the error bars are too sm all to be visible. y = -0.0053x 0.0186 y = -0.0128x + 0.0089 y = -0.0043x 0.0063-0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 015304560Time (min)log (C/Co)TiO2+ fresh PHF TiO2alone TiO2+ aged PHF y = -0.0053x 0.0186 y = -0.0128x + 0.0089 y = -0.0043x 0.0063-0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 015304560Time (min)log (C/Co)TiO2+ fresh PHF TiO2alone TiO2+ aged PHF
86 Figure 3-21. APCI-MS of fresh and aged Polyhydroxy Fullerenes. 700 710 720 730 740 750 760 770 780 790 800 810 820 830 840 850 / 0 10 20 30 40 50 60 70 80 90 100 720.6 721.5 776.5 777.5 722.7 775.1 778.7 723.1 736.6 752.3 792.1 779.1 743.7 846.3 843.2 8 4 770.3 802.5 761.8 735.4 808.5 718.9 827.9 820.5 700.6 700 710 720 730 740 750 760 770 780 790 800 810 820 830 840 850 0 10 20 30 40 50 60 70 80 90 100 Relative Abundance 720.6 721.5 776.4 777.5 841.6 792.1 737.2 795.1 778.3 790.1 727.5 849.3 742.6 715.3 754.5766.3 838.4 774.9 702.3 757.7 751.3 812.6 808.3 816.4 836.5 801.8 820.5 707.7 829.3 786.3 Fresh PHF Aged PHF 700 710 720 730 740 750 760 770 780 790 800 810 820 830 840 850 / 0 10 20 30 40 50 60 70 80 90 100 720.6 721.5 776.5 777.5 722.7 775.1 778.7 723.1 736.6 752.3 792.1 779.1 743.7 846.3 843.2 8 4 770.3 802.5 761.8 735.4 808.5 718.9 827.9 820.5 700.6 700 710 720 730 740 750 760 770 780 790 800 810 820 830 840 850 0 10 20 30 40 50 60 70 80 90 100 Relative Abundance 720.6 721.5 776.4 777.5 841.6 792.1 737.2 795.1 778.3 790.1 727.5 849.3 742.6 715.3 754.5766.3 838.4 774.9 702.3 757.7 751.3 812.6 808.3 816.4 836.5 801.8 820.5 707.7 829.3 786.3 Fresh PHF Aged PHF
87 Figure 3-22. FTIR spectrum of fresh Polyhydroxy Fullerenes. 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 400 1200 2000 2800 3600Wavenumber(cm-1)Kubelka-MunkOâ€“H C=C Hemi-ketal Câ€“Oâ€“H Câ€“O Câ€“C 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 400 1200 2000 2800 3600Wavenumber(cm-1)Kubelka-MunkOâ€“H C=C Hemi-ketal Câ€“Oâ€“H Câ€“O Câ€“C
88 Figure 3-23. FTIR spectrum of aged Polyhydroxy Fullerenes. 0 0.02 0.04 0.06 0.08 0.1 0.12 400 1200 2000 2800 3600Wavenumber(cm-1)Kubelka-MunkOâ€“H C=C Hemi-ketal Câ€“Oâ€“H Câ€“O 0 0.02 0.04 0.06 0.08 0.1 0.12 400 1200 2000 2800 3600Wavenumber(cm-1)Kubelka-MunkOâ€“H C=C Hemi-ketal Câ€“Oâ€“H Câ€“O
89 Figure 3-24. Gaussian simulation of C60(OH)24. Carbon Oxygen Hydrogen Carbon Oxygen Hydrogen
90 Figure 3-25. Gaussian simulation of vibrational spectrum for C60(OH)24. 3 500 3 000 2 500 2 000 1 500 1 000 500 0 2 800 2 600 2 400 2 200 2 000 1 800 1 600 1 400 1 200 1 000 800 600 400 200 0 Oâ€“H Oâ€“H Câ€“Oâ€“H Câ€“O Wavenumber(cm-1)IR Intensity 3 500 3 000 2 500 2 000 1 500 1 000 500 0 2 800 2 600 2 400 2 200 2 000 1 800 1 600 1 400 1 200 1 000 800 600 400 200 0 Oâ€“H Oâ€“H Câ€“Oâ€“H Câ€“O Wavenumber(cm-1)IR Intensity
91 Figure 3-26. Experimental C1s XPS spectrum (top curve) of fres h Polyhydroxy Fullerenes with fitted curves representing three different oxidation states of carbon. 30000 60000 90000 120000 150000 180000 282 284 286 288 290 292Binding Energy (eV)Counts per secondExperimental Fitted Hemi-ketal Non-oxygenated Carbon Mono-oxygenated Carbon 30000 60000 90000 120000 150000 180000 282 284 286 288 290 292Binding Energy (eV)Counts per secondExperimental Fitted Hemi-ketal Non-oxygenated Carbon Mono-oxygenated Carbon
92 Figure 3-27. Experimental C1s XPS spectrum (top curve) of aged Polyhydroxy Fullerenes with fitted curves representing three different oxidation states of carbon. 30000 60000 90000 120000 150000 180000 282 284 286 288 290 292Binding Energy (eV)Counts per secondExperimental Fitted Hemi-ketal Non-oxygenated Carbon Mono-oxygenated Carbon 30000 60000 90000 120000 150000 180000 282 284 286 288 290 292Binding Energy (eV)Counts per secondExperimental Fitted Hemi-ketal Non-oxygenated Carbon Mono-oxygenated Carbon
93 Figure 3-28. TGA spectra for fresh and ag ed Polyhydroxy Fullerenes. Numbers refer to different stages of weight loss. 0 10 20 30 40 50 60 70 80 90 100 20170320470620770920TemperatureWeight (%)Fresh PHF Aged PHF1a 4a 3a 2a 2b 3b 5b 1b 0 10 20 30 40 50 60 70 80 90 100 20170320470620770920TemperatureWeight (%)Fresh PHF Aged PHF1a 4a 3a 2a 2b 3b 5b 1b
94 Table 3-1. Zeta potential va lues of photocatalysts and E. coli . Systems Zeta Potential (mV) pH 6.0 pH 7.5 Escherichia coli -35 5 -40 5 TiO2 coated MWNT -20 5 Degussa P25 -15 5 -25 5
95 Table 3-2. Peak assignments of FTIR peaks for fresh and aged PHF based on results from Gaussian simulation of C60(OH)24 and literature. Peak Location based on Peak Assignments based on Gaussian simulation + literature Vibration modes Gaussian Simulation Literature FreshPHF Aged PHF Oâ€“H stretching 3450 3420a, 3430b, 3410c, 3300d, 3300 3300 Hemiketal 1658c, 1661 1680 C=C stretching 1595a, 1600b, 1593c, 1585d, 1591 1595 Câ€“Oâ€“H bending 1450 1392a, 1412c, 1450 1408 Epoxides 1376e 1357 Câ€“O stretching 1070 1084a, 1070b, 1065d, 1062 1074 Esters 1197e 1165 Câ€“C 490f 471 a â€“ Chiang et al., 1993; b â€“ Li et al., 1993; c â€“ Xi ng et al., 2004; d â€“ Vileno et al., 2006; e â€“ Schraff et al., 2005; f â€“ Alves et al., 2006.
96 Table 3-3. Elemental compositi on of fresh and aged PHF obtained with XPS analysis. Elements Relative Atomic Concentration Fresh PHF Aged PHF Carbon 61.23 % 47.33 % Oxygen 28.44 % 33.82 % Sodium 10.33 % 18.85 % Table 3-4. Peak position and elemental compos ition of fresh and aged PHF obtained with XPS analysis. Oxidation States of Carbon Peak Position Fresh PHF Aged PHF Relative Concentration Fresh PHF Aged PHF Non-oxygenated 284.8 eV 284.8 eV 40.2 % 57.1 % Mono-oxygenated 286.16 eV 286.14 eV 47.1 % 16.1 % Di-oxygenated 288.55 eV 287.91 eV 12.7 % 26.8 %
97 CHAPTER 4 CONCLUSIONS AND SUGGESTIONS FOR FUTURE RESEARCH Conclusions Titanium dioxide photocatalysis has immens e potential for pollution remediation and inactivation of microorganisms as it has severa l advantages over current technology including complete mineralization (complete oxidation of bacterial cell to carbon dioxide and water) of microorganisms and no release of toxic by-product s. The intrinsic disadvantage of low quantum efficiency for titanium dioxide has motivated re searchers to integrate titanium dioxide with metallic and organic components as enhancers w ith contradictory results reported (Vamathevan et al. , 2002; Arabatzis et al. , 2003a). Other materials which have been investigated for their unique electronic properties ar e carbon nanotubes and fullerenes. They possess electron accepting characterâ€”a requirement for enhancers. The present study utilizes the electronic pr operties of carbon nanot ubes and fullerenes by functionalization and integration with titaniu m dioxide to accelerate photocatalysis. TiO2 coated MWNT inactivated bacterial endospores twice as fast as the best commercially available photocatalyst (Degussa P25), suggesting that MW NT can enhance the photocatalytic activity of TiO2. Photocatalytic inactivation of microorganisms with TiO2 coated MWNT revealed that size and aspect ratio of photocatalyst are important pa rameters to be considered while designing the photocatalyst. TiO2 coated MWNT, which are high asp ect ratio photocatalyst with one dimension 2 m length, were not able to photocatalytically inactivate Escherichia coli . The surface appendages of Gram negative bacteria ( E. coli ) can prevent contact of TiO2 coated MWNT with cell-wall of the bacterium thereby hindering photocatalytic inactivation. Although the role of surface appendages (such as fimbriae and capsules) in evasion from recognition and phagocytosis by immune system is well known (Henderson et al. , 2003), its role in sterically
98 hindering photocatalysis was unknown until now. The present studies suggest that an important design criterion for developing photo catalysts is the size, which shoul d be less than 100 nm to be effective against microorganisms with surface appendages, such as E. coli . Consequently fullerenes, which are 1 nm spherical version of carbon nanotubes and therefore expected to share the same electron scavenging ability, were employed to enhance the photocatalytic activity of TiO2. Pristine fullerenes are not soluble in water and are reported to be toxic. However, hydroxylation of the fullerenes renders them water soluble and eliminates toxicity. Nanocomposites of the hydroxylated (i.e. polyhydroxy) fullerenes and TiO2 were formed by self-assembly. The PHF-TiO2 nanocomposites were successfully employed to enhance the photocatalytic efficacy of titanium di oxide nanoparticles for degradation of Procion red dye, which because of its aromatic struct ure is relatively resistant to oxidation. The affinity of PHF molecules towards tit anium dioxide surface enables the electron scavenging properties of PHF to be exploited without the need for chemical bonding between PHF molecule and titanium dioxi de through sol-gel or other coating process. Earlier work on improving the photocatalytic activity of titanium dioxide with various metals (Ag, Pt, Au) employed composite synthesis steps such as ch emical reduction or electron beam evaporation, which adds cost and unit operati ons for synthesis of modified photocatalysts. Our approach is simpler and the degree of enhancement observed is above average of reported values for metal/TiO2 or metal oxide/TiO2 composites (Li and Li, 2002; Vamathevan et al. , 2002; Arabatzis et al. , 2003a; Hong et al. , 2003; Sun et al. , 2003; Zhang et al. , 2003; Sung-Suh et al. , 2004; Ozawa et al. , 2005). Although the cost of fullerenes is pr esently high relative to other chemicals used for enhancing titanium dioxide photocatalys is, their cost has decreased since the beginning
99 of commercial large-scale production and we exp ect further reduction in cost. Furthermore, the quantity of PHF employed is 10 to 100 times lower than other enhancers. The present study suggests that there is an optimum ratio of PHF to TiO2 at which fastest rate for dye degradation is observed. The optimum ratio is much less than monolayer coverage of PHF on TiO2 surface. Fullerenes are known for thei r optical limiting ability, which can potentially block the UV light reaching TiO2 surface. The quite substantial degree of enhancement in photocatalysis achieved with PHF in the present study would seem to c ontradict previous literature. The role of PHF as electron scavenger was confirmed with measurement of production of hydroxyl radicals with electron paramagnetic resonance. The con centration of spin trapped hydr oxyl radicals was higher with PHF than TiO2 alone. PHF have also been reported to generate radicals upon irradiation with UVA or visible light (Kamat et al. , 2000; Vileno et al. , 2004; Pickering and Wiesner, 2005). As previously discussed, free radical generation by PHF alone is orders of magnitude lower than production by TiO2 alone and is thus not responsible for the observed enhancement. This leaves electron scavenging by PHF as th e most plausible explanation. PHF have been employed as antioxidants for ther apeutic applications due to their ability to neutralize both hydroxyl and s uperoxide radicals (Bogdanovic et al. , 2004; Djordjevic et al. , 2004a; Mirkov et al. , 2004). Scavenging of thes e radicals might be expected to decrease, not increase the rate of photocatalysis. However, th e lifetime of hydroxyl radi cals is very short (20 ns, (Grela et al., 1996)), ther efore only those radicals gene rated close to PHF could be scavenged. Superoxide radica ls, which are generated by photogenerated electrons, have a longer lifetime and thus may diffuse to the PH F. However, PHF is adsorbed to the TiO2 surface at the beginning of the experiment (Krishna et al., 2006), presenting an immediate and direct
100 route for electron transfer. Thus, it is likely that the preferred r oute of electron transfer to the PHF is directly from the TiO2, rather than through the diffusi on-mediated superoxide route. PHF usually contains impure groups such as hemiketal, epoxide a nd carbonyl, in addition to hydroxyl groups. The presence of impure groups has been show n to affect the electronic, physical and chemical properties of PHF (Xing et al. , 2004; Rodriguez-Zavala and GuiradoLopez, 2006). The present study with aged form of PHF and freshly synthesized form of PHF revealed that the higher ratio of impure groups to hydroxyl groups is detrimental to both electron accepting ability and stability of PHF molecules. The electron scavenging ability of PHF molecules can be maximized by employing a synthesis methodology, which completely eliminates impure groups on PHF molecules. Anothe r important parameter to be considered is the number of hydroxyl groups present on PHF molecules. Although higher number of hydroxyl groups is preferred to enhance wa ter solubility, numbers greater th an 36 has been experimentally and theoretically shown to reduce th e stability of PHF molecules. Based on the results obtained from present st udy, information about three important design parametersâ€”size of the photocatalytic nano composite, ratio of impure groups to hydroxyl groups in PHF and surface coverage of TiO2 by PHFâ€”was obtained. The design parameters are relevant to the case where either degradation of an organic dye or inactivation of microorganisms is the performance measure. There is a critical size of photocatalyst that is related to the su rface morphology of the target particles, for example th e spacing of fimbriae on a bacterial cell-wall. Photocatalyst that exceeds the critical size cannot co ntact the surface of the target particle because of steric hindrance and therefore would be in effective in photocatalytic degrad ation of that particle. In the case of E. coli as the target particle this size is 100 nm.
101 The degree of hydroxylation and the number of impure groups n eeds to be controlled and maintained in order to ensure favorable performa nce of the PHF. In the present study, a ratio of impure groups to hydroxyl groups of 0.27 was asso ciated with successful enhancement by PHF, whereas a ratio of 1.66 was associated with no enhancement. There is a critical surf ace coverage of the photocatalyst by the enhancer molecules. Where the enhancer molecules have high optical limiti ng property such as PHF, the surface coverage should be less than monolayer. In the case of PHF and a TiO2 agglomerate size of 80 nm, a surface coverage in the range of 2% give s the fastest rates of dye degradation. Overall, PHF-TiO2 combination has the potential to significantly enhance the photocatalytic activity of TiO2 on a commercial scale. Handli ng and processing of PHF does not pose any health or environmental hazard consider ing that hydroxylated fu llerenes are reported to be non-toxic in nature. Application of th is novel nanocomposite for inactivation of microorganisms that are harmful to human hea lth could improve the well-being of society. Suggestions for Future Research Based on the findings of the current research th e first proposed task for future research is optimization of PHF molecules based on theoretical and experimental approach to improve their electron scavenging ability. The impure groups can be eliminated from PHF molecules by employing suitable synthesis parame ters (such as alkali concentr ation, duration and temperature of reaction), which needs to be determined experimentally. Xing et al (2004) have experimentally shown that the concentration of impure groups can be reduced by increasing the concentration of alkali employed for hydroxylat ion. A theoretical approach, similar to Rodriguez-Zavala and Guirado-Lopez (2006), shoul d be employed to predict the stability and electron accepting ability of PH F molecules and to identify th e optimum number of hydroxyl
102 groups. Results of this work would have implicati ons for shelf-life of mate rials that incorporate PHF. The optimized PHF molecules can then be utilized to prepare self-assembled photocatalytic nanocomposites and tested for enhancement in inactiv ation of bacterial endospores. These results will corroborate the sh ape effect hypothesis sugg esting that nano-sized photocatalytic nanocomposites can accelerate inactiv ation of toughest microorganisms as well as microorganisms with surface appendages. The present study revealed that PHF enhances the photocatalytic activ ity of two different TiO2 photocatalysts. Since the mechanism of en hancement is electron scavenging, PHF should be applicable for enhancement in activity of ot her photocatalysts such as zinc oxide, vanadium oxide and cerium oxide (Liu and Yang, 2003; Hernandez-Alonso et al. , 2004; Karunakaran and Senthilvelan, 2005). Zinc oxide is commercially used as an antimicrobial agent in deodorants and air sanitizers. Considerable research has be en carried out on integrating zinc oxide with electron scavengers such gold, platinum and silv er to improve its photo catalytic activity (Gouvea et al. , 2000; Subramanian et al. , 2003b). Zinc oxide has also been coated on multi-wall carbon nanotubes to enhance its photocat alytic activity (Jiang and Gao, 2005). The potential of PHF to enhance these photocatalyst needs to be determined experimentally. PHF can also be applied as electron relay for improving th e efficacy of solar cells. Another potential area for application is in pho tocatalytic coatings. Th e spores of bacteria and fungi in the indoor environment are a major cause for allergy and respiratory problems and are also responsible for nosocomial infections (Utrup et al. , 2003; Schwab and Straus, 2004; Chauhan et al. , 2006; Paterson, 2006; Rice, 2006). The 2004 Center for Disease Control (CDC) summary health statistics for U.S. children stat es that nine million children under the age of 18
103 have been diagnosed with asthma (Dey et al. , 2004). Surfaces of windows, walls, tables etc. in buildings are repositories of bacterial and fungal spores, which can survive extreme environments and germinate on return of favorab le conditions. Our preliminary assessment with swab analysis of walls, windows and table-tops in our meeting r ooms and laboratories indicated that spores of bacteria and fungi are pr esent in a concentration range of 1 cfu/cm2. Common alcohol or hypochlorite disinfect ants are ineffective against s pores (Sagripanti and Bonifacino, 1999). Titanium dioxide has been commercially appl ied as a self-cleaning coating on buildings and glass materials, especially in Ja pan, South Korea and Singapore (Fujishima et al. , 1999). The current state-of-the-art photocat alyst coatings are active agains t specific organic pollutants. Preliminary studies with the best commercially available photocatalyst (D egussa P25) indicated that the photocatalysts can be successfully appl ied as a spray-on coating for degradation of organic pollutants. However they were not effect ive for inactivation of bacterial endospores. The self-assembled nanocomposites with optimized PHF molecules can be applied as spray-on coating, which should improve the rate of inac tivation of bioparticulates present on various illuminated surfaces. The photocatalyst coating on outdoor surfaces can potentially reduce atmospheric pollutants including greenhouse gases thereby contribu ting to a safer and cleaner environment. Furthermore, the self assembled photocatalytic nanocomposites can be applied as a coating on filter surfaces (illuminated with UVA light) present in air circulating ducts. The photocatalyst coating will degrade the bioparticulat es accumulated on the f ilter surface leading to reduction in pressure drop across the filter and therefore reducing the energy loss. Reduction in bioparticulates responsible for asthma, aller gy and sick building s yndrome will improve the indoor living conditions.
104 Another avenue for potentia l application of PHF-TiO2 self-assembled nanocomposite is visible light photocatalysis. Prel iminary dye degradation and bact erial inactivation experiments with these nanocomposites suggested that they are photocatalytica lly active in visible light and further optimization is needed to accelerate the rate of photocatalysis. Kamat and Gevaert have demonstrated charge transfer betw een fullerenes and titanium dioxide in visible light. In contrast to UVA photocatalysis, electrons are transferred from fullerene s to titanium dioxide under visible light illumination. Therefore the functiona l groups contributing to generation of charged state in PHF may be different and should be iden tified by theoretical and experimental approach. PHF with optimized composition can then be utilized for preparing self-assembled nanocomposites. Again optimization of the ratio of PHF to TiO2 is required since the mechanism of electron transfer is different . Alternatively PHF can be integrated with different electron scavengers. The visible light photoc atalytic nanocomposites can be th en tested for their efficacy in degradation of organic pollu tants and inactivation of micr oorganisms. These nanocomposites can also be applied as photocatalytic coatings for indoor and outdoor surfaces.
105 APPENDIX CALCULATION OF SURFACE COVERAGE The surface coverage of PHF on TiO2 was calculated on the basis of following assumptions: PHF molecules are in non-aggregated state. PHF molecules have affinity towards TiO2 surface and all PHF mo lecules adsorb on TiO2 particles. The molecular weight of PHF is 1626 Daltons , based on the average molecular weight determined from empirical formula of fresh and aged PHF. The molecular size of PHF is 1.3 nm (Jeng et al. , 1999). The surface coverage was calculated as follows Basis: 1 gm of TiO2 % 1002 TiO PHFTSA CS n erage SurfaceCov Where PHF PHFMW W n2310 023 . 6 22 PHF PHFd CS 2 2 2 21 210 1TiO TiO TiO TiOA V TSA 3 2 22 3 4 TiO TiOd V 2 2 22 4 TiO TiOd A WPHF = Dosed weight of PHF per gm of TiO2
106 MWPHF = Molecular weight of PHF = 1626 gm mole-1 dPHF = 1.3 nm TiO2 = 3.8 gm cm-3 dTiO2 = 80 nm (size of agglomerate as dete rmined by particle size measurement) Estimated surface coverage for different ratios of dosed PHF to TiO2 are presented in Figure A-1, along with enhancement observed for degradation of Procion Red MX 5B.
107 Figure A-1. Estimated surface coverage and observed enhan cement as a function of dosed ratios of PHF to TiO2. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 0.00010.0010.010.1PHF/TiO2 RatioEnhancement (A.U.)0 50 100 150 200 250Surface Coverage (%)
108 LIST OF REFERENCES Alves, G.C., Ladeira, L.O., Righi, A., Krambroc k, K., Calado, H.D., Gil, R.P.D.F. and Pinheiro, M.V.B. (2006) Synthesis of C-60(OH)(18-20 ) in aqueous alkaline solution under O-2atmosphere. Journal of the Brazil ian Chemical Society . 17 : 1186-1190. Anantram, M.P. and Leonard, F. (2006) Phys ics of carbon nanotube electronic devices. Reports on Progress in Physics . 69 : 507-561. Arabatzis, I.M., Stergiopoulos, T., Andreeva, D. , Kitova, S., Neophytides, S.G. and Falaras, P. (2003a) Characterization and photoc atalytic activity of Au/TiO2 thin films for azo-dye degradation. Journal of Catalysis . 220 : 127-135. Arabatzis, I.M., Stergiopoulos, T., Bernard, M. C., Labou, D., Neophytides, S.G. and Falaras, P. (2003b) Silver-modified titanium dioxide thin fi lms for efficient photodegradation of methyl orange. Applied Catalysis B-Environmental . 42 : 187-201. Atrih, A. and Foster, S.J. (2002) Bacteria l endospores the ultimate survivors. International Dairy Journal . 12 : 217-223. Berger, T., Sterrer, M., Diwal d, O., Knozinger, E., Panayotov, D., Thompson, T.L. and Yates, J.T. (2005) Light-induced charge separation in anatase TiO2 particles. Journal of Physical Chemistry B . 109 : 6061-6068. Block, S.S. (2001) Disinfection, steriliza tion, and preservation . Philadelphia, PA: Lippincott Williams & Wilkins. Block, S.S., Seng, V.P. and Goswami, D.W. (1997) Chemically enhanced sunlight for killing bacteria. Journal of Solar Energy Engin eering-Transactions of the Asme . 119 : 85-91. Bogdanovic, G., Kojic, V., Dordevic, A., Cana danovic-Brunet, J., Vojin ovic-Miloradov, M. and Baltic, V.V. (2004) Modulating activity of fullerol C-60(OH)(22) on doxorubicin-induced cytotoxicity. Toxicology in Vitro . 18 : 629-637. Brezova, V., Stasko, A. and Lapcik, L. (1991) Electron paramagnetic resonance study of photogenerated radicals in titanium-dioxi de powder and its aqueous suspensions. Journal of Photochemistry and Photobiology A-Chemistry . 59 : 115-121.
109 Brune, A., Jeong, G., Liddell, P.A., Sotomura, T., M oore, T.A., Moore, A.L. and Gust, D. (2004) Porphyrin-sensitized nanoparticulate TiO2 as the photoanode of a hybr id photoelectrochemical biofuel cell. Langmuir . 20 : 8366-8371. CDC (2006a) Healthcare-associated infections. http://www.cdc.gov/nci dod/dhqp/healthDis.html . CDC (2006b) Questions a nd answers about anthrax. http://www.bt.cdc.gov/agent/anthrax/faq . Chauhan, N., Latge, J.P. and Calderone, R. (2006) Signalling and oxidant adaptation in Candida albicans and Aspergillus fumigatus . Nature Reviews in Microbiology . 4 : 435-444. Chen, R.J., Bangsaruntip, S., Drouvalakis, K.A., Kam, N.W.S., Shim, M., Li, Y.M. , et al (2003) Noncovalent functionalization of carbon nanotube s for highly specific electronic biosensors. Proceedings of the National Academy of Sc iences of the United States of America . 100 : 49844989. Chen, Y., Cai, R.F., Chen, S.M. and Huang, Z. E. (2001) Synthesis a nd characterization of fullerol derived from C-60(n-) precursors. Journal of Physics and Chemistry of Solids . 62 : 9991001. Chen, Y.W., Hwang, K.C., Yen, C.C. and Lai, Y.L. (2004) Fullerene deriva tives protect against oxidative stress in RAW 264.7 cells and ischemia-reperfused lungs. American Journal of Physiology-Regulatory Integrat ive and Comparative Physiology . 287 : R21-R26. Chiang, L.Y., Upasani, R.B., Swirczewski, J.W. and Soled, S. (1993) Evidence of hemiketals incorporated in the structure of fullero ls derived from aqueous acid chemistry. Journal of the American Chemical Society . 115 : 5453-5457. Criado, J. and Real, C. (1983) Mechanism of th e inhibiting effect of phosphate on the anataserutile transformation induced by thermal and mechanical treatment of TiO2. Journal of Chemical Society, Faraday Transactions . 79 : 2765-2771. Deflandre, B. and Gagne, J.P. (2001) Es timation of dissolved organic carbon (DOC) concentrations in nanoliter samples using UV spectroscopy. Water Research . 35 : 3057-3062. Dey, A.N., Schiller, J.S. and Tai, D.A. (2004) Summary Health Statistics for U.S. Children: National Health Interview Survey, 2002. In: U.S. Department of Health and Human Services Centers for Disease Control and Prevention National Center for Health Statistics, Washington, D.C.
110 Djordjevic, A., CanadanovicBrunet, J.M., Vojinovic-Milo radov, M. and Bogdanovic, G. (2004a) Antioxidant properties and hypothetic radical mechanis m of fullerenol C-60(OH)(24). Oxidation Communications . 27 : 806-812. Djordjevic, A., CanadanovicBrunet, J.M., Vojinovic-Milo radov, M. and Bogdanovic, G. (2004b) Antioxidant properties and hypothetical radical mechanism of fullerenol C60(OH)24. Oxidation Communications . 27 : 806-812. Driks, A. (1999) Bacillus subtilis spore coat. Microbiology and Molecu lar Biology Reviews . 63 : 1. Dvoranova, D., Brezova, V., Mazur, M. and Mala ti, M.A. (2002) Investigations of metal-doped titanium dioxide photocatalysts. Applied Catalysis B-Environmental . 37 : 91-105. Ebbesen, T.W., Hiura, H., Bisher, M.E., Treacy, M.M.J., ShreeveKeyer, J.L. and Haushalter, R.C. (1996) Decoration of carbon nanotubes. Advanced Materials . 8 : 155-157. Foresman, J. (1996) Exploring chemistry with el ectronic structure methods . Pittsburg, PA: Gaussian, Inc. Fujishima, A. and Honda, K. (1972) Electroche mical Photolysis of Water at a Semiconductor Electrode. Nature . 238 : 37-&. Fujishima, A., Hashimoto, K. and Watanabe, T. (1999) TiO2 Photocatalysis: Fundamentals and Applications . Tokyo: BKC Inc. Goswami, T.H., Singh, R., Alam, S. and Mathur, G.N. (2004) Thermal analysis: a unique method to estimate the number of substituents in fullerene derivatives. Thermochimica Acta . 419 : 97104. Gouvea, C.A.K., Wypych, F., Moraes, S.G., Duran, N. and Peralta-Zamora, P. (2000) Semiconductor-assisted photodegradation of ligni n, dye, and kraft effluent by Ag-doped ZnO. Chemosphere . 40 : 427-432. Gratzel, M. (2001) Phot oelectrochemical cells. Nature . 414 : 338-344.
111 Grela, M.A., Coronel, M.E.J. and Colussi, A. J. (1996) Quantitative spin-trapping studies of weakly illuminated titanium dioxide sols. Impl ications for the mechanism of photocatalysts. Journal of Physical Chemistry . 100 : 16940-16946. Guirado-Lopez, R.A. and Rincon, M.E. (2006) St ructural and optical pr operties of highly hydroxylated fullerenes: Stability of molecular domains on the C-60 surface. Journal of Chemical Physics . 125 : -. Guiseppi-Elie, A., Lei, C.H. and Baughman, R.H. (2002) Direct electron transfer of glucose oxidase on carbon nanotubes. Nanotechnology . 13 : 559-564. Guldi, D.M. (2000a) Fullerenes: three di mensional electron acceptor materials. Chemical Communications . 5 : 321-327. Guldi, D.M. (2000b) Fullerenes: three di mensional electron acceptor materials. Chemical Communications : 321-327. Guldi, D.M. and Prato, M. (2000) Excited-st ate properties of C-60 fullerene derivatives. Accounts of Chemical Research . 33 : 695-703. Guldi, D.M., Rahman, G.M.A., Zerbetto, F. a nd Prato, M. (2005) Carbon nanotubes in electron donor-acceptor nanocomposites. Accounts of Chemical Research . 38 : 871-878. Guldi, D.M., Zilbermann, I., Anderson, G., Kotov, N.A., Tagmatarchis, N. and Prato, M. (2004) Versatile organic (fullerene) Inorga nic (CdTe nanoparticle) nanoensembles. Journal of the American Chemical Society . 126 : 14340-14341. Gust, D., Moore, T.A. and Moore, A.L. (2000) Photochemistry of supramolecular systems containing C-60. Journal of Photochemistry and Photobiology B-Biology . 58 : 63-71. Haley, R.W., Culver, D.H., White, J.W., Morgan, W.M. and Emori, T.G. (1985) The Nationwide Nosocomial Infection-Rate a New Need for Vital Statistics. American Journal of Epidemiology . 121 : 159-167. Harwood, C.R. and Cutting, S.M. (1990) Molecular biological methods for Bacillus . Chichester ; New York: Wiley.
112 Henderson, B., Oyston, P.C.F. and ebrary Inc. (2003) Bacterial evasion of host immune responses . Cambridge, UK ; New York, N.Y.: Cambridge University Press. Henriques, A.O. and Moran, C.P. (2000) Structure and assembly of the bacterial endospore coat. Methods . 20 : 95-110. Hernandez-Alonso, M.D., Hungria, A.B., MartinezArias, A., Fernandez-Garcia, M., Coronado, J.M., Conesa, J.C. and Soria, J. (2004) EPR study of the photoassisted formation of radicals on CeO2 nanoparticles employed for toluene photooxidation. Applied Catalysis B-Environmental . 50 : 167-175. Hiura, H., Ebbesen, T.W. and Tanigaki, K. (1995) Opening and Purification of Carbon Nanotubes in High Yields. Advanced Materials . 7 : 275-276. Hoffmann, M.R., Martin, S.T., Choi, W.Y. and Bahnemann, D.W. (1995) Environmental applications of semic onductor photocatalysis. Chemical Reviews . 95 : 69-96. Hong, S.S., Lee, M.S., Park, S.S. and Lee, G.D. (2003) Synthesis of nanosized TiO2/SiO2 particles in the microemulsion and their photo catalytic activity on the decomposition of pnitrophenol. Catalysis Today . 87 : 99-105. Horneck, G., Rettberg, P., Reitz, G., We hner, J., Eschweiler, U., Strauch, K. , et al (2001) Protection of bacterial spores in space, a contribution to the disc ussion on Panspermia. Origins of Life and Evolution of the Biosphere . 31 : 527-547. Hu, C., Yu, J.C., Hao, Z.P. and Wong, P.K. (2003a) Photocatalyt ic degradation of triazinecontaining azo dyes in aqueous TiO2 suspensions. Applied Catalysis B-Environmental . 42 : 4755. Hu, C., Yu, J.C., Hao, Z. and W ong, P.K. (2003b) Effects of acidity and inorganic ions on the photocatalytic degradation of different azo dyes. Applied Catalysis B-Environmental . 46 : 35-47. Hu, C., Tang, Y.H., Jiang, Z., Hao, Z.P., Tang, H.X. and Wong, P.K. (2003c) Characterization and photocatalytic activity of nobl e-metal-supported surface TiO2/SiO2. Applied Catalysis aGeneral . 253 : 389-396. Husebo, L.O., Sitharaman, B., Furukawa, K., Ka to, T. and Wilson, L.J. (2004) Fullerenols revisited as stable radical anions. Journal of the American Chemical Society . 126 : 12055-12064.
113 Iijima, S. (1991) Helical mi crotubules of graphitic carbon. Nature . 354 : 56-58. Ivanova, N., Sorokin, A., Anderson, I., Galleron, N., Candelon, B., Kapatral, V. , et al (2003) Genome sequence of Bacillus cereus and comparative analysis with Bacillus anthracis . Nature . 423 : 87-91. Jacoby, W.A., Maness, P.C., Wolfrum, E.J., Blake, D.M. and Fennell, J.A. (1998) Mineralization of bacterial cell mass on a phot ocatalytic surface in air. Environmental Science & Technology . 32 : 2650-2653. Jeng, U.S., Liu, W.J., Lin, T.L., Wang, L.Y. a nd Chiang, L.Y. (1999) Fractal structure of polyhydroxylated fullerenes in water solutions. Fullerene Science and Technology . 7 : 599-608. Jiang, L.Q. and Gao, L. (2005) Fabrication a nd characterization of ZnO-coated multi-walled carbon nanotubes with enhan ced photocatalytic activity. Materials Chemis try and Physics . 91 : 313-316. Kaiser, H. (2006) Water disinfection, treatment and purification (water quality securing) 20042005-2010-2015. http://www.hkc22.com/waterdisinfection.html . Kamat, J.P., Devasagayam, T.P.A., Priyadar sini, K.I. and Mohan, H. (2000) Reactive oxygen species mediated membrane damage induced by fu llerene derivatives and its possible biological implications. Toxicology . 155 : 55-61. Kamat, P.V., Bedja, I. and Hotchandani, S. (1994) Photoinduced charge-transfer between carbon and semiconductor clusters One-electr on reduction of C-60 in colloidal TiO2 semiconductor suspensions. Journal of Physical Chemistry . 98 : 9137-9142. Karunakaran, C. and Senthilvelan, S. (2005) Va nadia-catalyzed solar pho tooxidation of aniline. Journal of Colloid and Interface Science . 289 : 466-471. Keech, M., Scott, A.J. and Ryan, P.J.J. (1998) Th e impact of influenza and influenza-like illness on productivity and healthcare resource utilization in a working population. Occupational Medicine-Oxford . 48 : 85-90. Keleher, J., Bashant, J., Heldt, N., Johnson, L. a nd Li, Y.Z. (2002) Photo-cat alytic preparation of silver-coated TiO2 particles for antibacterial applications. World Journal of Microbiology & Biotechnology . 18 : 133-139.
114 Klemm, P. (1994) Fimbriae: Adhesion, geneti cs, biogenesis, and vaccines . Boca Raton, FL: CRC Press. Kosmulski, M. (2002) The significance of the diffe rence in the point of zero charge between rutile and anatase. Advances in Colloid and Interface Science . 99 : 255-264. Krishna, V., Noguchi, N., Koopman, B. and M oudgil, B. (2006) Enhancement of titanium dioxide photocatalysis by water-soluble fullerenes. Journal of Colloid and Interface Science . 304 : 166-171. Krishna, V., Pumprueg, S., Lee, S.H., Zhao, J., Sigmund, W., Koopman, B. and Moudgil, B.M. (2005) Photocatalytic disinfection with titani um dioxide coated multi-wall carbon nanotubes. Process Safety and Environmental Protection . 83 : 393-397. Kroto, H.W., Heath, J.R., Obrien, S.C., Cu rl, R.F. and Smalley, R.E. (1985) C-60 Buckminsterfullerene. Nature . 318 : 162-163. Kuhn, K.P., Chaberny, I.F., Massholder, K., S tickler, M., Benz, V.W., Sonntag, H.G. and Erdinger, L. (2003) Disinfection of surfaces by ph otocatalytic oxidation with titanium dioxide and UVA light. Chemosphere . 53 : 71-77. Kymakis, E. and Amaratunga, G.A.J. (2003) Phot ovoltaic cells based on dye-sensitization of single-wall carbon nanotubes in a polymer matrix. Solar Energy Materials and Solar Cells . 80 : 465-472. Lee, S.-H. (2004) Photocatalytic nanocomposites based on TiO2 and carbon nanotubes. University of Florida . Lee, S.H., Pumprueg, S., Moudgil, B. and Si gmund, W. (2005) Inactivation of bacterial endospores by photocatalytic nanocomposites. Colloids and Surfaces B-Biointerfaces . 40 : 93-98. Li, F.B. and Li, X.Z. (2002) The enhancement of photodegradation efficiency using Pt-TiO2 catalyst. Chemosphere . 48 : 1103-1111. Li, J., Takeuchi, A., Ozawa, M., Li, X.H., Sa igo, K. and Kitazawa, K. (1993) C-60 Fullerol formation catalyzed by quater nary ammonium hydroxides. Journal of the Chemical SocietyChemical Communications : 1784-1785.
115 Li, J.G., Ishigaki, T. and Sun, X. (2007) Anatas e, brookite, and rutile nanocrystals via redox reactions under mild hydrothermal conditions: Phase-selective synthesis and physicochemical properties. Journal of Physical Chemistry C . ASAP Article . Liu, H.L. and Yang, T.C.K. (2003) Photocatalytic inactivation of Escherichia coli and Lactobacillus helveticus by ZnO and TiO2 activated with ultraviolet light. Process Biochemistry . 39 : 475-481. Liufu, S.C., Mao, H.N. and Li, Y.P. (2005) Adso rption of poly(acrylic aci d) onto the surface of titanium dioxide and the colloidal stability of aqueous suspension. Journal of Colloid and Interface Science . 281 : 155-163. Lu, Z.X., Zhou, L., Zhang, Z.L., Shi, W.L., Xie, Z.X., Xie, H.Y. , et al (2003) Cell damage induced by photocatalysis of TiO2 thin films. Langmuir . 19 : 8765-8768. LungUSA (2005) Trends in asthma morbid ity and mortality. In. Americal Lung Association:Epidemiology and st atistics unit, R.a.p.s. (ed.). Makarova, T.L. (2001) Electrical and optical properties of pris tine and polymerized fullerenes. Semiconductors . 35 : 243-278. Maness, P.C., Smolinski, S., Blake, D.M., Hu ang, Z., Wolfrum, E.J. and Jacoby, W.A. (1999) Bactericidal activity of photocatalytic TiO2 reaction: Toward an understanding of its killing mechanism. Applied and Environmental Microbiology . 65 : 4094-4098. Mastrapa, R.M.E., Glanzberg, H., Head, J.N., Melo sh, H.J. and Nicholson, W.L. (2001) Survival of bacteria exposed to extreme acceleration: implications for panspermia. Earth and Planetary Science Letters . 189 : 1-8. Matos, J., Laine, J. and Herrmann, J.M. (2001) E ffect of the type of activated carbons on the photocatalytic degradation of aqueous or ganic pollutants by UV-irradiated titania. Journal of Catalysis . 200 : 10-20. Matsunaga, T., Tomoda, R., Nakajima, T. a nd Wake, H. (1985) Photoelectrochemical Sterilization of Microbial-Cells by Semiconductor Powders. Fems Microbiology Letters . 29 : 211-214.
116 McNeill, C.R., Frohne, H., Holdsworth, J.L., Furs t, J.E., King, B.V. and Dastoor, P.C. (2004) Direct photocurrent mapping of organic sola r cells using a near-f ield scanning optical microscope. Nano Letters . 4 : 219-223. Mileikowsky, C., Cucinotta, F.A., Wilson, J. W., Gladman, B., Horneck, G., Lindegren, L. , et al (2000) Natural transfer of viable microbes in sp ace 1. From Mars to Earth and Earth to Mars. Icarus . 145 : 391-427. Mirkov, S.M., Djordjevic, A.N., Andric, N.L., Andric, S.A., Kostic, T.S., Bogdanovic, G.M. , et al (2004) Nitric oxide-scavenging activity of polyhydroxylated fu llerenol C-60(OH)(24). Nitric Oxide-Biology and Chemistry . 11 : 201-207. Miyauchi, M., Ikezawa, A., Tobimatsu, H., Irie, H. and Hashimoto, K. (2004) Zeta potential and photocatalytic activity of nitrogen doped TiO2 thin films. Physical Chemistry Chemical Physics . 6 : 865-870. Mizuki, E., Ohba, M., Ichimatsu, T., Hwang, S.H ., Higuchi, K., Saitoh, H. and Akao, T. (1998) Unique appendages associ ated with spores of Bacillus cereus isolates. Journal of Basic Microbiology . 38 : 33-39. Mohan, H., Chiang, L.Y. and Mittal, J.P. (1997) Radiation chemical investigations on aqueous solutions of C-60(OH)(18). Research on Chemical Intermediates . 23 : 403-414. Mol, O. and Oudega, B. (1996) Molecular and st ructural aspects of fimbriae biosynthesis and assembly in Escherichia coli . FEMS Microbiology Reviews . 19 : 25-52. Montgomery, D.C. (2005) Design and analysis of experiments . Hoboken, NJ: John Wiley & Sons. Nicholson, W.L. and Galeano, B. (2003) UV resistance of Bacillus anthracis spores revisited: Validation of Bacillus subtilis spores as UV surrogates for spores of B. anthracis Sterne. Applied and Environmental Microbiology . 69 : 1327-1330. Nicholson, W.L., Munakata, N., Horneck, G., Melo sh, H.J. and Setlow, P. (2000) Resistance of Bacillus endospores to extreme terrestrial and extraterrestrial environments. Microbiology and Molecular Biology Reviews . 64 : 548-+. NSF Bizarre life forms thrive beneath earth's surface http://www.nsf.gov/news/frontie rs_archive/7-97/7extreme.jsp .
117 Oberdorster, E. (2004) Ma nufactured nanomaterials (Fullerene s, C-60) induce oxidative stress in the brain of juvenile largemouth bass. Environmental Health Perspectives . 112 : 1058-1062. Onoda, K., Watanabe, J., Yoshinori, N. and Izumi, I. (1988) Photocatalytic bactericidal effect of powdered TiO2 on Streptococcus mutans . Denki Kagaku . 56 : 1108-1109. Osawa, E. (1970) Superaromaticity. Kagaku . 25 : 854-863. Ozawa, M., Li, J., Kishio, K., Kitazawa, K., Tadano, A., Aogaki, R. (1995) Electrochemical investigation of adsorbed fullerenol on platinum electr ode in aqueous solutions. Physics and Chemistry of Fullerenes and Derivatives; Procee dings of the Interna tional Winterschool on Electronic Properties of Novel Materials : 171-174. Ozawa, T., Iwasaki, M., Tada, H., Akita, T., Tanaka, K. and Ito, S. (2005) Low-temperature synthesis of anatase-brookite composite nanocrystals: the junc tion effect on photocatalytic activity. Journal of Colloid and Interface Science . 281 : 510-513. Paterson, D.L. (2006) Resistance in Gram-negative bacteria: Enterobacteriaceae . American Journal of Infection Control . 34 : S20-28; discussion S64-73. Paunesku, T., Rajh, T., Wiederrecht, G ., Maser, J., Vogt, S., Stojicevic, N. , et al (2003) Biology of TiO2-oligonucleotide nanocomposites. Nature Materials . 2 : 343-346. Perry, J.J., Staley, J.T. and Lory, S. (2002) Microbial life . Sunderland, Mass.: Sinauer Associates. Pickering, K.D. and Wiesner, M.R. (2005) Fu llerol-sensitized production of reactive oxygen species in aqueous solution. Environmental Science & Technology . 39 : 1359-1365. Popov, V.N. (2004) Carbon nanotubes: properties and application. Materials Science & Engineering R-Reports . 43 : 61-102. Post, J.E. and Burnham, C.W. (1986) Ionic modeli ng of mineral structures and energies in the electron-gas approximation TiO2 polymorphs, quartz, forsterite, diopside. American Mineralogist . 71 : 142-150.
118 Potter, B.B., Wimsatt, J. C. (2005) Determin ation of total organi c carbon and specific UV absorbance at 254 nm in source water and drin king water. In Cinci nnati: National Exposure Research Laboratory, EPA, pp. 1-56. Pumprueg, S. (2004) Inactivation of Bacillus cereus endospores by high-intensity, tunable excimer UV lamp and photocatalytic nanocomposites with 350 nm UV. University of Florida . Pyrgiotakis, G. (2006) Titani a carbon nanotube composites fo r enhanced photocatalysis. University of Florida . Rajh, T., Nedeljkovic, J.M., Chen, L.X., Poluek tov, O. and Thurnauer, M.C. (1999) Improving optical and charge separation pr operties of nanocrystalline TiO2 by surface modification with vitamin C. Journal of Physical Chemistry B . 103 : 3515-3519. Read, T.D., Peterson, S.N., Tourasse, N ., Baillie, L.W., Paulsen, I.T., Nelson, K.E. , et al (2003) The genome sequence of Bacillus anthracis Ames and comparison to closely related bacteria. Nature . 423 : 81-86. Rice, L.B. (2006) Antimicrobial resi stance in Gram-positive bacteria. American Journal of Medicine . 119 : S11-19; discussion S62-70. Riesenman, P.J. and Nicholson, W.L. (2000) Role of the spore coat layers in Bacillus subtilis spore resistance to hydrogen peroxide, arti ficial UV-C, UV-B, a nd solar UV radiation. Applied and Environmental Microbiology . 66 : 620-626. Rincon, A.G. and Pulgarin, C. (2003) Photocatalytical inactivation of E. coli : effect of (continuous-intermittent) light intens ity and of (suspended-fixed) TiO2 concentration. Applied Catalysis B-Environmental . 44 : 263-284. Rodriguez-Zavala, J.G. and Guirado-Lopez, R. A. (2006) Stability of highly OH-covered C-60 fullerenes: Role of coadsorbed O impurities and of the charge stat e of the cage in the formation of carbon-opened structures. Journal of Physical Chemistry A . 110 : 9459-9468. Rohlfing, E.A., Cox, D.M. and Kaldor, A. (1984) Production and characte rization of supersonic carbon cluster beams. Journal of Chemical Physics . 81 : 3322-3330. Sagripanti, J.L. and Bonifacino, A. (1999) Bacter ial spores survive treatment with commercial sterilants and disinfectants. Applied and Environmental Microbiology . 65 : 4255-4260.
119 Saito, R., Fujita, M., Dresselhaus, G. and Dre sselhaus, M.S. (1992a) Electronic-structure of graphene tubules based on C-60. Physical Review B . 46 : 1804-1811. Saito, R., Fujita, M., Dresselhaus, G. and Dre sselhaus, M.S. (1992b) Electronic-structure of chiral graphene tubules. Applied Physics Letters . 60 : 2204-2206. Saito, T., Iwase, T., Horie, J. and Morioka, T. (1992c) Mode of Photocatalytic Bactericidal Action of Powdered Semiconducto r Tio2 on Mutans Streptococci. Journal of Photochemistry and Photobiology B-Biology . 14 : 369-379. Schraff, P., Siegmund, C. and Risch, K. (2005) Characterizati on of water-soluble fullerene C60 oxygen and hydroxyl group derivatives for photosensitizers. Fullerenes, Nanotubes, and Carbon Nanostructures . 13 : 497-509. Schwab, C.J. and Straus, D.C. (2004) The roles of Penicillium and Aspergillus in sick building syndrome. Advances in Applied Microbiology . 55 : 215-238. Sherigara, B.S., Kutner, W. and D'Souza, F. (2003) Electrocatalytic properties and sensor applications of fullerenes and carbon nanotubes. Electroanalysis . 15 : 753-772. Slanina, Z., Lee, S.L., Adamowicz, L. and Ch iang, L.Y. (1996) Structural computations of fullerols C60(OH)x. Proceedings of the Electrochemical Society . 96 : 987-998. So, C.M., Cheng, M.Y., Yu, J.C. and Wong, P.K. (2002) Degradation of azo dye Procion Red MX-5B by photocatalytic oxidation. Chemosphere . 46 : 905-912. Sokal, R.R. and Rohlf, F.J. (1997) Biometry: The principles a nd practice of statistics in biological research. New York: Freeman and Company. Solvay (2007) Cost of influenza. http://www.solvayinfluenza.com/aboutinfluenza/costofinfluenza . Sreethawong, T. and Yoshikawa, S. (2005) Co mparative investigation on photocatalytic hydrogen evolution over Cu-, Pd-, and Au-loaded mesoporous TiO2 photocatalysts. Catalysis Communications . 6 : 661-668.
120 Subramanian, V., Wolf, E.E. and Kamat, P. V. (2003a) Influence of metal/metal ion concentration on the photocat alytic activity of TiO2-Au composite nanoparticles. Langmuir . 19 : 469-474. Subramanian, V., Wolf, E.E. and Kamat, P.V. (2003b) Green emission to probe photoinduced charging events in ZnO-Au nanoparticles. Char ge distribution and ferm i-level equilibration. Journal of Physical Chemistry B . 107 : 7479-7485. Sun, B., Vorontsov, A.V. and Smirniotis, P.G. (2003) Role of platinum deposited on TiO2 in phenol photocatalytic oxidation. Langmuir . 19 : 3151-3156. Sunada, K., Kikuchi, Y., Hashimoto, K. and Fujis hima, A. (1998) Bacteric idal and detoxification effects of TiO2 thin film photocatalysts. Environmental Science & Technology . 32 : 726-728. Sung-Suh, H.M., Choi, J.R., Hah, H.J., Koo, S.M. and Bae, Y.C. (2004) Comparison of Ag deposition effects on the photocataly tic activity of nanoparticulate TiO2 under visible and UV light irradiation. Journal of Photochemistry and Photobiology A-Chemistry . 163 : 37-44. Taylor, R. (1999) Lecture notes on fullerene chemistry . London: Imperial College Press. Tsang, S.C., Chen, Y.K., Harris, P.J.F. and Gree n, M.L.H. (1994) A simple chemical method of opening and filling carbon nanotubes. Nature . 372 : 159-162. Uchino, T., Tokunaga, H., Ando, M. and Utsumi, H. (2002) Quantitative determination of OH radical generation and its cytotoxicity induced by TiO2-UVA treatment. Toxicology in Vitro . 16 : 629-635. USAID Avian influenza and its global implications. http://www.usaid.gov/about_us aid/acvfa/intro_ai.pdf . Utrup, L.J., Werner, K. and Frey, A.H. (2003) Minimizing pathogenic bact eria, including spores, in indoor air. Journal of Environmental Health . 66 : 19-26, 29. Vamathevan, V., Amal, R., Beydoun, D., Low, G. and McEvoy, S. (2002) Photocatalytic oxidation of organics in wate r using pure and silver-modified titanium dioxide particles. Journal of Photochemistry and P hotobiology a-Chemistry . 148 : 233-245.
121 Vileno, B., Sienkiewicz, A., Lekka, M., Kulik, A.J. and Forro, L. (2004) In vitro assay of singlet oxygen generation in the presence of water-soluble derivatives of C-60. Carbon . 42 : 1195-1198. Vileno, B., Marcoux, P.R., Lekka, M., Sienkiewi cz, A., Feher, T. and Forro, L. (2006) Spectroscopic and photophysical properties of a highly derivatized C-60 fullerol. Advanced Functional Materials . 16 : 120-128. Wang, C., Wang, T.M. and Zheng, S.K. (2002) Investigation of th e photoreactivity of nanocrystalline TiO2 thin film by ion-implantation technique. Physica E . 14 : 242-248. Wang, X.H., Li, J.G., Kamiyama, H. a nd Ishigaki, T. (2006) Fe-doped TiO2 nanopowders by oxidative pyrolysis of organom etallic precursors in induction thermal plasma: synthesis and structural characterization. Thin Solid Films . 506 : 278-282. WHO (2007) Spending on health: A global overview. http://www.who.int/mediacentre/ factsheets/fs319/en/index.html . Xie, Q.S., Perezcordero, E. and Echegoyen, L. ( 1992) Electrochemical detection of C-60(6-) and C-70(6-) Enhanced stability of fullerides in solution. Journal of the American Chemical Society . 114 : 3978-3980. Xing, G.M., Zhang, J., Zhao, Y.L., Tang, J., Zhang, B., Gao, X.F. , et al (2004) Influences of structural properties on stability of fullerenols. Journal of Physical Chemistry B . 108 : 1147311479. Yan, M.C., Chen, F., Zhang, J.L. and Anpo, M. ( 2005) Preparation of controllable crystalline titania and study on the phot ocatalytic properties. Journal of Physical Chemistry B . 109 : 86738678. Yu, Y., Yu, J.C., Chan, C.Y., Che, Y.K., Zhao, J.C., Ding, L. , et al (2005a) Enhancement of adsorption and photocatalytic activ ity of TiO2 by using carbon nanotubes for the treatment of azo dye. Applied Catalysis B-Environmental . 61 : 1-11. Yu, Y., Yu, J.C., Yu, J.G., Kw ok, Y.C., Che, Y.K., Zhao, J.C. , et al (2005b) Enhancement of photocatalytic activity of mesoporous TiO2 by using carbon nanotubes. Applied Catalysis AGeneral . 289 : 186-196. Zeng, Y., Liu, J.T., Wu, W. and Ding, C.X. ( 2005) Photocatalytic performance of plasma sprayed TiO2-ZnFe2O4 coatings. Surface & Coatings Technology . 200 : 2398-2402.
122 Zhang, L.Z., Yu, J.C., Yip, H.Y., Li, Q., Kwong, K.W., Xu, A.W. and Wong, P.K. (2003) Ambient light reduction strategy to synthesize silver nanoparticles and silver-coated TiO2 with enhanced photocatalytic a nd bactericidal activities. Langmuir . 19 : 10372-10380. Zhao, G.M. and Wang, Y.S. (2001) Possible su perconductivity above 400 K in carbon-based multiwall nanotubes. Condensed Matter . 2 . Zheng, H., Maness, P.C., Blake, D.M., Wolfrum, E.J., Smolinski, S.L. and Jacoby, W.A. (2000) Bactericidal mode of titani um dioxide photocatalysis. Journal of Photochemistry and Photobiology a-Chemistry . 130 : 163-170.
123 BIOGRAPHICAL SKETCH Vijay Krishna was born on August 28, 1979, in Ba ngalore, India. He completed his high school education in 1996 from Kendriya Vidyalaya, Delhi. In September 2000, he graduated from B.M.S. College of Engineering (BMSCE), Bangalore University, India with a bachelorâ€™s degree in chemical engineering. In October 2000, he joined Unile ver Research India and worked with the global laundry bar re search team. He conducted rese arch on development of novel structuring systems and density re duction of laundry bars, which re sulted in three patents. Vijay decided to pursue higher studies and joined the Department of Materials Science and Engineering at University of Florida in August 2002. He earned the degree of Master of Science in 2004 with specialization in biom aterials. He has been working on the topic of this dissertation and expects to graduate with a Doct or in Philosophy in the spring of 2007.