Oxidant Generation on Photolysis of Silver Chloride Suspensions: Implications to Organic Contaminant

Oxidant generation on photolysis of silver chloride suspensions: implications to organic contaminant

degradation

Tian Ma

Supervisor: Scientia Professor T. David Waite

A thesis in fulfilment of the requirements for the degree of

Doctor of Philosophy

School of Civil and Environmental Engineering

Faculty of Engineering

March, 2014

ORIGINALITY STATEMENT

‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’

Signed ……………………………………………......

Date ………………………......

‘I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only).

I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'

Signed ……………………………………………......

Date ………………………......

AUTHENTICITY STATEMENT

‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’

Signed ……………………………………………......

Date ………………………......

Abstract

The photochemistry of silver chloride and other silver halides has attracted intensive attention among researchers for the past twenty years, in part at least, because it is recognized as a promising technology for degradation of organic contaminants in waters and wastewaters.

Silver chloride displays semiconducting properties and absorbs incoming photons in the UV- visible region with resultant formation of several highly reactive species and, in many instances, the degradation of organics, if organics are present. The fact that the actual reaction mechanism remains uncertain hinders its practical application.

This thesis describes work investigating the photolysis of silver chloride under various solution conditions relevant to real life environmental water systems and describes the photocatalytic decay of formic acid and rhodamine B induced by silver chloride. The impact of solution matrix on the formation, size, aggregation behaviour and stability of silver chloride was also investigated by employing advanced analytical technologies including

Dynamic Light Scattering (DLS), SEM (STEM) and X-ray Diffraction (XRD), as well as spectrophotometry and spectrofluorimetry. These techniques allow determination of the concentration of species at micro- or nano- molar levels and particle sizes down to the nanometer size range. Kinetic modelling of the mechanism on photolytic activity of silver chloride and its photocatalytic decomposition of formic acid was done with Matlab-Acuchem.

The concentration change of formic acid was detected using a radioisotope-labelling technique coupled with scintillation counting.

Results show that reactive species superoxide, hydrogen peroxide and singlet oxygen are generated through irradiation of silver chloride colloid, as well as the generation of free chlorine and silver nanoparticles. The relative importance of these various species was influenced by the initial concentrations of Ag(I), Cl–,and bicarbonate as well as by the pH and the presence of oxygen. Simulation of laboratory results by kinetic modelling of the processes indicated that oxidation of Cl– by holes resulted in formation of Cl0 while scavenging of

0 •– photo-generated electrons by Ag(I) and O2 led to the formation of Ag and O2 respectively.

Results obtained indicate that holes are responsible for the oxidation of formic acid while for rhodamine B, reaction with holes and chlorine atoms was found to be responsible for its decay.

Table of Contents Page

Abstract ...... A

Table of Contents ...... I

ORIGINALITY STATEMENT ...... i

AUTHENTICITY STATEMENT ...... ii

Acknowledgements ...... iii

List of Figures ...... iv

List of Tables ...... xviii

Chapter 1: Introduction ...... 1

1.1 Problem statement and thesis purpose ...... 1

1.2 Photochemistry of AgCl(s) ...... 2

Properties of silver chloride ...... 3

Synthesis and characterization of silver chloride ...... 4

Mechanism of photocatalytic activities of Ag/AgX ...... 5

Mechanism of degradation of organics ...... 5

1.3 Silver nanoparticle ...... 7

Properties of Silver Nanoparticles ...... 7

Optical Properties...... 7

1.4 Key objectives and outcomes ...... 8

Chapter 2: Characterisation of AgCl(s) Particle ...... 9

2.1 Introduction ...... 9

2.2 Experimental Details ...... 20

2.2.1 General ...... 20

2.2.2 Particle Size and Zeta-Potential Measurement ...... 21

2.2.3 Sample Preparation for SEM/STEM Analysis ...... 22

2.2.4 Electronic Spectra of AgCl(s) Colloid ...... 23

2.2.5 XRD measurement ...... 23

2.3 Results and discussion ...... 23

2.3.1 Formation of AgCl(s) Colloid ...... 23

2.3.2 Aggregation of Silver Chloride in the Presence of Sodium Chloride ...... 25

2.3.3 Zeta Potential ...... 32

2.3.4 Silver Chloride Particle Surface Topography and Composition by SEM ...... 33

2.3.4.1 EDS Results from SEM Analysis ...... 34

2.3.4.2 SEM of AgCl(s) particles kept in the dark...... 36

2.3.4.3 SEM of AgCl(s) particles irradiated without organics ...... 38

2.3.4.4 SEM of AgCl(s) particles irradiated in the presence of rhodamine B ...... 38

2.3.5 Silver Chloride Particle Surface Topography and Composition by STEM ...... 40

2.3.5.1. STEM of irradiated AgCl(s) without organics...... 41

2.3.5.2. STEM of irradiated AgCl(s) with rhodamine B ...... 43

2.3.5.3. STEM of irradiated AgCl(s) with formate ...... 45

2.3.6. X-ray Diffraction Analysis ...... 47

2.3.7 Electronic Absorption Spectra ...... 48

2.4 Conclusion and Implications...... 57

Chapter 3. Generation of free chlorine and ROS on irradiation of AgCl(s) suspensions ...... 59

3.1 Introduction ...... 59

3.2 Experimental Methods ...... 60

3.2.1 Preparation of reagents ...... 60

3.2.2 Photochemical experimental setup ...... 61

3.2.3 Free chlorine measurement ...... 63

3.2.4 Hydrogen peroxide measurement ...... 64

III

3.2.5 Singlet oxygen measurement ...... 65

3.2.6 Hydroxyl radical measurement ...... 66

3.2.7 DO measurement ...... 68

3.2.8 AgNP measurement ...... 68

3.3 Results and Discussion ...... 70

3.3.1 Generation of AgNP ...... 70

3.3.2 Generation of free chlorine on photolysis of AgCl(s) suspension ...... 73

3.3.2.1 Role of oxygen in free chlorine generation ...... 75

3.3.2.2 Role of holes or HO• in free chlorine generation ...... 76

3.3.2.3 Effect of pH on free chlorine generation ...... 78

3.3.3 Decay of free chlorine in dark ...... 80

3.3.4 Generation of hydrogen peroxide on photolysis of AgCl(s) suspension ...... 82

3.3.5 Generation of singlet oxygen on photolysis of AgCl(s) suspension ...... 82

3.3.6 Generation of hydroxyl radical on irradiation of AgCl(s) ...... 85

3.4 Discussion ...... 86

3.4.1 Mechanism of photo-generation of ROS and free chlorine ...... 86

3.4.2 Kinetic model for generation of ROS and free chlorine on irradiation of AgCl(s) 87

3.4.2.1 Photon absorption and instantaneous establishment of steady-state concentration of holes and electrons ...... 89

3.4.2.2 Generation of free chlorine ...... 89

3.4.2.3 Generation of AgNPs and superoxide ...... 90

3.4.2.4 Decay of OCl- ...... 90

3.4.2.5 Aggregation of AgNPs ...... 91

3.5 Conclusions and Implications ...... 92

Chapter 4: Kinetics and mechanism of free chlorine decay in the presence of silver nanoparticles ...... 93

4.1 Introduction ...... 93

4.2 Experimental Methods ...... 93

4.2.1 Reagent Preparation ...... 93

4.2.2 AgNPs measurement ...... 94

4.2.3 Free Chlorine measurement ...... 96

4.3 Results and Discussion ...... 96

4.3.1 Decay of AgNPs in the absence of OCl– at pH 8 ...... 96

4.3.2 Decay of free chlorine in the absence of AgNPs at pH 8 ...... 98

4.3.3 Decay of AgNPs in the presence of OCl- at pH 8 ...... 99

4.3.4 Decay of free chlorine in the presence of AgNPs at pH 8.0 ...... 101

4.3.5 Mechanism of AgNP and OCl− reaction at pH 8 ...... 103

4.3.6 AgNP decay in absence of HOCl− at pH 4.0 ...... 103

4.3.7 AgNPs decay in presence of HOCl at pH 4 ...... 104

4.4 Conclusion ...... 105

Chapter 5: Degradation of formic acid in irradiated AgCl(s) suspensions ...... 106

5.1 Introduction ...... 106

5.2 Experimental Details ...... 110

5.2.1 Reagent preparation ...... 110

5.2.2 Photochemical experimental setup ...... 110

5.2.3 Formic acid measurement ...... 111

5.2.3.1: Effect of free chlorine and hydrogen peroxide addition ...... 112

5.2.3.2: Role of singlet oxygen ...... 112

5.2.3.3: Effect of dioxygen removal ...... 112

5.2.3.4: Role of superoxide ...... 113

5.2.3.5: Role of oxidizing intermediate formed on AgNP and H2O2 reaction ...... 113

5.3 Results and Discussion ...... 113

5.3.1 Formic acid degradation ...... 113

5.3.1.1 Role of hydrogen peroxide...... 117

5.3.1.2 Role of singlet oxygen ...... 117

5.3.1.3 Role of oxidizing intermediate formed on AgNP-H2O2 reaction ...... 119

5.3.1.4 Role of superoxide ...... 120

5.3.1.5 Role of dioxygen ...... 121

5.3.2 Mechanism of formic acid degradation in the presence of irradiated AgCl(s) ..... 122

5.3.2.1 Degradation of HCOOH by photo-generated holes ...... 122

5.3.2.2 Degradation of HCOOH by Ag(I) ...... 123

5.3.2.3 Degradation of HCOOH by oxidizing intermediate formed on AgNP-H2O2

reaction ...... 123

5.4 Conclusion ...... 124

Chapter 6: Degradation of Rhodamine B during Visible Light Irradiation of AgCl(s)

Suspensions ...... 125

6.1 Introduction ...... 125

6.2 Experimental Methods ...... 129

6.2.1 Reagents and equipment ...... 129

6.2.2 Experimental method ...... 129

VII

6.2.3 Data analysis methodology ...... 130

6.3 Results and Discussion ...... 131

6.3.1 Background Degradation of RhB...... 131

6.3.2 The effect of initial silver(I) concentration on RhB degradation ...... 133

6.3.3 The effect of initial chloride concentration ...... 137

6.3.4 Impact of rhodamine B concentration ...... 141

6.3.5 The effect of dissolved oxygen concentration ...... 144

6.3.6 The effect of bicarbonate concentration ...... 147

6.3.7 The effect of solution pH on rhodamine B degradation ...... 149

6.3.8 The effect of spiked hypochlorite ...... 151

6.4 Implications and Conclusions ...... 157

Chapter 7: Conclusions ...... 161

7.1 Main findings ...... 161

7.1.1 Characteristics of AgCl(s) Particles ...... 161

7.1.2 Generation of free chlorine and ROS on irradiation of AgCl(s) suspensions...... 162

7.1.3 Free chlorine decay in the presence of silver nanoparticles ...... 163

7.1.4 Photodegradation of formic acid ...... 163

VIII

7.1.5 Photodecomposition of rhodamine B...... 164

7.2 Limitations and future work...... 165

References ...... 168

Acknowledgements

First and foremost, I wish to thank my supervisor and friend Scientia Professor T. David

Waite for introducing me into the world of research, constantly motivating and inspiring me to achieve more with hard work, supporting the work of myself and fellow students through discussion, investing money in developing laboratory technologies and facilities. I cannot finish this thesis without his tireless input and encouragement for pursuit of knowledge for the good of environment. I gratefully thank my co-supervisor Dr. Shikha Garg for her strong support throughout the entire period of the research. Thanks to Dr. Christopher J. Miller for his invaluable help both in laboratory work and in analysing data. Thank our Mark

Wainwright Analytical Centre at UNSW for providing XRD, SEM and STEM instrument for sample analysis. I also want specifically thank Dr. Yu Wang from XRD Lab, Senior

Technical Officer Katie Levick and Dr. Karen Privat from Electron Microscope unit for examining samples with me.

Also I wish to acknowledge the contribution of Dr. Shikha Garg and Dr. Chris J. Miller to specific aspects of this thesis. The manuscript of Chapter 3 and 5 were co-authored with Dr

Shikha Garg, who contributed to the analysis of data. Chapter 2 and 6 were co-authored with

Dr. Chris J. Miller, with whom the data and ideas in the chapter were jointly analysed and developed, who also contributed to the writing.

I dedicate this thesis to my family for whom I am doing this.

iii

List of Figures

Chapter 2

Figure 2.1. Measured absorbance spectra of spherical AgNPs mounted on quartz posts in air.

Spectra are adapted from Figure 2 of Russell et al. (1987).

Figure 2.2: Measured absorbance spectra of AgNPs dispersed in aqueous solution. Spectra are adapted from Figure 3.1 of Kinnan (2008).

Figure 2.3: The impact of immobilizing particles upon a substrate such that they multiple particles can coherently couple; immobilized particles are shown as solid lines, with the analogous particles dispersed in aqueous solution shown as dashed lines. Spectra are adapted from Figures 3.1 and 3.2 in Kinnan (2008).

Figure 2.4: Influence of number of particles on the absorbance of a planar quadratic array of adjacent 40 nm diameter particles, including an infinite sheet of silver atoms. Adapted from

Figure 8 of Quinten and Kreibig (1993)

Figure 2.5: Influence of aggregate structure upon absorbance of 40 nm AgNPs. Adapted from

Figure 4 of Quinten and Kreibig (1993)

Figure 2.6: Changes in the absorbance spectra of atomic Ag0 entities as they undergo reaction to give AgNPs. Spectra are those reported by Tausch-Treml et al.(1978)

Figure 2.7: AgCl(s) equilibrium concentration as a function of initial chloride ion

– concentration, [Ag(I)]0 = 100 µM and [Cl ]0 = 20 – 200 mM.

Figure 2.8: AgCl(s) concentration as a function of [Ag(I)]0 (10 – 200 µM), initial

– – concentrations: [Cl ]0 = 100 mM , [HCO3 ]0 = 2 mM .

Figure 2.9: The AgCl(s) particle size as a function of time: [Ag(I)] = 100 µM (solid diamond), 50 µM (empty triangle), 25 µM (empty square) and 10 µM (solid triangle).

– – Common solution conditions: [Cl ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0.

Figure 2.10: (a) The size of silver chloride as a function of time. Initial concentrations of Cl–

= 200 mM (empty triangle), 100 mM (solid diamond), 90 mM (empty square), 70 mM (solid triangle) and 50 mM (empty diamond), 25 mM (solid square). Common solution conditions :

– [Ag(I)]0 = 100 µM, [HCO3 ]0 = 2 mM, pH 8.0. (b) shows the same data as (a) but on an expanded y-axis scale.

Figure 2.11: The plot of ddz/dt as a function of NaCl concentration. [Ag(I)]0 = 100 µM,

– – – [HCO3 ]0 = 2 mM (pH 8.0, triangles), [OH ]0 = 1 mM (pH 11.0, circles), [HNO3 ]0 = 0.1 mM

(pH 4.0, squares). The dashed lines respresent an estimate of the diffusion limited rate, obtained by the weighted average of the 100 mM and 200 mM NaCl data points.

– Figure 2.12: Growth rate of dZ as a function of silver plus concentration. [Cl ]0 = 100 mM ,

– − [HCO3 ]0 = 2 mM (pH 8.0, solid diamond and solid line), [OH ]0 = 1 mM (pH 11.0, empty

– square and dash line), [HNO3 ]0 = 0.1 mM (pH 4.0, empty triangle and long dash dot).

Figure 2.13: (a) Aggregation kinetics of AgCl(s) with irradiated 1.0 µM formate (empty diamond), irradiated 0.21 µM Rhodamine B, irradiated without added organic (solid square), without irradiation (empty square). Light was turned on after 35 minutes Ag(I) and Cl– ions

– reaction in the dark, time zero was the time point of spiking Ag(I) inot Cl solution. [Ag(I)]0

– – = 100 µM, [Cl ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0; (b) a close-up of size measurement between 30 to 60 min.

Figure 2.14: The ζ potential values of silver chloride colloid as a function of reaction time

– – between Ag(I) and Cl ion. Solution conditions: [Cl ]0 = 200 mM (solid square) , 100 mM

(empty diamond) , 50 mM (solid triangle) and 25 mM (empty square); [Ag(I)]0 = 100 µM , pH 8.0.

Figure 2.15: The ζ potential values of silver chloride colloid as a function of reaction time

– between Ag(I) and Cl ion. Solution conditions : [Ag(I)]0 = 25 µM (empty triangle) , 50 µM

– (solid diamond) , 100 µM (empty square) and 200 µM (solid triangle); [Cl ]0 = 100 mM , pH

8.0.

Figure 2.16: The EDS spectrum of AgCl(s) particles when imaged using SEM. For each condition two separate particles have been imaged. All spectra have been normalized to the

Ag Lα1 peak to allow for simpler comparison. The “Ag Component” trace is the calculated component for all Ag emissions using data from Thompson, Lindau et al. (2009)

Figure 2.17: The SEM image of AgCl(s) particles in the dark.

Figure 2.18: The typical silver chloride crystal structure (WebElements 2013) showing a fcc crystal structure.

Figure 2.19: The SEM image of AgCl(s) particles after being exposed to simulated sunlight irradiation for 30 minutes. Magnification ×20k (a) and ×10k (b).

Figure 2.20: AgCl(s) particles after being exposed to simulated sun light with 0.21 µM rhodamine B for 30 minutes, magnification ×20k (a) and ×10k (b)2.3.4.5 SEM of AgCl(s) particles irradiated in the presence of formate.

Figure 2.21: SEM image showing AgCl(s) crystals after being irradiated with 1.0 µM initial sodium formate, magnification ×30k (a) and ×19k (b).

Figure 2.22: STEM images of AgCl(s) after irradiation with visible light including secondary electron images before (a) and after (b) element mapping, a TEM image (c), element maps for Ag, Cl and Na (panels d, e, and f respectively) and a composite element map (g). These figures demonstrate no areas of obvious Ag(0) formation, but suggest the particle to primarily consist of AgCl with co-precipitated NaCl (or perhaps simply just adhesion of NaCl).

Figure 2.23: EDX spectra scaled to Ag Lβ1 (a) or Cl Kα1 (b) emission peaks. Point numbers correspond to those in Figure 2.23 a, b.

Figure 2.24: TEM images of AgCl(s) after irradiation in the presence of rhodamine B, a) element map, b) secondary electron image showing points analysed by EDX, c) TEM image, with broken white lines showing region analyzed by EDX and element mapping and d), e), f),

Ag, Cl and Na element maps, respectively

Figure 2.25: EDX images corresponding to the points shown above in Figure 2.25 b. Vertical lines show the expected emission lines for Ag, Cl and Na (for the transition labelled)

Figure 2.26: TEM images of AgCl after irradiation in the presence of formate, a) element map, b) secondary electron image showing points analysed by EDX, c) TEM image, d), e), f),

Ag, Cl and Na element maps, respectively

vii

Figure 2.27: EDX images corresponding to points shown above in figure 2.27. Vertical lines show the expected emission lines for Ag, Cl and Na (for the transition labelled)

Figure 2.28: The XRD graph of AgCl grown and collected in the dark.

Figure 2.29: The XRD graph of AgCl grown in the dark but irradiated 60 minutes’ with visible light afterwards.

Figure 2.30: The absorbance of AgCl(s) after aging 35 minutes with initial silver plus

– concentrations are : 200 µM (solid line), 100 µM (round dot) and 50 µM (dash). [Cl ]0 = 100

– mM, [HCO3 ]0 = 2 mM, pH 8.0.

Figure 2.31: The absorbance of AgCl(s) after aging 35 minutes with initial sodium chloride concentrations are : 25 mM (dash line), 50 mM (long dash dot), 100 mM (round dot) and 200 mM (solid line). [Ag(I)]0 = 100 µM, [NaHCO3]0 = 2 mM, pH8.0.

Figure 2.32: Measured absorbance spectra of 50 µM Ag(I) and 25 mM NaCl allowed to age in the dark for 20, 25, 30 and 35 minutes. Panel (a) shows this data on a natural scale, with panel (b) showing this same data on a log-log plot. The sloped gridlines in panel b are lines proportional to λ−4 and give the expected slopes for Rayleigh scattering.

Figure 2.33: The change in absorption spectrum upon irradiation of a 100 µM Ag(I), 1 mM

NaCl, 2 mM NaHCO3, pH 8 solution which was first allowed to age in the dark for 35 minutes. Spectra are recorded every five minutes, with reaction progressing in the direction of the arrow in panel (a). Panel (b) shows the same data on a log-log plot, with the sloped gridlines proportional to λ−4 showing the expected slopes for Rayleigh scattering. Panel (c) show the difference spectra and are determined by subtracting the 35 minute dark-aged spectrum from the subsequent irradiated spectra.

viii

Figure 2.34: absorbance spectra of 100 µM Ag(I), 100 mM NaCl, 2 mM NaHCO3, pH 8 system allwed to age in the dark. Panel (a) shows spectra on a natural scale with (b) showing the same data on a log-log plot with the sloping gridlines proportional to λ−4 to show the expected slopes for Rayleigh scattering.

Chapter 3

Figure 3.1: Experimental setup for irradiation of AgCl(s)

Figure 3.2: Incident spectral irradiance from the light source as function of wavelength.

Figure 3.3: The conversion of DPD to its radical cation DPD•+ .

Figure 3.4: Chemiluminiscence signal after adding 0.5µM 5–HO–Phth into irradiated AgCl(s) sol for 60 minutes.

Figure 3.5: Chemiluminescent signal from irradiated AgCl(s) sample containing (a) 1 µM and

(b) 50 nM 5–HO–Phth added before (open bar) and after centrifugation (closed bar).

Figure 3.6: Change in absorbance of solution containing AgCl(s) in presence of (a) 1 mM and

(b) 10 mM chloride on irradiation.

Figure 3.7: (a) Generation of free chlorine on irradiation of solution containing 100 mM chloride and 200 µM (squares), 100 µM (triangles) and 50 µM (circles) Ag(I) at pH 8.0; (b)

Generation of free chlorine on irradiation of solution containing 100 µM Ag(I) and 200 mM

(squares), 100 mM (diamonds), 50 mM (triangles) and 25 mM (circles) Cl– . Data represent

average of duplicate measurements. Shaded region represents the measured free chlorine concentration after the lamp was extinguished.

Figure 3.8: Concentration of free chlorine generated after 60 minutes of irradiation of air- saturated (open bar) and partially deoxygenated (closed bar) solution containing 500 µM

Ag(I) and 100 mM Cl– . Data represents average of duplicate measurements.

Figure 3.9: Generation of free chlorine on irradiation of solution containing 100 mM chloride and 100 µM Ag(I) in presence of 2 mM (squares), 4 mM (triangles) and 10 mM (circles)

NaHCO3 at pH 8.0. Data represent average of duplicate measurements. Shaded region represents the measured free chlorine concentration after the lamp was extinguished. The pH was controlled at 8.0 by addition of HNO3 and/or NaOH.

Figure 3.10: Absorbance of aged AgCl(s) formed on addition of 100 µM Ag(I) to solution

- – containing 100 mM Cl at pH 8.0 in presence of varying HCO3 concentration.

Figure 3.11: Generation of free chlorine on irradiation of solution containing 100 mM chloride and 100 µM Ag(I) at pH 4 (closed circles), 8 (open diamonds) 9 (open squares), 10

(open triangles) and 11 (open circles). Data represents average of duplicate measurements.

Ag(I) and 100 mM Cl-(open circles), 100 µM Ag(I) and 200 mM Cl-(open diamonds) and

100 µM Ag(I) and 50 mM Cl-(open triangles) at pH 8.

Figure 3.13: Generation of SOSG epoxide on irradiation of 3 mL of solution containing 100 mM Cl– , 100 µM Ag(I) and 8 µM SOSG in absence (open bar) and presence of catalase

(closed bar) or glycine (grey bar) for 10 minutes at pH 8. Data represents average of duplicate measurements.

Figure 3.14: Concentration of SOSG-EP generated on irradiation of 3 mL of solution containing Ag(I), 100 mM Cl- and 8 µM SOSG at pH 8.0. Data represents average and error bars are standard deviations from triplicate measurements.

Figure 3.15: Chemiluminiscence signal corresponding to Phth–OH on irradiation of solution containing 100 µM Ag(I) , 100 mM Cl- and 0.5 mM PhTh at pH 8.0.

Figure 3.16: Reaction mechanism for generation of ROS and free chlorine on irradiation of

AgCl(s).

Chapter 4

Figure 4.1: Measured absorbance of 8 µM AgNPs in absence (a) and presence (b) of 8 µM

OCl- at pH 8.

Figure 4.2: Oxidative-dissolution of AgNPs in absence of OCl− at pH 8. Initial AgNPs concentration used here are 24 µM (open squares), 16 µM (open triangles), 8 µM (open circles), 4 µM (open diamonds) and 2 µM (closed squares).

Figure 4.3: Decay in OCl− concentration in absence of AgNPs at pH 8. Initial OCl- concentration used here are 16 µM(open circles), 8 µM(open triangles), 4 µM(open diamonds) and 2 µM(open squares). Panel (a) shows the full decay, panel (b) shows the change in OCl− concentration.

Figure 4.4: Decay of AgNPs in presence of 8 µM free chlorine at pH 8. Initial AgNP concentrations used here are 24 µM (open squares), 16 µM (open triangles), 8 µM (open circles), 4 µM (open diamonds) and 2 µM (closed squares).

Figure 4.5: Decay of 8 µM AgNP in presence of 2 µM (open diamond), 4µM (open squares) and 16 µM (open triangle) free chlorine at pH 8.0.

Figure 4.6: Decay of free chlorine in presence of 0.8 µM AgNPs at pH 8.0. Initial OCl− concentration used here are 16 µM (open triangles), 8 µM (open squares), 4 µM (open diamonds) and 2 µM (open circles).

Figure 4.7: Decrease in OCl− concentration in presence of 0.2 µM (open squares), 0.4 µM

(open circles), 0.8 µM (open diamonds), 1.6 µM (open triangles) and 2.4 µM (closed squares)

AgNPs at pH 8.0.

Figure 4.8: Oxidative-dissolution of AgNPs in absence of OCl− at pH 4. Initial AgNPs concentration used here are 28 µM (open diamonds), 24 µM(open squares), 16 µM(open triangles) and 8 µM(open circles).

Chapter 5

Figure 5.1: The spectrum of incident irradiance for this study.

− Figure 5.2: Degradation of 1µM formic acid in irradiated 2 mM HCO3 solution containing

100 mM Cl− and 0 (open squares), 25 µM (open triangle), 50 µM (open diamond), 100 µM

(open circle), 200 µM (closed squares) and 400 µM (closed triangles) Ag(I) at pH 8. Data represents averages of duplicate measurements and lines represent model values.

xii

− Figure 5.3: Degradation of 1µM formic acid in irradiated 2 mM HCO3 solution containing

100 µM Ag(I) and 0 (open squares), 200mM (open triangle), 100mM (open diamond), 50 mM (closed circle), 25 mM (open circles), 10 mM (closed triangles) and 2 mM (closed squares) Cl−at pH 8. Data represent averages of duplicate measurements and lines represent model values.

− Figure 5.4: Degradation of 1µM formic acid in 2 mM HCO3 solution containing 100 µM

Ag(I) and 100 mM Cl− at pH 8 in absence (open triangles) and presence (open squares) of 25

µM OCl−. Data represents the average of duplicate measurements.

− Figure 5.5: Degradation of 1µM formic acid in 2 mM HCO3 solution containing 100 µM

Ag(I) and 100 mM Cl– at pH 8 in the absence (open triangles) and presence (open squares) of

10 mM H2O2. Data represent averages of duplicate measurements.

− Figure 5.6: Degradation of 1µM formic acid in 2 mM HCO3 solution containing 2 µM H2O2 and 1 µM OCl– at pH 8. Data represent the average of duplicate measurements.

Figure 5.7: Concentration of HCOOH remaining after 110 minutes in the presence of 1.0 µM

– OCl and 2.0 µM H2O2 in aqueous (open bar) and deuterium oxide (closed bar) solutions at pH 8.

− Figure 5.8: Degradation of 1µM formic acid in 2 mM HCO3 solution containing 10 mM

H2O2 and 20 µM AgNPs at pH 8. Data represent averages of duplicate measurements.

Figure 5.9: Degradation of 1µM formic acid in irradiated AgCl(s) solution in the absence

(solid triangles) and presence (open triangles) of 20 kU.L−1 SOD at pH 8.0. Data represent averages of duplicate measurements.

xiii

Figure 5.10: Degradation of 1µM formic acid in irradiated air-saturated (open squares) and partially deoxygenated (closed squares) solutions containing AgCl(s) at pH 8.0. Data represent averages of duplicate measurements and lines represent model values.

Chapter 6

Figure 6.1 Structure of rhodamine B.

Figure 6.2: Diagram showing the chemical structures of RhB (R1 – R4 = Et), TER (R1 – R3 =

Et, R4 = H), DER (R1, R4 = Et, R2, R3 = H), MER (R1 = Et, R2 – R4 = H) and Rh (R1 – R4 =

H).

Figure 6.3: Photo–degradation of RhB with simulated sunlight in the absence of AgCl(s) (a) and a close–up of low concentration products: Rh, MER, DER and TER (b). Initial solution

– – conditions : [Rhodamine B]0 = 2 µM, [Cl ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0.

Figure 6.4: Concentration profiles during degradation of RhB (a) and a close–up of the low concentration TER, DER, MER and Rh species (b). Initial solution conditions: [Ag(I)]0 = 100

– – µM, [Cl ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0.

Figure 6.5: Concentration profiles of TER (a) and DER (b) as a function of initial Ag(I) concentration.

Figure 6.6: Semi-logarithmic plot of the concentration of RhB as a function of initial Ag(I) concentration (a), with the pseudo-first order rate constant obtained from these plots, as well as the initial RhB degradation rate, shown in (b). Data is shown for 2.0 µM RhB,100 mM Cl–,

– 2 mM HCO3 , pH 8.0.

xiv

Figure 6.7: Rhodamine B initial degradation rate (a) and rate constant (b) as a function of initial concentration of chloride ion concentration. [Rhodamine B]0 = 2 µM, [Ag(I)]0 = 100

– µM, [HCO3 ]0 = 2 mM, pH 8.0.

Figure 6.8: The degradation profile of rhodamine B as a function of time (a) and a close–up of low concentration species (TER, DER, MER and Rh) (b). The light was turned off at 30

– minute point after sampling. [Rhodamine B]0 = 2 µM, [Ag(I)]0 = 100 µM, [Cl ]0 = 50 mM,

– [HCO3 ]0 = 2 mM, pH 8.0 .

Figure 6.9: The photo-degradation of rhodamine B as a function of time (a) and a close–up of low concentration species (TER, DER, MER and Rh). Initial solution conditions:

– – [Rhodamine B]0 = 1.0 µM, [Ag(I)]0 = 100 µM, [Cl ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0 .

Figure 6.10: Pseudo-first order degradation rate constant for rhodamine B as a function of initial rhodamine B concentration (a) and the concentration profile of the total degraded chromophoric rhodamine (b). The dashed line in panel B indicates the initial rhodamine B

– concentration for the 0.5 µM RhB data. Initial solution conditions:[Ag(I)]0 = 100 µM, [Cl ]0

– = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0.

Figure 6.11: TER concentration profile as a function of initial rhodamine B concentration

Figure 6.12: Rhodamine B (RhB) and total rhodamine (∑Rh) concentrations before and after irradiation either under an ambient atmosphere or whilst sparging with Ar (+ 300 ppm CO2):

(1) the mixture of RhB and AgCl(s) was pre–sparged and then irradiated; (2) the mixture was sparged but kept in the dark; (3) the mixture was irradiated but without pre–sparging; (4) the

– mixture was kept in the dark without sparging. [RhB]0 = 2.0 µM, [Ag(I)]0 = 100 µM, [Cl ]0 =

– 100 mM, [HCO3 ]0 = 2 mM, pH 8.0, irradiation time 30 minutes. Irradiations were performed in a 3 mL quartze fluorescence cuvette.

Figure 6.13: Relative concentration of rhodamine B species (sum Rh = 1): (1) the RhB and

AgCl(s) mixture was sparged and irradiated (solid blue bar); (2) the mixture was irradiated for the same period of time but did not undergo sparging (empty red bar).

Figure 6.14: Rhodamine B degradation rate (a) and rate constant (b) as a function of initial concentration of bicarbonate. Data is shown for 2.0 µM rhodamine B, 100 µM Ag(I), 100 mM Cl– at pH 8.0. AgCl(s) was aged for 35 minutes in the dark before irradiation was commenced.

Figure 6.15: Impact of solution pH upon the concentration profiles of rhodamine B (a) and total chromophoric rhodamine (b).

Figure 6.16: Impact of solution pH upon the concentration profiles of TER (a) and DER (b).

Figure 6.17: Chlorination of fluorescein by HOCl (from Hurst, Albrich et al. (1984)) as well as the mechanism considered likely for rhodamine B under the conditions of this study.

Figure 6.18: The concentration of intermediate products (ΣRh, RhB, TER, DER, MER and

Rh) as a function of irradiation time, the product of chloro–rhodamine concentration and its molar absorptivity (εC(ClRh)) is on the right–hand–side of secondary y–axis (a); a close–up of the low concentration DER, Rh, TER and MER, and εC of chloro–rhodamine (ClRh) (b).

Figure 6.19: The concentration of intermediate products (ΣRh, RhB, TER, DER, MER and

Rh) as a function of time, the product of chloro–rhodamine concentration and its molar absorptivity (εC(ClRh)) is on the right–hand–side of secondary y–axis (a); a close–up of the

xvi

low concentration DER, Rh, TER, and MER, εC of chloro–rhodamine (ClRh) (b).

Degradation happened in the dark.

Figure 6.20: Comparison of RhB degradation products (∑Rh, RhB and ClRh) as a function of time. The degradation was with simulated sunlight irradiation (empty marker and dash line) as contrast to that in the dark (solid marker and line).

Figure 6.21: The intensity of light (primary y–axis), absorbance of AgCl(s) and RhB

(secondary y–axis) as a function of wavelength.

xvii

List of Tables

Chapter 2

Table 2.1: Zeta Potential Values and Stability of Colloid

Table 2.2: The Speciation of Silver by Percentage.

Table 2.3: Quantitative XRD Analysis Results

Chapter 3

Table 3.1: Kinetic model to predict generation of ROS and OCl– on irradiation of AgCl(s)

Chapter 5

Table 5.1: Reactions showing degradation of HCOOH in presence of irradiated AgCl(s)

xviii

Chapter 1: Introduction

1.1 Problem statement and thesis purpose

Elimination of organic contamination from wastewaters is a major area in water treatment.

Some organic pollutants, such as azo dyes, have proven difficult to degrade with traditional secondary treatment methods (Shaul, Holdsworth et al. 1991, Chu and Chen 2002) and could do tremendous harm to the health of human beings (as well as plants and animals) if water containing such dyes, and/or their precursors and degradation products (some of which are suspected as carcinogenic) end up in natural waterways. On the other hand, for the past twenty years AgX(X = Cl, Br) related photocatalysts have drawn intensive attention for their effectiveness to decompose many water-bearing organic pollutants. Some preliminary research has demonstrated high removal efficiency for organic pollutants using AgCl related photocatalytic process. For instance,

 Chen, Deliang et al. (Chen, Liu et al. 2014) investigated the degradation of rhodamine

B (RhB), methyl orange (MO) and methyl blue (MB) with their Ag/AgCl

nanocrystals and found a high photodegradation rate for the above organic dyes, with

the degradation rate for RhB ~54 times greater than that using TiO2 nanocrystals.

 Hu, P. et al. (Hu, Hu et al. 2014) fabricated AgCl@Ag on agar gel support and

examined its ability to photodegrade MO. They suggested that its high degradation

efficiency was due to the localized surface plasmon resonance of the silver

nanoparticles.

 Lin, C.P. et al. (Liu, Lin et al. 2014) synthesized the photocatalyst AgI/AgCl/H2WO4

and evaluated its effect on MO degradation with visible light (lambda > 400 nm).

Compared to the photodecay results with AgI, AgCl, H2WO4 and AgCl/H2WO4 , they

tentatively reasoned that the increased degradation efficiency with AgI/AgCl/H2WO4

was due to a significant reduction in rate of electron-hole pair recombination.

Although the photocatalytic oxidation of organic pollutants under visible light illumination has shown great promise, no convincing mechanism for any reaction in the photochemical process has been established. Some tentative schemes have been suggested and various conceptual models presented (Zhang, Fan et al. 2011, Zhou, Cheng et al. 2011, Zhu, Chen et al. 2012, Zhu, Wang et al. 2013, Min, He et al. 2014, Sun, Zhang et al. 2014, Wang, Ming et al. 2014). There are some general trends in the conclusions drawn so far: (1) the existence of

Ag nanoparticles as part of the photocatalyst could increase the degradation strength due to its SPR characteristic; (2) the photogeneration of electron–hole pairs is essential; (3) reactive

•– • 1 oxygen species (O2 , H2O2, HO and O2) are generated.

This thesis is an attempt to clarify the photocatalytic reaction occurring in AgCl(s) suspensions. In so doing we hope to demonstrate the relationship between the species and especially their relative importance with regard to oxidizing organics. The motivation behind this research was to determine the solution conditions which would maximize the photo– decomposition potential of AgCl(s).

1.2 Photochemistry of AgCl(s)

Since the time when William Henry Fox Talbot, pioneer of photography, discovered the calotype process more than a century ago (1840s), people have never ceased investigating the theory underlying it, and the behaviour of silver chloride, bromide and iodide in the light has

created renewed interests from researchers around the world with the prospective generation of new energy (H2), as photo/sono–catalysts for the degradation of organic pollutants or making novel and more efficient antibacterial materials. (Logan 1989, Hoffmann, Martin et al.

1995, Lanz, Schurch et al. 1999, Glaus and Calzaferri 2003, Young, Lim et al. 2008, Yu, Dai et al. 2009, Zhang, Wang et al. 2010, Wen, Ma et al. 2011, Zhu, Wang et al. 2013). Although some plausible assumptions have been made as to the charge transfer processes occurring in water in the presence of Ag(I) and Cl– (Calzaferri 1997), no consistent experimental data is available to confirm the proposed process:

charge transfer transition   AgCl + hv Ag s,i + Cl s (1)

 2Cl s Cl2 (2)

 + cluster growth m nAg s,i + mAg [Ag] nm (3)

Light absorption by AgCl resulted in the formation of silver atom and chlorine radical at the

• surface of AgCl (reaction (1)); two radicals of Cl readily combined to yield Cl2 (reaction (2)); silver atom and silver plus combined to form clusters of unknown charge and size according to equation (3).

Properties of silver chloride

Silver chloride can be readily synthesized by combining aqueous solutions of silver nitrate and sodium chloride:

– + – + – Ag(I) + NO3 + Na + Cl → AgCl(s) + Na + NO3

−10 The solubility product (Ksp) for AgCl is 1.6 × 10 which, once exceeded, results in formation of a white precipitate. The solid adopts a structure in which each Ag(I) ion is surrounded by an octahedron of six chloride ligands. AgF and AgBr crystallize similarly.

AgCl darkens on exposure to light yielding elemental chlorine and metallic silver; i.e.

hv 2AgCl(s)  2Ag + Cl2

Synthesis and characterization of silver chloride

In recent years some research has been carried out to connect microstructure and photocatalytic properties of nanomaterials like silver nanoparticles with materials like silver chloride or silver bromide which display semiconductor properties (Moser, Nail et al. 1959,

Hamilton 1974, Glaus and Calzaferri 2003, Wang, Huang et al. 2010, Peng and Sun 2011).

Various fabricated AgCl and Ag photocatalytic materials (with support) have been shown to display enhanced organic degradation properties with light irradiation (Hoffmann, Martin et al. 1995, Kakuta, Goto et al. 1999, Kang, Han et al. 1999, Andersson, Birkedal et al. 2005,

Rodrigues, Uma et al. 2005, Wang, Huang et al. 2008, Xiong, Zhao et al. 2011, Qi, Yu et al.

2014, Sun, Zhang et al. 2014). The resulting photocatalysts have been extensively studied with a number of modern analytical techniques (X-ray powder diffraction (XRD), Fourier transform infrared spectrometry (FT-IR), field-emission scanning electron microscopy (FE-

SEM), transmission electron microscopy (TEM), selected area electron diffraction (SAED),

UV-Vis, thermogravimetric analysis (TGA), and differential thermal analysis (DTA) and the size/morphology–dependent optical properties described (Li, Pfanner et al. 1995, Kim, Chung et al. 2010, Hu, Jia et al. 2011, Tian and Zhang 2012). The implication for our work here is that the size distribution of AgCl(s) (and the associated AgNP) as a function of time could

have an effect on the photocatalytic behaviour of these particles though proper elucidation requires further investigation.

Mechanism of photocatalytic activities of Ag/AgX

Various mechanisms have been proposed with regard to the interaction of silver and silver halides with light. For example, Choi and co-workers synthesized plasmonic photocatalysts which exhibit enhanced photocatalytic performance compared to nitrogen-doped titania nanomaterials. The improved catalytic activity was thought to originate from the enhanced adsorption for visible light, electron–hole separation, and the formation of chlorine atoms in silver chloride/silver nanostructured materials (Choi, Shin et al. 2010). The photoactivity of

AgCl extends from the UV into the visible region in a process known as self-sensitization, which is due to the formation of silver during the photoreaction (Calzaferri, Bruhwiler et al.

2001, Calzaferri 2010).

As reasonable as these speculations are, limited experimental data is presented in their support. As such, we believe that further investigation into the photochemical transformation of silver halides to silver zero and chlorine is warranted.

Mechanism of degradation of organics

Investigations have been undertaken on the degradation of organic dyes or/and other environmental pollutants using Ag associated photocatalysts and, although the effectiveness of these materials in inducing contaminant degradation is evident, no mechanism has as yet been confirmed:

 Yu et al. synthesized Ag@AgCl/rGO (Yu, Miller et al. 2014) and studied its

photocatalytic degradation of formic acid in an attempt to identify the important

oxidants for the process and optimize the reaction.

 Hu and colleagues (Hu, Peng et al. 2010, Ma, Guo et al. 2011) investigated plasmon-

induced photodegradation of a range of toxic pollutants with Ag-Agl/Al(2)O(3)

including 2-chlorophenol (2-CP), 2,4-dichlorophenol (2,4-DCP), and trichlorophenol

(TCP) using either visible light or simulated solar radiation. On the basis of

application of electron spin resonance and cyclic voltammetry under a variety of

experimental conditions, two electron transfer processes were verified from the

excited AgNPs to AgI and from 2-CP to the AgNPs. A plasmon-induced

photocatalytic mechanism was proposed.

 Zhang et al. (Zhang, Fan et al. 2011) have fabricated graphene sheets grafted

Ag@AgCl hybrid and compared to plain Ag/AgCl hybrid with 4-fold enhancement in

terms of degradating of Rhodamine B and theoretically traced their observation to

enhanced surface plasmon resonance (SPR) absorption of Ag nanocrystal.

 Similar observations have been reported for Ag/AgCl/ZnO - mediated degradation of

dye pollutants (Xu, Xu et al. 2011), Ag/AgCl - mediated degradation of methyl

orange (MO) (Xu, Li et al. 2011) and Ag@AgCl/TiO2 nanotube-mediated degradation

of methylene blue dye (Wen and Ding 2011).

To attribute the degradation of pollutants to enhanced surface plasmon resonance (SPR) and increased absorbance within the UV-Vis spectral range with dosed noble metal or Ag0 formed through reducing Ag(I) and metal halides (Br/Cl) on solid substrate seems unanimous but supporting data is required to confirm such a mechanism.

1.3 Silver nanoparticle

Recent studies have indicated that AgNP could have influence on the photocatalytic degradation of interesting pollutants in the presence of AgCl or AgBr (Wang, Huang et al.

2008, Xiong, Zhao et al. 2011, Tian and Zhang 2012, Ye, Liu et al. 2012, Adhikari, Gyawali et al. 2013, Liang, Zhu et al. 2014, Sun, Zhang et al. 2014). While the photochemical mechanisms behind those phenomena are unclear, it seems universally recognised that the presence of silver nanoparticle has improved the photolysis of AgX (X = Cl, Br). Hence, studying the photochemistry of AgCl(s) is interesting since it could have significant impact on understanding nano–sized silver as well.

Properties of Silver Nanoparticles

Silver nanoparticles exhibit several unique physical, chemical, biological and spectroscopic properties. Many of these properties are attributed to the high surface area of the nanoparticles. This research could serve to extend the knowledge of AgNP, especially that concerning its optical properties.

Optical Properties

Ag nanocrystalline domains in the AgCl particles result in the efficient absorption of visible light and enable these particles to serve as visible-light-driven photocatalysts which is considered to be due to the strong localized Surface Plasmon Resonance (SPR) associated with the Ag nanocrystallites present (Moser, Nail et al. 1959). With this property silver nanoparticles exhibit different optical properties to that of the bulk material. Furthermore, the shape and size of AgNPs are recognised to influence their spectra (Dominguez-Vera, Galvez

et al. 2007). According to results from our laboratory, the solution chemistry also impacts the

AgNPs spectra (Jones, Garg et al. 2011).

1.4 Key objectives and outcomes

Overall, understanding the photochemistry of AgCl(s) requires knowledge about a variety of related processes and species. The primary oxidants that are expected to play important roles

+ – 0 in organic decay are h , free chlorine (Cl2, HOCl and OCl ) and its precursor Cl , ROS and oxygen, while AgNP is assumed to be involved in most of the redox reactions. The main focus of this thesis is dedicated to the following endeavours:

(1) Determine the species that are generated through photolysis of AgCl(s) suspensions

in defined conditions;

(2) Determine which species are responsible for formic acid and rhodamine B photo–

decomposition;

(3) Provide knowledge for better understanding the aggregation and dissolution

behaviour of AgNP and AgCl(s) under various solution/suspension conditions.

The chapters of this thesis are organised based on these aims. AgCl(s) nanoparticles with/without pre–irradiation and their salt–dependency aggregation as well as spectral change with initial solution conditions are examined in Chapter 2. In Chapter 3, attention is given to the nature of oxidising products produced on AgCl(s) photolysis and a kinetic model developed to describe this generation process. The reaction between AgNP and OCl– at pH

8.0 and 4.0 is examined in Chapter 4 while Chapters 5 and 6 present detailed analyses of the photo–degradation of formic acid and Rhodamine B in the presence of AgCl(s). The degradation mechanism is discussed and a kinetic model for formic acid decay provided.

Chapter 2: Characterisation of AgCl(s) Particle

2.1 Introduction

The physicochemical properties of suspensions of colloidal silver chloride particles will likely influence their reactivity, both photochemical and otherwise. This Chapter outlines preliminary investigation into how the solution conditions influence both particle formation rates, particle size distributions and the observed electronic spectra of the particles, as well as examining how these properties change during the course of the photochemical reactions to be described in subsequent chapters. In particular, many processes have been examined as a function of the ”aging time”, defined as the time elapsed since addition of Ag+ to the chloride-containing solutions. As will be demonstrated in Chapter 3 (Generation of Oxidants) the generation rates of free chlorine and singlet oxygen from irradiated AgCl(s) solution correspondingly changed when initial Ag(I) or Cl– ion concentrations changed. Since solution conditions obviously play a major role in this study, insights into the formation process of silver chloride and associated particle properties are necessary.

Zeta potential is recognized to represent electrokinetic potential in colloidal systems

(McNaught and Wilkinson 1997). It is broadly used to represent the stability of a colloid system (Greenwood and Kendall 1999). Colloids with high zeta potential (negative or positive) are electrically stabilized while colloids with low zeta potentials tend to coagulate or flocculate (Greenwood and Kendall 1999, Hanaor, Michelazzi et al. 2012). This relationship is outlined in Table 2.1.

Table 2.1: Zeta Potential Values and Stability of Colloid (Hunter 2001)

ζ potential (mV) Stability behaviour of the colloid From 0 to ±5 Rapid coagulation or flocculation From ±10 to ±30 Incipient instability From ±30 to ±40 Moderate stability From ±40 to ±60 Good stability More than ±61 Excellent stability

The study of many of these parameters has a long history stretching back many decades.

Earlier studies on the precipitation of silver chloride have shown that initial particle formation is quite rapid (within a few hundredths of a second), with the turbidity of these suspensions initially growing very rapidly, followed by a period of slower development to a maximum and then subsequent decrease (Greene and Frizzell 1936). The degree of turbidity of the suspension also depends to a large part on the concentration of reagent (i.e., either

Ag(I) or Cl– ions) present in excess; once the AgCl(s) particles are greater than 100 nm in size, this scattering follows Rayleigh’s law.

Some other studies on nucleation and crystal growth of silver chloride precipitants concluded that the precipitation rate is not constant throughout the precipitation from homogeneous solution and that the size distribution can be controlled to some extent by varying the concentrations of reactants (Klein and Gordon 1959).

In colloid aggregation two limiting kinetic regimes have been identified: rapid, diffusion- limited (DLCA) and slow, reaction-limited (RLCA) cluster aggregation. Rapid (DLCA) aggregation kinetics, where all collisions result in attachment and cluster growth is limited only by diffusion, has been shown to result in clusters with open structures and fractal

dimensions (Df) of ~1.8 . In comparison, when a repulsion barrier is present, reaction limited cluster aggregation (RLCA) occurs with a large number of collisions required for two particles to stick together. In this case, a relatively compact structure results with a higher Df than is the case for DLCA (Lin, Lindsay et al. 1990).

In recent years research has been carried out to investigate the morphology-dependency of photochemical properties of Ag/AgCl. For example, Zhu, Chen and Liu (Zhu, Chen et al.

2011) reported that quasi-cubic Ag/AgCl nanospecies displayed much higher photocatalytic activity than the corresponding nanospheres with regard to the degradation of MO (Methyl

Orange) under sunlight and visible-light irradiations. These investigators tentatively attributed their findings to i) higher adsorptive ability for target pollutant molecules (with some adsorption data comparing spherical and cubic shaped Ag/AgCl on MO); ii) more catalytically active sites (i.e. corners, edges, steps, etc) on cubic than on a spherical materials; iii) higher percentage of active facets in the cubic Ag/AgCl-based nanostructures compared to the spherical materials.

The search for photocatalytic nanomaterials that can better harness sunlight and decompose pollutants more efficiently is perhaps the most intensive activity in environment-oriented materials R&D. Thus the findings of Zhu and his colleagues are very exciting. Nevertheless, a lack of convincing explanation for results such as these remains. Not only are the supporting data often missing in such studies but the assumptions made in interpreting the data are often contradictory. For instance, in the discussion as to which shape of AgCl crystal was most active in the photo-degradation of organics, Zaizhu Lou and his co-workers (Lou,

Huang et al. 2011) arrived at a completely different conclusion to that of Zhu, Chen and Liu

(Zhu, Chen et al. 2011) who stated that the plasmonic photocatalyst Ag@AgCl prepared from

the near-spherical AgCl showed higher activity than that of its cubic counterpart with respect to methyl orange (MO) dye degradation!

There is also previous work examining the microscopic structure of Ag/AgCl nanocomposites in relation to their photolysis. For example, (Bi and Ye 2009) introduced

SEM images of Ag/AgCl core-shell nanowires which showed the AgCl wrapped in Ag nanowires and they discussed its degradation of MO; Jiang et al. (Jiang and Zhang

2011)fabricated Ag/AgCl hollow spheres by adding AgNO3 to micelles of C18N3 dispersed in

HCl solution and found the concentration of C18N3 played a key role in the morphology of the product; (Wang, Huang et al. 2008) synthesized Ag/AgCl photocatalysts with various morphological structures including microrods, mixtures of microrods and irregular spheres and hollow spheres and examined the ability of various Ag/AgCl assemblages to degrade MO and concluded that the hierarchical hollow spheres, which facilitated greater exposure to light, improved the photocatalytic activity.

In order to study the photochemistry of the AgCl(s) system, it is important to understand the nature of the absorption process. The fundamental driving force for chemical change in this work is the absorption of light by AgCl(s); clearly, the nature of the possible mechanisms that lead to light absorption (and also scattering) are of importance to understanding this interaction. There are four main ways in which light will be absorbed or scattered in this system, namely i) through elastic scattering, ii) the semiconductor absorbance of AgCl(s), iii) the plasmon absorbance of any Ag(0) that may be formed and iv) through electronic transitions of solvated silver atoms.

Simple elastic light scattering will occur as AgCl(s) particles are formed; when these particles are much smaller than the wavelength of the incident light they will exhibit Rayleigh

scattering with an intensity proportional to λ−4. When the size of the particles approaches the wavelength of light, the nature of this scattering process becomes more complex, with treatment of this process beyond the scope of this work. In addition to this elastic scattering,

AgCl(s) will also undergo an electronic transition when the incident light has an energy greater than its bandgap. AgCl(s) is a semiconductor with an indirect band gap of 3.25 eV

(Scop 1965), which corresponds to a wavelength of 381.5 nm. As a result, incident light with a wavelength lower than 381.5 nm will be absorbed by the AgCl(s), leading to an electron being transferred from the valence band to the conduction band, leaving an electron vacancy

+ − in the valence band, termed a “hole” (hvb ), and an electron in the conduction band (ecb ).

This is likely the fundamental excitation process, and it is the resulting chemistry of this photo-formed hole-electron pair that will drive all further reactions in this system.

A likely sink for the conduction band electrons is the reduction of some of the Ag(I) atoms within the AgCl(s) to Ag(0), potentially leading to the formation of silver nanoparticles

(AgNPs). Should these AgNP’s form in the system, they would be expected to exhibit plasmonic absorbances, which are derived by interaction of the electric field of the incident light with the free electrons in the silver metal, leading to oscillations in the charge distribution. The nature of this interaction is rather complex, with the structure of the resulting absorbance strongly dependent on the permittivity of the surrounding medium, the

AgNP diameter, the proximity of an AgNP to other AgNPs and also the structure of any

AgNP assemblage that may form (Russell, Mantovani et al. 1987, Quinten and Kreibig 1993,

Kumbhar, Kinnan et al. 2005, Kinnan 2008, Mahmoud and El-Sayed 2008). The position of the absorbance maximum for the AgNP surface plasmon resonance (SPR) is typically in the range of 375 – 405 nm (Mulvaney 1996).

The impact of size can be seen from the data of Russell et al. (1987), where the impact of the size of spherical particles mounted on a solid quartz support in air was examined (the inter- particle spacing was large enough to prevent coupling between particles). The general trend observed is that the peak both broadens and splits into multiple peaks (due to separation of dipole and quadrupole resonances) as the particle diameter increases (Figure 2.1). A similar result is seen in the solution phase studies undertaken by Kinnan (2008) (Figure 2.2), although in this case the splitting of the peaks is less apparent and the wavelengths of peak absorbance are shifted, probably as a result of the different dielectric medium (water c.f. air) and also due to the non-spherical morphology of the particles (formed from H2 reduction of

Ag2O).

Figure 2.1: Measured absorbance spectra of spherical AgNPs mounted on quartz posts in air. Spectra are adapted from Figure 2 of Russell et al. (1987).

Figure 2.2: Measured absorbance spectra of AgNPs dispersed in aqueous solution. Spectra are adapted from Figure 3.1 of Kinnan (2008).

The general broadening apparent with increased size observed in Figure 2.2 can be reduced significantly if the same particles are immobilized on some kind of support since, for larger particles (~100 nm and greater), if they are in close proximity, then the plasmon resonances are able to couple coherently, leading to much sharper peaks (see Figure 2.3 below). It is clear that the environment of the particles has a great impact upon the nature of absorption.

As well as the proximity of the particles, the nature of any aggregates that form also has significant impacts upon their absorption. This has been treated theoretically by Quinten and

Kreibig (1993) and subsequently validated experimentally by Quinten (1999). As shown in

Figure 2.4 below, as more AgNP particles are added to a planar sheet, the peaks increasingly broaden and shift to higher wavelength, approaching the spectrum of an infinite thin film of silver. At a fixed particle number, the structure of the aggregate also influences the nature of the coherent coupling, resulting in a differing spectrum for different particle arrangements

(Figure 2.5). Ultimately, it is clear that the shape of any AgNP spectrum is strongly dependent upon a large variety of factors, with the resulting observed spectrum quite variable.

Figure 2.4: Influence of number of particles on the absorbance of a planar quadratic array of adjacent 40 nm diameter particles, including an infinite sheet of silver atoms. Adapted from Figure 8 of Quinten and Kreibig (1993)

Figure 2.5: Influence of aggregate structure upon absorbance of 40 nm AgNPs. Adapted from Figure 4 of Quinten and Kreibig (1993)

Should isolated silver atoms (Ag0) form in solution, it would be expected that their

4 −1 −1 reasonably strong absorbance (360 = 1.4 × 10 M cm , Tausch-Treml, Henglein et al.

(1978)) would allow them to be observed at sub-micromolar concentrations (a concentration of 0.7 µM would lead to an absorbance of 0.01). Once formed however, they are likely to rapidly be complexed by any Ag+ ions in solution and also undergo precipitation to ultimately

+ m+ yield some form of Ag -associated AgNP (termed Agn ), which, as can be seen in Figure 2.6 below, has absorbance essentially similar to the AgNP spectra discussed previously, as would be expected considering they are the same chemical entity even though formed from a different pathway.

Figure 2.6: Changes in the absorbance spectra of atomic Ag0 entities as they undergo reaction to give AgNPs. Spectra are those reported by Tausch-Treml et al.(1978)

In order to understand the underlying behaviour of the AgCl system examined in this work, the relationship between cluster structure, aggregation kinetics and the light absorption of silver chloride was examined. The objectives of this chapter are dedicated to understanding i) how the initial concentrations of Ag(I) and Cl– affect the formation and the size of AgCl(s)

– particles ; ii) the aggregation behaviour of AgCl(s) particles in relation to [Ag(I)]0 and [Cl ]0 ; iii) the stability of AgCl(s) colloid and its associated solution conditions; iv) the compositional change of AgCl(s) nanoparticles after being exposed to light (in the presence of target organics formate or rhodamine B (RhB)).

2.2 Experimental Details

2.2.1 General

Solutions of 50 mM silver nitrate and 5 M sodium chloride were prepared. Citrate stabilized

AgNP solution was prepared according to the method described by (Liu and Hurt 2010). All pHs ≥ 8.0 solution were maintained with 2 mM NaHCO3 adjusted with HNO3/NaOH while pH 4.0 solutions were achieved using 0.1 mM HNO3. All glassware, plasticware and the reaction vessel were cleaned as described in Chapter 3. The initial concentrations of both Ag+ and Cl− used throughout this chapter are representative of those used in later chapters. To control aging time, fresh colloidal silver chloride suspensions were prepared for each experiment and allowed to aggregate for the desired duration in a 1.0 L glass reaction vessel wrapped completely in aluminium foil to prevent any environmental light penetration. In all cases the AgCl(s) suspension was generated by spiking AgNO3 stock solution into the NaCl- containing buffer solution. The pH value of the solution was pre-adjusted to the desired value immediately before spiking Ag(I). Incident light, when desired, was provided horizontally through the quartz side window (of approximately 16 cm2 area) of the 1.0 L glass vessel.

Total sample volume used for analysis was no more than 50 mL (5 % of 1.0 L), so that the error introduced by volume change (which alters the light energy flux per unit volume) was minimal.

Particles were characterised by SEM (scanning electron microscopy), STEM (scanning transmission electron microscopy) and associated Energy Dispersive X-ray (EDX) spectroscopy. The high magnification capability of the SEM/STEM enables elucidation of the shape, size and surface structure of the nano-sized specimens. Simultaneously acquired EDX spectra (enabling determination of elemental composition) were obtained by focusing the

electron beam into a narrow spot on the assemblage under investigation. X-ray diffraction

(XRD) analysis was carried out in order to determine the AgCl(s) crystal structure before and after light exposure while a Malvern Nano ZS instrument was used to monitor both the zeta potential and size development of AgCl(s) clusters in NaCl solutions. Size measurement was based on dynamic light scattering (DLS) with incident light produced by a 4 mW He–Ne laser with a wavelength (λ) of 663 nm. Scattered light was detected with an avalanche photodiode detector at a scattering angle (θ) of 173°. A digital correlator was used to develop an autocorrelation function which was analysed using the method of cumulants (Koppel

1972, Finsy 1994). An apparent intensity weighted mean hydrodynamic diameter (z–average diameter dz) was adopted to represent the size of aggregates. If the cluster were spherical dz would be equal to the actual intensity weighted hydrodynamic diameter, dh.

The absorbance of AgCl(s) suspensions was obtained using a highly sensitive Cary

+ - spectrophotometer. While the concentrations of silver species (Ag , Ag(OH)0, Ag(OH)2 ,

2- - AgCl3 , AgCl0, AgCl2 , AgCl(s)) were calculated with program “Visual minteq”.

2.2.2 Particle Size and Zeta-Potential Measurement

An AgCl(s) suspension was generated as described in §2.2.1. At desired times, a 1.5 mL sample was transferred to a disposable polystyrene cell for analysis using the Malvern

Zetasizer “Nano ZS” instrument. Sampling was undertaken every 5 minutes for 45 to 60 minutes (depending on the experiment). Zeta potential measurements were carried out similarly with the AgCl(s) solution stirred at a constant rate for all experiment in this Chapter

(as well as experiments in Chapter 3) during the measuring process.

2.2.3 Sample Preparation for SEM/STEM Analysis

The AgCl(s) suspension was created and aged in the 1.0 L vessel for 35 minutes after which time the AgCl(s) aggregates were filtered through 0.22 µm Millipore GV membranes using a peristaltic pump to drive the suspension through the filtration apparatus. Both the collecting flask and reaction vessel were covered with aluminium foil during the process. The filtration process was repeated as often as required in order to obtain sufficient AgCl(s). The filter paper was then placed in a -85 °C freezer overnight and then freeze dried for 3 days using a

MODULYOD Freeze Dryer from Thermo Electron Corporation.

The recovered solid material was then transferred into a 1.0 mL eppendorf tube and dispersed with approximately 300 µL of 100% ethanol for SEM analysis or with methanol for STEM analysis. This suspension was then sonicated until the specimen was well dispersed, yielding a slightly grey suspension. A small volume of suspension was then dropped onto a piece of mica and left in the fume hood for around 5 minutes to allow the residual ethanol/methanol to evaporate. The specimen was then cased (with the case covered with aluminium foil) and left at 35°C overnight until the day of observation with SEM/STEM.

Immediately prior to the SEM observation, the sample was sputter coated with a thin layer of chromium to improve the electrical conductivity and secondary electron emission of the sample. The SEM used in this work was a Hitachi S3400 while the STEM was a FEI Tecnai

G2 with both instruments housed within the Mark Wainwright Analytical Centre at the

University of New South Wales.

2.2.4 Electronic Spectra of AgCl(s) Colloid

The absorbance of AgCl(s) suspension was measured by scanning from 300 – 700 nm using a

Cary 50 Bio UV-Vis spectrophotometer using a 1 cm pathlength quartz cuvette. The scans were baseline corrected and zeroed against the appropriate buffer solution (NaCl + NaHCO3).

2.2.5 XRD measurement

Silver chloride samples in the dark and after visible light irradiation were prepared by first

– spiking Ag(I) into Cl containing NaHCO3 solution at pH 8.0; AgCl(s) was left to age for 35 minutes (in the dark); the AgCl(s) suspension was then irradiated for an hour with visible light. The sample was then collected on 0.22 µm GV Millipore membrane, freeze dried and used for XRD analysis.

2.3 Results and discussion

2.3.1 Formation of AgCl(s) Colloid

– When fixing [Ag(I)]0 = 100 µM but changing [Cl ]0, the equilibrium concentration of AgCl(s)

– remained relatively stable; however when fixing [Cl ]0 = 100 mM and continuously increasing [Ag(I)]0 , the [AgCl(s)]0 built up accordingly. In this system the ion activity product (IAP) of AgCl was at least 104-fold higher than the solubility product Ksp(AgCl) =

10−9.74 (25 °C , I = 0 M) (Morel and Hering 1993), suggesting that AgCl(s) formation occurred under all conditions examined. The predicted AgCl(s) concentrations as a function of initial chloride and silver(I) concentration assuming equilibrium conditions are shown in

Figures 2.7 and 2.8 and silver species distribution is shown in Table 2.2.

100

100 µ M AgNO3 Silver Chloride Concentration(µM) 70 0 25 50 75 100 125 150 175 200 225

Cl– ion concentration (mM) Figure 2.7: AgCl(s) equilibrium concentration as a function of initial chloride ion concentration,

– [Ag(I)]0 = 100 µM and [Cl ]0 = 20 – 200 mM.

200

160

120 (µM) 80

Silver Chloride Concentration Chloride Silver 100 m M NaCl 0 0 25 50 75 100 125 150 175 200 225 Ag(I) (µM)

– Figure 2.8: AgCl(s) concentration as a function of [Ag(I)]0 (10 – 200 µM), initial concentrations: [Cl – ]0 = 100 mM , [HCO3 ]0 = 2 mM .

Calculations at pH 4.0 and 11.0 yielded a similar distribution of dominant species with a consistent increase in AgCl(s) concentration with increase in initial silver(I) concentration,

– and a relatively stable AgCl(s) concentration at fixed [Ag(I)]0. Specifically, when [Cl ]0 was raised from 50 mM to 200 mM, AgCl(s) equilibrium concentration decreased by only 1.8%;

yet as [Ag(I)]0 increased from 50 µM to 200 µM, silver chloride concentration increased by over 360%.

Table 2.2: The Speciation of Silver by Percentage.

– [Cl ] [Ag(I)] + − 2− 0 − Ag Ag(OH)0 Ag(OH)2 AgCl3 AgCl AgCl2 AgCl(s) mM µM 200 100 1.8 × 10−3 1.3 × 10−7 1.8 × 10−11 1.6 2.0 8.0 88.4 100 100 3.2 × 10−3 2.5 × 10−7 3.3 × 10−11 0.40 4.0 4.0 91.6 90 100 3.5 × 10−3 2.8 × 10−7 3.6 × 10−11 0.33 4.4 3.6 91.7 70 100 4.3 × 10−3 3.5 × 10−7 4.4 × 10−11 0.20 5.6 2.8 91.4 50 100 5.8 × 10−3 4.8 × 10−7 5.9 × 10−11 0.10 7.8 2.0 90.0 20 100 1.3 × 10−2 1.1 × 10−6 1.3 × 10−10 1.8 × 10−2 19.0 0.84 80.2 100 10 3.2 × 10−2 2.5 × 10−6 3.3 × 10−10 4.05 39.5 40.2 16.3 100 25 1.3 × 10−2 1.0 × 10−6 1.3 × 10−10 1.62 15.8 16.1 66.5 100 50 6.4 × 10−3 5.1 × 10−7 6.5 × 10−11 0.81 7.9 8.0 83.3 100 100 3.2 × 10−3 2.5 × 10−7 3.3 × 10−11 0.40 4.0 4.0 91.6 100 200 1.6 × 10−3 1.3 × 10−7 1.6 × 10−11 0.20 2.0 2.0 95.8

It can be seen from Table 2.2 that when [Ag(I)]0 ≥ 25 µM, AgCl(s) was always dominant with more than 66 % of the silver added forming AgCl(s); under certain solution conditions,

~ 96 % of all the Ag added was present as AgCl(s). The next two abundant species were

– typically AgCl2 and AgCl(aq).

2.3.2 Aggregation of Silver Chloride in the Presence of Sodium Chloride

Silver chloride sol has the tendency to aggregate with resultant particle size growth, as shown by the results of size measurement. As can be seen from Figure 2.9, size increase was strongly influenced by [Ag(I)]0 with increase in initial Ag(I) concentration resulting in increase in the average diameter of AgCl(s) particles. When [Ag(I)]0 is 10 µM, after 5

minutes reaction between Ag(I) and Cl– the diameter of the AgCl(s) particle was 105 nm; while if [Ag(I)]0 was 100 µM, after the same period of time of reaction, the AgCl(s) diameter reached 168 nm.

800 100 µM 700 50 µ M 600 25 µM 10 µM 500

400

(d. nm) (d. Z

d 300

200

100

0 0 5 10 15 20 25 30 35 40 45 50 Time (min)

Figure 2.9: The AgCl(s) particle size as a function of time: [Ag(I)] = 100 µM (solid diamond), 50 µM (empty triangle), 25 µM (empty square) and 10 µM (solid triangle). Common solution conditions: [Cl–

– ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0.

The data also show that, with the passage of time, AgCl(s) particles kept growing in size, presumably due to aggregation. Since Ag(I) and Cl– combine to form AgCl(s), within 60 ms after mixing (Kimijima and Sugimoto 2004), significant growth of AgCl(s) has occurred prior to collecting the first datapoint. Earlier studies examining the nephelometric turbidity also confirm rapid formation of AgCl(s) particles of visible size within a few hundredths of a second (Greene and Frizzell 1936).

On changing the initial concentration of chloride at a fixed initial Ag(I) concentration, the

AgCl(s) particle size measurement displayed a similar trend as was the case for variations in

Ag(I) concentration, i.e., increasing rates of growth with increasing chloride concentration, as shown in Figure 2.10 below.

800 200 m M 100 mM 700 (a) 90 mM 70 mM 600 50 mM 25 m M 500

400

(d. nm) (d. Z

d 300

200

100

1600 200 m M 100 mM 1400 (b) 90 mM 70 mM 1200 50 mM 25 m M 1000

800

(d. nm) (d. Z

d 600

400

200

0 0 5 10 15 20 25 30 35 40 45 50 Time (min) Figure 2.10: (a) The size of silver chloride as a function of time. Initial concentrations of Cl– = 200 mM (empty triangle), 100 mM (solid diamond), 90 mM (empty square), 70 mM (solid triangle) and

50 mM (empty diamond), 25 mM (solid square). Common solution conditions : [Ag(I)]0 = 100 µM, – [HCO3 ]0 = 2 mM, pH 8.0. (b) shows the same data as (a) but on an expanded y-axis scale.

Figure 2.10 demonstrates that the AgCl(s) particle size initially increased at a linear rate. As shown in Figures 2.11, higher initial concentration of Cl– resulted in more rapid aggregation, with an apparent critical coagulation concentration on the order of 100 mM NaCl. This increased coagulation of AgCl(s) with increased [NaCl] is a result of shielding and compression of the electrical double layer of AgCl(s) particles, facilitating the approach of two particles resulting in collision and agglomeration. Similar patterns of aggregation dependence on Ag(I) and Cl– were also observed at pH 4 and pH 11.

At relatively low Cl– concentrations (25, 50, 70 and 90 mM), each increase in the electrolyte concentration led to a corresponding increase in the aggregation rate until a concentration of

≈100 mM Cl– was reached, at which point further increases in Cl– ion concentration did not lead to increases in the aggregation rate, suggesting that diffusion limited aggregation was occurring, i.e., 100 mM NaCl was the critical coagulation concentration.

A log-log plot such as that shown in Figure 2.11 below would typically display a linear trend below the CCC. That this is not the case here is likely due to the fact that the background electrolyte is actually intimately involved in the particle formation in the first place, leading to not only a change in the electric shielding in the system but also to a change in the formation of the initial clusters and quite likely their surface properties. It can be seen in

Table 2.1 that below 50 mM NaCl the proportion of Ag present as AgCl(s) begins to decrease from the relatively constant ~90% down to ~80% for 20 mM NaCl, suggesting this may impact the aggregation properties. The zeta potential measurements to be discussed in §2.3.3 below however suggest that there is only minimal change in the surface charge of the particles under these conditions.

– Figure 2.11: The plot of ddh/dt as a function of NaCl concentration. [Ag(I)]0 = 100 µM, [HCO3 ]0 = 2

– – mM (pH 8.0, triangles), [OH ]0 = 1 mM (pH 11.0, circles), [HNO3 ]0 = 0.1 mM (pH 4.0, squares). The dashed lines respresent an estimate of the diffusion limited rate, obtained by the weighted average of the 100 mM and 200 mM NaCl data points.

The z–average diameter change with time for fixed chloride ion concentration but increasing

[Ag(I)]0 is shown in Figure 2.12. The data indicate a constant increase in size with increase in silver(I) concentration. As shown in Eq. 2.3 below (rearranged from Eq. 2.1), the growth rate of dh is expected to be proportional to N0 with the constancy in rate of size change as a function of silver(I) concentration consistent with this (as higher [Ag(I)] necessarily leads to greater concentrations of AgCl(s)), although it is also likely that the initially formed number of AgCl(s) clusters will likely change with Ag(I) concentration. It was not possible however to reach any deductions about the initial number-concentration of AgCl(s) clusters from this data.

dd t Z  kN (2.3) dt 0

20 pH 8.0 pH 11.0 15

pH 4.0 t

/d 10

d d

0 0 20 40 60 80 100 120 Ag(I) concentration (µM)

– – Figure 2.12: Growth rate of dZ as a function of silver plus concentration. [Cl ]0 = 100 mM , [HCO3 ]0

− = 2 mM (pH 8.0, solid diamond and solid line), [OH ]0 = 1 mM (pH 11.0, empty square and dash line), – [HNO3 ]0 = 0.1 mM (pH 4.0, empty triangle and long dash dot).

The particle size measurements described above were all conducted in the absence of light and the absence of any added probe compounds, which may not be representative of the irradiated solutions containing probe compounds that will be considered in later chapters. In order to confirm that the previously described results are representative of these conditions,

AgCl(s) aggregation was also determined with visible light irradiation (after dark aging for

35 minutes) in the absence and presence of the orgaic compounds formate and rhodamine B.

The results demonstrate that irradiation did not appreciably change the rate of size increase compared to the results obtained in the dark; the results also indicate that adding organic probe compounds had no influence on silver chloride aggregation (Figure 2.13)

2000 add 1.0 µ M formate

1600 w/o added organic add 0.21 µ M RhB

1200 in the dark (d. nm) (d.

Z 800 d

400

(a) 0 30 50 70 90 110 130 150 170 Time (min)

700 add 1.0 µ M formate w/o added organic

add 0.21 µ M RhB in the dark 600

500

(d. nm) (d.

Z d

400

(b) 300 30 40 50 60 Time (min) Figure 2.13: (a) Aggregation kinetics of AgCl(s) with irradiated 1.0 µM formate (empty diamond), irradiated 0.21 µM Rhodamine B, irradiated without added organic (solid square), without irradiation (empty square). Light was turned on after 35 minutes Ag(I) and Cl– ions reaction in the dark, time

– – zero was the time point of spiking Ag(I) inot Cl solution. [Ag(I)]0 = 100 µM, [Cl ]0 = 100 mM,

– [HCO3 ]0 = 2 mM, pH 8.0; (b) a close-up of size measurement between 30 to 60 min.

2.3.3 Zeta Potential

As ζ potential can be used to identify the likely stability of a colloid system, we measured the

ζ potential values of AgCl(s) colloids under various solution conditions, with various initial concentrations of Ag (I) or Cl–. The results are shown in Figure 2.14 and 2.15.

0 0 10 20 30 40 50 200 mM -20 100 mM 50 mM 25 mM -40

Zeta Potential (mV) Potential Zeta -60

-80 Time (min)

Figure 2.14: The ζ potential values of silver chloride colloid as a function of reaction time between

– – Ag(I) and Cl ion. Solution conditions: [Cl ]0 = 200 mM (solid square) , 100 mM (empty diamond) ,

50 mM (solid triangle) and 25 mM (empty square); [Ag(I)]0 = 100 µM , pH 8.0.

Time (min) 0 0 25 µM 10 20 30 40 50 50 µM -20 100 µM 200 µM -40

Zeta Potential (mV) Potential Zeta -60

-80

Figure 2.15: The ζ potential values of silver chloride colloid as a function of reaction time between

– Ag(I) and Cl ion. Solution conditions : [Ag(I)]0 = 25 µM (empty triangle) , 50 µM (solid diamond) , – 100 µM (empty square) and 200 µM (solid triangle); [Cl ]0 = 100 mM , pH 8.0.

The ζ potential data in Figures 2.14 and 2.15 indicate that AgCl(s) colloids should have good stability as their ζ potential values ranged from −45 mV to −68 mV during the analysis period. The ζ potential fluctuated a little but did not really display notable changes, either as a function of time or of solution conditions. When the initial sodium chloride concentration was fixed at 100 mM, higher silver nitrate concentrations led to a small increase in the magnitude of the ζ potential, thus indicating likely greater stability. Despite the relatively high absolute value of the ζ potentials, the particle size increased successively from 5 minutes to 45 minutes of aging time.

2.3.4 Silver Chloride Particle Surface Topography and Composition by SEM

Electron microscopy was conducted for four different sample types, namely an un-irradiated

AgCl(s) that was kept in the dark during the 35 minute aging time, and three visible light irradiated, dark aged AgCl(s) samples that were irradiated either in the presence of 1.0 µM

formate, 0.21 µM rhodamine B or no added probe compound, with visible light irradiation durations of 110, 30 or 35 minutes, respectively. All these materials were prepared from 100

+ µM Ag and 100 mM NaCl in pH 8 2 mM NaHCO3 buffer. These conditions are representative of the material to be discussed in subsequent chapters. Both SEM and STEM imagery as well as accompanying Energy Dispersive X-ray Spectroscopy (EDS or EDX) spectra were obtained, which demonstrated distinct differences among the four specimens of

AgCl(s) with regard to size, shape, relative elemental composition ratios and surface structure.

The data showed that light exposure changed the initial particle size and shape (comparing

Figure 2.19 to Figure 2.17). The relative abundance of large particles decreased after being irradiated. The results also displayed variance among the three types of irradiated specimens

(with and without pre-added organics), which implied the organic probe compound impacted the AgCl(s) morphological transformations.

2.3.4.1 EDS Results from SEM Analysis

The EDS analysis of the AgCl(s) particles showed that the specimen displayed emissions typical of chlorine and silver, with some sodium also occasionally present, presumably due to some crystal formation of NaCl, which was co-crystallized with AgCl(s) but was not fully rinsed off during the sample preparation step. As expected, traces of chromium (Cr)

(specimen coating), aluminium (Al) and silicon (Si) (mica) were also detected. These elements had relatively low counts of emitted X-rays and are simply due to the support materials upon which the samples were mounted. Although the complexity of the X-ray emission process and the heterogeneity of the samples precludes reliable quantitative analysis, some semi-quantitative inference is possible.

The scaled EDS spectra are given in Figure 2.16 below, where it can be seen that the strength of the Cl-derived emissions, relative to the Ag-derived emissions, varies greatly between treatments. It can be seen that there is the greatest quantity of Cl in the dark sample, followed by the light-irradiated sample in the absence of organics, with the organic irradiated samples having similar composition. This suggests that the dark sample is likely to simply be AgCl, with increasing quantities of Ag(0) present in the other cases; the decrease in the Cl emission peak is a result of the concomitant removal of Cl (through processes to be described in subsequent chapters) during the reduction of Ag(I) to Ag(0), i.e., the reduction in the Cl peak is likely due to conversion of a portion of of the AgCl(s) to Ag(0) nanoparticles (AgNPs), presumably located upon the surface of the AgCl(s).

Figure 2.16: The EDS spectrum of AgCl(s) particles when imaged using SEM. For each condition two separate particles have been imaged. All spectra have been normalized to the Ag Lα1 peak to allow for simpler comparison. The “Ag Component” trace is the calculated component for all Ag emissions using data from Thompson, Lindau et al. (2009)

2.3.4.2 SEM of AgCl(s) particles kept in the dark

The SEM images of AgCl(s) in the dark (Figure 2.17) indicate the presence for the most part of cubic AgCl(s) particles of fairly uniform, similar size. It can be seen that the particles display a quasi-cubic structure, which is consistent with the fact that standard AgCl(s) crystals have a fcc (face-centered cubic) NaCl structure with each Ag(I) ion surrounded by an octahedron of six Cl− ligands (Figure 2.18). The sizes of the crystals are mostly in the range of 650 nm to 1.2 µm in diameter though some amorphous silver chloride particles are also formed. These sizes appear reasonably consistent with the hydrodynamic sizes reported earlier (e.g. Figure 2.10), although suggest that some additional aggregation has occurred

during the solids recovery procedure. These results provide information on the nature of the

AgCl(s) particles present prior to any photocatalytic reaction which act as the baseline for comparison with the other AgCl(s) irradiated treatments.

Figure 2.17: The SEM image of AgCl(s) particles in the dark.

Figure 2.18: The typical silver chloride crystal structure (WebElements 2013) showing a fcc crystal structure.

2.3.4.3 SEM of AgCl(s) particles irradiated without organics

Silver chloride particles irradiated with simulated sunlight were more polydisperse than those kept in the dark exhibiting many smaller quasi-spherical particles (sizes ranging around 130 nm to 300 nm) apparently attached on the surface of the larger particles (sizes about 650 nm to 1.6 µm, see Figure 2.19). The smaller quasi-spherically shaped particles are considered to be Ag(0) in the form of silver nanoparticles, however, conclusive evidence for this is lacking.

The larger sized particles have sizes and shapes similar to those observed in Figure 2.17 and are consistent with the earlier hydrodynamic diameter measurements (Figure 2.9). The images also show that after being exposed to light a number of particles appeared more smooth (Figure 2.19) losing their angles and becoming more semi-spherical or irregularly shaped. This is thought to be the result of photochemical reactions.

Figure 2.19: The SEM image of AgCl(s) particles after being exposed to simulated sunlight irradiation for 30 minutes. Magnification ×20k (a) and ×10k (b).

2.3.4.4 SEM of AgCl(s) particles irradiated in the presence of rhodamine B

When AgCl(s) was irradiated in the presence of rhodamine B for 30 minutes (after the usual

35 minute dark aging period) the particles seemingly retained their cubic shape. The growth of smaller assemblages on the surfaces of the AgCl(s) is also observed to some extent (Figure

2.20a). The particles generally have a reasonably uniform distribution of sizes in the range of

300 – 400 nm, with some larger particles also present (e.g. Figure 2.20a), agreeing well with the earlier hydrodynamic diameter measurements (Figure 2.13). These particles appear generally smaller than the light or dark treatments above, with the reason for this not clear.

When AgCl(s) was irradiated in the presence of 1.0 µM sodium formate, the particles were clearly divided into two size ranges (see Figure 2.21), one with diameters from 130 nm to 270 nm consisting of quasi-spherical particles, and the other with sizes from ~500 nm to 630 nm.

Some particles of diameters greater than 1 µm were also present, and were comprised mainly of cubic or semi-cubic particles. These larger particles appear similar to the dark-aged

AgCl(s) particles which had diameters around 640 nm to 1.2 µm. The smaller sized particle

(d < 300 nm) are once again attributed to reduced silver (Ag0) after photolysis of AgCl(s), and appear to be formed with much greater abundance than was the case for either the light or rhodamine B cases.

Although all the light-irradiated samples exhibited significant morphological changes, the changes were most extensive in the presence of formate. Although the more extensive

duration of irradiation no doubt plays a role, it is also considered that the ability of formate to react with the oxidizing entity formed under visible-light irradiation of AgCl(s) is a likely reason for this more extensive change. After irradiation, part of AgCl(s) transformed into

Ag0, with the presence of formate apparently aiding electron-hole pair separation (since photo-generated hole and/or its initial reaction product is a strong oxidant able to oxidize formate) thus leading to more silver nanoparticle generation by reduction of Ag(I) ions to

Ag0, which coalesces to form silver nanoparticles.

Figure 2.21: SEM image showing AgCl(s) crystals after being irradiated with 1.0 µM initial sodium formate, magnification ×30k (a) and ×19k (b).

2.3.5 Silver Chloride Particle Surface Topography and Composition by STEM

Although insightful, there were significant limitations with the SEM results presented in the previous section. In an attempt to overcome these limitations, Scanning Transmission

Electron Microscopy (STEM) was employed, with the higher resolution from this technique expected to help clarify the potential formation of Ag(0) domains. In this technique one individual representative particle was selected for detailed element map imaging and for EDS analysis at several points within the one particle.

2.3.5.1. STEM of irradiated AgCl(s) without organics

For the light irradiated sample six distinct points of the assemblage were analysed with EDS, however, as the particle moved during the EDS acquisition process, only 5 of these points could be used in the subsequent analysis (point 6 in Figure 2.22 a, b). The element maps

(Figure 2.22 d – g) seem to suggest that this particle was an AgCl(s) assemblage, perhaps coated in some NaCl.

The EDS spectra from the 5 points are shown below in Figure 2.23, demonstrating that points

1 and 4 have much more Ag relative to Cl than the other points. However, this is not conclusive for formation of Ag(0), since this could also simply be due to a thinner coating of

NaCl at these points (Figure 2.22 f demonstrates significant NaCl coating all over this particle).

Figure 2.23: EDX spectra scaled to Ag Lβ1 (a) or Cl Kα1 (b) emission peaks. Point numbers correspond to those in Figure 2.23 a, b.

2.3.5.2. STEM of irradiated AgCl(s) with rhodamine B

The results in the presence of rhodamine B are similar to those found for irradiation in the absence of added organic probe compounds, i.e., there is a significant quantity of NaCl which precludes conclusive evidence of Ag(0) formation. In this case there is even stronger evidence for NaCl formation as the Na Kα1,2 emission can now be seen quite clearly for point

3 (Figures 2.24 and 2.25).

Figure 2.25: EDX images corresponding to the points shown above in Figure 2.25 b. Vertical lines show the expected emission lines for Ag, Cl and Na (for the transition labelled)

2.3.5.3. STEM of irradiated AgCl(s) with formate

The results in the presence of formate are much more conclusive with regard to Ag(0) formation, in large part due to the absence of significant NaCl at many of the points sampled.

In this sample there are three quite clear NaCl crystals, however, the remainder of the particle is relatively free of any Na (Figure 2.26). One of the points for EDX was selected at one of the NaCl crystals, obtaining a strong Na signal, as expected (Figure 2.27). However, point 1 in this sample is quite interesting, as in the element maps it is apparent that there is a weakened Cl signal here relative to the remainder of the particle, yet a very strong Ag signal

(Figure 2.27 d, e). The EDX results further support this, showing a very strong signal for Ag at this point (relative to Cl). This is consistent with point 1 being elemental silver. Point 2 seems to be simply AgCl(s). In this particle there is strong evidence for the formation of

Ag(0) on the surface of an AgCl(s) particle.

Figure 2.26: TEM images of AgCl after irradiation in the presence of formate, a) element map, b) secondary electron image showing points analysed by EDX, c) TEM image, d), e), f), Ag, Cl and Na element maps, respectively

Figure 2.27: EDX images corresponding to points shown above in figure 2.27. Vertical lines show the expected emission lines for Ag, Cl and Na (for the transition labelled)

2.3.6. X-ray Diffraction Analysis

X-ray diffraction (XRD) results obtained indicated that there were essentially only two species in the system: AgCl and Ag, although the large uncertainties from this technique make it hard to determine whether or not the Ag portion of the sample has increased after irradiation, which was largely attributed to low quantity of the AgCl analysed (Figures 2.28,

2.29 and Table 2.3). None the less, it does provide further support for the notion that Ag(0) does indeed exist in this system.

Counts AgCl-2_exported Silver, syn 16.9 % Chlorargyrite, syn 78.9 % 10000 Silver Chlorate 4.2 %

5000

0 30 40 50 60 70 80 90

Position [°2Theta] (Copper (Cu))

Figure 2.28: The XRD graph of AgCl grown and collected in the dark.

Counts UV-Vis-2_exported Chlorargyrite, syn 87.7 % Silver, syn 12.3 %

0 0 2

15000

10000

0 2 2

1 1 1

5000

2 2 2

1 1 3

0 2 4

2 2 4

1 1 1

0 0 4

1 3 3

0 0 2

0 2 2

1 1 3

2 2 2

0 30 40 50 60 70 80 90

Position [°2Theta] (Copper (Cu))

Figure 2.29: The XRD graph of AgCl grown in the dark but irradiated 60 minutes’ with visible light afterwards.

Table 2.3: Quantitative XRD Analysis Results

Sample name AgCl a Ag Ag(ClO4) AgCl after irradiation 88 ± 10% 12 ± 21% 0 AgCl kept in the dark 79 ± 11% 17 ± 40% 4 ± 60% a values are given ± the estimated standard error

2.3.7 Electronic Absorption Spectra

The photolytic reaction of silver chloride derives from its light sensitivity, with the potential modes of light absorbance (and scattering) in this system described in §2.1. In order to better understand the photochemistry of silver chloride it is relevant to clarify how AgCl(s) colloid absorbs light and the implications of this absorption to photocatalytic reactions.

(Greene and Frizzell 1936) stated that illumination had no observable effect on the opalescence at a given stage of silver chloride development in their system. According to our observations however, the apparent absorbance of silver chloride colloids changed significantly during light irradiation. The variation of transmitted light through the AgCl(s) suspension before and after irradiation can be studied with a regular spectrophotometer with the absorbance related to its concentration provided certain conditions are satisfied according to the Beer-Lambert law:

–εLC Beer–Lambert law: I/I0 = 10

I = intensity of the radiation that has passed through the medium,

I0 = intensity of the incident radiation,

ε = molar absorptivity,

L = light path length,

C = molar concentration of the absorbing species,

A = εLC = log(I0/I)

A = absorbance.

Obviously the AgCl(s) colloidal suspension examined here did not satisfy the prerequisites of the Beer-Lambert law as the AgCl(s) particles scattered light (as was evident from the turbidity of the suspension). Additionally, the absorbing medium was not homogeneous. As such, the absorbance of the AgCl(s) suspension could not be related directly to its

concentration using the Beer–Lambert law. Despite this, the apparent absorbance of the

AgCl(s) suspension provides useful insights.

For example, the silver chloride absorbance changed over time and also changed with initial reagent concentrations. Because AgCl(s) particles scatter light, the apparent absorbance measured spectrophotometrically was the combined effect of particle scattering and absorbance. Based on the particle cluster size deduced by size measurement, two theories can be used to account for AgCl(s) particle scattering: Rayleigh scattering for particles much smaller than the wavelength of light with scattering ∝ λ−4 and Tyndall scattering for particles with a size of a similar magnitude as the wavelength of light with scattering ∝ λ−2. Scattering causes an apparent absorbance because less light reaches the detector (Owen 1995). As can be seen in Figure 2.30, an increase in initial Ag(I) concentration lead to a proportional increase in the absorbance, likely due to the proportional increase in scattering mass.

0.4 200 µM 0.35 100 µM 0.3 50 µM 0.25

0.2

0.15

0.1

AbsorbanceofAgCl(s) 0.05

0 300 400 500 600 700 Wavelength (nm)

Figure 2.30: The absorbance of AgCl(s) after aging 35 minutes with initial silver plus concentrations

– – are : 200 µM (solid line), 100 µM (round dot) and 50 µM (dash). [Cl ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0.

Similarly, an increase in absorbance is also observed if AgCl(s) is allowed to age for 35 minutes (in the dark) in the presence of different NaCl concentrations. As can be seen in

Figure 2.31, low concentrations display scattering that appears to obey Rayleigh scattering, but at concentrations of NaCl greater than 50 mM, the particles are too large for this law to hold, leading to the observed complicated scattering patterns.

0.4 25 mM 0.35 50mM 0.3 100mM 0.25 200mM 0.2

0.15

0.1 AbsorbanceofAgCl(s) 0.05

0 300 350 400 450 500 550 600 650 700 Wavelength (nm) Figure 2.31: The absorbance of AgCl(s) after aging 35 minutes with initial sodium chloride concentrations are : 25 mM (dash line), 50 mM (long dash dot), 100 mM (round dot) and 200 mM

(solid line). [Ag(I)]0 = 100 µM, [NaHCO3]0 = 2 mM, pH8.0.

As demonstrated earlier, at low chloride concentrations, the AgCl(s) particle size remains quite low (~150 nm) for extended periods of time; under these conditions Rayliegh scattering is expected to be operative. In Figure 2.32 below, the absorbance spectrum of 50 µM Ag(I) and 25 mM NaCl, which has been allowed to react in the dark, is shown on a log-log plot.

The sloping gridlines are plotted proportional to λ−4; the fact that the experimental spectra are parallel to these lines (at least at wavelengths greater than 300 nm) confirms that Rayleigh scattering is indeed the cause of the absorbance. There is no other feature apparent in these dark spectra.

When a similar low-chloride system (100 µM Ag(I), 1 mM NaCl) is allowed to age for 35 min and then irradiated (Figure 2.33), additional spectral features are observed after the commencement of irradiation. Although the initial absorbance can be attributed to Rayleigh

scattering, as expected, irradiation leads to rapid growth of peaks in the regions of ~390 nm and 500 – 550 nm, as well as a minima near 325 nm. These features would seem to be typical of those expected for either large (>200 nm) AgNPs or for a collection of smaller AgNPs in close enough proximity for coherent coupling to occur. These spectral changes are considered evidence for the formation of Ag(0) from AgCl(s) in this system.

Figure 2.33: The change in absorption spectrum upon irradiation of a 100 µM Ag(I), 1 mM NaCl, 2 mM NaHCO3, pH 8 solution which was first allowed to age in the dark for 35 minutes. Spectra are recorded every five minutes, with reaction progressing in the direction of the arrow in panel (a). Panel (b) shows the same data on a log-log plot, with the sloped gridlines proportional to λ−4 showing the expected slopes for Rayleigh scattering. Panel (c) show the difference spectra and are determined by subtracting the 35 minute dark-aged spectrum from the subsequent irradiated spectra.

When the concentration of chloride is increased, the rate of particle growth is significantly accelerated, resulting in particles with diameters on the order of 600 nm after 35 minutes

(Figure 2.10). Such particles are much too large to exhibit Rayleigh scattering for visible light with the observed absorbance spectra for dark-aged particles, although initially showing spectra typical of Rayleigh scattering until t ≈ 5 min, rapidly shifting instead to a broad maximum that shifts as the particles age (Figure 2.34). This absorbance structure greatly interferes with the region where AgNP absorbance would be expected to occur, and is also quite dynamic, which precludes the ability to observe any potential AgNP formation in such systems spectroscopically, even though AgNP is expected to occur (for example, the SEM and TEM data discussed previously was obtained from high-chloride systems).

2.4 Conclusion and Implications

The results from this Chapter indicate that:

1. Silver chloride aggregated over time and resulted in enlarged particle cluster size even

though zeta potential measurements yielded values suggesting a colloid system with

good stability.

2. The data showed that AgCl(s) rapidly aggregated (at rates expected of DLCA) when

NaCl concentrations were greater than ~100 mM, where all collisions result in

attachment and cluster growth was limited only by diffusion. This kind of aggregation

has been shown to lead to clusters with open structures and fractal dimensions (Df) of

~ 1.8 (Lin, Lindsay et al. 1990) compared to a relatively compact structure from a

slow RLCA (reaction-limited cluster aggregation) process. While compact AgCl(s) is

not particularly light sensitive (Glaus and Calzaferri 1999), the fact that DLCA

appears to predominate suggests that the AgCl(s) formed here will be relatively light

sensitive with resultant rapid photocatalytic reactions.

3. The absorbance of the AgCl(s) suspension does not always depend on the ion in

excess. When the concentration of Cl– ions was 103 times that of Ag(I), the

absorbance of AgCl(s) solution increased with increasing initial Ag(I) concentration

due to the increased concentration of AgCl(s) present.

4. The precipitation rate and size distribution of silver chloride can be controlled by

changing the initial Ag(I) and Cl– ion concentrations with the nature of the resultant

AgCl(s) influencing the photocatalytic properties of AgCl(s).

5. The microscopic analysis of the AgCl(s) nanoparticles suggested that photocatalytic

reaction changed the composition of the AgCl(s) particles and that silver

nanoparticles were likely generated. The silver chloride photolytic reaction also alteres the distribution of silver chloride particles in terms of size, shape and surface structure. As a whole, considering the EDS spectra of the four AgCl(s) specimens that were examined, the ratio of Ag to Cl elements increased in the order i) the specimen in the dark, ii) the specimen following visible light irradiation, iii) the specimen irradiated with formate and iv) the speciment irradiated with RhB with the increasing proportion of Ag compared to Cl attributed to the generation of Ag0 during the photochemical reaction.

Chapter 3. Generation of free chlorine and ROS on irradiation of AgCl(s) suspensions

3.1 Introduction

Silver/silver halides (Ag/AgX) based plasmonic material displays excellent photocatalytic performance under visible light irradiation. The metallic silver nanoparticles (AgNPs) formed on the surface of AgX can efficiently separate electron–hole pairs which enhances photocatalytic activity as well as the stability of AgX (Kakuta, Goto et al. 1999, Lanz,

Schurch et al. 1999, Jiang and Zhang 2011). Furthermore AgNPs exhibit strong UV–Visible absorption due to their plasmon resonance and hence can serve as an alternative type of sensitizer. Thus, recently, researchers have focused on use of Ag@AgX or Ag@AgX formed on the surface of other semiconductors (including AgBr/nanoAlMCM-41 by (Pourahmad,

Sohrabnezhad et al. 2010), Ag-AgI/Fe3O4@SiO2 by (Guo, Ma et al. 2011),

Fe3O4/SiO2/AgCl/Ag by (Zhang, Li et al. 2012)) with a smaller band gap for degradation of organic pollutants, NOx conversion, O2 evolution and sterilization and have shown that both photocatalyst loading as well as its morphology affects its photocatalytic ability (Lanz,

Schurch et al. 1999, Wang, Huang et al. 2008, Wang, Huang et al. 2008, Xu, Xu et al. 2011,

Miller, Yu et al. 2013).

Although the effectiveness of Ag@AgX has been demonstrated, few insights into the mechanism of degradation or factors important to optimize the process have been elucidated.

A number of investigators have reported that reactive oxygen species (ROS), including

•− • superoxide (O2 ), hydrogen peroxide (H2O2) and hydroxyl radicals (HO ) as well as free

– chlorine (Cl2, OCl and HOCl) are generated on photolysis of AgCl(s) and play a role in

degradation of organic pollutants (Zhou, Cheng et al. 2011, Tian and Zhang 2012, Wang, An et al. 2012, Zhou, Long et al. 2012, Xu, Xu et al. 2013, Yan, Kochuveedu et al. 2013, Zhu,

Wang et al. 2013) but little definitive information on either the mechanism or the kinetics of these reactions is available. In an attempt to fill this knowledge gap, we examine the

•− • generation of hydrogen peroxide (H2O2), superoxide (O2 ), hydroxyl radical (HO ), singlet

1 – oxygen ( O2) and free chlorine (Cl2, OCl and HOCl) from an irradiated AgCl(s) colloidal suspension under various solution conditions. This work will help in understanding the mechanism of photolytic production of ROS from AgCl(s) and will also provide insight into the role of ROS in degradation of organic pollutants by AgCl(s).

The photocatalyst, Ag@AgCl, used here was generated in situ by photochemical reduction of

AgCl(s) which was formed by addition of Ag(I) to solution containing excess NaCl. No other semiconductor was used as support in order to simplify the system to enable the study of reactions occurring due to the presence of AgCl(s) alone. Based on the experimental results obtained, we have determined the main reactive species produced on irradiation of AgCl(s) and also proposed a reaction mechanism for their generation as well as their likely role in degradation of organic pollutants.

3.2 Experimental Methods

3.2.1 Preparation of reagents

All reagent solutions were prepared using 18 MΩ.cm resistivity MilliQ water unless otherwise stated. All experiments were carried out in 2 mM NaHCO3 buffer containing specified concentrations of NaCl adjusted to pH 8.0 unless specifically stated. All pH values

were adjusted with 1M HNO3 and NaOH. A 6 mM N,Ndiethylpphenylenediamine (DPD;

Fluka Analytical) stock solution was prepared and stored in 50 mM H2SO4 solution (pH ≈ 1) to prevent DPD auto–oxidation. A 1.3 M stock solution of glycine was prepared in MilliQ water. An approximately 50 mM NaOCl stock solution was prepared by 20-fold dilution of concentrated sodium hypochlorite and was standardized using the Na2S2O3 titration method

(1998). The stock solution prepared usually yielded a concentration ~ 46.125 mM. A stock solution of 100 KU.L–1 horseradish peroxidase (HRP; Sigma) was prepared in MilliQ water and stored at −20°C prior to use. A 0.3 mM stock solution of singlet oxygen sensor green

1 (SOSG; Molecular Probes) was prepared for O2 measurement by first dissolving 100 µg

SOSG in 33.0 µL methanol (HPLC purity; Sigma) and quickly adding 520 µL MilliQ water.

The stock solution was stored in the freezer (≤ − 20 °C) until use with any thawed SOSG used within one day and the residue discarded. A 4 mM H2O2 stock solution was prepared weekly and was standardized by UV spectrometry (Morgan, Van Trieste et al. 1988). A stock solution of 0.55 mM phthalhydrazide (Phth; Sigma) was prepared in solution containing

100 mM NaCl and 2 mM NaHCO3 with the final solution pH adjusted to 8.0.

Since at pH 8, free chlorine mostly exists as OCl–, we have used OCl– to represent total free chlorine from here on in our discussion.

3.2.2 Photochemical experimental setup

Photochemistry experiments were performed in a water-jacketed 1 L glass reactor equipped with a quartz side window. The reactor was covered with aluminium foil to exclude light and reaction solution was maintained at 25±0.5°C with a recirculating water bath. A ThermoOriel

150 W Xe lamp (equipped with AM0 and AM1 filter to simulate the solar spectrum at the

Earth’s surface) was positioned horizontally adjacent to the quartz window to illuminate an

11 cm deep cross section of sample (Figure 3.1). The incident photon irradiance of the lamp is shown in Figure 3.2. Some experiments were carried out in 1 cm pathlength quartz cuvette and, in these instances, the quartz cuvette was positioned immediately in front of the lamp.

AgCl photolysis studies were typically undertaken for 60 minutes with samples continuing to be taken for analysis of products for a further 60 minutes in the dark after extinguishing the lamp.

Arc lamp Waterjacketed bottle Water bath

Quartz window

Figure 3.1: Experimental setup for irradiation of AgCl(s)

Figure 3.2: Incident spectral irradiance from the light source as function of wavelength.

3.2.3 Free chlorine measurement

Free chlorine was measured using the “DPD method” as described earlier (Bader,

Sturzenegger et al. 1988). In this method, DPD reacts with OCl– and produces the radical cation DPD•+ which exhibits an absorption peak at 551 nm with a molar extinction coefficient of 21,000 M1 cm1 (Bader, Sturzenegger et al. 1988) and a second peak at 510nm with molar extinction coefficient of 19,930 M –1 cm –1 (Gilbert 1981). This OCl– measurement method works well in the concentration range of 0.14 – 56 µM (Gilbert 1981).

– For measurement of OCl , 300 µL of 500 mM phosphate (NaH2PO4:Na2HPO4 = 3:1) buffer and 100 µL of 6 mM DPD stock solution were added to 2.6 mL of sample in a 1 cm quartz cuvette and the absorbance was measured at 553 nm using Cary UV–Visible

Spectrophotometer (Varian) after centrifugation to remove residual AgCl(s) particles (at

10,000 rpm for 30 seconds). The final pH of the solution was 6.3 at ~ 24 °C, which is within the ideal pH range (6.2 to 6.5) for the DPD colorimetric method (Bader, Sturzenegger et al.

1988). Calibration was performed by standard addition of NaOCl (in the concentration range of 0 to 6.0 µM) either before or after the sample measurement using the experimental procedure described above. A molar extinction coefficient of 24,930 M –1cm –1 was obtained at a wavelength of 553 nm which is slightly higher than the reported value of 21,000 M –1 cm

–1 (Bader, Sturzenegger et al. 1988) at 551 nm.

To account for DPD oxidation occurring due to the presence of other oxidants in our experimental matrix, a parallel experiment was designed, for every set of OCl– generation experiments, in which 24 mM glycine was added to our sample prior to addition of DPD since glycine converts OCl– instantaneously into chloroaminoacetic acid but has no effects on other DPD oxidants (1998). The absorbance observed in samples containing glycine corresponds to DPD•+ formed by all other oxidants except OCl–. Our data shows that these other oxidants account for about 10 – 50% of DPD oxidation observed in our experimental matrix, with the actual percentage varying as a function of solution conditions (such as the

– – concentrations of Cl , Ag(I), HCO3 and the pH).

3.2.4 Hydrogen peroxide measurement

The concentration of H2O2 generated on irradiation of AgCl(s) suspension was measured using DPD–based spectrophotometric technique similar to that reported by Bader and co– workers (Bader, Sturzenegger et al. 1988). Peroxidase from horseradish readily combines with hydrogen peroxide (H2O2) with the resultant HRP–H2O2 complex oxidizing DPD to

DPD•+ resulting in an increase in absorbance at 553 nm (Figure 3.3).

Figure 3.3 The conversion of DPD to its radical cation DPD•+ .

For measurement of H2O2, 24 µL HRP stock was added into a solution containing 300 µL of

500 mM phosphate (NaH2PO4:Na2HPO4 = 3:1) buffer, 100 µL of 6 mM DPD stock solution and 2.6 mL of sample in a 1 cm quartz cuvette and the absorbance measured at 553 nm using a Cary UV–Visible Spectrophotometer (Varian). Calibration was performed by standard addition of H2O2 stock solution to the experimental matrix used here using the method described above.

3.2.5 Singlet oxygen measurement

Singlet oxygen generated from photolysis of AgCl(s) was measured by a fluorescence technique using 8.0 µM SOSG as the probe. SOSG exhibits weak blue fluorescence with excitation peaks at 372 nm and 393 nm and emission peaks at 395 nm and 416 nm. In the

1 presence of O2, SOSG emits a green fluorescence with excitation and emission peaks at 504 nm and 525 nm, respectively, which can be readily detected (Gollmer, Arnbjerg et al. 2011).

For measurement, 3.0 mL of the aged AgCl(s) colloid suspension containing 8 µM SOSG was irradiated in a 1 cm quartz cuvette for 10 minutes. After irradiation, samples were

centrifuged in eppendorf tubes for 0.5 min at 10,000 rpm (which was tested to be able to successfully lower the absorbance of AgCl(s) colloid below 0.001) to remove residual AgCl(s) from solution. The fluorescence of the supernatant solution was then measured at an excitation wavelength of 504 nm and an emission wavelength of 525 nm using a Cary Eclipse fluorescence spectrophotometer (Varian Australia Pty Ltd). The instrument was zeroed with the same aged AgCl(s) solution containing SOSG following centrifugation but without irradiation.

1 – For calibration, O2 was generated from the reaction of OCl and H2O2 and the final concentration of fluorescent SOSG endoperoxide product (SOSG–EP) formed was obtained based on the output of a simple kinetic model with the following input reactions:

1 3 -1 -1 1, HOCl + H2O2 → O2 + HCl , k = 2.6 x 10 M s (Held, Halko et al. 1978)

1 5 -1 2, O2 → O2 , k = 2.4 x 10 s (Dalrymple, Carfagno et al. 2010)

1 8 -1 -1 3, O2 + SOSG → product , k = 4 x 10 M s (Gollmer, Arnbjerg et al. 2011)

3.2.6 Hydroxyl radical measurement

Hydroxyl radical generation from irradiation of AgCl(s) was measured using the phthalhydrazide method developed by Miller et al (Miller, Rose et al. 2011). For measurement, AgCl(s) was formed by adding 100 μM Ag(I) to solution containing 0.55 mM phthalhydrazide, 100 mM NaCl and 2 mM NaHCO3 at pH 8.0. After irradiation of AgCl(s),

1.2 mL of sample was added to 1.74 mL of 1 M Na2CO3 solution (at pH 11), centrifuged

(0.5 min at 10,000 rpm) and the supernatant removed for subsequent chemiluminescence measurement.

The stability of Phth-OH that may be formed under the conditions employed here was verified by using synthetic 5-HO-Phth standard. No change in Felume response due to presence of light and AgCl(s) (Figure 3.4) demonstrated that 5-HO-Phth was not degraded by

AgCl(s) on irradiation. Hence any HO• generated on photolysis of AgCl(s) can be detected using this method.

Figure 3.4: Chemiluminiscence signal after adding 0.5µM 5–HO–Phth into irradiated AgCl(s) sol for 60 minutes.

To determine whether or not 5–HO–Phth adsorbs significantly to the particles or if centrifugation removed the 5–HO–Phth product (i.e., if the chemiluminescent product was removed by adsorption to the solid particles during centrifugation), 5–HO–Phth was added to the reaction solution (after 1 hour irradiation) both before and after centrifugation and the

recovery of the added 5–HO–Phth determined. The results shown in Figure 3.5 indicate that the centrifugation process used here did not remove the chemiluminescent product.

The stability of Phth during experiments was verified by measuring its UV–Visible absorbance before and after irradiation.

3.2.7 DO measurement

The concentration of dissolved oxygen was measured using an Orion 081010 DO probe

(Thermo Fisher Scientific Inc.). Before use, the probe was calibrated according to the manufacturer’s instructions.

3.2.8 AgNP measurement

Generation of AgNP on irradiation of AgCl(s) was monitored spectrophotometrically by measuring the absorbance resulting from surface plasmon resonance with scans undertaken every 5-10 minutes during irradiation.

Figure 3.5: Chemiluminescent signal from irradiated AgCl(s) sample containing (a) 1 µM and (b) 50 nM 5–HO–Phth added before (open bar) and after centrifugation (closed bar).

3.3 Results and Discussion

3.3.1 Generation of AgNP

As observed in SEM images, AgCl(s) were mostly cubic and of similar size in the absence of irradiation with this observation consistent with the fact that the standard AgCl(s) crystal has a face–centered cubic NaCl structure with each Ag+ ion surrounded by an octahedron of six

Cl– ligands (Figure 2.11). On irradiation, smaller quasi–spherical nanoparticles in the range of 133 nm – 298 nm were deposited on the surface of AgCl(s) particles (sizes between 653 nm to 1.57 µm) with an increase in polydispersity of the overall particle sizes (Figure 2.15 and Figure 2.16). Further analysis in STEM images and EDS graphs (Figure 2.19 – 2.21) gave additional evidence for the formation of AgNP on visible light irradiation.

More evidence supporting the formation of AgNPs on irradiation of AgCl(s) is provided by the measured absorption spectra of AgCl(s) (Figure 3.6). As shown, two absorbance peaks are formed at 380 nm and 500 nm on irradiation which may correspond to the plasmon resonance of AgNPs. AgNPs usually exhibit plasmon resonance in the wavelength range 380

– 450 nm but the peak shifts to the red region with increase in particle size (Choi and Hu

2008). Thus it is possible that variation in size may cause formation of two peaks rather than one peak. The formation of AgNPs on the AgCl(s) surface is consistent with the findings of other studies in which Ag0 has been purported to be formed as a result of reduction of Ag(I) by photo–generated electrons on absorption of photon by AgCl(s) (Zhou, Cheng et al. 2011,

Tian and Zhang 2012, Wang, An et al. 2012, Zhou, Long et al. 2012, Xu, Xu et al. 2013, Yan,

Kochuveedu et al. 2013, Zhu, Wang et al. 2013). The possibility that the absorbance observed here is due to scattering of light caused by AgCl(s) is unlikely since the measured absorbance is corrected for absorbance caused due to nonphotolysed AgCl(s) particles. Furthermore, no

change in average AgCl(s) particle size was observed on irradiation (Figure 2.8) and hence no change in AgCl(s) absorbance is expected due to light scattering by the particles. If anything, irradiation of AgCl(s) particles would be expected to cause a decrease in absorbance as a result of AgCl(s) consumption rather than the increase observed here. Thus, in all likelihood, the increased absorbance in the visible wavelength region can be attributed to the formation of AgNPs.

Interestingly, no formation of an SPR peak is observed at high chloride concentrations (≥ 100 mM) though the absorbance decreases suggesting that at high chloride concentrations, an oxidative sink(s) of AgNPs is formed. Potential oxidizing sinks of AgNPs include OCl– and

1 reactive oxygen species such as O2 and H2O2 (Di, Jones et al. 2011). Alternatively, it is possible that the AgNPs formed undergo rapid aggregation in presence of high chloride concentration, thereby resulting in decrease in SPR peak absorbance. Earlier work done by Di et al had showed that unstabilized AgNPs are not stable at pH 8 (Di, Garg et al. 2012) and even the citrate-stabilized AgNPs undergo rapid aggregation with increase in chloride concentraion with diffusion–limited aggregation occuring for chloride concentration > 300 mM (Di, Bligh et al. 2013).

Figure 3.6: Change in absorbance of solution containing AgCl(s) in presence of (a) 1 mM and (b) 10 mM chloride on irradiation.

3.3.2 Generation of free chlorine on photolysis of AgCl(s) suspension

Irradiation of 1L suspensions of aged AgCl(s) produced micromolar concentrations of free

– – chlorine (Cl2 , HOCl and OCl ) (Figure 3.7). No OCl was generated in the dark or in solutions containing no silver and/or Cl– (data not shown), thereby supporting the conclusion that irradiation of AgCl(s) results in OCl– generation. Upon irradiation, the free chlorine concentration increased gradually with a rate closely related to the initial concentrations of

Ag(I) and Cl– present. The maximum concentration of free chlorine formed with 100 µM

Ag(I) was around 3.6 times higher than that in the presence of 50 µM Ag(I). On increasing

Ag(I) concentration from 100 µM to 200 µM, the free chlorine concentration after 60 minutes of irradiation increased by a factor of 1.6. While the dependence on initial Ag(I) concentration appears complex, an increase in concentration of AgCl(s) resulting from increased Ag(I) seems a reasonable explanation: the more AgCl(s) present, the more photo– generated hole–electron pairs would be expected upon light irradiation. Increase in total Cl– concentration also resulted in increase in OCl– concentration supporting the suggestion by other investigators that OCl– is most likely formed via oxidation of Cl– (Lanz, Schurch et al.

1999).

Figure 3.7: (a) Generation of free chlorine on irradiation of solution containing 100 mM chloride and 200 µM (squares), 100 µM (triangles) and 50 µM (circles) Ag(I) at pH 8.0; (b) Generation of free chlorine on irradiation of solution containing 100 µM Ag(I) and 200 mM (squares), 100 mM (diamonds), 50 mM (triangles) and 25 mM (circles) Cl– . Data represent average of duplicate measurements. Shaded region represents the measured free chlorine concentration after the lamp was extinguished.

In order to probe the mechanism of OCl– generation in irradiated AgCl(s) solution, a series of experiments were undertaken to determine the role of dioxygen, HO• and holes, as well as pH in the formation of OCl– .

3.3.2.1 Role of oxygen in free chlorine generation

– – In order to test the role of O2 in generation of OCl , we measured the generation of OCl in solutions sparged with argon containing 300 ppm CO2. As shown in Figure 3.8, decrease in

– O2 concentration led to a substantial decrease in the concentration of OCl generated

– suggesting that O2 is involved in OCl generation. One possible explanation is that O2 acts as an electron scavenger and hence prevents electron–hole pair recombination; with the free hole oxidizing Cl– ions to generate Cl0 atoms which subsequently hydrolyses to form OCl–.

Removal of O2 increases electron–hole pair recombination, thereby reducing the

– concentration of holes interacting with Cl . Furthermore, in the absence of O2, most of the electrons excited to the AgCl(s) conduction band will react with Ag+ to form AgNPs which may result in the decay of any OCl– formed, thereby decreasing the concentration of OCl– formed. It should be noted that while argon sparging effectively removes most of the oxygen present, residual oxygen (measured to be ~ 20 µM) remained.

Figure 3.8: Concentration of free chlorine generated after 60 minutes of irradiation of air-saturated (open bar) and partially deoxygenated (closed bar) solution containing 500 µM Ag(I) and 100 mM Cl– . Data represents average of duplicate measurements.

3.3.2.2 Role of holes or HO• in free chlorine generation

To verify that OCl– generation occurs by oxidation of Cl– by HO• or holes, we measured OCl– generation in the presence of excess bicarbonate ions, which is an excellent scavenger of holes or HO• (Buxton, Greenstock et al. 1988). As shown in Figure 3.9, addition of bicarbonate ions, substantially decreased OCl– generation supporting our hypothesis that OCl– is generated via oxidation of Cl– by holes or HO•.

Figure 3.9: Generation of free chlorine on irradiation of solution containing 100 mM chloride and 100

µM Ag(I) in presence of 2 mM (squares), 4 mM (triangles) and 10 mM (circles) NaHCO3 at pH 8.0. Data represent average of duplicate measurements. Shaded region represents the measured free chlorine concentration after the lamp was extinguished. The pH was controlled at 8.0 by addition of

HNO3 and/or NaOH.

As can be seen from Figure 3.10, the presence of different concentrations of NaHCO3 induced minimal effect on the absorbance of the AgCl(s) suspension over the wavelength region of interest with the extent of change unlikely to significantly influence the primary photon absorbance process and electron–hole pairs generation, further supporting the

– + • conclusion that the effect of HCO3 was related to h or HO scavenging with free chlorine generation originating from the oxidation of chloride ions to chlorine atom by holes or HO•.

Figure 3.10: Absorbance of aged AgCl(s) formed on addition of 100 µM Ag(I) to solution containing

- – 100 mM Cl at pH 8.0 in presence of varying HCO3 concentration.

3.3.2.3 Effect of pH on free chlorine generation

A change in pH may induce a number of effects including change in rates of key reactions and also change in key reactants and products. For example, the rate constant for disproportionation of superoxide (resulting in formation of H2O2) is strongly pH dependent, increasing from 50 M-1 s-1 at pH 11.0 to 4 х 104 M-1 s-1 at pH 8.0 and 1.3 х 107 M-1 s-1 at pH 4

(Bielski and Richter 1977). pH change may induce change in speciation, for example, HOCl exhibits an acidity constant of 10-7.53 and will thus exist predominantly in protonated form

(HOCl) below pH 7.53 and in anionic (OCl–) above this pH value. Carbonate speciation is

2– – also strongly pH dependent with CO3 dominant above pH 10.3, HCO3 dominant between pHs 6.3 and 10.3 and H2CO3 dominant below pH 6.3. These pH dependent changes in

2– speciation could influence product yield as CO3 is recognised to be a better hole and

– hydroxyl radical scavenger than HCO3 and H2CO3 (Buxton, Greenstock et al. 1988). To test the effect of pH on the generation of free chlorine from irradiated AgCl(s) suspensions, studies were undertaken at suspension pHs of 4.0 (0.1 mM nitric acid), 8.0 (2 mM NaHCO3 +

NaCl, adjusted with nitric acid) and 9.0, 10.0, 11.0 (2 mM Na2CO3 + NaOH).

As shown in Figure 3.11, increase in pH caused a decrease in rate and extent of OCl– generation from irradiation of AgCl(s) though the effect was relatively minor for pH up to 8 but more dramatic at higher pHs with chlorine generation ceased completely at pH 11.0.

Possibilities include the consumption of free chlorine by superoxide (Long and Bielski 1980) at the higher pH where superoxide is relatively stable or the preferential consumption of the oxidant produced on photolysis of AgCl(s) (presumably either holes or hydroxyl radicals) by carbonate rather than chloride ions given the predominance of carbonate at this high pH

(pKa2=10.3 for the carbonate system). This latter possibility is consistent with the proposal that holes or HO• are involved in OCl– generation with more oxidant scavenged at higher pH

– 2– due to increase in concentration of HCO3 and CO3 with increase in pH. Another possibility is that of change in size and thus photon absorption behaviour of AgCl(s) with pH, thereby affecting OCl– generation rate; however no effect of pH was found on AgCl(s) size (see

Figure 2.5).

Figure 3.11: Generation of free chlorine on irradiation of solution containing 100 mM chloride and 100 µM Ag(I) at pH 4 (closed circles), 8 (open diamonds) 9 (open squares), 10 (open triangles) and 11 (open circles). Data represents average of duplicate measurements.

3.3.3 Decay of free chlorine in dark

On cessation of photolysis, free chlorine generated on irradiation of AgCl(s) decayed with respect to a second–order reaction when the initial free chlorine concentrations are high (> 1

µM). A second order decay rate constant of 47.8±56 M –1s –1 was calculated from linear regression of measured 1/[OCl–] data versus time, as shown in Figure 3.12. This observation supports the conclusion that OCl– mostly decays via bimolecular dismutation, resulting in formation of Cl– and oxygen as suggested earlier (Schurch, Currao et al. 2002). The second– order decay rate constant is higher at lower initial OCl– concentration (when the decay started) suggesting that reactions other than bimolecular dismutation are important when OCl– is low

(Figure 3.12). Reaction with AgNPs which are generated on irradiation of AgCl(s) at low Cl–

concentration (where generated OCl– concentrations are low as well, can explain fast decay kinetics of OCl– in dark for low OCl– concentration observed here. HOCl decays rapidly in presence of AgNPs (see Figure 4.1 and Chapter 4 for more detailed consideration of the reaction between OCl– and AgNP). While this pathway of OCl– decay in dark is important at low chloride concentrations where AgNPs are generated; bimolecular dismutation reaction is important at high Cl– concentration where relatively little AgNPs are formed due to rapid aggregation, and OCl– concentration is high.

Figure 3.12: Plot of “1/[OCl–]” versus time where OCl– concentration was measured after extinguishing the lamp. OCl– was generated on irradiation of solution containing 50 µM Ag(I) and 100 mM Cl-(closed circles) , 100 µM Ag(I) and 100 mM Cl-(open squares), 200 µM Ag(I) and 100 mM Cl-(open circles), 100 µM Ag(I) and 200 mM Cl-(open diamonds) and 100 µM Ag(I) and 50 mM Cl-(open triangles) at pH 8.

3.3.4 Generation of hydrogen peroxide on photolysis of AgCl(s) suspension

As discussed above, O2 is an important scavenger of electrons excited to the AgCl(s) conduction band with resultant formation of superoxide and, following disproportionation of this reactive oxygen species, hydrogen peroxide. However, no measureable accumulation of

•+ H2O2 was observed. Indeed, the absorbance of DPD (corresponding to H2O2 concentration) was observed to decrease during the course of our experiments implying that the H2O2 concentration decreased from the ambient background level typically present in water (~ 30 nM).

This observation suggests that the H2O2 formed on irradiation of AgCl(s) decays rapidly due to its interaction with AgNP (Di, Jones et al. 2011) and/or OCl- thereby resulting in no net

– generation of H2O2. Though the decay rate of both H2O2 and OCl is expected to be similar due to their reaction with each other and AgNPs (note that the rate constant for reaction of

AgNPs with H2O2 is of similar magnitude to the rate constant for reaction of AgNPs with

– OCl ), the fact that no H2O2 (< 100 nM) was observed in our system while micromolar

- concentrations (~ 1–5 µM) of OCl were generated, suggests that the generation rate of H2O2 is at least an order of magnitude lower than that of OCl–. This is reasonable given that the

– concentration of Cl (>10 mM) is much higher than that of O2 (~0.26 mM). Furthermore,

Ag(I) and O2 will compete for photo-generated electrons further reducing the extent of O2 reduction during photolysis.

3.3.5 Generation of singlet oxygen on photolysis of AgCl(s) suspension

As suggested above, H2O2 generated on irradiation of AgCl(s) probably decays rapidly due to

– 1 its reaction with OCl . If this reaction indeed occurs, we would expect to observe some O2

1 generation since reaction of H2O2 and HOCl is known to generate O2 (Held, Halko et al.

1978);

1 3 -1 -1 HOCl + H2O2 → O2 + HCl , k = 2.6 x 10 M s

1 As shown in Figure 3.13, O2 is generated on irradiation of AgCl(s) with the concentration of

1 O2 decreasing in the presence of catalase, an enzyme known to catalyse the decay of H2O2.

1 The extent of generation of O2 is also reduced in the presence of glycine, a known scavenger

– – of OCl . These observations support our hypothesis that the reaction of OCl and H2O2 results

1 in formation of O2 in irradiated AgCl(s) solutions, where H2O2 is formed via superoxide disproportionation and OCl– is formed by oxidation of Cl–.

1 The concentration of O2 generated increased with increase in Ag(I) concentration in the presence of constant chloride concentration (Figure 3.14) which is reasonable given that generation of OCl– increases with increase in Ag(I) concentration.

Figure 3.13: Generation of SOSG epoxide on irradiation of 3 mL of solution containing 100 mM Cl– , 100 µM Ag(I) and 8 µM SOSG in absence (open bar) and presence of catalase (closed bar) or glycine (grey bar) for 10 minutes at pH 8. Data represents average of duplicate measurements.

3.3.6 Generation of hydroxyl radical on irradiation of AgCl(s)

As shown in Figure 3.15, no increase in the 5–Phth–OH chemiluminescence signal was observed during the 60 minutes of irradiation. Indeed, if anything, the signal dropped by more than 40 percent after 30 minutes of irradiation. After extinguishing the lamp, the signal oscillated a little but no substantial increase or decrease was observed.

Figure 3.15: Chemiluminiscence signal corresponding to Phth–OH on irradiation of solution containing 100 µM Ag(I) , 100 mM Cl- and 0.5 mM PhTh at pH 8.0.

These results suggest either that:

1. HO• is not formed in this system, or

2. HO• is formed in this system at a rate lower than 15 nM/hr (based upon a typical

detection limit using this method), or

3. The presence of Phth interferes with any potential HO• generation in this system

(e.g., HO• production is driven by UV absorbance of AgCl(s), which would be

significantly decreased in the presence of Phth due to the strong absorbance of

Phth in the UV region).

The most likely conclusion is that hydroxyl radicals are not generated to any significant extent on photolysis of AgCl(s) suspensions. Since, no HO• generation was observed in our experimental matrix, we can further conclude that holes are responsible for OCl– generation.

3.4 Discussion

3.4.1 Mechanism of photo-generation of ROS and free chlorine

– Based on the discussion presented above, the mechanism for photo-generation of OCl , H2O2 and AgNPs is shown in Figure 3.16.

Figure 3.16: Reaction mechanism for generation of ROS and free chlorine on irradiation of AgCl(s).

As shown, absorption of photon by AgCl(s) results in excitation of electron to the conduction band leaving a hole in the valence band. The photo-generated electron combines with Ag(I)

0 – 0 or O2 to form Ag atom or superoxide. The photo-generated holes oxidize Cl to form Cl

– which forms Cl2 and subsequently hydrolyses to form OCl . Superoxide disproportionates

– readily to form H2O2 which undergoes rapid decay due to its reaction with OCl and AgNPs.

– 1 The reaction of H2O2 and OCl results in formation of O2 which is a powerful oxidant and may cause rapid degradation of organic pollutants. We are currently undertaking work to study degradation of formic acid and Rhodamine B by AgCl(s) to determine the role of OCl–

1 and O2, the two main reactive species generated on AgCl(s) irradiation, in their degradation mechanism.

3.4.2 Kinetic model for generation of ROS and free chlorine on irradiation of

AgCl(s)

Based on the mechanism proposed above, we have developed a kinetic model to explain the generation of ROS and free chlorine on irradiation of AgCl(s) (see Table 3.1).

Table 3.1: Kinetic model to predict generation of ROS and OCl– on irradiation of AgCl(s)

Reaction Rate constant (M-1s-1 unless indicated) Used Published 1 ℎ휗 3×10-5 s-1 - AgCl(s) → ℎ+ + 푒− 2 1×1018 - ℎ+ + 푒− → 3 1×107 ℎ+ + Cl− → Cl0

H O 7 a 4 0 0 2 − − > 1×10 Cl + Cl → Cl2 → OCl + Cl 5 Ag(I) + 푒− → AgNP 1×107 − − 9 6 O2 + 푒 → O2 1×10 − − 4 4 e 7 O2 + O2 → H2O2 + O2 4×10 4×10 − − 10 10 f 8 AgNP + O2 → AgNP + O2 1×10 1×10 − − 8 8 10 f 9 AgNP + O2 → AgNP + O2 1×10 1×10 - 1×10 10 AgNP− + Ag(I) → AgNP + AgNP 1×108 11 AgNP + OCl− → Ag(I) + Cl− ~ 600b − − − c 12 OCl + OCl → O2 + Cl 47.9 f 13 H2O2 + AgNP → Ox 200 200 − 5 5 f 14 Ox + H2O2 → Ag(I) + O2 >1×10 >1×10 0 3 9 d 15 AgNP + AgNP → Agagg 1×10 - 1×10 a Not sensitive to model fits as long as the rate constant is higher than the pseudo-first rate constant for reaction 3 b roughly calculated from the data presented in Chapter 4 c calculated based on the measured decay of OCl– in dark for high OCl– concentration d varies with Cl– concentration e (Bielski, Cabelli et al. 1985) f (Di, Garg et al. 2012)

The main features of the kinetic model are discussed in detail below.

3.4.2.1 Photon absorption and instantaneous establishment of steady-state concentration of holes and electrons

Reaction 1 represents absorption of photon by AgCl(s) resulting in formation of electron in the conduction band and holes in the valence band. The holes and electron so formed can rapidly recombine (Reaction 2) thereby resulting in instantaneous establishment of steady- state concentration of holes and electrons.

The rate constant for reaction 1 was determined based on the best-fit to the measured OCl– data shown in Figure 3.7. The second-order rate constant for recombination of hole-electron pair was determined based on our experimental observations which showed that partial deoxygenation decreased OCl– generation rate. As described earlier, removal of dioxygen

– decreased OCl generation rate since O2 acts as an electron scavenger and hence prevents electron-hole pair recombination with the free hole oxidizing Cl– ions to generate Cl0 atoms which subsequently hydrolyses to form OCl–. This observation further suggests that the rate of electron-hole pair recombination is comparable to the rate of electron scavenging by dioxygen. Assuming, a diffusion-limited rate constant of 1×109 M-1s-1 for reaction of electrons and dioxygen, and that the steady-state concentration of electron and holes are comparable, we calculate that the rate constant for recombination of electron-hole pair has to be > 1×1018 M-1s-1 to explain our experimental observation.

3.4.2.2 Generation of free chlorine

Reaction 3 represents oxidation of Cl– by holes resulting in formation of Cl0 that combines to

– – – form Cl2 which subsequently hydrolyses to form OCl and Cl . The generation of OCl is

sensitive to the rate constant for reaction 3 and hence was determined based on best-fit to our experimental data. The rate constant for combination of Cl0 and subsequent hydrolyses has no effect on the model prediction as long as these reactions are faster than the rate of oxidation of Cl– by holes.

3.4.2.3 Generation of AgNPs and superoxide

Reaction 5 and 6 represent scavenging of photo-generated electrons by Ag(I) and O2, resulting in formation of AgNPs and superoxide respectively. Superoxide, so formed, can either disproportionate or reduce Ag(I) in the presence of AgNPs via charging-discharging mechanisms proposed by He et al (He, Jones et al. 2011) (Reaction 8-10). The rate constant for reaction 5 and 6 were determined based on best-fit to our measured OCl– generation data shown in Figure 3.7. The rate constant for superoxide decay via disproportionation reaction and/or charging-discharging of AgNPs, used here is consistent with previously reported values (Bielski, Cabelli et al. 1985, He, Jones et al. 2011, Jones, Garg et al. 2011).

3.4.2.4 Decay of OCl-

Free chlorine formed on irradiation of AgCl(s) decays either via its reaction with AgNPs

(Reaction 11) and/or second-order dismutation reaction (Reaction 12). The rate constants for reactions 11 and 12 were determined based on our measured decay rate of OCl– in irradiated

AgCl(s) solution in the dark. The rate constant for reaction 11 was determined based on the measured decay rate of OCl– in presence of citrate-stabilized AgNPs (Chapter 4). The rate

– constant for reaction 12 was determined by linear regression of measured 1/[OCl ]t versus

time data, in presence of high Cl– concentration, since the integrated rate law for second- order dismutation reaction is given by,

1 1 = + 2k t (1) [OCl- ] [OCl- ] disp t i

3.4.2.5 Aggregation of AgNPs

Reaction 15 shows the aggregation of photo-generated AgNPs. As shown earlier, in the presence of high Cl– concentrations, citrate-stabilized AgNPs aggregate rapidly with diffusion-limited aggregation occurring for Cl– concentration ≥ 300 mM (He, Bligh et al.

2013). Since the concentration of Cl– in our experimental matrix is very high and the AgNPs are not-stabilized, we expect that faster aggregation of AgNPs occurs than that observed earlier (He, Bligh et al. 2013). Aggregation of AgNPs in presence of high Cl- concentration is consistent with our observation that no AgNP SPR peak was observed in the presence of high

Cl– concentrations. The rate constant for the aggregation reaction was determined based on best-fit to our experimental data on OCl– generation and decay. A reaction showing two- particle collision resulting in aggregation probably does not represent the complex aggregation process completely however is simple and is also consistent with our experimental data and thus is used here.

As shown in the various figures throughout this chapter, a kinetic model based on the reactions shown in Table 3.1 and with rate constants as discussed above, performs well in describing the range of results presented here.

3.5 Conclusions and Implications

The current work uses unsupported AgCl(s) to study the generation mechanism of ROS and free chlorine on irradiation of AgCl(s). As shown, ROS including superoxide and hydrogen peroxide are generated on reduction of O2 by photo-generated electron while free chlorine is formed via oxidation of Cl– by photo-generated holes. Free chlorine so formed, although of interest, is unlikely to be important with respect to degradation of organic contaminants due to its low oxidative capacity. The more likely candidates for organic oxidation include Cl0

1 atom and/or O2 or the reactive intermediate formed on AgNP–H2O2 reaction. It is further possible that the role of these various oxidizing species vary with the solution conditions; for example, at low chloride concentration (≤ 1 mM) where we expect increased formation of

H2O2 and AgNPs, the reactive intermediate formed on AgNP–H2O2 reaction may be important while at high chloride concentration, Cl0 may play a role. Direct oxidation of organic pollutants by holes is also plausible. The degradation mechanism of organic contaminants will be investigated further in Chapters 5 and 6.

Chapter 4: Kinetics and mechanism of free chlorine decay in the presence of silver nanoparticles

4.1 Introduction

In chapter 3, we showed that the reaction of silver nanoparticle (AgNPs) and free chlorine

(OCl−) is important in controlling the rate of OCl− decay under both irradiated and non- irradiated conditions. This reaction may also be of significance in drinking water treatment processes since the disinfection efficiency of OCl− is expected to decrease in the presence

AgNPs, which are likely to be released to drinking water sources due to increasing use of

AgNPs in textiles, personal care products, food storage containers and even food supplements. The overall effect of AgNPs on the disinfection efficiency will be controlled by the rates of AgNPs aggregation, oxidative dissolution and interaction with OCl−. Thus, we investigated the kinetics and mechanism of the AgNP–OCl− reaction in detail in this chapter.

To the best of our knowledge this is the first investigation of this reaction.

4.2 Experimental Methods

4.2.1 Reagent Preparation

AgNP solution was prepared according to the method described earlier (Liu and Hurt 2010).

A 50 mM stock solution of trisodium citrate, sodium borohydride (NaBH4) and AgNO3 was prepared in Milli-Q water and stored at 4°C when not in use. A 59.8 mL solution containing

0.6 mM trisodium citrate and 0.4 mM NaBH4 was prepared in Milli-Q water and stirred vigorously in an ice bath. As 0.24 mL of 50.0 mM Ag(I) was added into the mixture the

solution turned greenish yellow, indicating formation of AgNPs. Following 3 hours of additional stirring at room temperature, soluble by-products were removed by centrifugal ultrafiltration (Amicon Ultra-15 3K, Millipore, MA) and MilliQ water addition in two cycles, after which the concentrations of AgNPs were characterized by inductively coupled plasma- optical emission spectrometry confirming a yield of ~100% after purification. The 0.2 mM

AgNPs stock suspensions were subsequently stored at 4°C for later use.

Sodium hypochlorite (6 – 14 % active chlorine; Sigma) was standardized using iodometric titration method (Clesceri, Greenberg et al. 1998) and was diluted in Milli-Q water to prepare a ~46 mM NaOCl working stock solution .

Experiments at pH 8 were performed in solutions containing 2 mM NaHCO3 buffer while experiments at pH 4 were carried out in solutions containing 0.1 mM HNO3. All experiments were performed in the presence of 100 mM NaCl in accord with the Cl– concentration used in

Chapter 3. The presence of Cl– can affect the aggregation and dissolution kinetics of AgNPs

(He, Bligh et al. 2013) as well as impact the important species of Ag(I) formed on oxidation of AgNPs by OCl–.

4.2.2 AgNPs measurement

Roughly spherical AgNPs absorb visible light between 400 – 450 nm (Mock, Barbic et al.

2002) and therefore visible spectrophotometry was used to characterize AgNPs. A Cary 50

(Varian Inc.) was employed to perform UV visible spectrophotometric scans over the wavelength range 320 – 700 nm. In the experimental system investigated here, the aggregation of AgNPs is negligible since no red- shift in AgNPs peak was observed which confirms that there is no change in particle size (see Figure 4.1 (a) and (b)), as the growth in

sizes of AgNPs is characterized by a shifting of the peak to larger wavelengths and by the widening of the SPR curve (Figure 2.1 and 2.2).

For measurement of AgNPs decay in the presence of OCl−, 3 mL of buffer solution

– – containing Cl and HCO3 was added to 1 a cm quartz cuvette and placed inside the UV-

Visible spectrophotometer, then appropriate volumes of OCl– and AgNP stock solution were added to the buffer solution simultaneously. A scanning kinetics program was used to measure the absorbance of AgNPs after addition over time for 20 minutes.

Figure 4.1: Measured absorbance of 8 µM AgNPs in absence (a) and presence (b) of 8 µM OCl- at pH 8.

4.2.3 Free Chlorine measurement

Concentrations of free chlorine were measured using the DPD method (Bader, Sturzenegger et al. 1988) as described in Chapter 3. For measurement of free chlorine decay in the presence of AgNPs, 2.7 mL of sample was withdrawn from the reactor at regular intervals, and added to a 1 cm quartz cuvette containing 0.3 mL of 0.5 M phosphate buffer and 0.1 mL of 6 mM

DPD stock solution and the absorbance was measured at 551 nm using a Cary UV-Visible

Spectrophotometer (Varian). Calibration was performed by standard addition of NaOCl (in the concentration range of 0 to 6.0 µM) either before or after the sample measurement using the experimental procedure described above.

4.3 Results and Discussion

4.3.1 Decay of AgNPs in the absence of OCl– at pH 8

As shown in Figure 4.2 AgNPs concentration decreases overtime due to oxidative-dissolution of AgNPs (eq.1) in the presence of Cl− as reported earlier (He, Bligh et al. 2013). Earlier studies have shown that oxygenation-dissolution of AgNPs is slow (He, Jones et al. 2011); however the presence of Cl− increases the rate of the dissolution reaction by inhibiting the back reduction of Ag(I) by superoxide via the charging-discharging reaction (eq.2-4) due to the formation of AgCl(s) (eq.5).

− AgNPs + O2 → Ag(I) + O2 eq.1

− ∗ AgNPs + O2 → AgNP + O2 eq.2

∗ − AgNP + O2 → AgNPs + O2 eq.3

AgNP∗ + Ag(I) → AgNPs + AgNPs eq.4

Ag(I) + Cl− → AgCls eq.5

As shown in Figure 4.2, the AgNP dissolution rate decreases over time with this effect most likely due to aggregation of AgNPs since decrease in specific surface area on aggregation would be expected to result in a decrease in the dissolution rate. However, no red-shift in

AgNPs absorbance was observed during the duration of our experiments (Figure 4.1) suggesting that aggregation of AgNPs does not occur in our experimental matrix. He et al.

(2013) recently reported that citrate-stabilized AgNPs undergo rapid aggregation in the presence of Cl− with diffusion-limited aggregation occurring for Cl− concentrations > 300 mM. Although, the Cl− concentration used here is comparable, the time-scales (~10-20 minutes) of our experiments are much smaller compared to the time-scale (~7 hours) of He et al. (2013). Hence, aggregation was considered unimportant here.

4.3.2 Decay of free chlorine in the absence of AgNPs at pH 8

Figure 4.3 shows the decay of OCl− in the absence of AgNPs. As shown, a small decrease in

OCl− concentration is observed in the first 30 seconds; however no change in OCl− concentration occurs for the next 30 minutes. This observation suggests that the uncatalysed self-decay of OCl− is unimportant here. We further suggest that the initial small decrease in

OCl− concentration observed here is probably due to the presence of some impurity in our experimental matrix. A readily oxidisable contaminant concentration of ~0.8 µM in the buffer solution would seem to be consistent with the approximately constant rapid, initial decrease in OCl− concentration in the absence of AgNPs (Figure 4.3(b)).

(a)

Figure 4.3: Decay in OCl− concentration in absence of AgNPs at pH 8. Initial OCl-concentration used here are 16 µM(open circles), 8 µM(open triangles), 4 µM(open diamonds) and 2 µM(open squares). Panel (a) shows the full decay, panel (b) shows the change in OCl− concentration.

4.3.3 Decay of AgNPs in the presence of OCl- at pH 8

Decay in AgNPs concentration in the presence of OCl– is shown in Figures 4.4 and 4.5. The extent of decay of AgNPs observed here is higher than that observed in the absence of OCl−, especially towards the later stages, with this observation supporting the conclusion that

AgNPs react with OCl–; however the initial rate of this reaction is slower than the oxygenation rate of AgNPs. The kinetics of AgNP decay shows two distinct phases with one possibly corresponding to oxygenation of AgNPs while the second phase corresponds to the oxidation of AgNPs by OCl−, with oxygenation of AgNPs important at low OCl− concentration while oxidation by OCl− is important at high OCl- concentration.

Based on the AgNP decay data shown here, it is difficult to determine the stoichiometry and mechanism of the AgNP–OCl− reaction, given that a significant proportion of the AgNPs present undergo oxygenation.

100

Figure 4.5: Decay of 8 µM AgNP in presence of 2 µM (open diamond), 4µM (open squares), 8 µM (open triangle) and 16 µM (open circle) free chlorine at pH 8.0.

The slow-down of AgNP decay at 16 µM OCl– suggests that there is an oxidation process, and also a secondary process driven by hypochlorite that somehow leads to the stabilisation of AgNPs.

4.3.4 Decay of free chlorine in the presence of AgNPs at pH 8.0

The decrease in OCl− concentration in the presence of AgNPs is shown in Figures 4.6 and

4.7. The decay rate of OCl– increases with increase in the initial concentration of AgNPs.

Although, as suggested earlier, minimal self-decay of OCl− was observed in the absence of

AgNPs, it is likely that the dismutation of OCl− becomes important, especially towards the later stages of our experiments, since Ag(I) (which is formed on oxidation of AgNPs) is known to catalyze the bimolecular dismutation of OCl−. Decay of OCl− occurring via bimolecular dismutation is supported by the observation that the decrease in OCl− concentration far exceeds (≤ 7 fold) the initial concentration of AgNPs used, thereby supporting a pathway for OCl− decay independent of AgNPs.

101

Figure 4.7: Decrease in OCl− concentration in presence of 0.2 µM (open squares), 0.4 µM (open circles), 0.8 µM (open diamonds), 1.6 µM (open triangles) and 2.4 µM (closed squares) AgNPs at pH 8.0.

102

4.3.5 Mechanism of AgNP and OCl− reaction at pH 8

The decrease in AgNP concentration observed here occurs either via oxygenation or as a result of AgNP reaction with OCl−. Similarly, decay of OCl− occurs due to its reaction with

AgNP as well Ag(I)-catalysed bimolecular dismutation, where Ag(I) is formed via oxidation of AgNPs. Although, the exact stoichiometry and mechanism of the AgNP-OCl− reaction is not clear from the data presented here due to presence of competing reactions, a possible reaction mechanism is as follows:

2 H+ − ∙ AgNPs + OCl → Ag(I) + Cl + H2O eq.6

퐴g(I) ∙ ∙ Cl + Cl → Cl2 eq.7

H O 2 − − Cl2 → OCl + Cl eq.8

4.3.6 AgNP decay in absence of HOCl− at pH 4.0

As shown in Figure 4.8, the AgNP decay rate in the absence of HOCl at pH 4.0 is the same as that observed at pH 8 (Figure 4.2). This observation suggest that oxygenation-dissolution of

AgNPs is not influenced by the solution pH.

103

5 AgNPs concentration (µM) concentration AgNPs 0 0 250 500 750 1000 1250 Time (s)

4.3.7 AgNPs decay in presence of HOCl at pH 4

Decay of 8 µM AgNPs in the presence of HOCl is shown in Figure 4.9. As shown, the AgNP decay rate increases with increase in initial HOCl concentration. The initial decay rate of

AgNP is much faster at pH 4.0 than at pH 8.0. Furthermore, the AgNP-HOCl(OCl–) reaction is more sensitive to free chlorine concentration change at pH 4.0 where HOCl is the dominant species suggesting that HOCl is more reactive than OCl-.

104

10 0.33 µ M 0.67 µ M 1.33 µ M 2.67 µ M

5 AgNPs concentration (µ M) concentration AgNPs

0 0 200 400 600 800 Time (s)

Figure 4.9 Decay of 8 µM AgNP in presence of 0.33 µM (open diamonds), 0.67 µM (open squares) and 1.33 µM (open triangles) and 2.67 µM (open circles) at pH 4.0.

4.4 Conclusion

The data presented here shows that AgNPs undergo rapid oxygenation-dissolution at both pH

4 and 8. The presence of free chlorine increases the dissolution rate of AgNPs albeit slowly.

At pH 8, increasing concentrations of HOCl lead to more rapid initial AgNP dissolution, although the nature and stoichiometry of this process is not clear. At higher concentrations of

OCl−, although the initial oxidation rate is accelerating, over time an apparent steady-state is reached suggesting that there is also some (unknown) OCl−-driven process that leads to

AgNP stabilization. The oxidation of AgNPs by HOCl/OCl− is greatly accelerated at the lower pH, suggesting that HOCl is more reactive than OCl−.

105

Chapter 5: Degradation of formic acid in irradiated

AgCl(s) suspensions

5.1 Introduction

Organic contaminant elimination from aqueous streams represents a major challenge, particularly as some organic pollutants have proven difficult to degrade with traditional treatment methods. For example, phenolic compounds and azo dyes are highly toxic and are resistant to biological degradation in an ordinary secondary treatment plant. Advanced oxidation processes are typically needed for degradation of such recalcitrant contaminants with particular interest shown recently in the use of AgCl(s)-containing materials given their ability to induce degradation on illumination with visible wavelength light (Andersson,

Birkedal et al. 2005, Xu, Xu et al. 2013, Yan, Kochuveedu et al. 2013, Sun, Zhang et al.

2014, Wang, Ming et al. 2014, Yu, Miller et al. 2014). Examples of recent reports of the ability of AgCl(s)-containing materials to degrade contaminants are given below.

 Xu and co-workers synthesized Ag@AgCl core-shell nanocomposites which

decomposed 96% of methyl orange (MO) dye in 40 minutes and decomposed 81.8%

of 4-chlorophenol in 30 minutes under visible light irradiation (Xu, Xu et al. 2013).

 Guo and co-workers prepared the composite photocatalyst Ag/AgCl/BiMg2VO6 by

deposing Ag/AgCl nanoparticles on a BiMg2VO6 substrate and subsequently

examined the ability of this material to photodegrade acid red G. These investigators

reported that 95.90% of the dye had degraded after 120 minutes of visible light

irradiation (Guo, Zhang et al. 2013).

106

 Similar observations have been reported for Ag/AgCl/ZnO - mediated degradation of

dye pollutants (Xu, Xu et al. 2011), Ag/AgCl - mediated degradation of MO (Xu, Li

et al. 2011) and Ag@AgCl/TiO2 nanotube-mediated degradation of methylene blue

dye (Wen and Ding 2011).

While AgCl(s)-based materials have been of particular focus, there is now ample evidence that, more generally, silver halides (AgX) are more capable of effectively harnessing visible light energy than traditional semiconductor photocatalysts such as TiO2, ZnO, WO3 and CdS.

AgX-based materials have been used to induce degradation of a range of contaminants, including:

 A range of dyes including (1) Rhodamine B (Jiang and Zhang 2011, Xiong, Zhao et

al. 2011, Wang, Bai et al. 2012, Ye, Liu et al. 2012, Adhikari, Gyawali et al. 2013,

Wang, An et al. 2013, Zhang, Tian et al. 2013, Zhu, Wang et al. 2013); (2) Methyl

orange (MO) (Feng, He et al. 2012, Wang, An et al. 2012, Xu, Song et al. 2012, Yi,

Du et al. 2012, Zhao, Yang et al. 2012, Zhou, Long et al. 2012, Fan, Zhu et al. 2013,

Wang, Du et al. 2013, Xu, Xu et al. 2013); (3) Azo dyes (acid orange 7 (AO7), acid

blue 92, acid orange 7, acid red G); (4) Methylene blue (Sun 2010, Sohrabnezhad and

Pourahmad 2012, Zhang, Liu et al. 2013, Zhu, Wang et al. 2013) and a variety of

other dyes (Chen, Yoo et al. 2012, Padervand, Salari et al. 2012, Yang, Zhang et al.

2012, Ye, Liu et al. 2012, Dong, Tian et al. 2013, Guo, Zhang et al. 2013, Hou, Wang

et al. 2013).

 A variety of chlorinated phenolic compounds including 4-chlorophenol (Zhou, Cheng

et al. , Xu, Xu et al. 2013), pentachlorophenol and others (Tang, Subramaniam et al.

107

2011, Guo, Ma et al. 2012, Ma, Guo et al. 2012, Yang, Zhang et al. 2012, Yu, Liu et

al. 2012, Dong, Tian et al. 2013, Fan, Zhu et al. 2013).

Escherichia coli (E. coli) (Wang, Tang et al. 2012, Rehan, Hartwig et al. 2013).

 Other aromatic compounds (Zhao, Kuai et al. 2012).

Although the effectiveness of these materials in inducing contaminant photodegradation is evident, no mechanism for the degradation process has as yet been confirmed though a variety of plausible mechanisms have been suggested (Zhang, Fan et al. 2011, Zhou, Cheng et al. 2011, Zhu, Chen et al. 2012, Zhu, Wang et al. 2013). While there are obvious disagreements with regard to how the photodegradation process proceeds, it is generally agreed that the degradation of pollutants is associated with the presence of noble metal Ag nanoparticles which improve electron-hole pair separation and, because of surface plasmon resonance behaviour of Ag nanoparticles, increased absorbance within the visible spectral range. There is a particular lack of information on the identity of the oxidising species inducing degradation, for example:

 Zhou and colleagues suggested that the main active species in degrading methyl

•– orange with Ag–AgCl/BiVO4 were O2 and holes after ruling out the possibility of

HO• using isopropanol and methanol as hydroxyl radical scavenger and getting little

influence on the photocatalytic activity. (Zhou, Long et al. 2012);

0 •– •  Xu et al suggested that Cl , O2 , H2O2 or HO could contribute to organic pollutant

oxidation but did not provide any justification for a role of any particular oxidising

species (Xu, Xu et al. 2013);

 Zhu et al implied that chlorine atoms were the most likely species responsible for

oxidation of R6G dye (Zhu, Liu et al. 2012).

108

A major problem with the above studies relates to the complexity of the systems under investigation. Dyes are used as target species in many of the investigations but dyes represent a rather unique contaminant in that they are themselves strong chromophores, with this property complicating both mechanistic interpretation and rendering the study a “special case”. Other contaminants such as chlorinated phenolics are certainly of environmental relevance but the production of oxidant scavenging intermediates and end products complicates interpretation considerably. For this reason, we have chosen to use the simple target organic compound formic acid which, on oxidation, degrades to CO2 thereby minimising complications with oxidant scavenging intermediates. Formic acid degradation is also of interest in its own right as this compound is used in fabric dyeing, printing processes and as a food additive. It also widely used in laboratories as a buffering agent in HPLC analysis.

In this chapter, the results of a range of studies on photo–degradation of formic acid in irradiated AgCl(s) solution are presented. Various experiments are performed to determine the role of free chlorine, hydrogen peroxide, singlet oxygen, dioxygen, hydroxyl radicals, holes, silver nanoparticles and the oxidizing intermediate formed on AgNP-H2O2 reaction in the degradation mechanism. Based on our experimental observations, we have proposed a mechanism of formic acid degradation in the presence of irradiated AgCl(s) and have also developed a kinetic model that is capable of predicting formic acid degradation rates under various solution conditions.

109

5.2 Experimental Details

5.2.1 Reagent preparation

Radioactive C–14 labelled sodium formate (H14COONa) at a final concentration of 0.1 µM was added to all solutions. A Liquid Scintillation Counter (LSC) (PerkinElmer) was used for analysis of beta energy emission from C–14 after mixing the aqueous samples with scintillation fluid (Beckman Coulter USA). Silver nitrate, sodium chloride and all other reagents solutions were prepared and treated as described in Chapter 3.

5.2.2 Photochemical experimental setup

Photochemistry experiments were performed in a water-jacketed 0.5 L glass reactor. The reactor was covered with aluminium foil to exclude light and reaction solution was maintained at 25±0.5°C with a recirculating water bath. A 300 W Xe arc lamp (Perfectlight

Inc, Beijing) equipped with a UV-cutoff filter (λ>400 nm) was positioned vertically above the reactor to irradiate the sample. The incident photon irradiance of the lamp is shown in

Figure 5.1.

110

Figure 5.1: The spectrum of incident irradiance for this study.

For formic acid degradation experiments, AgCl(s) was formed by addition of an appropriate volume of 50 mM Ag(I) stock solution to 200 mL of buffer solution at pH 8.0 containing 2 mM NaHCO3 and NaCl at particular concentrations. AgCl(s) so formed was continuously stirred and aged for 35 minutes after which radiolabelled and non-labelled sodium formate were added such that final concentrations of 0.1 µM H14COOH and 0.9 µM H12COOH were achieved and then irradiation was commenced. 1 mL of sample was withdrawn from the reactor every 5 or 10 minute for formic acid measurement.

5.2.3 Formic acid measurement

Samples for formic acid measurement were withdrawn from the reactor and immediately mixed with 4 mL of 10 mM HNO3, 3 mL of which was then centrifuged at 10,000 rpm for 1 minute in two 1.5 mL eppendorf tubes. Then 2 х 1.0 mL of the supernatant from the eppendorf tubes was removed and placed in scintillation vials, which was then sparged with

99+% argon to drive out any residual CO2 from the solution. Subsequently, 5 mL scintillation

111

fluid was added to the scintillation vial containing 1 mL of sample, thoroughly shaken and then was placed in the liquid scintillation counter for measurement. A calibration curve was obtained using the procedure described above prior to experiments.

More detailed description of the experimental procedure used for various control and quantitative experiments measuring formic acid degradation is provided below.

5.2.3.1: Effect of free chlorine and hydrogen peroxide addition

− To determine the role of OCl and H2O2 in the degradation of HCOOH, appropriate volumes

− of OCl or H2O2 stock solutions were added to the buffer solution containing AgCl(s) and 1

μM formic acid and then samples were withdrawn and analysed as described above.

5.2.3.2: Role of singlet oxygen

1 1 To measure the degradation of formic acid in the presence of O2, O2 was generated through

− the reaction of OCl with H2O2 (Held, Halko et al. 1978) in situ in buffer solution containing

1.0 µM formic acid at pH 8.0. H2O2 at final concentration of 2 mM was added first followed

− 1 by addition of 1 mM OCl for generation of O2.

5.2.3.3: Effect of dioxygen removal

The effect of partial removal of dioxygen on formic acid degradation was investigated by sparging the buffer solution containing AgCl(s) and formic acid with argon containing 300 ppm CO2 for 2 hours prior to irradiation (with sparging continuing during the entire degradation experiment).

112

5.2.3.4: Role of superoxide

To determine the role of superoxide, which is formed on irradiation of AgCl(s), in HCOOH degradation, 20 kU.L−1 of SOD was added to the buffer solution containing AgCl(s) and formic acid prior to irradiation.

5.2.3.5: Role of oxidizing intermediate formed on AgNP and H2O2 reaction

As shown in Chapter 2, AgNPs are generated on irradiation of AgCl(s). Based on the effect

•– of oxygen on free chlorine generation it is reasonably followed that O2 was also generated by reaction with conduction band electrons of AgCl(s). Superoxide quickly disproportionates to form H2O2, which is known to form an oxidative intermediate upon reaction with AgNP

(He, Jones et al. 2011) Although no evidence for H2O2 production could be found in Chapter

2, it may be conceivable, although unlikely, that this is because any H2O2 that does form reacts to form this oxidative intermediate prior to diffusion to the solution. Hence it is possible that the oxidizing intermediate formed on AgNP-H2O2 reaction (represented by OX here) may play a role in formic acid oxidation. To investigate if OX is involved in formic acid degradation, appropriate volumes of AgNP (prepared using the method described in

Chapter 3) and H2O2 stock solution were added to the buffer solution containing HCOOH and

AgCl(s) and the degradation of HCOOH examined.

5.3 Results and Discussion

5.3.1 Formic acid degradation

Figures 5.2 and 5.3 show the degradation of formic acid in the presence of irradiated AgCl(s) solution. Either no or very small degradation of formic acid was observed in solution

113

containing no Ag(I) and/or no Cl− suggesting that the presence of AgCl(s) is required for degradation of HCOOH. Furthermore, no degradation was observed in non-irradiated

AgCl(s) solution (Figure 5.3) suggesting that the HCOOH oxidant is generated on photolysis of AgCl(s).

As shown in Figure 5.2, increase in initial total Ag(I) concentration resulted in an increase in the degradation rate of HCOOH until [Ag(I)]0 ≈ 200 µM, above which the degradation rate of HCOOH did not increase significantly. Increase in total Ag(I) concentration increases the concentration of AgCl(s) formed which is apparently responsible for generation of the oxidant involved in formic acid degradation.

− Figure 5.2: Degradation of 1µM formic acid in irradiated 2 mM HCO3 solution containing 100 mM Cl− and 0 (open squares), 25 µM (open triangle), 50 µM (open diamond), 100 µM (open circle), 200 µM (closed squares) and 400 µM (closed triangles) Ag(I) at pH 8. Data represents averages of duplicate measurements and lines represent model values.

114

Figure 5.3 shows the effect of initial Cl− concentration on HCOOH degradation. Small but measureable HCOOH degradation is observed in the absence of Cl− which suggests that dissolved Ag(I) can oxidize HCOOH at a very slow rate, similar to the observation of silver acetate reacting with sodium formate by Arthur A. Noyes and George T. Cottle (Noyes and

Cottle 1898). The degradation rate of HCOOH decreases with increase in Cl− concentration suggesting that Cl− competes with HCOOH for the oxidant. One possible such oxidant are photo-generated holes which, as shown in chapter 3, are scavenged by Cl– to form Cl0.

Decrease in HCOOH degradation rate with increase in Cl− concentration further suggests that

Cl0 and/or OCl− is not an important oxidant for formic acid since increase in Cl− concentration results in increase in the concentration of Cl0 formed at the AgCl(s) surface

(see Chapter 3). The conclusion that OCl− plays a limited role in HCOOH degradation is further confirmed by the fact that little HCOOH was degraded in solutions containing excess

OCl− (Figure 5.4). As shown in Figure 5.4, even when the free chlorine concentration was raised to 25 µM, only around 9% of HCOOH was degraded in 2 hours. This is very slow compared to the degradation rate observed in the presence of irradiated AgCl(s) suspensions

(Figures 5.2 and 5.3).

115

− Figure 5.3: Degradation of 1µM formic acid in irradiated 2 mM HCO3 solution containing 100 µM Ag(I) and 0 (open squares), 200mM (open triangle), 100mM (open diamond), 50 mM (closed circle), 25 mM (open circles), 10 mM (closed triangles) and 2 mM (closed squares) Cl−at pH 8. Data represent averages of duplicate measurements and lines represent model values.

− Figure 5.4: Degradation of 1µM formic acid in 2 mM HCO3 solution containing 100 µM Ag(I) and 100 mM Cl− at pH 8 in absence (open triangles) and presence (open squares) of 25 µM OCl−. Data represents the average of duplicate measurements.

116

The results of various experiments investigating the role of holes and ROS in degradation of formic acid are presented below.

5.3.1.1 Role of hydrogen peroxide

As shown in Figure 5.5, no HCOOH degradation was observed in the presence of 10 mM

H2O2, suggesting that H2O2 cannot directly oxidize HCOOH.

− Figure 5.5: Degradation of 1µM formic acid in 2 mM HCO3 solution containing 100 µM Ag(I) and

– 100 mM Cl at pH 8 in the absence (open triangles) and presence (open squares) of 10 mM H2O2. Data represent averages of duplicate measurements.

5.3.1.2 Role of singlet oxygen

As shown in Figure 5.6, no degradation of HCOOH was observed in a solution containing

− 1 H2O2 and OCl , which is known to generate O2 (Held, Halko et al. 1978), thereby supporting

1 the conclusion that O2 is not involved in HCOOH oxidation. The lack of degradation of

117

1 1 HCOOH in D2O solution in which O2 is longer-lived (Turro 1991) further confirms that O2 does not degrade HCOOH (Figure 5.7).

− Figure 5.6: Degradation of 1µM formic acid in 2 mM HCO3 solution containing 2 µM H2O2 and 1 µM OCl– at pH 8. Data represent the average of duplicate measurements.

Figure 5.7: Concentration of HCOOH remaining after 110 minutes in the presence of 1.0 µM OCl– and 2.0 µM H2O2 in aqueous (open bar) and deuterium oxide (closed bar) solutions at pH 8.

118

5.3.1.3 Role of oxidizing intermediate formed on AgNP-H2O2 reaction

As shown in Figure 5.8, no degradation was observed in solution containing citrate-stabilized

AgNPs and H2O2, confirming that the oxidizing intermediate formed from the interaction of

AgNP and H2O2 plays no role in HCOOH degradation. This is reasonable given that AgNPs undergo relatively rapid oxidative dissolution in the presence of high Cl– concentrations at the pHs utilised here and, as such, the extent of generation of any oxidizing intermediate might be expected to be limited. Although, the oxidizing intermediate may not be an important HCOOH oxidant at high chloride concentrations, it may play some role in the presence of low Cl− concentrations since AgNP aggregation/dissolution will be very slow at

− low Cl concentration and hence most of the H2O2 formed will react with AgNP to form the oxidizing intermediate.

− Figure 5.8: Degradation of 1µM formic acid in 2 mM HCO3 solution containing 10 mM H2O2 and 20 µM AgNPs at pH 8. Data represent averages of duplicate measurements.

119

5.3.1.4 Role of superoxide

As shown in Figure 5.9 the spiked 20 kU.L–1 SOD quenched formic acid photo–degradation completely. This result may indicate either i) that superoxide, which would be expected to

- form on reaction of ecb with oxygen, plays a role in formic acid degradation or ii) that SOD outcompetes formic acid for holes (the likely major formic acid oxidant). With regard to the first possibility, superoxide is recognised to be capable of reducing Ag(I) to Ag0 thus, removal of superoxide may decrease the rate and extent of Ag0 formation (He, Jones et al.

2011, Jones, Garg et al. 2011). A reduced rate of Ag0 formation may, in turn, result in a

- + higher rate of ecb - hvb recombination and thus decreased rate of formate degradation. It is

+ more likely however that SOD simply competes with formic acid for hvb generated at the

AgCl(s) surface (the second option mentioned above).

Figure 5.9: Degradation of 1µM formic acid in irradiated AgCl(s) solution in the absence (solid triangles) and presence (open triangles) of 20 kU.L−1 SOD at pH 8.0. Data represent averages of duplicate measurements.

120

5.3.1.5 Role of dioxygen

As shown in Figure 5.10, removal of dioxygen decreased the rate of degradation of HCOOH.

This observation supports the conclusion that holes are involved in HCOOH oxidation since, as shown in chapter 3, dioxygen acts as an important electron scavenger and hence reduces the rate of electron-hole pair recombination, thereby increasing hole-availability. Decrease in dioxygen concentration decreases the hole availability thereby decreasing HCOOH degradation.

121

5.3.2 Mechanism of formic acid degradation in the presence of irradiated AgCl(s)

Based on the discussion presented above, we suggest that holes are the main oxidant of

HCOOH in irradiated AgCl(s) solution. The availability of holes depends on the dioxygen concentration since the presence of dioxygen reduces the rate of electron-hole pair recombination. The kinetic model presented in Chapter 3 is extended to predict HCOOH degradation rates under the experimental conditions used here with inclusion of oxidation of

HCOOH by photo-generated holes (see Table 5.1). Below we present details of the reactions used in the kinetic model:

Table 5.1: Reactions showing degradation of HCOOH in presence of irradiated AgCl(s)

Reaction Rate constant (M-1s-1 unless indicated) Used Published

9 16 + ∙- + 7×10 - HCOOH+ h → CO2 + 2 H 17 1.4×109 (Flyunt, Schuchmann CO∙− + CO∙− → Int 2 2 et al. 2001) 18 1×107 (Flyunt, Schuchmann Int +2 H+ → HCOOH + CO 2 et al. 2001) 19 Ag(I) 0.5 (Noyes and Cottle HCOOH → CO 2 1898) 20 ∙ 1×104 Ox + HCOOH → CO2 + Ag(I)

5.3.2.1 Degradation of HCOOH by photo-generated holes

Reactions 16-18 represent the degradation of HCOOH by photo-generated holes. As shown, photo-generated holes oxidize HCOOH to carboxyl radical which then forms a short-lived carbanion intermediate which subsequently decays to form HCOOH and CO2. The rate constant for reaction 17 and 18 were used as reported previously (Flyunt, Schuchmann et al.

2001) while the rate constant for reaction 16 was determined based on best-fit model results.

122

5.3.2.2 Degradation of HCOOH by Ag(I)

As shown in Figure 5.3, small concentration of HCOOH was degraded in irradiated solution containing Ag(I) only, hence we have included a reaction in our kinetic model to account for this observation (reaction 19). The mechanism of Ag(I) mediated oxidation is as reported by

(Noyes and Cottle 1898), the rate of which is proportional to one molecule of formate (or of its anion) and two molecules of silver acetate (or of the silver ion). A simple reaction resulting in oxidation of HCOOH to carbon dioxide explains our experimental observation very well. The rate constant for this reaction was determined based on best-fit to our experimental results.

5.3.2.3 Degradation of HCOOH by oxidizing intermediate formed on AgNP-H2O2

reaction

To account for faster degradation kinetics of HCOOH observed at lower Cl− concentration, we have included oxidation of HCOOH by oxidizing intermediate formed on AgNP-H2O2 reaction (reaction 20). This is reasonable given that AgNP aggregation will be very slow at

− low Cl concentration and hence most of the H2O2 formed will react with AgNP to form the oxidizing intermediate. The rate constant for reaction 20 was determined based on best-fit to our experimental data.

As shown in Figure 5.2 and 5.3, the kinetic model explains the degradation of HCOOH under various experimental conditions investigated here very well. The kinetic model presented here also predicts that the rate of formic acid degradation will decrease in partially- deoxygenated solutions however does not predict the complete inhibition of HCOOH degradation observed since some (albeit relatively slow) degradation by Ag(I) ions in irradiated solution has been observed.

123

5.4 Conclusion

We have shown here that the photo-generated hole is mainly responsible for oxidizing formic acid in irradiated AgCl(s) suspensions under the conditions investigated here. The relatively fast photo-decay in low salt concentration solutions was accounted for by including the oxidant intermediate formed through the Ag–H2O2 reaction. To the best of our knowledge this is the first work to provide a kinetic model as well as experimental evidence to explain the reaction mechanism of AgCl(s)–mediated photocatalytic decomposition of organics.

Scope exists for further clarification of aspects of the process described above. For example, the role of AgNP in formic acid photodecay would appear to warrant further investigation. If the presence of AgNP aids formate degradation by scavenging electrons and thereby reducing

− + ecb –hvb recombination, pre-added AgNP would be expected to lead to a higher rate of degradation of formic acid in irradiated AgCl(s) suspensions. Additionally, very little attention has been given to the dependency of formic acid degradation rate on initial formic acid concentration in the studies conducted to date.

Despite these shortcomings, the model presented adequately describes the results obtained and provides the basis for wider investigation of this interesting photo-oxidation process, particularly with regard to degradation of contaminant organics in waste streams.

124

Chapter 6: Degradation of Rhodamine B during Visible

Light Irradiation of AgCl(s) Suspensions

6.1 Introduction

The photo-degradation of organic pollutants using semiconducting silver halides (AgX, X =

Cl, Br) and their assemblages with other components, such as AgNPs, has been intensively studied in recent years, with rhodamine B (see Figure 6.1) frequently employed as the target organic contaminant for such research. Studying the degradation of dyes is attractive due to the simplicity of monitoring the loss of the dye using visible light spectrophotometry and also because the treatment of dye waste streams is itself a pressing problem. Some of the highlights studies into the photo-degradation rhodamine B utilizing silver halides are discussed as below.

Figure 6.1: Structure of rhodamine B.

Recent work by Miller, Yu et al. (2013) where partially reduced AgCl on reduced graphene

(rGO) was used to degrade rhodamine B (RhB) under visible light demonstrated that both chromophore destruction and N-deethylation processes occur during the photo–degradation of RhB. The concentration of the N-deethylated intermediates TER (N,N,N′- triethylrhodamine), DER (N,N′-diethylrhodamine), MER (N-ethylrhodamine) and Rh

(rhodamine) as well as RhB were estimated using a spectral deconvolution procedure. They

125

proposed that the generation of oxidants from the photocatalyst (in their case

Ag@AgCl/rGO) and the adsorption of contaminant to the sites responsible for oxidant production were important to rhodamine B degradation. They further discussed the role of direct injection of electrons from excited, adsorbed RhB to the photocatalyst, concluding that two primary mechanisms occur; the formation of some strong oxidant at the surface and direct electron injection from adsorbed dye to the Ag@AgCl/rGO photocatalyst. Earlier work by Watanabe, Takizawa et al. (1977) indicated that such a charge-transfer mechanism occured from excited dye molecules to the conduction band of a complex of dye–CdS

(adsorbate–adsorbent), after excluding the possibility for oxidant formation from photo- excited CdS. They also tentatively suggested the possible roles of oxygen in the destruction of dyes as a result of: i) enhancing electron transfer through scavenging of electrons from the

CdS surface, and ii) trapping surface electrons and forming superoxide which could react with cationic dye radicals.

Oxidants that are generated on irradiation of AgCl(s) that could possibly decompose RhB

0 • •– include the photo-generated hole, Cl , HO and O2 . While many studies have suggested the involvement of these various species in the observed decrease in RhB concentration on photolysis of AgCl(s), supporting evidence is generally missing. A variety of experiments have been undertaken using scavengers of these various species in order to assess their possible role in RhB degradation. For instance,

 Sun, Zhang et al. (2014) prepared Ag/AgCl/Zn-Cr photocatalysts and found they

displayed increased efficiency for photo–degradation of RhB compared to Ag/AgCl,

Zn-Cr layered double hydroxides (LDHs). These authors attributed this capacity to the

126

increased h+ generation resulting from a retardation in electron–hole recombination in

these particular materials.

 Adhikari, Gyawali et al. (2013) synthesized Ag/AgCl/WO3 photocatalyst and tested

its degradation under simulated sunlight with rhodamine B. This photocatalyst

showed better degradation ability compared to that of WO3 with the authors

attributing this increased efficiency to the local surface plasmon resonance of Ag

nanoparticle, which acted as carrier for the electron from the WO3 conduction band

with the photo–generated hole on the valence band of WO3 mainly responsible for

RhB degradation.

 Xiong, Zhao et al. (2011) compared the visible light photo–degradation efficiency of

RhB by Ag/AgCl/BiOCl compared to the simpler simpler catalysts Ag/AgCl and

•– 0 BiOCl. They hypothesized that superoxide (O2 ), chlorine radical (Cl ) and/or the

photo-generated hole (h+) could be the oxidant responsible for RhB photo–

decomposition. They used EDTA and tBuOH to trap h+ and HO• respectively, and

showed that EDTA addition reduced the degradation rate while adding tBuOH had no

effect. They also bubbled N2 into solution to reduce the concentration of dissolved

oxygen and observed a decrease in RhB degradation. Wang, Bai et al. (2012) also

used trapping experiments and concluded that the photo–generated hole was

responsible for RhB degradation.

 (Jiang and Zhang 2011) synthesized AgCl/Ag assemblages of controlled size and

shape and examined the degradation of RhB and methyl orange (MO) under daylight.

They concluded that direct hole oxidization of organic pollutants occurs by carrying

out a series of radical- and hole- trapping experiments with sodium bicarbonate used

+ • • to examine the role of h and adsorbed HO ,, tBuOH for HO in solution, and an

127

anaerobic experiment to clarify the role of O2. They found that oxygen did not play a

major role in either RhB or MO photo-degradation which contradicted the

observations of (Xiong, Zhao et al. 2011). tBuOH did not change the degradation rate

– perceivably while addition of 10 mM HCO3 decreased the extent of RhB and MO

degradation greatly. Based on these results they proposed a photo-degradation scheme

in which i) the combined electromagnetic field forced surface electrons of excited Ag0

away from AgCl(s) and hindered Ag+ reduction to Ag0 thus maintaining AgCl/Ag

•– • stable; ii) O2 and HO were not formed in this process and therefore did not photo-

degrade organic pollutants, hence they concluded that reaction of the organic dye with

holes led to dye degradation.

To summarize, existing literature suggests that photo-generated holes and Cl0 are the likely candidates for degradation of RhB (and other organic compounds) in a silver halide-driven irradiated solution system, while the effect of oxygen is a little ambiguous. Based on the study of Miller, Yu et al. (2013) and Watanabe, Takizawa et al. (1977), the adsorptivity of

RhB (organic dye) molecules on the surface of AgCl(s) and the availability of binding sites are important factors in determining the RhB degradation rate. This Chapter contributes to the knowledge gap in this field in terms of i) further clarifying the relative importance of oxidants from an irradiated AgCl(s) colloid system; and ii) explaining the photo-reaction process between RhB and AgCl(s) in an aerobic solution system. In this chapter understanding of the photochemistry of AgCl(s) and its ability to degrade RhB is advanced by measuring the kinetics of RhB degradation on irradiation of AgCl(s) over a range of suspension conditions.

128

6.2 Experimental Methods

6.2.1 Reagents and equipment

Rhodamine B (Laser grade, 99+ %) was purchased from ACROS, silver nitrate (reagent) from UNILAB and sodium chloride and sodium bicarbonate from Sigma–Aldrich. All other reagents are as described in Chapter 3 and were used without further purification. The same light source (ThermoOriel 150 W Xe lamp) used for investigations described in Chapter 3 was adopted for this work. The lamp was equipped with AM0 and AM1 filters to simulate the solar spectrum at the Earth’s surface.

6.2.2 Experimental method

Fresh silver chloride colloid was prepared by spiking an appropriate volume of silver nitrate stock solution into chloride ion containing buffer solution. AgCl(s) was allowed to age in the dark for 35 minutes after addition of Ag+. RhB solution was added after the aging of the

AgCl(s). The spectra of RhB were collected at fixed reaction time intervals after removal of residual AgCl(s) by centrifugation (10,000 rpm, 30 sec.) with a Cary 50 spectrophotometer, which was zeroed and baselined against 18.0 MΩ cm water (MQ) from a Millipore Milli–Q system.

The impact of varying the initial Ag(I), chloride and RhB concentrations was examined, as was the effect of the initial pH. The nature of the oxidant in the system was investigated by examining the impact of bicarbonate concentration as well as the effect of varying the oxygen concentration. The potential role of free chlorine, which is formed in these systems in micromolar quantities (Chapter 3), in driving the oxidation of rhodamine B was examined in

129

experiments where aliquots of free chlorine were added either in the dark or during the irradiation process.

6.2.3 Data analysis methodology

The data analysis method described by Miller, Yu et al. (2013) was used in this work. The spectrum of the N-deethylation products TER, DER, MER and Rh was approximated by scaling and shifting the measured spectrum of RhB at each time point based on the peak absorbance λmax and molar absorbance coefficient εmax reported by Watanabe, Takizawa et al.

(1977). At any given time point the spectrum of RhB will consist of a certain combination of

RhB, TER, DER, MER and Rh spectra. The simulation giving the minimum sum squared error between the calculated spectrum (from λ = 400 to 700 nm) and the experimentally observed spectrum was used for determining these unknown concentrations and the sum of all these products, the total chromophoric rhodamine (ΣRh) (Miller, Yu et al. 2013). The structures of the N-deethylation products are shown in Figure 6.2 below. All other RhB degradation products are assumed to have negligible absorbance in the visible region.

Figure 6.2: Diagram showing the chemical structures of RhB (R1 – R4 = Et), TER (R1 – R3 = Et, R4 = H), DER (R1, R4 = Et, R2, R3 = H), MER (R1 = Et, R2 – R4 = H) and Rh (R1 – R4 = H).

130

6.3 Results and Discussion

6.3.1 Background Degradation of RhB

Rhodamine B was found to be stable in the absence of silver chloride or light (Figure 6.3), as expected. Experimental data on RhB degradation with the presence of AgCl(s) solids but without light irradiation also showed no perceivable decrease of RhB concentration or generation of de-ethylated products. These results indicate that RhB photo-degradation was the direct result of photocatalytic reactions involving the AgCl(s) suspension.

131

2.0E-06 (a) Sum Rh [RhB] [TER] 1.9E-06 [DER] [MER] [Rh]

1.8E-06 Concentration (M) Concentration

1.7E-06 0 5 10 15 20 25 30 35 Time (min)

Time (min)

0 5 10 15 20 25 30 35 4.0E-08 (b) Sum Rh 3.0E-08 [RhB] [TER] 2.0E-08 [DER] [MER] 1.0E-08 [Rh]

Concentration (M) Concentration -1.0E-22

-1.0E-08

-2.0E-08

Figure 6.3: Photo–degradation of RhB with simulated sunlight in the absence of AgCl(s) (a) and a close–up of low concentration products: Rh, MER, DER and TER (b). Initial solution conditions :

– – [Rhodamine B]0 = 2 µM, [Cl ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0.

132

6.3.2 The effect of initial silver(I) concentration on RhB degradation

The impact of initial silver(I) concentration on the degradation of rhodamine B was examined at four initial concentrations of silver nitrate at a fixed sodium chloride concentration of 100 mM. As shown in Chapter 2, the concentration of AgCl(s) increases with increasing concentration of silver(I) (Figure 2.8), therefore increased Ag(I) concentration also leads to increased concentration of the photocatalyst (AgCl(s)). The results demonstrated that RhB degradation commenced quickly under irradiation, with little accumulation of intermediates

(TER, DER, MER and Rh).

A typical RhB degradation profile is presented in Figure 6.4, showing that the solution phase concentration of RhB rapidly decreased upon irradiation with AgCl(s). The results also indicate that N-deethylation occurred sequentially. The peak concentration of intermediate products generally occurred in the order of RhB > TER > DER > MER ≥ Rh. At the lowest

Ag(I) concentration of 20 µM this order was less obvious and appeared to follow the order

RhB > DER > Rh > MER > TER. The data suggested that the N-deethylated products do not build up to any appreciable concentration but are rapidly N-deethylated further, or undergo chromophore destruction. The rate of N-deethylated product formation increased with increasing silver(I) concentration, however, as can be seen in Figure 6.5, the removal process was also accelerated, leading to a similar peak concentration. The peak concentration of each

N-deethylated product was substantially lower as the N-deethylation progressed (i.e., the peak concentration of TER was greater than DER, etc). The RhB degradation profile was qualitatively similar at all silver(I) concentrations examined.

As can be seen from Figure 6.4(b) with the degradation of rhodamine B there was a gradual generation of de-ethylated products. Within the first 3 minutes the total de-ethylation

133

products constituted only 2 % of the ∑Rh, and at the end of 30 minutes’ reaction they made up around one third of the ∑Rh concentration, even though it was due in large part due to the reduced total concentration of RhB. Rhodamine B concentration was much higher than that of its N-deethylated counterparts at any time point of the reaction.

134

2.0E-06 Sum Rh (a) [RhB]

1.5E-06 [TER] [DER] [MER] 1.0E-06 [Rh]

Concentration (M) Concentration 5.0E-07

0.0E+00 0 5 10 15 20 25 30 Time (min)

2.0E-07 Sum Rh (b) [RhB]

1.5E-07 [TER] [DER] [MER] 1.0E-07 [Rh]

Concentration (M) Concentration 5.0E-08

0.0E+00 0 5 10 15 20 25 30 Time (min) Figure 6.4: Concentration profiles during degradation of RhB (a) and a close–up of the low concentration TER, DER, MER and Rh species (b). Initial solution conditions: [Ag(I)]0 = 100 µM, – – [Cl ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0.

135

Figure 6.5: Concentration profiles of TER (a) and DER (b) as a function of initial Ag(I) concentration.

The decay rate of rhodamine B concentration was able to be described by a pseudo-first order rate law, with plots of ln(C/C0) vs time demonstrating a linear relationship (Figure 6.6(a)).

The pseudo-first order rate constant, k′RhB increased with AgCl(s) concentration (Figure

6.6(b)). The initial rate of degradation of rhodamine B is also plotted in Figure 6.6 and shows that the initial degradation rate of RhB also increased linearly with increased Ag(I) loading, demonstrating the reliance of RhB degradation on the amount of catalyst in the solution: the more AgCl(s) the faster RhB degraded. This suggested that the entity (or entities if multiple paths existed) degrading rhodamine B was derived from the AgCl(s).

136

Figure 6.6: Semi-logarithmic plot of the concentration of RhB as a function of initial Ag(I) concentration (a), with the pseudo-first order rate constant obtained from these plots, as well as the

– – initial RhB degradation rate, shown in (b). Data is shown for 2.0 µM RhB, 100 mM Cl , 2 mM HCO3 , pH 8.0.

6.3.3 The effect of initial chloride concentration

Initial chloride concentration also had an effect on RhB photo-degradation; the RhB degradation rate was initially low at low chloride, then increased to a plateau value of ~3.6

µM·h−1 for chloride concentrations greater than ~75 mM (Figure 6.7). The results suggest that chloride ion and/or its oxidizing forms (Cl0, HOCl and OCl–) may play important roles in

137

RhB photo–decay. The concentration dependence here is reminiscent of that observed for the aggregation rate of AgCl(s) particles (Figure 2.11 and 2.12), suggesting that perhaps particle size or aggregation rate also influences the reaction. Alternatively, perhaps it indicates that the rhodamine oxidant generation process requires the interaction of like-charged entities

(which may or may not be localized on the AgCl(s) surface), with the higher ionic strength at higher chloride concentrations able to accelerate this unfavourable interaction in much the same way as it enhances the aggregation proceses. It would seem that the best correlation is found with the aggregation rate, although the reason for this is not clear.

Figure 6.7: Rhodamine B initial degradation rate (a) and rate constant (b) as a function of initial

– concentration of chloride ion concentration. [Rhodamine B]0 = 2 µM, [Ag(I)]0 = 100 µM, [HCO3 ]0 = 2 mM, pH 8.0.

The rhodamine B degradation profile with initial concentration of 100 µM Ag(I) and 50 mM

Cl– (Figure 6.8) was not much different from Figure 6.4 (RhB degradation with 100 µM

Ag(I) and 100 mM Cl–). The concentration of rhodamine B and total chromophoric rhodamine (sum Rh) showed a consistent decrease over irradiation time and stopped decreasing once the light was turned off. There was a very small percentage of rhodamine

138

(Rh) products (the species with intact chromophoric entity but being deprived of the four ethyls). At the end of the photolytic reaction (30 minutes time point) the Rh concentration constituted barely 3% of the total rhodamine products (RhB, TER, DER, MER and Rh), and the maximum concentration of Rh reached was only 2.74 х 10–8 M; while the peak concentration of all four de-ethylated products (TER, DER, MER and Rh) were below 30 % of the total, implying the major photo–reaction route led to the decolourization of rhodamine

B instead of deethylation. The peak concentration of RhB photo–degradation products at fixed Ag(I) concentration but with various Cl– concentration followed the same sequence as mentioned previously (RhB > TER > DER > MER ≥ Rh) except for initial chloride ion concentration of 20 mM in which RhB > DER > Rh > TER > MER.

139

2.0E-06 (a)

1.5E-06 Sum Rh [RhB] [TER] 1.0E-06 [DER]

[MER] Concentration (M) Concentration [Rh] 5.0E-07

0.0E+00 0 5 10 15 20 25 30 35 40 45 Time (min)

1.4E-07 (b) Sum Rh 1.2E-07 [RhB] [TER] 1.0E-07 [DER] [MER] 8.0E-08 [Rh]

6.0E-08 Concentration (M) Concentration 4.0E-08

2.0E-08

0.0E+00 0 5 10 15 20 25 30 35 40 45 Time (min) Figure 6.8: The degradation profile of rhodamine B as a function of time (a) and a close–up of low concentration species (TER, DER, MER and Rh) (b). The light was turned off at 30 minute point after

– – sampling. [Rhodamine B]0 = 2 µM, [Ag(I)]0 = 100 µM, [Cl ]0 = 50 mM, [HCO3 ]0 = 2 mM, pH 8.0 .

140

6.3.4 Impact of rhodamine B concentration

To test if different initial concentrations of RhB would degrade at the same rate, experiments were carried out with four different initial concentrations of RhB. The concentrations of rhodamine B and its photo–decay products decreased and increased respectively over time of irradiation, and reaction stopped once the light was cut off. The result with initial rhodamine

B concentration at 1.0 µM is displayed in Figure 6.9. The decay profiles for the other initial

RhB concentrations followed a similar trend.

1.00E-06 (a) Sum Rh 8.00E-07 [RhB] [TER] 6.00E-07 [DER] 4.00E-07 [MER] [Rh]

2.00E-07 Concentration (M) Concentration

0.00E+00 0 5 10 15 20 25 30 35 Time (min)

1.00E-07

(b) Sum Rh 8.00E-08 [RhB]

6.00E-08 [TER] [DER] 4.00E-08 [MER] [Rh]

Concentration (M) Concentration 2.00E-08

0.00E+00 0 5 10 15 20 25 30 35 Time (min)

141

Figure 6.9: The photo-degradation of rhodamine B as a function of time (a) and a close–up of low concentration species (TER, DER, MER and Rh). Initial solution conditions: [Rhodamine B]0 = 1.0 – – µM, [Ag(I)]0 = 100 µM, [Cl ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0 .

The concentration profile of rhodamine B was initially pseudo-first order, with the pseudo- first order rate constants, k′RhB shown in Figure 6.10. The value for k′RhB was found to steadily decrease with initial RhB concentration, however, if the total amount of chromophoric rhodamine that has been degraded (termed ∆∑Rh) is plotted as a function of time, it is clear that this apparently complex kinetics is simply due to a constant flux of oxidant being formed by the AgCl(s), seemingly independent of the presence of RhB (Figure 6.10b). In other words, with high enough initial concentration of rhodamine B its photodegradation rate with

AgCl(s) solution had the tendency to be more strongly limited by the supply of oxidizing entities from irradiated AgCl(s) nanoparticles than by rhodamine B concentration; while RhB concentration was still very low, the oxidant flux from AgCl(s) was surplus such that the observed pseudo-first order rate constant was strongly influenced by RhB concentration.

142

– – µM RhB data. Initial solution conditions: [Ag(I)]0 = 100 µM, [Cl ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0.

When the concentration profiles of the intermediate N-deethylated products are examined, it is clear that the initial rhodamine B concentration significantly impacts both the timing and magnitude of the peak concentrations (Figure 6.11). When the concentration of rhodamine B is increased, this leads to a proportional increase in both the length of irradiation time to

143

reach the peak, as well as the magnitude of the peak concentration; as all the different rhodamine B conditions are subject to the same oxidant flux, this would seem to suggest that the intermediates are in competition with RhB for the same oxidant, so they only undergo a net-decrease in concentration when they reach the same relative concentration ratio, which, necessarily, must occur at proportionally higher intermediate concentrations as the initial concentration of rhodamine B is increased.

Figure 6.11: TER concentration profile as a function of initial rhodamine B concentration

6.3.5 The effect of dissolved oxygen concentration

The role of oxygen in rhodamine B degradation was investigated as O2 is considered a probable candidate to scavenge photo-generated electrons from AgCl(s); therefore it could have the effect of hindering electron–hole pair recombination, which presumably would enhance the oxidizing capacity of the system towards rhodamine B degradation. The concentration decrease of rhodamine B and total rhodamine as a function of O2 condition is shown in Figure 6.12. The results indicated that, as expected, sparging did not lead to loss of rhodamine B, however, there was some impact of the O2 concentration upon the final

144

concentrations of both rhodamine B and total chromophoric rhodamine after an identical 30 min irradiation; the solution exposed to the ambient atmosphere resulted in final concentrations of ~10 nM and 50 nM for RhB and ∑RhB, respectively, whilst the solution that was sparged with Ar (which also contained 300 ppm CO2 to allow for pH control by the bicarbonate system) experienced a somewhat lesser degradation to final concentrations of

~300 nM and 550 nM, respectively. Apparently, with greatly decreased O2 concentration in the solution (about 10 ~ 30 µM after sparging, while the saturated O2 concentration was ~

260 µM at 25.0 °C, as measured with oxygen probe) the photo–decay of RhB was hampered, suggesting that O2 may indeed enhance hole-electron pair separation. It may also be possible that O2 is able to react with adsorbed one-electron oxidized dye radicals, preventing them from being reduced back to their initial state by conduction band electrons, thus also promoting the degradation process. Although it is not possible to distinguish which mechanism is operative under these conditions, it is clear that O2 does enhance the ability of the AgCl(s) system to degrade organic compounds.

145

2.5E-06

2.0E-06

1.5E-06 [RhB] 1.0E-06

Sum Rh Concentration (M) Concentration 5.0E-07

0.0E+00 sparged and sparged irradiated but RhB irradiated w/o sparging degradation in the dark Figure 6.12: Rhodamine B (RhB) and total rhodamine (∑Rh) concentrations before and after irradiation either under an ambient atmosphere or whilst sparging with Ar (+300 ppm CO2): (1) the mixture of RhB and AgCl(s) was pre–sparged and then irradiated; (2) the mixture was sparged but kept in the dark; (3) the mixture was irradiated but without pre–sparging; (4) the mixture was kept in

– – the dark without sparging. [RhB]0 = 2.0 µM, [Ag(I)]0 = 100 µM, [Cl ]0 = 100 mM, [HCO3 ]0 = 2 mM, pH 8.0, irradiation time 30 minutes. Irradiations were performed in a 3 mL quartze fluorescence cuvette.

The composition of the various species of rhodamine B and N-deethylation products relative to the total chromophoric rhodamine concentration (∑Rh) after the photocatalytic reaction is shown in Figure 6.13. The data are consistent with the previous observation that O2 enhances the degradation, as the data in the presence of O2 show that RhB is almost completely degraded, and that N-deethylation has proceeded substantially (to the point that TER is the dominant final product), with significant concentrations of all N-deethylated species observed. The data from the Ar-sparged solution show that RhB is still the dominant

(chromophoric) product, with N-deethylation occurring to a somewhat lesser extent. This data is entirely consistent with the notion that deprivation of O2 simply slows down the overall reaction, without any observable change in mechanism.

146

Figure 6.13: Relative concentration of rhodamine B species (sum Rh = 1): (1) the RhB and AgCl(s) mixture was sparged and irradiated (solid blue bar); (2) the mixture was irradiated for the same period of time but did not undergo sparging (empty red bar).

6.3.6 The effect of bicarbonate concentration

Bicarbonate ions may influence the system in many ways, the two most significant being that they may compete with RhB for surface sites, or that they may scavenge the holes (or other reactive intermediates, such as hydroxyl radicals, HO•) in competition with RhB. The experiment result, however, showed no significant impact upon rhodamine B degradation rate for bicarbonate concentrations between 2 to 8 mM (Figure 6.14). The results suggest the following three possibilities:

(1) No significant concentration of HO• was reached during the photolysis of AgCl(s), as

has been thoroughly demonstrated in Chapter 3.

(2) Bicarbonate is not able to compete with RhB for surface sites, or it is unable to

prevent an oxidant formed from AgCl(s) from diffusing and subsequently reacting

with RhB

147

•− (3) bicarbonate scavenged the reactive oxidant to form CO3 radicals, which are

themselves good oxidants and likely to also effectively oxidize RhB (i.e., bicarbonate

did intercept the reactive oxidant, but there was no net observable effect)

0.08

(a)

)

1 –

0.07

(µM.min 0

0.06 (d[RhB]/dt)

0.05 0 2 4 6 8 10 – [HCO3 ]0 (mM)

0.06

(b) )

1 0.04 –

k'RhB k'RhB (min 0.02

0.00 0 2 4 6 8 10 – [HCO3 ]0 (mM) Figure 6.14: Rhodamine B degradation rate (a) and rate constant (b) as a function of initial concentration of bicarbonate. Data is shown for 2.0 µM rhodamine B, 100 µM Ag(I), 100 mM Cl– at pH 8.0. AgCl(s) was aged for 35 minutes in the dark before irradiation was commenced.

148

6.3.7 The effect of solution pH on rhodamine B degradation

The pH value of a system is usually critical for its performance (Ungelenk and Feldmann

2012). Changes to pH are likely to lead to changes in the speciation of reactive components and also to the surface charge of the AgCl(s) particles. Changes to pH can also influence the reactivity of various components in the system; for example, as discussed earlier, pH changes the rate of disproportionation of superoxide due to the differing reactivity of its acid and base

• •− forms, HO2 and O2 , respectively (Bielski, Cabelli et al. 1985). The pH value also has a minor influence on the AgCl(s) aggregation rate and particle stability. Free chlorine

(hypochlorous acid, hypochlorite ion and chlorine) also change their dominant form with pH, suggesting that its reactivity will also likely vary with pH. Although RhB speciation should be relatively unchanged under the conditions investigated in this work, as the pKa for RhB is

3.22 (Mchedlov-petrossyan, Kukhtik et al. 1994), the reactivity of AgCl(s) towards rhodamine B could be impacted if the adsorbed dye molecule exhibited a different pKa to the solvated dye.

The experiment results demonstrated that although the solution pH did not have much impact on the rhodamine B degradation rate, which is similar at pH = 4, 8 and 10, the impact of pH upon the decay of total chromophoric rhodamine was pronounced; there is only minimal degradation of chromophoric rhodamine at pH 4 (i..e, N-deethylation competes favourably with chromophore destruction), whereas at pH = 8 and 10 N-deethylation is only a minor pathway (Figures 6.15 and 6.16). The reason for this behaviour is not immediately clear, however it does suggest that the N-deethylation and chromophore destruction may not occur through the same process. One possibility is that the change in the nature of the rhodamine-

AgCl(s) interaction allows for more efficient injection of an electron from the dye to the

149

conduction band of AgCl(s), promoting the N-deethylation process in a similar fashion to that reported by Watanabe, Takizawa et al. (1977) for CdS; perhaps the AgCl(s) surface charge is more amenable to efficient rhodamine coupling at pH 4, or perhaps bicarbonate is ordinarily able to outcompete the dye for suitable surface sites but, since at pH 4 its concentration will be much reduced, rhodamine is then able to efficiently adsorb to such sites. At this stage is not possible to conclude with certainty the exact nature of the mechanism.

Figure 6.15: Impact of solution pH upon the concentration profiles of rhodamine B (a) and total chromophoric rhodamine (b).

150

Figure 6.16: Impact of solution pH upon the concentration profiles of TER (a) and DER (b).

6.3.8 The effect of spiked hypochlorite

The effect of hypochlorite upon RhB degradation was examined by spiking 2.0 µM sodium hypochlorite into a RhB-containing AgCl(s) solution immediately prior to irradiation. Both the dark and light exposed solutions displayed the typical blue-shift in absorbance peak as N- deethylation occurred along with a general decrease in absorbance intensity, however, there is also a general broadening of the peak towards higher wavelength, which has not previously been observed in this work. None of the potential rhodamine intermediates considered thus far absorbs at a wavelength higher than rhodamine B itself, therefore, this broadening of the

151

peak towards higher wavelengths must be as a result of the formation of a previously- unconsidered component with significant absorbance at wavelengths higher than rhodamine

B. Unlike earlier experiments, significant degradation of rhodamine B occurred in the dark, implying that HOCl must be able to react with rhodamine B at an appreciable rate; this also suggests that the rhodamine-HOCl reaction may also be the precursor to this new absorbance component. From examination of literature data for the structurally similar fluoresceins, it is clear that HOCl is able to chlorinate the xanthene ring at quite a rapid rate (Hurst, Albrich et al. 1984, Best, Sattenapally et al. 2013). The chlorinated product typically undergoes a red- shift, consistent with these observations.

If the normal deconvolution procedure is applied to the final spectra after the reaction of RhB with HOCl, but with the addition of another component with a higher λmax and unknown molar absorbance coefficient (but the same shape spectrum as RhB), good fits to the data can be obtained. λmax was initially treated as an unknown parameter and allowed to vary, with

λmax = 568.2 nm found to best fit the data. Whilst this is clearly an approximate treatment of the data, it is the best available technique. Furthermore, as no information about the molar absorbance coefficient is available for this proposed species, all that can be fit for this species is the value of the product of the concentration and max.

As RhB is present in much greater quantities than the various N-deethylation products, it would seem most likely that chlorination product(s) derived from rhodamine B should dominate. Hurst, Albrich et al. (1984) show that fluorescein undergoes a very rapid monochlorination, with this monochlorinated product then undergoing a still rather fast additional reaction to form a dichloro product. Under the conditions of the current experiment, 2 µM

HOCl was reacted with 2 µM RhB. In the dark experiment more than 1 µM RhB is degraded,

152

which suggests that only a monochlorination process occurs (or is at least the dominant process). This suggests that assuming only one chlorination product is formed would be a reasonable assumption. The proposed reaction scheme, based upon the work of Hurst,

Albrich et al. (1984), is shown in Figure 6.17 below.

Figure 6.17: Chlorination of fluorescein by HOCl (from Hurst, Albrich et al. (1984)) as well as the mechanism considered likely for rhodamine B under the conditions of this study.

Applying this modified analysis scheme leads to the concentration profiles shown in Figures

6.18 and 6.19. The products of photo–degradation aside from the de–ethylated intermediates mentioned previously (TER, DER, MER and Rh) also includes the C value for chloro– rhodamine (ClRh). This ClRh species rapidly reached its maximum concentration in the first

5 minute of the reaction (in either the light or dark) and remained relatively stable throughout the process. Comparing the rhodamine B degradation with spiked OCl– in the dark to that with irradiated AgCl(s) (Figure 6.20) showed that there was one third more chloro– rhodamine (ClRh) generated when RhB degraded in the dark (with the whole reaction vessel

153

enwrapped with aluminium foil) in the same period of time. The dark degradation also resulted in slightly greater degradation of RhB and total chromophoric rhodamine. Also if the

RhB degradation with added OCl– occurred with irradiation, the reaction slowed down more quickly than the dark reaction, where RhB and sum Rh were seen to keep decreasing for a longer time. These differences indicate that OCl–, because of its own instability under irradiation, was also independently consumed in some photochemical process, leading to a slower RhB degradation rate (relative to the dark case). The results also suggest that the symmetrical loss of two or four ethyls was preferred to other N-deethylation processes in

RhB photo-decay with irradiated AgCl(s) solution in the presence of pre-added OCl– ions.

The concentration of DER and Rh was higher than TER and MER. However, the accumulated maximum concentration of any N-deethylated intermediate was generally below

130 nM, which suggests that N-deethylation did not constitute a major route of degradation.

154

2.0E-06 0.015 (a)

0.012 Sum Rh 1.5E-06

[RhB] 1) 0.009 – [DER] [Rh] 1.0E-06 [TER] 0.006

[MER] εC(ClRh) (cm

Concentration(M) [ClRh] 5.0E-07 0.003

0.0E+00 0.000 0 5 10 15 20 25 30

Irradiation time (min)

2.0E-07 0.015 (b)

0.012 Sum Rh 1.5E-07 [RhB]

[DER] 1) 0.009 – [Rh] 1.0E-07 [TER]

0.006 [MER]

εC(ClRh) (cm [ClRh] Concentration(M) 5.0E-08 0.003

0.0E+00 0.000 0 5 10 15 20 25 30 Irradiation time (min)

Figure 6.18: The concentration of intermediate products (ΣRh, RhB, TER, DER, MER and Rh) as a function of irradiation time, the product of chloro–rhodamine concentration and its molar absorptivity (εC(ClRh)) is on the right–hand–side of secondary y–axis (a); a close–up of the low concentration DER, Rh, TER and MER, and εC of chloro–rhodamine (ClRh) (b).

155

2.0E-06 0.020 (a) 0.018

0.016 Sum Rh 1.5E-06

0.014 [RhB] 1) 0.012 – [DER] [TER] 1.0E-06 0.010 [Rh] 0.008

[MER] εC(ClRh) (cm Concentration(M) 0.006 5.0E-07 [ClRh] 0.004

0.002

0.0E+00 0.000 0 5 10 15 20 25 30 35 Time (min)

2.0E-07 0.020 (b) 0.018 Sum Rh 0.016 [RhB] 1.5E-07 0.014 [DER]

1) [TER] 0.012 – [Rh] 1.0E-07 0.010 [MER]

0.008 [ClRh] εC(ClRh) (cm

Concentration(M) 0.006 5.0E-08 0.004

0.002

0.0E+00 0.000 0 5 10 15 20 25 30 35 Time (min)

Figure 6.19: The concentration of intermediate products (ΣRh, RhB, TER, DER, MER and Rh) as a function of time, the product of chloro–rhodamine concentration and its molar absorptivity (εC(ClRh)) is on the right–hand–side of secondary y–axis (a); a close–up of the low concentration DER, Rh, TER, and MER, εC of chloro–rhodamine (ClRh) (b). Degradation happened in the dark.

156

6.4 Implications and Conclusions

157

This proportional increase of RhB photo–degradation reactivity with increased catalyst concentration implied that at an initial RhB concentration of 2 µM the solution was optically thin at wavelengths responsible for oxidant production. Alternatively, degradation was limited by the capability of RhB to gain access to the sites where the acting oxidant(s) were generated, if the redox reaction between RhB and the oxidant decomposing it occurred on the surface of the catalyst instead of in solution. In the first scenario there was still a lot of incoming light intensity left after passing through the solution, in which both RhB and

AgCl(s) absorbed light. This matching of increase in RhB photo–decay rate and the increase in initial concentration of Ag(I) (and thus [AgCl(s)]0) is very much similar to what we observed in Figure 3.7 in Chapter 3, where the free chlorine generation rate increased

0 + correspondingly with [Ag(I)]0. This coincidence suggested that free chlorine, Cl or h were the main oxidants inducing RhB photo–decay.

Further insight into RhB photo–decomposition can be found in the effect of RhB concentration. Since the loss of RhB indicated the oxidative deprivation of the chromophoric entity and was connected with the generation of oxidant from excited AgCl(s) nanoparticles, combined with the fact that both RhB and AgCl(s) absorb photons strongly within the visible light region (especially given that the absorbance spectrum of AgCl(s) is extremely broad for the initial Ag(I) and Cl– concentration used), the data suggested a competition for photons between RhB and AgCl(s) (Figure 6.10) and increased RhB concentration accordingly decreased the photon density that could be absorbed by AgCl(s), resulting in reduced generation of reactive oxidant and subsequently lower R(RhB).

158

160 0.25

140 0.2

120 1)

– 100 0.15 2.nm

- 80 light source

60 0.1 aged AgCl(s)

(mW.m 2 µ M RhB

40 Absorbance(AgCl(s))

0.05 Incident Incident Spectral Irradiance 20

0 0 300 400 500 600 700 Wavelength (nm)

Figure 6.21: The intensity of light (primary y–axis), absorbance of AgCl(s) and RhB (secondary y– axis) as a function of wavelength.

Dissolved oxygen has been demonstrated to enhance the degradation of rhodamine B, either as a result of enhancing the separation of photo-formed electron-hole pairs or perhaps by driving the oxidation of intermediate dye-radical species. At pH 8, varying the concentration of bicarbonate over a moderate range did not lead to any significant change in reaction profile, however, at low pH, chromophore destruction was found to be greatly retarded, but

N-deethylation substantially more favourable, which may potentially be as a result of the lowered bicarbonate concentration at low pH, hence resulting in more available surface sites for reaction/binding of rhodamine molecules.

The dissolved oxygen effect offered additional evidence of the involvement of h+, Cl0 or free chlorine in RhB decay, as decreased O2 concentration also reduced free chlorine generation

(Figure 3.8) (as the presence of oxygen facilitates h+ – e– separation leading to more generation of free chlorine). Further experiments with spiked OCl– indicated that free chlorine could be one of the main species responsible for the concentration decrease of RhB and sum

Rh. As shown in Figure 6.20, spiking with OCl– resulted in a faster initial degradation rate of

159

RhB and total chromophoric Rh (sum Rh) than irradiated AgCl(s), which exhibited a maximum free chlorine generation of around 2 µM. After around 30 minutes’ reaction, the final concentrations of RhB and sum Rh were essentially the same under the three experiment conditions. With spiked OCl– the reaction of RhB degradation was no longer first order.

Experiments have also been conducted whereby aliquots of HOCl were spiked into the solution at concentrations typical of those shown to be formed in this system (in the absence of rhodamine B) in Chapter 3. In these experiments a new chlorinated rhodamine species is postulated to be formed, which is not observed in any other reaction; its absence from other experiments suggests that HOCl does not accumulate when rhodamine is present in the system, which would imply that the precursor of HOCl formation is being scavenged by rhodamine B. The simplest explanation for this process would seem to be that rhodamine B is itself directly oxidized by holes, or, perhaps, oxidized by AgCl(s)-bound Cl• at a rate faster than it is possible for Cl2 to be formed.

160

Chapter 7: Conclusions

7.1 Main findings

In this thesis the photochemistry of AgCl(s) aqueous dispersions is studied with attention given to both the generation of reactive oxygen species, free chlorine and silver nanoparticles and the impact of these species on organic contaminant degradation. Photocatalytic reaction kinetics of ROS generation and degradation of both formic acid and rhodamine B in the presence of AgCl(s) suspension were also investigated and mechanisms proposed. Further details of key outcomes are provided below.

7.1.1 Characteristics of AgCl(s) Particles

The results from characterization of AgCl(s) particles by a range of techniques have revealed that silver chloride particles aggregate rapidly (at rates expected of diffusion-limited cluster aggregation (DLCA)) when NaCl concentrations are greater than ~100 mM, While compact

AgCl(s) aggregates are recognised to not be particularly light sensitive (Glaus and Calzaferri

1999), the fact that DLCA appears to predominate suggests that the AgCl(s) formed here will be relatively light sensitive given the relatively open nature of the clusters formed. The precipitation rate and size distribution of silver chloride particles produced can be controlled by changing the initial Ag(I) and Cl– ion concentrations with the nature of the resultant

AgCl(s) influencing the photocatalytic properties of AgCl(s).

Microscopic analysis of the AgCl(s) nanoparticles suggested that the photocatalytic reaction changed the composition of the AgCl(s) particles with generation of silver nanoparticles. The silver chloride photolytic reaction also alters the distribution of silver chloride particles in terms of size, shape and surface structure. As a whole, considering the EDS spectra of the

161

four AgCl(s) specimens that were examined, the ratio of Ag to Cl elements increased in the order: i) the specimen in the dark, ii) the specimen following visible light irradiation, iii) the specimen irradiated with formate and iv) the speciment irradiated with RhB with the increasing proportion of Ag compared to Cl attributed to the generation of Ag0 during the photochemical reaction.

7.1.2 Generation of free chlorine and ROS on irradiation of AgCl(s) suspensions

Detailed studies have been undertaken to show that free chlorine is the most abundant product in irradiated AgCl(s) suspensions with micromolar concentrations generated over one hour of irradiation. Chlorine generation was influenced by a range of solution parameters including initial Ag(I) and Cl- concentrations, pH, the buffer used and the absence/presence of oxygen since all these parameters influence the fate of h+ and e- pairs generated at the

•- 1 0 AgCl(s) surface. Reactive oxygen species (O2 , H2O2, O2) and Ag were generated but consumed through a chain of electron transfer reactions. A detailed mechanism and associated kinetic model for the production of free chlorine, AgNP and ROS on photolysis of

AgCl(s) have been developed which satisfactorily explains all the data obtained.

Our results indicate that superoxide is formed on irradiation of AgCl(s) suspensions as a result of oxygen-mediated scavenging of electrons from the AgCl(s) conduction band.

•- Hydrogen peroxide is believed to form as a result of the disproportionation of O2 . The scavenging of hypochlorite with glycine and scavenging of hydrogen peroxide with catalase

1 respectively provided evidence for the generation of O2 as a result of the reaction between

1 hypochlorite and peroxide. Direct evidence of O2 generation was also found from measurement of singlet oxygen with SOSG reagent. No net hydroxyl radical generation was apparent in our AgCl(s) system. Free chlorine was generated as the direct result of reaction

162

of Cl- ions with photogenerated holes to yield Cl0, which subsequently hydrolyzed to form

HOCl.

7.1.3 Free chlorine decay in the presence of silver nanoparticles

AgNPs undergo rapid oxygenation-dissolution at both pH 4 and 8 with the presence of free chlorine increasing the dissolution rate of AgNPs. At pH 8, increasing concentrations of OCl- lead to more rapid initial AgNP dissolution, although the nature and stoichiometry of this process is not clear. At higher concentrations of OCl-, although the initial oxidation rate is accelerating, over time an apparent steady state is reached suggesting that there is also some

(unknown) OCl−-driven process that leads to AgNP stabilization. The oxidation of AgNPs by

HOCl/OCl− is greatly accelerated at lower pH, suggesting that HOCl is more reactive with

AgNP than OCl−.

7.1.4 Photodegradation of formic acid

•− Our results show that holes and their subsequent initial oxidation product, CO3 , generated on photo-excitation of AgCl(s) by visible light are responsible for oxidizing HCOO− under

•− the experimental conditions investigated here. The concentration of holes and CO3 available for HCOO− degradation depends on the dioxygen concentration, total carbonate concentration and chloride concentration as well as pH with more holes available in the presence of high dioxygen, low chloride and high bicarbonate concentration. A kinetic model was developed which was able to account for all results obtained. Although the efficiency of degradation was estimated to be low (0.5 ‒ 5%), the concentration of the photocatalyst,

AgCl(s), remains unchanged with minimal release of Ag0, a necessary requirement for this technology to be used in large-scale applications. Further work is required to optimize the solution conditions to improve the degradation efficiency. Furthermore, although the

163

mechanism proposed here adequately explains the rate and extent of HCOO− degradation, its applicability to other organic compounds needs to be further investigated.

7.1.5 Photodecomposition of rhodamine B

AgCl(s) has been demonstrated to efficiently photo-degrade Rhodamine B, with the degradation process consisting of reactions that both destroy the chromophore structure and also lead to N-deethylation. At fixed rhodamine B concentration, increasing concentrations of both chloride and silver(I) lead to enhanced degradation, with reasons for the chloride effect unclear while the effect of increasing silver concentration is associated with the formation of more AgCl(s). Over the concentration range studied, the initial concentration of rhodamine B did not alter the kinetics of oxidant production, with the same quantity of oxidant generated at high and low concentrations (although this necessarily leads to slower pseudo-first order degradation rate constants at higher rhodamine B concentrations). An observd proportional increase of rate of RhB photodegradation with increased catalyst concentration implied that at the micromolar concentrations of RhB used, the solution was optically thin at wavelengths responsible for oxidant production. Alternatively, degradation was limited by the capability of RhB to gain access to the sites where the acting oxidant(s) were generated, if the redox reaction between RhB and the oxidant occurred on the surface of the catalyst instead of in solution. Results obtained suggest that free chlorine, Cl0 or h+ were the main oxidants inducing RhB photo–decay. Increased RhB concentration decreased the photon density that could be absorbed by AgCl(s), resulting in reduced generation of oxidant and subsequently lower rate of RhB degradation.

Dissolved oxygen enhanced the degradation of rhodamine B, either as a result of enhancing the separation of photo-formed electron-hole pairs or by driving the oxidation of intermediate

164

dye-radical species. The dissolved oxygen effect offered additional evidence of the

+ 0 involvement of h , Cl or free chlorine in RhB decay, as decreased O2 concentration also reduced free chlorine generation (as the presence of oxygen facilitates h+ – e– separation leading to more generation of free chlorine). Further experiments with spiked OCl– indicated that free chlorine could be one of the main species responsible for RhB degradation since spiking with OCl– resulted in a faster initial rate of degradation of RhB than was the case in the presence of irradiated AgCl(s) alone. Further experiments in which aliquots of HOCl were spiked into the solution at concentrations typical of those shown to be formed in this system suggests that HOCl does not accumulate when rhodamine is present in the system, which would imply that the precursor of HOCl formation is being scavenged by rhodamine

B. The simplest explanation for this process would seem to be that rhodamine B is itself directly oxidized by holes, or, perhaps, oxidized by AgCl(s)-bound Cl• at a rate faster than it is possible for Cl2 to be formed.

7.2 Limitations and future work

The role of AgNP in formic acid photodecay warrants further investigation. Since silver nanoparticles were shown to be generated on irradiation of AgCl(s) suspensions and earlier results from He et al. (He, Jones et al. 2011) showed that AgNP was involved in electron

•- exchange with O2 , pre-spiking AgNP into AgCl(s) solution might increase the photodecomposition rate of formic acid as a result of a reduction in the rate of recombination of photo-generated holes and electrons. On the other hand, by adding AgNP and/or H2O2 the oxidising intermediate resulting from reaction of these two species (He, Garg et al. 2012) could lead to more formic acid decomposition.

165

Even though two different processes (chromophore destruction and deethylation) were proposed to account for RhB photodecomposition, at this stage no clear mechanism can be forwarded as to whether direct oxidation or electron injection is responsible for this process.

Even though carbonate radical was suggested to improve the photodecomposition of formic acid, similar rates of decay of RhB at pH values of 4, 8 and 10 suggest that carbonate radical does not play an important role in this case.

The effect of pH on the deethylation of rhodamine B suggested that the adsorption of rhodamine B to the surface of AgCl(s) might be an important step, especially in view of the recognized pH dependence of RhB adsorption to surfaces (Chen, Zhao et al. 2003).

Alternatively, the observed pH effect might be associated more with the pH dependence of the generation and fate of the e- and h+ generated at the AgCl(s) surface.

As the results obtained indicated a competition between RhB and AgCl(s) for incident photons yet, for the same solution matrix, the total degradation of Rh (ƩRh) at initial concentrations of 1.0 to 4.0 µM RhB was essentially the same, we concluded that at high enough RhB concentration, its degradation was not impacted by its initial concentration but by the oxidant influx from photocatalyst AgCl(s). An initial concentration of 200 µM Ag(I) would hopefully see faster degradation for RhB (with initial concentration of 1.0 to 4.0 µM) if the optically thin condition is still satisfied. Further study to maximize the generation of h+

(and thus Cl0) is suggested here as a good direction in which to proceed if efficient rhodamine

B decomposition with photo catalyst AgCl(s) is sought. In addition, more attention should be given to the role of free chlorine in RhB degradation and to answer the question of why a chlorinated rhodamine product only was formed with pre-added free chlorine.

166

It is also suggested that in future studies, the photocatalytic behavior of AgBr, AgI and AgCl be compared under similar reaction conditions. It would also be interesting to see whether substituting AgCl with AgBr (or AgI) would increase the photodegradation rate of formic acid or rhodamine B. Furthermore, other more recalcitrant organic chemicals (such as pharmaceuticals) could be considered as targets for future work with photolyzed AgBr.

Finally, the oxidative dissolution of AgNP by free chlorine also requires substantial additional investigation.

167

References

(1998). Standard Methods For the Examination of Water and Wastewater. the United States of America, American Public Health Association

American Water Works Association

Water Environment Federation.

Adhikari, R., G. Gyawali, T. Sekino and S. W. Lee (2013). "Microwave assisted hydrothermal synthesis of Ag/AgCl/WO3 photocatalyst and its photocatalytic activity under simulated solar light." Journal of Solid State Chemistry 197: 560- 565.

Andersson, M., H. Birkedal, N. R. Franklin, T. Ostomel, S. Boettcher, A. E. C. Palmqvist and G. D. Stucky (2005). "Ag/AgCl-loaded ordered mesoporous anatase for photocatalysis." Chemistry of Materials 17(6): 1409-1415.

Bader, H., V. Sturzenegger and J. Hoigne (1988). "Photometric-Method for the Determination of Low Concentrations of Hydrogen-Peroxide by the Peroxidase Catalyzed Oxidation of N,N-Diethyl-P-Phenylenediamine (Dpd)." Water Research 22(9): 1109-1115.

Best, Q. A., N. Sattenapally, D. J. Dyer, C. N. Scott and M. E. McCarroll (2013). "pH-dependent Si-fluorescein hypochlorous acid fluorescent probe: Spirocycle ring-opening and excess hypochlorous acid-induced chlorination." Journal of the American Chemical Society 135(36): 13365-13370.

Bi, Y. P. and J. H. Ye (2009). "In situ oxidation synthesis of Ag/AgCl core-shell nanowires and their photocatalytic properties." Chemical Communications(43): 6551-6553.

Bielski, B. H. J., D. E. Cabelli, R. L. Arudi and A. B. Ross (1985). "Reactivity - of HO2/O2 radicals in aqueous solution." Journal of Physical and Chemical Reference Data 14(4): 1041-1100.

Bielski, B. H. J., D. E. Cabelli, R. L. Arudi and A. B. Ross (1985). "Reactivity of Ho2/O-2 Radicals in Aqueous-Solution." Journal of Physical and Chemical Reference Data 14(4): 1041-1100.

Bielski, B. H. J. and H. W. Richter (1977). "Study of Superoxide Radical Chemistry by Stopped-Flow Radiolysis and Radiation-Induced Oxygen- Consumption." Journal of the American Chemical Society 99(9): 3019-3023.

168

Buxton, G. V., C. L. Greenstock, W. P. Helman and A. B. Ross (1988). "Critical-Review of Rate Constants for Reactions of Hydrated Electrons, Hydrogen-Atoms and Hydroxyl Radicals (.Oh/.O-) in Aqueous-Solution." Journal of Physical and Chemical Reference Data 17(2): 513-886.

Buxton, J. D., C. L. Greenstock, W. P. Helman and A. B. Ross (1988). "Critical review of rate constants for reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals (.OH/O-.) in aqueous solution." Journal of Physical and Chemical Reference Data 17: 513-531.

Calzaferri, G. (1997). "At the time he made the first photographs on paper: Did Henry Fox Talbot oxidize water to oxygen with sunlight?" Catalysis Today 39(3): 145-157.

Calzaferri, G. (2010). "Artificial photosynthesis." Topics in Catalysis 53(3-4): 130-140.

Calzaferri, G., D. Bruhwiler, S. Glaus, D. Schurch, A. Currao and C. Leiggener (2001). "Quantum-sized silver, silver chloride and silver sulfide clusters." Journal of Imaging Science and Technology 45(4): 331-339.

Chen, D. L., M. N. Liu, Q. Q. Chen, L. F. Ge, B. B. Fan, H. L. Wang, H. X. Lu, D. Y. Yang, R. Zhang, Q. S. Yan, G. S. Shao, J. Sun and L. Gao (2014). "Large- scale synthesis and enhanced visible-light-driven photocatalytic performance of hierarchical Ag/AgCl nanocrystals derived from freeze-dried PVP-Ag+ hybrid precursors with porosity." Applied Catalysis B-Environmental 144: 394-407.

Chen, D. L., S. H. Yoo, Q. S. Huang, G. Ali and S. O. Cho (2012). "Sonochemical Synthesis of Ag/AgCl Nanocubes and Their Efficient Visible- Light-Driven Photocatalytic Performance." Chemistry-a European Journal 18(17): 5192-5200.

Chen, F., J. Zhao and H. Hidaka (2003). "Highly selective deethylation of rhodamine B: Adsorption and photooxidation pathways of the dye on the TiO2/SiO2 composite photocatalyst." International Journal of Photoenergy 5: 9. Choi, M., K.-H. Shin and J. Jang (2010). "Plasmonic photocatalytic system using silver chloride/silver nanostructures under visible light." Journal of Colloid and Interface Science 341(1): 83-87.

Choi, O. and Z. Q. Hu (2008). "Size Dependent and Reactive Oxygen Species Related Nanosilver Toxicity to Nitrifying Bacteria." Environ. Sci. Technol. 47.

169

Chu, H. C. and K. M. Chen (2002). "Reuse of activated sludge biomass: I. Removal of basic dyes from wastewater by biomass." Process Biochemistry 37(6): 595-600.

Clesceri, L. S., A. E. Greenberg and A. D. Eaton (1998). Standard Methods. Baltimore, Maryland, the United States of America, American Public Health Association

American Water Works Association

Water Environment Federation.

Dalrymple, R. M., A. K. Carfagno and C. M. Sharpless (2010). "Correlations between Dissolved Organic Matter Optical Properties and Quantum Yields of Singlet Oxygen and Hydrogen Peroxide." Environmental Science & Technology 44(15): 5824-5829.

Di, H., M. W. Bligh and T. D. Waite (2013). "Effects of aggregate structure on dissolution kinetics of citrate stabilized Silver Nanoparticles." Environ. Sci. Technol 47: 9148-9156.

Di, H., S. Garg and T. D. Waite (2012). "H2O2-mediated Oxidation of Zero- valent Silver and Resultant Interactions between Silver Nanoparticles, Silver Ions and Reactive Oxygen Species." Langmuir 28: 10266-10275.

Di, H., A. M. Jones, S. Garg, A. N. Pham and T. D. Waite (2011). "Silver Nanoparticle−Reactive Oxygen Species Interactions: Application of a Charging−Discharging Model." The Journal of Physical Chemistry 115: 5461- 5468.

Dominguez-Vera, J. M., N. Galvez, P. Sanchez, A. J. Mota, S. Trasobares, J. C. Hernandez and J. J. Calvino (2007). "Size-controlled water-soluble Ag nanoparticles." European Journal of Inorganic Chemistry(30): 4823-4826.

Dong, R. F., B. Z. Tian, C. Y. Zeng, T. Y. Li, T. T. Wang and J. L. Zhang (2013). "Ecofriendly Synthesis and Photocatalytic Activity of Uniform Cubic Ag@AgCl Plasmonic Photocatalyst." Journal of Physical Chemistry C 117(1): 213-220.

Fan, H., J. Y. Zhu, J. C. Sun, S. X. Zhang and S. Y. Ai (2013). "Ag/AgBr/CoNiNO3 Layered Double Hydroxide Nanocomposites with Highly Adsorptive and Photocatalytic Properties." Chemistry-a European Journal 19(7): 2523-2530.

170

Feng, L., K. F. He and W. P. Chen (2012). "Study on stability of AgI/TiO2 visible light photocatalysts in solutions of various pH values." New Materials, Applications and Processes, Pts 1-3 399-401: 1272-1275.

Finsy, R. (1994). "Particle Sizing by Quasi-Elastic Light-Scattering." Advances in Colloid and Interface Science 52: 79-143.

Flyunt, R., M. N. Schuchmann and C. V. Sonntag (2001). "A common carbanion intermediate in recombination and proton-catalysed •- disproportionation of the caboxyl radical anion, CO2 , in aqueous solution." Chemistry-A European Journal 7: 796-799.

Gilbert, E. (1981). "Photometrische Bestimmung niedriger Ozonkonzentrationen in Wasser mit Hilfe von Diathyl-p-phenylendiamin (DPD)." GWF Wass./Abwass. (122): 6.

Glaus, S. and G. Calzaferri (1999). "Silver chloride clusters and surface states." Journal of Physical Chemistry B 103(27): 5622-5630.

Glaus, S. and G. Calzaferri (2003). "The band structures of the silver halides AgF, AgCl, and AgBr: A comparative study." Photochemical & Photobiological Sciences 2(4): 398-401.

Gollmer, A., J. Arnbjerg, F. H. Blaikie, B. W. Pedersen, T. Breitenbach, K. Daasbjerg, M. Glasius and P. R. Ogilby (2011). "Singlet Oxygen Sensor Green (R): Photochemical Behavior in Solution and in a Mammalian Cell." Photochemistry and Photobiology 87(3): 671-679.

Gollmer, A., J. Arnbjerg, F. H. Blaikie, B. W. Pedersen, T. Breitenbach, K. Daasbjerg, M. Glasius and P. R. Ogilby (2011). "Singlet Oxygen Sensor Green: Photochemical Behavior in Solution and in a Mammalian Cell." Photochemistry and Photobiology 87: 671-679.

Greene, C. H. and L. D. Frizzell (1936). "Studies of the precipitation of silver chloride. II. From silver nitrate and hydrochloric acid." Journal of the American Chemical Society 58: 516-522.

Greenwood, R. and K. Kendall (1999). "Selection of Suitable Dispersants for Aqueous Suspensions of Zirconia and Titania Powders using Acoustophoresis." Journal of the European Ceramic Society 19(4): 479-488.

Guo, J., B. Ma, A. Yin, K. Fan and W. Dai (2011). "Photodegradation of rhodamine B and 4-chlorophenol using plasmonic photocatalyst of Ag-

171

AgI/Fe3O4@SiO2 magnetic nanoparticle under visible light irradiation." Applied Catalysis B: Environmental 101(3-4): 7.

Guo, J. F., B. W. Ma, A. Y. Yin, K. N. Fan and W. L. Dai (2012). "Highly stable and efficient Ag/AgCl@TiO2 photocatalyst: Preparation, characterization, and application in the treatment of aqueous hazardous pollutants." Journal of Hazardous Materials 211: 77-82.

Guo, R., G. K. Zhang and J. Liu (2013). "Preparation of Ag/AgCl/BiMg2VO6 composite and its visible-light photocatalytic activity." Materials Research Bulletin 48(5): 1857-1863.

Hamilton, J. F. (1974). "Physical-Properties of Silver-Halide Microcrystals." Photographic Science and Engineering 18(5): 493-500.

Hanaor, D., M. Michelazzi, C. Leonelli and C. C. Sorrell (2012). "The effects of carboxylic acids on the aqueous dispersion and electrophoretic deposition of ZrO2." Journal of the European Ceramic Society 32(1): 235-244.

He, D., M. W. Bligh and T. D. Waite (2013). "Effects of Aggregate Structure on the Dissolution Kinetics of Citrate-Stabilized Silver Nanoparticles." Environmental Science & Technology 47(16): 9148-9156.

He, D., S. Garg, A. Jones and T. D. Waite (2012). "H2O2-mediated oxidation of zero-valent silver and resultant interactions between silver nanoparticles, silver ions and reactive oxygen species." Langmuir 28: 10.

He, D., A. M. Jones, S. Garg, A. N. Pham and T. D. Waite (2011). "Silver Nanoparticle-Reactive Oxygen Species Interactions: Application of a Charging- Discharging Model." Journal of Physical Chemistry C 115(13): 5461-5468.

He, D., A. M. Jones, S. Garg, A. N. Pham and T. D. i. p. Waite, accepted February 2011). (2011). "Silver nanoparticle - reactive oxygen species interactions: Application of an electron charging-discharging model." Journal of Physical Chemistry C 115: 5461-5468.

He, D., A. M. Jones and T. D. Waite (2011). "Silver Nanoparticle-Reactive Oxygen Species Interactions: Application of a Charging-Discharging Model." The Journal of Physical Chemistry 115: 8.

Held, A. M., D. J. Halko and J. K. Hurst (1978). "Mechanisms of Chlorine Oxidation of Hydrogen-Peroxide." Journal of the American Chemical Society 100(18): 5732-5740.

172

Hoffmann, M. R., S. T. Martin, W. Y. Choi and D. W. Bahnemann (1995). "Environmental Applications of Semiconductor Photocatalysis." Chemical Reviews 95(1): 69-96.

Hou, J. G., Z. Wang, C. Yang, L. Zhou, S. Q. Jiao and H. M. Zhu (2013). "Hierarchically Plasmonic Z-Scheme Photocatalyst of Ag/AgCl Nanocrystals Decorated Mesoporous Single-Crystalline Metastable Bi20TiO32 Nanosheets." Journal of Physical Chemistry C 117(10): 5132-5141.

Hu, C., T. W. Peng, X. X. Hu, Y. L. Nie, X. F. Zhou, J. H. Qu and H. He (2010). "Plasmon-Induced Photodegradation of Toxic Pollutants with Ag-Agl/Al(2)O(3) under Visible-Light Irradiation." Journal of the American Chemical Society 132(2): 857-862.

Hu, J., N. Jia, J. X. Jiang, M. G. Ma, J. F. Zhu, R. C. Sun and J. Z. Li (2011). "Hydrothermal preparation of boehmite-doped AgCl nanocubes and their characterization." Materials Letters 65(11): 1531-1534.

Hu, P., X. L. Hu, C. J. Chen, D. F. Hou and Y. H. Huang (2014). "Biomaterial- assisted synthesis of AgCl@Ag concave cubes with efficient visible-light- driven photocatalytic activity." Crystengcomm 16(4): 649-653.

Hunter, R. J. (2001). Foundations of Colloid Science. the United States, Oxford University Press Inc., New York.

Hurst, J. K., J. M. Albrich, T. R. Green, H. Rosen and S. Klebanoff (1984). "Myeloperoxidase-dependent fluorescein chlorination by stimulated neutrophils." Journal of Biological Chemistry 259(8): 4812-4821.

Jiang, J. and L. Z. Zhang (2011). "Rapid Microwave-Assisted Nonaqueous Synthesis and Growth Mechanism of AgCl/Ag, and Its Daylight-Driven Plasmonic Photocatalysis." Chemistry-a European Journal 17(13): 3710-3717.

Jones, A. M., S. Garg, D. He, A. N. Pham and T. D. Waite (2011). "Superoxide- Mediated Formation and Charging of Silver Nanoparticles." Environmental Science & Technology 45(4): 1428-1434.

Kakuta, N., N. Goto, H. Ohkita and T. Mizushima (1999). "Silver bromide as a photocatalyst for hydrogen generation from CH3OH/H2O solution." Journal of Physical Chemistry B 103(29): 5917-5919.

Kang, M. G., H. E. Han and K. J. Kim (1999). "Enhanced photodecomposition of 4-chlorophenol in aqueous solution by deposition of CdS on TiO2." Journal of Photochemistry and Photobiology a-Chemistry 125(1-3): 119-125.

173

Kim, S., H. Chung, J. H. Kwon, H. G. Yoon and W. Kim (2010). "Facile Synthesis of Silver Chloride Nanocubes and Their Derivatives." Bulletin of the Korean Chemical Society 31(10): 2918-2922.

Kimijima, K. and T. Sugimoto (2004). "Growth mechanism of AgCl nanoparticles in a reverse micelle system." Journal of Physical Chemistry B 108(12): 3735-3738.

Kinnan, M. K. (2008). Fabrication and optical properties of noble metal nanostructures PhD, Clemson University.

Klein, D. H. and L. Gordon (1959). "Nucleation and Precipitation of Silver Chloride from Homogeneous Solution." Talanta 3: 10.

Koppel, D. E. (1972). "Analysis of Macromolecular Polydispersity in Intensity Correlation Spectroscopy - Method of Cumulants." Journal of Chemical Physics 57(11): 4814-&.

Kumbhar, A. S., M. K. Kinnan and G. Chumanov (2005). "Multipole plasmon resonances of submicron silver particles." Journal of the American Chemical Society 127(36): 12444-12445.

Lanz, M., D. Schurch and G. Calzaferri (1999). "Photocatalytic oxidation of water to O2 on AgCl-coated electrodes " Journal of Photochemistry and Photobiology A : Chemistry 120: 105-117.

Lanz, M., D. Schurch and G. Calzaferri (1999). "Photocatalytic oxidation of water to O-2 on AgCl-coated electrodes." Journal of Photochemistry and Photobiology a-Chemistry 120(2): 105-117.

Li, J. W., K. Pfanner and G. Calzaferri (1995). "Silver-Zeolite-Modified Electrodes - an Intrazeolite Electron-Transport Mechanism." Journal of Physical Chemistry 99(7): 2119-2126.

Liang, J., G. Q. Zhu, P. Liu and C. Xu (2014). "Highly Visible-Light- Responsive Photocatalytic AgCl/BiOCl Hetero-Nanostructures Synthesized by a Chemical Coprecipitation Method." Journal of Nanoscience and Nanotechnology 14(6): 4185-4190.

Lin, M. Y., H. M. Lindsay, D. A. Weitz, R. C. Ball, R. Klein and P. Meakin (1990). "Universal Reaction-Limited Colloid Aggregation." Physical Review A 41(4): 2005-2020.

174

Lin, M. Y., H. M. Lindsay, D. A. Weitz, R. Klein, R. C. Ball and P. Meakin (1990). "Universal Diffusion-Limited Colloid Aggregation." Journal of Physics- Condensed Matter 2(13): 3093-3113.

Liu, C. P., H. L. Lin, S. M. Gao, P. Yin, L. Guo, B. B. Huang and Y. Dai (2014). "AgI/AgCl/H2WO4 Double Heterojunctions Composites: Preparation and Visible-Light Photocatalytic Performance." Bulletin of the Korean Chemical Society 35(2): 441-447.

Liu, J. and R. H. Hurt (2010). "Ion release kinetics and particle persistence in aqueous nano-silver colloids." Environmental Science and Technology 44(6): 2169-2175.

Liu, J. Y. and R. H. Hurt (2010). "Ion Release Kinetics and Particle Persistence in Aqueous Nano-Silver Colloids." Environmental Science & Technology 44(6): 2169-2175.

Logan, S. R. (1989). "Redox Reactions of Organic Radicals with Ferrocene Ferricenium Species in Aqueous-Solution .1. Radicals Derived from Carboxylic-Acids." Journal of the Chemical Society-Perkin Transactions 2(7): 751-754.

Long, C. A. and B. H. J. Bielski (1980). "Rate of Reaction of Superoxide Radical with Chloride-Containing Species." Journal of Physical Chemistry 84(5): 555-557.

Lou, Z. Z., B. B. Huang, P. Wang, Z. Y. Wang, X. Y. Qin, X. Y. Zhang, H. F. Cheng, Z. K. Zheng and Y. Dai (2011). "The synthesis of the near-spherical AgCl crystal for visible light photocatalytic applications." Dalton Transactions 40(16): 4104-4110.

Ma, B. W., J. F. Guo, W. L. Dai and K. N. Fan (2012). "Ag-AgCl/WO3 hollow sphere with flower-like structure and superior visible photocatalytic activity." Applied Catalysis B-Environmental 123: 193-199.

Ma, B. W., J. F. Guo, L. Y. Zou, W. L. Dai and K. N. Fan (2011). "Ag/AgCl@Cotton-fabric: A Highly Stable and Easy-recycling Plasmonic Photocatalyst under Visible Light Irradiation." Chinese Journal of Chemistry 29(4): 857-859.

Mahmoud, M. A. and M. A. El-Sayed (2008). "Comparative study of the assemblies and the resulting plasmon fields of Langmuir−Blodgett assembled

175

monolayers of silver nanocubes and gold nanocages." The Journal of Physical Chemistry C 112(37): 14618-14625.

Mchedlov-petrossyan, N. O., V. I. Kukhtik and V. I. Alekseeva (1994). "Ionization and Tautomerism of Fluorescein, Rhodamine-B, N,N-Diethylrhodol and Related Dyes in Mixed and Nonaqueous Solvents." Dyes and Pigments 24(1): 11-35.

McNaught, A. D. and A. Wilkinson. (1997). "IUPAC Compendium of Chemical Terminology - Electrokinetic potential." Retrieved April 19, 2014, from http://goldbook.iupac.org/E01968.html.

Miller, C. J., A. L. Rose and T. D. Waite (2011). "Phthalhydrazide Chemiluminescence Method for Determination of Hydroxyl Radical Production: Modifications and Adaptations for Use in Natural Systems." Analytical Chemistry 83(1): 261-268.

Miller, C. J., H. Yu and T. D. Waite (2013). "Degradation of rhodamine B during visible light photocatalysis employing Ag@AgCl embedded on reduced graphene oxide." Colloids and Surfaces A: Physiochem. Eng. Aspects 435: 147- 153.

Miller, C. J., H. J. Yu and T. D. Waite (2013). "Degradation of rhodamine B during visible light photocatalysis employing Ag@AgCl embedded on reduced graphene oxide." Colloids and Surfaces a-Physicochemical and Engineering Aspects 435: 147-153.

Min, Y. L., G. Q. He, Q. J. Xu and Y. C. Chen (2014). "Self-assembled encapsulation of graphene oxide/Ag@AgCl as a Z-scheme photocatalytic system for pollutant removal." Journal of Materials Chemistry A 2(5): 1294- 1301.

Mock, J. J., M. Barbic, D. R. Smith, D. A. Schultz and S. Schultz (2002). "Shape effects in plasmon resonance of individual colloidal silver nanoparticles." J. Chem. Phys. 116(15): 6755-6759.

Morel, F. M. M. and J. G. Hering (1993). Principles and Applications of Aquatic Chemistry. New York, the United States of America, John Wiley & Sons, Inc.

Morgan, M. S., P. F. Van Trieste, S. M. Garlick, M. J. Mahon and A. L. Smith (1988). "Ultraviolet molar absorptivities of aqueous hydrogen peroxide and hydroperoxyl ion." Analytica Chimica Acta 215: 325-329.

176

Moser, F., N. R. Nail and F. Urbach (1959). "Optical Absorption Studies of the Volume Photolysis of Large Silver Chloride Crystals." Journal of Physics and Chemistry of Solids 9(3-4): 217-&.

Mulvaney, P. (1996). "Surface plasmon spectroscopy of nanosized metal particles." Langmuir 12(3): 788-800.

Noyes, A. A. and G. T. Cottle (1898). The Velocity of the reaction between silver acetate and sodium formate: A reaction of third order. the New York meeting of the American Chemical Society.

Owen, A. J. (1995). "Quantitative UV-Visible Analysis in the Presence of Scattering " Retrieved 28th January 2014.

Padervand, M., H. Salari, S. Ahmadvand and M. R. Gholami (2012). "Removal of an organic pollutant from waste water by photocatalytic behavior of AgX/TiO2 loaded on mordenite nanocrystals." Research on Chemical Intermediates 38(8): 1975-1985.

Peng, S. and Y. G. Sun (2011). "Ripening of bimodally distributed AgCl nanoparticles." Journal of Materials Chemistry 21(31): 11644-11650.

Pourahmad, A., S. Sohrabnezhad and E. Kashefian (2010). "AgBr/nanoAlMCM-41 visible light photocatalyst for degradation of methylene blue dye." Spectrochimica Acta Part A 77: 7.

Qi, L. F., J. G. Yu, G. Liu and P. K. Wong (2014). "Synthesis and photocatalytic activity of plasmonic Ag@AgCl composite immobilized on titanate nanowire films." Catalysis Today 224: 193-199.

Quinten, M. (1999). "Optical effects associated with aggregates of clusters." Journal of Cluster Science 10(2): 319-358.

Quinten, M. and U. Kreibig (1993). "Absorption and elastic scattering of light by particle aggregates." Applied Optics 32(30): 6173-6182.

Rehan, M., A. Hartwig, M. Ott, L. Gatjen and R. Wilken (2013). "Enhancement of photocatalytic self-cleaning activity and antimicrobial properties of poly(ethylene terephthalate) fabrics." Surface & Coatings Technology 219: 50- 58.

Rodrigues, S., S. Uma, I. N. Martyanov and K. J. Klabunde (2005). "AgBr/Al- MCM-41 visible-light photocatalyst for gas-phase decomposition of CH3CHO." Journal of Catalysis 233(2): 405-410.

177

Russell, B. K., J. G. Mantovani, V. E. Anderson, R. J. Warmack and T. L. Ferrell (1987). "Experimental test of the Mie theory for microlithographically produced silver spheres." Physical Review B 35(5): 2151-2154.

Schurch, D., A. Currao, S. Sarkar, G. Hodes and G. Calzaferri (2002). "The silver chloride photoanode in photoelectrochemical water splitting." Journal of physical Chemistry B 106: 12764-12775.

Scop, P. M. (1965). "Band structure of silver chloride and silver bromide." Physical Review 139(3A): A934-A940.

Shaul, G. M., T. J. Holdsworth, C. R. Dempsey and K. A. Dostal (1991). "Fate of Water-Soluble Azo Dyes in the Activated-Sludge Process." Chemosphere 22(1-2): 107-119.

Sohrabnezhad, S. and A. Pourahmad (2012). "As-synthesis of nanostructure AgCl/Ag/MCM-41 composite." Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy 86: 271-275.

Sun, J. C., Y. B. Zhang, J. Cheng, H. Fan, J. Y. Zhu, X. Wang and S. Y. Ai (2014). "Synthesis of Ag/AgCl/Zn-Cr LDHs composite with enhanced visible- light photocatalytic performance." Journal of Molecular Catalysis a-Chemical 382: 146-153.

Sun, Y. G. (2010). "Conversion of Ag Nanowires to AgCl Nanowires Decorated with Au Nanoparticles and Their Photocatalytic Activity." Journal of Physical Chemistry C 114(5): 2127-2133.

Tang, Y. X., V. P. Subramaniam, T. H. Lau, Y. K. Lai, D. G. Gong, P. D. Kanhere, Y. H. Cheng, Z. Chen and Z. L. Dong (2011). "In situ formation of large-scale Ag/AgCl nanoparticles on layered titanate honeycomb by gas phase reaction for visible light degradation of phenol solution." Applied Catalysis B- Environmental 106(3-4): 577-585.

Tausch-Treml, R., A. Henglein and J. Lilie (1978). "Reactivity of silver atoms in aqueous solution II. A pulse radiolysis study." Berichte der Bunsengesellschaft für physikalische Chemie 82(12): 1335-1343.

Thompson, A., I. Lindau, D. Attwood, Y. Liu, E. Gullikson, P. Pianetta, M. Howells, A. Robinson, K.-J. Kim, J. Scofield, J. Kirz, J. Underwood, J. Kortright, G. Williams and H. Winick (2009). Center for X-Ray Optics and Advanced Light Source: X-Ray Data Booklet. Berkeley, CA, Lawrence Berkely National Laboratory, University of California.

178

Tian, B. Z. and J. L. Zhang (2012). "Morphology-Controlled Synthesis and Applications of Silver Halide Photocatalytic Materials." Catalysis Surveys from Asia 16(4): 210-230.

Turro, N. J. (1991). Singlet oxygen and chemiluminescent organic reactions in Modern Molecular Photochemistry. Sausalito, CA, University Science Books.

Ungelenk, J. and C. Feldmann (2012). "Adjustable kinetics in heterogeneous photocatalysis demonstrating the relevance of electrostatic interactions." Applied Catalysis B-Environmental 127: 11-17.

Wang, J. Z., C. H. An, J. X. Liu, G. Z. Xi, W. Jiang, S. T. Wang and Q. H. Zhang (2013). "Graphene oxide coupled AgBr nanosheets: an efficient dual- functional visible-light-responsive nanophotocatalyst with enhanced performance." Journal of Materials Chemistry A 1(8): 2827-2832.

Wang, J. Z., C. H. An, M. Y. Zhang, C. Qin, X. J. Ming and Q. H. Zhang (2012). "Photochemical conversion of AgCl nanocubes to hybrid AgCl-Ag nanoparticles with high activity and long-term stability towards photocatalytic degradation of organic dyes." Canadian Journal of Chemistry-Revue Canadienne De Chimie 90(10): 858-864.

Wang, P., B. Huang, X. Qin, X. Zhang, Y. Dai, J. Wei and M. H. Whangbo (2008). "Ag@AgCl: A highly efficient and stable photocatalyst active under visible light." Angewandte Chemie 47: 7931-7933.

Wang, P., B. Huang, X. Zhang, X. Qin, Y. Dai, H. Jin, J. Wei and M. H. Whangbo (2008). "Composite semiconductor H2WO4.H2O/AgCl as an efficient and stable photocatalyst under visible light." Chem.Eur.J. 14: 10543-10546.

Wang, P., B. B. Huang, X. Y. Qin, X. Y. Zhang, Y. Dai, J. Y. Wei and M. H. Whangbo (2008). "Ag@AgCl: A highly efficient and stable photocatalyst active under visible light." Angewandte Chemie-International Edition 47(41): 7931- 7933.

Wang, P., B. B. Huang, X. Y. Zhang, X. Y. Qin, Y. Dai, H. Jin, J. Y. Wei and M. H. Whangbo (2008). "Composite Semiconductor H(2)WO(4)center dot H(2)O/AgCl as an Efficient and Stable Photocatalyst under Visible Light." Chemistry-a European Journal 14(34): 10543-10546.

Wang, P., T. S. Ming, G. H. Wang, X. F. Wang, H. G. Yu and J. G. Yu (2014). "Cocatalyst modification and nanonization of Ag/AgCl photocatalyst with

179

enhanced photocatalytic performance." Journal of Molecular Catalysis a- Chemical 381: 114-119.

Wang, P. Q., Y. Bai, J. Y. Liu, Z. Fan and Y. Q. Hu (2012). "Facile synthesis and activity of daylight-driven plasmonic catalyser Ag/AgX (X = Cl, Br)." Micro & Nano Letters 7(8): 838-841.

Wang, W. S., H. Du, R. X. Wang, T. Wen and A. W. Xu (2013). "Heterostructured Ag3PO4/AgBr/Ag plasmonic photocatalyst with enhanced photocatalytic activity and stability under visible light." Nanoscale 5(8): 3315- 3321.

Wang, X. P., Y. X. Tang, Z. Chen and T. T. Lim (2012). "Highly stable heterostructured Ag-AgBr/TiO2 composite: a bifunctional visible-light active photocatalyst for destruction of ibuprofen and bacteria." Journal of Materials Chemistry 22(43): 23149-23158.

Wang, Z. Y., B. B. Huang, Y. Dai, P. Wang, Z. K. Zheng and H. F. Cheng (2010). "Relationship between microstructure and photocatalytic properties of nanomaterials." Zeitschrift Fur Kristallographie 225(11): 520-527.

Watanabe, T., T. Takizawa and K. Honda (1977). "Photo-Catalysis through Excitation of Adsorbates .1. Highly Efficient N-Deethylation of Rhodamine B Adsorbed to Cds." Journal of Physical Chemistry 81(19): 1845-1851.

WebElements. (2013). "Silver compounds : silver chloride." Retrieved the 17th September, 2013, from http://www.webelements.com/compounds/silver/silver_chloride.html.

Wen, B., J. Ma, C. Chen, W. Ma, H. Zhu and J. Zhao (2011). "Supported noble metal nanoparticles as photo/sono-catalysts for synthesis of chemicals and degradation of pollutants." Science China-Chemistry 54(6): 887-897.

Wen, Y. Y. and H. M. Ding (2011). "Preparation and Photocatalytic Activity of Ag@AgCl Modified Anatase TiO(2) Nanotubes." Chinese Journal of Catalysis 32(1): 36-45.

Xiong, W., Q. D. Zhao, X. Y. Li and D. K. Zhang (2011). "One-step synthesis of flower-like Ag/AgCl/BiOCl composite with enhanced visible-light photocatalytic activity." Catalysis Communications 16(1): 229-233.

Xu, H., H. M. Li, J. X. Xia, S. Yin, Z. J. Luo, L. Liu and L. Xu (2011). "One- Pot Synthesis of Visible-Light-Driven Plasmonic Photocatalyst Ag/AgCl in Ionic Liquid." Acs Applied Materials & Interfaces 3(1): 22-29.

180

Xu, H., Y. H. Song, L. Liu, H. M. Li, Y. G. Xu, J. X. Xia, X. Y. Wu and S. W. Zhao (2012). "Plasmonic-enhanced visible-light-driven photocatalytic activity of Ag-AgBr synthesized in reactable ionic liquid." Journal of Chemical Technology and Biotechnology 87(12): 1626-1633.

Xu, Y., H. Xu, H. Li, J. Xia, C. Liu and L. Liu (2011). "Enhanced photocatalytic activity of new photocatalayst Ag/AgCl/ZnO." Journal of Alloys and Compounds 509: 3286-3292.

Xu, Y. G., H. Xu, H. M. Li, J. X. Xia, C. T. Liu and L. Liu (2011). "Enhanced photocatalytic activity of new photocatalyst Ag/AgCl/ZnO." Journal of Alloys and Compounds 509(7): 3286-3292.

Xu, Y. G., H. Xu, H. M. Li, J. Yan, J. X. Xia, S. Yin and Q. Zhang (2013). "Ionic liquid oxidation synthesis of Ag@AgCl core-shell structure for photocatalytic application under visible-light irradiation." Colloids and Surfaces a-Physicochemical and Engineering Aspects 416: 80-85.

Xu, Y. G., H. Xu, J. Yan, H. M. Li, L. Y. Huang, Q. Zhang, C. J. Huang and H. L. Wan (2013). "A novel visible-light-response plasmonic photocatalyst CNT/Ag/AgBr and its photocatalytic properties." Physical Chemistry Chemical Physics 15(16): 5821-5830.

Yan, H. J., S. T. Kochuveedu, L. N. Quan, S. S. Lee and D. H. Kim (2013). "Enhanced photocatalytic activity of C, F-codoped TiO2 loaded with AgCl." Journal of Alloys and Compounds 560: 20-26.

Yang, Y. Q., G. K. Zhang and W. Xu (2012). "Facile synthesis and photocatalytic properties of Ag-AgCl-TiO2/rectorite composite." Journal of Colloid and Interface Science 376: 217-223.

Ye, L. Q., J. Y. Liu, C. Q. Gong, L. H. Tian, T. Y. Peng and L. Zan (2012). "Two Different Roles of Metallic Ag on Ag/AgX/BiOX (X = Cl, Br) Visible Light Photocatalysts: Surface Plasmon Resonance and Z-Scheme Bridge." Acs Catalysis 2(8): 1677-1683.

Yi, G., C. F. Du, Y. G. Su and Z. L. Liu (2012). "Synthesis, characterization and photocatalytic activity of AgCI loaded on TiO2/SBA-15 composites." Advanced Materials and Processes Ii, Pts 1-3 557-559: 249-253.

Young, C., T. M. Lim, K. Chiang, J. Scott and R. Amal (2008). "Photocatalytic oxidation of toluene and trichloroethylene in the gas-phase by metallised (Pt, Ag) titanium dioxide." Applied Catalysis B-Environmental 78(1-2): 1-10.

181

Yu, H. G., L. Liu, X. F. Wang, P. Wang, J. G. Yu and Y. H. Wang (2012). "The dependence of photocatalytic activity and photoinduced self-stability of photosensitive AgI nanoparticles." Dalton Transactions 41(34): 10405-10411.

Yu, H. J., C. J. Miller, A. Ikeda-Ohno and T. D. Waite (2014). "Photodegradation of contaminants using Ag@AgCl/rGO assemblages: Possibilities and limitations." Catalysis Today 224: 122-131.

Yu, J. G., G. P. Dai and B. B. Huang (2009). "Fabrication and Characterization of Visible-Light-Driven Plasmonic Photocatalyst Ag/AgCl/TiO2 TiO2 Nanotube Arrays." Journal of Physical Chemistry C 113(37): 16394-16401.

Zhang, H., X. F. Fan, X. Quan, S. Chen and H. T. Yu (2011). "Graphene Sheets Grafted Ag@AgCl Hybrid with Enhanced Plasmonic Photocatalytic Activity under Visible Light." Environmental Science & Technology 45(13): 5731-5736.

Zhang, J. Z., X. Liu, S. M. Gao, B. B. Huang, Y. Dai, Y. B. Xu, L. R. Grabstanowicz and T. Xu (2013). "From AgI/TiO2 to Ag/TiO2: effects of the annealing temperature on the compositions, porous nanostructures, and visible- light photocatalytic properties." Ceramics International 39(2): 1011-1019.

Zhang, Q., C. G. Tian, A. P. Wu, Y. Hong, M. X. Li and H. G. Fu (2013). "In situ oxidation of Ag/ZnO by bromine water to prepare ternary Ag-AgBr/ZnO sunlight-derived photocatalyst." Journal of Alloys and Compounds 563: 269- 273.

Zhang, X., J. Li, X. Lu, Z. Tang and G. Lu (2012). "Visible light induced CO2 reduction and RhB decolorization over electrostatic-assembled AgBr/palygorskite." Journal of Colloid and Interface Science 377: 7.

Zhang, Z. Z., X. X. Wang, J. L. Long, Z. X. Ding, X. L. Fu and X. Z. Fu (2010). "H-2-O-2 promoting effect on photocatalytic degradation of organic pollutants in an aqueous solution without an external H-2 supply." Applied Catalysis a- General 380(1-2): 178-184.

Zhao, C., S. Y. Yang, Z. T. Liu and Y. P. Fang (2012). "AgCl Loaded Mesoporous Graphitic Carbon Nitride as Visible Light Photocatalyst." Advances in Environmental Science and Engineering, Pts 1-6 518-523: 54-58.

Zhao, Y. Y., L. Kuai and B. Y. Geng (2012). "Low-cost and highly efficient composite visible light-driven Ag-AgBr/gamma-Al2O3 plasmonic photocatalyst for degrading organic pollutants." Catalysis Science & Technology 2(6): 1269-1274.

182

Zhou, J., Y. Cheng and J. Yu "Preparation and characterization of visible-light- driven plasmonic photocatalyst Ag/AgCl/TiO2 nanocomposite thin films." Journal of Photochemistry and Photobiology A: Chemistry.

Zhou, J. B., Y. Cheng and J. G. Yu (2011). "Preparation and characterization of visible-light-driven plasmonic photocatalyst Ag/AgCl/TiO2 nanocomposite thin films." Journal of Photochemistry and Photobiology a-Chemistry 223(2-3): 82- 87.

Zhou, Z. J., M. C. Long, W. M. Cai and J. Cai (2012). "Synthesis and photocatalytic performance of the efficient visible light photocatalyst Ag- AgCl/BiVO4." Journal of Molecular Catalysis a-Chemical 353: 22-28.

Zhu, M. S., P. L. Chen and M. H. Liu (2011). "Sunlight-driven plasmonic photocatalysts based on Ag/AgCl nanostructures synthesized via an oil-in-water medium: enhanced catalytic performance by morphology selection." Journal of Materials Chemistry 21(41): 16413-16419.

Zhu, M. S., P. L. Chen and M. H. Liu (2012). "Highly efficient visible-light- driven plasmonic photocatalysts based on graphene oxide-hybridized one- dimensional Ag/AgCl heteroarchitectures." Journal of Materials Chemistry 22(40): 21487-21494.

Zhu, Q., W. S. Wang, L. Lin, G. Q. Gao, H. L. Guo, H. Du and A. W. Xu (2013). "Facile Synthesis of the Novel Ag3VO4/AgBr/Ag Plasmonic Photocatalyst with Enhanced Photocatalytic Activity and Stability." Journal of Physical Chemistry C 117(11): 5894-5900.

Zhu, Y. Q., H. L. Liu, L. B. Yang and J. H. Liu (2012). "Study on the synthesis of Ag/AgCl nanoparticles and their photocatalytic properties." Materials Research Bulletin 47(11): 3452-3458.

183