Photocatalytic Reduction of Trace Aqueous Mercury Using a Visible Light Source

MISSING IMAGE

Material Information

Title:
Photocatalytic Reduction of Trace Aqueous Mercury Using a Visible Light Source
Physical Description:
1 online resource (72 p.)
Language:
english
Creator:
Gonzaga, Erica B W
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.E.)
Degree Grantor:
University of Florida
Degree Disciplines:
Environmental Engineering Sciences
Committee Chair:
MAZYCK,DAVID W
Committee Co-Chair:
CHADIK,PAUL A
Committee Members:
BONZONGO,JEAN-CLAUDE J
SIGMUND,WOLFGANG MICHAEL

Subjects

Subjects / Keywords:
aqueous -- light -- mercury -- photocatalysis -- visible
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre:
Environmental Engineering Sciences thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Mercury emissions may originate from both natural and anthropogenic sources. As regulations become stricter, refined removal processes may be required to reduce trace concentrations. The photocatalytic properties of semiconducting materials have been investigated extensively using ultraviolet, UV, light, but high costs associated with UV energy make photocatalysis an impractical treatment method. Studies were conducted to investigate the potential of using photocatalysis for the reduction of trace aqueous mercury using a visible light source. A silica titania composite, STC, was mixed with tungsten oxide, WO3, to determine if co-mixing semi-conductors may improve the photocatalytic potential. Removal of mercury increased slightly in the comixing studies compared to single catalysts, but it was determined that enhanced adsorption and not photocatalytic reduction contributed to better removal. However, positive results from STC experiments using visible light lead to further investigation of its photocatalytic potential. Experiments were conducted to understand the effect of contact time, catalyst dose, dissolved oxygen and the presence of an organic oxygen scavenging compound, formic acid, using STC and TiO2. More than 95 percent removal was achieved within 15 minutes using STC when formic acid and/or a nitrogen purge were utilized. Detailed 5 minute experiments demonstrated that the presence of formic acid is the most important factor affecting removal; acting as a radical a scavenger and preventing oxidation of mercury. With formic acid present, reduction of mercury ranged between 82 to 98 percent for different STC concentrations. The presence of dissolved oxygen affected removal negatively only when formic acid was used; reducing removal by approximately 40 percent. Lastly, the effect of initial concentration of mercury was investigated. Visible light was effective at photoreducing mercury by more than 80 percent at low concentrations of 50 and 100 parts per billion using STC. There was no statistical difference between removals with visible light compared to UV 254 nm at these concentrations. However, removal rates decreased drastically as mercury concentrations increased to 500 and 1000 ppb as hypothesized. At 1000 ppb, mercury removal was only about 15 percent using visible light compared to more than 95 percent using UV 254 nm. The reduction of mercury is directly proportion to available electrons from the photocatalyst.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Erica B W Gonzaga.
Thesis:
Thesis (M.E.)--University of Florida, 2014.
Local:
Adviser: MAZYCK,DAVID W.
Local:
Co-adviser: CHADIK,PAUL A.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2014
System ID:
UFE0046804:00001


This item is only available as the following downloads:


Full Text

PAGE 1

PHOTOCATALYTIC REDUCTIO N OF TRACE AQUEOUS MERCURY USING A VISIBLE LIGHT SOURCE By ERICA WALLACE GONZAGA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR TH E DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2014

PAGE 2

2014 Erica Wallace Gonzaga

PAGE 3

To my family

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my advisor, Dr. David Mazyck, for opening the doors to research and offering an opportunity to work with his graduate students when I will still completing my undergraduate degree. I am also grateful for the guidance he has offered in the past two years of graduate school. I would like to thank Dr. Jean Claude Bonzongo, Dr. Paul Chadik and Dr. Wolfgang Sigmund for serving on my committee. I am grateful for the wonderful research team and colleagues that have helped me throughout graduate school: Christine Valcarce, Amy Borello, Sanaa Jaman and Ana Maria Hagan. Thank you to all CCI employees, especially Regina Rodriguez, for offering assistance whenever I needed it. I am grateful for the help and hard work of m y undergraduate assistant Sutanya Singivipulya. Lastly, I would like to thank my husband Luiz Gonzaga for being patient and understanding. You have been the most supportive partner I could ask for and I am grateful for having you in my life.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 Openin g Statement ................................ ................................ ................................ 12 Objectives ................................ ................................ ................................ ............... 13 2 LITERATURE REVIEW ................................ ................................ .......................... 14 Mercury ................................ ................................ ................................ ................... 14 Sources ................................ ................................ ................................ ................... 14 Regulations ................................ ................................ ................................ ............. 15 Health Impacts ................................ ................................ ................................ ........ 16 Chemistry ................................ ................................ ................................ ................ 18 Photocatalysis ................................ ................................ ................................ ......... 19 Oxidation Reactions ................................ ................................ ................................ 20 Reduction Reactions ................................ ................................ ............................... 21 Creating a Reducing Environment ................................ ................................ .......... 22 Visible Light Photoc atalysis ................................ ................................ .................... 23 3 EXPERIMENTAL ................................ ................................ ................................ .... 27 STC Synthesis ................................ ................................ ................................ ........ 27 Co mixing Stud ies ................................ ................................ ................................ ... 27 STC and TiO 2 Comparison Studies ................................ ................................ ........ 28 Sample Collection and Analysis ................................ ................................ .............. 29 4 PHOTOCHEMICAL REMOVAL OF AQUEOUS MERCURY BY CO MIXING CATALYSTS ................................ ................................ ................................ ........... 33 Control Studies ................................ ................................ ................................ ....... 33 Effect of Visible Light and Nitrogen Purge ................................ .............................. 33 5 STC AND TiO 2 TIME DEPENDENT STUDIES ................................ ....................... 41 Control Studies ................................ ................................ ................................ ....... 41

PAGE 6

6 Effects of F ormic Acid ................................ ................................ ............................. 41 Effects of Nitrogen Purge ................................ ................................ ........................ 42 TiO 2 Comparison Studies ................................ ................................ ....................... 43 Comp arison of Visible Light and UV 254 ................................ ................................ 43 6 OPTIMIZATION DOSE DEPENDENT STUDIES ................................ .................... 50 Control Studies and Dose Determination ................................ ................................ 50 Effect of Formic Acid ................................ ................................ ............................... 50 Effect of Nitrogen Purge ................................ ................................ .......................... 52 7 MERCURY CONCENTRATION DEPENDENT STUDIES ................................ ...... 60 Control Studies ................................ ................................ ................................ ....... 60 Comparison of Visible Light vs. UV 254 ................................ ................................ .. 60 8 CONCLUSIONS ................................ ................................ ................................ ..... 65 LIST OF REFERENCES ................................ ................................ ............................... 67 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 72

PAGE 7

7 LIST OF TABLES Table page 2 1 Band gap energies f or different photocatalysts ................................ ................... 26 3 1 Catalyst Surface Characterization ................................ ................................ ...... 31 3 2 Equivalent catalyst dose. ................................ ................................ .................... 32 6 1 Catalyst doses for dose dependent studies. ................................ ....................... 54

PAGE 8

8 LIST OF FIGURES Figure page 3 1 Spectral power distribution provided by the manufacturer [57] ........................... 31 3 2 Visible light setup (profile view). ................................ ................................ ......... 31 3 3 Reactor setup employed in UV 254 studies ................................ ........................ 32 3 4 Experimental conditions for STC and TiO 2 comparison studies. ........................ 32 4 1 Time dependent control studies in the absen ce of light ................................ ...... 37 4 2 Photoreduction of mercury using STC under visible light in the presence of formic acid. ................................ ................................ ................................ ......... 37 4 3 Photoreduction of mercury using WO 3 under visible light in the presence of formic acid. ................................ ................................ ................................ ......... 38 4 4 Photoreduction of Hg: effects of co mixing silica titania composite and tungsten oxide in the presence of formic acid.. ................................ ................... 38 4 5 Comparison of catalysts in the absence of a nitrogen purge.. ............................ 39 4 6 Comparison of silica titania composite alone and co mixed silica titania and tungsten oxide under nitrogen purge in the presence of formic acid .................. 39 4 7 Photoreduction of mercury: comparison of all catalysts under nitrogen purge in the presence of formic acid ................................ ................................ ............. 40 5 1 STC and TiO 2 time dependent control studies ................................ .................... 45 5 2 Photoreduction of mercury using STC under nitrogen purge. ............................. 45 5 3 Mercury oxidation in the absence of formic acid. ................................ ................ 46 5 4 Photoreduction of mercury using S TC without nitrogen purge ........................... 46 5 5 Photoreduction of mercury using STC in the presence of formic acid. .............. 47 5 6 Photoreduction of mercury using STC without formic acid ................................ 47 5 7 Comparison of STC and TiO2 with nitrogen purge and formic acid .................... 48 5 8 Photoreduction of mercury using TiO 2 under ni trogen purge. ............................ 48 5 9 Photoreduction of mercury using STC under vis ible and UV 254 light. .............. 49

PAGE 9

9 6 1 Effects of formic acid on photoreduction of mercury using STC without a purge.. ................................ ................................ ................................ ................ 54 6 2 Effects of formic acid on photoreduction of mercury u sing STC under nitrogen purge ................................ ................................ ................................ ... 55 6 3 Effects of formic acid on photoreduction of mercury using TiO 2 without nitrogen purge. ................................ ................................ ................................ ... 55 6 4 Effects of formic acid on photoreduction of mercury using TiO 2 under nitrogen purge.. ................................ ................................ ................................ .. 56 6 5 Effects of nitrogen purge on photoreduction of mercury using STC with form ic acid ................................ ................................ ................................ .................... 56 6 6 Effe cts of nitrogen purge on photoreduction of mercury using STC with no formi c acid ................................ ................................ ................................ .......... 57 6 7 Effects of nitrogen purge of photoreductio n of mercury using TiO 2 with formic acid. ................................ ................................ ................................ .................... 57 6 8 Effects of nitrogen purge on photoreduction of mercury using TiO 2 without formic acid. ................................ ................................ ................................ ......... 58 6 9 Photoreduction of mercury using STC: comparison on all scenarios. ................. 58 6 10 Photoreduction of mercury using TiO 2 : comparison of all scenarios. .................. 59 7 1 Mercury concentrati on dependent control studies. ................................ ............ 62 7 2 Comparison of UV 254 vs. visible light using STC.. ................................ ............ 62 7 3 Comparison of UV 254 vs. visible light using TiO 2 ................................ .............. 63 7 4 Comparison of visible light vs. UV 254 using STC at lower Hg concentrations .. 63 7 5 Representation of photocatalytic reduction of high concentrations of aqueous mercury. ................................ ................................ ................................ ............. 64

PAGE 10

10 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requir ements for the Degree of Master of Engineering PHOTOCATALYTIC REDUCTION OF TRACE AQUEOUS MERCURY USING A VISIBLE LIGHT SOURCE By Erica Wallace Gonzaga May 2014 Chair: David Mazyck Major: Environmental Engineering Sciences Mercury emissions may originate from both natural and anthropogenic sources. As regulations beco me stricter refined removal processes may be required to reduce trace concentrations. The photocatalytic properties of semi conducting materials have been investigated extensively using ultra violet (UV) light, but high costs associate d with UV energy make photocatalysis an impractical treatment method. Studies were conducted to investigate the potential of using photocatalysis for the reduction of trace aqueous mercury using a visible light source. A silica titania composite (STC) was mixed with tungsten oxide (WO 3 ) to determine if co mixing semi conductors may improve the photocatalytic potential. Removal of mercury increased slightly in the co mixing studies compared to single catalysts, but it was determined that enhanced adsorption and not photocatalytic reduct ion contributed to better removal. However, positive results from STC experiments using visible light lead to furthe r investigation of its photocata lytic potential Experiments were conducted to unders tand the effect of contac t time, catalyst dose, dissolved oxygen and the presence of an organic oxygen scavenging compound (formic acid) using STC and TiO 2 More than 95% removal was achieved within 15 minutes

PAGE 11

11 using STC when formic acid and/or a nitrogen purge were utilized. Detai led 5 minute experiments demonstrated that the presence of formic acid is the most important factor affecting removal; acting as a radical a scavenger and preventing re oxidation of mercury. With formic acid present, reduction of mercury ranged between 82 98% for different STC concentrations. The presence of dissolved oxygen affected removal negatively only when formic acid was used; reducing removal by approximately 40%. Lastly, the effect of initial concentration of mercury was investigated. Visible lig ht was effective at photoreducing mercury by more than 80% at low concentrations of 50 and 100 parts per billion using STC. There was no statistical difference between removals with visible light compared to UV 254 nm at these concentrations. However, remo val rates decreased drastically as mercury concentrations increased to 500 and 1000 ppb as hypothesized. At 1000 ppb, mercury removal was only about 15% using visible light compared to more than 95% using UV 254 nm The reduction of mercury is directly pro portion to available electrons from the photocatalyst.

PAGE 12

12 CHAPTER 1 INTRODUCTION Opening Statement The toxicity of mercury was fir st recognized in the early 1950 s when a chemical plant discharged an estimated 70 150 tons of inorganic mercury into Minamata Bay in Japan. The fish in the bay displayed mercury concentrations of approximately 15 ppm [1] The local residents relied heavily on the consumption of fish and started developing methylmercury poisoning symptoms. Some of these symptoms included impairment of speech, muscle weakness, vision problems and numbness of hands and feet [2] Awareness about the toxicity of mercu ry has increased since the 1950 s and countries all around the world are working on implementing mercury emission regulations. In the United States, the regulatory limit of mercury in drinking water is 0.002 mg/L; a concentration believed to not affect human health [3] In 2013, an international treaty known as Minamata Convention was signed by 140 countries, including the United States, pledging to regulate and reduce merc ury emissions [4] Photocatalytic treatment of heavy metals has been gaining more attention from the scientific community in the past couple of decades. Unlike other treatment process such as precipitation and adsorption, phot ocatalytic oxidation/reduction is capable of reducing contaminants to trace or nearly undetectable levels. When a catalyst is exposed to ultra violet light, the photon energy excites an electron from the valence band causing it to jump to the conduction ba nd. This is known as the electron hole pair. This pair is responsible for generating redox reactions if recombination of the electron does not occur.

PAGE 13

13 Many studies have focused on the modification of a photocatalyst, such as titanium dioxide (TiO 2 ), for us e of visible light. Most of these studies have characterized the catalyst, measuring light absorbance/reflectance and binding energy, but there has not been substantial research on applying these modified catalysts in treatment processes. It has been noted that mercury has been reduced in natural systems exposed to sunlight, which contains less than 5% UV rays [5] Other studies have questioned whether just a small percentage of UV photons are necessary to induce photocatalytic reactions. The purpose of this study herein is to investigate the potential of photocatalytic reduction of trace aqueous mercury under visible light. Since most modification techniques include doping the catalyst with another metal, the effect of co mixin g catalysts (two semi conductor metals) will be investigated. Most other studies have tested visible light catalysis with high concentration of contaminants (ppm range). If the formation of electron hole pairs is directly proportional to the intensity of l ight, and the reduction of mercury depends on the number of electron hole pairs, then it is hypothesized that visible light (which emits low levels of UV light) may be effective at reducing low concentrations of mercury (ppb range). Objectives Investigate the potential of using co mixed catalysts in the photocatalytic reduction of aqueous mercury under visible light. Optimize the visible light photocatalytic process using a silica titani a composite by controlling dissolved oxygen, organic concentration an d catalyst concentration. Compare the performa nce of titania immobilized with silica to conventional titanium dioxide exposed to visible light Investigate the effects of initial concentration of mercury on visible ligh t photocatalysis.

PAGE 14

14 CHAPTER 2 LITERATURE REVIEW Mercury Mercury is present naturally in the environment and exits in three different forms that vary in toxicity: elemental, organic, and inorganic. The elemental form of mercury is released into the environment naturally through erosion and volcanic eruptions but anthropogenic sources are responsible for appro ximately 70% of mercury ( Hg ) released today [6] emissions are still debated today. Sources Most of the mercury transport in the atmosphere originates from anthropogenic sources. The main industrial activi ties that contribute to emissions include gold mining, pharmaceutical production, manufacturing of light bulbs and thermometers, metal processing and chlor alkali chlorine production facilities among others. By far, the largest source of anthropogenic Hg emissions is coal fired power plants, accounting for close to 50% of industrial discharges [6] [7] The combustion of coal is responsible for the emi ssion of both elemental and oxidized forms of mercury. Divalent Hg 2+ is highly reactive and has a lifetime in the atmosphere of only a few days. It is deposited on land or in water close to its point of emission. Once deposited, oxidize d mercury has the po tential to transform into methylmerucy; a highly toxic, organic form of mercury. On the other hand, elemental mercury has a lifetime in the atmosphere of approximately one year and can travel long distances, contaminating areas far from the point of emissi on.

PAGE 15

15 In the United States, mercury concentration in coal ranges from 0.8 to 1 g/g [7] During combustion, mercury is released from coal as its gaseous, elemental form. The total concentration of mercury in flue gas emissions r anges from 1 to 35 g/N m 3 [8] As the flue gas cools, part of the gaseous mercury gets oxidized to HgO, HgCl 2 or Hg 2 Cl 2 Wet scrubbers or dry sorbent injection are common treatment processes used to remove the oxidized mercury. If not captured or treated, the remaining 50% of elemental mercury gets released to the atmosphere. Contamination of water can occur via atmospheric transport or direct industrial discharge. Mercury emitted hundreds, or even thousands, of miles away can be discharged to the soil or a body of water through rainfall, thus contaminating drinking water sources. Elemental mercury released into the environment can get oxidized by ozone in the atmosphere to its divalent form from where it can be returned to the surface of the Earth through wet or dry deposition [9] Once deposited, it may bind with other compounds on the surface or is re emitted to the atmosphere as gaseous mercury. Regulations In the United States, mercury emissions are currently regulated by the Environmental Protection Agency (EPA) under the Safe Drinking Water Act (SDWA), the Clean Water Act (CWA), the Clean Air Act (CAA), and the Resource Conservation and Recovery Act (RCRA). In 2005, the EPA created the Clean A ir Mercury Rule under the CAA with the intention of reducing emissions for coal fired power plants by setting caps. The rule was voided in 2009 when the D.C. Circuit removed power plants from the list of hazardous contaminant sources of the CAA [10] In 2012, the EPA issued the Mercury and Air

PAGE 16

16 Toxic Standards (MATS) with the intention of reducing air pollutants for oil and coal fired power plants. The proposed rules will regulate emissions from plants generating 25 megawatts or more and allows each plant up to 4 years to comply with the technology based standards [11] In 1992, the EPA issued the Impaired Waters and Total Maximum Daily Load (TMDL) Regulations which limits the amount of contaminants t hat may be discharged into a body of water. Point source discharges are controlled by the National Pollutant Discharge Elimination System (NPDES) in which permits are issued by individual states [12] ; therefore the TMDL for me rcury varies from state to state. The drinking water in the United States is protected by the SDWA, which was passed in 1974 [3] It sets standards and ensures the quality of municipal drinking water around the country. The re gulation of mercury under the SDWA began in 1992 and the maximum contaminant level (MCL) in drinking water is 0.002 mg/L or 2 ppb [3] In 2013, the Minamata Convention on Mercury was created to reduce anthropogenic emissions of mercury around the globe. A 2001 global assessment of mercury conducted by the United Nations Environmental Programme (UNEP) provided sufficient evidence on the ne gative impacts of mercury. Negotiations started in 2009 and after 4 years the treaty was finalized and signed by over 140 countries including the United States [4] Health Impacts Exposure to mercury can occur via inhalation, ingestion, or through dermal contact. Mercury may exist as a cation which gives it the potential to react with most ligands in living cells. Elemental mercury (Hg 0 ) is the least toxic species of the heavy

PAGE 17

17 metal. If ingested, it is a stable, unreactive elem ent in the body that will pass through the digestive system and be eliminated without causing harm. In bodies of water, mercury undergoes biomethylation by specific organisms to form one of the most toxic species known as methylmercury (MeHg) [13] It is suggested th at methylmercury reacts with cysteine to form a compound that mimics the essential amino acid methionine [14] Methionine is capable of entering the brain barrier and is always the first amino acid to start the polypeptide chain [14] Therefore, when methylmercury enters the body, it is capable of penetrating the brain and inhibiting protein synthesis. MeHg is ingested by fish and other aquatic organisms whi ch in turn are consumed by humans. Mercury binds to proteins and lipids and can cause serious harm to the brain and nervous system. In adults, mercury contamination may cause headaches, nausea, vomiting, numbness, and difficulty walking, seeing, speaking a nd hearing among other symptoms [15] Fetuses and young children are more susceptible to the damaging effects of mercury and exposure may cause a series of developmental issues including cerebral palsy, retardation and seizures [16] The toxicity of MeHg wa s not recognized until the 1950 s when the population of Minamata, Japan began developing serious neurological problems. Contamination of the water occurred via wastewater discharge from a Chisso Corpo ration plant that manufactured fertilizers [17] The local population relied heavily on the contaminated fish as their source of protein. Adults developed serious symptoms affecting the neurological system, similar to epilepsy, known as Minamata disease. 30 cases of

PAGE 18

18 c erebral palsy were reported due to mothers who consumed the contaminated fish during pregnancy [18] Chemistry Mercury is a heavy metal that exists in three oxidation states, Hg 0 Hg + and Hg 2+ Gaseous mercury in the atmosphere exists primarily in its elemental state, Hg 0 The divalent form is found in dissolved w aters or attached to particles [19] Therefore the oxidized form of mercury, Hg 2+ tends to be discharged from the atmosphere through precipitation, close to its source of emission [20] On the other hand, Hg 0 is less soluble and will travel long distances across the globe before it is oxidized and released back to the earth. The lifetime of Hg 0 in the atmosphere could last up to 2 years [20] Across varying pH values and chloride concentrations, the divalent form of mercury will complex with hydroxide and chloride ions in surfa ce waters with an oxic environment. Hg will comple x with hydroxide ions at pH values greater than 3 or when there is no chloride present and will hydrolyze to HgOH + and Hg(OH) 2 according to Equations ( 2 1 and ( 2 2 [21] ( 2 1 ) ( 2 2 ) HgOH + is predominant in the lower pH range of 2.2 to 3.8 while Hg(OH) 2 occurs at pH above 3.8 and being the only hydrolyzed species at pH 6 and above. Examining the various hydrolyzed species of mercury suggests the degree of sol ubility of the heavy metal. At natural water pH ranges, mercury will be present as a soluble hydrolyzed complex rather than free ions. However, divalent mercury has a large affinity for complexation with chloride rather than hydroxide in oxic environments At lower pH values, the Hg Cl complexation

PAGE 19

19 will form four species depending on the concentration of Cl present in solution: HgCl + HgCl 2 HgCl 3 and HgCl 2 4 [22] where HgCl 2 is the predominant species. An intermediate speci es, HgClOH may be formed with in a small window of precise pH and chloride concentration values As discussed in subsequent sections, Hg(NO 3 ) 2 was used in this study. The compound ionizes into NO 3 and Hg 2+ Since no other ions are present in solution, the mercury ions will hydrolyze as seen in Figure 2 1 [23] Photocatalysis Photocatalytic oxidation has been widely used for the treatment of organic compounds in water and air streams. In the past 20 years, interest has been increasingly shifting to the potential of using photocatalysis in the reduction of heavy metals from aque ous waste streams. Many metal oxides display photocatalytic properties. These semiconductors contain a void area, where no energy levels are available, known as the band gap [24] When the semiconductor is exposed to light at a specific wavelength, an electron jumps the band gap from the valence band to the conduction band producing a positively charged hole ( h + ) in the valence band and a photoexcited electron (e ) in the conduction band. This is known as an electron hole pair ( Eq uation ( 2 3 ) [25] The amount of energy required to produce an electron hole pair must be equal to or greater than the band gap energy, in most cases UV or near UV light is required [24] Recombination of the electron to the valence band can occur almost instantaneously if no other reactions take place. Instead of recombination, the electron hole pair may parti cipate in redox reactions with molecules adsorbed to the surface of the catalyst. The positively charged hole becomes available to oxidize any

PAGE 20

20 nearby compounds while the negatively charged conduction band promotes a reduction reaction. Oxidation Reactions Valence band holes may directly oxidize a target pollutant or create hydroxyl radicals in the water. Hydroxyl radicals are believed to be the primary oxidants in photocatalytic oxidation and may be produced through various mechanisms. Water molecules adso rbed by the catalyst or OH groups on the surface of the catalyst may interact with the valence band to produce hydro xyl radicals as described by Equations ( 2 4 and ( 2 5 [26] [27] [28] ( 2 3 ) ( 2 4 ) ( 2 5 ) It is also suggested that hydroxyl radicals may be formed by the interacting of oxygen with the conduction band electron to form a superoxide ion which then transf orms into hydrogen pe roxide as described by Equations ( 2 6 through ( 2 10 [29] The hydrogen peroxide molecules are th en broken into hydroxyl radicals by interaction with the conduction band electron, by reacting with the superoxide ion or by direct photolysis (Equations ( 2 11 through ( 2 13 ) [28] [30] [31] ( 2 6 ) ( 2 7 ) ( 2 8 ) ( 2 9 ) ( 2 10 )

PAGE 21

21 ( 2 11 ) ( 2 12 ) ( 2 13 ) Reduction Reactions Many studies have focused on the photooxidation of organic contaminants, but slowly there has been an att ention shift to investigating the photoreduction of heavy metals and inorganic compounds [32] For photoreduction to be thermodynamically achievable, the reduction potential of the metal ion must greater than the energy of the conduction band of the photocatalyst [33] [34] The size of the gap between the conduction (cb) and valence (vb) bands will remain constant, but the energy levels will alter with a change in pH. T he di fference in energy levels is quantified by Eq uations ( 2 14 and ( 2 15 [35] [36] ( 2 14 ) ( 2 15 ) According to the equations above, the photoreduction of mercury should increase with an increasing pH. It has been discussed that the photoreduction of a metal ion is directly proportional to the adsorption of the metal on to the surface of the photocataly st [37] ; therefore increasing the amount of metal ions adsorbed should increase the rate of photoreduction. However, an increased in absorbed metal ions increases the chances of photooxidation by the valence band hole [34] [38] The following products are produced through the photocatalytic reductions of varying mercury salts in neutral or acidic pH (Equations 2 16 through 2 18) [39] :

PAGE 22

22 ( 2 16 ) ( 2 17 ) ( 2 18 ) At a high pH, elemental mercury is oxidized acc ording to Equation 2 19 [39] : ( 2 19 ) The reduced form of mercury may be eliminated from the system through volatilization or removed through adsorption on to the catalysts. Creating a Reducing Environment Treatment techniques may be tailored to favor oxidation or reduction reactions. A reducing environment is preferable when treating aqueous mercury. Organic materials may be added to act as a hole scavenger and/or radica l sink. In the presence of nitrogen purge, formic acid displayed the highest amount of carbon oxidized compared to acetic acid, methanol, ethanol, sucrose, and salicylic acid ; oxidizing twice as much carbon as acetic acid at the 40 minute interval [40] Formic acid is a small molecule that can be adsorbed on to the surface of TiO 2 impeding recombination by interaction with the valence band hole to produce carbon dioxide and water through the oxidation reaction. It will absorb only high energy light at wavelengths less than 260 nm [34] which means that formic acid will not be in competition with TiO 2 for the absorbance of lower energy UVA rays. These advantages make formic acid a suitable hole scavenger in the photo reduction of mercury as seen by Wang et al where the addition greatly increased the amount of mercury photoreduced [34] However, it is important to note that the amount of formic acid necessary is directly proportional to numb er of electrons necessary for reduction. In a study investigating the

PAGE 23

23 enhanced photocatalytic reduction of nitrate over Bi doped TiO 2 it was noted that 5 moles of electrons are required to completely reduce 1 mole of nitrate to nitrogen gas [41] Therefore, 5 moles of formic acid would be necessary. In the case of mercury reduction, two electrons are required for the formation of elemental mercury requiring two moles of formic acid per mole of mercury as seen in Equation 2 20 : ( 2 20 ) This is consistent with the results obtained by Wang et a l ; the addition of formic acid at concentrations more than double the concentration of mercury showed no improved photoreduction [34] As covered earlier, the presence of dissolved oxygen may hinder the reduction of heavy metals. Oxygen facilitates the oxidation reactions occurring at the valence band hole by trapping the photoexcited electron at the conduction band [42] competing with mercury for the reduction reaction. Additionally, the reaction of oxygen with the cb electron generates a superoxide radical which has the potential to re oxidize the reduced species of mercury Therefore, an inert ga s purge such as nitrogen, is desirable to reduce the dissolved oxygen in solution Visible Light Photocatalysis Table 2 1 lists a few possible photocatalysts and the ir band gap energies. A suitable catalyst must be photoreactive at an achievable energy level, non toxic, chemically stable and non photocorrosive [43] Titanium dioxide (TiO 2 ) is the most widely used photocatalytic material; h ighly suitable for environmental applications [44] Titanium dioxide exits in two crystalline structures: rutile (3.0 eV) and anatase (3.2 eV) [45] Degussa P25 is a commercially available TiO 2 comprised of a mixture of 70% anatase and 30% rutile with an average particle size of 30 nm and BET surface

PAGE 24

24 area of 40 70 m 2 /g [46] Although the band gap of TiO 2 requires photon energy at or below UVA wavelengths (< 290 nm), it has been argued that only a few photons at the required energy level are sufficient to promote photocatalysis [44] Therefore, the small amount of UVA photons emitted by indoor lighting could perhaps be capable of generatin g electron hole pairs on the surface of TiO 2 and treating trace amounts of contaminant. A series of studies investigated the photocatalytic degradation (PCD) rate of various organic compounds with respect to light intensity and were able to achieve degrada tion by solar radiation; however it was observed that PCD rate does increase with increasing light intensity [29] [47] [48] Various techniques have been conducted to improve the photocatalytic potential of TiO 2 under visible light such as doping the catalyst with metal and non metal ions [49] [50] [51] [52] Most of the literature has focused on characterizing the modified photocatalyst and measuring light absorption/reflection [45] [52] [53] but not many studies have be en published comparing side by side results of photocatalytic reduction of heavy metals under visible and ultra violet light. If indeed only a few photons are required to excite the valence band electron, it could be argued that a visible light source woul d be sufficient to reduce trace levels of contaminants. Additionally, the slightly increased absorbance of light for the doped catalysts may be negligible if successful results are obtained by using plain TiO 2 under visible light. Asahi et al investigated the effects of doping titanium dioxide with nitrogen. The absorbance of light in the visible range for doped catalysts is only enhanced marginally compared to plain TiO 2 [53] Other studies slightly improved the reflectance/absorbance of light in the

PAGE 25

25 visible light range of TiO 2 by doping techniques, demonstrating insignificant difference between doped and undoped catalysts [51] [54] [55] About 30% decomposition of rhodamine B and 2,4 dichlorophenol was achieved using pure TiO 2 under visible light [56] Nitrogen doping of the catalyst improved results to decompositi on rates between 45% and 98% depending on the nitrogen dose, however the study was conducted using high ppm concentrations of contaminants. It would be interesting to investigate the photocatalytic decomposition rate with pure TiO 2 at lower contaminant con centrations. Additionally, the contact times were as high as 5 hours, which would be unfeasible in application. If it can be demonstrated that the photoreduction of heavy metals is possible using a visible light source, the additional costs of doping techniques and high energy UV light sources would be unnecessary for the commercial application of photocatalytic treatment of trace met als.

PAGE 26

26 Figure 2 1 Speciation of 100 uL of Hg solution prepared from Hg(NO 3 ) 2 Adapted from: H. Byrne, Adsorption, Photocatalysis, and Photochemistry of Trace Level Aqueous Mercury, Gainesville: University of Florida, 2009. Table 2 1 Band gap energi es for different photocatalysts. Adapted from: D. Bhatkhand e, V. Pangarkar and A. Beenackers, "Photocatalytic degradation for environmental applications: a review," Journal of Chem. Tech. and Biotech. vol. 77, no. 1, pp. 102 116, 2001. Photocatalyst Bandgap Energy (eV) Photocatalyst Bandgap Energy (eV) Si 1.1 ZnO 3.2 TiO 2 (rutile) 3 TiO 2 (anatase) 3.2 WO 3 2.7 CdS 2.4 ZnS 3.7 SrTiO 2 3.4 SnO 2 3.5 WSe 2 1.2 Fe 2 O 3 2.2 Fe 2 O 3 3.1

PAGE 27

27 CHAPTER 3 EXPERIMENTAL STC Synthesi s The silica titania composite was synthesized using tetraethyl orthosilicate (TEOS) as the precursor in the sol gel method. 50 mL of DI water (18.2 M cm), 100 mL of ethanol (Fisher, 200 proof), 4 mL of 1 M nitric acid (made from concentrated Fisher) and 70 mL of TEOS were added in a HDPE container. 8 mL of 3% wt hydrofluoric acid (diluted from 49% Fishe r) was added to achieve a pore diameter of approximately 140 A. Lastly, a mass of 2.8 g of titanium dioxide (Degussa P25) was added for a TiO 2 loading of 4%. The mixture was magnetically stirred until gelation. It was then capped and allowed to age at room temperature for 4 hours and dried at 65 degrees Celsius for an additional 48 hours. The material was transferred to a Pyrex container and covered with pin sized holes to allow moisture to escape. It was heated at a temperature of 103 degrees Celsius for 1 8 hours followed by 180 degrees Celsius for 6 hours. The dried material was ground using a mortar and pestle and then sieved to achieve particles between 45 to 90 m. Table 3 1 summarizes the surface characteristics of each catalyst. Co mixing Studies Hg solutions were prepared by diluting 1000 mg/L of Hg(NO 3 ) 2 (Fisher ) with deionized water. Formic acid (Fisher) was added at a concentration of 1.5 ppm of carbon Expe riments were conducted in Pyrex flasks with a liquid volume of 150 mL and continuously stirred by magnetic stir bars. The STC was added at a loading of 1 g/L based on previous pseudo equilibrium batch studies [56] Catalysts fo r the co mixing experiments were added in equal amounts (0.5 mg/L of each) to achieve a total

PAGE 28

28 concentration of 1 mg/L. Illuminated experiments were conducted for 5, 15 and 30 minutes exposed to visible light at a distance of 1 inch from the reactors. The bulb used was a GE Ecolux Startcoat T 8 Linear Fluorescent Light with a peak wavelength close to 550 nm ( Figure 3 1 ) [57] Flasks were capped with rubber stoppers equipped with a purge glass tube (5 mm diameter) and a similar vent tube ( Figure 3 2 ) Experiments were conducted under no p urge (solely exposed to air) and under nitrogen purge (Airgas, ultra high purity) at a rate of 2 L per minute controlled by a flow meter. Control experiments were conducted in the dark following the same procedure described above. The flasks were covered i n aluminum foil to prevent exposure to any ambient light. Duplicate experiments were conducted for all sets and the ranges of values are expressed by error bars. STC and TiO 2 Comparison Studies Time dependent studies were conducted as described above. Pyrex flasks containing 150 mL of 100 ppb Hg solution were exposed to visible light under varying conditions The STC concentration was maintained at 1 g/L and the TiO 2 concentration was determined from the loading of the material in the silica titania com posite; approximately 76 m g/L Mixtures were exposed to one linear fluorescent visible light bulb at a distance of 1 inch from the flasks for time intervals of 5, 15 and 30 minutes. Flasks were covered with a rubber stopper equipped with 5 mm glass tubes f or gas purge and ventilation. The sto pper and glass tubes were present in dark experiments as well to maintain equal mixing gradients. Lastly, the same experiments were conducted under 254 nm ultra violet light in order to compare mercury removal in both scenarios. For these studies, a liquid volume of 100 mL was placed in a cylindrical Pyrex glass reactor and continuously stirred by

PAGE 29

29 magnetic mixing ( Figure 3 3 ). The r eactors were covered with a Teflon lid equipped with two inch glass tubes; one for delivery of a gas purge and the other for ventilation. A PL S Twin Tube Short Compact Fluorescent Lamp was inserted through the Teflon lid into the reactor. Illumination, addition of the catalyst, and the gas purge (when applicable) were initialized at the same time for all experiments. For dose dependent studies, experiments were conducted as described above for a period of 5 minutes under the same scenarios describe d i n Figure 3 4 The matr ix of catalyst doses is presented in Table 3 2 The 0 mg dose represents the experiments exposed to visible light only, without the addition of a catalyst. For all STC and TiO 2 comparison studies, control experiments were conducted in the dark. Additionally, comparison experiments were conducted under 254 nm ultra violet light. Duplicate experiments were conducted for all studies and results are reported as normalized concentration C/C 0 Error bars represent the range of values obtained. In all studies, the pH was measured but not control led. The values rang ed from 3.80 to 4.5 There was no significant change in removal between different pH values within the range. Sample Collection and Analysis membrane to separate the catalyst from the solution. Aliquots of 20 mL were taken from the filtrate and placed in 40 mL glass vials with Teflon lined caps. Nitric acid was added immediately in order to preserve the samples which were digested according to the EPA standard method 254.3 and analy zed within two weeks of collection using a Teledyne

PAGE 30

30 Instruments Hydra Atomic Adsorption Mercury Analyzer manufactured by Leeman Labs. The detection limit was determined to be 1 ppb of mercury.

PAGE 31

31 Table 3 1 Catalyst Surfa ce Characterization BET Pore Size Pore Vol. Mesopore Vol. m 2 /g cc/g cc/g WO 3 10.45 136.00 0.0336 0.03172 STC 174.67 182.60 0.7974 0.7826 TiO 2 45.22 194.90 0.2203 0.2046 Figure 3 1 Spectral power distribution provided by the manufacturer [57] Figure 3 2 Visible light setup (profile view). Stir Plate Light Source Purge Tube Vent Tube 150 mL Liquid Vol.

PAGE 32

32 Figure 3 3 Reactor setup employed in UV 254 studies Figure 3 4 Experimental conditions for STC and TiO 2 comparison studies. Table 3 2 Equivalent catalyst dose. Catalyst Dose (mg) TiO 2 0 2 4 6 8 10 STC 0 26 53 79 105 132

PAGE 33

33 CHAPTER 4 PHOTOCHEMICAL REMOVAL OF AQUEOUS MERCURY BY CO MIXING CATALYSTS Control Studies Control studies were conducted in the dark under nitrogen purge with the presence of formic acid to measure adsorption and ensure that photochemical reactions occurred when visible light was present. The removal of mercury in this scenario represents the a dsorption on to the surface of the silica titania composite. It is expected that a competition for adsorption exists between mercury and formic acid and that the removal of mercury via adsorption would increase if there was no formic acid present. However, the intention was to maintain the optimal conditions for removal under illumination to ensure the photoreduction in the presence of visible light. The highest removal overall in the absence of light was for the co mixed (STC+WO 3 ) catalyst indicating enha nced adsorption when more than one catalyst was used simultaneously ( Figure 4 1 ). The removal of mercury ranged from 21% to 54% for STC, from 44% to 59% for WO 3 and from 45% to 69% for co mixed STC+WO 3 for time intervals ranging from 5 minutes to 30 minutes. Effect of Visible Light and Nitrogen Purge The addition of visible light seems to drastically improve removal compared to the dark experiments ( Figure 4 2 ). Removal proceeds as a first order reaction dependent on the initial mercury concentration. This agrees with the hypothesis that higher concentrations of mercury would require more electrons produced from electron hole pairs, therefore requ iring a higher dose of UV light. As discussed in Chapter 2 (Equation 2 20), two electrons are required to reduce Hg 2+ to Hg 0 therefore higher concentrations of Hg 2+ w ould require more electrons for oxidation

PAGE 34

34 In this case, the photon energy in visible light is sufficient to decrease the concentration of mercury from 100 ppb at time zero to approximately 1 ppb at 30 minutes. The presence of a nitrogen purge slightly in creased removal at the 5 minute interval but there was no significant difference at 15 or 30 minutes ( Figure 4 2 ). Similar results were observed for W O 3 at 15 and 30 minutes ; there was a significant difference in the removal of mercury when comparing dark runs with illuminated experiments and no substantial difference in results between experiments with and without a nitrogen purge ( Figure 4 3 ). The presence of a nitrogen purge may not be as significant in these scenarios since there is an organic compound in solution. The formic acid present act s as a hole scavenger and radical sink preventing re oxidation of Hg 0 to Hg 2+ by either hydroxyl or superoxide radicals. However, this may not occur in the absence of formic acid since there would be no compound to absorb superoxide radicals formed by dissolved oxygen The effect of nitrogen purge without an organic compound present will be discussed in subsequent sections. The nitrogen purge, however, did affect the removal in t he co mixing experiments ( Figure 4 4 ). At the 5 minute interval, removal was about 68% with no purge compared to 94% under nitrogen purge ; showing almo st 30% more removal It was indicated in the literature that tungsten oxide has smaller band gap energy than titanium dioxide; 2.7 eV compared to 3.2 eV [43] This would mean that more electron hole pairs would be produced at t he same energy intensity as compared to TiO 2 therefore, increasing the potential for radical formation and requiring both a nitrogen purge and organic present for effective removal. At first glance, the removal of Hg using a co mixed catalyst seemed to i mprove when compared to STC or WO 3 alone However, comparing co

PAGE 35

35 mixed data to STC in the absence of N 2 purge, under visible light shows no difference in removal ( Figure 4 5 ). Additionally, there was a slight improvement in adsorption of mercury for co mixed catalysts compared to single catalysts as discussed in the control studies. Analyzing the 30 minute interval for STC and STC + WO 3 in the dark demons trates an improvement in removal of approximately 14% ( Figure 4 6 ). Co mixed STC+WO 3 demonstrated high removal at all time intervals. Studies were performed using half dose of STC to verify that removal improvement was actually due to the addition of WO 3 and not the decreased dose of STC. The results show that there was no significant difference in mercury removal using 1 g/L of STC compared to 0.5 g /L. In summary, the addition of WO 3 to STC improves removal of mercury through enhanced adsorption and there is no effect on photoreduction. Figure 4 7 summarizes the results of all experiments under nitrogen purge in the presence of formic acid. It is important to note that the silica titania composite was successful at removing mercury under illumination with more than 80% removal at only 5 minutes an d performed better than tungsten oxide. This could be explained by the same theory of why the nitrogen purge is effective for co mixed catalysts compared to single catalysts. WO 3 displays lower band gap energy than TiO 2 creating more electron hole pairs in visible light. The increased number of these pairs will escalate the chance of radical formation and lead to re oxidation of mercury. Increasing the initial concentration of mercury could possibly invert removal results leading to higher removal with WO 3 compared to STC.

PAGE 36

36 The next section s are devoted to understanding reduction of mercury using STC in more detail and optimizing the process by studying the effects of contaminant concentration, gas purge, presence of an organic and comparison to titanium diox ide.

PAGE 37

37 Figure 4 1 Time dependent control studies in the absence of light. Experimental conditions: 100 ppb of Hg, 1.5 ppmC of formic acid, 2 L/min nitrogen purge, 1 g/L total catalyst concentration. Figure 4 2 Photoreduction of mercury using STC under visible light in the presence of formic acid. Experimental conditions: 100 ppb Hg, 1.5 ppmC formic acid (FA), 2 L/min nitrogen purge, 1 g/L STC. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 5 15 30 C/C O Time (min) STC, purge, w/ FA, dark WO3, purge, w/ FA, dark STC+WO3, purge, w/ FA, dark 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 10 20 30 40 C/C O Time (min) STC, no purge, w/ FA STC, purge, w/ FA STC, purge, w/ FA, dark

PAGE 38

38 Figure 4 3 Photoreduction of mercury using WO 3 under visible light in the presence o f formic acid. Experimental conditions: 100 ppb Hg, 1.5 ppmC formic acid (FA), 2 L/min nitrogen purge, 1 g/L WO 3 Figure 4 4 Photoreduction of Hg: effects of co mixing silica titania composite and tungsten oxide in the presence of formic acid. Experimental conditions: 100 ppb Hg, 1.5 ppmC formic acid (FA), 0.5 g/L of each catalyst. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 10 20 30 40 C/Co Time (min) WO3, no purge, w/ FA WO3, purge, w/ FA WO3, purge, w/ FA, dark 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 10 20 30 40 C/Co Time (min) STC+WO3, no purge, w/ FA STC+WO3, purge, w/ FA STC+WO3, purge, w/ FA, dark

PAGE 39

39 Figure 4 5 Comparison of catalysts in the absence of a nitrogen purge. Experimental conditions: 100 ppb Hg, 1.5 ppmC formic acid (FA), 1 g/L single catalyst, 0. 5 g/L combined catalyst. Figure 4 6 Comparison of silica titania composite alone and co mixed silica titania and tungsten oxide under nitrogen purge in the presence of formic acid. Experimental conditions: 100 ppb Hg, 1.5 ppmC formic acid (FA), 2 L/m in nitrogen purge. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 10 20 30 40 C/Co Time (min) STC, no purge, w/ FA WO3, no purge, w/ FA STC+WO3, no purge, w/ FA 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 10 20 30 40 C/Co Time (min) STC, purge, w/ FA STC, purge, w/ FA, 1/2 dose STC+WO3, purge, w/ FA STC, purge, w/ FA, dark STC+WO3, purge, w/ FA, dark

PAGE 40

40 Figure 4 7 Photoreduction of mercury: comparison of all catalysts under nitrogen purge in the presence of formic acid. Experimental conditions: 100 ppb Hg, 1.5 ppmC formic acid (FA), 2 L/min nitrogen purge. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 10 20 30 40 C/C O Time (min) STC, purge, w/ FA WO3, purge, w/ FA STC+WO3, purge, w/ FA STC, purge, w/ FA, dark WO3, purge, w/ FA, dark STC+WO3, purge, w/ FA, dark

PAGE 41

41 CHAPTER 5 STC AND TiO 2 TIME DEPENDENT STUDIES Control Studies Control studies were performed in the dark as described in the previous section to verify that reduction of mercury was due to photochemical reactions. The results are summarized in Figure 5 1 The removal of mercury by adsorption was greater using TiO 2 compared to STC for all time intervals. Effects of Formic Acid The effects of formic acid are described by Figure 5 2 and Figure 5 4 Removal increases by approximately 50% in the presence of formic acid compared to no formic acid under nitrogen purge within 5 minutes of illumination. However, t he difference in removal is insignificant at the 15 and 30 minute intervals under the same conditions It is hypothesized that si nce purge starts at time 0 (simultaneously with illumination), 5 minutes is not sufficient to completely eliminate the dissolved oxygen. The oxygen still present is competing for the conduction band electron and forming superoxide radicals that may re oxid ize the elemental species of mercury [28] [30] On the other hand, there seems to be sufficient time at the 15 and 30 minute intervals to create a completely inert nitrogen enviro nment making the a ddition of formic acid insignificant. The purge was initiated at the same time as illumination to mimic how treatment would occur is application. It would not be feasible to purge pre treatment due to higher operational costs and longer retention times. T he importance of f ormic acid is emphasized in Figure 5 4 when no purge is used; r emoval increases significantly for a ll three time intervals. Formic ac id reacts with the valence band hole preventing hydroxyl radical formation and acts as a sink for any

PAGE 42

42 radicals present. No photocatalytic removal occurs in the absence of formic acid as seen in the comparison of dark experiments with illuminated experiment s. For every divalent mercury s pecies reduced by one electron, a radical is formed by the electron hole pair. Since no other compound is present to react with the radical, mercury is re oxidized to its divalent form, Hg 2+ ( Figure 5 3 ). Dark reactions with formic acid in the absence of a nitrogen purge were not conducted since the presence of formic acid did not significantly affect adsorption. Effects of Nitrogen Purg e Experiments with and without formic acid were conducted to investigate the effects of a nitrogen purge under both scenarios. In the presence of formic acid, the addition of a nitrogen purge does not significantly improve results ( Figure 5 5 ). The formic acid prevents both hydroxyl and superoxide radicals from re oxidizing Hg 0 to its divalent form. The nitrogen purge becomes significant at longer time intervals when there is no formic acid present. As seen in Figure 5 6 removal is the same for all scenarios at 5 minutes when there is still dissolved oxygen presen t. As the purge decreases the concentration of dissolved oxygen, removal of mercury improves drastically compared to no nitrogen purge (15 and 30 minutes). Reducing the concentration of dissolved oxygen prevents the formation of superoxide radicals and sin ce there is no formic acid present to prevent re oxidation of Hg 0 the effect a nitrogen purge becomes more significant. Interestingly, the silica titania composite performed quite well in the reduction of mercury unde r the appropriate conditions. The stud ies d iscussed above suggest two possible mechanisms: either the silica may be responsible for lowering the band gap energy of titanium dioxide or the few photons emitted by visible light in the UVA range

PAGE 43

43 are enough to cause the photocatalytic reduction of trace mercury levels Further studies comparing STC to TiO 2 were conducted to verify a conclusion TiO 2 Comparison Studies The same experiments were conducted using TiO 2 photocatalytic activity to a known, widely used catalyst. There was n o significant difference in removal using STC compared to TiO 2 in the presence of formic acid and a nitrogen purge ( Figure 5 7 ), suggesting that the sili ca does not alter the band gap energy of ti tanium dioxide. There is significance difference at the 95% confidence level for dark experiments at 30 minutes for STC and TiO 2 but this cannot be confirmed for 5 and 15 minute experiments. However, the difference in removal between the two catalysts varied greatly at 5 minutes when no formic acid was present, but this due to better adsorption of mercury onto TiO 2 nanoparticles ( Figure 5 8 ). The data points for TiO 2 illuminated studies compared to dark studies are not statistically different at the 95% confidence level. The presence of formic acid seems to greatly affect the removal mechanism when using titanium dioxide. If formic acid is present, photocatalysis occurs and the illuminated experiments are statistically different from the dark experiments. In the absence of formic acid, the only mechanism of mercury removal is adsorption, evident by similar dark and illuminated studies data. Comparison of Visible Light and UV 254 Studies were conducted comparing removal using STC exposed to visible light and UV 254 nm Removal seemed to slightly increase under UV 254 at the 5 minute inter val but the difference was not statistically significant at the 95% confidence level This suggests that a linear fluorescent bulb may be capable of generating the same

PAGE 44

44 electron hole pairs necessary to reduce mer cury at a trace concentration as a compact ultra violet bulb. Positive results from the studies presented require further investigation to understand the photcatalytic process under visible light. More detailed experiments were conducted and discussed in the subsequent sessions focusing on the effect of dissolved oxy gen, the presence of hole scavengers/radical sinks, the effect of catalyst dose contaminant initial concentration. Low doses can only be achieved by employing catalyst slurries; however, additions of dry catalysts were used for the dose dependent studies. Since the previous experiments suggest that photocatalysis may occur under visible light, it was concluded that forming a water/catalyst slurry prior to running experiments may result in the formation of radicals in the slurry.

PAGE 45

45 Figure 5 1 STC a nd TiO 2 time dependent control studies. Experimental conditions: 100 ppb of Hg, 1.5 ppmC of formic acid, 2 L/min nitrogen purge, 1 g/L STC, 76 m g/L TiO 2 Figure 5 2 Photoreduction of mercury using STC under nitrogen purge. Experimental conditions: 1 00 ppb of Hg, 1.5 ppmC of formic acid (FA), 2 L/min nitrogen purge, 1 g/L STC. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 5 15 30 C/C O Time (min) STC, purge, w/ FA, dark TiO2, purge, w/ FA, dark 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 10 20 30 40 C/C O Time (min) STC, purge, w/ FA STC, purge, no FA STC, purge, w/ FA, dark STC, purge, no FA, dark

PAGE 46

46 Figure 5 3 Mercury oxidation in the absence of formic acid. Figure 5 4 Photoreduction of mercury using STC without nitrogen purge. Experimental conditions: 100 ppb of Hg, 1.5 ppmC of formic acid (FA), 1 g/L STC, 76 m g/L TiO 2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 10 20 30 40 C/Co Time (min) STC, no purge, w/ FA STC, no purge, no FA STC, no purge, no FA, dark

PAGE 47

47 Figure 5 5 Photoreduction of mercury using STC in the presence of formic acid. Experimental conditions: 100 ppb of Hg, 1.5 ppmC of formic acid (FA), 2 L/min nitrogen purge, 1 g/L STC. Figure 5 6 Photoreduction of mercury using STC without formic acid. Experimental conditions: 100 ppb Hg, 2 L/min nitrogen purge, 1 g/L STC. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 10 20 30 40 C/C O Time (min) STC, no purge, w/ FA STC, purge, w/ FA STC, purge, w/ FA, dark 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 10 20 30 40 C/C O Time (min) STC, purge, no FA STC, no purge, no FA STC, no purge, no FA, dark

PAGE 48

48 Figure 5 7 Comparison of STC and TiO2 with nitrogen purge and formic acid. Experimental conditions: 100 ppb Hg, 1.5 ppmC of formic acid (FA), 2 L/min nitrogen purge, 1 g/L STC, 76 m g/L TiO 2. Figure 5 8 Photoreduction of mercury using TiO 2 under nitrogen purge. 100 ppb Hg, 2 L/min nitrogen purge, 1 g/L STC, 76 m g/L TiO 2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 10 20 30 40 C/Co Time (min) STC, purge, w/ FA TiO2, purge, w/ FA STC, purge, w/ FA, dark TiO2, purge, w/ FA, dark 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 20 40 C/C O Time (min) TiO2, purge, no FA STC, purge, no FA TiO2, purge, no FA, dark

PAGE 49

49 Figure 5 9 Photoreduction of mercury using STC under visible and UV 254 light. 100 ppb Hg, 1.5 ppmC formic acid (FA), 2 L/min nitrogen purge, 1 g/L STC. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 10 20 30 40 C/Co Time (min) STC, purge, w/ FA, UV 254 STC, purge, w/ FA STC, purge, w/ FA, dark

PAGE 50

50 CHAPTER 6 OPTIMIZATION DOSE DEPENDENT STUDIES Control Studies and Dose Determination Experiments were conducted in the dark with and without the addition of formic acid. It is evident from previous studies that the organic compound interferes with mercury removal (i n dark runs) by competing for adsorption. The previous doses used for STC and TiO2 were 150 mg and 11.4mg respectively. This corresponded to a concentration of 1 g/L of STC and an equivalent 76 mg/L of TiO2. It was determined that the lowest, accurate mass of dry TiO2 measurable with the available balance was 2 mg. Therefore, the TiO2 doses chosen were 2, 4, 6, 8, and 10 mg. The corresponding STC doses were 26, 53, 79, 105, and132 ( Table 6 1). The dose of 0 mg represents the experiments exposed to light only, without the addition of a catalyst Effect of Formic Acid The effects of formic acid on mercury reduction were investigated for both STC and TiO 2 with and without the presence of a nitrogen purge. The addition of formic acid did not drastically effect the removal of mercury using STC without a nitrogen purge ( Figure 6 1 ). There seems to be a slight improvement in removal for STC doses above 50 mg. However, removal improved drastically when formic acid was added to the solution in the presence of a nitrogen purge as seen in Figure 6 2 The nitrogen purge reduces the amount of dissolved oxygen in solution, limiting the formation of oxidizing radicals. As discussed earlier, dissolved oxygen comp etes with mercury for the conduction band electron. The oxygen molecule will react with the electron to form hydrogen peroxide which cleaves into hydroxyl radicals (Eq uations ( 2 6 through ( 2 12 ). Therefore, two mechanisms exits that hinder mercury reduction in the

PAGE 51

51 presence of oxygen: the competition fo r the conduction band electron and the formation of oxidizing agents which my re oxidize the reduced species of mercury. The addition of formic acid may reduce the amount of oxidizing agents in solution, but it will not prevent the competition for the cond uction band electron as described by Equations ( 2 6 through ( 2 10 in Chapter 2. When the amount of dissolved oxygen has been decreased by a nitrogen purge, the effects of formic acid are amplified. There is no longer a compound competing with mercury for the conduction band spot and the amount of oxidizing agents in solution is decreased. The formic acid interacts with the valence band hole to prevent the formation of radicals improving the reduction of mercury. Further research may be proposed to investigate whether increasing the concentration of formic acid may improve removal when there is dissolved oxygen present. The change in dose show ed little to no effect on the removal of mercury suggesting that lower concentrations must be tested. However, achieving an accurate lower concentration may only possible by using a pre mixed slurry which may result in the formation of additional radicals if this slurry is exposed to ambient light. The same results were not obtained using titanium dioxide. The addition of formic acid hindered removal drastically in the absence of a nitrogen purge ( Figure 6 3 ). As mentioned in the previous sections, formic acid competes for adsorption on to the catalyst, blocking the mercury from reaching the photo excited electrons from the conduction band. This is evid ent by the drastic difference in removal between the two dark reactions. The data were statistically different at the 95% confidence level for dark and illuminated studies with no purge in the presence of formic acid. For the comparison

PAGE 52

52 between illuminated and dark studies without a purge and without formic acid, all points were statistically different at the 95% confidence level except for 10 mg. Therefore, the photocatalytic reduction of mercury occurs using TiO 2 when there is no organic compound or nitro gen purge present. Figure 6 4 shows the effects of formic acid when a purge is present. It is interesting to note that the presence of formic acid did not hinder re moval in illuminated studies. The improvement of removal by the addition of formic acid under nitrogen purge was statistically significant for points 2, 4 and 10 mg but not for 6 and 8 mg (illuminated purge with no formic acid compared to illuminated purge with formic acid). Although formic acid competes with mercury for adsorption as seen in the dark studies, it does not impede the mercury compound from reaching the conduction band electron. In the absence of formic acid, removal was due to adsorption only even in illuminated studies. Effect of Nitrogen Purge When formic acid is present, the addition of a nitrogen purge greatly improves the reduction of mercury using STC as seen in Figure 6 5 most likely by reducing the oxidizing radicals as discussed earlier. Formic acid alone is not capable of combating radicals formed by both the valence band hole (hydroxyl radicals) and the conduction band electron (superoxide radicals) and removal never surpasses 52% in this scenario. When the nitrogen purge is added, formation of radicals occurs only by hydroxide ions at the valence band hole. However, when no formic acid is present, the ad dition of a nitrogen purge is insignificant ( Figure 6 6 ). Enough hydroxyl radicals are formed to re oxidize the elemental mercury species. This may su pport the theory that the main oxidizing agents

PAGE 53

53 generated by the electron hole pair are hydroxyl radicals formed by the interaction between water and/or hydroxide ions at the surface of the catalyst and the positive valen ce band hole as described by in Cha pter 2 by Equations ( 2 4 and ( 2 5 [26] Similar r esults were observed using TiO 2 The addition of a nitrogen purge made a drastic difference when formic acid was present ( Figure 6 7 ) but was insignificant in the absence of formic acid ( Figure 6 8 ) It was concluded that the generation of radicals by the dissolved oxygen affects removal neg atively when there are no additional hydroxyl radicals being formed. In the absence of formic acid, radicals are being produced by the positive valence band hole at the same rate that mercury is reduced by an electron from the conduction band. Therefore, r emoval of mercury reaches equilibrium whether or not additional radicals are formed by the dissolved oxygen. Figure 6 9 and Figure 6 10 summarize the results of all experiments using STC and TiO 2 respectively. By comparing results altogether it is evident that the best mercury removal scenario using STC includes the presence of formic acid exposed to a nitrogen purge. For TiO 2 the difference is not as drastic, but the best removal environment also includes the combination of formic acid with a nitrogen purge. For TiO 2 studies with no formic acid an d no nitrogen purge performed just as well as having both components present, but it was concluded that remo val was due to adsorption only.

PAGE 54

54 Table 6 1 Catalyst doses for dose dependent studies. Catalyst Dose (mg) TiO2 0 2 4 6 8 10 STC 0 26 53 79 105 132 Figure 6 1 Effects of formic acid on photoreduction of mercury using STC without a purge. Experimental conditions: 100 ppb of Hg, 1.5 ppmC formic acid (FA), 5 min. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 50 100 150 C/Co STC (mg) STC, no purge, no FA STC, no purge, no FA, dark STC, no purge, w/ FA, dark STC, no purge, w/FA

PAGE 55

55 Figure 6 2 Effects of formic acid on photoreduction of mercury using STC under nitrogen purge. Experimental conditions: 100 ppb Hg, 1.5 ppmC formic acid (FA), 2L/min purge, 5 min. Figure 6 3 Effects of formic acid on photoreduction of mercury using TiO 2 without nitrogen purge. Experimental conditions: 100 ppb Hg 1.5 ppmC formic acid (FA), 5 min. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 50 100 150 C/Co STC (mg) STC, purge, no FA STC, purge, w/ FA STC, purge, w/ FA, dark 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 2 4 6 8 10 12 C/Co Dose (mg) TiO2, no purge, no FA TiO2, no purge, no FA, dark TiO2, no purge, w/ FA TiO2, no purge, w/ FA, dark

PAGE 56

56 Figure 6 4 Effects of formic acid on photoreduction of mercury using TiO 2 under nitrogen purge. Experimental conditions: 100 ppb Hg, 1.5 ppmC formic acid (FA), 2 L/min nitrogen purge, 5 min. Figure 6 5 Effects of nitrogen purge on photoreduction of mercury using STC with formic acid. Experimental conditions: 100 ppb Hg, 1.5 ppmC of formic acid (FA), 2 L/min purge, 5 min. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 2 4 6 8 10 12 C/Co Dose (mg) TiO2, purge, no FA TiO2, purge, w/ FA TiO2, purge, w/ FA, dark TiO2, purge, no FA, dark 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 50 100 150 C/Co Dose (mg) STC, no purge, w/ FA, dark STC, no purge, w/FA STC, purge, w/ FA STC, purge, w/ FA, dark

PAGE 57

57 Figure 6 6 Effects of nitrogen purge on photoreduction of mercury using STC wi th no formic acid. Experimental conditions: 100 ppb Hg, 2 L/min nitrogen purge, 5 min. Figure 6 7 Effects of nitrogen purge of photoreduction of mercury using TiO 2 with formic acid. Experimental conditions: 100 ppb of Hg, 1.5 ppmC formic acid (FA), 2 L/min nitrogen purge, 5 min. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 50 100 150 C/Co Dose (mg) STC, no purge, no FA STC, purge, no FA STC, no purge, no FA, dark 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 2 4 6 8 10 12 C/Co Dose (mg) TiO2, no purge, w/ FA TiO2, purge, w/ FA TiO2, no purge, w/ FA, dark TiO2, purge, w/ FA, dark

PAGE 58

58 Figure 6 8 Effects of nitrogen purge on photoreduction of mercury using TiO 2 without formic acid. Experimental conditions: 100 ppb of Hg, 2 L/min nitrogen purge, 5 min. Figure 6 9 Photoreduction of mercury using S TC: comparison on all scenarios. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 2 4 6 8 10 12 C/Co Dose (mg) TiO2, purge, no FA TiO2, no purge, no FA TiO2, no purge, no FA, dark TiO2, purge, no FA, dark -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 50 100 150 C/Co Dose (mg) STC, no purge, no FA STC, purge, no FA STC, no purge, no FA, dark STC, no purge, w/ FA, dark STC, no purge, w/FA STC, purge, w/ FA

PAGE 59

59 Figure 6 10 Photoreduction of mercury using TiO 2 : comparison of all scenarios. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 2 4 6 8 10 12 C/Co Dose (mg) TiO2, purge, no FA TiO2, no purge, no FA TiO2, no purge, no FA, dark TiO2, no purge, w/ FA TiO2, no purge, w/ FA, dark TiO2, purge, w/ FA TiO2, purge, no FA, dark

PAGE 60

60 CHAPTER 7 MERCURY CONCENTRATION DEPENDENT STUDIES Control Studies Studies were conducted in the dark with mercury initial concentrations of 100, 500 and 1000 parts per billion. The catalyst concentration was 1 g/L for STC and the equivalent 76 m g/L for TiO 2 Experiments were conducted in the presence of formic acid and n itrogen purge for 5 minutes. STC and TiO2 performed similarly with an initial concentration of 500 ppb of Hg; removing less than 20%. STC outperformed TiO 2 at the higher concentration of 1000 ppb ( Figure 7 1 ). Comparison of Visible Light vs. UV 254 Experiments were conducted for all three concentrations using STC ( Figure 7 2 ) and TiO 2 ( Figure 7 3 ) exposed to visible light and UV 254. As hypothesized, removal of mercury increased as initial concentration decreased when us ing visible light. At an initial concentration of 100 ppb, there was no statistical difference in removal using visible light compared to UV 254. Utilizing visible light source is just as effective in reducing mercury at low concentrations compared to ultr a violet light. A lower concentration of 50 ppb was investigated using STC ( Figure 7 4 ). There was no statistical difference in removal between visible light and UV 25 4 at this concentration reinforcing the conclusion the visible light is capable of photoreducing low concentrations of mercury. Two electrons are required to reduce Hg 2+ to Hg 0 The electrons are provided by the surface of th e photocatalyst after excitation by light. Since the excitation of electrons is directly proportional to the number of photons in the required energy level (370 380 nm), the reduction of mercury will also be directly proportional to the amount of photons.

PAGE 61

61 If higher concentrations of mercury exist (more Hg 2+ ), then more electrons will be required for reduction. Visible light emits only a limited number of UVA photons, therefore, sufficient electrons would not be produced to reduce higher mercury concentratio ns. Figure 7 5 portrays a representation of the photocatalytic reduction of higher concentrations of aqueous mercury (above 100 ppb).

PAGE 62

62 Figure 7 1 Mercury concentration dependent control studies. Experimental conditions: 1 g/L STC, 76 g/L TiO 2 1.5 ppm formic acid, 2 L/min nitrogen purge, 5 min. Figure 7 2 Comparison of UV 254 vs. visible light using STC. Experimental conditions: 1g/L STC 1.5 ppmC formic acid, 2 L/min nitrogen purge, 5 min. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 100 500 1000 C/C O Initial Hg Conc. (ppb) STC, Dark TiO2, Dark 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 100 500 1000 C/C O Initial Hg Conc. (ppb) STC, UV 254 STC, Visible Light STC, Dark

PAGE 63

63 Figure 7 3 Comparison of UV 254 vs. visible light using TiO 2 Experimental conditions: 76 g/L TiO 2 1.5 ppmC formic acid, 2 L/min nitrogen purge, 5 min. Figure 7 4 Comparison of visibl e light vs. UV 254 using STC at lower Hg concentrations. Experimental conditions: 1g/L STC, 1.5 ppm formic acid, 2 L/min nitrogen purge, 5 min. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 100 500 1000 C/C O Initial Hg Conc. (ppb) TiO2, UV 254 TiO2, Visble Light TiO2, Dark 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 50 100 C/C O Initial Hg Conc. (ppb) STC, Visible Light STC, UV

PAGE 64

64 Figure 7 5 Representation of photocatalytic reduction of high concentrations of aqueous mercury.

PAGE 65

65 CHAPTER 8 CONCLUSIONS The purpose of this study was to enhance visible light photocatalysis by modifying the removal process rather than modifying the catalyst, The co mixing of different catalysts does not affect the photocatalytic removal of mercury. The slightly increased removal evident in the co mixed STC+WO 3 studies was due to enhanced adsorption. However, the silica titan ia composite performed quite well as a single catalyst under visible light and the positive results lead to further investigation, including comparison with regular titanium dioxide. STC and TiO 2 performed similarly in illuminated experiments with visible light suggesting that silica does not lower the band gap energy of TiO 2 Studies were conducted in different scenarios in order to understand the effect of dissolved oxygen, the presence of an organic compound and catalyst dose. It was concluded that the addition of formic acid greatly improved removal by acting as a radical sink, preventing the re oxidation of Hg 0 to Hg 2+ Dissolved oxygen only affected removal negatively if formic acid was present. In the absence of an organic compound, re oxidation of m ercury occurred whether or not there was dissolved oxygen. Lastly, it was determined that visible light is capable of emitting enough photons to reduce trace levels of mercury. Results showed similar removal using visible light compared to UV 254 at low i nitial concentrations of mercury (100 ppb), but poor removal at high initial concentrations of 1000 ppb. It is suggested that visible light emits few photons at a range slightly below 400 nm which may be capable of producing some electron hole pairs. Futur e work may be conducted to create a model of light intensity needed based on contaminant concentrations.

PAGE 66

66 Visible light photocatalysis may be employed as a polishing step to further reduce contaminant concentrations. Cheaper, more widely used process such as precipitation and carbon adsorption can reduce the bulk of contaminants found in various waste streams but these methods may not be feasible to reduce levels below the parts per billion range. Visible light photocatalysis provides a safe method of reduc ing multiple contaminants through both oxidation and reduction reactions.

PAGE 67

67 LIST OF REFERENCES [1] M. Fujiki and S. Tajima, "The pollution of Minamata Bay by mercury," Water Sci. Tech., vol. 25, no. 11, pp. 133 140, 1992. [2] K. Eto, "Minamata Disease," Neuropathology, vol. 20, no. 1, pp. 14 19, 2000. [3] U.S. Environmental Protection Agency, "Basic Information about Mercury in Drinking Water," May 2012. [Online]. Available: http://water.epa.gov/ drink/contaminants/basicinformation/mercury.cfm [Accessed January 2014]. [4] United Nations Environmental Programme, "Minamata Convention on Mercury," October 2013. [Online]. Available: http://www.mercuryconvention.org/Portals/ 11/documents/publications/MinamataConventiontextEn.pdf [Accessed January 2014]. [5] M. Amyot, G. Mierle, D. Lean and D. McQueen, "Sunlight induced formation of dissolved gaseous mercury in lake waters," Environ. Sci. Tech., vol. 28, pp. 2366 2371, 1994. [6] UNEP, "Global mercury assesment report," United Nations Environmental Programme, New York, 2002. [7] E. G. Pacyna, J. M. Pacyna, K. Sundseth, J. Munthe, K. Kindbom, S. Wilson, F. Steenhuisen and P. Maxson, "Global emission of mercury to the atmosphere from antropogenic sources in 2005 and projections to 2020," Atmospheric Environment, vol. 44, no. 20, pp. 2487 2499, 2010. [8] S. Serre and G. Silcox, "Adsorption of elemental mercury on the residual carbon in coal fly ash," Ind. Chem. Eng. Res., vol. 39, no. 6, pp. 1723 1730, 2000. [9] M. Galperin, M. Sofiev and E. Mantseva, "A model of the chemical transforma tion of mercury and its long range atmospheric transport," in Global and regional mercury cycles: sources, fluxes and mass balances, Dordretch, The Netherlands, Kluwer Academic Publishers, 1996. [10] U.S. Environmental Protection Agency, "Clean Air Mercury Rule," March 2011. [Online]. Available: http://www.epa.gov/camr/. [Accessed January 2014]. [11] U.S. Environmental Protection Agency, "Mercury and Air Toxics Standards Basic Information," April 2012. [Online]. Available: http://www.epa.gov/mats/ basic.html [Accessed January 2014]. [12] U.S. Environmental Protection Agency, "Impaired Water and Total Maximum Daily Loads," Dece mber 2013. [Online]. Available: http://water.epa.gov/ lawsregs/lawsguidance /cwa/tmdl/ [Accessed January 2014].

PAGE 68

68 [13] P. Tchounwou, W. Ayensu, N. Ninashvili and D. Sutton, "Environmental Exposure to Mercury and its Toxicopathologic Implications for Purblic Health," Env. Toxicology, vol. 18, no. 3, 2003. [14] T. Clarkson, "Metal toxicity in the central nervous system," Env. Health Perspectives, vol. 75, pp. 59 64, 1987. [15] S. Witt, "OSHA safety hazard information bulletin on dimethylmercury," [Online]. Available: http://www.osha.gov/dts/hib/hib_data/hib19980309.html [16] World Health Organization, "Methylmercury. International Program on Chemical Safety," Geneva 1990. [17] Harada, "Congenital Minamata disease: intrauterine methylmercury poisoning," Teratology, vol. 18, no. 2, pp. 285 288, 1978. [18] Y. Harada, "Congenital (or fetal) Minamata disease.," Kumamato University, Kumamato, Japan, 1968. [19] C. J. Lin and S. Pehkonen, "The chemistry of atmospheric mercury: a review," Atmospheric Environment, vol. 33, pp. 2067 2079, 1999. [20] O. Lindqvist and H. Rodhe, "Atmospheric Mercury a reveiw," Tellus, vol. 37, no. B, pp. 136 159, 1985. [21] F. Morel, A. Kraepiel and M. Amyot, "The Chemic al Cycle and Bioaccumulation of Mercury," Annual Review of Ecology and Systematics, vol. 29, pp. 543 566, 1998. [22] H. Hahne and W. Kronntje, "Significance of pH and Chloride Concentration on Behavior of Heavy Metal Pollutants: Mercury (II), Cadmium (II ), Zinc (II), and Lead (II)," Journal of Environmental Quality, vol. 2, no. 4, pp. 444 450, 1973. [23] H. Byrne, Adsorption, Photocatalysis, and Photochemistry of Trace Level Aqueous Mercury, Gainesville: University of Florida, 2009. [24] K. Kabra, R. Chaudhary and R. Sawhney, "Treatment of hazardous inorganic and organic compounds through aqueous phase photocatalysis," Ind. Eng. Chem. Res., vol. 43, pp. 7683 7696, 2004. [25] O. Legrini E. Oliveros and A. M. Braun, "Photochemical Processes for Water Treatment," Chem. Rev., vol. 93, pp. 671 698, 1993. [26] R. W. J. Matthews, "Hydroxylation reactions induced by near ultraviolet photolysis of aqueous titanium dioxide suspensions," Chem. Soc. Faraday Trans., vol. 80, p. 457, 1984.

PAGE 69

69 [27] R. Bickley, T. Gonzalez Carreno, J. Lees, L. Palmisiano and R. Tilley, "A structural inverstigation of titanium dioxide photocatalyts," Journal of Solid State Chem., vol. 92, pp. 178 190, 1990. [28] M. Litter, "Heterogeneous photocatalysis, transition metal ions in ph otocatalytic systems," Applied Catalysis B: Environmental vol. 23, pp. 89 114, 1999. [29] K. Okamoto, Y. Yamamoto, H. Tanaka and A. Itaya, "Kinetics of heterogeneous photocatalytic decomposition of phenol over anatase TiO2 powder," Bull Chem Soc., vol. 58, pp. 2023 2028, 1985. [30] R. Cundall, R. Rudham and M. Salim, "Photocatalytic oxidation of propan 2 ol in the liquid phase by rutile," J. Chem. Soc., Faraday Trans. vol. 72, pp. 1642 1651, 1976. [31] P. Harvey, R. Rudham and S. Ward, "Photocatal ytic oxidation of liquid propan 2 ol by titanium dioxide," J. Chem. Soc., Faraday Trans. 1, vol. 79, pp. 1381 1390, 1983. [32] L. B. Khalil, W. E. Mourad and M. W. Rophael, "Photocatalytic reduction of environmental pollutant Cr(VI) over some semiconduct ors under UV/visible light illumination," Applied Catalysis B: Environmental vol. 17, no. 3, pp. 267 273, 1998. [33] K. Rajeshwar, C. R. Chenthamarakshan and Y. M. Wenjian Sun, "Cathodic photoprocesses on titania films and in aqueous suspensions," Jour nal of Electoanalytical Chemistry, Vols. 538 539, pp. 173 182, 2002. [34] X. Wang, S. Pehkonen and A. Ray, "Photocatalytic reduction of Hg(II) on commercial TiO2 catalysts," Electrochemica Acta, vol. 49, no. 9 10, pp. 1435 1444, 2004. [35] M. Ward, J. White and A. Bard, "Electrochemical investigation of energetics of particulate titanium dioxide photcatalysts. The methyl viologen acetate system," J. Am. Chem. Soc., vol. 105, pp. 27 31, 1983. [36] Y. Xu and M. Shoonen, "The absolute energy positions of conduction and valence bands of selected semiconducting minerals," American Mineralogist vol. 85, no. 3 4, pp. 543 556, 2000. [37] S. Sanuki, K. Shako, S. Nagaoka and H. Majima, "Photocatalytic re duction of Se ions using suspended anatase powders," Mat. Trans., JIM, vol. 41, no. 7, pp. 799 805, 2000. [38] N. Serpone, Y. K. Ah You, T. P. Tran and R. Harris, "AM 1 simulated sunlight photoreduction and elimination of Hg (II) and CH3Hg(II) chloride s alts from aqueous suspensions of titanium dioxide," Solar Energy, vol. 39, no. 6, pp. 491 498, 1987.

PAGE 70

70 [39] S. Botta, D. Rodriguez, A. Leyva and M. Litter, "Feature of the transformation og Hg(II) by heterogeneous photocatalysis over TiO2," Catalysis Today vol. 76, no. 2 4, pp. 247 258, 2002. [40] T. Tan, D. Beydoun and R. Amal, "Effects of organic hole scavengers on the photocatalytic reduction of selenium ions," Journal of Photochem. and Photobio. A: Chemistry, vol. 159, pp. 273 280, 2003. [41] S. Re ngaraj and X. Z. Li, "Enhanced photocatalytic reduction reaction over Bi3+ TiO2 nantoparticles in the presence of formic acid as a hole scavenger," Chemosphere, vol. 66, no. 5, pp. 930 938, 2007. [42] M. A. Fox and M. Dulay, "Heterogenous photocatalysis, Chem. Rev., vol. 93, pp. 341 357, 1993. [43] D. Bhatkhande, V. Pangarkar and A. Beenackers, "Photocatalytic degradation for environmental applications: a review," Journal of Chem. Tech. and Biotech. vol. 77, no. 1, pp. 102 116, 2001. [44] A. Fujish ima, T. Rao and D. Tryk, "Titanium dioxide photocatalysis," Journal of Photochem. and Photobio. C: Photochem. Reviews, vol. 1, no. 1, pp. 1 21, 2000. [45] J. B. Varley, A. Janotti and C. G. Van de Walle, "Mechanism of visible light photocatalysis in nitr ogen doped TiO2," Advanced Materials, vol. 23, no. 20, pp. 2343 2347, 2011. [46] D. Chen and A. Ray, "Removal of toxic metal ions from wastewater by semiconductor photocatalysis," Chem. Eng. Sci., vol. 56, no. 4, pp. 1561 1570, 2001. [47] J. Critteden, J. Liu, D. Hand and D. Perram, "Photocatalytic oxidation of chlorinated hydrocarbons in water," Water Res. vol. 31, pp. 429 438, 1997. [48] D. Ollis, E. Pellizzetti and N. Serpone, "Destruction of water contaminants," Env. Sci. Technol., vol. 25, pp. 1523 1528, 1991. [49] M. Seery, R. George, P. Floris and S. Pillai, "Silver doped titanium dioxide nanomaterials for enhanced visible light photocatalysis," Journal of Photochem. and Photobio. A: Chemistry, vol. 189, no. 2 3, pp. 258 263, 2007. [50] D. Chatterjee and S. Dasgupta, "Visible light induced photocatalytic degradation of organic pollutants," Journal of Photochem. and Photobio. C: Photchem. Reviews, vol. 6, pp. 186 205, 2005. [51] T. Umebayashi, T. Yamaki, H. Itoh and K. Asai, "Bad gap narro wing of titanium dioxide by sulfur doping," Applied Physics Letters, vol. 81, no. 3, p. 454, 2002.

PAGE 71

71 [52] M. Mrowetz, W. Balcerski, A. J. Colussi and M. Hoffmann, "Oxidative power of nitrogen doped TiO2 photocatalysts under visible light illumination," J. Phys. Chem. B vol. 108, no. 45, pp. 17269 17273, 2004. [53] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki and Y. Taga, "Visible light photocatalysis in nitrogen doped titanium oxides," Science, vol. 293, p. 269, 2001. [54] C. y. Wany, C. Bottcher, D. Bah nemann and J. Dohrmann, "A comparative study of nanometer sized Fe(III) doped TiO2 photocatalysis: synthesis, characterization and activity," J. Mater. Chem., vol. 13, pp. 2322 2329, 2003. [55] M. Liu, X. Qiu, M. Miyauchi and K. Hashimoto, "Cu(II) oxide amorphous nanoclusters grafted Ti3+ self doped TiO2: an efficient visible light photocatalyst," Chem. of Mat., vol. 23, pp. 5282 5286, 2011. [56] H. Byrne and D. Mazyck, "Removal of trace levels of aqueous mercury by adsorption and photocatalysis on sili ca titania composites," Journal of Hazardous Materials pp. 915 919, 2009. [57] General Electric, "GE Lighting Product Catalog," 2013. [Online]. Available: http://genet.gelighting.com/LightProducts/Dispatcher?REQUEST=COMMERCIA LSPECPAGE&PRODUCTCODE=10415&BreadCrumbValues=Lamps_Linear% 20Fluorescent%5EFull%20Wattag e&SearchFieldCode=null [58] C. Senior, A. Sarofim, T. Zeng, J. Heble and R. Mamani Paco, "Gas phase transformations of mercury in coal fired power plants," Fuel Processing Tech., vol. 63, pp. 197 213, 1999.

PAGE 72

72 BIOGRAPHICAL SKETCH Erica Wallace Gonzaga was born in Saint Paul, MN. Her family moved to her American School of Brasilia, Brazil and moved to Fort Lauderdale, FL. after graduation in 2004. She started studying architect ure at Broward College but after 2 years, decided to change her major to engineering. Erica was accepted to the University of Florida Biological Engineering program as a transfer student in the spring of 2009. She cember 2011 and spent the following semester working as a lab assistant for a PhD candidate in the department of Environmental Engineering. Erica was accepted as a Master of Engineering student in the Department of Environmental Engineering in the fall of 2012. One week before starting graduate school, Erica married her husband, Luiz Gonzaga, who is currently finishing is residency in Prosthodontics in the College of Dentistry at the University of Florida.