Development of rapid and inexpensive derivatisation methods for methylphosphonic acid

By Jolene Anthony

This thesis is presented for the degree of Bachelor of Science Honours at Murdoch

University. College of Science, Health, Engineering and Education. Academic

Operations, Medical Molecular and Forensic Science, Perth, Western Australia.

Nov, 2019.

Supervisor:

Dr Kate Rowen

Co-supervisors:

Dr John Coumbaros & Associate Professor James Speers

Declaration

I declare that this thesis is my own account of my research and contains as its main content work which has not previously been submitted for a degree at any tertiary education institution.

Jolene Anthony

(Jolene Anthony)

Word Count: 17, 657.

1

Acknowledgements

First and foremost, I would like to thank Dr Kate Rowen for her continuous support throughout the year. Kate, your guidance and positivity have been invaluable in conducting this research.

You always managed to find the time for my questions and to teach me new techniques. Even in the face of several setbacks, when I found myself preparing to google ‘abyss near me’ to scream into, you helped me find a way to persevere. It has been such a rewarding experience and I have genuinely enjoyed almost every minute of it. You have far surpassed any expectation

I had for a supervisor. I feel truly honoured to have worked with and learnt from someone I consider to be “total science idol material”!

I would also like to express my gratitude to the SHEE technical staff, in particular Andrew

Foreman, Claire Tull, Goshia Kowalczyk, Sajiel Jani, and Juita Juita for having the patience to train me on a range of equipment. You are all incredibly hard workers! No matter how busy you were, you always managed to squeeze in some time to assist me. Without the help and knowledge provided by the technical team, this research would not have been possible.

There are several others I owe thanks to: Co-supervisors Dr John Coumbaros and Associate

Professor James Speers for their assistance. My friends/university family, for listening to me blabber on about my research for 8 months straight. Lastly, thank you to Murdoch University for funding this research.

2

Abstract

This study probed alternative methods towards rapid and inexpensive in-field derivatisation of methylphosphonic acid (MPA), the final hydrolysis product of several chemical warfare nerve agents, for detection by gas chromatography-mass spectrometry (GC-MS).

Initially, research focused on acetylation as a derivatisation method, with attempted preparation of a reference sample of diacetyl methylphosphonate. This was unsuccessful with reduction of the acetyl carbonyl group during synthesis and detection of diethyl methylphosphonate instead.

The research subsequently focused on esterification of methylphosphonic acid with methanol or ethanol. Preparations employed the use of dicyclohexyl carbodiimide as a coupling reagent, and silica chloride or sulfuric acid as catalysts. The coupling reagents and catalysts utilised had limited success, with no dialkyl methylphosphonate ester derivates detected via GC-MS.

Dialkyl derivatives were successfully prepared by conversion of MPA to methylphosphonyl dichloride using thionyl chloride, and subsequent reaction with alcohols. Conversion of methylphosphonyl dichloride to corresponding dialkyl methylphosphonates occurs rapidly with derivatives being detected within 10 minutes for primary alcohols (methanol, ethanol, and butan-1-ol) and within 20 minutes for secondary alcohols (propan-2-ol), while benzyl alcohol did not react. Dialkyl methylphosphonates were identified by analysis of reaction products using

GC-MS and nuclear magnetic resonance (NMR) spectroscopy. This method applied here acts as a proof of concept, providing a promising platform for future exploration. Alkylation of methylphosphonic acid with alcohols provides an inexpensive and relatively non-hazardous alternative to current widely accepted derivatisation methods in supporting the rapid in-field detection of chemical warfare nerve agents.

3

Table of contents

Development of rapid and inexpensive derivatisation methods for methylphosphonic acid ... 0

Declaration ...... 1

Acknowledgements ...... 2

Abstract ...... 3

Abbreviations ...... 6

Inventory of tables ...... 8

Inventory of figures ...... 9

Chapter 1. Introduction ...... 10

1.0 Introduction ...... 10

1.1 Chemical warfare agents ...... 10

1.2 Nerve agents ...... 11

1.2.1 Toxicity of nerve agents ...... 13

1.2.2 Organophosphate nerve agent degradation ...... 14

1.3 OPCW chemical analysis ...... 17

1.4. Derivatisation ...... 19

1.4.1 Types of derivatisation ...... 20

1.4.2 Derivatisation of phosphonic acids ...... 21

1.4.3 Derivatisation without the removal of the water ...... 25

1.5 Acetylation...... 27

1.5.1 Acetylation reagents ...... 28

1.5.2 Acetylation of methylphosphonic acid ...... 30

1.6 Concluding statements and research aims ...... 31

Chapter 2 Experimental ...... 33

2.1 General methods ...... 33

2.1.1 Instrumentation and instrumental parameters ...... 33

2.1.2 Reagents and solvents ...... 34

2.1.3 Solvent drying procedures...... 34

2.1.4 Preparation of methylphosphonyl dichloride ...... 35 4

2.1.5 Preparation of silver acetate ...... 35

2.1.6 Preparation of silica chloride ...... 35

2.2 Trialled preparation of diacetyl methylphosphonate ...... 36

2.3 Preparation of dialkyl methylphosphonates ...... 36

2.3.1 Conversion of methylphosphonyl dichloride to alkyl methylphosphonates...... 36

2.3.2 Trialed esterification of methylphosphonic acid with coupling reagents/ catalysts ...... 37

Chapter 3. Results and discussion ...... 38

3.1 Trialled preparation of diacetyl methylphosphonate ...... 38

3.2 Introduction to dialkyl methylphosphonates ...... 45

3.3 Revised research aims ...... 47

3.4 Preparation of dialkyl methylphosphonates ...... 49

3.4.1 Diethyl methylphosphonate...... 49

3.4.2 Dimethyl methylphosphonate ...... 53

3.4.3 Other dialkyl methylphosphonates alcohols ...... 55

3.4.6 Summary/ rate of reaction ...... 55

3.4.7 Trialled esterification using catalyst and coupling reagent...... 56

Chapter 4. Future direction and concluding statements ...... 58

References ...... 61

5

Abbreviations

(d, 2) doublet

(q, 4) quartet

(s, 1) singlet

(s, 5) quintet

(s, 7) septet

(t, 3) triplet

(u) unclear, could not be distinguished

AA acetic anhydride

AC acetyl chloride

Acetyl CoA acetyl coenzyme A

ACh acetylcholine

AChE acetylcholinesterase enzyme

AChR acetylcholine receptor

AED atomic emission detection

BSTFA N,O-bis(trimethylsilyl)trifluoroacetamide

CWA chemical warfare agent

CWC Chemical Weapons Convention

CWNA chemical warfare nerve agent

FPD flame photometric detection

GC-MS gas chromatography mass spectrometry

HF-LPME hollow fibre-based liquid-phase microextraction

LC-MS liquid chromatography mass spectrometry

MPA methylphosphonic acid

NMR nuclear magnetic resonance

NPD nitrogen-phosphorus detection

6

OP organophosphate

OPCW The Organisation for the Prohibition of Chemical Weapons

RI relative intensity

ROP recommended operating procedures

SPME solid-phase microextraction

TBDMS tertbutyldimethylsilyl

TMS trimethylsilyl

TMSCI trimethylsilyl chloride

7

Inventory of tables

Table 1. Names and structures of key organophosphate chemical warfare nerve agents...... 12 Table 2. Derivatives of methylphosphonic acid...... 24 Table 3. Acetyl donors and the compounds they acetylate...... 29 Table 4. Procedures for drying solvents...... 34 Table 5. Variations of acetate, solvent and purification method in trials of diacetyl methylphosphonate synthesis...... 36 Table 6. First sign of product formation as determined via GC-MS, names and structures of MPA derivatives detected...... 56

8

Inventory of figures

Figure 1. Hydrolysis of tabun (3, 35)...... 15 Figure 2. Hydrolysis of sarin, cyclosarin and soman (3, 35)...... 15 Figure 3. Hydrolysis of VX and VR (2, 15)...... 16 Figure 4. Acetylation of chlorophenols...... 26 Figure 5. Hydrolysis of acetic anhydride...... 29 Figure 6. Hydrolysis of acetyl chloride...... 30 Figure 7. Reaction scheme for the synthesis of diacetyl methylphosphonate Lazarus, Benkovic & Benkovic (1979)...... 39 Figure 8. Mono ethyl ester of MPA (ethyl methylphosphonate)...... 41 Figure 9. 1H NMR spectrum and expansions for the product mixture from trial 2...... 41 Figure 10. GC-MS chromatogram of the trial 3 reaction product...... 43 Figure 11. Acid-catalysed esterification of carboxylic acids Zimmermann and Rudolph, (1965) (133)...... 46 Figure 12.Synthesis of alkyl methylphosphonates using silica chloride (131)...... 46 Figure 13. Synthesis of dialkyl methylphosphonates using methylphosphonyl dichloride at ambient temperature. Where R’ represents an alcohol group...... 47 Figure 14. GC-MS EI+ fragmentation of diethyl methylphosphonate: (a) Common fragment ions of diethyl methylphosphonate⋅+, (b) McLafferty rearrangement mechanism in diethyl methylphosphonate. Schemes from Boateng et al.l(2019) (137)...... 50 Figure 15..1H NMR spectrum of products formed in the reaction of methylphosphonyl dichloride and ethanol...... 52

9

Chapter 1. Introduction

1.0 Introduction

Organophosphate based (OP) chemical warfare nerve agents (CWNAs) pose a considerable threat to humankind due to their lethality and ease of production (1). Methylphosphonic acid

(MPA) is the final degradant of several nerve agents (soman, sarin, cyclosarin, VX and VR), and if there is any suspicion of the use of nerve agents, it can be used as a key indicator in the forensic pursuit of CWNAs (2-5). Verification of the presence of MPA is commonly performed by gas chromatography-mass spectrometry (GC-MS); however, to be amenable to analysis by GC-MS,

MPA must be converted to a volatile derivative (5-7). Currently, this process involves a time- consuming step of evaporating water from aqueous samples and extracts to complete dryness

(5, 7). This project aims to develop a method for the derivatisation of aqueous MPA, thereby, increasing the rapidity of analysis in an area where the timely turnaround of results is crucial.

This review will cover background information regarding CWNAs, including their toxicity and degradation. Descriptions of chemical analysis related to CWNAs, the importance of MPA as an analyte and its derivatisation for GC-MS analysis will provide context for the research project. A review of acetyl derivatisation is also included as it relates to the research aims.

1.1 Chemical warfare agents

Chemical warfare agents are a broad classification assigned to any chemical substance whose toxic properties are utilised to kill, maim or injure a person in the context of warfare (8).

Chemical warfare agents (CWAs) can be classified into five distinct categories: Choking, blistering, blood, riot control, or nerve agents (9, 10). Although the method of dispersal differs between agents they commonly include liquid, aerosol, vapour and dust (9). Chemical warfare continues to be among the most feared form of conflict. In recent years, the use of such toxic

10 agents has extended beyond military use with confirmed use of various CWAs in high profile assassinations and in terror attacks (11-13). This literature review will focus on nerve agents.

1.2 Nerve agents

Commonly, nerve agents are classified as OP compounds. These compounds contain a phosphorus atom, which is involved in a double bond to an oxygen atom, and a carbon- phosphorus bond (8, 14). Nerve agents can principally be divided into two groups: G-series

(German) (2) agents and V-series (venomous) (2) agents, named in accordance with their military designations (9). Prominent nerve agents are inclusive of tabun, sarin, soman, and cyclosarin belonging to the G-series and VX and Russian VX (VR) in the V-series. The common names, military designations, chemical names, and structures of OP nerve agents are presented in Table

1. The common names will be used throughout this review. All nerve agents are viscous liquids; however, G-series agents tend to be more volatile and present more of a vapour hazard (15). In contrast, V-agents are more potent (16), stable and less water soluble (14). Additionally, they are less volatile than G-agents and can persist in the environment for several weeks after release

(14).

OP-based nerve agents do not occur naturally (14). The G-series agents were developed during the 1930s, starting with the synthesis of tabun by German chemist Gerhard Schrader (17, 18).

This was followed by the development of sarin and soman, which are both attributed to research by the group of Wolfgang Wirth (19). Although G-series nerve agents were being developed and stockpiled during World War II, they were not put to use (14, 20, 21). In subsequent years, a new series (V-series) of sulfur-containing nerve agent was developed (16). British research based on the work of Farben, saw the development of O,O-diethyl S-[2-

(diethylamino)ethyl]phosphorothioate) (VG) (16). Originally sold as a pesticide under the brand name Amiton, VG was later withdrawn from the market due to the high level of toxicity (16).

11

Arguably the most well-known V-agents are VX, manufactured by the Americans during 1958

and Russian VX (VR) which was developed by the Soviets thereafter (22).

Table 1. Names and structures of key organophosphate chemical warfare nerve agents.

Common Military Chemical Name Vapour Volatility Structure Name Designation Pressure (mg/m3) (mmHg 20 or 25 °C) tabun GA ethyl 0.037 at 20 °C 610 O N dimethylphosphoramido- 0.07 at 25 °C P O cyanidate N

2.10 at 20 °C 22, 000 O sarin GB isopropyl P O methylphosphonofluoridate F

cyclosarin GF cyclohexyl 0.044 at 20 °C 438 O methylphosphonofluoridate P O F

soman GD pinacolyl 0.40 at 25 °C 3, 900 O methylphosphonofluoridate P O F

VX VX O-ethyl, S-2- 0.0007 10.5 S P N diisopropylaminoethyl O methylphosphonothiolate O

Russian VR S-(N,N-diethylaminoethyl) 0.00062 at 25 °C 8.9

VX isobutyl O S P methylphosphothiolate O

N

Adapted from Munro et al., 1999 (15), with additional information sourced from Bajgar, 2004 (23). Nerve agents were first deployed during the Iran-Iraq war (1983–1988) (17, 24), and on the

Kurdish in northern Iraq (1987–1988) (17, 21, 25). They have also been incorporated

intermittently into terrorism and assassinations (11-13, 26-28). An emblematic example of nerve

agent use is the 1995 terror attacks involving the release of sarin on five separate lines in the

12

Tokyo subway (11). The subway attack resulted in the death of 11 commuters and an additional

5000 individuals requiring medical attention (11). Thereafter, the release of sarin in Damascus during the Syrian military conflict (2013) caused a large number of casualties (26-28). While more recently nerve agents have been utilised in assassinations, prominent examples of which include, the high profile VX assassination of Kim Jong-Nam (2017) (12) and the Novichok agent

(A-234) poisoning of former Russian spy Sergei V. Skripal and his daughter, Yulia Skripal (2018)

(13). Since less is known about the Novichok agents, their structures and respective degradation pathways, they will not be discussed any further in this review. Due to the threat OP nerve agents pose to life and the environment, and the broader horrors associated with the use of toxic chemicals in warfare, the Chemical Weapons Convention (CWC) was developed (29). As detailed by the Organisation for the Prohibition of Chemical Weapons (OPCW) (10) and

Chauchan et al. (2008) (8), the CWC is a convention on the prohibition of the development, production, acquisition, stockpiling and use of chemical weapons. The convention additionally covers monitoring and destruction, which later developed into a treaty, accepted by 188 countries, signed on January 1993 and came into effect on the 27th of April 1997. The OPCW is tasked with the administration of the treaty as well as the regulatory inspections to ensure compliance.

1.2.1 Toxicity of nerve agents

Exposure to a nerve agent results in rapid absorption of the respective agent into the bloodstream via inhalation, oral ingestion and/ or contact with the skin (17). Nerve agents exert their effects by inhibiting the enzyme acetylcholinesterase (AChE) (14). Acetylcholine (Ach) and acetylcholine receptors (AChR) are integral to normal biological functioning (14, 30). AChR is a membrane protein that responds to the binding of ACh, which is a neurotransmitter. (14, 30).

AChE is responsible for hydrolysing ACh liberated at nerve-synapse, nerve-gland and nerve- muscle (neuromuscular) junctions (14). In a normal functioning system, a small amount of ACh is continuously liberated and hydrolysed by AChE (14). Nerve agents inhibit AChE though the 13 phosphorylation (formation of a P-O covalent bond) of a serine hydroxyl group in the active site of the enzyme (17). This results in the hyperstimulation of muscarinic and nicotinic ACh receptors in the peripheral and central nervous systems through the accumulation of ACh in the synapses (17). The toxic effects of nerve agent exposure largely depend on the dosage, with associated symptoms ranging from miosis, or pinpoint pupils, and blurred vision at lower doses, to nausea, vomiting, involuntary defecation, muscular twitching, convulsions and weakness, and death at higher doses (31). OP nerve agents exert extreme toxicity and acute exposure can be lethal at low levels with the LD50 of tabun, soman, sarin and VX in humans estimated as 24.3, 5,

14.3 and 0.14 mg/kg respectively (29, 32).

1.2.2 Organophosphate nerve agent degradation

The principal degradation processes for nerve agents are inclusive of photolysis, oxidation, hydrolysis, and microbial degradation (15). Different conditions and operations influence the various processes (15). In general, photolysis is largely associated with spills, particularly on solid surfaces, water surfaces or airborne release of the agent (15). Oxidation is relevant to exposure of the agent to air and/or natural oxidants in the soil or water (15). Microbial degradation is predominantly associated with burials and spills (15). Hydrolysis is pertinent in aqueous environments such as bodies of water and moisture in the soil and air (2, 15, 33, 34). This section focuses on hydrolysis pathways of the common CWNAs.

For many nerve agents, hydrolysis occurs rapidly, which adds a layer of difficulty to the identification of target analytes during the initial phase of an incident (2, 33, 34). As nerve agents follow highly specific hydrolysis pathways, the presence of their respective degradation products can be used to indicate the prior presence of the parent compound (2).

14

HO O + P H O O O N N P O rate-pH dependentO-ethylphosphorylcyanidate HHO P OH N rate-pHOH dependent - tabun O OH N P phosphoric acid O HO O-ethyl N,N-dimethylamido phosphoric acid

Figure 1. Hydrolysis of tabun (3, 35).

The rate of hydrolysis differs among nerve agents, being dependent on temperature and pH (5,

15). Tabun has a half-life of 8.5 hrs (pH 7) (15) and the pH-dependent hydrolysis pathways are shown in Figure 1 (3). The final degradant in each case is phosphoric acid (inorganic phosphate)

(5). Tabun is the only common nerve agent that does not degrade to MPA.

O O

P O H2O P O Rapid F OH

sarin IMPA H 2 O slow O O P O P O OH F H2O O H O Rapid 2 P OH slow OH cyclosarin CMPA MPA

O H 2 O O slow

P O H2O P O Rapid F OHH

soman PMPA

Figure 2. Hydrolysis of sarin, cyclosarin and soman (3, 35).

15

Sarin, soman and cyclosarin undergo rapid hydrolysis of the P-F bond in an aqueous environment, producing alkylphosphonic acids (3, 36). While the complete hydrolysis occurring in the second step is much slower than the first step, the final product in each case is MPA (3,

36). The corresponding degradation pathways are shown in Figure 2. The detection of both the intermediary and final hydrolysis product(s) of these G-series agents is important since these analytes may reveal the potential presence of nerve agents in test samples, where the parent compounds are below the detection limit (33).

O S P O H O 2 O N rapid P H O 2 O slow iBuMPA Russian VX (VR) O

HSCH2CH2N(iPr)2 P OH + OH O O H 2 H O N 2 MPA rapid P OCH CH slow S 2 3

P OHOH O H O 2 O EMPA - VX acid rapid S VX P OCH CH HOCH CH N(iPr) H 2 3 + 2 2 2 2 O rapid OHOH EMPTA S

P SCH CH N(iPr) 2 2 2 + EtOH

OHOH EA 2192

Figure 3. Hydrolysis of VX and VR (2, 15).

Degradation pathways for V-series agents are influenced by a range of factors (15). VR undergoes rapid hydrolysis of the P-S bond to an intermediary degradation product, shown in

Figure 3, with complete hydrolysis to the final product, MPA, occurring at a slower rate (3, 6, 15,

36). While degradation of VX may proceed via numerous pathways as shown in Figure 3, the final degradant, MPA, is common to both V agents (2, 15).

16

MPA is the resultant product of complete hydrolysis of several nerve agents and it possesses many forensic attributes of interest. For example, MPA is unable to undergo further hydrolysis, and is intrinsically more stable and, therefore, persists longer than other degradation products

(5, 34). In addition, MPA is not naturally occurring, although it may be released into the environment as a result of the degradation of OP-pesticides such as O,O-bis(2,4,5- trichlorophenyl)methylphosphonate (5, 15). Similarly, it is believed to be the degradant product of some flame retardants such as dimethyl methylphosphonate (5, 15). Due to the aforementioned properties, MPA is a key analyte in chemical analysis related to ensuring compliance with the CWC (7).

1.3 OPCW chemical analysis

The OPCW’s mission is to implement the provisions of the CWC (37), and the chemical analysis related to this mission takes various forms. A significant role is investigating alleged use, which may either be rapid response and assistance missions immediately after an incident or detection long after an incident. Also important are site or compliance inspections. These inspections normally involve the monitoring of scheduled chemicals, such as, warfare agents, precursors, degradation products and products that may be formed upon decontamination of a scheduled chemical. The OPCW also require competency testing of both the accredited laboratories and related personnel. To maintain the designated laboratory status, a laboratory must undergo and pass a minimum of one proficiency test per annum (38). In proficiency testing, laboratories are required to determine if any of a very large set of chemicals relevant to the convention is present in blind samples (38).

The recommended operating procedures (ROPs) outline procedures for verification and analysis of chemical disarmament. These procedures were established via international collaboration with expert laboratories working in the field of chemical weapons convention-related analysis.

The ROPs are the basis for accreditation of OPCW laboratories, personnel training, and sampling 17 and analytical procedures. The ROPs by Vanninen (2017) (7) will underpin discussion in this section.

Samples for OPCW chemical analysis typically involve complex environmental matrices such as water (39-43), soil (43-46), air (43, 47, 48), munitions (43, 44, 49) and manufactured materials such as polymers, paint or clothing (43, 44, 50). Sample extraction can principally be divided into two sub-categories, which are designed to recover chemicals of different polarities and/ or for differing analysis methods. General procedures involve extraction of the matrix with an organic solvent to recover the non-polar analytes, such as undegraded nerve agents, and aqueous extraction of polar, water-soluble degradation products. It is recommended that a new subsample be prepared for each extraction method; however, this is not always possible as the sample may be too small or not homogenous. In these instances, it is imperative that the organic extraction is carried out prior to aqueous extraction to avoid further hydrolysis of the agents because of the extraction method.

The OPCW inspectorate requires two or more different spectrometric techniques and reference standards for the unambiguous identification of CWAs, their precursors and their degradants

(51). The choice of the analytical technique will depend on the properties associated with the analyte. Typical detection techniques employed are GC-MS, gas chromatography coupled to tandem mass spectrometry (GC-MS/MS), liquid chromatography coupled to mass spectrometry

(LC-MS), liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS), and nuclear magnetic resonance (NMR) spectroscopy. The most accessible and widely available of these around the world is GC-MS. However, the resultant products from hydrolysis of organophosphorus nerve agents, namely alkyl alkylphosphonic acids and alkylphosphonic acids, exhibit low volatility and moderate to high polarity. Therefore, they are required to undergo derivatisation to facilitate analysis by GC-MS (5). Derivatisation of analytes for GC-MS detection of analytes in OPCW chemical analysis is discussed in detail in the following section.

18

1.4. Derivatisation

In GC-MS, typically, the MS can only detect compounds that are stable, volatile, and able to travel successfully through the chromatographic column during run time (52). It is well known that polarity, molecular mass and thermal stability of native compounds can potentially limit the application of GC-MS (52-58). An analyte’s polarity can generally be attributed to the presence of –OH, –NH, –SH or –COOH groups (52), which means their thermal decomposition (if unstable at the column or inlet temperature) may be accompanied by such processes as decarboxylation, the formation of cyclic structures, and dehydration. It is these resultant products that pass through the chromatographic column being recorded by the MS, not the native compound (52,

53). Moreover, the analysis of many compounds may prove to be difficult due to the interactions between compounds in a mixture, or between compounds and the heated inlet and the column

(52-58). For example, those containing the above mentioned polar functional groups have a tendency to form intermolecular hydrogen bonds (52) that reduce their volatility, and may also cause interactions between the compounds and the stationary phase (52). Such interaction can cause peak broadening, which results in the formation of asymmetrical peaks and/ or poor peak resolution (52).

Derivatisation is a procedure that may be applied for the modification of a functional group of a polar chemical compound (52, 59). The technique produces a structurally similar derivative, which is amenable to separation by chromatographic techniques (52, 59). The primary objectives of derivatisation in GC-MS analysis are (52-57, 59):

1) Increase volatility, by reducing polarity.

2) Improve compound stability and chromatographic properties.

3) Improve selectivity and sensitivity, increasing detectability.

4) Enhance separation and thus structural information obtained via the spectra.

19

5) Improve peak resolution and/ or symmetry (increasing reproducibility of peak shapes,

height and area).

For these reasons, derivatisation plays an important role in chromatographic separation and MS analysis.

1.4.1 Types of derivatisation

Derivatisation reactions for GC or GC-MS can be classified into three general types: acylation, silylation and alkylation (52). These reactions can be represented by a general equation:

Y-H + R-X → Y-R + HX

In this equation, the reagent ‘R-X’ contains a leaving group ‘X’ and a group ‘R’, which imparts the desirable properties after derivatisation. Group ‘R’ donates to the respective analyte making it more amenable to chromatographic analysis.

As stated by Kumirska et al. (2013) (52), acylation involves the use of carboxylic acids or their derivative reagents to convert compounds with active hydrogens –OH, –NH or –SH into esters, amides and thioesters respectively. Silylation involves the replacement of a hydrogen in –OH, –

SH, –NH2, and –COOH groups with a silyl group. Alkylation includes the replacement of an active hydrogen by an aromatic group (such as benzyl) or aliphatic group. Alkylating reagents react with compounds containing acidic hydrogens such as carboxylic acids and phenols. These reagents can be used in the preparation of ethers from alcohols, thioethers from sulfur containing compounds, and N-alkylamines, amides, and sulfonamides from amines.

When selecting a derivatisation method for GC or GC-MS analysis, multiple aspects are to be considered, which have been outlined by Orata (2012) (59). These include:

1) Reagent produces a ≥95 % product yield.

2) Produce a stable derivative (with consideration to time). 20

3) Produce a derivative that does not interact with the chromatographic column.

4) Derivatisation agent does not rearrange or make structural alterations to the native

compound.

5) Does not contribute to the loss of sample during the reaction process.

Both alkylation (methylation and pentafluorobenzylation) and silylation are applied in OPCW chemical analysis of phosphonic acids, as explained in the following section. Acetylation of MPA will be attempted in the research project related to this review.

1.4.2 Derivatisation of phosphonic acids

The feature that makes nerve agent degradants suitable analytes, namely their persistence in the environment, is also the feature that makes them less amenable to GC-MS analyses. As they are less volatile than their parent analytes, phosphonic acids must undergo derivatisation to be detected by GC-MS (5).

The most common methods employed for derivatisation of phosphonic acids are alkylation (3,

6, 36, 60) and silylation (2, 52). Common derivatising reagents employed by the OPCW for formation of these derivatives are diazomethane for alkylation (specifically methylation), and

MTBSTFA (N-methyl-N-tert-butyldimethylsilyltrifluoroacetamide) and BSTFA (O- bis(trimethylsilyl)trifluoroacetamide) for silylation (5, 7, 51, 61, 62). The widely accepted ‘gold standard’ methylation agent is diazomethane, because it does not produce any GC-MS interfering by-products and is highly volatile and reactive even under mild conditions (5, 6, 59).

Diazomethane, in a dry organic solvent (such as methanol), rapidly converts (15 minutes, under appropriate conditions) alkyl phosphonic acids at ambient temperature to methyl esters (alkyl methyl methylphosphonic acids) as shown in Table 2, entry 1) (5, 63). Provided that a large excess of reagent is present, methylation of alkyl methylphosphonic acids can generate yields of

≥99 % (5). Methyl esters produced via the aforementioned reaction are not sensitive to trace

21 amounts of water and, where necessary, clean-up can be carried out through the use of silica on the column (5). There are several drawbacks associated with diazomethane derivatisation procedures (4, 5, 7, 51, 64). The most prominent disadvantages of diazomethane are:

1) The limited shelf-life (even when stored at 4 °C) of the reagent.

2) Explosive properties and high toxicity (possibly even carcinogenic).

3) Reportedly produces poor peak shapes.

4) Problematic in trace analysis as the highest mass ions fall below m/z of 200 [MPA has a

m/z of 95].

More recently, trimethyloxonium tetrafluoroborate (TMO·BF4) has been utilised as a safer alternative to diazomethane (6). TMO·BF4 produces optimal yields in less than two hours under mild conditions (6). However, it too has a limited shelf-life (6, 65).

As outlined by Black & Muir (2013) (5) and Vanninen (2017) (7), silyl derivatisation is widely considered to be a safer, and less problematic alternative to methylation. The most broadly applicable procedure for the derivatisation of CWNA degradation products is trimethylsilyl

(TMS) or tertbutyldimethylsilyl (TBDMS) esters or ethers, shown in Table 2, entries 2 and 3. The recommended operating procedures state that conversation to TMS derivatives is applicable to on-site analysis, while for off-site analysis TBDMS derivatives are preferentially used as they are more stable in storage and in the presence of trace amounts of water. Silyl derivatives are highly versatile, and able to be analysed by a variety of methods inclusive of atomic emission detection

(AED), flame photometric detection (FPD), nitrogen-phosphorus detection (NPD), and MS. The method is highly effective and relatively rapid for the derivatisation of phosphonic acids. For example, N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) (with 1 % trimethylsilyl chloride

(TMSCI)) can be used at 60 °C for 30 minutes producing satisfactory yields of MPA-TMS derivative. Although silylation is commonly accepted for GC-MS analyses of this class, the method is not without its disadvantages (66), the most prominent of which is the expense of

22 reagents and the sensitivity to the presence of magnesium and calcium ions extracted from environmental samples (41, 67, 68).

While silyl and methyl derivatisation are the most common methods, there are several other known derivatives of MPA. Pentafluorobenzyl esters (Table 2, entry 4) are typically slow to form and require particular reaction conditions to form (101). Less is known about the acetylated derivative of MPA, with only two references to diacetyl methylphosphonate (Table 2, entry 5).

The first of which, Hennig & Sartori (1983) describes research on bis(acyloxy)phosphine oxides

(69). The related research is unavailable in an English translation. The second, by Lazarus,

Benkovic & Benkovic (1979) (70), describes research on the synthesis and properties in enzymatic reactions of substrates containing the methylphosphonyl group. Acetylation has not been utilised for the qualitative or quantitative analysis of MPA relating to nerve agents.

There are numerous issues associated with derivatisation as described by Black and Muir (2003)

(5). Derivatisation methods can introduce error into chromatographic analysis. Major issues can arise due to extraneous materials extracted from the matrix, including water, which may suppress derivatisation or react with the derivatising agent producing a complex background.

Similarly, a large excess of reagent may be required to drive the reaction to completion during derivatisation. This potentially leads to a chemical background or reduced column lifetime, which limits analysis. Additionally, derivatisation of polar analytes like phosphonic acids often require concentration prior to analyses. Concentration involves the time-consuming process of evaporating water from aqueous samples and extracts to complete dryness.

In pursuit of a rapid, in-field, qualitative detection of MPA, the viability of bypassing the evaporation of water should be considered. The ideal derivatisation agent for this application would be inexpensive, stable at ambient temperature, pose minimal risk to health and, ideally, not overly susceptible to hydrolysis. As such, an extensive search of the literature aims to assess the viability of trialling alternatives to the silyl and methyl derivatisation methods. One such possibility is acetylation. Although significant previous reports on the effective, inexpensive, and

23 versatile derivatisation power of acetylation exist (71-78), there are no precedents for acetylation of phosphonic acids. The viability of trialling acetyl derivatisation for the qualitative analysis of MPA is discussed in Section 1.5 Acetylation.

Table 2. Derivatives of methylphosphonic acid.

Entry Derivative Reagent Structure

1 methyl diazomethane O O P O

2 tert- MTBSTFA

butyldimethylsilyl Si O O P O Si

3 trimethylsilyl BSTFA Si O O P O Si

4 pentafluorobenzyl PFBBr F F

F O F F O P O F F F

F F

5 acetyl AA or AC* O O O P O

O

*acetic anhydride and acetyl chloride as acetylation agents is theoretical and has yet to be trialled for viability (this is discussed in Section 1.5.1 Acetylation reagents).

24

1.4.3 Derivatisation without the removal of the water

Derivatisation of aqueous MPA without removing water would be a great time saver in OPCW chemical analysis. The only precedents for this involve solid-phase microextraction (SPME).

Developed and commercialised during the 1990s (79), SPME was originally utilised for the extraction of semi-volatile substances from a simple matrix for subsequent analysis by GC (80).

SPME consists of two processes (81):

1) Partitioning of the analytes between the sample and the coating.

2) Desorption of concentrated analytes into the analytical instrument.

Derivatisation can be carried out in conjunction with SPME, and there are three different techniques, as described by the developers, Pan and Pawliszyn (1997) (82):

1) Derivatisation in the sample matrix.

2) Derivatisation in the fibre coating, either following or simultaneous to the extraction.

3) Derivatisation in the (GC) injector port.

In-situ SPME involves derivatisation directly in the fibre coating of the SPME device (83). Sng &

Ng (1999) (83) applied this method to the analysis of CWA related compounds in water. Their research exposed SPME fibres to derivatisation reagent prior to insertion into aqueous solutions of CWA analytes. The solutions were stirred then re-exposed to the derivatisation reagent, and subsequently analysed via GC-MS completely bypassing the need to remove the water from samples. In-situ SPME resulted in minimal depletion of the analyte and successfully aided in the analysis of semi-volatile CWA-related compounds. As such, the technique was posed as an alternative to current ROPs for rapid on-site analysis. It was recently applied in proficiency testing, in which it met the OPCW standard.

An alternate method, hollow fibre-based liquid-phase microextraction (HF-LPME) utilises a hollow fibre, containing an organic solvent to extract analytes of interest from aqueous samples

(61, 84). Lee et al. (2007) (61) performed the in-situ derivatisation, extracting CWA degradation

25 products from aqueous samples using MTBSFTA with a solvent mixture (containing BSTFA and octane, 5:1 ratio). In this method, the reagent is protected within the hollow fibre, which overcomes potential moisture sensitivity. HF-LPME allows for the unhindered extraction and analysis via GC-MS and reportedly has better limits of detection than the in-situ SPME method.

In 2018, research carried out by Dival (85) and Chua (86) investigated tert-butyldimethylsilyl derivatisation of aqueous MPA without the removal of water. Derivatisation was carried out in a two-phase system consisting of aqueous MPA and an organic layer of n-hexane and the derivatisation reagent. Presumably, the reaction would occur at the interface of the two phases with the derivative product partitioning to the organic layer. The organic layer was analysed by

GC-MS and the TBDMS derivative detected for a 1000 ppm solution of MPA. Although the derivative was not detected for lower concentrations of aqueous MPA, the result showed the potential applicability of such a method to OPCW chemical analysis. Prior to this, there is no evidence in the literature to suggest that derivatisation via the two-phase method for analysis of aqueous MPA had been tested. This method provides an inexpensive and simple alterative to the aforementioned aqueous derivatisation methods.

Direct acetylation of aqueous chlorophenols, according to the equation shown in Figure 4, was carried out by Al-Janabi et al, (2012) (71).

Figure 4. Acetylation of chlorophenols.

The presence of sodium hydroxide would deprotonate the acidic phenols and enhance their reactivity with acetic anhydride. Acetylated chlorophenols were subsequently extracted with n- hexane, and the organic layer analysed by GC-MS. This precedent for acetylation in basic aqueous solution with subsequent organic extraction is potentially applicable to acetylation of

26 aqueous MPA. Both chlorophenols and MPA are acidic, but MPA may be less reactive due to the electron withdrawing effect of the phosphorus-oxygen bond.

1.5 Acetylation

The role of acetylation, commonly acetylated functional groups and the viability of trialling acetyl derivatisation in OPCW chemical analysis is presented in this section. While the acetyl derivative of MPA is a known compound (70), there are no reports of direct acetylation of methylphosphonic acid. The premise of this section is to review acetylation and assess the viability of acetyl derivatisation of MPA.

Acetylation refers to the addition of an acetyl (CH3CO) functional group onto an organic compound. Commonly acetylated functional groups include amino, sulfhydryl and hydroxyl groups (71, 77, 87). Acetyl donation is a general reaction that plays a major role in a diverse range of chemical and biochemical functions, such as:

• Protein modification, both co-translation and post-translation acetylation to control

their biological activities (88).

• The protection of phenols, thiols, alcohols, and amines in organic synthesis (87, 89).

• Modification of chemical properties for a diverse range of functions such as improving

thermal and structural stability, and amenability to analytical techniques (71, 75, 90,

91).

In biological systems, acetylation occurs during processes related to gene expression and metabolism (78, 92, 93). Acetyl coenzyme A (acetyl CoA) donates the functional group during post-translational or co-translational modification (78, 93-95). Which manifolds effects at both the protein and metabolome level, aiding in the regulation of gene expression and metabolic processes as part of a highly complex system (78, 92-95).

27

Acetylation is commonly carried out in the industrial food industry, often used for the stabilisation of starches (91). Acetic anhydride, vinyl acetate or acetic acid acts as acetyl donors at the hydroxyl group of the glucose monomers, which ultimately alters the molecular structure of the starch (76, 90, 91, 96). A prominent example of acetylated starches in the food industry is additive E 1420, which affects both the stability and consistency of products it is added to (74).

A common technique applied in organic synthesis is the use of protecting groups (89). The addition of an acetyl protecting group can overcome decomposition and/or prevent undesirable side reactions by blocking the reactivity of a functional group (87, 89).

Acetyl functional groups may be introduced to modify chemical properties for various purposes.

As an example, to enable the use of GC-based analytical techniques (97-99). For instance, acetylation is the most common method applied for derivatisation of phenols and has additionally been applied to carbamate and urea pesticides (97, 98). Acetylation reportedly increased thermostability, reducing polarity, as well as improving sensitivity and/or peak shape

(97-99).

1.5.1 Acetylation reagents

Among the most common acetylation reagents are acetic anhydride and acetyl chloride, which are frequently substituted in the place of acetic acid due to their higher reactivity and ability to produce higher yields (71, 77, 98, 100-102). Other known acetylation agents are acetyl CoA, ketene and vinyl acetate. Acetyl CoA carries out acetylation in biological systems (103, 104).

Ketene is versatile and theoretically does not produce any objectionable by-products, but is less frequently used because it is toxic and challenging to prepare (105). Vinyl acetate is a more novel agent, which has been applied in the place of acid anhydrides or acid chlorides as it is reportedly less sensitive to lower pH (75). The acetyl donors and compounds commonly used to acetylate are shown below in Table 3.

28

Table 3. Acetyl donors and the compounds they acetylate.

Acetyl Donor Acetylated Compound(s) Reference

acetic anhydride starches, chlorophenols, phenols, alcohols, (71, 106- glycerol, amino groups, amines, wood, 111) acetyl chloride proteins (101, 112)

acetyl coenzyme A proteins (78, 103, 113) acetic acid alcohol, phenols (102, 114)

ketene alcohols, mercaptans, carboxylic acids, glycols, (105, 115) polyhydroxy compounds, amides, hydrocarbons, nitroparaffins, carbohydrates, imides vinyl acetate wood, starch (75)

A common feature of acetylation agents is that they are moisture sensitive (116). This is discussed with respect to acetic anhydride and acetyl chloride since they are the most commonly used reagents. The equation for acetic anhydride hydrolysis is shown in Figure 5. Acetic anhydride hydrolyses to two equivalents of acetic acid (117, 118). The calculated half-life for acetic anhydride is 4.4 minutes at 25 °C, and 8.1 minutes at 15 °C (117, 118).

O O O

H2O

O HO

Acetic anhydride Acetic acid

Figure 5. Hydrolysis of acetic anhydride.

In the presence of water, acetyl chloride rapidly undergoes hydrolysis to form acetic acid and hydrochloric acid. The equation for this reaction is shown in Figure 6. Due to its high reactivity and concomitant instability, it is an ineffective reagent under aqueous conditions (112).

29

O O

H2O + HCl Cl HO

Acetyl chloride Acetic acid

Figure 6. Hydrolysis of acetyl chloride.

Although acetylation agents are highly sensitive to water, it is postulated that the use of excess reagent may overcome this issue and it may be possible to acetylate compounds in the presence of water.

1.5.2 Acetylation of methylphosphonic acid

Although it is evident that acetylation of the hydroxyl group (OH) is a general reaction, there is currently no literature directly pertaining to the acetylation of MPA. Additionally, there is limited research on the acetylation of similar compounds such as alkylphosphonic acids, sulfonic acids or carboxylic acids. Although Dunbar and Garven (1955) (119) carried out acetylation of monocarboxylic acids with ketene, little detail is provided on the associated procedures and subsequent outcomes. Direct acetylation of these functional groups may be limited by electron withdrawing effects that reduce the reactivity of the hydroxyl group oxygen atom with acetylation reagents. It is possible that this could be overcome by conducting the reactions in basic conditions to deprotonate the hydroxyl group and enhance its reactivity to acetylation.

With consideration of the literature reviewed, the ideal reagent for acetylation of MPA would have the following features:

1) Relatively inexpensive, easy and relatively safe to use.

2) Have established derivatisation ability in the presence of water.

3) Increase the volatility of MPA.

30

It was determined that acetic anhydride could be an appropriate reagent, as it is comparatively safer and/or relatively inexpensive to the currently recommended reagents for methyl and silyl derivatisation of MPA. Acetic anhydride is a versatile acetylation agent that has successfully been utilised for direct acetylation of a range of compounds. As discussed in Chapter 1. Section

1.4.3 Derivatisation without the removal of the water, aqueous chlorophenols were acetylated using excess acetic anhydride to overcome competing hydrolysis of the reagent (71). Acetyl group modification is a common technique used to adjust the chemical properties associated with an analyte (97-99). The addition of an acetyl group to increase the volatility of polar molecules to enable them to be detected via GC-MS is frequently carried out and should be applicable to the derivatisation of MPA (69, 70).

1.6 Concluding statements and research aims

Nerve agents are highly toxic and pose a considerable threat to both life and the environment.

The final degradation product of several OP nerve agents, MPA is a key analyte of interest in

OPCW chemical analysis. GC-MS is the most widely available and conventional method employed by the OPCW for the identification of chemical warfare nerve agents and their degradants. To be amenable to GC-MS analysis, MPA is required to undergo derivatisation, which includes the time-consuming evaporation of water from aqueous samples and extracts to complete dryness. The two-phase method evaluated in this review has real-world applications and could provide a rapid alternative to the current ROPs. The protocols investigated by Dival

(2018) (85) and Chua (2018) (86) act as a proof of concept for derivatisation of MPA via a two- phase system that is yet to be fully explored.

Currently, silyl and methyl derivatisation are the favoured methods of derivatisation for the analysis of polar degradation products in OPCW testing. These methods have many associated drawbacks, as reagents tend to be expensive, toxic, have a limited shelf-life or require specific reaction conditions. The versatility and inexpensive nature of common acetylation reagents,

31 such as acetic anhydride, are well known. Acetylation of the OH functional group to reduce polarity, making an analyte more amenable to GC-based techniques is widely accepted and has previously been applied to phenols in a two-phase system. Moisture sensitivity was found to be sufficiently overcome with use of excess reagent. However, the viability of diacetyl methylphosphonate as an analyte for OPCW analysis has yet to be researched. Because of the aforementioned advantages of acetylation as a method of derivatisation, it is intended that it will be trialled for MPA, as a safer and less expensive alternative to the current ROPs.

In the assessment of the viability of the diacetyl methylphosphonate as an analyte, it is appropriate that a reference sample is prepared via the protocols detailed in the synthesis by

Lazarus, Benkovic & Benkov (1979) (70). The reference sample will be characterised by MS and

NMR and used to develop GC methods. Thereafter, acetylation will be trialled for the qualitative detection of MPA in a two-phase system for analysis via GC-MS, in alignment with the following research aims.

The broad research focus of this study is effective derivatisation of aqueous MPA in a two-phase system. This can be separated into three explicit aims:

1. Prepare a reference sample of diacetyl methylphosphonate.

2. Develop GC-MS methods for detection of diacetyl methylphosphonate.

3. Trial acetylation of MPA in two-phase derivatisation.

32

Chapter 2 Experimental

2.1 General methods

2.1.1 Instrumentation and instrumental parameters

The determination of solvent water content was carried out using a Mettler Toledo C20

Coulometric Karl Fisher (KF) titrator using CombiCoulomat KF reagent.

GC-MS analysis was carried out on a Shimadzu GCMS-QP2010S (Shimadzu Australasia,

Rydalmere, NSW, Australia), with an SGE BPX5 GC capillary column (5 % phenyl polysilphenylene-siloxane) (length 30 m, I.D. 0.25 mm, film thickness 0.25 μm) and ultra-high purity helium used as carrier gas (BOC, Sydney, NSW, Australia).

GC analytical parameters: 3 sampler pre/post rinses, injector temperature 270 °C, initial column temperature 60 °C was held for 2-4 minutes then ramped to 250 °C at 16 °C/min and held for 6 minutes. The total run time was 20.07–22.17 minutes with 2.5 minutes solvent delay. The pressure was set at 8.0 psi, total flow 5.9 mL/min, column flow 0.98 mL/min, linear velocity 36.5 cm/sec, purge flow 3.0 mL/min, with a split ratio ranging between 2:1–50:1. The interface were

270 °C and the ion source was set to 200 °C. The MS scan ion masses in the range of m/z 40 to

250-500 at 70 volts.

All NMR spectra were acquired using a Varian 400 MHz spectrometer, regulated at 25 °C with

1 samples dissolved in acetone-d6/ DMSO-d6/ chloroform-d. H NMR spectra were measured with resonance frequency of 399.9 MHz. 13C NMR spectra were recorded at a resonance frequency of 100.6 MHz and 31P NMR spectra at 162.9 MHz resonance frequency. Chemical shifts were referenced to solvent residual signals using NMR Chemical Shifts of Common Laboratory

Solvents as Trace Impurities, Gottlieb, Kotlyar, and Nudelman (1997) (120), is the primary reference utilised for a comparative discussion of NMR solvent chemical shifts in this thesis. 31P

NMR spectra were recorded with 85 % phosphoric acid as an external chemical shift reference

(δ 0 ppm).

33

2.1.2 Reagents and solvents

Chemicals and solvents were used as supplied, unless otherwise specified. MPA (98 %), N,N′- dicyclohexylcarbodiimide (DCC) (99.0 %) and butan-1-ol (99.2 %) from Sigma-Aldrich. Thionyl chloride from Riedel-deHaën. Silver nitrate analytical reagent (AR) and ethanol, methanol, isopropyl alcohol AR (≥99.8 %) from UNIVAR. Sodium acetate anhydrous AR, potassium carbonate AR and chloroform AR from Chem-Supply. Laboratory grade diethyl ether was decanted from general stock. CombiCoulomat KF reagent (containing methanol) from

AquastarTM. Sulfuric acid AR (98 %) from Lab-Scan. Acetonitrile (99.9 %) from VWR Prolab

Chemicals. Silica gel (0.5-0.5 mm) from MERCK. Acetone-d6 (D, 99.9 %), chloroform-d (D, 99.9

%), and dimethyl sulfoxide-d6 (DMSO) (D, 99.9 %) from Cambridge Isotope Laboratories, Inc.

2.1.3 Solvent drying procedures

Following the respective drying processes specified in Table 4 (Vogel 1956) (121), a portion of the solvent was tested in triplicate via a KF titrator, using CombiCoulomat KF reagent. Water content <1 % was deemed to be sufficiently dry. Molecular sieves were regenerated by heating at 250 °C in a furnace overnight, then cooled in a desiccator.

Table 4. Procedures for drying solvents.

Solvent Drying Procedure Average Water Content by KF analysis diethyl ether Pre-dried over sodium wire (overnight) and distilled 0.007 %

chloroform Decanted into a dry amber bottle over 4A molecular 0.9 % sieves Pre-dried over potassium carbonate for 72 hrs and 0.05 % distilled into an amber bottle for storage

34

2.1.4 Preparation of methylphosphonyl dichloride

Thionyl chloride (1 mL, 1.63 g, 0.014 mol per 0.25 g, 0.002 mol MPA) was added dropwise to solid MPA over a period of 2 minutes, while rapidly stirring, then refluxed for 3 hours. Excess thionyl chloride was removed by rotary evaporation at 60 °C for 35 minutes under house vacuum. Where specified, methylphosphonyl dichloride was purified using reduced pressure distillation (55–70 °C, 8.4–8.9 Torr).

2.1.5 Preparation of silver acetate

Aqueous sodium acetate (1.76 mol/L, 50 mL, 0.088 mol) was added to aqueous silver nitrate

(0.88 mol/L, 100 mL, 0.088 mol). The white precipitate was isolated using vacuum filtration, washed with water, ethanol and diethyl ether then dried at 100 °C for 1 hour as per the method of Lazarus, Benkovic & Benkovic (1979) (70). The solid product was characterised by 1H NMR

(DMSO-d6) and stored in a cool, dry, tightly sealed, light-resistant container.

2.1.6 Preparation of silica chloride

Silica chloride was prepared via an adaptation of the method by Sathe et al. (2006) (122), a suspension of silica gel (5 g) in dry chloroform (25 mL) was added dropwise to thionyl chloride

(13 mL, 21.2 g, 0.18 mol) at room temperature. After stirring at room temperature for 2 hours, the solvent and excess thionyl chloride was removed under reduced pressure (8.7 Torr). The product presented as an off-white powder, which was stored in a tightly sealed vessel and used without characterisation.

35

2.2 Trialled preparation of diacetyl methylphosphonate

Synthesis of the reference compound diacetyl methylphosphonate was attempted using an adaptation of the method by Lazarus, Benkovic & Benkovic (1979) . Specifications for each trial are given in Table 5. A suspension of methylphosphonyl dichloride (0.1 g, 0.0008 mol) in diethyl ether/ chloroform (4 mL) was added dropwise over 1 minute to a stirred suspension of silver acetate (0.35 g, 0.0021 mol) or sodium acetate (0.18 g, 0.0022 mol) in diethyl ether/ chloroform

(4.5 mL) at room temperature. The reaction mixture was stirred at room temperature in the dark overnight, then gravity filtered to remove excess acetate. The solvent was removed from the filtrate via rotary evaporation under house vacuum.

Table 5. Variations of acetate, solvent and purification method in trials of diacetyl methylphosphonate synthesis.

Trial Acetate Solvent Purification methylphosphonyl dichloride 1 silver acetate Dry diethyl ether Rotary evaporation at 60 oC for 30 minutes under house vacuum 2 silver acetate Dry chloroform Rotary evaporation at 60 oC for 30 (molecular sieves) minutes under house vacuum 3 sodium acetate Dry chloroform Rotary evaporation at 60 oC for 30 (potassium minutes under house vacuum carbonate/ distilled) 4 silver acetate Dry chloroform Rotary evaporation at 60 oC for 30 (potassium minutes under house vacuum. carbonate/ distilled) Then Purified by reduced pressure distillation (55 oC, 8.6 Torr)

2.3 Preparation of dialkyl methylphosphonates

2.3.1 Conversion of methylphosphonyl dichloride to alkyl methylphosphonates

A stock of methylphosphonyl dichloride was prepared according to the general procedure given in Section 2.1.4 Preparation of methylphosphonyl dichloride and purified by reduced pressure distillation. Excess of dry alcohol (methanol, ethanol, propan-2-ol, butan-1-ol, or benzyl alcohol)

(3 mL) was added dropwise to methylphosphonyl dichloride (0.12 g, 0.0009 mol) over 1 minute.

The reaction vessel was agitated by hand initially to ensure proper mixing of reactants. The reaction mixtures were then magnetically stirred at room temperature for times ranging from

10 minutes to 24 hours. The reaction mixture was sampled (10 µL in chloroform 1.5 mL) after 36

10, 20, 30, and 40 minutes, then 12 and 24 hours to qualitatively monitor the reaction progress by GC-MS.

2.3.2 Trialed esterification of methylphosphonic acid with coupling reagents/ catalysts

An excess of dry ethanol or methanol (3 mL) was added dropwise to MPA (0.1 g, 0.001 mol) over

1 minute. Subsequently sulfuric acid (1.53 mol/L, 10 µL, 0.00002 mol), silica chloride (0.1 g), or

DCC (0.4 g, 0.002 mol) was added. The DCC trial additionally included acetonitrile (7 mL) solvent.

The reaction mixtures were then magnetically stirred at room temperature and sampled (10 µL in chloroform/ acetonitrile 1.5 mL) after 1, 2, 6, 12 and 24 hours to qualitatively monitor product formation by GC-MS. Samples were subsequently analysed by NMR if a phosphorus-containing species were detected via GC-MS using the similarity search function to compare the mass spectra of chromatogram peaks to a database.

37

Chapter 3. Results and discussion

This chapter outlines and discusses the various methods trialled in the pursuit of an inexpensive and rapid alternative to current derivatisation methods for MPA. Two acetates (sodium and silver acetate) and several alcohols (methanol, ethanol, propan-2-ol, butan-1-ol, and benzyl alcohol) were trialled as inexpensive derivatisation reagents. Different synthesis methods were employed, including chlorination, esterification and heterogeneous catalysis. The resultant outcomes of each preparation will be discussed herein.

When analysing the experimental data, it was important to consider all possible products that may have formed. MPA possesses the ability to couple and form anhydrides, compounds with

P-O-P bonds. Additionally, MPA can undergo substitution of one or both hydroxyl groups/acidic hydrogen atoms and mixtures of mono and diesters may form.

The analysis was further complicated by the rate of the reactions, along with the polarity and instability of intermediates and products. Some compounds were either too polar to be detected via GC-MS or too unstable to lend themselves to NMR analysis. Several reactions undertaken in this project occurred rapidly and acknowledging the time difference between preparation and analysis is crucial. The delay between GC-MS and NMR analysis often lead to observable differences between product formation seen via one analytical technique to the next.

3.1 Trialled preparation of diacetyl methylphosphonate

Four experimental trials were undertaken in pursuit of the first research aim, preparing a reference sample of diacetyl methylphosphonate. The procedures were adaptations of the

Lazarus, Benkovic & Benkovic (1979) (70) method shown in Figure 7 and described in Section

2.2. Trialled preparation of dialkyl methylphosphonate. Initial chlorination of MPA by thionyl chloride gives the intermediate product methylphosphonyl dichloride, which reacts with silver acetate to produce diacetyl methylphosphonate.

38

O O O O AgOAc, SOCl Et O (dry) O P OH 2 P Cl 2

O P O OH Cl

Figure 7. Reaction scheme for the synthesis of diacetyl methylphosphonate Lazarus, Benkovic & Benkovic (1979).

Successful preparation of silver acetate (preparation outlined in section 2.1.5) was verified by 1H

NMR. Silver acetate was compared to a predicted spectrum developed by ACD Labs Software

V11.01 (123), through which it was determined that the 1H NMR was consistent with the formation of silver acetate. MPA and methylphosphonyl dichloride were also analysed by NMR to obtain reference spectra that would enable identification of unreacted starting materials in the spectra of reaction products. The 1H NMR spectrum of MPA has a signal at δ 1.47 (d, 2) ppm for its methyl group protons and another at δ 9.36 (s, 1) ppm attributable to the –OH group proton. When analysed at various concentrations, the chemical shift associated with the –OH group proton ranges between δ 7.68-9.36 (s, 1) ppm, which may be due to the different extent of hydrogen bonding at different concentrations (124). The 13C NMR spectrum of MPA gave a signal at δ 12.07 ppm (d, 2) which is assigned to the methyl group carbon atom. The 31P NMR spectrum of MPA has one signal at δ 31.76 (s, 1) ppm due to the presence of only one phosphorus atom.

The 1H NMR, 13C NMR and 31P NMR spectra for the intermediate product methylphosphonyl dichloride were highly complex which may be attributed to its instability. The phosphorus spectrum indicated the presence of several phosphorus-containing compounds, with no signals being indicative of the presence of unreacted MPA. The mixture was presumed to contain methylphosphonyl dichloride and used in the next step of the reaction scheme.

The reaction product mixture for the first trial (Table 5, line 1) was analysed by GC-MS and the mass spectrum of each peak was compared with the NIST data base (125) linked to the

39 instrument processing software. This process only revealed the presence of silicon-based contaminants and no phosphorus compounds were evident.

1 NMR analysis showed only the presence of diethyl ether: H NMR, –CH3, δ 1.15 ppm (t, 3) and –

13 31 CH2, δ 3.44 ppm (q, 4), C NMR –CH3, δ 14.67 ppm and –CH2, δ 65.18 ppm. The P NMR spectrum had no signals, supporting the conclusion that no phosphorus compounds were present. The absence of phosphorus-compounds was attributed to observable issues that arose during preparation. There were notable solubility issues with the methylphosphonyl dichloride and dry diethyl ether, which resulted in most of the presumed product becoming suspended in the

Pasteur pipette. An additional issue arose as a result of the low boiling point of the solvent. The diethyl ether almost entirely evaporated during the preparation process. For subsequent preparations dry chloroform was used as an alternative solvent to circumvent such issues.

The second trial (Table 5, line 2) substituted dry chloroform for diethyl ether as the reaction solvent. GC-MS analysis of the product mixture revealed a complex chromatogram with a peak at retention time 7.08 minutes consistent with the presence of diethyl methylphosphonate

(structure shown in page 56, Table 6) (91 % NIST library database similarity). 31P and 1H NMR spectra were also consistent with a mixture of products. The 31P NMR spectrum displayed two signals, indicating the possible presence of two phosphorus-containing compounds.

The 1H NMR spectrum (shown in Figure 9) was the most informative and indicated the presence of acetic acid, with a signal at δ 1.99 (s, 1) ppm, and other methyl group proton signals overlapping at δ 1.46 (d, 2) ppm and δ 1.44 (d, 2) ppm. The chemical shift of these methyl group proton signals was inconsistent with the observed 1H NMR spectrum of MPA, indicating conversion of this starting material to products. The phosphorus compounds present could be a mixture of both diethyl methylphosphonate, identified by GC-MS, and the mono ester ethyl methylphosphonate (shown in Figure 8), based on the following observations. A signal at δ 8.05

(s, 1) ppm was assigned to the –OH group proton of ethyl methylphosphonate (mono ester).

40

O P O

OH

Figure 8. Mono ethyl ester of MPA (ethyl methylphosphonate).

The mono and diester would have the distinctive triplet and quartet signals for ethyl group –CH3 and –CH2 groups respectively. The presence of clearly overlapping triplet (δ 1.12-1.27 (t, 3) ppm) and quartet (δ 4.05-4.35 (q, 4) ppm) signals in the spectrum (Figure 9) is consistent with a mixture of componds containing ethyl groups.

Figure 9. 1H NMR spectrum and expansions for the product mixture from trial 2.

41

It was deemed important to corroborate these conclusions with a 1H NMR spectrum of pure diethyl methylphosphonate. Due to lag times with shipping and time constraints, a reference compound could not be purchased. For this reason, one was synthesised, and the GC retention time, MS and NMR data were used for comparison to those of the product mixture discussed here. Synthesis of the reference compound will be discussed in further detail in Chapter 3.

Results and discussion, Section 3.4. Preparation of dialkyl methylphosphonates.

Several issues arose during this synthesis and analysis. The GC-MS chromatogram had several peaks consistent with the presence of silica-based compounds. The NIST library matches ranging between 74-76 % were insufficient to identify any compounds. However, it was hypothesised that the silica compounds could have come from several sources. It is reported that such compounds may result from column bleed, silica-based lubricants, the septa of vials, or a contaminant in contact with the sample (126). Several measures were taken to address this issue.

A blank sample of the reaction solvent, chloroform dried over molecular sieves, was analysed by

GC-MS and similarly showed a range of silica-based contaminants. Herold and Mokhatab (2017)

(127) report that molecular sieves are typically composed of zeolite and a binder material.

Although molecular sieves are designed to withstand being subjected to high temperatures, as often required for activation, heating sieves may lead to some degradation of the binder material. Avoidance of dust formation resulting from surface abrasions is largely dependent on the thermal stability of its chemical components and the efficiency of the binding agent (128).

To remove the molecular sieves as a possible cause of the silica-based contaminants, chloroform was dried via an alternate method (dried over potassium carbonate and distilled) which is detailed in 2.1.3 Solvent drying procedures. GC-MS analysis showed reduced frequency of appearance, but not the intensity of contaminant peaks eluting. At the high temperatures used in GC, acidic compounds (126), are known to result in degradation of GC columns, increasing

42 column bleed. The pH of the carbonate-dried chloroform was approximately 6.5, indicating it was not a likely source of acidity for column degradation.

To further address the issue, approximately 25 cm was removed from both ends of the GC column as ends of the column tend to degrade first (126, 129). This drastically reduced the appearance of silica-based contamination when pure chloroform was analysed.

Another aspect to be addressed was the unexpected but apparent formation of diethyl methylphosphonate during the trialled synthesis of diacetyl methylphosphonate. At some point in the synthesis the carbonyl group of the acetate reactant or acetylated product was reduced to a –CH2 group. To rule out the unlikely possibility of silver ions or contaminants in the silver acetate from its preparation playing a role, sodium acetate was used in the place of silver acetate in trial three.

In the third trial (Table 5, line 3) to prepare diacetyl methylphosphonate, the GC chromatogram of the product (Figure 10) showed a peak at retention time of 7.18 minutes, with a NIST library match of 92 % to the mass spectrum of diethyl methylphosphonate. The repeated observation of this product meant that silver ions, and any contaminants that may have been present in the batch prepared, did not contribute to reduction of the carbonyl group. An additional peak at a retention time of 11.45 minutes, had a NIST library match of 90 % to the mass spectrum of diethyl dimethylpyrophosphonate, structure shown in Table 6. The chromatogram showed an additional peak at 8.58 minutes with low (78 %) similarity to the mass spectrum of sulfurous acid diethyl ester. Even though the similarity was low, this compound could have formed from thionyl chloride in a side reaction.

Figure 10. GC-MS chromatogram of the trial 3 reaction product. 43

The 31P NMR spectrum of the reaction product indicated the possible presence of two phosphorus-containing compounds, as indicated by the presence of two signals. As these chemical shifts were not indicative of MPA, it was concluded that two phosphorus-containing products had formed, consistent with the GC-MS data.

1H NMR spectrum had a prominent signal at δ 1.96 (s, 1) ppm, indicative of acetic acid and attributed to the protonation of sodium acetate during the synthesis. Additional signals at δ

1.41-1.61 (d, 2) ppm were not indicative of the methyl group protons of MPA. The phosphorus compounds present could be a mixture of both diethyl methylphosphonate and diethyl diemethylpyrophosphonate, identified by GC-MS. It was clear that the target compound had not formed and the carbonyl group of reactant acetate or acetylated product had again been reduced.

Commercially sourced thionyl chloride can contain several impurities, that may not be readily removed via rotary evaporation. To evaluate if these may have contributed to the reduction of the acetyl groups to an alcohol, subsequent preparations purified methylphosphonyl dichloride by reduced pressure distillation.

Trial four (Table 5, line 4) utilised purified methylphosphonyl dichloride and silver acetate. GC-

MS analysis of the product showed only two peaks in the chromatogram at retention times of

7.10 and 11.42 minutes, consistent with diethyl methylphosphonate (92 % NIST library similarity match) and diethyl dimethylpyrophosphonate (91 % NIST library similarity match), respectively.

However, the 31P NMR spectrum showed presence of three phosphorus-containing compounds, indicated by the presence of three signals, it was concluded that three phosphorus compounds had formed. The mixture could be diethyl methylphosphonate, diethyl dimethyldiphosphonate, consistent with the GC-MS, and a monoester ethyl methylphosphonate.

The 1H NMR spectrum had a signal at δ 8.02 (s, 1) ppm consistent with the –OH group proton of the monoester. Overlapping signals at δ 1.47 (d, 2) ppm, were inconsistent with MPA. Ethyl

44 group proton signals from δ 3.62-4.08 (q, 4) ppm (–CH2) and signals at δ 1.12-1.28 ppm (t, 3) (–

CH3) indicate the presence of ethyl derivatives.

As diethyl methylphosphonate appeared to be forming, it was clear that a reducing agent must have been present in the reaction mixture. It was beyond the scope of this project to determine the reducing agent and this will be further discussed in Chapter 4. Future direction and concluding statements.

3.2 Introduction to dialkyl methylphosphonates

Attempts to prepare diacetyl methylphosphonate appeared to produce diethyl methylphosphonate instead. It was decided that a reference sample of diethyl methylphosphonate would be prepared to confirm this finding. This section introduces preparation of dialkyl methylphosphonates, which are important derivatives in the forensic pursuit of CWNAs in themselves. Their preparation often requires the use of expensive reagents and harsh experimental conditions. There are several dialkyl methylphosphonates, of which dimethyl methylphosphonate is the most commonly employed. Dimethyl methylphosphonate, as it relates CWNA analysis is outlined in Chapter 1. Introduction, Section 1.4.2 Derivatisation of phosphonic acids. Less common dialkyl derivates of MPA include diethyl methylphosphonate, diisopropyl methylphosphonate, and dibutyl methylphosphonate.

Several aspects were assessed in the consideration of viable preparation methods for diethyl methylphosphonate:

1) Utilise MPA as a starting material.

2) Accessibility and cost of reagents.

3) Minimisation of hazards.

An extensive search of the literature was required to determine a viable preparation method for a diethyl methylphosphonate, with limited reports on the use of MPA as a staring material.

45

Existing synthetic methods involve the use of diazoethane (130) or alcohols (131). Diazoethane is not only highly toxic, but potentially explosive, use of this reagent was excluded due to its hazardous nature. Alcohols were surmised as a viable derivatisation reagent. Various methods for trialling the preparation of dialkyl methylphosphonate with the use of readily available and inexpensive alcohols will be discussed below.

Esterification of carboxylic acids with alcohols in the presence of an acid catalyst is a common reaction (132) (shown in Figure 11). Concentrated acid acts as a catalyst, serving a dual purpose protonating the C=O oxygen making it a better electrophile. The acid additionally acts as a dehydrating agent forcing the equilibrium forward, allowing for the loss of H2O as a leaving group. There are no precedents for esterification of MPA in this way, but acid catalysed esterification of carboxylic acids could be applicable to phosphonic acids.

O Acid (H+) O

+ R'OH + H2O

OH OR'

Figure 11. Acid-catalysed esterification of carboxylic acids Zimmermann and Rudolph, (1965) (133).

Sathe et al. (2006) (131) report the use of silica chloride as a heterogenous catalyst to convert

MPA to dialkyl methylphosphonates using various primary and secondary alcohols at room temperature (20-25 °C). Silica chloride is generated by the method outlined in Chapter 2.

Experimental, Section 2.1.6 Preparation of silica chloride. Esterification reportedly occurred rapidly forming the desired phosphonates within 20-25 minutes for primary alcohols and within

40 minutes for secondary alcohols. This method resulted in successful preparation of dialkyl

MPA methyl, ethyl, isopropyl and butyl esters Shown Figure 12.

O O

SiO2-Cl + R'OH HO P OH OR' P OR'

Figure 12.Synthesis of alkyl methylphosphonates using silica chloride (131).

46

Another method that could be trialled is by using a coupling reagent. DCC is a common coupling reagent for synthesis of esters from carboxylic acids and primary and secondary alcohols (134).

DCC activates the carboxylic acid by converting the –OH group to a better leaving group, allowing the reaction to proceed under more mild conditions (134-136). It is theorised that esterification of MPA may proceed under similar conditions although there is currently no literature reference to DCC as a coupling reagent for dialkylation of MPA.

Another method that could be trialled is substitution of the –OH groups functional groups through chlorination. This had been previously employed in this project and could be further probed for preparation of MPA dialkyl esters under the reaction scheme shown in Figure 13.

O O O SOCl R'OH P OH 2 P Cl P OR'

. OH Cl OR'

Figure 13. Synthesis of dialkyl methylphosphonates using methylphosphonyl dichloride at ambient temperature. Where R’ represents an alcohol group.

3.3 Revised research aims

Multiple trials to prepare diacetyl methylphosphonate were undertaken throughout the course of this project. Reduction of the acetate or acetylated product carbonyl groups to –CH2 groups was occurring during the synthesis. As it was not viable within the timeframe of this project to source commercially available diethyl methylphosphonate, a reference sample would be synthesised. The GC-MS and NMR data for such a sample would aid in the analysis of reaction products from the trialled preparation of diacetyl methylphosphonate and confirm reduction of the carbonyl group.

47

Due to the limited success of the trialled diacetyl methylphosphonate synthesis the initially proposed research aims could not be addressed (aims outlined in Chapter 1. Introduction,

Section 1.6. Concluding statements and research aims). In the pursuit of a reference compound, synthesis methods could also be assessed for their wider applicability to the rapid and inexpensive derivatisation of MPA

The dialkyl methylphosphonate preparation methods outlined in Chapter 3. Results and discussion, Section 3.2 Introduction to dialkyl methylphosphonates, move away from water tolerant derivatisation procedures, but still focus on relatively non-hazardous and inexpensive reagents. It is posed that such methods may provide a viable alternative to current CWNA methyl and silyl derivatisation methods, moving towards rapid and economic methods with infield potential.

The overarching aim was to investigate alternative procedures for the derivatisation of MPA, to further probe the aim of this project the initial aims were revised. This can be divided into two explicit aims:

1. Assess methods that utilise conversion of methylphosphonyl dichloride to

dialkyl methylphosphonates using readily available alcohols.

2. Assess potential esterification methods for MPA using catalysts or coupling

reagents.

Herein, the preparation of various dialkyl methylphosphonates will be reported. A comparison of the potential alkyl methylphosphonate products observed in trialled preparation of diacetyl methylphosphonate will be discussed. Additionally, this method will be assessed as a rapid and economic MPA derivatisation method with in-field potential.

48

3.4 Preparation of dialkyl methylphosphonates

This section will focus on the respective outcomes of trialling methanol, ethanol, propan-2-ol, butan-1-ol and benzyl alcohol as derivatization reagents for MPA.

3.4.1 Diethyl methylphosphonate

This preparation involved conversion of MPA to methylphosphonyl dichloride and subsequent treatment with ethanol. Two peaks were observed in the GC-MS chromatogram of the product, consistent with the presence of diethyl methylphosphonate (retention time 7.08 minutes) and diethyl dimethylpyrophosphonate (retention time 11.33 minutes). The peaks were identified by similarity search of their mass spectra, the appearance of which will be discussed below. These compounds were previously identified as reaction products in the attempted preparation of diacetyl methylphosphonate (Section 3.1 Trialled preparation of diacetyl methylphosphonate).

MS fragmentation patterns of experimental spectra were compared to the NIST library (2005) and the ROPs by Vanninen (2017) (7). The ROPs, as explained in Chapter 1. Introduction, underpin analysis of chemical warfare agents, for this reason it was used for comparative purposes to a determine and assign characteristic dialkyl methylphosphonate mass spectral ions to experimental spectra. It is important to address that not all fragment ion signals have been assigned and some signals have been assigned to fragments with aid of computerised conceptual models and have not be experimentally studied via high resolution MS. The NIST library similarly matches for the peak assigned as diethyl methylphosphonate in all preparations where it was detected ranged between 91-98 %. The most abundant fragment ion signals and their relative intensities for mass spectra of peaks identified as diethyl methylphosphonate

+ (peak retention time 7.08-7.20 minutes) are as follows, the base peak ion CH5O3P 97 m/z.(Relative Intensity [RI] 100 %) from the loss of two ethyl groups. A small molecular ion signal

+ + C5H13O3P 152 m/z (RI 2-3 %), PO3C3H10 125 m/z (RI 68-70 %), 108 m/z (RI 21-22 %), 80 m/z (RI

49

+ 39-43 %), PO2CH4 79 m/z (RI 99 %), 65 m/z (RI 19-24 %). Variation in RI of fragment signals between samples is within the acceptable range for ROPs chemical warfare analysis (±20 %) (7).

Boateng et al (2019) (137) outline common fragmentation patterns of diethyl methylphosphonate, explaining that fragmentation occurs along a sequential dissociation pathway beginning from the parent ion: m/z = 152 → 125 → 97 → 79 (shown in Figure 14(a)), with the first two steps involving McLafferty rearrangements (138) (shown in Figure 14(b)), where two hydrogen atoms migrate in the first step, resulting in simultaneous loss of C2H3,

+ forming PO3C3H10 . In the second step, a single hydrogen atom migration results in subsequent loss of C2H4, which forms the PO3CH6 ion, and can subsequently lose H2O to form PO2CH4

(a)

(b)

Figure 14. GC-MS EI+ fragmentation of diethyl methylphosphonate: (a) Common fragment ions of diethyl methylphosphonate⋅+, (b) McLafferty rearrangement mechanism in diethyl methylphosphonate. Schemes from Boateng et al.l(2019) (137).

The NIST library database similarity match for the peak identified as diethyl dimethylpyrophosphonate (retention time 11.33-11.42 minutes) ranged between 90-92 %.The

+ most abundant fragment ion signals are as follows, base signal PO2CH4 79 m/z (RI 100 %),

+ C2H8P2O6 203 m/z (RI 11-14 %) resulting from the loss of two ethyl groups with hydrogen

+ rearrangement, 175 m/z (RI 16-17 %), 157 m/z (RI 21-23 %), 143 m/z (RI 33-34 %), CH5O3P 97 m/z (RI 59-60 %). No signal presented for the molecular ion.

50

While GC-MS analysis indicated the formation of two phosphorus compounds, 31P NMR indicated the presence of three phosphorus species with display of three signals.

The 1H NMR spectrum showed distinctive overlapping signals for the phosphorus-bonded methyl group proton (δ 1.50-δ 1.51 (d, 2) ppm). As none of the chemical shifts were consistent with unreacted MPA, it was concluded that these signals belonged to phosphorus-containing products. Overlapping ethyl group signals at δ 1.13-1.28 (t, 3) (–CH3) and δ 3.39-4.01 ppm (q, 4)

(–CH2) were consistent with the formation of ethyl derivatives, shown in Figure 15. Additionally, the chemical shifts for these ethyl group signals did not correlate to those of ethanol, the other reactant and solvent for this process.

This was taken as an indication that ethanol was not present in a detectable amount. The signal at δ 8.60 (s, 1) ppm was attributed to the –OH group proton of the MPA monoester, ethyl methylphosphonate (structure shown in Figure 8). This compound was not detected via GC-MS, as it typically too polar and requires derivatisation prior to analysis (139). The remaining signals were consistent with the presence of diethyl methylphosphonate and diethyl dimethylpyrophosphonate, which were detected by GC-MS.

51

Figure 15..1H NMR spectrum of products formed in the reaction of methylphosphonyl dichloride and ethanol.

52

It is important to restate that the aim of this experiment was to obtain a reference sample of diethyl methylphosphonate to confirm its formation in the attempted preparation of diacetyl methylphosphonate. This was done in lieu of the time required to take delivery of commercially available diethyl methylphosphonate. It was beyond the scope of this project to isolate and fully characterise the three products formed in the experiment described here. The data on the product mixture obtained are consistent with the presence of the three phosphorus-containing compounds described. Comparison of signals in the NMR spectra from this experiment and those of the product from attempted preparation of diacetyl methylphosphonate show a similar profile. This supports the conclusion that the carbonyl group of acetate or acetylated products was reduced in the attempted preparation of diacetyl methylphosphonate.

As diethyl methylphosphonate formed with relative ease in this experiment, it was developed as a potential method for rapid and inexpensive derivatisation of MPA. Various other alcohols were trialled, and the results of these preparations will be discussed herein.

3.4.2 Dimethyl methylphosphonate

Dimethyl methylphosphonate (Table 6) was prepared using the same process described above, with methanol instead of ethanol added to methylphosphonyl dichloride. GC-MS analysis of the product identified a single peak at retention time 5.21 minutes. The mass spectrum of this peak had 98 % similarity to the NIST database spectrum of dimethyl methylphosphonate. A signal with 94 m/z (RI 100 %) is attributed to a spontaneous series of 1,4-hydrogen atom shifts, which is documented to occur during EI fragmentation of dimethyl methylphosphonate (140). This results in isomerization to its enol form, the intermediate product of this fragmentation reaction

+ can dissociate to the ion PO2C2H7 94 m/z via 1,4-hydrogen rearrangement and low energy loss

+ of formaldehyde (CH2O) (140). The second high intensity signal for fragment PO2CH4 with 79 m/z (RI 99 %) results from the loss of a methyl radical (CH3) and CH2O, without any additional rearrangement (141). Holtzclaw et al. (1982) (141) report these losses, suggesting that such

53 fragmentation patterns imply the presence of a phosphorus bound hydrogen acceptor site. This facilitates hydrogen migration required for the loss of CH2O, as the phosphorus site is already occupied, and the presence of a methoxy group in the 79 m/z fragment. The high intensity of signals for 94 m/z and 79 m/z suggests these fragments are the most energetically favoured

+ products relative to the parent molecular ion. A small molecular ion signal for C3H9O3P with 124

+ + m/z (RI 10 %), and additional signals for fragments PO3(CH3)2 109 m/z (RI 30 %), PO2(CH3)2 93

+ + m/z (RI 29 %), PO2 63 m/z (RI 20 %), PO 47 m/z (RI 21 %) were also observed.

1 NMR analysis of this product was carried out on both acetone-d6 and chloroform-d solutions. H

NMR of the acetone-d6 solution was analysed for comparison to the experimentally obtained

MPA spectrum, through which it was determined that no signals were consistent with the presence of unreacted MPA. 1H NMR chemical shifts for dimethyl methylphosphonate in chloroform-d have been reported by Wang et al. (2013) (142). The chloroform-d 1H NMR spectrum of the product showed signals at δ 3.84 (d, 2) ppm and δ 1.57 (d, 2) ppm which can be attributed to the H3C-O-P and P-CH3 group protons respectively, are consistent with those reported previously.

Dimethyl methylphosphonate, is the most common derivative of MPA employed in OPCW testing. The recommended operating procedures for preparation of this important derivative involve the use of dangerous reagents (see Chapter 1. Introduction, Section 1.4.2 Derivatisation of phosphonic acids). Formation of this derivative using readily available and much less hazardous reagents is an important finding of this project. Various lines of enquiry could be further pursued to optimise this method for in-field application, and these will be discussed further in Chapter 4. Future direction and concluding statements.

54

3.4.3 Other dialkyl methylphosphonates alcohols

Carrying out this procedure using propan-2-ol and butan-1-ol gave the corresponding dialkyl methylphosphonate derivatives with successful preparation probed using GC-MS and NMR as described above for ethyl and methyl derivatisation. Diisopropyl methylphosphonate and dibutyl methylphosphonate presented at GC-retention times of 7.77 minutes, 11.16 minutes, respectively. In the cases of methanol, propan-2-ol and butan-1-ol, the dialkyl methylphosphonate was the only product observed by GC-MS with formation of a dialkyl dimethylpyrophosphate only observed when using ethanol. Using benzyl alcohol in the process did not result in the formation of a derivative, presumably due to its bulkiness compared to the other alcohols tried. No further experiments were conducted with benzyl alcohol.

3.4.6 Summary/ rate of reaction

Product formation of all the trialled primary and secondary alcohols was assessed via GC-MS, to determine the time at which the corresponding diesters could be detected. Methylphosphonyl dichloride was converted via the use of methanol, ethanol, and butan-1-ol to corresponding diesters within 10 minutes at room temperature (20-25 °C) shown in Table 6. While diisopropyl methylphosphonate formed within 10-20 minutes, the slower rate is likely to be attributed to how crowded the secondary propan-2-ol is in comparison to the other straight chain alcohols.

Detection of diesters after 10-20 minutes of reaction time is another key factor in the in-field application of MPA derivatisation.

55

Table 6. First sign of product formation as determined via GC-MS, names and structures of MPA derivatives detected.

Reagent Room Temp Derivative Structure (20-25 °C) Time in Minutes methanol ≥10 dimethyl methylphosphonate O P O

O ethanol ≥10 diethyl methylphosphonate O P O O

ethanol ≥10 diethyl dimethylpyrophosphonate O O P O P O O

propan-2-ol 10-20 diisopropyl methylphosphonate O P O O

butan-1-ol ≥10 dibutal methylphosphonate O P O O

3.4.7 Trialled esterification using catalyst and coupling reagent.

Various experimental methods that utilised methanol or ethanol with a coupling agent or a catalyst were attempted. Preparations include the use of DCC, sulfuric acid in a catalytic amount, and silica chloride.

Acetonitrile was used as a solvent for DCC coupling, as both methanol and ethanol are miscible, but the by-product dicyclohexyl urea is less soluble than in chloroform (143). Residual urea may be removed by dissolving the crude product in a small volume of ethyl acetate, followed by

56 refrigeration, to precipitate out the rest of the urea (143). However, this step was not deemed necessary as the by-product urea does not have to be removed to check that the target derivative has formed using analytical techniques. Analysis of both methanol and ethanol-based preparations did not indicate the formation of a GC-MS detectable derivative. A small amount of dialkyl methylphosphonate derivative may have formed, but the amount present may be below the detection limit. It is also possible that this method did not form an ester of MPA as

DCC may simply not have been an efficient enough coupling reagent.

A simple, sulfuric acid catalysed Fischer-type esterification of MPA by methanol and ethanol was attempted. When analysed via GC-MS no phosphorus species were detected in the reaction product, suggesting that these conditions were not suitable. Crenshaw and Cummings (2004)

(144) reported on the phenylarsonic acid catalysed preparation of various alkyl mono MPA esters. However, this preparation includes a complicated work-up with multiple alcohols, heating under dry nitrogen for 28 hours with continuous azeotropic removal of water. A more tedious variation of this work up could be trialled as an alternative for derivatisation of MPA, but its in-field applications are incredibly limited.

Using silica chloride as a heterogenous catalyst for MPA dialkyl derivatisation was not successful as indicated by GC-MS analysis of the product. A low confidence similarity match to the NIST library presented for the mass spectrum of the peak with retention time 9.49 minutes, may be attributed to a phosphorus-compound which eluted with a silica-based contaminant or column bleed. Analysis by 1H NMR and 13C NMR presented signals consistent with unreacted ethanol.

31P NMR indicated the presence of two phosphorus species, but these signals could not be assigned. Further exploration into this line of inquiry is required to make conclusive statements.

The limitations of this approach and possible modifications will be discussed in the subsequent chapter.

57

Chapter 4. Future direction and concluding statements

MPA is the final hydrolysis product of several nerve agents and an important analyte for monitoring the stockpiling or use of CWNAs. The overarching aim of this project was to develop a viable, economic, rapid, and relatively non-hazardous derivatisation method with infield potential for the detection of MPA by GC-MS. Several derivatisation methods for MPA were probed throughout this project with variable success.

Research initially focused on acetylation as a derivatisation method. Several trials were undertaken to prepare a reference sample of diacetyl methylphosphonate. This was unsuccessful, with reduction of the acetyl carbonyl group during synthesis and detection of diethyl methylphosphonate instead.

The project aims were subsequently redirected to esterification of MPA with methanol or ethanol. Trialled preparations utilised silica chloride or sulfuric acid as catalysts, and DCC as a coupling reagent. These methods were also unsuccessful. The use of silica chloride as a reagent for dialkyl derivatisation of MPA could be probed in future work. This reagent could facilitate a viable MPA derivatisation technique if the preparation of silica chloride was verified. In this work a literature preparation for silica chloride was employed, but it was used without characterisation.

Various dialkyl methylphosphonates were successfully prepared by conversion of MPA to methylphosphonyl dichloride using thionyl chloride, and subsequent reaction with alcohols.

Conversion of methylphosphonyl dichloride to corresponding dialkyl derivatives occured rapidly at room temperature (20-25 °C) with derivatives being detected via GC-MS, within 10 minutes for primary alcohols (methanol, ethanol, and butan-1-ol) and within 20 minutes for secondary alcohols (propan-2-ol), while benzyl alcohol did not react.

Dialkyl methylphosphonates have not previously been prepared via this method, which offers an exciting alternate derivatisation method for MPA. Successful preparation of the dimethyl derivative using methanol is arguably the most significant outcome of this project. Dimethyl 58 methylphosphonate is the most common derivative utilised in CWNA infield analysis. The current favoured dimethyl esterification method employs the use of diazomethane which is an expensive, highly toxic and explosive reagent. Using the method described here, dimethyl methylphosphonate formed rapidly and the corresponding pyrophosphate by product was not detected.

In the reaction between methylphosphonyl dichloride and ethanol, the reaction proceeded rapidly to form diethyl methylphosphonate within 10 minutes, but formation of diethyl dimethylpyrophosphonate was favoured as time progressed.

Commercially available reference samples were not analysed because they could not be obtained in the timeframe of the project. Reference compounds should be utilised for comparison in future work. Sourcing commercially available reference compounds would enable identification of compounds in the NMR spectra of product mixtures and allow for a full quantitative study to be undertaken. Product mixtures were not separated in this project due to the focus on derivatisation method development rather than characterisation, and no novel compounds were prepared. Any novel compounds prepared in future work should be isolated, purified and fully characterised.

Future work could focus on optimisation of the successful dialkyl methylphosphonate preparation. MPA was converted to methylphosphonyl dichloride by thionyl chloride through a

3-hour reflux period. Shorter time periods could be trialled for effectiveness. Furthermore, a distillation step was carried out to remove the excess thionyl chloride. This was done to compensate for possible inefficiencies of the rotary evaporator system. Removal of excess thionyl chloride could be trialled via an alternative method (e.g. high vacuum) to bypass this step. Additionally, an alternative chlorinating agent, such as oxalyl chloride could be trialled.

Oxalyl chloride is often employed in the place of thionyl chloride as it produces by-products that are readily removed via evaporation and it does not require prolonged reflux periods (145).

59

The viability of synthesising a diacetyl methylphosphonate reference sample could also be further probed though use of alternative chlorination reagents. This may give a greater insight into the possible impurities in the thionyl chloride reducing the acetyl group.

In conclusion, the dialkyl methylphosphonate preparation method applied here provides a promising platform for future exploration. Alkylation of methylphosphonic acid with alcohols provides an inexpensive and relatively non-hazardous alternative to current widely accepted derivatisation methods in supporting the rapid in-field detection of chemical warfare nerve agents.

60

References

1. Guidotti M, Trifirò F. Chemical risk and chemical warfare agents: science and technology against humankind. Toxicological & Environmental Chemistry. 2016;98(9):1018-25. 2. Richardson D, Caruso J. Derivatization of organophosphorus nerve agent degradation products for gas chromatography with ICPMS and TOF-MS detection. Analytical and Bioanalytical Chemistry. 2007;388(4):809-23. 3. Popiel S, Sankowska M. Determination of chemical warfare agents and related compounds in environmental samples by solid-phase microextraction with gas chromatography. Journal of chromatography. 2011;1218(47):8457. 4. Valdez C, Marchioretto M, Leif R, Hok S. Efficient derivatization of methylphosphonic and aminoethylsulfonic acids related to nerve agents simultaneously in soils using trimethyloxonium tetrafluoroborate for their enhanced, qualitative detection and identification by EI-GC–MS and GC–FPD. Forensic Science International. 2018;288:159-68. 5. Black RM, Muir B. Derivatisation reactions in the chromatographic analysis of chemical warfare agents and their degradation products. Netherlands: Elsevier B.V; 2003. p. 253-81. 6. Valdez C, Leif R, Alcaraz A. Effective methylation of phosphonic acids related to chemical warfare agents mediated by trimethyloxonium tetrafluoroborate for their qualitative detection and identification by gas chromatography-mass spectrometry. Analytica Chimica Acta. 2016;933:134-43. 7. Vanninen P. Recommended operating procedures for analysis in the verification of chemical disarmament. University of Helsinki, Helsinki. 2017. 8. Chauhan S, Chauhan S, D’Cruz R, Faruqi S, Singh K, Varma S, et al. Chemical warfare agents. Environmental Toxicology and Pharmacology. 2008;26(2):113-22. 9. Organization for the Prohibition of Chemical Weapons, Chemical Weapon Convention [cited 26 Feb 2019]. Available from: www.opcw.org. 10. Convention on the prohibition of the development, production, stockpiling and use of chemical weapons and on their destruction. Signed in Janurary 1993, entered into force 29 April 1997. [cited 5 March 2019]. Available from: www.opcw.org 11. Okumura T, Takasu N, Ishimatsu S, Miyanoki S, Mitsuhashi A, Kumada K, et al. Report on 640 Victims of the Tokyo Subway Sarin Attack. Annals of Emergency Medicine. 1996;28(2):129-35. 12. Nakagawa T, Tu A. Murders with VX: Aum Shinrikyo in Japan and the assassination of Kim Jong-Nam in Malaysia. Forensic Toxicology. 2018;36(2):542-4. 13. Kloske M, Witkiewicz Z. Novichoks – The A group of organophosphorus chemical warfare agents. Chemosphere. 2019;221:672-82. 14. Ganesan K, Raza S, Vijayaraghavan R. Chemical warfare agents. Journal of pharmacy and bioallied sciences. 2010;2(3):166. 15. Munro NB, Talmage SS, Griffin GD, Waters LC, Watson AP, King JF, et al. The Sources, Fate, and Toxicity of Chemical Warfare Agent Degradation Products. Environmental Health Perspectives. 1999;107(12):933-74. 16. Toxicity of Chemical Warfare Agents and their Degradation Products. Chichester, UK: John Wiley & Sons, Ltd; 2011. p. 19-57. 17. Figueiredo T, Apland J, Braga M, Marini A. Acute and long‐term consequences of exposure to organophosphate nerve agents in humans. Epilepsia. 2018;59(S2):92-9. 18. López-Muñoz F, García-García P, Alamo C. The pharmaceutical industry and the German National Socialist Regime: I.G. Farben and pharmacological research. Journal of Clinical Pharmacy and Therapeutics. 2009;34(1):67-77. 19. Lopez-Munoz F, Alamo C, Guerra J, Garcia-Garcia P. The development of neurotoxic agents as chemical weapons during the National Socialist period in Germany. Revista De Neurologia 2008;47(2):99-106.

61

20. Lopez-Munoz F, Alamo C, Guerra JA, Garcia-Garcia P. The development of neurotoxic agents as chemical weapons during the National Socialist period in Germany. REVISTA DE Neurologia. 2008;47(2):99-106. 21. Szinicz L. History of chemical and biological warfare agents. Toxicology. 2005;214(3):167-81. 22. Vásárhelyi G, Földi L. History of Russia’s chemical weapons. Academic and applied research in military science 2007;6(1):135-46. . 23. Bajgar J. Prophylaxis against organophosphorus poisoning. J Med Chem Def. 2004;1:1- 16. 24. Balali-Mood M, Saber H. Recent advances in the treatment of organophosphorous poisonings. Iranian Journal of Medical Sciences. 2012;37(2):74-91. 25. Haines D, Fox S. Acute and long-term impact of chemical weapons: Lessons from the Iran-Iraq war. Forensic Science Review. 2014;26(2):97-114. 26. Pita R, Domingo J. The use of chemical weapons in the Syrian conflict. Toxics. 2014;2(3):391-402. 27. Government assessment of the syrian governments use chemical weapons [Internet]. White House. [cited 16 May 2019]. Available from: www.whitehouse.gov. 28. Chairman of the Joint Intelligence Committee. Syria: Reported chemical weapon use, 29 August 2013. [cited 15 May 2019]. Available from: www.gov.uk. 29. Chauhan S, Chauhan S, D’Cruz R, Faruqi S, Singh KK, Varma S, et al. Chemical warfare agents. Environmental Toxicology and Pharmacology. 2008;26(2):113-22. 30. Taylor P. Anticholinesterase agents. Goodman & Gilman’s The Pharmacological Basis of Therapeutics Twelfth edition New York, Macmillan. 2011:239-54. 31. Protocols. IoMCotEotDoVAUCA. Adequacy of the Comprehensive Clinical Evaluation Program: Nerve Agents. National Academies Press; 2000. p. 97-104. 32. Raber E, Jin A, Noonan K, McGuire R, Kirvel R. Decontamination issues for chemical and biological warfare agents: how clean is clean enough? International journal of environmental health research. 2001;11(2):128-48. 33. Popiel S, Sankowska M. Determination of chemical warfare agents and related compounds in environmental samples by solid-phase microextraction with gas chromatography. Journal of chromatography. 2011;1218(47):8457. 34. Willison S. Investigation of the persistence of nerve agent degradation analytes on surfaces through wipe sampling and detection with ultrahigh performance liquid chromatography-tandem mass spectrometry. Analytical chemistry. 2015;87(2):1034-41. 35. CA. V, Leif Roald N, Hok S, Hart Bradley R. Analysis of chemical warfare agents by gas chromatography-mass spectrometry: methods for their direct detection and derivatization approaches for the analysis of their degradation products. Reviews in Analytical Chemistry2018. 36. Richardson DD, Caruso JA. Derivatization of organophosphorus nerve agent degradation products for gas chromatography with ICPMS and TOF-MS detection. Analytical and Bioanalytical Chemistry. 2007;388(4):809-23. 37. Organisation for the Prohibition of Chemical Weapons [Internet]. OPCW. [cited 2 June 2019]. Available from: www.opcw.org. 38. Dubey V, Velikeloth S, Sliwakowski M, Mallard G. Official proficiency tests of the organisation for the prohibition of chemical weapons: current status and future directions. Accreditation and Quality Assurance. 2009;14(8-9):431-7. 39. Sega G, Tomkins B, Griest W. Analysis of methylphosphonic acid, ethyl methylphosphonic acid and isopropyl methylphosphonic acid at low microgram per liter levels in groundwater. Journal of Chromatography A. 1997;790(1):143-52. 40. Rohrbaugh DK, Sarver EW. Detection of alkyl methylphosphonic acids in complex matrices by gas chromatography–tandem mass spectrometry. Journal of Chromatography A. 1998;809(1):141-50.

62

41. Kataoka M, Tsuge K, Seto Y. Efficiency of pretreatment of aqueous samples using a macroporous strong anion-exchange resin on the determination of nerve gas hydrolysis products by gas chromatography-mass spectrometry after tert-butyldimethylsilylation. Journal of Chromatography A. 2000;891(2):295-304. 42. Tomkins B, Sega G. Determination of thiodiglycol in groundwater using solid-phase extraction followed by gas chromatography with mass spectrometric detection in the selected- ion mode. Journal of Chromatography A. 2001;911(1):85-96. 43. Weissberg A, Madmon M, Elgarisi M, Dagan S. Aqueous extraction followed by derivatization and liquid chromatography–mass spectrometry analysis: A unique strategy for trace detection and identification of G-nerve agents in environmental matrices. Journal of Chromatography A. 2018;1577:24-30. 44. Black RM, Clarke RJ, Read RW, Reid MT. Application of gas chromatography-mass spectrometry and gas chromatography-tandem mass spectrometry to the analysis of chemical warfare samples, found to contain residues of the nerve agent sarin, sulphur mustard and their degradation products. Journal of chromatography A. 1994;662(2):301. 45. D’Agostino PA, Hancock JR, Provost LR. Determination of sarin, soman and their hydrolysis products in soil by packed capillary liquid chromatography–electrospray mass spectrometry. Journal of Chromatography A. 2001;912(2):291-9. 46. Noami M, Kataoka M, Seto Y. Improved tert-butyldimethylsilylation gas chromatographic/mass spectrometric detection of nerve gas hydrolysis products from soils by pretreatment of aqueous alkaline extraction and strong anion-exchange solid-phase extraction. Analytical chemistry. 2002;74(18):4709. 47. Hancock J, McAndless J, Hicken R. Solid Adsorbent Based System for the Sampling and Analysis of Organic Compounds in Air: An Application to Compounds of Chemical Defense Interest. Journal of chromatographic science. 1991;29(1):40-5. 48. Stankov I, Sergeeva A, Sitnikov V, Derevyagina I, Morozova O, Mylova S, et al. Gas Chromatographic Determination of Sulfur Mustard and Lewisite in Community Air. Journal of Analytical Chemistry. 2004;59(5):447-51. 49. Mazurek M, Witkiewicz Z, Popiel S, Śliwakowski M. Capillary gas chromatography– atomic emission spectroscopy–mass spectrometry analysis of sulphur mustard and transformation products in a block recovered from the Baltic Sea. Journal of Chromatography A. 2001;919(1):133-45. 50. D’agostino P, Hancock J, Chenier C, Lepage CJ. Liquid chromatography electrospray tandem mass spectrometric and desorption electrospray ionization tandem mass spectrometric analysis of chemical warfare agents in office media typically collected during a forensic investigation. Journal of Chromatography A. 2006;1110(1-2):86-94. 51. Mesilaakso M. Chemical Weapons Convention chemicals analysis: sample collection, preparation and analytical methods: John Wiley & Sons; 2005. 52. Kumirska J, Plenis A, Łukaszewicz P, Caban M, Migowska N, Białk-Bielińska A, et al. Chemometric optimization of derivatization reactions prior to gas chromatography–mass spectrometry analysis. Journal of Chromatography A. 2013;1296:164-78. 53. Down S. Vladimir Zaikin and John Halket: A handbook of derivatives for mass spectrometry. Berlin/Heidelberg: Springer Berlin Heidelberg; 2014. p. 6453-4. 54. Majer J. Handbook of analytical derivatization reactions : D.R. Knapp, Wiley, Chichester, 1979. PP. 18-741. Talanta. 1980;27(10). 55. Grob RL, Barry EF. Modern practice of gas chromatography. 4th ed. Hoboken, N.J: Wiley-Interscience; 2004. 56. Sparkman O, Penton Z, Kitson F. Gas chromatography and mass spectrometry: a practical guide: Academic Press; 2011. 57. Encyclopedia of Analytical Chemistry, John Wiley & Sons, Ltd, Chichester, 2000, pp. 7042–7076. 58. Fowlis IA, (Ed.), Gas Chromatography, 2nd ed., John Wiley & Sons, Ltd., Chichester, New York, Brisbane, Toronto, Singapore, 1995.

63

59. Orata F. Derivatization reactions and reagents for gas chromatography analysis. Advanced gas chromatography-Progress in agricultural, biomedical and industrial applications: IntechOpen; 2012. 60. Talmage SS, Watson AP, Hauschild V, Munro NB, King J. Chemical warfare agent degradation and decontamination. Current Organic Chemistry. 2007 Feb 1;11(3):285-98. 61. Lee H, Lee H, Sng M, Basheer C. Determination of degradation products of chemical warfare agents in water using hollow fibre-protected liquid-phase microextraction with in-situ derivatisation followed by gas chromatography–mass spectrometry. Journal of Chromatography A. 2007;1148(1):8-15. 62. D'Agostino PA, Provost LR. Determination of chemical warfare agents, their hydrolysis products and related compounds in soil. Journal of Chromatography A. 1992;589(1-2):287-94. 63. Hirsjrvi P. Identification of Degradation Products of Potential Organophosphorus Warfare Agents. An Approach for the Standardization of Techniques and Reference Data. 1981. 64. Kemsley J. Firm Fined For Chemist's Death. American Chemical Society, NW, Washington, DC, 20036, USA.; 2011. 65. Trimethyloxonium Tetrafluoroborate. Encyclopedia of Reagents for Organic Synthesis. 66. Vasilevskii S, Kireev A, Rybal'chenko I, Suvorkin V. Identification of silylated derivatives of alkylphosphonic, alkylthiophosphonic, and dialkylamidophosphoric acids by mass spectrometry. Journal of Analytical Chemistry. 2002;57(6):491-7. 67. Kataoka M, Tsunoda N, Ohta H, Tsuge K, Takesako H, Seto Y. Effect of cation-exchange pretreatment of aqueous soil extracts on the gas chromatographic–mass spectrometric determination of nerve agent hydrolysis products after tert.-butyldimethylsilylation. Journal of Chromatography A. 1998;824(2):211-21. 68. Kataoka M, Tsuge K, Takesako H, Hamazaki T, Seto Y. Effect of pedological characteristics on aqueous soil extraction recovery and tert-butyldimethylsilylation yield for gas chromatography-mass spectrometry of nerve gas hydrolysis products from soils. Environmental science & technology. 2001;35(9):1823. 69. Hennig H, Sartori P. Bis(acyloxy)phosphine oxides. Chem-Ztg. 1983;107(9):257-60. 70. Lazarus R, Benkovic S, Benkovic P. The synthesis and properties in enzymic reactions of substrate analogs containing the methylphosphonyl group. Archives of Biochemistry and Biophysics. 1979;197(1):218-25. 71. Alazawi FN, Al-Janabi KWS, Mohammed MI, Kadhum AAH, Mohamad AB. Direct Acetylation and Determination of Chlorophenols in Aqueous Samples by Gas Chromatography Coupled with an Electron-Capture Detector. Journal of Chromatographic Science. 2012;50(7):564-8. 72. Brown J, Roberts W. Evidence that approximately eighty per cent of the soluble proteins from Ehrlich ascites cells are Nalpha-acetylated. Journal of biological Chemistry. 1976;251(4):1009-14. 73. Arata K, Hino M. The Acetylation of Toluene and Benzene with Acetyl Chloride and Bromide and Acetic Anhydride Catalyzed by Calcined Iron Sulfate Activated by Exposure to a Mixture of Benzyl Chloride and the Aromatics. Bulletin of the Chemical Society of Japan. 1980;53(2):446-9. 74. Golachowski A, Zięba T, Kapelko-Żeberska M, Drożdż W, Gryszkin A, Grzechac M. Current research addressing starch acetylation. Food Chemistry. 2015;176:350-6. 75. Mormann W, Al‐Higari M. Acylation of starch with vinyl acetate in water. Starch‐ Stärke. 2004;56(3‐4):118-21. 76. Jarowenko W. Acetylated starch and miscellaneous organic esters. Modified starches: Properties and uses. 1986;55. 77. Fritz J, Schenk G. Acid-Catalyzed Acetylation of Organic Hydroxyl Groups. Analytical Chemistry. 1959;31(11):1808-12. 78. Drazic A, Myklebust L, Ree R, Arnesen T. The world of protein acetylation. Biochimica et Biophysica Acta (BBA)-Proteins and Proteomics. 2016;1864(10):1372-401.

64

79. Lakso H, Ng W. Determination of Chemical Warfare Agents in Natural Water Samples by Solid-Phase Microextraction. Analytical Chemistry. 1997;69(10):1866-72. 80. Stashenko E, Martıneź J. Derivatization and solid-phase microextraction. Trends in Analytical Chemistry. 2004;23(8):553-61. 81. Zhang Z, Yang M, Pawliszyn J. Solid-Phase Microextraction. A Solvent-Free Alternative for Sample Preparation. Analytical Chemistry. 1994;66(17):844A-53A. 82. Pan L, Pawliszyn J. Derivatization/Solid-Phase Microextraction: New Approach to Polar Analytes. Analytical Chemistry. 1997;69(2):196-205. 83. Sng M, Ng W. In-situ derivatisation of degradation products of chemical warfare agents in water by solid-phase microextraction and gas chromatographic–mass spectrometric analysis. Journal of Chromatography A. 1999;832(1):173-82. 84. Desoubries C, Chapuis-Hugon F, Bossée A, Pichon V. Three-phase hollow fiber liquid- phase microextraction of organophosphorous nerve agent degradation products from complex samples. Journal of Chromatography B. 2012;900:48-58. 85. Dival, L. The development of an in-field rapid derivatisation technique for the analysis of chemical warfare agent degradants. Forensic Science, Masters by Coursework [dissertation], Western Australia: Murdoch University; 2018. Available from: Murdoch University Research Repository. 86. Chua, XL. Derivatisation of chemical warfare agent degradant without removal of water. Forensic Science, Masters by Coursework [dissertation], Western Australia: Murdoch University; 2018. Available from: Murdoch University Research Repository. 87. Firouzabadi H, Iranpoor N, Nowrouzi F, Amani K. Aluminium dodecatungstophosphate as a highly efficient catalyst for the selective acetylation of–OH,–SH and–NH2 functional groups in the absence of solvent at room temperature. Chemical Communications. 2003(6):764-5. 88. Bernal V, Castaño-Cerezo S, Gallego-Jara J, Écija-Conesa A, de Diego T, Iborra JL, et al. Regulation of bacterial physiology by lysine acetylation of proteins. New biotechnology. 2014;31(6):586-95. 89. Greene TW, Wuts PG. Protective groups in organic synthesis: Wiley; 1999. 90. Rutenberg MW, Solarek D. Starch derivatives: Production and uses. Starch: Chemistry and technology: Elsevier; 1984. p. 311-88. 91. Bello-Pérez L, Agama-Acevedo E, Zamudio-Flores P, Mendez-Montealvo G, Rodriguez- Ambriz S. Effect of low and high acetylation degree in the morphological, physicochemical and structural characteristics of barley starch. LWT-Food Science and Technology. 2010;43(9):1434- 40. 92. Aksnes H, Van Damme P, Goris M, Starheim K, Marie M, Støve S, et al. An organellar nα-acetyltransferase, naa60, acetylates cytosolic N termini of transmembrane proteins and maintains Golgi integrity. Cell reports. 2015;10(8):1362-74. 93. Helbig A, Gauci S, Raijmakers R, Van Breukelen B, Slijper M, Mohammed S, et al. Profiling of N-acetylated protein termini provides in-depth insights into the N-terminal nature of the proteome. Molecular & Cellular Proteomics. 2010;9(5):928-39. 94. Van Damme P, Evjenth R, Foyn H, Demeyer K, De Bock P, Lillehaug J, et al. Proteome- derived peptide libraries allow detailed analysis of the substrate specificities of Nα- acetyltransferases and point to hNaa10p as the post-translational actin Nα-acetyltransferase. Molecular & cellular proteomics. 2011;10(5):M110. 004580. 95. Helsens K, Van Damme P, Degroeve S, Martens L, Arnesen T, Vandekerckhove J, et al. Bioinformatics analysis of a Saccharomyces cerevisiae N-terminal proteome provides evidence of alternative translation initiation and post-translational N-terminal acetylation. Journal of proteome research. 2011;10(8):3578-89. 96. El Halal S, Colussi R, Pinto V, Bartz J, Radunz M, Carreño N, et al. Structure, morphology and functionality of acetylated and oxidised barley starches. Food chemistry. 2015;168:247-56.

65

97. Lyytikäinen M, Pellinen J. Some issues concerning the gas chromatographic determination of chlorinated phenolics in water. Toxicological & Environmental Chemistry. 1997;63(1-4):185-97. 98. Pokryshkin S, Kosyakov D, Kozhevnikov A, Lakhmanov D, Ul’yanovskii N. Highly Sensitive Determination of Chlorophenols in Sea Water by Gas Chromatography−Tandem Mass Spectrometry. Journal of Analytical Chemistry. 2018;73(10):991-8. 99. Stan H, Klaffenbach P. Determination of carbamate pesticides and some urea pesticides after derivatization with acetic anhydride by means of GC-MSD. Fresenius' journal of analytical chemistry. 1991;339(3):151-7. 100. Sadykova YM, Sadikova LM, Zalaltdinova AV, Dalmatova NV, Burilov AR, Pudovik MA, et al. Phosphorus-containing bicyclic phosphonates in silylation and acetylation reactions. Phosphorus, Sulfur, and Silicon and the Related Elements. 2016;191(3):493-5. 101. Riordan JF, Vallee BL. [66] Acetylation. Methods in Enzymology. 11: Academic Press; 1967. p. 565-70. 102. Batra P, Rozdon O. Acetylation of phenols using acetic acid. Proceedings Mathematical Sciences. 1949;29(5):349-51. 103. Glozak MA, Sengupta N, Zhang X, Seto E. Acetylation and deacetylation of non-histone proteins. Gene. 2005;363:15-23. 104. Cai L, Sutter B, Li B, Tu B. Acetyl-CoA induces cell growth and proliferation by promoting the acetylation of histones at growth genes. Molecular cell. 2011;42(4):426-37. 105. Dunbar R, Swenson W. Acetylation of imides with ketene. The Journal of Organic Chemistry. 1958;23(11):1793-4. 106. Chi H, Xu K, Wu X, Chen Q, Xue D, Song C, et al. Effect of acetylation on the properties of corn starch. Food Chemistry. 2008;106(3):923-8. 107. Iqbal J, Srivastava R. Cobalt (II) chloride catalyzed acylation of alcohols with acetic anhydride: scope and mechanism. The Journal of Organic Chemistry. 1992;57(7):2001-7. 108. Silva L, Gonçalves V, Mota C. Catalytic acetylation of glycerol with acetic anhydride. Catalysis Communications. 2010;11(12):1036-9. 109. Vogel’s textbook of practical organic chemistry, 5th edn (London: Pearson Education). 110. Rowell R, Simonson R, Hess S, Plackett D, Cronshaw D, Dunningham E. Acetyl distribution in acetylated whole wood and reactivity of isolated wood cell-wall components to acetic anhydride. Wood and Fiber Science. 2007;26(1):11-8. 111. Jin T, Ma Y, Zhang Z, Li T. Sulfamic acid catalysed acetylation of alcohols and phenols with acetic anhydride. Synthetic communications. 1998;28(17):3173-7. 112. Basu K, Chakraborty S, Sarkar A, Saha C. Efficient acetylation of primary amines and amino acids in environmentally benign brine solution using acetyl chloride. Journal of Chemical Sciences. 2013;125(3):607-13. 113. Tsunasawa S, Sakiyama F. Amino-terminal acetylation of proteins: An overview. Methods in enzymology. 106: Elsevier; 1984. p. 165-70. 114. Akçay M. The catalytic acylation of alcohols with acetic acid by using Lewis acid character pillared clays. Applied Catalysis A: General. 2004;269(1-2):157-60. 115. Hurd C, Cantor S, Roe A. Acetylation of Carbohydrates by Ketene. Journal of the American Chemical Society. 1939;61(2):426-8. 116. Ruiz-Matute A, Hernández-Hernández O, Rodríguez-Sánchez S, Sanz M, Martínez- Castro I. Derivatization of carbohydrates for GC and GC–MS analyses. Journal of Chromatography B. 2011;879(17):1226-40. 117. Sids O. Acetic anhydride SIAM; 1997. 118. Gold V. The hydrolysis of acetic anhydride. Transactions of the Faraday Society. 1948;44:506-18. 119. Dunbar RE, Garven FC. Acetylation of Monocarboxylic Acids with Ketene. Journal of the American Chemical Society. 1955;77(15):4161-2. 120. Gottlieb HE, Kotlyar V, Nudelman A. NMR chemical shifts of common laboratory solvents as trace impurities. The Journal of organic chemistry. 1997;62(21):7512-5.

66

121. Vogel AI. Practical organic chemistry. Longmans, London. 1956:168. 122. Sathe M, Gupta AK, Kaushik MP. An efficient method for the esterification of phosphonic and phosphoric acids using silica chloride. Tetrahedron Letters. 2006;47(18):3107- 9. 123. Scifinder, version 2019.1; RN 563-63-3. Predicted NMR data calculated using Advanced Chemistry Development, Inc. (ACD/Labs) Software V11.01 (1994-2019 ACD/Labs). 124. Foris A. On Hydrogen Bonding and OH Chemical Shifts. 125. Mass Spectral Standard Reference Database National Institute of Standards and Technology; 2005. 126. Rood D. Gas Chromatography Problem Solving and Troubleshooting. Journal of chromatographic science. 1998;36(9):476-7. 127. Herold R, Mokhatab S. Optimal design and operation of molecular sieve gas dehydration units—Part. Gas Processing & LNG. 128. Gleichmann K, Unger B, Brandt A. Industrial Zeolite Molecular Sieves'. Zeolites-Useful Minerals. 2016. 129. Kitson FG, Larsen BS, McEwen CN. Gas chromatography and mass spectrometry: a practical guide. San Diego: Academic Press; 1996. 130. Stan'kov I, Beresnev A, Lanin, S. Extraction-chromatographic determination of acids and their salts in water. Journal of analytical chemistry of the USSR. 1991;46(5):681-6. 131. Sathe M, Gupta AK, Kaushik M. An efficient method for the esterification of phosphonic and phosphoric acids using silica chloride. Tetrahedron letters. 2006;47(18):3107- 9. 132. Vos JM, De VR. Production of carboxylic acids. Google Patents; 1962. 133. Zimmermann H, Rudolph J. Protonic States and the Mechanism of Acid‐Catalysed Esterification. Angewandte Chemie International Edition in English. 1965;4(1):40-9. 134. Holmberg K, Hansen B. Ester synthesis with dicyclohexylcarbodiimide improved by acid catalysts. Acta Chem Scand B. 1979;33:410-2. 135. Valeur E, Bradley M. Amide bond formation: beyond the myth of coupling reagents. Chemical Society Reviews. 2009;38(2):606-31. 136. Sheehan JC, Hess GP. A new method of forming peptide bonds. Journal of the American Chemical Society. 1955;77(4):1067-8. 137. Ampadu Boateng D, Word MKD, Tibbetts KM. Probing Coherent Vibrations of Organic Phosphonate Radical Cations with Femtosecond Time-Resolved Mass Spectrometry. Molecules. 2019;24(3):509. 138. McLafferty FW. Mass spectrometric analysis. Molecular rearrangements. Analytical Chemistry. 1959;31(1):82-7. 139. Katagi M, Tsuchihashi H. VX and its decomposition products. Drugs and Poisons in Humans: Springer; 2005. p. 619-27. 140. Gutsev G, Ampadu Boateng D, Jena P, Tibbetts KM. A Theoretical and Mass Spectrometry Study of Dimethyl Methylphosphonate: New Isomers and Cation Decay Channels in an Intense Femtosecond Laser Field. The Journal of Physical Chemistry A. 2017;121(44):8414-24. 141. Holtzclaw JR, Wyatt JR, Campana JE. Structure and fragmentation of dimethyl methylphosphonate and trimethyl phosphite. Organic mass spectrometry. 1985;20(2):90-7. 142. Wang Z, Gao SJ, Che XX, Shen CH, editors. Synthesis and Characterization of a Flame Retardant Dimethyl Methyl Phosphonate (DMMP) and its Application in FRP. Advanced Materials Research; 2013: Trans Tech Publ. 143. Philippe R. Organic Reaction Workup Formulas for Specific Reagents. DCC: University of Rochester; 2019. p. 2-3. 144. Crenshaw MD, Cummings DB. Preparation, derivatization with trimethylsilyldiazomethane, and GC/MS analysis of a “pool” of alkyl methylphosphonic acids for use as qualitative standards in support of counterterrorism and the chemical weapons convention. Phosphorus, Sulfur, and Silicon. 2004;179(6):1009-18.

67

145. Stowell MHB, Ueland JM, McClard RW. The mild preparation of synthetically useful phosphonic dichlorides: Application to the synthesis of cyclic phosphonic diesters and diamides. Tetrahedron Letters. 1990;31(23):3261-2.

68