University of Calgary PRISM: University of Calgary's Digital Repository Graduate Studies The Vault: Electronic Theses and Dissertations 2014-04-28 Light Emitting Diode Based Photochemical Treatment of Contaminants in Aqueous Phase Yu, Linlong Yu, L. (2014). Light Emitting Diode Based Photochemical Treatment of Contaminants in Aqueous Phase (Unpublished doctoral thesis). University of Calgary, Calgary, AB. doi:10.11575/PRISM/26762 http://hdl.handle.net/11023/1447 doctoral thesis University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission. Downloaded from PRISM: https://prism.ucalgary.ca
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
University of Calgary
PRISM: University of Calgary's Digital Repository
Graduate Studies The Vault: Electronic Theses and Dissertations
2014-04-28
Light Emitting Diode Based Photochemical Treatment
of Contaminants in Aqueous Phase
Yu, Linlong
Yu, L. (2014). Light Emitting Diode Based Photochemical Treatment of Contaminants in Aqueous
Phase (Unpublished doctoral thesis). University of Calgary, Calgary, AB.
doi:10.11575/PRISM/26762
http://hdl.handle.net/11023/1447
doctoral thesis
University of Calgary graduate students retain copyright ownership and moral rights for their
thesis. You may use this material in any way that is permitted by the Copyright Act or through
licensing that has been assigned to the document. For uses that are not allowable under
copyright legislation or licensing, you are required to seek permission.
Downloaded from PRISM: https://prism.ucalgary.ca
UNIVERSITY OF CALGARY
Light Emitting Diode Based Photochemical Treatment of Contaminants in Aqueous Phase
chlorophenol (4-CP) and 2,4-dichlorophenol (2,4-DCP) was studied. Further, the impact
of photocatalyst loading and light intensity on the degradation rate was evaluated. The
degradation of 2,4-D under LED irradiation was compared to that with mercury discharge
lamp irradiation. The results show these compounds can be efficiently degraded using
LED based TiO2 photocatalysis. They are completely mineralized upon prolonged
irradiation. Our results indicate that LEDs are a better light source than the mercury
lamps.
iii
To design an efficient LED based photocatalytic reactor, a radiation field model was
developed in this research. The model was tested with experimental data and good
agreement between two was noted. The model can be used to optimize the photoreactor
and chose the optimal gap between adjacent LEDs, the irradiated distance and the light
output of LEDs for a homogenous radiation field.
Finally, an LED based photocatalytic reactor was designed and fabricated. The reactor
uses anodized TiO2 nanostructure as a photocatalyst. The performance of reactor was
evaluated and optimized by studying the degradation of 2,4-D. The effect of different
operational parameters on the reactor performance were investigated, including light
intensity, distance between the LED module and photocatalytic plate (DL-P), the flow rate
through the reactor, presence of external electron scavengers and photocatalyst
configuration. A power law relationship was observed between the light intensity (2.2
mW cm-2~17.3 mW cm-2) and the first order degradation rate constant for 2,4-D. A
suitable flow rate and DL-P was determined for the reactor. Enhanced performance of the
reactor was observed where electron scavengers were introduced.
iv
Acknowledgements
I would like to express my sincerest appreciation and gratitude to my supervisor Dr.
Gopal Achari and my co-supervisor Dr. Cooper H. Langford for their continuous
encouragement, intellectual advice, precious guidance and enthusiastic supports
throughout my doctoral program.
It is my fortune to have friendly colleagues, Dr. Jyoti Ghosh, Dr. Maryam Izadifard,
Jiansong Kong, Chien-Kai Kenneth Wang, Upasana Chamoli and Mitra mehrabani. I
greatly appreciate their helps. My gratitude is also extended to Mr. Daniel Larson for his
assistance with instruments and laboratory facilities during my research. Thanks to Mr.
Edward C. Cairns, Mr. Andrew Read, Mr. Mark Toonen and Robert Thomson for their
help on fabricating LED reactors.
I gratefully acknowledge the financial support provided by Samuel Hanen Foundation,
RES'EAU WaterNet Strategric Research Network, Natural Science and Engineering
Council of Canada and Department of Civil Engineering.
Finally, I would like to show my gratitude to my sister, my uncles and my aunties for
their supports in the past five years.
v
Dedication
This thesis is dedicated to my beloved parents.
vi
Table of Contents
Abstract ............................................................................................................................... ii Acknowledgements ............................................................................................................ iv Dedication ............................................................................................................................v Table of Contents ............................................................................................................... vi List of Tables .......................................................................................................................x List of Figures and Illustrations ......................................................................................... xi List of Symbols, Abbreviations and Nomenclature ...........................................................xv
CHAPTER TWO: LITERATURE REVIEW ......................................................................9 2.1 Principle of Photochemistry.......................................................................................9
2.1.1 Light and photon ................................................................................................9 2.1.2 The electronic excited states ............................................................................11 2.1.3 Quantum yield .................................................................................................11 2.1.4 Direct photolysis ..............................................................................................12 2.1.5 Photosensitized degradation ............................................................................13 2.1.6 Photocatalysis ..................................................................................................13 2.1.7 Advanced Oxidation Processes (AOPs) ..........................................................14
2.2 TiO2 Photocatalysis .................................................................................................14 2.2.1 TiO2 as a photocatalyst....................................................................................14 2.2.2 Mechanism of TiO2 photocatalysis .................................................................17 2.2.3 The kinetics of photocatalytic degradation ......................................................20 2.2.4 Factors affecting the photocatalytic degradation kinetics ...............................21
2.3.2 PCBs ................................................................................................................33 2.4 Photochemical treatment of pesticides and PCBs ....................................................36
2.4.1 Direct photolytic degradation of pesticides and PCBs ....................................36 2.4.2 Photosensitized degradation of pesticides and PCBs ......................................37 2.4.3 Photocatalytic degradation of pesticides and PCBs ........................................37
vii
2.5 Design of a photocatalytic reactor ...........................................................................38 2.5.1 State of Photocatalyst in the Reactor ...............................................................38
2.5.1.1 Slurry photocatalytic reactor vs immobilized photocatalytic reactor ....38 2.5.1.2 TiO2 immobilization through electrochemical anodization ..................40
3.2.2.1 PCB 138 solubilization with surfactants ................................................65 3.2.2.2 Photochemical reaction ..........................................................................65 3.2.2.3 Sampling, extraction and GC analysis ...................................................66
3.3 Results and Discussion ............................................................................................67 3.3.1 Selectivity of surfactants .................................................................................67 3.3.2 Dechlorination of PCBs in TEA and NaBH4 systems ....................................72
3.3.2.1 MB and TEA ..........................................................................................72 3.3.2.2 MB and NaBH4 .....................................................................................73 3.3.2.3 Photodegradation of Aroclor 1254 with NaBH4 and TEA ....................74
3.3.3 The dechlorination pathways of PCB 138 using CTAB and TWEEN 80 .......75 3.4 Conclusions ..............................................................................................................77
CHAPTER FOUR: LED-BASED PHOTOCATALYTIC TREATMENT OF PESTICIDES AND CHLOROPHENOLS ...............................................................79
4.1 Introduction ..............................................................................................................79 4.2 Methods and Materials .............................................................................................81
4.3 Results and Discussions ...........................................................................................86 4.3.1 Photocatalytic degradation of pesticides and chlorophenols ...........................86 4.3.2 Photocatalytic degradation of pesticides mixtures ..........................................89
viii
4.3.3 Effect of Photocatalyst Loading ......................................................................93 4.3.4 Effect of Light Intensity ..................................................................................95 4.3.5 Comparison between LED and Mercury Lamp Irradiation .............................97
CHAPTER FIVE: DESIGN A HOMOGENEOUS RADIATION FIELD IN A UV-LED BASED PHOTOCATALYTIC REACTOR ..................................................100
5.1 Introduction ............................................................................................................100 5.2 Advantage of homogeneous radiation field in a photocatalytic reactor ................101 5.3 Development of radiation field model ...................................................................103
5.3.1 UV-LED array and photocatalyst plate .........................................................103 5.3.2 Radiation field model without shielding glass plate ......................................104 5.3.3 Radiation field model with a shielding glass plate ........................................109
5.4 Calibration and validation of the radiation field model .........................................110 5.4.1 Light intensity measurement .........................................................................110 5.4.2 Model light intensities vs measured light intensities .....................................112
5.5 Design of a homogenous radiation filed ................................................................114 5.5.1 The effect of ID on the homogeneity of radiation field for a fixed gap ........114 5.5.2 Optimal combination of ID and gap ..............................................................116 5.5.3 Selection of the output of the UV-LED .........................................................117
6.2.1 Chemicals ......................................................................................................121 6.2.2 Design and fabrication of an LED based photocatalytic reactor ...................121
6.2.2.1 Preparation of anodized TiO2 photocatalytic plate. .............................121 6.2.2.2 UV-LEDs module ................................................................................122 6.2.2.3 Photocatalytic system ..........................................................................123
6.2.3 Radiation field and light intensity estimation ................................................124 6.2.4 Experimental set-up and sample analysis ......................................................126
6.3 Result and discussion .............................................................................................127 6.3.1 Degradation of phenoxy pesticides and chlorophenols in a flow-through
LED based photocatalytic reactor ..................................................................127 6.3.2 Degradation of 2,4-D with different combination of (UV, TiO2
photocatalyst plate, H2O2 and O2) in the UV-LED photoreactor . ...............128 6.3.3 Effect of DL-P .................................................................................................130 6.3.4 Effect of flow rates on the photocatalytic degradation of 2,4-D. ..................131 6.3.5 Effect of UV light intensity ...........................................................................133 6.3.6 Comparison of three different photocatalyst configurations .........................135
CHAPTER SEVEN: CONCLUSION AND RECOMMENDATION FOR FUTURE RESEARCH ............................................................................................................137
7.1.1 Photosensitized dechlorination of PCBs solubized in surfactant solution ....137 7.1.2 LED based photocatalytic treatment of pesticides and chlorophenols ..........138 7.1.3 Design a homogenous radiation field model for photocatalytic reactor ........138 7.1.4 A novel light emitting diode photocatalytic reactor for water treatment ......139
7.2 Recommendations for Future Research .................................................................140 7.2.1 Incorporating PCBs extraction using surfactants and PCBs
photodechlorination using sensitized visible light .........................................140 7.2.2 UVC-LED ......................................................................................................140 7.2.3 The decay of photocatalytic activity and its life time ....................................140 7.2.4 Hollow microsphere coated with TiO2 (HGMT)...........................................140 7.2.5 Scale-up of the reactor ...................................................................................141
APPENDIX B: INVESTIGATION OF PHOTODEGRADATION OF BIPHENYL IN ULTRAVIOLET WATER PURIFICATION SYSTEMS.................................181
B.1. Experimental ........................................................................................................181 B.1.1. Chemicals .....................................................................................................181 B.1.2. Photoreactor .................................................................................................181 B.1.3. Photodegradation of biphenyl in IPA ..........................................................181
B.2. Results and discussions ........................................................................................182
APPENDIX C: UV VIS ABSORPTION SPECTRUM OF DIFFERENT PESTICIDES ..........................................................................................................184
APPENDIX D: THE CALCULATION OF PERCENTAGE OF AVAILABLE PHOTONIC ENERGY FOR PHOTOCATALYTIC REACTION ........................186
x
List of Tables
Table 2-1: Definitions of the quantum yield (Oppenlander, 2003). ................................. 12
Table 2-2: Generation of hydroxyl radicals for different AOPs. ...................................... 15
Table 2-3: Sales/use of the top 20 pesticide active ingredient in Canada (Brimble et al., 2005). .................................................................................................................. 28
Table 2-4: Slurry versus Immobilized Photocatalytic Systems (Lasa et al., 2005). ........ 39
Table 3-1: First order rate coefficients (K) for PCB138 dechlorination in TEA-MB system with different concentration of surfactants. .................................................. 71
Table 3-2: First order rate coefficients (K) for PCB138 dechlorination in NaBH4-MB system. ...................................................................................................................... 74
Table 4-1: Light intensity of different photoreactors. ....................................................... 84
Table 4-2: First order rate coefficients (K) for photocatalytic pegradation of different pesticides. .................................................................................................................. 89
Table 4-3: Mixtures of Pesticides. .................................................................................... 89
Table 4-4: Percentage removal of pesticides at 0.028 kJ energy dosage. ......................... 91
Table 4-5: Percentage of pesticides adsorbed on the surface of TiO2 after 30 minutes of stirring in the dark. ................................................................................................ 93
Table 5-1: Parameters used for radiation field model calculation. ................................. 113
Table 5-2: Comparison of modeled light intensity and measured light intensity. .......... 113
Table 6-1: Average light intensity received by the photocatalytic plate. ....................... 124
Table 6-2: First order kinetic rate constants for different photocatalyst configurations . 135
xi
List of Figures and Illustrations
Figure 2-1: Classification of electromagnetic radiation in the wavelength range below 1200 nm. [Reproduced from (Oppenlander, 2003) with the permission]. ................ 10
Figure 2-2: Phenomenological subdivision of ultraviolet radiation into four sub-bands and their characteristic effects. [Reproduced from (Oppenlander, 2003) with the permission]. ............................................................................................................... 10
Figure 2-3: Photochemical activation of TiO2. ................................................................. 19
Figure 2-4: Molecular structure of 2,4-dichlorophenoxyacetic acid. ............................... 29
Figure 2-5: Molecular structure of 2-methyl-4-chlorophenoxyacetic acid. ...................... 31
Figure 2-6: Molecular structure of chlorophenols. ........................................................... 32
Figure 2-7: Molecular structure of polychlorinated biphenyls. ........................................ 34
Figure 2-8: Solar spectral irradiance distribution on the surface of earth. [Reproduced from (Hulstrom et al., 1985) with permission] ......................................................... 42
Figure 2-9: Fractional cumulative integrated irradiance vs. wavelength. [Reproduced from (Hulstrom et al., 1985) with permission] ......................................................... 43
Figure 2-10: An inner working on an LED. [Adapted from (Wikipedia, 2011)] ............. 46
Figure 2-11: Various photochemical reactor configurations. [Reproduced from (Pareek et al., 2008) with permission] ...................................................................... 48
Figure 2-12: Scheme of a multiple tube reactor. [Reproduced from (Ray and Beenackers, 1998) with permission] ......................................................................... 50
Figure 2-13: Scheme of an optical fibre photocatalytic reactor. [Reproduced from (Nguyen and Wu, 2008) with the permission] .......................................................... 51
Figure 2-14: Scheme of a rotating disk reactor. [Reproduced from (Hamill et al., 2001) with the permission] ....................................................................................... 51
Figure 2-15: Top view of a distributive photocatalytic reactor. [Reproduced from (Ray and Beenackers, 1998) with permission] ......................................................... 52
Figure 2-16: Experimental setup of taylor vortex photocatalytic reactor:(1) motor, (2) speed controller, (3) gear coupling, (4) UV lamp (5) sample collection point (6) lamp holder (7) outer cylinder and (8) catalyst-coated inner cylinder. [Reproduced from (Dutta and Ray, 2004) with the permission]............................... 52
xii
Figure 2-17: Scheme of a fluidized photocatalytic reactor. [Reproduced from (Vaisman et al., 2005) with permission] ................................................................... 53
Figure 2-18: Solar photocatalytic reactor: (a) parabolic trough reactor (PTR) (b) compound parabolic collector (CPC). [Reproduced from (Braham and Harris, 2009) with permission] ............................................................................................. 54
Figure 2-19. Typical reactor layout for an (a) inclined plate collector and (b) double skin sheet photoreactor. [Reproduced from (Braham and Harris, 2009) with permission] ................................................................................................................ 55
Figure 2-20: Typical reactor layout for (a) horizontal rotating disk reactor and (b) water bell reactor. [Reproduced from (Braham and Harris, 2009) with permission] ................................................................................................................ 56
Figure 2-21: Schematic for photon transport. .................................................................. 57
Figure 3-1: Reductive dechlorination of PCB 138 using LMB with TEA as the reducing agent; [PCB 138] = 6.6 mg L-1, [SDS] = [CTAB] = [TWEEN 80] = 3.2 g L-1, [MB] = 750 mg L-1, [TEA] = 68 g L-1, Io = 5.2×1016 photon s-1 . ................... 68
Figure 3-2: Reductive dechlorination of PCB 138 using LMB with NaBH4 as the reducing agent; [PCB 138] = 20 mg L-1, [SDS] = [CTAB] = [TWEEN 80] = 3.2 g L-1; [MB] = 600 mg L-1; [NaBH4] = 20 g L-1, Io = 3.0×1016 photon s-1
. ................ 69
Figure 3-4: Dechlorination of Aroclor1254 solubilized with TWEEN 80 in the presence of MB and TEA or NaBH4: [Aroclor1254] = 10 mg L-1, [MB] = 600 mg L-1, [TWEEN80] = 1.6 g L-1, [TEA] = 108 g L-1, [NaBH4] = 20 g L-1, Io = 3.0 ×1016 photon s-1. ........................................................................................................ 75
Figure 3-5: The product distribution (after six minutes irradiation) for dechlorination of PCB 138 solubilized by TWEEN 80 (1.6 g L-1) or CTAB (1.6 g L-1) in the presence of MB (600 mg L-1) and TEA (68 g L-1), Io = 3.0 ×1016 photon/s, P: peak area of each congener from GC, Po: the peak area of initial PCB 138. ........... 76
Figure 3-6: The product distribution (after six minutes irradiation) for dechlorination of PCB 138 solubilized by TWEEN 80 (1.6 g L-1) or CTAB (1.6 g L-1) in the presence of MB (600 mg L-1) and NaBH4 (10 g L-1); Io = 3.0 ×1016 photon s-1, P: peak area of each congener from GC, Po: the peak area of initial PCB 138. ........... 77
Figure 4-1: LED photoreactor and insert. ......................................................................... 82
Figure 4-2: Photocatalytic degradation of different pesticides with UV-LED photoreactor (Io = 8.55×1016 photon s-1, CTiO2=2.0 g L-1, Co=20 mg L-1): (a) loss of parent pesticides; and (b) loss of total organic carbon. ........................................ 87
Figure 4-3: Photocatalytic degradation of pesticides mixture with UV-LED photoreactor based on the loss of pesticides detected by HPLC (Io =8.55×1016
xiii
photon s-1, CTiO2=2 g L-1, Co=20 mg L-1): (a) mixture containing 4-CP and 2,4-DCP; (b) mixture containing 4-CP and 2,4-D; (c) mixture containing 2,4-DCP and 2,4-D. .................................................................................................................. 90
Figure 4-4: Photocatalytic degradation of 2,4-D with different TiO2 loadings and LED irradiation (Io =8.55×1016 photon s-1, Co=20 mg L-1). ..................................... 94
Figure 4-5: Photocatalytic degradation of 2,4-D with UV-LED photoreactor under different light conditions (Co=20 mg L-1, CTiO2=2 g L-1). ......................................... 96
Figure 4-6: Photocatalytic degradation of 2,4-D in the two photoreactors: Co=20 mg L-1, CTiO2=2 g L-1. (a): LED reactor, Io =8.55×1016 photon s-1; (b): Rayonet reactor, Io =8.25×1016 photon s-1. ............................................................................. 98
Figure 5-1: UV-LED array and photocatalyst plate. ....................................................... 103
Figure 5-2: Directivity of radiation (NICHIA, 2013). ................................................... 105
Figure 5-3: Cartesian and polar coordinates in radiation system. ................................... 107
Figure 5-4: Scheme of UV-LED radiation. .................................................................... 109
Figure 5-5: Geometry of sensor. ..................................................................................... 111
Figure 5-6: The radiation field with different ID: (a) ID=0.01 m, gap=0.025 m; (b) ID= 0.04 m, gap=0.025 m. ...................................................................................... 115
Figure 5-7: The effect of irradiated distance (ID) on Maximum Error........................... 116
Figure 5-8: Optimal combination of ID and gap. ........................................................... 116
Figure 5-9: Selection of light output of UV-LED. .......................................................... 117
Figure 6- 1: SEM image of anodized TiO2 nanostructrure. ............................................ 122
Figure 6-2: Scheme of an LED based photocatalytic reactor. ........................................ 123
Figure 6-3: Radiation field on a photocatalyst plate under different conditions; (a) DL-
P = 0.014 m, 4 by 4 LEDs panel; (b) DL-P = 0.034 m, 4 by 4 LEDs panel; (c) DL-P = 0.054 m, 4 by 4 LEDs panel. ............................................................................... 125
Figure 6-4: Photodegradation of MCPA, 2,4-D, 2,4-DCP and 4-CP in a UV-LED photoreactor: flow rate = 2.03 L min-1; DL-P = 0.54 cm; Ia=17.3 mW cm-2. .......... 128
Figure 6-5: Photodegradation of 2,4-D in a flow-through UV-LED photoreactor: flow rate = 2.03 L min-1; DL-P = 0.54 cm ; Ia=17.3 mW cm-2. ........................................ 129
Figure 6- 6: The effect of DL-P on 2,4-D degradation: flow rate =2.03 L min-1, Ia=17.3 mW cm-2. ................................................................................................................. 131
xiv
Figure 6-7: The effect of flow rate on degradation of 2,4-D: DL-P = 5.4 cm, Iaverage=17.3 mW cm-2. ............................................................................................ 132
Figure 6-8: The effect of light intensity on degradation of 2,4-D: DL-P = 5.4 cm, Flow rate=2.03 L min-1. ................................................................................................... 134
Figure 7-1: Scheme of a scale-up LED based photocatalytic reactor. ............................ 141
Figure A-A-1: Ultrasonic extraction efficiency of PCBs from 10 g soil under different experimental conditions. ......................................................................................... 180
Figure A-B-1: Ultraviolet water purification system. ..................................................... 181
Figure A-B-2: Degradation of biphenyl under different flow rate. ................................ 183
Figure A-B-3: The pseudo first order kinetics of biphenyl degradation under different flow rate. ................................................................................................................. 183
Figure A-C-1: UV-Vis absorption spectra of 40 mg/L of 2,4-D in water. ..................... 184
Figure A-C-2: UV-Vis absorption spectra of 40 mg/L of 2,4-DCP in water. ................ 184
Figure A-C-3: UV-Vis absorption spectra of 40 mg/L of 4-CP in water. ...................... 185
Figure A-C-4: UV-Vis absorption spectra of 40 mg/L of MCPA in water. ................... 185
Figure A-D-1: The emission spectrum and TiO2 band edge. ......................................... 187
xv
List of Symbols, Abbreviations and Nomenclature
Symbol Definition
A Cross surface area, m2
Ap Area of photocatalyst plate, m2
c Light speed in vacuum, m s-1
C Concentration of substrate, mole L-1
Co Initial concentration of parent compound, mg L-1
d Distance, m
Ds-p The distance between shielding glass and photocatalyst, cm
Dl-p The distance between LEDs and photocatalyst, cm
do Specific distance, m
E Energy of photon, J
g Distance between point (x, y) and point (xo, yo), m
Gv Incident light intensity, photon s-1 m-2
h Plank constant, 6.62*10-34 J s
I Light intensity, mw cm-2
Ia Average light intensity, mw cm-2
Imax Maximum of light intensity, mw cm-2
Imeasured Light intensity measured by UV meter, mw cm-2
Imin Minimum of light intensity, mw cm-2
Imodel Light intensity calculated by model, mw cm-2
Io Light intensity measured by actinometry, photon s-1
It Light output of an LED lamp, mw
Iλ Specific light intensity, photon s-1 m-2
k Apparent reaction rate constant, mol L-1 s-1
K First order rate coefficient, s-1
Ka The average of first order rate coefficient, s-1
Kad Adsorption coefficient, L mol-1
lg The thickness of glass plate, m
Me Max error
xvi
p(Ω-Ω') Phase function for scattering in RTE
Q Number of photons
qa Rate of photon absorption, photon s-1
qe Rate of photon emission, photon s-1
qin Rate of photon in-scattered, photon s-1
qout Rate of photon out-scattered, photon s-1
r Radius, m
R Radial distance, m
Re Radiation directivity function
Re' Modified radiation directivity function
rp Kinetic reaction rate, mole L-1 s-1
rs Radius of the sensor, m
s Surface area, m2
S Direction vector, m
t Time, s
T Transmittance
v Frequency, s-1
V Elementary control volume, m3
Wa Volumetric rate of photon absorption, photon s-1 m-3
We Volumetric rate of photon emission, photon s-1 m-3
Win Volumetric rate of photon in-scattered, photon s-1 m-3
Wout Volumetric rate of photon out-scattered, photon s-1 m-3
x x-coordinates
xo x-coordinates of LED position
y y-coordinates
yo y-coordinates of LED position
z z-coordinates
α Volumetric absorption coefficient, m-1
β Extinction coefficient, m-1
γ Constant
η Constant
xvii
θ View angle, radian
λ Wavelength
λmax Wavelength of maximum emission, nm
σ Volumetric scattering coefficient, m-1
τ Asymmetry factor
ψ Scattering angle, radian
Ω Solid angle, steradian
Abbreviations Definition
2,4-D 2,4-Dichlorophenoxyacetic Acid
2,4-DCP 2,4-Dichlorophenol
4-CP 4-Chlorophenol
AOPs Advanced Oxidation Processes
ARPs Advanced Reduction Processes
CCA Chromated Copper Arsenate
CMC Critical Micelle Concentration
CTAB Cetyltrimethylammonium Bromide
DDT Dichlorodiphenyltrichloroethane
ECD Electron Captured Detector
EQE External Quantum Efficiency
GC Gas Chromatography
HGMT Hollow Glass Microspheres Coated with Anatase TiO2
HPLC High Performance Liquid Chromatography
IARC International Agency for Research Cancer
ID Irradiated Distance
LED Light Emitting Diode
LMB Leuco-methylene Blue
LVREA Local Volumetric Rate of Energy Absorption
MB Methylene Blue
MC Monte Carlo
MCPA 2-methyl-4-chlorophenoxyacetic acid
xviii
PCB138 2,2',3,4,4',5'-Hexachlorobiphenyl
PCBs Polychlorinated biphenyls
PEPO Photon Energy per Order
PFP Pentafluorophenyl
PVC Polyvinyl Chloride
RTE Radiation Transport Equation
SDS Sodium Dodecyl Sulfate
SEM Scan Electron Microscopy
TEA Triethylamine
TOC Total Organic Carbon
TWEEN80 Polyoxyethylene (80) Sorbitanmonooleate
USEPA United States Environmental Protection Agency
UV Ultraviolet
UVA Ultraviolet, subtype A
UVB Ultraviolet, subtype B
UVC Ultraviolet, subtype C
UV-Vis Ultraviolet-visible
VUV Vacuum Ultraviolet
1
Chapter One: INTRODUCTION
1.1 Background
In the past several decades, increased population, industrialization and agricultural
activities have led to an increase in the level of water contamination of receiving water
bodies. Pesticides, a major category of pollutants causing water contamination, pose a
potential threat to human health and the environment. Using pesticides is almost a
necessary way to maintain and improve the food production for an ever increasing world
population. However, extensive use of pesticides has resulted in water pollution in
different ways such as runoffs, run-ins and leaching (Polyrakis, 2009). The primary focus
of this thesis is to study the photochemical treatment of pesticides in water and to design
an efficient light emitting diode (LED) based photocatalytic reactor. Polychlorinated
biphenyls (PCBs) form a secondary interest in this thesis.
Pesticides exposure can cause different acute and chronic effects on human health
(Younes and Galal-Gorchev, 2000). A large number of pesticides, such as mancozeb,
dithiocarbamate and organophosphorus compound, manifest their toxicity through
functional and biochemical action in the central and peripheral nervous system (Kimura
et al., 2005). Several chronic diseases have been linked to the long-term exposure to
pesticides. Examples include porphyria following exposure to hexachlorobenzene,
delayed neuropathy from exposure to organophosphates and chloracne due to long-term
exposure to chlorophenoxy acid derivatives and chrolophenols (Younes and Galal-
Gorchev, 2000). Besides, cancers of the soft tissue, lung, gonads, liver, brain, the urinary
2
tract and the digestive system have been associated with long-term exposure to some
pesticides, although the association is not firm (Younes and Galal-Gorchev, 2000).
Polychlorinated biphenyls (PCBs) are toxic contaminants, which are less soluble in
water, but can bind to sediments of aquatic systems or adsorb on suspended particulates
(Sullivan et al., 1983, Manchester-Neesvig et al., 1996, Bergen et al., 1998). Once they
are released to the environment, they are difficult to remediate. The occurrence of water
contamination by PCBs is due to desorption from sediments or leaching from landfills
and contaminated soil. PCBs have been demonstrated to cause a variety of adverse health
effects. Data on animal experiments have provided conclusive evidence that PCBs are
carcinogenic to animals and can cause a number of non carcinogenic health effects,
including effects on the immune system, nervous system, endocrine system, reproductive
system and others (USEPA, 2013a). The studies also support that PCBs can cause
potential carcinogenic and non-carcinogenic effects to human beings (USEPA, 2013a).
Long term exposure to PCBs can cause damages to heart, kidney, liver and central
nervous systems (Erickson, 1997).
To alleviate water pollution with these two categories of pollutants, a variety of
techniques has been developed: bio-treatment (Hussain et al., 2009, Portier et al., 1990,
Zhang et al., 2004, Natarajan et al., 1996), membrane separation (Bhattacharya, 2006,
Boussahel et al., 2000), activated carbon adsorption (Foo and Hameed, 2010, Sotelo et
al., 2002), coagulation followed by settling (Dempsey and O'Melia, 1984) and others.
3
Although these methods, to some extent, can reduce the water contamination, several
drawbacks limit them from wider applications. Bio-technologies may require specific
pollutant resistant microbes as well as appropriate environmental conditions such as pH,
nutrients and temperature. Physical separation can remove contaminants from water and
transfer them to other phases. Nevertheless, the disposal of the concentrate or sludge can
be a serious problem. Besides, regeneration of adsorbents and fouling of membranes limit
the application of these techniques. To detoxify and degrade these compounds,
technologies based on photochemical processes are considered as good choices
(Devipriya and Yesodharan, 2005, Ollis et al., 1991, Parsons, 2004, Izadifard et al.,
2010a, Achari et al., 2003, Chu et al., 2005). These methods are quite fast and mostly
lead to complete degradation of the contaminants.
1.2 Photochemical Treatment Processes
Most photochemical treatment processes are based on advanced oxidation processes
(AOPs), using the generated hydroxyl radicals, positive holes, oxygen species and other
strong oxidants to degrade the organic compounds. They have been widely used in
removing organic contaminants in water and wastewater such as disinfection by-
products, pesticides, endocrine disruptors and so on (Parsons, 2004). Besides, during
some photochemical processes, highly reactive reducing radicals, such as free electrons,
may be formed. These strong reducing agents can be used to degrade the oxidized
contaminants such as nitrate, perchlorate, dichlorophenols and perfluorooctanoic acid
(Vellanki et al., 2013). Such photochemical treatment techniques are called the advanced
reduction processes (ARPs).
4
In this research, the major interest is focused on the removal of aqueous contaminants
such as pesticides. The most commonly used pesticides such as 2,4-D, MCPA and
chlorophenols were selected as the studied compounds. TiO2 photocatalysis based on
AOPs was chosen for degrading these compounds, as it does not consume large amount
of chemicals and is able to use the longer wavelength domain of ultraviolet light, which is
a part of UV region in the solar spectrum received on the surface of earth. Application of
TiO2 photocatalysis can lead to usage of longer (less energy) light sources. To treat PCBs
in aqueous medium, photosensitization based on ARPs are used.
1.3 Light emitting diode (LED) in photocatalytic reactors
To apply TiO2 photocatalysis in water and wastewater treatment, an efficient
photocatalytic reactor need to be designed and fabricated. The rapid development of LED
technology has made it a promising light source in photochemical applications. This
mercury-free light source is able to provide monochromatic light, has a longer lifetime,
and efficient electricity to light conversion (Würtele et al., 2011). Furthermore, the small
size of LEDs does not limit the geometry of the reactor. All these advantages have made
LEDs favourable in photocatalytic reactor designs. The application of LEDs has been
reported in photochemical treatment of air and water by several researchers (Huang et al.,
2009, Shie et al., 2008, Chen et al., 2005, Wang and Ku, 2006, Ghosh et al., 2008, Ghosh
et al., 2009). In this research, ultraviolet-light emitting diodes (UV- LEDs) are selected
for reactor design and fabrication.
5
1.4 Research Objectives and Scopes
The goal of this research is twofold: (1) develop efficient photochemical technologies to
treat contaminants (e.g. pesticides) in aqueous medium and design an efficient LED
based photocatalytic reactor; (2) study photochemical treatment of PCBs in aqueous
medium. To achieve this goal, four objectives are defined:
Dechlorinate PCBs in aqueous surfactant solutions using photosensitized visible
light irradiation.
o Investigate the photodechlorination of PCBs using Leuco-methylene blue
as a photosensitizer and cool white lamps as a light source
o Determine the effect of the type and the concentration of surfactants on the
photodechlorination of PCBs.
o Optimize the PCBs dechlorination conditions.
Investigate the photocatalytic degradation of certain pesticides in a batch UV-
LED photoreactor.
o Design a batch UVA-LED based photoreactor.
o Investigate the photocatalytic degradation of pesticides mixtures.
o Study the effect of photocatalyst loading and light intensity on the
photocatalytic degradation rate.
o Compare the photocatalytic degradation of pesticides in the LED
photoreactor with the mercury lamps.
Develop a radiation field model for a UV-LED photocatalytic reactor and design a
homogenous radiation field.
o Determine the most efficient radiation field for a photocatalytic reactor.
6
o Develop and validate a mathematical radiation field model for LED
arrays.
o Develop a method for designing a homogenous radiation field generated
by LED arrays.
Design, fabricate and test an efficient UV-LED based photocatalytic reactor, as
well as optimize the reactor performance and provide useful information for the
scale-up of the reactor.
o Design a novel photocatalytic reactor using UV-LED and TiO2 nanotubes.
o Evaluate the performance of the photocatalytic reactor by studying the
degradation of pesticides.
o Optimize the photocatalytic reactor performance through the study of the
effect of different operational parameters on the photocatalytic
degradation rate of pesticides.
1.5 Thesis Overview
This thesis contains seven chapters as outlined here.
Chapter one provides the general background information, research scope and objectives,
and an outline of the dissertation.
Chapter two provides a review of principles of photochemistry, photocatalysis,
photochemical treatment of PCBs and pesticides, designs of photocatalytic reactors and
radiation field modelling.
7
Chapter three describes a study of dechlorination of PCBs in surfactant solution with
visible light irradiation using a photosensitizer-Leuco methylene blue (LMB). In this
chapter, the generation of LMB through two different ways are studied. The impact of
surfactant type and surfactant concentration on PCBs photodechlorination efficiency is
investigated.
Chapter four describes photocatalytic degradation of phenoxy herbicides and
chlorophenols with a UV-LED light source in a TiO2 slurry system. During this research,
a batch UV-LED photoreactor is fabricated. The impact of light intensity and TiO2
loading on photocatalytic degradation is investigated.
Chapter five describes the development of a radiation field model for a LED based
photocatalytic reactor and the design of a homogenous radiation field.
Chapter six describes the design, fabrication and optimization of a flow-through LED
based photocatalytic reactor. Parameters such as flow rate, light intensity, and
photocatalyst configuration are studied.
Chapter seven provides a summary of research results as well as recommendations for
future research.
This thesis is written in a paper format where chapter 3, 4, 5 and 6 comprise separate
papers. Chapter 3 and 4 have been published as "Yu, Linlong; Izadifard Maryam; Achari,
8
Gopal; Langford, Cooper H., 2013. Electron transfer sensitized photodechlorination of
surfactant solubilized PCB 138. Chemosphere, 90, 2347-2351." and "Yu, Linlong;
Achari, Gopal; Langford, Cooper H., 2013. LED-Based Photocatalytic Treatment of
Pesticides and Chlorophenols. Journal of Environmental Engineering, 139, 1146-1151.",
respectively. Chapter 5 has been submitted to Journal of Environmental Engineering and
Science.
9
Chapter Two: LITERATURE REVIEW
2.1 Principle of Photochemistry
2.1.1 Light and photon
Photochemistry is the science of light-induced chemical reactions. The modern theory of
quantum mechanics considers light beam as consisted a number of photons which possess
the property of both waves and particles (Turro, 1991). Each photon has energy related to
its wavelength (Plank's Equation, Equation [2-1]). The shorter the wavelength the higher
the energy it carries.
E = hν =hcλ
[2-1]
Where E is the radiant energy of the photon (J), h is Plank constant (6.62*10-34 J·s), ν is
the frequency of photon (s-1), λ is the wavelength of photon (m) and c is the velocity of
photon travelling in vacuum (m s-1).
The wavelength range generally utilized in photochemistry lies between 170 nm and 1000
nm (Figure 2-1) (Oppenlander, 2003), which is divided into five sub region: the vacuum-
nm), VIS (380-850 nm) and infrared (800-1000nm) The subdivisions of the UV spectral
domain are related to physical, chemical, biological or biochemical effects showed in
Figure 2-2.
10
Figure 2-1: Classification of electromagnetic radiation in the wavelength range below 1200 nm. [Reproduced from (Oppenlander, 2003) with the permission].
Figure 2-2: Phenomenological subdivision of ultraviolet radiation into four sub-bands and their characteristic effects. [Reproduced from (Oppenlander, 2003) with the permission].
Photoionization
M→M++e
0 200 400 600 800 1000
Wavelength (nm)
IR
Visible light UVA
UVB
UVC
VUV X-ray
γ-ray
Vibrational excitation
M→Mvib
Photoexcitation
M→M*
400 100 150 200 250 300 350
Wavelength (nm)
UVA (315-380 nm)
UVB (280-315 nm)
UVC(200-280 nm)
VUV (100-200nm)
Absorbed by Organic Chromophores
Absorbed by all substances including H2O,O2,CO2
Absorbed by all Proteins, DNA, RNA,O2
Sunburn Skin Cancer
Sun Tanning
11
2.1.2 The electronic excited states
In photochemistry, only the absorbed photon can cause a photochemical reaction, and
each photon is absorbed by a single molecule to initiate the reaction (Turro, 1991).
Absorption in the wavelength region of photochemical interest promotes the absorber
from its ground state to its excited state. Absorption at longer wavelengths (infra-red)
usually leads to the excitation of vibrations or rotations of a molecule in its ground state;
generally, only electronically excited states are involved in photochemical processes
(Wayne, 1988). The fates of excited species 'A*' are shown as below (Wayne, 1988) :
The excited species 'A*' can lose its energy by emitting a photon, which gives rise
to the phenomenon of luminescence.
The excess energy of 'A*' can also be relieved by an atom or molecule 'M' in the
form of physical quenching. Normally, in this process, the excess energy of 'A*' is
converted to translational or vibrational excitation of 'M*' at lower energy.
The excited species A* can transfer energy to other molecules to generate other
excited species, which can then participate in any of the processes including
relaxation to the ground state (radiationless decay);
The excited species A* may undergo dissociation, direct reaction, ionization or
spontaneous isomerization.
2.1.3 Quantum yield
The absorption of photons can cause other processes rather than the desired reaction. To
determine the efficiency of the photochemical reaction, the concept of quantum yield was
12
developed. Four commonly used definitions of quantum yield are shown in Table 2-1.
The quantum yield is a unitless constant, usually ranging from zero to one; the value of
quantum yield larger than one indicate a photo-induced chain reaction involving radicals
species or photo-generated catalysis (Oppenlander, 2003).
Table 2-1: Definitions of the quantum yield (Oppenlander, 2003).
Mathematical
expression
Definition
𝜙𝜆 =dn(event)/dt
Φpabs
Universally valid: Number n of events per unit time divided by the
number of photons absorbed during this period
𝜙𝜆 =dn(M)/dt
Φpabs
Number n of reactant molecules M consumed per unit time divided
by the number of photons absorbed during this period
𝜙𝜆 =dn(P′)/dt
Φpabs
Number n of photoproduct molecules P' formed per unit time
divided by the number of photons absorbed during this period
𝜙λ1−λ2
=dn(P′)/dt
Φpabs(λ1−λ2) ≠ 𝜙𝜆
Ratio of the number m of photoproduct molecules formed per unit
time to the total number of photons absorbed in the spectral region
λ1- λ2 during this period
note: Φpabs and Φp
abs(λ1−λ2)are the absorption rates of photons.
2.1.4 Direct photolysis
Direct photolysis involves the transformation of a chemical resulting from the direct
absorption of a photon. Absorption of photons with high energy can promote the
contaminants (e.g.2-chloro-N-methylacetanilide ) to their excited singlet states from
13
electronic ground state. This excited state can then undergo, among other processes; (i)
homolysis (ii) heterolysis or (iii) photoionization (Burrows et al., 2002). Most organic
compound show absorption bands at relatively short UV wavelengths capable of
producing direct photolytic degradation of these compounds.
2.1.5 Photosensitized degradation
Photosensitization is the process of initiating a reaction through the use of a
photosensitizer capable of absorbing light and transferring the energy or exchanging
electrons with the reactants (Burrows et al., 2002). The major advantage of
photosensitized photodegradation is its possibility to use light of wavelengths longer than
those corresponding to the absorption characteristics of the pollutants.
2.1.6 Photocatalysis
Photocatalysis is a chemical reaction induced by absorption light by a photocatalyst
(Ohtani, 2008a). With solid photocatalyst, the reaction is activated by absorption of a
photon with sufficient energy, i.e. equal or higher than the band-gap energy of the
photocatalysts (Fox and Dulay, 1993, Herrmann, 2005, Hoffmann et al., 1995). The
band-gap energy is the energy difference between the bottom of conduction band (lowest
unoccupied molecular orbital) and the top of the valance band (highest occupied
molecular orbital) related to the electronic structures of semiconducting materials.
Various semiconductors such as TiO2, CdO, ZnO, WO3, CdS, CdSe, GaP, GaAs, ZnS,
SnO2, Fe2O3, SrTiO3, BaTiO3 etc, have been used as photocatalysts. Generally, the best
photocatalytic performances are obtained with titanium dioxide as catalyst (Herrmann,
14
2005). The details of TiO2 photocatalysis fundamental and mechanism will be described
in the section 2.2.
2.1.7 Advanced Oxidation Processes (AOPs)
AOPs are processes designed to degrade recalcitrant organic compounds using chemical
oxidants. Most organic contaminants can be completely mineralized or partially
mineralized to innocuous compounds using appropriate AOPs. Currently, the major light
induced AOPs include UV&H2O2, Ozone&UV, vacuum UV, TiO2 photocatalysis, and
others. Besides, there are several non-light induced AOPs such as H2O2, Fenton’s reagent
and ozonation. Although different AOPs make use of different reaction systems (Table 2-
2), the chemistry of these reaction systems are similar: generation of highly reactive
oxidative species, such as hydroxyl radicals (OH•), positive holes and singlet oxygen
(Andreozzi et al., 1999). The oxidation potential of hydroxyl radicals are greater than that
of most conventional oxidants such as chlorine, oxygen, ozone, etc. (Parsons, 2004).
2.2 TiO2 Photocatalysis
2.2.1 TiO2 as a photocatalyst
TiO2 is considered to be the most successful photocatalyst as it has several advantages
such as: (1) photo active (2) low toxicity (3) biologically and chemically stable (4) able to
utilize near UV light and (5) economic (Bhatkhande et al., 2002, Linsebigler et al., 1995,
Hoffmann et al., 1995). Titanium dioxide naturally exists in three crystal forms: anatase,
rutile and brookite. Brookite is extremely difficult to synthesize, while anatase and rutile
15
Table 2-2: Generation of hydroxyl radicals for different AOPs.
Type of AOPs Spectral domain Reactions
Vacuum UV VUV 𝐻2𝑂ℎ𝑣��𝐻𝑂 ∙ +𝐻 ∙
UV/H2O2 UVC 𝐻2𝑂2ℎ𝑣��𝐻𝑂 ∙ +𝐻𝑂 ∙
TiO2/UV UVA-UVC
𝑇𝑖𝑂2ℎ𝑣�� 𝑒− + ℎ+
ℎ+ + 𝐻2𝑂 → 𝐻𝑂 ∙ +𝐻+
ℎ+ + 𝑂𝐻− → 𝐻𝑂 ∙
O3/UV UVC
𝑂3ℎ𝑣��𝑂(𝐷) + 𝑂2
𝑂(𝐷) + 𝐻2𝑂 → 𝐻2𝑂2
𝐻2𝑂2ℎ𝑣��𝐻𝑂 ∙ +𝐻𝑂 ∙
Ozonation No UV
𝐻𝑂− + 𝑂3 → 𝑂2 + 𝐻𝑂2−
𝐻𝑂2− + 𝑂3 ⇄ 𝐻𝑂2 ∙ +𝑂3 ∙−
𝐻𝑂2 ∙⇄ 𝐻+ + 𝑂2 ∙−
𝑂2 ∙−+ 𝑂3 → 𝑂2 + 𝑂3 ∙−
𝑂3 ∙−+ 𝐻+ → 𝐻𝑂3 ∙
𝐻𝑂3 ∙→ 𝐻𝑂 ∙ +𝑂2
𝐻𝑂 ∙ +𝑂3 → 𝐻𝑂2 ∙ +𝑂2
Fenton process No UV
𝐹𝑒2+ + 𝐻2𝑂2 ⇄ 𝐹𝑒𝑂2+ + 𝐻2𝑂
𝐹𝑒2+ + 𝐻2𝑂2 → 𝐹𝑒3+ + 𝑂𝐻− + 𝑂𝐻 ∙
𝐹𝑒3+ + 𝐻2𝑂2 ⇄ 𝐹𝑒𝑂𝑂𝐻2+ + 𝐻+
𝐹𝑒𝑂𝑂𝐻2+ ⟶ 𝐻𝑂2 ∙ +𝐹𝑒2+
16
can be produced easily in the laboratory (Bickley et al., 1991). Among these three crystal
forms, anatase and rutile are the two most commonly used types and have been employed
most in the photocatalytic study. The band-gap energy are, respectively, 3.0 eV, 3.2 eV,
for rutile phase and anatase phase, and its amorphous form is reported to have the band-
gap energy varying from 3.2 to 3.5 eV (Roy et al., 2011). Even though the most active
form of titanium dioxide is believed to be anatase, a mixed phase of anatase and rutile
appears to achieve better photocatalytic efficiency (Bickley et al., 1991). The co-presence
of anatase and rutile phase introduce mesoporosity and a wider pore size distribution,
which may be responsible for the high level of photocatalytic activity (Thiruvenkatachari
et al., 2008). Hurum et al. (2003) proposed three possible reasons for the greater
photocatalytic activity of TiO2 mixed phase: (1) the band-gap of rutile is smaller than that
of anatase and extends the useful wavelength range of photoactivity; (2) the transfer of
photoexcited electrons between rutile/anatase phase enhance the charge separation and
slows down electron-hole recombination; (3) the small size of the rutile crystallites
enhance the photocatalyst activity.
Degussa (Evonik) P25, Aeroxide TiO2 P25, via the chloride technology method is
currently the de-facto commercial reference TiO2 photocatalyst (Alonso-Tellez et al.,
2012). It is widely used in potocatalytic reaction systems because of its high
photocatalytic activity, and has been reported in more than one thousand papers since
1900 (Ohtani et al., 2010). P25 has a large surface area (50 m2 g-1) (Zertal et al., 2004)
and small crystal size (20 nm). Theoretically, a photocatalyst with larger surface area and
smaller particle size can provide more active sites for illumination and adsorption of the
17
reactants, leading to a higher expected photocatalytic activity. The composition of P25 is
reported to be 70% anatase and 30% rutile or 80% anatase and 20% of rutile, however,
the exact crystalline composition seems to be unknown, presumably due to a lack of
determination techniques for crystalline contents in nano-sized particle samples (Ohtani
et al., 2010, Ohtani, 2008b). Except Degussa P25, other commercial TiO2, such as
products from Millennium and Hombikat also show their high photocatalytic activities
(Zertal et al., 2004, Alonso-Tellez et al., 2012) .
2.2.2 Mechanism of TiO2 photocatalysis
The fundamentals and mechanism of TiO2 photocatalysis have been intensively reported
in many literatures (Fujishima et al., 2000, Gaya and Abdullah, 2008, Fox and Dulay,
1993, Herrmann, 1999). The overall process of TiO2 photocatalysis can be broken into
five independent steps (Herrmann, 2005, Mozia, 2010) :
Transfer of the reactants in the bulk solution to the TiO2 surface;
Adsorption of the reactants on the surface of TiO2;
Reaction in the adsorbed phase;
Desorption of the products;
Removal of by-products from the interface region.
The third step includes all the photochemical processes (Herrmann, 2005) and is
summarized in Equations [2-2]~[2-14] (Mozia, 2010) and Figure 2-3. The initial step of
photon-induced reaction is the excitation of TiO2 by absorbing photon with formation of
electron-hole pair. Once TiO2 absorbs photons with sufficient energy, i.e. equal or larger
than its band-gap energy, electrons are promoted from the valence band to the conduction
18
band, while positive holes (h+) are left in the valence band (Equation [2-2]). The electron
and the hole can migrate to the catalyst surface and participate in the redox reactions
(Equations [2-4] ~ [2-14]) with different species adsorbed on the catalyst surface. A
recombination of the electron and hole will occur if no suitable electron and hole
scavengers are present (Equation [2-3]). If oxygen is present in the water (e.g. water
open to air), it will capture the electron in the conduction band to form the superoxide
radical ion while the remaining hole can react with surface-bond H2O molecule or
hydroxide ion to produce hydroxyl radicals. Hydroxyl radicals can also be generated
following the pathways through reactions shown in Equations [2-7] ~ [2-11]. The
hydroxyl radicals generated on the surface of illuminated TiO2 are supposed to be the
primary oxidizing species in the photocatalytic oxidation processes, which are highly
reactive and can degrade most organic compound and eventually convert them to CO2,
reactors, double-skin sheet photocatalytic reactors, horizontal rotating disk reactors and
water bell reactors. A parabolic trough reactor (Figure 2-18 a) is a light concentrating-
type unit, which uses a long parabolic reflecting trough to concentrate solar radiation on a
transparent tubular reactor placed on the parabolic focal line. Compound parabolic
reactors (Figure 2-18 b) are trough reactors without light concentrating devices. The
reflector in compound parabolic reactor is characterized with a two half-cylinders of
54
Figure 2-18: Solar photocatalytic reactor: (a) parabolic trough reactor (PTR) (b) compound parabolic collector (CPC). [Reproduced from (Braham and Harris, 2009) with permission]
parabolic profile which allow indirect light to be reflected onto the tubular reactor. An
inclined plate photocatalytic reactor (Figure 2-19 a) consists of an inclined surface coated
55
with photocatalyst. The reactant fluid flows through the inclined surface to form a thin
film. An double-skin sheet photoreactor (Figure 2-19 b) uses a double-skin transparent
plexiglass to construct a long, convoluted back and forth channel on a flat plane through
which the reactant fluid flow. A horizontal rotating disk reactor (Figure 2-20 a) has a
rotating disk of which the surface is exposed to sunlight. The reactant fluid is injected up
from the center of the disk and forms a fluid film on the surface of the disk. The rotation
of disk generates a turbulent flow regime in the fluid film. A water bell reactor (Figure 2-
20 b) features a nozzle which sprays a continuous and unsupported thin film of liquid
exposed to solar irradiation.
Figure 2-19. Typical reactor layout for an (a) inclined plate collector and (b) double skin sheet photoreactor. [Reproduced from (Braham and Harris, 2009) with permission]
56
Figure 2-20: Typical reactor layout for (a) horizontal rotating disk reactor and (b) water bell reactor. [Reproduced from (Braham and Harris, 2009) with permission]
2.6 Radiation-field modelling
A uniform distribution of light is necessary for an efficient photocatalytic treatment
system. In this context, a radiation field model is very useful. The model need to compute
the rate of photon absorption at any position within the reactor or 'local volumetric rate of
energy absorption' (LVREA) (Cassano et al., 1995). In homogenous media, the change of
the light intensity along the direction of photon propagation is due to absorption process
in the reaction media, while, in heterogeneous media scattering of radiation by particles
should also be accounted in the variation of light intensity. Various radiation field models
in homogenous or heterogeneous environments have been described in several papers
(Imoberdorf et al., 2008b, Jacob and Dranoff, 1969, Jacobm and Dranoff, 1970, Irazoqui
et al., 1973, Alfano et al., 1986a, Alfano et al., 1986b, Alfano et al., 1986c).
57
2.6.1 The Radiation transport equation (RTE)
In heterogeneous media (slurry TiO2 photocatalysis), the light intensity may change
along its light path due to photon absorption, scattering and emission. In-scattering and
emission from reacting mixture can increase the light intensity, while, the photon
absorption and out-scattering can reduce the light intensity (Figure 2-21).
Figure 2-21: Schematic for photon transport.
In a control volume V (m3), the photon balance equation can be described as (Cassano et
al., 1995):
𝑑𝑄𝑑𝑡
+ 𝐴 ∗ 𝑑𝐼𝜆 = −𝑞𝑎 + 𝑞𝑒 + 𝑞𝑖𝑛 − 𝑞𝑜𝑢𝑡 [2-25]
Where is the number of photon in control volume (photon), t is time (s), A is the cross
section surface area (m2), is the spectral light intensity (photon m-2 s-1), is rate of
58
photon absorption (photon s-1), is rate of photon emission (photon s-1), is rate of
photon in-scattered (photon s-1), is rate of photon out-scattered (photon s-1).
At steady state: =0, then the equation can be written as:
𝐴 ∗ 𝑑𝐼𝜆 = −𝑞𝑎 + 𝑞𝑒 + 𝑞𝑖𝑛 − 𝑞𝑜𝑢𝑡 [2-26]
Equation [2-27] is obtained by dividing Equation [2-26] with V; V=A*ds,
𝐴𝑑𝐼𝜆𝑉
= −𝑞𝑎/V + 𝑞𝑒/v + 𝑞𝑖𝑛/v − 𝑞𝑜𝑢𝑡/v [2-27]
Finally, the photon balance in a control volume can be described by Equation [2-28]
(Pareek et al., 2008, Cassano et al., 1995).
𝑑𝐼𝜆(𝑠,𝛺)𝑑𝑠
= −𝑊𝑎 + 𝑊𝑒 + 𝑊𝑖𝑛 −𝑊𝑜𝑢𝑡 [2-28]
Where is the solid angle (steradian), is volumetric rate of photon absorption
(photon m-3 s-1), is volumetric rate of photon emission (photon m-3 s-1), is
volumetric rate of photon in-scattered (photon m-3 s-1), is volumetric rate of photon
out-scattered (photon m-3 s-1).
At normal or low temperature, the spontaneous radiation emission can be neglected:
𝑊𝑒 = 0 [2-29]
59
In general, linear isotropic constitutive equations (Equation [2-30] &[2-31])can be used
to characterize absorption and out-scattering (Cassano et al., 1995),
𝑊𝑎 = 𝛼𝜆𝐼𝜆(𝑠,𝛺) [2-30]
𝑊𝑜𝑢𝑡 = 𝜎𝜆𝐼𝜆(𝑠,𝛺) [2-31]
Where is the volumetric absorption coefficient (m-1) and is the volumetric
scattering coefficient (m-1), usually, the combination of the absorption coefficient and the
scattering coefficient is defined as the extinction coefficient:
βλ = 𝛼𝜆 + 𝜎𝜆 [2-32]
Where is the extinction coefficient (m-1).
In the RTE, In-scattering is a more complicated term, which can be described by
Equation [2-33] (Cassano et al., 1995). Two assumptions were made here:
• Every scattering is independent of each other.
• The scattering is elastic, which means that the frequency of scattered radiation is
the same as incident radiation.
𝑊𝑖𝑛 =1
4𝜋𝜎𝜆 � 𝑝(𝛺′ ⟶ 𝛺)𝐼𝜆(𝑠,𝛺′)𝑑𝛺′
4𝜋
0 [2-33]
Where is the phase function for the in-scattering of photons, usually, the
phase function is normalized according to Equation [2-34]:
14𝜋
� 𝑝(𝛺′ ⟶ 𝛺)𝑑𝛺′ = 14𝜋
0 [2-34]
60
The phase function is the one that make the RTE difficult to solve. To simulate the model
more efficiently, the Henyey-Greenstein (HG) phase function (Equation [37]) can be an
appropriate choice (Marugán et al., 2006).
𝑃𝐻𝐺(𝜑) =1
4π∗
1 − τ2
(1 + τ2 − 2τ ∗ cos(φ))3/2 [2-35]
Where is the scattering angle and τ is the asymmetry factor of the scattered radiation
distribution. The value of τ varies smoothly from -1 to +1. When τ =0, it represents an
isotropic phase function.
According to Equations [2-25] ~ [2-35], the final RTE can be written as
𝑑𝐼𝜆(𝑠,𝛺)𝑑𝑠
= −βλ𝐼𝜆(𝑠,𝛺) +1
4𝜋𝜎𝜆 � 𝑝(𝛺′ ⟶ 𝛺)𝐼𝜆(𝑠,𝛺′)𝑑𝛺′
4𝜋
0 [2-36]
For homogenous media (when a photocatalyst is immobilized), the scattering term on
Equation [2-36] can be excluded.
Then the incident light intensity from all direction:
Gv(s) = � 𝐼𝜆(𝑠,𝛺)𝑑𝛺4𝜋
0 [2-37]
Where is the incident light intensity (photon m-2 s-1).
And the LVREA at any point is given by
LVREA= αλGv(S) [2-38]
2.6.2 Numerical methods to solve the RTE
The RTE is an integral-differential equation, an exact analytical solution is impossible
except for homogeneous photoreaction systems, where scattering phenomenon is not
taken into account (Pareek et al., 2008). Numerical method offers a viable alternative to
61
solve the RTE. Carvalho and Farias (1998) reviewed a variety of methods developed to
numerically solve the RTE, including Zone method, Monte Carlo (MC) method, flux
method and hybrid method. Among these methods, the MC method is been accepted as
efficient and reliable. (Pareek et al., 2008, Pasquali et al., 1996). MC is a statistical
method, which is based on following the probable trajectories and fates of photons inside
the reaction zone, until their final absorption in the system or existing out of the system.
Both the trajectories and fates are decided with the help of random numbers, which are
generated by a random function through computer (Pareek et al., 2008). Consider a
photon entering into the reaction zone, it may get absorbed by the particle and its
trajectory ends there or it may be scattered by the particle to a new direction and the
trajectory continues until being absorbed by other particles or exiting out of the reaction
zone (Pareek et al., 2008). Whether absorption or scattering is determined by a random
choice based on the absorption coefficient and the scattering coefficient. To solve the
RTE with the Monte Carlo method, the optical properties of the reaction medium like
absorption coefficient scattering coefficient and the phase function should be obtained.
Further, the boundary conditions are needed.
2.6.3 Radiation source models
Radiation source model are the functions predicting the light intensity emitted from the
lamps. When the scattering effect is negligible such as in homogeneous medium, the
radiation source model can be directly used as a radiation filed model. The radiation
source model also play an important role in solving the boundary condition of RTE.
Alfano et al. (1986b) reviewed a number of radiation source models and classified them
62
into two categories: incidence models and emission models. Three incidence models have
been proposed (Pareek et al., 2008, Alfano et al., 1986b), including the radial incidence
model, the partially diffuse incidence model and the diffuse incidence model. The
development of incidence models requires an existing specific radiant energy distribution
to be assumed in the reactor space. Besides, these incidence model always need one or
more experimentally adjustable parameters, which is dependent on the size and the
configuration of the reactor (Alfano et al., 1986b). To overcome this problem, emission
models based on the lamp emission are developed and regarded as a preferred choice for
radiation source modeling. A lamp may be regarded as a point (LED), a line, a surface, or
a volume source. Depending upon the nature of the lamp, different emission models such
as line source model, surface source model, volume source model have been developed
(Pareek et al., 2008, Alfano et al., 1986b). The line source models were considered as
appropriate methods to simulate the light intensity distribution over the photocatalytic
plate when the photocatalytic reactor is equipped with tubular lamps (the geometry of
conventional mercury lamp) (Salvadó-Estivill et al., 2007).
A good radiation field model can accurately predicted the light intensity distribution
within photoreactors. Such model can be used as a tool to figure out the optimal
arrangement of light source and reactor geometry. Furthermore, radiation field model is
very important to the mathematically simulating photochemical treatment processes.
63
Chapter Three: ELECTRON TRANSFER SENSITIZED PHOTODECHLORINATION OF SURFACTANT SOLUBILIZED PCB138
3.1 Introduction
Direct and sensitized photodechlorination of polychlorinated biphenyls (PCBs) dissolved
in organic or organic/water mixtures have been the focus of various investigations (Bunce
et al., 1978, Ruzo et al., 1974, Hawari et al., 1992, Hawari et al., 1991, Miao et al., 1996,
Izadifard et al., 2008, Lin et al., 1996, Jakher et al., 2007). The low solubility of PCBs in
water necessitates the presence of an organic solvent. Surfactants also have been used to
solubilize PCBs in water (Yao et al., 2000, Bunce et al., 1978, Hawari et al., 1992,
Hawari et al., 1991, Miao et al., 1999). Using a surfactant instead of an organic solvent is
advantageous for two reasons: lower cost and minimized side reactions (Chu et al., 1998).
To-date investigations on PCBs dissolved in water by surfactants has been restricted to
direct UV photolysis, which requires high energy photons with wavelengths less than 300
nm (Chu et al., 2005, Chu et al., 1998). The application of sensitized dechlorination can
make use of photons with longer wavelengths, eventually leading to using sunlight for
dechlorination. Besides, sensitized dechlorination of PCBs can also result in a high
photodegradation rate (Dhol, 2005). To the best of our knowledge, no research is reported
on sensitized dechlorination of PCBs dissolved in water using surfactants; though there
are a few reports on sensitized reaction, where an aliphatic amine which cannot function
as a sensitizer is used to enhance the efficiency of the reaction (Chu and Kwan, 2002,
Chu and Kwan, 2003). In this case, PCB itself must be excited so that an electron transfer
from the aliphatic amine to PCBs becomes favorable.
64
This paper presents the results of an investigation on sensitized dechlorination of a PCB
congener - 2,2',3,4,4',5'-hexachlorobiphenyl (PCB 138) and a commercial mixture of
PCBs (Aroclor 1254). PCBs were dissolved in water using an anionic surfactant (sodium
dodecyl sulfate, SDS), a nonionic surfactant (polyoxyethylene (80) sorbitan monooleate,
TWEEN 80) or a cationic surfactant (cetyltrimethylammonium bromide, CTAB). The
sensitizer of choice is leuco-methylene blue (LMB), which has been reported to
effectively dechlorinate PCBs under blue and near UV irradiation (Izadifard et al.,
2010a). LMB, while being stable in an oxygen devoid environment, is produced
efficiently upon reduction of methylene blue (MB). MB can be reduced under red light
and in the presence of an aliphatic amine (such as triethylamine, TEA), or by a thermal
reaction using sodium borohydride (NaBH4). Both approaches are studied in this paper.
Air oxidation of LMB closes a ‘catalytic’ cycle that consumes the reductant.
3.2 Materials and methods
3.2.1 Materials
All PCB congeners and Aroclor 1254 were purchased from Chromatographic Specialties
Inc.; MB, CTAB, and 99.5% pure TEA were obtained from Sigma; 99.8% pure hexane
and 98% pure sodium borohydride were procured from EMD; ultra pure SDS was
obtained from MP Biomedicals and TWEEN 80 was purchased from VWR. All reagents
were used as received. Milli-Q ultrapure water was used in the experiments.
65
3.2.2 Methods
3.2.2.1 PCB 138 solubilization with surfactants
For the selected surfactant (SDS, CTAB and TWEEN 80) a stock solution with 4 g L-1
concentration was prepared. Each surfactant stock solution was prepared by dissolving 4
g of pure surfactant in 1 L milli-Q water with the aid of sonication. The PCB stock
solution was prepared by dissolving 10 mg PCB138 or 10 mg Aroclor 1254 in 10 ml
acetonitrile. For each photolysis experiment, a certain amount of PCB stock solution,
PCB 138 or Aroclor 1254, in acetonitrile (1000 mg L-1) was pipetted into a 20 ml Pyrex
glass vial and, left at room temperature in a fume hood for one day to evaporate the
acetonitrile completely. Then a known amount of surfactant stock solution was added to
the vial and the mixture was left in the sonicator for 4 hours to prepare the PCB
surfactant solution.
3.2.2.2 Photochemical reaction
The sensitizer of choice, LMB, was generated in two ways: (1) reaction between MB and
TEA under visible light irradiation and (2) thermal reaction between MB and sodium
borohydride. In each case, to the prepared PCB surfactant mixture was added a MB
solution along with either TEA or NaBH4. The final solution volume was made equal to
20 ml with milli-Q water. Uniform mixing during irradiation was achieved by placing the
sample vial on a magnetic stirring plate. The samples (contained in a Pyrex glass vial)
were irradiated in a Rayonet photoreactor equipped with either 8 or 14 cool white
fluorescent lamps. The intensity of light (Io) was measured using ferrioxalate actinometry
66
that monitored wavelengths from 254 to 500 nm, covering the region of LMB absorption
(Calvert and Pitts, 1966). Values were 5.2×1016 photon s-1 for 14 lamps and 3.0×1016
photon s-1 for 8 lamps. This convenient actinometer measures the light intensity inside the
reaction vessel and provides an order of magnitude value of intensity in the region of
LMB absorbance. Our irradiation source, the cool white lamps emit light in visible range,
up to 700 nm. MB absorption peak is at 660 nm, so it can be effectively reduced to LMB
with the cool white lamps. It is the actinometry method used here that measures
wavelengths between 254 nm to 500 nm. Since our sensitizer is LMB, whose absorption
is within this range, we used this actinometry.
3.2.2.3 Sampling, extraction and GC analysis
Exactly 0.5 ml illuminated samples were taken at different irradiation times and
transferred into a small glass vial. To each aliquot was added 2 ml of hexane, the vial was
covered by aluminum foil and was left in the wrist shaker for 1 hour to extract PCBs
from the water mixture. Around 80% extraction efficiency was obtained following this
procedure.
The extracted PCBs were analyzed using an Agilent 6890 gas chromatograph equipped
with auto-sampler and electron capture detector (ECD), using a fused silica capillary
column DB608. Helium was used as the carrier gas with a flow rate of 1.5 ml min-1 and
the temperature of the injection port was 280 oC. The GC/ECD temperature programming
was set up as follow: The initial temperature for each run was set at 80 oC, which was
67
ramped up to 180 oC at a rate of 10 oC min-1; the temperature beyond 180 oC was ramped
at a rate of 3 oC until it reached 270 oC where it was held for 15 minutes (USEPA,
1996b). For Aroclor 1254, six major peaks were chosen and a multipoint calibration
curve was made by using those six peaks.
Each experiment was conducted in duplicate and results were reported as an error bar
along with the average.
3.3 Results and Discussion
3.3.1 Selectivity of surfactants
The critical micelle concentration (CMC) in water for SDS, CTAB and TWEEN 80 were,
respectively, 2300 mg L-1 (Mandal et al., 1988), 324 mg L-1 (Paredes et al., 1984) and 15
mg L-1 (Hillgren et al., 2002). The sensitized dechlorination of PCB 138 in these
surfactant solutions are shown in Figure 3-1 and Figure 3-2. Figure 3-1 presents results
where LMB was produced under irradiation, while Figure 3-2 where it was produced
thermally. Based on our previous experiments published elsewhere (Izadifard et al.
2010b), the NaBH4 based reactions are faster than TEA based reactions. Consequently, to
ensure that we can reliably measure the concentration of PCB (above the detection limit)
within the irradiation time, a higher concentration of PCB and a lower light intensity was
applied in the NaBH4 system. We are fully aware that two different initial concentrations
were used for the two experiments, that is why we are careful in the paper not to compare
68
the results of Figure 3-1 with those of Figure 3-2. We have instead compared the
performance of different surfactants amongst themselves, presented in either Figure 3-1
or Figure 3-2. Of the three surfactants investigated, the dechlorination efficiency of PCB
138 solubilized with TWEEN 80 or with CTAB are similar, when TEA is used. However,
in the NaBH4 system the performance of CTAB is better than that of TWEEN 80. In
addition, it takes only half the irradiation intensity to achieve similar results with NaBH4
reduction.
Figure 3-1: Reductive dechlorination of PCB 138 using LMB with TEA as the
note: DL-P is larger than Ds-p due to the thickness of shielding glass plate and the gap
between lamps and the shielding glass.
125
Figure 6-3: Radiation field on a photocatalyst plate under different conditions; (a) DL-P = 0.014 m, 4 by 4 LEDs panel; (b) DL-P = 0.034 m, 4 by 4 LEDs panel; (c) DL-P = 0.054 m, 4 by 4 LEDs panel.
126
The average light intensity received by photocatalyst plate was estimated based on an
emission radiation field model developed for this design (see Chapter Five). The average
light intensities (Ia) for different situations are listed in Table 6-1. And the light intensity
distribution for different distance between the shielding glass and the photocatalyst plate
(DL-P ) with an input current of 500 mA is shown in Figure 6-3. Light intensity received
by the centre of photocatalyst plate was verified by a Silver Line UV radiometer
(M007153, Geneq Inc. Canada). When DL-P is 5.4 cm and the current is 500 mA, the
value reading from radiometer was 23 mW cm-2, which is in agreement with the model
data (Figure 6-3c).
6.2.4 Experimental set-up and sample analysis
A stock solution of 1.50 L of 20 mg L-1 pesticides or chlorophenol solution was prepared
by dissolving 30 mg of pure compound in water using ultrasonication. A peristaltic pump
(LaSalle Scientific Inc, Model: 400-205) circulated the solution containing the target
compounds between a reservoir covered with aluminum foil and the photocatalytic
reactor. The solution in the reservoir is continuously mixed with a magnetic bar.
Experiments were conducted at variable flow rate, variable light intensity, variable DL-P
and different photocatalyst configuration. All experimental conditions are described in
the captions of the figures. Prior to irradiation, the solution containing pesticides or
chlorophenols was circulated for 30 minutes to ensure that the adsorption of the
investigated compound onto the surface of photocatalyst reached equilibrium. After
different irradiation periods, a 1.0 mL sample from the reservoir was collected and
analyzed using a Varian Prostar 210 high performance liquid chromatography equipped
127
with a 325 LC UV-Vis detector (Yu et al., 2013). Variations in the obtained data in are
shown by error bars.
6.3 Result and discussion
6.3.1 Degradation of phenoxy pesticides and chlorophenols in a flow-through LED based photocatalytic reactor
To better understand the reactor performance from an energy perspective, the results are
expressed as the change of normalized concentration of parent compound versus the
energy dosage per unit volume. Energy dosage per unit volume is defined as the energy
of photons entering the reaction zone divided by the volume of sample treated (1.50 L).
The degradation of two phenoxy pesticides (2,4-D, MCPA) and two chlorophenols (4-
CP, 2,4-DCP) in the photocatalytic reactor are shown in Figure 6-4. The normalized
concentrations of parent compounds decreased as the energy dosage increased. With an
energy dosage of 25 kJ L-1, 91% of 2,4-D 85% of MCPA, 85% of 2,4-DCP and 81% of
4-CP were removed from the solution (approximately one log reduction). As reported in
Yu et al. (2013), these four compound can be efficiently photocatalytically decomposed
with slurry TiO2 in a UVA-LED batch reactor. In our immobilized photocatalytic reactor,
the TiO2 nanotubes growing on the titanium plate can also efficiently capture the UV
photons (365nm). The results show that both phenoxy pesticides and chlorophenol can be
degraded efficiently under the operational conditions shown in the caption of Figure 6-4.
128
Figure 6-4: Photodegradation of MCPA, 2,4-D, 2,4-DCP and 4-CP in a UV-LED photoreactor: flow rate = 2.03 L min-1; DL-P = 0.54 cm; Ia=17.3 mW cm-2.
6.3.2 Degradation of 2,4-D with different combination of (UV, TiO2 photocatalyst plate, H2O2 and O2) in the UV-LED photoreactor .
Photodegradation of 2,4-D under different experimental conditions is shown in Figure 6-
5. With only UV-LED irradiation, 13% of 2,4-D was removed at an energy dosage of
6.22 kJ L-1. In the UV-visible spectrum range from 250 nm-700 nm, 2,4-D has a single
peak with a maximum at around 280 nm and does not have an absorption at 365 nm
(peak wavelength for the LEDs). However, there is still a small amount of photons in the
UV-LED emission spectrum, leading to a direct photolysis of 2,4-D. The presence of
H2O2 (0.1%) in LED reactor did improve the degradation efficiency and 54% of 2,4-D
was removed from bulk solution with the same energy dosage. H2O2 has weak
absorption on the emission spectrum of LED, which may cause the photolysis of
hydrogen peroxide and generate hydroxyl radicals.
0.00
0.20
0.40
0.60
0.80
1.00
1.20
0 5 10 15 20 25 30
C/C
o
Energy dosage per volume (kJ/L)
2,4-DCP 4-CP
MCPA 2,4-D
129
Figure 6-5: Photodegradation of 2,4-D in a flow-through UV-LED photoreactor: flow rate = 2.03 L min-1; DL-P = 0.54 cm ; Ia=17.3 mW cm-2.
The reactor was considered as a photocatalytic reactor while being mounted with a TiO2
photocatalyst plate. Such photocatalytic reactor can eliminate 40 % of 2,4-D at an energy
dosage of 6.22 kJ L-1. Bubbling oxygen in this photocatalytic reactor did not significantly
improve the degradation efficiency. In the experiments, the solution in reservoir is
thoroughly mixed using a magnetic stirrer. The aeration led to the sample getting
saturated with oxygen activity near 0.2 atm. This provided enough oxygen for
photocatalytic reactions. Therefore, further addition of oxygen into the system did not
boost the photocatalytic degradation. Apparently, the presence of 0.1% H2O2 in the
photocatalytic system resulted in a better degradation efficiency. At an energy dosage of
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0 1 2 3 4 5 6 7
C/C
o
Energy dosage per volume (kJ/L)
UV only UV+H2O2 (0.1%) UV+TiO2 UV+TiO2+bubbling Oxygen UV+TiO2+ H2O2 (0.1%)
130
6.22 kJ L-1, 80% of 2,4-D is degraded. Hydrogen peroxide serves as a good electron
scavenger and accelerates the photocatalytic reaction.
6.3.3 Effect of DL-P
DL-P is a key factor for system scale-up. To study its effect on photocatalytic degradation,
experiments were conducted at three DL-P (1.4 cm, 3.4 cm and 5.4 cm). The results (see
Figure 6-6) show that the degradation of 2,4-D is relatively slow when DL-P is set to be
1.4 cm. The degradation efficiency improved as DL-P was increased to 3.4 cm, while
further increment of DL-P to 5.4 cm did not enhance the degradation efficiency. DL-P can
impact the photocatalytic degradation in two ways: (1) for the same photon energy input,
a uniform radiation field can result in more efficient distribution of activity over the
photocatalyst. In this reactor, a less uniform radiation field is obtained at shorter DL-P
(Figure 6-3); (2) for an immobilized photocatalytic reactor, the photocatalytic
degradation is limited by the mass transfer of the contaminants between the photocatalyst
surface and the bulk solution (Chen et al., 2001). At the same flow rate, the Reynolds
number decreases with DL-P, and hinders mass transfer. Therefore, from a kinetics
perspective, an optimal DL-P should make the light intensity received on the photocatalyst
plate uniform and not inhibit mass transfer.
One way to scale-up this system is to use a baffle reactor design which contains multiple
modules composed of UV-LED plate and photocatalytic plate. Each module has a
reaction zone and dead zone accommodating the electronics. The reaction zone volume
can be adjusted by changing DL-P, while the dead zone volume is limited to the electronic
131
part. The advantage of larger DL-P is that fewer modules are required for the same
reaction zone volume, and the total volume of reactor occupied is reduced. Therefore, in
this research, DL-P of 5.4 cm is superior to a DL-P of 3.4 cm.
Figure 6- 6: The effect of DL-P on 2,4-D degradation: flow rate =2.03 L min-1, Ia=17.3 mW cm-2.
6.3.4 Effect of flow rates on the photocatalytic degradation of 2,4-D.
To investigate the effect of flow rate on performance of LED photocatalytic reactor,
experiments were conducted at four different flow rates (0.72 L min-1, 1.50 L min-1, 2.03
L min-1 and 2.87 L min-1) and the results are shown in Figure 6-7. At the lowest flow rate
(0.72 L min-1), only 28 % 2,4-D removal was achieved with an energy dosage of 6.22
kJ/L. The removal percentage was improved as the flow rate increased and 40% of 2,4-D