Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Capillary electrophoresis-mass spectrometry and complementary approaches for semi-quantitative analysis… MacLennan, Matthew S. 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_november_maclennan_matthew.pdf [ 3.91MB ]
Metadata
JSON: 24-1.0371931.json
JSON-LD: 24-1.0371931-ld.json
RDF/XML (Pretty): 24-1.0371931-rdf.xml
RDF/JSON: 24-1.0371931-rdf.json
Turtle: 24-1.0371931-turtle.txt
N-Triples: 24-1.0371931-rdf-ntriples.txt
Original Record: 24-1.0371931-source.json
Full Text
24-1.0371931-fulltext.txt
Citation
24-1.0371931.ris

Full Text

  CAPILLARY ELECTROPHORESIS-MASS SPECTROMETRY AND COMPLEMENTARY APPROACHES FOR SEMI-QUANTITATIVE ANALYSIS OF SAMPLES WITH COMPLEX MATRICES by  Matthew S. MacLennan  B.Sc. Honours, St. Francis Xavier University, 2008  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Chemistry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  September 2018  © Matthew S. MacLennan  ii  Committee The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:  CAPILLARY ELECTROPHORESIS-MASS SPECTROMETRY AND COMPLEMENTARY APPROACHES FOR SEMI-QUANTITATIVE ANALYSIS OF SAMPLES WITH COMPLEX MATRICES  submitted by Matthew S. MacLennan  in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Chemistry  Examining Committee: David D.Y. Chen, Chemistry Co-supervisor John V. Headley, Chemistry Co-supervisor   Supervisory Committee Member Ed Grant, Chemistry University Examiner Suzana Straus, Chemistry University Examiner  Additional Supervisory Committee Members:  Supervisory Committee Member  Supervisory Committee Member iii  Abstract Capillary electrophoresis-electrospray ionization-mass spectrometry (CE-ESI-MS) combines the superior separation efficiency of CE with the detection capability of MS to provide information-rich data on samples with complex matrices. CE-ESI-MS is applied to a range of complex mixtures to semi-quantify unknown components and find patterns among groups of samples. The first part of this thesis tells how the industrial generation of oil sands process-affected waters (OSPW) poses an environmental threat, and a review is given on the state-of-the-art of analytical chemistry in the semi-quantification of toxic naphthenic acid fraction compounds (NAFC) in OSPW. The chapter ends with a technical description of the CE-ESI-MS system which was utilized. CE-ESI-MS was demonstrated to produce effective analyses of a well-known complex mixture: human urine. CE-ESI-MS was used to quantify and “fingerprint” components in human urine via targeted and untargeted analyses of the sub-5 kDa urine metabolome of patients with prostate and/or bladder cancer. For targeted analysis, endogenous levels of sarcosine and 5 other metabolites were quantified in four patients and in a pooled healthy urine sample. An untargeted analysis of patient urine was also performed identifying over 400 distinct molecular features per patient.   Next, a CE-ESI-MS method was developed for the analysis of a relatively unknown complex mixture: NAFCs in OSPW. A standard mixture of amine-derivatized naphthenic acids was analyzed in under 15 min, detecting NAFCs between m/z 250 and 800. Derivatization of NAFCs consisted of two-step amidation reactions mediated by 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and N-hydroxysuccinimide in three different iv  solvents. The optimum BGE composition was determined to be 30% (V/V) methanol in water and 2% (V/V) formic acid.  The optimized CE-ESI-MS method was then used to find patterns in NAFCs across porewater samples impacted by oil sands waste. CE-ESI-MS, Orbitrap MS, and a standard GC-FID method were used to characterize porewater samples. Differences in measured amounts of total petroleum hydrocarbon by GC-FID and NAFC by HRMS indicate that the two methods provide complementary information about dissolved organic species in water leachate samples. CE-ESI-MS also provides complementary information and is a feasible and practical option for source evaluation of NAFCs in water.  v  Lay Summary A method was created to chemically fingerprint complex mixtures using a technique called capillary electrophoresis-electrospray ionization-mass spectrometry (CE-ESI-MS), combined with statistical data analysis. The first complex mixture that was studied was human urine, which demonstrated the suitability of CE-ESI-MS. The second complex mixture that was studied was naphthenic acids in process water. The Alberta oil sands industries use large volumes of water to obtain bitumen. After the water has been used, it becomes toxic and is stored in large tailings ponds. This water is called “oilsands process affected water” and it contains a mixture of substances called “naphthenic acids” which are toxic and very difficult to measure. The method was used to track naphthenic acids that were carried out of a waste pile and into the environment by running water. Compared to other existing methods, CE-ESI-MS provides new insights into complex mixtures and produces less waste.   vi  Preface Summary of material in thesis chapters The majority of the research conducted in this dissertation was conducted by M.S. MacLennan.  Contributions from other researchers Chapter 1: The review for the semi-quantification of naphthenic acids was summarized primarily by Kevin Kovalchik and secondarily by M.S. MacLennan.  Publications arising from work presented in this thesis Chapter 1 contains the following published material Kovalchik, K. A.; MacLennan, M. S.; Peru, K. M.; Headley, J. V.; Chen, D. D. Y. Standard Method Design Considerations for Semi-Quantification of Total Naphthenic Acids in Oil Sands Process Affected Water by Mass Spectrometry: A Review. Front. Chem. Sci. Eng. 2017, 11 (3), 497–507.  Chapter 2 was written by me and contains the following published material. Targeted studies were performed by me, 95% of manuscript was written by M.S. MacLennan. MacLennan, M. S.; Kok, M. G. M.; Soliman, L.; So, A.; Hurtado-Coll, A.; Chen, D. D. Y. Capillary Electrophoresis-Mass Spectrometry for Targeted and Untargeted Analysis of the Sub-5 kDa Urine Metabolome of Patients with Prostate or Bladder Cancer: A Feasibility Study. J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci. 2018, 1074–1075 (January), 79–85. vii   Chapter 3 was written by M.S. MacLennan and contains the following published material. 100% of experimental work and almost all writing was performed by M.S. MacLennan. MacLennan, M. S.; Tie, C.; Kovalchik, K.; Peru, K. M.; Zhang, X.; Headley, J. V.; Chen, D. D. Y. Potential of Capillary Electrophoresis Mass Spectrometry for the Characterization and Monitoring of Amine-Derivatized Naphthenic Acids from Oil Sands Process-Affected Water. J. Environ. Sci. 2016, 49, 203–212.  Chapter 4 was written by me and contains the following published material. M.S. MacLennan completed the CESI-MS analyses of samples and 99% of the data analysis.  MacLennan, M. S.; Peru, K. M.; Swyngedouw, C.; Fleming, I.; Chen, D. D. Y.; Headley, J. V. Characterization of Athabasca Lean Oil Sands and Mixed Surficial Materials: Comparison of Capillary Electrophoresis/Low-Resolution Mass Spectrometry and High-Resolution Mass Spectrometry. Rapid Commun. Mass Spectrom. 2018, 32 (9), 695–702.  Chapter 5 was written by M.S. MacLennan and contains mention of the following published material. M.S. MacLennan performed approximately 1/3 of the total experimentation.  Kovalchik, K. A.; MacLennan, M. S.; Peru, K. M.; Ajaero, C.; McMartin, D. W.; Headley, J. V.; Chen, D. D. Y. Characterization of Dicarboxylic Naphthenic Acid Fraction Compounds Utilizing Amide Derivatization: Proof of Concept. Rapid Commun. Mass Spectrom. 2017, 31 (24).  viii  Table of Contents  Abstract ......................................................................................................................................... iii Lay Summary ................................................................................................................................ v Preface ........................................................................................................................................... vi Table of Contents ....................................................................................................................... viii List of Tables ................................................................................................................................ xi List of Figures ............................................................................................................................. xiii List of Symbols .......................................................................................................................... xvii List of Equations ........................................................................................................................ xxi List of Abbreviations ................................................................................................................ xxii Glossary .................................................................................................................................... xxvi Acknowledgements .................................................................................................................. xxxi Dedication ............................................................................................................................... xxxiii 1.1 Oil sands process-affected waters (OSPW) and naphthenic acids fraction compounds (NAFC) .................................................................................................................................... 1 1.2 Standard method design considerations for semi-quantification of total naphthenic acids in oil sands process affected water by mass spectrometry: a review ............................. 3 1.2.1 Toward a standard mass spectrometric method for quantifying total naphthenic acids        .............................................................................................................................. 4 1.2.2 Method design considerations for mass spectrometric semi-quantification of total NAs in water ........................................................................................................................ 6 1.2.2.1 Working definition of “total NAs” .................................................................... 7 1.2.2.2 Extraction of NAs and sample clean-up ............................................................ 9 1.2.2.3 Matrix effects and extraction efficiency .......................................................... 14 1.2.2.4 Minimum resolving power .............................................................................. 15 1.2.2.5 Derivatization vs. no derivatization ................................................................. 16 1.2.2.6 Polarity and mode of ionization ....................................................................... 16 1.2.2.7 Choice of calibration standards and use of internal standards ......................... 17 1.2.2.8 Online or offline fractionation of samples ....................................................... 19 1.2.3 Conclusions ......................................................................................................... 20 1.3 Capillary electrophoresis-electrospray ionization-mass spectrometry ....................... 21 ix  2.1 Introduction ................................................................................................................ 29 2.1.1 Urine sampling in the surgical workflow ............................................................ 30 2.1.2 Comparison of analytical methodologies ............................................................ 31 2.2 Materials and Methods ............................................................................................... 33 2.2.1 Ethical statement ................................................................................................. 33 2.2.2 Urine sample collection ....................................................................................... 33 2.2.3 Pathological analysis of cancer tissues................................................................ 35 2.2.4 Patient Diagnoses ................................................................................................ 36 2.2.5 Chemicals and reagents ....................................................................................... 37 2.2.6 Preparation of standard stock solutions ............................................................... 37 2.2.7 Calibration by standard addition method ............................................................ 38 2.2.8 Measurement of creatinine .................................................................................. 39 2.2.9 Instrumentation.................................................................................................... 41 2.2.9.1 Capillary electrophoresis system and software ............................................... 41 2.2.9.2 Electrophoretic procedure ................................................................................ 41 2.2.9.3 ESI-MS/MS and ESI-MS ................................................................................ 42 2.2.10 Data Analysis ...................................................................................................... 43 2.2.10.1 Targeted metabolite data analysis .................................................................... 43 2.2.10.2 Non-targeted metabolite data analysis ............................................................. 44 2.3 Results ........................................................................................................................ 44 2.3.1 Targeted analysis ................................................................................................. 44 2.4 Untargeted Analysis ................................................................................................... 48 2.5 Conclusions ................................................................................................................ 50 3.1 Introduction ................................................................................................................ 52 3.1.1 EDC/NHS-mediated chemical derivatization of naphthenic acids ..................... 52 3.2 Experimental ............................................................................................................... 55 3.2.1 Materials and Chemicals ..................................................................................... 55 3.2.2 Derivatization procedure ..................................................................................... 56 3.2.3 CE-ESI-MS system ............................................................................................. 57 3.2.4 Capillary electrophoresis ..................................................................................... 58 3.2.5 Other Software .................................................................................................... 58 3.3 Results and Discussion ............................................................................................... 59 3.3.1 Impact of derivatization on separation and mass spectra .................................... 59 3.3.1.1 EDC derivatives of naphthenic acids prior to aminolysis ............................... 59 3.3.1.2 General characteristics of CE-ESI-MS data of amine-derivatized naphthenic acids        ......................................................................................................................... 60 x  3.3.1.3 Derivatization with diamine using EDC/NHS in dimethyl sulfoxide ............. 62 3.3.1.4 Derivatization using EDC under different conditions ..................................... 63 3.3.2 Optimization of capillary electrophoresis separation parameters ....................... 65 3.3.2.1 An Introduction to Rough Global Optimization .............................................. 65 3.3.2.2 Set-up of rough optimization of CE-ESI-MS data of derivatizated naphthenic acids        ......................................................................................................................... 66 3.3.2.3 Simplex optimization of BGE composition and rough global optimization of responses ........................................................................................................................ 69 3.3.2.4 Peak swath slope as a representative parameter for multicomponent   resolution ........................................................................................................................ 70 3.3.2.5 Rough Global Optimization............................................................................. 70 3.3.3 CE-ESI-MS of naphthenic acid fraction compounds derived from oil sands process waters .................................................................................................................... 76 3.4 Conclusions ................................................................................................................ 77 4.1 Introduction ................................................................................................................ 79 4.2 Experimental Methods ................................................................................................ 80 4.2.1 Orbitrap mass spectrometry characterization of LOS and surficial materials. ... 81 4.2.2 Capillary electrophoresis-mass spectrometry characterization of LOS and surficial materials ............................................................................................................... 81 4.2.3 Data processing and factor/components analysis ................................................ 82 4.2.4 Data visualization ................................................................................................ 83 4.3 Results and Discussion ............................................................................................... 84 4.4 Conclusions ................................................................................................................ 96 5.1 Conclusions ................................................................................................................ 98 5.2 Future work................................................................................................................. 99 5.2.1 Complex mixtures and semi-quantification ........................................................ 99 5.2.2 Analytical complementarity and data fusion ..................................................... 101 5.2.3 Chemical derivatization and tandem MS .......................................................... 102 5.2.4 Artificial neural networks for pattern recognition............................................. 103 Bibliography .............................................................................................................................. 104 Appendices ................................................................................................................................. 124 Appendix A: Naphthenic acids definitions and vocabulary ................................. 124 Appendix B: Urine metabolomics supplementary material .................................. 129 Appendix C: Data processing scripts and computational supplement.................. 136 Appendix D: Script for enumerating parameter combinations in factor/principal components analysis ............................................................................................. 138  xi  List of Tables Table 1.1 Considerations discussed for the proposal of a standard classical NAs semi-quantification method. .................................................................................................................... 6  Table 1.2 Summary of the 244 target ions for classical NA (O2 species) quantification analysis. Masses represent the anions formed by deprotonation in negative-ion mode ESI. ........................ 8  Table 2.1 A brief comparison of some important aspects of analytical chemistry in two large studies as compared to the present study. ..................................................................................... 32  Table 2.2 Summary of urine sample collection and preparation protocols for patient urine used in this study. ...................................................................................................................................... 34  Table 2.3 Patient diagnoses at prostatectomy/cystectomy as pertaining to this study. Patient C was not analyzed due to the presence of blood in the urine.......................................................... 36  Table 2.4 Stock standard concentrations, general make up, and concentrations. All standards were made up in water, except for leucine and kynurenine, which were dissolved in 0.1 M HCl........................................................................................................................................................ 38  Table 2.5 Multiple reaction monitoring conditions for targeted metabolite analysis via QQQ-MS........................................................................................................................................................ 43  Table 2.6 Endogenous concentrations of the various metabolites in patient urine derived from the x-intercept of a three-point standard addition method with duplicates, using multiple reaction monitoring (MRM) CE-ESI-MS/MS. ........................................................................................... 46  Table 2.7 The regression line equations and R2 values calculated from the standard addition calibration used to calculate endogenous concentrations of the various metabolites in patient urine. ............................................................................................................................................. 46  Table 2.8 The endogenous concentrations of metabolites in patient urine (μmol/L) divided by the measured creatinine (mmol/L). ..................................................................................................... 47  Table 2.9 A list of urinary compounds that were lower (↓) or higher (↑) in concentration in patient samples as compared to control samples. *Intensities of these m/z peaks for one of the patient samples were similar to intensities of the same compounds in the control urine samples........................................................................................................................................................ 50   xii  Table 3.1 Comparison of salient features of CE-ESI-MS electropherograms of derivatized NAs with derivatization solvent properties. DMSO stands out as high viscosity, high dielectric constant compared to ethyl acetate and dichloromethane and produces different separation characteristics under the same BGE conditions. ........................................................................... 64  Table 3.2 Computer filenames used for the global optimization data and the concentrations of formic acid and methanol in the BGE used during each run. ....................................................... 67  Table 3.3 Cut-off values for rough global optimization for each factor. The global minimum value of a factor is denoted as minglobal(factor) and global standard deviation is denoted as σglobal(factor). ................................................................................................................................. 68  Table 3.4 Summary of the relevant system equations for the effects of BGE composition on migration time of first peak, peak width, average peak resolution and total analysis time. ......... 73  Table 3.5 Effects of varying % volume of methanol and formic acid in the BGE for separations of amine-derivatized NAs (EDC/NHS+DMSO) .......................................................................... 75  Table 4.1 Measured values of NAFC (via Orbitrap MS) and TPH (via GC-FID) in samples of LOS and surficial materials. ......................................................................................................... 85  Table B.0.1 The grading of BCa according to the WHO 1973, dubbed the “traditional” system, as well as the WHO 2004 system.142 .......................................................................................... 132  Table B.0.2 The staging of BCa according to the 2002 Tumour-node metastases classification system.142 Common stages include T0, Ta, and Tis.81 ................................................................ 133  Table B.0.3 Gleason score (grade) for prostate cancer based on the VACURG data.82,143 ........ 134    xiii  List of Figures Figure 1.1 (Left) A small subset of possible naphthenic acid fraction compounds (NAFC) structures based on C9-28, a single 6-membered ring, 3 heteroatom types at 2 locations (L in the diagram) and a single carboxylic acid group. This small subset of all possible NAFC contains approximately 259,000 molecules, not including stereoisomers. A sample NAFC molecule in this subset is shown on the right. ........................................................................................................... 2 Figure 1.2 Three examples of naphthenic acid molecules which fall under the definition of “total NAs” ............................................................................................................................................... 7  Figure 1.3 Relative total NAFC extraction using selected solvent systems or SPE. Solvent polarity index is given along the x-axis. Areas represent total area under the TIC curves observed by negative ion electrospray ionization (ESI) Orbitrap MS. [Reprinted with permission from Figure 3, Headley et al. 2013.36 Copyright (2013) Elsevier.] ....................................................... 10  Figure 1.4 Distribution of selected components of NAFC. Extraction using selected solvent systems or SPE was carried out prior to analysis by negative-ion ESI Orbitrap MS. The NAFC component classes refer specifically to the heteroatoms O and S. [Reprinted with permission from Figure 4, Headley et al. 2013.36 Copyright (2013) Elsevier.] .............................................. 11  Figure 1.5 Extracted amounts of selected NAFC classes using six solvents. Extractions were carried out at pH values of a) 12.0, b) 8.5, and c) 2.0. [Reprinted with permission from Figure 1, Huang et al. 2016.41 Copyright (2016) Elsevier.] ......................................................................... 12  Figure 1.6 Predicted log DOW values with changing pH for 19 classical NAs. Numbers reference the compounds described in Celsie et al. 2016 SI. [Data taken from Celsie et al. 2016 supplemental information.]40 ........................................................................................................ 13  Figure 1.7 Predicted log KOW values with changing temperature for 19 classical NAs. Numbers reference the compounds described in Celsie et al. 2016 SI. [Data taken from Celsie et al. 2016 supplemental information.]40 ........................................................................................................ 14  Figure 1.8 Relative abundance of compounds matching the formula CnH2n+ZOx, summed for n = 8 to 30 and Z = 0 to -12. [Reprinted with permission from Figure 4, Grewer et al. 2010.21 Copyright (2010) Elsevier.] .......................................................................................................... 18  Figure 1.9 Bar plots and contour diagrams showing the number of homologues detected for the formula CnH2n+ZOx for: a) OSPW sample extracted by liquid-liquid extraction and directly injected into FTICR-MS; b) OSPW sample extracted as in (a) and fractionated by UHPLC prior to FTICR-MS; and c) combination of two OSPW samples processed as in (b) to compensate for dilution effects. [Reprinted with permission from Figure 3, Nyakas et al. 2013.29 Copyright (2013) American Chemical Society.] ........................................................................................... 20 xiv   Figure 1.10 Schematic diagram of the CE-ESI-MS setup as used in this thesis. The ESI device acts as the positive electrode in CE-ESI-MS. Large NAs are detected first and small NAs are detected last due to their electrophoretic mobilities (blue arrow) being in the opposite direction of the EOF-directed bulk flow (red arrows). ..................................................................................... 25  Figure 1.11 Schematic of electrospray ionization apparatus which utilizes a flow-through microvial and bevelled tip geometry.52 ......................................................................................... 26  Figure 2.1 Descriptive flowchart of the hierarchical data structure of this study. Each node in the flowchart represents a type of information. The green nodes refer to medical information directly used for the diagnosis of cancers. The orange node represents the biological information accessible via chemical methods. The yellow nodes represent the analytical chemical information produced in this study. The grey, dashed arrow refers to how CE-ESI-MS information can be seen as contextual biological information, and to show the connection between chemical analysis and the grander goal of diagnosis. ................................................................................................ 31  Figure 2.2 Molecular structure of the Jaffé red chromogen depicted as a 2:3 Meisenheimer σ-complex. The kinetics of this reaction depend non-linearly on hydroxide at high pH. ................ 40  Figure 2.3 Set of extracted ion chromatograms (XIC) from Patient B obtained by CE-ESI-MS/MS, multiple reaction monitoring (MRM) mode. Notice the relative intensities of sarcosine and alanine and the magnified inset. ............................................................................................. 45  Figure 2.4 Plot of first 2 principal components (6 PC’s total). The control pooled urine samples (blue circles) were identical in composition; whereas the diseased urine samples (red squares) are each from patients A, B, D, and E. ............................................................................................... 49  Figure 3.1 General reaction equation for the EDC (and optional NHS) mediated amidation of carboxylic acids (e.g., NAs) with a primary-tertiary diamine to form (naphthenic) amines. Figure 3.2 provides additional information on the reaction mechanism. ................................................. 53  Figure 3.2 (A) The electrophilic carbodiimide carbon in EDC accepts electrons from R-carboxylic acid to (B) produce the O-acylisourea derivative of EDC. Subsequent aminolysis of the O-acylisourea releases (C) the EDC urea by-product (far right, lower structure) to produce the diamine-derivatized carboxylic acid (far right, upper structure). ........................................... 54  Figure 3.3 Isomerization of O-acylisourea EDC derivatives of NAs to N-acylurea derivatives based on naphthenic-acyl transfer. ................................................................................................ 55    xv  Figure 3.4 CE-ESI-QMS data generated from EDC-derivatized standard NAs mixture prior to aminolysis. The vertical axis has units of m/z and the horizontal axis has units of seconds (time). The colours of the peaks represent peak intensity, shown in the legend on the right-hand side of the mass electropherogram. This mixture showed three swaths of peaks representing one group of parent ions and two groups of fragmentations. Tentative structural assignments for neutral losses are shown to the right. ........................................................................................................ 60  Figure 3.5 Sample mass electropherogram for amine-derivatized NAs from NA standard (Aldrich). The zoomed inset on the right demonstrates the typical spacing of NAs data: Swaths of peaks with 14 Da difference represent series of -CH2- and within each carbon number class, there is a Z-series which represents increasing double bond equivalents (double bonds or rings), each producing a hydrogen deficiency of 2. ................................................................................. 61  Figure 3.6 A sample base peak electropherogram (BPE) produced by analysis of amine-derivatized NAs (mediated by EDC/NHS in DMSO) .................................................................. 63  Figure 3.7 (Left) Optimal criteria for each of the 5 factors. (Right) Size of circle graphics used in Figure 3.8 to represent optimum response of each of the 5 factors. The size of the circle correlates with the weight put on each factor. .............................................................................. 71  Figure 3.8 Diagram describing global optima for 5 CE separation responses. A circle means an optimum was reached for a factor response. The diameter of the circle indicates the relative importance (weight) of the response. The most circles at a particular BGE composition indicate a point of roughly global optimality. The red triangle outlines a region of global optimality. The black arrow indicates the direction of the system from initial BGE composition toward global optimality upon natural methanol evaporation (starting at 30% and evaporating to 0%). ........... 72  Figure 3.9 Response surface plots for 4 factors. The response surfaces were approximated by polynomial regression on the rough global optimization data to approximate the effects of changing % formic acid and % methanol on the optimum factors described in Table 3.3. ......... 74  Figure 3.10 CE-ESI-TOFMS of derivatized NAFCs derived from oil sands process waters (OSPW). The data shown here are mean-centred and normalized to standard deviation by m/z channel. The colour bar illustrates relative peak intensity (%). .................................................... 76  Figure 4.1 Example of GC-FID of LOS and surficial materials analyzed according to a standard CCME method. The vertical axis is “picoamperes” of current and horizontal axis is “Time” in minutes. The fractions F2, F3, F4 are also delineated with equivalent carbon number (ECN) range given for each fraction (C10, C16, C34, C50). ................................................................... 84   xvi  Figure 4.2 Orbitrap MS characterization of the distribution (relative abundance) of Ox (x=1-5), OxSy (x=1-4) classes of LOS and surficial materials across all samples. .................................... 87  Figure 4.3 Relative abundance of negative-ion Orbitrap MS double bond equivalents (DBE) versus carbon number for the O2 species across all samples. ...................................................... 88  Figure 4.4 3-D plot of 3-factor factor analysis (FA) using principal axis (PA) factoring of infusion negative-ion Orbitrap MS of all ions observed in the raw data for LOS and surficial materials. ....................................................................................................................................... 90  Figure 4.5 Loadings plot of FA principal axes 1 and 2 of Orbitrap MS data with a Gaussian-blurred contour overlay approximating the total NAFCs (mg/L) concentration of each sample. Higher concentration samples cluster together and have similar chemical “fingerprints” in principal axis loadings space, whereas lower concentrated samples have a range of chemical “fingerprints”. ............................................................................................................................... 92  Figure 4.6 Two-panel plot of typical CE-MS data of NAFCs extract sample of LOS-affected water: Total ion electropherogram (top, blue) and CE-MS ion map (bottom, yellow). Peak intensity cut-off was applied. The mass-electropherogram can be divided into 5 general zones that appear in every sample: A=dip in background signal at electroosmotic flow marker, which is more visible when peak intensity cut-off is not applied, B=various amines or derivatized acids, C=amine-derivatized naphthenic acids, D and E=derivatization by-products. ............................ 93  Figure 4.7 Three-component principal component analysis (PCA) plot (first two components) of total ion mass electropherogram peak intensity (left) and mean migration time of selected analytes (right) for positive-ion CE-LRMS of derivatized naphthenic acid ions observed in the raw data for LOS and surficial materials. The three-component PCA explains 97% of the variance in peak intensity (PC1=43%, PC2=30%, PC3=23%; left) and 55% of the variance in mean migration time (PC1=31%, PC2=14%, PC3=10%; right). ................................................. 94  Figure B.0.1 Three-point standard addition calibration curves for proline, cystine, leucine, glutamic acid, kynurenine and sarcosine in patient urine samples. ............................................ 131  Figure B.0.2. Approximately linear decrease in the three-point calibration coefficient of determination (R2). Patient designation (A, B, D, E) is placed next to the blue diamond which corresponds to its data. Note that patient E’s creatinine level is actually 3.5 times greater than Patient B, but they are placed together since both R2 show very little correlation. This trend was only seen for patient cystine data. ............................................................................................... 131    xvii  List of Symbols % v/v Volume percent [M–OH]+ Mass spectral cation which is composed of the parent ion (M) minus one hydroxide ion [OH-] Hydroxide ion concentration ↑ Increase in metabolite concentration, relative to patient sample ↓ Decrease in metabolite concentration, relative to patient sample °C Degrees Celsius µL Microlitre µmol Micromole C=O Carbonyl moiety cps Counts per second D Defined as ñΣxi2 − (Σ𝑥𝑖)2 for a standard addition calibration.  Da Dalton (unit) DOW Octanol-water distribution coefficient eV Electron Volt E Electric field strength (Volts/cm) ε The product of the permittivity of the vacuum and the dielectric constant of a fluid (ε0εr).  g Gram ġ Unit of gravitational acceleration h Hour Hz Hertz η Viscosity of solution xviii  k Number of times the sample was measured (in the context of a calibration). For duplicate measurements, k = 2; for triplicate measurement, k = 3. k3 Drag force on a charged particle moving in an electrolyte solution under an applied electric field due to the surrounding ionic cloud. k4 Electrophoretic relaxation force on a charged particle moving in an electrolyte solution under an applied electric field. KOW Octanol-water partition coefficient kV Kilovolt Lc Length of capillary (cm) used in capillary electrophoresis Ld Length to detector (cm) in capillary electrophoresis L/h Litres per hour m Value of slope of a calibration curve; Instrument sensitivity M Molarity; M = mol/L m/z Mass-to-charge ratio mg/kg Milligrams petroleum hydrocarbon per kilogram lean oil sands mg/L Milligrams solute per litre solvent min Minute mL Millilitre mmol Millimole mol Mole mPa*s Millipascal seconds; unit of viscosity (𝑚𝑧)𝑎 m/z value for centroided MS peak a xix  µep Electrophoretic mobility N Normality; effective molarity; activity in mol/L  n Number of (carbon) atoms in a molecule ñ Population or sample size (statistical) N-H N-H bond nL Nanolitre nm Nanometre pH Power of Hydrogen; pH = log([H+]) ppm Parts per million psi Pounds per square inch Qeff Effective charge of a spherical particle in solution R Hydrodynamic radius of a particle R2 The coefficient of determination for a fitted calibration curve Rm Mass resolving power of a mass spectrometer s Second sx Square root of error in x-intercept of calibration curve sy Square root of error in y values of calibration curve, as measured by the instrument t Migration time (sec) of an analyte V Volts veof Electroosmotic flow velocity vep Electrophoretic velocity of analyte (cm/sec) xi x-value of measurement i xx  x-int Numerical value of the x-intercept  Z Total hydrogen atom discrepancy in a chemical formula due to specific characteristics of the molecule such as cyclic structures, double or triple bonds and nitrogen atoms.  ζ Zeta potential    xxi  List of Equations Equation 1.1 Mass resolution of a mass spectrometer, Rm, calculated from the ratio of the m/z value of a single peak, (m/z)a, to the difference between (m/z)a and another peak (m/z)b. .......... 15  Equation 1.2 Electrophoretic mobility as a function of analysis time (t), capillary length to detector (Ld) and electric field strength (V/Lc). This equation can be used to modulate data between the time and mobility domains. ...................................................................................... 22  Equation 1.3 Electrophoretic velocity as a balance of four forces exerted by the applied electric field (QeffE), Stokes drag on a spherical particle (6πηRv), and two forces, k3 and k4, which are drag and relaxation forces caused by the ionic atmosphere in solution. ....................................... 22  Equation 1.4 The electroosmotic flow of a charged layer is proportional to the zeta potential and the applied electric field strength and is inversely proportional to the solution viscosity. ........... 23  Equation 2.1 Formula used to calculate standard deviation of the extrapolated x-intercept of a standard addition calibration, where k is a constant that equals 2. ............................................... 47  Equation 2.2 Formula to calculate relative standard deviation (RSD) in a standard calibration, where sx is the standard deviation of the x-intercept, as calculated in Equation 2.1, asnd x-int is the value of the x-intercept. .......................................................................................................... 47   xxii  List of Abbreviations The following table contains a list of abbreviations used in the main body of the text.  Abbreviation Long-form AEO Acid-extractable organics APPI Atmospheric pressure photoionization BCa Bladder cancer BGE Background electrolyte CCME Canadian Council of Ministers of the Environment CE Capillary electrophoresis  CE(V) Collision energy (Volts) CE-ESI-MS Capillary electrophoresis-electrospray ionization-mass spectrometry CE-ESI-MS/MS Capillary electrophoresis-electrospray ionization-tandem mass spectrometry CE-ESI-QMS Capillary electrophoresis-electrospray ionization-quadrupole mass spectrometry CE-ESI-TOFMS Capillary electrophoresis-electrospray ionization-time of flight mass spectrometry CE-LRMS Capillary electrophoresis-low resolution mass spectrometry CE-MS Capillary electrophoresis-mass spectrometry  CMP Chemical measurement process CXP(V) Collision cell exit potential (Volts) Cys Cysteine (amino acid) CysCys Cystine (amino acid dimer) d3-Sarc Deuterated sarcosine (amino acid standard) xxiii  DBE Double bond equivalent DCM Dichloromethane (methyl chloride) DMSO Dimethyl sulfoxide DP(V) Declustering potential (Volts) ECCC Environment and Climate Change Canada ECN Equivalent carbon number EDC 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide ENV+ Mixed polystyrene-divinylbenzene solid phase extraction cartridge EOF Electroosmotic flow EP(V) Entrance potential (Volts) ESI Electrospray ionization EtOAc Ethyl acetate FA Factor analysis FTICR-MS Fourier-transform ion cyclotron resonance-mass spectrometry FTIR Fourier-transform infrared spectroscopy GC Gas chromatography GC-MS Gas chromatography-mass spectrometry Glu Glutamic acid (amino acid) GS Gleason Score GU Genitourinary H+E Hematoxalin + Eosin stain HCD Higher-energy C-trap dissociation xxiv  HRMS High-resolution mass spectrometry I.D. Inner diameter (of capillary) Kyn Kynurenine (amino acid) L Length (of capillary) LC Liquid chromatography LC-MS Liquid chromatography-mass spectrometry LC-MS/MS Liquid chromatography-tandem mass spectrometry Leu Leucine (amino acid) LOS Lean oil sands MRM Multiple reaction monitoring MWCO Molecular weight cut-off N,N-DEEDA N,N-diethylethylenediamine or N,N-diethylethane-1,2-diamine N,N-DMEDA N,N-dimethylethylenediamine or N,N-dimethylethane-1,2-diamine NA Naphthenic acids NAFC Naphthenic acids fraction compounds NHS N-hydroxysuccinimide O.D. Outer diameter (of capillary) OSPW Oil sands process-affected waters OSTWAEO Oil sands tailings water acid-extractable organics PA Principal axis PAR Peak area ratio PC Principal component xxv  PCa Prostate cancer PCA Principal components analysis PEI Polyethyleneimine PHC Petroleum hydrocarbons Pro Proline (amino acid) PSA Prostate-specific antigen Q1 The mass of a representative ion which is transmitted through quadrupole 1 in QQQ-MS. The value of this mass is chosen by the MS operator. Q3 The mass of a representative ion which is transmitted through quadrupole 3 in QQQ-MS. The value of this mass is chosen by the MS operator. QQQ-MS Triple quadrupole mass spectrometry QTOF Quadrupole-time-of-flight (mass spectrometer) RSD Relative standard deviation Sarc Sarcosine or N-methylglycine (amino acid) SPE Solid-phase extraction TCC Transitional cell carcinoma TIC Total ion chromatogram TIE Total ion electropherogram TOFMS Time of flight mass spectrometry UHP Ultra-high purity UHPLC Ultra-high-performance liquid chromatography  xxvi  Glossary The following table contains definitions for terms found throughout the body of the text which are specific to medical, geographical, statistical-mathematical, analysis and electrophoresis domains. Background electrolyte A liquid-phase ionic solution, often aqueous, which is used as the mobile bulk phase in capillary electrophoresis separations. Carcinoma A type of cancer that develops from epithelial cells. Caustic hot water Aqueous NaOH at 82° C, conditions which are used in the Clark hot water extraction of bitumen from oil sands ore. Chemical measurement process (CMP) The subprocess of analytical chemistry which involves handling and pretreatment of samples, instrumental analysis, signal generation and data analysis. The CMP provides critical information regarding chemical systems. Congener Two or more chemicals are congeners if they have related origin, structure or function. In a complex mixture, several isomers or stereoisomers are considered as congeners of the same formula. Double bond equivalent The number of times 2 hydrogen atoms are removed from a chemical formula due to the presence of double bonds or rings. Signifies the total number of double bonds or rings in a molecule. xxvii  Equivalent carbon number The value of the quotient of detected molecular mass, or approximated molecular mass, divided by 12.  Factor analysis, (FA) A statistical method of realigning a dataset onto a new set of coordinates according to common variance in the dataset. The assumption is that variance in the dataset is caused by real but unknown factors. Factor analysis finds those factors but does not interpret them. The usage of common variance (i.e. removing outliers) allows for generalizability of the factors into a model. Fingerprinting Generating chemical signals which are unique representations of samples. Grading (pathology) Assessment of how abnormal tumour tissues look under microscopic examination and is used to indicate how likely a tumour is to grow and spread. Groundwater Water in a large underground body. Histopathology Microscopic examination of tissue to understand disease. Lean oil sands Soil which contains < 5% bitumen that is not economically viable. Lean oil sands is a common waste material at mining sites. Mass electropherogram The name for CE-MS full-scan data. CE produces an “electropherogram” and MS produces a “mass spectrum”. “Mass electropherogram” is a compound word containing both terms. xxviii  Midstream The point in time between beginning urination and finishing urination where a constant stream of urine is exiting the patient, effective because of low probability of bacterial interference. Multi-source natural system An environmental system (e.g. body of water) which has many sources of contamination. Naphthenic acids fraction compounds (NAFC) All chemical compounds present in the naphthenic acid fraction of any OSPW sample. Neoplasm Abnormal growth of tissue, commonly referred to as a “tumour”. Node A graph theoretic term that signifies an “entity” and is visually represented by a shape or point. The relationships between entities are visually represented by lines or arrows (called “edges”) connecting the shapes (nodes). Oil sands process-affected waters (OSPW) Any water derived from any tailings ponds at any time in the Alberta oil sands industry operations or any water which has come in contact with bitumen ore via industrial processes.  Perched water Perched water is a type of groundwater that is separated from the main body of groundwater by rock or some other material. In the case of lean oil sands, groundwater is “perched” inside the LOS pile. Porewater Water which is held in the pores of soil. xxix  Principal component analysis, (PCA) A statistical method of realigning a dataset onto a new set of coordinates according to all variance in the dataset (i.e. including “outliers”). The assumption is that variance in the dataset is caused by real but unknown components. All data points contribute to variance, and PCA can be used to reduce dataset noise.  Rag layer The layer of unresolved emulsified oil sands, clay and surfactants which forms from froth during bitumen separation from sand. The rag layer contains naphthenic acids. Seepage Water which flows from one place to another, usually carrying particles or compounds of concern. Seepage water has the potential to come into contact with many materials such as bitumen, OSPW and river water. Selectivity The selectivity of an analytical method is its ability to measure one type of analyte over another. A method is “selective” for acids if it can measure acids above the matrix background signal. Sensitivity The sensitivity of an instrument is its ability of instrument signal to change relative to changes in input sample amounts. In the context of linear calibration, sensitivity is equal to the slope of a fitted calibration line.  Solvent-polarity index An index of the polarity of a solvent. The higher the value, the more polar the solvent. Staging (pathology) Assessment of the size and reach of a tumour in the body. xxx  Swath The shape naphthenic acids signals make on a mass-electropherogram when separated and analyzed by full scan CE-MS. Traceability The ability of a chemical substance to be tracked (identity and quantity) throughout an analytical chemical measurement process. Tracer A chemical compound, or compounds, which is representative of a target quality (e.g. toxicity, industrial activity) and whose presence represents the presence of the target quality. Transitional cell carcinoma A type of cancer that occurs in the urinary system. It is the most common type of bladder cancer.          xxxi  Acknowledgements  First off, I wish to thank my supervisor Dr. David Chen for accepting me into his research group some 6 years ago. David, you believed in me and provided me with more challenges and opportunities than I could ask for! I have learned a lot from you on how to nurture important connections across the globe, and the way you care for your students inspires me. You also introduced me to Dr. John Headley, who graciously agreed to be my co-supervisor. I want to thank Dr. John Headley, who has also journeyed with me the past few years. John, your strong belief in my abilities and tangible encouragement have always meant a lot to me. Your perseverance and work ethic have been inspiring, and I know you care about my progress, even though we weren’t often in the same province! Thank you for introducing me into Saskatoon where I got to meet Kerry and Jon and learn Portuguese from Emeka.   I wish to thank my lab-mates—the many who have gone before me and the new students who have arrived—for putting up with me and feigning interest in my crazy ideas! We should have gone on more coffee breaks. And I should have listened more. Thanks to Kevin for being a great co-author! Special thanks go to my crew: Caitlyn, Cheng, Jess and Ling Yu, and past members Hong, Sherry and Zhen.  Thanks to Marcelo Bravo for being my best friend on campus. You introduced me to the world of research impact beyond the academy and we explored it together! My approach to research and scholarship have changed forever. Thank you, Geoff Woollard for being a best friend to both of us.    Mom and Dad… uill, sin agad e. Thank you for working so hard to raise me and help me become well-rounded. I can’t imagine how hard it was at times. I know faith, family and xxxii  education are important because of how you raised me. You kept me connected to our roots, to the land, to our people. I am proud to have you as my parents and I hope I make you proud.  Finally, and firstly, to my wife, Danielle. Thank you for being with me in all this. Thank you for taking care of our children, for keeping me focused when things were blurry and for pushing me to be a better version of myself everyday. Thanks to my children Damien, Veronica and Finlay for always cheering me up and for reminding me that joy is in the everyday little things. I am excited for what life holds around the corner for us.            xxxiii  Dedication This thesis is dedicated to my grandmothers who passed away during my time at University: Adelaide LeBlanc and Jessie Ann MacLennan.  I’ll never forget the ways you cared for me. Cha bhi e fad’ a-nis.1   Introduction 1.1 Oil sands process-affected waters (OSPW) and naphthenic acids fraction compounds (NAFC)  More than 170 billion barrels of crude oil are considered to be economically viable for recovery in the Athabasca oil sands in Northern Alberta, Canada.1 The bitumen extraction process involves mixing oil sands bitumen with caustic hot water, mobilizing the bitumen off the sand, causing the partitioning of naphthenic acids (NAs) and other molecules into the water phase. This water, termed oil sands process affected water (OSPW), is recycled several times and ultimately stored in large tailings ponds whose total volume has already surpassed 720 billion L.2 By law, OSPW must be stored until it is converted to a quality which is safe for reintroduction into the environment. For every 1 barrel of synthetic crude extracted from the oil sands, 3.3 barrels of OSPW “tailings” are produced.3 Fresh OSPW has been demonstrated to be acutely toxic toward a number of animal species, due in part to the presence of NAs at concentrations as high as 100 mg/L in tailing ponds.2,4 Although “oil sands process-affected waters” and “naphthenic acid fraction compounds” are commonly used today, they have been called by many names in engineering, industry and research over the last century. Appendix A1 describes the development of terminology and definitions of these substances.  NAs are organic acids naturally present in crude oil and bitumen and are classically defined as the family of alicyclic monocarboxylic acids with the general chemical formula CnH2n+ZO2, where Z is a negative, even integer representing hydrogen deficiency in the molecule due to the presence of ring structures. OSPW contains high levels of classical NAs, but also N and S heteroatomic acids, aromatic acids, and heavily oxygenated acids. This larger class of acids, of which classical NAs are a subset, is often referred to as “naphthenic acid fraction 2  compounds” (NAFCs). NAFCs are defined as the acid-extractable fraction of OSPW.5 The generalized molecular formula for NAFCs is CnH2n+3y+ZOwSxNy.1,6 Here, the subscripts n, w, x and y indicate the number of carbon, oxygen, sulfur and nitrogen atoms respectively, and Z represents the hydrogen atom deficiency due to the presence of double bonds and/or rings.7 A small subset of possible NAFC structures with C8-26 is shown in Figure 1.1.                 SCH3OHOOH Figure 1.1 (Left) A small subset of possible naphthenic acid fraction compounds (NAFC) structures based on C9-28, a single 6-membered ring, 3 heteroatom types at 2 locations (L in the diagram) and a single carboxylic acid group. This small subset of all possible NAFC contains approximately 259,000 molecules, not including stereoisomers. A sample NAFC molecule in this subset is shown on the right.  The number of rings, the nature of the alkyl chains or number and type of heteroatoms in NAFC molecules influence NAFC solubility and ability to leach into the surrounding environment.8,9 Biodegradation of some NAFCs can be relatively slow with half-lives of ~14 years.10,11 NAFCs are among the principal toxicants in OSPW and thus increased attention has been given to monitoring the levels of oil sands acids in the environment.12–16 NAFC mixture compositions vary according to the geographical location of the crude deposit. Additionally, NAFCs found in OSPW may have altered molecular structures, including norbornene and adamantane hydrocarbon chains, and toxicity from those found naturally in bitumen.17,18 NAFCs are toxic and persistent surfactants with largely unidentified molecular structures existing in unquantified 3  amounts. Identification of the source of a particular OSPW sample is therefore an important challenge for environmental forensics. Currently, there is no standard monitoring system for NA levels in water. This is due in part to the fact that current chromatographic separation methods cannot resolve individual NAs species, including isomers, from the complex mixtures that exist in the environment,19 and in part due to the lack of analytical standard compounds for calibration.   1.2 Standard method design considerations for semi-quantification of total naphthenic acids in oil sands process affected water by mass spectrometry: a review There are a wide variety of methods used for the determination of total NAs in OSPW and environmental samples.1,15,20 Extracting NAs from aqueous OSPW samples has largely developed along the lines of liquid-liquid extractions and solid-phase extractions. Some recent methods dispense with extraction altogether and rely only on pH adjustment prior to NAs analysis.21 Detailed reviews are given elsewhere1,15,20,22 covering the pros and cons of existing methods in use by practitioners along with emerging methods for fingerprinting or environmental forensics.1 The variety of methods used for analyses of total NAs include Fourier transform infrared (FTIR)23 and fluorescence spectroscopic techniques,24,25 along with low resolution26 or high resolution13,26–28 mass spectrometry. Many mass spectrometric methods have been developed which utilize direct injection,29 gas chromatography/mass spectrometry (GC-MS),23,30 liquid chromatography/mass spectrometry (LC-MS),28,31–33 and LC-MS/MS,34,35 employing either negative-ion34 or positive-ion35 detection and numerous ionization platforms.15 Likewise, some methods use off-line chromatography for sample clean up or preconcentration prior to MS analysis.29 Some methods analyze the sample without derivatization whereas others utilize 4  derivatization steps.33,35 Instrument calibration methods vary between laboratories and depend on the availability or choice of commercial standards, along with the limited access to actual standards of NAFCs extracted from different sources of OSPW. Finally, the data analysis for some methods is based on integrated peaks of total naphthenic acid fraction compounds (“NAFCs” or the “acid extractable fraction”)36 while others select extracted ions that correspond to NA congeners.26  1.2.1 Toward a standard mass spectrometric method for quantifying total naphthenic acids As more laboratories engage in analyses of NAs and other OSPW components, there is demand to assess the comparability of results from various methods.37 Furthermore, the need for interlaboratory comparison was identified by practitioners as a high priority activity by specialists at an international workshop on analytical strategies for NAs.1  Evaluation of an interlaboratory study on semi-quantifying total NAs in water was reported for 15 participating laboratories.38 Methods included (number of laboratories is given in parentheses): FTIR (3), along with MS methods (12) using either low resolution (9) or high resolution (3) with direct injection (2), GC-MS (3), LC-MS (3), and LC-MS/MS (1) employing either negative-ion (11) or positive-ion detection (1). Four methods utilized derivatization steps and one laboratory used off-line chromatography for sample cleanup and preconcentration prior to MS analyses. Quantitative data analysis in the FTIR methods was based on integrated peaks for total carbonyl group signal, while the mass spectrometry methods utilized specific ions which corresponded to CnH2n+ZO2 congeners. A neat Merichem NAs mixture (a gift received from Merichem Chemicals and Refinery Services LLC, Houston, TX) was also provided to the 5  participating laboratories for use as a reference to minimize variability between laboratory standards.38 Despite these measures, variable results were reported and this led to subsequent intralaboratory studies by Environment and Climate Change Canada (ECCC), with 4 participating labs, and steps were taken to better understand and control the factors contributing to variability in measurements. The activities are on-going and have prompted the establishment of an ECCC taskforce of analytical chemists to improve the measurement of total NAs in environmental samples. All of this information suggests there is a need for a standard method against which laboratories would be required to demonstrate performance and traceability of a given method. Indeed, this need has been previously identified,37 but at present there is still no standard method. The intent of this review is to discuss the design considerations for such a method for the semi-quantification of classical NAs. A range of currently used methods for the analysis of NAs has been compiled. Additionally, studies on specific aspects of NAFCs analysis are considered (e.g. extraction method, ionization polarity, etc.). While this chapter will discuss the important features of a method for semi-quantification of classical NAs, it is acknowledged that there will likely be a series of standard methods for analyses of other NAFCs, depending on the end use of the data. Prior to discussion of the design considerations, the choice of detection method will be discussed.  Traditionally, the method for total NA quantification has been Fourier transform infrared spectroscopy (FTIR), in which quantification is correlated to total carboxylic acid functional groups. FTIR is thus sensitive not only to classical NAs, but also to any compound containing one or more carboxylic acid moieties.15 As such, FTIR is not suitable for the selective analysis of classical NAs. There is consensus among the authors that mass spectrometry is the preferred 6  instrument for the measurement of NAs: with sufficient mass resolution, mass spectrometry is able to distinguish the different classes of OSPW organic acids.13 The selection of mass spectrometry as an analysis method will inform much of the following discussion.   1.2.2 Method design considerations for mass spectrometric semi-quantification of total NAs in water  In this section we will review 8 important factors in developing method guidelines for the analysis of total NAs in OSPW. The factors are summarized in Table 1.1.    Factors  (organized by phase of CMP) Steering Group Conclusions References 1. Definition of total NAs Use the classical definition of NAs 13 2. Extraction phase, pH, temperature Liquid-liquid extraction at pH 2 and room temperature with DCM as organic phase, or use ENV+ SPE 36,39–41 3. Use of surrogate standards Use isotopically labelled model compounds as surrogate standards 21 4. Use of derivatization Not necessary to utilize derivatization but considered important approach 31,35 5. Suitable calibration standard and internal standard Use commercially available Merichem NA mixture and at least one isotopically labelled internal standard 13,21,26 6. Use of on-line or off-line fractionation of sample Employ on-line chromatography prior to MS detection 21,29,42,43 7. Polarity and mode of ionization Negative-ion mode ESI 15,44 8. Minimum resolving power of instrument 50,000 at m/z 200, acknowledging that potential interferences contribute to method uncertainty 1,13,26,29  Table 1.1 Considerations discussed for the proposal of a standard classical NAs semi-quantification method.  7  1.2.2.1 Working definition of “total NAs” While complete characterization and quantification of all OSPW NAFC is a desirable goal, we will restrict our scope by using a working definition of NAs which only includes classical NAs (formula class CnH2n+ZO2) and those only within a specified carbon number and hydrogen deficiency range (n = 6 to 40; Z = 0 to -12). Therefore, “total NAs” refers to the measured quantity of those NAs within the scope of the working definition. This provides a practical starting point for future quantification methods for total NAFCs and simplifies analysis considerations. Table 1.2 contains a summary of the target NAs for analysis. OHOCH3OHOCH3OHOC10H18O2C11H20O2C11H18O2Z = -2Z = -2Z = -4 Figure 1.2 Three examples of naphthenic acid molecules which fall under the definition of “total NAs”      8  Z-value 0 -2 -4 -6 -8 -10 -12 Carbon # Target ion accurate mass (Da) 6 115.07645 113.0608 111.04515 109.0295 107.01385 104.9982  7 129.0921 127.07645 125.0608 123.04515 121.0295 119.01385 116.9982 8 143.10775 141.0921 139.07645 137.0608 135.04515 133.0295 131.01385 9 157.1234 155.10775 153.0921 151.07645 149.0608 147.04515 145.0295 10 171.13905 169.1234 167.10775 165.0921 163.07645 161.0608 159.04515 11 185.1547 183.13905 181.1234 179.10775 177.0921 175.07645 173.0608 12 199.17035 197.1547 195.13905 193.1234 191.10775 189.0921 187.07645 13 213.186 211.17035 209.1547 207.13905 205.1234 203.10775 201.0921 14 227.20165 225.186 223.17035 221.1547 219.13905 217.1234 215.10775 15 241.2173 239.20165 237.186 235.17035 233.1547 231.13905 229.1234 16 255.23295 253.2173 251.20165 249.186 247.17035 245.1547 243.13905 17 269.2486 267.23295 265.2173 263.20165 261.186 259.17035 257.1547 18 283.26425 281.2486 279.23295 277.2173 275.20165 273.186 271.17035 19 297.2799 295.26425 293.2486 291.23295 289.2173 287.20165 285.186 20 311.29555 309.2799 307.26425 305.2486 303.23295 301.2173 299.20165 21 325.3112 323.29555 321.2799 319.26425 317.2486 315.23295 313.2173 22 339.32685 337.3112 335.29555 333.2799 331.26425 329.2486 327.23295 23 353.3425 351.32685 349.3112 347.29555 345.2799 343.26425 341.2486 24 367.35815 365.3425 363.32685 361.3112 359.29555 357.2799 355.26425 25 381.3738 379.35815 377.3425 375.32685 373.3112 371.29555 369.2799 26 395.38945 393.3738 391.35815 389.3425 387.32685 385.3112 383.29555 27 409.4051 407.38945 405.3738 403.35815 401.3425 399.32685 397.3112 28 423.42075 421.4051 419.38945 417.3738 415.35815 413.3425 411.32685 29 437.4364 435.42075 433.4051 431.38945 429.3738 427.35815 425.3425 30 451.45205 449.4364 447.42075 445.4051 443.38945 441.3738 439.35815 31 465.4677 463.45205 461.4364 459.42075 457.4051 455.38945 453.3738 32 479.48335 477.4677 475.45205 473.4364 471.42075 469.4051 467.38945 33 493.499 491.48335 489.4677 487.45205 485.4364 483.42075 481.4051 34 507.51465 505.499 503.48335 501.4677 499.45205 497.4364 495.42075 35 521.5303 519.51465 517.499 515.48335 513.4677 511.45205 509.4364 36 535.54595 533.5303 531.51465 529.499 527.48335 525.4677 523.45205 37 549.5616 547.54595 545.5303 543.51465 541.499 539.48335 537.4677 38 563.57725 561.5616 559.54595 557.5303 555.51465 553.499 551.48335 39 577.5929 575.57725 573.5616 571.54595 569.5303 567.51465 565.499 40 591.60855 589.5929 587.57725 585.5616 583.54595 581.5303 579.51465  Table 1.2 Summary of the 244 target ions for classical NA (O2 species) quantification analysis. Masses represent the anions formed by deprotonation in negative-ion mode ESI. 9  1.2.2.2 Extraction of NAs and sample clean-up When extracting NAs from a sample, the choice of extraction phase, solvent, temperature, and pH can play significant roles in the types and quantities of organic acids obtained.36,39–41 There are examples of methods which do not utilize extraction prior to analysis,21,34,45,46 but for the purposes of this chapter, the use of extraction will be considered. Headley et al.36 examined various phases and solvents for extraction of NAFCs from OSPW. For liquid-liquid extraction, dichloromethane (DCM) was observed to have the highest total extraction of NAFCs (see Figure 1.3). Hexane was observed to be most selective for classical NAs (see Figure 1.3), as was also observed by Huang et al.41 However, the total amount of components extracted by hexane was approximately 2/3 that of DCM. ENV+ solid phase extraction (SPE) performed relatively well for all NAFCs and had total extraction similar to that of liquid-liquid extraction with DCM (Figure 1.3).36 ENV+ is a solid phase composed of hyper-crosslinked and hydroxylated polystyrene-divinylbenzene. Given this evidence, liquid-liquid extraction with DCM as the organic phase or ENV+ SPE would be good choices for use in a standard quantification method for total NAs. It should be noted that the use of ENV+ SPE would have the additional advantage of aiding in desalting the sample prior to MS analysis, which would benefit quantification.47 10   Figure 1.3 Relative total NAFC extraction using selected solvent systems or SPE. Solvent polarity index is given along the x-axis. Areas represent total area under the TIC curves observed by negative ion electrospray ionization (ESI) Orbitrap MS. [Reprinted with permission from Figure 3, Headley et al. 2013.36 Copyright (2013) Elsevier.]   11   Figure 1.4 Distribution of selected components of NAFC. Extraction using selected solvent systems or SPE was carried out prior to analysis by negative-ion ESI Orbitrap MS. The NAFC component classes refer specifically to the heteroatoms O and S. [Reprinted with permission from Figure 4, Headley et al. 2013.36 Copyright (2013) Elsevier.]  It is commonly known that pH affects the distribution of an ionizable molecule between aqueous and organic phases. This was recently reported for NAs.39–41 Huang et al.41 demonstrated relatively high distribution of O2 NAs into DCM at pH 2.0 and 8.5 and negligible extraction at pH 12.0 (Figure 1.5). Theoretical data from Celsie et al.40 indicates the octanol-water distribution (DOW) of classical NAs is largely unchanged below a pH of 4 and changes significantly at pH values greater than ~6 (Figure 1.6). This indicates that if liquid-liquid extraction is used, both extraction efficiency and repeatability will benefit from a low extraction pH. 12   Figure 1.5 Extracted amounts of selected NAFC classes using six solvents. Extractions were carried out at pH values of a) 12.0, b) 8.5, and c) 2.0. [Reprinted with permission from Figure 1, Huang et al. 2016.41 Copyright (2016) Elsevier.] 13   Figure 1.6 Predicted log DOW values with changing pH for 19 classical NAs. Numbers reference the compounds described in Celsie et al. 2016 SI. [Data taken from Celsie et al. 2016 supplemental information.]40   Furthermore, Celsie et al.40 provide calculated data on octanol-water partitioning of organic acids with varying temperature (Figure 1.7). The data shows changes in partitioning behavior (Kow) with changing temperature, but the effects are considerably less than those occurring with changing pH and do not indicate a need for stringent temperature regulation. According to this information, room temperature (~20 °C) would be appropriate for carrying out sample extractions. 14   Figure 1.7 Predicted log KOW values with changing temperature for 19 classical NAs. Numbers reference the compounds described in Celsie et al. 2016 SI. [Data taken from Celsie et al. 2016 supplemental information.]40  1.2.2.3 Matrix effects and extraction efficiency When sample analysis requires extraction and extensive handling, it is commonplace to spike the original sample with surrogate standards. Surrogate standards allow for an estimate of sample loss and extraction efficiency. For a complex mixture like NAs, surrogate standards are not available for all components. As a result, extraction and ionization efficiencies can only be approximated using a limited number of surrogate standards.21 However, for design considerations, the addition of surrogate compounds to the sample prior to extraction is recommended to provide a guide to the extraction efficiencies of a given method. 15   1.2.2.4 Minimum resolving power  The mass resolving power of a mass spectrometer is known to have a significant impact on the identification and quantification of OSPW organic acids.26,31 In order to resolve classical NAs from other OSPW organic acids, high mass resolving power is necessary.1,21,29 Nominal mass resolution is insufficient. The NA anions C24H41O3 and C25H45O2 have the following monoisotopic m/z values (rounded) in negative mode MS: 377.3061 and 377.3425, respectively. To a unit resolution instrument these would appear as a single m/z peak. Based on the definition of mass resolving power in Equation 1.1, Rm =(mz )a|(mz )a− (𝑚𝑧 )𝑏| Equation 1.1 Mass resolution of a mass spectrometer, Rm, calculated from the ratio of the m/z value of a single peak, (m/z)a, to the difference between (m/z)a and another peak (m/z)b.   the minimum resolving power required for the pair of peaks at m/z 377.3061 and 377.3425 is 10,367.  In a study of OSPW extract utilizing Fourier transform ion cyclotron resonance mass spectrometry (FTICR-MS), Nyakas et al. reported sulfur-containing compounds made up 23% of all identified compounds.29 At lower resolution, sulfur containing compounds present as isobaric with O2 compounds. For example, the anions C25H45O2, C25H45S and C24H41OS have the following monoisotopic m/z values: 377.3425, 377.3247 and 377.2883. To resolve the O2 compound from the two S-containing compounds would require minimum mass resolving 16  powers of 21,199 and 6,962, respectively. Because S- and Ox-species are not the only possible interferences, characterization and fingerprinting studies often make use of ultra-high resolution instruments.20,29,42,44,48   To minimize the possibility of other mass interferences and decreases in resolving power accompanying instrument drift, a minimum resolving power of 50,000 at m/z 200 is suggested for a standard semi-quantification method for total NAs. Both medium-high-resolution (similar to 50,000) and high-resolution (similar to or greater than 100,000) mass spectrometers satisfy this requirement. Both quadrupole-time-of-flight (QTOF) and Orbitrap instruments meet this specification, allowing method design to be realized by a large number of laboratories.   1.2.2.5 Derivatization vs. no derivatization Derivatization of NAs has been carried out in both characterization and quantification efforts.31,35,48 A novel semi-quantification method presented by Woudneh et al. relies on carbodiimide derivatization and exhibits high analytical sensitivity and quantifies specific NA isomer groups as equivalents of a single standard.35 However, the extra step and cost involved in derivatization may distract from adoption of the method by some laboratories. It is noted that the method by Woudneh et al. could potentially play a role in the future quantification of individual NA classes present in calibration standards.  1.2.2.6 Polarity and mode of ionization Because derivatization has not been recommended, available ionization techniques are limited to those associated with direct-injection-MS or LC-MS. For the characterization of classical NAs, 17  ESI has received much attention due to the production of molecular ions from polar compounds with little fragmentation. Atmospheric pressure photoionization (APPI) has also been demonstrated as a useful technique in analysis of NAFCs, especially analysis of less polar compounds.27,39 Thus APPI may play an important role in the quantification of non-acidic classes of NAFCs. Classical NAs, however, being polar and easily ionizable in solution are well suited for ESI.  Pereira et al. demonstrated that positive- and negative-mode ESI result in different populations of NAs being measured, and chromatographic separation suggested the O2 species detected in positive-mode ESI were not classical NAs.44 Thus positive-ion ESI would be unfavorable for the specific analysis of classical NAs. Negative-ion ESI spectra, however, tend to be dominated by Ox species,27 indicating that with sufficient resolving power it is well suited to the analysis of classical NAs. Furthermore, negative-ion mode is already commonly used in analysis of NAs in OSPW, which would ease adoption of the standard method.   1.2.2.7 Choice of calibration standards and use of internal standards The selection of a calibration standard for use in total NAs quantification is not a straightforward decision. Two options for calibration standards are commercial NAs mixtures or NAFCs extracted from OSPW. Commercial mixtures are readily available, but they can differ greatly in composition from NAs in OSPW. Additionally, OSPW NAFCs differ from source to source, and the abundance of compounds other than classical NAs complicates standardization. Grewer et al. showed that commercially available Merichem NA mixtures are dominated by classical NAs, which potentially makes them well suited for a method semi-quantifying O2 species (Figure 1.8).21 Martin et al. compared two high-resolution MS calibration curves for classical NAs, one prepared with Merichem NAs and one prepared with OSPW extract. Both curves exhibited good 18  linearity, but the response of the Merichem NA curve, relative to the internal standard, was approximately three times that of the OSPW extract curve.26 It was suggested this might be due to the fact that OSPW extract contains many compounds other than classical NAs. Calibration curves made from Merichem NAs and OSPW extract might be similar to each other by gravimetric or FTIR quantification, but the responses will differ considerably when analyzed by a method such as high-resolution MS which can distinguish classical NAs from other compounds in the mixture. While this is an area that requires further investigation, the predominant O2 nature of Merichem NAs observed by Grewer et al. might lend them well to the semi-quantification of classical NAs in OSPW.   Figure 1.8 Relative abundance of compounds matching the formula CnH2n+ZOx, summed for n = 8 to 30 and Z = 0 to -12. [Reprinted with permission from Figure 4, Grewer et al. 2010.21 Copyright (2010) Elsevier.]  19  The use of at least one internal standard during MS analysis is proposed. The reasoning of Brunswick et al.21 has been adopted, in that the variety of compounds in OSPW is too great to be matched with any reasonable number of internal standards, and that a single standard may be used for monitoring instrument performance. Because of the complexity of the mixture, it is not possible to completely account for the effects of ion suppression. However, these effects can be significantly reduced using chromatographic separation.  1.2.2.8 Online or offline fractionation of samples Prefractionation of OSPW extract has been shown to enhance characterization and is commonly used in the analysis of OSPW.21,29,42,43 Although a minimum mass resolving power for mass spectrometers has been suggested that would ensure reduced interference with classical NAs, the use of a separation technique could help further resolve potential molecular interferences, reduce matrix effects and aid in sample desalting prior to MS, considerably reducing the effects of ion suppression. On this point, Nyakas et al. demonstrated that off-line prefractionation of OSPW extract by UHPLC prior to analysis by FTICR-MS resulted in a nearly 200% increase in the number of assigned compounds relative to direct injection (Figure 1.9).29 Off-line fractionation was used by Nyakas et al. only because of software limitations. HPLC is commonly interfaced with the suggested mass spectrometers (QTOF or Orbitrap), and software compatibility with on-line detection on these instruments should not be an issue. Choice of column, mobile phases and gradients is not considered here, but it is recognized that they are important factors to be considered in method development.  20   Figure 1.9 Bar plots and contour diagrams showing the number of homologues detected for the formula CnH2n+ZOx for: a) OSPW sample extracted by liquid-liquid extraction and directly injected into FTICR-MS; b) OSPW sample extracted as in (a) and fractionated by UHPLC prior to FTICR-MS; and c) combination of two OSPW samples processed as in (b) to compensate for dilution effects. [Reprinted with permission from Figure 3, Nyakas et al. 2013.29 Copyright (2013) American Chemical Society.]  1.2.3 Conclusions A standard method for the semi-quantification of total NAs is needed to provide a reference point for the multitude of different methods used in NA quantification. A range of studies on a) b) c) 21  specific aspects of NA and NAFC analysis were compiled, and important details of currently used methods were discussed. Requisite features were suggested for the design of future standard MS methods for quantifying total NAs. The design considerations proposed are suitable for use with both QTOF and Orbitrap mass spectrometers.  1.3 Capillary electrophoresis-electrospray ionization-mass spectrometry 1.3.1 Electrophoretic mobility and electroosmotic flow In the previous section, consensus was reached among a group of scientists and experts on LC-MS as being an acceptable technique for the development of semi-quantification methods for total NAs. In this section, capillary electrophoresis-mass spectrometry (CE-MS) is introduced as a powerful, complementary technique for the analysis of NAFCs. Capillary electrophoresis (CE) is an open-tube liquid phase separation technique which uses a high electric voltage to drive separation of analytes. CE can be used in a variety of modes that can separate positively charged, negatively charged and neutral compounds, and employ a variety of additives, injection types and liquid junctions. CE is also known to have very high separation power and creates significantly less operational waste (nanolitres) than other separation methods. For the conventional operating mode of CE, termed “capillary zone electrophoresis” (CZE), a fused-silica capillary of inner diameter between 25 and 75 micrometers is filled with a liquid called background electrolyte (BGE), several nanolitres of analyte mixture are injected onto the capillary column, and high voltage (e.g. 30 kV) is applied across the capillary, with the simultaneous introduction of more BGE solution. The application of voltage through the BGE causes the analytes in the sample plug to migrate in solution—cations moving toward the cathode, anions toward the anode and neutral compounds 22  stationary—in a process called electrophoresis. The electrophoretic mobility, µep, of a charged particle in solution can be calculated from the electropherogram thus 𝜇𝑒𝑝 =𝑣𝑒𝑝𝐸=𝐿𝑑𝐿𝑐𝑡𝑉 , Equation 1.2 Electrophoretic mobility as a function of analysis time (t), capillary length to detector (Ld) and electric field strength (V/Lc). This equation can be used to modulate data between the time and mobility domains. where v is the velocity of the analyte through the capillary (cm/sec) and E is the electric field strength. For complex mixtures, each analyte has a unique electrophoretic mobility. The analyte velocity can be replaced with the length (cm) to detector (Ld) per analyte migration time (t), and electric field strength can be replaced with the applied voltage (V) per capillary length in cm (Lc). This equation can be used to modulate CE-MS data between the time domain and the mobility domain.49 The electrophoretic velocity of an analyte can also be described on a particle level:50 𝑣𝑒𝑝 =16𝜋𝜂𝑅(𝑄𝑒𝑓𝑓𝐸 + 𝑘3 + 𝑘4) Equation 1.3 Electrophoretic velocity as a balance of four forces exerted by the applied electric field (QeffE), Stokes drag on a spherical particle (6πηRv), and two forces, k3 and k4, which are drag and relaxation forces caused by the ionic atmosphere in solution.  where Qeff is the effective charge of the spherical particle, E is the strength of the applied electric field, R is the hydrodynamic (Stokes) radius of the particle and 𝜂 is the dynamic viscosity of the liquid surrounding the particle. The equation is a force balance equation of four forces: Stokes drag on a spherical particle, movement of ions due to the applied electric field, and two forces, k3 23  and k4, which are drag and ionic relaxation forces caused by the ionic atmosphere in solution. This equation illustrates that the electrophoretic mobility of a small molecule is roughly proportional to its charge-to-size ratio. For complex mixtures of unknown composition, the vep values of all analytes are affected not only by the applied electric field strength, but also by dynamic complexation with other molecules or phases (e.g. micelles) in the mixture. If the capillary wall bears excess charges, a “sheath” of opposite charges in the BGE will assemble near the capillary wall. With the onset of voltage, the charged sheath will mobilize in the direction of the oppositely charged electrode, dragging the bulk solution with it. This phenomenon is called “electroosmotic flow” (EOF), the velocity of which, veof, has been famously described by von Helmholtz and Smoluchowski and is shown in Equation 1.4.51 𝑣𝑒𝑜𝑓 =𝜖𝜁𝜂𝐸 Equation 1.4 The electroosmotic flow of a charged layer is proportional to the zeta potential and the applied electric field strength and is inversely proportional to the solution viscosity.  Here, ε is the product of the solution dielectric constant and the permittivity of the vacuum, ζ is the zeta potential (i.e. the electric potential at the boundary of the charged layer) and η is the viscosity of the solution. The velocity of the EOF is proportional to the zeta potential and the applied electric field strength and is inversely proportional to the solution viscosity. Often, EOF is difficult to control, so EOF is eliminated by neutralizing the charge of the capillary wall, and, in that case, a pressure would be applied to move the bulk solution toward the detector at the outlet. However, EOF can be advantageous: if the magnitude of EOF (i.e. the bulk flow) toward the outlet or mass spectrometer is larger than the magnitude of electrophoretic mobilities of 24  analytes in the opposite direction, the analytes will still be detected without having to introduce a pressure push along the capillary, thus eliminating peak broadening due to a laminar flow profile.  The work described in this thesis uses a CE-ESI-MS setup where  1. the analytes of interest (amine-derivatized naphthenic acids) are cationic in acidic BGE, meaning their net electrophoretic mobilities are away from the positive voltage electrospray (ESI) device, 2. given the complex mixture of amine-derivatized NAs all bear the same charge, and via Equation 1.3, the large amine-derivatized NAs will migrate slower away from the positive ESI and small amine-derivatized NAs will migrate faster away from the positive ESI. 3. However, the bulk flow of the contents in the capillary is toward the positive ESI device due to EOF. An immobile cationic polymer coating on the inside of the capillary walls caused a layer of anions in solution to form close to the capillary wall during CE separations, causing EOF to drive the bulk solution toward the positively charged ESI device and into the MS for detection. 4. The resulting data form is a swath of MS peaks of amine-derivatized NAs where large amine-derivatized NAs are detected by MS first, and the smallest amine-derivatized NAs are detected last. Examples of the swath of peaks can be found in Figure 3.4, Figure 3.5, Figure 3.10 and Figure 4.6 (zone C). A schematic of the CE-ESI-MS setup is shown in Figure 1.10.  25   Figure 1.10 Schematic diagram of the CE-ESI-MS setup as used in this thesis. The ESI device acts as the positive electrode in CE-ESI-MS. Large NAs are detected first and small NAs are detected last due to their electrophoretic mobilities (blue arrow) being in the opposite direction of the EOF-directed bulk flow (red arrows).  1.3.2 Electrospray ionization-mass spectrometry apparatus  In order to introduce the CE separation materials at the outlet into the mass spectrometer, an electrospray apparatus developed in our lab was used.52–58 The apparatus is a three-piece, stainless-steel electrospray needle with a bevelled tip, and it accommodates two capillaries: the separation capillary and the modifier capillary. The separation capillary is inserted through the T-union and into the needle tip. The modifier capillary is introduced into the lower part of the T-union. The insertion of a flat-cut capillary tip into the bevelled electrospray needle results in the creation of a small dead space. This small dead space is termed the “flow-through microvial” (Figure 1.11). Upon introduction of separation and modifier solutions to the electrospray needle apparatus, the dead space is filled and the two solutions meet in the flow-through microvial, but engage in very minimal mixing.58 From here, the solution exits the needle and undergoes 26  electrospray ionization at the sharpest point on the needle tip, subsequently entering the mass spectrometer.   Figure 1.11 Schematic of electrospray ionization apparatus which utilizes a flow-through microvial and bevelled tip geometry.52     CE-ESI-MS analysis, as outlined in this thesis, produces time-dependent full-scan MS data (i.e. 3-dimensional data) which consists of a series of mass scans with predefined mass-to-charge ratio (m/z) range (y-axis) over the course of usually 10 to 40 minutes (x-axis) producing variations in ion counts or “peaks” (z-axis). The resulting data is termed a “mass electropherogram” and contains a map of ion signals detected by the mass spectrometer over the duration of CE separation time. Analyte migration time in CE is inversely proportional to analyte electrophoretic mobility. Analyte electrophoretic mobility is roughly proportional to the charge-Flow-through microvial Separation capillary T-union Modifier capillary Inlet vial Modifier vial 27  to-size ratio of the analyte. This is in contrast with liquid chromatography (LC), where analyte migration time (termed “retention time” in chromatography) is not correlated with the application of a field but is related only to the differential partitioning of the analyte between the mobile and stationary phases. Therefore, the migration time values of analytes in CE will not match up with LC retention times, nor with order of elution, for the same compounds, and CE will produce complementary information to LC (and other separation techniques) concerning analytes in the mixture. In terms of the analysis of unknown complex mixtures, data complementarity is an essential prerequisite for effectively directing research programs toward the timely and cost-effective identification and quantification of molecular components.  The coupling of CE to mass spectrometers via electrospray ionization (ESI) sources has made an impact in many fields of research including genomics, metabolomics and pharmaceuticals.59–62 CE-MS has shown important complementarity with LC-MS in proteomics and metabolomics research.63 The most important advantage of CE, when compared to separation techniques such as LC or GC, is that its mechanisms of separation are rooted in electrokinetic theory which is based on fundamental parameters, whereas LC and GC understand separations in terms of empirically-derived scales (e.g. Kovats). Other practical advantages of CE include (a) the possibility of high separation efficiency in order to resolve complex isomeric mixtures; (b) versatility in its ability to resolve a wide range of analytes (including neutral species) through the addition of additives to the BGE and the use of on-column liquid junctions; (c) low sample consumption (nanolitres) and low solvent consumption, which make it suitable for high-volume, routine analyses, method optimizations and environmental applications. Further, CE has high potential for miniaturization in the development of microfluidics devices.64  28   The contribution of this thesis to the field of complex mixtures analysis includes the first capillary electrophoresis method, which relies on fundamental electrokinetic parameters, for the analysis of NAFCs. The CE-ESI-MS analysis of NAFCs produces new, complementary information to other analytical techniques and aids in the semi-quantification of NAFCs by separating them from non-NAFCs in complex mixtures. This allows a variety of data processing approaches (e.g. PCA or FA) to be utilized to increase our understanding of complex mixtures of unknown environmental contaminants.  Before illustrating CE-ESI-MS method development for the analysis of unknown complex mixtures of NAFCs, the feasibility of CE-ESI-MS for the analysis of a relatively well-known complex mixture, human urine, will be demonstrated.   29   Capillary electrophoresis-mass spectrometry for targeted and untargeted analysis of the sub-5 kDa urine metabolome of patients with prostate or bladder cancer: A feasibility study   2.1 Introduction  With the advent of metabolomic profiling, the search for relevant biomarkers for a variety of diseases has intensified dramatically. In particular, the effectiveness of using metabolomic biomarkers as screening tools in various fields of oncology is being continuously investigated.65–67 In 1987, the measurement of prostate-specific antigen (PSA) was introduced as a diagnostic biomarker for prostate cancer (PCa) by Stamey et al.68 Although there have been mixed reports on the diagnostic effectiveness of PSA, the general consensus is that PSA is useful for the diagnosis of PCa and it has become a mainstay for screening.69,70 Although a variety of biofluids have been considered, urine remains an attractive source of potential biomarkers because of its abundance and non-invasive sampling. For the past decade, sarcosine (N-methylglycine) has been the target of multiple investigations concerning its relationship to the presence and progression of PCa. The first comprehensive report of the correlation between increased urine concentration of sarcosine and PCa progression was made by Sreekumar et al.71 In their study, over 1,000 metabolites were profiled on prostate tissue, urine and plasma samples, using LC-MS and GC-MS. Since then, a stronger push has been made to incorporate sarcosine into a multiplex marker in urine for detecting the presence and progress of PCa.71,72  30   In contrast, an epidemiological study from Norway published in 2014 indicated that higher serum concentrations of sarcosine and glycine were associated with reduced PCa risk in a patient pool of 6,000 individuals.73 For many other cancers, such as bladder cancer (BCa), no single biomarker has maintained significant diagnostic status, and studies have implicated a wide range of metabolites.67,74 In 2012, Soliman et al. published a validated capillary electrophoresis-electrospray-tandem mass spectrometry (CE-ESI-MS/MS) method used for quantifying endogenous levels of six underivatized amino acid metabolites, including sarcosine, in pooled healthy urine samples, employing only a single centrifugation step.55 Applying the same methodology to individual urine samples from subjects with pathologically confirmed cancers poses a different analytical challenge, from sample collection to analysis to diagnosis. In the presented study, underivatized urine metabolites in patient urine samples are separated by CE and detected by MS.   2.1.1 Urine sampling in the surgical workflow  The Vancouver Prostate Centre implements a Genitourinary Biobanking Management System for storing tissue samples from prostatectomy and integrates digital histopathology with epidemiological information. The urine samples in this study were collected from patients undergoing prostatectomy or cystectomy. Figure 2.1 shows a schematic for the integration of pathological and diagnostic data in the Genitourinary Biobanking Management System with metabolomics data obtained from this study. The inclusion of digital pathological information into the overall data structure allows for relevant spatial pathological data, such as tumour proximity to ureter (in the case of BCa), to be correlated to urine metabolite data, increasing the diagnostic accuracy of metabolomics. 31   Figure 2.1 Descriptive flowchart of the hierarchical data structure of this study. Each node in the flowchart represents a type of information. The green nodes refer to medical information directly used for the diagnosis of cancers. The orange node represents the biological information accessible via chemical methods. The yellow nodes represent the analytical chemical information produced in this study. The grey, dashed arrow refers to how CE-ESI-MS information can be seen as contextual biological information, and to show the connection between chemical analysis and the grander goal of diagnosis.  2.1.2 Comparison of analytical methodologies  In the two studies most relevant to the present study,71,73 the analytical methodologies used differ from each other and from those presented here, despite all being some type of separation-mass spectrometry. We place the current methodology alongside for comparison (See Table 2.1).  32  Sreekumar et al. (2009)71 de Vogel et al. (2014)73 The present study 110 patients studied  (59 PCa + 51 control) 6,000 patients studied  (3,000 PCa + 3,000 control) 4 patients studied (PCa, BCa) and 1 control (20 samples pooled)  Prostate tissue, serum and urine studied (archival samples) Blood serum studied (archival samples) Urine from PCa and BCa studied Creatinine not used as marker of renal function -- Metabolite variation calculated as a per patient  Serum creatinine used as marker of renal function Urine creatinine used as marker of renal function   Samples obtained from University Tissue Banks. Storage information not available, but assuming appropriate measures taken. Samples collected and stored for up to 29 years with appropriate measures taken.  Patient samples collected via catheter and stored for less than 3 months; healthy volunteers collected midstream and stored for less than 1 year. Appropriate measures taken. LC/GC-MS and bistrimethylsilyl-trifluoroacetamide derivatized GC-MS for untargeted metabolomics -- isotope dilution with t-butyl dimethylsilyl derivatized metabolites for targeted metabolomics LC-MS for some metabolites (C platform) -- methylchloroformate derivatized GC-MS/MS (B platform) for other metabolites -- Only targeted metabolomics CE-ESI-MS used for all metabolites in both targeted and untargeted metabolomics -- No derivatization -- SPE for sarcosine enrichment;  No storage information available [Choline] increase due to storage; [Serine] increase due to storage No suspected decrease/increase in metabolite levels Gleason scores and staging available Gleason scores not available; Disease staging available in 61% of cases Gleason scores, staging, and grading available  Table 2.1 A brief comparison of some important aspects of analytical chemistry in two large studies as compared to the present study. 33   2.2 Materials and Methods  2.2.1 Ethical statement  All patients with PCa and/or BCa who were involved in the Vancouver Prostate Centre Genitourinary (GU) Biobanking Project signed an ethically approved consent form for collection and subsequent analysis of urine. These consent forms were collected by the Vancouver Prostate Centre.  2.2.2 Urine sample collection   According to Saude and Sykes, the bacterial levels and the number of freeze-thaw cycles constitute two major confounding factors to the chemical integrity of urine, especially in terms of metabolomic studies.75 Urine samples were self-obtained from 20 healthy volunteers between the ages of 23 to 30 years and were mixed to form the pooled urine sample. The samples were self-collected presumably midstream urination with no extra chemical sterilization of the vessel prior to sample collection. Each of the 20 healthy volunteer samples were mixed together. The pooled urine sample was divided into 1 mL aliquots and stored at -20 °C on-site (Table 2.2).     34    Pooled urine sample from 20 healthy patients Urine samples from 5 patients with pathologically confirmed cancers Subject preparation No preparation pre-surgical 2% chlorhexidine topical rub Sampling method One-time, midstream, no chemical sterilization of vessel -- Approximately 50 mL One-time, cathetered, midstream after first 5 mL, no chemical sterilization of vessel -- Approximately 50 mL Time between sample collection and freezing < 1 hr Cooling “on ice” for Prostate patients: 3 hr Bladder patients: 2 hr Transported frozen Freezing temperature  -20 °C -80 °C Freeze-thaw cycles prior to analysis Collect-freeze-thaw-filter (5 kDa MWCO)-freeze-thaw-analyze  -- 2 full cycles Collect-freeze-thaw-filter (5 kDa MWCO) -freeze-thaw-analyze -- 2 full cycles Sodium azide (NaN3) preservative No -- Electrophoresis buffer likely preserved sample during analysis No -- Electrophoresis buffer likely preserved sample during analysis  Table 2.2 Summary of urine sample collection and preparation protocols for patient urine used in this study.   The five cancer patient urine samples were collected at prostatectomy or cystectomy by urologists at Vancouver General Hospital, obtained from patients possessing pathologically confirmed carcinoma of either the bladder or the prostate, after application of surgical skin rub of 2% chlorhexidine in alcohol. Urine was divided into 1.5 mL aliquots and stored at -80 °C. The 35  urine actively used in analysis was delivered into 2 mL vials and stored at -20 °C for practical purposes.   2.2.3 Pathological analysis of cancer tissues  The pathological analysis of neoplasms constitutes a type of spatial information. This fact has been taken advantage of with the recent rise in applications of imaging technology to histopathological samples and diagnosis.76–80 Histopathological evaluation is generally divided into staging and grading, where staging refers to an assessment of the development of cancer spreading throughout the body and grading refers to the cells’ appearance in a tumour, or other carcinoma-specific qualities.  The localized BCa were excised from the bladder and sectioned. Entire prostate glands were sectioned. Tissue sections were stained with hematoxylin and eosin (H+E) and analyzed under light microscope. Digital images of the paraffin mounts were stored in the Distiller SlidePath data management system (Leica Biosystems). Patient A had superficial BCa (transitional cell carcinoma, or TCC), the most common form of BCa in Americans, which follows three general stages.81 It is also graded pathologically from I (least developed) to III (fully developed). Squamous cell carcinoma of the bladder, exhibited in Patient B accounts for about 5% of American cases and represents cancer present in the squamous epithelium of the bladder.81 For Patients C, D, and E, the Gleason score was applied to the stained prostate tissue used to estimate the grade of PCa.82 Patient C was not studied here due to the presence of blood in the urine sample. Table 2.3 shows the diagnoses of the patients in this study.   36  2.2.4 Patient Diagnoses Patient diagnoses at prostatectomy/cystectomy as pertaining to this study are given in Table 2.3. Patient C was not analyzed due to the presence of blood in the urine. Reference tables for BCa staging, BCa grading and Gleason score (GS) grading system for PCa can be found in Appendix B3, Staging and Grading of Cancer Histopathology.  Patient Diagnosis A Low-grade papillary adenocarcinoma of the bladder (TCC) B Squamous cell carcinoma of the bladder;  Invasion of prostate gland  C* Prostate Cancer; GS 3+4= 7/10, PSA = 11; T2b Lymphnodes = negative D Prostate Cancer; GS 4+3 = 7/10, PSA = 6.9; T2b Lymphnodes = negative E Prostate Cancer; GS 3 + 4 = 7/10, PSA =16.5; T2c Lymphnodes = negative Invasion of seminal vesicles  Table 2.3 Patient diagnoses at prostatectomy/cystectomy as pertaining to this study. Patient C was not analyzed due to the presence of blood in the urine.  37  2.2.5 Chemicals and reagents All metabolite standards with minimum of 98% purity were purchased from Sigma Chemical Co. (St. Louis, MO, USA). Internal standard sarcosine-d3⋅hydrochloride (“d3-sarcosine”) was obtained from Toronto Research Chemicals Inc. (North York, ON, Canada) and Sigma Chemical Co., respectively. Formic acid (88%), NaOH, HCl, methanol (HPLC grade), and glacial acetic acid were purchased from Fisher Scientific (Nepean, ON, Canada). Reagent grade picric acid (98%, 35% water) for creatinine analysis was also obtained from Sigma Chemical Co. (St. Louis, MO, USA). Capillary coating reagent of trimethoxysilylpropyl-modified polyethylenimine (PEI), 50% in isopropanol, was purchased from Gelest Inc. (Morrisville, PA, USA).   2.2.6 Preparation of standard stock solutions Leucine and kynurenine were dissolved in varying volumes of 0.1 M HCl and diluted to volume with water. Because the analysis of sarcosine is separate from the analysis of the other amino acids, there were two standard solutions series: The series of standard solutions for sarcosine—containing sarcosine, and internal standard d3-sarcosine—and the standard solution mixture containing Pro, Cys, Leu, Glu, Kyn, and internal standard d3-Sarc. The pH of these solutions was 2.3. The concentrations of the stock solutions are listed in Table 2.4.     38  Metabolite Stock Concentration x200 x100 pKa Sarc  1.000 mM 0.005 mM 0.010 mM 2.06 Metabolite Stock Concentration x64 x32 pKa Pro 0.0397 ± 0.0001 mM 0.00062 mM 0.001241 mM 1.99 Cys 0.2002 ± 0.0006 mM 0.003128 mM 0.006256 mM 1.96 Leu  (HCl, < 0.1 M) 0.1598 ± 0.0004 mM 0.002497 mM 0.004994 mM 2.36 Glu 0.1999 ± 0.0000(2) mM 0.003123 mM 0.006247 mM 2.19, 4.25 Kyn  (HCl, <0.1 M) 0.007995 ± 0.000624 mM 0.000125 mM 0.00025 mM 1.19 Table 2.4 Stock standard concentrations, general make up, and concentrations. All standards were made up in water, except for leucine and kynurenine, which were dissolved in 0.1 M HCl.  2.2.7 Calibration by standard addition method  The analysis of sarcosine in patient urine required solid-phase extraction, generation of standard curve, and a three point calibration, as described previously.55 Calibration standard mixtures of the metabolites were prepared. For the non-sarcosine metabolite analysis, all urine samples were filtered through Amicon® Ultra-15 5 kDa Centrifugal Filter Units (Billerica, MA, USA) in a fixed-angle centrifuge at 4000 ġ for 20 min at around 25 °C. An extra several minutes at 10,000 ġ was applied to decrease retentate volume.  39  2.2.8 Measurement of creatinine  A Jaffé reaction-based assay on a 96-well plate (8 x 12) was employed using an alkaline picrate solution of 0.00833 M picric acid in 1.67 M NaOH (excess base) to produce a bright yellow solution. Serial dilutions of creatinine standard were added in triplicate to each patient urine sample. The samples were incubated at room temperature for 30 min and subsequently analyzed on a Beckman Coulter DTX880 multimode reader set at 450 nm in absorption mode. The Jaffé reaction between creatinine and alkaline picrate and has been known for over 100 years and has been used for over 50 years in clinical urinalysis.83 With an excess of hydroxide ions, the reaction between creatinine and picrate forms a mixture of stereoisomers of a Meisenheimer σ-complex with 2:3 ratio of creatinine to picrate. The reaction is thought to proceed via deprotonation of the creatinine methylene group to form a creatinine anion, followed by nucleophilic addition at the picrate meta position to form a creatinine-picrate σ-complex.84 The likely molecular structure of the complex is shown in Figure 2.2. Pseudo-first order rate constants of the Jaffé reaction depend linearly on picrate and creatinine concentrations, but nonlinearly on hydroxide concentration, suggesting the possibility of a mixture of several compounds at high [OH-].85  40  Alkaline Picrate CreatinineO–NOONOONO ONNHNHOCH3HHOH–NC–NHNHOCH3HO–NOONOONO OO–NOONOONO OHHNC–NHNHOCH3HNNHNHOCH3O–NOONOONO O–HOH2NNHNHOCH3NNHNHOCH3ONOONOONO OHH–––7Na+NNNOCH3NNNOCH3ONOONOONO OO–NOO NOONO OO–NOONOONO OHH–––––2Na+3Na+Na+Na+Creatinine, OH-Alkaline Picrate, OH- Figure 2.2 Molecular structure of the Jaffé red chromogen depicted as a 2:3 Meisenheimer σ-complex. The kinetics of this reaction depend non-linearly on hydroxide at high pH.  41  2.2.9 Instrumentation 2.2.9.1 Capillary electrophoresis system and software  All experiments were carried out on a PA 800 Plus capillary electrophoresis system (Beckman Coulter, Brea, CA, USA) connected to an AB SCIEX API 4000 triple quadrupole mass spectrometer (AB SCIEX, Framingham, MA, USA). Nitrogen (UHP) was used as curtain and collision gas. A modified capillary cartridge that could accommodate the unique electrospray setup was used for CE.52 All data acquisition, system control, and integration were performed with Analyst® 1.4.2 software (AB SCIEX, Framingham, MA, USA).  2.2.9.2 Electrophoretic procedure   CE separations were carried out on a 50 µm inner diameter (I.D.) × 365 µm outer diameter (O.D.) × 85 cm length (L) fused silica capillary (Polymicro Technologies, Phoenix, AZ, USA) coated with cationic polymer trimethoxysilylpropyl polyethyleneimine-HCl in isopropanol (50% v/v). For targeted CE-ESI-MS/MS, a chemical modifier was introduced through a 75 µm (I.D.) × 365 µm (O.D.) × 80 cm (L) bare fused silica capillary set at 266 nL/min flow rate. For untargeted CE-ESI-MS analysis, the chemical modifier was introduced via a bare fused-silica capillary (75 μm internal diameter, total length of 80 cm) at a flow rate of ~300 nL/min.   For detection of sarcosine in patient samples by CE-ESI-MS, the BGE consisted of 0.5% v/v formic acid, 50% v/v methanol and 49.5% v/v water, and the modifier solution consisted of the same. For detection of non-sarcosine amino acid metabolites, the BGE consisted of 2% v/v formic acid, 50% v/v methanol and 48% v/v water, with a modifying solution of the same composition. The same composition was used for untargeted CE-ESI-MS analyses. The 0.5% v/v formic acid buffer has the capability to fully resolve sarcosine from its alanine isomers.55 The 42  separation of non-sarcosine amino acid metabolites was performed in a 2% formic acid solution in 50% methanol. The samples were injected at 1 psi for 10 sec and a voltage of -30 kV was applied. Between injections, the separation capillary was rinsed at 40 psi with methanol for 5 min, air for 1 min, 0.1 M HCl for 5 min, H2O for 5 min, and buffer for 5 min.   2.2.9.3 ESI-MS/MS and ESI-MS Positive mode electrospray ionization was achieved using a bevelled needle tip geometry with a flow-through microvial.52 The multiple ion monitoring (MRM) experiment was carried out on API 4000 triple quad mass spectrometer (AB SCIEX). The MRM parameters were manually set according to Soliman et al (2012) for targeted analyte quantification and are listed in Table 2.5. Q1 and Q3 are the masses of ions accepted into Q1 and Q3 and which are unique to the compound of interest. The dwell time is how long ions are held between transitions. DP is the declustering potential parameter which is applied to the MS orifice to help prevent ions from clustering together during ESI. EP is the entrance potential which is set to focus the ion beam before it enters Q1. CE is the collision energy parameter that is applied to Q2 and accelerates ions to collide with the collision gas in order to cause fragmentation before entering Q3. CXP is the collision cell exit potential parameter that focuses and accelerates ions out of Q2 and into Q3.      43   Q1 Q3 Dwell Time (msec) DP (V) EP (V) CE (V) CXP (V) Sarc 90 44 20 40 10 19 3 Pro 116 70 20 35 10 20 10 CysCys 241.2 152.2 20 35 10 19 10 Leu 132.1 43 20 40 10 37 8 Glu 148.1 84.1 20 45 10 21 7 Kyn 209.2 94.1 20 55 10 19 7 d3-Sarc 93 47 20 40 10 19 3  Table 2.5 Multiple reaction monitoring conditions for targeted metabolite analysis via QQQ-MS.  2.2.10 Data Analysis 2.2.10.1 Targeted metabolite data analysis  For the targeted metabolite analysis, peak areas, migration times, and peak height were calculated using a manual algorithm in the Analyst® 1.4.2 software. Peak area ratios of metabolite to internal standard d3-sarcosine were used to construct a three-point calibration curve (no dilution, 64x dilution and 32x dilution) and the x-intercept was interpreted as the vial concentration. The endogenous metabolite concentration was calculated from this.  44  2.2.10.2 Non-targeted metabolite data analysis  For untargeted analyses, MsXelerator software (MsMetrix, Maarssen, The Netherlands) was used to process the full scan data. The obtained metabolic profiles were aligned using the reference peak warping function with the m/z values 76.3, 90.5, 106.5, 122.6, 144.5, 147.6, 162.7 and 186.4 as reference compounds. The maximum allowed difference in m/z between the different metabolite profiles was 0.2 Da. A peak picking procedure was used to determine compounds present in a reference urine sample with a signal-to-noise ratio of 3 and having a migration time between 5 and 30 min. Peak matching was carried out to determine peak areas and heights of common compounds present in all urine samples. Compounds were considered matching if the migration time difference was maximum 0.2 min. Peak areas were normalized to the sum of all peak areas to correct for the difference in urine volume. A table with all normalized areas for the detected urinary compounds was used for principal component analysis (PCA). The PCA loading plot was used to determine discriminatory compounds responsible for the separation of urine obtained from healthy people and patients with PCa or BCa.  2.3 Results 2.3.1 Targeted analysis  Figure 2.3 shows, as an example, the set of extracted ion chromatograms (XIC) for Patient B, including the XIC for sarcosine. For each standard addition duplicate run (n = 2), peak area ratios (PAR) of targeted metabolites to internal standard (d3-sarcosine) were calculated, averaged, and plotted against the standard concentration.   45   Figure 2.3 Set of extracted ion chromatograms (XIC) from Patient B obtained by CE-ESI-MS/MS, multiple reaction monitoring (MRM) mode. Notice the relative intensities of sarcosine and alanine and the magnified inset.  The absolute value of the x-intercept was interpreted as the vial concentration, which was used to calculate the patient’s endogenous metabolite concentration, shown in Table 2.6, along with standard addition calibration curve equations and R2 values shown in Table 2.7. Calibration curves for each metabolite are listed in Appendix B1.  46   Endo (μmol/L) Sarc RSD L-Pro RSD L,L-CysCys RSD L-Leu RSD L-Glu RSD L-Kyn RSD A 13.18 9% 7.69 16% 132.8 23% 73.01 15% 11.17 28% 0.11 105% B 5.21 21% 4.50 25% 118.1 46% 16.92 41% 21.71 19% 0.11 252% D 9.81 14% 2.94 18% 268.2 15% 17.93 19% 11.94 28% 0.092 413% E 36.21 20% 11.34 19% 648.0 76% 58.89 10% 42.66 21% 1.65 31% POOLED 0.81   5.99   57.67   17.74   21.55   1.78    Table 2.6 Endogenous concentrations of the various metabolites in patient urine derived from the x-intercept of a three-point standard addition method with duplicates, using multiple reaction monitoring (MRM) CE-ESI-MS/MS.  Metabolite Patient Slope Y-intercept R2 Metabolite Patient Slope Y-intercept R2 Pro A 1.67 0.86 0.9697 Glu A 0.80 0.60 0.9936 B 1.97 0.59 0.9745 B 0.70 1.02 0.9869 D 1.72 0.34 0.9947 D 0.80 0.64 0.9923 E 1.61 1.21 0.9186 E 0.85 2.41 0.9391 CysCys A 0.05 0.41 0.7803 Kyn A 3.83 0.03 0.9944 B 0.04 0.35 0.4205 B 3.68 0.03 0.9710 D 0.03 0.57 0.8913 D 3.86 0.02 0.9471 E 0.03 1.40 0.2774 E 2.53 0.28 0.8796 Leu A 0.10 0.46 0.9039 Sarc A 0.12 0.11 0.9804 B 0.19 0.22 0.9434 B 0.13 0.04 0.9850 D 0.14 0.17 0.9868 D 0.13 0.08 0.9773 E 0.14 0.55 0.9622 E 0.10 0.23 0.8177  Table 2.7 The regression line equations and R2 values calculated from the standard addition calibration used to calculate endogenous concentrations of the various metabolites in patient urine.  The standard deviation of the x-intercept was calculated according to the following equation:86   47  sx =sy|m|√1k+ñ(x-int)2 + ∑(xi2) − 2(x-int) ∑ xiD Equation 2.1 Formula used to calculate standard deviation of the extrapolated x-intercept of a standard addition calibration, where k is a constant that equals 2.   Where m is the slope of the calibration curve, ñ is the number of data pairs in the regression line, D is ñΣxi2 − (Σ𝑥𝑖)2, and k is the number of replicate measurements of the unknown. For this experiment, k = 2, and RSD was calculated as  RSD =sxx-int . Equation 2.2 Formula to calculate relative standard deviation (RSD) in a standard calibration, where sx is the standard deviation of the x-intercept, as calculated in Equation 2.1, and x-int is the value of the x-intercept.  Endogenous metabolite concentrations normalized to creatinine are shown in Table 2.8.  μmol/mmol Creatinine Pro CysCys Leu Glu Kyn Sarc A 1.07 18.48 10.15 1.55 0.016 1.83 B 0.416 10.91 1.56 2.01 0.010 0.48 D 0.621 56.70 3.79 2.52 0.019 2.07 E 0.322 18.40 1.67 1.21 0.047 1.03 POOLED Healthy 0.740 7.140 2.190 2.660 0.220 0.100  Table 2.8 The endogenous concentrations of metabolites in patient urine (μmol/L) divided by the measured creatinine (mmol/L). 48   2.3.2 Untargeted Analysis  The untargeted analysis of the urine samples resulted in the detection of 468 compounds. The repeatability of the CE-MS method for untargeted urine analysis was assessed by calculating relative standard deviations (RSDs) for the peak area of seven representative urinary compounds spanning the complete metabolomic profile. RSDs were acceptable with values ranging from 2.4 to 16.9%. Peak areas corresponding to compounds were normalized based on the sum of the areas of all detected peaks within a urine sample. Principal component analysis (PCA) using 6 components was performed on the normalized peak areas, and the score plot is shown in Figure 2.4. A clear distinction between the urine samples from the healthy volunteers and from the PCa patients is evident. Alterations in urine concentration of at least nine compounds are responsible for this discrimination. 49   Figure 2.4 Plot of first 2 principal components (6 PC’s total). The control pooled urine samples (blue circles) were identical in composition; whereas the diseased urine samples (red squares) are each from patients A, B, D, and E.  Four compounds showed a lower concentration, whereas the concentrations of five compounds were higher in the urine samples from the patients (Table 2.9). However, the normalized intensities of m/z value 180 and 212 for one patient appeared to be similar to the intensities of these compounds in the control samples.   50  Compound (m/z) Migration time (min) ↓/↑ 132 16.25 ↓ 144 11.41 ↓ 156 19.80 ↓ 162 17.76 ↓ 180* 6.50 ↑ 212* 6.53 ↑ 329 8.58 ↑ 346 8.61 ↑ 389 8.69 ↑  Table 2.9 A list of urinary compounds that were lower (↓) or higher (↑) in concentration in patient samples as compared to control samples. *Intensities of these m/z peaks for one of the patient samples were similar to intensities of the same compounds in the control urine samples.  2.4 Conclusions  Untargeted metabolomics using CE-MS analysis and subsequent principal components analysis (PCA) revealed a clear difference between urine from healthy volunteers and urine obtained from patients with PCa or BCa. Nine m/z values were singled out as having a higher or lower intensity signal in patient urine than in urine from healthy volunteers. Accurate masses for these compounds could be obtained with high-resolution MS.  51  CE-MS was also capable of measuring the levels of proline, cysteine, glutamic acid, leucine, kynurenine and sarcosine in the urine samples with acceptable linearity for all amino acids except cysteine. The variance in peak width of cysteine in the targeted analyses correlated positively with creatinine concentration, suggesting creatinine and like compounds could be responsible for the unreliability of cysteine concentration as reported in this study (see Appendix B2).  The number of patients in this study (n = 4) was too small to make any population-level correlations between metabolite levels and progression of cancer. Usually, a statistical test, such as Student’s t-test, requires that the underlying distribution be known, not unexpectedly skewed and the variances uniform across samples. Although the t-test can be performed on small sample sizes, the smallest accepted sample size for the t-test is 30. Rank based methods do not give reliable results either.87 Therefore, no statistical tests were performed in this study, and no p-values were generated. We have demonstrated that urine sample collection can occur in the surgical workflow antecedent to the biopsy of cancerous prostate and bladder tissue in order to perform targeted and untargeted metabolomics studies using CE-ESI-MS. Future studies will aim to strengthen the connection between urine metabolomic data and pathological information of prostate and associated cancers. CE-ESI-MS is a powerful analytical technique that has demonstrated success with a complex aqueous sample the scientific community is very familiar with: human urine. With this success, we move on to develop a CE-ESI-MS method to characterize a complex aqueous sample the scientific community is less familiar with: NAFC in OSPW. 52   Potential of capillary electrophoresis mass spectrometry for the characterization and monitoring of amine-derivatized naphthenic acids from oil-sands process affected water  3.1 Introduction NAs exist as negatively charged ions in alkaline medium; however, the electrophoretic separation and negative-mode ESI-MS analysis of these molecules in basic aqueous medium has been challenging with the instrument we have. Although the steering committee of scientists referred to in Table 1.1 recommended to not utilize derivatization, derivatization was considered an important approach for future naphthenic acids research. In the present study, three different types of chemical derivatization that convert the NAs into naphthenic amines were investigated to impart greater selectivity of acidic compounds from the sample matrix. The naphthenic amines, which existed as cations in acidic solution, were characterized by CE separation and ESI-MS analysis under positive ion mode. To our knowledge, this work is the first known application of CE-MS to the analysis of NAs.  3.1.1 EDC/NHS-mediated chemical derivatization of naphthenic acids Chemical derivatization is a common technique for increasing the selectivity, separation and sensitivity of detection of target analytes. The coupling of a carboxylic acid to a primary amine can be facilitated by the presence of 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and N-hydroxysuccinimide (NHS). Both EDC and NHS are widely used in bioconjugate 53  chemistry and peptide chemistry. The general outline of such derivatizations is shown below. (Figure 3.1)  Figure 3.1 General reaction equation for the EDC (and optional NHS) mediated amidation of carboxylic acids (e.g., NAs) with a primary-tertiary diamine to form (naphthenic) amines. Figure 3.2 provides additional information on the reaction mechanism.  EDC and NHS activate carboxylic acids via different mechanisms. In the case of EDC, the carboxylic oxygen atoms donate electrons to the carbodiimide carbon on EDC, forming a C-O bond to make a transient O-acylisourea (Figure 3.2 A).88,89 Subsequent aminolysis of the O-acylisourea affords the amine-derivatized acid (Figure 3.2 B). In aqueous media, EDC degrades to the urea by-product. This degradation of EDC occurs rapidly in acidic, aqueous solutions.90,91     54   Figure 3.2 (A) The electrophilic carbodiimide carbon in EDC accepts electrons from R-carboxylic acid to (B) produce the O-acylisourea derivative of EDC. Subsequent aminolysis of the O-acylisourea releases (C) the EDC urea by-product (far right, lower structure) to produce the diamine-derivatized carboxylic acid (far right, upper structure).  By contrast, the mechanism of N-hydroxysuccinimide (NHS) activation involves a substitution to form a N-succinimidyl ester, with subsequent aminolysis of the ester to afford the amine-derivatized product. The aminolysis of the N-succinimidyl ester is first order with respect to [OH-].92 In concert, EDC and NHS, buffered at approximately pH 7, enhance each other’s ability to mediate the amidation of carboxylic acids.93,94  Previous strategies for characterizing NAs have leveraged the N-acylurea by-product of EDC derivatizations, sometimes referred to as “N-acylisourea”, suggestive of an imidic acid tautomer 95, that forms when there is a large excess of EDC in aqueous protein chemistry, given by IR data of C=O and N-H stretches 90, an excess of tertiary amine reacting aromatic carbodiimides with monocarboxylic acids, 96 in the absence of faster reactions in pH 7 medium, by pH monitoring, 97 or in high dielectric organic solvents.98 The formation of N-acylurea is relatively slow and is thought to proceed via intramolecular acyl transfer from the ureic O to 55  ureic N (Figure 3.3), or in other words an isomeric O-acylisourea→N-acylurea rearrangement, and has been substantiated with FTIR and NMR data.99–101  Figure 3.3 Isomerization of O-acylisourea EDC derivatives of NAs to N-acylurea derivatives based on naphthenic-acyl transfer.   Once the N-acylurea derivative is formed, it is recalcitrant to aminolysis and has shown to be generally stable at temperatures up to 140 °C.102 Woudneh and coworkers reported that N-acylurea EDC derivatives of NAs give rise to a single fragment (m/z 129) in LC-(+ESI)-MS/MS that can be monitored for quantification.35 In contrast, the present study focuses on the full scan MS data, produced with both CE-ESI-TOFMS (optimized for high molecular weight) and CE-ESI-QMS, of NAs derivatized to form amines via EDC or EDC/NHS-mediated amidation.  3.2 Experimental 3.2.1 Materials and Chemicals NAs standard mixture, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), N-hydroxysuccinimide (NHS), N,N-dimethylethylenediamine (N,N-DMEDA), N,N-56  diethylethylenediamine (N,N-DEEDA) and NAs standard mixture were all purchased from Sigma-Aldrich, St. Louis, MO. Both EDC and NHS were stored at -20 °C in tightly sealed containers. DCM, ethyl acetate (EtOAc) and dimethylsulfoxide (DMSO) were purchased from Fisher Scientific. Oil sands process water (OSPW) extracts were provided by Environment Canada. The electrophoresis BGE consisted of methanol (LC-MS reagent grade, J.T. Baker, Avantor, USA), formic acid (Anachemia, Canada), and deionized water.   3.2.2 Derivatization procedure  This study implemented three types of derivatizations for the analysis of naphthenic acid compounds using capillary electrophoresis-electrospray ionization-(low resolution) mass spectrometry (CE-ESI-LRMS): Isolated NAs were derivatized with a primary-tertiary diamine using (1) EDC and NHS in DMSO, (2) EDC in DMSO, or (3) EDC in dichloromethane. EDC/Sulfo-NHS-mediated amidations in DMSO were also explored, but the analytical advantages compared to more economic alternatives were not significant enough to warrant its use at this time. Since the efficacy of derivatization and associated side-reactions appear to be highly dependent on the lability of protons in the reaction solution, aprotic solvents were used in this work to suppress side reactions.   For the NAs standard solution, 1 mL of 100 ppm solution of Aldrich Standard NAs in 0.1 N NaOH(aq) was acidified by an equal volume of 1.0 M HCl. NAs were extracted from the acidified solution using three extractions of 0.5 mL EtOAc. EtOAc was chosen over DCM or hexanes because of its low cost and toxicity. For the environmental tailings water samples, 50 μL of large-volume OSPW-extracted NAFCs in 0.1 N NaOH(aq) were acidified with 50 μL 1.0 M HCl and extracted into 50 μL ethyl acetate three times. Following the extraction, the combined 57  ethyl acetate layers were evaporated to dryness by vacuum centrifugation (Vacufuge, Eppendorf, Wesseling-Berzdorf, Germany).   The extracted NAs were derivatized by adding 5 μL of 10 mM of a reagent solution containing EDC and 10 mM NHS (or only 10 mM EDC) in DMSO or dichloromethane (DCM), followed by 5 μL of a second reagent solution containing 0.2 M N,N-DMEDA in DMSO or DCM. After 20 min, the reaction mixture was diluted with 40 μL BGE and injected directly into the CE-ESI-MS system.  3.2.3 CE-ESI-MS system  CE-ESI-MS of amine-derivatized NAs was carried out on P/ACE MDQ Glycoprotein System (Beckman-Coulter, Inc.) coupled to Micromass QTOF (Waters-Micromass) with mass resolution of approximately 500 (i.e., “low resolution”). The CE and MS instruments were coupled using the low-dilution “flow-through microvial” electrospray interface described in Section 1.3. Data acquisition and visualization, as well as electrospray ionization and mass spectrometer parameters were controlled via MassLynx Software V4.0 (Waters-Micromass). The electrospray voltage was set at +4500 V and the MS inlet temperature was 150 °C. The modifier solution was delivered through a separate capillary with inner diameter of 75 µm located at the CE instrument inlet. The modifier solution was identical in composition to the BGE and was kept at 0.2 psi (approximately 100 nL/min) to stabilize the electrospray. The mass spectrometer was operated in positive, full scan mode scanning between 50 and 2000 m/z at a spectrum acquisition rate of 1 Hz.   58  3.2.4 Capillary electrophoresis All CE separations were carried out using a 75 cm long fused-silica capillary (50 µm I.D., 365 µm O.D., Polymicro Technologies, Pheonix, AZ). Prior to use, the interior walls of the capillary were treated with trimethoxysilylpropyl modified polyethyleneimine (Gelest, Morrisville, PA) to form a positively-charged covalently bound surface layer.  CE separations were controlled using 32 Karat software, version 7.0 (Beckman-Coulter, Inc., Brea, CA). Samples were injected by applying a positive pressure (0.5 psi for 5 s) to the sample inlet vial, corresponding to an estimated sample injection volume of 4.4 nL. Separations were carried out in reverse-polarity mode at -30 kV. Under these conditions, the bulk flow (EOF) in the covalently treated capillary was towards the capillary outlet, whereas the electrophoretic migration of the positively charged derivatized NAs opposed the direction of the bulk flow.  The composition of the BGE was optimized using a Simplex protocol based on the empirical evaluation of 5 factors (see Section 3.3.2 below). Optimization experiments were carried out by varying the proportions of methanol and formic acid while maintaining a constant capillary temperature (15 °C). The optimum BGE composition was determined to be 30% methanol, 2% formic acid in water; this BGE composition was used for generating the CE-ESI-MS data displayed in the figures below.  3.2.5 Other Software  MSConvert (ProteoWizard) was used to convert MS data files to open source formats.103 TOPP view software (OpenMS) was used for fast visualization and generation of detailed CE-ESI-MS electropherograms.104,105 Data analysis for simplex optimization, finding unique m/z 59  peaks, polynomial regression and plot diagram creation were carried out in R statistical programming language 106 using in-house scripts.107 The usage of in-house scripts is explained further in Appendix C.  3.3 Results and Discussion 3.3.1 Impact of derivatization on separation and mass spectra 3.3.1.1 EDC derivatives of naphthenic acids prior to aminolysis  Prior to the amidation step of the derivatization, CE-ESI-MS data was obtained for the EDC O-acylisoureas of NAs. This data, obtained by scanning the quadrupole mass analyzer, showed three swaths of peaks (Figure 3.4). The groups of peaks had m/z differences of approximately 71 Da and 45 Da. If we allow that rearrangement to N-acylurea has occurred, the most probable fragmentation series is neutral loss of ethylisocyanate (71 Da) to form naphthenic tertiary amine derivatives and subsequent neutral loss of dimethylamine (45 Da) to form the 2-naphthenic-5,6-dihydro-1,3-oxazinium derivatives. In fact, the formation of N-acylurea and loss of isocyanate during electrospray mirrors the reverse of the two-step synthesis of EDC reported by Sheehan et al. 108 When the mixture was left overnight in separation BGE (formic acid, methanol, water), CE-ESI-MS analysis showed no discernable peaks, suggesting the EDC had largely degraded in acidic, aqueous BGE, as expected.   The fragment m/z 129 previously reported by Woudneh 35 as diagnostic of EDC-derivatized NAs in LC-(+ESI)-MS/MS was also clearly present in CE-ESI-MS of derivatized NAs mixtures when scanning the quadrupole.  60   Figure 3.4 CE-ESI-QMS data generated from EDC-derivatized standard NAs mixture prior to aminolysis. The vertical axis has units of m/z and the horizontal axis has units of seconds (time). The colours of the peaks represent peak intensity, shown in the legend on the right-hand side of the mass electropherogram. This mixture showed three swaths of peaks representing one group of parent ions and two groups of fragmentations. Tentative structural assignments for neutral losses are shown to the right.   A probable assignment for this fragment is loss of an EDC urea species from the naphthenic acid with subsequent loss of neutral dimethylamine and cyclization of the remaining fragment. This fragmentation could occur reasonably in both O-acylisourea and N-acylurea species, but not reasonably in the amine-derivatized naphthenic acid species.   3.3.1.2 General characteristics of CE-ESI-MS data of amine-derivatized naphthenic acids For all cases described below, the data generated by CE-ESI-MS of amine-derivatized Aldrich NAs revealed the presence of up to 5 swaths of peaks representing compounds Δ=71  Δ=45  [Naphthenic] [Naphthenic] MT (sec) 50 400 m/z 300 50 10 100 61  (carboxylic acids) between 200 and 800 Da when using the TOF mass analyzer. When scanning with the quadrupole, two swaths of peaks (and fragmentations) representing low-molecular weight NAs between 150 and 450 Da were visible. Each swath of peaks likely represented a distinct class of naphthenic acid fraction compounds (NAFCs). Due to differences in total number of –CH2– groups in NAFC molecules, there was clear mass resolution between groups of molecules with different carbon number—i.e., the difference of 14 Da between classes in the carbon-series was easily observed (see Figure 3.5). The migration time resolution of naphthenic acid fraction compound (NAFC) groups adjacent to each other in the carbon-series ranged from 0 for high molecular weight NAFCs to baseline separation for low molecular weight NAFCs.  Figure 3.5 Sample mass electropherogram for amine-derivatized NAs from NA standard (Aldrich). The zoomed inset on the right demonstrates the typical spacing of NAs data: Swaths of peaks with 14 Da difference represent series of -CH2- and within each carbon number class, there is a Z-series which represents increasing double bond equivalents (double bonds or rings), each producing a hydrogen deficiency of 2. C-series; 14 Da Z-series; 2 Da 62   Within each molecular group of the carbon-series, there were varying levels of C-C bond saturation: For each double bond or ring structure, there was a mass loss of 2 Da due to the loss of two H atoms. The numbers of peaks in this “Z-series” were difficult to discern at low mass resolution and the migration time resolution of adjacent NAFCs in the Z-series was low, but non-zero.   3.3.1.3 Derivatization with diamine using EDC/NHS in dimethyl sulfoxide Despite the commonalities between CE-ESI-MS datasets of amine-derivatized NAFCs in different derivatization conditions, there were several salient differences worth noting. Derivatization of NAs using EDC, NHS, and diamine in dimethyl sulfoxide (DMSO) produced the largest number of m/z and total ion peak features of all methods described herein. The base peak electropherogram (Figure 3.6) shows baseline separation between the derivatized NAs and the derivatization by-products as well as good resolution within naphthenic acid classes.  An in-house algorithm written in R was applied to the CE-ESI-TOFMS data to identify the m/z values of peaks whose intensity exceeds a predefined threshold. This identification produced approximately 350 unique m/z values belonging to NAFCs. This large number of peaks is likely due in part to the presence of up to 13 reactions extraneous to the amidation of NAs that have been described in several reviews.88,89,109 Two of these side reactions are likely to form new analytes with significantly different mobilities: ring opening of N-hydroxysuccinimide by primary amine and reaction of EDC with DMSO. 63   Figure 3.6 A sample base peak electropherogram (BPE) produced by analysis of amine-derivatized NAs (mediated by EDC/NHS in DMSO)  Ring opening of NHS forms a dipeptide-like molecule ([M–OH]+ at m/z 187) that could participate in ester aminolysis, while EDC forms a reactive adduct with DMSO. It is unclear as to whether these reactions impart a greater selectivity to the target analytes in the sample or are entirely redundant in terms of information. High-resolution MS is expected to clarify the issue.   3.3.1.4 Derivatization using EDC under different conditions Derivatization without NHS was performed in order to eliminate the possibility of ring-opening reaction between NHS and the primary amine. The resulting CE-ESI-TOFMS data revealed a smaller total number of peaks in the same swath pattern as with EDC/NHS in DMSO, and at lower intensity. Approximately 250 unique m/z peaks were found using the same identification algorithm. We replaced DMSO with DCM as the reaction solvent in order to Signal Intensity 64  reduce the formation of adducts between DMSO and EDC. CE-ESI-TOFMS of the resulting mixture of N,N-dimethylethylenediamine (N,N-DMEDA) derivatized NAs resulted in a separation profile with less background. Approximately 300 unique m/z values were matched between the two derivatization methods.  Additionally, CE-ESI-MS of the mixture derivatized with N,N-DMEDA in ethyl acetate revealed a virtually identical electropherogram to the reaction performed in dichloromethane with the exception of differing migration time.  Table 3.1 Comparison of salient features of CE-ESI-MS electropherograms of derivatized NAs with derivatization solvent properties. DMSO stands out as high viscosity, high dielectric constant compared to ethyl acetate and dichloromethane and produces different separation characteristics under the same BGE conditions.  The swath of peaks with a higher m/z likely represent the parent ions of amine-derivatized NAs and the lower m/z swath of peaks are a combination of fragmentation products of amine- EDC+diamine in DMSO EDC+diamine in dichloromethane EDC+diamine in ethyl acetate Average NAFC peak width 25 sec 65 sec 65 sec Migration time range of NAFC peaks 100 sec 250 sec 250 sec Derivatization solvent dielectric constant 4.7 0.89 0.6 Viscosity (mPa*s) 2.20 0.45 0.46 EOF marker inverse peak Absent 3.5 min 3.5 min Relative vapour pressure Low High High 65  derivatized NAs and the stable N-acylureas which resisted aminolysis. The neutral loss of dimethylamine (45 Da) from N,N-DMEDA-derivatized NAs was confirmed by the neutral loss of diethylamine (73 Da) from N,N-diethylethylenediamine (N,N-DEEDA) derivatized NAs, as shown in Figure 3.4, and was confirmed using density functional theory calculations (Appendix C.2). Table 3.1 summarizes the principal differences in peak characteristics for the derivatized NAs in terms of derivatization solvent properties. For ethyl acetate and dichloromethane, a prominent inverse peak was consistently seen at 3.5 min and can be used as the EOF marker. For DMSO, the EOF marker peak was less prominent and often absent. DMSO is about twice as viscous as the BGE and has a lower dielectric constant than the BGE, causing an increase in local field strength experienced by the analytes for a longer period of time than with dichloromethane or ethyl acetate, resulting in sharper (less broad) peaks.   3.3.2 Optimization of capillary electrophoresis separation parameters  In addition to evaluating the EDC-based derivatizations of NAs in organic solvents, the effects of BGE composition on various separation parameters were assessed using a simplex-type design of experiment approach called rough global optimization.  3.3.2.1 An Introduction to Rough Global Optimization Multidimensional datasets are encountered in many areas of science. In these datasets, there is an intricate interplay of up to thousands of variables, each with complex covariances and interactions. In this situation, solving for the values of all variables in the system which allow for 66  an optimum condition is a daunting task. Global optimization is the process of computing the values of variables which produce an optimum effect in a system of variables. Global optimization is currently an active area of research and much important computational work has been done. However, in a relatively straightforward analytical system such as capillary electrophoresis-electrospray-mass spectrometry (CE-ESI-MS), a computational method which is geared toward very subtle changes in data may be superfluous. Rough global optimization, essentially a manual method, is a strategic approach to global optimization requiring little computational expertise and operates on approximations of available data. For the system in question—CE-ESI-MS of derivatized NAFCs—there are a small enough number of system parameters to vary and a single outcome, peak resolution, to monitor.  3.3.2.2 Set-up of rough optimization of CE-ESI-MS data of derivatizated naphthenic acids Any optimization method needs parameters to tune according to some criteria and an outcome to monitor for optimization. In the case of CE-ESI-MS of NAs, the parameters that were tuned were the methanol and formic acid concentrations in the running BGE of the CE separation (see Table 3.2). The rationale for choosing these concentrations was practical: the concentration range represented here was known to be stable in the CE-ESI-MS system and concentrations outside this range would not result in a significant increase in separation quality.   67  %v/v Formic acid %v/v Methanol Computer filename 3.5% v/v  Formic acid 0% v/v  Methanol 20140703 30KV15C_-2NHALF_+3_0 EDCNHS BIG A.mzML 2% v/v  Formic acid 0% v/v  Methanol 20140703 30KV15C_-2NHALF_+1_0 EDCNHS BIG A.mzML 20140703 30KV15C_-2NHALF_+1_0 EDCNHS BIG B.mzML 20140703 30KV15C_-2NHALF_+1_0 EDCNHS BIG C.mzML 2% v/v  Formic acid 10% v/v  Methanol 20140703 30KV15C_-2_+1_0 EDCNHS BIG A.mzML 2% v/v  Formic acid 30% v/v  Methanol 20140703 30KV15C_-1_+1_0 EDCNHS BIG A.mzML 20140703 30KV15C_-1_+1_0 EDCNHS BIG B.mzML 2% v/v  Formic acid 70% v/v  Methanol 20140703 30KV15C_+1_+1_0 EDCNHS BIG A.mzML 0.5% v/v Formic acid 30% v/v  Methanol 20140702 30KV15C_-1_-1_0_EDCNHS BIG A.mzML 20140702 30KV15C_-1_-1_0_EDCNHS BIG C.mzML 0.5% v/v Formic acid 50% v/v  Methanol 20140702 30KV15C_0_-1_0_EDCNHS BIG A.mzML 20140702 30KV15C_0_-1_0_EDCNHS BIG B.mzML 20140702 30KV15C_0_-1_0_EDCNHS BIG C.mzML 0.5% v/v Formic acid 70% v/v  Methanol 20140702 30KV15C_+1_-1_0_EDCNHS BIG A.mzML 20140702 30KV15C_+1_-1_0_EDCNHS BIG C.mzML  Table 3.2 Computer filenames used for the global optimization data and the concentrations of formic acid and methanol in the BGE used during each run.  For the eight BGE compositions used, five chemometric factors were monitored and are listed here in the order of importance: peak swath slope, average peak width, electrophoresis current, first peak analysis time, total analysis time (Table 3.3). The 5 factors which were monitored were obtained by observing the CE-ESI-MS data generated on the EDC/NHS-mediated amine-derivatized Aldrich NAFCs in DMSO, as well as observing the CE current data.  68   Global optimization factors for a single analysis “nickname” Definition Units Criteria for analysis x to be a member Average peak resolution “swath” slope of line bisecting swath of peaks (m/z)/min -5 < x < 0 Average peak width “pwidth” Average peak width of a sample of several peaks from a single run min x ≥ minglobal(pwidth) &  x < σglobal(pwidth) Electrophoresis current “current” Stable current value read from CE software µA x < -10 First peak analysis time “migz” Migration time of first peak read from data output min x ≥ minglobal(migz) &  x < [ minglobal(migz) +  σglobal(migz) ] Total analysis time “atime” Total time from beginning of analysis to migration of final peak  sec  (min) x ≤ 900  (x ≤ 15)  Table 3.3 Cut-off values for rough global optimization for each factor. The global minimum value of a factor is denoted as minglobal(factor) and global standard deviation is denoted as σglobal(factor).  The choice of these factors depended on knowledge of the physical chemistry of the CE-ESI-MS system: parameters were chosen which were known to have significant effects on a separation quality. The cut-off criteria were chosen to achieve an acceptable quality trade-off, as is often the case in analytical chemistry: good analyte peak resolution is good for analysis, but a separation under 15 minutes running time is good for a realistic industrial protocol. For each of the 15 runs performed in all BGE conditions ( Table 3.2), 5 values for the global optimization factors were obtained, resulting in 75 such values total. Arranging the values in matrix form, one matrix for each global optimization factor, with column and row indices as the formic acid and methanol concentrations, allowed for the 69  cut-offs in Table 3.3 (far right column) to be applied. The cut-off criterion was only applied to the matrix it directly referred to—i.e. the total analysis time optimum criterion of < 900 seconds was only applied to the total analysis time matrix. Every value in the matrix which obeyed the member criterion was given a +1 point. Every value in the matrix which did not obey the member criterion was given 0 points.   3.3.2.3 Simplex optimization of BGE composition and rough global optimization of responses The optimal composition of the BGE for the separation and detection of amine-derivatized NAs by CE-ESI-MS was assessed by varying the proportions of formic acid and methanol in the BGE ( Table 3.2). The concentration of formic acid in the BGE determines the ionic strength and acidity, which in turn will impact the magnitude of the electroosmotic (bulk) flow in the capillary, the electrophoretic mobilities of the solutes, and the solubility of the amine-derivatized NAs. The concentration of methanol in the BGE also affects the dissociation of the formic acid, and therefore also plays a role in determining the ionic strength and pH of the BGE.  However, it plays a larger role in determining the solubility of the analytes in the BGE, and in determining the surface tension of the solution, which is an important factor in establishing a stable electrospray ionization. Initially, all combinations of formic acid and methanol concentrations were considered (factorial design), but a simplex optimization of the general separation resolution in the form of “peak swath slope” (explained below) was performed instead and experiments were halted when optimum was reached.    70  3.3.2.4 Peak swath slope as a representative parameter for multicomponent resolution Generally, the peak resolution is the most important metric for the success of a separation. It is conventionally defined in relation to two analyte peaks. However, in the case of full scan CE-ESI-MS data with hundreds of peaks, the resolution will be best described as a multicomponent resolution matrix of values that describes the resolution of each two-peak combination. Here, we define the “peak swath slope” as an approximated subset of the more informative multicomponent resolution matrix suitable for complex mixtures. The derivatized NAs reliably produce a swath, or swaths, of peaks that describes the arrangement of analyte peaks in the space of m/z and migration time and follows the carbon series of NAs where peak groupings are separated by m/z 14 (-CH2-). The swath of peaks extends from high m/z to low m/z as migration time increases (i.e. negative slope, downward to the right) and can be seen in the electropherograms shown in Figure 3.4, Figure 3.5 and Figure 3.10. Excellent peak-to-peak resolution along the carbon series, then, would be approximated by a negatively sloped peak swath approaching a value of zero. Therefore, the slope of the swath of the peaks was the main CE-ESI-MS data characteristic monitored for the simplex optimization. It should be noted that, in contrast to the conventional definition of resolution, peak swath slope does not take into account the peak width. Peak swath slope may be a useful proxy for chromatographic resolution in the case of very complex mixtures which are difficult to separate.  3.3.2.5 Rough Global Optimization  A rough global optimization was manually performed on the separation data produced from the CE-ESI-QMS runs of derivatized NAs to determine a range of BGE compositions that could produce the desired overall separation effect. The five separation factors and their response 71  characteristics for optimization, listed in order of importance, are maximum peak resolution (i.e. negative swath slope), sharper peaks, medium-high separation current (< -10 μA), fast migration of first peak (or fast EOF) and a total analysis time of 15 minutes or less (Figure 3.7).  Figure 3.7 (Left) Optimal criteria for each of the 5 factors. (Right) Size of circle graphics used in Figure 3.8 to represent optimum response of each of the 5 factors. The size of the circle correlates with the weight put on each factor.   After having assigned membership points to all 5 factor matrices, the matrices were superimposed and the BGE compositions which gave rise to the highest number of points suggests the formic acid/methanol composition (% v/v) which produces the most optimum responses in the 5 factors (Figure 3.8).  72    Figure 3.8 Diagram describing global optima for 5 CE separation responses. A circle means an optimum was reached for a factor response. The diameter of the circle indicates the relative importance (weight) of the response. The most circles at a particular BGE composition indicate a point of roughly global optimality. The red triangle outlines a region of global optimality. The black arrow indicates the direction of the system from initial BGE composition toward global optimality upon natural methanol evaporation (starting at 30% and evaporating to 0%).   Of course, no condition fulfilled the optimum for every factor, but a choice could still be made because the 5 factors were weighted by importance. If the data is plotted and the weighting information is incorporated into the plot in the form of data-point character radius, then the plot directs attention to those points which have the largest radii in addition to the largest number of points. Given that the optimum area included a range of methanol concentrations, we chose the highest % methanol within that area to ensure solubility of the analytes and to stabilize the electrospray. In this case, methanol evaporation, which is a significant effect for this BGE, would 73  cause a decrease in % methanol and a subsequent increase in % formic acid (denoted by the curved black arrow in Figure 3.8) and would propel the system deeper into the global optimum region (i.e., the region within the red triangle in Figure 3.8). The BGE composition of 30% methanol and 2% formic acid in water was chosen as the optimum for future separations.  Finally, polynomial regression was performed on each data matrix, assuming a simple polynomial expression as a function of methanol and formic acid concentrations. Polynomial regression was performed in R using the stats package.106 Table 3.4 shows some relevant polynomial equations fitted to the CE-ESI-MS data. Relevant system equations obtained by polynomial regression  x = % v/v Methanol; y = % v/v Formic acid  migration time(𝐱, 𝐲) = 𝟒𝟑. 𝟔𝟗 + 𝟖. 𝟔𝟕𝟔𝐱 + 𝟗𝟐. 𝟐𝟖𝐲 − 𝟎. 𝟔𝟒𝟗𝐱𝐲 − 𝟎. 𝟎𝟓𝟗𝟖𝐱𝟐   − 𝟏𝟒. 𝟕𝟖𝟓𝐲𝟐 peak width(𝐱, 𝐲) = 𝟑𝟐𝟐. 𝟑𝟓 − 𝟎. 𝟎𝟕𝟖𝟎𝐱 − 𝟐𝟐𝟐. 𝟗𝟔𝐲 − 𝟎. 𝟏𝟓𝟐𝐱𝐲 + 𝟎. 𝟎𝟎𝟔 𝐱𝟐 + 𝟑𝟖. 𝟔𝟐𝐲𝟐 average peak resolution(𝐱, 𝐲)= −𝟏𝟎. 𝟎𝟏𝟔 + 𝟎. 𝟕𝟑𝟐𝐱 − 𝟏𝟎. 𝟎𝟒𝐲 + 𝟎. 𝟎𝟑𝟏𝐱𝐲 − 𝟎. 𝟎𝟎𝟖𝐱𝟐 + 𝟑. 𝟑𝟔𝐲𝟐 analysis time(𝐱, 𝐲) = 𝟏𝟐𝟏𝟒. 𝟒 − 𝟒. 𝟓𝟒𝟖𝐱 − 𝟔𝟔𝟐. 𝟐𝟖𝐲 + 𝟕. 𝟕𝟑𝟖𝐱𝐲 − 𝟎. 𝟎𝟏𝟖𝐱𝟐 + 𝟏𝟑𝟕. 𝟔𝐲𝟐  Table 3.4 Summary of the relevant system equations for the effects of BGE composition on migration time of first peak, peak width, average peak resolution and total analysis time.  3-D plots of the four most important system equations obtained by polynomial regression are illustrated in Figure 3.9.  74              Figure 3.9 Response surface plots for 4 factors. The response surfaces were approximated by polynomial regression on the rough global optimization data to approximate the effects of changing % formic acid and % methanol on the optimum factors described in Table 3.3.  Although there are no relevant underlying system physics represented by these polynomial regression equations, the approximations hold loosely within the boundaries of the BGE 75  concentrations and are illustrative of the effects of changing % formic acid and % methanol on the quality of the CE-ESI-MS data. Here, the size of the coefficient correlates to the effects of methanol (x, x2), formic acid (y, y2), and their coupled influence (xy) on the five factors explored.  The data for the 5 responses (z) were fit to polynomial models with two BGE variables, % methanol (x) and % formic acid (y). For all responses, the effects of methanol were between 1 and 4 orders of magnitude weaker than the effects of formic acid. An increase in formic acid led to decrease in peak width, a decrease in average peak resolution in the time domain (i.e., decrease in “peak swath slope” which is used as a proxy for “chromatographic resolution”), and decreased analysis time (due to much faster electroosmotic flow). An increase in the percentage of methanol in the BGE increased total analysis time. It is unclear as to whether or not there is better separation resolution of isomeric naphthenic acid species at higher methanol percentage. The effects of formic acid and methanol on the CE-MS data are qualitatively summarized in  Table 3.5.   Effect of methanol % (x) Effect of Formic acid % (y) Migration time of first peak Weak positive correlation; weakly coupled to formic acid Positive correlation Peak width N/A Strong negative correlation Average peak resolution Very weak negative correlation Strong negative correlation Total analysis time Weak positive correlation; weakly coupled to formic acid Strong negative correlation  Table 3.5 Effects of varying % volume of methanol and formic acid in the BGE for separations of amine-derivatized NAs (EDC/NHS+DMSO)  76  3.3.3 CE-ESI-MS of naphthenic acid fraction compounds derived from oil sands process waters  The optimized separation conditions were applied to a sample of naphthenic acid fraction compounds (NAFCs) extracted from oil sands process waters (OSPW) which were derivatized with N,N-DMEDA, mediated by EDC/NHS in DMSO. The mass-electropherogram revealed a single large swath of peaks between 400 and 500 s migration time and extending from m/z 1500 to m/z 350 (Figure 3.10), as well as two smaller swaths of peaks, one swath with a significantly different migration time, but much less intense.   Figure 3.10 CE-ESI-TOFMS of derivatized NAFCs derived from oil sands process waters (OSPW). The data shown here are mean-centred and normalized to standard deviation by m/z channel. The colour bar illustrates relative peak intensity (%).  77  3.4 Conclusions  Amine-derivatized standard NA mixture shows considerable tolerance for changes in the composition of the BGE for CE separation. When optimizing the concentrations of formic acid and methanol, formic acid concentration has the largest impact on the electroosmotic flow, separation time, peak shape, and global resolution. It was determined that 30% methanol, 2% formic acid in water was an optimum BGE composition for the separation and analysis of amine-derivatized NAs, based on a rough global optimization of 5 separation criteria.  Different EDC/NHS derivatization schemes result in very different two-dimensional spectra, due to the presence or absence of reaction side-products. Reactions without NHS and DMSO as solvent resulted in spectra with a lower degree of complexity. Further studies are needed to determine if spectrum complexity is an advantage or disadvantage in the effort to characterize persistent water-soluble oil sands contaminants such as NAs. Our data confirms the presence of m/z 129 ion in CE-ESI-QMS coeluting with EDC-derivatized NAs 35 and absence of m/z 129 for amidated NAs. We provide further evidence for the ring-opening reaction of NHS with a primary amine.109 We also have evidence to support the formation of N-acylurea by-products in the low-dielectric organic solvents dichloromethane and ethyl acetate, supplementing a previous report.98 Preliminary results for the EDC/NHS+diamine derivatization in DMSO on naphthenic acid fraction compounds (NAFCs) derived from OSPW reveal the presence of a swath of compounds between m/z 350 and 1500. Further optimization of the reaction kinetics of derivatization and application to a wider range of environmental NAs samples are needed in order to develop this method as a “turn-key” operation for the routine analysis of NAFCs. Application of high-resolution mass spectrometry (HRMS) to the capillary electrophoresis 78  separations should allow for accurate chemical formula assignment in order to identify individual components present in NAFCs samples.   79   Characterization of Athabasca Lean Oil Sands and Mixed Surficial Materials: Comparison of Capillary Electrophoresis-Low Resolution Mass Spectrometry and High-Resolution Mass Spectrometry   4.1 Introduction The Athabasca oil sands deposits in Alberta, Canada are one of the world’s largest reserves of petroleum.110 The lean oil sands (LOS) overburden, representing low-grade ore (< 7% hydrocarbon by weight) that is removed and stockpiled prior to surface mining, results in a significant volume of material that is placed in large, above-ground deposits. These materials contain low grade bitumen and thus petroleum hydrocarbons. Water that comes in contact with stockpiled LOS has the potential to transport polar organic compounds throughout the ecosystem. The levels and distribution of potentially water soluble organic compounds, such as NAs, have not yet been established.111 Likewise, there has not yet been an exhaustive characterization of the physical-chemical properties that govern seepage within and leaching of dissolved constituents from these materials.    The preliminary studies reported herein are intended to provide a demonstration of the complementarity of CE-LRMS and HRMS for assessment of chemical profiling of LOS and related materials. The capillary electrophoresis-MS migration times, analyte information not provided by HRMS, and infusion high-resolution MS speciation of NAFCs according to compound class and double bond equivalence are all demonstrated to have high potential for assessment of the fate and transformation of oil sands derived components in the materials investigated. A two-level approach to analyzing petroleum compounds in water is recommended: 80  the analysis of total petroleum hydrocarbons (PHC) acts as Level 1 screening, and the speciation of NAFCs represents a Level 2 screening. Preliminary evidence is also presented for use of OxSy heteroatomic species at the molecular level as tracers of possible seepage from the LOS materials.   4.2 Experimental Methods Porewater samples were collected at the mine site and shipped to the laboratories at WSTD, Saskatoon, and Exova, Calgary. Full sample description and the physical-chemical properties of the solid soil/LOS material from which the porewater samples were taken have been summarized in references Korbas112 and Scale et al.113 The material consists of silty sand (average 10% clay-size fraction, 35% silt and 55% sand).  4.2.1 Gas chromatography flame ionization detection for total petroleum hydrocarbons  A standard GC-FID method (CCME) was used to characterize porewater from Athabasca lean oil sands (LOS) and mixed surficial materials.114 50 mL porewater samples were extracted in the sample container with hexane while placed on a rotary tumbler for 1h. The hexane fraction was selectively removed, and an aliquot analyzed. The bulk petroleum hydrocarbon (PHC) content of the LOS water samples (in units of mg bulk petroleum per kg LOS) was determined by an Agilent 6890 Series gas chromatography (Agilent Technologies, Santa Clara, CA, USA) coupled to HP 5890 Series II plus flame ionization detector (Hewlett-Packard, Palo Alto, CA, USA).   81  4.2.2 Orbitrap mass spectrometry characterization of LOS and surficial materials.   Total NAFCs were measured by negative-ion ESI MS as described previously by Headley et al.1,6 The calibration is based on actual Athabasca oil sands acid extract generated in-house as opposed to using commercial standards such as Fluka or Acros NA mixtures. The measurements are semi-quantitative and are best used to assess the relative differences in NAFCs obtained under the same experimental conditions. A Thermo-Scientific LTQ Orbitrap Elite mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) was used for determination of NAFC concentrations. The mass resolution of the Orbitrap Elite mass spectrometer was set to 240,000 at m/z 250 and full-scan mass spectra were acquired over an m/z range of 100-600. The electrospray ionization (ESI) interface was set to negative ionization mode. Ionization and full scan MS conditions were optimized by use of the automated tune function for transmission of ions of m/z 212.07507. The ion at m/z 212.07507 is derived from n-butyl benzenesulfonamide which is a well-known background ion present in the negative-ion mass spectra.116 Parameters for the heated ESI interface were: source heater temperature, 53 °C; spray voltage, 2.86 kV; capillary temperature, 275 °C; sheath gas flow rate, 25 L/h; aux gas flow rate, 5 L/h. All formula assignments were < 2 ppm mass error.  4.2.3 Capillary electrophoresis-mass spectrometry characterization of LOS and surficial materials Porewater samples (50 mL aliquots) were acidified with 40 mL 1 M HCl and extracted into 25 mL ethyl acetate twice. Residue acids were derivatized to form amines according to the procedure described in Chapter 3: which uses a carbodiimide intermediate and a diamine to convert NAs to naphthenic amines. The derivatized extracts were diluted to 50 μL with an 82  aqueous CE electrolyte solution (2% formic acid in 30% methanol). Sample volume of 4 nL was injected onto a polyethyleneimine (PEI) coated capillary and electrophoresis was carried out at 30 kV for 15 minutes using a Beckman Coulter P/ACE MDQ system (Beckman Coulter, Inc., Brea, CA, USA). The flow-through microvial ESI interface56 was set to positive ionization mode, and the TOF-MS (Waters Corporation, Milford, MA, USA) inlet voltage was set to transmit ions m/z 100 to approximately m/z 1500.  4.2.4 Data processing and factor/components analysis All data processing was carried out using in-house scripts written in R Statistical programming language (version 3.1.2).106 Raw HRMS and CE-MS data were converted to open source formats using mzConvert117 and processed into list objects in R using the mzR package.103,117–120 The lists were reformatted into matrices using custom code. Principal components analysis (PCA) and factor analysis (FA) algorithms were provided by the psych package in R.121 An in-house script was used to loop through rotational and matrix factoring options to select feasible combinations of component analysis criteria (See Appendix D). Guassian blur for contour overlay in Figure 4.5 was provided by the spatstat122, lattice123 and plot3D124 packages.  A linear cut-off was applied to the sample data peak intensity, values were rounded to three decimal places and mean centred sample-wise. CE-MS data also had linear cut-offs applied to migration time and m/z and the additional application of a rudimentary peak picking algorithm. The number of factors/components used in each case were chosen by visual analysis of a Scree plot. For both FA and PCA, the minimum obliqueness rotational criterion, which sacrifices some orthogonality of components/factors for better clustering, was applied.125,126 Factors/components obtained with minimum obliqueness rotational criteria are usually correlated (i.e. not orthogonal) 83  which can make interpretation challenging. For FA, the principal axis (PA) factoring method was employed because of its similarity to PCA factoring algorithms.127   For high-resolution MS data, factor analysis was used. Factor analysis algorithms calculate using the data in the sample dataset contributing to common variance (i.e. outliers removed) and use a matrix with approximated “communalities” for factorization, allowing for the results to be generalized into a model. By contrast, PCA uses all data (i.e. all variance) in the dataset and factorizes a matrix based on the calculated covariances. PCA was also used for HRMS data, and produced nearly identical components to FA, but the results could not be generalized into a model. For higher dimensional CE-MS data, PCA is used to explore the structure of the dataset and reduce dimensionality.  4.2.5 Data visualization  Figure 4.2 and Figure 4.3 were generated using Microsoft Excel. Figure 4.4 and Figure 4.5 were generated in R. Figure 4.6 image of CE-MS TIC was generated using OpenMS TOPPView.104,105 Figure 4.7 was also generated in R. It is important to note that there are many ways to visualize the results of PCA or FA and the visualizations in Figure 4.7 are one among many options. For models with dimensionality greater than 3, such as the models described below, it can be difficult to effectively visualize clustering. No experimental data were available for samples not included in Figure 4.7.  84  4.3 Results and Discussion  The CCME GC-FID method groups the PHCs into 4 fractions (Figure 4.1). These fractions are related to PHC types that one may find in PHC contaminated soil samples. The fractions are determined by their equivalent carbon number (ECN) range, based on GC retention times obtained from a standard reference mixture of hydrocarbons: F1 corresponds more or less to gasoline-type PHC, F2 and F3 correspond to middle distillates such as diesel and motor oil, and F4 contains very heavy PHC such as asphalt.   Figure 4.1 Example of GC-FID of LOS and surficial materials analyzed according to a standard CCME method. The vertical axis is “picoamperes” of current and horizontal axis is “Time” in minutes. The fractions F2, F3, F4 are also delineated with equivalent carbon number (ECN) range given for each fraction (C10, C16, C34, C50).  For the LOS samples, F2 hydrocarbons (ECN: C10-C16) ranged between 1700-3000 mg/kg and F3 hydrocarbons (ECN: C16-C34) had a range between 9000-16000 mg/kg. For the heaviest 85  hydrocarbons, group F4 (ECN: C34-C50+) PHC was measured as 9000-53000 mg/kg using high-temperature gas chromatography. Gravimetric heavy hydrocarbon analysis confirmed the total heavy hydrocarbons to be between 23000 and 53000 mg/kg.114  Sample Description Sample Name NAFCs Conc. (mg/L) Total Petroleum Hydrocarbons (mg/L) Groundwater from the surficial aquifer near the interface of the original ground and the present overlying dumped LOS/overburden pile 1A-1 16.8 < 0.1 1A-2 16.2 1A-3 15.5 SP2A-1 7.0 1.9 SP2A-2 5.3 SP2A-3 7.0 Shallow perched water* within dumped LOS/overburden pile 1B - Organics-1 65.5 15.1 1B - Organics-2 72.5 1B - Organics-3 75.1 2B-1 82.3 3.3 2B-2 81.3 Subsequent to infiltration through soil cover, shallow “interflow” collected from the interface of the soil cover and underlying dumped LOS/overburden material    TP4-1 38.7 0.9 TP4-2 37.5 TP14 26.1 n/a TP18-1 26.7 < 0.2 TP18-2 26.5 TP30-1 41.2 0.2 TP30-2 36.8 TP30-3 44.3 Seepage from a spring located on the side-slope of the dumped LOS/overburden pile. SEEP-1 40.8 2.7 SEEP-2 49.0 SEEP-3 48.1 *Perched water is a type of groundwater that is separated from the main body of groundwater by rock or some other material. In this case the groundwater is “perched” inside the LOS pile.  Table 4.1 Measured values of NAFC (via Orbitrap MS) and TPH (via GC-FID) in samples of LOS and surficial materials. 86   As illustrated in Table 4.1, all samples contained a preponderance of petroleum hydrocarbons (<0.1-15.1 mg/L) with a continuum of components across the duration of the GC run, corresponding to the equivalent carbon number range of C7-C30 (Figure 4.1). Here, GC-FID for total petroleum hydrocarbons is an example of Level 1 screening.115  Table 4.1 also shows the levels of acid extractable organics or NAFCs measured for the LOS and surficial materials via infusion Orbitrap MS and were in the range 5.3-82.3 mg/L. The expected differences in values between mg/L “petroleum hydrocarbons” (PHC), measured by GC-FID, and mg/L NAFCs, measured by Orbitrap MS, are due to the differences in definition between PHC and NAFCs, and that the two techniques use different sample preparation and analytical detection. Although PHC and NAFC are similarly defined as complex mixtures being derived from petroleum sources, and containing hydrocarbons with heteroatoms, NAFC is an acid fraction injected into MS as aqueous solution whereas PHC is a hexane fraction injected onto GC, which mobilizes molecules based on boiling point. The GC-FID method for PHC would therefore be selective for non-polar PHC and it is likely that most of the acidic compounds present in the naphthenic acids fraction (NAFCs) would not be seen in GC-FID. Conversely, non-polar species would not be observed in the Orbitrap MS of NAFC. The total petroleum hydrocarbon values (Level 1 screening) in a sample may not necessarily reflect the levels of oil sands derived organics. For this reason, an approach based on both total PHCs (Level 1) and speciation of NAFCs (Level 2) is recommended. To illustrate how speciation of porewater NAFCs provides important complementary information, the ratio of compound classes and the corresponding distribution of double bond equivalence for a given class can be used as a “fingerprint” for the fate and transformation of a given sample as evident in Figure 4.2, Figure 4.3, and Figure 4.4 below.  87   Figure 4.2 Orbitrap MS characterization of the distribution (relative abundance) of Ox (x=1-5), OxSy (x=1-4) classes of LOS and surficial materials across all samples.   For the Orbitrap MS data, no significant difference was generally apparent within triplicate field samples. The Orbitrap MS characterization revealed the distribution of Ox (x=1-5), OxSy (x=1-4) classes (Figure 4.2), of which the dataset was almost entirely comprised. The O2 species are presumably classical NAs. Figure 4.3 shows representative double-bond equivalent (DBE) values for O2 species of LOS and surficial materials. Collectively, the Orbitrap MS results revealed a relative difference in the DBE plot of the O2 class for the LOS and surficial materials. The nature of the differences between samples was further explored using factor analysis. 88   Figure 4.3 Relative abundance of negative-ion Orbitrap MS double bond equivalents (DBE) versus carbon number for the O2 species across all samples.   Factor analysis (FA), using the principal axis factoring method, was performed on the sample full scan MS data, which incorporates all Ox species, OxSy species and DBE information. The data did not include separate columns related to DBE. Factor analysis differs from principal component analysis (PCA) mainly in that FA only considers variance shared across the sample set (i.e. covariance), putting the communalities in the matrix diagonal. Consideration of variance common to all samples allows the results to be generalized into models. PCA, on the other hand, considers all variance across samples and is not generalizable beyond the sample set (no model assumed).128   89   The factor analysis (FA) loadings on the raw Orbitrap MS data were plotted (Figure 4.4) to reveal clusters of points, clearly grouped according to similar sample types. For example, the groundwater triplicate samples named “1A”, represented by the blue circle symbols in Figure 4.4, are clustered closely together, and perched water samples “1B - Organics” (red squares) are clustered closely together but apart from the “1A” samples, and similarly with the other samples in the set (see Table 4.1). Samples that occupy similar space in the factor analysis plots have similar MS signal “fingerprints” and are interpreted to be of similar origin or share similar fate and behavior in the environment. The fact that replicate samples group together was expected and was indeed observed. This demonstrated that Ox and OxSy compound classes were diagnostic of the water sources from which the samples were collected.  90   Figure 4.4 3-D plot of 3-factor factor analysis (FA) using principal axis (PA) factoring of infusion negative-ion Orbitrap MS of all ions observed in the raw data for LOS and surficial materials.   Seepage water in contact with LOS has the potential to transport NAFCs. Therefore, an important water sample metric could be “extent of exposure to LOS”, which could be measured by total NAFCs concentrations and by comparing sample chemical “fingerprints” in FA. The factor loadings plot in Figure 4.4 contains information about the samples based on the HRMS 91  peak intensities at each m/z value. The quantitative Orbitrap analysis confirmed the different NAFC concentrations (mg/L) in the samples (Table 4.1). The NAFC concentrations were then overlaid as a contour map onto the FA analysis in order to determine if any patterns were present (Figure 4.5). Samples with higher exposure to NAFCs (mg/L) on the contour map appear to cluster together on the FA principal axis loadings space and lower concentration samples on the periphery of loadings space. This confirmed the assumptions that porewater collected from or exiting the shallow saturated zone within the LOS/overburden pile (1B, 2B & seeps) had higher exposure to NAFCs and similar chemical “fingerprints”, shallow aquifer samples had the lowest exposure and test pit groundwater samples had varying levels of exposure to NAFCs reflecting the variable extent of LOS contact of this interflow water and a range of chemical “fingerprints”. However, shallow aquifer sample sets 1A and SP2A did not cluster well on FA principal axis 1, suggesting similar total NAFCs concentrations, but different NAFC “fingerprints”. A similar pattern is seen with contour overlay of total PHC measured by GC-FID.   92   Figure 4.5 Loadings plot of FA principal axes 1 and 2 of Orbitrap MS data with a Gaussian-blurred contour overlay approximating the total NAFCs (mg/L) concentration of each sample. Higher concentration samples cluster together and have similar chemical “fingerprints” in principal axis loadings space, whereas lower concentrated samples have a range of chemical “fingerprints”.  CE-MS of the extracted porewater samples consistently produces mass electropherograms with predictable series of peaks. First to appear are distinct peak shapes at the electroosmotic flow marker across all samples which should correspond to neutral species, possibly neutral hydrocarbon, polymeric species (Figure 4.6, zone A). Next, a variety of analytes are detected (Figure 4.6, zone B), many baseline-separated compounds within ±0.05 m/z. Immediately following, a large swath of peaks appears extending from m/z 500 to m/z 100 and 5 to 10 minutes wide representing amine-derivatized NAs (Figure 4.6, zone C). This swath is followed by the derivatization by-products appearing in two large peaks (Figure 4.6, zones D & E). CE-MS has the potential to provide simultaneous information on polar and non-polar petroleum 93  hydrocarbons, dividing the sample into distinct classes of compounds. Infusion Orbitrap MS cannot distinguish different “zones” of peaks, as there is no online separation prior to MS, and there is a possibility of interference of non-NAFC compounds that could affect measured total NAFCs concentrations.   Figure 4.6 Two-panel plot of typical CE-MS data of NAFCs extract sample of LOS-affected water: Total ion electropherogram (top, blue) and CE-MS ion map (bottom, yellow). Peak intensity cut-off was applied. The mass-electropherogram can be divided into 5 general zones that appear in every sample: A=dip in background signal at electroosmotic flow marker, which is more visible when peak intensity cut-off is not applied, B=various amines or derivatized acids, C=amine-derivatized naphthenic acids, D and E=derivatization by-products. 400 300 50 10 100 94   CE-MS data, like all separation-mass spectrometric data, contains information about mass-to-charge ratio of analytes, analyte peak intensity and information about analyte migration time. As such, there are more approaches for clustering and fingerprinting when compared to infusion MS. Figure 4.7, left, shows a 2D plot of a three-component principal component analysis (PCA), with minimum obliqueness criteria, performed on the total ion mass electropherogram of samples according to summed peak intensity and m/z value.   Figure 4.7 Three-component principal component analysis (PCA) plot (first two components) of total ion mass electropherogram peak intensity (left) and mean migration time of selected analytes (right) for positive-ion CE-LRMS of derivatized naphthenic acid ions observed in the raw data for LOS and surficial materials. The three-component PCA explains 97% of the variance in peak intensity (PC1=43%, PC2=30%, PC3=23%; left) and 55% of the variance in mean migration time (PC1=31%, PC2=14%, PC3=10%; right).  95  It appears from the loadings plots that perched water groups with infiltrated water, and groundwater groups with seepage water. These results were not anticipated prior to this investigation. A three-component PCA accounts for 97% (PC1=43%, PC2=30%, PC3=23%) of the variance in the sample dataset and results in a similar clustering to the Orbitrap MS data.  Separation-mass spectrometry data can also be clustered by migration time. In capillary electrophoresis, analyte migration time is inversely proportional to analyte mobility, and analyte mobility is proportional to the charge-to-size ratio of the analyte in question. Therefore, analyte migration times contain complementary information about analyte charge-to-size ratios. Average migration times of selected peaks across the mass-electropherogram were calculated and subject to PCA. Figure 4.7, right, is a plot of the results of PCA on mean migration time data for selected peaks across samples. It also appears from the loadings plots that perched water groups with infiltrated water, and groundwater groups with seepage water. However, three-component PCA accounts for only 55% of the variance in the data set (PC1=31%, PC2=14%, PC3=10%). This suggests that any component or factor analysis of CE-MS migration time data of LOS porewater samples would contain more than three important variables to explain the migration time variation across samples, which could aid in establishing sample CE-MS “fingerprints”. Indeed, a ten-component PCA accounts for 81% of total variance. The values of the m/z and migration times which lead to this discrimination are a topic of on-going investigations.   Furthermore, there are a variety of costlier algorithms that can simultaneously take m/z and migration time data into account and these are currently being investigated. Preliminary evidence indicates both migration time and m/z data from CE-LRMS analyses are important in the characterization of NAs in LOS and surficial water samples and clustering approaches on higher-dimensional, low resolution data may still provide fine-grained sample classification. It is 96  important to note that there are many ways to visualize the results of PCA or FA and the visualizations in Figure 4.7 are one among many options. No experimental data were available for samples not included in Figure 4.7.  4.4 Conclusions  Standard GC-FID method for measuring total petroleum hydrocarbons (PHC) can be effectively supplemented by the analysis of total naphthenic acid fraction compounds (NAFCs): GC-FID of hexane extracts of lean oil sands (LOS) porewater samples provide information on non-polar PHC, whereas analysis of NAFCs in an acid extract of the same samples by Orbitrap MS provides concentrations and speciation on more polar compounds. Both approaches (“level 1” and “level 2”) are recommended for a comprehensive analysis of organic compounds in water samples.  The speciation of NAFCs according to compound class and double bond equivalence is demonstrated to have high potential for assessment of the fate and transformation of oil sands derived components in the materials investigated. Total NAFCs measured by Orbitrap MS overlaid on the factor loadings plot of the raw Orbitrap MS data on the same samples revealed higher concentrated samples clustering together and confirmed prior knowledge concerning the samples’ exposure to LOS NAFCs. Preliminary evidence was observed for use of OxSy heteroatomic species at the molecular level as tracers of possible seepage from the LOS materials. Future work will involve using GC-MS to characterize PHC in hexane extracts, providing complementary information. 97   CE-MS of NAFCs allows for the analysis of all charged analytes in a sample mixture. The sample extraction procedure described, although prescribed for acidic compounds, did not rule out the presence of polymeric and non-polar neutral species, allowing larger coverage of compounds compared to GC-FID and HRMS. Acidic NAFCs were separated from neutral species and the total ion electropherograms from all samples can be divided into 5 zones. The higher-dimensional data obtained by CE-MS can be plotted using component analysis methods according to peak intensity (m/z), by analyte migration time or by both simultaneously. The complementarity of CE-LRMS thus represents a feasible and practical tool to aide in oil sands forensics.  98   Conclusions and Future Work 5.1 Conclusions OSPW is a very complex mixture whose history spans across Indigenous and non-Indigenous cultures, as well as the studies of engineering, chemistry, public health and politics. Analytical chemistry is the science of generating chemical information in the service of the public good, incorporating research, development, innovation (R&D&I), service, knowledge transfer and exchange, and education. Analytical approaches used for known mixtures do not produce expected outcomes when applied to complex unknown mixtures. Therefore, it is paramount for analytical chemistry to establish reliable methodologies using appropriate techniques for characterizing OSPW and NAFCs for the sake of public and environmental health. This thesis reports the establishment of semi-quantitative methods for the analysis of NAFCs in OSPW and other environmental waters using CE-ESI-MS.  CE-ESI-MS was demonstrated to provide targeted and untargeted analyses of urine samples from human patients with PCa or BCa. Human urine is a complex mixture of many components and, via the techniques of standard addition, tandem MS, and SPE, CE-ESI-MS was able to provide information regarding the quantities of select metabolites known to be present in the samples and compare chemical signatures across patients. This work demonstrated the effectiveness of CE-ESI-MS in the analysis of complex mixtures.  For the analysis of NAFCs, standard Aldrich NAs were extracted and derivatized to form amines. The naphthenic amines produced a 3-D chemical signature in CE-ESI-MS under acidic conditions. The method was optimized for BGE composition and was applied to an OSPW sample. This was the first reported optimized capillary electrophoresis-mass spectrometry method for the analysis of NAs. 99  Finally, an application of the optimized CE-ESI-MS method is demonstrated to provide high-dimensional data which is complementary to other techniques, allowing increased insight into complex NAFC mixtures in OSPW. Porewater samples from a lean oil sands waste pile (composed of soil and bitumen) were analyzed by a standard GC-FID method for total petroleum hydrocarbons (TPH), an Orbitrap MS method for total NAFCs, and the developed CE-ESI-MS method. GC-FID and Orbitrap MS results drew attention to the differences between NAFCs and TPH and that both measurements provide complementary data on porewater sample levels. Orbitrap MS provided high-resolution signatures for each sample which were clustered according using factor analysis and linked to a potential distribution model. CE-ESI-MS provided the highest dimensional 3-D chemical signatures (“fingerprints”) which were also explored using principal component analysis. The results suggested that NAFCs could be carried by runoff water from lean oil sands waste sites into the environment. Although there is a movement toward high-resolution MS in the analysis of NAFCs, CE coupled to low-resolution MS was demonstrated to be a powerful, complementary approach with potential to provide more accurate characterization of NAFCs.  5.2 Future work 5.2.1 Complex mixtures and semi-quantification  Complex unknown mixtures such as NAFC in OSPW suffer the limitation of not having appropriate reference standards for calibration, confounding efforts to quantify mixture components. This is a major concern in scientific disciplines which employ analytical chemistry techniques to measure complex mixtures of time-dependent composition, such as nanoparticle 100  ecotoxicology and metabolomics of diseases, and much effort is spent developing and maintaining these materials.129,130 Absolute quantification of an analyte suspected to be present in a sample depends on multiple, subsequent steps of an applied chemical measurement process (CMP) being robust, rugged and well-controlled over time. With a complex mixture like NAFCs containing thousands of different components, most of which are unidentified, many steps in an applied CMP change the samples in unexpected ways leading to differences between actual and expected measurements. An example of this would be WAX SPE of NAs selectively retaining naphthenic acid sulfates and other highly oxygenated anionic species, decreasing the amounts of the more toxic classical NAs, resulting in the incorrect judgment of sample composition. Because the sample is an unknown complex mixture, experiments like this need to be run in order to develop sets of hypotheses that could be tested to lend evidence to the presence of sulfates, which in turn would strengthen the suspicion that NA sulfates were outcompeting classical NAs at some step in the CMP.  Essentially, semi-quantification is what occurs when one or more steps in an applied CMP have a level of uncertainty that is difficult to measure. Since uncertainty is propagated throughout the steps of CMP, the extent of error propagation is also uncertain. Therefore, it is not a matter of high uncertainty, but of high uncertainty in measurement uncertainty. Therefore, the term “semi-quantification” is reserved: the result is not absolutely quantitative, nor is it entirely qualitative. Semi-quantitative analyses are necessary prerequisites on route to absolute quantifications of the components of complex mixtures.  101  5.2.2 Analytical complementarity and data fusion Two or more data sets obtained on the same objects of study via different techniques are often described as “complementary” to each other. There is a danger in seeing this statement as a kind of truism: if a technique is complementary to others because it is not identical to the others, then “complementary” becomes another way of stating something that is obvious—that two different techniques are different—and the understanding of a complex mixture sample is no further ahead. Instead, complementarity is useful, especially for semi-quantification, and there are different types of analytical complementarity that should be distinguished A chemical measurement process (CMP) is often a multi-step process. The complementarity of two CMPs on the same sample depends on the precise differences between the two methodologies and the expected data type, which fundamentally depends on exploited interactions between materials and energy. In some cases, two CMPs are exclusive because each provides a different type of data. An example of two techniques having exclusive complementarity would be NMR and MS: NMR provides atomic connectivity (i.e. bonding) in molecules, and MS provides identification of atoms in ions. In other cases, two techniques have non-exclusive complementarity, where although steps in the respective CMPs differ, causing exclusivity at some steps in the CMPs, the output data type is the same (MS). An example of this is CE-MS and LC-MS: although CZE in acidic buffer and reverse-phase LC are orthogonal techniques in the domain of chemical separation, the two techniques share the use of MS and the output data type is still m/z versus analyte detection time. Any two CMPs can have different numbers of steps, one or more differences in techniques, and can be hyphenated or non-hyphenated—an excellent combinatorics problem! 102  Data fusion is an active area of chemometrics which employs a combination of multivariate statistics and machine learning techniques to create multi-dimensional fingerprints of complex samples that are often coupled to automated decision-making processes. Many research reports exist of consolidating separation-mass spectrometric and spectroscopic data from multiple techniques on the same samples.131,132 Often the instrumentation used in data fusion contains detectors that can switch through multiple modes, generating data of different dimensionality in the same apparatus (e.g. LC-diode array-MS). Spectral imaging makes routine use of data fusion, food quality chemistry, petroleum chemistry and metabolomics have all forayed into this area,131,133–135 and data fusion techniques have been used in the analytical characterization of waste streams.136 The characterization of NAFCs in OSPW is important in understanding its environmental and health impacts, which provide the major inputs for responsible policymaking. Since OSPW is a complex mixture, data fusion chemometric techniques hold promise.  5.2.3 Chemical derivatization and tandem MS In terms of identifying individual NAFCs in OSPW, which is important from toxicological and remediation perspectives, chemical labelling will be a route for active research. Just as proteins and glycans are “sequenced” with reagents and tandem MS, NAFCs will also need to employ chemical tools for their identification as individual NAFCs or as sub-classes. Recently, we published a chemical derivatization MS/MS method to reveal the presence of dicarboxylic O4 NAFCs.137 A next step would be to use an isobaric labelling series and targeted MS/MS fragmentation to explore differences among complex NAFCs mixtures.  103  5.2.4 Artificial neural networks for pattern recognition Finally, for a more robust classification of 3-D or higher-dimensional chemical signatures produced by CE-ESI-MS, stronger classification algorithms such as artificial neural networks (ANN) are needed. More research also needs to be done in terms of how to preprocess the high-dimensional CE-ESI-MS data of NAFCs prior to statistical comparisons: do spectra need to be aligned? Is classification more robust in the mobility domain than in the time domain? What type of classification functions are necessary? Since capillary electrophoresis is amenable to miniaturization, perhaps microfluidic monitoring devices capable of continuously classifying environmental water in-situ are not far off.            104      Bibliography  (1)  Headley, J. V.; Peru, K. M.; Mohamed, M. H.; Frank, R. A.; Martin, J. W.; Hazewinkel, R. R. O.; Humphries, D.; Gurprasad, N. P.; Hewitt, L. M.; Muir, D. C. G.; et al. Chemical Fingerprinting of Naphthenic Acids and Oil Sands Process Waters—A Review of Analytical Methods for Environmental Samples. J. Environ. Sci. Heal. Part A 2013, 48 (10), 1145–1163. (2)  McKenzie, N.; Yue, S.; Liu, X.; Ramsay, B. A.; Ramsay, J. A. Biodegradation of Naphthenic Acids in Oils Sands Process Waters in an Immobilized Soil/Sediment Bioreactor. Chemosphere 2014, 109, 164–172. (3)  Headley, J. V.; Peru, K. M.; Fahlman, B.; Colodey, A.; McMartin, D. W. Selective Solvent Extraction and Characterization of the Acid Extractable Fraction of Athabasca Oils Sands Process Waters by Orbitrap Mass Spectrometry. Int. J. Mass Spectrom. 2013, 345–347, 104–108. (4)  Allen, E. W. Process Water Treatment in Canada’s Oil Sands Industry: I. Target Pollutants and Treatment Objectives. J. Environ. Eng. Sci. 2008, 7 (2), 123–138. (5)  Hao, C.; Headley, J. V.; Peru, K. M.; Frank, R.; Yang, P.; Solomon, K. R. 105  Characterization and Pattern Recognition of Oil-Sand Naphthenic Acids Using Comprehensive Two-Dimensional Gas Chromatography/Time-of-Flight Mass Spectrometry. J. Chromatogr. A 2005, 1067 (1–2), 277–284. (6)  Headley, J. V.; Peru, K. M.; Mohamed, M. H.; Wilson, L. D.; McMartin, D. W.; Mapolelo, M. M.; Lobodin, V. V.; Rodgers, R. P.; Marshall, A. G. Atmospheric Pressure Photoionization Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Characterization of Tunable Carbohydrate-Based Materials for Sorption of Oil Sands Naphthenic Acids. Energy & Fuels 2014, 28 (3), 1611–1616. (7)  Headley, J. V.; McMartin, D. W. A Review of the Occurrence and Fate of Naphthenic Acids in Aquatic Environments. J. Environ. Sci. Heal. Part A 2004, 39 (8), 1989–2010. (8)  Mohamed, M. H.; Wilson, L. D.; Headley, J. V.; Peru, K. M. Sequestration of Naphthenic Acids from Aqueous Solution Using β-Cyclodextrin-Based Polyurethanes. Phys. Chem. Chem. Phys. 2011, 13 (3), 1112–1122. (9)  Janfada, A.; Headley, J. V.; Peru, K. M.; Barbour, S. L. A Laboratory Evaluation of the Sorption of Oil Sands Naphthenic Acids on Organic Rich Soils. J. Environ. Sci. Health. A. Tox. Hazard. Subst. Environ. Eng. 2006, 41 (6), 985–997. (10)  Han, X.; Scott, A. C.; Fedorak, P. M.; Bataineh, M.; Martin, J. W. Influence of Molecular Structure on the Biodegradability of Naphthenic Acids. Environ. Sci. Technol. 2008, 42 (4), 1290–1295. (11)  Han, X.; MacKinnon, M. D.; Martin, J. W. Estimating the in Situ Biodegradation of Naphthenic Acids in Oil Sands Process Waters by HPLC/HRMS. Chemosphere 2009, 76 (1), 63–70. 106  (12)  Frank, R. A.; Fischer, K.; Kavanagh, R.; Kent Burnison, B.; Arsenault, G.; Headley, J. V; Peru, K. M.; Van Glen Kraak, D. E. R.; Solomon, K. R. Effect of Carboxylic Acid Content on the Acute Toxicity of Oil Sands Naphthenic Acids. Environ. Sci. Technol. 2009, 43 (2), 266–271. (13)  Grewer, D. M.; Young, R. F.; Whittal, R. M.; Fedorak, P. M. Naphthenic Acids and Other Acid-Extractables in Water Samples from Alberta: What Is Being Measured? Sci. Total Environ. 2010, 408 (23), 5997–6010. (14)  Headley, J. V; McMartin, D. W. A Review of the Occurrence and Fate of Naphthenic Acids in Aquatic Environments. J. Environ. Sci. Heal. Part a-Toxic/Hazardous Subst. Environ. Eng. 2004, 39 (8), 1989–2010. (15)  Headley, J. J. V.; Peru, K. M. K.; Barrow, M. M. P. Mass Spectrometric Characterization of Naphthenic Acids in Environmental Samples: A Review. Mass Spectrometry Reviews, 2009, 28, 121–134. (16)  Morandi, G. D.; Wiseman, S. B.; Pereira, A.; Mankidy, R.; Gault, I. G. M.; Martin, J. W.; Giesy, J. P. Effects-Directed Analysis of Dissolved Organic Compounds in Oil Sands Process-Affected Water. Environ. Sci. Technol. 2015, 49 (20), 12395–12404. (17)  Rowland, S. J.; Scarlett, A. G.; Jones, D.; West, C. E.; Frank, R. A. Diamonds in the Rough: Identification of Individual Naphthenic Acids in Oil Sands Process Water. Environ. Sci. Technol. 2011, 45 (7), 3154–3159. (18)  Wilde, M. J.; West, C. E.; Scarlett, A. G.; Jones, D.; Frank, R. A.; Hewitt, L. M.; Rowland, S. J. Bicyclic Naphthenic Acids in Oil Sands Process Water: Identification by Comprehensive Multidimensional Gas Chromatography-Mass Spectrometry. J. 107  Chromatogr. A 2015, 1378, 74–87. (19)  Proposal to Add Naphthenic Acids to the NPRI for the 2012 Reporting Year Recommendations of the NPRI Multi-Stakeholder Work Group; 2012. (20)  Headley, J. V.; Peru, K. M.; Barrow, M. P. Advances in Mass Spectrometric Characterization of Naphthenic Acids Fraction Compounds in Oil Sands Environmental Samples and Crude Oil-A Review. Mass Spectrom. Rev. 2016, 35, 311–328. (21)  Brunswick, P.; Shang, D.; van Aggelen, G.; Hindle, R.; Hewitt, L. M.; Frank, R. A.; Haberl, M.; Kim, M. Trace Analysis of Total Naphthenic Acids in Aqueous Environmental Matrices by Liquid Chromatography/Mass Spectrometry-Quadrupole Time of Flight Mass Spectrometry Direct Injection. J. Chromatogr. A 2015, 1405 (1), 49–71. (22)  Rodgers, R. P.; McKenna, A. M. Petroleum Analysis. Anal. Chem. 2011, 83 (12), 4665–4687. (23)  Scott, A. C.; Young, R. F.; Fedorak, P. M. Comparison of GC-MS and FTIR Methods for Quantifying Naphthenic Acids in Water Samples. Chemosphere 2008, 73 (8), 1258–1264. (24)  Kavanagh, R. J.; Burnison, B. K.; Frank, R. A.; Solomon, K. R.; Van Der Kraak, G. Detecting Oil Sands Process-Affected Waters in the Alberta Oil Sands Region Using Synchronous Fluorescence Spectroscopy. Chemosphere 2009, 76 (1), 120–126. (25)  Mohamed, M. H.; Wilson, L. D.; Headley, J. V; Peru, K. M. Screening of Oil Sands Naphthenic Acids by UV-Vis Absorption and Fluorescence Emission Spectrophotometry. J. Environ. Sci. Heal. Part a-Toxic/Hazardous Subst. Environ. Eng. 2008, 43 (14), 1700–1705. 108  (26)  Martin, J. W.; Han, X.; Headley, J. V.; Peru, K. M.; Headley, J. V. Comparison of High- and Low-Resolution Electrospray Ionization Mass Spectrometry for the Analysis of Naphthenic Acid Mixtures in Oil Sands Process Water. Rapid Commun. Mass Spectrom. 2008, 22 (12), 1919–1924. (27)  Barrow, M. P.; Witt, M.; Headley, J. V.; Peru, K. M. Athabasca Oil Sands Process Water: Characterization by Atmospheric Pressure Photoionization and Electrospray Ionization Fourier Transform Ion Cyclotron Resonance Mass Spectrometry. Anal. Chem. 2010, 82 (9), 3727–3735. (28)  Han, X.; MacKinnon, M. D.; Martin, J. W. Estimating the in Situ Biodegradation of Naphthenic Acids in Oil Sands Process Waters by HPLC/HRMS. Chemosphere 2009, 76 (1), 63–70. (29)  Nyakas, A.; Han, J.; Peru, K. M.; Headley, J. V.; Borchers, C. H. Comprehensive Analysis of Oil Sands Processed Water by Direct-Infusion Fourier-Transform Ion Cyclotron Resonance Mass Spectrometry with and without Offline UHPLC Sample Prefractionation. Environ. Sci. Technol. 2013, 47 (9), 4471–4479. (30)  Holowenko, F. M.; MacKinnon, M. D.; Fedorak, P. M. Characterization of Naphthenic Acids in Oil Sands Wastewaters by Gas Chromatography-Mass Spectrometry. Water Res. 2002, 36 (11), 2843–2855. (31)  Bataineh, M.; Scott, A. C.; Fedorak, P. M.; Martin, J. W. Capillary HPLC/QTOF-MS for Characterizing Complex Naphthenic Acid Mixtures and Their Microbial Transformation. Anal. Chem. 2006, 78 (24), 8354–8361. (32)  Hindle, R.; Noestheden, M.; Peru, K. M.; Headley, J. V. Quantitative Analysis of 109  Naphthenic Acids in Water by Liquid Chromatography-Accurate Mass Time-of-Flight Mass Spectrometry. J. Chromatogr. A 2013, 1286, 166–174. (33)  Smith, B. E.; Rowland, S. J. A Derivatisation and Liquid Chromatography/Electrospray Ionisation Multistage Mass Spectrometry Method for the Characterisation of Naphthenic Acids. Rapid Commun. Mass Spectrom. 2008, 22 (23), 3909–3927. (34)  Shang, D.; Kim, M.; Haberl, M.; Legzdins, A. Development of a Rapid Liquid Chromatography Tandem Mass Spectrometry Method for Screening of Trace Naphthenic Acids in Aqueous Environments. J. Chromatogr. A 2013, 1278, 98–107. (35)  Woudneh, M. B.; Coreen Hamilton, M.; Benskin, J. P.; Wang, G.; McEachern, P.; Cosgrove, J. R. A Novel Derivatization-Based Liquid Chromatography Tandem Mass Spectrometry Method for Quantitative Characterization of Naphthenic Acid Isomer Profiles in Environmental Waters. J. Chromatogr. A 2013, 1293 (1), 36–43. (36)  Headley, J. V.; Peru, K. M.; Fahlman, B.; Colodey, A.; McMartin, D. W. Selective Solvent Extraction and Characterization of the Acid Extractable Fraction of Athabasca Oils Sands Process Waters by Orbitrap Mass Spectrometry. Int. J. Mass Spectrom. 2013, 345–347, 104–108. (37)  Wrona, F. J.; di Cenzo, P. Lower Athabasca Water Quality Monitoring Program: PHASE 1, Athabasca River Mainstem and Major Tributaries; 2011. (38)  Malle, H.; Simser, J. Final Report for Naphthenic Acids - Interlaboratory Calibration Study; 2014. (39)  Barrow, M. P.; Peru, K. M.; McMartin, D. W.; Headley, J. V. Effects of Extraction PH on 110  the Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Profiles of Athabasca Oil Sands Process Water. Energy & Fuels 2015, 30 (5), acs.energyfuels.5b02086. (40)  Celsie, A.; Parnis, J. M.; Mackay, D. Impact of Temperature, PH, and Salinity Changes on the Physico-Chemical Properties of Model Naphthenic Acids. Chemosphere 2016, 146, 40–50. (41)  Huang, R.; McPhedran, K. N.; Sun, N.; Chelme-Ayala, P.; El-Din, M. G. Investigation of the Impact of Organic Solvent Type and Solution PH on the Extraction Efficiency of Naphthenic Acids from Oil Sands Process-Affected Water. Chemosphere 2016, 146, 472–477. (42)  Pereira, A. S.; Martin, J. W. Exploring the Complexity of Oil Sands Process-Affected Water by High Efficiency Supercritical Fluid Chromatography/Orbitrap Mass Spectrometry. Rapid Commun. Mass Spectrom. 2015, 29 (8), 735–744. (43)  Ross, M. S.; Pereira, A. D.; Fennell, J.; Davies, M.; Johnson, J.; Sliva, L.; Martin, J. W. Quantitative and Qualitative Analysis of Naphthenic Acids in Natural Waters Surrounding the Canadian Oil Sands Industry. Environ. Sci. Technol. 2012, 46 (23), 12796–12805. (44)  Pereira, A. S.; Bhattacharjee, S.; Martin, J. W. Characterization of Oil Sands Process-Affected Waters by Liquid Chromatography Orbitrap Mass Spectrometry. Environ. Sci. Technol. 2013, 47 (10), 5504–5513. (45)  Duncan, K. D.; Letourneau, D. R.; Vandergrift, G. W.; Jobst, K.; Reiner, E.; Gill, C. G.; Krogh, E. T. A Semi-Quantitative Approach for the Rapid Screening and Mass Profiling of Naphthenic Acids Directly in Contaminated Aqueous Samples. J. Mass Spectrom. 2016, 51 (1), 44–52. 111  (46)  Duncan, K. D.; Volmer, D. A.; Gill, C. G.; Krogh, E. T. Rapid Screening of Carboxylic Acids from Waste and Surface Waters by ESI-MS/MS Using Barium Ion Chemistry and On-Line Membrane Sampling. J. Am. Soc. Mass Spectrom. 2016, 27 (3), 443–450. (47)  Headley, J. V; Peru, K. M.; McMartin, D. W.; Winkler, M. Determination of Dissolved Naphthenic Acids in Natural Waters by Using Negative-Ion Electrospray Mass Spectrometry. J. AOAC Int. 2002, 85 (1), 182–187. (48)  Ortiz, X.; Jobst, K. J.; Reiner, E. J.; Backus, S. M.; Peru, K. M.; McMartin, D. W.; O’Sullivan, G.; Taguchi, V. Y.; Headley, J. V. Characterization of Naphthenic Acids by Gas Chromatography-Fourier Transform Ion Cyclotron Resonance Mass Spectrometry. Anal. Chem. 2014, 86 (15), 7666–7673. (49)  Schmitt-Kopplin, P.; Garmash, A. V.; Kudryavtsev, A. V.; Menzinger, F.; Perminova, I. V.; Hertkorn, N.; Freitag, D.; Petrosyan, V. S.; Kettrup, A. Quantitative and Qualitative Precision Improvements by Effective Mobility-Scale Data Transformation in Capillary Electrophoresis Analysis. Electrophoresis 2001, 22, 77–87. (50)  Wiersema, P. H.; Loeb, A. L.; Overbeek, J. T. G. Calculation of the Electrophoretic Mobility of a Spherical Colloid Particle. J. Colloid Interface Sci. 1966, 22 (1), 78–99. (51)  Lyklema, J. Electrokinetics after Smoluchowski. Colloids Surfaces A Physicochem. Eng. Asp. 2003, 222 (1–3), 5–14. (52)  Maxwell, E. J.; Zhong, X.; Zhang, H.; van Zeijl, N.; Chen, D. D. Y. Decoupling CE and ESI for a More Robust Interface with MS. Electrophoresis 2010, 31 (7), 1130–1137. (53)  Maxwell, E. J.; Ratnayake, C.; Jayo, R.; Zhong, X.; Chen, D. D. Y. A Promising Capillary 112  Electrophoresis-Electrospray Ionization-Mass Spectrometry Method for Carbohydrate Analysis. Electrophoresis 2011, 32 (16), 2161–2166. (54)  Jayo, R. G.; Thaysen-Andersen, M.; Lindenburg, P. W.; Haselberg, R.; Hankemeier, T.; Ramautar, R.; Chen, D. D. Y. Simple Capillary Electrophoresis-Mass Spectrometry Method for Complex Glycan Analysis Using a Flow-through Microvial Interface. Anal. Chem. 2014, 86 (13), 6479–6486. (55)  Soliman, L. C.; Hui, Y.; Hewavitharana, A. K.; Chen, D. D. Y. Monitoring Potential Prostate Cancer Biomarkers in Urine by Capillary Electrophoresis-Tandem Mass Spectrometry. J. Chromatogr. A 2012, 1267, 162–169. (56)  Lindenburg, P. W.; Ramautar, R.; Jayo, R. G.; Chen, D. D. Y.; Hankemeier, T. Capillary Electrophoresis-Mass Spectrometry Using a Flow-through Microvial Interface for Cationic Metabolome Analysis. Electrophoresis 2013, 1308–1314. (57)  Zhong, X.; Maxwell, E. J.; Ratnayake, C.; Mack, S.; Chen, D. D. Y. Flow-through Microvial Facilitating Interface of Capillary Isoelectric Focusing and Electrospray Ionization Mass Spectrometry. Anal. Chem. 2011, 83 (22), 8748–8755. (58)  Zhong, X.; Maxwell, E. J.; Chen, D. D. Y. Mass Transport in a Micro Flow-through Vial of a Junction-at-the-Tip Capillary Electrophoresis-Mass Spectrometry Interface. Anal. Chem. 2011, 83 (12), 4916–4923. (59)  Maxwell, E. J.; Chen, D. D. Y. Twenty Years of Interface Development for Capillary Electrophoresis-Electrospray Ionization-Mass Spectrometry. Anal. Chim. Acta 2008, 627 (1), 25–33. 113  (60)  Klepárník, K. Recent Advances in Combination of Capillary Electrophoresis with Mass Spectrometry: Methodology and Theory. Electrophoresis 2015, 36 (1), 159–178. (61)  Ramos-Payán, M.; Ocaña-Gonzalez, J. A.; Fernández-Torres, R. M.; Llobera, A.; Bello-López, M. Á. Recent Trends in Capillary Electrophoresis for Complex Samples Analysis: A Review. Electrophoresis 2018, 39 (1), 111–125. (62)  Hamdan, I. I. Capillary Electrophoresis in the Analysis of Pharmaceuticals in Environmental Water: A Critical Review. J. Liq. Chromatogr. Relat. Technol. 2017, 40 (3), 111–125. (63)  Sun, L.; Zhu, G.; Yan, X.; Zhang, Z.; Wojcik, R.; Champion, M. M.; Dovichi, N. J. Capillary Zone Electrophoresis for Bottom-up Analysis of Complex Proteomes. Proteomics 2016, 16 (2), 188–196. (64)  Liu, C.; Luo, Y.; Maxwell, E. J.; Fang, N.; Chen, D. D. Y. Letters to Analytical Chemistry Reverse of Mixing Process with a Two-Dimensional Electro-Fluid-Dynamic Device. Anal. Chem. 2010, 82 (6), 2182–2185. (65)  Spratlin, J. L.; Serkova, N. J.; Eckhardt, S. G. Clinical Applications of Metabolomics in Oncology: A Review. Clin. cancer Res. 2009, 15 (2), 431–440. (66)  Davis, V. W.; Bathe, O. F.; Schiller, D. E.; Slupsky, C. M.; Sawyer, M. B. Metabolomics and Surgical Oncology: Potential Role for Small Molecule Biomarkers. J. Surg. Oncol. 2011, 103 (5), 451–459. (67)  Hyndman, M. E.; Mullins, J. K.; Bivalacqua, T. J. Metabolomics and Bladder Cancer. Urol. Oncol. 2011, 29 (5), 558–561. 114  (68)  Rao, A. R.; Motiwala, H. G.; Karim, O. M. a. The Discovery of Prostate-Specific Antigen. BJU Int. 2008, 101 (1), 5–10. (69)  Wallner, L. P.; Jacobsen, S. J. Prostate-Specific Antigen and Prostate Cancer Mortality. Am. J. Prev. Med. 2013, 45 (3), 318–326. (70)  Kattan, M. W.; Eastham, J. A.; Stapleton, A. M. F.; Wheeler, T. M.; Scardino, P. T. A Preoperative Nomogram for Disease Recurrence Following Radical Prostatectomy for Prostate Cancer. J. Natl. Cancer Inst. 1998, 90 (10), 766–771. (71)  Sreekumar, A.; Poisson, L. M.; Rajendiran, T. M.; Khan, A. P.; Cao, Q.; Yu, J.; Laxman, B.; Mehra, R.; Lonigro, R. J.; Li, Y.; et al. Metabolomic Profiles Delineate Potential Role for Sarcosine in Prostate Cancer Progression. Nature 2009, 457 (7231), 910–914. (72)  Khan, A. P.; Rajendiran, T. M.; Ateeq, B.; Asangani, I. A.; Athanikar, J. N.; Yocum, A. K.; Mehra, R.; Siddiqui, J.; Palapattu, G.; Wei, J. T.; et al. The Role of Sarcosine Metabolism in Prostate Cancer Progression. Neoplasia 2013, 15 (5), 491–501. (73)  Vogel, S. de; Ulvik, A.; Meyer, K.; Ueland, P. M.; Nygård, O.; Vollset, S. E.; Tell, G. S.; Gregory III, J. F.; Tretli, S.; Bjørge, T. Sarcosine and Other Metabolites along the Choline Oxidation Pathway in Relation to Prostate Cancer - a Large Nested Case-Control Study within the JANUS Cohort in Norway. Int. J. Cancer 2013, No. Accepted Article, 2–26. (74)  Tripathi, P.; Somashekar, B. S.; Ponnusamy, M.; Gursky, A.; Dailey, S.; Kunju, P.; Lee, C. T.; Chinnaiyan, A. M.; Rajendiran, T. M.; Ramamoorthy, A. HR-MAS NMR Tissue Metabolomic Signatures Cross-Validated by Mass Spectrometry Distinguish Bladder Cancer from Benign Disease. J. Proteome Res. 2013, 12, 3519–3528. 115  (75)  Saude, E. J.; Sykes, B. D. Urine Stability for Metabolomic Studies: Effects of Preparation and Storage. Metabolomics 2007, 3 (1), 19–27. (76)  Gurcan, M. N.; Boucheron, L.; Can, A.; Madabhushi, A.; Rajpoot, N.; Yener, B. Histopathological Image Analysis: A Review. IEEE Rev. Biomed. Eng. 2009, 2, 147–171. (77)  Shukla-Dave, A.; Hricak, H.; Ishill, N.; Moskowitz, C. S.; Drobnjak, M.; Reuter, V. E.; Zakian, K. L.; Scardino, P. T.; Cordon-Cardo, C. Prediction of Prostate Cancer Recurrence Using Magnetic Resonance Imaging and Molecular Profiles. Clin. Cancer Res. 2009, 15 (11), 3842–3849. (78)  Kurhanewicz, J. Magnetic Resonance Imaging Biomarkers of Prostate Cancer. Am. Assoc. Cancer Res. Educ. B. 2008, 2008, 523–526. (79)  Zakian, K. L.; Sircar, K.; Hricak, H.; Chen, H.-N.; Shukla-Dave, A.; Eberhardt, S.; Muruganandham, M.; Ebora, L.; Kattan, M. W.; Reuter, V. E.; et al. Correlation of Proton MR Spectroscopic Imaging with Gleason Score Based on Step-Section Pathologic Analysis after Radical Prostatectomy. Radiology 2005, 234 (3), 804–814. (80)  Nimesh, S.; Mohottalage, S.; Vincent, R.; Kumarathasan, P. Current Status and Future Perspectives of Mass Spectrometry Imaging. Int. J. Mol. Sci. 2013, 14, 11277–11301. (81)  Amling, C. L. Diagnosis and Management of Superficial Bladder Cancer. Curr. Probl. Cancer 2001, 25 (4), 219–278. (82)  Gleason, D. F. Histologic Grading of Prostate Cancer: A Perspective. Hum. Pathol. 1992, 23 (3), 273–279. (83)  Clarke, J. T. Colorimetric Determination and Distribution of Urinary Creatinine and 116  Creatine. Clin. Chem. 1961, 7 (4), 271–283. (84)  Blass, B. K. G.; Thibert, R. J.; Lam, K. A Study of the Mechanism of the Jaffé Reaction. Z. Klin. Chem. Klin. Biochem. 1974, 12, 336–343. (85)  Vasiliades, J. Reaction of Alkaline Sodium Picrate with Creatinine: I. Kinetics and Mechanism of Formation of the Mono-Creatinine Picric Acid Complex. Clin. Chem. 1976, 22 (10), 1664–1671. (86)  Salter, C. Error Analysis Using the Variance-Covariance Matrix. J. Chem. Educ. 2000, 77 (9), 1239. (87)  Bland, J. M.; Altman, D. G. Analysis of Continuous Data from Small Samples. BMJ 2009, 338. (88)  Williams, A.; Ibrahim, I. T. Carbodiimide Chemistry: Recent Advances. Chem. Rev. 1981, 81 (6), 589–636. (89)  Mikołajczyk, M.; Kiełbasiński, P. Recent Developments in the Carbodiimide Chemistry. Tetrahedron 1981, 37, 233–284. (90)  Nakajima, N.; Ikada, Y. Mechanism of Amide Formation by Carbodiimide for Bioconjugation in Aqueous Media. Bioconjug. Chem. 1995, 6 (1), 123–130. (91)  Gilles, M. A.; Hudson, A. Q.; Borders, C. L. Stability of Water-Soluble Carbodiimides in Aqueous Solution. Anal. Biochem. 1990, 184 (2), 244–248. (92)  Cline, G. W.; Hanna, S. B. Kinetics and Mechanisms of the Aminolysis of N-Hydroxysuccinimide Esters in Aqueous Buffers. J. Org. Chem. 1988, 53 (15), 3583–3586. (93)  Staros, J. V.; Wright, R. W.; Swingle, D. M. Enhancement by N-117  Hydroxysulfosuccinimide of Water-Soluble Carbodiimide-Mediated Coupling Reactions. Anal. Biochem. 1986, 156 (1), 220–222. (94)  Grabarek, Z.; Gergely, J. Zero-Length Crosslinking Procedure with the Use of Active Esters. Anal. Biochem. 1990, 185 (1), 131–135. (95)  Hermanson, G. T. Bioconjugate Techniques, 2nd ed.; Elsevier Inc., 2008. (96)  Kurzer, F.; Douraghi-Zadeh, K. Advances in the Chemistry of Carbodiimides. Chem. Rev. 1967, 67 (2), 107–152. (97)  Hoare, D. G.; Koshland, D. E. A Method for the Quantitative Modification and Estimation of Carboxylic Acid Groups in Proteins. J. Biol. Chem. 1967, 242 (10), 2447–2453. (98)  Rich, D. H.; Singh, J. The Carbodiimide Method. In The Peptides Vol. 1; 1979; pp 241–261. (99)  Sam, S.; Touahir, L.; Salvador Andresa, J.; Allongue, P.; Chazalviel, J. N.; Gouget-Laemmel, A. C.; De Villeneuve, C. H.; Moraillon, A.; Ozanam, F.; Gabouze, N.; et al. Semiquantitative Study of the EDC/NHS Activation of Acid Terminal Groups at Modified Porous Silicon Surfaces. Langmuir 2010, 26 (2), 809–814. (100)  Wang, C.; Yan, Q.; Liu, H. B.; Zhou, X. H.; Xiao, S. J. Different EDC/NHS Activation Mechanisms between PAA and PMAA Brushes and the Following Amidation Reactions. Langmuir 2011, 27 (19), 12058–12068. (101)  Volonterio, A.; Zanda, M. Domino Condensation / Aza-Michael / O “ N Acyl Migration of Carbodiimides with Activated a , b-Unsaturated Carboxylic Acids to Form Hydantoins. J. Org. Chem. 2005, 70, 2161–2170. 118  (102)  Wei, K. L.; Wu, C. H.; Huang, W. H.; Lin, J. J.; Dai, S. A. N-Aryl Acylureas as Intermediates in Sequential Self-Repetitive Reactions to Form Poly (Amide-Imide)S. Macromolecules 2006, 39 (1), 12–14. (103)  Chambers, M. C.; Maclean, B.; Burke, R.; Amodei, D.; Ruderman, D. L.; Neumann, S.; Gatto, L.; Fischer, B.; Pratt, B.; Egertson, J.; et al. A Cross-Platform Toolkit for Mass Spectrometry and Proteomics. Nat. Biotechnol. 2012, 30 (10), 918–920. (104)  Sturm, M.; Bertsch, A.; Gröpl, C.; Hildebrandt, A.; Hussong, R.; Lange, E.; Pfeifer, N.; Schulz-Trieglaff, O.; Zerck, A.; Reinert, K.; et al. OpenMS – An Open-Source Software Framework for Mass Spectrometry. BMC Bioinformatics 2008, 9 (1), 163. (105)  Kohlbacher, O.; Reinert, K.; Gropl, C.; Lange, E.; Pfeifer, N.; Schulz-Trieglaff, O.; Sturm, M. TOPP--the OpenMS Proteomics Pipeline. Bioinformatics 2007, 23 (2), e191–e197. (106)  R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing: Vienna, Austria 2014. (107)  MacLennan, M. S. 2015-06-29-EDC-NAs-CE-ESI-MS: For Publication. 2016. (108)  Sheehan, J. C.; Cruickshank, P. A.; Boshart, G. L. Convenient Synthesis of Water-Soluble Carbodiimides. J. Org. Chem. 1961, 26 (7), 2525–2528. (109)  Katritzky, A. R.; Yao, J.; Qi, M.; Chou, Y.; Sikora, D. J.; Davis, S. Ring Opening Reactions of Succinimides. Heterocycles 1998, 48 (12), 2677–2691. (110)  Burrowes, A.; Marsh, R.; Evans, C.; Teare, M.; Ramos, S.; Rahnama, F.; Kirsch, M.-A.; Philp, L.; Stenson, J.; Yemane, M.; et al. Alberta’s Energy Reserves 2008 and Supply/Demand Outlook 2009−2018; Hill, C., Crowfoot, C., MacGillivray, J., Rahnama, 119  F., Eds.; Energy Resources Conservation Board: Calgary, Alberta, Canada, 2009. (111)  Rogers, V. V.; Wickstrom, M.; Liber, K.; MacKinnon, M. D. Acute and Subchronic Mammalian Toxicity of Naphthenic Acids from Oil Sands Tailings. Toxicol. Sci. 2002, 66 (2), 347–355. (112)  Korbas, T. S. Degradation and Mobility of Petroleum Hydrocarbons in Oilsands Waste at the Aurora Fort Hills Disposal Area; University of Saskatchewan: Saskatoon, SK, 2013. (113)  Scale, K. O.; Korbas, T. S.; Fleming, I. R. Degradation and Mobility of Petroleum Hydrocarbons in Oil Sand Waste. Environ. Geotech. 2017, 4 (6), 402–414. (114)  Canadian Council of Ministers of the Environment; (CCME). Reference Method for the Canada-Wide Standard for Petroleum Hydrocarbons in Soil: Tier 1 Method, Including Addendum 1; Canadian Council of Ministers of the Environment Inc.: Winnipeg, Manitoba, 2001. (115)  Frank, R. A.; Roy, J. W.; Bickerton, G.; Rowland, S. J.; Headley, J. V.; Scarlett, A. G.; West, C. E.; Peru, K. M.; Parrott, J. L.; Conly, F. M.; et al. Profiling Oil Sands Mixtures from Industrial Developments and Natural Groundwaters for Source Identification. Environ. Sci. Technol. 2014, 48 (5), 2660–2670. (116)  Keller, B. O.; Sui, J.; Young, A. B.; Whittal, R. M. Interferences and Contaminants Encountered in Modern Mass Spectrometry. Anal. Chim. Acta 2008, 627 (1), 71–81. (117)  Kessner, D.; Chambers, M.; Burke, R.; Agus, D.; Mallick, P. ProteoWizard: Open Source Software for Rapid Proteomics Tools Development. Bioinformatics 2008, 24 (21), 2534–2536. 120  (118)  Martens, L.; Chambers, M.; Sturm, M.; Kessner, D.; Levander, F.; Shofstahl, J.; Tang, W. H.; Römpp, A.; Neumann, S.; Pizarro, A. D.; et al. MzML—a Community Standard for Mass Spectrometry Data. Mol. Cell. Proteomics 2011, 10 (1), R110.000133. (119)  Pedrioli, P. G. A.; Eng, J. K.; Hubley, R.; Vogelzang, M.; Deutsch, E. W.; Raught, B.; Pratt, B.; Nilsson, E.; Angeletti, R. H.; Apweiler, R.; et al. A Common Open Representation of Mass Spectrometry Data and Its Application to Proteomics Research. Nat. Biotechnol. 2004, 22 (11), 1459–1466. (120)  Keller, A.; Eng, J.; Zhang, N.; Li, X.; Aebersold, R. A Uniform Proteomics MS/MS Analysis Platform Utilizing Open XML File Formats. Mol. Syst. Biol. 2005, 1 (1), E1–E8. (121)  Revelle, W. Psych: Procedures for Psychological, Psychometric, and Personality Research. Evanston, Illinois 2016. (122)  Baddeley, A.; Rubak, E.; Turner, R. Spatial Point Patterns: Methodology and Applications with R; Chapman & Hall/CRC: London, 2015. (123)  Sarkar, D. Lattice: Multivariate Data Visualization with R; Springer: New York, 2008. (124)  Soetaert, K. Plot3D: Plotting Multi-Dimensional Data. 2016. (125)  Gorsuch, R. L. Factor Analysis, 2nd ed.; Routledge: New York, NY, 2015. (126)  Bernaards, C. A.; Jennrich, R. I. Gradient Projection Algorithms and Software for Arbitrary Rotation Criteria in Factor Analysis. Educ. Psychol. Meas. 2005, 65 (5), 676–696. (127)  Spicer, J. Making Sense of Multivariate Data Analysis; SAGE Publications, Inc.: Thousand Oaks, CA, 2005. 121  (128)  Jolliffe, I. T. Principal Components Analysis, 2nd ed.; 2, Ed.; Springer: New York, 2002. (129)  Aitken, R. J.; Hankin, S. M.; Lang Tran, C.; Donaldson, K.; Stone, V.; Cumpson, P.; Johnstone, J.; Chaudhry, Q.; Cash, S.; Garrod, J. A Multidisciplinary Approach to the Identification of Reference Materials for Engineered Nanoparticle Toxicology. Nanotoxicology 2008, 2 (2), 71–78. (130)  Simón-Manso, Y.; Lowenthal, M. S.; Kilpatrick, L. E.; Sampson, M. L.; Telu, K. H.; Rudnick, P. A.; Mallard, W. G.; Bearden, D. W.; Schock, T. B.; Tchekhovskoi, D. V; et al. Metabolite Profiling of a NIST Standard Reference Material for Human Plasma (SRM 1950): GC-MS, LC-MS, NMR, and Clinical Laboratory Analyses, Libraries, and Web-Based Resources. Anal. Chem. 2013, 85 (24), 11725–11731. (131)  Ortiz-Villanueva, E.; Benavente, F.; Piña, B.; Sanz-Nebot, V.; Tauler, R.; Jaumot, J. Knowledge Integration Strategies for Untargeted Metabolomics Based on MCR-ALS Analysis of CE-MS and LC-MS Data. Anal. Chim. Acta 2017, 978, 10–23. (132)  Rabe, J. H.; Sammour, D. A.; Schulz, S.; Munteanu, B.; Ott, M.; Ochs, K.; Hohenberger, P.; Marx, A.; Platten, M.; Opitz, C. A.; et al. Fourier Transform Infrared Microscopy Enables Guidance of Automated Mass Spectrometry Imaging to Predefined Tissue Morphologies. Sci. Rep. 2018, 8 (1), 313. (133)  Dearing, T. I.; Thompson, W. J.; Rechsteiner, C. E.; Marquardt, B. J. Characterization of Crude Oil Products Using Data Fusion of Process Raman, Infrared, and Nuclear Magnetic Resonance (NMR) Spectra. Appl. Spectrosc. 2011, 65 (2), 181–186. (134)  Bitar, T.; Mavel, S.; Emond, P.; Nadal-Desbarats, L.; Lefèvre, A.; Mattar, H.; Soufia, M.; Blasco, H.; Vourc’h, P.; Hleihel, W.; et al. Identification of Metabolic Pathway 122  Disturbances Using Multimodal Metabolomics in Autistic Disorders in a Middle Eastern Population. J. Pharm. Biomed. Anal. 2018, 152, 57–65. (135)  Borràs, E.; Ferré, J.; Boqué, R.; Mestres, M.; Aceña, L.; Busto, O. Data Fusion Methodologies for Food and Beverage Authentication and Quality Assessment – A Review. Anal. Chim. Acta 2015, 891, 1–14. (136)  Razuc, M.; Fernández Band, B. S.; Garrido, M. Data Fusion Applied to the Photodegradation Study of Ciprofloxacin Using Hyphenated Detection Systems (UV–Vis and Fluorescence) and Multivariate Curve Resolution. Microchem. J. 2018, 138, 180–189. (137)  Kovalchik, K. A.; MacLennan, M. S.; Peru, K. M.; Ajaero, C.; McMartin, D. W.; Headley, J. V.; Chen, D. D. Y. Characterization of Dicarboxylic Naphthenic Acid Fraction Compounds Utilizing Amide Derivatization: Proof of Concept. Rapid Commun. Mass Spectrom. 2017, 31 (24). (138)  Mahaffey, A.; Dubé, M. Review of the Composition and Toxicity of Oil Sands Process-Affected Water. Environ. Rev. 2016, 25 (1), 1–18. (139)  Senning, A. No. Dictionary of Chemoetymology; Elsevier, 2007. (140)  IUPAC. IUPAC Compendium of Chemical Terminology (the “Gold Book”), 2nd ed.; McNaught, A. D., Wilkinson, A., Eds.; Blackwell Scientific Publications: Oxford, 1997. (141)  Nic, M.; Jirat, J.; Kosata, B.; Jenkins, A. D. IUPAC Compendium of Chemical Terminology (the “Gold Book”) - XML on-line corrected version. (142)  Colombel, M.; Soloway, M.; Akaza, H.; Böhle, A.; Palou, J.; Buckley, R.; Lamm, D.; Brausi, M.; Witjes, J. A.; Persad, R. Epidemiology, Staging, Grading, and Risk 123  Stratification of Bladder Cancer. Eur. Urol. Suppl. 2008, 7 (10), 618–626. (143)  Gleason, D. F. Urologic Pathology: The Prostate; Tannenbaum, M., Ed.; Lea & Febiger: Philadelphia, 1977.    124  Appendices Appendix A:   Naphthenic acids definitions and vocabulary  The evolution of the term “Naphthenic acids” The terms “naphthenic acids fraction compounds” and “oil sands process-affected waters” were not always the preferred terms to describe water samples that have come in contact with bitumen. There are cases where the same term was used to refer to different samples and where different terms were being used for the same sample. Because both naphthenic acids and oil sands process-affected waters are important in the engineering, chemical, environmental and social spheres, a number of terms have arisen to describe these samples. The term “oil sands process-affected waters” (OSPW) has gone through some changes and research can be found using multiple names derived from top-down definitions for the same material such as “oil sands process waters”, “oil sands tailings waters”, “bitumen affected water”, and “oil sands wastewater”. OSPW refers to the water in the tailings ponds and any other water that has come in contact with oil sands industrial processes. It is also important to note that the term “OSPW” has been used for at least 10 mixtures with very different top-down sample descriptions (i.e. highly likely to have different molecular compositions).138  The term “naphthenic acids” (NA) was originally used to describe the lightweight acidic compounds in the acid-extracted fraction from petroleum (top-down definition). Basic chemical research provided some bottom-up information: most of the components of that fraction were organic carboxylic acids, aliphatic, and had ring structures (the word “naphthenic” means “lightweight, aliphatic, cyclic hydrocarbon”). 139–141 Today, you can find “Naphthenic acids” in the CAS registry with a single submitted chemical structure, and a unique molecular weight, shown in Figure A.1. 125   Figure A.1 A chemical structure for “naphthenic acids” downloaded from PubChem. The molecule shown is a single, lightweight naphthenic acid and should not be interpreted as the molecular structure of all naphthenic acids. This molecule would now be considered a “classical naphthenic acid”. Image accessed 2018-01-12.  The chemical structure in Figure A.1 is, of course, a representative of the class of naphthenic acids and is not equivalent to the chemical structure of most naphthenic acids. Still, it conveys a picture that naphthenic acids are small, aliphatic cyclic monocarboxylic acids extracted from petroleum. The situation changed when “naphthenic acids” were also found in oil sands process-affected water (OSPW) and posed a threat to the health of many aquatic species. It turned out that there were molecules in the acid fraction of OSPW that did not follow the classical molecular definition of “naphthenic acids”: many molecules contained S and N heteroatoms, hydrocarbon scaffolds that did not exist abundantly in petroleum, and aromatic groups. As a result, OSPW naphthenic acids went through a series of name changes from “naphthenic acids” 126  to “naphthenic acid fraction compounds” (NAFC) to “oil sands tailings water acid extractable organics” (OSTWAEO), which has been shortened to “acid-extractable organics” (AEO). NAFC is the most commonly used term in naphthenic acids research. The class of aliphatic, cyclic monocarboxylic acids, without N or S heteroatoms, are now referred to as “classical naphthenic acids” or just “naphthenic acids” (NA) for short. Without a doubt, OSPW contains classical naphthenic acids, but the acid fraction of OSPW was found to contain a wider variety of functional groups than that described by the classical naphthenic acids definition, even though extractive procedures for naphthenic acids were the same for OSPW and petroleum. In the body of the thesis text, the term “NAFC” is used to describe naphthenic acids and associated compounds in the acid fraction, and “OSPW” is used to describe tailings pond water, from which NAFC are extracted. Occasionally, the term “naphthenic acids” (NA or NAs) is used to refer to the classical naphthenic acids.  “Top-down” and “bottom-up” definitions of complex mixtures A hallmark of chemical science is its ability to define substances according to the molecules, atoms, ions and even isotopes of which they are comprised (“bottom-up”). However, the information most readily available when dealing with samples of complex industrial and environmental mixtures is the information regarding the geochemical origins of the material and the provenance of its industrial handling. This unique information describes a sample according to how it was obtained and constitutes a “top-down definition” of the sample substance. Both “bottom-up” and “top-down” definitions of OSPW and NAFC can be found. An example of a top-down definition for naphthenic acids could be,  127  “Naphthenic acids are the molecules in the organic acid fraction extracted from petroleum using alkaline water or using alternating extractions of organic solvent and alkaline water.”  The only molecular information in this definition is that naphthenic acids are molecules, presumably acids. A top-down definition can have many levels of description, but includes the sample’s salient properties, the context in which it was found and the method by which it was obtained from something else. A commonly used name for the fraction obtained by alkaline extraction is “acid-extractable organics” (AEO). Top-down definitions are refined by the inclusion of new information from any intersecting sciences. In addition to a top-down approach to defining substances, there is the bottom-up approach: the “molecular definition”, which is the definition of a substance according to the molecular components of which it is comprised. Molecular definitions are developed based on evidence from chemical methods. As an example: “Naphthenic acids are the class of organic carboxylic acids containing one or more saturated rings with formula CnH2n+ZO2, where Z is an even integer less than 30 in magnitude, and components have mass between 100 and 500 Da.”  Molecular definitions of unknown samples are always progressing toward increasingly fine-grained definitions of molecular properties (e.g. “Monocarboxylic acids with less than 6 rings under 450 Da”). The molecular definition of sample components is refined by using analytical methods which can better describe molecular structures, or find new molecules, in a complex mixture. Whereas the names “acid-extractable organics” was used above to describe the acid 128  fraction derived from petroleum, the name “naphthenic acids” is an attempt to use molecular definitions to describe the same lightweight acid fraction of petroleum.  A sample’s top-down definition and its bottom-up molecular definition are connected by the fact that the bulk sample is comprised of molecules, ions and atoms: whatever events have happened to the bulk material in question necessarily have happened to the molecules which make up the bulk material. Although chemical analysis progresses toward a complete description of all molecules in a sample, a comprehensive definition of complex sample mixtures necessitates the inclusion of both the top-down and bottom-up definitions. This has a very practical repercussion in the oil sands industry and tailings water chemistry. The fact is that each oil company mines chemically unique samples of bitumen, using unique proportions of additive chemicals, implementing unique proprietary technologies to create chemically unique waste water that is held in company-owned tailings ponds. Therefore, each individual tailings pond water sample contains a different suite of chemical components which is changing over time due to fresh input, biodegradation and weathering. While the water in each tailings pond sample is called “OSPW”, the reality is that each sample of OSPW contains different chemical species and at different quantities, which justifies attempts in the research community to create unique “chemical fingerprints” of samples. Another important goal for the environmental assessment of NAs is their quantification in environmental waters. However, since NAs are a complex mixture of compounds, exact quantification of each NA species is currently not possible. Instead, the MS measurement of total NAs (i.e. “semi-quantification”) has become a useful strategy. 129  Appendix B:   Urine metabolomics supplementary material  Three-point calibration curves for targeted metabolites Three-point standard addition calibration curves for targeted amino acids in patient urine samples. R2 values are for averaged duplicates.  130    131   Figure B.0.1 Three-point standard addition calibration curves for proline, cystine, leucine, glutamic acid, kynurenine and sarcosine in patient urine samples.  Cystine Gaussian Regression  Figure B.0.2. Approximately linear decrease in the three-point calibration coefficient of determination (R2). Patient designation (A, B, D, E) is placed next to the blue diamond which corresponds to its data. Note that patient E’s creatinine level is actually 3.5 times greater than Patient B, but they are placed together since both R2 show very little correlation. This trend was only seen for patient cystine data.  132   Staging and Grading of Cancer Histopathology  From an analytical perspective, histopathology provides a type of spatio-temporal data, which can be correlated to diagnosis. A particular pathology will be divided into grades, each representing a sort of severity of disease. Staging tables are also compiled, describing the spread of a particular cancer. Here are provided some staging and grading tables for BCa and PCa. WHO 1973 (“Traditional” system)  Urothelial papilloma  Grade 1: well differentiated  Grade 2: moderately differentiated  Grade 3: poorly differentiated WHO 2004  Urothelial papilloma  PUNLMP  Low-grade papillary urothelial carcinoma  High-grade papillary urothelial carcinoma PUNLMP = papillary urothelial neoplasms of low malignant potential  Table B.0.1 The grading of BCa according to the WHO 1973, dubbed the “traditional” system, as well as the WHO 2004 system.142     133  T: Primary tumour TX  Primary tumour cannot be assessed T0  No evidence of primary tumour Ta  Non-invasive papillary carcinoma Tis  Carcinoma in situ: ‘‘flat tumour’’ T1  Tumour invades subepithelial connective tissue T2  Tumour invades muscle  T2a:   Tumour invades superficial muscle (inner half)  T2b:   Tumour invades deep muscle (outer half) T3  Tumour invades perivesical tissue:  T3a:   Microscopically  T3b:   Macroscopically T4: Tumour invades any of the following: prostate, uterus, vagina, pelvic wall, abdominal wall  T4a:  Tumour invades prostate, uterus, or vagina  T4b:  Tumour invades pelvic wall or abdominal wall N: Lymph nodes NX  Regional lymph nodes cannot be assessed N0  No regional lymph node metastasis N1  Metastasis in a single lymph node < 2 cm in greatest dimension N2  Metastasis in a single lymph node > 2 cm but not > 5 cm in greatest dimension, or multiple lymph nodes, none > 5 cm in greatest dimension N3  Metastasis in a lymph node > 5 cm in greatest dimension M: Distant metastasis MX  Distant metastasis cannot be assessed M0  No distant metastasis M1 Distant metastasis Table B.0.2 The staging of BCa according to the 2002 Tumour-node metastases classification system.142 Common stages include T0, Ta, and Tis.81 134  Pattern Margins of tumour areas Gland size Gland pattern Gland distribution Stromal invasion 1 Well defined Medium Single, separate, round Closely packed Minimal expansile 2 Less definite Medium Single, separate, rounded but more variable Spaced up to one gland diameter, average Mild, in larger stromal planes 3 Poorly defined Small, medium, or large Single, separate, more irregular Spaced more than one gland diameter, rarely packed Moderate, in larger or smaller stromal planes Or 3 Poorly defined Medium or large Rounded masses of cribriform or papillary epithelium Rounded masses with smooth, sharp edges Expansile masses 4 Ragged infiltrating Small Fused glandular masses or “hypernephroid” Fused in ragged masses Marked, through smaller planes 5 Ragged infiltrating Small Almost absent, few tiny glands or signet ring cells Ragged anaplastic masses of epithelium Severe, between stromal fibres or destructive Or 5 Poorly defined Small Few small lumina in rounded masses of solid epithelium; central necrosis? Rounded masses and cords with smooth, sharp edges Expansile masses  Table B.0.3 Gleason score (grade) for prostate cancer based on the VACURG data.82,143  135  As a corollary, the diagnostic power of this so-called “Gleason score” is enhanced when the presence of two grades is reported and added together. Thus, if grade 3 is present in more than 50% of the sample, and grade 4 is present in less than 50%, the score is reported as 3+4 = 7, where 7 is the corresponding Gleason score. Consequently, the Gleason score is usually reported on a scale of 1 to 10. This score has had better correlation to mortality and thus better diagnostic power.143             136  Appendix C:   Data processing scripts and computational supplement   In-house R scripts for processing data  All R scripts used in this section have been published online in the Zenodo repository at https://dx.doi.org/10.5281/zenodo.46867.107 The specific scripts used include the following: R script name Usage Arguments and notes mzUnpackBin2.R 2-D histogram binning of data Bins along retention time and m/z simultaneously either dividing vector into equal bin sizes, or by dividing vector into a specified number of bins. nagenmz.R Generating chemical formula assignments to m/z values Loops through the various combinations of atoms in the target molecule versatile.plot.R versatile.points.R versatile.plot.list.R Assigning the intensity cutoff for CE-ESI-MS data Cutoff is usually applied as a single value empirically determined using plots; can also be applied as a formula surfaces.R Assembling the 5 factor data used in the rough global optimization Requires data matrices (not given) RGOptMeOHevap.R Generating the global optimization plot Requires surfaces.R to be complete versatile.plot.R Generating plot used in Figure of report Requires data processed by package mzR functions peaks() and header() data2colours.R Generating colour palette for colour bar used in Figure of report   Table C.1 Scripts written in the R Statistical Computing language which were used in the work of Chapter 3:.  137   DFT calculation of optimized molecular structure Density functional theory B3LYP 6-31G(d) basis set was used to optimize the molecular structure of a N,N-DMEDA-derivatized acetate (as a model naphthenic acid). The input structure (left) was a linear conformer of the model NA. The optimized structure (right) shows the amine hydrogen pointing toward the carbonyl oxygen.  The N-C bond length between the dimethylamine moiety and the rest of the molecule is 1.524 Å, suggesting it would be more easily broken and lending evidence to the loss of 45 Da in ESI-MS.   138   Appendix D:   Script for enumerating parameter combinations in factor/principal components analysis  Testing 420 component/factor analysis criteria The following set of scripts, written in R, were used to loop through all 420 options available in the R package psych for the functions principal() and fa() which are principal components analysis and factor analysis, respectively, and rely on the package GPArotation. The input values include matrix factoring method, rotational criteria, and scores estimation for factor analyses. The data matrix is ‘em’. After generating all PCA and FA results, the estimated FA scores were compared to all other scores by covariance.  rotate<-c("Varimax","oblimin","promax","quartimax","quartimin","biquartimin","simplimax","equamax","varimin","geominT","geominQ","bentlerT","bentlerQ","bifactor") #"cluster" scores<-c("regression","Harman","Bartlett","Anderson","tenBerge") fm<-c("mle","minres","minrank","pa","wls","uls")  pcpf<-function(nfactors,rotate,fm,scores,fa=TRUE,maxit=maxit){ if(fa==TRUE){ falist<-list() for(i in 1:length(rotate)){ falist[[i]]<-list()    for(j in 1:length(fm)){ falist[[i]][[j]]<-list()     for(k in 1:length(scores)){ falist[[i]][[j]][[k]]<-fa(em,nfactors=nfactors,rotate=rotate[i],scores=scores[k],fm=fm[j],maxit=maxit) 139  #falist[[i]][[k]]<-principal(em,nfactors=nfactors,rotate=rotate[i],scores=T,method=scores[k])     } falist    } falist } falist} else {  falist<-list() for(i in 1:length(rotate)){ falist[[i]]<-list() # for(j in 1:length(fm)){ #  falist[[i]][[j]]<-list()    for(k in 1:length(scores)){ #falist[[i]][[j]][[k]]<-fa(em,nfactors=nfactors,rotate=rotate[i],scores=scores[k],fm=fm[j]) falist[[i]][[k]]<-principal(em,nfactors=nfactors,rotate=rotate[i],scores=T,method=scores[k],maxit=maxit)    } #falist #  } falist } falist} falist }  falist1<-unlist(unlist(falist,recursive=F),recursive=F) rnames<-apply(cbind(rep(rotate,each=30),unlist(lapply(falist1,function(x) x$fm)),rep(scores,14*6)),1,function(x) paste0(x,collapse="")) names(falist1)<-rnames falist1df<-data.frame(lapply(falist1,function(x) x$scores)) 140  scorfa<-round(cor(falist1df),2) diffident<-cbind(rownames(cor(falist1df))[which(cor(falist1df)==1,arr.in=T)[,1]],colnames(cor(falist1df))[which(cor(falist1df)==1,arr.in=T)[,2]])[(rownames(cor(falist1df))[which(cor(falist1df)==1,arr.in=T)[,1]]!=colnames(cor(falist1df))[which(cor(falist1df)==1,arr.in=T)[,2]]),] scorall<-round(cor(cbind(falist1df,pclist1df)),2)  sgorr<-function(scorfa,val){ #finds score correlations of 1 in a score correlation matrix a<-cbind(rownames(scorfa)[which(scorfa==val,arr.in=T)[,1]],colnames(scorfa)[which(scorfa==val,arr.in=T)[,2]])[(rownames(scorfa)[which(scorfa==val,arr.in=T)[,1]]!=colnames(scorfa)[which(scorfa==val,arr.in=T)[,2]]),] a } ########################################################## cn<-unique(gsub("\\.[A-Z]+[1-3]","",colnames(scorall))) rn<-unique(gsub("\\.[A-Z]+[1-3]","",rownames(scorall))) ln<-apply(expand.grid(rn,cn),1,function(x) paste0(x,collapse="-x-")) names(cv)<-ln ########################################################## #table of factoring/component methods (fm) that have correlation 1 to #PCA component regex: "C" (=TC,RC) in total scores correlation matrix #absolute values returned, but percentages would be useful #e.g. 35% of FA with PA method have 1-to-1 correlation to PCA component scores. # table(unlist(regmatches(sgorr(scorall,1)[grep("C",sgorr(scorall,1)[,2]),1], gregexpr("[A-Z]+[0-9]",sgorr(scorall,1)[grep("C",sgorr(scorall,1)[,2]),1]))))  #Divide by relevant totals to obtain percentage 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0371931/manifest

Comment

Related Items