Analysis of Dissolved Organic Matter and Inorganic Arsenic III/V in Drinking Water

A thesis submitted to the Graduate School of the University of Cincinnati in partial fulfillment of the requirements for the degree of

Masters of Science

in the Department of Chemistry of The McMicken College of Arts and Sciences by

Eugenia C. Riddick

B. S.: University of Cincinnati June 2009

Committee Chair: Apryll M. Stalcup, Ph. D.

Abstract

An increase in water conservation awareness and more frequent occurrences of water shortages and high levels of water contaminants across the globe highlight the need for accurate methods of detecting changes in water quality that may affect both existing and future potential contaminants in water. In an effort to elucidate the reactive structures within humic substances (HSs) and their potential role in arsenic

(As) speciation in water, a method of HA analysis in dilute solutions of sodium bicarbonate (NaHCO3) was developed that increased the intensity of the electronic transitions detectable by UV-Vis spectroscopy. The

UV-Vis results indicating the HA analyzed contained chromophores, similar to carbonyl groups, and/or unconjugated aromatic groups, like benzene or phenols, was confirmed using fluorescence spectroscopy.

The potential role humic substances play in the fate of inorganic As III/V in drinking water was investigated using RP-HPLC-ICP-MS to separate and detect the different species of arsenic in water spiked at 1 ppm with dissolved organic matter (cDOM). The results obtained indicated cDOM acted as a preservative, stabilizing species of As III/V in water, even at low arsenic concentrations, similar to ethylenediamine tetraacetic acid (EDTA). These results may be helpful in elucidating persisting contamination issues in both drinking and source waters.

iii

Acknowledgements

This is dedicated to my mother. You’ve always been there, in the background, supporting and encouraging me, putting up with me when I’m surly, distracted and scatterbrained. I know you think I don’t notice and/or appreciate, but I know I wouldn’t have gotten here without you. So, in case you ever read this, I want you to know that, even though you were here first, you’re the best thing to ever happen to me.

I love you.

I would like to thank Dr. Stalcup, for all of her assistance in “helping” me “get it.” Just like my parental unit, I know you don’t think I appreciate it, but I really do, so, thanks!

I would like to thank Dr. Ridgeway and Dr. Heineman for their assistance over the years and especially for agreeing to be on my committee.

Lastly, I would like to thank the EPA and all my co-workers, especially Keith Kelty, Darren Lytle, Heath

Mash, Stephanie Brown, David Griffith, and multiple others who know who they are. You make coming to work every day an adventure!

Table of Contents

Abstract ii

Acknowledgements iv

Table of Contents v

List of Figures ix

List of Tables xii

List of Abbreviations xiii

Chapter 1: Introduction 1

1.0 Introduction 1

1.1 Humic Substances in the Environment 2

1.1.1 Characterization of Humic Substances 4

1.1.2 UV-Vis Spectroscopy 5

1.1.3 Fluorescence Spectroscopy 7

1.1.4 Fourier Transform - Infra-Red (FT-IR), Nuclear Magnetic Resonance (NMR) and Mass Spectroscopy (MS) 12

1.1.5 Variation in Chromatography Analysis Results 14

1.2 Arsenic in the Environment 17

1.2.1 Arsenic Removal During Water Treatment and Processing 19

1.2.2 Interferences in Arsenic Removal During Water Treatment 21

1.2.3 Potential Health Effects of Arsenic Exposure in Mammals 22

1.2.4 Methods of Arsenic Speciation Analysis 23

1.2.5 Spectroscopic Methods of Detecting Arsenic Species 24

1.2.6. Molecular Absorptive Spectroscopic Arsenic Speciation 24

1.2.7 Atomic Adsorptive Spectroscopic Arsenic Speciation 25

1.2.8 Electrochemical Arsenic Speciation 26

1.2.9 High Performance Liquid Chromatography (HPLC) - Inductively Coupled Plasma (ICP) 28 1.3 Research Performed 29

Chapter 2: Developing a Method for Dissolved Organic Matter Characterization and Arsenic

III/V Detection in Water Matrices 34

2.0 Experimental Introduction 34

2.1 Instrumentation 34

2.1.1 UV-Vis Spectroscopy 34

2.1.2 Fluorescence Spectroscopy 35

2.1.3 Total Organic Carbon (TOC) 35

2.1.4 Inductively-Coupled Plasma Atomic Emission (ICP-AES) Spectroscopy 35

2.1.5 HPLC-ICP-MS 37

2.2 Reagents, Solutions and Miscellaneous Labware 37

2.2.1 NaHCO3 Solutions (10 and 50 mM) - Dissolved Humic Acid/cDOM

Analysis 38

2.2.2 Mobile Phase (2.5 mM (NH4)H2PO4, 5 mM TBAH, pH 6.0) - Arsenic III/V

Speciation 38

2.3 Sample Preparation 39

2.3.1 Humic acid (HA)/cDOM analysis samples 39

2.3.2 Arsenic III/V speciation analysis samples 39

2.4 Analysis Methods 41

2.4.1 Humic acid/cDOM UV-Vis and Fluorescence Spectroscopy 40

2.4.2 Arsenic III/V HPLC-ICP-MS Speciation 41

2.5 Safety 46

2.5.1 Personal Protective Equipment (PPE) 46

2.5.2 Health and Safety Plans (HASP) 46

2.5.3 Waste Collection and/or Removal 47

Chapter 3: Results 48

3.0 Introduction 48

3.1 UV-Vis and Fluorescence of Humic Substances 48

3.1.1 Humic Acid UV-Vis Spectroscopy 48

3.1.2 Model compounds UV-Vis Spectroscopy 56

3.1.3 Humic Acid Calibration 86

3.1.4 Humic Acid Fluorescence 87

3.1.5 Conclusion 93

3.2 Arsenic III/V Speciation Analysis 95

vii

3.2.1 Detection of As III/V 96

3.2.2 Preservation of As III/V 99

3.2.3 Effect of DOM on As III/V Detection 113

3.2.4 Arsenic III/V Detection in Water 123

Chapter 4: Conclusions and Future Work 129

4.0 Introduction 129

4.1 UV-Vis and Fluorescence Spectroscopy 129

4.2 Arsenic III/V Speciation 132

4.3 Future Work 133

Chapter 5: Supplemental Information 134

Appendix A EPA Methods and Health and Safety Plans (HASPs) -1-

A1-1 EPA Method 415.2 Total Organic Carbon Analysis -2-

A1-2 EPA Method 200.5 ICP-AES Trace Elements Analysis -10-

A1-3 HASP # 2010-056 Speciation of Metals Complexed to Humic Acid in

Natural/Drinking Water -49-

A1-4 HASP # 2012-072 Trace Metal Analysis of Drinking Water -65-

viii

List of Figures

Chapter 1

Figure 1.1-1 Molecular structures of several known DBPs believed to be formed from NOM precursors 4

Figure 1.1.2-1 UV-Vis absorbance of 3 HA fractions 5

Figure 1.1.2-2 Differential spectroscopy UV-Vis of natural water 6

Figure 1.1.3-1 Comparison of fluorescence analysis methods of HSs obtained from International Humic Substances Society (IHSS) 8

Figure 1.1.3-2 UV-Vis absorbance corrected EEMs of HSs obtained from International Humic Substances Society (IHSS) 10

Figure 1.1.3-3 UV-Vis absorbance corrected EEMs of unfractionated wastewater (MWW) and Suwannee River fulvic acid (SRF) obtained from International Humic Substances Society (IHSS) 11

Figure 1.1.3-4 Synchronous fluorescence correlation of Mesa Wastewater (MWW) blended with a Suwannee River fulvic acid (SRF) standard, IHSS 12

Figure 1.1.4-1 UV-Vis and Solid State CP-MAS 13C NMR comparison of FA fractionated using DAX-8 and DEAE resins 13

Figure 1.2-1 Redox potential – pH (Eh-pH) diagram of various arsenic species in solution at pH 0-14 18

Figure 1.2.1-1 Potential water processing route at a water treatment facility 20

Figure 1.2.6-1 Molybdoarsenate – Malachite Green aggregate formation over time at decreasing phosphate concentrations 24

Figure 1.2.7-1 Ion-exchange (cation/anion) chromatography schematic for the separation and detection of arsenic (total), As V, MMAA and DMAA 25

Figure 1.2.8-1 Electropherogram of capillary electrophoresis inductively couple plasma mass spectrometry (CE-ICP-MS) separation and detection of chromium (Cr) III/VI, As III/V, strontium (Sr) and tin (Sn II/IV) 27

Figure 1.2.9-1 Chromatogram of AEC-ICP-MS separation and detection of standard solutions of As III/V, MMAA and DMAA 28

Chapter 2

Figure 2.1.4-1 Energy transition energy level diagram 36

Chapter 3

Figure 3.1.1-1 A & B Changing UV-Vis absorbance detected in HA, varying the pH, in 10 and 50 mM NaHCO3 49

Figure 3.1.1-2 A-D UV-Vis spectra of HA in 10 and 50 mM NaHCO3, varying solution pH 51

Figure 3.1.1-3 A-D Comparison the UV-Vis absorbance detected for HA in 10 and 50 mM NaHCO3, varying solution pH 53

Figure 3.1.2-1 A-F UV-Vis analysis spectra of D-(-)-ribose in 10 and 50 mM NaHCO3 solutions, varying pH 57

Figure 3.1.2-2 1H-decoupled 13C NMR spectra of 13C (C-1) labeled D-ribose at varying NaOH concentrations 61

Figure 3.1.2-3 A-D UV-Vis analysis spectra of γ-butyrolactone (GBL) in 10 and 50 mM NaHCO3 solutions, varying pH 62

Figure 3.1.2-4 A-F UV-Vis analysis spectra of mandelic acid in 10 and 50 mM NaHCO3 solutions, varying pH 65

Figure 3.1.2-5 A-F UV-Vis analysis spectra of catechin in 10 and 50 mM NaHCO3 solutions, varying pH 69

Figure 3.1.2-6 A-F UV-Vis analysis of DL-tyrosine in 10 and 50 mM NaHCO3 solutions, varying pH 73

Figure 3.1.2-7 A-C UV-Vis analysis spectra of 10 and 50 mM NaHCO3 “method” blanks, varying pH 77

Figure 3.1.2-8 A & B “Blank” corrected UV-Vis analysis spectra of ribose in DI water and/or 10 and 50 mM NaHCO3, varying pH 79

Figure 3.2.1-9 A & B “Blank” corrected UV-Vis analysis spectra of γ-butyrolactone (GBL) in 10 and 50 mM NaHCO3, varying pH 80

Figure 3.1.2-10 A & B “Blank” corrected UV-Vis analysis spectra of mandelic acid in 10 and 50 mM NaHCO3, varying pH 82

Figure 3.1.2-11 A & B “Blank” corrected UV-Vis analysis spectra of catechin in 10 and 50 mM NaHCO3, varying pH 83

Figure 3.1.2-12 A & B “Blank” corrected UV-Vis analysis spectra of tyrosine in 10 and 50 mM NaHCO3, varying pH 84

Figure 3.1.3-1 UV-Vis Humic Acid calibration curve in 50 mM NaHCO3, pH 7 at

254 nm 86

Figure 3.1.3-2 Comparing UV-Vis HA detection at varying wavelengths 87

Figure 3.1.4-1 A-C Uncorrected Humic Acid fluorescence detected at 260, 300 and 390 nm in 50 mM NaHCO3 88

Figure 3.1.4-2 A-C “Blank” Corrected Humic Acid fluorescence detected at 260, 300 and 390 nm in 50 mM NaHCO3 91

Figure 3.2.1-1 Ion-Pair-Reversed Phase Chromatography (IP-RPC) separation and ICP-MS instrument detection and calibration of As III and As V at ppb concentrations in solution 96

Figure 3.2.1-2 RP-HPLC-ICP-MS As III/V Instrument Calibration 98

Figure 3.2.2-1 A-D Four week arsenic III stability analyzed by RP-HPLC-ICP-MS 100

Figure 3.2.2-2 A-C Analysis of the stability of As III in unpreserved, ascorbic acid, EDTA and HCl preserved samples over four weeks 104

Figure 3.2.2-3 A-C Comparison the % of oxidation detected in 1, 50 and 150 ppb As III samples over four weeks 106

Figure 3.2.2-4 A-E Individual comparison of the % oxidation detected in 1, 50 and 150 ppb As III samples over four weeks 108

Figure 3.2.2-5 RP-HPLC-ICP-MS As III/V Speciation Method Optimization 112

Figure 3.2.3-1 A-D Four week analysis of arsenic III spiked with concentrated dissolved organic matter (cDOM) analyzed by RP-HPLC-ICP-MS 114

Figure 3.2.3-2 Arsenic III/V ascorbic acid/cDOM blank – showing potential organoarsenic peak elution 117

Figure 3.2.3-3 A-C Four week analysis of arsenic III/V with and without cDOM analyzed by RP- HPLC-ICP-MS 121

Figure 3.2.4-1 A-C RP-HPLC-ICP-MS separation, detection and quantification of As III/V in in drinking water 123

Chapter 5

S.I. Figure 5.1-1 A-C Analysis of the stability of As III spiked with cDOM in unpreserved, ascorbic acid, EDTA and HCl preserved samples over four weeks 135

S. I. Figure 5.1-2 A-C Comparing the % of As III oxidation detected in 1, 50 and 150 ppb As III samples, spiked with 1 ppm cDOM, over four weeks 136

S.I. Figure 5.1-3 Analysis of 1 ppb As III/V in solution spiked with 1 ppm cystine 137

S.I. Figure 5.1-4 A-D Four week arsenic V stability analyzed by RP-HPLC-ICP-MS 138

S.I. Figure 5.1-5 A-D Four week analysis of arsenic V spiked with concentrated dissolved organic matter (cDOM) analyzed by RP-HPLC-ICP-MS 139

List of Tables

Chapter 1

Table 1.2.9-1 Method detection limits in µg∙L-1, calculated as three times the baseline noise at 100 µL injections of a mixture containing As III, V, MMAA, DMAA, AsB and AsC 29

Chapter 2

Table 2.3.2-1 Inorganic arsenic species typically detected in water and/or environmental 42

samples

Table 2.3.2-2 RP-HPLC-ICP-MS As (III/V) speciation/detection conditions 43

Chapter 3

Table 3.2.2-1 RP-HPLC-ICP-MS Analysis of As III stability in unpreserved, ascorbic acid, EDTA and HCl preserved samples over four weeks 102

Table 3.2.2-2 Percentage of As III oxidation detected over time during four week analysis in unpreserved, ascorbic acid, EDTA and HCl preserved samples 105

Table 3.2.2-3 Change in As V concentration detected over time during four week analysis in unpreserved, ascorbic acid, EDTA and HCl preserved samples 111

Table 3.2.3-1 Comparison of the ICP-MS signal response (cps) in As III samples spiked with cDOM to As III samples without cDOM over a four week analysis period in unpreserved, ascorbic acid, EDTA and HCl preserved samples 119

Table 3.2.4-1 Method validation showing the As III/V method applied to drinking and source

water 125

Chapter 5

S.I. Table 5.1-1 ICP-AES metals analysis of concentrated dissolved organic matter (cDOM) in water and humic acid (HA) in DI water and 10 and 50 mM NaHCO3 134

xii

List of Abbreviations:

A. A.: Ascorbic acid

+ As V: arsenate/arsenic 5 (H3AsO4)

+ As III: arsenite/arsenic 3 (HAsO2)

C: carbon

CCA: chromated copper arsenate

CPMAS: Cross-Polarization Magnetic Angle Spinning cps: counts per second

DBPs: disinfection by-products

DDT: dichlorodiphenyltrichloroethane

DMAA: dimethylarsinic acid

DOC: dissolved organic carbon

DOM: dissolved organic matter

EDTA: ethylenediamine tetraacetic acid

Eh: oxidation/reduction potential (V)

EPA: Environmental Protection Agency

ESI: Electro-Spray Ionization

FA: fulvic acid

xiii

GSH: glutathione

GBL: γ-butyrolactone

HPLC: High Performance Liquid Chromatography

HA: humic acid

HSs: humic substances

ICP: Inductively-Coupled Plasma

IHSS: International Humic Substance Society

NaHCO3: sodium bicarbonate

MS: Mass Spectrometry

MDL: method detection limit

MCL: maximum contaminant level

MMAA: mono-methylarsonic acid

NOM: natural organic matter

NMR: nuclear magnetic resonance

O: oxygen

OC: organic carbon ppb: parts per billion ppm: parts per million

xiv

PFA: perfluoro-alkoxy

PVP: polyvinyl pyrrolidine redox: oxidation/reduction

RO: reverse osmosis

RPC: reverse-phased chromatography

TOC: total organic carbon

UF: ultrafiltration

UV-Vis: ultraviolet-visible

WHO: World Health Organization.

Chapter 1

Humic Substances and Arsenic in the Environment

1.0 Introduction

The Safe Drinking Water Act (SDWA) was originally established by Congress in 1974 and allowed the Environmental Protection Agency (EPA) to regulate contaminants in both drinking and drinking source water in the United States. Drinking water contaminants currently monitored by the EPA include microbial contaminants, several heavy metals & metalloids, including arsenic and several disinfection by-products

(DBPs), molecules that form during the disinfection step of water treatment processing prior to final distribution. Humic substances (HSs) dissolved in water, commonly referred to as dissolved organic matter

(DOM), are believed to be the main precursor in DBP formation when these molecules are in source water prior to treatment and disinfection. Interestingly, despite decades of research on HSs, no single molecular structure has ever been confirmed, dissolved or otherwise, with most researchers citing the heterogeneity of HSs molecules as the culprit [1]. In addition to their potential to form DBPs, current research also indicates DOM may mobilize certain metals and metalloids, including inorganic arsenic (As) III and V [2] in water, and force cycling between different oxidation states [3] depending on the analytes’ redox potential.

Work presented in this thesis was performed to determine a general method of characterizing DOM and to elucidate the potential effect, if any, of concentrated DOM (cDOM) [4], on the ability to separate and speciate As III/V using a HPLC-ICP-MS speciation method proposed for drinking water matrices.

Current EPA regulations limiting the concentration of arsenic and DBPs in processed water were imposed, in part, due to research linking the contaminants to adverse human health effects including some neurological effects, skin lesions, anemia, liver problems and cancer [5] when exposure occurred in water.

Increasing concern regarding the severity of these health effects combined with the relatively small cumulative amounts needed for them to occur, ~ 200 ppb As [6] and between 0.010 and 1 ppm [5b]

1 depending upon the DBP, suggest quick, reliable analysis methods that provide accurate water quality measurements with minimal sample preparation in a short period of time are essential.

Research performed during the completion of the work contained in this thesis on DOM and As

III/V in drinking water encompassed a broad range of topics including general water treatment methods, practices and the various analytical measurements performed during treatment. In addition, a variety of commonly used humic substance characterization methods and multiple methods of metal speciation, including various methods employed to preserve metal species, in drinking water matrices for improved analytical analysis were also reviewed.

1.1 Humic Substances

Found in soil, sediment and most natural water sources, humic substances (HSs), also referred to as dissolved and/or natural organic matter (DOM or NOM), were first isolated from water sources in 1919.

Believed to be derivatives of microbial decomposition of plant and animal life, these molecules are widely distributed in the environment. Because HSs are so widely represented in nature they may be further classified, based on their levels of solubility in water, as humin, fulvic (FA) or humic acid (HA). Humins were largely insoluble at all pH’s and generally remained segregated in soils and sediment while FAs were infinitely soluble at all pHs and could be found in any environment. Humic acids (HAs), however, were alkaline soluble, precipitating at or below pH 2.0, and were therefore more reliant on the surrounding redox environment (Eh-pH) to dictate their solubility [3, 7].

Early attempts to characterize HSs revealed highly heterogeneous, amphiphilic molecules with structures and molecular weights that varied greatly with changing season, their surrounding climate and location. This high degree of variability in the form and ultimately function of HSs has been cited as one of the primary reasons it has been so difficult to analyze and characterize these molecules in their native form. Also contributing was HS’s ability to both interact with and complex various metals, including

2 arsenate and arsenite [2, 8], in solution with the reaction, again, being dependent on the surrounding redox environment (Eh-pH), concentration and/or charge or oxidation state of the metal in solution.

Dissolved HSs (DOM and/or NOM) are believed to be active participants in numerous other chemical reactions, many of which are essential to the environment, including redox, aggregation, ion- mobility and photochemical. However, their most notorious potential reaction occurs when disinfectants

(chlorine, chloramines, ozone, etc), are introduced during water treatment, resulting in the formation of disinfection by-products (DBPs) in distributed water. When the first DBPs were isolated from water in

1975, there were only 13 reported. In contrast, Krasner et al. indicated more than six hundred DBPs have been characterized to date, several of which have been implicated in the occurrence of cancer and other adverse health effects in humans [9]. Currently, the most reported instrument utilized in quantifying samples of DOM, and thus indicating potential DBP formation, is the total organic carbon (TOC) analyzer which measures the concentration of organic carbon dissolved in water. Theoretically, this instrument has the potential to indicate the reactive carbon [1] when analysis is applied to HSs isolated from a particular water supply. Chiou et al. used TOC analysis to compare humic and fulvic acids extracted from 3 different rivers, using XAD-8 resin, to similar commercial samples showing higher concentrations of organic carbon

(OC) within a sample of DOM led to increased partitioning of organic molecules into solution [10]. In addition, they showed higher C concentrations in DOM promoted interaction with non-ionic molecules while DOM containing higher concentrations of oxygen (O) had increased interaction with ionic solute molecules [10]. However, the inability to differentiate between potential sources of organic carbon during

TOC analysis, highlight the obvious limitations using to this method alone to fully characterizing native

HSs. Despite these limitations, compiled research obtained using TOC analysis has indicated DOM accounts for 50-90 % of the organic carbon found in natural water sources and several of the DBPs shown in Figure 1.1-1 adapted from [11] .

Figure 1.1-1 Molecular structures of several known DBPs believed to be formed from NOM precursors. Figure adapted from Xie et al. [11]. A. Tri-halomethanes (THMs); B. Haloacetic acids (HAAs); C. (MX) 3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)- furanone; (E-MX) 2-chloro-3-(dichloromethyl)-4-oxobutanoic acid; D. Aldehydes; E. Keto acids.

Increasing occurrence of DBPs combined with the extent and severity of the health effects associated with exposure to these molecules has re-ignited a desire to fully characterize HSs in hopes of elucidating mechanisms of DBP formation to eliminate them from treated water [12].

1.1.1 Characterization of Humic Substances

Investigation of HSs and their role in the environment has led to a general division of the research performed during attempted characterization of these molecules, with two general bodies of research

4 emerging: 1. Analysis of fractionated and 2. Analysis of unfractionated HSs. Regardless of division, multiple spectroscopic methods have been utilized to differentiate and characterize both native and fractionated HSs since their early discovery. Results obtained using several of the analytical methods commonly employed during HS analysis are reviewed and described below.

1.1.2 UV-Vis Spectroscopy

Ultraviolet-visible (UV-Vis) spectroscopy is a popular method for detecting DOM in water [13], however broad, undefined peaks similar to those shown in Figure 1.1.2-1, adapted from Chen et al. [12], were characteristic of the spectra resulting from UV-Vis analysis of DOM.

Figure 1.1.2-1 UV-Vis absorbance of 3 HA fractions. Figure adapted from Chen et al. [12].

Chen et al. used UV-Vis to analyze samples of DOM fractionated based on their suspected concentrations of aromatic carbon-carbon double bond character absorbed onto cross-linked polyvinyl pyrrolidone (PVP) polymer [12]. Chen suggested the spectra, shown in Figure 1.1.2-1 adapted from Chen et al., obtained 260 nm could be used to quantify the relative abundances of aromatic carbon-carbon double bonds in samples of fractionated DOM. Chen theorized, if present, the Π to Π* transitions of substituted benzenes and polyphenols in DOM could be monitored [12] in solution. However, the lack of discernible peaks detected

5 in Figure 1.1.2-1 appear to limit its use beyond determining apparent concentration of DOM samples prepared from solid (commercial) HSs and/or potentially confirming the presence of DOC in water. In addition, most UV-Vis analysis of DOM requires concentrations greater than those normally found in natural water sources, further limiting its potential usefulness in characterizing native HSs in situ.

Korshin et al. reported successful peak resolution analyzing low concentrations of FA in natural water sources using a differential spectroscopy method of UV-Vis analysis in which data, collected at one specific environmental parameter, was used to normalize all other data sets obtained during analysis. Figure

1.1.2-2, adapted from Korshin et al., shows spectra obtained from analysis of natural water, varying pH, with the final analysis results normalized using the absorbance collected at pH 1.0 [14].

Figure 1.1.2-2 Differential spectroscopy UV-Vis of natural water. UV-Vis analysis at pH 1.0 used to calculate the differential absorbance shown for pH 3.0, 5.0 and 7.0. Water collected from Judy Reservoir, DOC concentration of 3.6 mg∙L-1, at a cell length of 5 cm. Figure adapted from Korshin et al. [14].

Korshin proposed this method could be used to model the reactivity of HSs in situ and/or at the low concentrations typical of natural waters potentially predicting DBP formation. However the method used perchloric acid, a known oxidizer of organic matter, during pH manipulation which may have affected the final results and similar spectra could not be generated during the course of this thesis work.

Specific ultraviolet (SUVA) analysis at 254 nm, a method established that normalizes the UV absorbance with concentrations of DOC (mg∙L-1) in samples of bulk water and/or fractionated DOM to determine relative aromaticity of DOM, is another popular method of characterizing DOM in water [11].

Weishaar et al. showed a strong correlation existed between the SUVA254 analysis of 13 DOM samples, isolated using XAD-8 resin, and their corresponding liquid state 13C NMR spectra, indicating SUVA may be a fast, reliable, inexpensive alternative to NMR for determining aromatic content of DOM [15].

1.1.3 Fluorescence Spectroscopy

Fluorescence spectroscopy has also been widely used in the characterization of native HSs [16].

However, the complexity of DOM and the potential variables affecting fluorescence in environmental samples including temperature, solvent, hydrogen bonding, pH, metal ions, etc. highlight potential difficulties when choosing experimental conditions and/or interpreting spectra obtained during analysis.

This is particularly apparent when reviewing fluorescence spectra from literature. One example in Figure,

1.1.3-1 adapted from Senesi et al., shows the variation that can occur in fluorescence spectra when different methods of fluorescence analysis are applied to samples of HSs [16a].

Figure 1.1.3-1 Comparison of fluorescence analysis methods of HSs obtained from International Humic Substances Society (IHSS). Top: Fluorescence excitation spectra. Bottom: Fluorescence synchronous-scan excitation spectra. IHSS HSs displayed: Suwanee River (a – standard; a1 – reference), Nordic aquatic (b), soil (c), peat (d) and leonardite (e). Left: Fulvic acid spectra. Right: Humic acid spectra. Excitation spectra obtained scanning from 270-500 nm fixed at 520 nm emission; synchronous-scan excitation spectra obtained scanning simultaneously λexc 290-550 nm, Δλ = λem – λexc = 18 nm. Figure adapted from Senesi et al. [16a].

Senesi et al. analyzed eleven different HS samples obtained from the International Humic Substances

Society (IHSS) using various fluorescence analysis methods. Figure 1.1.3-1 (top) [16a], of the fluorescence excitation shows, with the exception of the solid HS samples, little to no difference in the peak position and/or appearance/shape of HSs. While the spectra obtained during emission fluorescence (not shown) analysis of the same HS samples revealed even less peak differentiation, showing broad, featureless peaks at a variety of different intensities. However, when Senesi et al. applied synchronous-scan excitation fluorescence, Figure 1.1.3-1 (bottom) [16a], they were able to differentiate, not only between different sources (solid, river and/or aquatic) of FA and HA, but spectra also revealed differences in peak position and intensity that distinguished between FA and HA from the same source. However, it was unclear from the method published by Senesi et al. whether the IHSS sample solutions analyzed were merely dissolved or subjected to some type of fractionation prior to analysis.

Mobed et al. were also able to distinguish between HA and FA using UV-Vis absorbance corrected fluorescence excitation-emission matrices (EEMs) when they analyzed solid and dissolved samples of HSs prepared using IHSS reference and standard materials. Figure 1.1.3-2, adapted from Mobed et al., shows bathochromic shifts occurred along the EEM emission axis of HA compared to FA obtained from the same source [16b].

Figure 1.1.3-2 UV-Vis absorbance corrected EEMs of HSs obtained from International Humic Substances Society (IHSS). Top figures correspond to aquatic HSs: NA (Nordic aquatic) FAR – fulvic acid reference; NAHAR – humic acid reference; SR (Suwannee River) FAS – fulvic acid standard; SRHAS – humic acid standard. Bottom figures correspond to soil HSs: P (peat) FAR – fulvic acid reference; PHAR – humic acid reference; S (soil) HAR – humic acid reference. Samples analyzed at pH 6.0. Figure adapted from Mobed et al. [16b].

Mobed suggested the red shift detected for HA may have resulted from certain molecular effects, e.g. electron withdrawing functional groups (carbonyl, amino, hydroxides etc.), higher conjugation and/or the presence of high molecular weights groups in HA compared to FA [16b]. Various solvent effects, including changes in polarity and/or pH, have also been shown to shift the fluorescence wavelength [17], affecting charge distribution of molecules in solution. However the aforementioned shift in fluorescence was detected in both solid and dissolved HS samples which made the assumptions posed by Mobed et al. reasonable and indicated methods of EEM fluorescence may also be useful in differentiating dissolved HSs from various locations.

Westerhoff et al. also used UV-Vis absorbance corrected fluorescence EEMs while comparing

Suwannee River standards obtained from IHSS to wastewater samples collected from the Northeast Mesa

Wastewater Reclamation Facility. Westerhoff theorized EEM fluorescence could be used, in conjunction with synchronous fluorescence, to differentiate between sources of DOC based on dominant fluorophores

10 detected within samples of HSs similar to those detected in the EEMs shown in Figure 1.1.3-3, adapted from Westerhoff et al. [16c].

Figure 1.1.3-3. UV-Vis absorbance corrected EEMs of unfractionated wastewater (MWW) and Suwannee River fulvic acid (SRF) obtained from International Humic Substances Society (IHSS). Samples (MWW & SRF) were diluted to 1 mg∙L-1 organic carbon in 0.01 M KCl and filtered prior to analysis. EEMs shown are in order with and without UV-Vis correction: MWW (top, A. & B.) pH 3.0; SRF (middle, C. & D.) pH 3.0; and SRF (bottom, E. & F.) pH 7.0. Figure adapted from Westerhoff et al. [16c]

Westerhoff et al. a fluorescence intensity maxima, detected along the excitation axis at ~ 280 and 330 nm in the Mesa Wastewater EEM, in Figure 1.6, could be consistent with protein-like bases and/or neutral fractions that contained organic nitrogen groups consistent with wastewater [16c]. They further suggested the fluorescence intensity maxima, corresponding to dominant fluorophores in MWW and SRF EEMs,

11 obtained using synchronous fluorescence analysis of MWW/SRF mixtures could plotted, as shown in

Figure 1.1.3-4, and used to estimate the origin of DOC in an unknown mixture when compared to an IHSS standard [16c].

Figure 1.1.3-4 Synchronous fluorescence correlation of Mesa Wastewater (MWW) blended with a Suwannee River fulvic acid (SRF) standard, IHSS. Fluorescence obtained at 228, 278, 344 and 400 nm (Δλ = 20 nm) with a slit width of 10 nm, using samples prepared at final concentrations of 1 mg∙L-1 DOC, pH 3.0. Figure adapted from Westerhoff et al. [16c].

However, Westerhoff et al. used KCl throughout analysis to dilute all samples analyzed, and while analysis of SRF was shown at pH 3.0 and 7.0, similar analysis of MWW was only shown at pH 3.0. Since HSs are known to react with various forms of chlorine, with the final reaction products dependent upon pH [11] this may have affected the final analysis results. In addition, despite suggestions that EEM/synchronous fluorescence could potentially be used to estimate the percentage of DOC from wastewater in a “clean” water sample if contamination were to occur, it was unclear how this technique would be applied in “real- world” situations.

1.1.4 Fourier Transform - Infra-Red (FT- IR), Nuclear Magnetic Resonance (NMR) and Mass

Spectroscopy (MS)

Nuclear magnetic resonance (NMR) and Mass Spectroscopy (MS) analysis of native and fractionated DOM has also provided some structural characterization of DOM, confirming several of the

12 functional groups believed to be present within these molecule. McDonald et al. used solid state cross- polarization magnetic angle spinning carbon nuclear magnetic resonance (CP-MAS 13C NMR) spectroscopy to compare DOM fractionated using two different size exclusion chromatography (SEC) resins DAX-8 and diethylaminoethyl cellulose (DEAE) [18]. Their analysis showed, despite differences in

UV-Vis spectra, seen in Figure 1.1.4-1 (top) [18], obtained from SEC fractions, four main NMR bands re- occurred, seen in Figure 1.1.4-1 (bottom) [18], indicating aliphatic, o-alkyl, aromatic [19] and carboxylic acid groups [19], were present.

Figure 1.1.4-1 UV-Vis and Solid State CP-MAS 13C NMR comparison of FA fractionated using DAX-8 and DEAE resins. Top : UV-Vis spectra of FA; Bottom: Solid State CP-MAS 13C NMR spectra of FA fractions. Water Samples obtained from Berry Jerry Lagoon and Murrumbidgee River. Figure adapted from McDonald et al. [18].

In addition to aliphatic, o-alkyl, aromatic and carboxylic acid groups, research has indicated several other functional groups may be present in DOM including amino [19a], phenolic [19a], and esteric [19c] groups.

The varying concentration of these functional groups within DOM molecules is often cited as a reason for their reactivity, interactions and/or tendency to complex with and/or bind contaminants [10], potentially mobilizing them within water sources.

1.1.5 Variation in Chromatography Analysis Results

The majority of the methods discussed that have been utilized during the analysis of dissolved HSs required the samples be fractionated and/or separated, using chromatographic methods, prior to analysis.

Chromatography methods employed to fractionate HSs prior to analysis using one of the methods previously mentioned often separate HSs based on molecular size and/or HS hydrophilic/hydrophobicity in solution. However, as previously mentioned HS molecules are highly heterogeneous. This heterogeneity, combined with variation in analysis results obtained with regards to their structures and/or reactive functional groups, highlight potential difficulties choosing appropriate separation parameters. Conte et al. showed some of these variations when they found discrepancies occurred in both UV-Vis and/or refractive index (RI) detection and the fraction volume obtained during high performance size exclusion chromatography (HP-SEC) separation of HSs when small changes were made to the mobile phase (MP) composition [19b]. Examples of chromatograms obtained for two of the three HA samples analyzed are displayed in Figure 1.1.5-1, adapted from Conte et al. [19b], showing the differences in HS detection and volume that occurred when small changes were made to the mobile phase (MP) composition used during separation.

Figure 1.1.5-1 Comparison of UV-Vis (top) to (RI) Refractive Index (bottom) detection of variations in High Performance - Size Exclusion Chromatography (HP-SEC) fractions of humic acid from two different sources. HA1 – North Dakota Leonardite (top/bottom left); HA2 – Danish agricultural soil (top/bottom right). A – control mobile phase: 0.05 M NaNO3, pH 7, I – 0.05; B – same as A + 4.6E-07 M methanol (MeOH), pH 6.97; C – same as A, pH adjusted to 5.54 with HCl; D – same as A + 4.6E-07 M acetic acid, pH 5.69. Figure adapted from Conte et al. [19b].

Conte compared the UV-Vis detection (Figure 1.1.5-1 – top) to the RI detection (Figure 1.1.5-1 – bottom) of North Dakota Leonardite HA (left figures) and HA obtained from a Danish agricultural soil sample (right

15 figures). Conte et al. proposed variation to the volume obtained and the signal detected were due to molecular effects, including association between methanol (MeOH) and HA groups when the MeOH concentration was varied and disruption to negatively charged (carboxylic) functional groups when the pH of the mobile phase was lowered resulting in aggregation and/or disruption of HA association in solution

[19b].

Phillips et al. found differences in the HA fractions obtained by liquid chromatography at the critical condition (LCCC) when they used LCCC to determine the critical conditions of several polymers

(globular proteins, polysaccharides, sodium polystyrene sulfonates and polyacrylic acids) typically used as

SEC calibration standards. After the critical conditions were isolated for each SEC polymer, Phillips et al. proposed chromatograms obtained from HAs injected, and separated, under those same conditions could be analyzed to determine if their elution patterns correlated to those of the SEC polymers. Using Electro-

Spray Ionization - Mass Spectroscopy (ESI-MS) analysis Phillips et al. showed there was some discrepancy between the MS and the SEC isolated fractions with the MS showing molecular weights that were lower than the weights determined by SEC [20]. They suggested these discrepancies, often seen in SEC separation, may have resulted from disruption of HA molecules during ionization prior to mass detection during ESI-MS analysis.

Anion exchange and reversed-phase chromatography (AEC & RPC) have also been used to separate and analyze HSs with researchers reporting irreversible adsorption of HSs to stationary phases. In addition to potential discrepancies introduced analyzing HSs separated based on their size and/or charge, fractionation may irreversibly alter the structure of native HSs, further hindering characterization attempts and conclusions drawn from analysis. Overall, despite multiple years and massive amounts of research on

HSs, DOM and/or NOM, no single analysis method has successfully characterized nor has a single structure for these molecules ever been confirmed for either native or fractionated [21]. This further highlights the difficulty associated with analyzing these molecules and the potential for error when utilizing results obtained from said analyses especially when using only fractionated samples.

1.2 Arsenic in the Environment

For centuries, the diverse chemical properties associated with arsenic (As) led to its use in multiple disciplines including medicine, industrial manufacturing and agriculture [22]. However, the toxicity of arsenic, combined with increasing evidence of hazards it posed to human health, ultimately resulted in its discontinued use in most manufacturing by the late 1980’s [5a]. More recently, the World Health

Organization (WHO) classified arsenic as one of only a few chemicals known to cause large scale “health effects” in humans when repeatedly exposed to the element in water [23]. The severity of the potential

“health effects” associated with chronic exposure to arsenic in water, also led the Environmental Protection

Agency (EPA) to lower its Maximum Contaminant Level (MCL) for the element, from 50 to 10 parts per billion (ppb), in drinking water in 2001 [24].

Widely occurring in nature, though small “native” deposits can be found, most arsenic formed

“cooperatively” on solid supports of sulfur (S), iron (Fe), manganese (Mn), aluminum (Al) or nickel (Ni) in rocks, sediment and soil [22-23, 25]. These cooperatively formed arsenic deposits, located in rivers and creek beds, have been cited as the most abundant potential source of arsenic contamination in surface and drinking waters [22]. Several less common arsenic contamination sources included volcanic ash, airborne dust from smelting plants or contaminated runoff from mines, chromated copper arsenate (CCA) treated wood and/or soil treated with arsenic-based pesticides or fertilizers [22, 25-26].

The stability of solid arsenic deposits in surface and drinking source waters is controlled by a combination of complex, localized, environmental conditions including chemical contamination, levels of aeration, microbial activity, pH and redox potential (Eh). However, the majority of arsenic speciation research performed on rivers and source waters cited Eh-pH as the most important indicator of arsenic species stability and, by default, indicators of potential arsenic mobilization in water [22, 27]. A recent kinetics study, performed by Amirbahman et al., described arsenic speciation as a control of its mobility and toxicity in water, with forms of As III prevalent in reducing and As V prevalent in oxidizing

17 environments [28]. Redox potential and pH (Eh-pH) (see Figure 1.2-1) [29] also dictated which dissolved arsenic species (III or V) was likely to be found within an aqueous environment.

Figure 1.2-1 Redox potential – pH (Eh-pH) diagram of various arsenic species in solution at pH 0-14, adapted from The Atlas of Eh-pH Diagrams [29].

The majority of metals and metalloids surrounding arsenic in the periodic table form positively charged species (cations) when solvated. Arsenic, however, is one of only a few elements that forms inorganic oxyanions. In water, arsenite (As III, +3), has no charge while arsenate (As V, +5), has a negative charge. The Eh-pH diagram for arsenic in Figure 1.2-1 confirms dissolved species of arsenate, As V,

- 2- (H2AsO4 and HAsO4 ; pKa – 6.76 and 11.29 [30]), at drinking water pH (between 6.5-8.5) [31], were typically negatively charged in solution; while solvated inorganic species of arsenite, As III, (HAsO2; pKa

– 9.29 [30]), were neutral or had zero charge [29]. Small amounts of several methylated arsenic forms, most commonly di-methylarsinic acid (DMAA) (CH3)2AsO(OH); pKa – 6.2 [30] and mono-methylarsonic acid (MMAA) (CH3AsO(OH)2; pKa – 4.1 and 8.7 [30], have also been detected in water and some marine life. However inorganic III/V species of arsenic, As III and V, were prevalent in most natural waters.

Research has shown Eh-pH was not only an important predictor of potential arsenic species and/or its

18 dissolution from solid deposits in water but changes in Eh-pH could promote cycling (oxidation-reduction) between the various arsenic species increasing destabilization of solid deposits in water [32]. Further, if less than optimum levels for any of the aforementioned environmental conditions affecting arsenic stability were reached, dissolution and mobilization of arsenic from its solid support into the surrounding water supply might occur.

1.2.1 Arsenic Removal in Water Processing and Processing

Most water processing plans at water treatment facilities include a filtration step in which water comes into contact with some type of filtration media, usually sand or silica, theoretically attracting charged metal ions in solution and retaining them [33]. An example of a typical water processing plan, displayed in Figure 1.2.1-1, showed one potential route a drop of water might take from source to tap.

Figure 1.2.1-1 Potential water processing route at a water treatment facility [34].

The differences in the behavior of solvated arsenic species previously discussed makes the complete removal of these molecules from water during water treatment challenging. Media filtration has been shown to be pH selective in its adsorption of As III/V, when As V in natural waters was shown to adsorb to mineral oxides more readily at low pH while As III adsorbed at neutral pH [28, 35]. These results were not surprising and highlighted the potential difficulty in completely removing As III/V simultaneously during water processing. Silica (sand) based filtration media may contain de-protonated silanols at neutral or more basic pH which would not readily adsorb negatively charged As V molecules in solution but could

20 adsorb As III. The reverse could prove true at lower pH, filtration would adsorb As V more readily than

As III depending upon the filtration media used. However, in general, under normal circumstances and typical drinking water pH, the majority of As V are retained/adsorbed onto multiple filtration surfaces including silica, sand, ferric and/or anion-exchange [25, 33b, 36] similar to the scenario shown in Figure

1.2.1-1. In contrast, under these same conditions, uncharged As III was mostly un-retained by typical filtration surfaces utilized during water treatment [25, 33b, 36]. This posed a unique problem during water treatment if both As III and V were detected in pre-treated water. The tendency of arsenic to form solid supports when interacting with other analytes has been used to potentially alleviate this problem by generating modified filtration media that facilitated complete arsenic removal by oxidizing or retaining As

III during processing [33b]. However, Amirbahman et al. showed the oxidation of arsenic, from As III to

As V, versus its adsorption by certain aquifer materials depended solely upon the filtration medias’ content of manganese (Mn) or iron (Fe) oxides [28]. Also, including these modified medias in a water treatment plan could be time consuming and potentially expensive to maintain as they often required frequent backwashing to remove built-up As III, a step which generated unwanted toxic waste, and/or regeneration of filtration media [33b]. As a potential alternative, chlorination was found to immediately oxidize As III to V in water, facilitating its removal under typical media filtration as shown in Figure 1.2.1-1, and could be added in a pre-treatment step in areas of high As III concentration. Therefore, depending upon a water sources redox environment, pH and potential As III/V concentrations, adding a chemical oxidation step before media filtration could facilitate maximum arsenic removal, regardless of original species distribution, during water treatment [33b, 36a].

1.2.2 Interferences in Arsenic Removal During Water Treatment

Phosphates, sulfates and carbonates have all been shown to compete with the adsorption of dissolved As III/V onto filtration media in water with the level of interference dependent upon the adsorption media, redox potential and pH [22, 33b, 36a]. The presence of natural/dissolved organic matter

(NOM/DOM) also interfered with dissolved arsenic III/V adsorption onto solid supports during water treatment and has been shown, in some cases, to mobilize metals in solution [4, 37].

1.2.3 Potential Health Effects of Arsenic Exposure in Mammals

Research has indicated inorganic species of arsenic, namely As III and As V, were more toxic when ingested than methylated species, DMAA or MMAA [38]. Exposure to As III/V in drinking and source water has been linked to the occurrence of skin lesions, certain neurological effects and multiple types of cancer in humans and other mammals [5a, 26, 27b, 30]. Research performed by Hauchman indicated concentrations as small as 200 µg∙L-1 in drinking water were shown to be carcinogenic in humans [6]. In addition, As III and As V both accumulated and had toxic effects in plants [12], therefore consumable plant life could be a compounding source of arsenic exposure to humans and mammals [5a] upon ingestion. The exact mode of arsenic metabolism in mammals remains unclear. However, the toxicity associated with As

III and V may be due to an apparent lack of mechanisms, in most mammals, for the immediate excretion of these two arsenic species [32, 39]. Currently the only known routes of As III/V excretion from the body were metabolism where, upon ingestion, As III and V were carried to the liver and/or kidneys via blood cells. Once in the liver/kidneys, arsenic was further processed in a glutathione (GSH) reduction metabolism pathway where, if present, As V was reduced to As III and methylated to form either DMAA and/or

MMAA, the two arsenic forms preferentially excreted by mammals in urine [39a, 40].

The majority of arsenic speciation methods apply some form of High Performance Liquid

Chromatography (HPLC) to separate arsenic species prior to detection and quantification, often by

Inductively Coupled Plasma (ICP) Mass Spectrometry (MS) [22, 28, 35], however multiple detection methods exist. Though numerous, reliable arsenic speciation methods were available neither the WHO nor the EPA imposed maximum contaminant level (MCL) differentiated between the potential toxicity associated with different species of arsenic viable in processed water [24]. The aforementioned mandates on arsenic in distributed drinking water applied only to total arsenic concentrations [24]. However, the varying effect of arsenic in the environment, the potential difficulty in removing it from water prior to

22 distribution and its impact upon mammals and plant life depending upon the species detected highlighted the potential need for standardized speciation methods during water processing. In addition, recent studies have indicated methylated arsenic species may be more toxic, in vivo, than inorganic species of As III or

As V as was originally believed [32]. Since mammals relied upon methylation for to excrete arsenic, this information further supported both the need for water treatment plans that removed the maximum amounts of arsenic prior to water distribution and a potential need for speciation analysis of treated waters when elevated levels are arsenic were detected.

1.2.4 Methods of Arsenic Speciation Analysis

While numerous analytical methods for the separation and detection of the different arsenic species exist, the methods reviewed in this section will focus primarily on those dealing with the detection of inorganic species of arsenic in aqueous (water) matrices. As previously mentioned, inorganic forms of arsenic in water exist primarily in two different oxidation states (species), As III, zero charge, and As V, -1 charge. Typically, choosing a method to speciate arsenic in water depends upon several different variables.

Including, but not limited to, arsenic form (solid or dissolved), matrix (fresh, marine, wastewater, etc.), pH, redox potential, temperature, potentially interfering and/or competing analytes (humics, sulfur, phosphates, iron, nitrates, etc.), species of interest (inorganic, organic) and/or its concentration (ppt, ppb, ppm, etc.).

Arsenic speciation, typically requires two major steps: 1. Separation (of the arsenic species of interest) and

2. Detection (of the arsenic species in solution). Depending on their surrounding redox environment and the aforementioned species dependent variables, solvated inorganic arsenic species (As III/V), may become unstable when transported. This potential species instability could result in the need for a third step in speciation, 3. Preservation (e.g. acidification, addition of chelating and/or anti-oxidants, thermal control, filtration, etc.). However, to provide the most accurate, reproducible analysis results that are representative of arsenic species in-situ, analytical speciation methods will generally encompass all three steps.

1.2.5 Spectroscopic Methods of Detecting Arsenic Species

Due to their ease of application, low detection limits and minimal sample preparation required, spectroscopic detection methods including UV-Vis, inductively coupled plasma (mass and atomic emission) spectroscopy, atomic adsorption, graphite furnace and hydride generation, among others, are often applied to quantify arsenic species in water.

1.2.6. Molecular Absorptive Spectroscopic Arsenic Speciation

Matsubara et al. was able to simultaneously quantitate both As V and phosphate (PO4) in water using a UV-Vis spectroscopy method. The method reacted PO4 and As V with ammonium molybdate and malachite green, under acidic conditions, to form aggregates which were detectable at 627 nm [41]. Figure

1.2.6-1, adapted from Matsubara et al., shows the UV-Vis instrument response as a function of time and

PO4 concentration [41]. Matsubara et al. indicated the method was able to detect As V in solution at ppb concentrations [41].

Figure 1.2.6-1 Molybdoarsenate – Malachite Green aggregate formation over time at decreasing phosphate concentrations. -06 -06 -07 -07 Arsenate concentration: 2E M in 0.45 M sulfuric acid H2SO4. Phosphate concentrations: A. 2E M; B. 4E M; C. 2E M; D. 8E-07 M; E. 0.0 M. Figure adapted from Matsubara et al. [41].

A similar method developed by Palanivelu et al. measured the UV-Vis intensity of As III, extracted from a potassium iodate-salt mixture in the presence of benzene, reacting with Rhodamine 6G under acidic conditions at 535 nm to speciate arsenite in water at µg∙mL-1 concentrations [42].

1.2.7 Atomic Adsorptive Spectroscopic Arsenic Speciation

At pH 5, sodium borohydride arsine generation only occurs when As III is present in solution while arsine generation in 1 M HCl solutions consumes both As III and V in solution [43]. Capitalizing on this pH dependent difference in arsine formation, Howard et al. developed a sodium borohydride method that used atomic adsorption spectroscopy (AAS) for detecting both inorganic and organic (As III/V, MMAA and DMAA) species of arsenic in solution [43a]. Optimizing the method at a carrier gas flow rate of 150 ml∙min-1 Howard et al. indicated ng∙mL-1 detection limits of As III/V, MMAA and DMAA were possible

[43a]. In contrast, Pacey et al. coupled ion-exchange chromatography (IEC) with graphite furnace atomic adsorption to detect As III, V, MMAA and DMAA in solution according to the schematic shown in Figure

1.2.7-1, adapted from Pacey et al. [44].

Figure 1.2.7-1 Ion-exchange (cation/anion) chromatography schematic for the separation and detection of arsenic (total), As V, MMAA and DMAA. Figure adapted from Pacey et al. [44].

Coupling anion and cation IEC chromatography to AAS detection, Pacey et al. showed limits of detection for the method of 4.0, 0.4, 2.0 and 0.02 ppb for total arsenic, As V, MMAA and DMAA respectively, with

As III calculated by taking the difference in total arsenic and the remaining arsenic species in solution [44].

1.2.8 Electrochemical Arsenic Speciation

Similar to spectroscopic methods, electrochemical methods are potentially inexpensive, sensitive, easily applied selective speciation methods. In addition, use of applied electric fields to achieve species separation has the potential to provide superior results for multiple elements as, once applicable parameters are determined, solute separation occurs due to charge differences within the electric field. Several different electrochemical methods have been applied to arsenic speciation analysis including multiple forms of voltammetry [22], anodic/cathodic stripping [45], differential pulse [46] and capillary electrophoresis [47].

Henry et al. reported that they were able to successfully speciate and detect total arsenic, As III and

V by differential pulse polarography (DPP) using a hanging mercury drop electrode (HMDE) [48].

However, the HMDE-DPP method of speciation required splitting samples prior to analysis as only As III could be measured at the electrode surface which may have led to a degree of error in analysis. In addition, interferences from lead (Pb) and tin (Sn) in solution were found to interfere with As III detection at the electrode surface when HCl was utilized during analysis. However, Henry et al. later reported eliminating these interferences using HClO4 instead of HCl and successfully detecting As III/V, MMAA and DMAA using their HMDE-DPP method, incorporating ion-exchange chromatography (IEC) for MMAA/DMAA separation prior to reduction to As III and detection at the electrode surface [46].

Using capillary electrophoresis (CE) Casiot et al. reported simultaneously speciation of As III/V,

MMAA, DMAA, selenium (Se), antimony (Sb) and tellurium (Te), at -20 kV, with indirect UV-Vis detection at 254 nm [49]. Results published by Casiot et al. showed CE in the electro-migrative injection mode provided the greatest sensitivity for each analyte injected at concentrations of 1 ppb As III/V, MMAA,

DMAA, Se and Sb and 5 ppb Te, with reported detection limits of 124, 46, 44 and 130 ppb for As III, V,

MMAA and DMAA respectively [49].

Initial attempts to couple capillary electrophoresis (CE) to inductively coupled plasma mass spectrometry (ICP-MS) were unsuccessful with results showing little to no separation of inorganic species

26 of arsenic. Olesik et al. used a silver coated capillary interfaced directly into a concentric nebulizer when they coupled CE with an ICP-MS in an attempt to achieve simultaneous detection of As III and V at ppb concentrations [47]. However the electropherogram seen in Figure 1.2.8-1, adapted from Olesik et al., displays the minimal separation was achieved during analysis [47].

Figure 1.2.8-1. Electropherogram of capillary electrophoresis inductively couple plasma mass spectrometry (CE-ICP-MS) separation and detection of chromium (Cr) III/VI, As III/V, strontium (Sr) and tin (Sn II/IV). Elemental species concentration was 1 ppb. Electropherograms were obtained using an applied voltage of 10 kV, with a 0.06 M calcium chloride electrolyte solution at pH 6.7. Figure adapted from Olesik et al. [47].

Olesik et al. indicated the inability to achieve separation was a result of inefficient sample introduction due to the pneumatic nebulizer sample interface between the CE and the ICP-MS. Van Holderbeke et al. cited similar concerns and instead coupled CE to an ICP-MS, outfitted with a micro-concentric nebulizer, using a stainless steel capillary that routed the speciation sample directly into the ICP-MS nebulizer. Employing an osmotic flow modifier (OFM) which effectively created a positively charged capillary surface, Van

Holderbeke et al. were able to simultaneously separate both arsenic anions (As III, As V, MMAA and

DMAA) and cations arsenobetaine, arsenochloine (AsB and AsC) with detection limits between 1.3 and

2.1 ppb [50].

1.2.9 High Performance Liquid Chromatography (HPLC) – Inductively Coupled Plasma (ICP)

Within the category of chromatography there are multiple types of separation, including, chiral, reversed phase, ion exchange, supercritical fluid, size exclusion, etc. Chromatography separation of different analytes, and multi-species elements in solution is widely utilized due to the availability of multiple commercial separation columns and a wide array of standardized separation techniques. In addition, advances within HPLC instrumentation and the methods applied have resulted in decreased separation times, increased detection limits and improved reproducibility [51]. Coupling these highly specific modes of species separation to the highly selective, sensitive detection of inductively coupled plasma mass and atomic emission spectroscopy (ICP-MS and ICP-AES), has led to the simultaneous separation and detection of multiple different elemental species including arsenic [51].

Evaluating anion-exchange HPLC methods of arsenic speciation coupled to ICP-MS Day et al. was able to demonstrate successful species separation of As III/V, MMAA and DMAA with 67, 89, 35 and 74 ng∙L-1 limits of detection as shown in Figure 1.2.9-1, adapted from Day et al. [52].

Figure 1.2.9-1 Chromatogram of AEC-ICP-MS separation and detection of standard solutions of As III/V, MMAA and DMAA. Solutions prepared at concentrations of 1 µg∙L-1 for each individual arsenic species. Figure adapted from Day et al. [52]. Hansen et al. used anion and cation exchange chromatography to separate anionic species of arsenic

(As III/V, MMAA and DMAA) from cation species (AsB and AsC) prior to flame atomic adsorption (FAA) and ICP-MS [53]. Following method optimization, Hansen et al. reported detection limits of detection for

ICP-MS detection that were between 150 to 470 times greater than those reported using FAA, as shown in

Table 1.2.9-1, adapted from Hansen et al. [53].

Table 1.2.9-1. Method detection limits in µg∙L-1, calculated as three times the baseline noise at 100 µL injections of a mixture containing As III, V, MMAA, DMAA, AsB and AsC. Table adapted from Hansen et al. [53].

Arsenic Compound HPLC-FAAS HPLC-ICP-MS As III 1.1 0.007 As V 1.4 0.003 MMAA 1.4 0.003 DMAA 0.7 0.003 AsB 0.3 0.002 AsC 0.5 0.002

1.3 Research Performed

This work details initial attempts to determine potential UV-Vis and/or fluorescence spectroscopy methods to quantify and characterize DOM in solution by varying the pH of the matrix. These methods may be applied to future research with DOM obtained from natural waters. Details of a proposed HPLC-

ICP-MS speciation method for the separation and detection of inorganic As III and V, in drinking water matrices, is also detailed in this work. The proposed As III/V speciation method was used to analyze both samples prepared “in house” and samples obtained from natural water sources containing As III and V.

Optimization of the arsenic speciation method included a four week analysis of both the stability of As

III/V and any potential impact from DOM on arsenic stability or its detection in water. The DOM used during analysis was donated by Dr. Jonathan Pressman at the US EPA and was concentrated using a novel method detailed in Pressman et al. [4].

References

1. Hessen, D. O.; Tranvik, L. J., Aquatic Humic Substances: Ecology and Biogeochemistry. Springer: 1998. 2. Buschmann, J.; Kappeler, A.; Lindauer, U.; Kistler, D.; Berg, M.; Sigg, L., Arsenite and Arsenate Binding to Dissolved Humic Acids: Influence of pH, Type of Humic Acid, and Aluminum. Environmental Science & Technology 2006, 40 (19), 6015-6020. 3. Maurer, F.; Christl, I.; Kretzschmar, R., Reduction and Reoxidation of Humic Acid: Influence on Spectroscopic Properties and Proton Binding. Environmental Science & Technology 2010, 44 (15), 5787- 5792. 4. Pressman, J. G.; Richardson, S. D.; Speth, T. F.; Miltner, R. J.; Narotsky, M. G.; Hunter, E. S., 3rd; Rice, G. E.; Teuschler, L. K.; McDonald, A.; Parvez, S.; Krasner, S. W.; Weinberg, H. S.; McKague, A. B.; Parrett, C. J.; Bodin, N.; Chinn, R.; Lee, C. F.; Simmons, J. E., Concentration, chlorination, and chemical analysis of drinking water for disinfection byproduct mixtures health effects research: U.S. EPA's Four Lab Study. Environ Sci Technol 2010, 44 (19), 7184-92. 5. (a) Nriagu, J. O., Arsenic in the Environment: Human health and ecosystem effects. Wiley: 1994; (b) Agency, U. S. E. P. Drinking Water Contaminants. http://water.epa.gov/drink/contaminants/index.cfm. 6. Hauchman, F. S., Drinking water health effects research at the US Environmental Protection Agency. The International Association of Hydrological Sciences 1995, (233), 55-62. 7. Kashem, M. A.; Singh, B. R., Metal availability in contaminated soils: I. Effects of floodingand organic matter on changes in Eh, pH and solubility of Cd, Ni andZn. Nutrient Cycling in Agroecosystems 2001, 61 (3), 247-255. 8. Warwick, P.; Inam, E.; Evans, N., Arsenic’s Interaction with Humic Acid. Environmental Chemistry 2005, 2 (2), 119-124. 9. Krasner, S. W.; Weinberg, H. S.; Richardson, S. D.; Pastor, S. J.; Chinn, R.; Sclimenti, M. J.; Onstad, G. D.; Thruston, A. D., Occurrence of a New Generation of Disinfection Byproducts. Environmental Science & Technology 2006, 40 (23), 7175-7185. 10. Chiou, C. T.; Kile, D. E.; Brinton, T. I.; Malcolm, R. L.; Leenheer, J. A.; MacCarthy, P., A comparison of water solubility enhancements of organic solutes by aquatic humic materials and commercial humic acids. Environmental Science & Technology 1987, 21 (12), 1231-1234. 11. Xie, Y., Disinfection Byproducts in Drinking Water: Formation, Analysis, and Control. Taylor & Francis: 2004. 12. Chen, J., Spectroscopic characterization of the structural and functional properties of natural organic matter fractions. Chemosphere 2002, 48 (1), 59. 13. McDonald, S.; Bishop, A. G.; Prenzler, P. D.; Robards, K., Analytical chemistry of freshwater humic substances. Analytica Chimica Acta 2004, 527 (2), 105-124. 14. Korshin, G. V.; Benjamin, M. M.; Li, C.-W., Use of differential spectroscopy to evaluate the structure and reactivity of humics. Water Science and Technology 1999, 40 (9), 9-16. 15. Weishaar, J. L.; Aiken, G. R.; Bergamaschi, B. A.; Fram, M. S.; Fujii, R.; Mopper, K., Evaluation of Specific Ultraviolet Absorbance as an Indicator of the Chemical Composition and Reactivity of Dissolved Organic Carbon. Environmental Science & Technology 2003, 37 (20), 4702-4708. 16. (a) Senesi, N.; Miano, T. M.; Provenzano, M. R.; Brunetti, G., Spectroscopic and compositional comparative characterization of I.H.S.S. reference and standard fulvic and humic acids of various origin. Science of the Total Environment 1989, 81-82, 143-156; (b) Mobed, J. J.; Hemmingsen, S. L.; Autry, J. L.; McGown, L. B., Fluorescence Characterization of IHSS Humic Substances: Total Luminescence Spectra with Absorbance Correction. Environmental Science & Technology 1996, 30 (10), 3061-3065; (c) Westerhoff, P.; Chen, W.; Esparza, M., Fluorescence Analysis of a Standard Fulvic Acid and Tertiary Treated Wastewater. J Environ Qual 2001, 30 (6), 2037-2046. 17. Senesi, N., Molecular and quantitative aspects of the chemistry of fulvic acid and its interactions with metal ions and organic chemicals. Analytica Chimica Acta 1990, 232, 77-106.

18. McDonald, S.; Pringle, J. M.; Bishop, A. G.; Prenzler, P. D.; Robards, K., Isolation and seasonal effects on characteristics of fulvic acid isolated from an Australian floodplain river and billabong. Journal of Chromatography A 2007, 1153 (1-2), 203-213. 19. (a) Abdulla, H. A. N.; Minor, E. C.; Hatcher, P. G., Using Two-Dimensional Correlations of 13C NMR and FTIR To Investigate Changes in the Chemical Composition of Dissolved Organic Matter along an Estuarine Transect. Environmental Science & Technology 2010, 44 (21), 8044-8049; (b) Conte, P.; Piccolo, A., Conformational Arrangement of Dissolved Humic Substances. Influence of Solution Composition on Association of Humic Molecules. Environmental Science & Technology 1999, 33 (10), 1682-1690; (c) De Paolis, F.; Kukkonen, J., Binding of organic pollutants to humic and fulvic acids: Influence of pH and the structure of humic material. Chemosphere 1997, 34 (8), 1693-1704. 20. Phillips, S. L.; Olesik, S. V., Initial Characterization of Humic Acids Using Liquid Chromatography at the Critical Condition Followed by Size-Exclusion Chromatography and Electrospray Ionization Mass Spectrometry. Analytical Chemistry 2003, 75 (20), 5544-5553. 21. Hundt, T. R.; O'Melia, C. R., Aluminum-fulvic acid interactions: Mechanisms and applications. Journal of the American Water Works Association 1988, 88 (4), 176-186. 22. Nriagu, J. O., Arsenic in the Environment, Cycling and Characterization. Wiley: 1994. 23. Organization, W. H., Guidelines for Drinking-Water Quality. Stylus Pub Llc: 2011. 24. Agency, U. S. E. P. Safe Drinking Water Act (SWDA) Laws and Regulations. http://water.epa.gov/lawsregs/rulesregs/sdwa/arsenic/index.cfm. 25. Manning, B. A.; Goldberg, S., Adsorption and Stability of Arsenic(III) at the Clay Mineral−Water Interface. Environmental Science & Technology 1997, 31 (7), 2005-2011. 26. Kimberly Gehle MD, M. S. C., PhD Case Studies in Environmental Medicine - Arsenic Toxicity. http://www.atsdr.cdc.gov/. 27. (a) Cheng, H.; Hu, Y.; Luo, J.; Xu, B.; Zhao, J., Geochemical processes controlling fate and transport of arsenic in acid mine drainage (AMD) and natural systems. Journal of Hazardous Materials 2009, 165 (1-3), 13-26; (b) Sharma, V. K.; Sohn, M., Aquatic arsenic: toxicity, speciation, transformations, and remediation. Environment international 2009, 35 (4), 743-59. 28. Amirbahman, A.; Kent, D. B.; Curtis, G. P.; Davis, J. A., Kinetics of sorption and abiotic oxidation of arsenic(III) by aquifer materials. Geochimica et Cosmochimica Acta 2006, 70 (3), 533-547. 29. Technology, N. I. o. A. I. S. a., Atlas of Eh-pH diagrams - Geological Survey of Japan Open File Report No.419. 2005. 30. Lehr, J.; Keeley, J.; Lehr, J., Water Encyclopedia, Volumes 1-5. John Wiley & Sons: 2005. 31. (a) Agency, E. P., National Secondary Drinking Water Regulations. Office, C. o. F. R., Ed. Office of the Federal Register National Archives and Records Administration: 2002; Vol. 19; (b) Smedley, P. L.; Kinniburgh, D. G., A review of the source, behaviour and distribution of arsenic in natural waters. Applied Geochemistry 2002, 17 (5), 517-568. 32. Vahter, M., Mechanisms of arsenic biotransformation. Toxicology 2002, 181-182, 211-217. 33. (a) Lytle, D. A.; Chen, A. S.; Sorg, T. J.; Phillips, S.; French, K., Microbial As(III) oxidation in water treatment plant filters. American Water Works Association. Journal 2007, 99 (12), 72-86,12; (b) Lytle, D. A., Sorg, T. J. and Snoeyink, V. L. , Optimizing arsenic removal during iron removal: Theoretical and practical considerations. Journal of Water Supply: Research and Technology - Aqua 2005, 54 (8), 545- 560. 34. Agency, U. S. E. P. http://www.epa.gov/enviro/html/icr/index.html. 35. Dixit, S.; Hering, J. G., Comparison of arsenic(V) and arsenic(III) sorption onto iron oxide minerals: implications for arsenic mobility. Environ Sci Technol 2003, 37 (18), 4182-9. 36. (a) Lytle, D. A.; Sorg, T. J.; Muhlen, C.; Wang, L., Particulate arsenic release in a drinking water distribution system. American Water Works Association. Journal 2010, 102 (3), 87-98,16; (b) Emett, M. T.; Khoe, G. H., Photochemical oxidation of arsenic by oxygen and iron in acidic solutions. Water Research 2001, 35 (3), 649-656. 37. (a) McCurry, D.; Speth, T.; Pressman, J., Lyophilization and Reconstitution of Reverse-Osmosis Concentrated Natural Organic Matter from a Drinking Water Source. Journal of Environmental

Engineering 2012, 138 (4), 402-410; (b) Weng, L.; Van Riemsdijk, W. H.; Hiemstra, T., Effects of Fulvic and Humic Acids on Arsenate Adsorption to Goethite: Experiments and Modeling. Environmental Science & Technology 2009, 43 (19), 7198-7204. 38. Chappell, W. R.; Abernathy, C. O.; Calderon, R. L., Arsenic Exposure and Health Effects III. Elsevier Science: 1999. 39. (a) Delnomdedieu, M.; Basti, M. M.; Otvos, J. D.; Thomas, D. J., Transfer of arsenite from glutathione to dithiols: A model of interaction. Chemical Research in Toxicology 1993, 6 (5), 598-602; (b) Delnomdedieu, M.; Basti, M. M.; Otvos, J. D.; Thomas, D. J., Reduction and binding of arsenate and dimethylarsinate by glutathione: a magnetic resonance study. Chemico-Biological Interactions 1994, 90 (2), 139-155. 40. Vega, L.; Styblo, M.; Patterson, R.; Cullen, W.; Wang, C.; Germolec, D., Differential Effects of Trivalent and Pentavalent Arsenicals on Cell Proliferation and Cytokine Secretion in Normal Human Epidermal Keratinocytes. Toxicology and Applied Pharmacology 2001, 172 (3), 225-232. 41. Matsubara, C.; Yamamoto, Y.; Takamura, K., Rapid determination of trace amounts of phosphate and arsenate in water by spectrophotometric detection of their heteropoly acid-malachite green aggregates following pre-concentration by membrane filtration. The Analyst 1987, 112 (9), 1257-1260. 42. Palanivelu, K.; Balasubramanian, N.; Ramakrishna, T. V., A chemical enhancement method for the spectrophotometric determination of trace amounts of arsenic. Talanta 1992, 39 (5), 555-561. 43. (a) Howard, A. G.; Arbab-Zavar, M. H., Determination of inorganic arsenic(III) and arsenic(V), methylarsenic and dimethylarsenic species by selective hydride evolution atomic-absorption spectroscopy. The Analyst 1981, 106 (1259), 213-220; (b) Analytical methods support document for arsenic in drinking water. DIANE Publishing. 44. Pacey, G. E.; Ford, J. A., Arsenic speciation by ion-exchange separation and graphite-furnace atomic-absorption spectrophotometry. Talanta 1981, 28 (12), 935-938. 45. Holak, W., Determination of arsenic by cathodic stripping voltammetry with a hanging mercury drop electrode. Analytical Chemistry 1980, 52 (13), 2189-2192. 46. Henry, F. T.; Thorpe, T. M., Determination of arsenic(III), arsenic(V), monomethylarsonate, and dimethylarsinate by differential pulse polarography after separation by ion exchange chromatography. Analytical Chemistry 1980, 52 (1), 80-83. 47. Olesik, J. W.; Kinzer, J. A.; Olesik, S. V., Capillary Electrophoresis Inductively Coupled Plasma Spectrometry for Rapid Elemental Speciation. Analytical Chemistry 1995, 67 (1), 1-12. 48. Henry, F. T.; Kirch, T. O.; Thorpe, T. M., Determination of trace level arsenic(III), arsenic(V), and total inorganic arsenic by differential pulse polarography. Analytical Chemistry 1979, 51 (2), 215-218. 49. Casiot, C.; Barciela Alonso, M. C.; Boisson, J.; Donard, O. F. X.; Potin-Gautier, M., Simultaneous speciation of arsenic, selenium, antimony and tellurium species in waters and soil extracts by capillary electrophoresis and UV detection. The Analyst 1998, 123 (12), 2887-2893. 50. Van Holderbeke, M.; Zhao, Y.; Vanhaecke, F.; Moens, L.; Sandra, P., Speciation of six arsenic compounds using capillary electrophoresis-inductively coupled plasma mass spectrometry. Journal of Analytical Atomic Spectrometry 1999, 14 (2), 229-234. 51. Ali, I.; Aboul-Enein, H. Y., Instrumental Methods in Metal Ion Speciation. Taylor & Francis: 2006. 52. Day, J. A.; Montes-Bayón, M.; iacute; Vonderheide, A. P.; Caruso, J. A., A study of method robustness for arsenic speciation in drinking water samples by anion exchange HPLC-ICP-MS. Analytical and Bioanalytical Chemistry 2002, 373 (7), 664-668. 53. Hansen, S. H.; eacute; Larsen, E. H.; Pritzl, G.; Cornett, C., Separation of seven arsenic compounds by high-performance liquid chromatography with on-line detection by hydrogen–argon flame atomic absorption spectrometry and inductively coupled plasma mass spectrometry. Journal of Analytical Atomic Spectrometry 1992, 7 (4), 629-634.

Chapter 2

Developing a Method for Dissolved Organic Matter Characterization and Arsenic III/V

Detection in Water Matrices

2.0 Experimental Introduction

Details of the analytical work performed over the course of completing this thesis are described in the current chapter. Analyses performed included UV-Vis and fluorescence spectroscopy characterization of samples of dissolved humic acid (HA), obtained from Sigma Aldrich, and solutions of concentrated dissolved organic matter (cDOM), concentrated via a method published by Pressman et al. [4] at the US

EPA. Prior to spectroscopic analysis, samples of dissolved HA and cDOM were subjected to total organic carbon (TOC) and trace metal analysis. Also included in this chapter are details of the high performance liquid chromatography inductively coupled plasma mass spectrometry (HPLC-ICP-MS) speciation method proposed for the separation, detection and quantification of inorganic arsenic III and V (As III/V) in drinking water. The separation of multi-species elements and complex mixtures using HPLC chromatographic methods is a technique often applied to environmental samples [54]. The chosen HPLC separation method is often combined with detection and quantification by ICP-MS due to the highly specific and sensitive detection capability of the ICP-MS, which detects molecules based on their mass-to-charge

(m/z) ratio [55]. Coupling HPLC separation to an ICP-MS allows for the element-specific detection of analytes in solution at the extremely low concentrations typically found in drinking water matrices.

2.1 Instrumentation

2.1.1 UV-Vis Spectroscopy

Ultraviolet-visible (UV-Vis) Spectroscopy was utilized in two separate venues during the analyses described here. The first was analysis of DOM samples using a Perkin Elmer Spec Lambda 20 spectrophotometer (Waltham, MA). Samples were analyzed in 1 cm quartz cuvettes scanning from 200 to

800 nm at a scan rate of 60 nm per minute at a slit width of 4 nm. The second was to detect dissolved organic matter (DOM) removal from the HPLC column using an Agilent Technologies (Santa Clara, CA)

1100 Series Multiple Wavelength Detector (MWD), model G1365B, placed in-line, between the HPLC column compartment and the ICP-MS. The HPLC UV-Vis MWD was controlled by a separate computer, in a “stand-alone” capacity, utilizing Agilent Technologies (Santa Clara, CA) ChemStation software, model

G2170AA, version A.10.01.

2.1.2 Fluorescence Spectroscopy

Fluorescence analysis was performed using a Varian Cary Eclipse Fluorometer (Palo Alto, CA) with a xenon pulse light source. The fluorometer was fitted with a 370 blaze Czerny Turner excitation monochromator with 30x35 mm grating (1200 lines/mm) and an X2 photo-multiplier and 92 - 8 % beam splitter. Sample spectra were obtained using 1 cm quartz cuvettes.

2.1.3 Total Organic Carbon (TOC)

The dissolved organic carbon (DOC) content of the DOM samples utilized during analysis was obtained using a Shimadzu TOC (TOC-Veph) analyzer (Columbia, MD) according to US EPA method

415.2 (See Appendix A). All TOC analyses were performed by analyst Stephanie Brown. The instrument utilized a catalytic combustion reaction to isolate, detect and quantify the dissolved organic carbon within sample solutions. Total organic carbon analysis was performed solely to verify the final concentration of injected DOM samples (results not published).

2.1.4 Inductively Coupled Plasma - Atomic Emission Spectroscopy (ICP-AES)

The concentrations of various trace metals within the DOM samples utilized during analysis were analyzed using an iCAP 6000 Series ICP-AES, Thermo Scientific (Waltham, MA) according to US EPA method 200.7 (See Appendix A). Solutions of DOM were diluted 1:100 and 1:1000 and acidified with double distilled nitric acid (HNO3) for at least 24 hours prior to analysis. During analysis, samples are drawn, via a peristaltic pump, into the instrument nebulizer where samples are mixed with argon gas and

34 aerosolized into the spray chamber. From there, the aerosolized sample is carried into the ICP torch chamber where the plasma evaporates, atomizes and ionizes the sample. As the ionized sample travels beyond the plasma, electrons, in an excited state due to atomization relax, emitting photons as they transition to the ground state specific to the elements in the periodic table. The difference between the transition states is equal to the wavelength of the element as seen in Figure 2.1.4-1.

Figure 2.1.4-1. Energy transition energy level diagram. The various transitions shown: a & b - excitation; c - ionization; d - excitation/ionization; e - ion emission; and f, g & h - atom emission. Figure adapted from Boss et al.

The emitted photons are UV-Visible radiation which are directed from the torch to the monochromator which diffracts and separates the energy into its characteristic wavelength prior to the detector. The instrument detector assigns the individual analyte concentration detected in each sample based on the monochromator separated wavelengths and the intensity of the photon energy detected corresponding to the individual elements within a sample. Similar to the TOC analysis described above,

ICP-AES trace metal analysis of DOM samples was performed solely to determine the presence and concentration of metals which could potentially interfere and/or enhance detection of DOM and/or As III/V speciation.

2.1.5 HPLC-ICP-MS

Separation of As III and V was achieved utilizing an Agilent Technologies (Santa Clara, CA) 1260

Infinity Series High Performance Liquid Chromatography (HPLC) system coupled to a Agilent

Technologies (Santa Clara, CA) 7700x Inductively Coupled Mass Spectrometer (ICP-MS) for subsequent mass detection and quantification of As (III/V).

The HPLC set-up consisted of a quaternary pump with integrated solvent degasser, model G1311B, auto-sampler (ALS), model G1329B and thermostatted column compartment (TCC), model G1316A. The

ICP-MS, model G3281A, was configured utilizing Agilent Technologies (Santa Clara, CA) standard environmental set-up and control of all components of the HPLC-ICP-MS, including data analysis, was achieved simultaneously utilizing Agilent Technologies (Santa Clara, CA) MassHunter software, model

G7201A, version A.01.02.

Reversed-phase chromatography (RPC) was performed using a Zorbax Eclipse XDB C-18, 4.6 x

150 mm 5 µm, analytical column, equipped with the corresponding Eclipse XDB-C18 guard column, 4.6 x

12.5 mm 5 µm Agilent Technologies (Santa Clara, CA). The HPLC system was optimized routing the sample volume from column using perfluoroalkoxy (PFA) tubing, 0.5 mm ID, 1.6 mm OD Agilent

Technologies (Santa Clara, CA), via a low dead volume tee Agilent Technologies (Santa Clara, CA), through tygon pump tubing, 1.02 mm ID, 0.86 mm wall IDEX Health and Science (Oak Harbor, WA) via the ICP-MS peristaltic pump into the nebulizer.

2.2 Reagents, Solutions and Miscellaneous Labware

Solutions were prepared in either Milli-Q distilled, de-ionized (DI) water obtained from a Millipore filtration system equipped with a Quantum ultrapure Ionex water filtration cartridge purchased from

Millipore (Billerica, MA) or in DI water obtained from a Barnstead B-Pure H2O filtration system purchased from Thermo Scientific (Waltham, MA). Arsenic speciation standard solutions at 1000 parts per million

(ppm) arsenic III (in 2 % HCl) and arsenic V (in H2O), were purchased from SPEX CertiPrep (Metuchen,

NJ). Whatman syringe filters, 25 mm GD/X, 0.45 µm pore size and J. T. Baker sodium bi-carbonate

(NaHCO3) were purchased from Fisher Scientific (Pittsburgh, PA). Concentrated, double distilled, hydrochloric acid (HCl) and solutions of ammonium hydroxide (28-30 % NH3) and 99.999 % phosphoric acid (85 % solution) were purchased from GFS Chemicals (Columbus, OH) and a fifty percent solution of sodium hydroxide (NaOH) was obtained from Spectrum Chemicals (Los Angeles, CA). Ammonium phosphate monobasic (99.99 % trace metal basis), Fluka Analytical tetrabutyl ammonium hydroxide (40 % solution in H2O for ion chromatography) and a commercial sample of humic acid were purchased from

Sigma Aldrich (St. Louis, MO). In addition, a concentrated solution of DOM (cDOM) [4] was donated by

Dr. Jonathan Pressman, EPA Cincinnati.

2.2.1 NaHCO3 Solutions (10 and 50 mM) - Dissolved Humic Acid/cDOM Analysis

Separate10 and 50 mM NaHCO3 solutions were prepared at pH’s of 2, 4, 6, 8, 10 and 11 by dissolving NaHCO3 in the appropriate amount of DI water and adjusting to the final pH with either HCl and/or NaOH.

2.2.2 Mobile Phase (2.5 mM (NH4)H2PO4, 5 mM TBAH, pH 6.0) - Arsenic III/V Speciation

The HPLC mobile phase was prepared at a concentration of 2.5 mM ammonium phosphate monobasic ((NH4)H2PO4) and 5 mM tetrabutyl-ammonium hydroxide (TBAH). The mobile phase was adjusted to pH 6.0 with concentrated phosphoric acid, H3PO4 and filtered utilizing a MilliSolve™ filtration system, P/N XX1604700, with 0.22 µm Millipore GS membrane filters, P/N GSWG047S6, EMD Millipore

(Billerica MA), prior to introduction at the HPLC. The HPLC mobile phase was stored in acid washed, amber bottles at room temperature between sample analyses.

2.3 Sample Preparation

2.3.1 Humic acid (HA)/cDOM analysis samples

Humic acid samples analyzed by UV-Vis and fluorescence spectroscopy were prepared by either dissolving and/or diluting the sample in DI water and/or 10 or 50 mM sodium bicarbonate solutions, described above, to the desired organic carbon content, verified by TOC analysis as previously described.

If necessary, the pH of final solution was adjusted using HCl and/or NaOH to the final desired pH.

Sample blanks were prepared for all of the analyses performed, as described above, utilizing humic acid and/or cDOM in either DI water and/or NaHCO3, for each pH monitored and analyzed by either UV-

Vis and/or fluorescence spectroscopy.

2.3.2 Arsenic III/V speciation analysis samples

Arsenic III/V instrument calibration solutions and any necessary sample dilutions were prepared daily, diluting samples to their final concentration with the 2.5 mM (NH4)H2PO4, 5 mM TBAH mobile phase, pH 6, utilized during HPLC separation. Unless noted, water samples submitted for As III/V analysis were processed “as is” with dilutions, in mobile phase, occurring as needed, to bring samples within the limits of the instrument calibration. All arsenic III/V samples were refrigerated, at 3 °C, until analysis was performed. In addition to refrigeration, the As III/V samples used to monitor the stability of both arsenic species in solution were also stored in black vial storage boxes, Cole Parmer (Vernon Hills, IL) in between periods of analysis.

Arsenic III/V instrument calibration standards were prepared, diluting 1000 ppm As III and As V speciation standards, Spex Certi-Prep®, in 2.5 mM (NH4)H2PO4, 5 mM TBAH, pH 6, HPLC mobile phase.

The instrument calibration standards were prepared at final concentrations of 1, 10, 25, 50, 75, 100 and 250 parts per billion (ppb).

The As III/V “stability” samples were prepared at concentrations of 0, 1, 50 and 150 ppb in 2.5 mM (NH4)H2PO4, 5 mM TBAH, pH 6.0 HPLC mobile phase for each of the four As III/V anti-oxidant schemes analyzed. 1. No preservative - samples contained only As III/V; 2. Ascorbic acid (A. A.) - samples contained 1.7E-04 M A. A. at 1 ppb As III/V, 2.84E-03 M A. A. at 50 ppb As III/V and 8.52E-03 M A. A. at

150 ppb As III/V (A. A. to As III/V concentrations were determined from HPLC-ICP-MS analysis of

Cincinnati tap water spiked with A. A. and As III/V - results not shown); 3. Hydrochloric (HCl) acid - samples were prepared at a final concentration of 0.05 % HCl; and 4. Ethylenediamine tetraacetic (EDTA) acid - samples were prepared at a final concentration of 1.25E-03 M EDTA [56].

The As III/V samples spiked with concentrated dissolved organic matter (cDOM) were prepared as described above for the As III/V “stability” samples (at concentrations of 0, 1, 50 and 150 ppb As III/V) in HPLC mobile phase at each anti-oxidant scheme. However, each DOM/As III/V sample was spiked to a final concentration of 1 ppm DOM, verified by TOC analysis, as previously described. The concentrated

DOM stock sample contained a total organic carbon (TOC) content of 23.63 ppm.

The As III/V microbial water samples were filter sterilized utilizing 0.2 µm Captiva filter kits,

Agilent Technologies (Santa Clara, CA). Following bacteria removal, samples were diluted to an approximate concentration of 75 ppb As III/V in 2.5 mM (NH4)H2PO4, 5 mM TBAH, pH 6, HPLC mobile phase.

All drinking water samples submitted for As III/V speciation from “outside” sources, e.g. EPA researcher submitted, were analyzed without dilution unless HPLC-ICP-MS results were outside the instrument calibration (1-150 ppb As III/V). All samples analyzed filtered using 0.20 µm syringe filters and verified to be within drinking water pH range (6.5-8.5) [31] prior to analysis. A duplicate of every third sample and a 10 ppb As III/V spike of every fourth sample were prepared daily. Duplicates of each

10 ppb As III/V sample spike and a single duplicate of the 1 ppb As III/V calibration standard were also

0.20µm filtered and analyzed to detect any potential losses to filtration media. In addition, a 10 ppb As

III/V “check standard” was prepared daily and analyzed every 10 samples to detect any potential instrument

39 drift and/or need for instrument recalibration during analysis. Any dilutions and/or re-runs were performed the same day as the initial sample analysis to eliminate potential error/discrepancy in As III/V concentrations.

If the approximate concentration of arsenic in samples submitted for analysis was unknown, samples were diluted 1:100, in 2.5 mM (NH4)H2PO4, 5 mM TBAH, pH 6, HPLC mobile phase, prior to analysis. Following analysis, if the calculated arsenic concentration of diluted samples was outside, or at the extreme low end of the calibration curve, samples were rerun the same day undiluted and/or at the appropriate dilution to verify results.

2.4 Analysis Methods

2.4.1 Humic acid/cDOM UV-Vis and Fluorescence Spectroscopy

As previously mentioned, both UV-Vis and fluorescence analysis have been applied to the characterization of DOM. Researchers have used these methods, in part, to verify the reactive functional groups believed to be present within HA (ketone, carboxylic acid, phenolic, etc.) based on characteristic electronic transitions and/or the wavelength or intensity of fluorophore detection occurring during the analysis of samples of DOM.

Humic acid/cDOM samples prepared, as previously described, for UV-Vis and/or fluorescence spectroscopy were analyzed without additional sample preparation in 1 cm quartz cuvettes. Samples were placed into the respective cuvette holder in either the UV-Vis or fluorometer, the lid was closed to prevent introduction of incident light and analyses were performed.

Instrument data generated during analysis was manipulated using Microsoft Excel software. In excel, the UV-Vis and/or fluorescence absorbance detected in the samples blanks, prepared as previously described, were manually subtracted from the absorbance detected in samples.

퐴 = 퐴푠 − 퐴푏

Where: As = sample absorbance

Ab = blank absorbance

This “corrected” sample absorbance was then plotted versus the wavelength of detection in nanometers for each sample analyzed.

2.4.2 Arsenic III/V HPLC-ICP-MS Speciation

As previously mentioned, inorganic species of arsenic in water exist primarily as oxyanions of arsenite (As III) and/or arsenate (As V). Based on their respective pKa’s, seen in Table 2.3.2-1, species of

As III have zero charge while species of As V have a -1 charge in solution.

Table 2.3.2-1 Inorganic arsenic species typically detected in water and/or environmental samples. Table adapted from Almassalkhi [57].

Compound Structure pKa Reference

Arsenous acid (As III) O As OH 9.3 [57]

Arsenic acid (As V) O 2.2 HO As OH 7.0 [57] OH 11.5

Monomethylarsonic acid (MMAA) O 4.00

H3C As OH 8.7 [57] OH

Dimethylarsinic acid (DMAA) O 1.8

H3C As OH 6.1 [57]

CH3

Capitalizing on the charge difference between As III and V in solution allowed for isocratic, reversed phase separation of the two species applying a variation of the arsenic speciation method utilized by Almassalkhi [57]. The proposed arsenic speciation method used an ammonium phosphate, (NH4)H2PO4, tetrabutyl ammonium hydroxide (TBAH), ion pairing agent, HPLC mobile phase to achieve simultaneous separation of As III/V in water [57]. Optimization of the method showed a flow rate of 1.75 mL·min-1 resulted in separation of inorganic As III/V species within six minutes. The HPLC-ICP-MS arsenic speciation and detection was achieved according to the method parameters detailed in Table 2.3.2-2.

Table 2.3.2-2 HPLC-ICP-MS As (III/V) speciation/detection conditions. ICP-MS Forward Power 1549 W Plasma gas flow 15.0 L·min-1 Carrier gas flow 1.0 L·min-1 Collision gas (He) flow 4.6 mL·min-1 Quadrupole bias -15.0 V Octopole bias -18.0 V Energy discrimination 3.0 V Sampler cone nickel, 1 mm Skimmer cone nickel, 0.4 mm Nebulizer pump 0.3 rps Monitored isotope As75

HPLC

Mobile phase 2.5 mM (NH4)H2PO4, 5 mM TBAH (pH 6.0) Flow rate 1.75 mL·min-1 Injection volume 100 µL Integration time/Mass 1.5 min

Prior to As III/V speciation analysis, the HPLC column was equilibrated with (NH4)H2PO4, TBAH

(pH 6.0) mobile phase for approximately 30 minutes at a flow rate of 0.5 mL·min-1 and 15 additional minutes at 1.75 mL·min-1.

Following calibration standard and/or sample preparation, as described above, samples were placed into 2 mL amber, glass HPLC vials and placed into the ALS rack according to the pre-determined sample run log stored into the instruments MassHunter® software. Samples were injected at the ALS, routed through the HPLC column for As III/V separation and detected at the ICP-MS.

The MassHunter® software generated instrument integrated data correlating to ppb As III/V concentrations and the ICP-MS signal detected in counts per second (cps). The ppb As III/V concentrations and/or ICP-MS cps were manipulated using Microsoft Excel software. The consistency in both HPLC column separation and ICP-MS detection was verified using sample duplicates, sample spikes and a 10 ppb

As III/V “check standard” prepared as previously described. In addition, sample spikes were filtered using

0.2 μm syringe filters and the filtered sample was analyzed to detect any potential As III/V adsorption to the filters utilized during sample preparation. Sample duplicates were analyzed every 10 samples to verify

HPLC separation of As III/V in speciation samples maintained at least 90 % consistency. The sample duplicate recoveries were calculated:

퐶푑 푅푑 = × 100 ((퐶 + 퐶푑)⁄2)

Where: Rd = percentage of As III/V recovered in the duplicate sample

C = concentration of As III/V in the original sample analyzed

Cd = concentration of As III/V detected in the duplicate sample analyzed

Sample spikes were analyzed every 11 samples to detect potential interferences in sample solutions which could affect As III/V separation and/or detection at the HPLC-ICP-MS. Sample spike recoveries were calculated:

(퐶푠 − 퐶) 푅푠 = × 100 퐶10

Where: Rs = percentage of As III/V recovered in the sample spike

C = concentration of As III/V in the sample

Cs = concentration of As III/V in the sample spike

C10 = concentration of As III/V in the 10 ppb calibration standard

Sample spikes prepared every 11 samples were filtered and analyzed to detect for any As III/V adsorption to the filters used during sample preparation. The sample spikes after filtration recoveries were calculated:

퐶푓 푅푓 = × 100 퐶푠

Where: Rf = percentage of As III/V recovered in the sample spike after filtration

Cs = concentration of As III/V in the sample spike

Cf = concentration of As III/V in the sample spike after filtration

Ten ppb “check standards” were prepared along with the instrument calibration standards prior to analysis each day. These check standards were placed in the ALS after every 10th sample and analyzed to detect for instrument drift at the ICP-MS during analysis. The recoveries from the 10 ppb check standards were calculated:

퐶푐 푅푐 = × 100 퐶10

Where: Rc = percentage of As III/V recovered in the 10 ppb check standard

Cc = concentration of As III/V in the 10 ppb check standard

C10 = concentration of As III/V in the 10 ppb calibration standard

After arsenic speciation analysis was completed, all data obtained was corrected, as described above, and the final analysis results were transmitted, in ppb (μg∙L-1) concentrations, to the primary researcher who requested analysis. Any results obtained from sample dilutions were also treated, as described above, with the appropriate dilution factor:

푉푓 퐹푑 = 푉푠

applied prior to reporting the final analysis to the primary researcher.

Where: Fd = dilution factor calculated

Vf = final sample volume

Vs = volume of sample added

At the completion of each HPLC-ICP-MS speciation analysis, a 90:10 methanol:water

-1 (MeOH:H2O) solution was flushed through the column for ~ 10 minutes at a flow rate of 1.75 mL·min and an additional 10 minutes at 0.5 mL·min-1 to inhibit any potential bacterial growth. The column was stored in 90:10 MeOH:H2O solution until the next analysis was performed.

2.5 Safety

2.5.1 Personal Protective Equipment (PPE)

Work performed within the analytical lab required the use of personal protective equipment including (at minimum) lab safety glasses/goggles, latex/nitrile gloves and a lab coat. The acidification and/or pH adjustment of samples and/or solutions was performed within a fume hood to eliminate the potential for harsh/harmful fumes within the lab.

2.5.2 Health and Safety Plans (HASP)

Prior to work being performed in any analytical lab, a detailed reporting of the work, including the chemicals, equipment and/or compressed gases involved, the method utilized and all MSDS information

45 pertaining to all chemicals utilized over the course of completing the analyses, were required. The required

HASP(s) written authorizing the use of both humic acid and/or arsenic III/V speciation standards during analysis can be found in the Supplemental Information section.

2.5.3 Waste Collection and/or Removal

As previously mentioned, arsenic is an EPA regulated analyte with an allowable maximum contaminant level, in water, of 10 ppb. All HPLC-ICP-MS instrument waste generated during arsenic speciation analysis and all humic acid waste generated during UV-Vis/fluorescence spectroscopy analysis was collected in HDPE carboys to eliminate the potential for high concentration arsenic and/or heavy metal introduction into the Cincinnati sewer system. Prior to waste elimination, via the sink, samples were collected from the carboy(s), acidified for at least 24 hours prior to analysis and analyzed by ICP-AES.

Once “waste” samples were confirmed to be within EPA allowable limits for regulated analytes, carboys containing instrument waste were neutralized between pH 6-8 and eliminated.

References

4. Pressman, J. G.; Richardson, S. D.; Speth, T. F.; Miltner, R. J.; Narotsky, M. G.; Hunter, E. S., 3rd; Rice, G. E.; Teuschler, L. K.; McDonald, A.; Parvez, S.; Krasner, S. W.; Weinberg, H. S.; McKague, A. B.; Parrett, C. J.; Bodin, N.; Chinn, R.; Lee, C. F.; Simmons, J. E., Concentration, chlorination, and chemical analysis of drinking water for disinfection byproduct mixtures health effects research: U.S. EPA's Four Lab Study. Environ Sci Technol 2010, 44 (19), 7184-92. 31. (a) Agency, E. P., National Secondary Drinking Water Regulations. Office, C. o. F. R., Ed. Office of the Federal Register National Archives and Records Administration: 2002; Vol. 19; (b) Smedley, P. L.; Kinniburgh, D. G., A review of the source, behaviour and distribution of arsenic in natural waters. Applied Geochemistry 2002, 17 (5), 517-568. 54. Corradini, D.; Eksteen, E.; Eksteen, R.; Schoenmakers, P.; Miller, N., Handbook of HPLC. Taylor & Francis: 2011. 55. Frank M Dunnivant, J. W. G., Flame Atomic Absorbance and Emission Spectrometry and Inductively Coupled Plasma - Mass Spectrometry. 2008. 56. Bednar, A. J.; Garbarino, J. R.; Ranville, J. F.; Wildeman, T. R., Preserving the Distribution of Inorganic Arsenic Species in Groundwater and Acid Mine Drainage Samples. Environmental Science & Technology 2002, 36 (10), 2213-2218. 57. Almassalkhi, B. A. Arsenic Speciation, Detection, and Quantification in Drinking Water using High Performance Liquid Chromatography and Inductively Coupled Plasma Mass Spectrometry. University of Cincinnati, 2009.

Chapter 3

Results

3.0 Introduction This chapter includes a detailed reporting of all the analysis results obtained during the completion of the work contained within this thesis. The reported results include the UV-Vis spectroscopy analysis of humic acid and several “model” compounds at two different sodium bicarbonate (NaHCO3) concentrations and various solution pH. In addition, results obtained from fluorescence analysis performed to confirm the presence of a chromophore in HA samples analyzed are shown. Concluding this chapter will be the analysis results verifying a proposed HPLC-ICP-MS As III/V speciation method, including results from an As III/V stability analysis and analyses of As III/V samples spiked with dissolved organic matter (cDOM). Also included within this chapter are brief discussions of the aforementioned results providing insight, affirmation and/or alternate reasoning using both the results reported herein and results reported in peer-reviewed journals.

3.1 UV-Vis and Fluorescence of Humic Substances 3.1.1 Humic acid UV-Vis Spectroscopy:

Initial attempts to analyze HA using Ultraviolet-Visible (UV-Vis) Spectroscopy in water showed, in general, the absorbance of HA increased with both increasing pH and increasing amounts of sodium bicarbonate

(NaHCO3) added. This can be seen in Figure 3.1.1-1 A and B showing the UV-Vis spectra obtained during analysis of 10 ppm solutions of HA in 10 mM NaHCO3 (Figure 3.1.1-1 A) and 50 mM NaHCO3 (Figure 3.1.1-

1 B).

Humic Acid in 10 mM NaHCO3 3

2.5

1.5

1 Absorbance (AU) 0.5

0 200 250 300 350 400 Wavelength (nm)

Humic Acid in 50 mM NaHCO3 6

2 Absorbance (AU)

0 200 250 300 350 400 Wavelength (nm)

Figure 3.1.1-1 A & B. Changing UV-Vis absorbance detected in HA, varying the pH, in 10 and 50 mM NaHCO3. The 10 mM (A) and 50 mM (B) NaHCO3 HA solutions analyzed are: pH 2 ( ); 6 ( ); 8 ( ); and 10 ( ). Spectra shown are not “blank” corrected.

The UV-Vis spectra shown in Figure 3.1.1-1 A and B are displayed prior to the application of any

“blank” correction to highlight the differences between the spectra before and after. In general, the UV-Vis spectra obtained for HA in 10 mM NaHCO3, show the maximum absorbance occurs somewhere in the UV region, below 200 nm with minimal absorbance detected above 200 nm, which may suggest limited conjugation

48 within the HA analyzed. However the spectra, of HA in 50 mM NaHCO3 shows not only a red shift toward higher wavelengths and increasing absorbance intensity but also some peak definition with increasing pH which may suggest solvent strength plays a role in HA UV-Vis detection in solution. Research performed by Conte et al. [19b] suggested molecules of HA molecules, in solution, associate through weak, hydrophobic, electrostatic interactions instead of polymerization. Therefore at lower pH, dissolved HA and/or DOM molecules may contract or form weak intra-molecular bonds with neighboring atoms within the molecule, possibly to protect hydrophobic groups, similar to proteins folding to protect active sites when dissolved in solution. This could account for the decreased absorbance detected for HA in 10 mM NaHCO3 and may instead show a hypsochromic shift toward lower wavelengths as HA molecules “folding” in solution decrease their surface area, limiting exposure of relevant functional groups to the instrument beam, decreasing the absorbance and detection above 200 nm. This may indicate the increased absorbance intensity detected at both higher

NaHCO3 concentrations and pH may be caused by increased diffusion, and exposure, of the electronic structures responsible for UV-Vis absorbance in solution and/or interaction between HA molecules and the solvent. [1].

Comparing the HA spectra in Figure 3.1.1-1 A and B show similar, increases in the intensity of HA absorbance, at both 10 and 50 mM NaHCO3, as the solution pH increases from pH 2 to pH 10. The HA absorbance at pH

2 and 6, in both samples sets appears to almost overlap, showing little difference between 10 or 50 mM NaHCO3

HA samples. This is more easily seen in Figure 3.1.1-2 A-D, showing the individual pH comparison of HA

UV-Vis spectra obtained in solutions of 10 and 50 mM NaHCO3.

Humic Acid - pH 2 (uncorrected) 1.8 1.6 1.4 1.2 1 0.8 0.6

Absorbance (AU) 0.4 0.2 0 200 250 300 350 400 Wavelength (nm)

Humic Acid - pH 6 (uncorrected) 3

2.5

1.5

1 Absorbance (AU) 0.5

0 200 250 300 350 400 Wavelength (nm)

Figure 3.1.1-2 A-D. UV-Vis spectra of HA in 10 and 50 mM NaHCO3, varying solution pH. The 10 mM NaHCO3 ( ) and 50 mM NaHCO3 ( ) solutions shown were analyzed at pH: 2 (A), 6 (B), 8 (C) and 10 (D). Spectra shown are not “blank” corrected.

Humic Acid - pH 8 (uncorrected) 3

2.5

1.5

1 Absorbance (AU) 0.5

0 200 250 300 350 400 Wavelength (nm)

Humic Acid - pH 10 (uncorrected) 7

2 Absorbance (AU) 1

0 200 250 300 350 400 Wavelength (nm)

The individual HA spectra, separated by solution pH, more clearly show the increased absorbance with increasing pH and while there are minimal features seen in the HA spectra, at all pH’s analyzed in 10 mM

NaHCO3, the spectra suggested the appearance of a shoulder starting at approximately 215 nm. This could be an indication of the presence of single chromophoric, carbonyl and/or nitro, [58], groups and/or alternate electronic structures in HA that absorb primarily below 200 nm. In addition, the hypsochromic, or blue, shift toward shorter wavelengths seen in HA at 10 mM NaHCO3 may be caused by increased solvent polarity.

However, while the UV-Vis spectra obtained from HA in 50 mM NaHCO3, in Figure 3.1.1-2 A-D, show similar hyperchromic effects at pH 2, 6 and 8, as seen in 10 mM NaHCO3, the increase in absorbance intensity at pH

10 is larger in 50, compared to 10, mM NaHCO3, and includes the appearance of a portion of a peak starting at

~ 234 nm, cresting at ~ 206 nm. While there are minimal features seen in the UV-Vis spectra of HA shown in

Figures 3.1.1-1 and 3.1.1-2 to conclusively indicate the electronic structures contained within the HA analyzed here, there is enough information, or lack of information, to suggest which structures may not contribute to the

UV-Vis spectra seen here. For example, the Π to Π* transitions typical of K-Banks show extremely high absorbance intensities and very little change in absorbance due to changes in solvent polarity. The spectra seen here, show minimal absorbance intensity compared to those typical of the Π to Π* transitions seen in K-Band transitions [58]. This lack of absorbance is further apparent in the “blank” corrected UV-Vis spectra of HA at

10 and 50 mM NaHCO3 seen in Figure 3.1.1-3 A-D.

Humic Acid - pH 2 (corrected) 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 Absorbance (AU) -0.05 -0.1

-0.15

280 200 220 240 260 300 320 340 360 380 400 Wavelength (nm)

Figure 3.1.1-3 A-D. Comparison the UV-Vis absorbance detected for HA in 10 and 50 mM NaHCO3, varying solution pH. The 10 mM NaHCO3 ( ) and 50 mM NaHCO3 ( ) HA solutions were analyzed at pH 2 (A), 6 (B), 8 (C) and 10 (D).

Humic Acid - pH 6 (corrected) 0.35 0.3 0.25 0.2 0.15 0.1 0.05

Absorbance (AU) 0 -0.05

-0.1

200 220 240 260 280 300 320 340 360 380 400

Wavelength (nm)

Humic Acid - pH 8 (corrected) 0.8

0.6

0.4

Absorbance Absorbance (AU) 0.2

200 220 240 260 280 300 320 340 360 380 400

Wavelength (nm)

Humic Acid - pH 10 (corrected) 4.0

3.0

2.0

Absorbance (AU) 1.0

0.0

260 200 220 240 280 300 320 340 360 380 400 Wavelength (nm)

In addition to the limited absorbance intensity seen in the HA samples analyzed, highlighted by the

“blank” corrected HA spectra in Figure 3.1.1-3 A-D, there was also a definite change in the absorbance intensities seen with both increasing solution pH, especially at pH 10 in 50 mM NaHCO3, and increasing

NaHCO3 concentrations. Granted, increasing the concentration of NaHCO3 in solution from 10 to 50 mM is small when compared to the molarity of water, 55 M, it still represents at 5 fold increase in carbonate/carbonic acid depending upon the final solution pH. These UV-Vis features all suggest electronic structures capable of

K-Band transitions are not present within the samples of HA analyzed.

While, the minimal absorbance detected for the HA samples described may still suggest the electronic structures present are those with single chromophore groups capable of n to Π* transitions, or R-Bands [58], as previously mentioned. The presence of functional groups capable of the Π to Π* transitions seen in B-Bands and/or E-Bands (auxochrome substituted benzenes) cannot be excluded based on the UV-Vis seen here

54 however, like K-Band transitions, the high absorbance intensities typical of E-Band transitions make these type of structures seem unlikely.

Due to the limited and inconclusive results obtained from the analysis of HA in NaHCO3, and the previously mentioned belief that multiple different functional groups are present in HA molecules, several different model molecules that included some of these functional groups were analyzed, using similar solvent conditions, and compared to those obtained from HA. Five different model compounds, D-(-)-ribose, mandelic acid, DL-tyrosine, γ-butyrolactone (BGL) and (+)-catechin, were chosen for analysis. These molecules were chosen in part because they contained functional groups believed to be present in HA molecules including carboxylic acid, carbonyl, aldehyde, phenyl and hydroxyl groups. In addition, they provided a varied mix of molecular structures ranging from saturated hydrocarbons, (D-(-)-ribose and BGL, capable of n to Π* transitions, to aromatic molecules, mandelic acid, catechin and DL-tyrosine, capable of Π to Π* transitions, that allowed for the examination/comparison of most of the potential electronic transitions detectable by UV-Vis.

3.1.2 Model compounds UV-Vis Spectroscopy:

Similar to the HA analyses described above, the uncorrected UV-Vis spectra showing the model compounds analyzed will be discussed prior to the “corrected” spectra to demonstrate the effect, if any, of the solvent on the UV-Vis obtained for each molecule. Figure 3.1.2-1 A-F shows the UV-Vis spectra obtained for

D-(-)-ribose in solution at a concentration of 75 ppm.

D-Ribose - pH 2 (uncorrected) 1.8 1.6 1.4 1.2 1.0 0.8 0.6

Absorbance (AU) 0.4 0.2 0.0 200 250 300 350 400 Wavelength (nm)

D-Ribose - pH 4 (uncorrected) 3.0

2.5

2.0

1.5

1.0 Absorbance (AU) 0.5

0.0 200 250 300 350 400 Wavelength (nm)

Figure 3.1.2-1 A-F. UV-Vis analysis spectra of D-(-)-ribose in 10 and 50 mM NaHCO3 solutions, varying pH. The 10 mM NaHCO3 ( ) and 50 mM NaHCO3 ( ) ribose solutions were analyzed at pH 2 (A), 4 (B), 6 (C), 8 (D), 10 (E) and 11 (F). The concentration of ribose in solution was 75 ppm. Spectra shown are not “blank” corrected.

D-Ribose - pH 6 (uncorrected) 3.0

2.5

2.0

1.5

1.0 Absorbance (AU) 0.5

0.0 200 250 300 350 400 Wavelength (nm)

D-Ribose - pH 8 (uncorrected) 2.5

2.0

1.5

1.0

Absorbance (AU) 0.5

0.0 200 250 300 350 400 Wavelength (nm)

D-Ribose - pH 10 (uncorrected) 4.0 3.5 3.0 2.5 2.0 1.5

Absorbance (AU) 1.0 0.5 0.0 200 250 300 350 400 Wavelength (nm)

D-Ribose - pH 11 (uncorrected) 4.0 3.5 3.0 2.5 2.0 1.5

Absorbance (AU) 1.0 0.5 0.0 200 250 300 350 400 Wavelength (nm)

Ribose contains both hydroxyls and an aldehyde group, in the Fischer projection, located at carbon 1 of the molecule however only the (CO) of the aldehyde group is a potential chromophore. This carbonyl group indicates ribose is capable of n to Π* transitions which may have been a result of the electronic transitions detected in the HA samples analyzed. Theoretically, if the solvent were transparent at the wavelengths analyzed and the molecule in solution was in its stable ground state, any electronic transitions detected following

58 exposure of the molecule, and others like it containing chromophores like carbonyl groups, to UV excitation capable of promoting electrons to an excited state would be due to the carbonyl group alone. However, since the molecules analyzed were dissolved in a polar solvent, the spectra could also show certain solvent effects.

Similar to the HA spectra previously described, the UV-Vis spectra for ribose shows minimal absorbance features above 250 nm. In general, ribose in 10 mM NaHCO3 shows minimal absorbance intensity, typical of unconjugated molecules. Initial comparison of the UV-Vis spectra obtained for ribose in 10 mM

NaHCO3 and the spectra obtained for HA in the same solution, shows some similarity between the two samples sets. Both the HA and ribose spectra display similar features with potentially emerging peak characteristics as the spectra approach 200 nm. However, the spectra for ribose in 10 mM NaHCO3 shows an initial increase in absorbance intensity as the pH increases from pH 2 to 4, then shows a slight decrease in absorbance at pH 6.

This absorbance intensity appears to be stable through pH 8 and can be seen to increase at pH 10 with the highest absorbance intensity detected in ribose at pH 11. Also, the potential beginning of a shoulder seen in the

HA spectra at 10 mM NaHCO3 at pH 10 does not appear in the ribose spectra until pH 11 at ~ 203 nm. However, in ribose the feature is more apparent, appearing more like the beginning of a peak rather than the shoulder detected in HA at the same NaHCO3 concentration.

The features seen in the UV-Vis spectra for ribose in 50 mM NaHCO3 also show a fair amount of similarity to the HA spectra previously seen at the same NaHCO3 concentration. However, similar to ribose in

10 mM NaHCO3, the absorbance intensity increases from pH 2 to pH 6, then a slight decrease in absorbance is seen at pH 8 with the intensity increasing again at pH 10 with the highest detected absorbance intensity seen at pH 11. Similar to the 10 mM NaHCO3 ribose spectra, there is a slight discrepancy in the solution pH when the beginning of a peak at ~ 203 nm occurs in the 50 mM NaHCO3 solution. The HA spectra shows a peak appearing at pH 10, while ribose, at the same pH and NaHCO3 concentration only shows the beginning of a shoulder. The appearance of a peak in ribose at 50 mM NaHCO3 does not occur until the sample solution reaches pH 11.

The variation seen in the UV-Vis absorbance intensity for ribose at several of the pH’s monitored in 10 mM NaHCO3, and to a smaller extent at 50 mM NaHCO3, may reflect de-protonation of the number 4 carbon hydroxyl group, causing the molecule to cycle into its ring conformation in solution. Research performed by

Sen et al. suggests ribose, in solution, is an equilibrium mixture of conformations of both furanose and pyranose

[59] which may support this hypothesis. Sen et al used 13C-NMR to show this equilibrium mix of ribose was stable up to pH 7.3 after which the equilibrium cycled between containing either conformations of furanose or conformations of pyranose, not both, until pyranose became the primary conformation detected at pH 10.4 [59].

This was followed by de-protonation of the C-1 aldehyde (COH) group at ~ pH 11.7 as seen in Figure 3.1.2-2, adapted from Sen et al. [59].

Figure 3.1.2-2. 1H-decoupled 13C NMR spectra of 13C (C-1) labeled D-ribose at varying NaOH concentrations. The final pH of each solution is shown in parentheses. Figure shows only enriched carbon. Peaks are assigned to α- and β-ribo-furanose and α-and β-ribo- pyranose products. Samples contained 1 mM labeled (13C) ribose. Chemical shift values are marked in the spectrum based on a previously published report. Figure adapted from Sen et al. [59].

The variations detected in the UV-Vis spectra of ribose, in Figure 3.1.2-1 A-F, including the variation in absorbance intensities and the hint of peak resolution seen at pH 10 and 11 along with increasing absorbance

60 intensity could correlate to the changing conformation equilibrium described by Sen et al. and the subsequent de-protonation of the C-1 (COH) group seen in the 13C-NMR spectra in Figure 3.1.2-2.

The UV-Vis spectra seen for γ-butyrolactone (GBL) in Figure 3.1.2-3 A-D closely resembles the spectra obtained for ribose, Figure 3.1.2-1 A-F, and by default, also shows a fair amount of similarity to the HA spectra seen in Figure 3.1.1-1.

γ-Butyrolactone - pH 2 (uncorrected) 1.6 1.4 1.2 1.0 0.8 0.6

Absorbance (AU) 0.4 0.2 0.0 200 220 240 260 280 300 Wavelength (nm)

γ-Butyrolactone - pH 6 (uncorrected) 2.5

2.0

1.5

1.0

Absorbance (AU) 0.5

0.0 200 220 240 260 280 300 Wavelength (nm)

Figure 3.1.2-3 A-D. UV-Vis analysis spectra of γ-butyrolactone (GBL) in 10 and 50 mM NaHCO3 solutions, varying pH. The 10 mM NaHCO3 ( ) and 50 mM NaHCO3 ( ) GBL solutions were analyzed at pH 2 (A), 6 (B), 8 (C) and 10 (D). The concentration of GBL in solution was 5 ppm. Spectra shown are not “blank” corrected.

γ-Butyrolactone - pH 8 (uncorrected) 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 Absorbance (AU) 0.4 0.2 0.0 200 220 240 260 280 300 Wavelength (nm)

γ-Butyrolactone - pH 10 (uncorrected) 4.5 4.0 3.5 3.0 2.5 2.0 1.5

Absorbance (AU) 1.0 0.5 0.0 200 220 240 260 280 300 Wavelength (nm)

Like ribose and HA, the majority of the features seen for GBL occur below ~ 250 nm and the spectra for GBL shows the same initial decrease in absorbance intensity, detected in ribose in both 10 and 50 mM

NaHCO3 between pH 2 and 6, followed by increasing absorbance thereafter at pH 8 and 10. This includes the appearance of a peak, seen at ~ 205 nm in GBL at pH 10, however in GBL, this peak appearance can be seen in both 10 and 50 mM NaHCO3. In ribose at the same pH the spectra showed only the appearance of a shoulder,

62 with the peak not appearing until pH 11. This is presumably due to the (CO) group, as the lone chromophore in both ribose and GBL, performs the same electronic transition under the applied UV excitation energy. There was also a difference between the intensity of absorbance detected in the spectra for ribose at pH 10 in 50 mM

NaHCO3 which led to an increase in absorbance from approximately 2 AU to over 10 AU. Between pH 8 and

10 in GBL the intensity in absorbance only increased from ~ 1.7 AU to ~ 4 AU. The larger increase in absorbance intensity detected in ribose at pH 10 (and 11), not seen in GBL may be due to increased interaction between the hydroxyl groups in the molecule and the solvent allowing for increased diffusion of absorbed energy into the surrounding solvent.

In addition, the changing absorbance spectrum seen between pH 2 and 11, in GBL compared to the changes seen in ribose, may show hydrolysis of the ketone group [60] followed by ring opening and conversion of GBL to γ-hydroxybutyric (GBH) acid at increasing solution pH. Hennessy et al. demonstrated similar chemistry, showing the hydrolysis of GBL to GHB in alcohol and/or methanol solutions, occurred slowly under acidic conditions with rapid conversion occurring with increasing pH in solution [61].

Similar to the spectra obtained for ribose, the GBL spectra at both 10 and 50 mM NaHCO3 show a good deal of similarity to the HA spectra previously described, however, since ribose and GBL contained the same chromophore this information remains inconclusive as to the potential source of the spectral features detected in HA.

The UV-Vis absorbance spectra obtained during analysis of mandelic acid in 10 and 50 mM NaHCO3 are shown in Figure 3.1.2-4 A-F.

Mandelic Acid - pH 2 (uncorrected) 3.5

3.0

2.5

2.0

1.5

1.0 Absorbance (AU) 0.5

0.0 200 250 300 350 400 Wavelength (nm)

Mandelic Acid - pH 4 (uncorrected) 3.5

3.0

2.5

2.0

1.5

1.0 Absorbance (AU) 0.5

0.0 200 250 300 350 400 Wavelength (nm)

Figure 3.1.2-4 A-F. UV-Vis analysis spectra of mandelic acid in 10 and 50 mM NaHCO3 solutions, varying pH. The 10 mM NaHCO3 ( ) and 50 mM NaHCO3 ( ) mandelic acid solutions were analyzed at pH 2 (A), 4 (B), 6 (C), 8 (D), 10 (E) and 11 (F). The concentration of mandelic acid in solution was 44 ppm. Spectra shown are not “blank” corrected.

Mandelic Acid - pH 6 (uncorrected) 3.5

3.0

2.5

2.0

1.5

1.0 Absorbance (AU) 0.5

0.0 200 250 300 350 400 Wavelength (nm)

Mandelic Acid - pH 8 (uncorrected) 3.5

3.0

2.5

2.0

1.5

1.0 Absorbance (AU) 0.5

0.0 200 250 300 350 400 Wavelength (nm)

Mandelic Acid - pH 10 (uncorrected) 3.5

3.0

2.5

2.0

1.5

1.0 Absorbance (AU) 0.5

0.0 200 250 300 350 400 Wavelength (nm)

Mandelic Acid - pH 11 (uncorrected) 4.0 3.5 3.0 2.5 2.0 1.5

Absorbance (AU) 1.0 0.5 0.0 200 250 300 350 400 Wavelength (nm)

Mandelic acid (pKa 3.41), contains two chromophores, a phenyl (benzene) group capable of Π to Π* transitions and a carboxylic acid group capable of an n to Π* transition. The spectra seen in Figure 3.2.1-4 also shows some similarity to the HA spectra previously described, in both 10 and 50 mM NaHCO3, with the majority of its features seen below 250 nm (a broad shoulder beginning at ~ 236 nm showing increasing absorbance intensity as the trace approaches 200 nm). However very few differences in both the features and/or

66 the absorbance intensity of the spectra obtained at both NaHCO3 concentrations are seen for any of the pH’s analyzed until pH 10 when the 10 and 50 mM NaHCO3 traces begin to diverge slightly with the appearance of a peak at ~ 206 nm in the 50 mM NaHCO3 trace of mandelic acid at pH 10. The sharper peak seen in the spectrum of mandelic acid at pH 10, in 50 mM NaHCO3, had a smoother appearance in the spectra obtained at pH 11, however at the same pH the trace for mandelic acid in 10 mM NaHCO3 now shows a sharp peak atop the broader peak appearing at ~ 206 nm. Presumably the peaks, which appeared more as spikes atop of the much broader peak, may be due to some vibrational excitation of the benzene ring which overlaps the electronic

Π to Π* (B-Band) transitions [58]. These sharp vibrational bands typical of benzene UV spectra are reported to be smoother when UV-Vis spectral analysis occurs in polar solvents [58]. The slight bathochromic shift detected in the peak at ~ 206 nm seen at pH 10 and 11 for mandelic acid in 50 mM NaHCO3 may be due to de- protonation of the meso hydroxyl group. The n to Π* transition of the carboxylic acid group, detectable at approximately 200 nm, was absent potentially as a result of the much broader and more intense absorbance band assigned to the benzene ring. The stability of benzene may be responsible for the lack of changes seen in the UV-Vis spectra obtained for mandelic acid at the majority of the pH’s analyzed. In general, there is enough similarity in the spectra obtained for mandelic acid, features below 250 nm with absorbance intensity increasing as the spectra approach 200 nm, and the HA spectra that the presence phenyl groups capable of Π to Π* transitions like benzene still cannot be discounted.

Similar to mandelic acid, the UV-Vis spectra for catechin (pKa 8.64, 9.70, 11.18 and 13.25 [62]), seen in Figure 3.1.2-5, also shows some similarity to the HA spectra previously described, possibly in part due to the phenyl groups present within both mandelic acid and catechin.

Catechin - pH 2 (uncorrected) 4.0 3.5 3.0 2.5 2.0 1.5

Absorbance (AU) 1.0 0.5 0.0 200 250 300 350 400 Wavelength (nm)

Catechin - pH 4 (uncorrected) 4.0 3.5 3.0 2.5 2.0 1.5

Absorbance (AU) 1.0 0.5 0.0 200 250 300 350 400 Wavelength (nm)

Figure 3.1.2-5 A-F. UV-Vis analysis spectra of catechin in 10 and 50 mM NaHCO3 solutions, varying pH. The 10 mM NaHCO3 ( ) and 50 mM NaHCO3 ( ) catechin solutions were analyzed at pH 2 (A), 4 (B), 6 (C), 8 (D), 10 (E) and 11 (F). The concentration of catechin in solution was 15 ppm. Spectra shown are not “blank” corrected.

Catechin - pH 6 (uncorrected) 4.0 3.5 3.0 2.5 2.0 1.5

Absorbance (AU) 1.0 0.5 0.0 200 250 300 350 400 Wavelength (nm)

Catechin - pH 8 (uncorrected) 4.0 3.5 3.0 2.5 2.0 1.5

Absorbance (AU) 1.0 0.5 0.0 200 250 300 350 400 Wavelength (nm)

Catechin - pH 10 (uncorrected) 4.5 4.0 3.5 3.0 2.5 2.0 1.5

Absorbance (AU) 1.0 0.5 0.0 200 250 300 350 400 Wavelength (nm)

Catechin - pH 11 (uncorrected) 4.5 4.0 3.5 3.0 2.5 2.0 1.5

Absorbance (AU) 1.0 0.5 0.0 200 250 300 350 400 Wavelength (nm)

Like the UV-Vis spectra obtained for mandelic acid, the spectra seen for catechin show the most intense absorbance intensity occurring below 250 nm with increasing absorbance as the spectra approach 200 nm. The spectra for catechin also show the same stability as mandelic acid, with little to no variation in the traces for the molecule seen in the spectra until pH 10 and 11. However, the spectra for catechin show increasing levels of sharp peaks at ~ 203 and 210 nm, previously attributed to vibrational excitation, appearing at the top of the

70 broader peak, which is slightly red shifted in the spectra of catechin, appearing at ~ 210 nm for catechin at pH

2, 4, 6 and 8 in 50 mM NaHCO3. This peak shifts to approximately 219 nm in the spectra of catechin at pH 10 and 11 in 50 mM NaHCO3. In addition, the sharp “vibrational” peaks seen atop the broad peak assigned to benzene, which were only apparent in the 10 mM NaHCO3 mandelic acid sample at pH 11, are seen in catechin at both pH 10 and 11 in 10 mM NaHCO3. The changes, or enhancement, of the features seen in the spectra obtained from catechin, compared to those seen in mandelic acid, are attributed to there being two benzene rings instead of one, theoretically increasing the features detected rather than the absorbance intensity.

However, since there was limited absorbance detected in the spectra seen here than typically expected from a benzene Π to Π* transition, increased absorbance intensity may have also been a potential result.

Also seen in the spectra of catechin, are slight peaks at ~ 250 and 296 nm that appear at all of the pH’s monitored in catechin. A similar peak at ~ 250 nm was also seen in the spectra of mandelic acid, however its absorbance intensity was so weak it offered limited peak assignment (electronic transition) information. Even the spectra obtained from catechin, where the peaks were readily apparent, shows minimal absorbance in both

10 and 50 mM NaHCO3 at pH 2, 4, 6, 10 and 11. At pH 8 the peaks show decreased absorbance intensity similar to that detected in mandelic acid. These peaks may have been a result of the ortho/meta hydroxyl groups attached to the benzene rings. The difference in absorbance intensity seen at pH 8 could have been a result of protonation/de-protonation of one of the hydroxyl groups, however the 8.64 was the closest pKa to the pH of the solution analyzed so there may be an alternate explanation that is not currently apparent for the decreased absorbance seen.

In general, the appearance of features attributed to the second benzene ring, and possibly the endocyclic oxygen seen in the spectra obtained for catechin shows decreasing similarity to the spectra obtained from HA suggesting multiple chromophore groups similar to those in catechin are most likely not contained within the

HA analyzed in this work.

The final model compound analyzed using UV-Vis was DL-tyrosine (pKa 2.20, 9.11 and 10.07) whose spectra is shown in Figure 3.1.2-6.

Tyrosine - pH 2 (uncorrected) 4 3.5 3 2.5 2 1.5

Absorbance (AU) 1 0.5 0 200 250 300 350 400 Wavelength (nm)

Tyrosine - pH 4 (uncorrected) 4 3.5 3 2.5 2 1.5

Absorbance (AU) 1 0.5 0 200 250 300 350 400 Wavelength (nm)

Figure 3.1.2-6 A-F. UV-Vis analysis of DL-tyrosine in 10 and 50 mM NaHCO3 solutions, varying pH. The 10 mM NaHCO3 ( ) and 50 mM NaHCO3 ( ) DL-tyrosine solutions were analyzed at pH 2 (A), 4 (B), 6 (C), 8 (D), 10 (E) and 11 (F). The concentration of DL-tyrosine in solution was 15 ppm. Spectra shown are not “blank” corrected.

Tyrosine - pH 6 (uncorrected) 4 3.5 3 2.5 2 1.5

Absorbance (AU) 1 0.5 0 200 250 300 350 400 Wavelength (nm)

Tyrosine - pH 8 (uncorrected) 4 3.5 3 2.5 2 1.5

Absorbance (AU) 1 0.5 0 200 250 300 350 400 Wavelength (nm)

Tyrosine - pH 10 (uncorrected) 4.5 4 3.5 3 2.5 2 1.5

Absorbance (AU) 1 0.5 0 200 250 300 350 400 Wavelength (nm)

Tyrosine - pH 11 (uncorrected) 4.5 4 3.5 3 2.5 2 1.5

Absorbance (AU) 1 0.5 0 200 250 300 350 400 Wavelength (nm)

Similar to the spectra seen for both mandelic acid and catechin, the UV-Vis spectra for tyrosine shows a high degree of stability in the both the features, peaks at ~ 213 and 238 nm, and the absorbance intensity at all of the pH’s analyzed, in both 10 and 50 mM NaHCO3, until pH 10 and 11. At pH 10 and 11, the trace for tyrosine in 50 mM NaHCO3 begins to diverge slightly showing both a slight red shift and the appearance of a sharp peak at ~ 206 nm. The sharp peak at ~ 206 nm, may again be due to vibrational rotation of the benzene

74 ring overlapping the Π to Π* electronic transition detected. While the peaks seen at 213 and 238 may be due to the para hydroxyl group and the amine. While the amine isn’t close enough to the benzene ring to participate in any type of resonance, it is close enough to the carboxyl group which may have resulted in the peak seen at

~ 213 nm.

Similar to the spectra seen for catechin, correlating the features seen within the UV-Vis spectra to the structure of the molecule suggests a similar combination of molecular structure was not part of the HA samples analyzed by UV-Vis spectroscopy.

Examination of the spectra obtained from the model compounds up until this point indicate there could be single chromophore groups, e.g. carbonyls, and/or aromatic chromophores, e.g. benzene or phenols, within the HA analyzed here. The broad features, appearing below 250 nm, with minimal absorbance that increases towards 200 nm while showing some degree of reaction to changes in both the concentration of NaHCO3 and the pH of the solvent that was seen in ribose, GBL and mandelic acid suggest these groups may be present in

HA. In contrast, the spectra for catechin and tyrosine suggest, while an aromatic group may be present in the

HA analyzed in this work, multiple aromatic groups with various substituents close enough to potentially diffuse the energy absorbed, resulting in wavelength “shifts” similar to those seen for catechin and tyrosine, would most likely result in UV-Vis spectra not seen for the HA described here.

For comparison, the UV-Vis spectra obtained for the 10 and 50 mM NaHCO3 blanks are shown in

Figure 3.1.2-7 A and B and the “blank” corrected UV-Vis spectra for ribose, GBL, mandelic acid, catechin and tyrosine are shown in Figures 3.1.2-8, 9, 10, 11 & 12 (A and B).

10 mM NaHCO3 Blank 3.5

2.5

1.5

1 Absorbance (AU) 0.5

0 200 250 300 350 400 Wavelength (nm)

50 mM NaHCO3 Blank 4 3.5 3 2.5 2 1.5

Absorbance (AU) 1 0.5 0 200 250 300 350 400 Wavelength (nm)

Figure 3.1.2-7 A-C. UV-Vis analysis spectra of 10 and 50 mM NaHCO3 “method” blanks, varying pH. The pH of the 10 mM NaHCO3 (top) and the 50 mM NaHCO3 (bottom) solutions shown: pH 2 ( ); 4 ( ); 6 ( ); 8 ( ); 10 ( ) and 11 ( ).

Similar to the UV-Vis spectra seen for HA, ribose, GBL, mandelic acid, catechin and tyrosine, the

NaHCO3 blanks show some features appearing below 250 nm, with increasing absorbance as the spectra approaches 200 nm. The spectra seen in Figure 3.1.2-7 also show some increasing absorbance intensity as the pH in solution increases. This is most likely a result of n to Π* transitions of the carbonyl group present within the molecule [58]. The spectra seen for 10 mM NaHCO3 shows minimal absorbance at pH 2, 4, 6 and 8, with the absorbance increasing at pH 10 and 11 to intensities which could potentially interfere with the assignment

76 of functional groups observed during UV-Vis analysis of molecules prepared in a solvent containing NaHCO3 at the concentrations shown here. Similarly, the spectra obtained from the 50 mM NaHCO3 blank also show minimal features at pH 2, 4, 6 and 8 however, the spectra at all the pH’s analyzed, except pH 8, show increased absorbance intensities compared to those detected in the 10 mM NaHCO3 blank. In addition, both the 10 and the 50 mM NaHCO3 blanks show the appearance of a shoulder at ~ 203 nm (10 mM NaHCO3) and at ~ 206 nm

(50 mM NaHCO3), corresponding to the carbonyl group within the molecule. However, examination of the spectra obtained for the 10 mM NaHCO3 blank at pH 6 shows minimal absorbance and features in the trace at pH 6, while the 50 mM NaHCO3 blank at pH 8 shows almost no absorbance and/or features. If this had been seen in all the samples, it may have been an indication of contamination in the solution that inhibited and/or quenched the absorbance detected at the other pH’s monitored. However, the NaHCO3 were prepared at the same time using the same supplies.

Despite the detections seen here for both 10 and 50 mM NaHCO3, use of carbonate in solutions used to perform UV-Vis may not have an adverse effect on the spectra obtained, as long as the absorbance was sufficiently intense to negate the “corrections” applied. However, as shown in Figures 3.1.2-8 and 9 for ribose and GBL, applying “blank” corrections to the spectra obtained during analysis of these molecules resulted in negative absorbances at most of the pH’s monitored.

Ribose in 10 mM NaHCO3 - corrected 2.5 2 1.5 1 0.5 0 -0.5200 250 300 350 400 -1

Absorbance (AU) -1.5 -2 -2.5 -3 Wavelength (nm)

Ribose in 50 mM NaHCO3 - corrected 8

Absorbance (AU) 200 250 300 350 400 -2

-4 Wavelength (nm)

Figure 3.1.2-8 A & B. UV-Vis analysis spectra of ribose in DI water and/or 10 and 50 mM NaHCO3, varying pH. The solutions analyzed were: 756 ppm ribose stock in DI water ( ); and 10 mM NaHCO3 (top) and 50 mM NaHCO3 solutions at pH: 2 ( ); 4 ( ); 6 ( ); 8 ( ); 10 ( ) and 11 ( ). Spectra shown are “blank” corrected.

Butyrolactone in 10 mM NaHCO3 - corrected 0.15 0.1 0.05 0 200 250 300 350 400 450 -0.05 -0.1 -0.15

-0.2 Absorbance (AU) -0.25 -0.3 -0.35 Wavelength (nm)

Butyrolactone in 50 mM NaHCO3 - corrected 1

0.5

0 200 250 300 350 400 450 -0.5

-1

-1.5

Absorbance (AU) -2

-2.5

-3 Wavelength (nm)

Figure 3.2.1-9 A & B. UV-Vis analysis spectra of γ-butyrolactone (GBL) in 10 and 50 mM NaHCO3, varying pH. The 10 mM NaHCO3 (top) and 50 mM NaHCO3 solutions analyzed were at pH: 2 ( ); 4 ( ); 6 ( ); 8 ( ); 10 ( ) and 11 ( ). The concentration of catechin in solution was 5 ppm. The spectra shown are “blank” corrected.

The UV-Vis spectra of the ribose stock solution, prepared at 756 ppm, seen in Figure 3.1.2-8 A and B shows no features and extremely weak absorbance. Internal “quenching”, inhibiting the chromophores absorption of UV radiation decreasing both the appearance of features and the absorbance intensity typical of the carbonyl group transition, could have been the cause of the decreased detection seen in the concentrated stock solution. However, the UV-Vis spectra of a similar stock solution, prepared for DL-tyrosine at 730 ppm

(not shown) was also analyzed and showed the majority of the features, previously described for the molecule at diluted concentrations. In addition, the solution analyzed was not cloudy and several different cuvettes were used over the course of completing the scans seen here. The blank corrected spectra seen for GBL in 10 mM

NaHCO3 in Figure 3.1.2-9 A shows negative absorbance only for the sample analyzed at pH 11, which was the blank that showed the highest absorbance intensity at both 10 and 50 mM NaHCO3. In addition, the “blank” corrected spectra seen for mandelic acid (Figure 3.1.2-10 A and B), catechin (Figure 3.1.2-11 A and B) and tyrosine (Figure 3.1.2-12 A and B) show both the absorbance intensities and features comparable to those seen in the “uncorrected” spectra.

Mandelic Acid in 10 mM NaHCO3 - corrected 3.0

2.5

2.0

1.5

1.0

Absorbance (AU) 0.5

0.0 200 250 300 350 400 -0.5 Wavelength (nm)

Mandelic Acid in 50 mM NaHCO3 - corrected 2.5

2.0

1.5

1.0

0.5 Absorbance (AU)

0.0 200 250 300 350 400 -0.5 Wavelength (nm)

Figure 3.1.2-10 A & B. UV-Vis analysis spectra of mandelic acid in 10 and 50 mM NaHCO3, varying pH. The 10 mM NaHCO3 (top) and 50 mM NaHCO3 solutions analyzed were at pH: 2 ( ); 4 ( ); 6 ( ); 8 ( ); 10 ( ) and 11 ( ). The concentration of catechin in solution was 44 ppm. The spectra shown are “blank” corrected.

Catechin in 10 mM NaHCO3 - corrected 3.5

2.5

1.5

Absorbance (AU) 0.5

0 200 250 300 350 400 -0.5 Wavelength (nm)

Catechin in 50 mM NaHCO3 - corrected 3.5

2.5

1.5

Absorbance (AU) 0.5

0 200 250 300 350 400 -0.5 Wavelength (nm)

Figure 3.1.2-11 A & B. UV-Vis analysis spectra of catechin in 10 and 50 mM NaHCO3, varying pH. The 10 mM NaHCO3 (top) and 50 mM NaHCO3 solutions analyzed were at pH: 2 ( ); 4 ( ); 6 ( ); 8 ( ); 10 ( ) and 11 ( ). The concentration of catechin in solution was 15 ppm. The spectra shown are “blank” corrected.

Tyrosine in 10 mM NaHCO3 - corrected 3

2.5

1.5

Absorbance (AU) 0.5

0 200 250 300 350 400 -0.5 Wavelength (nm)

Tyrosine in 50 mM NaHCO3 - corrected 2 1.8 1.6 1.4 1.2 1 0.8 0.6

Absorbance (AU) 0.4 0.2 0 -0.2200 250 300 350 400 Wavelength (nm)

Figure 3.1.2-12 A & B. UV-Vis analysis spectra of tyrosine in 10 and 50 mM NaHCO3, varying pH. The 10 mM NaHCO3 (top) and 50 mM NaHCO3 solutions analyzed were at pH: 2 ( ); 4 ( ); 6 ( ); 8 ( ); 10 ( ) and 11 ( ). The tyrosine concentration in solution was 15 ppm. The spectra shown are “blank” corrected.

However, examination of the “blank” corrected UV-Vis spectra shown for mandelic acid, catechin and tyrosine in both 10 and 50 mM NaHCO3 show, a general red shift in the features detected, potentially due to increased solvent interaction allowing the molecules greater diffusion of the absorbed energy into the surrounding solvent. This was, again, more pronounced in the samples at pH 10 and 11 in both 10 and 50 mM

NaHCO3. Depending on the molecule being analyzed and the venue, qualitative and/or quantitative, applying

UV-Vis spectroscopy to molecules in solvents containing NaHCO3 may negatively impact the results obtained.

However, as seen in the spectra described here, in general, the addition of NaHCO3 revealed some features in

HA which were otherwise obscured when the sample was analyzed in water.

Following “blank” correction, comparison of the results seen at both 10 and 50 mM NaHCO3 for ribose,

GBL, mandelic acid, catechin and tyrosine to those detected for HA at the same solvent concentration now show more similarity between HA and both GBL (at pH 2, 6 and 8 in 10 mM NaHCO3) and mandelic acid than the other model compounds analyzed. This could indicate, again, the presence of either single chromophores and/or aromatic groups with minimal conjugation and/or a large degree of separation between chromophores within the molecule.

Typically, the application of “blank” correction reduces the level of potential error due to experimental conditions which are unavoidable, e.g. solvents/chemicals used during sample preparation, in the case of UV-

Vis distortion in the quartz cuvette used during analysis and/or contamination from chemicals/containers etc. used during analysis. However, while the application of “blank” correction did not alter the overall final conclusions described here, the differences seen in the spectra obtained highlight the importance of applying some form of blank correction to limit potential misinterpretation of data obtained during analysis.

In conclusion, analysis of ribose, GBL, mandelic acid, catechin and tyrosine to evaluate and compare the UV-Vis transitions detected from their various functional groups (carbonyl, carboxylic acid, hydroxyl, phenyl and amine) showed some similarity between the spectra obtained for HA and both GBL (at pH 2, 6 and

8 in 10 mM NaHCO3) and mandelic acid. The similarities detected could indicate the presence of single chromophores and/or aromatic groups, similar to benzene or phenols, with minimal conjugation and/or well separated chromophores within the molecule. Also, the addition of dilute concentrations of NaHCO3 in the solvent used to dissolve and/or perform UV-Vis analysis of samples of HA may result in wavelength shifts and/or changes in the absorbance intensity detected which may assist in assigning functional groups responsible for the transitions detected during analysis. This may be a potential area of manipulation for developing future methods to analyze HSs in their native form. However, additional analysis comparing different potential

84 solvents, polar and non-polar, and several sources of DOM, both commercial and naturally occurring, at various organic carbon concentrations needs to be performed to verify the results and conclusions made over the course of performing the work described here.

3.1.3 Humic Acid Calibration

To determine if UV-Vis analysis could be used to calculate the relative concentration of “prepared” samples of HA in NaHCO3, the absorbance detected at 254 nm for solutions prepared in 50 mM NaHCO3 at concentrations of 2, 4, 6, 8, 10, 12 and 15 ppm HA was analyzed. A calibration curve showing the absorbance for each HA sample at 254 nm was displayed in Figure 3.1.1-1.

0.4 0.35 0.3 0.25 y = 0.0253x - 0.0007 0.2 R² = 0.9986 0.15

0.1 Absorbance (AU) 0.05 0 0 2 4 6 8 10 12 14 16 Concentration (ppm)

254 nm Linear (254 nm)

Figure 3.1.3-1. UV-Vis Humic Acid calibration curve in 50 mM NaHCO3, pH 7 at 254 nm. Solutions were prepared at HA concentrations of 0, 2, 4, 6, 8, 10, 12 and 15 ppm in 50 mM NaHCO3, pH 7.06 and scanned from 200 to 800 nm at 280 nm per minute in a 1 cm quartz cuvette and a 1 nm slit width.

The calibration curve in Figure 3.1.3-1 shows a linear response to the prepared HA solutions indicating adequate correlation between the UV-Vis absorbance detected at the instrument and the concentration of HA in solution. To confirm that UV-Vis absorbance at 254 nm provided the most accurate calibration of HA the absorbance detected at 254 nm was compared to absorbance detected at 200, 215, 260, 300 and 350 nm and displayed in Figure 3.1.3-2.

0.5

0.4

0.3

0.2

Absorbance (AU) 0.1

0 0 5 10 15 Concentration (ppm)

200 nm 215 nm 254 nm 260 nm 300 nm 350 nm

Figure 3.1.3-2. Comparing UV-Vis HA detection at varying wavelengths. Humic Acid calibration in 50 mM NaHCO3, pH 7.00 at 200, 215, 254, 260, 300 and 350 nm. Solutions were prepared at HA concentrations of 0, 2, 4, 6, 8, 10, 12 and 15 ppm in 50 mM NaHCO3, pH 7.00 and scanned from 200 to 800 nm at 280 nm per minute in a 1 cm quartz cuvette and a 1 nm slit width.

Figure 3.1.3-2 shows calibration of the instrument at 254, 260 and 300 nm provided comparable results for the HA solutions prepared, displaying calculated R2 values of 0.9986 at each wavelength. However, there was a large amount of variation in the slope calculated at 300 nm (0.0188x) compared to the slopes at 254 and

260 nm (0.0253x and 0.0246x) potentially due to a decrease in absorbance intensity by HA in solution as the energy of the wavelength applied decreased. This may suggest that calibration and subsequent detection of HA using a standard additions method of UV-Vis analysis could provide some indication of the concentration of

HA in unknown samples. However as previously mentioned similar analysis performed under varying solvent conditions would be required to confirm these results.

3.1.4 Humic Acid Fluorescence

The maximum UV-Vis absorbance detected for all of the HA/DOM samples analyzed was between

200 and 240 nm, which could support the UV-Vis analysis that aromatic functional groups may be responsible for the UV-Vis spectra obtained as previously described. The UV-Vis spectra, seen in Figure 3.1.1-1, show the results obtained during analysis of HA in dilute NaHCO3. The results show absorbance, in the aromatic region, which may be a result of electronic transitions typical of single chromophores including carbonyl and/or phenyl

86 groups, like benzene, unconjugated aromatic chromophores and/or aromatic chromophores that are well spaced within the samples analyzed. In an attempt to confirm the UV-Vis results obtained, 5 ppm HA samples were prepared in 50 mM NaHCO3 at pH 2, 6, 7, 8, 9, 10 and 11 and subjected to fluorescence analysis. Since the number and/or type of aromatic groups within the HA samples analyzed was unknown, the excitation/emission

(ex/em) in the samples was monitored at 3 separate wavelengths: 249/260, 280/300 and 360/380 nm, excitation wavelength shown on the left, first wavelength monitored for fluorescence emission on the right, and an ex/em slit width of 10/10. The uncorrected emission spectra obtained during analysis are shown in Figure 3.1.4-1 A-

Humic Acid 260 nm Emission - uncorrected 2.0E+02 1.8E+02 1.6E+02 1.4E+02 1.2E+02 1.0E+02 8.0E+01 6.0E+01 4.0E+01

Emission Emission Intensity (AU) 2.0E+01 0.0E+00 250 300 350 400 450 500 550 600 Wavelength (nm)

pH 2 pH 6 pH 7 pH 8 pH 9 pH 10 pH 11

Figure 3.1.4-1 A-C. Humic Acid fluorescence detected at 260, 300 and 390 nm in 50 mM NaHCO3. The detection of HA fluorescence emission occurred at 260 nm (A), 300 nm (B) and 390 nm (C), analyzed in 50 mM NaHCO3 at varying pH: 2 ( ), 6 ( ), 7 ( ), 8 ( ), 9 ( ), 10 ( ) and 11 ( ). The HA concentration analyzed was 5 ppm. The spetra shown are uncorrected.

Humic Acid 300 nm Emission - uncorrected 1.6E+02 1.4E+02 1.2E+02 1.0E+02 8.0E+01 6.0E+01 4.0E+01

Emission Emission Intensity (AU) 2.0E+01 0.0E+00 290 340 390 440 490 540 590 Wavelength (nm)

Humic Acid 390 nm Emission - uncorrected 1.0E+02 9.0E+01 8.0E+01 7.0E+01 6.0E+01 5.0E+01 4.0E+01 3.0E+01

2.0E+01 Emission Emission Intensity (AU) 1.0E+01 0.0E+00 380 430 480 530 580 630 680 Wavelength (nm)

Figure 3.1.4-1 A-C. Uncorrected Humic Acid fluorescence detected at 260, 300 and 390 nm in 50 mM NaHCO3. The detection of HA fluorescence emission occurred at 260 nm (A), 300 nm (B) and 390 nm (C), analyzed in 50 mM NaHCO3 at varying pH: 2 ( ), 6 ( ), 7 ( ), 8 ( ), 9 ( ), 10 ( ) and 11 ( ). The HA concentration analyzed was 5 ppm. The spectra shown are uncorrected.

The HA samples were analyzed in triplicate, with several minutes “rest” time in between analysis, and the average wavelength showing maximum fluorescence emission was reported. The emission wavelength maxima was 474.55 ± 8.32 exciting at 249 nm, 484.00 ± 7.83 exciting at 280 nm and 474.80 ± 2.42 exciting at

360 nm, however the spectra seen in Figure 3.1.4-1 A-C show minimal difference in both the wavelength and

88 intensity of the fluorescence emission detected at 260 and 300 nm. Similarities between the average ex/em wavelengths indicated the fluorescence detected may be due to a single fluorophore present in all samples which supported the electronic transition data obtained during UV-Vis analysis.

Fluorescence theory [63] suggested the bathochromic shift toward higher wavelengths seen in the emission spectra in Figures 3.1.4-1 A-C was a typical indicator of increased conjugation in fluorescent molecules. However, while the UV-Vis spectra indicated an aromatic molecule could be responsible for the electronic transitions detected during analysis, the limited absorbance intensity detected and the limited features seen suggested the functional group(s) in HA were lone and/or unconjugated chromophores. Chen et al. reported similar results, showing red shifts occurred in the fluorescence detected during the analysis of samples of soil HA with little known conjugation [12]. However, this may or may not correlate to HA in solution [12].

Blank corrections applied to the UV-Vis spectra obtained during analysis of both HA and the model compounds, previously described resulted in some discrepancy in the spectra obtained. For comparison, the fluorescence emission spectra in Figure 3.1.4-1 A-C were uncorrected. The spectra obtained applying “blank” corrections to the fluorescence emission results previously describe are shown in Figure 3.1.4-2.

Humic Acid Fluorescence at 260 nm - corrected 2.0E+02 1.8E+02 1.6E+02 1.4E+02 1.2E+02 1.0E+02 8.0E+01 6.0E+01

4.0E+01 Emission Emission Intensity (AU) 2.0E+01 0.0E+00 260 300 340 380 420 460 500 540 580 Wavelength (nm)

Humic Acid Fluorescence at 300 nm - corrected 1.5E+02

1.2E+02

9.0E+01

6.0E+01

3.0E+01 Emission Emission Intensity (AU)

0.0E+00 300 350 400 450 500 550 600 Wavelength (nm)

Figure 3.1.4-2 A-C. “Blank” Corrected Humic Acid fluorescence detected at 260, 300 and 390 nm in 50 mM NaHCO3. The detection of HA fluorescence emission occurred at 260 nm (A), 300 nm (B) and 390 nm (C), analyzed in 50 mM NaHCO3 at varying pH: 2 ( ), 6 ( ), 7 ( ), 8 ( ), 9 ( ), 10 ( ) and 11 ( ). The HA concentration analyzed was 5 ppm. The spectra shown are “blank” corrected.

Humic Acid Fluorescence at 300 nm - corrected 9.0E+01 8.0E+01 7.0E+01 6.0E+01 5.0E+01 4.0E+01 3.0E+01

2.0E+01 Emission Emission Intensity (AU) 1.0E+01 0.0E+00 390 420 450 480 510 540 570 600 630 660 Wavelength (nm)

The “blank” corrected fluorescence emission spectra show smoother features and a decrease in the stokes shift detected in each spectra. In addition, minimal emission intensity was lost, at any of the wavelengths monitored, due to the corrections suggesting the use of “blank” correction during fluorescence analysis seen here was beneficial.

The excitation/emission spectra at each wavelength analyzed shows the lowest fluorescence intensity was seen in samples of HA samples at pH 2 and pH 8. The fluorescence intensity shown for the remaining HA samples overlapped, showing minimal separation of peaks and no clear order based on the pH of the sample.

This is most likely an indication of protonation/de-protonation occurring within the samples analyzed and the wavelength variation could be a result of the molecules increased and/or decreased capability to diffuse energy to its surrounding once absorbed. Mobed et al. concluded that changes in solution pH affected fluorescence maxima of certain acidic functional groups in HA [16b]. While Westerhoff et al. demonstrated certain functional groups (amines, carboxyl’s, etc.), believed to be present in HA, decreased the intensity of HA fluorescence in solution [16c]. In addition, certain solvent effects or increased interaction between HA and the

91 polar solvent used during dissolution of the samples analyzed here may have contributed to the variation in fluorescence emission detected. However, as previously mentioned, additional analysis of HA, both commercial and from natural water sources, applying several different solvent conditions, both polar and non- polar, are needed to confirm the findings reported herein.

3.1.5 Conclusion

Due to the severity of the potential health concerns associated with DBPs in water, current EPA guidelines detailed in the Information Collection Rule (ICR) require ongoing research be performed in drinking water treatment [34]. Fueling this push for research was an increased awareness that alternative water sources, namely water reclamation and wastewater treatment, may be necessary in the future to provide adequate water to an ever increasing population with increased urban development and shrinking natural water sources. Waste and reclaimed waters contained significantly higher DOM concentrations than natural waters, and could require specialized techniques to remove enough DOM prior to disinfection to reduce DBP formation [64] to levels similar to natural waters. Detection of the HSs in these alternative water sources throughout the treatment process would require standardized, accurate analytical methods for characterizing and determining the levels of NOM and DOM. The research described herein was performed to determine potential method parameters that would assist in HS characterization and analysis.

Humic acid dissolved in solutions at higher solvent concentrations (50 mM NaHCO3) and higher pH

> pH 8, exhibit increased absorbance intensity when analyzed by UV-Vis Spectroscopy. In general, the spectra obtained over the course of the analyses described here show decreased UV-Vis absorbance at lower pH with increasing absorbance intensity at higher pH at HA concentrations of 10 ppm in 50 mM NaHCO3. Additional

UV-Vis analysis of several model compounds containing carbonyl, carboxylic acid, phenyl, hydroxyl and amine groups suggested either lone chromophores or aromatic molecules, similar to benzene, that were unconjugated and/or well-spaced in the molecules may have been responsible for the electronic transitions detected in the samples of HA analyzed. The increases detected in the absorbance intensity as a result of changing solution pH were most likely due to protonation/de-protonation of certain functional groups which resulted in either

92 increased or decreased interaction between the molecule and the polar solvent affecting its ability to diffuse energy following UV-Vis exposure.

In addition, a calibration curve was generated for HA samples at concentrations of 2, 4, 6, 8, 10, 12 and

2 15 ppm in 50 mM NaHCO3, at pH 7, using spectra obtained at 254 nm that showed a calculated R value of

0.9986. This calibration indicated adequate response to prepared samples of HA at several different concentrations when monitored using UV-Vis analysis at 254 nm. This could potentially be useful in calculating the HA concentration, utilizing some form of standard additions analysis, of an unknown sample of

HA. However, repeat analysis of DOM, from both the same source originally analyzed here as well as several additional sources of different origin, containing both FAs and HAs, under similar experimental conditions, would need to be performed to confirm any of the results reported herein and further elucidate the potential source(s) of the variations detected in the UV-Vis spectra shown in this work.

The detection of UV-Vis absorbance between 200 and ~ 250 nm, suggesting the HA analyzed may contained aromatic groups led to the application of fluorescence analysis on the samples of HA in an attempt to confirm this hypothesis. The detection of fluorescence with the samples of HA analyzed at each pH monitored confirmed the presence of a chromophore within the samples analyzed. Additional analysis of the spectra obtained indicated aromatic groups, like benzene and/or phenols, could be responsible for the fluorescence detected within the HA samples, potentially confirming the UV-Vis results obtained previously mentioned. While the results obtained during fluorescence analysis may indicate the HA analyzed was highly conjugated, this conclusion contradicts the UV-Vis results obtained during the analysis of several model compounds, which suggested the UV-Vis transitions detected were due to lone chromophores and/or unconjugated aromatic molecules. However, as previously stated, additional analysis of HA, both commercial and from natural waters, potentially including non-humic organic carbon sources obtained from natural waters, would be needed to confirm and/or compare these results.

3.2 Arsenic III/V Speciation Analysis Increasing occurrence of high concentrations of arsenic in multiple countries, including the U.S., and the severity of health hazards associated with exposure to arsenic in water (skin lesions, neurological defects and cancer, etc.) [32], led the EPA to lower the MCL for the element from 50 to 10 ppb in 2001 [24]. These developments, combined with recent evidence suggesting species of As III may be more toxic than other species of arsenic, suggest the addition of arsenic speciation analysis to water processing SOPs in high risk areas may be beneficial [5a, 26, 27b, 30]. In addition, the need to assess and analyze the effectiveness of biological arsenic oxidation treatment plans, currently in place or planned for future application, on drinking and source waters requires the implementation of rapid, reliable speciation methods that provides accurate analysis representative of conditions found in-situ. This was extremely important in high risk areas where elevated levels of arsenic are known to occur.

At the pH (6.5-8.5) [31] and redox potential (Figure 1.2-1), prevalent in most drinking/source waters,

As III and V is detected primarily in the form of either arsenous (pKa 9.3) and/or arsenic (pKa’s 2.2, 7 & 11.5) acids, respectively. Based on the Eh-pH diagram for arsenic (Figure 1.2-1) and the aforementioned pKa’s for

As III and V, hydrated As III species are uncharged (+/-0) while As V species carry a negative (-1) charge in water. This charge difference, between As III and V, was successfully exploited to separate the two species using a variation of the Ion-Pair Reversed Phase Chromatography (IP-RPC) method introduced by Almassalkhi in 2009 [57]. The method variation utilized isocratic Reversed-Phase (RP) chromatography to simultaneously separate As III and V in drinking water using separation parameters detailed in Table 2.4-2. Buffering the mobile phase to pH 6.0 both maintained the charge difference between As III and V and resulted in a net positive charge on the ion-pairing agent, tetrabutylammonium hydroxide (TBAH), in the mobile phase. The net positive charge on TBAH in solution supported its interaction with any negatively charged As V molecules in water increasing their retention time on the RP column effectively separating As V from As III in solution.

In addition to confirming the methods detection of As III and V in various water samples the stability

(preservation) of each species and the effect, if any, of concentrated dissolved organic matter (cDOM) on their separation and detection by HPLC-ICP-MS was investigated in this phase of the work.

3.2.1 Detection of As III/V

Chromatograms in Figures 3.2.1-1 A-G, show the HPLC separation and ICP-MS detection of As III and As V in calibration solutions prepared at 1, 10, 25, 50, 75, 100 and 150 ppb of each species.

1 ppb As (III/V) 10 ppb As (III/V) 4.0E+02 4.0E+03 3.0E+02 3.0E+03

2.0E+02 MSSignal(cps) 2.0E+03

MSSignal(cps) -

1.0E+02 ICP 1.0E+03 ICP 0.0E+00 0.0E+00 -1.0E+02 -1.0E+03 0 1 2 3 4 5 0 1 2 3 4 5 Elution Time (min) Elution Time (min)

25 ppb As (III/V) 50 ppb As (III/V) 1.0E+04 2.0E+04 8.0E+03 1.5E+04 6.0E+03

MSSignal(cps) 1.0E+04

MSSignal(cps) - 4.0E+03 -

5.0E+03 ICP 2.0E+03 ICP 0.0E+00 0.0E+00 -2.0E+03 -5.0E+03 0 1 2 3 4 5 0 1 2 3 4 5 Elution Time (min) Elution Time (min)

Figure 3.2.1-1 A-G. Ion-Pair-Reversed Phase Chromatography (IP-RPC) separation and ICP-MS instrument detection and calibration of As III and As V at ppb concentrations in solution. Arsenic III was un-retained and eluted at ~ 1.15 minutes before As V shown eluting at ~ 4.15 minutes. The calibration was obtained using solutions prepared at (ppb) As III/V concentrations of: A. 1; B. 10; C. 25; D. 50; E. 75; F. 100 and G. 150.

75 ppb As (III/V) 100 ppb As (III/V) 3.0E+04 4.0E+04

2.0E+04 3.0E+04

2.0E+04

MSSignal(cps)

MSSignal(cps) - 1.0E+04 -

1.0E+04

ICP ICP 0.0E+00 0.0E+00 -1.0E+04 -1.0E+04 0 1 2 3 4 5 0 1 2 3 4 5 Elution Time (min) Elution Time (min)

150 ppb As (III/V) 6.0E+04 5.0E+04 4.0E+04

3.0E+04 MSSignal(cps) - 2.0E+04

ICP 1.0E+04 0.0E+00 -1.0E+04 0 1 2 3 4 5 Elution Time (min)

The chromatograms in Figure 3.2.1-1 A-G indicate the proposed RP-HPLC-ICP-MS method was successful at both separating and detecting the two arsenic species in solution at the concentrations analyzed.

In solution, at pH 6.0, As III was uncharged and un-retained on the RP C-18 column and is shown eluting before

As V with an average experimental retention time of 1.14 ± 0.02 minutes. While As V, with a -1 charge in solution at the same pH, interacted with the positively charged TBAH in the mobile phase. This led to increased retention of As V in solution due to hydrostatic interaction between As V and TBAH in the mobile phase resulting in average retention times of 4.15 ± 0.42 minutes for As V.

Calibration of the HPLC-ICP-MS was performed daily, prior to analysis with calibration curves generated, one example is shown in Figure 3.2.1-2, utilizing ICP-MS signal response data points integrated by

Agilents’ MassHunter software.

1.4E+06

1.2E+06

1.0E+06

8.0E+05

6.0E+05 MSSignal(cps)

- 4.0E+05 ICP 2.0E+05

0.0E+00 0 50 100 150 Arsenic III Concentration (ppb)

Figure 3.2.1-2. RP-HPLC-ICP-MS As III/V Instrument Calibration. The HPLC-ICP-MS was calibrated using As III/V solutions prepared at concentrations of 1, 10, 25, 50, 75, 100 and 150 ppb of each species. As III - (X); As V - (□).

To date, the proposed method has been used to separate and quantify As III and V during the analysis of more than 3000 samples some of which were performed to corroborate results previously obtained on submitted samples. The average R2 value, calculated from instrument calibration performed during these speciation analyses, was 0.9997 ± 0.0004 for As III and 0.9998 ± 0.0002 for As V. Demonstration of consistently linear instrument response during arsenic calibration further supported the proposed method adequately separated As III (X) and As V (□) allowing for the detection and quantification of both species in solution by HPLC-ICP-MS. The detection limit for the method (MDL) was calculated, taking the average of seven repeat ICP-MS signal measurements for the lowest concentration calibration standard (1 ppb). The MDLs were 0.22 and 0.23 ppb, respectively, for As III and As V.

3.2.2 Preservation of As III/V

As previously stated, the +5 oxidation state of arsenic (As V) is generally stable in water while the +3 oxidation state (As III) is less stable and shows greater dependence upon the surrounding environment including pH, redox [25, 65] and the presence of contaminants including DOM [56, 65b, 66] to maintain its +3 oxidation state.

Current research indicates ingestion of As III, in drinking water, may be more toxic than ingestion of

As V [32]. This information, combined with the Eh-pH diagram showing the stability of As III in water, is more reliant upon the surrounding redox environment, compared to As V which is stable in water, led to the primary focus of the results reported herein to be on any potential changes in As III concentration, except where noted. Theoretically, focusing the reported analyses on As III and its stability in water could facilitate in providing the most accurate, consistent analytical results representative of conditions found in-situ. To date multiple methods of stabilizing the various arsenic species detected in sampled drinking/source water have been employed. Several include controlling the temperature of the sample (0-4° C) and/or adding potential anti- oxidants e.g. hydrochloric (HCl), nitric (HNO3), sulfuric (H2SO4), ethylene-diamine tetra-acetic (EDTA), phosphoric (H3PO4) and ascorbic (A. A.) acids [65b, 67] to As speciation samples. To determine the most reliable As III/V preservative for the proposed speciation method, a four week analysis [67a] of samples prepared at concentrations of 1, 50 and 150 ppb of both species, As III and As V, was performed. For the duration of the four week analysis period, before and after analysis, arsenic preservation samples were stored in 2 mL amber, glass HPLC vials at 4.0 ° C, in “light eliminating” vial storage boxes.

The four As III anti-oxidant scenarios chosen to analyze were: 1. No preservation (Figure 3.2.2-1 A);

2. Ascorbic acid (A. A.) (Figure 3.2.2-1 B); 3. EDTA (Figure 3.2.2-1 C) and 4. HCl (Figure 3.2.2-1 D). Previous analysis of samples of Cincinnati tap water spiked with As III/V and ascorbic acid resulted in the use of 1.7E-

04 M A. A. at 1 ppb As III/V, 2.84E-03 M A. A. at 50 ppb As III/V and 8.52E-03 M A. A. at 150 ppb As III/V.

HCl and EDTA preserved samples were prepared at final concentrations of either 0.05 % HCl or 1.25E-03 M

EDTA [56]. To facilitate interpretation and comparison, results collected from arsenic samples analyzed at

98 concentrations of 1 ppb were placed on the primary y-axis, while the 50 and 150 ppb results were placed on the secondary y-axis. Results displayed in Figures 3.2.2-1 A-D show As III in samples prepared at concentrations of 1, 50 and 150 ppb analyzed at various times, as described in [67a], over a period of four weeks.

Arsenic III - No Preservative

1.0E+04 1.5E+06 ICP

8.0E+03

- MS (cps) Signal MS 6.0E+03 1.0E+06 4.0E+03 5.0E+05

2.0E+03 MSSignal(cps) - 0.0E+00 0.0E+00

ICP 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

Arsenic III in Ascorbic Acid

1.0E+04 1.6E+06 ICP 8.0E+03 1.1E+06 - 6.0E+03 (cps) Signal MS 4.0E+03 6.0E+05 2.0E+03 1.0E+05

0.0E+00 MSSignal(cps) - -2.0E+03 -4.0E+05

0 100 200 300 400 500 600 700 ICP Time (hours)

1 50 150

Figure 3.2.2-1 A-D. Four week arsenic III stability analyzed by RP-HPLC-ICP-MS. Arsenic concentrations (ppb) were: 1 (○) - primary y-axis, 50 (□) and 150 (Δ) - secondary y-axis. Arsenic III samples preservation schemes were: A. Control (no preservative), B. Ascorbic Acid (A. A.), C. 1.25E-03 M EDTA and D. 0.05 % HCl.

Arsenic III in EDTA

1.0E+04 1.5E+06 ICP

8.0E+03 - 1.0E+06 (cps) Signal MS 6.0E+03 4.0E+03 5.0E+05

2.0E+03

MSSignal(cps) - 0.0E+00 0.0E+00 ICP 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

Arsenic III in HCl

1.0E+04 1.6E+06 ICP

8.0E+03 -

1.1E+06 (cps) Signal MS 6.0E+03 6.0E+05 4.0E+03

2.0E+03 1.0E+05

MSSignal(cps) - 0.0E+00 -4.0E+05 ICP 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

Figure 3.2.2-1 A-D. Four week arsenic III stability analyzed by HPLC-ICP-MS. Arsenic concentrations (ppb) were: 1 (○) - primary y-axis, 50 (□) and 150 (Δ) - secondary y-axis. Arsenic III samples preservation schemes were: A. Control (no preservative), B. Ascorbic Acid (A. A.), C. 1.25E-03 M EDTA and D. 0.05 % HCl. Immediately apparent was the large difference in ICP-MS signal counts detected for 1 ppb As III preserved with HCl, seen in Figure 3.2.2-1 D, compared to the corresponding signals in unpreserved, A. A. and

EDTA preserved samples, seen in Figure 3.2.2-1 A, B and D, for As III at the same concentration. Table 3.2.2-

1 lists the counts detected at the ICP-MS for all of the As III samples, analyzed at 1, 50 and 150 ppb As III, over the four week analysis period at each of the preservation schemes analyzed corresponding to data shown in Figures 3.2.2-1 A-D.

100

Table 3.2.2-1. RP-HPLC-ICP-MS Analysis of As III stability in unpreserved, ascorbic acid, EDTA and HCl preserved samples over four weeks. ICP-MS As III Detection in counts per second (cps).

ICP-MS Counts Per Second (cps) Detected As Preservation 0 Hours 24 Hours 48 Hours 1 Week 2 Weeks 3 Weeks 4 Weeks III Unpreserved 9.22E+03 6.39E+03 4.35E+03 3.84E+03 3.21E+03 2.67E+03 2.61E+03

Ascorbic Acid 7.97E+03 6.01E+03 2.15E+03 0.00E+00 0.00E+00 -8.71E+02 -7.24E+02 EDTA 8.86E+03 9.24E+03 8.63E+03 9.58E+03 9.56E+03 9.18E+03 8.42E+03

1 ppb HCl 3.36E+03 1.46E+03 1.70E+03 1.19E+03 1.09E+03 1.69E+03 1.52E+03

Unpreserved 4.36E+05 4.22E+05 4.35E+05 3.77E+05 3.58E+05 3.59E+05 3.53E+05 Ascorbic Acid 4.09E+05 3.71E+05 3.42E+05 3.19E+05 2.78E+05 2.46E+05 2.25E+05 EDTA 4.30E+05 4.20E+05 4.20E+05 4.43E+05 4.42E+05 4.36E+05 4.34E+05 50 ppb 50 HCl 4.27E+05 4.145+05 4.11E+05 4.31E+05 4.14E+05 4.10E+05 4.04E+05

Unpreserved 1.32E+06 1.32E+06 1.35E+06 1.22E+06 1.21E+06 1.24E+06 1.23E+06 Ascorbic Acid 1.23E+06 1.18E+06 1.12E+06 1.14E+06 1.05E+06 9.68E+05 8.76E+05 EDTA 1.31E+06 1.29E+06 1.29E+06 1.35E+06 1.37E+06 1.34E+06 1.33E+06

150 ppb HCl 1.29E+06 1.27E+06 1.27E+06 1.32E+06 1.32E+06 1.30E+06 1.28E+06

Calculating the difference in cps, at “time zero”, between the preservation schemes at 1 ppb As III

showed the signal detected in the HCl sample was 64 % lower than the unpreserved sample, and 58 or 62 %

lower than either the A. A. or EDTA samples, respectively. The large difference in ICP-MS signal at 1 ppb As

III in HCl was not corroborated in other speciation research to the extent seen here potentially due to the

majority of speciation protocols using HCl preservation recommending acidification to pH ≤ 2.0 [65b, 67a].

However, multiple researchers have shown rapid oxidation of As III to As V occurs during speciation analysis.

Analysis of As III/V in groundwater and acid mine drainage performed by Bednar et al. [56] showed rapid

oxidation of As III to As V occurred when samples containing 10 ppb As III were acidified to 0.06 M HCl and

analyzed by HPLC-ICP-MS. Bednar hypothesized the rapid oxidation resulted from trace metal contaminants

in the hydrochloric acid utilized during preservation. However McCleskey et al. [65b], during an extensive

arsenic speciation study, suggested both HCl and/or bacteria contamination of the HPLC mobile phase could

be potential culprits for As III oxidation detected by Bednar. McCleskeys’ [65b] results supported the

prevailing theme of HCl preservation protocols, for inorganic arsenic speciation, showing samples containing

101

As III acidified with HCl to pH ≤ 2.0 were stable up to 45 days. Similarly, Almassalkhi [57], while optimizing the As III/V method that was varied to perform the speciation analysis described herein, concluded HCl prevented As III oxidation better than EDTA in samples analyzed over a period of one week. However if the apparent oxidation detected here, at 1 ppb As III in HCl, and possibly in Bednars’ work, resulted from bacteria and/or trace metal contaminated HCl, similar oxidation should have been detected in all samples analyzed including those preserved in HCl. Analysis of the raw data in Table 3.2-1, for the 50 and 150 ppb As III samples preserved in HCl along with those for the other preservations schemes, at 1, 50 and 150 ppb As III, show no similar drop in As III was detected in any of the samples analyzed. On the contrary, similar comparison of

“time zero” ICP-MS signal counts in Table 3.2-1 show the 50 ppb As III sample in HCl differed by only 2, 4 and 0.8 % to those detected in the unpreserved, A. A. and EDTA preserved samples, respectively. At 150 ppb

As III, the difference calculated between the results detected in HCl compared to the unpreserved, A. A. and

EDTA preserved samples also show minimal differences, only 2.2, 5 and 1.4 %, respectively. The changing

As III concentration, over time, for each of the preservation schemes analyzed at 1, 50 and 150 ppb As III are displayed in Figure 3.2.2-2 A-C.

102

1 ppb Arsenic III over time 1.2E+04 1.0E+04 8.0E+03 6.0E+03 4.0E+03

2.0E+03 MSSignal(cps) - 0.0E+00

-2.0E+03 ICP 0 100 200 300 400 500 600 700 Time (hours)

(NH4)H2PO4 A. A. EDTA HCl

50 ppb Arsenic III stability 5.0E+05 4.0E+05 3.0E+05 2.0E+05

MSSignal(cps) 1.0E+05 -

0.0E+00 ICP 0 100 200 300 400 500 600 700 Time (hours)

(NH4)H2PO4 A. A. EDTA HCl

150 ppb Arsenic III stability 1.5E+06

1.0E+06

5.0E+05

MSSignal(cps) - 0.0E+00 ICP 0 100 200 300 400 500 600 700 Time (hours)

(NH4)H2PO4 A. A. EDTA HCl

Figure 3.2.2-2 A-C. Analysis of the stability of As III in unpreserved, ascorbic acid, EDTA and HCl preserved samples over four weeks. (◊) No preservation; (□) ascorbic acid; (∆) EDTA; (○) HCl.

103

The ICP-MS results in Figure 3.2.2-2 A-C highlight both the “time zero” difference, described above, in ICP-MS detection observed at 1 ppb As III in HCl compared to the other preservation schemes as well as the absence of any similar signal differences at 50 and 150 ppb As III in HCl. In addition, analysis of the “time zero” results in Table 3.2.2-1 and Figure 3.2.2-1, at 1, 50 and 150 ppb As III across the other preservation schemes (unpreserved, A. A. and EDTA), show minimal differences in instrument response occurred. The greatest difference in ICP-MS signal, 15.6 %, was calculated when the unpreserved 1 ppb As III sample counts were compared to those detected in the A. A. preserved sample. These results suggest the oxidation or instability detected in As III at 1 ppb preserved with dilute HCl may not have been the result of bacteria and/or trace metal contamination as suggested in McCleskey’s research. Instead, the results shown indicate the stability of dissolved As III, or its potential to oxidize, was somehow related to its initial concentration in solution, with lower concentrations more susceptible to oxidation than higher concentrations. The percent of apparent As III oxidation calculated over time in the 1, 50 and 150 ppb As III samples at each of the preservation schemes, shown in Table 3.2.2-2, confirms that the amount of As III oxidation detected decreased with increasing concentrations of As III in solution in all of the samples analyzed.

Table 3.2.2-2. Percent of As III oxidation detected over time during four week analysis in unpreserved, ascorbic acid, EDTA and HCl preserved samples. ICP-MS As III/V Detection in counts per second (cps).

As III Preservation 24 Hours 48 Hours 1 Week 2 Weeks 3 Weeks 4 Weeks Concentration Unpreserved 31 53 58 65 71 72

A. A. 25 73 100 100 111 109 EDTA -4 3 -8 -8 -4 5

1 ppb HCl 56 49 65 67 50 55

Unpreserved 3 0 13 18 18 19 A. A. 9 16 22 32 40 45 EDTA 2 2 -3 -3 -1 -1 50 ppb 50 HCl 3 4 -1 3 4 5

Unpreserved 0 -2 8 8 6 7

ppb A. A. 4 9 7 15 21 29

EDTA 2 1 -3 -4 -2 -1

150 HCl 2 2 -2 -2 -0 1

104

Hall et al. noticed similar dependence of As III oxidation on the actual concentration in solution during a recent stability study performed on samples of As III and As V in water [68]. Hall et al. suggested at higher As III species concentration a kinetic equilibrium was reached that reduced the cycling between As III and V species in solution reducing levels of oxidation/reduction.

In an attempt to verify the potential relationship between the concentration of As III and its oxidation in solution, results from the 1, 50 and 150 ppb As III analyses over four weeks, at each preservation scheme, were plotted and compared (e.g. 1-50 ppb As III; 1-150 ppb As III and 50-150 ppb As III) as seen in Figure

3.2.2-3 A-C.

Comparing 1 and 50 ppb As III LC- ICP-MS signal

1.2E+04 5.0E+05 ICP 1.0E+04

4.0E+05 - 8.0E+03 (cps) Signal MS 6.0E+03 3.0E+05

4.0E+03 2.0E+05 MSSignal(cps)

- 2.0E+03 1.0E+05

ICP 0.0E+00 -2.0E+03 0.0E+00 0 100 200 300 400 500 600 700 Time (hours)

(NH4)H2PO4 A. A. EDTA HCl (NH4)H2PO4 A. A. EDTA HCl

Figure 3.2.2-3 A-C. Comparing the % of oxidation detected in 1, 50 and 150 ppb As III samples over four weeks. 1 ppb As III - primary y-axis; 50 and 150 ppb As III - secondary y-axis. A. 1 vs 50 ppb As III: 1 ppb (♦) No preservation; (■) ascorbic acid; (▲) EDTA; (●) HCl; 50 ppb (◊) No preservation; (□) ascorbic acid; (∆) EDTA; (○) HCl; B. 1 vs 150 ppb As III: 1 ppb (♦) No preservation; (■) ascorbic acid; (▲) EDTA; (●) HCl; 150 ppb (◊) No preservation; (□) ascorbic acid; (∆) EDTA; (○) HCl and C. 50 vs 150 ppb: 50 ppb (♦) No preservation; (■) ascorbic acid; (▲) EDTA; (●) HCl; 150 ppb (◊) No preservation; (□) ascorbic acid; (∆) EDTA; (○) HCl.

105

Comparing 1 and 150 ppb As III LC- ICP-MS signal

1.2E+04 1.5E+06 ICP

1.0E+04 1.3E+06 - 8.0E+03 (cps) Signal MS 1.0E+06 6.0E+03 7.5E+05 4.0E+03

5.0E+05 MSSignal(cps)

- 2.0E+03

0.0E+00 2.5E+05 ICP -2.0E+03 0.0E+00 0 100 200 300 400 500 600 700 Time (hours)

(NH4)H2PO4 A. A. EDTA HCl (NH4)H2PO4 A. A. EDTA HCl

Comparing 50 and 150 ppb As III LC- ICP-MS signal

5.0E+05 1.5E+06 ICP

4.0E+05 1.3E+06 - 1.0E+06 (cps) Signal MS 3.0E+05 7.5E+05 2.0E+05

MSSignal(cps) 5.0E+05 -

1.0E+05 2.5E+05 ICP 0.0E+00 0.0E+00 0 100 200 300 400 500 600 700 Time (hours)

(NH4)H2PO4 A. A. EDTA HCl (NH4)H2PO4 A. A. EDTA HCl

The results seen in Figure 3.2.2-3 A-C show the 1 ppb As III results on the primary y-axis with the 50 and 150 ppb results appearing on the secondary y-axis. With the exception of the EDTA preserved As III

106 samples, the results shown comparing the 1 ppb As III results to the 50 ppb results (Figure 3.2.2-3 A) and to the 150 ppb As III results (Figure 3.2.2-3 B), show no correlation between the stability of 1 ppb As III and the stability of either 50 or 150 ppb As III over time. However, subsequent comparison of the 50 and 150 ppb As

III results seen in Figure 3.2.2-3 C shows close grouping of points with the data points for each preservation scheme falling either on top or close to the corresponding point at the comparable concentration. The high degree of correlation between the EDTA samples in Figure 3.2.2-3 A and B (at 1, 50 and 150 ppb As III) and the unpreserved, EDTA and HCl preserved samples in Figure 3.2.2-3 C (at 50 and 150 ppb As III) are difficult to differentiate, due to overlap in the data points, and were therefore separated and shown in Figure 3.2.2-4 A-

E for easier analysis.

Comparing 50 and 150 ppb As III LC- ICP-MS signal

6.0E+05 1.5E+06 ICP

4.0E+05 1.0E+06

- MS (cps) Signal MS 2.0E+05 5.0E+05 0.0E+00 0.0E+00

0 100 200 300 400 500 600 700 MSSignal(cps)

- Time (hours)

ICP (NH4)H2PO4 (NH4)H2PO4

Figure 3.2.2-4 A-E. Individual comparison of the % oxidation detected in 1, 50 and 150 ppb As III samples over four weeks. 1 ppb As III - primary y-axis; 50 and 150 ppb As III - secondary y-axis. A. 50 vs 150 ppb: 50 ppb (♦) No preservation; 150 ppb (◊) No preservation; B. 1 vs 50 ppb As III: 1 ppb (▲) EDTA; 50 ppb (∆) EDTA; C. 1 vs 150 ppb As III: 1 ppb (▲) EDTA; 150 ppb (∆) EDTA; D. 50 vs 150 ppb: 50 ppb (▲) EDTA; 150 ppb ; (∆) EDTA and E. 50 vs 150 ppb: 50 ppb (●) HCl; 150 ppb (○) HCl.

107

Comparing 1 and 50 ppb As III LC- ICP-MS signal

1.5E+04 6.0E+05 ICP

1.0E+04 4.0E+05

- MS (cps) Signal MS 5.0E+03 2.0E+05 0.0E+00 0.0E+00

0 100 200 300 400 500 600 700 MSSignal(cps)

- Time (hours) ICP EDTA EDTA

Comparing 1 and 150 ppb As III LC- ICP-MS signal

1.5E+04 1.5E+06 ICP

1.0E+04 1.0E+06 - 5.0E+03 5.0E+05 (cps) Signal MS 0.0E+00 0.0E+00 0 100 200 300 400 500 600 700

MSSignal(cps) Time (hours) -

ICP EDTA EDTA

Comparing 50 and 150 ppb As III LC- ICP-MS signal

6.0E+05 1.5E+06 ICP

4.0E+05 1.0E+06 - MS (cps) Signal MS 2.0E+05 5.0E+05 0.0E+00 0.0E+00

0 100 200 300 400 500 600 700 MSSignal(cps)

- Time (hours) ICP EDTA EDTA

Figure 3.2.2-4 A-E. Individual comparison of the % oxidation detected in 1, 50 and 150 ppb As III samples over four weeks. 1 ppb As III - primary y-axis; 50 and 150 ppb As III - secondary y-axis. A. 1 vs 50 ppb As III: 1 ppb (▲) EDTA; 50 ppb (∆) EDTA; B. 1 vs 150 ppb As III: 1 ppb (▲) EDTA; 150 ppb (∆) EDTA; C. 50 vs 150 ppb: 50 ppb (♦) No preservation; 150 ppb (◊) No preservation; D. 50 vs 150 ppb: 50 ppb (▲) EDTA; 150 ppb ; (∆) EDTA and E. 50 vs 150 ppb: 50 ppb (●) HCl; 150 ppb (○) HCl.

108

Comparing 50 and 150 ppb As III LC- ICP-MS signal

6.0E+05 1.5E+06 ICP -

4.0E+05 1.0E+06 (cps) Signal MS 2.0E+05 5.0E+05

0.0E+00 0.0E+00 MSSignal(cps)

- 0 100 200 300 400 500 600 700

Time (hours) ICP

HCl HCl

Figure 3.2.2-4 A-E. Individual comparison of the % oxidation detected in 1, 50 and 150 ppb As III samples over four weeks. 1 ppb As III - primary y-axis; 50 and 150 ppb As III - secondary y-axis. A. 1 vs 50 ppb As III: 1 ppb (▲) EDTA; 50 ppb (∆) EDTA; B. 1 vs 150 ppb As III: 1 ppb (▲) EDTA; 150 ppb (∆) EDTA; C. 50 vs 150 ppb: 50 ppb (♦) No preservation; 150 ppb (◊) No preservation; D. 50 vs 150 ppb: 50 ppb (▲) EDTA; 150 ppb ; (∆) EDTA and E. 50 vs 150 ppb: 50 ppb (●) HCl; 150 ppb (○) HCl.

Based on final results from the four week analysis of As III/V stability, shown in Table 3.2.2-1, Figures

3.2.2-3 A-C and 3.2.2-4 A-E, EDTA provided the greatest, with A. A. providing the least, anti-oxidant protection of As III in the samples analyzed over the course of this work, especially at a concentration of 1 ppb

As III/V. The high amount of As III oxidation detected in A. A. preserved samples (100 % of 1 ppb As III was oxidized after 1 week) may be due to increased levels of bacterial growth stimulated by A. A. in the samples

[69]. At higher As III concentrations (50 and 150 ppb), despite the limited overall oxidation that was detected, for all the preservation schemes analyzed, the samples preserved with EDTA continued to show the least amount of variation to the initial concentrations of As III detected at the ICP-MS. However, it is important to note, at

50 and 150 ppb As III, both the unpreserved and the dilute HCl preserved arsenic samples show percent oxidation results similar to those detected in the samples preserved with EDTA. This further supported the hypothesis of a concentration-dependent stabilization effect limiting the oxidation of As III when higher concentrations were found in solution.

Over the course of the stability analysis, slight increases in the amount of As III detected were also noticed, primarily in EDTA preserved samples, and can be seen in the results in Table 3.2-1 with the potential

109 reduction of As V to As III occurring in solution shown by the negative values recorded for percent oxidation.

This apparent change in arsenic oxidation state may have been “falsely” reported due to changes in instrument response as a result of daily tuning of the ICP-MS and were within the acceptable ± 10 % recovery allowed for approved analytical methods. However, by the end of the four week analysis period, the majority of samples initially showing increased As III concentrations stabilized and show limited or no reduction of As V to As III.

To detect any discrepancy in the concentration of As III and V detected, over time, the difference between the two species was calculated for the 1, 50 and 150 ppb As III/V samples, for each preservation scheme analyzed, and the results are shown in Table 3.2.2-3.

Table 3.2.2-3. Change in As V concentration detected over time during four week analysis in unpreserved, ascorbic acid,

EDTA and HCl preserved samples. ICP-MS As III/V Detection in counts per second (cps).

Difference Between the Concentration of Arsenic III + Arsenic V Over Time (%) As V Preservation 24 Hours 48 Hours 1 Week 2 Weeks 3 Weeks 4 Weeks Concentration Unpreserved 113 100 106 110 99 108 A. A. 102 104 99 100 99 103 EDTA 102 103 97 98 100 102

1 ppb HCl 105 104 108 113 96 100

Unpreserved 102 107 103 102 108 102 A. A. 101 107 105 102 105 101

ppb EDTA 105 120 106 101 99 105

50 HCl 100 106 104 105 103 100

Unpreserved 101 106 106 104 104 101 A. A. 120 111 119 113 113 120 EDTA 100 105 103 101 101 100

150 ppb HCl 102 105 104 102 102 102

With the exception of the 1 ppb 24 hour, unpreserved, 2 week, HCl preserved; the 50 ppb 1 week EDTA preserved and all 150 ppb A. A. preserved samples correlation of the oxidation/reduction of As III/V in solution was within ± 10 % of the original concentration detected for each preservation scheme analyzed. Blanks, for each of the preservation schemes, were also prepared and analyzed (results not shown) as samples during each run and their concentration of either As III or V were manually deducted from the appropriate set of

110 experimental results obtained during analysis. These manual “blank” adjustments to As III/V concentration may also have contributed to discrepancy detected between the initial and final concentrations values reported for the As III/V stability samples.

Finally, while the results reported herein indicate EDTA provided the best anti-oxidant protection of

As III in solution, multiple factors including, but not limited to, pH, interfering analytes and other contaminants in solution should be considered when choosing an anti-oxidant for any type of metal/metalloid speciation analysis. The anti-oxidant schemes analyzed in this work were chosen due to initial optimization results for the

Reversed Phase method (0.75 mL·min-1 flow rate) which showed the maximum As III/V peak separation occurred when samples were maintained between pH 5.0 and 7.0. Acidification of samples below pH 3.0 was found to result in either peaks that eluted simultaneously, as seen in Figure 3.2.2-5, or decreased separation [70] between As III/V peaks that could not be integrated (results not shown).

50 ppb As III/V Cal Std, pH 3.0 1.6E+05 1.4E+05 1.2E+05 1.0E+05 8.0E+04

6.0E+04 MSSignal(cps) - 4.0E+04

ICP 2.0E+04 0.0E+00 0 1 2 3 4 5 Elution Time (min)

Figure 3.2.2-5. RP-HPLC-ICP-MS As III/V Speciation Method Optimization. Simultaneous elution of As III/V peaks using 50 ppb As III/V calibration standard acidified to pH 3.0 eluted at 0.75 mL∙min-1.

The simultaneous elution of As III and As V shown in Figure 3.2.2-5 may have been caused by a combination of factors including low sample pH and increased ionic strength resulting in a reducing environment where arsenic in the +5 oxidation state was converted to a +3 oxidation. Research performed by

111

Hall et al. [68] showed 0.1 and 0.4 % HCl matrices spiked to concentrations between 2-10 ppb As III/V reduced

As V to As III in as little as 2 hours when analyzed by HPLC-ICP-MS. However, an alternate hypothesis to the lack of separation of As III/V peaks may also be the low sample pH led to interaction between negatively charged As V and H+ ions instead of As V and TBAH in the mobile phase effectively interfering with, reducing and/or eliminating separation of As III and V on the column prior to detection.

However, as previously stated, the majority of arsenic speciation protocols indicated acidification of samples to pH ≤ 2.0 [65b, 67a, 71] was necessary to stabilize arsenic species in solution. In his previously mentioned work, McCleskey further stressed acidification of speciation samples with weak acid solutions offered no As III anti-oxidant protection, reiterating HCl preservation to pH ≤ 2.0 provided the most reliable speciation results. However, the large number of methods available for the preservation and speciation of arsenic in solution, combined with the high degree of variability in results obtained, including those shown here, show there is no simple answer nor any single method that currently applies to every potential scenario when applying these methods [72].

3.2.3 Effect of DOM on As III/V Detection

As previously stated, several contaminants in water affect the absorption and detection of arsenic species in solution including DOM [22]. Dissolved humic substances or DOM formed heterogeneous, amphoteric molecules believed to contain various functional groups, including hydroxyl, ketone, nitroso and carboxylic [38, 40, 66], several of which could potentially interact with As III and/or V in water depending upon the surrounding redox potential and pH. These features of DOM molecules may be responsible for their proposed reaction with and potential to mobilize metals and metalloids, including arsenic, in water [22, 37a,

73].

To analyze the effect, if any, of concentrated DOM (cDOM) [73] on As III/V speciation and detection a concurrent four week analysis was performed spiking samples of As III/V, prepared as described in the As

III/V stability analysis, with 1 ppm cDOM. The results reported in this work, from the analysis of the

112 arsenic/cDOM samples, emphasize any deviation(s) from the As III/V stability results described previously as the two sample “sets” were compared and contrasted. The As III/cDOM results obtained, over the four week analysis period, in unpreserved, A. A., EDTA and HCl preserved samples of As III/V at concentrations of 1, 50 and 150 ppb spiked to a final concentration of 1 ppm cDOM are displayed in Figure 3.2.3-1 A-D.

Arsenic III - Unpreserved with cDOM

1.0E+04 1.5E+06 ICP

8.0E+03

- MS (cps) Signal MS 6.0E+03 1.0E+06 4.0E+03 5.0E+05

2.0E+03 MSSignal(cps) - 0.0E+00 0.0E+00

ICP 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

Arsenic III - Ascorbic Acid with cDOM

6.0E+03 1.5E+06 ICP

4.0E+03 - 1.0E+06 (cps) Signal MS 2.0E+03 0.0E+00 5.0E+05

-2.0E+03 MSSignal(cps) - -4.0E+03 0.0E+00

ICP 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

Figure 3.2.3-1 A-D. Four week analysis of arsenic III spiked with concentrated dissolved organic matter (cDOM) analyzed by RP-HPLC-ICP-MS. Concentrated dissolved organic matter concentration was 1 ppm. Arsenic concentrations (ppb) were: 1 (○) - primary y-axis, 50 (□) and 150 (Δ) - secondary y-axis. Arsenic III/cDOM samples were: A. Control (no preservative), B. 3:1 As III:Ascorbic Acid (A. A.), C. 1.25E-03 M EDTA and D. 0.05 % HCl.

113

Arsenic III - EDTA with cDOM

1.2E+04 1.5E+06 ICP

1.0E+04 - 8.0E+03 1.0E+06 (cps) Signal MS 6.0E+03 4.0E+03 5.0E+05

MSSignal(cps) 2.0E+03 - 0.0E+00 0.0E+00

ICP 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

Arsenic III - HCl with cDOM

1.0E+04 1.5E+06 ICP

8.0E+03 - 1.0E+06 (cps) Signal MS 6.0E+03 4.0E+03 5.0E+05

2.0E+03 MSSignal(cps) - 0.0E+00 0.0E+00

ICP 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

Several interesting variations were detected when concentrated dissolved organic matter (cDOM) was added to samples of As III/V at the different concentrations analyzed. The first variation can be seen in Figure

3.2.3-1 A, B and D, which, similar to the As III/V stability samples previously described, shows an immediate decrease in the 1 ppb As III “time zero” ICP-MS signal detected. However, with cDOM added to the arsenic samples, the decrease in ICP-MS signal was seen in all sample sets, at 1 ppb As III/V, except those preserved with HCl when compared to samples sets without cDOM at the same arsenic concentration shown in Figure

3.2.3-1 A, B and D. The second variation seen was, at the same 1 ppb As III concentration, the HCl preserved

114 sample spiked with cDOM, at “time zero” instead show an increase in the signal counts detected at the ICP-MS when compared to the results seen without cDOM. The counts detected at 1 ppb As III, spiked with cDOM are

28, 41 and 15 % lower in the unpreserved, A. A. and EDTA preserved samples, respectively, and 53 % higher in the HCl preserved sample than they were in samples without cDOM added. The variability of the decrease in signal in three out of four sample sets, however, only seen at the lowest arsenic concentration in addition to the contrasting increase seen for HCl suggested there may be multiple causes for the discrepancy noticed in

Figure 3.2.3-1.

Bacterial oxidation/reduction [8, 33a, 56, 65b, 74] of As III/V in water is a well-known phenomenon and could also be a plausible explanation for a reduction in ICP-MS signal especially in samples with DOM added. However the absence of a similar reduction in the ICP-MS signal at higher arsenic concentrations, combined with the increased signal detected for HCl preserved samples suggested bacterial interference may not be the cause of the drop in signal seen here. In addition, not only wasn’t there a corresponding increase in

As V concentrations seen, there was instead an accompanying decrease in those concentrations as well which further indicated bacterial interference was not responsible. This did however lead to the third interesting variation noted as further analysis of the chromatograms obtained over the course of the four week analysis period, particularly the cDOM blanks with ascorbic acid added after the second week of analysis, show the appearance of a third peak at ~ 3.35 minutes (201 seconds) as shown in one example seen Figure 3.2.3-2.

115

As III/V Ascorbic acid/cDOM blank 4.0E+02

3.0E+02

2.0E+02 MSSignal(cps)

- 1.0E+02

ICP 0.0E+00 0 1 2 3 4 5 6 Elution Time (min)

Figure 3.2.3-2. Arsenic III/V Ascorbic acid/cDOM blank – showing potential organoarsenic peak elution. Ascorbic acid/cDOM blank was prepared in 2.5 mM (NH4)H2PO4 + 5 mM TBAH mobile phase, pH 6.0 at 1 ppb ascorbic acid spiked with 1 ppm cDOM analyzed by RP-HPLC-ICP-MS. Chromatogram was obtained during week 3 of the As III/V-cDOM analysis.

The chromatogram in Figure 3.2.3-2, showing the ascorbic acid/cDOM blank analyzed during the third week of the four week analysis period, was chosen as an example as it clearly shows the elution of a third arsenic peak between the As III/V peaks which suggests, at some point in solution, conversion of inorganic As

III and/or V to an organo-arsenic species occurred. This may have been the result of increased bacterial activity within the samples due to the inclusion of both ascorbic acid and cDOM both known to promote bacterial activity in solution. It is important to note that, with the exception of a few of the A. A. preserved blanks

(observed towards the end of the four week analysis period), the majority of the chromatograms containing cDOM show little to no indication of any additional arsenic peaks eluting between the As III/V peak positions.

This being said, conversion of inorganic As III/V to arsenic species outside the 6 minute analysis window could also be a viable explanation for a decrease in ICP-MS signal detection. However, while this could potentially explain the simultaneous decrease seen in As V concentrations it would not explain the increase in ICP-MS signal detected for As III/V preserved in HCl. In addition, ideally, conversion to a different arsenic form, even one outside the six minute analysis time should be detectable. Theoretically, the elution and detection at the

ICP-MS of As75 beyond the six minute “monitored” window, should result in random and/or potentially

116 overlapping peaks which, in the case of overlapping peaks, would result in increased concentrations of As III.

This further indicated bacterial oxidation/reduction was not the cause of the decrease/increase in ICP-MS signal.

However, the decrease could also indicate portions of the As III/V in solution may have formed a complex with cDOM in solution and/or possibly the addition of cDOM resulted in dissolution of a portion of the As III/V, which would be more apparent at low concentrations, in the samples before injection occurred at the HPLC. Prior to the injection cDOM onto the HPLC column, a UV-Vis diode array detector (DAD) was placed in line, after the column compartment, before the ICP-MS, to monitor/ensure cDOM injected at the

HPLC was not irreversibly adsorbed onto the column. Preliminary results (not shown), injecting cDOM with and without the column in place, indicated all cDOM that was injected at the HPLC passed through the column almost immediately with none of it retained on the HPLC column. In support of this hypothesis, results published by Redman et al. showed similar discrepancy in the concentration of As III and V in solution before and after various DOM samples were added [66] Redman et al. concluded, depending upon the source of DOM, rapid complexation of both As III and As V in solution could occur, inhibiting the detection of As III/V when analyzed by HPLC-ICP-MS [66] Initially findings reported by Redman et al. seemed to be the most plausible explanation to the results detected in the As III/V cDOM samples reported here, however, results from Redman et al. further concluded prolonged incubation of DOM with As III and V prior to analysis increased the level leeching arsenic species from solution [66] This was not the case in the samples analyzed over the course of this work. The samples analyzed here indicate when cDOM is spiked into samples of As III/V it appears to stabilize and or preserve the individual species in solution.

Disregarding the discrepancy originally noted upon comparing the As III/V samples with and without cDOM, with the exception of the A. A. preserved samples, the chromatograms previously described in Figure

3.2.3-1 A, B and D show all of the As III samples, regardless of concentration and/or preservation, remained stable over the four week analysis period when cDOM was added. Table 3.2.3-1 shows the percent oxidation of As III, over the four week analysis period, in unpreserved, A. A., EDTA and HCl preserved samples spiked with 1 ppm cDOM.

117

Table 3.2.3-1. Comparison of the ICP-MS signal response (cps) in As III samples spiked with cDOM to As III samples without cDOM over a four week analysis period in unpreserved, ascorbic acid, EDTA and HCl preserved samples. Results shown are calculated as percent difference.

As III Preservation 24 Hours 48 Hours 1 Week 2 Weeks 3 Weeks 4 Weeks Concentration NOM 10 12 -8 6 11 19

NOM/A.A. 33 70 133 124 155 123 NOM/EDTA -12 -8 -13 -33 -20 -12

1 ppb NOM/HCl -16 -5 3 -1 -1 -12

NOM 2 3 0 2 2 5 NOM/A.A. 10 17 27 48 65 77 NOM/EDTA 1 -1 -4 -2 -4 -1 50 ppb 50 NOM/HCl 2 2 -2 2 -1 2

NOM 3 2 -3 0 0 2 NOM/A.A. 6 8 12 27 37 49 NOM/EDTA 2 0 -5 0 -3 -1

150 ppb NOM/HCl 2 1 -2 1 -2 0

The percent oxidation results in Table 3.2.3-1 show that after four weeks, with the exception of the 1 and 150 ppb As III/V samples preserved in A. A., the samples spiked with cDOM show less deviation from the original concentration of As III detected in solution. However, the 1 ppb As III/V samples preserved in EDTA and HCl also show an ~ 12 % increase in As III compared to its original concentration by the end of the four week analysis period which could have been the result of trace levels of arsenic in the samples of cDOM.

However subsequent Inductively Coupled Plasma (ICP) Atomic Emission Spectroscopy (AES) analysis of the cDOM showed no arsenic detected in the sample used during analysis.

The preservation of As III/V in the presence of cDOM in solution could potentially be due to cDOM binding arsenic through some type of electrostatic interaction that effectively formed a “protecting” group around As III/V molecules in solution. However, it may also be the result of the much larger concentration of cDOM in solution acted as an anti-oxidant of sorts, as bacteria and other contaminants in solution that would normally be responsible for oxidizing or reducing inorganic arsenic species in solution, interacted with and oxidized/reduced cDOM instead. While it is known that DOM in natural waters can result in the dissolution and/or mobilization of species of arsenic and inhibit and/or compete with arsenic for adherence onto various

118 solid supports including hematite and various clays [2, 8, 66, 74a]. These results suggest that once mobilized, surrounding DOM in solution may also maintain and/or preserve the different species of arsenic detected in natural waters. This could potentially lead to elevated levels of arsenic, outside the EPA allowable limits, in certain areas and during certain seasons where increased levels of DOM are found.

Multiple explanations have been suggested to explain the ability of DOM, including FAs and HAs, to associate with metals and metalloids like arsenic with most suggesting the extensive array of potential

- - functional groups (COOH, SH , OH , NH2 and C6H5) believed to be present within individual samples of DOM

[2, 8, 66, 74a, 75]. However the inclusion of metal mediated functional groups (Fe, Al, Mn, etc.) within DOM active sites has also been suggested as an excuse for DOM binding and interacting with inorganic arsenic species in water [8, 76]. Liu et al. demonstrated the potential for DOM to bind As III was wholly pH dependent, showing only small amounts were bound or formed associations in solution at pH < 5 with progressively larger amounts being bound with increasing pH until pH > 9 when the amounts of As III bound to DOM were significantly higher [74a]. Liu et al. proposed the significantly higher levels of As III complexed to DOM was

- due to As III forming both neutral As(OH)3 and negatively charged H2AsO3 species in solution at pH > 9 [74a].

A review published by Wang et al. reported species of arsenic preferentially bound FAs compared to HAs due to FAs lower molecular weights, increased acidity and functional groups capable of binding arsenic in solution

[74b, 74c]. They went on to suggest, in solution, NOM complexation occurred preferentially with As III compared to As V [74b, 74c]. Buschmann et al. published similar findings in a study performed to determine conditional distribution coefficients for the binding of Suwanee River Humic Acid (SRHA) to As III and V, showing HA bound both As III and V during dialysis experiments using SRHA in the presence of As III/V, analyzed by atomic fluorescence spectrometry (AFS) and ICP-MS [2]. Buschmann also demonstrated the potential for HA binding to As V was 6-10 times stronger than its potential to bind As III [2]. Redman et al. also suggested interaction between species of arsenic and DOM in solution could be detected using HPLC as arsenic complexed to organic matter would alter its elution pattern potentially resulting in a change in elution

119 time. The chromatograms seen in Figure 3.2.3-3 may show this phenomenon occurring in the As III/V samples spiked with cDOM.

1 ppb As III/V 50 ppb As III/V 5.0E+02 2.0E+04 4.0E+02 1.5E+04 3.0E+02 1.0E+04 2.0E+02 5.0E+03 1.0E+02 0.0E+00 0.0E+00

-1.0E+02 -5.0E+03 As (III/V) (III/V) As Concentration(cps) Elution Time (min) (III/V) As Concentration(cps) Elution Time (min)

150 ppb As III/V 6.0E+04

4.0E+04

2.0E+04

0.0E+00

-2.0E+04

Elution Time (min) As (III/V) (III/V) As Concentration(cps)

Figure 3.2.3-3 A - C. Four week analysis of arsenic III/V with and without cDOM analyzed by RP-HPLC-ICP-MS. Arsenic samples were prepared in HPLC mobile phase of 2.5 mM (NH4)H2PO4 + 5 mM TBAH at pH 6. Concentrated dissolved organic matter concentration was 1 ppm. Samples: arsenic III/V - ( ); arsenic III/V + cDOM - ( ).

While several of the chromatograms obtained over the course of the four week analysis of As III/V spiked with cDOM show a small degree of change in the retention time for As V, the chromatograms seen in

Figure 3.2.3-3, showing unpreserved As III/V samples with cDOM on the second day of the four week period, showed the greatest change in retention time the two samples sets. However, despite appearing pronounced, the difference between the samples with cDOM and without was only ~ 30 seconds and was therefore more likely due to limited equilibration time of the column prior to the start of analysis effecting the elution of As V

120 in solution. However, the idea of detecting DOM complexation via HPLC was intriguing therefore, in an attempt to verify if interaction between arsenic and molecules known to bind either As III or V would alter the

HPLC retention times, samples of As III/V at 1 ppb were prepared with 1 ppm cystine, believed to preferentially bind As III due to its di-sulfide group, and analyzed over a period of four weeks. Comparison of the results, however showed no change to the retention times of either As III or As V in any of the samples analyzed over the four weeks. However, while the results shown here display a slight change in the retention of As V in solutions containing cDOM and the analysis of As III/V in cystine showed no apparent binding of As III or V in solution multiple different variables and additional analyses would need to be performed to determine if

DOM complexation could be detected using a RP method of HPLC.

In conclusion, while the addition of cDOM in samples of As III/V initially resulted in decreased ICP-

MS detection in unpreserved, A. A. and EDTA preserved samples, overall it had a stabilizing effect with the majority of samples analyzed maintaining their original species concentration than those samples analyzed without cDOM. In addition, inclusion of cDOM in As III/V samples containing HCl at 1 ppb As III/V resulted in an increase in ICP-MS detection, a 53 % increase compared to the analysis results reported in samples that did not contain cDOM. The stabilization effect that cDOM appears to have on inorganic species of arsenic in solution may be of potential concern in areas where arsenic levels are close to the EPA regulated maximum contaminant level and/or where there is greater probability of DOM. In these situations, the presence of DOM may result in levels of arsenic outside the EPA mandated limits which could result in the imposition of fines and/or time consuming and costly methods to return a water source to allowable levels. The reason for the variation in ICP-MS response to As III/V samples with cDOM compared to those without remains unclear however cDOM in solution, especially at higher concentrations may act as an anti-oxidant and/or protect inorganic species of arsenic from contaminants that would otherwise result in its oxidation and/or reduction in solution. Multiple mechanisms for DOM interaction and/or binding have been proposed, including metal mediated binding and/or binding via the functional groups believed to be present within molecules of DOM.

121

3.2.4 Arsenic III/V Detection in Water

In an effort to validate the proposed As III/V method it was applied to the speciation of arsenic in drinking and source waters. The water samples analyzed were submitted by various researchers from multiple different areas both inside and outside the state of Ohio. Figure 3.2.4-1 A-C shows the HPLC-ICP-MS results obtained applying the proposed As III/V speciation method to the analysis of samples of drinking and/or source water.

Arsenic III/V Speciation Analysis of Drinking Water 2.5E+04

2.0E+04

1.5E+04

1.0E+04 MSSignal(cps)

- 5.0E+03 ICP 0.0E+00 0 1 2 3 4 5 6 Elution Time (min)

137521 137521_dup

Figure 3.2.4-1 A-C. RP-HPLC-ICP-MS separation, detection and quantification of As III/V in in drinking water. Chromatograms displayed show the As III/V analysis of “unknown” samples of drinking water submitted for As III/V detection over the course of several months.

122

Arsenic III/V Speciation Analysis of Drinking Water 7.0E+04 6.0E+04 5.0E+04 4.0E+04

3.0E+04 MSSignal(cps) - 2.0E+04

ICP 1.0E+04 0.0E+00 0 1 2 3 4 5 6 Time (min)

Arsenic III/V Speciation Analysis of Drinking Water 3.0E+04 2.5E+04 2.0E+04 1.5E+04

1.0E+04

MSSignal(cps) -

5.0E+03 ICP 0.0E+00 0 1 2 3 4 5 6 Elution Time (min)

140104 140104_dup

The samples seen in Figure 3.2.4-1 A-C were chosen, at random to highlight the ability of the proposed method to both separate and quantify As III and V in drinking water matrices. The chromatograms seen in

Figure 3.2.4-1 A and B show the analysis of samples corresponding to sample ID’s 137521 and 137521_dup

(Figure 3.2.4-1 A) and samples ID 137670 (Figure 3.2.4-1 B) also shown in Table 3.2.4-1, below. The chromatogram seen in Figure 3.2.4-1 C shows a sample and its duplicate analyzed several months prior to the

123 analysis seen in Figure 3.2.4-1 A. These chromatograms indicate the proposed method was successful in separating and detecting As III/V in natural water sources. In addition, the chromatograms displayed indicate both reproducibility and limited variation, in the detected retention times for As III/V over the course of several months, was detected when the proposed method was applied to the analysis of As III/V in water.

Table 3.2.4-1 shows various As III/V results obtained on multiple different analysis days.

Table 3.2.4-1. Method validation showing the As III/V method applied to drinking and source water.

Sample 75 As(III) 75 As(V) Sample Name Conc. [ μg/l ] Count (cps) Conc. [ μg/l ] Count (cps) 137493 35.19 6.31E+05 137493_dup 35.15 6.30E+05 137494 0.00 0.00 81.23 1.45E+06 137494_spk 10.38 1.74E+05 92.14 1.65E+06 137494_spk @ fil 10.23 1.71E+05 91.95 1.64E+06 137521 37.20 6.67E+05 137521_dup 37.51 6.72E+05 137522 0.00 0.00 84.75 1.51E+06 137522_spk 10.44 1.75E+05 95.75 1.71E+06 137522_spk @ fil 10.35 1.73E+05 95.62 1.71E+06 137671 0.087 1.47E+03 35.40 6.35E+05 137671_dup 35.51 6.37E+05 137672 0.00 0.00 83.28 1.49E+06 137672_spk 10.40 1.74E+05 92.58 1.65E+06 137672_spk @ fil 10.60 1.78E+05 92.39 1.65E+06 137708 35.76 6.41E+05 137708_dup 36.16 6.48E+05 137709 0.00 0.00 82.37 1.47E+06 137709_spk 10.44 1.75E+05 92.78 1.66E+06 137709_spk @ fil 10.68 1.79E+05 92.61 1.65E+06

% Recovery 75 % Recovery 75 As(III) As(V) Sample Name Conc. [ ug/l ] Count Conc. [ ug/l ] Count 137493_dup N/A N/A 100 100 137494_spk 104 104 115 113 137494_spk @ fil 99 99 100 100 137521_dup N/A N/A 99 99

124

137522_spk 105 105 116 114 137522_spk @ fil 99 99 100 100 137671_dup 100 100 100 100 137672_spk 105 105 98 96 137672_spk @ fil 102 102 100 100 137708_dup N/A N/A 99 99 137709_spk 105 105 110 108 137709_spk @ fil 102 102 100 100

The As III/V results seen in Table 3.2.4-1 show the proposed method of As III/V speciation was successfully applied to the quantification and detection of arsenic in drinking and source waters. The results seen here include both the actual, instrument reported values and several calculated recoveries for both

“duplicate” sample analysis performed and As III/V “spike” analysis in samples spiked with an additional 10 ppb As III/V prior to analysis. The results also show the recoveries for As III/V “spike after (@) filter sample analysis for samples spikes that were filtered, using 0.2 μm syringe filters, after they were spiked to detect potential arsenic retention on the filters used during sample preparation. An acceptable As III/V recovery is ±

10 % of the original value. Values obtained outside of the acceptable range of ± 10 % are marked in yellow, as shown in Table 3.2.4-1, and re-analyzed. Columns with missing values or those marked N/A, correspond to samples where the species concentration detected was zero.

The analysis results shown here indicate the proposed As III/V speciation method was successful in the separation, detection and quantification of both inorganic arsenic III and V in drinking and source waters. The chromatograms displayed show the separation and detection of both species in drinking water in samples analyzed months apart further indicating the method could be reliably used, with minimal variation seen in the separation of species on the RP-HPLC column. In addition, the results listed, representing multiple samples analyzed over multiple months, indicate minimal variation was detected in the analysis results obtained, including sample duplicates and sample spikes as indicated by the recoveries seen in Table 3.2.4-1.

125

References

1. Hessen, D. O.; Tranvik, L. J., Aquatic Humic Substances: Ecology and Biogeochemistry. Springer: 1998. 2. Buschmann, J.; Kappeler, A.; Lindauer, U.; Kistler, D.; Berg, M.; Sigg, L., Arsenite and Arsenate Binding to Dissolved Humic Acids: Influence of pH, Type of Humic Acid, and Aluminum. Environmental Science & Technology 2006, 40 (19), 6015-6020. 5. (a) Nriagu, J. O., Arsenic in the Environment: Human health and ecosystem effects. Wiley: 1994; (b) Agency, U. S. E. P. Drinking Water Contaminants. http://water.epa.gov/drink/contaminants/index.cfm. 8. Warwick, P.; Inam, E.; Evans, N., Arsenic’s Interaction with Humic Acid. Environmental Chemistry 2005, 2 (2), 119-124. 12. Chen, J., Spectroscopic characterization of the structural and functional properties of natural organic matter fractions. Chemosphere 2002, 48 (1), 59. 16. (a) Senesi, N.; Miano, T. M.; Provenzano, M. R.; Brunetti, G., Spectroscopic and compositional comparative characterization of I.H.S.S. reference and standard fulvic and humic acids of various origin. Science of the Total Environment 1989, 81-82, 143-156; (b) Mobed, J. J.; Hemmingsen, S. L.; Autry, J. L.; McGown, L. B., Fluorescence Characterization of IHSS Humic Substances: Total Luminescence Spectra with Absorbance Correction. Environmental Science & Technology 1996, 30 (10), 3061-3065; (c) Westerhoff, P.; Chen, W.; Esparza, M., Fluorescence Analysis of a Standard Fulvic Acid and Tertiary Treated Wastewater. J Environ Qual 2001, 30 (6), 2037-2046. 19. (a) Abdulla, H. A. N.; Minor, E. C.; Hatcher, P. G., Using Two-Dimensional Correlations of 13C NMR and FTIR To Investigate Changes in the Chemical Composition of Dissolved Organic Matter along an Estuarine Transect. Environmental Science & Technology 2010, 44 (21), 8044-8049; (b) Conte, P.; Piccolo, A., Conformational Arrangement of Dissolved Humic Substances. Influence of Solution Composition on Association of Humic Molecules. Environmental Science & Technology 1999, 33 (10), 1682-1690; (c) De Paolis, F.; Kukkonen, J., Binding of organic pollutants to humic and fulvic acids: Influence of pH and the structure of humic material. Chemosphere 1997, 34 (8), 1693-1704. 22. Nriagu, J. O., Arsenic in the Environment, Cycling and Characterization. Wiley: 1994. 24. Agency, U. S. E. P. Safe Drinking Water Act (SWDA) Laws and Regulations. http://water.epa.gov/lawsregs/rulesregs/sdwa/arsenic/index.cfm. 25. Manning, B. A.; Goldberg, S., Adsorption and Stability of Arsenic(III) at the Clay Mineral−Water Interface. Environmental Science & Technology 1997, 31 (7), 2005-2011. 26. Kimberly Gehle MD, M. S. C., PhD Case Studies in Environmental Medicine - Arsenic Toxicity. http://www.atsdr.cdc.gov/. 27. (a) Cheng, H.; Hu, Y.; Luo, J.; Xu, B.; Zhao, J., Geochemical processes controlling fate and transport of arsenic in acid mine drainage (AMD) and natural systems. Journal of Hazardous Materials 2009, 165 (1-3), 13-26; (b) Sharma, V. K.; Sohn, M., Aquatic arsenic: toxicity, speciation, transformations, and remediation. Environment international 2009, 35 (4), 743-59. 30. Lehr, J.; Keeley, J.; Lehr, J., Water Encyclopedia, Volumes 1-5. John Wiley & Sons: 2005. 31. (a) Agency, E. P., National Secondary Drinking Water Regulations. Office, C. o. F. R., Ed. Office of the Federal Register National Archives and Records Administration: 2002; Vol. 19; (b) Smedley, P. L.; Kinniburgh, D. G., A review of the source, behaviour and distribution of arsenic in natural waters. Applied Geochemistry 2002, 17 (5), 517-568. 32. Vahter, M., Mechanisms of arsenic biotransformation. Toxicology 2002, 181-182, 211-217. 33. (a) Lytle, D. A.; Chen, A. S.; Sorg, T. J.; Phillips, S.; French, K., Microbial As(III) oxidation in water treatment plant filters. American Water Works Association. Journal 2007, 99 (12), 72-86,12; (b) Lytle, D. A., Sorg, T. J. and Snoeyink, V. L. , Optimizing arsenic removal during iron removal: Theoretical and practical considerations. Journal of Water Supply: Research and Technology - Aqua 2005, 54 (8), 545-560. 34. Agency, U. S. E. P. http://www.epa.gov/enviro/html/icr/index.html. 37. (a) McCurry, D.; Speth, T.; Pressman, J., Lyophilization and Reconstitution of Reverse-Osmosis Concentrated Natural Organic Matter from a Drinking Water Source. Journal of Environmental Engineering

126

2012, 138 (4), 402-410; (b) Weng, L.; Van Riemsdijk, W. H.; Hiemstra, T., Effects of Fulvic and Humic Acids on Arsenate Adsorption to Goethite: Experiments and Modeling. Environmental Science & Technology 2009, 43 (19), 7198-7204. 38. Chappell, W. R.; Abernathy, C. O.; Calderon, R. L., Arsenic Exposure and Health Effects III. Elsevier Science: 1999. 40. Vega, L.; Styblo, M.; Patterson, R.; Cullen, W.; Wang, C.; Germolec, D., Differential Effects of Trivalent and Pentavalent Arsenicals on Cell Proliferation and Cytokine Secretion in Normal Human Epidermal Keratinocytes. Toxicology and Applied Pharmacology 2001, 172 (3), 225-232. 56. Bednar, A. J.; Garbarino, J. R.; Ranville, J. F.; Wildeman, T. R., Preserving the Distribution of Inorganic Arsenic Species in Groundwater and Acid Mine Drainage Samples. Environmental Science & Technology 2002, 36 (10), 2213-2218. 57. Almassalkhi, B. A. Arsenic Speciation, Detection, and Quantification in Drinking Water using High Performance Liquid Chromatography and Inductively Coupled Plasma Mass Spectrometry. University of Cincinnati, 2009. 58. Silverstein, R. M.; Bassler, G. C.; Morrill, T. C., Spectrometric identification of organic compounds. Wiley: 1974. 59. Sen, S.; Pal, U.; Maiti, N. C., pKa determination of D-ribose by Raman spectroscopy. The journal of physical chemistry. B 2014, 118 (4), 909-14. 60. Brown, W. H.; Foote, C. S.; Iverson, B. L., Organic Chemistry. Brooks/Cole: 2005. 61. Hennessy, S. A.; Moane, S. M.; McDermott, S. D., The reactivity of gamma-hydroxybutyric acid (GHB) and gamma-butyrolactone (GBL) in alcoholic solutions. Journal of forensic sciences 2004, 49 (6), 1220- 9. 62. Muzolf-Panek, M.; Gliszczyńska-Świgło, A.; Szymusiak, H.; Tyrakowska, B., The influence of stereochemistry on the antioxidant properties of catechin epimers. Eur Food Res Technol 2012, 235 (6), 1001- 1009. 63. Guilbault, G. G.; Chemistry, A. C. S. D. o. A., Fluorescence: Theory, Instrumentation, and Practice. E. Arnold: 1967. 64. (a) Wu, Q.-Y.; Hu, H.-Y.; Zhao, X.; Sun, Y.-X., Effect of Chlorination on the Estrogenic/Antiestrogenic Activities of Biologically Treated Wastewater. Environmental Science & Technology 2009, 43 (13), 4940-4945; (b) Lim, M.; Kim, M.-J., Removal of Natural Organic Matter from River Water Using Potassium Ferrate(VI). Water, Air, and Soil Pollution 2009, 200 (1-4), 181-189. 65. (a) Daus, B.; Mattusch, J.; uuml; rgen; Wennrich, R.; Weiss, H., Investigation on stability and preservation of arsenic species in iron rich water samples. Talanta 2002, 58 (1), 57-65; (b) McCleskey, R. B.; Nordstrom, D. K.; Maest, A. S., Preservation of water samples for arsenic(III/V) determinations: an evaluation of the literature and new analytical results. Applied Geochemistry 2004, 19 (7), 995-1009. 66. Redman, A. D.; Macalady, D. L.; Ahmann, D., Natural Organic Matter Affects Arsenic Speciation and Sorption onto Hematite. Environmental Science & Technology 2002, 36 (13), 2889-2896. 67. (a) Aggett, J.; Kriegman, M. R., Preservation of arsenic(III) and arsenic(V) in samples of sediment interstitial water. The Analyst 1987, 112 (2), 153-157; (b) Feldman, C., Improvements in the arsine accumulation-helium glow detector procedure for determining traces of arsenic. Analytical Chemistry 1979, 51 (6), 664-669. 68. E. M. Hall, G.; C. Pelchat, J.; Gauthier, G., Stability of inorganic arsenic (III) and arsenic (V) in water samples. Journal of Analytical Atomic Spectrometry 1999, 14 (2), 205-213. 69. EDDY, B. P.; INGRAM, M., Interactions between ascorbic acid and bacteria. 1953; Vol. 17, p 93- 107. 70. Farzana Akter, K.; Chen, Z.; Smith, L.; Davey, D.; Naidu, R., Speciation of arsenic in ground water samples: A comparative study of CE-UV, HG-AAS and LC-ICP-MS. Talanta 2005, 68 (2), 406-415. 71. Rajakovic, L. V., Todorovic, Zaklina N., Rajakovic-Ognjanovic, Vladana N. and Onjia, Antonije E., Analytical methods for arsenic speciation analysis. Journal of the Serbian Chemical Society 2013, 78 (10), 1461-1479.

127

72. Watts, J.; O'Reilly, J. P.; Smiles, C. A.; Cook, J. M.; Survey, B. G., Measurement of Arsenic Compounds in Water by HPLC-ICP-MS: British Geological Survey Report OR/07/021. 2007. 73. Pressman, J. G.; Richardson, S. D.; Speth, T. F.; Miltner, R. J.; Narotsky, M. G.; Hunter, I. I. I. E. S.; Rice, G. E.; Teuschler, L. K.; McDonald, A.; Parvez, S.; Krasner, S. W.; Weinberg, H. S.; McKague, A. B.; Parrett, C. J.; Bodin, N.; Chinn, R.; Lee, C.-F. T.; Simmons, J. E., Concentration, Chlorination, and Chemical Analysis of Drinking Water for Disinfection Byproduct Mixtures Health Effects Research: U.S. EPA’s Four Lab Study. Environmental Science & Technology 2010, 44 (19), 7184-7192. 74. (a) Liu, G.; Cai, Y., Complexation of arsenite with dissolved organic matter: Conditional distribution coefficients and apparent stability constants. Chemosphere 2010, 81 (7), 890-896; (b) Wang, S.; Mulligan, C. N., Natural attenuation processes for remediation of arsenic contaminated soils and groundwater. Journal of Hazardous Materials 2006, 138 (3), 459-470; (c) Wang, S.; Mulligan, C. N., Effect of natural organic matter on arsenic release from soilsand sediments into groundwater. Environmental Geochemistry and Health 2006, 28 (3), 197-214. 75. Lin, H.-T.; Wang, M. C.; Li, G.-C., Complexation of arsenate with humic substance in water extract of compost. Chemosphere 2004, 56 (11), 1105-1112. 76. (a) Ritter, K.; Aiken, G. R.; Ranville, J. F.; Bauer, M.; Macalady, D. L., Evidence for the Aquatic Binding of Arsenate by Natural Organic Matter−Suspended Fe(III). Environmental Science & Technology 2006, 40 (17), 5380-5387; (b) Chen, Z.; Cai, Y.; Solo-Gabriele, H.; Snyder, G. H.; Cisar, J. L., Interactions of Arsenic and the Dissolved Substances Derived from Turf Soils. Environmental Science & Technology 2006, 40 (15), 4659-4665.

128

Chapter 4

Conclusions and Future Work

4.0 Introduction

The conclusions contained within this chapter summarize the entirety of results obtained over the course of completing the work detailed within this thesis. The results obtained using various methods of analytical measurement including UV-Vis Spectroscopy, Fluorescence, ICP-AES, HPLC-ICP-MS and

TOC analysis were used to elucidate the potential functional groups and/or structures contained within samples of organic matter and correlate its effect, if any, on the species of inorganic arsenic (As III/V) in drinking and source water. This information is essential in both maintaining and developing new methods to ensure the quality of current and future sources of drinking water continue in the face of increasing reports of water shortages and high levels of harmful contaminants including arsenic and several regulated

DBP’s, which are believed to be byproducts of organic matter in water.

4.1 UV-Vis and Fluorescence Spectroscopy

UV-Vis spectroscopy analysis of samples of organic matter result in generally inconclusive spectra which show minimal features above 250 nm, with increasing absorbance intensity as the spectra approaches

200 nm. The addition of NaHCO3 at concentrations of 10 and 50 mM, varying the solution pH between 2 and 10 resulted in both increased UV-Vis absorbance intensity and the appearance of features in the HA spectra obtained during the analysis. In general, the HA spectra obtained showed the most intense absorbance and features when analysis was performed at pH 8 or higher, in 50 mM NaHCO3. The features detected included the appearance of a shoulder at ~ 215 nm, in 10 mM NaHCO3, and the appearance of a peak starting at ~ 234 nm, cresting at ~ 206 nm in 50 mM NaHCO3 at pH 10. The UV-Vis wavelength and absorbance intensity suggested the electronic transitions detected in HA could be the result of the n to Π* transitions typical of lone, unsaturated chromophores, like carbonyl groups, and/or the Π to Π* transitions

129 typical of aromatic molecules, like benzene [58]. However, despite the additional features detected in HA spectra by modifying the solvent environment, the results obtained remained inconclusive.

To potentially elucidate the aforementioned findings, 5 “model” compounds, D-(-)-ribose, γ- butyrolactone (GBL), mandelic acid, catechin and DL-tyrosine, were analyzed using UV-Vis, under similar solvent conditions, and the spectra obtained were compared to the HA spectra previously obtained. To some extent, the UV-Vis spectra obtained from all 5 of the model compounds showed similarities to the spectra obtained from HA. However the spectra obtained from GBL, Figures 3.1.2-3 and 3.1.2-9, which contained a lone carbonyl group capable of n to Π* transitions, and the spectra from mandelic acid, Figures

3.1.2-4 and 3.1.2-10, which contained a carboxylic acid group, capable of n to Π* as well as a phenyl group, capable of Π to Π* transitions in solution [58] showed the most similarity to the HA spectra. While the spectra of the electronic transitions seen in catechin and tyrosine showed minimal similarities to the spectra of HA, they did help to confirm the transitions detected in HA were most likely the result of saturated chromophores or a single aromatic group within the molecule. The UV-Vis spectra obtained for catechin,

Figures 3.1.2-5 and 3.1.2-11, which contains two phenyl groups, each containing two hydroxyl groups, one

“ortho” and one “meta”, showed indications of the “vibrational” transitions, typical of benzene rings [58], overlapping the Π to Π* transitions at all of the pH’s analyzed. The intensity of the “vibrational” transitions detected in catechin, which was only seen in the mandelic acid spectra, to a lesser extent, at pH 11 in 10

* mM NaHCO3, were most likely caused by the compounding Π to Π transition overlap detected from both phenyl groups. In addition, the spectra of catechin showed the appearance of two smaller, red shifted, peaks at ~ 250 and 296 nm, which may have been a result of the hydroxyl group substitutions on the benzene rings. The spectra for tyrosine, Figures 3.1.2-6 and 3.1.2-12, showed the least amount of similarity to HA, with the appearance of similar “vibrational” rotation overlap of the benzene Π to Π* transition detected at

~ 206 nm. The spectra for tyrosine also showed well-defined peaks at ~ 213 and 238 nm, possibly due to electronic transitions resulting from the para-hydroxyl substituted phenyl group and the carboxylic acid substituted meso-amine within the molecule. Combined, the features detected in the spectra for both

130 catechin and tyrosine indicated that if aromatic groups, like benzene, were present in the samples of HA analyzed, there was most likely only one or they were unconjugated and/or extremely well-spaced within the molecule.

The analysis of HA and the 5 model compounds in dilute solutions of NaHCO3 also highlighted the potential for error that can occur when applying “blank” corrections during analysis. The n to Π* transition of the carbonyl group in NaHCO3 resulted in absorbance detected in all the “blanks” that was intense enough to result in manually “calculated” negative absorbance values for GBL, Figure 3.2.1-9, in all of the 50 mM NaHCO3 samples and at pH 11 in 10 mM NaHCO3, and for all ribose samples after the corrections were applied to the data obtained. However, comparison of the uncorrected and the “blank” corrected spectra obtained for HA and the 5 model compounds ultimately resulted in the same final conclusions: that the electronic transitions seen in the UV-Vis spectra for HA were most likely the result of a lone, saturated chromophore or a single and/or unconjugated aromatic group like benzene.

Fluorescence spectroscopy potentially confirmed these UV-Vis findings, when samples of HA were analyzed in 50 mM NaHCO3 at various solution pH. The detection of fluorescence in all of the samples confirmed the presence of a chromophore within the HA analyzed. Scanning the samples at three different excitation/emission wavelengths resulted in very similar maximum ex/em wavelengths that indicated the fluorescence was due to a single chromophore in solution, confirming the electronic transition detected during UV-Vis analysis.

In addition, the UV-Vis calibration, Figure 3.1.3-1, using a commercially available HA, that was performed at 254 nm, at concentrations of 2, 4, 6, 8, 10, 12 and 15 ppm, showed a linear response to the concentrations of HA in solution, with a calculated R2 values of 0.9986. This suggesting a method of instrument calibration, using commercially available and HA/FA/DOM, and subsequent “unknown” HA concentration determination using a “standard additions” method of detection may be useful in areas at high risk for DBP formation due to elevated DOM concentrations in water.

131

4.2 Arsenic III/V Speciation

The separation, detection and quantification of As III/V in water was achieved using a proposed

RP-HPLC-ICP-MS method that used a 2.5 mM (NH4)H2PO4 + 5 mM TBAH mobile phase at pH 6 to separate inorganic As III and V in solution. Using the method, As III and V was separated within six minutes with average retention times of 1.145 ± 0.023 minutes, for As III, and 4.146 ± 0.417 minutes, for

As V. Instrument calibration using calibration standards prepared at As III/V concentrations of 1, 10, 25,

50, 75, 100 and 150 ppb showed linear instrument response and calculated R2 values of 0.9997 ± 0.0004 for As III and 0.9998 ± 0.0002 for As V.

A four week analysis of potential As III preservatives (no preservative, ascorbic acid, 0.05 % HCl

& 1.25E-03 M EDTA) showed samples preserved in EDTA had the lowest detected As III oxidation, 5, -1.0

& -1.3 %, respectively, at 1, 50 and 150 ppb As III, compared to their original instrument reported concentrations. While preservation of As III with dilute HCl resulted in an immediate decrease in the ICP-

MS signal response which may have been due to immediate oxidation of As III to V. Most surprising was, at the conclusion of this analysis, the unpreserved As III/V samples prepared at higher arsenic concentrations (50 & 150 ppb) also showed minimal As III oxidation, only 19 & 7 %, respectively, at 50 and 150 ppb As III. This suggested the level of As III oxidation occurring in solution may be dependent upon the initial arsenic concentration, with the level of As III oxidation decreasing with increasing species concentration. Hall et al. noticed similar reductions in As III oxidation at higher solutions concentrations, suggesting at higher As III concentrations a kinetic equilibrium existed that reduced the cycling between

As III and V species in solution reducing levels of oxidation/reduction [68]. If accurate, and higher levels of the purportedly more toxic As III are equilibrium stabilized, this information suggests the occurrence of elevated As III in natural waters could require chemical and/or biological intervention to convert As III to

V, facilitating its removal from a contaminated water source.

A similar four week analysis evaluating the potential effect, if any, of concentrated organic matter

(cDOM) on the speciation of As III/V in water showed As III was preserved, even at low concentrations,

132 when 1, 50 and 150 ppb As III/V samples were spiked with 1 ppm cDOM. After four weeks, the 1 ppb As

III sample showed only 18.8 % oxidation from its original instrument calculated concentration. This information indicates, natural water sources with elevated concentration of DOM may inhibit As III oxidation, even at low concentrations.

4.3 Future Work

While the analyses performed, and reported, within this work have provided a large degree of information, they have also led to several new questions which may lead to additional analysis performed in the future. While the analyses performed over the course of completing this work evaluated several different potential interferences in the separation and detection of As III/V using HPLC-ICP-MS, several known inhibitors/interferences were not analyzed, including iron (Fe), manganese (Mn), etc. In addition, there may be a future need to provide Cr III/VI speciation analysis in water which indicates future research evaluating and developing a potential method of simultaneous As III/V-Cr III/VI speciation and detection may be beneficial.

References

58. Silverstein, R. M.; Bassler, G. C.; Morrill, T. C., Spectrometric identification of organic compounds. Wiley: 1974. 68. E. M. Hall, G.; C. Pelchat, J.; Gauthier, G., Stability of inorganic arsenic (III) and arsenic (V) in water samples. Journal of Analytical Atomic Spectrometry 1999, 14 (2), 205-213.

133

Chapter 5

Supplemental Information

S.I. Table 5.1-1. ICP-AES metals analysis of concentrated dissolved organic matter (cDOM) in water and humic acid (HA) in DI water and 10 and 50 mM NaHCO3.

HA_10 mM HA_50 mM HA_DI-1 HA_DI-2 NaHCO3 NaHCO3 cDOM_d100 Al 0.44 0.45 0.39 0.45 1.29 As 0.023 0.049 0.11 0.12 -0.11 Ba 1.54 1.52 1.50 1.50 2.47 Be 0.001 0.001 0.001 0.000 0.003 Ca 1.69 1.59 1.77 1.73 4.17 Cd 0.003 0.001 0.001 0.003 0.000 Cr 0.014 0.017 0.011 0.012 -0.021 Cu 0.030 0.022 0.033 0.024 -0.017 Fe 0.38 0.36 0.37 0.40 1.36 K 3.71 4.96 3.71 3.27 0.26 Li 0.006 0.008 -0.005 0.038 -0.30 Mg 0.23 0.22 0.25 0.25 0.59 Mn 0.014 0.011 0.012 0.023 -0.003 Na 1.25E+03 1.25E+03 2.24E+03 2.24E+03 1.31E+03 Ni -0.010 -0.016 -0.015 -0.005 -0.003 P -0.36 -0.27 -0.18 -0.22 -0.20 Pb 0.033 0.042 0.005 0.032 -0.10 S 2.00 1.80 2.45 3.03 5.54 Sb -0.020 0.008 -0.008 -0.004 -0.090 Si 0.49 0.051 -0.19 0.53 -0.72 Sn -0.011 0.009 0.003 -0.002 -0.032 Sr 0.013 0.017 0.013 0.014 0.002 V -0.011 0.005 -0.001 0.001 0.023 Zn 1.04 0.58 0.88 0.97 0.58

- 134 -

1 ppb Arsenic III/cDOM 1.20E+04 1.00E+04 8.00E+03 6.00E+03 4.00E+03

2.00E+03 MSSignal(cps)

- 0.00E+00

-2.00E+03 ICP -4.00E+03 0 100 200 300 400 500 600 700 Time (hours)

50 ppb Arsenic III/cDOM 5.00E+05 4.00E+05 3.00E+05

2.00E+05 MSSignal(cps)

- 1.00E+05 ICP 0.00E+00 0 100 200 300 400 500 600 700 Time (hours)

150 ppb Arsenic III/cDOM stability 1.60E+06 1.40E+06 1.20E+06 1.00E+06 8.00E+05

6.00E+05 MSSignal(cps) - 4.00E+05

ICP 2.00E+05 0.00E+00 0 100 200 300 400 500 600 700 Time (hours)

S.I. Figure 5.1-1 A-C. Analysis of the stability of As III spiked with cDOM in unpreserved, ascorbic acid, EDTA and HCl preserved samples over four weeks. (◊) No preservation; (□) ascorbic acid; (∆) EDTA; (○) HCl.

- 135 -

Comparing 1 and 50 ppb As III with cDOM 1.2E+04 5.0E+05

1.0E+04 4.5E+05 ICP 4.0E+05

8.0E+03 - 3.5E+05 (cps) Signal MS 6.0E+03 3.0E+05 4.0E+03 2.5E+05

2.0E+03 2.0E+05 MSSignal(cps)

- 1.5E+05 0.0E+00 1.0E+05 ICP -2.0E+03 5.0E+04 -4.0E+03 0.0E+00 0 100 200 300 400 500 600 700 Time (hours)

Comparing 1 and 150 ppb As III with cDOM 1.2E+04 1.6E+06

1.0E+04 1.4E+06 ICP

8.0E+03 1.2E+06 - MS (cps) Signal MS 6.0E+03 1.0E+06 4.0E+03 8.0E+05

2.0E+03 6.0E+05 MSSignal(cps) - 0.0E+00 4.0E+05

ICP -2.0E+03 2.0E+05 -4.0E+03 0.0E+00 0 100 200 300 400 500 600 700 Time (hours)

S.I. Figure 5.1-2 A-C. Comparing the % of As III oxidation detected in 1, 50 and 150 ppb As III samples, spiked with 1 ppm cDOM, over four weeks. 1 ppb As III - primary y-axis; 50 and 150 ppb As III - secondary y-axis. Final concentration of cDOM – 1 ppm A. 1 vs 50 ppb As III: 1 ppb (♦) No preservation; (■) ascorbic acid; (▲) EDTA; (●) HCl; 50 ppb (◊) No preservation; (□) ascorbic acid; (∆) EDTA; (○) HCl; B. 1 vs 150 ppb As III: 1 ppb (♦) No preservation; (■) ascorbic acid; (▲) EDTA; (●) HCl; 150 ppb (◊) No preservation; (□) ascorbic acid; (∆) EDTA; (○) HCl and C. 50 vs 150 ppb: 50 ppb (♦) No preservation; (■) ascorbic acid; (▲) EDTA; (●) HCl; 150 ppb (◊) No preservation; (□) ascorbic acid; (∆) EDTA; (○) HCl.

- 136 -

Comparing 50 and 150 ppb As III with cDOM 5.0E+05 1.6E+06 4.5E+05 1.4E+06 4.0E+05 ICP

1.2E+06 -

3.5E+05 (cps) Signal MS 3.0E+05 1.0E+06 2.5E+05 8.0E+05

2.0E+05 6.0E+05 MSSignal(cps) - 1.5E+05 4.0E+05

ICP 1.0E+05 5.0E+04 2.0E+05 0.0E+00 0.0E+00 0 100 200 300 400 500 600 700 Time (hours)

S.I. Figure 3.3-2 A-C. Comparing the % of As III oxidation detected in 1, 50 and 150 ppb As III samples, spiked with 1 ppm cDOM, over four weeks. 1 ppb As III - primary y-axis; 50 and 150 ppb As III - secondary y-axis. Final concentration of cDOM – 1 ppm A. 1 vs 50 ppb As III: 1 ppb (♦) No preservation; (■) ascorbic acid; (▲) EDTA; (●) HCl; 50 ppb (◊) No preservation; (□) ascorbic acid; (∆) EDTA; (○) HCl; B. 1 vs 150 ppb As III: 1 ppb (♦) No preservation; (■) ascorbic acid; (▲) EDTA; (●) HCl; 150 ppb (◊) No preservation; (□) ascorbic acid; (∆) EDTA; (○) HCl and C. 50 vs 150 ppb: 50 ppb (♦) No preservation; (■) ascorbic acid; (▲) EDTA; (●) HCl; 150 ppb (◊) No preservation; (□) ascorbic acid; (∆) EDTA; (○) HCl.

Week 4: 1 ppb As III/V - Cystine 8.0E+02 7.0E+02 6.0E+02 5.0E+02 4.0E+02 3.0E+02

2.0E+02 MS Signal MS (cps)Signal - 1.0E+02 0.0E+00 ICP -1.0E+02

Time (min)

S.I. Figure 5.1-3. Analysis of 1 ppb As III/V in solution spiked with 1 ppm cystine. The sample was prepared in 2.5 mM (NH4)H2PO4 + 5 mM TBAH, pH 6 HPLC-ICP-MS mobile phase.

- 137 -

Arsenic V - Unpreserved 2.00E+04 1.60E+06 1.60E+04 1.20E+06 1.20E+04 8.00E+05 8.00E+03

4.00E+03 4.00E+05 MSSignal(cps) - 0.00E+00 0.00E+00

ICP 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

Arsenic V - Ascorbic Acid 2.50E+04 2.00E+06 2.00E+04 1.60E+06 1.50E+04 1.20E+06 1.00E+04 8.00E+05

5.00E+03 4.00E+05 MSSignal(cps) - 0.00E+00 0.00E+00

ICP 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

Arsenic V - EDTA 1.50E+04 1.50E+06

1.00E+04 1.00E+06

5.00E+03 5.00E+05

MSSignal(cps) - 0.00E+00 0.00E+00 ICP 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

S.I. Figure 5.1-4 A-D. Four week arsenic V stability analyzed by RP-HPLC-ICP-MS. Arsenic concentrations (ppb) were: 1 (○) - primary y-axis, 50 (□) and 150 (Δ) - secondary y-axis. Arsenic V samples preservation schemes were: A. Control (no preservative), B. 3:1 As III:Ascorbic Acid (A. A.), C. 1.25E-03 M EDTA and D. 0.05 % HCl.

- 138 -

Arsenic V - HCl 2.00E+04 1.60E+06 1.40E+06 1.60E+04 1.20E+06 1.20E+04 1.00E+06 8.00E+05

8.00E+03 6.00E+05 MSSignal(cps) - 4.00E+03 4.00E+05 2.00E+05 ICP 0.00E+00 0.00E+00 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

Arsenic V - Unpreserved with cDOM 1.00E+04 1.60E+06 8.00E+03 1.20E+06 6.00E+03 8.00E+05 4.00E+03

2.00E+03 4.00E+05 MSSignal(cps)

- 0.00E+00 0.00E+00

0 100 200 300 400 500 600 700 ICP Time (hours)

1 50 150

S.I. Figure 5.1-5 A-D. Four week analysis of arsenic V spiked with concentrated dissolved organic matter (cDOM) analyzed by RP-HPLC-ICP-MS. Concentrated dissolved organic matter concentration was 1 ppm. Arsenic concentrations (ppb) were: 1 (○) - primary y-axis, 50 (□) and 150 (Δ) - secondary y-axis. Arsenic V samples were: A. Control (no preservative), B. 3:1 As III:Ascorbic Acid (A. A.), C. 1.25E-03 M EDTA and D. 0.05 % HCl.

- 139 -

Arsenic V - Ascorbic Acid with cDOM 2.00E+04 2.50E+06 1.50E+04 2.00E+06 1.50E+06 1.00E+04 1.00E+06

5.00E+03 5.00E+05

MSSignal(cps) - 0.00E+00 0.00E+00

ICP 0 100 200 300 400 500 600 700 Time (hours)

1 50 150

Arsenic V - EDTA with cDOM 1.00E+04 1.50E+06 8.00E+03 6.00E+03 1.00E+06 4.00E+03 5.00E+05 2.00E+03

MSSignal(cps) 0.00E+00 0.00E+00 - 0 100 200 300 400 500 600 700

ICP Time (hours)

1 50 150

Arsenic V - HCl with cDOM 1.50E+04 1.50E+06

1.00E+04 1.00E+06

5.00E+03 5.00E+05

MSSignal(cps) 0.00E+00 0.00E+00 - 0 100 200 300 400 500 600 700 ICP Time (hours)

1 50 150

- 140 -

Appendix A

EPA Methods/Health and Safety Plans

METHOD #: 415.2 (Issued December 1982)

TITLE: Organic Carbon, Total (Low Level) (UV Promoted, Persulfate Oxidation)

ANALYTE: CAS # Total Organic Carbon TOC C 7440-44-0

INSTRUMENTATION: Carbon Analyzer

1.0 Scope and Application

1.1 This method covers the determination of total organic carbon in drinking water and other waters subject to the limitations in 1.3 and 5.1.

1.2 This instrument is designed for a two-step operation to distinguish between purgeable and non-purgeable organic carbon. These separate values are not pertinent to this method. 1.3 This method is applicable only to the carbonaceous matter which is either soluble or has a particle size of 0.2 mm or less.

1.4 The applicable range is from approximately 50 μg/ L to 10 mg/ L. Higher concentrations may be determined by sample dilution.

2.0 Summary of Method

A sample is combined with 1 mL of acidified persulfate reagent and placed in a sparger. The sample is purged with helium which transfers inorganic CO and2 purgeable organics to a CO scrubber. The CO is removed with at least 99.9% efficiency with a 2.5-minute purge. The purgeable organics proceed through a reduction system where the gas stream is joined by hydrogen and passed over a nickel catalyst which converts the purgeable organic carbon to methane. The methane is measured by a flame ionization detector. The detector signal is integrated and displayed as the concentration of purgeable organic carbon.

The sample is then transferred to a quartz ultraviolet reaction coil where the nonpurgeable organics are subjected to intense ultraviolet illumination in the presence of the acidified persulfate reagent. The nonpurgeables are converted to CO and2 transferred to a second sparger where a helium purge transfers the CO to the2 reduction system and into the detector. The signal is integrated, added to the purgeable organic carbon value, and displayed as the concentration of total organic carbon:

3.0 Definitions

- 1 -

3.1 Total organic carbon measured by this procedure is the sum of the purgeable organic carbon and the nonpurgeable organic carbon as defined in 3.2 and 3.3.

3.2 Purgeable organic carbon is the organic carbon matter that is transferred to the gas phase when the sample is purged with helium and which passes through the CO

scrubber. The definition is instrument- condition dependent.2

3.3 Nonpurgeable organic carbon is defined as that which remains after removal of the purgeable organic carbon from the sample containing acidified persulfate reagent and which is converted to CO under the instrument2 conditions. 3.4 The system blank is the value obtained in 8.2 for an irradiated, recirculated reagent distilled water sample.

4.0 Sample Handling and Preservation

4.1 Sampling and storage of samples must be done in glass bottles. Caution: Do not leave any headspace in the sample bottle as this may contribute to loss of purgeable organics. 4.2 Because of the possibility of oxidation or bacterial decomposition of some components of aqueous samples, the lapse of time between collection of samples and start of analysis should be kept to a minimum. Also, samples should be kept cool (4°C) and protected from sunlight and atmospheric oxygen.

4.3 When analysis cannot be performed within two hours from time of sampling, the sample should be acidified to pH 2 with H SO.

Note: HCl should not be used because it is converted to chlorine during the analysis. This causes damage to the instrument.

5.0 Interferences

5.1 If a sample is homogenized to reduce the size of the particulate matter, the homogenizing may cause loss of purgeable organic carbon, thus yielding erroneously low results.

6.0 Apparatus

6.1 Apparatus for blending or homogenizing samples: A household blender or similar device that will reduce particles in the sample to less than 0.2 mm.

6.2 Apparatus for Total Organic Carbon: The essential components for the apparatus used in this method are: A sparge assembly, flow switching valves, a pyrolysis furnace, quartz ultraviolet reactor coil, reducing column, flame ionization detector, electrometer and integrator. This method is based on the Dohrmann Envirotech DC-54 Carbon Analyzer. Other instruments having similar performance characteristics may be used.

- 2 -

6.3 Sampling Devices: Any apparatus that will reliably transfer 10 mL of sample to the sparger. A 50 mL glass syringe is recommended when analyzing samples with easily purgeable organics so as to minimize losses.

7.0 Reagents

7.1 Reagent Distilled Water: Distilled water used in preparation of standards and for dilution of samples should be ultra-pure to reduce the magnitude of the blank. Carbon dioxide-free, double distilled water is recommended. The water should be distilled from permanganate or be obtained from a system involving distillation and carbon treatment. The reagent distilled water value must be compared to a system blank determined on a recirculated distilled water sample. The total organic carbon value of the reagent distilled water should be less than 60 g/ L. Purgeable organic carbon values of the reagent distilled water should be less than 4 g/ L.

7.2 Potassium hydrogen phthalate, stock solution, 500 mg carbon/liter: Dissolve 1.063 g of potassium hydrogen phthalate (Primary Standard Grade) in reagent distilled water (7.1) and dilute to 1 liter.

7.3 Potassium hydrogen phthalate (2 mg/L): Pipet 4 mL of potassium hydrogen phthalate stock solution (7.2) into a one liter volumetric flask and dilute to the mark with reagent distilled water (7.1).

7.4 Potassium hydrogen phthalate (5 mg/L): Pipet 1 mL of potassium hydrogen phthalate stock solution (7.2) into a 100 mL volumetric flask and dilute to the mark with reagent distilled water (7.1).

7.5 Potassium hydrogen phthalate (10 mg/L): Pipet 2 mL of potassium hydrogen phthalate stock solution (7.2) into a 100 mL volumetric flask and dilute to the mark with reagent distilled water (7.1).

7.6 Acidified Persulfate Reagent: Place 100 mL of reagent distilled water (7.1) in a container. Add 5 g of potassium persulfate. Add 5 g (3 mL) of concentrated (85%) phosphoric acid.

7.7 Carbonate-bicarbonate, stock solution, 1000 mg carbon/liter: Place 0.3500 g of sodium bicarbonate and 0.4418 g of sodium carbonate in a 100 mL volumetric flask. Dissolve with reagent distilled water (7.1) and dilute to the mark.

7.8 Carbonate-bicarbonate, standard solution 50 mg/L: Place 5 mL of the carbonate- bicarbonate stock solution in a 100 mL volumetric flask and dilute to the mark with reagent distilled water (7.1).

8.0 Procedure

- 3 -

8.1 Allow at least 30 minutes warm-up time. Leave instrument console on continuously when in daily use, except for the ultraviolet light source, which should be turned off when not in use for more than a few hours.

8.2 Adjust all gas flows, temperatures and cycle times to manufacturer's specifications. Perform the "System Cleanup and Calibration" procedure in the manufacturer's specifications each day. Recirculate a sample of irradiated distilled water until two consecutive readings within 10% of each other are obtained. Record the last value for the system blank. This value is a function of the total instrument operation and should not vary significantly from previous runs. Reasons for significant changes in the value should be identified.

8.3 Check the effectiveness of the CO scrubber by analyzing the2 carbonate- bicarbonate standard solution(7.8). Add 1 mL of acidified persulfate reagent (7.6) to 50 mL of the solution. Transfer 10 mL of the solution-with-reagent to the first sparger and start the analysis cycle. No response, or a very minor reading, should be obtained from this solution.

8.4 Add 1 mL of acidified persulfate reagent (7.6) to 50 mL of reagent distilled water (7.1) blank, standards 7.3, 7.4, and 7.5 and the samples.

8.5 Calibrate the analyzer as follows:

8.5.1 Run the reagent distilled water (7.1) and 5.0 mg/L standard (7.4): Transfer 10 mL of the solution-with-reagent to the first sparger and start analyzer cycle.

Ignore the meter reading for the first cycle

Transfer a second 10 mL of the solution-with-reagent to the first sparger and start the analysis cycle

Record the meter reading (see 9.1) of the final carbon value for each of the reagent distilled water (7.1) and the standard (7.4).

If the meter reading is more than 25% above or below the calculated value of standard 7.4, reanalyze the standard and set the calibration within 25% (8.5.4), reanalyze the system blank, and then begin 8.5.1 again. If the meter reading (see 9.1) is within 25% of the calculated value, continue to next step. The calculated value is defined in 8.5.2.

8.5.2 Calculate the factor for the deviation of the instrument reading (see 9.1) for the standard (7.4) from the calculated value by:

where the calculated value is that value obtained by using the weight of potassium hydrogen phthalate and does not include the carbon contributed by the reagent distilled water (7.1) with which it has been diluted. 8.5.3 Calculate the adjusted reading by:

- 4 -

calculated value + (RDW - (FACTOR X RDW)) = ADJUSTED READING. where RDW = mean reagent distilled water (7.1) value.

8.5.4 Push in CALIBRATE button after READY light comes on and adjust the SPAN control to the ADJUSTED READING calculated in 8.5.3.

8.6 Analyze the standards 7.3 and 7.5 in order to check the linearity of the instrument at least once each day:

Transfer 10 mL of the solution-with-reagent to the first sparger and start analyzer cycle

Ignore the meter reading for the first cycle

Transfer a second 10 mL of the solution-with-reagent to the first sparger and start the analyzer cycle

Record the meter reading (see 9.1) of the final carbon value for each of the standards 7.3 and 7.5.

The range of concentration used for calibrating the instrument and checking the linearity of the instrument should be ascertained from a knowledge of the range of concentrations expected from the samples.

Standards for lower ranges can be prepared by diluting standards 7.2, 7.3, and 7.4.

8.7 Analyze the samples. Transfer 10 mL of sample with reagent to the first sparger and start the analysis cycle.

Transfer 10 mL of the solution-with-reagent to the first sparger and start analyzer cycle

Ignore the meter reading for the first cycle

Transfer a second 10 mL of the solution-with-reagent to the first sparger and start the analyzer cycle

Record the meter reading (see 9.1) of the final carbon value for each of the samples.

9.0 Calculations

9.1 The values are read off the final digital readout in g/ L. The system blank reading obtained in 8.2 must be subtracted from all reagent distilled water, standard and sample readings.

10.0 Precision and Accuracy

10.1 In a single laboratory (MERL), using raw river water, centrifuged river water, drinking water. and the effluent from a carbon column which had concentrations of 3.11, 3.10, 1.79, and 0.07 mg/ L total organic carbon respectively, the standard deviations from ten replicates were ±0.13, ±0.03, ±0.02, and ±0.02 mg/ L, respectively.

- 5 -

10.2 In a single laboratory (MERL), using potassium hydrogen phthalate in distilled water at concentrations of 5.0 and 1.0 mg/ L total organic carbon, recoveries were 80% and 91%, respectively.

Bibliography

1. Proposed Standard Method for Purgeable and Nonpurgeable Organic Carbon in Water (UV- promoted, persulfate oxidation method). ASTM Committee D-19, Task Group 19.06.02.03 (Chairman R. J. Joyce), January 1978. 2. Operating Instruction Dohrmann Envirotech, 3420 Scott Boulevard, Santa Clara, California 95050. 3. Takahashi, Y., "Ultra Low Level TOC Analysis of Potable Waters." Presented at Water Quality Technology Conference. AWWA, Dec. 5-8, 1976.

- 6 -

- 7 -

EPA Document #: EPA/600/R-06/115

METHOD 200.5 DETERMINATION OF TRACE ELEMENTS IN DRINKING WATER BY AXIALLY VIEWED INDUCTIVELY COUPLED PLASMA - ATOMIC EMISSION SPECTROMETRY

Revision 4.2

October 2003

Theodore D. Martin

NATIONAL EXPOSURE RESEARCH LABORATORY OFFICE OF RESEARCH AND DEVELOPMENT U. S. ENVIRONMENTAL PROTECTION AGENCY CINCINNATI, OHIO 45268

- 8 -

200.5-1 METHOD 200.5

DETERMINATION OF TRACE ELEMENTS IN DRINKING WATER BY AXIALLY VIEWED INDUCTIVELY COUPLED PLASMA-ATOMIC EMISSION SPECTROMETRY

1.0 SCOPE AND APPLICATION

1.1 Axially viewed inductively coupled plasma-atomic emission spectrometry (AVICP-AES) is used to determine trace elements, as well as water matrix elements, in drinking water and drinking water supplies. This method is applicable to the following analytes:

Chemical Abstract Services Analyte Abbreviation Registry Numbers (CASRN)

Aluminum (Al) 7429-90-5

Antimony* (Sb) 7440-36-0

Arsenic* (As) 7440-38-2

Barium* (Ba) 7440-39-3

Beryllium* (Be) 7440-41-7

Boron (B) 7440-42-8

Cadmium* (Cd) 7440-43-9

Calcium (Ca) 7440-70-2

Chromium* (Cr) 7440-47-3

Copper* (Cu) 7440-50-8

Iron (Fe) 7439-89-6

Lead* (Pb) 7439-92-1

Magnesium (Mg) 7439-95-4

Manganese (Mn) 7439-96-5

Nickel (Ni) 7440-02-0

Selenium* (Se) 7782-49-2

- 9 -

Silica (SiO2) 7631-86-9

Silver (Ag) 7440-22-4

Sodium (Na) 7440-23-5

Tin (Sn) 7440-31-5

Vanadium (V) 7440-62-2

Zinc (Zn) 7440-66-6

* Designated primary drinking water contaminant.

1.2 For reference where this method is approved for use in compliance monitoring program (e.g., Safe Drinking Water Act [SDWA]) consult both the appropriate sections of the Code of Federal Regulation (40 CFR Part 141 § 141.23) and the latest Federal Register announcements.

200.5-2 1.3 This method provides a specific procedure utilizing axially-viewed plasma atomic emission signals generated only by pneumatic nebulization for the analysis of all analytes. Some AVICP-AES instruments are so configured that the emitted signal can also be viewed alternately or simultaneously in a radial manner. Radially viewed signals for the determination of drinking water matrix elements (Ca, Mg and Na) and silica are acceptable. The Ca and Mg data can be used in the calculation of hardness.

1.4 When viewing sodium emission from the axial configuration, the ratio of signal intensity to analyte concentration is not a linear response. Therefore, sodium should be calibrated using multiple standard solutions of increasing concentration to properly define the response ratio at various levels of concentration (see Sect. 7.8.2).

1.5 For drinking water compliance monitoring, a “total” element determination (dissolved + suspended fractions) is required. When the measured turbidity on an acid preserved sample is < 1 NTU, direct analysis, without sample digestion, is permitted with the use of some approved spectrochemical methods. However, when using this method, all samples are digested and pre-concentrated prior to analysis using the total recoverable digestion(1) step described in Section 11.1. Pre-concentrating the sample prior to analysis increases analytical sensitivity for meeting the method detection limit (MDL) requirements given in Section 1.12. Thus, when using this method, the need to measure sample turbidity prior to metal analysis is eliminated.

- 10 -

1.6 Operative matrix effects can occur from elevated dissolved solids. Using this technique, matrix effects have been observed when the concentration of calcium and/or the combined concentrations of the matrix elements (Ca, Mg, and Na) and silica exceed 125 mg/L and 250 mg/L, respectively. To verify that a matrix effect is not operative, an LFM (see Sect. 9.4) must be analyzed when a primary contaminant (see Sect. 1.1) concentration exceeds 80% of the established maximum contaminant level (MCL) or action level. If the absence of a matrix interference can not be verified, the sample must be analyzed by method of standard additions (MSA; see Sect. 11.3) or another approved method (see Sect. 1.7 for special requirements for lead).

1.7 When determining lead by this method, the instrument must be capable of analyzing silica as well. Levels of silica that exceed 30 mg/L, when pre-concentrated 2X, cause a suppressive effect on lead determinations. For samples containing silica above 30 mg/L and lead concentrations $ 10 µg/L, lead must be determined by method of standard additions (MSA; see Sect. 11.3) or by another approved compliance monitoring method. If the laboratory can not determine silica when using this method, this method can not be used for compliance monitoring of lead.

1.8 When determining boron and silica, only plastic or PTFE labware should be used from time of sample collection to completion of analysis. In this method, glassware is specifically avoided and only the use of metal-free plastic labware is recommended. Borosilicate glass should be avoided to prevent contamination of these analytes.

1.9 The total recoverable sample digestion procedure given in Section 11.1 is suitable for the determination of silver in aqueous samples containing concentrations up to 0.1 mg/L. Also, samples prepared using the procedure may be analyzed for thallium using EPA Method 200.9(2).

1.10 Compliance monitoring data for metal contaminants are normally reported in units of mg/L; however, the data for the total recoverable analytes in this method are noted in units of µg/L. This difference is done to reduce or eliminate the listing of non-significant zeros. When data are reported for compliance monitoring, the data should be reported in the same units used to express the established MCL and to the appropriate numerical level of significance.

1.11 MDLs for trace elements and linear ranges for the drinking water matrix elements will vary with the wavelength selected and the spectrometer configuration and operating conditions. Table 4 provides determined MDLs for the listed wavelengths utilizing the instrument operating conditions given in Table 3. These values are provided for comparative purposes for user self-evaluation when completing the mandatory initial demonstration of performance. Meeting the exact same limits listed in Table 4 is not necessarily a requirement for the use of this method. Users of this method must document

- 11 -

and have on file the required initial demonstration performance data described in Section 9.2 prior to using the method for analysis (see Sect. 1.12).

1.12 Users of this method for the purpose of SDWA compliance monitoring must achieve and document MDLs for As, Be, Cd, Sb, Se, and Pb that are # a value of 1/5 their respective MCL or action level. 2.0 SUMMARY OF METHOD

2.1 A 50 mL aliquot of a well-mixed, non-filtered, acid preserved aqueous sample is accurately transferred to a clean 50-mL plastic disposable digestion tube containing a mixture of nitric and hydrochloric acids. The aliquot is heated to 95 °C (± 2 °C), evaporated to approximately 25 mL, covered with a ribbed plastic watch glass and subjected to total recoverable solubilization with gentle refluxing for 30 minutes. The sample is allowed to cool and diluted to 25 mL with reagent water to effect a 2X pre-concentration. The sample is capped, mixed and now ready for analysis (The time required to complete the sample preparation step is approximately 2.5 hours).

2.2 The analytical determinative step described in this method involves multi-elemental determinations by AVICP-AES using sequential or simultaneous instruments. The instruments measure characteristic atomic-line emission spectra by optical spectrometry. Standard and sample solutions are nebulized by pneumatic nebulization and the resulting aerosol is transported by argon carrier-gas to the plasma torch. Element specific emission spectra are produced by a radio-frequency inductively coupled plasma. The spectra are dispersed by a grating spectrometer, and the intensities of the line spectra are monitored at specific wavelengths by a photosensitive device. Photo currents from the photosensitive device are processed and controlled by a computer system. A background correction technique is required to compensate for variable background contribution to the determination of the analytes. Background should be measured adjacent to the analyte wavelength during analysis. Possible interferences that can occur must be considered and addressed appropriately as discussed in Section 4. 3.0 DEFINITIONS

3.1 CALIBRATION BLANK - A volume of reagent water acidified with the same acid matrix reagents as in the calibration standards. The calibration blank is a zero standard and is used to calibrate the AVICP instrument (Sects. 7.9.1).

3.2 CALIBRATION STANDARD (CAL) - A solution prepared from the dilution of stock standard solutions. The CAL solutions contain the acid matrix reagents and are used to calibrate the instrument response with respect to analyte concentration (Sect. 7.8.1).

3.3 DISSOLVED ANALYTE - The concentration of analyte in an aqueous sample that will pass through a 0.45-μm membrane filter assembly prior to sample acidification.

- 12 -

3.4 FIELD REAGENT BLANK (FRB) - An aliquot of reagent water that is placed in a sample container in the laboratory and treated as a sample in all respects, including shipment to the sampling site, exposure to the sampling site conditions, storage, acid preservation, and all analytical procedures. The FRB is used to determine if method analytes or interferences are present in the field environment (Sect. 8.2).

3.5 INSTRUMENT DETECTION LIMIT (IDL) - The concentration equivalent to the analyte signal which is equal to three times the standard deviation of a series of ten replicate measurements of the calibration blank signal at the same wavelength.

3.6 INSTRUMENT PERFORMANCE CHECK (IPC) SOLUTION - A solution of method analytes in the acid matrix reagents used to evaluate the performance of the instrument system with respect to a defined set of method criteria (Sects. 7.10.3 & 9.3.4).

3.7 LABORATORY DUPLICATES (LD1 and LD2) - Two aliquots of the same sample taken in the laboratory and analyzed separately with identical procedures. Analyses of LD1 and LD2 indicates precision associated with laboratory procedures, but not with sample collection, preservation, or storage procedures.

3.8 LABORATORY FORTIFIED BLANK (LFB) - An aliquot of LRB to which known quantities of the method analytes are added in the laboratory. The LFB is analyzed exactly like a sample, and its purpose is to determine whether the methodology is in control and whether the laboratory is capable of making accurate and precise measurements (Sects. 7.9.3 & 9.3.2).

3.9 LABORATORY FORTIFIED SAMPLE MATRIX (LFM) - An aliquot of a drinking water or drinking water supply sample to which known quantities of the method analytes are added in the laboratory. The LFM is analyzed exactly like a sample, and its purpose is to determine whether the sample matrix contributes bias to the analytical results. The background concentrations of the analytes in the sample matrix must be determined in a separate aliquot and the measured values in the LFM corrected for background concentrations (Sects. 1.6, 1.7 & 9.4).

3.10 LABORATORY REAGENT BLANK (LRB) - An aliquot of reagent water that is treated exactly as a sample including exposure to all labware, equipment, and reagents, that are used with other samples. The LRB is used to determine if method analytes or other interferences are present in the laboratory environment, reagents, or apparatus (Sects. 7.9.2 & 9.3.1).

3.11 LINEAR DYNAMIC RANGE (LDR) - The concentration range over which the instrument response to an analyte is linear (Sect. 9.2.2).

3.12 MAXIMUM CONTAMINANT LEVEL (MCL) - The maximum permissible level of a contaminant in water which is delivered to any user of a public water system.

- 13 -

3.13 METHOD DETECTION LIMIT (MDL) - The minimum concentration of an analyte that can be identified, measured, and reported with 99% confidence that the analyte concentration is greater than zero (Sects. 1.11, 1.12, 9.2.5 and Table 4).

3.14 PLASMA SOLUTION - A solution used to determine the nebulizer argon flow rate or gas pressure that will produce the optimum net-signal-to-noise (S-B/B) needed for the most requiring analyte included in the analytical scheme (Sects. 7.11 & 10.2).

3.15 QUALITY CONTROL SAMPLE (QCS) - A solution of method analytes of known concentrations which is used to fortify an aliquot of LRB or sample matrix. The QCS is obtained from a source external to the laboratory and different from the source of calibration standards. It is used to check either laboratory or instrument performance (Sects. 7.10.4 & 9.2.4).

3.16 SPECTRAL INTERFERENCE CHECK (SIC) SOLUTION - A solution of selected method analytes of higher concentrations which is used to evaluate the procedural routine for correcting known inter=element spectral interferences with respect to a defined set of method criteria (Sects. 4.1 & 9.3.5).

3.17 STANDARD ADDITION - The addition of a known amount of analyte to the sample in order to determine the relative response of the detector to an analyte within the sample matrix. The relative response is then used to assess either an operative matrix effect or the sample analyte concentration (Sects. 1.6 & 11.3).

3.18 STOCK STANDARD SOLUTION - A concentrated solution containing one or more method analytes prepared in the laboratory using assayed reference materials or purchased from a reputable commercial source (Sect. 7.7).

3.19 TOTAL RECOVERABLE ANALYTE - For this method, the concentration of analyte determined by the analysis of an unfiltered acid preserved drinking water sample following digestion by refluxing with hot dilute mineral acid(s) as specified in the method (Sects. 1.5 & 11.1). Data are reported as a “total” element determination - the combined concentrations of the dissolved and suspended fractions of the sample. 4.0 INTERFERENCES

4.1 Spectral interferences are caused by background emission from continuous or recombination phenomena, stray light from the line emission of high concentration elements, overlap of a spectral line from another element, or unresolved overlap of molecular band spectra. Except for interference from background emission and possible stray light, which can usually be compensated for by subtracting the background emission adjacent to the analyte wavelength peak, spectral interferences associated with the analysis of drinking water are minimal. However, the absence of inter-element spectral interference should be verified. Criteria for determining an inter-element spectral interference is an apparent positive or negative concentration on the analyte that is

- 14 -

outside the 3-sigma control limits of the calibration blank for the analyte. When an instrument equipped with a conventional diffraction grating that provides 0.016 nm first order resolution is used with the wavelengths in the noted spectral order and background correction locations given in Table 1, no detectable concomitant inter-element spectral interferences occurs between the trace element analytes listed in this method at concentrations # 20 mg/L. Since concentrations of trace elements in drinking water and drinking water supplies are far below the level of 20 mg/L, an inter-element correction routine for trace analytes would be unnecessary for an instrument so configured. On the other hand, the concentration of the water matrix elements can be in excess of 100 mg/L. Fortunately, the matrix elements are not spectrally rich and have few prominent lines to cause inter-element spectral interference. Using this method and analyzing single element solutions of 300 mg/L Ca, 200 mg/L Mg, 200 mg/L Na, and 100 mg/L Si, no spectral enhancement of other method analytes were observed, thus not requiring inter-element corrections. However, there are three concerns worth noting: (1) yttrium, a commonly used internal standard, proved an interference in the spectral region recommended for background correction on the listed Ag wavelength (328.068 nm), (2) a similar situation occurs from molybdenum on the spectral region recommended for background correction on the V wavelength (292.402 nm), and (3) the listed Fe wavelength (271.441 nm) experiences an apparent concentration increase of approximately 8% from cobalt (271.442 nm) when the two analytes are present in equal concentration. Therefore, a quality control check sample containing both iron and cobalt or both vanadium and molybdenum should not be used to confirm the calibration standards when the Fe 271.441 nm and V 292.402 nm wavelengths are utilized.

Note: If wavelengths, noted spectral order, and background correction locations different from those listed in Table 1 are used with this method, and/or the optical resolution of the instrument utilized does not provide 0.016 nm first order resolution or better, the absence of inter-element spectral interference must be confirmed by completing spectral scans over the wavelength area and background correction locations to be utilized. The spectral scans should be completed using single element solutions of both trace and water matrix elements of concentrations noted above that will verify nonexistent apparent analyte concentrations. If an inter-element spectral interference is detected, a correction routine that is operative during analysis must be used with daily verification using SIC solutions, to demonstrate that the routine meets the above criteria.

4.2 Physical interferences are effects associated with the sample nebulization and transport processes. Changes in viscosity and surface tension can cause significant inaccuracies, especially in samples containing high dissolved solids or high acid concentrations. Physical interferences of the types described above have not been evident in the analysis of most drinking waters. However, the use of a peristaltic pump to regulate solution uptake rate and the use of mass flow controllers that provide better control of the argon flow rates, especially for the nebulizer, improve instrument stability and precision.

- 15 -

4.3 Chemical interferences include molecular-compound formation, ionization effects, and solute-vaporization effects. In general, chemical interferences are highly dependent on matrix type and the specific element. In radial ICP-AES, one way of controlling these effects is careful selection of the observation height in the plasma. However, for increased sensitivity, the total emission of the plasma is observed in AVICP-AES, thus eliminating this useful option. To counteract ionization and matrix interferences in AVICP-AES, some laboratories routinely use an ionization buffer along with an internal standard added to both standards and samples alike in the sample train using a peristaltic pump and mixing tee. Use of an ionization buffer is permitted with this method provided the addition does not cause an inter-element spectral interference with a method analyte. However, in drinking water analyses the use of an internal standard with pneumatic nebulization is discouraged because it is not necessary and adds additional variance to the determination. The above stated chemical interferences have not been observed using this method for drinking water analyses when the operating conditions and preparations procedures are followed as written.

4.4 Memory interferences result when analytes in a previous sample contribute to the signals measured in a new sample. Memory effects can result from sample deposition on the uptake tubing to the nebulizer and from the buildup of sample material in the plasma torch and spray chamber. The site where these effects occur is dependent on the element and can be minimized by flushing the system with a rinse blank between samples (Sect. 7.9.4). The rinse times necessary for a particular element should be estimated prior to analysis. For the water matrix elements this may be achieved by aspirating a single element standard solution corresponding to their LDRs (Sect. 3.11), while for the trace contaminants single element solutions containing 10 mg/L are sufficient. The aspiration time should be the same as a normal sample analysis period, followed by analysis of the rinse blank at designated intervals. The length of time required to reduce analyte signal to within a factor of 10 times the calibration blank should be noted. Until the required rinse time is established, this method requires a rinse period using the rinse blank of at least 30 sec between samples and standards. If a memory interference is suspected, the sample should be re-analyzed after a long rinse period. 5.0 SAFETY

5.1 The toxicity or carcinogenicity of each reagent used in this method have not been fully established. Each chemical should be regarded as a potential health hazard and exposure to these compounds should be as low as reasonably achievable. Each laboratory is responsible for maintaining a current awareness file of OSHA regulations regarding the safe handling of the chemicals specified in this method (3-5). A reference file of material data handling sheets should also be made available to all personnel involved in the chemical analysis. Specifically, concentrated nitric and hydrochloric acids are moderately toxic and extremely irritating to skin and mucus membranes. Use these reagents in a fume hood whenever possible and if eye or skin contact occurs, flush with large volumes of

- 16 -

water. Always wear safety glasses or a shield for eye protection, protective clothing and observe proper mixing when working with these reagents.

5.2 Acidification of samples should be done in a fume hood.

5.3 The inductively coupled plasma should only be viewed with proper eye protection from the ultraviolet emissions.

5.4 It is the responsibility of the user of this method to comply with relevant disposal and waste regulations. For guidance see Sections 14.0 and 15.0. 6.0 EQUIPMENT AND SUPPLIES

6.1 Axially viewed inductively coupled plasma emission spectrometer:

6.1.1 Computer-controlled emission spectrometer with background-correction capability. The spectrometer must be capable of meeting and complying with the requirements described and referenced in Section 2.2.

6.1.2 Radio-frequency generator compliant with FCC regulations.

6.1.3 Argon gas supply - High purity grade (99.99%). When analyses are conducted frequently, liquid argon is more economical and requires less frequent replacement of tanks than compressed argon in conventional cylinders.

6.1.4 A variable speed peristaltic pump is required to deliver both standard and sample solutions to the nebulizer.

Use of a mass flow controller will provide more exacting control of reproducible plasma conditions.

6.2 An analytical balance with capability to measure to 0.1 mg, for use in preparing

standards and for weighing samples as may be required.

6.3 A temperature adjustable hot plate for preparing stock standard solutions.

6.4 A temperature adjustable block digester capable of maintaining a temperature of 95 oC for use with 50-mL plastic disposable digestion tube.

6.5 A gravity convection drying oven with thermostatic control capable of maintaining 180 oC ± 5 oC.

6.6 Air displacement pipetters capable of delivering volumes ranging from 50 μL to 10.0 mL with an assortment of high quality disposable pipet tips.

6.7 Labware - For determination of trace levels of elements, contamination and loss are of prime consideration. Potential contamination sources include improperly cleaned

- 17 -

laboratory apparatus and general contamination within the laboratory environment from dust, etc. A clean laboratory work area designated for trace element sample handling should be used. Sample containers can introduce positive and negative errors in the determination of trace elements by (1) contributing contaminants through surface desorption or leaching, and (2) depleting element concentrations through adsorption processes. All reusable labware (polyethylene, polymethylpentene, PTFE, FEP, etc.) and plastic disposable digestion tubes, caps, and watch glasses should be sufficiently clean for the task objectives. Several cleaning procedures can provide clean labware. The procedure recommended for reusable labware includes washing with a detergent solution, rinsing with tap water, and soaking for 4 h or

more in a mixture of 5% (v/v) HNO3 and 5% (v/v) HCl, rinsing with reagent water and storing clean. (If digested LRBs indicate random contamination, the plastic disposable

digestion tubes, caps, and watch glasses should be cleaned with 2% (v/v) HNO3 and rinsed with reagent water prior to use.) Chromic acid cleaning solutions must be avoided because chromium is an analyte.

6.7.1 Plastic volumetric labware - PMP (polymethylpentene) or equivalent metal free plastic volumetric flasks (50-mL to 500-mL capacities), graduated cylinders (50- mL), and disposable metal-free plastic digestion tubes with caps and watch glass covers.

6.7.2 (optional) PTFE Griffin beakers, 250-mL with PTFE covers for preparing stock standards and reagents.

6.7.3 Narrow-mouth storage bottles, FEP (fluorinated ethylene propylene) and LDPE (low density polyethylene) with screw closure, 60-mL to 500-mL capacities.

6.7.4 One-piece stem FEP wash bottle with screw closure, 125-mL capacity.

7.0 REAGENTS AND STANDARDS

7.1 Reagents may contain elemental impurities which might affect analytical data. Only high-purity reagents that conform to the American Chemical Society specifications must be used whenever possible. If the purity of a reagent is in question, analyze for contamination. All acids used for this method should be of ultra high-purity grade or equivalent. Trace metal grade acid may also be used if it can be verified by analysis to be free of contamination. Suitable acids are available from a number of manufacturers. Redistilled acids prepared by sub-boiling distillation are acceptable.

7.2 Hydrochloric acid, concentrated (sp.gr. 1.19) - HCl.

7.2.1 Hydrochloric acid (1+1) - Add 250 mL concentrated HCl to 200 mL reagent water and dilute to 500 mL.

- 18 -

7.2.2 Hydrochloric acid (1+20) - Add 10 mL concentrated HCl to 200 mL reagent water.

7.3 Nitric acid, concentrated (sp.gr. 1.41) - HNO3.

7.3.1 Nitric acid (1+1) - Add 250 mL concentrated HNO3 to 200 mL reagent water and dilute to 500 mL.

7.3.2 Nitric acid (1+2) - Add 100 mL concentrated HNO3 to 200 mL reagent water. 7.3.3 Nitric

acid (1+5) - Add 50 mL concentrated HNO3 to 250 mL reagent water. 7.3.4 Nitric acid (1+9) -

Add 10 mL concentrated HNO3 to 90 mL reagent water.

7.4 Reagent water. All references to reagent water in this method refer to ASTM Type I grade water (6).

7.5 Ammonium hydroxide, concentrated (sp. gr. 0.902).

7.6 Tartaric acid, ACS reagent grade.

7.7 Standard Stock Solutions - Stock standards may be purchased or prepared from ultra-high purity grade chemicals (99.99 to 99.999% pure). All compounds must be dried for 1 h at 105 oC, unless otherwise specified. It is recommended that stock solutions be stored in acid-cleaned, never-used LDPE bottles for storage. Replace stock standards when succeeding dilutions for preparation of calibration standards cannot be verified (see Sect. 9.2.4).

CAUTION: Many of these chemicals are extremely toxic if inhaled or swallowed (Sect. 5.1). Wash hands thoroughly after handling.

Typical stock solution preparation procedures follow for 500-mL quantities, but for the purpose of pollution prevention, the analyst is encouraged to prepare smaller quantities when possible. Concentrations are calculated based upon the weight of the pure element or upon the weight of the compound multiplied by the fraction of the analyte in the compound.

7.7.1 Aluminum solution, stock, 1 mL = 1000 μg Al: Dissolve 0.500 g of aluminum metal, weighed accurately to at least three significant figures, in an acid mixture

of 4.0 mL of (1+1) HCl and 1.0 mL of concentrated HN03 in a beaker. Warm beaker slowly to effect solution. When dissolution is complete, transfer solution quantitatively to a 500-mL PMP flask, add an additional 5.0 mL of (1+1) HCl and dilute to volume with reagent water.

7.7.2 Antimony solution, stock, 1 mL = 1000 μg Sb: Dissolve 0.500 g of antimony powder, weighed accurately to at least three significant figures, in 10.0 mL (1+1)

HNO3 and 5.0 mL concentrated HCl in a beaker. Add 50 mL reagent water and - 19 -

0.75 g tartaric acid. Warm solution slightly to effect complete dissolution. Cool solution and add reagent water to volume in a 500-mL PMP volumetric flask.

7.7.3 Arsenic solution, stock, 1 mL = 1000 μg As: Dissolve 0.660 g of As2O3 (As fraction = 0.7574), weighed accurately to at least three significant figures, in 50 mL of

reagent water containing 5.0 mL concentrated NH4OH in a beaker. Warm the solution gently to effect dissolution. Acidify the solution with 10.0 mL

concentrated HNO3 and dilute to volume in a 500-mL PMP volumetric flask with reagent water.

7.7.4 Barium solution, stock, 1 mL = 1000 μg Ba: Dissolve 0.719 g BaCO3 (Ba fraction = 0.6960), weighed accurately to at least three significant figures, in a beaker

containing 75 mL (1+2) HNO3 with heating and stirring to degas and dissolve compound. Let solution cool and dilute with reagent water in 500-mL PMP volumetric flask.

7.7.5 Beryllium solution, stock, 1 mL = 1000 μg Be: DO NOT DRY. Dissolve 9.823 g

BeSO4•4H2O (Be fraction = 0.0509), weighed accurately to at least four

significant figures, in reagent water, add 5.0 mL concentrated HNO3, and dilute to volume in a 500-mL PMP volumetric flask with reagent water.

7.7.6 Boron solution, stock, 1 mL = 1000 μg B: DO NOT DRY. Dissolve 2.859 g

anhydrous H3BO3 (B fraction = 0.1749), weighed accurately to at least four significant figures, in reagent water and dilute in a 500-mL PMP volumetric flask with reagent water.

7.7.7 Cadmium solution, stock, 1 mL = 1000 μg Cd: Dissolve 0.500 g Cd metal, acid

cleaned with (1+9) HNO3, weighed accurately to at least three significant figures,

in 25 mL (1+1) HNO3 in a beaker with heating to effect dissolution. Let solution cool and dilute with reagent water in a 500-mL PMP volumetric flask.

7.7.8 Calcium solution, stock, 1 mL = 1000 μg Ca: Suspend 1.249 g CaCO3 (Ca fraction = 0.4005), dried at 180 °C for 1 h before weighing, weighed accurately to at least four significant figures, in reagent water and dissolve cautiously with a minimum

amount of (1+1) HNO3. Add 5.0 mL concentrated HNO3 and dilute in a 500-mL PMP volumetric flask with reagent water.

7.7.9 Chromium solution, stock, 1 mL = 1000 μg Cr: Dissolve 0.962 g CrO3 (Cr fraction = 0.5200), weighed accurately to at least three significant figures, in 60 mL (1+5)

HNO3. When dissolution is complete, dilute in a 500-mL PMP volumetric flask with reagent water.

7.7.10 Copper solution, stock, 1 mL = 1000 μg Cu: Dissolve 0.500 g Cu metal, acid

cleaned with (1+9) HNO3, weighed accurately to at least three significant figures,

- 20 -

in 25 mL (1+1) HNO3 in a beaker with heating to effect dissolution. Let solution cool and dilute in a 500-mL PMP volumetric flask with reagent water.

7.7.11 Iron solution, stock, 1 mL = 1000 μg Fe: Dissolve 0.500 g Fe metal, acid cleaned with (1+1) HCl, weighed accurately to three significant figures, in 50 mL (1+1) HCl in a beaker with heating to effect dissolution. Let solution cool and dilute in a 500-mL PMP volumetric flask with reagent water.

7.7.12 Lead solution, stock, 1 mL = 1000 μg Pb: Dissolve 0.799 g Pb(NO3)2 (Pb fraction = 0.6256), weighed accurately to at least three significant figures, in a minimum

amount of (1+1) HNO3. Add 10.0 mL (1+1) HNO3 and dilute in a 500-mL PMP volumetric flask with reagent water.

7.7.13 Magnesium solution, stock, 1 mL = 1000 μg Mg: Dissolve 0.500 g of cleanly polished Mg ribbon, accurately weighed to at least three significant figures, in slowly added 2.5 mL (1+1) HCl (CAUTION: reaction is vigorous). Add 10.0 mL

(1+1) HNO3 and dilute in a 500-mL PMP volumetric flask with reagent water.

7.7.14 Manganese solution, stock, 1 mL = 1000 μg Mn: Dissolve 0.500 g of manganese metal, weighed accurately to at least three significant figures, in 25 mL (1+1)

HNO3 and dilute to volume in a 500-mL PMP volumetric flask with reagent water.

7.7.15 Nickel solution, stock, 1 mL = 1000 μg Ni: Dissolve 0.500 g of nickel metal, weighed accurately to at least three significant figures, in 10.0 mL hot

concentrated HNO3, cool, and dilute in a 500-mL PMP volumetric flask with reagent water.

7.7.16 Selenium solution, stock, 1 mL = 1000 μg Se: Dissolve 0.703 g SeO2 (Se fraction = 0.7116), weighed accurately to at least three significant figures, in 100 mL reagent water and dilute to volume in a 500-mL PMP volumetric flask with reagent water.

7.7.17 Silica solution, stock, 1 mL = 1000 μg SiO2: DO NOT DRY. Dissolve

1.482 g (NH4)2SiF6, weighed accurately to at least four significant figures, in 100 mL (1+20) HCl in a beaker with heating at 85 °C to effect dissolution. Let solution cool and dilute in a 500-mL PMP volumetric flask with reagent water.

7.7.18 Silver solution, stock, 1 mL = 1000 μg Ag: Dissolve 0.500 g Ag metal, weighed

accurately to at least three significant figures, in 50 mL (1+1) HNO3 in a beaker with heating to effect dissolution. Let solution cool and dilute in a 500-mL PMP volumetric flask with reagent water.

- 21 -

7.7.19 Sodium solution, stock, 1 mL = 1000 μg Na: Dissolve 1.271 g NaCl (Na fraction = 0.3934), weighed accurately to at least four significant figures, in reagent

water. Add 5.0 mL concentrated HNO3 and dilute to volume in a 500-mL PMP volumetric flask with reagent water.

7.7.20 Tin solution, stock, 1 mL = 1000 μg Sn: Dissolve 0.500 g Sn shot, weighed accurately to at least three significant figures, in an acid mixture of 5.0 mL

concentrated HCl and 1.0 mL (1+1) HNO3 in a beaker with heating to effect dissolution. Let solution cool, add 100 mL concentrated HCl, and dilute to volume in a 500-mL PMP volumetric flask with reagent water.

7.7.21 Vanadium solution, stock, 1 mL = 1000 μg V: Dissolve 0.500 g V metal, acid

cleaned with (1+9) HNO3, weighed accurately to at least three significant figures,

in 25 mL (1+1) HNO3 in a beaker with heating to effect dissolution. Let solution cool and dilute in a 500-mL PMP volumetric flask with reagent water.

7.7.22 Zinc solution, stock, 1 mL = 1000 μg Zn: Dissolve 0.500 g Zn metal, acid cleaned

with (1+9) HNO3, weighed accurately to at least three significant figures, in 25

mL (1+1) HNO3 in a beaker with heating to effect dissolution. Let solution cool and dilute in a 500-mL PMP volumetric flask with reagent water.

7.8 Calibration Standard Solutions.

7.8.1 Mixed Calibration Standard Solutions - For total recoverable analyses prepare mixed calibration standard solutions by combining appropriate volumes of the

stock solutions in 500-mL PMP volumetric flasks containing 20 mL (1+1) HNO3 and 10 mL (1+1) HCl and dilute to volume with reagent water. Prior to preparing the mixed standards, each stock solution should be analyzed separately to determine possible spectral interferences or the presence of impurities. Care should be taken when preparing the mixed standards to ensure that the elements are compatible and stable together. To minimize the opportunity for contamination by the containers, it is recommended to transfer the mixed- standard solutions to acid-cleaned, never-used FEP fluorocarbon (FEP) bottles for storage. Fresh mixed standards should be prepared, as needed, with the realization that concentrations can change on aging. Calibration standards not prepared from primary standards must be initially verified using a certified reference solution. Recommended wavelengths and calibration concentrations are listed in Table 1. Typical calibration standard combinations are given in Table 2.

7.8.2 Sodium Multi-point Calibration Standards - To determine elevated concentrations of Na in drinking water and drinking water supplies using the recommended wavelength, Na is determined using a multi-point analytical

calibration usable to 160 mg/L. Prepare in a mixture of 2% HNO3 (v/v) and 1%

- 22 -

HCl (v/v) a calibration blank and six calibration standards at concentrations of 5, 10, 20, 40, 80, and 160 mg/L, respectively. To create the multi-point calibration, curve-fit the response of the blank and standards and store as a computer file. This calibration is standardized before each period of analysis using the calibration blank and the mixed calibration standard solution containing 20 mg/L Na (Sect. 7.8.1). A new multi-point calibration should be prepared whenever there is a change in analytical performance caused by either a change in instrument hardware or operating conditions. (Of the 990 ground water samples analyzed in the National Inorganic Radionuclide Survey, Na was reported below 90 mg/L in 84% of the samples.)

7.9 Blanks - Four types of blanks are required for this method. The calibration blank is used in establishing the analytical curve, the laboratory reagent blank is used to assess possible contamination from the laboratory procedure, the laboratory fortified blank is used to assess routine laboratory performance and a rinse blank is used to flush the uptake system to reduce memory interferences.

7.9.1 The calibration blank is prepared by diluting 20 mL (1+1) HNO3 and 10 mL (1+1) HCl in a 500-mL PMP volumetric flask to volume with reagent water. Store the prepared blank solution to an acid-cleaned, never-used 500-mL FEP bottle. This bottle should be dedicated for reuse and storage of this solution.

7.9.2 The laboratory reagent blank (LRB) is prepared by carrying 50 mL of reagent water through the entire analytical procedure. The LRB must contain all the reagents in the same volumes as used in processing the samples.

7.9.3 The laboratory fortified blank (LFB) is prepared in the same manner as the LRB, and fortified by adding 1.0 mL of the fortifying solution (7.10.1) to 50 mL of LRB. The LFB must be carried through the entire analytical procedure. The analyte concentrations fortified in the LFB are as follows: 4 µg/L Be; 5 µg/L Cd; 6 µg/L Sb; 10 µg/L As; 15 µg/L Pb; 50 µg/L Ag, Mn, Se, Sn and V; 100 µg/L B, Cr and Ni; 200 µg/L Al; 300 µg/L Fe; 1000 µg/L Ba, Cu and Zn.

7.9.4 The rinse blank is prepared by acidifying reagent water in an acid-cleaned LDPE

bottle to concentrations of 2% (v/v) HNO3 + 2% (v/v) HCl.

7.10 Quality Control Solutions.

7.10.1 Fortifying solution - The fortifying solution is used to prepare the laboratory fortified blank (LFB) and laboratory fortified matrix (LFM) solutions. The fortifying solution should be prepared in a 100-mL PMP volumetric flask,

containing a mixture of 4 mL (1+1) HNO3 and 2 mL (1+1) HCl, by combining the following listed aliquot volumes of each stock standard and the low-level stock fortifying solution (7.10.2) and diluting to volume with reagent water: 5 mL Ba,

- 23 -

Cu & Zn and the lowlevel stock fortifying solution; 1.5 mL Fe; 1.0 mL Al; 0.5 mL B, Cr & Ni; and 0.25 mL Ag, Mn, Se, Sn & V. Store in a new, acid-cleaned LDPE bottle. The analyte concentrations in the fortifying solution are as follows: 50 mg/L Ba, Cu & Zn; 15 mg/L Fe; 10 mg/L Al ; 5.0 mg/L B, Cr & Ni; 2.5 mg/L Ag, Mn, Se, Sn & V; 0.75 mg/L Pb; 0.50 mg/L As; 0.30 mg/L Sb; 0.25 mg/L Cd; and 0.20 mg/L Be.

7.10.2 Low-level stock fortifying solution - The low-level stock fortifying solution is used to prepare the fortifying solution described in Section 7.10.1. The low- level stock fortifying solution is prepared in a 50-mL PMP volumetric flask,

containing a mixture of 2 mL (1+1) HNO3 and 1 mL (1+1) HCl, by combining the following listed aliquot volumes of each stock standard and diluting to volume with reagent water: 750 µL Pb (15 mg/L); 500 µL As (10 mg/L); 300 µL Sb (6 mg/L); 250 µL Cd (5 mg/L); and 200 µL Be (4 mg/L). (The concentration in parenthesis is that of the analyte in the low-level stock.) Store in a new, acid- cleaned LDPE bottle dedicated for reuse and repeated storage of this solution.

7.10.3 Instrument Performance Check (IPC) Solution - The IPC solution is used to periodically verify instrument performance during analysis. It should be prepared by combining method analytes at appropriate concentrations in the

same acid mixture (2% HNO3 + 1% HCl) as the calibration standards. Silver should be limited to < 100 µg/L; while Al and Fe should be made to a

concentration of 2 mg/L; Ca, Mg, and SiO2 made to 5 mg/L; and Na to a concentration of 10 mg/L. For all other analytes, a concentration of 200 µg/L is recommended. The IPC should be prepared in a PMP (metal-free plastic)

volumetric flask to avoid B and SiO2 contamination. Store the IPC solution in a new, acid-cleaned FEP bottle dedicated for reuse and repeated storage of this solution.

NOTE: If the instrument readout system incorporates the use of a dilution factor (0.5) to report original sample concentration prior to processing, and the IPC solution is analyzed in the same manner as the samples, the reported IPC concentrations will be half the concentrations listed above.

7.10.4 Quality Control Sample (QCS) - Analysis of a QCS is required for initial

and periodic verification of calibration standards or stock standard solutions in order to verify instrument performance. The QCS must be obtained from outside source different from the standard stock solutions and prepared in a PMP volumetric flask using the same acid mixture as the calibration standards. The concentration of the analytes in the QCS solution should be sufficient to meet the data quality objectives given in Section 9.2.4. The concentrations may range from 200 µg/L for sensitive analytes such as Be and Cd to > 2.0 mg/L for

Al, Ca, Fe, Mg, Na, and SiO2. However, the concentration of Ag should be limited

- 24 -

to 0.1 mg/L or less to ensure complete solubility and stability. The QCS solution should be stored in a new, acid-cleaned LDPE bottle and analyzed as needed to meet data-quality needs. A fresh solution should be prepared quarterly or more frequently as needed.

7.11 Plasma Solution - The plasma solution is used for determining the nebulizer argon flow rate or gas pressure that will produce the optimum net-signal-to-background noise (S-B/B) ratio needed for the most requiring analytes included in the method without degrading the performance of the other analytes. The two analytes that present the greatest challenge are Sb and As because of the low MCLs and limited analytical sensitivity. The combined 1 mg/L solution is prepared by adding 100 µL of the Sb stock standard (7.7.2)

and a 100 µL of the As stock standard (7.7.3) to a mixture of 4 mL (1+1) HNO3 and 2 mL (1+1) HCL in a 100-mL PMP volumetric and diluting to volume with reagent water. Store the solution in a new, acid-cleaned LPDE bottle for repeated use as necessary. 8.0 SAMPLE COLLECTION, PRESERVATION, AND STORAGE

8.1 For the determination of trace and water matrix elements in drinking water and drinking water supplies, samples are not filtered, but acidified with (1+1) nitric acid to a pH < 2 (3 mL of [1+1] acid per liter of sample should be sufficient). Preservation may be done at the time of collection; however, to avoid the hazards of strong acids in the field, transport restrictions and possible contamination, it is recommended that the samples be returned to the laboratory within two weeks of collection and acid preserved upon receipt in the laboratory. Following acidification, the sample should be mixed, held for sixteen hours, and then verified to be pH < 2 just prior to withdrawing an aliquot for sample processing. If for some reason, such as high alkalinity, the sample pH is verified to be > 2, more acid must be added and the sample held for sixteen hours until verified to be pH < 2. If properly preserved, the sample can be held up to 6 months.

8.2 If required by the data user, a field reagent blank (Sect. 3.4) should be prepared and analyzed in the same manner as a collected sample. Use the same type of container and acid as used in sample collection. 9.0 QUALITY CONTROL

9.1 Each laboratory using this method is required to operate a formal quality control (QC) program. The minimum requirements of this program consists of an initial demonstration of laboratory capability, and the periodic analysis of laboratory reagent blanks, fortified blanks and other laboratory solutions as a continuing check on performance. The laboratory is required to maintain performance records that define the quality of the data thus generated.

- 25 -

9.2 Initial Demonstration of Performance (mandatory).

9.2.1 The initial demonstration of performance is used to characterize instrument performance (determination of linear dynamic ranges and analysis of quality control samples) and laboratory performance (determination of method detection limits) prior to analyses conducted by this method.

9.2.2 Linear dynamic range (LDR) - The upper limit of the LDR must be established for each wavelength used for the analysis of the drinking water matrix analytes: Ca, Mg and silica. It must be determined from a linear calibration prepared using the established instrument operating conditions. The LDR should be determined by analyzing succeedingly higher standard concentrations of the analyte until the observed analyte concentration is no more than 10% below the stated concentration of the standard. Determined LDRs must be documented and kept on file. The LDR which may be used for the analysis of samples should be judged by the analyst from the resulting data. Determined sample analyte concentrations that are greater than 90% of the determined upper LDR limit must be diluted and reanalyzed. The LDRs should be verified as required for certification or whenever, in the judgment of the analyst, a change in analytical performance caused by either a change in instrument hardware or operating conditions would dictate they be re-determined.

9.2.3 Non-linear dynamic range - The upper limit of the non-linear calibration used for the determination of Na is the highest standard used to describe the calibration curve. The non-linear calibration must be established using the same instrument operating conditions used for analysis. Determined sample concentrations that are > 10% above the upper limit for Na must be diluted and reanalyzed. The upper limit should be verified as required for certification or whenever, in the judgment of the analyst, a change in analytical performance caused by either a change in instrument hardware or operating conditions would dictate they be re- determined.

9.2.4 Quality control sample (QCS) - When beginning the use of this method, on a quarterly basis, after the preparation of stock or calibration standard solutions, or as required to meet data-quality needs, verify the calibration standards and acceptable instrument performance with the preparation and analysis of a QCS (Sect. 7.10.4.) To verify the calibration standards, the determined mean concentrations from 3 analyses of the QCS must be within ± 5% of the stated values. If the calibration standard can not be verified, performance of the determinative step of the method is unacceptable. The source of the problem must be identified and corrected before either proceeding on with the initial determination of method detection limits or continuing with on-going analyses.

- 26 -

9.2.5 Method detection limit (MDL) - MDLs must be established for all wavelengths utilized for trace element determinations in total recoverable digestates. MDLs are determined using reagent water (blank) fortified to a concentration ranging from the instrument detection limit (IDL) to approximately two times the IDL(7) (see Table 4 for typical levels). To determine MDL values, take seven replicate aliquots of the fortified reagent water and process through the entire total recoverable analytical procedure. Perform all calculations defined in the method and report the concentration values in the appropriate units. Calculate the MDL as follows:

MDL = (t) x (S)

where: t = students' t value for a 99% confidence level and a standard deviation estimate with n-1 degrees of freedom (t = 3.14 for seven replicates). S = standard deviation of the replicate analyses. Note: If the relative standard deviation (RSD) from the analyses of the seven aliquots is < 10% and neither random nor reagent contamination is operative, the concentration used to determine the analyte MDL may have been inappropriately high for the determination. If so, this could result in the calculation of an unrealistically low MDL. In this case the MDL determination should be repeated using a lower concentration. Although data can be reported down to the MDL, the associated variability (RSD $ 30%) at the MDL concentration is very high. A more realistic reporting limit is the estimated upper limit of the 95% confidence interval about the MDL. This limit is based on the 97.5 percentile of chi square over associated 6 degrees of freedom and is computed by multiplying the MDL by a factor of 2.2(8). Typical single laboratory MDLs and reporting limits values using this method are given in Table 4.

The MDLs must be sufficient to meet data quality needs and detect analytes at the required levels according to compliance monitoring regulation (Sect. 1.2). Specifically, the determined MDLs for the analytes: As, Be, Cd, Sb, Se and Pb must be # 1/5 their respective MCL or action level before this method can be used for compliance monitoring. MDLs should be determined as required for laboratory certification, when a new operator begins work or whenever, in the judgment of the analyst, a change in analytical performance caused by either a change in instrument hardware or operating conditions would dictate they be re- determined.

9.3 Assessing Laboratory Performance (mandatory)

9.3.1 Laboratory reagent blank (LRB) - The laboratory must analyze at least one LRB (Sect. 7.9.2) with each batch of 20 or fewer samples. LRB data are used to assess contamination from the laboratory environment. LRB values that exceed

- 27 -

the MDL indicate laboratory or reagent contamination should be suspected. When LRB values for the trace analytes are above the calculated reporting limit (2.2 times the analyte MDL), fresh aliquots of the samples must be prepared and analyzed again for the affected analytes after the source of contamination has been corrected and acceptable LRB values have been obtained.

9.3.2 Laboratory fortified blank (LFB) - The laboratory must analyze at least one LFB (Sect. 7.9.3) with each batch of samples. Calculate accuracy as percent recovery using the following equation: 퐿퐹퐵 푅 = × 100 푠 where: R = percent recovery. LFB = laboratory fortified blank determined concentration. s = concentration equivalent of analyte added to fortify the LRB solution.

If the recovery of any analyte falls outside the required control limits of 90-110%, that analyte is judged to be out of control, and the source of the problem should be identified and resolved before continuing analyses.

9.3.3 The laboratory must use LFB analyses data to assess laboratory performance against the appropriate required control limits of 90-110% (see Sect.9.3.2). When sufficient internal performance data become available (usually a minimum of twenty to thirty analyses), optional control limits can be developed from the mean percent recovery (x) and the standard deviation (S) of the mean percent recovery. These data can be used to establish the upper and lower control limits as follows:

UPPER CONTROL LIMIT = x + 3S LOWER CONTROL LIMIT = x - 3S

The optional control limits must be equal to or better than the appropriate required control limits. After each five to ten new recovery measurements, new control limits can be calculated using only the most recent twenty to thirty data points. Also, the standard deviation (S) data should be used to establish an ongoing precision statement for the level of concentrations included in the LFB. These data must be kept on file and be available for review.

9.3.4 Instrument performance check (IPC) solution - The laboratory must initially and periodically verify that the instrument calibration is within required control limits. For all determinations the laboratory must analyze the IPC solution (Sect. 7.10.3) and a portion of the calibration blank immediately following calibration, after every tenth sample and at the end of the sample run. Analysis of the calibration

- 28 -

blank should always be less than the calculated reporting limit (2.2 times the analyte MDL) for the trace elements. Analysis of the IPC solution immediately following calibration must verify that the instrument is within ± 5% of calibration. Subsequent analyses of the IPC solution must be within ± 10% of calibration. If the calibration cannot be verified within the specified limits, reanalyze both the IPC solution and the calibration blank. If the second analysis confirms calibration to be outside the limits, sample analysis must be discontinued, the cause determined, corrected and/or the instrument recalibrated. All samples following the last acceptable IPC solution analysis must be reanalyzed. The analyses data of the IPC solution should be kept on file with the sample analyses data.

9.3.5 Spectral interference check (SIC) solution - For this method using the listed wavelengths, the specified background locations, and an instrument with first order resolution of 0.016 nm or better, verification of inter-element spectral interference is not required. However, if method flexibility, allowing the use of different wavelengths, spectral orders, and background correction locations requires an inter-element correction routine for spectral interference, it must be verified daily with the use of SIC solutions (see Sect. 4.1 for listed criteria and description of required testing).

9.4 Assessing Total Recoverable Analyte Recovery and Data Quality

9.4.1 Sample non-homogeneity and the chemical nature of the sample matrix can affect analyte recovery and the quality of the data. In the analysis of finished drinking water, these aspects are rarely an issue. However, source water for a drinking water supply can have varying turbidity. Taking separate aliquots from the sample for replicate and fortified analyses can, in some cases, assess the effect. Unless otherwise specified by the data user, laboratory or program, the following laboratory fortified matrix (LFM) procedure (Sect. 9.4.2) is required.

9.4.2 The laboratory must add a known amount of each analyte to a minimum of 10% of the routine samples. The LFM aliquot must be a duplicate of the aliquot used for sample analysis and fortified prior to sample preparation. The added analyte concentration must be the same as that used in the laboratory fortified blank (Sect. 7.9.3). Over time, samples from all routine sample sources should be fortified.

9.4.3 Calculate the percent recovery for each analyte, corrected for analyte background concentrations greater than the calculated reporting limit measured in the unfortified sample, and compare these values to the designated LFM recovery range of 85-115%. Percent recovery may be calculated using the following equation:

- 29 -

퐶푠 − 퐶 푅 = × 100 푠 where: R = percent recovery.

Cs = fortified sample concentration. C = sample background concentration (>2.2 X MDL). s = concentration equivalent of analyte added to fortify the sample.

9.4.4 If the recovery of any analyte falls outside the designated LFM recovery range, and the laboratory performance for that analyte is shown to be in control (Sect. 9.3), the recovery problem encountered with the fortified sample is judged to be matrix related, not system related. If the analyte in question is a primary contaminant (Sect. 1.1), under certain circumstances additional analyses may be required (see Sects. 1.6 and 1.7). For a primary contaminant not requiring additional analysis, for a secondary contaminant, or non-regulated analyte, the data user should be informed that the result for that analyte is suspect due to matrix effects.

9.4.5 Where reference materials are available, they should be analyzed to provide additional performance data. The analysis of reference samples is a valuable tool for demonstrating the ability to perform the method acceptably. 10.0 CALIBRATION AND STANDARDIZATION

10.1 Specific wavelengths and calibration concentrations are listed in Table 1. Other wavelengths may be substituted if they can provide the needed sensitivity and are corrected for spectral interference (see Sect. 4.1). However, because of the difference among various makes and models of spectrometers, specific instrument operating conditions are not required. The instrument and operating conditions utilized for determination must be capable of providing data of acceptable quality (see Sect. 1.12) to the drinking water program and data user. The analyst should follow the instructions provided by the instrument manufacturer; however, instrument operating conditions used to collect the single laboratory performance data included in this method are listed in Table 3 and are provided as a recommendation. Once operating conditions are established, it is intended that daily calibration will be accomplished using a calibration blank and a single analyte concentration.

10.2 Prior to using this method, optimize the plasma operating conditions. The purpose of plasma optimization is to provide a maximum net signal-to-background ratio (S-B/B) for the determination of As and Sb, the most requiring elements in the analytical array. The use of a mass flow controller to regulate or monitor the nebulizer gas flow rate greatly facilitates the procedure. The following procedure is recommended:

- 30 -

10.2.1 Ignite the plasma, and using the conditions listed in Table 3 as a guide, select appropriate incident rf power and plasma gas flows. Allow the instrument to become thermally stable before beginning. This usually requires approximately 30 minutes of operation. Set the aerosol argon flow rate through the nebulizer at approximately 650 mL per minute or at the instrument manufacturer’s recommended pressure setting if the flow rate can not be measured. Following the instrument manufacture’s instructions, optically profile the instrument to provide maximum signal for all wavelengths. While aspirating reagent water and using the As channel signal, adjust the horizontal and vertical position of the torch to provide minimum signal intensity. This should align the optics with the center of the sample channel of the plasma and minimize background noise.

10.2.2 After profiling the torch, aspirate the plasma solution (Sect. 7.11) and while following the instrument manufacturer's instructions, adjust the aerosol carrier gas flow rate through the nebulizer and collect signal intensity readings (S) at equal incremental flow settings on either side of the initial flow rate setting. Suggested flow rates (mL/min) settings are: 590, 600, 610, 620, 630, 640, 650, 660, 670, 680, 690 and 700. (NOTE: If nebulizer flow rates can not be measured, incremental pressure settings that control flow should be used.) After acid rinsing to eliminate any possible memory effect, repeat the same operation using an acid blank solution and collect the blank signal intensity readings (B) at the same respective flow settings. Calculate the S-B/B ratio for As and Sb at each flow setting. Plot, on the same graph, the calculated ratios and the blank intensity readings versus the argon flow rates. The intensity counts for the blank signal should decrease at a uniform rate as the argon flow rate increases, while the calculated S-B/B ratios for Sb should increase. At the lower flow rate settings, the As ratios should remain nearly constant; however, at some point the As ratio will start to decrease with an increase in flow rate. The flow rate where the As ratio begins to decrease (2% or more) is the limiting flow and the flow rate just prior to the limiting flow should be selected for routine operation. Record the nebulizer gas flow rate or pressure setting for future reference. If the nebulizer is replaced with a new or different nebulizer, repeat this optimization procedure.

10.2.3 After establishing the nebulizer gas flow rate, determine the solution uptake rate of the nebulizer in mL/min by aspirating a known volume calibration blank for a period of at least 3 minutes. Divide the spent volume by the aspiration time (in minutes) and record the uptake rate. Set the peristaltic pump to deliver the uptake rate in a steady even flow.

10.2.4 The final instrument operating condition, selected as being optimum, should provide acceptable instrument detection limits and method detection limits for all trace analytes. Refer to Table 4 for comparison of IDLs and MDLs, respectively.

- 31 -

10.2.5 Before daily calibration and after the instrument warmup period, the nebulizer gas flow should be reset to the determined optimized flow. If a mass flow controller is being used, it should be reset to the recorded optimized flow rate. In order to maintain reliable MDLs, the nebulizer gas flow rate should be the same from day-to-day (<2% change).

10.3 Before using the procedure (Sect. 11.0) to analyze samples, there must be data available documenting initial demonstration of performance. The required data and procedures are described in Sections 9.2.2, 9.2.3, 9.2.4, and 9.2.5. These data must be generated using the same instrument operating conditions and respective calibration routines used for sample analysis (see Sect. 11.2). These documented data must be kept on file and be available for review by the data user.

10.4 After completing the initial demonstration of performance, but before analyzing samples, the laboratory, if needed, must establish and initially verify the interelement spectral interference correction routine to be used during sample analysis. A general description concerning spectral interference and the analytical requirements for background correction are given in Sections 4.1 and 9.3.5. Once established, the entire routine must be verified on a daily basis by analyzing SIC solution(s) resulting in response data that falls within the 3-sigma control limits of the calibration blank of the analyte (Sect. 4.1). 11.0 PROCEDURE

11.1 Sample Preparation (Total Recoverable Digestion) - For the determination of trace analytes and water matrix elements in drinking water and source water supply, using a 50-mL PMP graduated cylinder, transfer a 50 mL (± 0.5 mL) aliquot from a well mixed, acid preserved

sample to a 50-mL clean digestion tube containing a mixture of 1.0 mL (1+1) HNO3 (Sect. 7.3.1) and 0.5 mL (1+1) HCl (Sect. 7.2.1). (The acids should be added to the digestion tube using an air displacement pipetter - see Sect. 6.6.) Place the digestion tube in the block digester (Sect. 6.4). (The block digester should be located in a clean fume hood.) Power the digestion block to preselected settings to evaporate the sample at a temperature of 95 °C (± 2 °C). Preconcentrate the sample until the volume has been reduced to approximately 25 mL. Cover the digestion tube with a plastic watch glass and reflux the sample for 30 minutes. (The time required to complete this step should approximate 2.5 h.) Once the refluxing step is complete, remove the digestion tube from the block digester and allow the sample to cool. When cool, using the volume gradation marks on the digestion tube, adjust the sample volume to 25 mL with reagent water (Sect. 7.4). Cap the digestion tube and mix. The sample is now ready for analysis. Because the effects of various matrices on the stability of analytes in low concentration can not be characterized, all analyses should be performed as soon as possible after the completed preparation.

- 32 -

11.2 Sample Analysis

11.2.1 Prior to daily calibration of the instrument, inspect the sample introduction system, including the nebulizer, torch, injector tube and uptake tubing, for salt deposits, dirt and debris that would restrict solution flow and affect instrument performance. Clean the system when needed or on a daily basis.

11.2.2 Configure the instrument system to the selected power and operating conditions as determined in Sections 10.1 and 10.2.

11.2.3 The instrument should be allowed to become thermally stable before calibration and analyses. This usually requires at least 30 minutes of operation. After instrument warmup, complete any required optical profiling or alignment routines particular to the instrument.

11.2.4 For initial and daily operation, calibrate the instrument according to the instrument manufacturer's recommended procedures, using mixed calibration standard solutions (Sect. 7.8.1) and the calibration blank (Sect. 7.9.1). A peristaltic pump must be used to introduce all solutions to the nebulizer. To allow equilibrium to be reached in the plasma, aspirate all solutions for 20 sec after reaching the plasma before beginning integration of the background corrected signal to accumulate data. To reduce measurement variance, use the average value of replicate integration periods of the signal to be correlated to analyte concentration. (Suggested data collection period for all determinations: 5 replicate 24 sec periods [8 sec on the wavelength peak and 8 sec on each BKGD location] = 120 sec.) Flush the system with the rinse blank (Sect. 7.9.4) for a minimum of 30 seconds (Sect. 4.4) between each standard.

11.2.5 After completion of the initial requirements of this method (Sects. 10.3 and 10.4), samples should be analyzed in the same operational manner used in the calibration routine with the rinse blank also being used between all sample solutions and quality control check solutions.

11.2.6 During sample analysis the laboratory must comply with the required quality control described in Sections 9.3 and 9.4.

11.2.7 Determined water matrix element concentrations that are 90% or more of the upper limit of the analyte LDR, or in the case of Na above the multipoint calibration range, must be diluted with reagent water that has been acidified in the same manner as the calibration blank and reanalyzed.

11.2.8 To ensure an accurate determination for compliance monitoring, a primary contaminant must be reanalyzed by either method of standard additions (see Sect.11.3), or by another approved method, when the concentration of that primary contaminant determined by the normal analytical routine (Sect. 11.2) is $

- 33 -

80% of the established MCL, or action level, and the required LFM analysis does not verify the absence of a matrix interference (see Sects. 1.6 & 9.4).

11.2.9 Report data as directed in Section 12.

11.3 If the method of standard additions (MSA) is used, standards are added at one or more levels to portions of a prepared sample. This technique (9) compensates for enhancement or depression of an analyte signal by a matrix. It will not correct for additive interferences such as contamination, inter-element interferences, or baseline shifts. This technique is valid in the linear range when the interference effect is constant over the range, the added analyte responds the same as the endogenous analyte, and the signal is corrected for additive interferences. The simplest version of this technique is the single-addition method. This procedure calls for two identical aliquots of the sample solution to be taken. To the first aliquot, a small volume of standard is added, while to the second aliquot, a volume of acid blank is added equal to the standard addition. The sample concentration is calculated by the following:

푆2 × 푉1 × 퐶

(푆1 − 푆2) × 푉2 where: C = Concentration of the standard solution (µg/L)

S1 = Signal for fortified aliquot

S2 = Signal for unfortified aliquot

V1 = Volume of the standard addition (L)

V2 = Volume of the sample aliquot (L) used for MSA

12.0 DATA ANALYSIS AND CALCULATIONS

12.1 Sample data for the water matrix elements (Ca, Mg and Na) and silica should be reported in units of mg/L. For compliance monitoring, total recoverable trace elements should be reported in the same units used to express the MCL or action level. If there is no established MCL, the trace element should be reported in µg/L.

12.2 For water matrix analytes, multiply the solution analyte concentrations by the dilution factor, 0.5, and report the data with allowance for sample dilution when analyte concentrations exceed 90% or more of the LDR upper limit, and in the case of Na when the analytical range is exceeded. Round the data to the thousandth place and report up to three significant figures. Do not report analyte concentrations below the IDL.

12.3 For total recoverable trace element analytes, multiply solution analyte concentrations by the dilution factor 0.5, round off the data values (µg/L) to the nearest tenths place and report analyte concentrations up to three significant figures. For drinking water

- 34 -

compliance monitoring, do not report data below the analyte reporting limit calculated from the laboratory determined MDL data (see Sect. 9.2.5). Typical MDLs and calculated reporting limits are given in Table 4.

12.4 The QC data obtained during the analyses provide an indication of the quality of the sample data and should be provided with the sample results. 13.0 METHOD PERFORMANCE

13.1 Listed in Table 4 are typical single laboratory total recoverable MDLs for the sample procedure given in Sections 11.1, followed by analysis using pneumatic nebulization. They were determined for the recommended wavelengths using simultaneous AVICP-AES and the operating conditions given in Table 3. The MDLs were determined in reagent blank matrix (best case situation) fortified with the respective analyte concentration also listed in Table 4.

13.2 Data obtained from single laboratory method testing are summarized in Tables 5 and 6. Table 5 lists precision (RSD) and average recovery data for SRM 1643c that was analyzed along with the drinking water samples listed in Table 6. The drinking water samples were prepared using the procedure given in Section 11.1. Table 6 lists data for 4 different tap water matrices (two well water supplies, one surface water supply, and a home cistern supply). Five unfortified aliquots were prepared to determine sample background concentrations and four aliquots for each LFM. For primary and secondary contaminants, the LFMs were fortified to a concentration equivalent to the respective analyte MCL. Data for the analysis of the water matrix elements and silica are listed at the bottom of sample data sheet. 14.0 POLLUTION PREVENTION

14.1 Pollution prevention encompasses any technique that reduces or eliminates the quantity or toxicity of waste at the point of generation. Numerous opportunities for pollution prevention exist in laboratory operation. The EPA has established a preferred hierarchy of environmental management techniques that places pollution prevention as the management option of first choice. Whenever feasible, laboratory personnel should use pollution prevention techniques to address their waste generation (e.g., Sect. 7.7). When wastes cannot be feasibly reduced at the source, the Agency recommends recycling as the next best option.

14.2 For information about pollution prevention that may be applicable to laboratories and research institutions, consult Less is Better: Laboratory Chemical Management for Waste Reduction, available from the American Chemical Society's Department of Government Relations and Science Policy, 1155 16th Street N.W., Washington D.C. 20036, or on-line at http://membership.acs.org/c/ccs/pub_9.htm.

- 35 -

15.0 WASTE MANAGEMENT

15.1 The Environmental Protection Agency requires that laboratory waste management practices be conducted consistent with all applicable rules and regulations. The Agency urges laboratories to protect the air, water, and land by minimizing and controlling all releases from hoods and bench operations, complying with the letter and spirit of any sewer discharge permits and regulations, and by complying with all solid and hazardous waste regulations, particularly the hazardous waste identification rules and land disposal restrictions. For further information on waste management consult The Waste Management Manual for Laboratory Personnel, available from the American Chemical Society at the address listed in the

Section 14.2. 16.0 REFERENCES

1. U.S. Environmental Protection Agency. Sample Preparation Procedure for Spectrochemical Determination of Total Recoverable Elements - Method 200.2, Revision 2.8, May 1994 (EPA-600/R-94/111).

2. U.S. Environmental Protection Agency. Determination of Trace Elements by Stabilized Temperature Graphite Furnace Atomic Absorption - Method 200.9, Revision 2.2, May 1994 (EPA-600/R-94/111).

3. Carcinogens - Working With Carcinogens, Department of Health, Education, and Welfare, Public Health Service, Center for Disease Control, National Institute for Occupational Safety and Health, Publication No. 77-206, Aug. 1977. Available from the National Technical Information Service (NTIS) as PB-277256.

4. OSHA Safety and Health Standards, General Industry, (29 CFR 1910), Occupational Safety and Health Administration, OSHA 2206, (Revised, January 1976).

5. Safety in Academic Chemistry Laboratories, American Chemical Society Publication, Committee on Chemical Safety, 3rd Edition, 1979.

6. American Society for Testing and Materials. Standard Specification for Reagent Water, D1193-77. Annual Book of ASTM Standards, Vol. 11.01. Philadelphia, PA, 1991.

7. Code of Federal Regulations 40, Ch. 1, Pt. 136 Appendix B.

8. Glaser, J.A.,D.L. Foerst, G.D. McKee, S.A. Quave, and W.L. Budde, “Trace Analyses for Wastewaters,” Environ. Sci. Technol., 15 (1981) 1426-1435.

9. Winefordner, J.D., Trace Analysis: Spectroscopic Methods for Elements, Chemical Analysis, Vol. 46, pp. 41-42.

- 36 -

17.0 TABLES, DIAGRAMS, FLOWCHARTS, AND VALIDATION DATA

TABLE 1. WAVELENGTHS, BACKGROUND CORRECTION LOCATIONS, AND RECOMMENDED CALIBRATION

Analyte Wavelengtha Location Calibrate b (nm) For BKGD. to

Correction (mg/L)

(nm)

Aluminum 308.215 +0.033 5 Antimony 206.833x2 -0.009 0.5 Arsenic 189.042x2 -0.009 0.5 Barium 493.409 +0.033 1 Beryllium 313.042 +0.033 0.1 Boron 249.678x2 +0.016 1 Cadmium 226.502x2 +0.016 0.1 Calcium 315.887 +0.033 20 Chromium 267.716 +0.033 0.5 Copper 324.754 +0.033 2 Iron 271.441 +0.033 5 Lead 220.353x2 +0.016 0.5 Magnesium 279.079 +0.033 20 Manganese 257.610 +0.033 0.5 Nickel 231.604x2 +0.016 0.5 Selenium 196.090x2 -0.009 0.5

Silica (SiO2) 251.611 +0.033 21.4 Silver 328.068 +0.033 0.1 Sodium 330.232 +0.033 20 Tin 189.980 -0.018 0.2 Vanadium 292.402 +0.033 0.5 Zinc 206.200 +0.033 0.5

aThe wavelengths listed in the noted spectral order are recommended because of their sensitivity and overall acceptability. Other wavelengths may be substituted if they can provide the needed sensitivity and are treated with the same corrective techniques for spectral interference (see Sect. 4.1). b Suggested concentration for instrument calibration. Other calibration limits in the linear ranges may be used.

- 37 -

200.5-29 TABLE 2. MIXED STANDARD SOLUTIONS

Solutiona Analytes

I As, Be, Cd, Pb, Sb, Se, V, and Zn

II III Ba, Cr, Fe, Mn, Ni, Sn, and SiO Al, Cu, Ca, Mg, and Na 2 Ag and B IV

a See Section 7.8.1

TABLE 3. AXIALLY VIEWED INDUCTIVELY COUPLED PLASMA INSTRUMENT OPERATING CONDITIONS

rf power 950 watts

Argon supply liquid argon

Argon pressure 60 psi

Coolant argon flow rate 20 L/min

Aerosol carrier 27 psi argon pressure 635 flow rate mL/min

Auxiliary (plasma)

argon flow rate 0.5 L/min

Sample uptake rate 1.6 controlled to mL/min

- 38 -

200.5-30 TABLE 4. DETECTION AND REPORTING LIMITS

INSTRUMENT(a) MDL METHOD CALCULATED(b) DETECT. LIMIT SPIKE DETECTION LIMIT REPORTING ANALYTE mg/L µg/L µg/L (MDL), µg/L LIMIT, µg/L

0.2

2 0.5

Ag 0.5 4.0 - 0.5 Al 2.5 - 0.03 0.2 As 0.8 4.9 B - 0.03 2.2 3.1 - 0.08 - 1.4 Ba 0.04 0.7 - 0.2 0.3 Be 0.2 Ca - - 0.2 0.05 0.4 0.1 Cd 0.02 0.2 0.02 - 0.4 Cr - 5 - 0.3 0.4 Cu - - 0.1 0.5 10 0.7 Fe - 0.07 0.2 0.3 - Mg - - 3.3 7.3 0.08 Mn 0.02 0.6 - 0.06 0.2 1.3 1.4 Na - 1 - 0.6 Ni 3.0 2.5 0.4 1 1.1 0.9 Pb 2.0 2.0 - Sb 2 1.3 Se - 3.0 2.9 SiO2 - - - - - Sn 1 1.0 0.5 1.1 0.01 V 0.2 0.6 0.2 0.5 - Zn - 0.4 0.6 0.4 0.9

a Instrument detection limits are used as reporting limits for matrix elements. b The listed calculated reporting limits have been rounded up to the tenths place to fully meet the 2.2 multiple criteria and to eliminate the listing of insignificant numbers. Because of rounding up to the tenths place, the reporting limits listed for Ba and Be are multiples of 4 and 5 times their respective MDLs.

- 39 -

200.5-31 TABLE 5. SRM(1643c) PRECISION AND ACCURACY DATA

NIST - 1643c DETERMINED AVERAGE CERTIFIED VALUE CONC. STD. RECOVERY RSD ANALYTE µg/L µg/L DEV. % %

Ag 2.21 ± 0.30 2.1 0.82 95% 3.9% Al 114.6 ± 5.1 125 5.1 109% 4.1% As 82.1 ± 1.2 83.7 1.9 102% 2.2% B 119.0 ± 1.4 114 0.5 96% 0.4% Ba 49.6 ± 3.1 49.2 0.52 99% 1.0% Be 23.2 ± 2.2 22.5 0.33 97% 1.5%

Cd 12.2 ± 1.0 11.9 0.13 98% 1.1% Cr 19.0 ± 0.6 18.0 0.21 103% 1.2% Cu 22.3 ± 2.8 22.9 0.74 99% 3.2%

Fe 106.9 ± 3.0 106 2.8 99% 2.7% Mn 35.1 ± 2.2 34.5 0.4 98% 1.2% Ni 60.6 ± 7.3 57.4 0.58 95% 1.0%

Pb 35.3 ± 0.9 34.4 1.2 97% 3.5% Se 12.7 ± 0.7 11.9 0.7 94% 5.9%

V 31.4 ± 2.8 28.3 0.15 93% 0.5% Zn 73.9 ± 0.9 74.4 0.52 101% 0.7%

NIST - 1643c DETERMINED AVERAGE CERTIFIED VALUE CONC. STD. RECOVERY RSD ANALYTE mg/L mg/L DEV. % %

Ca 36.8 ± 1.4 37.0 0.32 101% 0.9% 9.61 102% 1.6% Mg 9.45 ± 0.27 0.16 103% Na 12.6 0.7% 12.19 ± 0.36 0.09

200.5-32

- 40 -

TABLE 6. TRACE ELEMENT PRECISION AND RECOVERY DATA

TAP WATER - REGION 5 SURFACE WATER SUPPLY

______

SAMPLE AVERAGE CONC. STD. Spike RECOVERY RSD ANALYTE µg/L DEV. µg/L (%) (%)

100 1.2

Ag Al <0.5 18 1.3 200 1.5 100 105 As <3.1 - 10.0 105 4.2

B 34.8 37.7 0.9 100 101 100 1.5 <0.1 0.4 Ba 2000 100 1.3 - Be 4.0 1.4

Cd <0.3 <0.5 - 5.0 98 99 1.8 101 Cr 3.3 - 100 1.3

Cu 0.09 1000 1.3

Fe Mn <7.3 - 300 99 98 1.5 99 Ni 0.4 0.01 50 1.1 100 1.5 0.1 1.0

Pb <2.5 <2.0 - 15.0 100 100 2.2

Sb Se <2.9 - 6.0 104 5.3

50 2.0

Sn V <1.1 - 50 102 2.0

Zn <0.5 0.1 50 100 99 1.4

5.6 2000 1.0

< Sample concentration below the calculated reporting limit.

- 41 -

Analysis of Water Matrix Elements Sample RSD Analyte Conc. mg/L (%)

Ca 34.6 1.4 9.64 1.6 26.4 2.9 5.23 1.6 TAP WATER - REGION 5 WELL WATER SUPPLY

SAMPLE AVERAGE CONC. STD. Spike RECOVERY RSD ANALYTE µg/L DEV. µg/L (%) (%)

100 1.2

Ag Al - 200 97 103 1.1 <0.5 <4.9 As 17.7 0.3 10.0 101 1.2

B 96.0 107 0.7 100 102 97 0.9 0.7 Ba <0.1 2000 97 1.0 - Be 4.0 1.4

Cd <0.3 <0.5 - 5.0 96 1.9

Cr <0.7 - 100 93 1.3 1000 Cu 98 1.1

Fe Mn 552 2.7 300 103 95 1.1 0.1 94 Ni 15.0 <1.4 50 1.1 - 100 1.3

Pb <2.5 <2.0 - 15.0 97 97 4.7 101 Sb Se <2.9 - 6.0 7.4

50 1.2

- 42 -

Sn V <1.1 - 50 102 95 1.2 95 Zn <0.5 - 50 1.3 2000 6.2 0.36 1.5

< Sample concentration below the calculated reporting limit.

Analysis of Water Matrix Elements Sample RSD Analyte Conc. mg/L (%)

Ca 69.9 0.6

28.0 0.7

57.9 1.9

12.3 0.8 TAP WATER - REGION 6 WELL WATER SUPPLY

SAMPLE AVERAGE CONC. STD. Spike RECOVERY RSD ANALYTE µg/L DEV. µg/L (%) (%)

<0.5 - 100 100 1.2

Ag Al <4.9 - 200 102 1.6 As <3.1 - 10.0 101 2.3

B 56.4 203 1.2 100 104 100 1.4 3.6 Ba <0.1 2000 102 1.4 - Be 4.0 1.4

Cd <0.3 - 5.0 96 1.9

Cr 1.3 0.07 100 96 100 1.1

Cu 155 2.7 1000 1.3

- 43 -

Fe Mn 8.6 0.6 1.9 300 96 97 1.7 98 Ni 3.3 0.02 50 1.3 100 0.2 1.4

Pb <2.5 <2.0 - 15.0 105 98 2.5 100 Sb Se 4.1 0.5 6.0 3.6

50 1.1

Sn V 4.4 0.1 50 102 98 1.0 98 Zn 11.5 0.2 50 1.4 2000 26.4 0.5 1.4

< Sample concentration below the calculated reporting limit.

Analysis of Water Matrix Elements Sample RSD Analyte Conc. mg/L (%)

Ca 44.6 1.6 9.26 1.6 41.0 1.7 26.2 1.6

TAP WATER - CISTERN WATER SUPPLY

SAMPLE AVERAGE CONC. STD. Spike RECOVERY RSD ANALYTE µg/L DEV. µg/L (%) (%) 100 1.2

Ag Al <0.5 0.9 200 98 104 1.7 As 31.3 <3.1 - 10.0 102 7.8

B 6.6 0.2 100 100 98 1.2

Ba 3.3 <0.1 0.05 2000 96 1.4

- 44 -

Be - 4.0 1.2

Cd <0.3 - 5.0 94 95 1.1

Cr <0.5 1.5 100 99 1.1 1000 Cu 262 1.7

Fe Mn 13 1.9 300 95 94 1.3 95 Ni 0.6 <1.4 0.01 50 1.2 100 - 1.7

Pb <2.5 <2.0 - 15.0 101 1.6 103 95 Sb Se <2.9 - 6.0 3.2

50 1.2

Sn V <1.1 - 50 95 97 1.6 50 2000 95 Zn <0.5 14.2 3.4 1.3

1.2

< Sample concentration below the calculated reporting limit.

Analysis of Water Matrix Elements Sample RSD Analyte Conc. mg/L (%)

Ca 10.1 0.7 0.68 1.0 2.3 1.1 2.07 0.6

- 45 -

- 46 -

- 47 -

Health and Safety Plan

Title: Speciation of Metals Complexed to Humic Acid in Natural/Drinking Water

Principal Investigator(s): Eugenia C. Riddick

Office: ORD

Laboratory: NRMRL

Division: WSWRD

Branch: Treatment Technology Evaluation Branch

Building: AWBERC

Room/Lab #: 667

Estimated Research Completion Date (Month / Year): On going

Approvals

I have read and approve the attached Health and Safety Plan in conformance with the ORD Facility Chemical Hygiene Plan and Health & Safety Plan Policy.

Name Phone Signature Date

Eugenia C. Riddick 569-7507

PREPARER

Eugenia C. Riddick 569-7507

PRINCIPAL INVESTIGATOR

Darren Lytle 569-7432

BRANCH CHIEF

(Additional Approvals: such as CO-PI or Contractor Manager)

Steve Musson 569-7969

CHEMICAL HYGIENE OFFICER

- 48 -

Additional information on the completion of a Health & Safety may be found at the SHEM website – Link available at: http://cincinnati.epa.gov/services/safety/

- 49 -

Project Description

The Water Supply and Water Resources Division, Treatment Technology Evaluation Branch is focused on the current and future technical/engineering problems facing both the Office of Water and the American public relative to drinking water treatment. In support of this effort, the WSWRD, TTEB carries out in-house research projects at the A.W.Breidenbach Environmental Research Center. The main task of this in-house, preliminary research is to develop methodology for use in determining potential metal species preferentially bound/complexed to humic acid in source samples of drinking/natural waters. Because this research is preliminary, both SOPs and a project QAPP will be in place prior to formal continuation of research.

Methodology:

Metal speciation will be determined by voltammetric analysis utilizing a Princeton Applied Research static mercury drop electrode, model 303A, equipped with model 305 stirrer, via voltammetric analysis utilizing an Autolab modular potentiostat/galvanostat.

Field Activities (if applicable)

N/A

Laboratory Activities

Hanging Drop Mercury Electrode (HDME) Anodic Stripping Voltammetry (ASV)

Trace metal analyses are performed to determine the species of metal bound to humic substances in natural and drinking water samples. Samples are prepared in solutions of potassium nitrate buffered at pHs 5.0, 6.0 and 8.0 with either acetate, 2-(N-morpholino)ethanesulfonic acid or 3-(N-morpholino)propanesulfonic acid respectively. Nitrogen gas is bubbled through the sample for several minutes prior to analysis to remove dissolved oxygen and a layer of nitrogen gas is maintained above the solution for the remainder of the analysis. A droplet of mercury is dispensed from capillary tubing connected to the instrument and the potential of the working electrode

By reacting alkaline phenol and hypochlorite with ammonia, Indophenol blue is formed in proportion to the ammonia concentration, (Berthelot reaction). The blue sample color is then intensified with sodium nitroprusside and measured colorimetrically.

Cyanates and ogano-nitrogen compounds are not a significant source of false positive results in a drinking water matrix. Therefore boric acid and sodium tetraborate use as described in 350.1 is not necessary.

- 50 -

Standard preparation. Standards are prepared from dilutions of a primary standard solution purchased from chemical supply companies.

Sample preparation. Samples are prepared by aliquoting a portion of the sample, typically 2 mL, into autosampler vials.

Automated Colorimetry (PO4) EPA Method 365.1

Automated Colorimetry analyses are performed to determine the phosphorus in drinking water samples. Ammonium molybdate and antimony potassium tartrate react in an acid medium with dilute solutions of phosophorus to form an antimony-phospho-molybdate complex. The complex is reduced to an intensely blue- colored complex by ascorbic acid. The color is proportional to the phosphorus concentration. Polyphosphates may be converted to the orthophosphate form by manual sulfuric acid hydrolysis. Organic phosophorus compounds may be converted to the orthophosphate form by manual persulfate digestion and the resulting color is then measured automatically.

Method 365.1 requires ammonium persulfate, phenolphthalein and isopropyl alcohol. These chemicals are not required for this testing. In addition, sample boiling is not necessary within current protocols.

Standard preparation. Standards are prepared from dilutions of primary standard solutions purchased from chemical supply companies.

Sample preparation. Samples are prepared by aliquoting a portion of the sample, typically 2 mL, into autosampler vials.

System preparation. The Brij 35 surfactant can interfere with some PO4 determinations. Substituting Aerosol 22 or FL-70 surfactants can overcome this problem.

Automated Colorimetry (NO3/NO2) EPA Method 353.2

Automated Colorimetry analyses are performed to determine nitrate/nitrites in drinking water samples. For nitrite values, filtered samples are diazotized with sulfanilamide and coupled with N-(1-naphtyl) ethylenediamine dihydrochloride to form a highly colored azo dye that is then measured colorimetrically. Separately, for nitrate the filtered samples are passed through an Open Tubular Cadmium Reduction (OTCR) column to reduce nitrate to nitrite. The nitrate component of the sample is then determined by carrying out the above procedure for nitrite analysis. The cadmium metal of 353.2 has been replaced with the (OTCR).

- 51 -

Standard preparation. Standards are prepared from dilutions of primary standard solutions purchased from chemical supply companies. In addition, standards will be prepared with sodium nitrate and sodium nitrite vice the potassium compounds listed in the method and chloroform will not be used as listed in the method.

Sample preparation. Samples are prepared by aliquoting a portion of the sample, typically 2 mL, into autosampler vials.

System preparation. The OTCR is conditioned according to the OTCR preparation procedure in the attached appendix.

Glassware cleaning

Glassware used for the purposes of sample dilutions and reagent preparation are cleaned in a 10% solution of Contrad 70 followed by repeated rinsing in distilled water and then air dried.

Physical Hazards Summary

(The physical hazards marked below(X) have been identified as present during the performance of the project. Job hazards for specific steps are described in the Job Hazard Analysis Table at the end of the HASP.) The RSO shall be included in the list of reviewers/approval for all plans incorporating radioactive materials, radioactive devices, or radiation sources.

Physical Hazards Physical Hazards

Electrical Hazards Noise

Radioactive Materials Temperature

Non-Ionizing Radiation Illumination

Ionizing Radiation Compressed Gas X

Heavy Lifting Sharp Objects / Tools

Vibration Slips, Trips, Falls

UV light/radiation

Other (Specify)

PPE Summary

- 52 -

(The PPE items marked below(X) are required to be utilized during performance of the project. PPE requirements for specific steps are described in the Job Hazard Analysis Table at the end of the HASP.) Click here to review full PPE descriptions at OSHA's website

Face / Eye Protection

Safety Glasses w/ Side Shields X

Chemical Splash Goggles

Face Shield

Other (specify)

Ear Protection

Ear Plugs (Foam Inserts)

Both Ear Plugs and Ear Muffs

Ear Muffs

Other (specify)

Hand Protection

Chemical – Nitrile disposable exam X

Chemical - Latex disposable exam

Chemical – Butyl disposable exam

Chemical - Silver Shield®

Chemical – Ansel Barrier®

Cotton

Leather

Cut Resistant (Kevlar ®)

Other (specify)

Protective Clothing

Lab Coat X

- 53 -

Lab Apron

Jumpsuit

Shoe covers

Oversleeves

Other (specify)

Respiratory Protection

Employees Wearing Respiratory Protection must be enrolled in the Respiratory Protection Program must be medically cleared to wear a respirator and have current, annual training. Note: If respirator use is not required, this section may be deleted from the HASP.

The respirators marked below (X) are required to be utilized during performance of the project. Respirator requirements for specific project steps are described in the Job Hazard Analysis Table at the end of the HASP.

Nuisance Dusk Mask

N-95 Particulate Respirator

P-100 Particulate Respirator

Air Purifying Half Face Respirator

Air Purifying Full Face Respirator

Airline Supplied Air Respirator

SCBA

The following cartridges shall be used:

The cartridges shall be changed/removed from service on the following schedule:

- 54 -

Equipment Requirements

The safety equipment/engineering controls marked below(X) are required to be utilized during performance of the project. Requirements for specific steps are described in the Job Hazard Analysis Table at the end of the HASP.

Chemical Fume Hood X

Biological Safety Cabinet

Walk-in / Bulking Hood

Radiological Fume Hood

Balance Enclosure

Clear Air Bench (laminar flow hood

Spot Ventilation Unit (Snorkel)

Local Exhaust Ventilation X

Canopy Hood

Liquid Scintillation Counter

Refrigerator / Freezer

Deep Freezer

Other (specify)

Chemicals to be Used

Federal law under SARA, Title III, (also known as the Emergency Planning and Community Right to Know Act) requires that Material Safety Data Sheets (MSDS) be readily available for every chemical in the facility. EPA utilizes an online service to assist in meeting this requirement: http://epa.chemwatch.us User Name:epa; Password: cfr1910CW. If the MSDS is not available through Chemwatch, a hardcopy of the manufacturer supplied MSDS must accompany the HASP.

- 55 -

Item Chemical Name CAS# Project Use Disposal MSDS # Method for Availability: Unused Chemicals

Ex. Reagent, S = Sink C:Chemwatch Standard, or T = Trash H:Hardcopy Specific task # W = Chemical Waste Program

1 Aerosol 22 38916-42-6 Reagent W C

2 Ammonia Standard 12125-02-9 Standard W C

3 Ammonium Chloride 12125-02-9 Reagent W C

4 Ammonium Hydroxide 1336-21-6 Reagent W C

5 Ammonium Molybdate 12027-67-7 Reagent W C

6 Antimony Potassium Reagent W C Tartrate 28300-74-5

7 Ascorbic Acid 50-81-7 Reagent W C

8 Brij-35 9002-92-0 Reagent W C

9 Cupric Sulfate 7758-98-7 Reagent (OTCR) W C

10 Disodium EDTA 6381-92-6 Reagent W C

11 FL-70 Detergent (88.8% Reagent W C water) 7732-18-5

12 Hydrochloric Acid 7674-01-0 Reagent (OTCR) W C

13 N-(1- Reagent W C Naphtyl)ethylenediamine dihydrochloride 1465-25-04

14 Imidazole 288-32-4 Reagent (OTCR) W C

15 Nitrate Standard Standard W C Solution (Sodium nitrate) 7631-99-4

16 Nitric Acid 7697-37-2 Reagent (OTCR) W C

17 Nitrite Standard Solution Standard W C (Sodium nitrite) 7632-00-0

18 Phenol 108-95-2 Reagent W C

- 56 -

Item Chemical Name CAS# Project Use Disposal MSDS # Method for Availability: Unused Chemicals

Ex. Reagent, S = Sink C:Chemwatch Standard, or T = Trash H:Hardcopy Specific task # W = Chemical Waste Program

19 Phosphate Standard Standard W C Solution (Potassium phosphate monobasic) 7778-77-0

20 Phosphoric Acid 7664-38-2 Reagent W C

21 Sodium Hydroxide 1310-73-2 Reagent (OTCR) W C

22 Sodium Hypochlorite 7681-52-9 Reagent W C

23 Sodium Nitroprusside 13755-38-9 Reagent W C

24 Sodium Sulfite 7757-83-7 Reagent W C

25 Sodium Thiosulfate 10102-17-7 Reagent W C

26 Sulfanilamide 63-74-1 Reagent W C

27 Sulfuric Acid 7664-93-9 Reagent W C

28 Contrad 70 (Sodium Reagent W C hydroxide) 1310-58-3

Biological Research

Does the project in any way involve manipulation of recombinant DNA? No

If yes, are all proposed activities specifically exempted from the NIH Guidelines for Research Involving

Recombinant DNA Molecules?

Does the project in any way involve human subjects or biological materials obtained from human No subjects?

If yes, is the project exempt from the Health and Human Services Policy for Protection of Human Subjects?

Does the project involve animals requiring Institutional Animal Care & Use Committee (IACUC) approval? No (includes vertebrate & invertebrates animals)

- 57 -

Biological Agents

(The Biosafety Level (BSL) and Animal Biosafety Level (ABSL) refer to specific combinations of work practices, safety equipment, and facility design elements utilized to minimize exposure of workers and the environment to infectious agents. Principal Investigators must perform an agent risk assessment to determine the BSL.)

Item Vaccination

# Biological Agent (list all that apply) BSL # Source of Biological Agent Required?

Waste Management

 Sample waste stream from the instrument is collected in 2 L carboys. The waste stream is then adjusted to pH 6 – 8 with sufficient NaHCO3 or dilute (5%) HCl to neutralize the acid or base respectively. Final pH is checked with pH tests strips prior to sink disposal. Metal content has been determined (Trace Metal analysis, EPA Method 200.7, HASP 2009-80) to be below RCRA levels for sink disposal.

 Samples and standards represent unpreserved drinking water matrixes and are directly sink disposed.

 When expended the OTCR column must be disposed as hazardous waste through the chemical waste management program.

Will Hazardous Waste Be Generated? No Will the Treatability Exemption be Utilized? No

Sample Management

Samples are delivered to the lab from ongoing studies. Samples are logged in, assigned an unique internal laboratory ID number and information regarding sample type, location and PI is recorded. Samples are primarily raw and/or finished drinking water.

Samples analyzed at AWBERC are stored in a refrigerator located 6 Floor Lobby. Max holding time is six weeks at 4˚C.

- 58 -

Spill Response

Acid spills represent the primary laboratory hazard. Acid spill kits are located within the lab. Spent sorbent will be disposed of in accordance with kit manufacturer’s instructions. Waste carboys are positioned within spill control trays in case of overflows.

The chemical spill kit is located: (Provide lab # and location within the lab)

Lab 667, on shelving near the door.

The biological spill kit is located: (Provide lab # and location within the lab)

- 59 -

Authorized Personnel

Training and medical monitoring requirements will vary depending on the complexity and materials used in the process. Therefore, only personnel appropriately trained and monitored will be permitted to work under this plan. To be “authorized”, employees must have completed the training and screenings selected below.

Mandatory for all researchers

Initial Laboratory Safety X

Current Chemical Hygiene Plan Laboratory Safety Refresher X

Hazardous Waste Management (RCRA) X

Medical Surveillance X

Project/Task Dependent

Respiratory Protection

Biosafety / Blood borne Pathogens

Initial Field Safety and/or 8 hour field safety refresher training in the last 12 months

40 - hour HAZWOPER and/or 8 hour HAZWOPER refresher in the last 12 months

Hearing Protection

First Aid / CPR / AED

DOT Hazardous Materials Awareness/Shipment

Radiation Safety

Other (specify)

Forklift Operator

Climbing/Scaffolding

References:

- 60 -

- 61 -

Laboratory Staff Concurrence

I have read, understood and will comply with all the requirements of the attached Health and Safety Plan, MSDS’s, and the rules contained in the U. S. EPA- Facilities Chemical Hygiene Plan. I have also had the opportunity to ask any questions, and had my questions satisfactorily answered prior to my beginning work under this plan.

Government Other Date Name (Print) Signature Employee (Specify)

- 62 -

Job Hazard Analysis, Controls, and PPE

Recommended PPE Job Step/Operation Room / Area Potential Hazards/Risks Action/Procedure Required

Preparation of standard and Chemical splash due to pipetting – Use of required PPE as listed above. Gloves, eye protection, lab sample solutions 667 chemical burns – chemical –including chemical splash coats. dermatitis

Use of required PPE as listed above. Preparation of chemical Chemical splash due to pipetting – –including chemical splash Gloves, eye protection, lab reagents 667 chemical burns – chemical Use of CFH for pH buffer coats. dermatitis preparations

Chemical splash due to pipetting – Use of required PPE as listed above. Gloves, eye protection, lab Column preparation 667 chemical burns – chemical –including chemical splash coats. dermatitis

- 63 -

- 64 -

Health and Safety Plan

Title: Trace metal analysis of drinking Water

Principal Investigator(s): Keith C. Kelty

Office: ORD

Laboratory: NRMRL

Division: WSWRD

Branch: TTEB

Building: AWBERC

Room/Lab #: B21, B27

Estimated Research Completion Date (Month / Year): Ongoing

Approvals

I have read and approve the attached Health and Safety Plan in conformance with the ORD Facility Chemical Hygiene Plan and Health & Safety Plan Policy. I certify that the workplace hazards, routinely and non-routinely encountered by employees, during the described activities, and for which Personal Protective Equipment has been provided, have been assessed for the determination of Personal Protective Equipment appropriateness, in compliance with 29 CFR 1910 Subpart I.

Name Phone Signature Date

Keith C. Kelty 569-7414

PREPARER

Keith C. Kelty 569-7414

PRINCIPAL INVESTIGATOR

Darren Lytle 569-7432

BRANCH CHIEF

(Additional Approvals: such as CO-PI or Contractor Manager)

Steve Musson 569-7969

- 65 -

CHEMICAL HYGIENE OFFICER

Additional information on the completion of a Health & Safety Plan may be found at the SHEM website – Link available at: http://cincinnati.epa.gov/services/safety/

- 66 -

Project Description (Provide a brief synopsis/abstract of the research project. This can include background information and a general description of the research goals. Borrowing from Quality Assurance reports is acceptable and encouraged in each section of the description.)

Methodology:

These trace metal determinations are performed by Inductively Coupled Argon Plasma Atomic Emission Spectroscopy ( ICP-AES ) and Inductively Coupled Argon Plasma Mass Spectroscopy ( ICP-MS )/ Liquid Chromatography ICP-MS primarily utilizing EPA Methods 200.7 and 200.8 respectively.

Field Activities (if applicable) (Provide a description of the field activities to include methodology and processes used in the research and a description of the use of all chemical.)

N/A

Laboratory Activities (Provide a description of the laboratory activities to include methodology and processes used in the research and a description of the use of all chemicals. If referencing an SOP, please include a copy for review.)

Inductively Coupled Argon Plasma Atomic Emission Spectroscopy ( ICP-AES ) EPA Method 200.7

ICP-AES analyses are performed on drinking water samples to determine various metals concentrations in the ug/L to mg/L range. Samples are nebulized, the resultant aerosol is then thermally excited, by the Argon plasma, and the intensity of the spectral emissions are quantified by a spectrophotometer. The sample intensities are then compared to those of prepared standards for quantitation.

Standard preparation. Standards are prepared from dilutions of primary standard solutions purchased from chemical supply companies. Standards are preserved with 2% V/V HNO3 and HCl prior to analysis.

- 67 -

Sample preparation. Samples are prepared by aliquoting a portion of the sample, typically 5 mL, into autosampler tubes and preserved with 2% V/V HNO3 and HCl prior to analysis.

Inductively Coupled Argon Plasma Mass Spectroscopy ( ICP-MS ) EPA Method 200.8

ICP-MS analyses are performed on drinking water samples to determine various metals concentrations in the ng/L to ug/L range. Samples are nebulized, the resultant aerosol is then thermally excited, by the Argon plasma, and the ionized metals are quantified by a mass spectrophotometer. The sample mass intensities are then compared to those of prepared standards for quantitation.

Standard preparation. Standards are prepared from dilutions of primary standard solutions purchased from chemical supply companies. Standards are preserved with 2% V/V HNO3 prior to analysis.

Sample preparation. Samples are prepared by aliquoting a portion of the sample, typically 5 mL, into autosampler tubes and preserved with 2% V/V HNO3 prior to analysis.

Liquid Chromatography Inductively Coupled Argon Plasma Mass Spectroscopy (LC- ICP-MS ) EPA Method 200.8(modified)

LC-ICP-MS analyses are performed on drinking water samples to determine oxidative speciation of various metal concentrations in the ng/L to ug/L range. Samples are speciated chromatographically then nebulized, the resultant aerosol is then thermally excited, by the Argon plasma, and the ionized metals are quantified by a mass spectrophotometer. The sample mass intensities are then compared to those of prepared standards for quantitation. The chromatographic separation process typically consists of separation on C18, anion or cation columns with an ion pair, buffered mobile phase. The ion pairs are typically 5mM solutions of (Tetramethyl, Tetrapropyl or Tetrabutyl)-Ammonium Hydroxide with the buffers consisting 10 – 20 mM concentrations of either Ammonium Carbonate, Ammonium Phosphate (Mono and Dibasic), or Ammonium Sulfate at a pH range of pH 6 – 8. Phosphoric acid is used for final pH adjustments. Columns are stored in 50% – 90% solutions of Methanol/water or Acetonitrile/water.

Standard preparation. Standards are prepared from dilutions of primary standard solutions purchased from chemical supply companies. Standards are preserved with 2% V/V HNO3 prior to analysis.

Sample preparation. Samples are prepared by aliquoting a portion of the sample, typically 5 mL, into autosampler tubes and preserved with 2% V/V HNO3 prior to analysis. Sample metal speciation is preserved with EDTA, Hydrochloric acid or Nitric acid prior to analysis.

Operation

- 68 -

The following tasks are performed in day to day operation of the laboratory:

1. Standards preparation. Serial dilutions of purchased standard solutions. 2. Sample preparation. Dilution, spiking and matrix adjustments of submitted samples. Matrix adjustments are primarily pH adjustments with Nitric and/or Hydrochloric acid to a 0.25M concentration. 3. Analysis of the standards and solutions by automated ICP instrumentation.

- 69 -

Physical Hazards Summary

Physical Hazards Physical Hazards

Electrical Hazards X Noise

Radioactive Materials Temperature X

Non-Ionizing Radiation Illumination

Ionizing Radiation Compressed Gas X

Heavy Lifting Sharp Objects / Tools

Vibration Slips, Trips, Falls

UV light/radiation X

Other (Specify)

PPE Summary

Face / Eye Protection

Chemicals/Dust: Safety Glasses w/ Side Shields X

Chemicals/Dust: Chemical Splash Goggles

Chemicals/Compressed Gas: Face Shield

Welding Face Shields/Helmets

Other (specify)

Ear Protection

Ear Plugs (Foam Inserts)

Both Ear Plugs and Ear Muffs

Ear Muffs

- 70 -

Other (specify)

Hand Protection

Chemical – Nitrile disposable exam X

Chemical - Latex disposable exam

Chemical – Butyl disposable exam

Chemical - Silver Shield®

Chemical – Ansell Barrier®

Temperature – Thermal Gloves

Temperature – Cryogen Gloves X

Electric Arc/Voltage Protection

General Use/Work Gloves – Cotton

General Use/Work Gloves - Leather

Cut Resistant (Kevlar ®)

Other (specify)

Protective Clothing

Lab Coat X

Lab Apron

Jumpsuit/Coveralls

Traffic Safety Vests

Shoe covers

Safety Shoes: Steel Toe Boots and Shoes

Safety Shoes: Metatarsal Boots

Safety Shoes: Slip Resistant Boots and Shoes

Oversleeves

Other (specify)

- 71 -

Respiratory Protection

N-95 Particulate Respirator

P-100 Particulate Respirator

Air Purifying Half Face Respirator

Air Purifying Full Face Respirator

Airline Supplied Air Respirator

SCBA

Powered Air Purifying Respirator (PAPR)

The following cartridges shall be used:

N/A

The cartridges shall be changed/removed from service on the following schedule:

N/A

- 72 -

Equipment Requirements

Chemical Fume Hood

Biological Safety Cabinet

Walk-in / Bulking Hood

Radiological Fume Hood

Balance Enclosure

Clear Air Bench (laminar flow hood)

Spot Ventilation Unit (Snorkel)

Local Exhaust Ventilation X

Canopy Hood

Liquid Scintillation Counter

Refrigerator / Freezer

Deep Freezer

Other (specify): Fluid Pumps X

Chemicals to be Used

EPA utilizes an online service, Chemwatch, to provide Material Safety Data Sheets (MSDS) to employees. http://jr.chemwatch.net/chemgold3 User Name:epa; Password: cfr1910CW. If the MSDS is not available through Chemwatch, a hardcopy of the manufacturer supplied MSDS must be submitted to the SHEM office for upload to the Chemwatch system. . ALL fields must be completed in the table below for all chemicals used in the project.

- 73 -

Item Chemical Name CAS# Project Use Disposal MSDS # Method for Availability: Unused Chemicals

Ex. Reagent, S = Sink C:Chemwatch Standard, or T = Trash H:Hardcopy Specific task # W = Chemical Waste Program

R = Return to Vendor

1 Nitric Acid (HNO3) Sample W C 7697-37-2 Preservation

2 Hydrochloric Acid (HCl) Sample W C 7647-01-0 Preservation

3 Sodium bicarbonate pH adjustment, S C (NaHCO3) 144-55-8 Waste neutralization

4 Hydrogen peroxide Sample W C 7722-84-1 preparation

5 Aluminum 7429-90-5 Standard W C

6 Antimony 7440-36-0 Standard W C

7 Arsenic 7440-38-2 Standard W C

8 Barium 7440-39-3 Standard W C

9 Beryllium 7440-41-7 Standard W C

10 Bismuth 7440-69-9 Standard W C

11 Boron 7440-42-8 Standard W C

12 Bromine 7726-95-6 Standard W C

13 Cadmium 7440-43-9 Standard W C

14 Calcium 7440-70-2 Standard W C

15 Cerium 7440-45-1 Standard W C

16 Cesium 7440-46-2 Standard W C

17 Chromium 7440-47-3 Standard W C

- 74 -

Item Chemical Name CAS# Project Use Disposal MSDS # Method for Availability: Unused Chemicals

Ex. Reagent, S = Sink C:Chemwatch Standard, or T = Trash H:Hardcopy Specific task # W = Chemical Waste Program

R = Return to Vendor

18 Cobalt 7440-48-4 Standard W C

19 Copper 7440-50-8 Standard W C

20 Gallium 7440-55-3 Standard W C

21 Germanium 7440-56-4 Standard W C

22 Iron 7439-89-6 Standard W C

23 Indium 7440-74-6 Standard W C

24 Iodine 7553-56-2 Standard W C

25 Lead 7439-92-1 Standard W C

26 Lithium 7439-93-2 Standard W C

27 Lutetium 7439-94-3 Standard W C

28 Magnesium 7439-95-4 Standard W C

29 Manganese 7439-96-5 Standard W C

30 Mercury 7439-97-6 Standard W C

31 Molybdenum 7439-98-7 Standard W C

32 Multi-elemental solution none Standard W C

33 Nickel 7440-02-0 Standard W C

34 Phosphorus 7723-14-0 Standard W C

35 Platinum 7440-06-4 Standard W C

36 Potassium 7440-09-7 Standard W C

37 Scandium 7440-20-2 Standard W C

- 75 -

Item Chemical Name CAS# Project Use Disposal MSDS # Method for Availability: Unused Chemicals

Ex. Reagent, S = Sink C:Chemwatch Standard, or T = Trash H:Hardcopy Specific task # W = Chemical Waste Program

R = Return to Vendor

38 Selenium 7782-49-2 Standard W C

39 Silica 7631-86-9 Standard W C

40 Silver 7440-22-4 Standard W C

41 Sodium 7440-23-5 Standard W C

42 Strontium 7440-24-6 Standard W C

43 Sulfur 7704-34-9 Standard W C

44 Thallium 7440-28-0 Standard W C

45 Tin 7440-31-5 Standard W C

46 Titanium 7440-32-6 Standard W C

47 Tungsten 7440-33-7 Standard W C

48 Uranium 7440-61-1 Standard W C

49 Vanadium 7440-62-2 Standard W C

50 Yttrium 7440-65-5 Standard W C

51 Zinc 7440-66-6 Standard W C

52 Tetramethyl Ammonium Reagent W C Hydroxide 75-59-2

53 Tetrapropyl Ammonium Reagent W C Hydroxide 4499-86-9

54 Tetrabutyl Ammonium Reagent W C Hydroxide 2052-49-5

55 Ammonium Carbonate 506-87-6 Reagent W C

- 76 -

Item Chemical Name CAS# Project Use Disposal MSDS # Method for Availability: Unused Chemicals

Ex. Reagent, S = Sink C:Chemwatch Standard, or T = Trash H:Hardcopy Specific task # W = Chemical Waste Program

R = Return to Vendor

56 Ammonium Monobasic Reagent W C Phosphate 7722-76-1

57 Ammonium Dibasic Reagent W C Phosphate 7783-28-0

58 Ammonium Sulfate 7783-20-2 Reagent W C

59 Ammonium Hydroxide 1336-21-6 Reagent W C

60 Sample W C Phosphoric acid 7664-38-2 Preservation

61 Methanol 67-56-1 Reagent W C

62 Acetonitrile 75-05-8 Reagent W C

63 Sample W C EDTA 60-00-4 Preservation

64 Vacuum Pump Oil 68037-01-4 Vacuum Pump W C

65 Citranox Mixture Cleaner S C

66 Lab Detergent Mixture Cleaner S C

67 Lead Carbonate (PbCO3) 598-63-0 Standard W C

68 Lead Oxide (PbO2) 1309-60-0 Standard W C

69 Lead Phosphate Standard W C Pb3(PO4)2 7446-27-7

70 Copper Oxide (CuO) 1344-70-3 Standard W C

71 Liquid Argon 7440-37-1 Reagent R C

Biological Research

- 77 -

Does the project in any way involve manipulation of recombinant DNA? No

If yes, are all proposed activities specifically exempted from the NIH Guidelines for Research Involving N/A Recombinant DNA Molecules?

Does the project in any way involve human subjects or biological materials obtained from human No subjects?

If yes, is the project exempt from the Health and Human Services Policy for Protection of Human Subjects? N/A

Does the project involve animals requiring Institutional Animal Care & Use Committee (IACUC) approval? No (includes vertebrate & invertebrates animals)

Biological Agents

Item Vaccination

# Biological Agent (list all that apply) BSL # Source of Biological Agent Required?

Waste Management (Identify process/research derived samples and wastes and indicate the intended disposal method. Hazardous Waste identification and Treatability study exemptions per 40 CFR Part 261 as reviewed in annual SHEM RCRA training.)

Will Hazardous Waste Be Generated? Will the Treatability Exemption be Utilized?

 Sample waste stream from instrument is collected in 5 gal carboys containing sufficient NaHCO3 to neutralize acids as they are collected. Carboys are pH checked to ensure pH is between 6 and 10 prior to sink disposal. Sample waste stream will be monitored for RCRA metals As,Ag, Ba, Cd, Cr, Pb Hg and Se and analytes are all below 0.1 mg/L.  Standards are diluted as necessary and neutralized to a pH between 6 and 10 prior to sink disposal.

- 78 -

 Hydrogen peroxide (50uL) is used occasionally for samples with detectable hydrogen sulfide. These treated samples may be sink disposed. Concentrated hydrogen peroxide will be disposed through the SHEM chemical waste program.  Acidified samples meeting RCRA limits may be neutralized to a pH between 6 and 10 and sink disposed.  Chromatography gradient solutions containing 10 – 20 mM levels of buffers and ion pair reagents and meeting RCRA limits, will be neutralized to a pH between 6 and 10 prior to sink disposal.

Sample Management (Explain how samples will be identified and labeled for storage (if not immediately discarded) and eventual disposal. Sample contents must be clearly displayed. Include storage location and how long samples must be retained).

Samples are delivered to the lab from ongoing studies.

Samples analyzed at AWBERC are stored in B-22. Max holding time is six month at room temperature.

Spill Response (Describe procedures for managing spills of specific hazardous chemicals, both small and large. General spills may be addresses by reference to the Chemical Hygiene Plan).

Spills will be handled per the ORD-Cincinnati Chemical Hygiene Plan.

Acid spills represent the primary laboratory hazard. Spent sorbent will be collected disposed of through the SHEM chemical waste management program. Waste carboys will be positioned within spill control trays in case of overflows.

The chemical spill kit is located: (Provide lab # and location within the lab)

B21 – Beside of sink

- 79 -

B22 – Under lab sink

B27 – Under lab bench, left of sink

The biological spill kit is located: (Provide lab # and location within the lab)

N/A

- 80 -

Authorized Personnel

Mandatory for all researchers

Initial Laboratory Safety X

Current Chemical Hygiene Plan Laboratory Safety Refresher X

Hazardous Waste Management (RCRA) X

Medical Surveillance X

Project/Task Dependent

Respiratory Protection

Biosafety / Blood borne Pathogens

Initial Field Safety and/or 8 hour field safety refresher training in the last 12 months

40 - hour HAZWOPER and/or 8 hour HAZWOPER refresher in the last 12 months

Hearing Protection

First Aid / CPR / AED

DOT Hazardous Materials Awareness/Shipment

Radiation Safety

Other (specify)

EPA Driver’s Training

EPA Boat Safety Training

EPA Nanomaterials Health and Safety Awareness Training

References:

- 81 -

- 82 -

Laboratory Staff Concurrence

Government Other Date Name (Print) Signature Employee (Specify)

- 83 -

Job Hazard Analysis, Controls, and PPE

Job Room / Potential Recommended PPE Step/Operation Area Hazards/Risks Action/Procedure Required

Nitrile gloves, Safety glasses Dilution, Acidify samples and (ANSI Z87), Acidification, and Chemical Exposure, Handle and pour Laboratory Coats; Pipetting of Water B21/27 Acid burns to skin and concentrated acid use double gloves Samples eyes solutions in a chemical when handling fume hood. concentrated acids or neoprene gloves

Nitrile gloves, Safety glasses Dilution, (ANSI Z87), Handle and pour Acidification, and Chemical Exposure, Laboratory Coats; concentrated acid Pipetting of B21/27 Acid burns to skin and use double gloves solutions in a chemical Standard eyes when handling fume hood. Solutions concentrated acids or neoprene gloves

Nitrile gloves, UV Wear UV protective Safety glasses ICP Analysis B21/27 UV emission from ICP eyewear when viewing (ANSI Z87), plasma. Laboratory Coats

Cryogenic Liquid See cryogenic liquid See cryogenic liquid See cryogenic Handling (Liquid B21/27 JHA JHA liquid JHA Argon)

- 84 -

Cryogenic Liquid Handling

Sequence of Potential Hazards Recommended Basic Job Action or Procedure Steps

- Read MSDS of cryogen and any SOP’s related to the - Potential for ice formation; tasks involving the use of frostbite, burns, etc. General cryogens - Inadequate ventilation coupled - Ensure cylinders rented or with leakage or release causing purchased from gas potential oxygen deficiency suppliers meet regulations - Exposure to skin for a short period - Liquid nitrogen should never is not likely to cause injury, be used except in a well- ventilated area. This is however, prolonged contact will especially true when filling a result in freezing warm container or transfer - Contact with items that have been tube or inserting a warm cooled by the cryogenic material object, as large volumes of could cause damage to skin if in nitrogen gas is evolved. contact - Ensure appropriate PPE including standard laboratory PPE with the addition of cryogenic safety gloves and a face shield - Only containers or fittings (pipes, tongs etc.) that have Dispensing of liquid - extremely low temperature can been designed specifically nitrogen into smaller cause severe burn-like damage to for use with cryogenic the skin either by contact with the liquids may be used vessels fluid, surfaces cooled by the fluid or - ensure proper PPE is worn for material handling evolving gases - Failure of glass apparatus or brittle (minimum of protective failure of items cooled by liquid eyewear with face shield, laboratory coat, and nitrogen appropriate cryogenic safety - Potential breakage of glassware gloves) from cracks or defects – cuts / - Care must be taken with lacerations / contamination gloves, wrist-bands or bracelets which may trap liquid nitrogen close to the

skin - if any glassware is broken– only remote means should be used to pick up any broken glass

- 85 -

Cryogenic Liquid Handling

Sequence of Potential Hazards Recommended Basic Job Action or Procedure Steps

Filling dewars - potential for thermal stress on - Warm dewars should be containers resulting in cracks, filled slowly to reduce splintering, etc temperature shock - exposure to concentrated nitric effects and to minimize acid – acid burns from skin or splashing contact with mucous membranes – - All glass Dewars must be vaporous exposure to respiratory protected against the tract causing burns possibility of flying glass fragments, arising from failure by mechanical or temperature stress damage - be careful not to spill the material during transfer - Wear appropriate PPE such as eye protection and thermal gloves as listed above

Immersing objects into cryogens or use of - Wear appropriate PPE cryogens on cold fingers - soft materials such as rubber and such as eye protection (vacuum lines) and thermal gloves as plastics become brittle when listed above cooled by liquid nitrogen and may - Only people trained on shatter unexpectedly the use of a vacuum line - liquid oxygen may condense in should use cryogens for containers of liquid nitrogen or cooling cold fingers or vessels cooled by liquid nitrogen traps

- 86 -

Cryogenic Liquid Handling

Sequence of Potential Hazards Recommended Basic Job Action or Procedure Steps

forming a potential explosive - When necessary, wrap reaction with oxidizer glass/dewars with black - thermal stress damage can be tape to avoid injury caused to containers because of should the vessel shatter large, rapid changes of

temperature

Transporting cylinders - Pressure relief valves can off gas - proper personal and cause potential harm to ears protective equipment and / or face depending on shall be worn as listed proximity above; safety shoes - Tipping hazard; could cause a should be utilized when crushing / caught between objects moving cylinders injury - cylinders should be - Ensure that the cylinders are transported on carts to transported with caps on where increase stability applicable - ensure to point the pressure relief valves away from your body while in transport Cryogenic Liquid Handling:

General Precautions:

 Fire hazards may be greatly increased when gases normally thought to be non-flammable are used. Liquid nitrogen or liquid helium are capable of condensing oxygen from the atmosphere causing oxygen enrichment in unsuspected areas. Extremely cold metal surfaces are also capable of condensing oxygen from the atmosphere  When there is a possibility of personal contact with a cryogenic fluid, full face protection, an impervious apron or coat, cuffless trousers, and high-topped shoes should be worn. Watches, rings, bracelets, or other jewelry should not be permitted when personnel are working with cryogenic fluids. Basically, personnel should avoid wearing anything capable of trapping or holding a cryogenic fluid in close proximity to the flesh  The large expansion ratio from liquid to gas provides a source for the build-up of high pressures causing a potential for catastrophic release and subsequent displacement of ambient air in confined areas  Use caution whenever handling and glassware or other materials that have come in contact with a cryogenic liquid – the material will be very extremely cold or have that potential where skin

- 87 -

Cryogenic Liquid Handling

Sequence of Potential Hazards Recommended Basic Job Action or Procedure Steps

injury could occur on contact; there is also a potential for stress on that material causing brittleness, cracking, etc  All equipment should be kept very clean to avoid reactions or contamination  Ensure control of ignition sources  Ensure that all applicable MSDSs and this Job Hazard Analysis are reviewed prior to handling any cryogenic liquid  Cryogenic safety gloves can be purchase through many vendors. Some examples are listed below:

http://www.cryogloves.com/

http://www.airgas.com/browse/product.aspx?Msg=RecID&recIds=183074

http://www.labsafety.com/search/cryogen+safety+gloves/56365/

- 88 -