Development of Functionalized Paper-Based Sample Collection and Direct Mass Spectrometry Analysis Platforms by Deidre Erin Damon

Development of Functionalized Paper-Based Sample Collection and Direct

Mass Spectrometry Analysis Platforms

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree of Doctor of

Philosophy in the Graduate School of The Ohio State University

By Deidre Erin Damon

Graduate Program in Chemistry

The Ohio State University

2019

Dissertation Committee:

Dr. Abraham K. Badu-Tawiah, Advisor

Dr. Vicki H. Wysocki

Dr. James V. Coe

Copyrighted by

Deidre Erin Damon

2019

Abstract

The goal of this dissertation is to develop a paper-based mass spectrometry (MS) ionization method for detection of small molecules from biological fluids and water samples. In biofluids, highly sensitive analytical methods are required to monitor small biologically active molecules during treatment. When analyzing both biofluids and industrial water samples, rapid and inexpensive detection platforms streamline screening for harmful concentrations of chemicals. Paper spray ionization shortens analysis time to less than one minute. Current analytical detection methods for these molecules include gas chromatography (GC)-MS, which is useful for creating databases for small molecules. However, GC-MS often requires extensive sample preparation and analyte derivatization prior to analysis. Liquid chromatography (LC)-MS offers better sensitivity without the need for derivatization. However, LC-MS is also limited by additional sample preparation/handling procedures. In particular, matrix components must be removed prior to any chromatographic work, which can be time consuming. For these reasons, immunoassay screening is still the most commonly used detection method for these small molecules despite their well-recognized limitations with regards to sensitivity and selectivity. Alternatively, paper spray (PS) ionization MS is capable of direct analysis of small biological sample volumes. However, PS often suffers from low sensitivity due to inefficient extraction. The second chapter in this document utilizes treated hydrophobic paper substrates prepared by a gas-phase silanization reaction. Hydrophobic PS utilizes online liquid/liquid extraction from a drop of biological fluid without drying steps or sample pretreatment, which is applicable in point-of-care analyses. Preliminary experiments demonstrated high sensitivity in the analysis of drugs cocaine, benzoylecgonine, methamphetamine, and

ii amphetamine in 4 μL of raw blood, serum, and urine. When the sample is dried, a 3- dimensional dried spheroid is formed, as opposed to a dried blood spot. This spheroid decreases interaction between the bulk of the blood sample and ambient air, decreasing oxidative stress, increasing the lifetime of the analyte in the sample. Experiments show that labile organic compounds diazepam and cocaine have increased stability up to 28 days, where sample signal decreases by 90% within a day when stored in dried blood spots. The third chapter discusses the use of wax printing on paper. Wax printing creates microfluidic channels in which the designer can control solvent flow. Through this solvent control, spray time was increased from 1.5 minutes to 10 minutes. Additionally, because the solvent acts as the charge carrier, manipulation of the electric field decreased the required voltage applied to the paper triangle from 3-5 kV to 0.5-1 kV. The fourth chapter outlines the detection of low concentration corrosion inhibitor Duomeen O and molluscicide metaldehyde in industrial and environmental water samples. On-site analysis of Duomeen O would facilitate the maintaining of appropriate levels to prevent corrosion in water tube boiler plants. Because metaldehyde is readily soluble in water, runoff during periods of heavy rainfall ultimately introduces metaldehyde into drinking water. Paper spray ionization has been demonstrated as a viable technique for rapid screening and quantification of these molecules without sample preconcentration with detection limits below the threshold set by the World Health Organization.

iii

Dedication

This dissertation is dedicated to my father, Darrel Damon, for his advice in research and life.

Acknowledgements

I would like to thank my advisor, Dr. Abraham Badu-Tawiah, for his support and guidance, as well as the opportunities he gave me to develop my scientific potential. I also want to thank him for his unyielding optimism and ability to see much broader impacts that I was not capable of noticing. Additionally, I want to thank my group members, including Colbert Miller, Dmytro Kulyk, Qiongqiong Wan, and Suming Chen for joining me in the founding of the Badu group and giving me advice, support, and distraction necessary to survive graduate school while maintaining sanity. I also want to thank the undergraduate student researchers who worked with me, including Yosef Maher, Mengzhen Yin, Jill Baker, Christian Tanny, and Danyelle Baker for helping with research and persevering when experiments never seemed to work. Further, I would like to thank the rest of my group for giving me support and reprieve from daily research life, including Savithra Jayaraj, Sierra Jackson, Tatiana Velez, and Benji Frey. I want to also thank Fred Jjunju, Simon Maher, and Steven Taylor for their advice and partnership in research that is discussed in this document. Finally, I would like to thank my father and step mother, Darrel and Paula Damon, and my mother Cindy Damon for their unceasing support and advice through graduate school and beyond.

Vita

2009 ‒ 2014 ……………………………..… B.Sc. Chemistry, Biochemistry Concentration, Department of Chemistry, University of Wisconsin – Parkside

2014 ‒ 2019 …………………………..…… Graduate Associate, Department of Chemistry and Biochemistry, The Ohio State University

Publications Deidre E. Damon, Mengzhen Yin, Christian J. Tanny, Yosef S. Maher, Stephanie Oyoloa- Reynoso, Barry L. Smith, Simon Maher, Martin M. Thuo, Abraham K. Badu-Tawiah, “Dried Blood Spheroids for Dry-state Room Temperature Stabilization of Microliter Blood Samples” Analytical Chemistry 2018, 90 (15), 9353-9358.

Simon Maher, Fred P. M. Jjunju, Deidre E. Damon, Yosef S. Maher, Safaraz U. Syed, Ron M. A. Heeren, Stephen Taylor, Iain S. Young and Abraham K. Badu-Tawiah, “Direct Analysis and Quantification of Metaldehyde in Water using Reactive Paper Spray Mass Spectrometry” Scientific Reports 2016, 6, 35643.

Deidre E. Damon, Yosef S. Maher, M. Yin, Fred P. M. Jjunju, Simon Maher, Stephen Taylor, and A. K. Badu-Tawiah, “2D wax-printed paper substrates with extended solvent supply capabilities allow enhanced ion signal in paper spray ionization” Analyst 2016,141, 3866-3873.

Deidre E. Damon, Kathryn M. Davis, Camila R. Moreira, Patricia Capone, Riley Cruttenden, and Abraham K. Badu-Tawiah, “Direct Biofluid Analysis using Hydrophobic Paper Spray Mass Spectrometry” Analytical Chemistry 2016, 88 (3), 1878–1884.

Fred P. M. Jjunju, Simon Maher, Deidre E. Damon, Dick. R. Barrett, S. U. Syed, R. M. A. Heere, Stephen Taylor, Abraham K. Badu-Tawiah, “Screening and Quantification of Aliphatic Primary Alkyl Corrosion Inhibitor Amines in Water Samples by Paper Spray Mass Spectrometry” Analytical Chemistry 2015, 88 (2), 1391–1400.

Fields of Study

Major Field: Chemistry

Analytical Chemistry

vii

Table of Contents

Abstract ...... ii

Dedication ...... iv

Acknowledgements ...... v

Vita ...... vi

Table of Contents ...... viii

List of Tables ...... xxv

List of Equations ...... xxvi

List of Abbreviations ...... xxvii

Chapter 1. Introduction and Background ...... 1

1.1 Overview ...... 1

1.2 Dried Blood Spots on Paper as a Biofluid Collection Platform...... 2

1.3 Ionization Methods for Mass Spectrometry ...... 3 1.3.1 Electrospray Ionization ...... 4 1.3.2 Ambient Ionization ...... 6 1.3.2.1 Paper Spray ...... 7

1.4 Linear Ion Trap Mass Spectrometer Detection ...... 8

Chapter 2. Hydrophobic Paper Spray Ionization ...... 11 viii

2.1 Introduction: Dried Blood Spots ...... 11 2.1.1 Paper Spray Strategies for Biofluid Analysis ...... 12 2.1.2 Dried Blood Spheroids83 ...... 13

2.2 Materials and Methods ...... 16

2.3 Hydrophobic Paper Preparation for Paper Spray Analysis ...... 18

2.4 Characterization of Paper Strips for Paper Spray Analysis ...... 22

2.5 Analyte Stability in Dried Blood Spots and Spheroids ...... 26

2.6 Dried Urine Sample Analysis ...... 33

2.7 Quantification of Drugs of Abuse in Dried Blood Spheroids ...... 35

2.8 Analysis of Illicit Drugs in Fresh Blood ...... 43

2.9 Analysis of Alanine Transaminase Enzyme Activity ...... 48

2.10 Direct Analysis of Nonbiological Samples ...... 51

2.11 Surface Energy Analysis ...... 54

2.12 Ionization Mechanism ...... 63

2.13 Summary ...... 65

Chapter 3. 2D Wax-Printed Paper Spray Ionization ...... 67

3.1 Introduction ...... 67

3.2 Materials and Methods ...... 69

3.3 Wax Pattern Optimization...... 70

3.4 Characterization of Spray from Wax-Printed Paper Substrate ...... 76 ix

3.5 Analysis of Illicit Drugs ...... 78

3.6 Analysis of Corrosion Inhibitors and Pesticides in Water ...... 84

3.7 Summary ...... 88

Chapter 4. Water System Analysis with Paper Spray Ionization ...... 91

4.1 Introduction ...... 91 4.1.1 Amine Analysis Introduction ...... 93 4.1.2 Metaldehyde Analysis Introduction ...... 96

4.2 Materials and Methods ...... 99

4.3 Paper Spray Analysis and Characterization of Duomeen O and Metaldehyde ...... 102 4.3.1 Paper Spray Mass Spectrometry and Characterization of Duomeen O Using Positive Ion Mode ...... 103 4.3.2 Structure Characterization and Confirmation of Duomeen O ...... 105 4.3.3 Analysis of Metaldehyde using Paper Spray Mass Spectrometry ...... 109

4.4 Reactive Paper Spray Mass Spectrometry ...... 112 4.4.1 Duomeen O Detection Using Schiff-Base Reaction with Acetone ...... 112 4.4.2 Characterization and Identifications of Protonated Metaldehyde Molecular Ion Species using Formic Acid ...... 115

4.5 Analysis of Environmental and Industrial Samples ...... 122 4.5.1 Paper Spray Mass Spectrometry Analysis of Duomeen O in a Mixture of Polyamine Corrosion Inhibitors ...... 122 4.5.2 Direct Metaldehyde Quantitation in Environmental Water Samples Using Paper Spray Mass Spectrometry ...... 125

4.6 Metaldehyde Fragmentation Pathway Discussion ...... 127

4.7 Summary ...... 130 x

Chapter 5. Conclusions ...... 133

5.1 Summary ...... 133

5.2 Future Directions ...... 134

References ...... 137

Appendix A. List of Videos ...... 157

List of Figures

Figure 1.1. Electrospray ionization schematic...... 5 Figure 1.2. Schematic for paper spray ionization...... 8 Figure 1.3. Schematic of a Thermo Scientific LTQ Velos Pro mass spectrometer...... 9 Figure 1.4. Schematic of a linear ion trap...... 10 Figure 2.1. Experimental setup using (A) paper triangles and (B) paper rectangles. (C) Image showing a 4 L dried blood spot/spheroid on an untreated (left) and treated (right) paper substrates, including the front (top) and back (bottom). Scale bars show 0.5 mm. (D) Workflow of direct on-surface dried blood analysis...... 14 Figure 2.2. 4 µL blood was pipetted onto untreated and 30 minute treated paper. The following areas of wetting were measuring using ImageJ...... 14 Figure 2.3. (A) Functionalization of paper by modification of fiber surface using trichloro(3,3,3-trifluoropropyl) silane vapor to create a hydrophobic oligomeric silylated layer. Gaseous HCl released as a byproduct of the reaction is removed by vacuum in situ. (B) Photograph showing interaction of a dye solution on untreated versus treated (hydrophobic) paper. (C) Hydrophobic paper spray in which the biological sample (e.g., blood) prevents the spreading of the organic solvent beyond the point where the sample was deposited...... 15 Figure 2.4. Functionalization of paper by modification of surfaces hydroxyl groups using chlorosilane vapor to create a hydrophobic oligomeric silated layer. Gaseous HCl released as a byproduct of the reaction is removed by vacuum in situ...... 16 Figure 2.5. Effect of silanization on stiffness of filter paper and chromatography paper at 50% relative humidity salinization...... 19 103,104 Figure 2.6. Zisman plot showing surface energies of CH2CH2CF3-funsctionalized paper. Ɵ is measured contact angles. A series of selected solvents was used (surface

xii tension, mN/m) ranging from hexane (18.4), heptane (20.1), octane (21.6), decane (23.8), dodecane (25.4), hexadecane (27.5), nitrobenzene (43.9), ethylene glycol (47.7), glycerol (63.3) to water (72.2)...... 20 Figure 2.7. Stability of (A) cocaine in dried blood, (B) neat dried diazepam prepared in water, and (C) diazepam in dried blood. Both dried blood spots (untreated) and spheroids (treated) samples were stored under ambient conditions at 25°C. Internal standard was spiked into the spray solvent to normalize between samples and days. Three replicates were used. Error bars show one standard deviation...... 21 Figure 2.8. (Columns) show spectra grouped by paper treatments, and (Rows) show spectra grouped by drugs fragmented. (A-D) Fragmentation of (A) amphetamine, (B) methamphetamine, (C) benzoylecgonine, and (D) cocaine on untreated paper strips. (E-H) Fragmentation of (E) amphetamine, (F) methamphetamine, (G) benzoylecgonine, and (H) cocaine on 2 hour treated paper strips. Characteristic fragments include: cocaine (304→182), benzoylecgonine (290→168), methamphetamine (150→119) and amphetamine (136→119)...... 22 Figure 2.9. (Columns) show spectra grouped by paper treatments, and (Rows) show spectra grouped by drugs fragmented. (A-D) Fragmentation of (A) amphetamine, (B) methamphetamine, (C) benzoylecgonine, and (D) cocaine on paper strips treated for 30 minutes. (E-H) Fragmentation of (E) amphetamine, (F) methamphetamine, (G) benzoylecgonine, and (H) cocaine on paper strips treated for 2 hours. Characteristic fragments include: cocaine (304→182), benzoylecgonine (290→168), methamphetamine (150→119) and amphetamine (136→119)...... 23 Figure 2.10. (A) Shows setup of a paper strip in front of the mass spectrometer with a dried blood spot immobilized on the surface. (B and C) Images from a Watec camera showing a closer look at the loose paper fibers on the edge of the paper strip. (B) is a side on view and (C) is a top-town view. Fibers measure to be approximately 0.04 mm in diameter. Scale bars show 1 mm...... 24 Figure 2.11. Total ion chromatogram of paper strip with alternating 3 kV and 0 kV applied. Sample is a dried blood spot on 2 hour treated paper with 20 L ethyl acetate applied. . 26

xiii

Figure 2.12. Offline analysis of 2 g/mL cocaine in dried blood spot on untreated paper and dried blood spheroid on 30 minute and 2 hour treated paper. Sample was spotted and stored for 1 and 2 days at 25 ͦ C. The samples were then extracted in 500 ng/mL D3 cocaine in ethyl acetate for 30 minutes in a sonicator. Extract was then nanosprayed, and m/z 304 (cocaine) and 290 (benzoylecgonine, possible cocaine degradation product) were fragmented...... 28 Figure 2.13. Stability of benzoylecgonine in dried blood spots (untreated) and spheroids (treated) stored in ambient conditions at 25° C. Internal standard was spiked into the spray solvent to normalize between samples and days. Error bars show one standard deviation...... 29 Figure 2.14. Stability of 2 g/mL cocaine and benzoylecgonine in dried blood spots/spheroids on untreated, 30 minute treated paper, and 2 hour treated paper. Samples were stored in a desiccator for 15 days and then analyzed with 5 kV and 10 L ethyl acetate containing 500 ng/mL deuterated internal standard to normalize between samples and between days. Error bars show one standard deviation...... 30

Figure 2.15. (A) CID of neat diazepam, m/z 285. (B) CID of O2 adduct of diazepam (m/z

317) in water immediately after depositing on a paper triangle. (C) CID of O2 adduct of diazepam in water 4 days after depositing on a paper triangle...... 32 Figure 2.16. Cut-away view of both geometries through their respective centers, illustrating the heat transfer in a DBS compared to spheroid based on their geometric properties after 18 seconds...... 33 Figure 2.17. Heat transfer transient simulation analysis. Both blood storage geometries (DBS versus spheroid) had an initial temperature of 30 °C and were subjected to a constant ambient air temperature of 40 °C. Temperature is measured at the geometric center for each case ...... 33 Figure 2.18. Urine samples spiked to give a final concentration of (A) 0.96 ng/mL (B) 3.9 ng/mL (C) 10 ng/mL and (D) 50 ng/mL amphetamine. Aliquots of 4 µL were dried on the paper surface and then analyzed with ethyl acetate...... 34

xiv

Figure 2.19. Pictures of (A) front and (B) back of untreated and treated paper with 4 L whole blood dried for 24 hours. Time listed is the amount of time gas phase silane is allowed to react with the paper surface...... 36 Figure 2.20. Extraction from dried blood spots with 20 L ethyl acetate. Absolute intensity of 500 ng/mL amphetamine, methamphetamine, cocaine, and benzoylecgonine on the surface of paper triangles. Characteristic fragments of cocaine (304→182), benzoylecgonine (290→168), methamphetamine (150→119) and amphetamine (136→119) were used for quantification. These triangles were untreated and treated for 5, 30, 120, 240, 720, and 1440 minutes with silane. Error bars show one standard deviation...... 36 Figure 2.21. Optimization of treatment time of paper using common illicit drugs and extraction from dried blood spots with 20 L acetonitrile. Absolute intensity of 500 ng/mL amphetamine, meth-amphetamine, cocaine, and benzoylecgonine on the surface of paper triangles. Quantification of characteristic fragments of cocaine (304→182), benzoylecgonine (290→168), methamphetamine (150→119) and amphetamine (136→119) was performed. These triangles were untreated and treated for 5, 30, 120, 240, 720, and 1440 minutes with silane. Error bars show one standard deviation...... 37 Figure 2.22. (A, C, E) Calibration of cocaine ranging from 10-500 ng/mL in dried blood spots, and (B, D, F) representative mass spectra of fragmentation of cocaine with a concentration of 10 ng/mL on untreated paper (A and B), paper treated for 30 minutes (C and D), and paper treated for 2 hours (E and F). Mass spectra show the increased signal to noise of cocaine on paper treated for 30 minutes when compared to the untreated and 2 hour treated paper, which was expected, as shown by the optimization in Figure 2.20. Error bars show one standard deviation of trials performed in triplicate...... 39 Figure 2.23. (A, C, E) Calibrations of benzoylecgonine in dried blood on (A) untreated paper triangles, (C) 30 minute treated paper triangles, and (E) 2 hour treated paper triangles. Error bars are one standard deviation. (B, D, F) sample MS/MS spectra from the respective paper treatments at 10 ng/mL concentration of benzoylecgonine. Error bars show one standard deviation of trials performed in triplicate...... 40

Figure 2.24. Calibrations of methamphetamine in dried blood on (A) untreated paper triangles, (B) 30 minute treated paper triangles, and (C) 2 hour treated paper triangles. Error bars show one standard deviation of trials performed in triplicate...... 41 Figure 2.25. (A, C, E) Calibrations of amphetamine in dried blood on (A) untreated paper triangles, (C) 30 minute treated paper triangles, and (E) 2 hour treated paper triangles. Error bars are one standard deviation. (B, D, F) sample MS/MS spectra from the respective paper treatments at 10 ng/mL concentration of amphetamine. Error bars show one standard deviation of trials performed in triplicate...... 42 Figure 2.26. Tandem MS analysis of drugs extracted from urine and blood. Representative product ion spectra illustrate (A) the methamphetamine transitions 150 → 119 (primary product ion) as well as 150 → 91 (benzylic cation) and (B) the benzoylecgonine transitions + 290 → 168 (primary product ion) as well as 290 → 272 ([M+H−H2O] ). (C) Effect of the number of liquid/liquid extraction cycles on ion signal. MS/MS product ion intensities were monitored for cocaine (m/z 304 → 182), amphetamine (m/z 136 → 119), and benzoylecgonine (m/z 290 →168), each at 125 ng/mL spiked into urine (6 µL sample volume was used). MS/MS product ion intensities for corresponding internal standards at 50 ng/mL were also monitored: cocaine-d3, amphetamine-d5 and benzoylecgonine-d3. 44 Figure 2.27. Quantitative analysis of (A) undiluted human serum spiked with methamphetamine, and its internal standard methamphetamine-d5 (50 ng/mL) and (B) dried blood spot spiked with cocaine, and it internal standard cocaine-d3. Analyte concentration of 0.24 – 500 ng/mL was used. Error bars represent the standard deviation of analyses for three replicates with independent hydrophobic paper triangles. Ethyl acetate (10 µL for wet samples and 20 µL for dried samples) was used as the ex-traction/spray solvent...... 45 Figure 2.28. Photographs showing various stages during electrostatic-spray ionization from dry hydrophobic paper: (A) voltage off, (B) on-set of applied voltage, (C) on-set of spray, (D) stable spray formed just before 0.02 seconds, (E) stable spray ...... 46 Figure 2.29. Signal time compared to sample volume of methamphetamine (125 ng/mL) spiked in serum. All samples were extracted with 10 µL of ethyl acetate. Sample sizes of

xvi greater than 4 µL showed a significant increase in analysis lifetime. Inset: Spectrum obtained using 1 µL of human serum spiked with methamphetamine (125 ng/mL)...... 48 Figure 2.30. (A) Reaction scheme for the enzymatic conversion of alanine into pyruvate catalyzed by alanine transaminase (ALT). Tandem MS of pyruvate (m/z 87) was monitored with representative spectra shown for (B) control raw blood sample without spiked ALT and (C) 400 U/L ALT spiked in blood. (D) Quantitative analysis using intensity of m/z 43 and 41 fragment ions over enzyme concentration of 150 – 400 U/L. Error bars represent the standard deviation of analyses for three replicates with independent hydrophobic paper triangles and three different assays. Ethyl acetate (10 µL) was used as the extraction/spray solvent, with 3.5 kV spray voltage...... 49 Figure 2.31. (A) Dry hydrophobic paper spray analysis of 5 µg/mL methamphetamine prepared in methanol/water (50/50) solution is comparted with (B) mass spectrum recorded using the traditional wet paper spray...... 50 Figure 2.32. The electrostatic-spray generated from a 4 µL solution of 5 µg/mL methamphetamine (full MS) in MeOH/H2O (50/50) using (A) a wetted paper triangle charged at 3 kV lasted for 30 seconds compared with (B) electrospray time of 2 minutes from a dry hydrophobic paper triangle. Representative mass spectra recorded for both experiments (electrostatic-spray from dry hydrophobic paper, and electrospray from wet hydrophilic paper) are as shown in Figure 2.31A and B, respectively ...... 51 Figure 2.33. Analysis of denatured myoglobin (25 µM in 0.5% acetic acid in methanol/water, (50/50) solution) using: (A) traditional untreated paper spray, (B) dry hydrophobic (treated) paper spray, and (C) nanospray. Spray voltage was 3 kV for both of the paper spray experiments, and 1.8 kV for nanospray ...... 53 Figure 2.34. Extracted ion chromatogram for amphetamine (m/z 150) showing the dependence of ion signal on applied voltage...... 53 Figure 2.35. Effect of electrostatic-spray voltage on signal intensity. Fragmentation of cocaine (125 ng/mL) in MeOH/H2O (50/50) was monitored in MS/MS experiment using the major fragment ion at m/z 182. Superimposed MS/MS mass spectra are plotted in the mass range m/z 180 – 184...... 54

xvii

Figure 2.36. Contact angle of DI water deposited onto filter paper with varying treatment times of vapor phase silane. Error bars show one standard deviation...... 55 Figure 2.37. (A) Observation of ion intensity varying with the change of surface tension of

ACN/H2O spray solvents (Table 2.2). Peak surface tensions are used as values for y-axis in plot B. (B) Calibration of cellulose acetate and polycarbonate, with treated and untreated paper projected onto the line. The determined surface energies of paper substrates are provided in Table 2.3...... 57 Figure 2.38. Data found on Figure 2.37A fitted with Equation 5. Fitting parameters were found to be: a = 210.3 m2/mN2, b =62.9 mN/m, c = 2497.1 mN2/m2 ...... 60 Figure 2.39. A Mathematica plot of Equation 5 using parameters found for fitting in Figure 2.38...... 61 Figure 2.40. Acetonitrile/water droplets of varying ratios (see Table 2.2) resting on a paper strip treated for 2 hours when 5 kV is applied. (A) Droplet consists of solvent 7 (pure acetonitrile, surface tension 29 mN/m). (B) Consists of solvent 5 (surface tension 38 mN/m). (C) Consist of solvent 4 (surface tension 41 mN/m). (D) Consists of solvent 1 (surface tension 62 mN/m)...... 62 Figure 2.41. (A) Cellulose, (B) Polycarbonate, (C) Cellulose Acetate, (D) Polyacrylonitrile...... 63 Figure 3.1. Procedure for preparing wax-printed paper...... 70 Figure 3.2. (A) Wax-printed micro-fluid channels/patterns tested (B) Comparison of absolute intensity of major fragment ions. I = absolute fragment ion intensities ...... 72 Figure 3.3. Charge density (V/mm2) is shown to increases as the area available to solvent decreases...... 72 Figure 3.4. Detail geometrical differences between all wax-printed patterns is provided. All parameters are given in millimeters (mm). Overall geometry of all paper triangles are approximately measured 9 mm base and 16 mm height...... 74 Figure 3.5. Comparison of s2, s3, and t0 (waxless) paper triangles as a function of voltage using (A) amphetamine, (B) cocaine, and (C) methamphetamine diluted in water at 250 ng/mL. Error bars show standard deviation for three replicates. I = absolute product ion intensities...... 75 xviii

Figure 3.6. Selected ion (m/z 304) chromatogram (XIC) of 100 ng per mL cocaine solution; 4 μL sample was dried onto s3 wax-printed paper and sprayed with increasing volumes of

MeOH/H2O (1:1, v/v) at 3 kV. Arrows show where each signal ceased. Inset (i): 4 μL of 100 ng per mL cocaine solution was dried onto an un-waxed paper triangle and sprayed with 20 μL of MeOH/H2O (1:1, v/v). Spray time was approximately 1.5 minutes (0.2–1.7 minutes). Inset (ii): signal lifetime varies with spray solvent volume applied to the paper triangle. Un-waxed paper signal lifetime does not increase after approximately 7 μL of solvent, but wax-printed paper signal lifetime increases to ∼10 minutes after 20 μL is added to the triangle. Inset (iii): zoomed-in XIC in 0.2–1.3 minute time range when using 10 μL spray solvent...... 76 Figure 3.7. Total ion chromatogram recorded after spraying of 1 ppm 6-methoxy-1,2,3,4- tetrahydroquinoline with 10 L 4:1 MeOH/H2O spray solvent at 3 kV. Inserts shown mass spectra taken from (i) the beginning of spray lifetime, (ii) stable spray region, and (iii) increased spray current region near the end of spray lifetime. No significant difference is seen in the mass spectra...... 78 Figure 3.8. Calibration of methamphetamine standard solutions (3–250 ng mL−1) analyzed with MeOH/H2O (1:1, v/v) solution using (A) 1 kV and (B) 0.5 kV spray voltages. Error bars show one standard deviation for three replicates. Representative spectra of methamphetamine fragmentation used for quantification are shown for (C) 1 kV and (D) 0.5 kV. RI = relative intensity; A/IS = ratio of analyte-to-internal standard signal. Internal standard used for methamphetamine was methamphetamine d5 with MS/MS transition m/z 155 → 123...... 79 Figure 3.9. Samples of 4 μL fresh urine spiked with 3 – 500 ng/mL drug, sprayed with 10 μL of 100% acetonitrile on s3 wax paper. Drugs were quantified with absolute intensity of fragmentation (A) divided by their respective internal standards (IS) of (A) cocaine (m/z

304 → 182) and IS d3 (m/z 307 → 185), (B) amphetamine (m/z 136 → 119) and IS d5 (m/z

141 → 123), (C) benzoylecgonine (m/z 290 → 168), and IS d3 (m/z 293 → 171) (D) methamphetamine (m/z 150 →119) and IS d5 (m/z 155 → 123)...... 81

xix

304 → 182) and IS d3 (m/z 307 → 185), (B) amphetamine (m/z 136 → 119) and IS d5 (m/z

141 → 123), (C) benzoylecgonine (m/z 290 → 168), and IS d3 (m/z 293 → 171) (D) methamphetamine (m/z 150 →119) and IS d5 (m/z 155 → 123)...... 82 Figure 3.11. Samples of 4 μL dried urine spiked with 3 – 500 ng/mL drug, sprayed with 20 μL of 100% acetonitrile on un-waxed paper. Drugs were quantified with absolute intensity of fragmentation (A) divided by their respective internal standards (IS) of (A) cocaine (m/z

304 → 182) and IS d3 (m/z 307 → 185), (B) methamphetamine (m/z 150 →119) and IS d5

(m/z 155 → 123), and (C) benzoylecgonine (m/z 290 → 168), and IS d3 (m/z 293 → 171)...... 83 Figure 3.12. 500 ng/mL cocaine sprayed with 20 L and 10 L 100% acetonitrile with 3 kV spray voltage on un-waxed paper. When 10 L spray solvent was used, signal lasted approximately 15 seconds, which was determined to be inadequate for more than one MS/MS observation. When the spray solvent volume was doubled, the signal lifetime was increased to nearly 1 minute, which was more suitable for calibration, as shown in Figure 3.11 above...... 84 Figure 3.13. Calibration of (A) Duomeen sprayed at 3 kV, (B) Duomeen at 1 kV, (C) metaldehyde at 3 kV, and (D) metaldehyde at 1 kV in water samples.All samples were sprayed with 4:1 MeOH/H2O. AI = absolute fragment ion (m/z 308) intensity, A/IS = ratio of analyte-to-internal standard signal. Internal standard used for metaldehyde was atrazine with MS/MS transition m/z 221 → 179...... 85 Figure 3.14. Analysis of water samples for detection of duomeen. Samples were taken from (A) pre-treatment, and (B) post-treatment stages in the boiler system cycle. Duomeen (m/z 325) was only detected in post-treatment sample, as expected...... 85 Figure 3.15. Analysis of water samples containing metaldehyde, sprayed at (A) analysis of 300 ng/mL sprayed at 1 kV, fragmentation m/z 199 [M+Na]+, and (B) analysis of 150 ng/mL sprayed at 3 kV, fragmentation m/z 177 [M+H]+. Intensities of fragment ions at m/z

111 and 149 were used to construct calibration curves found in Figure 3.13 C and D, respectively...... 86 Figure 3.16. Depiction of the effect of spray voltage on metaldehyde ionization. Higher voltage (3 kV; green, right) generates high proton abundance at wax-printed paper tip producing protonated metaldehyde [M+H]+ ions with unique fragmentation pathway. Alternatively, at low voltage (1 kV; blue, left), the ionization process is dominated by sodium adduction forming [M+Na]+ ions at m/z 199. MS inlet was grounded...... 87 Figure 3.17. Comparison of PS-MS mass spectra for metaldehyde recorded at different pHs of (A) 7, and (B) 3. Spectrum in B was recorded after 5 min of adding acid. Inserts show product ion mass spectra for [M+Na]+ at m/z 199 and for [M+H]+ at m/z 177 achieved using collision induced dissociation...... 88 Figure 4.1. Schematic of the paper spray mass spectrometry experimental setup used for rapid detection of metaldehyde in water samples...... 98 Figure 4.2. Schematic of the paper spray mass spectrometry experimental setup used in rapid screening of Duomeen O in the boiler system water samples...... 101 Figure 4.3. Positive ion mode paper spray mass spectrum for Duomeen O corrosion inhibitor model compound analyzed using a benchtop ion trap mass spectrometer. Absolute amounts of analyte were spotted onto filter paper and ionized in the open air by application of an electric potential, 2 μL, viz., 10 ppb: (A) Duomeen O (MW 324) in methanol solution and (B) exact mass measurement of Duomeen O. Insert (i) shows the isotopic distribution of the Duomeen O protonated molecular ion [M+H]+ at m/z 325, and inserts (ii) and (iii) show the MS/MS CID data for the selected ions. Insert (iv) shows the corresponding exact mass MS/MS CID data...... 104 Figure 4.4. Positive ion mode paper spray mass spectrum for amine corrosion inhibitor model compound analyzed using a bench-top ion trap mass spectrometer. Absolute amounts of analyte spotted onto a filter paper and ionized in open air by application of an electric potential, 2 μL, viz 10 ppb with methanol spray solvent; (A) morpholine (Mw 87), (B) cyclohexylamine (MW 99), (C) diethyl amino ethanol (MW 117) Insert (i)-(iii) shows the MS/MS CID data for the selected ions at m/z 88, 100 and 118 for morpholine, cyclohexylamine and diethyl amino ethanol respectively...... 106 xxi

Figure 4.5. Duomeen O calibration curve for the qualitative analysis polyamine in boiler system water samples using paper spray mass spectrometry in positive ion mode...... 107 Figure 4.6. Positive ion mode paper spray mass spectrum of metaldehyde recorded using a bench-top ion trap mass spectrometer. 5 μg of the analyte in 1 μL of deionized water was spotted onto filter paper and ionized in air by application of a positive electric potential (3.5 kV) using methanol as the paper spray solvent. (A) The sodiated molecular ion [M+Na]+ peak of metaldehyde (MW 176) in deionized water produced the dominant ion signal intensity (m/z 199), and (B) Sodiated molecular ion [M+Na]+ of deuterated metaldehyde-d16 (MW 192) in deionized water produced the dominant ion peak (m/z 215). Inserts (i–ii) show the isotopic distribution of the metaldehyde and metadehyde-d16 sodiated [M + Na]+ ion adducts at m/z 199 and 215 respectively. Note that in insert (ii) the relatively large signal intensity for m/z 214 is likely a consequence of D-H back-exchanges occurring in the ambient environment (and 99% isotopic enrichment). Inserts (iii–v) show the tandem MS CID data for the selected ions of metaldehyde and metadehyde-d16. .... 108 Figure 4.7. Positive ion mode paper spray mass spectrum of metaldehyde recorded using a bench-top ion trap mass spectrometer. 5 μg of the analyte in 1 μL deionized water was spotted onto filter paper and ionized in air by application of a positive electric potential (3.5 kV) using methanol as the paper spray solvent. The figure shows the CID data for the precursor ion at m/z 194...... 110 Figure 4.8. Calibration curve for quantification of metaldehyde in water using PS-MS/MS when analyzing (A) sodiated ion types and (B) ammoniated ion types produced in neutral MeOH spray solvent. Error bars indicate standard deviation from three replicates...... 111 Figure 4.9. Positive ion mode reactive-PS mass spectrum Duomeen O analyzed using a benchtop instrument: (A,B) typical Duomeen O mass spectrum analyzed without the acetone reagent and MS/MS CID data, respectively, while parts B and D show the product of the Duomeen O reaction with acetone detected in open air...... 113 Figure 4.10. Schiff-Base Condensation Reaction of the Primary Amines. (i) Nucleophilic reaction between the primary amine and ketone and (ii) reaction between Duomeen O (n- oleyl-1,3-diamine propane) (MW 324) and acetone in gas phase under ambient conditions using Reactive-PS-MS...... 114 xxii

Figure 4.11. Positive ion mode paper spray mass spectrum using a bench-top ion trap mass spectrometer with MeOH:(H2O+0.1% formic acid) (1:1, v/v) spray solvent application. 5 μg of the analyte in 1 μL of deionized water was spotted onto filter paper and ionized in air by application of a positive electric potential (3.5 kV); (A) metaldehyde and (B) paraldehyde. Tandem MS CID data for the m/z 177 and m/z 133 ions are shown in inserts (i) and (ii) respectively...... 116

Figure 4.12. Positive ion mode paper spray mass spectrum of metaldehyde-d16 using acidified spray solvent. Inset (i) shows CID data for the precursor ion at m/z 193...... 117 Figure 4.13. Comparison of CID of m/z 149 formed (A) directly from the solution with (B) the fragmentation of gas-phase m/z 149 formed from the MS2 of m/z 177 (insert (i)). 119 Figure 4.14. Calibration curve for quantification of metaldehyde in water using PS-MS/MS when analyzing protonated ion types produced in acidified spray solvent. Error bars indicate standard deviation from three replicates...... 120 Figure 4.15. Positive ion mode paper mass spectrum for polyamine and amine corrosion inhibitor formulation complex mixture (competitor product A) analyzed using a benchtop mass spectrometer. (A) Mass spectrum of competitor product A corrosion inhibitor mixture analyzed without acetone reagent. A volume of 2 μL of the corrosion inhibitor mixtures was deposited onto the surface and ionized and analyzed in the open air by application of an electric potential of +3.5 kV positive ion mode. Insert (i)–(iii) are the MS/MS CID mass spectra for the m/z 325, m/z 337, m/z 351, respectively. (B) Mass spectrum of competitor product A corrosion inhibitor mixture analyzed with acetone reagent. The protonated ion of the reaction product is subsequently detected at m/z 365. Insert (iv) is the MS/MS CID mass spectra for the m/z 365...... 121 Figure 4.16. Positive ion mode paper mass spectrum for polyamine and amine corrosion inhibitor formulation complex mixture analyzed using a bench-top mass spectrometer; (A) mass spectrum ascameen corrosion inhibitor mixture, (B) mass spectrum of naylamul S II corrosion inhibitor mixture. About 2 μL of the corrosion inhibitor mixtures was deposited onto the surface and ionized and analyzed in the open air by application of an electric potential of + 3.5 kV positive ion mode. Insert (i)-(ii) are the MS/MS CID mass spectra for the m/z 325...... 123 xxiii

Figure 4.17. Positive ion mode paper spray mass spectrum for rapid detection of Duomeen O corrosion inhibitor boiler system water samples: (A) condensate water, (B) feedwater, (C) boiler water. A volume of 2 μL of the sample was deposited onto the surface and ionized in the open environment by application of an electric potential of +3.5 kV positive ion mode. Inserts (i)–(iii) are the MS/MS CID mass spectra for the protonated Duomeen O at m/z 325 detection from each sample...... 124 Figure 4.18. Positive ion mode paper spray mass spectra for rapid detection of metaldehyde in raw water samples (supplied by Northumbrian Water) whereby a volume of ~10 μL of the sample was deposited onto the paper substrate and ionized in the open environment by application of an electric potential of +3.5 kV. Abberton Raw was analyzed according to (A) the ‘normal PS-MS’ method and (B) with reactive PSMS. Similarly for Chigwell Raw, ‘normal PS-MS’ analysis is shown in (C) and reactive PS-MS in (D). Inserts (i) & (ii) are the MS/MS CID mass spectra for the protonated metaldehyde ion at m/z 177 from each water sample analyzed using the reactive methodology...... 126 Figure 4.19. Proposed mechanism of acid catalyzed metaldehyde ring opening...... 127 Figure 4.20. Illustrative diagram showing reactive and “normal” PS-MS analysis of metaldehyde generating different ion types...... 128

xxiv

List of Tables

Table 2.1. Limits of detection (LODs) of analytes in blood, serum and urine using in-situ liquid/liquid during hydrophobic paper spray...... 42 Table 2.2. Reported surface tensions of acetonitrile/water mixtures...... 56 Table 2.3. Surface energy determination using peak surface tension solvent...... 57 Table 3.1. Limit of detection (LOD) and limit of quantitation (LOQ) of selected illicit drugs spiked into urine and analyzed at 3 kV with 100% acetonitrile spray solvent...... 80 Table 4.1. Analytical performance of PS-MS/MS for analysis of metaldehyde in water...... 111

xxv

List of Equations

Equation 1 ...... 5 Equation 2 ...... 6 Equation 3 ...... 6 Equation 4 ...... 58 Equation 5 ...... 59 Equation 6 ...... 64 Equation 7 ...... 64 Equation 8 ...... 64 Equation 9 ...... 64

xxvi

List of Abbreviations

2D 2-dimensional 3D 3-dimensional L Microliter m Micrometer PADs Microfluidic paper-based analytical devices ACN Acetonitrile AI Ambient ionization A/IS Analyte intensity divided by internal standard signal intensity CID Collision induced dissociation CSD Charge state distribution DAPCI Desorption atmospheric pressure chemical ionization DART Direct analysis in real time DC Direct current DESI Desorption electrospray ionization ESI Electrospray ionization GC Gas chromatography HPLC High performance liquid chromatography kDa Kilodaltons kV Kilovolts LOD Limit of detection LogP Log of partition coefficient LOQ Limit of quantification LTP Low temperature plasma

xxvii m/z Mass to charge ratio mL Milliliter mm Millimeter MS Mass spectrometry MW Molecular weight ng Nanogram ng mL-1 or ng/mL Nanogram per milliliter NTU Nephelometric turbidity unit PAN Polyacrylonitrile PS Paper spray pg picogram ppb Part per billion ppm Part per million s second RSD Relative standard deviation U/L Units per liter V Volt(s) v/v Volume per volume

xxviii

Chapter 1. Introduction and Background

1.1 Overview In this dissertation, detection of low abundance organic molecules present in biofluids and water samples using variations of paper spray mass spectrometry (MS) will be discussed. Overall, goals of this dissertation include analyte detection untreated samples without prior extraction, separation, or preconcentration using mass spectrometry as a detector. By utilizing paper spray, a biofluid or water-based sample can be deposited, possibly dried, and then analyzed within minutes or seconds (excluding drying time), which contrasts to current methods using gas chromatography or liquid chromatography that may take tens of minutes to hours. Chapter 2 describes the analysis of drugs of abuse (e.g., cocaine, benzoylecgonine, methamphetamine, and amphetamine) from blood, serum, and urine on hydrophobic paper. By decreasing interaction between the sample and paper substrate, increased sensitivity and the possibility of fresh sample analysis without drying was demonstrated. In addition to detection, the stability of labile molecules cocaine, benzoylecgonine, and diazepam during sampling and storage was analyzed. It was found that storage of these molecules in 2D dried blood spots causes signal to degrade within 1-4 days. Alternatively, storage in dried blood spheroids on hydrophobic paper causes and increase in analyte stability to several weeks. Finally, alternate ionization mechanism electrostatic paper spray is observed and explored, including utilizing this ionization method to determine surface energy of the treated paper substrate, typically a difficult task with current methods due to paper’s porous nature. Chapter 3 uses wax printing to manipulate solvent wetting area in order to increase spray time while decreasing voltage requirements when using paper spray. This is of import due to traditional paper spray’s short analysis time, especially when multiple fragmentation events are required. Because a lower voltage is required, a smaller power source is

1 theoretically required for ionization to occur, which would increase potential synergy with a portable mass spectrometer for on-site analysis of water or biofluid samples. In chapter 4, paper spray is utilized to detect metaldehyde and Duomeen O, which are a type of molluscicide found in agricultural settings and a corrosion inhibitor found in high pressure boiler systems respectively. Both molecules are primarily found in water in-situ and analysis of the raw sample without sample pretreatment or preconcentration was performed as both screening and quantification using paper spray ionization mass spectrometry. As an additional confirmation of presence, in addition to high resolution mass spectrometry and fragmentation, reactive paper spray ionization was performed, in which acetone or an acid was deposited with the sample on the paper triangle during ionization.

1.2 Dried Blood Spots on Paper as a Biofluid Collection Platform Paper as a biofluid sample collection platform serves as an inexpensive and convenient alternative to liquid sample storage. In particular, blood stored as dried blood spots is utilized in newborn screening, infectious disease monitoring, forensics, and illicit drug analysis. Dried blood spots (DBS) are less invasive, as sample can be collected from a heel or earlobe prick as opposed to venous sample collection. Additionally, sample storage and transportation are easier. DBS do not require lowered temperature (dependent on target analyte), have less stringent requirements when mailed, poses a lower infection risk, and is not breakable. Additionally, DBS require minimal training for collection. However, concerns with sample stability in uncontrollable environments such as during shipment, concerns of the required small sample size, and hematocrit effects that may artificially decrease analyte response when measuring a punched volume of blood are all possible deficits for DBS. As a way to correct for variable sampling volume, a center punch is taken from the sample prior to offline extraction and analysis. However, it has been found that when blood samples containing varying levels of hematocrit have varying rates of diffusion when being absorbed into the paper fiber. Blood with more hematocrit results in a higher viscosity. Because of this, wetting area may change between two samples that have the same blood

2 volume, and a center punch of equal area would have different amount of sample. Additionally, chromatographic effects result in inhomogeneity throughout the DBS, where analyte travels preferentially to the edge of the spot similar to a coffee stain. This effect results in an artificially lowered concentration of analyte in the center punch. 1–7 These effects have been minimized through use of pre-cut dried blood spots.8 This approach uses the entire dried blood spot. Other methods utilizing diffuse reflectance spectroscopy9 and monitoring of potassium levels.10,11 In this dissertation, we propose a chemical treatment to decrease interaction between blood and paper, thereby completely eliminating chromatographic effects.12,13 By pairing DBS with mass spectrometry, multiplex detection of biomarkers and biologically relevant molecules with a high degree of certainty. Typically, an off-line workflow pairing DBS and mass spectrometry requires an extraction step prior to either liquid chromatography (LC)-MS, gas chromatography (GC)-MS, or capillary electrophoresis (CE)-MS. Commonly, mass spectrometers also have the option for tandem mass spectrometry (MS/MS) in which fragmentation of a target precursor ion produces a reproducible fragment ion that can be used for identification and quantification.

1.3 Ionization Methods for Mass Spectrometry The ion source of a mass spectrometer is one of the three key components of the instrument. It determines what kind of sample can be introduced and analyzed, so developments of ion sources have contributed to revolutionizing strategies for mass spectrometric analysis. Common ionization methods include electron ionization, matrix- assisted laser desorption ionization, and electrospray ionization. All three of these ion sources are responsible for drastic increase to utility of mass spectrometers. However, all three sources require stringent sample preparation/introduction before analysis can occur. The development of ambient ionization has eliminated this sample preparation requirement, decreasing sample analysis time and cost. In this dissertation, development of alternate ion sources based on paper spray ionization, based on electrospray ionization, is discussed.

1.3.1 Electrospray Ionization Electrospray ionization (ESI) is a method of creating and transferring ions dissolved in the liquid phase in atmospheric air to the low-pressure environment of a mass spectrometer. ESI was patented by John Fenn in the 1980s.14–16 As opposed to EI, ESI is a “soft” ionization method, in which energy is applied gradually to the analyte in solution due to the surrounding solvent molecules, which prevents fragmentation from occurring simultaneously. This is contrasted by “hard” ionization methods such as EI, in which energy is applied abruptly enough to break C-C bonds in a molecule, initiating fragmentation.17 A condensed explanation of the ESI process is as follows: a voltage is applied to a metal capillary containing a solution of the sample and is positioned in front of a mass spectrometer inlet held at a different voltage (generally ground). A Taylor cone is formed due to (i) the electrostatic forces drawing the solution toward the inlet, and (ii) surface tension pulling the solution back. A jet of liquid is emitted from the tip of the Taylor cone, producing charged droplets travelling toward the mass spectrometer inlet. The charged droplets, exposed to ambient air, undergo evaporation until the charge density in the droplet reaches the Rayleigh limit, and coulombic explosion occurs, producing many smaller charged droplets. This process is repeated until gas phase ions remain, which travel into the mass spectrometer inlet. (Figure 1.1)

Figure 1.1. Electrospray ionization schematic.

More specifically regarding the electric potentials required for ESI, when voltage is applied 17 to the metal capillary, the electric field at the tip of the capillary, Ec can be evaluated as:

2푉푐 퐸푐 = 푟푐ln⁡(4푑⁄푟푐) Equation 1

Where Vc is the applied potential, rc is the capillary outer radius, and d is the distance from the capillary tip to the counter electrode. The equation describing the required electric field at the capillary tip Eon for the formation of a charged jet and therefore charged droplets to form is:18

2훾 cos 휃 퐸표푛 ≈ ( ) 2휖0

Equation 2

Where γ is the surface tension of the solvent, θ is the half angle of the Taylor cone, and ϵ0 is the permittivity of vacuum. By combining Equation 1 and Equation 2 and substituting θ 19 -12 2 -1 -1 = 49.3° and ϵ0 = 8.8 x 10 C J m , we can estimate the required onset voltage for ESI to occur.

5 1⁄2 푉표푛 ≈ 2⁡푥⁡10 (훾푟푐) ln⁡(4푑⁄푟푐) Equation 3

This onset voltage is experimentally typically found to be 1-5 kV depending on the solvents and capillary used. Typical solvent systems are organic/water mixtures, such as acetonitrile, methanol, or ethanol. Because ESI requires the sample to be dilute in order for travel through a thin metal capillary and direct spray into the mass spectrometer, direct analysis of samples such as biofluids is not possible. In order to analyze these samples, a prior offline extraction must take place, followed by an optional separation through a chromatograph. Due to the large time requirements of sample preparation, alternative ionization sources have been developed for the direct analysis of samples with minimal preparation, coined as ambient ionization.20

1.3.2 Ambient Ionization Ambient ionization (AI) is a form of ionization that is performed on unmodified samples in open air and the method is capable of providing almost instantaneous data while minimizing sample preparation.21–28 AI has gained traction due to decreased analysis time, resulting in decreased cost-per-sample analysis and increased throughput for heavy- workload laboratories. Some of the most popular AI techniques include desorption electrospray ionization (DESI),29 extractive electrospray ionization (EESI),30–35 desorption atmospheric pressure chemical ionization (DAPCI),36–38 and direct analysis in real time (DART).39,40 AI-MS shows promise as an analytical tool for in-situ applications and has

6 been demonstrated in a variety of fields where timely intervention is highly desirable such as: homeland security,22, food safety,41 pharmaceutical drug development,42 and environmental monitoring.43 There are several advantages to using in-situ AI methods capable of onsite analysis. The foremost advantage is the provision of data in real-time (or near real-time) at the point of interest allowing key management decisions to be taken in a timely manner. Subsidiary advantages relate to the chain of custody: by effectively taking the lab to the sample rather than the sample to the lab, the sample integrity is maintained, and sampling/handling costs are significantly reduced.

1.3.2.1 Paper Spray The objective of the present study is to develop a new method for rapid detection and quantitative analysis of small molecules, found in either water or complex matrices such as blood, serum, and urine, using AI-MS, based on paper spray (PS) ionization. PS- MS is a relatively new AI technique, first reported by Cooks, Ouyang & co-workers44 in 2010. (Figure 1.2) PS has since been demonstrated for the analysis of a wide range of samples including biofluids,45–47 bio-tissues,48 protein complexes,49 foodstuffs,50–52 beverages,53,54 bacteria,55 and biocides.56 The technique has undergone various developments such as high throughput implementation,57 application of carbon nanotube impregnated paper enabling low voltage application,58 integration with solid phase extraction,59 solid substrate deposition,60–64 and chemical modification65 for manipulation of surface properties. The use of paper as a substrate material in analytical chemistry has been demonstrated for several decades and has many advantages such as: it has high surface area-to-volume ratio, it is readily available at low-cost, it wicks aqueous fluids, it is biodegradable and lightweight allowing for easy transportation and storage. In a typical PS experiment, a cellulose chromatographic paper is cut into equilateral triangles with ~5 mm sides using scissors and is wetted with a solvent. Charged droplets are emitted from the paper tip when a high DC voltage (± 3–5 kV) is applied. Droplet emission occurs via Taylor cone formation, which leads to analyte(s) ionization through electrospray-like (and/or other unidentified) mechanisms.66 Moreover, analysis by PS-MS requires little or no sample preparation and the entire full MS or MS/MS experiment can be completed within seconds

(< 1 minute). In comparison to other ambient ionization methods, PS integrates three analytical procedures: sample collection, separation, and ionization into a single experimental step making it more attractive for rapid and direct analysis of analyte(s) in complex mixtures. In addition, no nebulizer gases are required so the technique can be more readily used with portable MS in the field.

Figure 1.2. Schematic for paper spray ionization.

1.4 Linear Ion Trap Mass Spectrometer Detection For the most part, a linear ion trap mass spectrometer (LTQ Velos Pro, Thermo Scientific, Figure 1.3) was used for the analyses presented in this dissertation. A linear ion trap (Figure 1.4) is similar to a quadrupole, in which four metal rod electrodes are arranged in parallel square formation in the x-y plane. Opposite rods are treated as a pair, and each pair has an AC and DC voltage applied. Ions travel between the rods in the z-axis, and ions that have a stable trajectory that travel the length of the analyzer without crashing into the

8 instrument are detected. One pair of rods acts as a “high-mass filter,” in which ions above a set m/z do not have a stable trajectory. Similarly, the opposite pair acts as a “low-mass filter,” where the opposite is true. When paired together, ions of a selected m/z within 1 amu are able to be detected.

Figure 1.3. Schematic of a Thermo Scientific LTQ Velos Pro mass spectrometer from a Thermo Scientific brochure.67

Linear ion traps utilize these metal rod electrodes, but also have a trapping electrode across the would-be entrance and exit of the quadrupole. By setting the quadrupole to allow all ions within a certain m/z range and using both electrodes to prevent the escape of ions, these ions are stored inside of the trap. Ions can then be ejected from the trap based on their m/z, and detection can occur. Fragmentation of these ions can be performed by colliding with a neutral gas at elevated energy, called collision induced dissociation. Fragmentation provides an additional identification mechanism for ions. Precursor m/z is not specific enough to solely identify an ion, as a given m/z may have several possible chemical formulas or several isotopes associated. Fragmentation provides the information of where a molecule may have weaker bonds, which may help elucidate structure or chemical formulas in some cases. Additional information would be needed when identifying a 9 completely unknown sample, but when performing screening of known samples or quantification, fragmentation is typically sufficient when characteristic fragments of a given ion are identified.

Figure 1.4. Schematic of a linear ion trap.

Linear ion traps are a common mass analyzer when using a portable mass spectrometer. This is because they are able to operate at higher pressures than other mass analyzers such as time of flight and sector. Also, as discussed previously, linear ion traps can perform MSn experiments, making them ideal for identification of ions in the field. Because of this, experiments performed on benchtop linear ion trap instruments can ideally translate to their portable counterparts for on-demand mass spectrometry. Portable mass spectrometry paired with paper spray would lead to acquiring mass analysis from complex samples nearly instantaneously, resulting in rapid decision making and addressing issues such as diagnostic results or abnormal chemical levels in industrial settings.

Chapter 2. Hydrophobic Paper Spray Ionization

2.1 Introduction: Dried Blood Spots For more than a century, biological fluid samples (e.g., blood) have been collected onto paper substrates as dried blood spots (DBS),68,69 1 and yet all noted advantages are still valid today – including the simplification of (small) sample collection, transportation, storage and processing. Interests in the DBS method have been rekindled in recent years because of the emergence of personalized healthcare. This includes the recent introduction of an on-demand diagnostic strategy,70,71 which is expected to enable timely initiation of treatment and long-term disease monitoring within the context of a medical home and subspecialty center. These efforts are important especially for newborn screening programs,72–75 the development of low-cost analytical diagnostic methods for use in resource-limited settings,76–81 environmental research,82,83 and drug analysis.45,84,85 Therefore, certain aspects of DBS need improvement, including the preservation of labile compounds during storage, while also maintaining the simplicity of the approach.1 As a facile method, the basic precautions for DBS collection include limiting sample exposure to moisture, sunlight and heat. As will be shown, however, exposure of DBS to ambient air can also substantially affect analyte integrity. Because current major focus of DBS collection is re-testing at a reference laboratory (which may be a part of external equality assessment plan), it has become critical to know the exact volume of blood in the punched sample to enable effective comparison to results recorded at the testing site. This seemingly simple task is complicated by i) volcanic effects – cause concentration gradient in DBS with higher analyte concentrations detected toward the edge; ii) chromatographic effects – the choice of paper substrate impacts DBS sampling by

1 Reproduced with permission from Damon, D. E.; Yin, M.; Allen, D. M.; Maher, Y. S.; Tanny, C. J.; Oyola- Reynoso, S.; Smith, B. L.; Maher, S.; Thuo, M. M.; Badu-Tawiah, A. K. Dried Blood Spheroids for Dry- State Room Temperature Stabilization of Microliter Blood Samples. Anal. Chem. 2018, 90 (15), 9353–9358. 83 Copyright 2018 American Chemical Society.

11 altering blood diffusion and adsorption; and iii) hematocrit effects – varied red blood cells in patients’ blood (e.g., anemic sample) cause variable blood diffusion on paper, altering volume sampled in a punch. As such, the determination of blood volume in punched paper samples is currently achieved using mathematical calculations5,8,86,87 or via radioactive chemical tracers.87,88

2.1.1 Paper Spray Strategies for Biofluid Analysis Established protocols exist for biological sample loading, storage, and extraction from paper,89 but the ability to perform in situ chemical detection directly from the inexpensive paper substrate has profound implications on clinical analyses, especially in resource-limited settings.90 The relatively new paper spray (PS) ambient ionization technique enables handheld mass spectrometers to directly characterize untreated biofluid samples with no need for sample preparation.91 In PS mass spectrometry (MS), sample is simply placed on a paper triangle, and a high DC voltage (∼3−5 kV) is applied to the wet paper triangle. This process releases charged microdroplets that contain the analyte of interest, which are then transported to the mass spectrometer for characterization. PS-MS has simplified and expanded the utility of mass spectrometric analysis to include the quantitative detection of drugs and their metabolites directly from biofluids,92,93 detection of food contaminants,52 lipid profiling from tissues and bacteria samples,55,94 and analysis of noncovalent protein complexes,49 all requiring minimal sample pretreatment. Unfortunately, detection limits of PS analyses from complex samples such as blood and urine are often inadequate. These higher limits of detection have been attributed to inefficient analyte extraction from the sample matrix,45 and recent efforts have focused on developing solid-phase extraction (SPE) methods to enhance the sensitivity of the PS experiment by concentrating the analyte onto the SPE surface (e.g., a delrin plastic cartridge and C18-coated metal blade).59,95 Although lower detection limits are obtained, the separate extraction and cleaning steps required for these SPE methods can limit field

12 analysis when using portable mass spectrometers and in the analysis of small sample volumes.2

2.1.2 Dried Blood Spheroids83 Herein, a new paper-based blood collection platform is reported that is based on three-dimensional dried blood spheroids as opposed to the traditional two-dimensional DBS sample collection procedure. This new dried blood sample collection procedure uses functionalized hydrophobic paper substrates (prepared in-house) to overcome major challenges associated with the traditional DBS procedure. The advantages and attributes of this approach include: i) blood sample applied on the hydrophobic paper forms a spherical drop due to a mismatch in surface energies, which dries to yield a dried blood spheroid. Experiments have shown that hydrolytically labile chemicals such as cocaine and diazepam trapped in the 3D dried blood spheroid are stabilized, compared with storage done under the porous DBS conditions where a major portion of the sample becomes susceptible to oxidative stress from ambient air; ii) because the origin of volcanic, chromatographic and hematocrit effects can all be traced to a common source – uneven biofluid/analyte adsorption – controlling wetting on hydrophobic paper should enable easy validation of results without the of use chemical tracers to estimate sample volume in dried blood punch; iii) the hydrophobic paper strips also provide a direct mass spectrometry (MS) detection through paper spray (PS) ionization13,70,91,96–101 for sensitive analyte quantification. In-situ extraction of illicit drugs (methamphetamine, benzoylecgonine, amphetamine and cocaine) from the dried blood spheroids resulted in sub-ng/mL limit of detections using ethyl acetate spray solvent; and iv) proper control of the analyte desorption from the paper substrate enabled the development of a new electrostatic spray-based method to estimate the surface energies of the hydrophobic paper strips, which was found to be more effective than conventional approach based on contact angle measurements.

2 Reproduced with permission from Damon, D. E.; Davis, K. M.; Moreira, C. R.; Capone, P.; Cruttenden, R.; Badu-Tawiah, A. K. Direct Biofluid Analysis Using Hydrophobic Paper Spray Mass Spectrometry. Anal. Chem. 2016, 88 (3), 1878–1884.94 Copyright 2016 American Chemical Society.

Figure 2.1. Experimental setup using (A) paper triangles and (B) paper rectangles. (C) Image showing a 4 L dried blood spot/spheroid on an untreated (left) and treated (right) paper substrates, including the front (top) and back (bottom). Scale bars show 0.5 mm. (D) Workflow of direct on- surface dried blood analysis.

Untreated Paper Treated Paper

0.3 0.03

) ) 2 0.2 0.02

0.1 0.01

0 0

Area (cm Area 0 5 10 15 20 0 5 10 15 20 Number of Measurement Number of Measurement

Ave: 0.13 ± 0.05 cm2 Ave: 0.013 ± 0.002 cm2

Figure 2.2. 4 µL blood was pipetted onto untreated and 30 minute treated paper. The following areas of wetting were measuring using ImageJ.

To enable dried blood spheroid collection, pre-cut hydrophilic filter papers (Figure 2.1A, B) were converted into hydrophobic paper substrates through a gas-phase silanization procedure.94 Further discussion on paper treatment and geometry can be found on page 22. As observed from Figure 2.1C, the fresh blood penetrated to the back of the untreated paper forming a blood spot of area 0.13 ± 0.05 cm2, with >2X relative standard deviations between samples. In contrast, no traces of blood were observed on the back of the hydrophobic paper; instead the entire 4 μL blood volume bead, which rested on top of the hydrophobic paper strip, confined to a reproducible area of 0.013 ± 0.002 cm2 (Figure 2.2). Although not the focus of the current work, the whole dried blood spheroid could be punched for subsequent extraction without regard to volcanic, chromatographic or hematocrit effects.

Figure 2.3. (A) Functionalization of paper by modification of fiber surface using trichloro(3,3,3- trifluoropropyl) silane vapor to create a hydrophobic oligomeric silylated layer. Gaseous HCl released as a byproduct of the reaction is removed by vacuum in situ. (B) Photograph showing interaction of a dye solution on untreated versus treated (hydrophobic) paper. (C) Hydrophobic paper spray in which the biological sample (e.g., blood) prevents the spreading of the organic solvent beyond the point where the sample was deposited. 15

-3HCl +nH2O -nH2O RSiCl3 RSi(OH)3 + +3H O 2 Paper Substrate Functionalized Paper Substrate

Figure 2.4. Functionalization of paper by modification of surfaces hydroxyl groups using chlorosilane vapor to create a hydrophobic oligomeric silated layer. Gaseous HCl released as a byproduct of the reaction is removed by vacuum in situ.

2.2 Materials and Methods Chemicals and Reagents: Standard solutions (1.0 mg/mL) of benzoylecgonine, cocaine, amphetamine, and (±)-methamphetamine were obtained from Cerilliant (Round Rock, TX). Human blood was purchased from Innovative Research (Novi, MI). Phosphate-buffered saline (PBS) tablets were purchased from AMRESCO (Solon, OH). Glacial acetic acid was purchased from Thermo Fischer Scientific (Waltham, MA). Methanol (99.9%, HPLC grade), (3,3,3- trifluoropropyl)- silane, bovine serum albumin (BSA) (10%) in PBS, myoglobin from equine skeletal muscle, human serum, α-ketoglutaric acid disodium salt hydrate, and L- Alanine were purchased from Sigma-Aldrich (St. Louis, MO), including standards (hexadecane, tridecane, dodecane, decane, octane, benzene, toluene, p-xylene, nitrobenzene, glycerol, ethylene glycol) used in the estimation of surface energy. Lyophilized alanine aminotransferase from human liver was purchased from Lee Biosolutions (Maryland Heights, MO). Whatman filter paper (24 cm, grade 1) was purchased from Whatman (Little Chalfont, England).

Hydrophobic Paper Preparation: Using a digital template, paper triangles were cut from filter paper with an Epilog Legend 36EXT laser with 15% power at 1000 Hz. Typically, 0.5 mL of silanization reagent (Trichloro(3,3,3- trifluoropropyl) silane) was used for 4−5 sheets of filter paper. For all experiments, the required paper size was cut before silanization. Paper size was approximately 80 mm2 (base width of 9.5 mm, height of 16.6 mm), and paper/polymer rectangles were 4 mm base x 20 mm height to maintain similar surface area. Whole blood 16 samples were pipetted onto the paper surface and allowed to dry overnight unless otherwise stated. Because this approach utilizes a gas-phase preparation procedure to impact changes in surface properties, many of the physical/chemical characteristics (e.g., color, weight, porosity, tensile strength, malleability, flammability) of the filter paper remain unchanged. However, wettability of the paper is altered controllably by varying silane vapor exposure time. As a result of the lowered surface energy, aqueous-based samples such as blood and serum bead when applied onto the hydrophobic paper, and as a consequence, form 3D spheroids (molds, Figure 2.1C) upon drying due to concentration-driven self-assembly of bio-macromolecules in the biofluid. Only the outermost layer of the dried blood spheroid is exposed to air during storage, preserving the integrity of majority of analytes inside the dried blood. Two shapes of hydrophobic paper strips were used (triangular and rectangular (Figure 2.1A and B); pre-cut before silanization) to investigate if dedicated tips are necessary during direct, in-situ MS analysis of the dried blood spheroids.

Mass Spectrometry, Paper Spray Video, and Contact Angle: Samples were analyzed by a Thermo Fisher Scientific Velos Pro LTQ linear ion trap mass spectrometer (San Jose, CA, U.S.A.). MS parameters used were as follows: 150 °C capillary temperature, 3 microscans, and 60% Slens voltage. Spray voltage was 5 kV unless otherwise specified. Thermo Fisher Scientific Xcalibur 2.2 SP1 software was applied for MS data collecting and processing. Tandem MS with collision-induced dissociation (CID) was utilized for analyte identification. Dry hydrophobic PS spray plume was observed using a Watec camera (WAT- 704R). Color pictures were taken with a Canon PowerShot SX410 IS. Contact angles were observed using a Rame-Hart goniometer.

Statistical Analysis: Signal intensities obtained from Xcalibur were used for data analysis. For calibrations and stability experiments, intensity of the analyte of interest primary fragment was divided by the intensity of the deuterated internal standard primary fragment that was also spiked into the whole blood or into the spray solvent respectively. These experiments

17 were performed in triplicate and all error bars show one standard deviation. Limits of detection and quantification were calculated using a blank containing only internal standard in whole blood. The analyte of interest signal in this blank added to three times the standard deviation of that blank signal was used to find the corresponding concentration.

Enzyme Preparation: Lyophilized alanine aminotransferase was prepared in 1× PBS solution with 2% BSA in DI water. Samples were directly spiked into 30 μL human blood with 2 μL each of L-alanine (1.75 M) and α-ketoglutarate (60 mM) solutions. Triplicate solutions were prepared in a temporally staggered manner such that each sample was analyzed after the reaction was allowed to occur for 10 min.

Thermal Analysis Simulation Model: Transient natural thermal transfer analysis (natural convection and conduction) was carried out using SolidWorks Simulation 2014 (SolidWorks Corp., Massachusetts, USA). A sphere (radius 5 mm) and circular disc (radius 5 mm, depth 1 mm) were drawn in CAD to approximate the geometry of the blood spheroid and DBS respectively. A finite element analysis was applied to the CAD model using the thermal analysis FFEPlus solver with the following parameters: triangular mesh with 10234 nodes and 6929 elements each of size 0.378 mm, 30°C object initialization temperature, 40°C bulk ambient temperature, with air as the natural convection medium with a heat transfer coefficient of 25 W/(m2·°C). Transient analysis was conducted over 90 seconds with 3 second measurement intervals. Thermal properties were chosen so as to approximate whole human blood: 3617 J/kg/°C specific heat capacity, 0.492 W/m-°C thermal conductivity and 1060 kg/m3 density.

2.3 Hydrophobic Paper Preparation for Paper Spray Analysis The hydrophobic paper was prepared in-house through self-assembly chemistry of silanization (Figure 2.3A, Figure 2.4)102,103 by exposing laser-cut filter paper triangles to the vapor of trichloro(3,3,3-trifluoropropyl)silane reagent under vacuum (∼20 Torr) for approximately 4 h. Typically, 0.5 mL of silanization reagent (Trichloro(3,3,3-

18 trifluoropropyl) silane) was used for 4-5 sheets of filter paper. For all experiments, the required paper size was cut before silanization. Slow silanization (with no heating) was performed to ensure only exposed fabric surface hydroxyl (OH) groups are derivatized while leaving most of the OH groups involved in intermolecular hydrogen bonding within the fiber core unreacted. This prevents significant alteration of mechanical properties of the paper. (Figure 2.5). The surface energy of the 3,3,3-trifluoropropyl functionalized paper was estimated to be 44 mN/m (Figure 2.6).

200 Untreated Paper

) Treated Paper

160 Nmm

/ 120 mN 80

Stiffness Stiffness (mN/mm) Stiffness ( Stiffness

0 FilterGB003 Paper ChromGB004. Paper GB005

Figure 2.5. Effect of silanization on stiffness ofPaper filter paperType and chromatography paper at 50% relative humidity salinization.

It was determined that extended silanization time (> 12 h) resulted in paper material that is brittle due to substantial reduction in the number of hydrogen bonding which give the paper its mechanical strength. This observation lends support to the expectation that within typical silanization time (<4 h), OH groups located inside the fiber core are not silanized. This is supported by measurement of the stiffness of the functionalized-paper, and comparing it with that of the untreated (filter and chromatography papers). In all cases, any change in paper stiffness after silanization was not observed (Figure 2.5), indicating 19 that the mechanical strength remained unchanged after treatment. Silane paper does not retain its structural stability indefinitely, and is not suitable for analysis after approximately 4 months, by which time the paper becomes brittle and breaks easily during regular handling.

104,105 Figure 2.6. Zisman plot showing surface energies of CH2CH2CF3-funsctionalized paper. Ɵ is measured contact angles. A series of selected solvents was used (surface tension, mN/m) ranging from hexane (18.4), heptane (20.1), octane (21.6), decane (23.8), dodecane (25.4), hexadecane (27.5), nitrobenzene (43.9), ethylene glycol (47.7), glycerol (63.3) to water (72.2).

Total wetting occurs when the surface tension of the wetting liquid is less than the critical energy of the surface. To estimate the surface energy of the silanized paper, the method of bracketing was used. The basic idea in this bracketing experiment is that a liquid drop will wet a surface only when the wetted surface has a lower energy than the initial dry surface. Only such an exothermic reaction will proceed, and thus casting a drop of selected “inert” liquids onto a surface will lead to only two outcomes: category one, liquid droplet wets the surface and so its surface tension is lower than the critical energy of the surface, and category two, liquid droplet does not wet the surface meaning the surface tension of the liquid is higher than the critical energy of the surface. The result of a series of such bracketing experiments is a quantitative measure of the surface energy of the paper if the 20 surface tensions of the selected liquids are known. To do this, different solvents were tested (surface tensions in mN/m) including hexane (18.4), heptane (20.1), octane (21.6), decane (23.8), dodecane (25.4), hexadecane (27.5), nitrobenzene (43.9), ethylene glycol (47.7), glycerol (63.3), and water (72.2) on –CH2CH2CF3 functionalized paper. The critical surface energy was found to be in the range 43.9 – 47.7 mN/m (i.e., wets nitrobenzene but not ethylene glycol). Zisman plot (Figure 2.6) confirms the range of critical energies estimated by the bracketing approach. Unlike a previous report that used hydrophobic surfaces to differentially extract analytes from a matrix,106 the underlying principle for the current experiment is to prevent wetting by creating a mismatch in surface energies, where aqueous-based samples (surface tension ∼72 mN/m) can be organized as spherical droplets (Figure 2.3B) on the low-energy (hydrophobic) paper surface for enhanced sampling. 2.5 (A) 2

2.5 1.5 2 untreated 1.5

A/IS 30 min treated 1 1 0.5 2 hr treated 0.5 0 0 1 2 3 4 5 0 0 10 20 30 Days Since Blood Application 60 40 (B) untreated (C) 30 min treated 30 40 2 hr treated

A/IS A/IS 20 10

0 0 1 2 3 4 5 6 7 1 2 3 4 5 6 7 Days Since Water Application Days Since Blood Application

Figure 2.7. Stability of (A) cocaine in dried blood, (B) neat dried diazepam prepared in water, and (C) diazepam in dried blood. Both dried blood spots (untreated) and spheroids (treated) samples were stored under ambient conditions at 25°C. Internal standard was spiked into the spray solvent to normalize between samples and days. Three replicates were used. Error bars show one standard deviation. 21

2.4 Characterization of Paper Strips for Paper Spray Analysis The overall workflow for blood collection and analysis from hydrophobic paper is as illustrated in Figure 2.1D, where 4 μL of blood was deposited onto the hydrophobic paper strip, dried for a specified time, and analytes were detected using hydrophobic PS- MS. Using ethyl acetate as spray solvent, small organic compounds (e.g., amphetamine and methamphetamine) were selectively extracted and detected from blood and neat water- based samples dried on hydrophobic paper rectangles (Figure 2.8 and Figure 2.9), albeit lower ion intensity compared with hydrophobic paper tringles due to the absence of a dedicated macroscopic tip.

Figure 2.8. (Columns) show spectra grouped by paper treatments, and (Rows) show spectra grouped by drugs fragmented. (A-D) Fragmentation of (A) amphetamine, (B) methamphetamine, (C) benzoylecgonine, and (D) cocaine on untreated paper strips. (E-H) Fragmentation of (E) amphetamine, (F) methamphetamine, (G) benzoylecgonine, and (H) cocaine on 2 hour treated paper strips. Characteristic fragments include: cocaine (304→182), benzoylecgonine (290→168), methamphetamine (150→119) and amphetamine (136→119). 22

Figure 2.9. (Columns) show spectra grouped by paper treatments, and (Rows) show spectra grouped by drugs fragmented. (A-D) Fragmentation of (A) amphetamine, (B) methamphetamine, (C) benzoylecgonine, and (D) cocaine on paper strips treated for 30 minutes. (E-H) Fragmentation of (E) amphetamine, (F) methamphetamine, (G) benzoylecgonine, and (H) cocaine on paper strips treated for 2 hours. Characteristic fragments include: cocaine (304→182), benzoylecgonine (290→168), methamphetamine (150→119) and amphetamine (136→119).

500 ng/mL of cocaine, benzoylecgonine, amphetamine, and methamphetamine were spiked into water and then dried onto the paper strip surface that was untreated or previously treated for 2 hours with silane. 3 kV and 20 L of ethyl acetate were applied, and collision-induced dissociation (CID) of the target drugs was performed. In comparison to untreated paper rectangles, however, an enhancement (>10X) in ion yield was observed 23 when using the hydrophobic paper, with signal increasing with paper hydrophobicity (Figure 2.8). 500 ng/mL of cocaine, benzoylecgonine, amphetamine, and methamphetamine were spiked into whole human blood and then dried onto the paper strip surface that was untreated or previously treated for 30 minutes or 2 hours with silane. 3 kV and 20 L of ethyl acetate were applied, and CID of the target drugs was performed. Untreated paper strips showed no characteristic fragments of the target drugs, and therefore were excluded. This is due to increased matrix effects present from the blood in addition to the increased drug binding to the paper surface when compared to treated paper. Untreated paper is not able to overcome the matrix effects during extraction and also ionizes the analytes to a lesser extent. Treated paper performs a more efficient extraction and ionization step, so the target analyte is observed.

(A) (C)

1 mm (B)

1 mm 1 mm

Figure 2.10. (A) Shows setup of a paper strip in front of the mass spectrometer with a dried blood spot immobilized on the surface. (B and C) Images from a Watec camera showing a closer look at the loose paper fibers on the edge of the paper strip. (B) is a side on view and (C) is a top-town view. Fibers measure to be approximately 0.04 mm in diameter. Scale bars show 1 mm.

In the absence of a dedicated sharp tip, electrospray occurs from randomly oriented fibers protruding from the edges of the paper. (Figure 2.10) For untreated paper strips, these fibers easily bundle up, reducing the electric field needed to support electrospray-like ionization. The reduced wetting on hydrophobic paper in turn decreases the probability of fiber collapse,107 providing individual fibers that support ionization with 3 kV of direct current (DC) voltage. (Figure 2.11) In order to determine the basis of ionization, whether it be from pressure difference at the mass spectrometer (MS) inlet or from applied voltage, 3 kV (against the grounded MS inlet) was applied to the paper strip for a short time. The voltage was then changed to 0 kV. The total ion chromatogram shows the signal is only present at times when voltage is applied to the paper strip. This process shows that ionization is dependent on the applied voltage, and therefore the most likely method of ionization through electrospray-like mechanism from the paper strip. Because no tip is present on the strip (such as one present on paper triangles), ionization most likely occurs when a Taylor cone is formed on individual paper fibers that protrude from the blunt end of the paper strip.

3 kV 3 kV 100 0 kV

Relative Abundance Relative 0 0.2 0.4 0.6 Time (min)

Figure 2.11. Total ion chromatogram of paper strip with alternating 3 kV and 0 kV applied. Sample is a dried blood spot on 2 hour treated paper with 20 L ethyl acetate applied.

2.5 Analyte Stability in Dried Blood Spots and Spheroids To investigate the possibility of reducing oxidative stress during dried blood spheroid storage, two hydrolytically labile compounds were selected: cocaine and diazepam. Each compound (2 μg/mL) was spiked separately into whole human blood, and 4 μL aliquots were spotted onto the as-prepared hydrophobic paper and stored in ambient air for a maximum of 28 days. Similar blood samples were stored using the conventional DBS method on untreated, hydrophilic paper strips. Results from these analyses are summarized in Figure 2.7, which indicate that both cocaine (Figure 2.7A) and diazepam (Figure 2.7B) trapped inside the 3D dried blood spheroid are stabilized compared with storage done under the porous DBS conditions. About 90% of cocaine is hydrolyzed within a day of storage on untreated hydrophilic paper (insert, Figure 2.7A). Significant cocaine oxidation was observed after 40 days of storage under the dried blood spheroid storage condition. Benzoylecgonine is a metabolite of cocaine, but benzoylecgonine can also be a degradation product of cocaine.108 This degradation could be the cause of decrease of

26 cocaine intensity found in Figure 2.7. To monitor this, an offline extraction of the dried blood spots/spheroids in ethyl acetate was performed and analyzed via nanospray. Between days 1 and 2, benzoylecgonine intensity found in the cocaine-spiked blood sample increased relative to the cocaine intensity, indicating cocaine likely degraded to become benzoylecgonine while the sample was stored. Paper treated for 30 minutes and 2 hours did not experience this sharp increase in benzoylecgonine. (Figure 2.12). Therefore, the stability of benzoylecgonine under DBS versus dried blood spheroid conditions was investigated. (Figure 2.13). Here, greater than 95% of benzoylecgonine was oxidized (via the direct addition of oxygen) after the 12th day of storage under the typical DBS condition. In contrast, stable ion signal was detected for benzoylecgonine stored in the spheroid even after the 46th day. Unlike cocaine, the signal loss for diazepam on DBS was gradual, and 21% of the analyte remained in DBS after one week (untreated, Figure 2.7C). As expected, signal was relatively stable when stored using the dried blood spheroid methodology.

Day 1 Day 2

182 100 100 212 100% = 5.70E2 100% = 2.89E3 Cocaine m/z 304 m/z 304 50 212 50

91 304 91 182 304 Relative Abundance Relative Untreated Abundance Relative 0 0 100 200 300 400 100 200 300 400 m/z m/z 168 100 100 168 100% = 5.87E1 100% = 2.05E4 Benzoylecgonine m/z 290 m/z 290

in Cocaine Sample 50 50 Relative Abundance Relative 290 Abundance Relative 290 0 0 100 200 300 100 200 300 m/z m/z 182 100 182 100 100% = 5.34E3 100% = 1.21E3 m/z 304 212 m/z 304 Cocaine 50 50 212 304

91 Relative Abundance Relative

Relative Abundance Relative 91 304 0 0 100 200 300 400 100 200 300 400 30 Minute m/z m/z 168 168 100 100 Treated 100% = 9.19E3 100% = 2.96E2 m/z 290 m/z 290 Benzoylecgonine 50 50 in Cocaine Sample 290

290 Relative Abundance Relative Relative Abundance Relative 0 0 100 200 300 100 200 300 m/z m/z

182 100 100% = 5.48E3 100 212 182 100% = 1.70E4 m/z 304 m/z 304 Cocaine 50 212 50 304

91 91 Relative Abundance Relative

Relative Abundance Relative 304 2 Hour 0 0 100 200 300 400 100 200 300 400 m/z m/z Treated 100 168 100% = 1.66E2 100 168 100% = 3.14E1 m/z 290 Benzoylecgonine m/z 290 50 50 in Cocaine Sample 290

290 Relative Abundance Relative

0 Abundance Relative 100 200 300 400 0 m/z 100 200 300 m/z Figure 2.12. Offline analysis of 2 g/mL cocaine in dried blood spot on untreated paper and dried blood spheroid on 30 minute and 2 hour treated paper. Sample was spotted and stored for 1 and 2 days at 25 ͦ C. The samples were then extracted in 500 ng/mL D3 cocaine in ethyl acetate for 30 minutes in a sonicator. Extract was then nanosprayed, and m/z 304 (cocaine) and 290 (benzoylecgonine, possible cocaine degradation product) were fragmented.

0.25 untreated 30 min treated 0.2 2 hr treated 0.15

A/IS 0.1

0.05

0 1 2 3 5 12 21 28 39 46 Days Since Blood Application Figure 2.13. Stability of benzoylecgonine in dried blood spots (untreated) and spheroids (treated) stored in ambient conditions at 25° C. Internal standard was spiked into the spray solvent to normalize between samples and days. Error bars show one standard deviation.

To ensure that the decomposition processes observed here are due to oxidative stress from atmospheric air, a control experiment was performed in which both DBS and dried blood spheroids were stored in a vacuum desiccator for 15 days. Under this airtight condition, both cocaine and benzoylecgonine were found to be stable in DBS and spheroids (Figure 2.14). This confirms analyte degradation is due to ambient air oxidation. The involvement of oxygen is also confirmed by the direct detection of O2 adducts (+32 Da increase) in MS analysis (Figure 2.15). Neat diazepam was deposited onto paper triangles and analyzed immediately (day 0) or after 4 days of ambient storage. Lack of signal from the +32 peak (Figure 2.15B) on day 0 compared with signal on day 4 (Figure 2.15C), which yields water loss (m/z 299), CO2 loss (m/z 273), and a common ion with pure diazepam

(m/z 257). (Figure 2.15A) This indicates an increase in O2 addition to diazepam as it rests in ambient conditions for several days, contributing to the decreasing intensity of diazepam noted in Figure 2.7B and Figure 2.7C. To investigate a possible “wall effect” in the stabilization of analytes in dried blood spheroids, the stability of neat, dry diazepam (prepared in water, as opposed to blood) on both treated (no spheroid was formed) and

29 untreated paper substrates were compared, and found to be similar (Figure 2.7B). That is, neat diazepam analytes gradually degraded at comparable rates on both hydrophobic and hydrophilic paper strips. Ion intensities recorded from treated paper strips were relatively higher than those from untreated paper because of higher ionization efficiency of hydrophobic paper substrates.94

0.7

0.6

0.5 untreated 30 min treated 0.4 2 hr treated 0.3

0.2

0.1

0 Cocaine Benzoylecgonine

Figure 2.14. Stability of 2 g/mL cocaine and benzoylecgonine in dried blood spots/spheroids on untreated, 30 minute treated paper, and 2 hour treated paper. Samples were stored in a desiccator for 15 days and then analyzed with 5 kV and 10 L ethyl acetate containing 500 ng/mL deuterated internal standard to normalize between samples and between days. Error bars show one standard deviation.

Collectively, these results suggest that the creation of 3D spheroid from a viscous sample like blood is essential in preventing oxidation in air, and that the interior of the spheroid was protected by providing a possible critical radius of insulation109,110 that increases the spheroid’s resistance to thermal conduction and oxidative degradation. Finite element analysis of thermal energy flux from surrounding ambient air for spheroid and DBS approximated geometry confirms the spheroid’s enhanced thermal protection over a given time period (Figure 2.16). Transient thermal analysis (conduction/natural convection) was simulated in Solidworks, 2014 (MA, USA). Finite element analysis consisted of a triangular mesh with 10234 nodes and 6929 elements of size 0.378 mm. The 30 dried blood spot radius was approximated by a disc, having a depth of 1 mm, with a radius of 5 mm equal to that of a sphere approximating a blood spheroid. A transient analysis was conducted over 90 seconds with 3 second intervals. Both geometries were assumed as homogenous and initialized at 30°C, being subject to atmospheric air at an elevated temperature of 40°C over the transient period, with a heat transfer coefficient of 25 W/(m2·°C). Identical thermal properties which average that of whole human blood were used to describe both objects: specific heat: 3617 J/kg/°C,111 thermal conductivity: 0.492 W/m-°C112 and density: 1060 kg/m3. The reduced surface area-to-volume ratio of the spheroid limits bulk exposure to the ambient environment (Figure 2.17), which also improves resistance to oxidative degradation over time.

257 100 (A) 100% = 4.80E3 Pure m/z 285 Diazepam, 50 paper spray 154 immediately 222 182 285

Relative Abundance Relative 0 100 200 300 400 500 m/z

285 100 (B) 100% = 8.13E1 m/z 303 Day 0 50 Water adduct

Relative Abundance Relative 0 100 200 300 400 500 m/z (C) 257 100 100% = 1.91E4 m/z 303

Day 4 50 154 Water adduct 222 182 285

Relative Abundance Relative 0 100 200 300 400 500 m/z

Figure 2.15. (A) CID of neat diazepam, m/z 285. (B) CID of O2 adduct of diazepam (m/z 317) in water immediately after depositing on a paper triangle. (C) CID of O2 adduct of diazepam in water 4 days after depositing on a paper triangle.

Figure 2.16. Cut-away view of both geometries through their respective centers, illustrating the heat transfer in a DBS compared to spheroid based on their geometric properties after 18 seconds.

Figure 2.17. Heat transfer transient simulation analysis. Both blood storage geometries (DBS versus spheroid) had an initial temperature of 30 °C and were subjected to a constant ambient air temperature of 40 °C. Temperature is measured at the geometric center for each case

2.6 Dried Urine Sample Analysis The procedure described in this work is also applicable to the analysis of illicit drugs in dried biological samples. Using the hydrophobic paper, dried urine spots were prepared using a portable hand-pumped desiccator with 4 μL spiked raw samples. Dried samples were then analyzed by PS-MS after adding 20 μL of ethyl acetate to the

33 hydrophobic paper. MS/MS spectra were obtained almost instantaneously when ethyl acetate solvent was applied to the paper bearing the dried spot without any further sample processing steps. Amphetamine in dried urine at a concentration of 0.96 ng/mL can be readily identified with fragment ion peaks of substantial intensity visible in the MS/ MS spectrum (Figure 2.18). When dried urine samples spiked with cocaine were analyzed on untreated paper, a larger LOD of 3.5 ng/mL was achieved, which is 2 orders of magnitude higher than LOD recorded from treated hydrophobic paper. Our analysis is synergistic with portable, minimally invasive, and high-throughput biochemical analysis systems because small sample volumes do not spread on the hydrophobic paper and are able to be stored for later analysis.

Figure 2.18. Urine samples spiked to give a final concentration of (A) 0.96 ng/mL (B) 3.9 ng/mL (C) 10 ng/mL and (D) 50 ng/mL amphetamine. Aliquots of 4 µL were dried on the paper surface and then analyzed with ethyl acetate.

2.7 Quantification of Drugs of Abuse in Dried Blood Spheroids The quantitative abilities of the direct hydrophobic PS MS method was assessed using dried blood spheroid samples containing amphetamine, methamphetamine, cocaine, or benzoylecgonine. The initial investigations involved the use untreated paper and hydrophobic paper triangles treated with silane vapor at 5, 30, 120, 240, 720, and 1440 min exposure times. (Figure 2.19) These samples were analyzed with ethyl acetate as the spray solvent, and the absolute intensities of the fragment ions derived from collision-induced dissociation were quantified. Overall, the paper treated for 30 and 120 min produced the highest intensity responses and were selected for further testing (Figure 2.20,Figure 2.21). These responses were found to be influenced by i) drug binding affinity to the paper surface versus its solubility in the spray solvent (partitioning); ii) ionization efficiency – the impact of treatment time on PS performance; and iii) extraction efficiency of the analyte from the dried blood. Treatment time was not observed to affect analyte ionization due to comparable wetting of ethyl acetate on all paper triangles, which produced protonated ions via electrospray-based mechanism (as opposed to electrostatic ionization). Properties of the blood spheroids appeared identical (e.g. size and interaction with paper surface) on all paper treated for >30 min. Therefore, partitioning of the analyte between the paper and the solvent post-extraction is expected to be the major contributing factor affecting ion yields from treated paper substrates. This partitioning factor is in turn controlled by the logP of drug and paper treatment time (Table S3). For example, cocaine is the most hydrophobic drug tested (logP 2.28), hence, gave a higher ion intensity on paper with a shorter treatment time (i.e., less hydrophobic paper substrate: Figure 2.20). Similarly, benzoylecgonine, the most hydrophilic drug tested (LogP –0.59), showed a higher ion signal on paper with a longer treatment time (i.e., more hydrophobic). These results may be explained by that fact that molecules with high logP values prefer hydrophobic medium (and vice versa), and thus binding capacity is low on paper substrates prepared over shorter treatment times, enabling enhanced ion yield from such surfaces.

(A)

0 min 5 min 30 min 2 hrs 4 hrs 12 hrs 24 hrs (B)

30000 0 minutes treated 5 minutes treated 30 minutes treated 120 minutes treated 15000

AI 240 minutes treated 720 minutes treated 1440 minutes treated

0 Amphetamine Benzoylecgonine

700000 AI 350000

0 Cocaine Methamphetamine Figure 2.20. Extraction from dried blood spots with 20 L ethyl acetate. Absolute intensity of 500 ng/mL amphetamine, methamphetamine, cocaine, and benzoylecgonine on the surface of paper triangles. Characteristic fragments of cocaine (304→182), benzoylecgonine (290→168), methamphetamine (150→119) and amphetamine (136→119) were used for quantification. These triangles were untreated and treated for 5, 30, 120, 240, 720, and 1440 minutes with silane. Error bars show one standard deviation. 36

60000 0 minutes treated 5 minutes treated 30 minutes treated 40000 120 minutes treated

AI 240 minutes treated 720 minutes treated 20000 1440 minutes treated

0 Amphetamine Methamphetamine 350000

280000

AI 210000

140000

70000

0 Benzoylecgonine Cocaine Figure 2.21. Optimization of treatment time of paper using common illicit drugs and extraction from dried blood spots with 20 L acetonitrile. Absolute intensity of 500 ng/mL amphetamine, meth-amphetamine, cocaine, and benzoylecgonine on the surface of paper triangles. Quantification of characteristic fragments of cocaine (304→182), benzoylecgonine (290→168), methamphetamine (150→119) and amphetamine (136→119) was performed. These triangles were untreated and treated for 5, 30, 120, 240, 720, and 1440 minutes with silane. Error bars show one standard deviation.

Quantification was performed by spiking amphetamine, methamphetamine, cocaine, or benzoylecgonine and their internal standards separately into human whole blood at concentrations ranging from 10 to 500 ng/mL. 4 L of blood was deposited onto untreated paper triangles and paper triangles treated with silane for 30 minutes and 2 hours. The blood spots were allowed to dry for 24 hours. 3 kV was applied to the paper triangles, and 20 L ethyl acetate was applied to the triangle. Quantification of each drug was performed by analyzing the main fragment from each drug: cocaine (304→182), benzoylecgonine (290→168), methamphetamine (150→119) and amphetamine (136→119). Not only did the ion signal last approximately twice as long when hydrophobic

37 paper (30 and 120 min treated paper) was used, but limits of detection (LODs) as low as 0.12 ng/mL (corresponding to 10X reduction in LOD and LOQ for amphetamine on 30 min treated paper) were observed, including a more linear response to concentration (R2 > 0.999; Figure 2.22 – Figure 2.25 and Table 2.1). Compared with untreated paper substrates, the lower LODs calculated for hydrophobic paper are mainly attributed to: i) the inability of the blood sample to wet through the paper – the fact that the aqueous-based blood samples are unable to wet through the fiber core of the porous hydrophobic paper and spread suggests interactions between drug and the paper surface prior to extraction is decreased. This results in a greater number of free drug analytes available in the dried spheroid, increasing analyte signal; ii) the more uniform spot size for the dried spheroids – this contributes to the observed quantitative abilities (i.e., lower LODs and improved linearity) by creating a more reproducible extraction area and decreasing variations in analyte signal; and iii) the decreased analyte binding capacity to the paper post extraction. That is, redistribution of extracted analyte back into the hydrophobic paper is reduced compared with hydrophilic paper substrates.

100 (B)(b 6 (A) m/z 304 100%B) = 304 R² = 0.9835 8.47E1 untreated 3

A/IS 50

0 182

Relative Abundance Relative 0 0 100 200 300 400 500 100 200 300 400 500 Concentration Cocaine (ng/mL) m/z 5 (D)(D) 304 (C) 100 100% = 5.19E2 m/z 304 R² = 0.9993 182 30 minute treated 2.5 A/IS 50

0 0 0 100 200 300 400 500 Abundance Relative 100 200 300 400 500 Concentration Cocaine (ng/mL) m/z 5 (E) 100 (F)(F) m/z 304 304 R² = 0.9998 100% = 3.03E2 2 hour treated 2.5 50 A/IS 182

0 Abundance Relative 0 100 200 300 400 500 0 100 200 300 400 500 m/z Concentration Cocaine (ng/mL) Figure 2.22. (A, C, E) Calibration of cocaine ranging from 10-500 ng/mL in dried blood spots, and (B, D, F) representative mass spectra of fragmentation of cocaine with a concentration of 10 ng/mL on untreated paper (A and B), paper treated for 30 minutes (C and D), and paper treated for 2 hours (E and F). Mass spectra show the increased signal to noise of cocaine on paper treated for 30 minutes when compared to the untreated and 2 hour treated paper, which was expected, as shown by the optimization in Figure 2.20. Error bars show one standard deviation of trials performed in triplicate.

Figure 2.23. (A, C, E) Calibrations of benzoylecgonine in dried blood on (A) untreated paper triangles, (C) 30 minute treated paper triangles, and (E) 2 hour treated paper triangles. Error bars are one standard deviation. (B, D, F) sample MS/MS spectra from the respective paper treatments at 10 ng/mL concentration of benzoylecgonine. Error bars show one standard deviation of trials performed in triplicate.

2 (A)

1 A/IS R² = 0.9999 untreated

0 0 100 200 300 400 500 Concentration Methamphetamine (ng/mL)

2.5 (B) A/IS 30 minute treated R² = 0.9991

0 0 100 200 300 400 500 Concentration Methamphetamine (ng/mL)

1.5 (C)

2 hour treated A/IS 0.5 R² = 0.9995

0 0 100 200 300 400 500 Concentration Methamphetamine (ng/mL) Figure 2.24. Calibrations of methamphetamine in dried blood on (A) untreated paper triangles, (B) 30 minute treated paper triangles, and (C) 2 hour treated paper triangles. Error bars show one standard deviation of trials performed in triplicate.

Figure 2.25. (A, C, E) Calibrations of amphetamine in dried blood on (A) untreated paper triangles, (C) 30 minute treated paper triangles, and (E) 2 hour treated paper triangles. Error bars are one standard deviation. (B, D, F) sample MS/MS spectra from the respective paper treatments at 10 ng/mL concentration of amphetamine. Error bars show one standard deviation of trials performed in triplicate.

Table 2.1. Limits of detection (LODs) of analytes in blood, serum and urine using in-situ liquid/liquid during hydrophobic paper spray

Analyte LogP LODs (ng/mL)* in biological matrices Blood Serum Urine Cocaine 2.28 0.17 0.06 0.04 Amphetamine 1.80 0.06 0.02 0.08 Methamphetamine 2.24 0.18 0.09 0.05 Benzoylecgonine -0.59 0.13 0.08 0.2

*LODs were calculated from respective calibration curves using signal corresponding to Sblank +3×σblank; where Sblank is the average blank signal and σblank is the standard deviation of the signal from 3 replicates

2.8 Analysis of Illicit Drugs in Fresh Blood Here is presented a liquid/liquid extraction approach performed on a hydrophobic paper substrate. For complex biological samples, MS analysis of the sample droplet present on the hydrophobic paper triangle is achieved by adding an organic solvent to the paper and applying a DC potential (3 kV) (Figure 2.3C). The organic solvents suitable for in situ analyte extraction from the sample droplet must have the following properties: (i) surface tension less than that of the hydrophobic paper to allow wetting, (ii) immiscible with the biofluid sample, (iii) good solubilizing power for the target analytes, and (iv) suitable for electrospray. Ethyl acetate (surface tension 20 mN/m) satisfied these criteria113 and facilitated the analysis of both wet and dried biological samples. Contact with the organic solvent causes the fresh 4 μL sample droplet to collapse into a thin film, which allows several extraction cycles to be performed. This alternative paper spray ionization approach to biological sample analysis requires no sample preparation, where small organic molecules are extracted directly from undiluted biofluids; enzyme analysis is possible via direct detection of product without coupled reactions. The liquid/liquid extraction process occurring on the hydrophobic paper surface after application of 10 μL of ethyl acetate and 3 kV is very efficient, as demonstrated for the extraction of cocaine, methamphetamine, amphetamine, and benzoylecgonine from whole human blood, serum, and urine samples (Table 2.1). Limits of detection (LODs) as low as 0.06 ng/mL for amphetamine were recorded for undiluted human blood samples using the hydrophobic paper spray experiment. Analyte signal for LOD measurements were based on tandem MS (MS/MS) characteristic transitions. Representative product ion spectra for methamphetamine and benzoylecgonine at 3.9 ng/mL concentrations are shown in panels A and B of Figure 2.26, respectively. For urine samples, extraction efficiency correlated with hydrophobicity (log P) of the drugs (Figure 2.26C). LOD data, however, reveal that the absolute analyte amount in the organic solvent is not the limiting factor in MS sensitivity determination. For example, benzoylecgonine is a hydrophilic metabolite from cocaine with limited solubility in organic solvents (logP = −0.6) but it is detected with high sensitivity (Table 2.1). Limit of quantification ranged from 0.2−6.4 ng/mL for blood samples to 0.1−12.3 ng/ mL for serum samples to 0.1−18 ng/mL for urine samples,

43 with benzoylecgonine registering the highest quantities in all cases. For all biofluids tested, relative standard deviations less than 10% were obtained for samples with concentrations higher than 3.9 ng/mL. This performance was achieved via the use of internal standards, and monitoring analyte-to-internal standard ratios (A/IS) as a function of analyte concentration; this yielded good linearity for both wet and dried biological samples (Figure 2.27).

Figure 2.26. Tandem MS analysis of drugs extracted from urine and blood. Representative product ion spectra illustrate (A) the methamphetamine transitions 150 → 119 (primary product ion) as well as 150 → 91 (benzylic cation) and (B) the benzoylecgonine transitions 290 → 168 (primary + product ion) as well as 290 → 272 ([M+H−H2O] ). (C) Effect of the number of liquid/liquid extraction cycles on ion signal. MS/MS product ion intensities were monitored for cocaine (m/z 304 → 182), amphetamine (m/z 136 → 119), and benzoylecgonine (m/z 290 →168), each at 125 ng/mL spiked into urine (6 µL sample volume was used). MS/MS product ion intensities for corresponding internal standards at 50 ng/mL were also monitored: cocaine-d3, amphetamine-d5 and benzoylecgonine-d3.

Figure 2.27. Quantitative analysis of (A) undiluted human serum spiked with methamphetamine, and its internal standard methamphetamine-d5 (50 ng/mL) and (B) dried blood spot spiked with cocaine, and it internal standard cocaine-d3. Analyte concentration of 0.24 – 500 ng/mL was used. Error bars represent the standard deviation of analyses for three replicates with independent hydrophobic paper triangles. Ethyl acetate (10 µL for wet samples and 20 µL for dried samples) was used as the ex-traction/spray solvent.

The main determining factor for drug LOD was observed to be its protein binding capacity.114 In our experiment, only free drug molecules are detected, which are in equilibrium with protein-bound drug and drug that has diffused into red blood cells in the matrix. LOD was significantly improved for serum samples where the cellular and fibrinogen contents of blood have been removed (Table 2.1). Thus, accurate quantitation of pharmaceutical drugs or illicit compounds in biological samples can only be achieved after the free and bound drugs have been determined in a separate preparatory experiment.115,116 Another factor identified to influence the sensitivity of our liquid/liquid extraction approach is matrix effect. Direct MS analysis of raw urine using either electrospray ionization or atmospheric pressure chemical ionization (APCI) methods 45 typically do not provide sufficient sensitivity because of strong interference by the high concentration of salts and other chemicals in the mixture. The use of organic spray solvent allows the extraction of only small organic compounds while leaving behind the majority of proteins and salt components in the biological sample. This reduces ion suppression effects due to endogenous matrices and significantly enhances the ionization efficiency of the extracted analytes.

Figure 2.28. Photographs showing various stages during electrostatic-spray ionization from dry hydrophobic paper: (A) voltage off, (B) on-set of applied voltage, (C) on-set of spray, (D) stable spray formed just before 0.02 seconds, (E) stable spray

Viscosity of the sample was also observed to influence drug extraction efficiency and hence the analytical sensitivity of the method. For example, recorded LODs are comparable for drugs spiked into serum and urine (Table 2.1), although more drug molecules are expected to be available for sampling when spiked into urine. This result may be attributed to a biofluid thin film that is formed on the hydrophobic paper for more viscous samples (e.g., serum and blood). This thin film prevents further spreading of the 10 μL spray solvent and enables efficient and extended analyte extraction. Under this condition of limited solvent spreading, much of the area of the hydrophobic paper triangle

46 is dry, and yet, ion signal is observed when DC potential is applied (Videos 1 and 2, See Appendix A). During the period of the applied voltage, electrostatic charging of the dry paper causes charges to accumulate at the solution−air interface. Progeny of charged microdroplets (Figure 2.28) are observed to be emitted when the electrostatic pressure is greater than the pressure difference between the interior and exterior of the droplet (i.e., Laplace pressure).117 Further experiments with optical microscope (Videos 3 and 4, See Appendix A) have shown strong vibrations within the droplet when the potential is applied.118 This vibration is believed to further enhance the extraction process, which contributes to the high sensitivity observed for serum and blood samples. For blood and serum, sample volumes of greater than 4 μL are required to enable the formation of the solvent bridge (involving the thin film of the biofluid), which produces an ion signal over a longer period of time (Figure 2.29). (This required sample volume is dependent on paper geometry and the position of the loaded sample with respect to the paper tip; smaller droplet sizes can be used with narrower paper triangles or when samples are loaded nearer the tip and still form a solvent bridge.) In the current experiment, the 15 s signal generated from <1 μL sample is adequate for analyte characterization (inset, Figure 2.29) and will provide a method of analysis for samples of small volumes taken by minimally invasive methods.

Figure 2.29. Signal time compared to sample volume of methamphetamine (125 ng/mL) spiked in serum. All samples were extracted with 10 µL of ethyl acetate. Sample sizes of greater than 4 µL showed a significant increase in analysis lifetime. Inset: Spectrum obtained using 1 µL of human serum spiked with methamphetamine (125 ng/mL).

2.9 Analysis of Alanine Transaminase Enzyme Activity Recent interest in paper-based sensors will also benefit from the capacity to monitor enzymatic activities directly from biological samples for disease diagnosis.119,120 For example, liver transaminases (e.g., alanine transaminase (ALT)) are biomarkers for liver injury and are regularly used in laboratory tests to monitor patients under treatment for HIV and tuberculosis.121,122 With the exception of chromatographic methods, all liver function tests rely on complex and coupled reactions to enable colorimetric, spectrophotometric, chemiluminscence, and radiochemical detection of the transaminase enzyme activity.123 The hydrophobic liquid/liquid extraction method was used to monitor ALT levels in whole human blood via direct MS detection of pyruvate, one of the reaction products formed after ALT catalyzes the transfer of an amino group from L-alanine to α- ketoglutarate (Figure 2.30A).

(A)

(m/z 87)

(B) 100% = 2.22E1 (C) 100% = 1.43E2 m/z 87 87 100 41 100 m/z 87 43

41 - CO2 50 50 - H O 2 - CO 43 87 59 0

Relative Abundance Relative 0 20 40 60 80 100 20 40 60 80 100 m/z m/z

1.5 (D)

R² = 0.9961

43 I 0.5

0 140 240 340 440 Activity of ALT Enzyme (U/L)

Figure 2.30. (A) Reaction scheme for the enzymatic conversion of alanine into pyruvate catalyzed by alanine transaminase (ALT). Tandem MS of pyruvate (m/z 87) was monitored with representative spectra shown for (B) control raw blood sample without spiked ALT and (C) 400 U/L ALT spiked in blood. (D) Quantitative analysis using intensity of m/z 43 and 41 fragment ions over enzyme concentration of 150 – 400 U/L. Error bars represent the standard deviation of analyses for three replicates with independent hydrophobic paper triangles and three different assays. Ethyl acetate (10 µL) was used as the extraction/spray solvent, with 3.5 kV spray voltage.

Specifically, 30 μL blood samples previously spiked with various amounts of ALT (0−400 U/L) were mixed with 2 μL each of L-alanine (1.75 M) and α-ketoglutarate (60 mM) solutions. The mixtures were incubated for 10 min at 37 °C, and 4 μL blood aliquots were deposited on the hydrophobic paper triangle for MS analysis. Resultant pyruvate product in the blood aliquots was analyzed after applying 10 μL ethyl acetate and 3.5 kV DC voltage; representative tandem mass spectra are shown for control and 400 U/L ATL-

49 spiked blood samples in Figure 2.30B,C, respectively. Unlike chromatography, separation of the four closely related compounds (L-alanine, pyruvate, α-ketoglutarate, and L- glutamate) is not required. Instead, the use of molecular weight information and MS/MS fragmentation of pyruvate enables direct detection of ALT activity (Figure 2.30C). Amount of pyruvate produced varied linearly with ALT enzyme concentration (Figure 2.30D); this calibration curve was obtained by using the ratio of ion intensities resulting from the two distinct fragmentation pathways leading to MS/MS product ions m/z 43 (via loss of CO2, MW 44 Da) and m/z 41 (via sequential loss of CO (MW 28 Da) and water (MW 18 Da)). LOD as low as 60 U/L was recorded from this calibration curve. ALT concentrations used in this study were selected to mimic levels typical at the onset of liver injury (2 to 10-fold increase from the 40 U/L levels in healthy people).124,125 The new MS method described here, in which enzymatic products are transferred from liquid-phase to gas-phase ions through the simple exposure of biofluid to organic solvent, is an advancement through which liver function can be monitored immediately after the administration of a specific drug.

(A) 100% = 7.58E6 (B) 100% = 2.66E6

100 150 100

150

Rel. Int.Rel. Rel. Int.Rel. 0 0 100 220 340 460 100 220 340 460 m/z m/z

Figure 2.31. (A) Dry hydrophobic paper spray analysis of 5 µg/mL methamphetamine prepared in methanol/water (50/50) solution is comparted with (B) mass spectrum recorded using the traditional wet paper spray.

(A) 100% = 1.53E6 100

0 (B) 100% = 1.54E8

100 Relative Abundance Relative 50

0 0 0.3 0.6 0.9 1.2 1.5 1.8 Time (min)

Figure 2.32. The electrostatic-spray generated from a 4 µL solution of 5 µg/mL methamphetamine (full MS) in MeOH/H2O (50/50) using (A) a wetted paper triangle charged at 3 kV lasted for 30 seconds compared with (B) electrospray time of 2 minutes from a dry hydrophobic paper triangle. Representative mass spectra recorded for both experiments (electrostatic-spray from dry hydrophobic paper, and electrospray from wet hydrophilic paper) are as shown in Figure 2.31A and B, respectively

2.10 Direct Analysis of Nonbiological Samples The dry electrostatic paper spray phenomena observed here represents a simpler and more sensitive ambient ionization method for field analysis. More than 1 order of magnitude enhancement in ion intensity and signal-to-noise ratio (Figure 2.31A and B), and two times signal lifetime (Figure 2.32) are achieved with dry hydrophobic paper spray for nonbiological aqueous-based samples – in this case, no extra spray solvents are required. The redox active myoglobin protein was analyzed to determine the occurrence of corona discharge126 when the dry hydrophobic paper is charged at 3 kV. Such an oxidation process usually ends in the formation of covalent oxygen (+16 Da) adducts. The performance of the dry electrostatic-spray experiment was compared to traditional wet PS and nanospray experiments. Oxidation peaks were absent in all cases (Figure 2.33), but there were substantial differences in observed peak widths and charge-state distributions (CSD) between the three techniques. The electrospray-based methods (i.e., paper spray and nanospray) demonstrated comparable CSD, but higher myoglobin charge state was

51 detected from electrostatic-spray on dry hydrophobic paper. This higher charge state is attributed to to droplet vibration, which can cause further unfolding of the protein. Compared with untreated hydrophilic paper, the expected increase in hydrophobic interaction between the protein and hydrophobic paper can reduce the ease of desorption, and cause myoglobin to unfold. Adducted alkali metal cations were detected in all three cases, but wet PS showed broader peak width (i.e., drifted baseline), indicating higher currents during low solvent flow rates.66 A sharp onset voltage of 3 kV (applied over 3 mm distance) was observed for dry hydrophobic paper spray (Figure 2.34), and ion signal was strongly dependent on applied voltage (Figure 2.35). Because no ion signal was observed in the absence of the applied voltage, suction (pressure) effects were eliminated due to the MS vacuum during ionization.127 At least 3 kV of spray voltage is required, below which no signal is observed (Figure 2.34). Intensity of ion signal showed strong dependence on applied voltage (Figure 2.35). That is, there was instantaneous jumping of the droplet when ≥4 kV of DC voltage was applied.

(A) Untreated Paper 17+ 100 17+ +K 19+ +(Na+2K) 15+ heme 21+ Rel. Int.Rel. 23+ 13+ 0 600 800 1000 1200 1400 (B) Treated Paper 17+ 100

+(Na+2K) Rel. IntRel.

0 600 800 1000 1200 1400 (C) Nanospray 17+ 100

+Na Rel. Int.Rel. 0 600 800 1000 1200 1400 m/z

Figure 2.33. Analysis of denatured myoglobin (25 µM in 0.5% acetic acid in methanol/water, (50/50) solution) using: (A) traditional untreated paper spray, (B) dry hydrophobic (treated) paper spray, and (C) nanospray. Spray voltage was 3 kV for both of the paper spray experiments, and 1.8 kV for nanospray

3 kV 0 kV

100 Rel. Int.

0 0 2.0 4.0 Time (min)

Figure 2.34. Extracted ion chromatogram for amphetamine (m/z 150) showing the dependence of ion signal on applied voltage.

2.0E+05 3 kV 1.6E+05 4 kV 5 kV 1.2E+05

8.0E+04

4.0E+04 Absolute Ion Intensity Ion Absolute

0.0E+00 180 181 182 183 184 m/z

Figure 2.35. Effect of electrostatic-spray voltage on signal intensity. Fragmentation of cocaine (125 ng/mL) in MeOH/H2O (50/50) was monitored in MS/MS experiment using the major fragment ion at m/z 182. Superimposed MS/MS mass spectra are plotted in the mass range m/z 180 – 184.

2.11 Surface Energy Analysis Efficient and widespread application of new analytical methods depend on proper characterization of experimental parameters. In this study, the specific surface energy of the prepared hydrophobic paper is the most important parameter, and if not controlled, can lead to inconsistencies in spheroid size causing variations in analyte stability in a specified blood volume. Contact angle measurements with water is the most prominent method used in estimating surface energies of planar substrates. DI water was deposited on the surface of filter paper treated for 0, 5, 30, 120, 240, 720, and 1440 minutes. The contact angle of this water drop was observed using a Rame-Hart goniometer. The contact angle corresponds to the relative surface energy per area of the paper when compared to the surface tension of the water. If the surface energy of the paper exceeds the surface tension of the water (72 mN/m), the water will completely wet the paper and the contact angle will be 0°. If the surface energy does not exceed the surface tension of the water, the water drop will bead up, and the contact angle between the water drop and the paper will be some angle θ.

However, for as anisotropic surfaces like paper, contact angle measurements yielded inconsistent results.128 As seen in Figure 2.36, the contact angle θ was 0° for untreated paper and paper treated for 5 minutes. For paper treated for 30 minutes or greater, the contact angle θ was approximately 125°. These results indicate that paper treated for 5 minutes or less have a surface energy per area greater than 72 mN/m, but is unable to specify the actual surface energy. It is also surmised that paper treated for 30 minutes or greater have surface energies less than 72 mN/m, and treatment time does not lower surface energy past a certain plateau point that occurs at or before 30 minutes of treatment. These results do not agree with experimental observations.

140 120 100 80 60 40

Contact Angle Angle (degrees) Contact 20 0 0 300 600 900 1200 1500 Treatment Time (minutes)

Figure 2.36. Contact angle of DI water deposited onto filter paper with varying treatment times of vapor phase silane. Error bars show one standard deviation.

Therefore, a novel electrostatic spray-based method was developed in which a simple multimeter is used to measure total ion current or via selected ion monitoring by MS. Here, solutions consisting of water/acetonitrile mixtures were prepared. The specific proportion of water/acetonitrile used was varied, which yields known surface tensions for each prepared solution (Table 2.2). Acetonitrile/water mixture surface tensions were originally determined by Rafati et. al.129 These solution mixtures were used as spray 55 solvents in electrostatic spay where neat benzoylecgonine analyte dried on the hydrophobic paper was ionized in the process.

Table 2.2. Reported surface tensions of acetonitrile/water mixtures.

Solvent number Mole Fraction Mole Fraction of Does the solvent wet Literature Surface of Acetonitrile Water 2-hour treated paper? Tension (mN/m)130 1 0.0149 0.9851 No 62.36 2 0.0298 0.9702 No 55.92 3 0.0516 0.9484 No 49.39 4 0.0950 0.9050 No 40.54 5 0.1227 0.8773 Partially 37.97 6 0.2541 0.7459 Yes 32.92 7 1 0 Yes 29.3

Because the electrostatic spay ionization of dry samples is a function of solubility and wettability, a maximum ion signal was anticipated to be recorded when the surface tension of the spray solvent approximately equaled the surface energy of the hydrophobic paper. This expectation has been met (Figure 2.37A). Maximum/peak ion currents were observed at solvent surface tensions of 38, 40, and 52 mN/m for hydrophobic paper substrates prepared by 4 h, 2 h, and 30 min silane exposure times, respectively. Polymeric membranes of known surface energies were also employed: cellulose acetate (37 mN/m), polycarbonate (44 mN/m), and polyacrylonitrile (PAN) (48 mN/m). The corresponding peak currents were observed at 33, 40, and 49 mN/m (Figure 2.37A), respectively, which correlated well with the known surface energies of the membranes. These results suggest that the position of the peak current may be used to determine the surface energy of the paper/membrane from which the electrostatic spray is derived. Therefore, the three membranes were used as standards in order to estimate the unknown surface energies of treated paper (Figure 2.37B). Through this interpolation, the surface energies of the as- prepared hydrophobic paper substrates were estimated to be 41, 42.5, and 51 mN/m for 4 h, 2 h, and 30 min treatment times, respectively (Table 2.3).

Table 2.3. Surface energy determination using peak surface tension solvent.

Surface name Reported Surface Peak Surface Tension Calculated Surface Energy (mN/m) Solvent (mN/m) Energy (mN/m) Cellulose Acetate 37 33 -- Polycarbonate 44 40 -- Polyacrylonitrile 48 49 -- 4 Hour Treated Paper -- 38 41 2 Hour Treated Paper -- 40 42.5 30 Minute Treated Paper -- 44 51

(A) 30 minute treated 1.0 2 hour treated 4 hour treated Cellulose Acetate 0.5 Polycarbonate

Polyacrylonitrile Relative Intensity Relative 0.0 25 35 45 55 65 (B) Surface Tension of Solvent (mN/m)

60 30 Min Treated /m) PAN

mN 40 Polycarbonate

Cellulose Acetate Peak Surface Surface Peak

Tension ( Tension 20 35 40 45 50 55

Surface Energy (mN/m) Figure 2.37. (A) Observation of ion intensity varying with the change of surface tension of ACN/H2O spray solvents (Table 2.2). Peak surface tensions are used as values for y-axis in plot B. (B) Calibration of cellulose acetate and polycarbonate, with treated and untreated paper projected onto the line. The determined surface energies of paper substrates are provided in Table 2.3.

Optimum ion current is expected when solvent surface tension is approximately equal to the surface energy of the paper/polymer surface. Therefore, evaporation rate post- Taylor cone is negligible in determining the solvent surface tension that yield highest ion signal, when compared to the effect of wetting and slight partitioning effects. Three regions in Figure 2.37A can be distinguished: (1) region before the maximum current, involving solvents with lower surface tension than surface energy of the surface, (2) the point at which the ion current is maximum or peaks; the corresponding solvent surface tension is expected to equal the surface energy of the paper substrate, and (3) region after the peak current where solvent surface tension is great that surface energy of paper. If the partition coefficient is defined as:

푚푠표푙푣푒푛푡 퐾 = ⁡ 푚푝푎푝푒푟 Equation 4

where msolvent is the moles of the target analyte in the spray/extraction solvent, and mpaper is the moles of the target analyte on the paper surface. Then, each of the regions will have the following properties:

(1) Low surface tension solvent, high degree of wetting, high degree of paper-solvent- analyte interaction resulting in possible redistribution of analyte back into the paper substrate post-extraction, high evaporation rate (because of spreading), very small K value

(2) Solvent for the peak ion current must have surface tension that allows intermediate wetting, less paper-solvent-analyte interaction and less analyte re-deposition post- extraction. It will also have moderate evaporation rate and large K value

(3) High surface tension solvent, low degree of wetting, low degree of paper-solvent- analyte interaction resulting in low amount of extraction and low amount of re- deposition, low evaporation rate, and small K value

Using this logic, and using fitting parameters to correct for a changing K value, the following empirical equation is proposed to account to the shape of the ion current observed in Figure 2.37A:

푎 퐾 퐼 ≈ |푆퐸2 − 훾2| + 푏 훾 − 푐 Equation 5

where I is ion intensity; a, b, c are fitting parameters; K is the partition coefficient; SE is the surface energy of paper; and γ is the surface tension of solvent. By setting K to 1 and SE as 42.5 mN/m, this equation was fitted to the data collected for 2 hour treated paper and the result is shown below showing a good fit between theoretical and experimental data. (Figure 2.38 and Figure 2.39)

Figure 2.38. Data found on Figure 2.37A fitted with Equation 5. Fitting parameters were found to be: a = 210.3 m2/mN2, b =62.9 mN/m, c = 2497.1 mN2/m2

The shape of this function, when plotted with these same fitting parameter values, is shown in Figure 2.39. This shape/function implies there is an actual peak surface tension that is approximately equal to surface energy of the paper and is possible to be found experimentally.

1.2

1.0

0.8

0.6

Relative Intensity Relative 0.4

0.2

0 25 30 35 40 45 50 55 60 Surface Tension of Solvent (mN/m)

Figure 2.39. A Mathematica plot of Equation 5 using parameters found for fitting in Figure 2.38.

Videos showing various spray modes occurring when surface tension of a spray solvent is varied (Videos 5-8, See Appendix A) were recorded, which ranged from Taylor cone formation (wetting solvents) in electrospray to droplet vibration (non-wetting solvents) in electrostatic-spray (Figure 2.40). 20 mL of acetonitrile/water solutions of varying ratios were deposited onto the front of paper strips that had been previously treated for 2 hours each with trichloro(3,3,3-trifluoropropyl) silane. 5 kV DC voltage was applied to the back of the paper strip (away from the solvent), and the strip was pointed toward a mass spectrometer inlet.

(A) (B)

Figure 2.40. Acetonitrile/water droplets of varying ratios (see Table 2.2) resting on a paper strip treated for 2 hours when 5 kV is applied. (A) Droplet consists of solvent 7 (pure acetonitrile, surface tension 29 mN/m). (B) Consists of solvent 5 (surface tension 38 mN/m). (C) Consist of solvent 4 (surface tension 41 mN/m). (D) Consists of solvent 1 (surface tension 62 mN/m).

We note that some of the acetonitrile/water solvent mixtures used for surface energy estimations did not wet the hydrophobic paper, and yet ions were detected. Under this non-wetting condition, an electrostatic spray mode has been proposed,94 where capacitive charging at droplet surface causes analyte ions to oscillate. Ions break free from the liquid droplet surface at a sufficiently high kinetic energy, determined by the applied DC voltage (onset voltage determined to be 3 kV). To evaluate the effect of surface tension of the solvent on this electrostatic spray mode, a camera was used to image the spray dynamics of solutions comprising of varying mole ratios of acetonitrile in water. Figure 2.40 (see Videos 5-8, See Appendix A) reveals droplet oscillation was reduced with decreasing solvent surface tension, from Figure 2.40C to Figure 2.40A, at which point spray mode becomes electrospray. The corresponding ion current data (Figure 2.37A) indicate low abundance of ion yield was recorded for both very high and low surface tension solvents. This is because solvent with high surface tension is less likely to form a stable Taylor cone while low surface tension solvent suffers from unfavorable partitioning.

Figure 2.41. (A) Cellulose, (B) Polycarbonate, (C) Cellulose Acetate, (D) Polyacrylonitrile.

2.12 Ionization Mechanism Oscillating droplets have been under intense investigation for more than a century. Mathematical model for non-viscous droplets in vacuum was first described by Rayleigh,131 which was later generalized by Lamb to include the influence of a surrounding medium.132 A brief discussion on this subject is presented here to explain why vibrations were observed in Videos 3 and 4, See Appendix A. Surface charges of a droplet cause an outward electrostatic pressure acting in opposition to the inner pressure, which arise from surface tension. Forces resulting from surface tension stabilize the droplet shape, while the electrostatic force resulting from the charge disintegrate the droplet. With increased charge, the droplet becomes instable. The droplet can be considered as a spherical electrical capacitor, with a total potential energy of

2 푄 2 푊푡표푡 = + 4휋휎푟 8휋휖0푟 Equation 6

where the first term in the above equation represents electrostatic energy corresponding to charge Q. ε0 is permittivity in vacuum, and r is radius of droplet. The second term in the equation represents surface energy where σ is the surface tension.

Total energy (Wtot) is minimum at Rayleigh radius

3 푄2 √ 푟푅푎푦 = 2 64휋 휀0휎

Equation 7

The charge at Rayleigh limit is

3 푄푅푎푦 = 8휋√휀0휎푟 Equation 8

By considering Stoke’s law (i.e., QE = 6πµrv, where µ is the dynamic viscosity), the mobility b of a charged particle moving with a velocity v in an electric field E is defined by

푣 푄 푏 = ⁡ = 퐸 6휋휇푟 Equation 9

Maximum mobility is obtained at Rayleigh limit, and ejection of highly charged tiny droplets occurs.

2.13 Summary In summary, by using a hydrophobic paper substrate, a dried blood spheroid collection platform was established that can potentially eliminate chromatographic/volcanic effects associated with traditional 2D dried blood spot samples. The dried blood spheroid sample collection platform showed increased stability for hydrolytically labile compounds against oxidative stress, increasing the lifetime of diazepam, cocaine and benzoylecgonine (the main metabolite of cocaine), from days to several weeks under ambient conditions and without cold storage. Through manipulation of the surface energy of the paper and the use of organic spray solvent, selective extraction of target analytes may be performed, which allows enhanced PS MS detection of cocaine, benzoylecgonine, amphetamine, and methamphetamine from the dried blood spheroids, resulting in sub-ng/mL limits of detection. Tuning solvent surface tension allows a more sensitive determination of surface energy of the porous hydrophobic paper substrate compared to conventional tensiometric (contact angle) measurements. This novel electrostatic method employed a simple multimeter detector. Because of its close resembles DBS, the implementation of dried blood spheroid sample collection in clinical settings can be accomplished with no changes in blood collection procedures. Direct analysis of small organic compounds was demonstrated in biological samples by performing liquid/ liquid extraction on a hydrophobic paper surface. In various biomolecular applications, this technique shows the advantages of (1) greatly facilitating sample loading and extraction from paper and potentially increasing the sensitivity of PS ionization, (2) sample economy (compared with SPE methods), and (3) compatibility with raw blood, serum, and whole urine. Like the traditional PS experiment, the hydrophobic paper spray methodology represents a simple approach to reduce sample complexity compared with the traditional liquid- or gas chromatography MS. The hydrophobic paper spray methodology was applied for the direct quantitation of illicit drugs in both wet and dried biological fluids, and with high sensitivity and precision. Both viscous and non- viscous biofluids can be analyzed−the solvent bridge (without wicking) formed with viscous samples yielded a novel form of PS in which progeny of charged droplets are generated from the dry hydrophobic paper. This electrostatic-spray experiment is capable

65 of ionizing both small organic and large protein analytes using either positive or negative ion mode and without the need for extra spray solvents and nebulizer gases. This feature is particularly attractive for field analysis using hand-held mass spectrometers, including point-of-care and clinical applications. For example, this work has shown that the electrostatic paper spray methodology can be an efficient, rapid, and sensitive way to diagnose the onset of liver injury by monitoring the conversion of L-alanine to pyruvate by alanine transaminase enzyme, in the presence of α-ketoglutarate. When fully developed, the approach presented here could eliminate traditional coupled reaction detection strategies for liver function test which require complex setup and expertise. The biodegradable paper substrate can also serve as a container for on-demand ambient ionization by exposing the sample droplet present on the hydrophobic paper to discrete charge pulses.133

Chapter 3. 2D Wax-Printed Paper Spray Ionization

3.1 Introduction Microfluidic paper-based analytical devices (μPADs) have emerged as a promising technology to develop simple, low cost, portable, and disposable detection platforms for resource-limited settings.90,134,135 As a variant of conventional microfluidics typically constructed from glass or polymer substrates,80,136,137 μPADs are made from paper. The use of paper brings three mains benefits in microfluidic platforms: (i) it serves to pump fluids passively across the device by capillary wicking and eliminates external pumps that are typically not micro in size; (ii) like other porous substrates, paper allow the analysis of complex samples with little or no sample preparation; and (iii) using a paper substrate enables on-chip/on-surface detection. In effect, the ideal μPAD analytical systems are self- sustaining with all components necessary to perform an analytical assay (e.g. sample transport, sample pre-treatment, assay reagents, and signaling system) integrated into the device. Since their inception, μPADs have been utilized mainly for healthcare-related diagnostics;81,120,138–145 subsidiary applications have included environmental monitoring,146–150 explosive detection151–154 and detection of food contaminants.141,147,155– 157 In order to maintain simplicity and portability, low power detection techniques, such as photometric, electrochemical, electrical conductivity, chemiluminescence, and electrochemiluminescence have been used for analyte signal transduction.158–164 Unfortunately, detection limits of μPADs based on these detection platforms are often inadequate and give only semi-quantitative results. The motivation behind this work is to generate next-generation μPADs that can enable the direct, rapid, sensitive, and on-surface detection of a variety of analytes – including small organics and large (kDa) biomolecules – through the use of handheld mass spectrometers. Specifically, interest lies in the establishment of protocols that allow on-surface splitting of analyte solution for multiplexed detection, rapid immuno-extraction and

67 concentration of analyte, and ambient paper spray (PS)91,165 ionization mass spectrometry (MS), all from a single paper device.70 In pursuit of this significant goal, the current study focuses on the optimization and characterization of small molecule ionization from 2D wax-printed paper substrates using low spray voltages under ambient conditions. Solid wax printing represents an efficient approach to create microfluidic channels on paper in which the working hydrophilic regions are surrounded by hydrophobic wax barriers.142,166 The well-defined hydrophilic channels generated from the wax-printing methodology are hypothesized to be able to be utilized to confine DC potentials in narrow regions of the paper for efficient use of electrical power. Unlike other ambient ionization techniques,23,29,39,167–173 PS is well suited for on-site in situ sample analysis, because no pneumatic assistance is needed to transport the analyte to the inlet of the mass spectrometer. Transfer of analytes occurs when the sample present on the paper substrate is solubilized by applying a spray solvent, which is typically a methanol/water mixture (1:1, v/v). Under this condition, charged micro-droplets are emitted from the tip of the wet paper triangle after applying 3–5 kV DC voltage to the paper triangle. Performance of PS-MS has been shown to depend on the geometry of the paper.99 Recently, Colletes et al. described a novel PS-MS approach in which paper substrates with paraffin hydrophobic barriers (prepared by manual stamping of microfluidic shapes) allowed ∼10× increase in PS signal from mono- and disaccharides.174 The present study shows that electronic solid wax-printing is an efficient way to optimize the generation of hydrophobic barriers for use in PS-MS. The resultant wax-printed paper substrates enable (i) the use of lower voltages (0.5 kV) for analyte ionization, and (ii) extended signal lifetime which permits both improved signal averaging and provides an increased timeframe to monitor multiple fragmentation/reaction pathways in tandem MS (MS/MS) experiments. Chemical analysis at zero volts from plain paper substrates was recently reported by the Cooks’ group175 in a solvent assisted inlet ionization176 experiment. PS-MS of various analytes was also achieved by spraying analyte solution from a carbon nanotube- impregnated paper substrate under the influence of only 3 V.58 In both cases, however, high quality mass spectra were obtained at ppm analyte concentrations. Attempts to increase analyte signal beyond the typical 1 minute lifetime93 have included the use of excess spray

68 solvent,101 a hydrophilic wick,177 and by connecting the paper substrate to a solvent reservoir created in 3D printed cartridge178 or the nib of a fountain pen.179 In the present study, 2D solid wax patterns were created and optimized on paper that allowed sensitive (at sub-ppb levels) detection of illicit drugs (e.g., methamphetamine cocaine, amphetamine and benzoylecgonine), corrosion inhibitors (e.g., Duomeen), and pesticides (e.g., metaldehyde) in water samples using a spray voltage ≥0.5 kV. Up to 5× increase in sensitivity was achieved when analyzing dried and wet urine samples at 3 kV from the 2D wax-printed paper substrate compared with analysis done on a plain/un-waxed paper surfaces. In this case, analyte signal lifetime lasted up to 6 times longer for wax-printed paper spray compared with the traditional paper spray experiments.

3.2 Materials and Methods Chemicals: Standard solutions (1.0 mg mL−1) of cocaine, methamphetamine and amphetamine were obtained from Cerilliant (Round Rock, TX). Methanol, metaldehyde, and atrazine-d5 were purchased from Sigma-Aldrich (St Louis, MO). N-Oleyl-1,3-diaminopropane (Duomeen) was supplied by B&V Water Treatment, (Lamport Drive Heartlands Business Park Daventry Northamptonshire, NN11 8YH, UK) through the Department of Electrical Engineering and Electronics, Liverpool University.

Wax printing: A Xerox ColorQube 8870 wax printer (Norwalk, CT) was used to print patterns on Whatman (Maidstone, UK) grade 1 cellulose chromatography paper. The paper was then heated to 130 °C for 30 seconds to allow the wax to permeate the paper fibers. All wax printed and un-waxed triangles were cut into approximately 70 mm2 triangles (9 mm base and 16 mm height).

Mass spectrometry and current measurement: Samples were analyzed by a Thermo Fisher Scientific Velos Pro LTQ linear ion trap mass spectrometer (San Jose, CA, USA). MS parameters used were as follows: 150

°C capillary temperature, 3 microscans, and 60% S-lens voltage. Thermo Fisher Scientific Xcalibur 2.2 SP1 software was applied for MS data collecting and processing. Tandem MS with collision-induced dissociation (CID) was utilized for analyte identification. Current was measured in a separate experiment using a Keithley 485 autoranging picoammeter (Cleveland, OH).

(A) (B) Back of paper: (i) wax printing

Wax layer

4 mm

heat 130⁰C (ii) 30 seconds

4 mm Apply cut sample/solvent 300 m

(C-ii) HV MS

wax-printed paper spray mass spectrometry 300 m Figure 3.1. Procedure for preparing wax-printed paper.

3.3 Wax Pattern Optimization Since, for a given potential, electric fields are more intense for small objects (i.e., where there is a smaller distance between the applied potential and a charge), solid hydrophobic wax material was applied on chromatographic paper in order to control the area of paper substrate to be wetted by the spray solvent. In theory, manipulation of

70 available area for wetting should be able to alter the electric field generated at the tip. Common aqueous-based electrospray solvent compositions containing up to 80% organic (e.g., methanol) and 100% acetonitrile could not penetrate the wax hydrophobic barrier, and are therefore suitable spray solvents for ionization of various organic analytes in the wax-printed paper spray methodology. The process for generating the 2D wax-printed paper triangles and subsequently using the resultant paper triangles in PS-MS is as illustrated in Figure 3.1. Figure 3.1A shows computer generated patterns printed onto filter paper, followed by heating at 130 °C for 30 seconds to melt the wax; insert (i) shows the back of the filter paper after wax printing, but before heating and insert (ii) shows the back of the filter paper after heating the wax-printed paper. Finally, the wax-printed paper is cut into a triangle and used in paper spray ionization; MS inlet was grounded. Figure 3.1B shows the cross-section illustrations of paper before printing, after printing, and after melting, respectively. Printed wax rests on the surface of one face of the filter paper. Subjecting the paper to 130 °C causes the wax to melt into the paper pores, permeating the paper. Figure 3.1C shows SEM images of the wax-paper barrier: (C-i) printed wax before melting, (C-ii) printed wax after melting. No morphological differences were observed in waxed areas after heating when compared to un-waxed areas.

(A)

t0 t1 t2 t0 t1 t2

s1s1 ss22 s3s3

(B)

waxless 2E+6 s1

s2 I s3 1E+6 t1 t2

0E+0 Amphetamine Cocaine Methamphetamine

Figure 3.2. (A) Wax-printed micro-fluid channels/patterns tested (B) Comparison of absolute intensity of major fragment ions. I = absolute fragment ion intensities

600

400

Volt (V) 1 kV 3 kV Area (mm2) 200 5 kV

0 0% 56% 54% 71% 60% 52% t0 t1 t2 s1 s2 s3 % area of paper with solid wax

Figure 3.3. Charge density (V/mm2) is shown to increases as the area available to solvent decreases.

First, five different designs of computer generated microfluidic channels were constructed using adobe illustrator (t1 to s3, with t0 representing a plain, un-waxed paper substrate; Figure 3.2; t0 represents un-waxed paper whereas t1 and t2 have tapered channels (blue); s1–s3 represent straight channels with increasing width from s1 to s3). The shapes of the channels (blue regions in triangles) were made to (i) have different total wax coverage on the paper (Figure 3.3), (ii) represent two main groups of geometries: tapered (t1 and t2) and straight (s1, s2, s3) channels, (iii) vary the channel width within each group (width of t2 > t1 at the paper tip; width of s3 > s2 > s1; see details in Figure 3.4), and (iv) allow comparison between tapered and straight channels with respects to channel width and percent area of paper covered by solid wax. As equal voltage is applied to each paper triangle pattern, available area shape and magnitude varies, which alters the charge density (given here as volt per area). When the available voltage per area was analyzed (Figure 3.3), it is predicted that charge density should be greater for paper with printed wax patterns. Detail geometrical differences of all patterns is provided in Figure 3.4.After printing the desired channel onto the paper, the printed paper sheets were heated to melt the wax (Figure 3.1A), which then spreads through the fiber core of the filter paper to produce hydrophobic barriers that extend through the thickness (180 μm) of the paper and effectively confines the spray solvent. After melting, no morphological difference is seen between wax and un-waxed regions (Figure 3.1C), indicating a uniform spread of wax in the paper matrix.

s t t0 1 s2 s3 1 t2

a = 8.0 a = 7.0 a = 6.0 a = 7.9 a = 7.3 a = 16 b = 2.7 b = 2.5 b = 2.4 b = 2.1 b = 2.1 c = 8.7 c = 0.51 c = 1.2 c = 1.8 c = 2.0 c = 2.1 d = 1.4 e = 2.5 g = 0.71 f = 2.0 g = 1.2

Figure 3.4. Detail geometrical differences between all wax-printed patterns is provided. All parameters are given in millimeters (mm). Overall geometry of all paper triangles are approximately measured 9 mm base and 16 mm height.

Interestingly, the performance of the wax-printed paper substrates did not follow any particular pattern when utilized in PS-MS analysis of amphetamine (MW 135, log P 1.80), cocaine (MW 303, log P 2.28), and methamphetamine (MW 149, log P 2.24) using methanol/water (1:1, v/v) spray solvent charged at 3 kV. The results of these experiments are summarized in Figure 3.2B (amphetamine (m/z 136 → 119), methamphetamine (m/z 150 → 119), and cocaine (m/z 304 → 182) sprayed from a normal un-waxed paper triangle with the corresponding wax patterns generated shown in (Figure 3.2A). Wax pattern s3 was selected for further analysis.), which indicate that the wax-printed substrates generated relatively higher absolute ion signal compared with the plain paper substrate. When comparing the performance among the wax-printed substrates, s1 paper substrate (channel width 0.51 mm) was inferior in all cases compared with s2 (channel width 1.2 mm) and s3 (channel width 1.8 mm). These results encouraged us to further optimize the two best wax-printed substrates s2 and s3 for PS-MS experiments using lower spray voltages. Here, the wax-printed paper substrate s3 was superior to s2 and un-waxed (t0) paper substrates at all potentials tested, including 0.5 and 1 kV spray voltages (Figure 3.5). Based on these results, the s3 pattern was chosen to be used for all other experiments 74 discussed in this study. At all voltages, s3 wax-printed paper substrate produced higher signal intensity than un-waxed paper.

3E+5

800 waxless

400 s2 2E+5 (A) 0 s3 3E+5 0 0.5 1 1.5

2E+5 1E+5 I 1E+5 0E+0 0E+0 0 1 2 3 4 0 1 2 3 4 (B) 300 4E+5 150 3E+5

0 I 2E+5 0 0.5 1 1.5 1E+5

0E+0 0 1 2 3 4

1E+6 0 I 0 0.5 1 1.5 5E+5

0E+0 0 1 2 3 4 Spray Voltage (kV)

Figure 3.5. Comparison of s2, s3, and t0 (waxless) paper triangles as a function of voltage using (A) amphetamine, (B) cocaine, and (C) methamphetamine diluted in water at 250 ng/mL. Error bars show standard deviation for three replicates. I = absolute product ion intensities.

(i) 20 µL solvent on un-waxed paper 12 (ii) 8 waxWax 4 waxlessUn-waxed 0 0 5 10 15 20 25 0.2 0.6 1.0 1.4 1.8 (min) Lifetime Signal Solvent Volume (L)

20 µL (iii)

0.3 0.7 1.1 Volume of Spray Solvent

10 µL

4 µL 0 1 2 3 4 5 6 7 8 9 10 Time (min)

Figure 3.6. Selected ion (m/z 304) chromatogram (XIC) of 100 ng per mL cocaine solution; 4 μL sample was dried onto s3 wax-printed paper and sprayed with increasing volumes of MeOH/H2O (1:1, v/v) at 3 kV. Arrows show where each signal ceased. Inset (i): 4 μL of 100 ng per mL cocaine solution was dried onto an un-waxed paper triangle and sprayed with 20 μL of MeOH/H2O (1:1, v/v). Spray time was approximately 1.5 minutes (0.2–1.7 minutes). Inset (ii): signal lifetime varies with spray solvent volume applied to the paper triangle. Un-waxed paper signal lifetime does not increase after approximately 7 μL of solvent, but wax-printed paper signal lifetime increases to ∼10 minutes after 20 μL is added to the triangle. Inset (iii): zoomed-in XIC in 0.2–1.3 minute time range when using 10 μL spray solvent.

3.4 Characterization of Spray from Wax-Printed Paper Substrate Before applying the selected wax-printed paper substrate (s3) in PS-MS for real sample analysis, differences in spray dynamics and loading capacity of wax paper were compared with the traditional paper spray experiment. For this, 4 μL of cocaine (100 ng mL−1) prepared in methanol was deposited onto the wax-printed paper triangle, allowed to dry in ambient air after which the dried cocaine was sampled using methanol/ water (1:1, v/v) spray solvent. Typical selected ion (m/z 304) chromatograms generated in this experiment are shown in Figure 3.6 for varying the volumes of the methanol/water spray solvent, applied in direct dumping mode.101 In each case, 3 kV of spray voltage was used.

Using the typical 10 μL spray solvent, the signal from the wax-printed paper substrate lasted for more than 1.6 minutes. This spray lifetime was substantially extended to 10 minutes when 20 μL of solvent volume was used. A maximum solvent loading capacity of 25 μL was determined for the paper size (9 mm × 16.5 mm) used in this experiment (insert (ii), Figure 3.6). A similar experiment was performed on un-waxed paper; here, signal lasted for only 1.5 minutes when using 20 μL spray solvent (insert (i), Figure 3.6). Initial ion currents from the un-waxed paper triangle were two times higher than that from the wax-printed paper substrate. This observation suggests a higher flow rate, but the subtle differences in ion currents alone cannot explain the rapid depletion of solvent from the un-waxed paper. It is believed that solvent loss/depletion attributed to evaporation and electrospray are more effectively controlled when using the wax-printed paper triangle. The high surface area available in un-waxed paper enables easy spreading which enhances rate of solvent evaporation. Reducing the surface area of the paper triangle with solid wax printing allows a more stable electrospray for extended time by reducing solvent evaporation. The flow rate for the electrospray at the wax-printed paper tip varied as the solvent was consumed. Solvent flow was high at the onset of voltage application, which lasted for only a few seconds followed by a long stable spray period (insert (iii), Figure 3.6).

(i) 164 (ii) 164 (iii) 100 100 100 164

50 50 50

163 165 163 165 Abundance Relative 165 Relative Abundance Relative 163

Relative Abundance Relative 166 166 166 0 0 0 146 152 158 164 170 176 146 152 158 164 170 176 146 152 158 164 170 176 m/z m/z m/z 100

Relative Abundance Relative 0 0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 2.7 Time (min) Figure 3.7. Total ion chromatogram recorded after spraying of 1 ppm 6-methoxy-1,2,3,4- tetrahydroquinoline with 10 L 4:1 MeOH/H2O spray solvent at 3 kV. Inserts shown mass spectra taken from (i) the beginning of spray lifetime, (ii) stable spray region, and (iii) increased spray current region near the end of spray lifetime. No significant difference is seen in the mass spectra.

A spike in ion current was consistently detected towards the end of the spray (Figure 3.7); this observation is in accordance with a recently proposed paper spray mechanism66 in which corona discharge is thought to contribute to ion formation at lower solvent flow rates, typically occurring at that point where solvent is depleted. That no changes in ion type (protonated species [M+H]+ at m/z 164 versus radical cations M•+ at m/z 163) was detected in our wax-printed paper spray experiments when using 3 kV of spray voltage suggests minimal contribution from corona discharge (Figure 3.7).

3.5 Analysis of Illicit Drugs Illicit drug quantitation using the wax-printed paper spray method was first optimized using pure methanolic solutions of methamphetamine. Standard solutions were prepared in 3–250 ng mL−1 range, and analyzed both with un-waxed and wax-printed paper substrates using 0.5 and 1 kV spray voltages. All attempts to generate a linear calibration curve from unwaxed paper triangle at these low spray voltages failed.58 In contrast, good linearity and high precision were easily recorded for methamphetamine when analyzed at the same voltages using wax-printed paper substrate (Figure 3.8A and B). Detection limits

(LODs) as low as 0.36 ng mL−1 and 2.53 ng mL−1 were recorded for 1 kV and 0.5 kV spray voltages, respectively. This performance was achieved in MS/MS mode using m/z 150 → 119 transition and with the use of internal standard (methamphetamine-d5, with MS/MS transition m/z 155 → 121), and monitoring analyte-to-internal standard ratios (A/IS) as a function of analyte concentration. Representative MS/MS product ion spectra for methamphetamine at 250 ng mL−1 concentration are shown in Figure 3.8C and D for 0.5 kV and 1 kV spray voltages, respectively.

(A) (C) 3 119 100 100% = 7.04E1 m/z 150

RI A/IS 1 R² = 0.9987 150 91 0 0 0 50 100 150 200 250 50 90 130 170 m/z (B) (D) 119 100 100% = 1.94E1 2 m/z 150

1 R² = 0.9991 RI A/IS 150 91 0 0 0 50 100 150 200 250 50 90 130 170 Concentration Methamphetamine (ng/mL) m/z

Figure 3.8. Calibration of methamphetamine standard solutions (3–250 ng mL−1) analyzed with MeOH/H2O (1:1, v/v) solution using (A) 1 kV and (B) 0.5 kV spray voltages. Error bars show one standard deviation for three replicates. Representative spectra of methamphetamine fragmentation used for quantification are shown for (C) 1 kV and (D) 0.5 kV. RI = relative intensity; A/IS = ratio of analyte-to-internal standard signal. Internal standard used for methamphetamine was methamphetamine d5 with MS/MS transition m/z 155 → 123.

The method was then extended to detect other illicit drugs such as cocaine, amphetamine, and benzoylecgonine in raw urine (wet and dry). Direct analysis of small

79 molecules in dried urine is possible with the traditional paper spray,93 but their quantification in fresh/wet samples is challenging due to ion suppression effects associated with sample extraction and ionization with aqueous-based solvents. Organic spray solvents have been used,13,92 but rapid evaporation makes it difficult to achieve good quantitation precision, especially in MS/MS experiments. Pure acetonitrile was identified to not penetrate the hydrophobic wax barrier and so permits extended analysis time.

Table 3.1. Limit of detection (LOD) and limit of quantitation (LOQ) of selected illicit drugs spiked into urine and analyzed at 3 kV with 100% acetonitrile spray solvent.

Analyte LODs (LOQs)* in urine matrix Fresh (ng/mL) Dry (ng/mL)

Waxed paper Waxed paper Plain paper Cocaine 0.10 (0.87) 0.62 (1.6) 2.6 (13) Benzoylecgonine 0.21 (0.45) 0.97 (2.5) 5.9 (8.7) Methamphetamine 0.13 (0.60) 0.51 (2.27) 1.5 (8.7) Amphetamine 0.33 (0.38) 0.76 (1.4) - *LODs and LOQs were calculated from respective calibration curves using signal corresponding to (Sblank) + 3×σblank, and (Sblank ) + 10×σblank, respectively where Sblank is the average blank signal and σblank is the standard deviation of the signal from 3 replicates

The results recorded for the detection of the selected illicit drugs from dry and fresh whole human urine samples using wax-printed paper and 100% acetonitrile spray solvent are shown in Table 3.1 (see Figure 3.9 – Figure 3.11 for calibration curves). LODs and LOQs ranged from 0.51–1.2 ng mL−1 and 1.4–3.2 ng mL−1, respectively for fresh samples and 0.10–0.33 ng mL−1 and 0.38–0.87 ng mL−1 for dry samples. Relative standard deviations less than 10% were obtained for both wet and dry samples with concentrations higher than 3.0 ng mL−1. Increased sensitivity (up to >5×) was observed when analyzing dried urine samples compared with fresh urine (Table 3.1), and this effect is attributed to reduced ion suppression effects. Resolubilization of endogenous salts in dried urine is not

80 efficient with acetonitrile and so their transfer to the mass spectrometer during PS ionization is limited. Identical conditions (10 μL and 3 kV) were used when comparing the performance of wax-printed paper detection of drugs in fresh urine to that of un-waxed paper (Table 3.1). When using 10 μL pure acetonitrile as a spray solvent, the signal lasted for only 15 seconds on un-waxed paper (due to faster solvent evaporation over a larger wetting surface area); this time period was inadequate for signal acquisition of more than one ion (Figure 3.12). Increasing solvent volume to 20 μL increased signal lifetime to ∼1 minute and enabled linear regression analysis. Results are shown in Table 3.1 where LODs recorded from the un-waxed PS-MS analysis were at least 3× higher than LOD achieved on wax-printed paper substrates.

(A) (B) 1.4 y = 0.0024x - 0.0002 6 R² = 0.9999 y = 0.0108x + 0.0034 R² = 0.9999 4

0.7 A/IS 2

0 0 0 200 400 0 200 400 Concentration cocaine (ng/mL) Concentration amphetamine (ng/mL) (C) (D) 4 18

3 y = 0.0325x + 0.0189 y = 0.0074x - 0.0015 R² = 0.9999 R² = 1

2 9 A/IS

0 0 0 200 400 0 200 400 Concentration benzoylecgonine Concentration methamphetamine (ng/mL) (ng/mL)

Figure 3.9. Samples of 4 μL fresh urine spiked with 3 – 500 ng/mL drug, sprayed with 10 μL of 100% acetonitrile on s3 wax paper. Drugs were quantified with absolute intensity of fragmentation (A) divided by their respective internal standards (IS) of (A) cocaine (m/z 304 → 182) and IS d3 (m/z 307 → 185), (B) amphetamine (m/z 136 → 119) and IS d5 (m/z 141 → 123), (C) benzoylecgonine (m/z 290 → 168), and IS d3 (m/z 293 → 171) (D) methamphetamine (m/z 150 →119) and IS d5 (m/z 155 → 123). 81

(A) (B) 20 1.8 y = 0.0377x - 0.0009 R² = 0.9998 y = 0.0031x + 0.0052 1.2 R² = 0.9995

10 A/IS 0.6

0 0 0 200 400 0 200 400 Concentration cocaine (ng/mL) Concentration amphetamine (C) (D) (ng/mL) 8 50

y = 0.0133x - 0.001 y = 0.0789x + 0.0143 R² = 0.9991 R² = 0.9999

4 25 A/IS

0 0 0 200 400 0 200 400 Concentration benzoylecgonine Concentration methamphetamine (ng/mL) (ng/mL)

Figure 3.10. Samples of 4 μL dried urine spiked with 3 – 500 ng/mL drug, sprayed with 10 μL of 100% acetonitrile on s3 wax paper. Drugs were quantified with absolute intensity of fragmentation (A) divided by their respective internal standards (IS) of (A) cocaine (m/z 304 → 182) and IS d3 (m/z 307 → 185), (B) amphetamine (m/z 136 → 119) and IS d5 (m/z 141 → 123), (C) benzoylecgonine (m/z 290 → 168), and IS d3 (m/z 293 → 171) (D) methamphetamine (m/z 150 →119) and IS d5 (m/z 155 → 123).

1.2 (A)

0.8 y = 0.002x + 0.0063

R² = 0.9997 A/IS 0.4

0 0 200 400 Concentration cocaine (ng/mL)

12 (B) y = 0.0203x + 0.0555 R² = 0.9996

8 A/IS 4

0 0 100 200 300 400 500 Concentration methamphetamine (ng/mL)

5 (C) y = 0.009x - 0.0399 R² = 0.9998

2.5 A/IS

0 0 200 400 Concentration benzoylecgonine (ng/mL) Figure 3.11. Samples of 4 μL dried urine spiked with 3 – 500 ng/mL drug, sprayed with 20 μL of 100% acetonitrile on un-waxed paper. Drugs were quantified with absolute intensity of fragmentation (A) divided by their respective internal standards (IS) of (A) cocaine (m/z 304 → 182) and IS d3 (m/z 307 → 185), (B) methamphetamine (m/z 150 →119) and IS d5 (m/z 155 → 123), and (C) benzoylecgonine (m/z 290 → 168), and IS d3 (m/z 293 → 171).

20 µL

10 µL 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 Time (min) Figure 3.12. 500 ng/mL cocaine sprayed with 20 L and 10 L 100% acetonitrile with 3 kV spray voltage on un-waxed paper. When 10 L spray solvent was used, signal lasted approximately 15 seconds, which was determined to be inadequate for more than one MS/MS observation. When the spray solvent volume was doubled, the signal lifetime was increased to nearly 1 minute, which was more suitable for calibration, as shown in Figure 3.11 above.

3.6 Analysis of Corrosion Inhibitors and Pesticides in Water The wax-printed PS-MS method was also utilized to detect residual levels of Duomeen (a polyamine corrosion inhibitor) and metaldehyde (a tetracyclic acetaldehyde molluscicide) in water. Duomeen is a widely used chemical substance used to control corrosion in high pressure (HP) water-tube boilers.180 Direct detection and monitoring is necessary for boiler maintenance, and a reactive paper spray method was recently described for on-site in situ detection of Duomeen.181 The use of wax-printed paper substrates can improve the quantitative capabilities of the PS-MS method and that the use of low spray voltages will facilitate field analysis. These expectations have been met, and as shown in Figure 3.13A and B. Duomeen is sensitively detected with calculated LODs of 0.09 pg mL−1 and 0.68 pg mL−1 for 3 and 1 kV spray voltages, respectively. Excellent linearity (R2 = 0.9997) and precision (%RSD = 7.5%) were achieved without internal standards. This performance is due in part to the high ionization efficiency of the Duomeen and the occurrence of a long lasting stable spray from the wax-printed paper substrate, which enables ensemble averaging of many spectra. The method was applied to analyze two real water samples (pre- and post-treatment) collected from a large HP boiler system at Coventry waste treatment facility in the UK. As expected Duomeen (m/z 325) was detected in the post treatment water sample (Figure 3.14).

7E+6 (A) 3 (C) y = 678083x - 10526 R² = 0.9997 y = 0.0142x - 0.0073 2 R² = 0.9988

AI 4E+6 A/IS 1

0E+0 0 0 5 10 0 50 100 150 Concentration duomeen (pg/mL) Concentration metaldehyde (ng/mL) 4E+4 (B) 1.6 (D) y = 0.0091x + 0.0072 y = 343.67x - 316.56 R² = 0.9983 R² = 0.9991

AI 2E+4 0.8 A/IS

0E+0 0 0 50 100 0 50 100 150 Concentration duomeen (pg/mL) Concentration metaldehyde (ng/mL)

Figure 3.13. Calibration of (A) Duomeen sprayed at 3 kV, (B) Duomeen at 1 kV, (C) metaldehyde at 3 kV, and (D) metaldehyde at 1 kV in water samples.All samples were sprayed with 4:1 MeOH/H2O. AI = absolute fragment ion (m/z 308) intensity, A/IS = ratio of analyte-to-internal standard signal. Internal standard used for metaldehyde was atrazine with MS/MS transition m/z 221 → 179.

(A) (B) 100

325

325 Relative Abundance Relative

0 60 120 180 240 300 360 420 60 120 180 240 300 360 420 m/z m/z

Figure 3.14. Analysis of water samples for detection of duomeen. Samples were taken from (A) pre-treatment, and (B) post-treatment stages in the boiler system cycle. Duomeen (m/z 325) was only detected in post-treatment sample, as expected.

Figure 3.13C and D represent calibration curves recorded for the analysis of metaldehyde in water using methanol/water (4:1, v/v) spray solvent charged at 3 and 1 kV, respectively. Analysis of metaldehyde is necessitated by recent reports of metaldehyde poisoning and death in both animals and humans.182,183 The molluscicide is used in agriculture to control slugs in order to protect crops. However, large residues of metaldehyde are mobilized during heavy rainfalls which end up in rivers and groundwater and finally in drinking water. In fact, the European Commission and U.S. Environmental Protection Agency have both issued instructions on permissible level for metaldehyde, restricting its use as a pesticide.184,185

199 (A) 1 kV (B) 3 kV 149 100 m/z 199 m/z 177 111

50 Relative Abundance Relative 177 0 60 120 m/z 180 240 60 120 m/z 180 240

Figure 3.15. Analysis of water samples containing metaldehyde, sprayed at (A) analysis of 300 ng/mL sprayed at 1 kV, fragmentation m/z 199 [M+Na]+, and (B) analysis of 150 ng/mL sprayed at 3 kV, fragmentation m/z 177 [M+H]+. Intensities of fragment ions at m/z 111 and 149 were used to construct calibration curves found in Figure 3.13 C and D, respectively.

In the current study, our interest was in the use of lower spray voltage for direct metaldehyde analysis. Metaldehyde formed sodiated ions in high abundance when analyzed at 1 kV spray voltage, in which case calibration was obtained using m/z 199 [M+Na]+ → 111 MS/MS transition (through the elimination of neutral acetaldehyde dimer (MW 88); Figure 3.15). At 3 kV, however, the predominant peak was the protonated species [M+H]+; m/z 177, which fragmented to give product ion at m/z 149 via CH2=CH2 86

(MW 28) neutral loss. The formation of different ion types from metaldehyde when using different spray voltages is unique (Figure 3.16), and ascribed to increased charged density in the microfluidic channel when using higher DC voltage (3 kV). Presumably, the increased charge causes increased proton abundance, which in turn suppresses sodium adduction. The different fragmentation pathways for [M+Na]+ (m/z 199 → 111) versus [M+H]+ (m/z 177 → 149) suggests different structures exist for both ions. It is proposed that metaldehyde ring opening occurs in the presence of high proton abundance, at higher spray voltage, which subsequently fragments through eliminating ethylene in MS/MS. Similar effect is observed in acidic solution (Figure 3.17). LODs were determined to be 4.9 ng mL−1 and 5.2 ng mL−1 for 3 kV and 1 kV spray voltages using [M+H]+ and [M+Na]+ ions, respectively. The ability to generate different ion types at different spray voltages will aid easy differentiation of metaldehyde from other interfering ions (having the same nominal mass), especially during field analysis.

Figure 3.16. Depiction of the effect of spray voltage on metaldehyde ionization. Higher voltage (3 kV; green, right) generates high proton abundance at wax-printed paper tip producing protonated metaldehyde [M+H]+ ions with unique fragmentation pathway. Alternatively, at low voltage (1 kV; blue, left), the ionization process is dominated by sodium adduction forming [M+Na]+ ions at m/z 199. MS inlet was grounded.

Figure 3.17. Comparison of PS-MS mass spectra for metaldehyde recorded at different pHs of (A) 7, and (B) 3. Spectrum in B was recorded after 5 min of adding acid. Inserts show product ion mass spectra for [M+Na]+ at m/z 199 and for [M+H]+ at m/z 177 achieved using collision induced dissociation.

3.7 Summary In summary, a demonstration of 2D wax-printed substrates for expanding the applicability of paper spray ionization has been provided. By reducing the area of wetting on the paper triangle, spray solvent evaporation is minimized and stable electrospray can be generated that allows longer analysis times (∼6× increase). Up to 80% methanol could be used for aqueous-based solvent systems and 100% acetonitrile spray solvents could be 88 employed for the wax-printed paper spray experiment without penetrating the wax hydrophobic barrier. Microfluidic patterns suitable for low voltage paper spray ionization were determined empirically using five different designs. The spray dynamics and loading capacity of the selected wax-printed paper substrate were characterized and compared with plain, un-waxed paper triangles. It is believed that the reduction in channel width improves the likelihood of analyte transport to the paper tip region where both the DC potential is more confined and proximity (i.e., distance) to the MS atmospheric interface is favorable for successful ion entry. Spray could be sustained for more than ten minutes, enabling tandem MS analysis of various analytes to be performed with high precision. The 2D wax- printed paper substrates do not require external solvent reservoirs or pumps to achieve continuous wetting. The wax-printed PS-MS methodology was tested by analyzing water-based and whole human urine samples. Analytes detected include corrosion inhibitor Duomeen, pesticide metaldehyde, and illicit drugs such as cocaine and its metabolite benzoylecgonine, methamphetamine and amphetamine. Unlike un-waxed paper triangles which were unable to generate acceptable calibration curves at <1 kV spray voltages, good linearity and precision were obtained for water samples analyzed from wax-printed paper substrates. As low as 0.68 pg per mL sensitivity was recorded for Duomeen in water when using 1 kV spray voltage. A strong voltage dependence was observed for the analysis of the cyclic metaldehyde pesticide in water. Sodiated metaldehyde ions were detected from water at the reduced charged of 1 kV and fragmented differently upon collisional activation when compared with protonated species generated at high electrical charging of 3 kV. Such capability will aid field analysis with high selectivity. Higher voltage (>3 kV) was required for illicit drug analysis from raw urine and 5× improvement in signal sensitivity was recorded for dried urine samples compared with wet samples, which were in turn more sensitive for similar fresh urine samples analyzed from un-waxed paper. These results suggest that the wax-printing methodology can serve as an efficient approach to modify paper substrates for improved PS-MS analysis. The wax hydrophobic barrier may also serve to ease technical challenges (e.g., inaccuracies in quantitation) associated with sample collection in dried blood (biofluid) spot preparation by allowing a

89 more uniform fluid distribution.89 Future work will focus on mathematical modeling and computer simulation to investigate the mechanism of fluid flow and how confined DC potentials influence analyte ionization and transport from the tip of the wax-printed paper substrates.

Chapter 4. Water System Analysis with Paper Spray Ionization

4.1 Introduction Monitoring of surface waters and water in industrial applications is necessary due to the addition of potentially toxic, disruptive or otherwise environmentally destructive additives that may be introduced through water runoff.186–188 Examples of these additives and impacts include: (i) pesticides– increase in herbicides and fungicides resulting from runoff of agricultural centers, resulting in large concentrations found in surface water;189 (ii) pharmaceuticals and endocrine disrupters– unused or expired drugs that are improperly disposed of in sewers or landfills that ultimately are reintroduced into drinking water through groundwater or surface water;190,191 and (iii) eutrophication– the anthropogenic increase concentrations of nutrients, such as phosphate, in waste water sources, increasing bacterial and algal blooms that contribute to the destabilization of native ecosystems.192–194 Examples of these small molecules that are potentially harmful to the environment and human consumption are Duomeen O, a corrosion inhibitor used in boiler systems that is used as an alternative to more toxic aromatic amines, and metaldehyde, a molluscicide used on crops whose concentration is regulated due to its toxic nature, which are discussed in this chapter. Direct analysis and identification of long chain aliphatic primary diamine Duomeen O (n-oleyl-1,3-diaminopropane), corrosion inhibitor in raw water samples taken from a large medium pressure water tube boiler plant water samples at low LODs (<0.1 pg) has been demonstrated for the first time, without any sample preparation using paper spray mass spectrometry (PS-MS). The presence of Duomeen O in water samples was confirmed via tandem mass spectrometry using collision-induced dissociation and supported by exact mass measurement and reactive paper spray experiments using an LTQ Orbitrap Exactive instrument. Data shown herein indicate that paper spray ambient ionization can be readily

91 used as a rapid and robust method for in situ direct analysis of polyamine corrosion inhibitors in an industrial water boiler plant and other related samples in the water treatment industry. This approach was applied for the analysis of three complex water samples including feedwater, condensate water, and boiler water, all collected from large medium pressure (MP) water tube boiler plants, known to be dosed with varying amounts of polyamine and amine corrosion inhibitor components. Polyamine chemistry is widely used for example in large high pressure (HP) boilers operating in municipal waste and recycling facilities to prevent corrosion of metals. The samples used in this study are from such a facility in Coventry waste treatment facility, U.K., which has 3 × 40 ton/hour boilers operating at 17.5 bar.3 Metaldehyde is extensively used worldwide as a contact and systemic molluscicide for controlling slugs and snails in a wide range of agricultural and horticultural crops. Contamination of surface waters due to run-off, coupled with its moderate solubility in water, has led to increased concentration of the pesticide in the environment. In this study, for the first time, rapid analysis (<~1 minute) of metaldehyde residues in water is demonstrated using paper spray mass spectrometry (PS-MS). The observed precursor molecular ions of metaldehyde were confirmed from tandem mass spectrometry (MS/MS) experiments by studying the fragmentation patterns produced via collision-induced dissociation. The signal intensity ratios of the most abundant MS/MS transitions for metaldehyde (177→149 for protonated ion) and atrazine (221→179) were found to be linear in the range 0.01 to 5 ng/mL. Metaldehyde residues were detectable in environmental water samples at low concentration (LOD < 0.1 ng/mL using reactive PS-MS), with a relative standard deviation <10% and an R2 value >0.99, without any pre-concentration/ separation steps. This result is of particular importance for environmental monitoring and water quality analysis providing a potential means of rapid screening to ensure safe drinking water.

3 Reproduced with permission from Jjunju, F. P. M.; Maher, S.; Damon, D. E.; Barrett, R. M.; Syed, S. U.; Heeren, R. M. A.; Taylor, S.; Badu-Tawiah, A. K. Screening and Quantification of Aliphatic Primary Alkyl Corrosion Inhibitor Amines in Water Samples by Paper Spray Mass Spectrometry. Anal. Chem. 2016, 88 (2), 1391–1400.181 Copyright 2016 American Chemical Society.

4.1.1 Amine Analysis Introduction The addition of corrosion inhibitors to corrosive systems is a well-established preventative approach worldwide.195–198 Neutralizing agents such as aliphatic and aromatic amines, mono, di-, or poly amines and their salts when added in small amounts to a corrosive water boiler system are known to reduce, slow down, or prevent corrosion to metal.199–203 In agreement with green chemistry aims, new corrosion inhibitor formulations should be less toxic, soluble in aqueous medium, and biodegradable,187,188,204–207 especially when they are to be used in portable water transfer systems. Therefore, toxic aromatic amines and their salts should be avoided and replaced with greener long-chain aliphatic mono-, di-, or polyamines or their salts.180 Polyamine corrosion inhibitor formulations are widely used in large high pressure (HP) water tube boiler plants, e.g., refineries, power generating plants, steel works, chemical plants where the operating pressure is >45 bar. There is a strong need for analytical methods for on-site analysis and quantification of corrosion inhibitor residues in the large HP water tube boiler plants systems with fast response times, preferably with little or no special sample preparation.208–216 From such samples, the analytical data obtained should be useful in maintaining the appropriate levels of the inhibitor in the water tube boiler plants. This is useful not only for quality control but in the development of new effective corrosion inhibitors to combat corrosion in large high pressure (HP) water tube boiler plants.217,218 Currently, extraction procedures based on solid-phase extraction followed by gas chromatography (GC) or high-performance liquid chromatography (HPLC) coupled with mass spectrometry (MS) methods,219–222 have been successfully used in the identification and quantification of long-chain aliphatic primary polyamines in boiler water samples below part-per-billion (ppb) levels.221,223,224 These analytical methods are reliable, and low limits of detection (LODs) with high specificity and sensitivity can be achieved. However, direct identification and quantification of long chain polyamines corrosion inhibitor formulations is not possible at present using GC/MS or HPLC−MS due to the fundamental instrument characteristics and a large dipole of the polyamine.218,225–227 To overcome these challenges, polyamine corrosion inhibitors are first derivatized and preconcentrated (using either solid-phase or liquid−liquid extraction) to improve the GC/HPLC detection

93 properties.225,228–230 While these analytical methods have proven successful in the analysis of the long chain aliphatic primary polyamines corrosion inhibitor formulations in large high pressure (HP) water tube boiler plants, they can be time consuming. Moreover, these methods are limited by the need for manual transfer of samples to the laboratory before analysis. Therefore, there is a strong interest in rapid screening methods for long chain aliphatic primary polyamines inhibitor formulations in large HP water tube boiler plants that requires no sample preparation and yet provides specific information regarding the corrosion inhibitor levels in the large HP water tube boiler plant. Such methods would have several important applications in the water treatment industry, energy sector, and for environmental monitoring and hygiene.210,231,232 As will be shown in this study, ambient ionization mass spectrometry (AIMS) combined with tandem mass spectrometry (MS/MS) and exact mass measurements can meet such criteria.233–235 AI-MS is an experiment in which ion generation is performed from untreated samples, in air outside the vacuum environment of the mass spectrometer. The fact that no sample preparation or prior extraction steps are needed in AI-MS means that experiments are simple, which ultimately reduces the total MS analysis time.23,236–240 Ambient ionization methods include desorption electrospray ionization (DESI),99,235,241 direct analysis in real time (DART),242,243 laser ablation electrospray ionization,170 desorption atmospheric pressure chemical ionization (DAPCI),36,37,244 nanodesoption electrospray ionization (nanoDESI),245 and low temperature plasma (LTP)246 among others. These methods have been successfully applied to the analysis and quantification of a wide range of samples such as environmental pollutants,247 animal tissues,245 and in complex mixtures without any sample pretreatment. High molecular specificity and sensitivity have been successfully achieved with AI-MS analysis through MS/MS, in situ ion/molecule reactions, and high resolution mass measurements.248–251 Paper spray (PS) ionization is a relatively new AI-MS method, which has been successfully applied in the analysis and quantification of complex molecules, ranging from small organics to large biological molecules including dried blood under ordinary ambient conditions.44 When using PS, the sample is usually loaded onto paper (or another porous medium) that has been cut to a fine point (tip). The paper is wetted with a solvent and

94 charged liquid droplets are emitted from the paper tip when a high dc voltage (±3.5 kV) is applied. Droplet emission occurs presumably by field emission and/or other unidentified mechanisms.66,93,175 Analysis by PS-MS requires little or no sample preparation and the entire experiment can be completed in times on the order of a few seconds (i.e., less than 1 min). In comparison to other ambient ionization methods, PS integrates three analytical steps: sample collection, separation, and ionization into a single experimental step making it more attractive for rapid and direct analysis of analyte(s) in complex mixtures. In addition, no nebulizer gases are required and so the technique can be used with portable MS in the field. Recently the successful analysis of aromatic quaternary ammonium corrosion inhibitor formulations in petroleum oil samples using PS with a miniature mass spectrometer (Mini 12) has been reported.252 Reactive PS-MS is a variant of the normal PS-MS experiment that incorporates rapid chemical reactions into the PS ionization process. Reactions occur at the sampling spot concurrently with mass spectra acquisition to improve sensitivity and selectivity for target molecules present in the complex mixtures by doping a reactive reagent into the spray solvent.253–256 In this chapter, PS-MS is used as a sensitive and selective ionization method for the rapid detection and quantification of the aliphatic long chain primary polyamine (n-oleyl- 1,3-diaminopropane (Duomeen O)) corrosion inhibitor formulation in a variety of complex sample matrixes including polyamine and amine mixtures collected from a large water tube boiler plant operated at medium pressure (17.5 bar). Reactive PS-MS is also used in which acetone is doped with the spray solvent to aid in characterization and selective detection of the n-oleyl-1,3-diaminopropane (Duomeen O) from a mixture of polyamines and amines. The samples studied include competitor product A, naylamul S11 and ascamine DW BR1 (mixture of polyamine and amines), and three water samples (feedwater, condensate water, and boiler water) collected from a large HP boiler system at Coventry waste treatment facility in the U.K. that was previously dosed by a six-component polyamine and amine corrosion inhibitor. To successfully characterize and confirm the presence of Duomeen O analyte(s) in crude complex water samples, it was necessary to first analyze a standard with high- resolution MS and tandem MS using collision induced dissociation (CID) to determine the

95 molecular formula and structure, respectively. Sample preparation was reduced to dilution of the standard model compounds in methanol while real water samples were analyzed directly as supplied without any dilution. As shown, PS retains the advantages of high sensitivity and specificity typical of MS experiments, plus short (<1 min) total analysis times with, no sample pretreatment; the ability to identify corrosion inhibitor formulations can be achieved readily at trace levels with a limit of detection (LOD) of 0.1 pg (absolute concentration) with acceptable reproducibility (RSD of <10%) in a variety of untreated, complex samples. In this chapter, direct identification, structure characterization, and confirmation of the presence of long chain n-oleyl-1,3-diaminopropane (Duomeen O) corrosion inhibitor in water samples using chemical reactions, PS-MS(/MS), and exact mass measurement were performed. The characterization of purified Duomeen O samples is first presented, followed by quantitative/ analytical performance measurements, and finally the analysis of a variety of complex water boiler samples collected from large HP water tube boiler plant (Coventry Waste Treatment facility U.K.). The long chain Duomeen O corrosion inhibitor formulation was investigated in crude water samples because its identification and quantification is essential in the optimization of the corrosive system,203,206,226 and current efforts have focused on developing new, green, and efficient corrosion inhibitors for water treatment plants.257–259 There is also the need to monitor the level of residual corrosion inhibitors to prevent run away reactions. Water transfer pipelines are often carbonated to remove dissolved carbon dioxide species, but the process in turn generates carbonic acid that leads to reduced pH and consequently corrosion. Corrosion inhibitor formulations, when added in small amounts to a corrosive water boiler system, neutralize the carbonic acid and bring the pH to a normal value.

4.1.2 Metaldehyde Analysis Introduction Detection and quantification of contaminants or pollutants in surface waters is of great importance to ensure safety of drinking water and for the aquatic 182,186,260–263 environment. Metaldehyde (CH3CHO)4 is a cyclic tetramer of acetaldehyde and is used extensively around the world as a molluscicide in agriculture for the control of

96 slugs to protect crops. Large amounts of metaldehyde residues (from ‘slug pellets’) become mobilized, especially during periods of rainfall, seeping into reservoirs, rivers and groundwater, from which drinking water is sourced. Although metaldehyde has low toxicity, cases of metaldehyde poisoning and death in both humans and animals have been reported.182,183,264 The United States Environmental Protection Agency (EPA) re-registered metaldehyde as a ‘restricted use pesticide’ and required risk-reduction measures to be adopted due to the potential short-term and long-term effects on wildelife185,265 The World Health Organization (WHO) classifies metaldehyde as a “moderately hazardous” pesticide (class II).266 In Europe, the European Commission has adopted a directive that restricts pesticides levels to 0.1 μg/L in drinking water.184,267 Water companies and environmental agencies are under increasing pressure to routinely monitor levels of metaldehyde residues in water courses as part of their legal obligation.268 As such there is an increasing need to develop effective analytical methods for detecting and quantifying metaldehyde in water samples at the source. In particular, in-situ monitoring is required to ensure water management practices are based on empirical, up-to-date information which provides a better understanding of competing factors, risk and requirement. Rapid analytical methods for in-situ analysis of metaldehyde in water, if available, would provide critical information on water quality for water companies and regulation bodies to manage exposures. Quantitative analysis of metaldehyde has been reported using various ex-situ methods based on solid-phase extraction183,269 followed by gas chromatography (GC) or high performance liquid chromatography (HPLC) with mass spectrometry (MS).165,264,268–271 However, each of these analytical methods involves extensive sample preparation including extraction, separation, and derivatization, resulting in increased cost and time of analysis. As will be demonstrated in this study, ambient ionization (AI) combined with tandem mass spectrometry (MS/MS) can overcome such limitations.21,237,238,272

Figure 4.1. Schematic of the paper spray mass spectrometry experimental setup used for rapid detection of metaldehyde in water samples.

Here, experiments were carried out using a commercial benchtop ion trap mass spectrometer coupled with PS ionization (Figure 4.1). Sample preparation was reduced to dissolving the model compounds (metaldehyde and paraldehyde) in methanol/water to form a stock solution (1000 ppm), that was serially diluted with water to the desired concentration before analysis, while raw environmental water samples (Abberton Raw & Chigwell Raw) were analyzed directly as supplied (from Northumbrian Water, UK) without any dilution. The results show that < 0.1 ng mL−1 of metaldehyde in environmental water placed onto paper can be detected using a commercial benchtop mass spectrometer. The limit of detection (LOD) obtained was 0.05 ng mL−1 and below the permitted minimum EU levels for drinking water; good linearity (R2 = 0.9986) and accuracy (relative standard deviation ~7%) were also achieved. The analyte(s) identity was then characterized by analyzing the fragmentation patterns of metaldehyde in water using tandem mass spectrometry (MS/MS). The cyclic nature of metaldehyde can encourage the inclusion of + + + different ions (H , Na and NH4 ) to enable the formation of corresponding metaldehyde ion types when analyzed using appropriate spray solvents. This capability was assessed in reactive paper spray experiments, offering more than an order of magnitude enhancement in detection limits. When collisionally activated, each ion type ([M+H]+, [M+Na]+, and + [M+NH4] ) dissociated through unique pathways leading to the generation of distinctive

98 product ions. These fragmentation patterns were fully characterized through MS/MS experiments.

4.2 Materials and Methods Chemicals and Standards: The PS organic solvents; methanol (HPLC grade) and acetone were purchased from Mallinckrodt Baker Inc. (Phillipsburg, NJ). The chromatography paper used as the sample substrate was grade I cellulose Whatman (Whatman International Ltd., Maidstone, U.K.). The standard model compound, n-oleyl-1,3-diaminopropane (Duomeen O), cyclohexylamine, morpholine, diethyl amino ethanol, and the polyamine and amine mixture corrosion inhibitor formulations (competitor product A, naylamul S11 and ascamine DW BR1), used in this study were supplied by B&V Water Treatment Company (Lamport Drive Heartlands Business Park Daventry Northamptonshire, NN11 8YH, U.K.). The crude water samples (i.e., feedwater, condensate water, and boiler water) were collected from a large HP water tube boiler plant at the Coventry waste treatment facility U.K., that was previously dosed by a six-component mixture of cyclohexylamine, diethyl amino ethanol, mono ethanol amine, methyl ethyl ketonoxime, Duomeen O, and tallow S

11 corrosion inhibitor formulation. The deuterium labeled standards, metaldehyde-d16 and atrazine-d5, were purchased from QMX laboratories (Essex, UK) while laboratory grade deionized water was purchased from Reagent Chemicals (Cheshire, UK). The raw environmental water samples (Abberton Raw & Chigwell Raw) were supplied by Northumbrian Water (Durham, UK).

Sample Preparation: Sample preparation was reduced to the dilution of the model compounds to the desired concentration while no sample preparation was performed for the raw boiler water samples and raw water samples (Abberton Raw & Chigwell Raw). Each model compound was diluted in methanol (HPLC grade) to a desired concentration. Environmental water samples were used as supplied, from Northumbrian Water Ltd. (Durham, UK), without any modification or pre-concentration. From each solution pertaining to amine analysis, 2 μL was deposited using a pipet onto cellulose paper substrate and then analyzed using PS-MS. 99

The boiler water sample mixtures (i.e., feedwater, condensate water, boiler water), and polyamine and amine mixture (i.e., competitor product A, naylamul S11, and ascamine DW BR1), were used as supplied without any modification or preconcentration. A volume of 2 μL of each sample deposited using a pipet onto cellulose paper substrate analyzed using normal PS-MS and reactive-PS-MS. For samples pertaining to metaldehyde analysis, the 10 L sample was deposited onto the filter paper surface, using a pipette and analyzed directly without any sample preparation. In all of the metaldehyde PS-MS experiments performed approximately 10 μL of pure methanol was used as the spray solvent (unless otherwise stated).

PS-MS Instrumentation. All experiments were performed using a linear ion trap (LTQ) mass spectrometer (Thermo Fisher Scientific, San Jose, CA), tuned for optimum detection of the precursor ion of interest. The temperature of the MS capillary inlet was set at 200-250 °C. The tube lens voltage was set at 65-70 V and the capillary voltage maintained at 15 V in both positive and negative modes, respectively. The paper spray ion source was placed 3 mm in front of the inlet the LTQ instrument in all the experiments. An electric potential of ±3.5 kV was used for all the PS experiments in both positive and negative mode. It is important to note that in the paper spray experiments no carrier gas is required, instead a plume of ions is generated only with the application of a potential on the paper with the sample and the spray solvent as shown in Figure 4.2. Tandem mass spectrometry (MS/MS) was performed on the molecular ions of interest for structural elucidation allowing analyte identification using collision-induced dissociation (CID). An isolation window of 0.1–1.5 Th (mass/charge units) and normalized collision energy of 15–40 (manufacturers unit) were used. Approximately 2 μL of each sample was deposited on a filter paper surface and analyzed directly without any sample preparation. Tandem mass spectrometry (MS/MS) was used for the structural elucidation, and analyte identification was performed on the molecular ions of interest using collision induced dissociation (CID). An isolation window of 0.1−1.5 Th (mass/charge units) and normalized collision energy of 15− 40%

100

(manufacturers unit) was used. Furthermore, the identities of studied long-chain polyamine and other corrosion inhibitor formulations were confirmed using a high-resolution mass measurement Orbitrap mass spectrometer (Exactive, Thermo Fisher Scientific, San Jose, CA). The experimental conditions on the Orbitrap were as follows: maximum injection time of 50 ms, two microscans, and activated automatic gain control (AGC).

Figure 4.2. Schematic of the paper spray mass spectrometry experimental setup used in rapid screening of Duomeen O in the boiler system water samples.

PS-MS: A cellulose chromatography paper (Whatman, Maidstone, U.K. grade I) was used as the paper substrate, and equilateral triangles with ∼5 mm sides were cut manually with scissors. The tips of the base angles were cut off and the vertex angle was kept sharp. The paper substrate was held by a copper clip (Figure 4.2) so that the vertex was ∼3 mm away from the inlet capillary of the mass spectrometer with an atmospheric pressure interface that transports the spray plume of ionized analyte(s) into the vacuum system of the mass spectrometer for analysis. The sample solution was applied to the paper triangle followed 101 by application of a high voltage. The typical experimental parameters used were as follows: paper spray solvent 10 μL of acetonitrile; and the voltage applied to the paper was in the range of +3.5 kV in positive and −3.5 kV negative modes. In all experiments (unless noted) the instrument was set to record spectra in the AGC mode for a maximum ion trap injection time of 100 μs and 3 microscans were combined per spectrum. Figure 4.2 shows the experimental protocol that was followed in this study; first, a blank spectrum of 10 μL of methanol was taken before sample was applied onto the paper substrate. The analysis was performed in both full MS mode for analyte identification and tandem MS mode for structure elucidation.

Reactive-PS-MS: In the Reactive-PS-MS experiment of Schiff-base reaction, pure acetone was utilized to enhance the selectivity and specificity of the long chain n-oleyl-1,3- diaminopropane in a variety of sample matrixes. In this experiment, 10 μL of the pure acetone reagent was added to cellulose paper with long chain n-oleyl-1,3-diaminopropane using a pipette. All the reactive-PS-MS experiments were performed using a commercial LTQ instrument (as shown in Figure 4.2) following the same settings and procedures as used in the normal PS-MS experiment described above.

4.3 Paper Spray Analysis and Characterization of Duomeen O and Metaldehyde Paper spray often produces protonated [M+H]+ ions, but often, sodiated ([M+Na]+), potassiated ([M+K]+), or alternate progeny ions with various adducts are observed. Because of the wide variety of ions possible when analyzing a given sample, characterization of the most abundant ion is required under each set of conditions. The following section examines these ions and allows the determination of the optimum ion to further study.

102

4.3.1 Paper Spray Mass Spectrometry and Characterization of Duomeen O Using Positive Ion Mode The positive ion PS-MS molecular analysis of Duomeen O, using 2 μL of samples deposited on the chromatography paper triangle was achieved after the application of 10 μL of methanol as the PS spray solvent. The resultant mass spectrum is as shown in Figure 4.3, which is dominated by intact protonated molecular ion [M+H]+ at m/z 325 in the mass range of 100−1000 Da, with little or no fragmentation (Figure 4.3A). The insert (i), in Figure 4.3A shows the isotopic distribution at m/z 325, and the high proton affinity of n- oleyl-1, 3-diaminopropane (Duomeen O) allows for its protonation. The remarkable absence of signal due to the paper spray ionization background is consistent with the high proton affinities of diamine compounds, a well-known ionization feature of many chemical ionization methods.

103

(A)

(B)

Figure 4.3. Positive ion mode paper spray mass spectrum for Duomeen O corrosion inhibitor model compound analyzed using a benchtop ion trap mass spectrometer. Absolute amounts of analyte were spotted onto filter paper and ionized in the open air by application of an electric potential, 2 μL, viz., 10 ppb: (A) Duomeen O (MW 324) in methanol solution and (B) exact mass measurement of Duomeen O. Insert (i) shows the isotopic distribution of the Duomeen O protonated molecular ion [M+H]+ at m/z 325, and inserts (ii) and (iii) show the MS/MS CID data for the selected ions. Insert (iv) shows the corresponding exact mass MS/MS CID data.

104

4.3.2 Structure Characterization and Confirmation of Duomeen O Tandem MS via multistage CID was employed for the initial structural characterization of the intact protonated [M+H]+ , Duomeen O cation at m/z 325. The insert (iii) in Figure 4.3 shows the product ion scan MS2 mass spectrum obtained in the positive ion mode using PS-MS where the CID dissociation leads to a single fragment ion at m/z 308 owing to ammonia (MW 17 Da) neutral loss as a result of heterolytic cleavage of the low energy C−H−NH2 bond. The stability and abundance of the product ion allows three stage (MS3) tandem MS experiments to be performed. In this particular case, CID of the product ion at m/z 308 yielded further fragment ions at m/z 280 through the loss of ethylene (MW 28 Da) neutral molecule as shown in insert (iii) of Figure 4.3A. With the molecular weight of 324.6 Da, a major concern about Duomeen O in paper spray was the actual ion type generated (i.e., protonated or radical molecular cation). Nominal mass measurement produced 325.5 Da as the molecular ion (Figure 4.3A). The MS/MS experiment described above was useful but further verification was needed to confirm the structure of this long chain C8−Duomeen O compound. For this, tandem MS was combined with exact mass measurements, which provided the chemical formula assignment in the Xcalibur 3.1 software. The use of 50 000 resolution and lock mass proved to be sufficient to determine the molecular formula of Duomeen O with error considerably below 1 ppm (Figure 4.3B). The proposed molecular formula based on the exact mass measurement confirmed that the detected long chain Duomeen O formed a protonated molecule [M+H]+ upon ionization by paper spray ionization, exact mass of CID fragments and neutral loss (insert (iv), Figure 4.3B) all confirm this assignment which is consistent with the CID data interpretation described in insert (iii) Figure 4.3A. Other corrosion inhibitor formulation model compounds analyzed by the PS-MS method included cyclohexylamine (MW 99), morpholine (MW 87), and diethyl amino ethanol (MW 117). These compounds also gave intact protonated molecules [M+H]+, and their identities were confirmed using their MS/MS CID fragmentation patterns (see Figure 4.4, for detailed characterization).

105

(A)

(B)

(C)

Figure 4.4. Positive ion mode paper spray mass spectrum for amine corrosion inhibitor model compound analyzed using a bench-top ion trap mass spectrometer. Absolute amounts of analyte spotted onto a filter paper and ionized in open air by application of an electric potential, 2 μL, viz 10 ppb with methanol spray solvent; (A) morpholine (Mw 87), (B) cyclohexylamine (MW 99), (C) diethyl amino ethanol (MW 117) Insert (i)-(iii) shows the MS/MS CID data for the selected ions at m/z 88, 100 and 118 for morpholine, cyclohexylamine and diethyl amino ethanol respectively.

106

The Duomeen O showed a limit of detection (LOD) of 0.1 pg (absolute concentration) when analyzed using PS-MS. The LOD was determined as the concentration that produces a signal more than three times greater than the standard deviation plus the mean value of the blank, in MS/MS mode. The signal intensity ratios of the most abundant MS/MS transitions (at m/z 325.5 → 308) were found to be linear (regression parameters, y = 0.0056x + 0.001234, with R2 value 0.999; see Figure 4.5) in the range of absolute amounts from 0.1 to 1000 ppb and showed good reproducibility (relative standard deviation, RSD < 10% for 1 pg samples deposited on the paper substrate).

9E+07

Calibration of Duomeen O

6E+07

Signal y = 7E+06x R² = 0.9999

3E+07

0E+00 0 1 2 3 4 5 6 7 8 9 10 Concentration Duomeen O (ppb)

Figure 4.5. Duomeen O calibration curve for the qualitative analysis polyamine in boiler system water samples using paper spray mass spectrometry in positive ion mode.

107

(A)

(B)

Figure 4.6. Positive ion mode paper spray mass spectrum of metaldehyde recorded using a benchtop ion trap mass spectrometer. 5 μg of the analyte in 1 μL of deionized water was spotted onto filter paper and ionized in air by application of a positive electric potential (3.5 kV) using methanol as the paper spray solvent. (A) The sodiated molecular ion [M+Na]+ peak of metaldehyde (MW 176) in deionized water produced the dominant ion signal intensity (m/z 199), and (B) Sodiated + molecular ion [M+Na] of deuterated metaldehyde-d16 (MW 192) in deionized water produced the dominant ion peak (m/z 215). Inserts (i–ii) show the isotopic distribution of the metaldehyde and metadehyde-d16 sodiated [M + Na]+ ion adducts at m/z 199 and 215 respectively. Note that in insert (ii) the relatively large signal intensity for m/z 214 is likely a consequence of D-H back- exchanges occurring in the ambient environment (and 99% isotopic enrichment). Inserts (iii–v) show the tandem MS CID data for the selected ions of metaldehyde and metadehyde-d16.

108

4.3.3 Analysis of Metaldehyde using Paper Spray Mass Spectrometry In this section, the direct detection of residues of metaldehyde in water using paper spray- mass spectrometry (PS-MS) is described. Figure 4.6 shows the mass spectra of metaldehyde (MW 176) obtained in positive ion mode using paper spray ionization with methanol as the spray solvent. A dominant sodium adduct ion [M+Na]+ of metaldehyde at + m/z 199 and a less intense ammonium adduct ion [M+NH4] at m/z 194 were observed (Figure 4.6A). Insert (i) in Figure 4.6A shows the isotopic distribution of the metaldehyde sodiated adduct [M+Na]+ at m/z 199. To confirm the identity of the molecular sodiated ion [M+Na]+ attributed to m/z 199, product ion MS/MS spectra were recorded using collision- induced dissociation (CID). The result from this experiment is shown in insert (iii), Figure 4.6A, which indicates that, upon CID activation, the ion at m/z 199 yields a predominant fragment ion at m/z 67. This ion corresponds to sodiated acetaldehyde (MW 44) formed from the sequential loss of neutral dimer (MW 88) and monomer (MW 44) of acetaldehyde. Indeed, the intermediate fragment ion formed after the dissociation of the acetaldehyde dimer is observed at m/z 111, followed by the elimination of the monomer. A competing fragmentation pathway to the loss of the dimeric acetaldehyde was deemed to correspond to the elimination of a water (18 Da) molecule to give a less intense fragment ion peak at m/z 181. The less intense ammoniated molecular ion peak observed at m/z 194 was also confirmed via CID (Figure 4.7). Upon CID activation, the ion at m/z 194 yields a fragment through sequential loss of two water (18 Da) molecules, yielding intense product ions at m/z 176 and m/z 158 (major).

109

Figure 4.7. Positive ion mode paper spray mass spectrum of metaldehyde recorded using a benchtop ion trap mass spectrometer. 5 μg of the analyte in 1 μL deionized water was spotted onto filter paper and ionized in air by application of a positive electric potential (3.5 kV) using methanol as the paper spray solvent. The figure shows the CID data for the precursor ion at m/z 194.

The sodiated molecular ion and fragmentation assignments were further investigated using deuterated metaldehyde-d16 (MW 192) as a model compound sample. Here too, a dominant sodiated molecular ion [M+Na]+ at m/z 215 was observed demonstrating that adduction with the Na+ ion was unaffected by isotopic substitution (Figure 4.6B). Insert (ii) in Figure 4.6B shows the isotopic distribution of the metaldehyde sodiated adduct [M+Na]+ at m/z 215. These sodium adducts were formed with relatively low internal energy as indicated by the absence of associated fragmentation observed in the full mass spectrum (Figure 4.6). Insert (iv), Figure 4.6B, shows the CID data for the intact metaldehyde-d16 sodiated molecular [M+Na]+ ion at m/z 215, which upon collisional + activation dissociates yielding a more intense fragment ion at m/z 71 [CD3CDO+Na] via sequential elimination of dimeric (96 Da) and monomeric (48 Da) acetaldehyde-d4 without H/D scrambling. The stability and abundance of the precursor [M+Na]+ molecular ion from metaldehyde-d16 allowed multi-stage MS/MS/MS experiments to be performed and the result is as shown in insert (v), Figure 4.6B, which unambiguously confirms the source of

110 the m/z 71 product ion. Like metaldehyde, the deuterated metaldehyde-d16 species also formed adducts with ammonium ions at m/z 210.

(A)

(B)

Figure 4.8. Calibration curve for quantification of metaldehyde in water using PS-MS/MS when analyzing (A) sodiated ion types and (B) ammoniated ion types produced in neutral MeOH spray solvent. Error bars indicate standard deviation from three replicates.

Table 4.1. Analytical performance of PS-MS/MS for analysis of metaldehyde in water.

Figure of Merit PS-MS/MS LOD: [M+H]+ ion type 0.05 ng/mL LOD: [M+Na]+ ion type 2.69 ng/mL Estimated time of sample preparation <~ 60 seconds In-situ analysis Yes

111

The characterized sodiated [M+Na]+ molecular ions (m/z 199) provided a direct means to detect and quantify metaldehyde in water. This was accomplished by generating a calibration curve obtained using the collisionally activated fragment ion (m/z 71) intensity of the sodiated [M+Na]+ metaldehyde molecular ion at m/z 199 (Figure 4.8). The limit of detection (LOD) was determined to be 2.69 ng/mL, which is above the EU regulated LOD value for metaldehyde in water (Table 4.1). The LOD was determined as the concentration that produces a signal more than three times greater than the standard deviation plus the mean value of the blank (in MS/MS mode). The sensitivity and selectivity of the PS-MS method can be enhanced by exploring chemical reactions that form stable adducts. To this end a more robust ionization mechanism of metaldehyde was developed in the form of reactive paper spray ionization.

4.4 Reactive Paper Spray Mass Spectrometry Although observing an ion’s precursor m/z and the product ions from CID provides enough information to elucidate structure and confirm presence of a specific molecule in a solution, the presence of matrix molecules in real samples (such as environmental and industrial) increases the likelihood of false positive detection of the molecule of interest. Because of this, another confirmatory step may be taken in order to decrease the likelihood of a matrix ion of a different structure giving similar ion readings in the mass spectrometer. By introducing specific reagent on the paper during paper spray ionization, strategic reaction monitoring can be utilized to confirm not only mass, but molecular structure of the ion of interest. This strategic reaction monitoring has been termed “Reactive Paper Spray Mass Spectrometry.”

4.4.1 Duomeen O Detection Using Schiff-Base Reaction with Acetone In addition to exploring the direct detection of Duomeen O using PS ionization, chemical reactions that form stable adducts can be used in conjunction with PS-MS to enhance the selectivity and detection of analyte(s) in complex mixtures. As such, experiments of this type (reactive-PS-MS) were employed in this study to improve the analysis of Duomeen O in complex water samples. A volume of 10 μL of acetone was

112 spotted in situ onto the paper simultaneously with application of 10 μL of methanol solvent as shown in Figure 4.2. Intense mass spectra containing protonated molecular ion [M+H]+ of Duomeen O at m/z 325 were observed (Figure 4.9A) when only methanol was applied on a filter paper to which Duomeen O had previously been applied. In contrast, applying acetone in tandem with methanol resulted in a completely different mass spectrum (Figure 4.9C) where the nucleophilic attachment of the carbonyl group in acetone by the primary amine group in Duomeen O yielded a reaction product with MW 364 Da and concomitant loss of water (Figure 4.10). The protonated ion of the reaction product is subsequently detected at m/z 365. Collisional activation of the ion at m/z 365 in CID affords product ions m/z 322 (minor) and m/z 294 (major) through sequential elimination of ethenamine (MW 43 Da) and ethylene (MW 28 Da), respectively, as shown in Figure 4.9D. This reactive PS experiment provides reliable complementary chemical information which facilitates polyamine and amine corrosion inhibitor formulation identification in complex matrixes with enhanced selectivity.

(A) (B)

Figure 4.9. Positive ion mode reactive-PS mass spectrum Duomeen O analyzed using a benchtop instrument: (A,B) typical Duomeen O mass spectrum analyzed without the acetone reagent and MS/MS CID data, respectively, while parts B and D show the product of the Duomeen O reaction with acetone detected in open air. 113

Figure 4.10. Schiff-Base Condensation Reaction of the Primary Amines. (i) Nucleophilic reaction between the primary amine and ketone and (ii) reaction between Duomeen O (n-oleyl-1,3-diamine propane) (MW 324) and acetone in gas phase under ambient conditions using Reactive-PS-MS.

The introduction of reagents in normal PS-MS experiments produce selective detection; when used in combination with tandem MS, this approach provides the confirmation needed to identify the presence of a particular substance. From these experiments, two reactions occurred on the surface in open air: (1) the nonspecific proton transfer reaction forming protonated molecules [M+H]+ (Figure 4.9A), and (2) the Schiff- base reaction (Figure 4.9B). It is interesting to note that this condensation reaction between acetone and the amine proceeded rapidly (in less than 5 s) to enable analysis in real time. This reaction time scale is consistent with accelerated reaction rates observed for thin film/charged microdroplet reaction conditions.273,274 Paper spray ionization is a particularly simple ambient ionization technique which can be employed in the field to measure trace constituents of complex mixtures. Although analysis in MS/MS mode adequately removes matrix effects, a decision needs to be made as to what analyte ion within the mixture should be subjected to collisional activation. In this respect, preforming real time chemical reactions onsite will offer an efficient means to eliminate unrelated matrix ions. The

114 generation of a charged product is expected to improve ionization efficiency in a process once known as “reverse derivatization”.254 The combined derivatization/ionization process is tested in this study for the analysis of Duomeen O. As such both ionization efficiency and molecular selectivity can be improved by chemical derivatization such as the Schiff base reaction.

4.4.2 Characterization and Identifications of Protonated Metaldehyde Molecular Ion Species using Formic Acid From the results observed in Figure 4.6, it can be hypothesized that the sodium + + [M+Na] and ammonium [M+NH4] ions masked the protonation of metaldehyde. The introduction of reactive reagents in the PS spray solvent can improve the selective detection of metaldehyde in water; when used in combination with tandem MS, this approach can provide the confirmation needed to identify the presence of a particular substance in a complex mixture. This objective was achieved by adding acidified water (0.1% formic acid) to the methanol spray solvent MeOH:(H2O+0.1% formic acid) (1:1, v/v). The addition of the acidified water greatly suppressed cationization (i.e. [M+Na]+ and + [M+NH4] adduction) and aided protonation. The resultant PS-MS mass spectrum recorded when 5 μ g of metaldehyde in 1 μ L of deionized water was deposited on the paper substrate using MeOH:(H2O+0.1% formic acid) (1:1, v/v) as the PS solvent is shown in Figure 4.11. An intense, intact protonated molecular ion [M+H]+ of metaldehyde at m/z 177, including a major fragment ion at m/z 149, were observed in the single stage MS analysis (Figure 4.11A). This fragment ion (m/z 149) appears to be formed from the elimination of ethylene

(CH2=CH2, MW 28 Da), even prior to collisional activation suggesting a ring opening/rearrangement process in the presence of formic acid.

115

(A)

(B)

116

Figure 4.12. Positive ion mode paper spray mass spectrum of metaldehyde-d16 using acidified spray solvent. Inset (i) shows CID data for the precursor ion at m/z 193.

This observation was further investigated in two experiments: (i) studies of gas- phase fragmentation patterns in tandem MS experiments and (ii) detection of paraldehyde under acidified spray solvent conditions. First, the structure of the protonated metaldehyde and its dissociation behavior were characterized after collisional activation. Insert (i) of Figure 4.11A shows the product ion MS/MS mass spectra of the protonated metaldehyde. Unlike the sodiated molecular ion, which fragmented to give sodiated acetaldehyde (Figure 4.6), the protonated metaldehyde species dissociates predominantly via the loss of

CH2=CH2 to yield a product ion peak at m/z 149. This fragmentation pathway indicates that the ion at m/z 149, observed in the full MS spectrum, is related to the metaldehyde protonated species and supports our suggestion that the sodiated molecular ions are formed with minimal internal energy deposition. This behavior was also observed using acidified spray solvent for the isotopically labelled metaldehyde-d16, yielding protonated molecular ions with a fragmentation pathway that also suggests a similar ring opening has occurred

117

(Figure 4.12). In addition, CID of m/z 149 directly from the solution was compared with the fragmentation of gas-phase m/z 149 ion formed from MS2 of m/z 177 ion. The two spectra are similar where the mass of the main neutral loss is 28 Da providing an abundant ion peak at m/z 121 (Figure 4.13). This result suggests that the m/z 149 ion generated in solution is the same in structure as the m/z 149 ion produced in gas-phase under CID. The second experiment to confirm the observed behavior of the protonated metaldehyde involved the use of paraldehyde, a cyclic trimer of acetaldehyde (metaldehyde being the corresponding tetramer). Figure 4.11B shows the positive ion mode mass spectrum of paraldehyde obtained when 5 μg of the sample was deposited on the paper substrate and sampled by using MeOH:(H2O+0.1% formic acid) (1:1, v/v) as the spray solvent. A stable protonated molecular ion [M+H]+ of paraldehyde at m/z 133 was observed. The structure of the protonated paraldehyde species was confirmed from CID fragmentation patterns as shown in insert (ii) (Fig. 3B) where the molecular ion yields an intense fragment ion at m/z 89 owing to the neutral loss of acetaldehyde (MW 44 Da). Like metaldehyde, the fragment ion at m/z 89 was observed in the single stage MS experiment. Other signals were also observed such as at m/z 223 and m/z 164 in Figure 4.11A,B respectively and their origins are not known.

118

Figure 4.13. Comparison of CID of m/z 149 formed (A) directly from the solution with (B) the fragmentation of gas-phase m/z 149 formed from the MS2 of m/z 177 (insert (i)).

The fragmentation pathway (m/z 177→149) for the protonated ion type was used to quantify metaldehyde in water (Figure 4.14). Using a commercial linear ion trap mass spectrometer, the LOD was determined to be 0.05 ng/mL. This quantitative analysis of metaldehyde in water was achieved from the metaldehyde calibration curve obtained with

PS-MS using MeOH:(H2O+0.1% formic acid) (1:1, v/v) as the PS solvent (Figure 4.14).

Following procedures established using LC-MS, deuterated atrazine-d5 (3 ppb, m/z 221→179) was chosen as the internal standard.271,275 Monitoring the analyte-to-internal standard ratio (A/IS) as a function of analyte concentration yielded good linearity (R2 > 0.99), precision (RSD < 10%) and > fifty-fold decrease in the detection limit for metaldehyde in water compared with normal PS-MS, which utilized sodiated ions in the quantification process (Table 4.1).

119

0.03

y = 0.0048x + 0.001 R² = 0.9986

0.02 A/IS

0.01

0 0 1 2 3 4 5 Concentration of Metaldehyde (ng/mL)

Figure 4.14. Calibration curve for quantification of metaldehyde in water using PS-MS/MS when analyzing protonated ion types produced in acidified spray solvent. Error bars indicate standard deviation from three replicates.

120

(A)

(B)

Figure 4.15. Positive ion mode paper mass spectrum for polyamine and amine corrosion inhibitor formulation complex mixture (competitor product A) analyzed using a benchtop mass spectrometer. (A) Mass spectrum of competitor product A corrosion inhibitor mixture analyzed without acetone reagent. A volume of 2 μL of the corrosion inhibitor mixtures was deposited onto the surface and ionized and analyzed in the open air by application of an electric potential of +3.5 kV positive ion mode. Insert (i)–(iii) are the MS/MS CID mass spectra for the m/z 325, m/z 337, m/z 351, respectively. (B) Mass spectrum of competitor product A corrosion inhibitor mixture analyzed with acetone reagent. The protonated ion of the reaction product is subsequently detected at m/z 365. Insert (iv) is the MS/MS CID mass spectra for the m/z 365.

121

4.5 Analysis of Environmental and Industrial Samples The goal of this study is to provide a robust and rapid analysis procedure for environmental and industrial water samples for the trace analysis of contaminants. The following section performs the analysis of water samples from high pressure boiler systems that contain Duomeen O and environmental samples from surface water sources in England containing metaldehyde.

4.5.1 Paper Spray Mass Spectrometry Analysis of Duomeen O in a Mixture of Polyamine Corrosion Inhibitors Direct analysis of the long chain Duomeen O in complex polyamine and amine mixtures using PS-MS was investigated without any sample preparation. Polyamine and amine complex mixtures including competitor product A, naylamul S11, and ascamine DW BRI were analyzed as supplied without further pretreatment. A volume of 2 μL from each sample was deposited onto the paper triangle and analyzed using a commercial benchtop mass spectrometer in positive ion mode as described in Figure 4.2. Figure 4.15A shows the recorded mass spectrum for the competitor product A (polyamine and amine mixture) (mass range 200− 500) using only methanol as the PS spray solvent. Intense protonated molecular ions of Duomeen O [M+H]+ at m/z 325 were observed and confirmed by MS/MS CID experiments (insert (i) in Figure 4.15). Two unidentified peaks at m/z 337 and 351 were also observed, and MS/MS experiments (inserts (ii) and (iii), Figure 4.15) showed that they are unrelated to Duomeen O. This decision was supported by reactive paper spray experiments in which only the peak corresponding to Duomeen O (m/z 325) was observed to be affected by the presence of acetone, with the concomitant appearance of an ion at m/z 365 (Figure 4.15B). This product ion has previously been identified as coming from a reaction between acetone and Duomeen O (Figure 4.9B) using purified samples. Similarly, the remaining polyamine and amine corrosion inhibitor mixtures (i.e., naylamul S11 and ascamine) were analyzed using PS-MS and Duomeen O was detected and characterized (Figure 4.16). The ability to detect and characterize Duomeen O in a variety of different raw crude boiler water samples collected from a water tube boiler plant waste treatment facility

122

(Coventry, U.K.) has been demonstrated. In this experiment 2 μL from each sample was deposited on the paper substrate and analyzed using PS-MS. Figure 4.17 shows the recorded mass spectra for (a) condensate water, (b) feedwater, and (c) boiler water. Moderately intense protonated molecular ions [M+H]+ of Duomeen O were observed and confirmed using MS/MS CID data as shown in Figure 4.17, inserts (i)−(iii) in condensate, feed, and boiler water samples. The identification of the Duomeen O molecule in a variety of crude water samples collected from a large HP water tube boiler plant demonstrate the utility of the PS-MS method for direct, rapid screening with little or no sample preparation. It is important to note that other protonated molecules for amine compounds such as cyclohexylamine (MW 99), diethyl amino ethanol (MW 117), were also detected and confirmed using MS/MS CID data in the feedwater and boiler water at m/z 100, 118 (Figure 4.17B,C).

(A)

(B)

Figure 4.16. Positive ion mode paper mass spectrum for polyamine and amine corrosion inhibitor formulation complex mixture analyzed using a bench-top mass spectrometer; (A) mass spectrum ascameen corrosion inhibitor mixture, (B) mass spectrum of naylamul S II corrosion inhibitor mixture. About 2 μL of the corrosion inhibitor mixtures was deposited onto the surface and ionized and analyzed in the open air by application of an electric potential of + 3.5 kV positive ion mode. Insert (i)-(ii) are the MS/MS CID mass spectra for the m/z 325. 123

(A)

(B)

(C)

124

One advantage of ambient ionization methods is their compatibility with high- throughput rapid screening. To implement successful screening experiments, the analytes of interest need to be carefully evaluated with respect to the matrix due to possible complications of ionization suppression and isobaric ion interference. In our experiments Duomeen O was among the low-abundance ions detected in the full mass spectra from the water samples (Figure 4.17). The same analyte concentrations were sensitively detected in MS/MS mode in which matrix effects are completely eliminated. As demonstrated in other PSMS experiments,235,238 the porous cellulose paper substrate used for ionization reduces/filters a large proportion of the particulate present in complex samples and reduces ion suppression effects without extensive sample preparation. Direct analysis of Duomeen O at very low concentrations (<0.1 pg absolute) in complex mixtures has been demonstrated using paper spray mass spectrometry. The MS/MS experiments, complimented by the reactive PS-MS method, provide a powerful means of qualitative analysis with little or no sample preparation. As demonstrated above, either tandem MS or reactive PS-MS can be used to analyze Duomeen O in complex mixtures. Since quantification was carried out in MS/MS mode, it was required to establish the fragmentation pattern of Duomeen O in collision-induced dissociation experiments. For complex mixtures, it is often difficult to identify species of interest; to improve the efficacy of the identification process for Duomeen O, a reactive paper spray approach was implemented in which acetone was added to the methanol/water spray solvent. Any mass shifts observed after the in-situ reaction with acetone signified the presence of an amine functional group, potentially from Duomeen O analyte in water, which can then be quantified in subsequent MS/MS experiments.

4.5.2 Direct Metaldehyde Quantitation in Environmental Water Samples Using Paper Spray Mass Spectrometry The two water samples were collected directly from Abberton reservoir (Essex, UK) and Chigwell brook (Essex, UK) without any filtration except for large objects (> 3 cm). Each sample had water turbidity levels of ~1.8 and ~0.79 NTU, total organic carbon ~6.5 and ~3.5 mg/L, and, pH ~8.35 and ~8.36, respectively. A volume of ~10 μL from

125 each raw sample was deposited onto the paper substrate and analyzed using a commercial benchtop mass spectrometer in positive ion mode. Figure 4.18 shows the recorded mass spectrum for analysis of the raw water samples (Chigwell Raw and Abberton Raw supplied by Northumbrian Water Ltd.) using either MeOH or MeOH:(H2O+0.1% formic acid) (1:1, v/v) as the PS spray solvent. Moderately intense protonated molecular ions of metaldehyde [M+H]+ at m/z 177 were observed in both water samples, and confirmed by MS/MS CID experiments (insert (i) & (ii) in Figure 4.18) for the reactive experiment, which utilized an acidic spray solvent. Expectedly the presence of metaldehyde could not be confirmed in the same water samples when analyzed with the ‘normal PS-MS’ using methanol as the spray solvent.

(A) (B)

Figure 4.18. Positive ion mode paper spray mass spectra for rapid detection of metaldehyde in raw water samples (supplied by Northumbrian Water) whereby a volume of ~10 μL of the sample was deposited onto the paper substrate and ionized in the open environment by application of an electric potential of +3.5 kV. Abberton Raw was analyzed according to (A) the ‘normal PS-MS’ method and (B) with reactive PSMS. Similarly for Chigwell Raw, ‘normal PS-MS’ analysis is shown in (C) and reactive PS-MS in (D). Inserts (i) & (ii) are the MS/MS CID mass spectra for the protonated metaldehyde ion at m/z 177 from each water sample analyzed using the reactive methodology. 126

Figure 4.19. Proposed mechanism of acid catalyzed metaldehyde ring opening.

4.6 Metaldehyde Fragmentation Pathway Discussion The results obtained for the direct analysis of metaldehyde in water are apparent of the relatively lower sensitivity (higher LOD) of metaldehyde detection using the normal PS-MS method. The detection limit for sodiated metaldehyde is < 3 ng/mL but is not suitable for in situ analysis due to the relatively high detection limit and the potential for salt concentration variations, which are likely to be encountered in the environment. Doping a reactive agent into the PS spray solvent enables reactions to occur at the sampling spot concurrently with mass spectra acquisition to aid both sensitivity and selectivity for target molecules present in complex mixtures. As such, experiments of this type (reactive PS-MS) were employed in this study to improve the detection of metaldehyde in water samples by more than an order of magnitude. Although the inclusion of a trace amount of acid is common practice for MS techniques to aid protonation; the addition of the acidified reagent in this case leads to ring opening (Figure 4.19), hence the reactive nature of this process. The high sensitivity of the protonated ion type is attributed to the occurrence of only one major fragment ion in CID. The ability to form new ion type(s) from metaldehyde simply by adding reactive reagents (i.e. formic acid) into the PS spray solvent introduces an opportunity to differentiate metaldehyde from other potentially interfering ions having the same nominal mass. This advantage is particularly important for field metaldehyde analysis in which the selectivity of the paper spray method can be increased by studying

127 the fragmentation patterns of sodiated (formed using neutral spray solvent) and protonated (formed using acidified spray solvent) metaldehyde species. To understand this process (i.e., why [M+H]+ fragments differently than [M+Na]+) it was necessary to investigate the structure/nature of the suspected ring “opening” product formed in the presence of formic acid. As indicated in the results section, the elimination of 28 Da from metaldehyde was assigned to a loss of CH2=CH2 neutral species as illustrated in Figure 4.19. This proposal is supported by the failure of acidified metaldehyde to react with hydroxylamine, both in solution and in-situ during reactive paper spray experiments. Product C is presumably formed via an internal proton hopping process and explains why both gas-phase CID and solution-phase rearrangements occur via a common ethylene loss.

Figure 4.20. Illustrative diagram showing reactive and “normal” PS-MS analysis of metaldehyde generating different ion types.

128

PS-MS performed in the tandem mass spectrometry mode can reduce the effect of matrix ion suppression. For quantification purposes, a decision needs to be made as to which ions within the mixture should be subjected to collisional activation. In this respect, performing real time chemical reactions onsite will offer an efficient means to eliminate unrelated matrix ions. The generation of a charged product is expected to improve ionization efficiency of analyte(s) of interest in a complex mixture such as the protonation of metaldehyde in water. The combined reaction/ionization process is tested in this study for the analysis of metaldehyde (Figure 4.20). As such both ionization efficiency and molecular selectivity can be improved by the addition of acidified reagents that can yield protonated molecular ions [M+H]+ for targeted analysis and quantification of metaldehyde in water samples (Figure 4.11). The ability to detect and characterize metaldehyde in raw water samples collected from natural water courses has been demonstrated. The concentration of metaldehyde in both water samples was cross-validated and confirmed to be < 0.1 μ g/L (the EU limit) using LC-MS. The detection limit obtained by PS-MS (Table 4.1) suggests that it could be suitable for the rapid detection of metaldehyde in raw water, although it registers higher values than those that can be obtained from GC- and LC-MS analytical methods.270,271 With the capability of PS-MS to perform in-situ analyses of unmodified samples, the methodology described in this study shows promise for use in routine onsite investigative applications where regular monitoring or rapid screening is required. In this experiment ~10 μL from each sample was deposited on the paper substrate and analyzed using PS-MS. Figure 4.18 shows the recorded mass spectra for Abberton Raw (A,B) and Chigwell Raw (C,D). Moderately intense protonated molecular ions [M+H]+ of metaldehyde were observed for the reactive experiment and confirmed using MS/MS CID data as shown in Figure 4.18, inserts (i) & (ii). The identification of the metaldehyde molecule in both water samples demonstrates the utility of the PS-MS method for direct, rapid screening with little or no sample preparation.

129

4.7 Summary Direct analysis of long chain aliphatic primary polyamines by PS-MS has been demonstrated in a variety of boiler water samples with little or no sample pretreatment in open air. The use of tandem mass spectrometry analysis assisted in confirming the identity of aliphatic primary amine (Duomeen O) in various boiler system water samples. Exact mass measurements using LTQ-Orbitrap further confirmed the Duomeen O molecule formula observed within 1 ppm mass accuracy. PS-MS ambient ionization is both sensitive and selective for the analysis of corrosion inhibitor formulations in boiler water samples. Linear signal responses with a dynamic range of 5 orders of magnitude were obtained. The LOD of 0.1 pg (absolute concentration) with reproducibility of RSD of <10% is noteworthy for the direct analysis of aliphatic primary polyamine and amine corrosion inhibitor formulations in crude large medium pressure (MP) water tube boiler plant samples. Furthermore, the Schiff-base reaction between the aliphatic primary polyamine (Duomeen O) and acetone complements the usefulness of PS-MS analyte molecules in complex sample mixtures. The simplicity of paper spray ionization and the ability to analyze raw boiler water samples without sample preparation further enhances the potential for coupling to a portable or miniaturized mass spectrometer for on-site analysis. Such a system in operation would be of great value in the water industry for quality control. Future work will consider PS ionization coupled to a portable miniature mass spectrometer for in- field characterization of different boiler water samples under ambient conditions. Online in situ monitoring of the water boiler system is the ultimate aim. Rapid and direct analysis of metaldehyde has been described using paper spray mass spectrometry. Sodiated [M+Na]+ and protonated [M+H]+ molecular ions produced under two different spray conditions (i.e. acidified MeOH:(H2O+0.1% formic acid) (1:1, v/v) and normal MeOH PS solvents) were characterized in which [M+Na]+ species were identified to fragment through sequential losses of dimeric and monomeric acetaldehyde neutral species, whereas [M+H]+ dissociates via the elimination of ethylene. Quantitation of metaldehyde was achieved at low concentration (0.05 ng/mL for [M+H]+ and 2.69 ng/mL for [M+Na]+) in water using the reactive PS ionization method with acidified spray solvent. The MS/MS experiment provides a powerful means of qualitative analysis and

130 confirmation of metaldehyde in water. The generation of different ion types in the specified spray conditions can offer an opportunity to readily discriminate (in the field) against other background ions with similar molecular weights since it is unlikely for a particular ion to fragment in a similar fashion as metaldehyde when using sodiated versus protonated parent ions in MS/MS. The demonstrated detection limit shows promise for the direct detection of metaldehyde in water at regulatory levels. Future work will involve further investigation/validation with untreated environmental water samples and pre-determined mock samples of varying water quality to determine potential ion suppression effects with a view to onsite in-situ analysis of metaldehyde and related environmental contaminants using a portable mass spectrometer. The objective is to translate the reactive methodology demonstrated with a commercial benchtop system, to a portable MS platform. Recent reports that couple ambient ionization methods, including PS, with portable mass spectrometers are promising.233,252,276–278 In this respect, portable ion trap technology is often preferred as tandem analysis can be performed in time, without increase in instrument footprint. However, with any portable system there is an inevitable trade-off between portability/field ruggedness and performance. For the purpose of onsite testing, in the context of water analysis, the trade-off is not as severe since the mass spectrometer can be confined to a vehicle without stringent restrictions on weight and power. The overarching goal is to achieve timely analysis, allowing near instant decisions to be made. Under current procedures, it can take up to 48 hours from sample collection until a determination is made. It is expected that such a portable setup will provide rapid analysis, being suitable for pre-screening and identifying local sources of pesticide contamination to inform operational decisions. Furthermore, due to the generic nature of MS, this methodology can be extended to other water pollutants that are of concern in the environment. As such, the results are significant beyond the analysis of metaldehyde discussed herein as they represent a means for rapid analysis of other environmental contaminants in water. When coupled with a miniature mass spectrometer, the directness of the PS-MS experiment itself and the reactive alternative make for a potentially attractive on-site technique for water analysis and environmental monitoring. Other techniques such as the

131

‘leaf spray’ variant of the paper spray experiment39,40 can benefit by adopting this method for the analysis and determination of metaldehyde and other chemicals of concern on crops such as vegetables that may have been treated with pesticides and/or molluscicide.36,270

132

Chapter 5. Conclusions

5.1 Summary The ionization methods in the preceding chapters outline new methods for effective microsampling, more sensitive and rapid analysis, and enhanced ambient storage of target compounds. Each project is summarized below: In chapter 2, a hydrophobic paper substrate serves as a fresh and dried blood spheroid collection and analysis platform. Direct analysis of small organic compounds from fresh blood was demonstrated by performing on-line liquid/ liquid extraction on the hydrophobic paper surface. The dried blood spheroid sample collection platform showed increased stability for hydrolytically labile compounds (diazepam, cocaine and benzoylecgonine) against oxidative stress. In the case of cocaine, signal lifetime increased from days to several weeks under ambient conditions and without cold storage. Because cold storage is not required, use of this storage strategy is viable in rural settings where accessibility to cold storage is not as available. Additionally, the act of shipping samples does not require cold packs, which allows for longer-distance shipping and less resource consumption. In chapter 3, 2D wax-printed paper substrates was described. Wax printing creates hydrophobic barriers, which reduce the area of wetting on the paper triangle, resulting in the minimization of spray solvent evaporation. Because solvent is maintained, spray time is increased from 1.5 minutes to 10 minutes without increase in applied solvent volume, external solvent reservoirs, or pumps. Additionally, analyte quantification was able to be performed with 0.5-1 kV spray voltage. Analytes detected include corrosion inhibitor Duomeen, pesticide metaldehyde, and illicit drugs such as cocaine, benzoylecgonine, methamphetamine, and amphetamine. Finally, in chapter 4, direct analysis of long chain aliphatic primary polyamines by PS-MS has been demonstrated in a variety of boiler water. Additionally, detection of

133 metaldehyde in a sodiated and protonated form was performed from waters samples. The use of tandem mass spectrometry analysis assisted in confirming the identity of aliphatic primary amine (Duomeen O) in various boiler system water samples. By using reactive paper spray, the Schiff-base reaction between the aliphatic primary polyamine (Duomeen O) and acetone provides a secondary confirmation for detection.

5.2 Future Directions With regard to chapter 2, further work in hydrophobic paper spray can be expanded to analyze other small molecules or activity of other bloodborne enzymes that are commonly used in diagnostic situations. Analysis of enhance stability of acylcarnitines in dried blood spheroids can be more medically relevant. These acylcarnitines indicate presence of metabolic disorders in newborns before symptoms manifest that may be irreversible. Because hydrophobic paper spray ionization does not require a drying step or sample pretreatment or separation, turnaround time from drawing a patient’s sample to analysis and test results could be minutes instead of days. By manipulating solvent surface tension determination of surface energy of the porous hydrophobic paper substrate is possible. A major pitfall of data shown in Chapter 2 may be as follows: the data pertaining to surface energy determination found in 2.11 Surface Energy Analysis is possibly inaccurate due to inconsistent molecular structures of the fibrous medium from which ionization occurs. (Figure 2.41) The assumption molecular structure of the support does not affect surface tension determination is due to Taylor cone formation on the liquid surface rather than the polymer or paper surface. This results in the assumption that the support’s only effect on ion intensity is the degree of wetting of the acetonitrile/water spray solvent due to the surface energy of the molecular structure but does not include possible binding of benzoylecgonine due to molecular affinity to the substrate resulting in partitioning between the solid support phase and the liquid extraction phase. For example: polychloroprene has a very similar surface energy compared to cellulose acetate. However, an organic molecule such as benzoylecgonine may have a higher affinity to the cellulose acetate than the polychloroprene. Therefore, this method is only valid if the target molecule benzoylecgonine binds similarly to all polymers. Similar

134 binding is very unlikely; for example, benzoylecgonine has a protruding benzene ring that may interact more with polycarbonate compared to polyacrylonitrile, increasing analyte binding and decreasing ion intensity varying with acetonitrile/water composition. Additionally, because electrostatic spray depends on a capacitive coupling between the high voltage source and the sample,117 polymer composition may affect electrostatic spray due to varying capacitances of the support substrate. However, these effects may be accounted for in the undefined fitting parameters, and further experiments and calculations would be required to define these variables. It is clear that the overall shape of the function follows the 1/(a2-b2) pattern, and the physical significance of each variable described in the related section does follow current understanding of paper spray ionization. Additionally, although molecular binding between benzoylecgonine and the polymer substrate is likely to increase/decrease ion intensity not as a function of surface tension of solvent, but rather as a constant amount or a slope that will not shift the peak surface tension, resulting in no change to the above surface tension determination. However, future continuation of this work in the study of the effect should still be explored at least experimentally by comparing ion intensity on equal surface energy polymers/substrates and by using additional analytes for surface energy study. Overall, a systematic approach is needed to fully mathematically characterize each individual process contributing to ion intensity, including: (i) effect of surface energy on analyte binding, (ii) partitioning of analyte between surface and solvent, (iii) ionization efficiency, and (iv) Taylor cone/electrostatic stability and effects. Additionally, if the molecule is in a matrix, in addition to the analyte- surface and analyte- solvent interaction, the analyte- matrix effect must be studied in wetting and non-wetting situations. It is likely that most of these processes can be performed offline in an analytical approach for later equation fitting. In chapter 4, low voltage analysis of paper spray was demonstrated, which will aid field analysis. Future work will focus on mathematical modeling and computer simulation to investigate the mechanism of fluid flow and how confined DC potentials influence analyte ionization and transport from the tip of the wax-printed paper substrates. Additionally, development of a 3D wax-printed microfluidic platform is possible.

135

Similarly, chapter 5 outlines an objective toward portable mass spectrometry for the analysis of environmental and industrial water samples. Further work in this subject can include analysis of raw boiler water samples without sample preparation while coupled to a portable or miniaturized mass spectrometer for on-site analysis. Other future work will include investigation/validation of real environmental water samples that may have been contaminated with pesticides or other environmental contaminants. This is important, because ion suppression effects from real samples containing interfering molecules is possible, especially if pairing with a portable or miniature mass spectrometer In general, the major objective is to modify paper spray such that methodologies demonstrated using a commercial benchtop system can be used on a portable or miniature mass spectrometer. While such a portable setup provides rapid on-site analysis, lower power requirements, limited mass range, and decreased sensitivity requires enhanced ionization for a suitable reading to be produced. For example, benefits include pre- screening and identification of pesticide contamination, which allows real-time operational decisions and reactionary measures to decrease environmental impact. Another example may be to increase or decrease dosage of pharmaceuticals for patients who are being treated for a disease or diagnosis of that disease while the patient has been admitted minutes prior or may be asymptomatic.

136

References

(1) Zimmermann, E. The Dried Blood Test for Syphilis. Munch. Med. Wochenschr. 1939, 86, 1732–1733. (2) Adam, B. W.; Alexander, J. R.; Smith, S. J.; Chace, D. H.; Loeber, J. G.; Elvers, L. H.; Hannon, W. H. Recoveries of Phenylalanine from Two Sets of Dried-Blood Spot Reference Materials: Predition from Hematocrit, Spot Volume, and Paper Matrix. Clin. Chem. 2000, 46 (1), 126– 128. (3) Mei, J. V.; Alexander, J. R.; Adam, B. W.; Hannon, W. H. Use of Filter Paper for the Collection and Analysis of Human Whole Blood Specimens. J. Nutr. 2001, 131 (5), 1631S- 1636S. https://doi.org/10.1093/jn/131.5.1631S. (4) Holub, M.; Tuschl, K.; Ratschmann, R.; Strnadová, K. A.; Mühl, A.; Heinze, G.; Sperl, W.; Bodamer, O. A. Influence of Hematocrit and Localisation of Punch in Dried Blood Spots on Levels of Amino Acids and Acylcarnitines Measured by Tandem Mass Spectrometry. Clin. Chim. Acta 2006, 373 (1–2), 27–31. https://doi.org/10.1016/j.cca.2006.04.013. (5) Vu, D. H.; Koster, R. A.; Alffenaar, J. W. C.; Brouwers, J. R. B. J.; Uges, D. R. A. Determination of Moxifloxacin in Dried Blood Spots Using LC–MS/MS and the Impact of the Hematocrit and Blood Volume. J. Chromatogr. B 2011, 879 (15–16), 1063–1070. https://doi.org/10.1016/j.jchromb.2011.03.017. (6) Cobb, Z.; de Vries, R.; Spooner, N.; Williams, S.; Staelens, L.; Doig, M.; Broadhurst, R.; Barfield, M.; van de Merbel, N.; Schmid, B.; et al. In-Depth Study of Homogeneity in DBS Using Two Different Techniques: Results from the EBF DBS-Microsampling Consortium. Bioanalysis 2013, 5 (17), 2161–2169. https://doi.org/10.4155/bio.13.171. (7) Hannon, W. H.; Clinical and Laboratory Standards Institute. Blood Collection on Filter Paper for Newborn Screening Programs: Approved Standard; Clinical and Laboratory Standards Institute: Wayne, PA, 2013. (8) Youhnovski, N.; Bergeron, A.; Furtado, M.; Garofolo, F. Pre-Cut Dried Blood Spot (PCDBS): An Alternative to Dried Blood Spot (DBS) Technique to Overcome Hematocrit Impact: Pre- Cut Dried Blood Spot: An Alternative to Dried Blood Spot. Rapid Commun. Mass Spectrom. 2011, 25 (19), 2951–2958. https://doi.org/10.1002/rcm.5182. (9) Capiau, S.; Wilk, L. S.; Aalders, M. C. G.; Stove, C. P. A Novel, Nondestructive, Dried Blood Spot-Based Hematocrit Prediction Method Using Noncontact Diffuse Reflectance Spectroscopy. Anal. Chem. 2016, 88 (12), 6538–6546. https://doi.org/10.1021/acs.analchem.6b01321. (10) Capiau, S.; Stove, V. V.; Lambert, W. E.; Stove, C. P. Prediction of the Hematocrit of Dried Blood Spots via Potassium Measurement on a Routine Clinical Chemistry Analyzer. Anal. Chem. 2013, 85 (1), 404–410. https://doi.org/10.1021/ac303014b. (11) De Kesel, P. M. M.; Capiau, S.; Stove, V. V.; Lambert, W. E.; Stove, C. P. Potassium-Based Algorithm Allows Correction for the Hematocrit Bias in Quantitative Analysis of Caffeine

137

and Its Major Metabolite in Dried Blood Spots. Anal. Bioanal. Chem. 2014, 406 (26), 6749– 6755. https://doi.org/10.1007/s00216-014-8114-z. (12) Damon, D. E.; Yin, M.; Allen, D. M.; Maher, Y. S.; Tanny, C. J.; Oyola-Reynoso, S.; Smith, B. L.; Maher, S.; Thuo, M. M.; Badu-Tawiah, A. K. Dried Blood Spheroids for Dry-State Room Temperature Stabilization of Microliter Blood Samples. Anal. Chem. 2018, 90 (15), 9353– 9358. https://doi.org/10.1021/acs.analchem.8b01962. (13) Damon, D. E.; Davis, K. M.; Moreira, C. R.; Capone, P.; Cruttenden, R.; Badu-Tawiah, A. K. Direct Biofluid Analysis Using Hydrophobic Paper Spray Mass Spectrometry. Anal. Chem. 2016, 88 (3), 1878–1884. https://doi.org/10.1021/acs.analchem.5b04278. (14) Fenn, J.; Mann, M.; Meng, C.; Wong, S.; Whitehouse, C. Electrospray Ionization for Mass Spectrometry of Large Biomolecules. Science 1989, 246 (4926), 64–71. https://doi.org/10.1126/science.2675315. (15) Yamashita, M.; Fenn, J. B. Electrospray Ion Source. Another Variation on the Free-Jet Theme. J. Phys. Chem. 1984, 88 (20), 4451–4459. https://doi.org/10.1021/j150664a002. (16) Yamashita, M.; Fenn, J. B. Negative Ion Production with the Electrospray Ion Source. J. Phys. Chem. 1984, 88 (20), 4671–4675. https://doi.org/10.1021/j150664a046. (17) Electrospray Ionization Mass Spectrometry: Fundamentals, Instrumentation, and Applications; Cole, R. B., Ed.; Wiley: New York, 1997. (18) Smith, D. P. H. The Electrohydrodynamic Atomization of Liquids. IEEE Trans. Ind. Appl. 1986, IA-22 (3), 527–535. https://doi.org/10.1109/TIA.1986.4504754. (19) Disintegration of Water Drops in an Electric Field. Proc. R. Soc. Lond. Math. Phys. Eng. Sci. 1964, 280 (1382), 383–397. https://doi.org/10.1098/rspa.1964.0151. (20) Ambient Ionization Mass Spectrometry; Domin, M., Cody, R., Royal Society of Chemistry (Great Britain), Eds.; New developments in mass spectrometry; Royal Society of Chemistry: Cambridge, UK, 2015. (21) Wei, Y.; Chen, L.; Zhou, W.; Chingin, K.; Ouyang, Y.; Zhu, T.; Wen, H.; Ding, J.; Xu, J.; Chen, H. Tissue Spray Ionization Mass Spectrometry for Rapid Recognition of Human Lung Squamous Cell Carcinoma. Sci. Rep. 2015, 5 (1). https://doi.org/10.1038/srep10077. (22) Harris, G. A.; Galhena, A. S.; Fernández, F. M. Ambient Sampling/Ionization Mass Spectrometry: Applications and Current Trends. Anal. Chem. 2011, 83 (12), 4508–4538. https://doi.org/10.1021/ac200918u. (23) Monge, M. E.; Harris, G. A.; Dwivedi, P.; Fernández, F. M. Mass Spectrometry: Recent Advances in Direct Open Air Surface Sampling/Ionization. Chem. Rev. 2013, 113 (4), 2269– 2308. https://doi.org/10.1021/cr300309q. (24) Nyadong, L.; Green, M. D.; De Jesus, V. R.; Newton, P. N.; Fernández, F. M. Reactive Desorption Electrospray Ionization Linear Ion Trap Mass Spectrometry of Latest- Generation Counterfeit Antimalarials via Noncovalent Complex Formation. Anal Chem 2007, 79, 2150–2157. (25) Harris, G. A.; Nyadong, L.; Fernandez, F. M. Recent Developments in Ambient Ionization Techniques for Analytical Mass Spectrometry. Analyst 2008, 133, 1297–1301. (26) Liu, X.-P.; Wang, H.-Y.; Zhang, J.-T.; Wu, M.-X.; Qi, W.-S.; Zhu, H.; Guo, Y.-L. Direct and Convenient Mass Spectrometry Sampling with Ambient Flame Ionization. Sci. Rep. 2015, 5 (1). https://doi.org/10.1038/srep16893. (27) Kozlowski, R. L.; Mitchell, T. W.; Blanksby, S. J. A Rapid Ambient Ionization-Mass Spectrometry Approach to Monitoring the Relative Abundance of Isomeric Glycerophospholipids. Sci. Rep. 2015, 5 (1). https://doi.org/10.1038/srep09243.

138

(28) Musah, R. A.; Espinoza, E. O.; Cody, R. B.; Lesiak, A. D.; Christensen, E. D.; Moore, H. E.; Maleknia, S.; Drijfhout, F. P. A High Throughput Ambient Mass Spectrometric Approach to Species Identification and Classification from Chemical Fingerprint Signatures. Sci. Rep. 2015, 5, 11520. (29) Takáts, Z.; Wiseman, J. M.; Gologan, B.; Cooks, R. G. Mass Spectrometry Sampling Under Ambient Conditions with Desorption Electrospray Ionization. Science 2004, 306 (5695), 471–473. (30) Chen, H.; Wortmann, A.; Zhang, W.; Zenobi, R. Rapid In Vivo Fingerprinting of Nonvolatile Compounds in Breath by Extractive Electrospray Ionization Quadrupole Time‐of‐Flight Mass Spectrometry. Angew Chem 2007, 119, 586–589. (31) Zhang, H.; Gu, H.; Yan, F.; Wang, N.; Wei, Y.; Xu, J.; Chen, H. Direct Characterization of Bulk Samples by Internal Extractive Electrospray Ionization Mass Spectrometry. Sci. Rep. 2013, 3 (1). https://doi.org/10.1038/srep02495. (32) Chen, H.; Yang, S.; Wortmann, A.; Zenobi, R. Neutral Desorption Sampling of Living Objects for Rapid Analysis by Extractive Electrospray Ionization Mass Spectrometry. Angew Chem 2007, 46, 7591–7594. (33) Chen, H.; Yang, S.; Li, M.; Hu, B.; Li, J.; Wang, J. Sensitive Detection of Native Proteins Using Extractive Electrospray Ionization Mass Spectrometry. Angew. Chem. Int. Ed. 2010, 49 (17), 3053–3056. https://doi.org/10.1002/anie.200906886. (34) Li, M.; Ding, J.; Gu, H.; Zhang, Y.; Pan, S.; Xu, N.; Chen, H.; Li, H. Facilitated Diffusion of Acetonitrile Revealed by Quantitative Breath Analysis Using Extractive Electrospray Ionization Mass Spectrometry. Sci. Rep. 2013, 3 (1). https://doi.org/10.1038/srep01205. (35) Li, X.; Hu, B.; Ding, J.; Chen, H. Rapid Characterization of Complex Viscous Samples at Molecular Levels by Neutral Desorption Extractive Electrospray Ionization Mass Spectrometry. Nat Protoc 2011, 6, 1010–1025. (36) Jjunju, F. P. M.; Maher, S.; Li, A.; Badu-Tawiah, A. K.; Taylor, S.; Graham Cooks, R. Analysis of Polycyclic Aromatic Hydrocarbons Using Desorption Atmospheric Pressure Chemical Ionization Coupled to a Portable Mass Spectrometer. J. Am. Soc. Mass Spectrom. 2015, 26 (2), 271–280. https://doi.org/10.1007/s13361-014-1029-2. (37) Jjunju, F. P. M.; Maher, S.; Li, A.; Syed, S. U.; Smith, B.; Heeren, R. M. A.; Taylor, S.; Cooks, R. G. Hand-Held Portable Desorption Atmospheric Pressure Chemical Ionization Ion Source for in Situ Analysis of Nitroaromatic Explosives. Anal. Chem. 2015, 87 (19), 10047–10055. https://doi.org/10.1021/acs.analchem.5b02684. (38) Yang, S.; Ding, J.; Zheng, J.; Hu, B.; Li, J.; Chen, H.; Zhou, Z.; Qiao, X. Detection of Melamine in Milk Products by Surface Desorption Atmospheric Pressure Chemical Ionization Mass Spectrometry. Anal. Chem. 2009, 81 (7), 2426–2436. https://doi.org/10.1021/ac900063u. (39) Cody, R. B.; Laramée, J. A.; Durst, H. D. Versatile New Ion Source for the Analysis of Materials in Open Air under Ambient Conditions. Anal. Chem. 2005, 77 (8), 2297–2302. https://doi.org/10.1021/ac050162j. (40) Nilles, J. M.; Connell, T. R.; Durst, H. D. Quantitation of Chemical Warfare Agents Using the Direct Analysis in Real Time (DART) Technique. Anal Chem 2009, 81, 6744–6749. (41) Li, L.-P.; Feng, B.-S.; Yang, J.-W.; Chang, C.-L.; Bai, Y.; Liu, H.-W. Applications of Ambient Mass Spectrometry in High-Throughput Screening. The Analyst 2013, 138 (11), 3097–3103. https://doi.org/10.1039/c3an00119a. (42) Wu, C.; Dill, A. L.; Eberlin, L. S.; Cooks, R. G.; Ifa, D. R. Mass Spectrometry Imaging under Ambient Conditions. Mass Spectrom Rev 2013, 32, 218–243.

139

(43) Soparawalla, S.; Tadjimukhamedov, F. K.; Wiley, J. S.; Ouyang, Z.; Cooks, R. G. In Situ Analysis of Agrochemical Residues on Fruit Using Ambient Ionization on a Handheld Mass Spectrometer. Analyst 2011, 136, 4392–4396. (44) Wang, H.; Liu, J.; Cooks, R. G.; Ouyang, Z. Paper Spray for Direct Analysis of Complex Mixtures Using Mass Spectrometry. Angew. Chem. 2010, 122 (5), 889–892. https://doi.org/10.1002/ange.200906314. (45) Manicke, N.; Abu-Rabie, P.; Spooner, N.; Ouyang, Z.; Cooks, R. G. Quantitative Analysis of Therapeutic Drugs in Dried Blood Spot Samples by Paper Spray Mass Spectrometry: An Avenue to Therapeutic Drug Monitoring. J. Am. Soc. Mass Spectrom. 2011, 22 (9), 1501– 1507. https://doi.org/10.1007/s13361-011-0177-x. (46) Espy, R. D.; Manicke, N. E.; Ouyang, Z.; Cooks, R. G. Rapid Analysis of Whole Blood by Paper Spray Mass Spectrometry for Point-of-Care Therapeutic Drug Monitoring. Analyst 2012, 137, 2344–2349. (47) Yang, Q.; Manicke, N. E.; Wang, H.; Petucci, C.; Cooks, R. G.; Ouyang, Z. Direct and Quantitative Analysis of Underivatized Acylcarnitines in Serum and Whole Blood Using Paper Spray Mass Spectrometry. Anal. Bioanal. Chem. 2012, 404 (5), 1389–1397. https://doi.org/10.1007/s00216-012-6211-4. (48) Wang, H.; Manicke, N. E.; Yang, Q.; Zheng, L.; Shi, R.; Cooks, R. G.; Ouyang, Z. Direct Analysis of Biological Tissue by Paper Spray Mass Spectrometry. Anal. Chem. 2011, 83 (4), 1197–1201. https://doi.org/10.1021/ac103150a. (49) Zhang, Y.; Ju, Y.; Huang, C.; Wysocki, V. H. Paper Spray Ionization of Noncovalent Protein Complexes. Anal. Chem. 2014, 86 (3), 1342–1346. https://doi.org/10.1021/ac403383d. (50) Taverna, D.; Di Donna, L.; Mazzotti, F.; Policicchio, B.; Sindona, G. High‐throughput Determination of Sudan Azo‐dyes within Powdered Chili Pepper by Paper Spray Mass Spectrometry. J Mass Spectrom 2013, 48, 544–547. (51) Mazzotti, F.; Di Donna, L.; Taverna, D.; Nardi, M.; Aiello, D.; Napoli, A.; Sindona, G. Evaluation of Dialdehydic Anti-Inflammatory Active Principles in Extra-Virgin Olive Oil by Reactive Paper Spray Mass Spectrometry. Int. J. Mass Spectrom. 2013, 352, 87–91. https://doi.org/10.1016/j.ijms.2013.07.012. (52) Zhang, Z.; Cooks, R. G.; Ouyang, Z. Paper Spray: A Simple and Efficient Means of Analysis of Different Contaminants in Foodstuffs. Analyst 2012, 137 (11), 2556–2258. https://doi.org/10.1039/c2an35196j. (53) Deng, J.; Yang, Y. Chemical Fingerprint Analysis for Quality Assessment and Control of Bansha Herbal Tea Using Paper Spray Mass Spectrometry. Anal Chim Acta 2013, 785, 82– 90. (54) Li, A.; Wei, P.; Hsu, H.-C.; Cooks, R. G. Direct Analysis of 4-Methylimidazole in Foods Using Paper Spray Mass Spectrometry. Analyst 2013, 138, 4624–4630. (55) Hamid, A. M.; Jarmusch, A. K.; Pirro, V.; Pincus, D. H.; Clay, B. G.; Gervasi, G.; Cooks, R. G. Rapid Discrimination of Bacteria by Paper Spray Mass Spectrometry. Anal. Chem. 2014, 86 (15), 7500–7507. https://doi.org/10.1021/ac501254b. (56) Reeber, S. L.; Gadi, S.; Huang, S.-B.; Glish, G. L. Direct Analysis of Herbicides by Paper Spray Ionization Mass Spectrometry. Anal Methods 2015, 7, 9808–9816. (57) Shen, L.; Zhang, J.; Yang, Q.; Manicke, N. E.; Ouyang, Z. High Throughput Paper Spray Mass Spectrometry Analysis. Clin Chim Acta 2013, 420, 28–33. (58) Narayanan, R.; Sarkar, D.; Cooks, R. G.; Pradeep, T. Molecular Ionization from Carbon Nanotube Paper. Angew. Chem. Int. Ed. 2014, 53 (23), 5936–5940. https://doi.org/10.1002/anie.201311053. 140

(59) Zhang, C.; Manicke, N. E. Development of a Paper Spray Mass Spectrometry Cartridge with Integrated Solid Phase Extraction for Bioanalysis. Anal. Chem. 2015, 87 (12), 6212–6219. https://doi.org/10.1021/acs.analchem.5b00884. (60) Zheng, Y.; Zhang, X.; Yang, H.; Liu, X.; Zhang, X.; Wang, Q.; Zhang, Z. Facile Preparation of Paper Substrates Coated with Different Materials and Their Applications in Paper Spray Mass Spectrometry. Anal. Methods 2015, 7 (13), 5381–5386. https://doi.org/10.1039/C5AY00874C. (61) Wang, Q.; Zheng, Y.; Zhang, X.; Han, X.; Wang, T.; Zhang, Z. A Silica Coated Paper Substrate: Development and Its Application in Paper Spray Mass Spectrometry for Rapid Analysis of Pesticides in Milk. Analyst 2015, 140 (23), 8048–8056. https://doi.org/10.1039/C5AN01823D. (62) Wang, X.; Zheng, Y.; Wang, T.; Xiong, X.; Fang, X.; Zhang, Z. Metal–Organic Framework Coated Paper Substrates for Paper Spray Mass Spectrometry. Anal Methods 2016, 8 (45), 8004–8014. https://doi.org/10.1039/C6AY02123A. (63) Wang, X.; Chen, Y.; Zheng, Y.; Zhang, Z. Study of Adsorption and Desorption Performances of Zr-Based Metal–Organic Frameworks Using Paper Spray Mass Spectrometry. Materials 2017, 10 (7), 769. https://doi.org/10.3390/ma10070769. (64) Wang, T.; Zheng, Y.; Wang, X.; Austin, D. E.; Zhang, Z. Sub-Ppt Mass Spectrometric Detection of Therapeutic Drugs in Complex Biological Matrixes Using Polystyrene- Microsphere-Coated Paper Spray. Anal. Chem. 2017, 89 (15), 7988–7995. https://doi.org/10.1021/acs.analchem.7b01296. (65) Liu, J.; He, Y.; Chen, S.; Ma, M.; Yao, S.; Chen, B. New Urea-Modified Paper Substrate for Enhanced Analytical Performance of Negative Ion Mode Paper Spray Mass Spectrometry. Talanta 2017, 166, 306–314. https://doi.org/10.1016/j.talanta.2017.01.076. (66) Espy, R. D.; Muliadi, A. R.; Ouyang, Z.; Cooks, R. G. Spray Mechanism in Paper Spray Ionization. Int. J. Mass Spectrom. 2012, 325–327, 167–171. https://doi.org/10.1016/j.ijms.2012.06.017. (67) Thermo Fisher Scientific. Thermo Scientific LTQ Velos Dual-Pressure Linear Ion Trap http://www.scispec.co.th/brochure/LCMS/Velos_PRO_BR.pdf. (68) Schmidt, V. Ivar Christian Bang (1869-1918), Founder of Modern Clinical Microchemistry. Clin. Chem. 1986, 32 (1), 213–215. (69) Chapman, O. D. THE COMPLEMENT-FIXATION TEST FOR SYPHILIS: USE OF PATIENT’S WHOLE BLOOD DRIED ON FILTER PAPER. Arch. Dermatol. Syphilol. 1924, 9 (5), 607. https://doi.org/10.1001/archderm.1924.02360230067010. (70) Chen, S.; Wan, Q.; Badu-Tawiah, A. K. Mass Spectrometry for Paper-Based Immunoassays: Toward On-Demand Diagnosis. J. Am. Chem. Soc. 2016, 138 (20), 6356–6359. https://doi.org/10.1021/jacs.6b02232. (71) Seidenberg, P.; Nicholson, S.; Schaefer, M.; Semrau, K.; Bweupe, M.; Masese, N.; Bonawitz, R.; Chitembo, L.; Goggin, C.; Thea, D. M. Early Infant Diagnosis of HIV Infection in Zambia through Mobile Phone Texting of Blood Test Results. Bull. World Health Organ. 2012, 90 (5), 348–356. (72) Crossle, J. R.; Elliot, R. B.; Smith, P. A. DRIED-BLOOD SPOT SCREENING FOR CYSTIC FIBROSIS IN THE NEWBORN. The Lancet 1979, 313 (8114), 472–474. https://doi.org/http://dx.doi.org/10.1016/S0140-6736(79)90825-0. (73) Zytkovicz, T. H.; Fitzgerald, E. F.; Marsden, D.; Larson, C. A.; Shih, V. E.; Johnson, D. M.; Strauss, A. W.; Comeau, A. M.; Eaton, R. B.; Grady, G. F. Tandem Mass Spectrometric

141

Analysis for Amino, Organic, and Fatty Acid Disorders in Newborn Dried Blood Spots. Clin. Chem. 2001, 47 (11), 1945–1955. (74) Li, Y.; Scott, C. R.; Chamoles, N. A.; Ghavami, A.; Pinto, B. M.; Turecek, F.; Gelb, M. H. Direct Multiplex Assay of Lysosomal Enzymes in Dried Blood Spots for Newborn Screening. Clin. Chem. 2004, 50 (10), 1785–1796. https://doi.org/10.1373/clinchem.2004.035907. (75) Gelb, M. H.; Turecek, F.; Scott, C. R.; Chamoles, N. A. Direct Multiplex Assay of Enzymes in Dried Blood Spots by Tandem Mass Spectrometry for the Newborn Screening of Lysosomal Storage Disorders. J. Inherit. Metab. Dis. 2006, 29 (2), 397–404. https://doi.org/10.1007/s10545-006-0265-4. (76) Sherman, G. G.; Stevens, G.; Jones, S. A.; Horsfield, P.; Stevens, W. S. Dried Blood Spots Improve Access to HIV Diagnosis and Care for Infants in Low-Resource Settings. J. Acquir. Immune Defic. Syndr. 2005, 38 (5). (77) Sia, S. K.; Linder, V.; Parviz, B. A.; Siegel, A.; Whitesides, G. M. An Integrated Approach to a Portable and Low-Cost Immunoassay for Resource-Poor Settings. Angew. Chem. Int. Ed. 2004, 43 (4), 498–502. https://doi.org/10.1002/anie.200353016. (78) Nemiroski, A.; Christodouleas, D. C.; Hennek, J. W.; Kumar, A. A.; Maxwell, E. J.; Fernandez- Abedul, M. T.; Whitesides, G. M. Universal Mobile Electrochemical Detector Designed for Use in Resource-Limited Applications. Proc. Natl. Acad. Sci. 2014, 111 (33), 11984–11989. https://doi.org/10.1073/pnas.1405679111. (79) Fu, E.; Liang, T.; Spicar-Mihalic, P.; Houghtaling, J.; Ramachandran, S.; Yager, P. Two- Dimensional Paper Network Format That Enables Simple Multistep Assays for Use in Low- Resource Settings in the Context of Malaria Antigen Detection. Anal. Chem. 2012, 84 (10), 4574–4579. https://doi.org/10.1021/ac300689s. (80) Yager, P.; Edwards, T.; Fu, E.; Helton, K.; Nelson, K.; Tam, M. R.; Weigl, B. H. Microfluidic Diagnostic Technologies for Global Public Health. Nature 2006, 442 (7101), 412–418. https://doi.org/10.1038/nature05064. (81) Li, M.; Tian, J.; Al-Tamimi, M.; Shen, W. Paper-Based Blood Typing Device That Reports Patient’s Blood Type “in Writing.” Angew. Chem. Int. Ed. 2012, 51 (22), 5497–5501. https://doi.org/10.1002/anie.201201822. (82) McDade, T. W.; Williams, S.; Snodgrass, J. J. What a Drop Can Do: Dried Blood Spots as a Minimally Invasive Method for Integrating Biomarkers into Population-Based Research. Demography 2007, 44 (4), 899–925. https://doi.org/10.1353/dem.2007.0038. (83) Burse, V. W.; Deguzman, M. R.; Korver, M. P.; Najam, A. R.; Williams, C. C.; Hannon, W. H.; Therrell, B. L. Preliminary Investigation of the Use of Dried-Blood Spots for the Assessment Ofin UteroExposure to Environmental Pollutants. Biochem. Mol. Med. 1997, 61 (2), 236– 239. https://doi.org/10.1006/bmme.1997.2603. (84) Edelbroek, P. M.; Heijden, J. van der; Stolk, L. M. L. Dried Blood Spot Methods in Therapeutic Drug Monitoring: Methods, Assays, and Pitfalls. Ther. Drug Monit. 2009, 31 (3). (85) Hoogtanders, K.; van der Heijden, J.; Christiaans, M.; Edelbroek, P.; van Hooff, J. P.; Stolk, L. M. L. Therapeutic Drug Monitoring of Tacrolimus with the Dried Blood Spot Method. J. Pharm. Biomed. Anal. 2007, 44 (3), 658–664. https://doi.org/10.1016/j.jpba.2006.11.023. (86) Denniff, P.; Spooner, N. The Effect of Hematocrit on Assay Bias When Using DBS Samples for the Quantitative Bioanalysis of Drugs. Bioanalysis 2010, 2 (8), 1385–1395. https://doi.org/10.4155/bio.10.103.

142

(87) Wagner, M.; Tonoli, D.; Varesio, E.; Hopfgartner, G. The Use of Mass Spectrometry to Analyze Dried Blood Spots. Mass Spectrom. Rev. 2016, 35 (3), 361–438. https://doi.org/10.1002/mas.21441. (88) Mei, J. V.; Zobel, S. D.; Hall, E. M.; De Jesús, V. R.; Adam, B. W.; Hannon, W. H. Performance Properties of Filter Paper Devices for Whole Blood Collection. Bioanalysis 2010, 2 (8), 1397–1403. (89) Demirev, P. A. Dried Blood Spots: Analysis and Applications. Anal. Chem. 2013, 85 (2), 779– 789. https://doi.org/10.1021/ac303205m. (90) Martinez, A. W.; Phillips, S. T.; Whitesides, G. M. Diagnostics for the Developing World: Microfluidic Paper-Based Analytical Devices. Anal. Chem. 2010, 82 (1), 3–10. https://doi.org/10.1021/ac9013989. (91) Wang, H.; Liu, J.; Cooks, R. G.; Ouyang, Z. Paper Spray for Direct Analysis of Complex Mixtures Using Mass Spectrometry. Angew Chem Int Ed 2010, 49 (5), 877–880. https://doi.org/10.1002/anie.200906314. (92) Wang, H.; Ren, Y.; McLuckey, M. N.; Manicke, N. E.; Park, J.; Zheng, L.; Shi, R.; Cooks, R. G.; Ouyang, Z. Direct Quantitative Analysis of Nicotine Alkaloids from Biofluid Samples Using Paper Spray Mass Spectrometry. Anal. Chem. 2013, 85 (23), 11540–11544. https://doi.org/10.1021/ac402798m. (93) Liu, J.; Wang, H.; Manicke, N. E.; Lin, J.-M.; Cooks, R. G.; Ouyang, Z. Development, Characterization, and Application of Paper Spray Ionization. Anal. Chem. 2010, 82 (6), 2463–2471. https://doi.org/10.1021/ac902854g. (94) Oradu, S. A.; Cooks, R. G. Multistep Mass Spectrometry Methodology for Direct Characterization of Polar Lipids in Green Microalgae Using Paper Spray Ionization. Anal. Chem. 2012, 84 (24), 10576–10585. https://doi.org/10.1021/ac301709r. (95) Gómez-Ríos, G. A.; Pawliszyn, J. Development of Coated Blade Spray Ionization Mass Spectrometry for the Quantitation of Target Analytes Present in Complex Matrices. Angew. Chem. Int. Ed. 2014, 53 (52), 14503–14507. https://doi.org/10.1002/anie.201407057. (96) Davis, K. M.; Badu-Tawiah, A. K. Direct and Efficient Dehydrogenation of Tetrahydroquinolines and Primary Amines Using Corona Discharge Generated on Ambient Hydrophobic Paper Substrate. J. Am. Soc. Mass Spectrom. 2017, 28 (4), 647–654. https://doi.org/10.1007/s13361-016-1516-8. (97) Damon, D. E.; Maher, Y. S.; Yin, M.; Jjunju, F. P. M.; Young, I. S.; Taylor, S.; Maher, S.; Badu- Tawiah, A. K. 2D Wax-Printed Paper Substrates with Extended Solvent Supply Capabilities Allow Enhanced Ion Signal in Paper Spray Ionization. Analyst 2016, 141 (12), 3866–3873. https://doi.org/10.1039/C6AN00168H. (98) Maher, S.; Jjunju, F. P. M.; Damon, D. E.; Gorton, H.; Maher, Y. S.; Syed, S. U.; Heeren, R. M. A.; Young, I. S.; Taylor, S.; Badu-Tawiah, A. K. Direct Analysis and Quantification of Metaldehyde in Water Using Reactive Paper Spray Mass Spectrometry. Sci. Rep. 2016, 6, 35643. (99) Yang, Q.; Wang, H.; Maas, J. D.; Chappell, W. J.; Manicke, N. E.; Cooks, R. G.; Ouyang, Z. Paper Spray Ionization Devices for Direct, Biomedical Analysis Using Mass Spectrometry. Int. J. Mass Spectrom. 2012, 312, 201–207. https://doi.org/10.1016/j.ijms.2011.05.013. (100) Bills, B. J.; Manicke, N. E. Development of a Prototype Blood Fractionation Cartridge for Plasma Analysis by Paper Spray Mass Spectrometry. Clin. Mass Spectrom. 2016, 2, 18–24. https://doi.org/10.1016/j.clinms.2016.12.002.

143

(101) Ren, Y.; Wang, H.; Liu, J.; Zhang, Z.; McLuckey, M. N.; Ouyang, Z. Analysis of Biological Samples Using Paper Spray Mass Spectrometry: An Investigation of Impacts by the Substrates, Solvents and Elution Methods. Chromatographia 2013, 76 (19), 1339–1346. https://doi.org/10.1007/s10337-013-2458-y. (102) Glavan, A. C.; Martinez, R. V.; Subramaniam, A. B.; Yoon, H. J.; Nunes, R. M. D.; Lange, H.; Thuo, M. M.; Whitesides, G. M. Omniphobic “RF Paper” Produced by Silanization of Paper with Fluoroalkyltrichlorosilanes. Adv. Funct. Mater. 2014, 24 (1), 60–70. https://doi.org/10.1002/adfm.201300780. (103) Glavan, A. C.; Christodouleas, D. C.; Mosadegh, B.; Yu, H. D.; Smith, B. S.; Lessing, J.; Fernández-Abedul, M. T.; Whitesides, G. M. Folding Analytical Devices for Electrochemical ELISA in Hydrophobic RH Paper. Anal. Chem. 2014, 86 (24), 11999–12007. https://doi.org/10.1021/ac5020782. (104) Fowkes, F. M. Contact Angle, Wettability, and Adhesion; Gould, R. F., Ed.; Advances in Chemistry; American Chemical Society: Washington D.C., 1964; Vol. 43. (105) de Gennes, P. G. Wetting: Statics and Dynamics. Rev. Mod. Phys. 1985, 57 (3), 827–863. https://doi.org/10.1103/RevModPhys.57.827. (106) Wong, M. Y.-M.; Tang, H.-W.; Man, S.-H.; Lam, C.-W.; Che, C.-M.; Ng, K.-M. Electrospray Ionization on Porous Spraying Tips for Direct Sample Analysis by Mass Spectrometry: Enhanced Detection Sensitivity and Selectivity Using Hydrophobic/Hydrophilic Materials as Spraying Tips. Rapid Commun. Mass Spectrom. 2013, 27 (6), 713–721. https://doi.org/10.1002/rcm.6497. (107) Duprat, C.; Protiere, S.; Beebe, A. Y.; Stone, H. A. Wetting of Flexible Fibre Arrays. Nature 2012, 482 (7386), 510–513. https://doi.org/10.1038/nature10779. (108) Bijlsma, L.; Boix, C.; Niessen, W. M. A.; Ibáñez, M.; Sancho, J. V.; Hernández, F. Investigation of Degradation Products of Cocaine and Benzoylecgonine in the Aquatic Environment. Sci. Total Environ. 2013, 443, 200–208. https://doi.org/10.1016/j.scitotenv.2012.11.006. (109) Sahin, A. Z.; Kalyon, M. The Critical Radius of Insulation in Thermal Radiation Environment. Heat Mass Transf. 2004, 40 (5), 377–382. https://doi.org/10.1007/s00231- 003-0471-7. (110) Aziz, A. The Critical Thickness of Insulation. Heat Transf. Eng. 1997, 18 (2), 61–91. https://doi.org/10.1080/01457639708939897. (111) Mendlowitz, M. The Specific Heat of Human Blood. Science 1948, 107 (2769), 97–98. https://doi.org/10.1126/science.107.2769.97. (112) Balasubramaniam, T. A.; Bowman, H. F. Thermal Conductivity and Thermal Diffusivity of Biomaterials: A Simultaneous Measurement Technique. J. Biomech. Eng. 1977, 99 (3), 148– 154. https://doi.org/10.1115/1.3426282. (113) Ren, Y.; McLuckey, M. N.; Liu, J.; Ouyang, Z. Direct Mass Spectrometry Analysis of Biofluid Samples Using Slug-Flow Microextraction Nano-Electrospray Ionization. Angew. Chem. Int. Ed. 2014, 53 (51), 14124–14127. https://doi.org/10.1002/anie.201408338. (114) Zeitlinger, M. A.; Derendorf, H.; Mouton, J. W.; Cars, O.; Craig, W. A.; Andes, D.; Theuretzbacher, U. Protein Binding: Do We Ever Learn? Antimicrob. Agents Chemother. 2011, 55 (7), 3067–3074. https://doi.org/10.1128/AAC.01433-10. (115) Schuhmacher, J.; B�hner, K.; Witt-Laido, A. Determination of the Free Fraction and Relative Free Fraction of Drugs Strongly Bound to Plasma Proteins. J. Pharm. Sci. 2000, 89 (8), 1008–1021. https://doi.org/10.1002/1520-6017(200008)89:8<1008::AID- JPS5>3.0.CO;2-B. 144

(116) Banker, M. J.; Clark, T. H.; Williams, J. A. Development and Validation of a 96-Well Equilibrium Dialysis Apparatus for Measuring Plasma Protein Binding. J. Pharm. Sci. 2003, 92 (5), 967–974. https://doi.org/10.1002/jps.10332. (117) Qiao, L.; Sartor, R.; Gasilova, N.; Lu, Y.; Tobolkina, E.; Liu, B.; Girault, H. H. Electrostatic- Spray Ionization Mass Spectrometry. Anal. Chem. 2012, 84 (17), 7422–7430. https://doi.org/10.1021/ac301332k. (118) Becker, E.; Hiller, W. J.; Kowalewski, T. A. Experimental and Theoretical Investigation of Large-Amplitude Oscillations of Liquid Droplets. J. Fluid Mech. 1991, 231 (1), 189. https://doi.org/10.1017/S0022112091003361. (119) Parolo, C.; Merkoçi, A. Paper-Based Nanobiosensors for Diagnostics. Chem Soc Rev 2013, 42 (2), 450–457. https://doi.org/10.1039/C2CS35255A. (120) Pollock, N. R.; Rolland, J. P.; Kumar, S.; Beattie, P. D.; Jain, S.; Noubary, F.; Wong, V. L.; Pohlmann, R. A.; Ryan, U. S.; Whitesides, G. M. A Paper-Based Multiplexed Transaminase Test for Low-Cost, Point-of-Care Liver Function Testing. Sci. Transl. Med. 2012, 4 (152), 152ra129. https://doi.org/10.1126/scitranslmed.3003981. (121) Saukkonen, J. J.; Cohn, D. L.; Jasmer, R. M.; Schenker, S.; Jereb, J. A.; Nolan, C. M.; Peloquin, C. A.; Gordin, F. M.; Nunes, D.; Strader, D. B.; et al. An Official ATS Statement: Hepatotoxicity of Antituberculosis Therapy. Am. J. Respir. Crit. Care Med. 2006, 174 (8), 935–952. https://doi.org/10.1164/rccm.200510-1666ST. (122) Price, J. C.; Thio, C. L. Liver Disease in the HIV-Infected Individual. Clin. Gastroenterol. Hepatol. Off. Clin. Pract. J. Am. Gastroenterol. Assoc. 2010, 8 (12), 1002–1012. https://doi.org/10.1016/j.cgh.2010.08.024. (123) Huang, X.-J.; Choi, Y.-K.; Im, H.-S.; Yarimaga, O.; Yoon, E.; Kim, H.-S. Aspartate Aminotransferase (AST/GOT) and Alanine Aminotransferase (ALT/GPT) Detection Techniques. Sensors 2006, 6 (7), 756–782. https://doi.org/10.3390/s6070756. (124) Sorbi, D.; Boynton, J.; Lindor, K. D. The Ratio of Aspartate Aminotransferase to Alanine Aminotransferase: Potential Value in Differentiating Nonalcoholic Steatohepatitis from Alcoholic Liver Disease. Am. J. Gastroenterol. 1999, 94 (4), 1018–1022. https://doi.org/10.1111/j.1572-0241.1999.01006.x. (125) Stephens, M. D. B. Asymptomatic Abnormal Liver Function Tests in Clinical Trials. Pharmacoepidemiol. Drug Saf. 1994, 3 (2), 91–103. https://doi.org/10.1002/pds.2630030206. (126) Boys, B. L.; Kuprowski, M. C.; Noël, J. J.; Konermann, L. Protein Oxidative Modifications During Electrospray Ionization: Solution Phase Electrochemistry or Corona Discharge- Induced Radical Attack? Anal. Chem. 2009, 81 (10), 4027–4034. https://doi.org/10.1021/ac900243p. (127) Inutan, E. D.; Trimpin, S. Matrix Assisted Ionization Vacuum (MAIV), a New Ionization Method for Biological Materials Analysis Using Mass Spectrometry. Mol. Cell. Proteomics 2013, 12 (3), 792–796. https://doi.org/10.1074/mcp.M112.023663. (128) Zenkiewicz, M. Methods for the Calculation of Surface Free Energy of Solids. J. Achiev. Mater. Manuf. Eng. 2007, 24 (1), 137–145. (129) Rafati, A. A.; Bagheri, A.; Najafi, M. Experimental Data and Correlation of Surface Tensions of the Binary and Ternary Systems of Water + Acetonitrile + 2-Propanol at 298.15 K and Atmospheric Pressure. J. Chem. Eng. Data 2010, 55 (9), 4039–4043. https://doi.org/10.1021/je1001329. (130) Rafati, A. A.; Bagheri, A.; Najafi, M. Experimental Data and Correlation of Surface Tensions of the Binary and Ternary Systems of Water + Acetonitrile + 2-Propanol at 298.15 145

K and Atmospheric Pressure. J. Chem. Eng. Data 2010, 55 (9), 4039–4043. https://doi.org/10.1021/je1001329. (131) Rayleigh, L. On the Capillary Phenomena of Jets. Proc. R. Soc. Lond. 1879, 29 (196–199), 71–97. https://doi.org/10.1098/rspl.1879.0015. (132) Lamb, H. Hydrodynamics, 6. ed., 3. Dover paperback ed.; Dover: New York, 1993. (133) Li, A.; Hollerbach, A.; Luo, Q.; Cooks, R. G. On-Demand Ambient Ionization of Picoliter Samples Using Charge Pulses. Angew. Chem. Int. Ed. 2015, 54 (23), 6893–6895. https://doi.org/10.1002/anie.201501895. (134) Martinez, A. W.; Phillips, S. T.; Butte, M. J.; Whitesides, G. M. Patterned Paper as a Platform for Inexpensive, Low-Volume, Portable Bioassays. Angew. Chem. Int. Ed. 2007, 46 (8), 1318–1320. https://doi.org/10.1002/anie.200603817. (135) Martinez, A. W.; Phillips, S. T.; Whitesides, G. M. Three-Dimensional Microfluidic Devices Fabricated in Layered Paper and Tape. Proc. Natl. Acad. Sci. 2008, 105 (50), 19606–19611. (136) Whitesides, G. M. The Origins and the Future of Microfluidics. Nature 2006, 442 (7101), 368–373. https://doi.org/10.1038/nature05058. (137) Sackmann, E. K.; Fulton, A. L.; Beebe, D. J. The Present and Future Role of Microfluidics in Biomedical Research. Nature 2014, 507 (7491), 181–189. (138) Badu-Tawiah, A. K.; Lathwal, S.; Kaastrup, K.; Al-Sayah, M.; Christodouleas, D. C.; Smith, B. S.; Whitesides, G. M.; Sikes, H. D. Polymerization-Based Signal Amplification for Paper- Based Immunoassays. Lab. Chip 2015, 15 (3), 655–659. https://doi.org/10.1039/C4LC01239A. (139) Klasner, S. A.; Price, A. K.; Hoeman, K. W.; Wilson, R. S.; Bell, K. J.; Culbertson, C. T. Paper-Based Microfluidic Devices for Analysis of Clinically Relevant Analytes Present in Urine and Saliva. Anal. Bioanal. Chem. 2010, 397 (5), 1821–1829. https://doi.org/10.1007/s00216-010-3718-4. (140) Abe, K.; Suzuki, K.; Citterio, D. Inkjet-Printed Microfluidic Multianalyte Chemical Sensing Paper. Anal. Chem. 2008, 80 (18), 6928–6934. https://doi.org/10.1021/ac800604v. (141) Nie, Z.; Deiss, F.; Liu, X.; Akbulut, O.; Whitesides, G. M. Integration of Paper-Based Microfluidic Devices with Commercial Electrochemical Readers. Lab. Chip 2010, 10 (22), 3163–3169. https://doi.org/10.1039/C0LC00237B. (142) Lu, Y.; Shi, W.; Jiang, L.; Qin, J.; Lin, B. Rapid Prototyping of Paper-Based Microfluidics with Wax for Low-Cost, Portable Bioassay. Electrophoresis 2009, 30 (9), 1497–1500. https://doi.org/10.1002/elps.200800563. (143) Rohrman, B. A.; Richards-Kortum, R. R. A Paper and Plastic Device for Performing Recombinase Polymerase Amplification of HIV DNA. Lab. Chip 2012, 12 (17), 3082–3088. https://doi.org/10.1039/c2lc40423k. (144) Li, C.; Vandenberg, K.; Prabhulkar, S.; Zhu, X.; Schneper, L.; Methee, K.; Rosser, C. J.; Almeide, E. Paper Based Point-of-Care Testing Disc for Multiplex Whole Cell Bacteria Analysis. Biosens. Bioelectron. 2011, 26 (11), 4342–4348. https://doi.org/10.1016/j.bios.2011.04.035. (145) Fu, E.; Liang, T.; Houghtaling, J.; Ramachandran, S.; Ramsey, S. A.; Lutz, B.; Yager, P. Enhanced Sensitivity of Lateral Flow Tests Using a Two-Dimensional Paper Network Format. Anal. Chem. 2011, 83 (20), 7941–7946. https://doi.org/10.1021/ac201950g. (146) Wang, L.; Chen, W.; Xu, D.; Shim, B. S.; Zhu, Y.; Sun, F.; Liu, L.; Peng, C.; Jin, Z.; Xu, C.; et al. Simple, Rapid, Sensitive, and Versatile SWNT-Paper Sensor for Environmental Toxin Detection Competitive with ELISA. Nano Lett. 2009, 9 (12), 4147–4152. https://doi.org/10.1021/nl902368r. 146

(147) Hossain, S. M. Z.; Luckham, R. E.; McFadden, M. J.; Brennan, J. D. Reagentless Bidirectional Lateral Flow Bioactive Paper Sensors for Detection of Pesticides in Beverage and Food Samples. Anal. Chem. 2009, 81 (21), 9055–9064. https://doi.org/10.1021/ac901714h. (148) Cate, D. M.; Nanthasurasak, P.; Riwkulkajorn, P.; L’Orange, C.; Henry, C. S.; Volckens, J. Rapid Detection of Transition Metals in Welding Fumes Using Paper-Based Analytical Devices. Ann. Occup. Hyg. 2014, 58 (4), 413–423. https://doi.org/10.1093/annhyg/met078. (149) Apilux, A.; Dungchai, W.; Siangproh, W.; Praphairaksit, N.; Henry, C. S.; Chailapakul, O. Lab-on-Paper with Dual Electrochemical/Colorimetric Detection for Simultaneous Determination of Gold and Iron. Anal. Chem. 2010, 82 (5), 1727–1732. https://doi.org/10.1021/ac9022555. (150) Lewis, G. G.; Robbins, J. S.; Phillips, S. T. A Prototype Point-of-Use Assay for Measuring Heavy Metal Contamination in Water Using Time as a Quantitative Readout. Chem Commun 2014, 50 (40), 5352–5354. https://doi.org/10.1039/C3CC47698G. (151) Peters, K. L.; Corbin, I.; Kaufman, L. M.; Zreibe, K.; Blanes, L.; McCord, B. R. Simultaneous Colorimetric Detection of Improvised Explosive Compounds Using Microfluidic Paper- Based Analytical Devices (UPADs). Anal. Methods 2015, 7 (1), 63–70. https://doi.org/10.1039/C4AY01677G. (152) Taudte, R. V.; Beavis, A.; Wilson-Wilde, L.; Roux, C.; Doble, P.; Blanes, L. A Portable Explosive Detector Based on Fluorescence Quenching of Pyrene Deposited on Coloured Wax-Printed UPADs. Lab. Chip 2013, 13 (21), 4164–4172. https://doi.org/10.1039/C3LC50609F. (153) Salles, M. O.; Meloni, G. N.; de Araujo, W. R.; Paixao, T. R. L. C. Explosive Colorimetric Discrimination Using a Smartphone, Paper Device and Chemometrical Approach. Analalytical Methods 2014, 6 (7), 2047–2052. https://doi.org/10.1039/C3AY41727A. (154) Pesenti, A.; Taudte, R. V.; McCord, B.; Doble, P.; Roux, C.; Blanes, L. Coupling Paper- Based Microfluidics and Lab on a Chip Technologies for Confirmatory Analysis of Trinitro Aromatic Explosives. Anal. Chem. 2014, 86 (10), 4707–4714. https://doi.org/10.1021/ac403062y. (155) Lankelma, J.; Nie, Z.; Carrilho, E.; Whitesides, G. M. Paper-Based Analytical Device for Electrochemical Flow-Injection Analysis of Glucose in Urine. Anal. Chem. 2012, 84 (9), 4147–4152. https://doi.org/10.1021/ac3003648. (156) Zhang, Y.; Zuo, P.; Ye, B.-C. A Low-Cost and Simple Paper-Based Microfluidic Device for Simultaneous Multiplex Determination of Different Types of Chemical Contaminants in Food. Biosens. Bioelectron. 2015, 68, 14–19. https://doi.org/10.1016/j.bios.2014.12.042. (157) Jin, S.-Q.; Guo, S.-M.; Zuo, P.; Ye, B.-C. A Cost-Effective Z-Folding Controlled Liquid Handling Microfluidic Paper Analysis Device for Pathogen Detection via ATP Quantification. Biosens. Bioelectron. 2015, 63, 379—383. https://doi.org/10.1016/j.bios.2014.07.070. (158) Ellerbee, A. K.; Phillips, S. T.; Siegel, A. C.; Mirica, K. A.; Martinez, A. W.; Striehl, P.; Jain, N.; Prentiss, M.; Whitesides, G. M. Quantifying Colorimetric Assays in Paper-Based Microfluidic Devices by Measuring the Transmission of Light through Paper. Anal. Chem. 2009, 81 (20), 8447–8452. https://doi.org/10.1021/ac901307q. (159) Lan, W.-J.; Maxwell, E. J.; Parolo, C.; Bwambok, D. K.; Subramaniam, A. B.; Whitesides, G. M. Paper-Based Electroanalytical Devices with an Integrated, Stable Reference Electrode. Lab. Chip 2013, 13 (20), 4103–4108. https://doi.org/10.1039/C3LC50771H.

147

(160) Arena, A.; Donato, N.; Saitta, G.; Bonavita, A.; Rizzo, G.; Neri, G. Flexible Ethanol Sensors on Glossy Paper Substrates Operating at Room Temperature. Sens. Actuators B Chem. 2010, 145 (1), 488–494. https://doi.org/10.1016/j.snb.2009.12.053. (161) Steffens, C.; Manzoli, A.; Francheschi, E.; Corazza, M. L.; Corazza, F. C.; Oliveira, J. V.; Herrmann, P. S. P. Low-Cost Sensors Developed on Paper by Line Patterning with Graphite and Polyaniline Coating with Supercritical CO2. Int. Conf. Sci. Technol. Synth. Met. Porto Galinhas Pernamb. Braz. July 6-11 2008ICSM 2008 2009, 159 (21–22), 2329–2332. https://doi.org/10.1016/j.synthmet.2009.08.045. (162) Delaney, J. L.; Hogan, C. F.; Tian, J.; Shen, W. Electrogenerated Chemiluminescence Detection in Paper-Based Microfluidic Sensors. Anal. Chem. 2011, 83 (4), 1300–1306. https://doi.org/10.1021/ac102392t. (163) Ge, L.; Wang, S.; Song, X.; Ge, S.; Yu, J. 3D Origami-Based Multifunction-Integrated Immunodevice: Low-Cost and Multiplexed Sandwich Chemiluminescence Immunoassay on Microfluidic Paper-Based Analytical Device. Lab. Chip 2012, 12 (17), 3150–3158. https://doi.org/10.1039/C2LC40325K. (164) Thom, N. K.; Lewis, G. G.; Yeung, K.; Phillips, S. T. Quantitative Fluorescence Assays Using a Self-Powered Paper-Based Microfluidic Device and a Camera-Equipped Cellular Phone. RSC Adv 2014, 4 (3), 1334–1340. https://doi.org/10.1039/C3RA44717K. (165) Maher, S.; Jjunju, F. P. M.; Taylor, S. Colloquium : 100 Years of Mass Spectrometry: Perspectives and Future Trends. Rev. Mod. Phys. 2015, 87 (1), 113–135. https://doi.org/10.1103/RevModPhys.87.113. (166) Carrilho, E.; Martinez, A. W.; Whitesides, G. M. Understanding Wax Printing: A Simple Micropatterning Process for Paper-Based Microfluidics. Anal. Chem. 2009, 81 (16), 7091– 7095. https://doi.org/10.1021/ac901071p. (167) Cooks, R. G.; Ouyang, Z.; Takats, Z.; Wiseman, J. M. Ambient Mass Spectrometry. Science 2006, 311 (5767), 1566–1570. (168) Ford, M. J.; Van Berkel, G. J. An Improved Thin-Layer Chromatography/Mass Spectrometry Coupling Using a Surface Sampling Probe Electrospray Ion Trap System. Rapid Commun. Mass Spectrom. 2004, 18 (12), 1303–1309. https://doi.org/10.1002/rcm.1486. (169) Shiea, J.; Huang, M.-Z.; HSu, H.-J.; Lee, C.-Y.; Yuan, C.-H.; Beech, I.; Sunner, J. Electrospray-Assisted Laser Desorption/Ionization Mass Spectrometry for Direct Ambient Analysis of Solids. Rapid Commun. Mass Spectrom. 2005, 19 (24), 3701–3704. https://doi.org/10.1002/rcm.2243. (170) Nemes, P.; Vertes, A. Laser Ablation Electrospray Ionization for Atmospheric Pressure, in Vivo, and Imaging Mass Spectrometry. Anal. Chem. 2007, 79 (21), 8098–8106. https://doi.org/10.1021/ac071181r. (171) Roach, P. J.; Laskin, J.; Laskin, A. Nanospray Desorption Electrospray Ionization: An Ambient Method for Liquid-Extraction Surface Sampling in Mass Spectrometry. Analyst 2010, 135 (9), 2233–2236. https://doi.org/10.1039/C0AN00312C. (172) Badu-Tawiah, A. K.; Eberlin, L. S.; Ouyang, Z.; Cooks, R. G. Chemical Aspects of the Extractive Methods of Ambient Ionization Mass Spectrometry. Annu. Rev. Phys. Chem. 2013, 64 (1), 481–505. https://doi.org/10.1146/annurev-physchem-040412-110026. (173) Weston, D. J. Ambient Ionization Mass Spectrometry: Current Understanding of Mechanistic Theory; Analytical Performance and Application Areas. Analyst 2010, 135 (4), 661–668. https://doi.org/10.1039/B925579F.

148

(174) Colletes, T. C.; Garcia, P. T.; Campanha, R. B.; Abdelnur, P. V.; Romão, W.; Coltro, W. K. T.; Vaz, B. G. A New Insert Sample Approach to Paper Spray Mass Spectrometry: A Paper Substrate with Paraffin Barriers. Analyst 2016, 141 (5), 1707–1713. https://doi.org/10.1039/C5AN01954K. (175) Wleklinski, M.; Li, Y.; Bag, S.; Sarkar, D.; Narayanan, R.; Pradeep, T.; Cooks, R. G. Zero Volt Paper Spray Ionization and Its Mechanism. Anal. Chem. 2015, 87 (13), 6786–6793. https://doi.org/10.1021/acs.analchem.5b01225. (176) Pagnotti, V. S.; Chubatyi, N. D.; McEwen, C. N. Solvent Assisted Inlet Ionization: An Ultrasensitive New Liquid Introduction Ionization Method for Mass Spectrometry. Anal. Chem. 2011, 83 (11), 3981–3985. https://doi.org/10.1021/ac200556z. (177) Manicke, N. E.; Yang, Q.; Wang, H.; Oradu, S.; Ouyang, Z.; Cooks, R. G. Assessment of Paper Spray Ionization for Quantitation of Pharmaceuticals in Blood Spots. Int. J. Mass Spectrom. 2011, 300 (2–3), 123–129. https://doi.org/10.1016/j.ijms.2010.06.037. (178) Salentijn, G. I.; Permentier, H. P.; Verpoorte, E. 3D-Printed Paper Spray Ionization Cartridge with Fast Wetting and Continuous Solvent Supply Features. Anal. Chem. 2014, 86 (23), 11657–11665. https://doi.org/10.1021/ac502785j. (179) Lee, H.; Jhang, C.-S.; Liu, J.-T.; Lin, C.-H. Rapid Screening and Determination of Designer Drugs in Saliva by a Nib-Assisted Paper Spray-Mass Spectrometry and Separation Technique. J. Sep. Sci. 2012, 35 (20), 2822–2825. https://doi.org/10.1002/jssc.201200480. (180) Behpour, M.; Ghoreishi, S. M.; Khayatkashani, M.; Soltani, N. Green Approach to Corrosion Inhibition of Mild Steel in Two Acidic Solutions by the Extract of Punica Granatum Peel and Main Constituents. Mater. Chem. Phys. 2012, 131 (3), 621–633. https://doi.org/10.1016/j.matchemphys.2011.10.027. (181) Jjunju, F. P. M.; Maher, S.; Damon, D. E.; Barrett, R. M.; Syed, S. U.; Heeren, R. M. A.; Taylor, S.; Badu-Tawiah, A. K. Screening and Quantification of Aliphatic Primary Alkyl Corrosion Inhibitor Amines in Water Samples by Paper Spray Mass Spectrometry. Anal. Chem. 2016, 88 (2), 1391–1400. https://doi.org/10.1021/acs.analchem.5b03992. (182) Hallett, K. C.; Atfield, A.; Comber, S.; Hutchinson, T. H. Developmental Toxicity of Metaldehyde in the Embryos of Lymnaea Stagnalis (Gastropoda: Pulmonata) Co-Exposed to the Synergist Piperonyl Butoxide. Sci. Total Environ. 2016, 543, Part A, 37–43. https://doi.org/10.1016/j.scitotenv.2015.11.040. (183) Ruiz-Suárez, N.; Boada, L. D.; Henríquez-Hernández, L. A.; González-Moreo, F.; Suárez- Pérez, A.; Camacho, M.; Zumbado, M.; Almeida-González, M.; del Mar Travieso-Aja, M.; Luzardo, O. P. Continued Implication of the Banned Pesticides Carbofuran and Aldicarb in the Poisoning of Domestic and Wild Animals of the Canary Islands (Spain). Sci. Total Environ. 2015, 505, 1093–1099. https://doi.org/10.1016/j.scitotenv.2014.10.093. (184) Dolan, T.; Howsam, P.; Parsons, D. J.; Whelan, M. J. Is the EU Drinking Water Directive Standard for Pesticides in Drinking Water Consistent with the Precautionary Principle? Environ. Sci. Technol. 2013, 47 (10), 4999–5006. https://doi.org/10.1021/es304955g. (185) United States Environmental Protection Agency. Reregistration Eligibility Decision for Metaldehyde; Prevention, Pesticides, and Toxic Substances; OPP-2005-0231; 2006; pp 1– 53. (186) Stuart, M.; Lapworth, D.; Crane, E.; Hart, A. Review of Risk from Potential Emerging Contaminants in UK Groundwater. Sci Total Env. 2012, 416, 1–21. (187) El-Meligi, A. A. Corrosion of Materials in Polluted Environment and Effect on World Economy. Recent Pat. Corros. Sci. 2011, 1 (1), 144–155. https://doi.org/10.2174/1877610811101020144. 149

(188) Eddy, N. O.; Ebenso, E. E. Adsorption and Inhibitive Properties of Ethanol Extracts of Musa Sapientum Peels as a Green Corrosion Inhibitor for Mild Steel in H2SO4. African J. Pure Appl. Chem. 2008, 2, 46–54. (189) Rodriguez-Mozaz, S.; López de Alda, M. J.; Barceló, D. Monitoring of Estrogens, Pesticides and Bisphenol A in Natural Waters and Drinking Water Treatment Plants by Solid-Phase Extraction–Liquid Chromatography–Mass Spectrometry. J. Chromatogr. A 2004, 1045 (1–2), 85–92. https://doi.org/10.1016/j.chroma.2004.06.040. (190) Mompelat, S.; Le Bot, B.; Thomas, O. Occurrence and Fate of Pharmaceutical Products and By-Products, from Resource to Drinking Water. Environ. Int. 2009, 35 (5), 803–814. https://doi.org/10.1016/j.envint.2008.10.008. (191) Benotti, M. J.; Trenholm, R. A.; Vanderford, B. J.; Holady, J. C.; Stanford, B. D.; Snyder, S. A. Pharmaceuticals and Endocrine Disrupting Compounds in U.S. Drinking Water. Environ. Sci. Technol. 2009, 43 (3), 597–603. (192) Paerl, H. W. Controlling Eutrophication along the Freshwater–Marine Continuum: Dual Nutrient (N and P) Reductions Are Essential. Estuaries Coasts 2009, 32 (4), 593–601. https://doi.org/10.1007/s12237-009-9158-8. (193) Conley, D. J.; Paerl, H. W.; Howarth, R. W.; Boesch, D. F.; Seitzinger, S. P.; Havens, K. E.; Lancelot, C.; Likens, G. E. ECOLOGY: Controlling Eutrophication: Nitrogen and Phosphorus. Science 2009, 323 (5917), 1014–1015. https://doi.org/10.1126/science.1167755. (194) Harmful Cyanobacteria; Huisman, J., Matthijs, H. C. P., Visser, P. M., Eds.; Aquatic ecology series; Springer: Dordrecht ; Norwell, MA, 2005. (195) Alsabagh, A. M.; Migahed, M. A.; Awad, H. S. Reactivity of Polyester Aliphatic Amine Surfactants as Corrosion Inhibitors for Carbon Steel in Formation Water (Deep Well Water). Corros. Sci. 2006, 48 (4), 813–828. https://doi.org/10.1016/j.corsci.2005.04.009. (196) Okafor, P. C.; Liu, C. B.; Liu, X.; Zheng, Y. G. Inhibition of CO2 Corrosion of N80 Carbon Steel by Carboxylic Quaternary Imidazoline and Halide Ions Additives. J. Appl. Electrochem. 2009, 39 (12), 2535–2543. https://doi.org/10.1007/s10800-009-9948-5. (197) Almeida, E. Surface Treatments and Coatings for Metals. A General Overview. 1. Surface Treatments, Surface Preparation, and the Nature of Coatings. Ind. Eng. Chem. Res. 2001, 40 (1), 3–14. https://doi.org/10.1021/ie000209l. (198) Zheludkevich, M. L.; Shchukin, D. G.; Yasakau, K. A.; Möhwald, H.; Ferreira, M. G. S. Anticorrosion Coatings with Self-Healing Effect Based on Nanocontainers Impregnated with Corrosion Inhibitor. Chem. Mater. 2007, 19 (3), 402–411. https://doi.org/10.1021/cm062066k. (199) Asipita, S. A.; Ismail, M.; Majid, M. Z. A.; Majid, Z. A.; Abdullah, C.; Mirza, J. Green Bambusa Arundinacea Leaves Extract as a Sustainable Corrosion Inhibitor in Steel Reinforced Concrete. J. Clean. Prod. 2014, 67, 139–146. https://doi.org/10.1016/j.jclepro.2013.12.033. (200) Finšgar, M.; Jackson, J. Application of Corrosion Inhibitors for Steels in Acidic Media for the Oil and Gas Industry: A Review. Corros. Sci. 2014, 86, 17–41. https://doi.org/10.1016/j.corsci.2014.04.044. (201) Podobaev, N. I.; Avdeev, Y. G. A Review of Acetylene Compounds as Inhibitors of Acid Corrosion of Iron. Prot. Met. 2004, 40 (1), 7–13. https://doi.org/10.1023/B:PROM.0000013105.48781.86. (202) S. M. A. Shibli; V. S. Saji. Corrosion Inhibitors in Cooling Towers; Chemical Industry Digest, 2002; pp 74–80.

150

(203) Cai, J.; Chen, C.; Liu, J.; Liu, J. Corrosion Resistance of Carbon Steel in Simulated Concrete Pore Solution in Presence of 1-Dihydroxyethylamino-3-Dipropylamino-2-Propanol as Corrosion Inhibitor. Corros. Eng. Sci. Technol. 2014, 49 (1), 66–72. https://doi.org/10.1179/1743278213Y.0000000109. (204) El-Shamy, A. M.; Zakaria, K.; Abbas, M. A.; Zein El Abedin, S. Anti-Bacterial and Anti- Corrosion Effects of the Ionic Liquid 1-Butyl-1-Methylpyrrolidinium Trifluoromethylsulfonate. J. Mol. Liq. 2015, 211, 363–369. https://doi.org/10.1016/j.molliq.2015.07.028. (205) Quraishi, M. A.; Singh, A.; Singh, V. K.; Yadav, D. K.; Singh, A. K. Green Approach to Corrosion Inhibition of Mild Steel in Hydrochloric Acid and Sulphuric Acid Solutions by the Extract of Murraya Koenigii Leaves. Mater. Chem. Phys. 2010, 122 (1), 114–122. https://doi.org/10.1016/j.matchemphys.2010.02.066. (206) Rani, B. E. A.; Basu, B. B. J. Green Inhibitors for Corrosion Protection of Metals and Alloys: An Overview. Int. J. Corros. 2012, 2012, 1–15. https://doi.org/10.1155/2012/380217. (207) Videla, H. A.; Herrera, L. K. Microbiologically Influenced Corrosion: Looking to the Future. Int. Microbiol. 2005, 6 (3), 169–180. (208) Li, S.; Kim, Y.-G.; Jung, S.; Song, H.-S.; Lee, S.-M. Application of Steel Thin Film Electrical Resistance Sensor for in Situ Corrosion Monitoring. Sens. Actuators B Chem. 2007, 120 (2), 368–377. https://doi.org/10.1016/j.snb.2006.02.029. (209) Chen, C.-H.; Chen, T.-C.; Zhou, X.; Kline-Schoder, R.; Sorensen, P.; Cooks, R. G.; Ouyang, Z. Design of Portable Mass Spectrometers with Handheld Probes: Aspects of the Sampling and Miniature Pumping Systems. J. Am. Soc. Mass Spectrom. 2015, 26 (2), 240–247. https://doi.org/10.1007/s13361-014-1026-5. (210) Boonham, N. On-Site Testing: Moving Decision Making from the Lab to the Field. In Detection and Diagnostics of Plant Pathogens; Gullino, M. L., Bonants, P. J. M., Eds.; Springer Netherlands: Dordrecht, 2014; pp 135–146. https://doi.org/10.1007/978-94-017- 9020-8_9. (211) Sullivan, J.; Cooze, N.; Gallagher, C.; Lewis, T.; Prosek, T.; Thierry, D. In Situ Monitoring of Corrosion Mechanisms and Phosphate Inhibitor Surface Deposition during Corrosion of Zinc–Magnesium–Aluminium (ZMA) Alloys Using Novel Time-Lapse Microscopy. Faraday Discuss. 2015, 180, 361–379. https://doi.org/10.1039/C4FD00251B. (212) Noor Amizan Noor; Ruzairi Abdul Rahim; Leow Pei Ling; Jaysuman Pusppanathan; Mohd Hafiz Fazalul Rahiman; Nor Muzakkir Nor Ayob; Fazlul Rahman Mohd Yunusa. A Review of Ultrasonic Tomography for Monitoring the Corrosion of Steel Pipes. J. Teknol. 2015, 73 (6), 151–158. https://doi.org/10.11113/jt.v73.4468. (213) Khakharia, P.; Mertens, J.; Huizinga, A.; De Vroey, S.; Sanchez Fernandez, E.; Srinivasan, S.; Vlugt, T. J. H.; Goetheer, E. Online Corrosion Monitoring in a Postcombustion CO 2 Capture Pilot Plant and Its Relation to Solvent Degradation and Ammonia Emissions. Ind. Eng. Chem. Res. 2015, 54 (19), 5336–5344. https://doi.org/10.1021/acs.iecr.5b00729. (214) Mach, P. M.; McBride, E. M.; Sasiene, Z. J.; Brigance, K. R.; Kennard, S. K.; Wright, K. C.; Verbeck, G. F. Vehicle-Mounted Portable Mass Spectrometry System for the Covert Detection via Spatial Analysis of Clandestine Methamphetamine Laboratories. Anal. Chem. 2015, 87 (22), 11501–11508. https://doi.org/10.1021/acs.analchem.5b03269. (215) Verbeck, G. F.; Bierbaum, V. M. Focus on Harsh Environment and Field-Portable Mass Spectrometry: Editorial. J. Am. Soc. Mass Spectrom. 2015, 26 (2), 199–200. https://doi.org/10.1007/s13361-014-1057-y. 151

(216) Huynh, V.; Joshi, U.; Leveille, J. M.; Golden, T. D.; Verbeck, G. F. Nanomanipulation- Coupled to Nanospray Mass Spectrometry Applied to Document and Ink Analysis. Forensic Sci. Int. 2014, 242, 150–156. https://doi.org/10.1016/j.forsciint.2014.06.037. (217) Kolliopoulos, A. V.; Metters, J. P.; Banks, C. E. Quantification of Corrosion Inhibitors Used in the Water Industry for Steam Condensate Treatment: The Indirect Electroanalytical Sensing of Morpholine and Cyclohexylamine. Environ. Sci. Water Res. Technol. 2015, 1 (1), 40–46. https://doi.org/10.1039/C4EW00033A. (218) Kusch, P.; Knupp, G.; Kozupa, M.; Iowska, J.; Majchrzak, M. Identification and Application of Corrosion Inhibiting Long- Chain Primary Alkyl Amines in Water Treatment in the Power Industry. In Developments in Corrosion Protection; Aliofkhazraei, M., Ed.; InTech, 2014. https://doi.org/10.5772/57355. (219) Weiss, S.; Reemtsma, T. Determination of Benzotriazole Corrosion Inhibitors from Aqueous Environmental Samples by Liquid Chromatography-Electrospray Ionization- Tandem Mass Spectrometry. Anal. Chem. 2005, 77 (22), 7415–7420. https://doi.org/10.1021/ac051203e. (220) Oliveira, S. M. de; Siguemura, A.; Lima, H. O.; Souza, F. C. de; Magalhães, A. A. O.; Toledo, R. M.; D’Elia, E. Flow Injection Analysis with Amperometric Detection for Morpholine Determination in Corrosion Inhibitors. J. Braz. Chem. Soc. 2014. https://doi.org/10.5935/0103-5053.20140122. (221) Llop, A.; Pocurull, E.; Borrull, F. Automated Determination of Aliphatic Primary Amines in Wastewater by Simultaneous Derivatization and Headspace Solid-Phase Microextraction Followed by Gas Chromatography–Tandem Mass Spectrometry. J. Chromatogr. A 2010, 1217 (4), 575–581. https://doi.org/10.1016/j.chroma.2009.11.087. (222) Magnes, C.; Fauland, A.; Gander, E.; Narath, S.; Ratzer, M.; Eisenberg, T.; Madeo, F.; Pieber, T.; Sinner, F. Polyamines in Biological Samples: Rapid and Robust Quantification by Solid-Phase Extraction Online-Coupled to Liquid Chromatography–Tandem Mass Spectrometry. J. Chromatogr. A 2014, 1331, 44–51. https://doi.org/10.1016/j.chroma.2013.12.061. (223) Farajzadeh, M. A.; Nouri, N.; Khorram, P. Derivatization and Microextraction Methods for Determination of Organic Compounds by Gas Chromatography. TrAC Trends Anal. Chem. 2014, 55, 14–23. https://doi.org/10.1016/j.trac.2013.11.006. (224) Kusch, P.; Knupp, G.; Hergarten, M.; Kozupa, M.; Majchrzak, M. Solid-Phase Extraction- Gas Chromatography and Solid-Phase Extraction-Gas Chromatography–Mass Spectrometry Determination of Corrosion Inhibiting Long-Chain Primary Alkyl Amines in Chemical Treatment of Boiler Water in Water-Steam Systems of Power Plants. J. Chromatogr. A 2006, 1113 (1–2), 198–205. https://doi.org/10.1016/j.chroma.2006.01.118. (225) Dai, Z.; Wu, Z.; Wang, J.; Wang, X.; Jia, S.; Bazer, F. W.; Wu, G. Analysis of Polyamines in Biological Samples by HPLC Involving Pre-Column Derivatization with o-Phthalaldehyde and N-Acetyl-l-Cysteine. Amino Acids 2014, 46 (6), 1557–1564. https://doi.org/10.1007/s00726-014-1717-z. (226) Kusch, P.; Knupp, G.; Hergarten, M.; Kozupa, M.; Majchrzak, M. Identification of Corrosion Inhibiting Long-Chain Primary Alkyl Amines by Gas Chromatography and Gas Chromatography–Mass Spectrometry. Int. J. Mass Spectrom. 2007, 263 (1), 45–53. https://doi.org/10.1016/j.ijms.2006.12.006. (227) Jurado-Sánchez, B.; Ballesteros, E.; Gallego, M. Comparison of Several Solid-Phase Extraction Sorbents for Continuous Determination of Amines in Water by Gas

152

Chromatography–Mass Spectrometry. Talanta 2009, 79 (3), 613–620. https://doi.org/10.1016/j.talanta.2009.04.035. (228) Płotka-Wasylka, J. M.; Morrison, C.; Biziuk, M.; Namieśnik, J. Chemical Derivatization Processes Applied to Amine Determination in Samples of Different Matrix Composition. Chem. Rev. 2015, 115 (11), 4693–4718. https://doi.org/10.1021/cr4006999. (229) Smith, J. R. L.; Smart, A. U.; Twigg, M. V. The Reactions of Amine, Polyamine and Amino Alcohol Corrosion Inhibitors in Water at High Temperature. J. Chem. Soc. Perkin Trans. 2 1992, No. 6, 939. https://doi.org/10.1039/p29920000939. (230) O’Leary, A. E.; Hall, S. E.; Vircks, K. E.; Mulligan, C. C. Monitoring the Clandestine Synthesis of Methamphetamine in Real-Time with Ambient Sampling, Portable Mass Spectrometry. Anal. Methods 2015, 7 (17), 7156–7163. https://doi.org/10.1039/C5AY00511F. (231) Alberici, R. M.; Simas, R. C.; Sanvido, G. B.; Romão, W.; Lalli, P. M.; Benassi, M.; Cunha, I. B. S.; Eberlin, M. N. Ambient Mass Spectrometry: Bringing MS into the “Real World.” Anal. Bioanal. Chem. 2010, 398 (1), 265–294. https://doi.org/10.1007/s00216-010-3808-3. (232) Maher, S.; Jjunju, F. P. M.; Taylor, S. Colloquium : 100 Years of Mass Spectrometry: Perspectives and Future Trends. Rev. Mod. Phys. 2015, 87 (1), 113–135. https://doi.org/10.1103/RevModPhys.87.113. (233) Snyder, D. T.; Pulliam, C. J.; Ouyang, Z.; Cooks, R. G. Miniature and Fieldable Mass Spectrometers: Recent Advances. Anal Chem 2015, 88, 2–29. (234) Culzoni, M. J.; Dwivedi, P.; Green, M. D.; Newton, P. N.; Fernández, F. M. Ambient Mass Spectrometry Technologies for the Detection of Falsified Drugs. Med Chem Commun 2014, 5 (1), 9–19. https://doi.org/10.1039/C3MD00235G. (235) Ifa, D. R.; Wu, C.; Ouyang, Z.; Cooks, R. G. Desorption Electrospray Ionization and Other Ambient Ionization Methods: Current Progress and Preview. The Analyst 2010, 135 (4), 669–681. https://doi.org/10.1039/b925257f. (236) Takats, Z. Mass Spectrometry Sampling Under Ambient Conditions with Desorption Electrospray Ionization. Science 2004, 306 (5695), 471–473. https://doi.org/10.1126/science.1104404. (237) Cooks, R. G.; Ouyang, Z.; Takats, Z.; Wiseman, J. M. Ambient Mass Spectrometry. Science 2006, 311, 1566–1570. (238) Nemes, P.; Vertes, A. Ambient Mass Spectrometry for in Vivo Local Analysis and in Situ Molecular Tissue Imaging. TrAC Trends Anal. Chem. 2012, 34, 22–34. https://doi.org/10.1016/j.trac.2011.11.006. (239) Harris, G. A.; Galhena, A. S.; Fernandez, F. M. Ambient Sampling/Ionization Mass Spectrometry: Applications and Current Trends. Anal Chem 2011, 83, 4508–4538. (240) Huang, M.-Z.; Yuan, C.-H.; Cheng, S.-C.; Cho, Y.-T.; Shiea, J. Ambient Ionization Mass Spectrometry. Annu. Rev. Anal. Chem. 2010, 3 (1), 43–65. https://doi.org/10.1146/annurev.anchem.111808.073702. (241) Bennett, R. V.; Gamage, C. M.; Galhena, A. S.; Fernández, F. M. Contrast-Enhanced Differential Mobility-Desorption Electrospray Ionization-Mass Spectrometry Imaging of Biological Tissues. Anal. Chem. 2014, 86 (8), 3756–3763. https://doi.org/10.1021/ac5007816. (242) Robert B. Cody; James A. Laramée; J. Michael Nilles; H. Dupont Durst. Cody, R. B.; Laramée, J. A.; Nilles, J. M.; Durst, H. D. JEOL News 2005, 40, 8– 12. JEOL News 2005, 40, 8– 12.

153

(243) Cody, R. B.; Laramée, J. A.; Durst, H. D. Versatile New Ion Source for the Analysis of Materials in Open Air under Ambient Conditions. Anal Chem 2005, 77, 2297–2302. (244) Jjunju, F. P. M.; Badu-Tawiah, A. K.; Li, A.; Soparawalla, S.; Roqan, I. S.; Cooks, R. G. Hydrocarbon Analysis Using Desorption Atmospheric Pressure Chemical Ionization. Int. J. Mass Spectrom. 2013, 345–347, 80–88. https://doi.org/10.1016/j.ijms.2012.08.030. (245) Laskin, J.; Heath, B. S.; Roach, P. J.; Cazares, L.; Semmes, O. J. Tissue Imaging Using Nanospray Desorption Electrospray Ionization Mass Spectrometry. Anal. Chem. 2012, 84 (1), 141–148. https://doi.org/10.1021/ac2021322. (246) Harper, J. D.; Charipar, N. A.; Mulligan, C. C.; Zhang, X.; Cooks, R. G.; Ouyang, Z. Low- Temperature Plasma Probe for Ambient Desorption Ionization. Anal. Chem. 2008, 80 (23), 9097–9104. https://doi.org/10.1021/ac801641a. (247) Mulligan, C. C.; Talaty, N.; Cooks, R. G. Desorption Electrospray Ionization with a Portable Mass Spectrometer: In Situ Analysis of Ambient Surfaces. Chem. Commun. 2006, No. 16, 1709–1711. https://doi.org/10.1039/b517357d. (248) McLafferty, F. Tandem Mass Spectrometry. Science 1981, 214 (4518), 280–287. https://doi.org/10.1126/science.7280693. (249) Sleno, L.; Volmer, D. A. Ion Activation Methods for Tandem Mass Spectrometry. J. Mass Spectrom. 2004, 39 (10), 1091–1112. https://doi.org/10.1002/jms.703. (250) Syed, S. U. A. H.; Maher, S.; Eijkel, G. B.; Ellis, S. R.; Jjunju, F.; Taylor, S.; Heeren, R. M. A. Direct Ion Imaging Approach for Investigation of Ion Dynamics in Multipole Ion Guides. Anal. Chem. 2015, 87 (7), 3714–3720. https://doi.org/10.1021/ac5041764. (251) Smith, R. T.; Jjunju, F. P. M.; Maher, S. EVALUATION OF ELECTRON BEAM DEFLECTIONS ACROSS A SOLENOID USING WEBER-RITZ AND MAXWELL-LORENTZ ELECTRODYNAMICS. Prog. Electromagn. Res. 2015, 151, 83–93. https://doi.org/10.2528/PIER15021106. (252) Jjunju, F. P. M.; Li, A.; Badu-Tawiah, A.; Wei, P.; Li, L.; Ouyang, Z.; Roqan, I. S.; Cooks, R. G. In Situ Analysis of Corrosion Inhibitors Using a Portable Mass Spectrometer with Paper Spray Ionization. The Analyst 2013, 138 (13), 3740–3748. https://doi.org/10.1039/c3an00249g. (253) Perry, R. H.; Splendore, M.; Chien, A.; Davis, N. K.; Zare, R. N. Detecting Reaction Intermediates in Liquids on the Millisecond Time Scale Using Desorption Electrospray Ionization. Angew. Chem. 2011, 123 (1), 264–268. https://doi.org/10.1002/ange.201004861. (254) Espy, R. D.; Wleklinski, M.; Yan, X.; Cooks, R. G. Beyond the Flask: Reactions on the Fly in Ambient Mass Spectrometry. TrAC Trends Anal. Chem. 2014, 57, 135–146. https://doi.org/10.1016/j.trac.2014.02.008. (255) Badu-Tawiah, A. K.; Li, A.; Jjunju, F. P. M.; Cooks, R. G. Peptide Cross-Linking at Ambient Surfaces by Reactions of Nanosprayed Molecular Cations. Angew. Chem. 2012, 124 (37), 9551–9555. https://doi.org/10.1002/ange.201205044. (256) Li, A.; Jjunju, F. P. M.; Cooks, R. G. Nucleophilic Addition of Nitrogen to Aryl Cations: Mimicking Titan Chemistry. J. Am. Soc. Mass Spectrom. 2013, 24 (11), 1745–1754. https://doi.org/10.1007/s13361-013-0710-1. (257) Hasson, D.; Shemer, H.; Sher, A. State of the Art of Friendly “Green” Scale Control Inhibitors: A Review Article. Ind. Eng. Chem. Res. 2011, 50 (12), 7601–7607. https://doi.org/10.1021/ie200370v. (258) Dargahi, M.; Olsson, A. L. J.; Tufenkji, N.; Gaudreault, R. Green Technology: Tannin- Based Corrosion Inhibitor for Protection of Mild Steel. Corrosion 2015, 71 (11), 1321–1329. https://doi.org/10.5006/1777. 154

(259) Kamal, C.; Sethuraman, M. G. Spirulina Platensis – A Novel Green Inhibitor for Acid Corrosion of Mild Steel. Arab. J. Chem. 2012, 5 (2), 155–161. https://doi.org/10.1016/j.arabjc.2010.08.006. (260) Caloni, F.; Cortinovis, C.; Rivolta, M.; Davanzo, F. Suspected Poisoning of Domestic Animals by Pesticides. Sci Total Env. 2016, 539, 331–336. (261) Bleakley, C.; Ferrie, E.; Collum, N.; Burke, L. Self-Poisoning with Metaldehyde. Emerg Med J 2008, 25, 381–382. (262) Richardson, S. D.; Ternes, T. A. Water Analysis: Emerging Contaminants and Current Issues. Anal Chem 2014, 86, 2813–2848. (263) Alastuey, A. Emerging Organic Contaminants and Human Health; Barceló, D., Ed.; The handbook of environmental chemistry; Springer: Berlin, 2012; Vol. 20. (264) Luzardo, O. P.; Ruiz-Suarez, N.; Valeron, P. F.; Camacho, M.; Zumbado, M.; Henriquez- Hernandez, L. A.; Boada, L. D. Methodology for the Identification of 117 Pesticides Commonly Involved in the Poisoning of Wildlife Using GC-MS-MS and LC-MS-MS. J. Anal. Toxicol. 2014, 38 (3), 155–163. https://doi.org/10.1093/jat/bku009. (265) Jones, A.; Charlton, A. Determination of Metaldehyde in Suspected Cases of Animal Poisoning Using Gas Chromatography-Ion Trap Mass Spectrometry. J Agric Food Chem 1999, 47, 4675–4677. (266) WHO Recommended Classification of Pesticides by Hazard and Guidelines to Classification 2009; World Health Organization: Stuttgart, Germany, 2010. (267) Kallis, G.; Butler, D. The EU Water Framework Directive: Measures and Implications. Water Policy 2001, 3, 125–142. (268) Kay, P.; Grayson, R. Using Water Industry Data to Assess the Metaldehyde Pollution Problem. Water Env. J 2014, 28, 410–417. (269) Zhang, H.; Wang, C.; Xu, P.; Ma, Y. Analysis of Molluscicide Metaldehyde in Vegetables by Dispersive Solid-Phase Extraction and Liquid Chromatography-Tandem Mass Spectrometry. Food Addit Contam A 2011, 28, 1034–1040. (270) Li, C.; Wu, Y.-L.; Yang, T.; Zhang, Y. Determination of Metaldehyde in Water by SPE and UPLC–MS–MS. Chromatographia 2010, 72, 987–991. (271) The Determination of Metaldehyde in Waters Using Chromatography with Mass Spectrometric Detection (2009); Environment Agency, 2009. (272) Cheng, S.-C.; Jhang, S.-S.; Huang, M.-Z.; Shiea, J. Simultaneous Detection of Polar and Nonpolar Compounds by Ambient Mass Spectrometry with a Dual Electrospray and Atmospheric Pressure Chemical Ionization Source. Anal Chem 2015, 87, 1743–1748. (273) Badu-Tawiah, A. K.; Campbell, D. I.; Cooks, R. G. Reactions of Microsolvated Organic Compounds at Ambient Surfaces: Droplet Velocity, Charge State, and Solvent Effects. J. Am. Soc. Mass Spectrom. 2012, 23 (6), 1077–1084. https://doi.org/10.1007/s13361-012- 0365-3. (274) Yan, X.; Augusti, R.; Li, X.; Cooks, R. G. Chemical Reactivity Assessment Using Reactive Paper Spray Ionization Mass Spectrometry: The Katritzky Reaction. ChemPlusChem 2013, 78 (9), 1142–1148. https://doi.org/10.1002/cplu.201300172. (275) Busquets, R.; Kozynchenko, O. P.; Whitby, R. L.; Tennison, S. R.; Cundy, A. B. Phenolic Carbon Tailored for the Removal of Polar Organic Contaminants from Water: A Solution to the Metaldehyde Problem? Water Res 2014, 61, 46–56. (276) Bag, S.; Hendricks, P.; Reynolds, J. C.; Cooks, R. Biogenic Aldehyde Determination by Reactive Paper Spray Ionization Mass Spectrometry. Anal Chim Acta 2015, 860, 37–42.

155

(277) Ma, Q.; Bai, H.; Li, W.; Wang, C.; Cooks, R. G.; Ouyang, Z. Rapid Analysis of Synthetic Cannabinoids Using a Miniature Mass Spectrometer with Ambient Ionization Capability. Talanta 2015, 142, 190–196. https://doi.org/10.1016/j.talanta.2015.04.044. (278) Ma, X.; Ouyang, Z. Ambient Ionization and Miniature Mass Spectrometry System for Chemical and Biological Analysis. TrAC Trends Anal. Chem. 2016, 85, 10–19. https://doi.org/10.1016/j.trac.2016.04.009.

156

Appendix A. List of Videos

Blank solution of methanol/water (1:1) were sprayed across a distance of 5 mm with a DC voltage of 4 kV applied directly to the dry hydrophobic paper triangle. Two spray geometries were tested: (1) Grounded plate was placed in front of the hydrophobic triangle (Video 1), and (2) Ground plate was placed above the sample (Video 2). In the latter case, upwards spray was observed towards the grounded plate. Microvibrations were observed within the methanol/water (1:1) droplet when the same experiment was performed under a microscope (Video 3). These vibrations became stronger at higher electric filed so much that the entire droplet is physically observed to jump from the hydrophobic paper surface (Video 4). See Section 2.12 Ionization Mechanism for a brief discussion on droplet vibrations under the influence of an electric field.

Video 1. Ground was placed in front of the hydrophobic triangle: https://research.cbc.osu.edu/badu-tawiah.1/ms-techniques/ambient-ionization- techniques/paper-spray-ionization/25-speed_front/

Video 2. Ground was placed above the sample: https://research.cbc.osu.edu/badu-tawiah.1/ms-techniques/ambient-ionization- techniques/paper-spray-ionization/25-speed_above/

Video 3. Droplet vibrations were recorded under high magnifying microscope (Leitz Ergolux AMC Optical Microscope): https://research.cbc.osu.edu/badu-tawiah.1/ms-techniques/ambient-ionization- techniques/paper-spray-ionization/vibrating-droplet-1/

157

Video 4. Droplet vibrates, then jumps from paper surface captured under microscope: https://research.cbc.osu.edu/badu-tawiah.1/ms-techniques/ambient-ionization- techniques/paper-spray-ionization/droplet/

The following videos show that as surface tension increases (with water being a larger constituent), the solvent beads onto the surface of the paper and vibrates when voltage is applied (Video 5). As surface tension of the solvent decreases (when the acetonitrile proportion grows), the solvent bead vibrates more violently (Video 6). An even greater amount of acetonitrile in the solvent will cause the solvent drop to both vibrate and form momentary Taylor cones (Video 7). Finally, when acetonitrile is a large enough constituent of the solvent, the solvent is able to completely wet the paper, and droplet vibrating ceases. Instead, a stable Taylor cone is observed (Video 8). See section 2.11 Surface Energy Analysis for a discussion of wetting influencing ion intensity.

Video 5. Solvent 2 (Surface tension 62 mN/m). When voltage is applied, the solvent vibrates. https://research.cbc.osu.edu/badu-tawiah.1/wp-content/uploads/2019/02/Video-S5.mp4

Video 6. Solvent 4 (Surface tension 41 mN/m). When voltage is applied, the solvent vibrates more vigorously. https://research.cbc.osu.edu/badu-tawiah.1/wp-content/uploads/2019/02/Video-S6.mp4

Video 7. Solvent 5 (Surface tension 38 mN/m). When voltage is applied, the solvent vibrates, momentary Taylor cones are observed. https://research.cbc.osu.edu/badu-tawiah.1/wp-content/uploads/2019/02/Video-S7.mp4

Video 8. Solvent 7 (Surface tension 29 mN/m). When voltage is applied, the solvent forms multiple Taylor cones from the paper fibers protruding from the flat edge of the paper strip. https://research.cbc.osu.edu/badu-tawiah.1/wp-content/uploads/2019/02/Video-S8.mp4

158