Thesis - Final Submission Diphedine

“Using Isotopic Fractionation to Link Precursor to Product
in the Synthesis of Diphenidine Hydrochloride: A Tool for

Combating New Psychoactive Substances (NPSs)”
Faye Hannah Owen
A thesis submitted in fulfilment of the requirements of the Manchester Metropolitan

University for the degree of Master of Science (by Research)
The Faculty of Science and Engineering

School of Science and the Environment
2017
1
“I declare that none of the work detailed herein has been submitted for any other
award at Manchester Metropolitan University or any other Institution.”
“I declare that, except where specifically indicated, all the work presented in this
report is my own and I am the sole author of all parts. I understand that any evidence
of plagiarism and/or the use of unacknowledged third part data will be dealt with as
a very serious matter”
Signature F. H. Owen Date: 30-Jan-2017
2
Acknowledgements
The author would like to give thanks to Drs Oliver Sutcliffe and Ryan Mewis, along
with their research teams for their continued support throughout the project and to
Manchester Metropolitan University for the use of their facilities and the technical
staff for help in the laboratory. Also many thanks to Professor Niamh Nic Daéid and
Dr. Craig McKenzie at the Centre for Anatomy and Human Identification (CAHID),
Dundee University for the use of their facilities and processing of the Isotopic Ratio
Mass Spectrometry (IRMS) data.
3
Contents Page
1. Abstract........................................................................................................................... 6
2. Introduction .................................................................................................................... 8
2.1. New Psychoactive Substances ............................................................................... 8
2.2. Drug Classification and Legislation ................................................................... 10
2.3. Clinically Abused Drugs ...................................................................................... 12
2.4. Dissociative Agents............................................................................................... 13
2.4.1. NMDA Receptor Activity............................................................................... 13
2.4.2. Phencyclidine ................................................................................................ 14
2.4.3. Ketamine ........................................................................................................ 15
2.4.4. Methoxetamine .............................................................................................. 17
2.5. Diphenidine........................................................................................................... 18
2.6. Isotopic Ratio Mass Spectrometry...................................................................... 20
3. Aims............................................................................................................................... 23
4. Experimental ................................................................................................................ 24
4.1. Materials and Equipment.................................................................................... 24
4.2. The synthesis of diphenidine hydrochloride (8) ................................................ 24
4.2.1. Exemplar synthesis of diphenidine hydrochloride, Method A .................... 24
4.2.2. Synthesis of N-benzylpiperidine hydrochloride ........................................... 25
5. Results and Discussion ................................................................................................. 28
5.1. Synthesis................................................................................................................ 28
5.2. Gas-Chromatography Mass-Spectrometry (GCMS) ........................................ 33
5.3. Quantitative Gas-Chromatography Validation Method .................................. 37
5.4. Percentage Yield................................................................................................... 40
5.5. Isotopic Ration Mass Spectrometry ................................................................... 44
5.5.1. Natural Abundances and Delta ....................................................................... 44
5.5.2. Batches A1, A2 and A3, effect of changing supplier ................................... 45
5.5.3. Methods A and B, effect of order of addition in air ..................................... 48
5.5.4. Methods C and D, effect of order of addition under argon ......................... 50
5.5.5. Batches E and F, effect of mmol piperidine................................................. 52
5.5.6. Batches G and H, effect of stirring time ....................................................... 54
6. Conclusion .................................................................................................................... 56
7. Future Work ................................................................................................................. 58
8. References ..................................................................................................................... 59
4
List of Abbreviations
AbbreviationColumn1 Definition
ACMD Advisory Council on the Misuse of Drugs
AKB-48 N-(1-adamantyl)-1-pentyl-1H-indazole-3-carboxamide
CAHID Centre for Anatomy and Human Identifications
CD2Cl2 Deuterated dichloromethane
CH2Cl2 Dichloromethane
CN Cyanide Ion
CNS Central Nervous System
EI Electron Impact Ionisation Energy
EMCDDA The European Monitoring Centre for Drugs and Drug
Addiction
EWS EU Early Warning System
GCMS Gas chromatography-Mass Spectrometry
H2SO4 Sulphuric Acid
HCl Hydrochloric Acid
IR Infrared Spectroscopy
IRMS Isotopic Ration Mass Spectrometry
LCMS Liquid chromatography-mass spectrometry
MDMA 3,4-methylenedioxymethamphetamine
MgBr Magnesium bromide
MgSO4 Magnesium sulphate
MXE methoxetamine
MXP Methoxphenidine
NaOH Sodium Hydroxide
NH4Cl Ammonium Chloride
NMDA N-methyl-D-aspartate
NMDAR N-methyl-D-aspartate receptor
NPS New Psychoactive Substances
NRG-1 Naphyrone
PCA 1-(1-phenylcyclohexyl)amine
PCP phencyclidine
PMMA para-methoxymethamphetamine
pTSA p-toluene sulphuric acid
TFA Trifluoroacetic acid
U.K. United Kingdom
WHO World Health Organisation
5
1. Abstract
In recent years, there has been an unprecedented growth in the emergence of “legal
highs”, or as they are now commonly known, New Psychoactive Substances (NPS).
In most cases, they mimic the effects of illicit substances, however, they circumvent
legislation through changes in their chemical structures. NPS are being successfully
sold through online venders under the titles “bath salts” and “research chemicals” and
along with their low cost, are becoming an increasingly popular, convenient,
alternative to controlled substances.
Diphenidine, is an example of a fourth generation NPS that has increased in popularity

as a recreational dissociative drug since its appearance in 2013. Diphenidine poses a
threat to public health and safety due to lack of legislative control but also through the
lack of literature concerning its toxicity, chemical and physical data, with evidence
showing diphenidine has attributed to several deaths already. It is becoming apparent
that there is a need for a rapid and selective detection method in order to identify and
control not only diphenidine, but other legal and illicit substances too. Isotopic Ratio
Mass Spectrometry (IRMS) is an analytical technique that has flourished within the
forensic community, having the potential to not only link diphenidine samples of the
same manufacturer, but also differentiate between sample seizures. IRMS looks
beyond chemical composition but instead, at the relative isotopic abundances of the
elements composing the starting materials and final product. Trace impurities could be
used as a method of source identification, as the impurity will be unique to the drug
and its history of origin. This could potentially allow discrimination between synthetic
routes used.
The main objective of the study is to determine whether IRMS can be used as tool to
link product to precursor in the case of diphenidine hydrochloride, which if successful,
would could have a wide variety of forensic applications for further NPS and illicit
drugs.
After designing and implementing a new GC-MS method to determine the purity of
diphenidine hydrochloride, whilst considering peak area normalisation calculations
and percentage yield data, the optimum method for the synthesis of diphenidine
6
hydrochloride is F. This method uses 30 mmol piperidine and 1 hour stirring time
under an inert atmosphere to yield diphenidine hydrochloride in much quantities with
minimal trace impurities compared to the other methods employed.
From IRMS data it is possible to distinguish between sources of precursors except

when two sources are owned by the same brand, i.e. Sigma Aldrich can be
distinguished from Alfa or Acros, but Alfa and Acros products cannot be
discriminated. There are also significant changes the isotopic rations for both carbon
and nitrogen when the physical parameters for the synthesis are changed, notably when
the reaction is performed under argon and the wen the molar equivalents of the
precursors are increased, i.e. the mmol ration of piperidine/ Like the source
differentiation results, there are substantial different between 10 mmol and 20 mmol
ratios, but it is difficult to distinguish between 20 mmol and 30 mmol. In conclusion,
it would be possible to link precursor to product and also the synthetic pathway used,
but not with 100% certainty.
7
2. Introduction
2.1. New Psychoactive Substances
New Psychoactive Substances (NPS) is a term now used by the scientific community
to identify substances, formally known as “legal highs”, which are abused on the
recreational drugs market.1 In most cases, they mimic the effects of illicit and
controlled drugs, for example, cathinones (1) (which mimic amphetamine substances
such as 3,4-methylenedioxymethamphetamine (MDMA)) and synthetic cannabinoids
(which mimic the effects of ∆9-tetrahydrocannabinol, the principle active ingredient in
cannabis), yet are lacking definite legislation control.1
Drugs of abuse can fall into one of three main categories: natural illicit drugs that are
misused, for example cocaine and morphine; synthetic designer drugs specifically
made to mimic illicit drugs, for example methoxetamine (MXE, 2); and drugs that
have previously been used for clinical purposes, for example, benzodiazepines.2
NPS, or as they were more commonly known as “legal highs”, tend to fall into either
the designer drugs or the clinical drugs category. Currently on the recreational drugs
market, there are many classes of new psychoactive substances, including but not
limited to: cathinones (e.g. mephedrone) phenethylamines (3, e.g. 4-
methoxymethamphetamine, PMMA), piperazines (4, e.g. benzylpiperazine), synthetic
cannabinoids (e.g. N-(1-adamantyl)-1-pentyl-1H-indazole-3-carboxamide), AKB-48)
and dissociative agents (e.g. piperidine (phencyclidine, PCP, 5) and ketamine (6)
derivatives), shown in Figure 1.1, 3-5
The fundamental concept when initially designing an NPS is to expand upon a

previously known structure and by subtly changing it, for example, the addition of
functional groups or side-chains, causing the drug to no longer be illegal by
circumventing legislation, for example, the Misuse of Drugs Act (1971) in the UK.1, 6-
9
Identifying the psychoactive element of the drug and using the structure as a
replicable template, the new psychoactive substance is designed to possess the same
pharmacophore and target, consequently exhibiting similar psychoactive properties as
8
the well-known and studied illicit drugs. By this process, it is possible to create a
limitless library of new psychoactive substances.
Figure 1. Several of the more common new psychoactive substances on the recreational drugs market.
With marketing playing a crucial role in NPS popularity,8 they are sold under the titles
“bath salts”, “plant food”, “research chemicals” and “not for human consumption”
tactfully avoiding marketing regulations and therefore being legally sold, exploiting
laws such as the United Kingdom Medicines Act (1968).1, 2, 6-14 Because they are
branded as “legal highs” and their wide availability through online suppliers and head
shops combined with their low cost means they are becoming an increasingly popular
and convenient alternative to controlled substances.1, 2, 6, 7
The European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) through
the EU Early Warning System (EWS), have reported a dramatic growth of the amount
of legal highs emerging and their availability. With an increase of the amount of online
venders now selling NPS, 693 documented in 2012, which is a 300% increase since
2010, it is not surprising that the EMCDDA documented 101 new substances in 2014
and a further 100 in 2015.8, 9
9
Lack of literature concerning their pharmacology, toxicity and analytical
characteristics, for example physical and chemical data, means they pose a threat to
society and human health.8, 12-15 In most cases, their mechanisms of action are not fully
understood and with evidence showing that they are commonly found in conjunction
with other new or even controlled drugs, their potential for harm is increased.8, 10, 14
Not only does this indicate that the contents are different to those stated, but if higher
doses are required to achieve the same psychedelic effects, this could result in adverse
effects being exhibited; including cognitive impairment, paranoia and possibly
fatalities.15 With limited analytical data available on these new psychoactive
substances, they are not being detected in standard drug screenings.3, 8 This adds to
their appeal and broadens the consumer market, identifying that 8% of 15-24 year old
in the U.K. have used “legal highs”.8
Several recent deaths in the U.K.11, 13, 16, 17 supported by toxicology reports and autopsy
findings,18, 19 suggest that new psychoactive substances have contributed considerably
to the deaths; it is now apparent that legislation needs to adapt to this constantly
evolving drug market. Through such evidence, it is now recommended that law
enforcement agencies constantly monitor their use and prevalence,20 prompting more
rapid identification to the EWS and in turn their control.8
2.2. Drug Classification and Legislation
In the U.K., the first form of legislation began with the Misuse of Drugs Act (1971).
Brought around in an attempt to control drugs, it categorises scheduled substances into
either Class A, B, or C. Each class comes with varying levels of legal control based on
the substances’ potential for being misused, whether it is being misused and the
harmful effects associated with the substance and whether these would constitute to be
a social problem, with the overall aim to limit their distribution and usage.21
As of April 2010, many cathinones and synthetic cannabinoids were brought under the
Misuse of Drugs Act (1971) in the U.K.1, 2, 6-11 This change in legislation was expected
to decrease the popularity and use of such drugs, however, prompted the emergence of
second, (e.g. Naphyrone, NRG-1, 7), third (e.g. MXE, 2) and fourth generation (e.g.
10
diphenidine, 8 and methoxphenidine, 9) NPS, shown in Figure 2, to flood the
recreational drugs scene.
Figure 2. Examples of second (NRG-1, 7) and fourth generation (diphenidine, 8 and

methoxphenidine, 9) NPS that have begun to populate the recreational drugs scene.
With NPS emerging in great diversity and at speed, it has caused legislation to evolve.
Initially managing NPS under existing drug legislation and consumer safety laws
allowed temporary drug orders to come into place quickly and can last up to one year,
in which time thorough investigation into their harm and threat to society can occur.
The U.K. has initiated a “generic legislation system” where a precise definition of a
family of substances is created in order to group substances together in terms of their
chemical structure. This attempts to control the emergence of newly emerging
derivatives and analogues.8
In 2013, the U.K. implemented an arylcyclohexylamine ban, causing recreational

dissociative agents to become illegal, consequently creating a gap in the market. This
11
gap prompted a new class of dissociative agent to emerge, diarylethylamines, for
example, diphenidine (8) and 2-methoxphenidine (9), which currently remain legal
under U.K. law.
As of January 2016, the U.K. brought in the Psychoactive Substances Act, defining a
psychoactive substance as “a substance produces a psychoactive effect in a person if,
by stimulating or depressing the person’s central nervous system, it affects the person’s
mental functioning or emotional state”. Though the Act doesn’t specifically
criminalise possession (unless within a custodial setting), it effectively makes it a
criminal offence to produce, supply, offer to supply, possess with intent to supply, or
import or export psychoactive substances thereby hopefully restricting the circulation
of potentially dangerous psychoactive substances. Though the Act has successfully
shutdown the sale of NPS through headshops and/or specialist internet sites, it is
unclear, at present, how effective the legislative change will be on curtailing the illegal
sale (through organised crime) and the variety of substances in circulation.
2.3. Clinically Abused Drugs
Many pharmaceuticals, such as anaesthetics and benzodiazepines, are clinically

abused drugs, which have become increasingly popular in recent years. Usually
obtained via prescription or are readily available over the counter at pharmacists, they
are being used as recreational drugs and not for their originally intended use. Clinical
drugs that have been previously researched and tested are now emerging under the title
“research chemicals” and are contributing heavily to the research chemical market.
One of the more notable classification of clinical drugs to be abused began to emerge
in the 1950s, the arylcyclohexylamine family. Structurally composed of a
cyclohexamine unit with an aryl unit attached geminally, this group of substances is
collectively known as “dissociative anaesthetics”, originally employed as
pharmaceutical anaesthetics before becoming drugs of abuse.
12
2.4. Dissociative Agents
2.4.1. NMDA Receptor Activity
The most recognised pharmacological feature associated with dissociative agents is

their ability to act as N-methyl-D-aspartate receptor antagonists (NMDAR), acting
upon the NMDA receptors, shown in Figure 3, inducing anaesthesia. NMDA receptors
are fundamental for mediating neurotransmitters within the central nervous system
(CNS), in particular glutamate. Here, they can initiate intracellular pathways,
consequently affecting neuronal activity.10, 22
10
Figure 3. An example of the activated NMDA receptor (10) showing how glutamate is within the
glutamate -binding site.
It is well documented that dissociative agents, like PCP and its derivatives, act as
competitive inhibitors, binding to these receptors and causing normal neurotransmitter
activity to cease.22
NMDAR activity plays a vital role in physiological processes, for example

neurophysiology.23 It is the NMDAR antagonism via open channel blockade10, 14, 22, 24
that is suspected to be the fundamental pharmacological mechanism contributing
towards the “dissociative” and dream like states,25 however, the degree of dissociation
experienced is suspected to be dependant entirely on dose.14, 16, 22
At low doses, memory improvement and stimulation occurs, however, the well-known
adverse effects include short-term amnesia (memory loss), analgesia (pain relief) and
altered cognition are experienced at high doses. It is also documented that perceptual
alterations, hallucinations, ataxia (loss of control of bodily movements) and
paraesthesia (pins and needles/tingling) occur, regardless of dose.25 In some cases,
13
tachycardia and hypertensions have been reported26 suggesting that the NMDAR
action is not limited to the CNS but stretches to the circulatory system too.
2.4.2. Phencyclidine
1-(1-phenylcyclohexyl)piperidine hydrochloride salt (11), also known as

phencyclidine or PCP, was the first arylcyclohexylamine synthesised in 1956 by Victor
Maddox of Parke-Davis, a subsidiary of the pharmaceutical corporation, Pfizer.25, 26
Maddox unknowingly performed a Bruylants amination, however, literature disclosed
only the synthesis of the immonium intermediate, and in order proceed, the
intermediate must undergo a substitution reaction with potassium cyanide and
phenylmagnesium. The suspected synthetic pathway is shown in Scheme 1.25, 26
Scheme 1. A possible synthetic pathway for the original synthesis of PCP in 1956.26
The original analogue of (11), 1-(1-phenylcyclohexyl)amine (PCA, 12), shown in

Figure 4, was synthesised as early as 1907 where it was established to be a potent
sedative, however, studies were discontinued and (12) was consequently dismissed as
a pharmaceutical agent.
Figure 4. PCA, the original analogue which led to the synthesis of the derivative, PCP (11).
14
It was not until 1957 when clinical trials for (11) started, where the potential of such
dissociative agents as effective anaesthetics became apparent. Soon afterwards, (11)
became a surgical anaesthesia, due to its actions on NMDA receptors. However, in
1965, it was discontinued as a pharmaceutical drug because the majority of patients
experienced adverse side effects, including psychosis and agitation.14, 25
Since PCP’s withdrawal as an anaesthetic, it became regarded as a hallucinogenic drug

and has thrived as a drug of abuse since the 1960s, where it is often mixed with other
illicit substances and subsequently intensifying the toxicity. With idealistic side effects
such as auditory hallucinations, altered perception and euphoria being experienced, its
popularity has shown an overall increase.14
Evidence also reports PCP users experiencing convulsions, delirium, along with
serious neurological, cardiovascular and psychotic reactions, ultimately leading to
psychotic behaviour.14, 25 Through this, and in addition to, a vast amount of media
coverage, a change in legislation was prompted, with PCP becoming classified as a
Class A drug in the U.K, and subsequently, derivatives of this illicit drug, such as
ketamine (6), has begun to populate the recreational drug scene.25
2.4.3. Ketamine
An example of such a PCP derivative is the ketamine hydrochloride salt (2-(2-

chlorophenyl)-2-(2-methylamine)cyclohexanone), (6), which was first synthesised in
1962 by Calvin Steven’s of Parke-Davis, see Scheme 2. Once pharmacologically
evaluated, (6) was established to be a fast, though short-acting, general anaesthetic to
be used for both human, in particular paediatric, and veterinary practises.25, 27
However, it was not until 1969 that (6) became a commercial anaesthetic, sold under
the brand name Ketalar.25
Categorised in 1985 as an essential drug by the World Health Organisation (WHO),

(6) is a popular anaesthetic agent and pain relief in developing countries. With having
a rounded safety profile,27 in addition to being easily administered, it is often utilised
in the sedation of both adults and children, particularly where medical equipment and
conditions are lacking. Ketamine has a wide range of therapeutic applications
15
including depression and psychiatric treatment, which has led to multiple studies
investigating the use of ketamine as a treatment for depression and anxiety.1, 25, 29
Scheme 2. 1-hydoxycyclopentyl-(o-chlorophenyl)-ketone-N-methylamine (obtained via a Grignard

reagent) reacting with decalin under acidic conditions to produce 6.28
Despite the apparent positive applications of ketamine, it has become one of the most
popular drugs of abuse not only in the U.K., but globally, and has grown in popularity
over recent years.1, 25
Ketamine’s primary pharmacological mechanism is via the
antagonism of the NMDA receptor,22 similar to that of PCP, which has been a target
of exploitation when used as a recreational drug.
Ketamine users report feelings of “depersonalisation” once consumed, commonly

referred to as a “k-hole”, often experienced in conjunction with hallucinations and
altered perception of realities. These idealistic effects, along with being readily and
cheaply available, has caused ketamine’s abuse to increase greatly over recent years.
Nevertheless, as with all drugs, adverse effects are common. When high doses of
ketamine are consumed, adverse effects include; tachycardia, hypertension, and
amnesia. Recent literature has linked ketamine use to haematuria, bladder shrinkage
and degradation.1, 25, 27, 29
Through the discovery of the latest adverse effects and with the Advisory Council on
the Misuse of Drugs (ACMD) reporting a recent increase in the illicit use of ketamine,
it regained its Class B status in the U.K.25
16
2.4.4. Methoxetamine
The scheduling of ketamine (6) caused a third generation of NPS to emerge, 2-(3-
methoxyphenyl)-2-(ethylamino)cyclohexanone, more commonly referred to as
methoxetamine (MXE, 2), which is marketed as the new, legal and bladder safe
alternative27 to its predecessor with its appearance in the U.K. first reported in 2010
by the EMCDDA. 25, 30
In terms of chemical structure, it remains an arylcyclohexylamine class and is

considered to be a structural derivative of ketamine, whereby a 3-methoxy group has
replaced the 2-chloro attached to the phenyl ring and the N-methyl group replaced by
an N-ethyl group.10, 27 Although to the eye they are minor modifications, the entire
pharmacology is changed, causing methoxetamine (2) to be a more potent and longer
lasting hallucinogenic drug than the ketamine parent.10, 25
Figure 5 shows the
development of structures from PCP through to ketamine followed by MXE.
Figure 5. The development of structures from the original dissociative anaesthetic 11, followed by
derivative 6, and then to the designer drug 2.
Initially marketed as a “research chemicals” or “not for human consumption”,10

methoxetamine (2) quickly increased in popularity on a global scale through users
experiencing similar psychedelic effects produced by the previous recreational
dissociatives. For example, hallucinations, depersonalisation and euphoria are
commonly reported.10, 27
Although designed to have a similar safety profile to that of ketamine by retaining the
2-keto functional group,25 there have been an increasing amount of scientific
17
publications documenting evidence on (2) intoxication and related fatalities,16 with
110 non-fatal intoxications reported by the EMCDDA in 2012 alone.27, 31 Adverse
effects associated with higher doses of this designer drug include; anxiety, aggression,
ataxia, hypertension, violence and agitation.10, 27
A potential, yet significant, concern for public health and safety is that (2) has an
increased potency and if consumed in the same dose as ketamine, accidental
intoxication could occur ultimately resulting in fatalities.
With evidence mounting and the growing need for legislation to control this NPS, it
has since been classified as a Class B drug in the U.K. as of February 2013.25 Illegal
across Asia and Europe too, it was predicted that the appearance of more legal highs
would start to influx the recreational drugs scene.10
2.5. Diphenidine
Change in legislation through the restriction and banning of arylcyclohexylamine

derivatives in February 2013 prompted a new dissociative class of diarylethylamines
to emerge onto the research chemical (RC) market, as little as 2 weeks later.25 1-(1,2-
diphenylethyl)piperidine or more commonly known as diphenidine hydrochloride (8)
and its derivative 1-(1-(2-methoxyphenyl)-2-phenylethyl)piperidine, MXP (9, see
Figure 2) both fall under this classification and are known as a legal replacement for
MXE and other PCP-derived drugs.
Scheme 3. A predicted scheme of how the diphenidine hydrochloride salt would be synthesised if
following the Bruylants amination, showing the predicted experimental conditions and intermediate.
Diphenidine hydrochloride was originally synthesised in 1924 via the Bruylants

amination, utilising organometallic Grignard reagents, shown in Scheme 3, which is a
18
similar preparation to that which was later used by Maddox in the synthesis of PCP in
1956, see Scheme 1.25
Similar in structure to that of illicit dissociatives, it is not surprising that

pharmacological studies have identified NMDAR activity as the primary method of
action of (8). However, the exact binding has not been fully explored but evidence
does suggest that the S-enantiomer, shown in Figure 6, is more potent, showing a
higher affinity for binding.32
Figure 6. The two enantiomers of diphenidine hydrochloride (8), S and R
The subtle differences in terms of structure that cause diphenidine hydrochloride to be

classified as a diarylethylamine also allows diphenidine hydrochloride to circumvent
legislation, remaining legal to possess and sell. Being sold by online vendors as a legal
replacement for MXE, with users reporting similar psychedelic effects that of
arylcyclohexylamines, such as strong dissociative effects and “k-hole” like
experiences.25
Of recent years, since the appearance of diphenidine in 2013,25 its popularity as a drug
of abuse has increased significantly, especially in Japan.18, 19, 33
When taken in
relatively low doses, diphenidine seems to present a low threat to public health and
safety with mild effects reported, but when administered in higher doses, severe
adverse effects are experienced which could even result in death. This is supported by
toxicology and post-mortem results that report diphenidine to be a causative agent in
several fatalities.18, 19
19
An example of such an outcome occurred in Japan, where diphenidine was found in a
sample of “herbal incense”, amongst with several synthetic cannabinoids, including
AB-CHMINACA. After a 30-year old male died from inhaling “herbal incense, the
super lemon”, post-mortem results determined that the cause of death was through
diphenidine poisoning. There was a significantly high concentration found in the
adipose tissue (~11,000 ng/g mL-1), where the diphenidine remained to be in its
unchanged form and in a concentration which was obviously too high for the body to
metabolise, especially when in conjunction with other synthetic drugs.19
Through such evidence, it is being recommended that law enforcement agencies

should make a conscious effort to monitor the usage of diphenidine for harm-reduction
purposes, as there is clearly a large risk to public health and safety.18, 20 With little
analytical data available in scientific literature,1, 6, 25
the trafficking of this legal
substance is becoming an increasingly difficult task. To meet demand, there is a need
for a rapid and selective detection method in an attempt to identify and control not
only diphenidine, but also other derivatives and novel drugs of abuse.
2.6. Isotopic Ratio Mass Spectrometry
One method by which the reduction in trafficking, not only diphenidine, but also illicit
substances, can potentially be achieved is through the application of Isotopic Ratio
Mass Spectrometry (IRMS) as a method of source identification. Until recently, IRMS
is a technique that has been previously overlooked as a tool for the analysis of
controlled drugs. However, as this is one of the largest areas of active development,
along with a significant increase of publications focusing on IRMS, it has become
widely use amongst many disciplines and has particularly flourished within the
forensic science community. 34-36
With the unprecedented growth of illicit drugs and NPS on not only a national, but a
global scale, the demand for a technique that would allow seized drugs to be
differentiated by source, has significantly increased.11, 36, 37 The development of drug
profiling and intelligence gathering, which are methods used to aid and support crime
scene investigations, have the potential to identify samples of the same, and
theoretically, different sample seizures.
20
More advanced than simpler analytical methods, such as Gas Chromatography-Mass
Spectrometry (GC-MS) and Liquid Chromatography-Mass Spectrometry (LC-MS),
IRMS has opened the possibility of not only discriminating the geographical
provenance of the drug but also the precursors used, synthetic route and even product
batches, all of which data that before has been unobtainable. A similarity of these
applications is that they depend on the exploitation of stable isotopes in order to deduce
a trace source.11, 34-36
The fundamental concept of IRMS as an analytical technique is that it looks beyond

the chemical composition of the compound, but rather at the isotopic composition, in
this case, a synthetic drug, which is extremely desirable as a forensic tool.34, 36
The relative natural isotopic abundances of the chemical elements in the precursors
13
used to synthesis the compound, for example C/12C and 15
N/14N are expected to
remain constant throughout the synthesis, until a chemical or physical process alters
them. If changes are observed, this would potentially be due to reaction specific
isotopic fractionation.11, 35, 36
Trace impurities have the potential to play a crucial role in changing the isotopic ratios
of the compound and can be created from an array of sources: i.e. in the original
precursors, during the synthesis, handling and environmental factors. All of these
factors, if they change the relative isotopic abundances, could potentially be traced and
are common amongst drugs synthesised in clandestine laboratories, where health and
safety and quality control standards are lax in comparison to pharmaceutical
companies.11
A notable example of such methods was in North-Korea, whereby 20 samples of

heroin were seized. The nitrogen isotopic ratios of the samples were calculated then
compared to a library of over 200 authentic samples. Despite there being no matches
to that on record, it could be determined that the seized samples were likely to have
been from a new source.36, 38
Another area of investigation using IRMS is linking precursors to product in the

synthesis of mephedrone. It was suggested that the natural abundance of the chemical
elements in the precursors would be integrated into the intermediate and consequently,
21
the final product. Samples of mephedrone were synthesised using precursors from two
separate suppliers and throughout each synthesis, the δ13C and δ2H values were
measured for each of the precursor, intermediate and final product. Although there
were little differences in the δ13C (%) between the precursors, there was a notable
difference between the δ2H (%). The changes have possibly given a quantifiable link
between precursor and product, providing a foundation for future work in which IRMS
could provide more than just the synthetic route, but a potential link to the precursor
and/or manufacturer.11
Other examples of this methodology in the forensic discipline has already been applied
to the geo-location of heroin and cocaine,39 and the profiling of methamphetamine
samples.40.
22
3. Aims
Diphenidine (isolated as its corresponding hydrochloride salt) is isolated through a

zinc-mediated Barbier-type reaction involving benzyl bromide, benzaldehyde and
piperidine. Although very similar to a Grignard reaction, this reaction can, however,
be performed in “one pot” with the organometallic intermediate created in-situ.
The main objective of the study is to determine whether IRMS can be used as tool to
link product to precursor which if successful, would could have a wide variety of
forensic applications. Diphenidine hydrochloride will be synthesised via the Barbier
reaction, using only three different sources of the precursors (batch specific). Changing
only these variables whilst keeping the physical paraments controlled could allow for
the discrimination between different supplies of the precursors and thereby potentially
determine the source of precursors within a seized sample through the application of
IRMS. The physical paraments of the synthesis can then also be investigated, example,
moisture/oxygen, stirring time and the order of addition of the reagents and the affect
these changes have on first the yield then isotopic ratio of the products. This will
provide insight into whether the physical parameters used in a chemical synthesis
affect the isotopic ratio and whether these changes can be used as a “fingerprint” by
which we can determine the method by which a sample of diphenidine hydrochloride
has been produced. With only two publicised methods to synthesise diphenidine
hydrochloride and currently clandestine methods are unknown, the impurities and
IRMS produced for both synthetic pathways need to identified which would provide a
reference for when seized samples are analysed. Impurity profiling via this method
would enable the synthetic pathway used by clandestine laboratories to be identified
and the intrinsic characteristics associated with each sample.
In each variation of the parameter conditions, the product will be prepared in batches
of six replicates, each of the samples will be analysed by GC-MS using an in-house
developed method before being sent for IRMS analysis. Though the precursors and
products are not controlled under current U.K. legislation, the synthesis will be carried
out in strict accordance with Home Office guidelines (under licence) and the samples
stored securely in compliance with School of Science and Environment procedures
and practices.
23
4. Experimental
4.1. Materials and Equipment
All chemicals were obtained from either Sigma Aldrich, Acros Organics or Alfa Aesar
and were between 98% and 100% pure and were used without purification. NMR
spectra, including 1H and 13
C spectra, were acquired on a JEOL AS-400 (JEOL,
Tokyo, Japan) NMR spectrometer operating at a proton resonance frequency of 400
MHz. Samples were prepared via 10-20 mg of sample in 750 μL of deuterated solvent/
TFA 0.03 %. Infrared spectra were obtained in the range of 400 – 4000 cm-1 (8 scans)
using a ThermoScientific Nicolet iS10ATR-FTIR instrument (ThermoScientific,
Rochester, USA). High-resolution mass spectra were recorded on an Agilent 1260
infinity LC coupled to a 6540 UHV accurate mass Q-TOF mass spectrometer by
looped injection using electrospray ionisation (ESI, collision energy: 15 eV). Melting
points were acquired using Gallenkamp 5A 6797 apparatus and are uncorrected.
4.2. The synthesis of diphenidine hydrochloride (8)

4.2.1. Exemplar synthesis of diphenidine hydrochloride, Method A
An oven dried 100 mL round-bottom flask was charged with acetonitrile (40 mL) and
zinc dust (2.0 g, 30 mmol). Trifluoroacetic acid (TFA, 0.2 mL) and benzyl bromide
(0.4 mL, 3 mmol) were added and the resulting solution stirred for 5 mins. Benzyl
bromide (3.0 mL, 25 mmol), benzaldehyde (1.2 mL, 11 mmol) and piperidine (0.9 mL,
10 mmol) were introduced to the mixture and the solution stirred at room temperature
for 1 h. The resulting solution was then quenched with saturated aqueous ammonium
chloride solution (NH4Cl, 150 mL) and extracted with dichloromethane (CH2Cl2, 2 x
100 mL). The organic layers were combined, dried over anhydrous magnesium
sulphate (MgSO4) and concentrated in vacuo. Diethyl ether (150 mL) was added to the
resultant oil and after complete dissolution, concentrated sulphuric acid (H2SO4, 0.75
mL) was added dropwise to the stirred solution and allowed to react for 5 min. After
decanting the sulphuric acid, the remaining precipitate was then dissolved upon
addition of 5% sodium hydroxide aqueous solution (NaOH, 100 mL) before being
extracted with dichloromethane (CH2Cl2, 2 x 100 mL). The combined organic
fractions were dried over magnesium sulphate (MgSO4) and concentrated in vacuo to
give an amber oil. The resultant yellow oil was then dissolved in diethyl ether (200
24
mL) before adding hydrochloric acid in dioxane (4 M, 3 mL) dropwise, whilst stirring.
Upon repeated trituration with ice cold acetone, a white powder was yielded (0.45 g,
17%).
Mpt: 210 – 213 °C. IR (ATR-FTIR, cm -1): 3644 (N-H stretch), 3030 (alkyl C - H
stretch), 2933 (C-H stretch), 2365 (NH+), 1604 (aromatic C=C bend), 1494 – 1453
(aromatic C-C stretch). LC-MS (ESI+, 70 eV: m/z = 266.1909. GC-MS (EI, 70 eV):
tR 5.70 min; m/z = 91, tR 14.61 min; m/z = 174. 1H NMR (400 MHz, CD2Cl2)  ppm:
12.45 (NH+, br, 1H), 7.02 – 7.39 (Aromatic, CH, m, 10H), 4.10 – 4.23 (CH, d, 1H),
3.8 – 3.9 (CH2, t, 2H), 3.3 – 3.6 (CH2, m, 3H), 2.18 – 2.60 (CH2, m, 4H), 1.80 – 1.88
(CH2, m, 4H), 1.25 – 1.28 (CH2, t, 2H). 13C NMR (400 MHz, CD2Cl2)  ppm: 126.91
– 135.90 (Ar), 73.05 (CH), 48.88 (CH2), 36.88 (CH2), 22.86 (CH2), 22.39 (CH2).
The specific methods for batches (B – H) are provided in the supplementary

information. A summary of the modifications is provided in Table 1, below.
4.2.2. Synthesis of N-benzylpiperidine hydrochloride
To an oven dried 50 mL round bottom flask, equipped with stirrer bar, diethyl ether
(25 mL) was added before benzylpiperidine (0.5 mL, 11.4 mmol) was introduced,
dropwise. The mixture was concentrated in vacuo to yield a viscous, colourless oil.
The resultant oil was fully dissolved in diethyl ether (10 mL) before adding HCl in
dioxane (4 M, 0.5 mL) dropwise, whilst stirring. The resultant mixture filtered to yield
a white crystalline powder (0.8 g, 40.1%).
Mpt: 6 - 7 °C. GC-MS (EI, 70 eV): tR 5.80 min; m/z = 91. 1H NMR (400 MHz, CDCl3)
 ppm: 12.44 (NH+, br, 1H), 7.2 – 7.8 (Aromatic, m, 5H), 4.10 (d, 2H, J = 8 Hz), 3.36
(d, 2H, J = 12 Hz), 2.57 (q, 2H, 12 Hz), 2.16 (t, 2H, J = 16 Hz), 1.80 (t, 3H, J = 20 Hz,
J = 16 Hz,), 1.3 (q, 1H, 4 Hz). 13C NMR (400 MHz, CD2Cl2)  ppm: 129.12 – 131.95
(Ar), 60.92 (CH), 52.84 – 54.38 (CH2), 36.88 (CH2), 22.86 (CH2), 22.40 (CH2).
25
Table 1. All the parameters that were changed but also kept constant throughout the different methods.
For the detailed experimental method, please see the supplementary information.
Method Physical Parameters

A Order of addition without argon flush,
piperidine added last, stirring time = 60
mins, temp = RT. Changing sources of
precursors.
B Order of addition without argon flush
benzaldehyde, stirring time = 60 mins,
temp = RT.
C Order of addition with argon flush,
piperidine added last, stirring time = 60
mins, temp = RT
D Order of addition with argon flush,
benzaldehyde added last, stirring time =
60 mins, temp = RT.
E Order of addition: benzaldehyde last
with argon flush, stirring time = 60 mins,
temp = RT, mmol piperidine = 20 mmol.
F Order of addition: benzaldehyde last
temp = RT, mmol piperidine = 30 mmol.
G Order of addition: benzaldehyde last
with argon flush, stirring time = 120
mins, temp = RT.
H Order of addition: benzaldehyde last
temp = RT.
26
4.2.3. Gas Chromatography Mass Spectroscopy
Gas Chromatography Mass Spectrometry spectra were recorded on an Agilent 6890

gas chromatograph with split injection (sample volume: 1 μL, split ratio 20 mL min-1)
and a HP-5MS column (30 m x 0.25 mm, 0.25 μm film thickness). Sample preparation:
1 mg of sample in 1 mL of methanol (spiked with an internal standard, eicosane at 1
mg in 1 mL methanol). Helium was used at the carrier gas at a flow rate of 1 mL min-
1
. The gas chromatogram was coupled to an Agilent 5973 (El, 70 eV, TIC mode
scanning m/z 50-500) and injector port was set at 280 °C and transfer line at 280°C.
The oven temperature programme was 120 °C for 3 min, 20 °C min-1 to 170 °C, 170
°C for 1 min, 10 °C min-1 to 210 °C, 210 °C for 1 min, 20 °C min-1 to 250 °C, 250 °C
for 1 min. A 1 μL aliquot of the samples were manually injected with a split ratio of
20:1. The injector and the GC interface temperatures were both maintained at 280 °C
respectively. The MS source and quadrupole temperatures were set at 230 °C and 150
°C respectively. Mass spectra were obtained in full scan mode (50-550 amu).
Sample preparation: 100.00 mg of analytes were weighed accurately into a 100.00 mL

clear glass volumetric flask and diluted to volume with methanol to give a soluation
containing all components at 1000 μg mL-1. This solution was further diluted with
methanol and 5 mL of eicosane (1.0 mg mL-1 in methanol).
27
5. Results and Discussion
5.1. Synthesis
Using a modified version of Le Gall’s experimental method,41 the synthesis is an

example of a three-component Barbier-type reaction, shown in Scheme 4, whereby
benzyl bromide, benzaldehyde and piperidine are reacted in a single reaction vessel.
To create the organozinc Grignard reagent, a small amount of benzyl bromide and
trifluoroacetic acid (TFA) are added to the reaction vessel containing zinc powder
suspended in acetonitrile, and allowed to stir for 5 minutes. The remaining benzyl
bromide, benzaldehyde and piperidine were added, where a condensation reaction
occurs generating an iminium ion in situ, which subsequently reacts with the
organozinc reagent via nucleophilic addition, forming the free-based diphenidine
complex. The reaction is left to stir for a further hour to ensure reaction completion,
leaving no un-reacted precursors in the reaction mixture. To purify the crude
diphenidine, an acid-base work up was employed rather than the alternative
purification over a neutral alumina column, which yielded a yellow-colourless oil. The
oil was then dissolved in diethyl ether before HCl in dioxane (4 M) was added,
dropwise, to isolate the hydrochloride salt. Upon the addition of the hydrogen chloride,
there were two potential scenarios that followed. Either, a white solid was produced
and no further purification was required or an oil was obtained instead. The oil is
suspected to form through dioxane becoming trapped within the matrix of the
diphenidine complex and in order to yield a powder, the oil was triturated several times
with cold acetone and then further purified by recrystallization using a solution of
diethyl ether and acetone (3:1) to remove any final impurities. After synthesis, one
sample was fully characterised and each subsequent sample was characterised through
GC-MS analysis.
28
Step 1
Step 2
Scheme 4. The detailed mechanism for the synthesis of diphenidine hydrochloride.
29
The infrared spectroscopy (IR) spectrum for diphenidine hydrochloride, shown in
Figure 7, illustrates the main components stereotypical of a diarylethylamine
compound, summarised in Table 2.
The peaks located between 2900 – 3030 cm-1 suggests the presence of the ethyl linker,
whilst peak 3644 cm-1 identifies that the piperidine molecule has been successfully
incorporated. Although the IR spectrum substantiates that diphenidine has been
synthesised, it is the peak at 2365 cm-1 that indicates the hydrochloride salt rather than
the free base.
Table 2. The infrared bands identified in diphenidine hydrochloride.
Peak (cm-1) Band

3644 N-H (stretch)
3030 Alkyl C-H (stretch)
2933 C-H (stretch)
2365 NH+ (stretch)
1604 Aromatic C=C (bend)
1494 Aromatic C-C (stretch)
N – H (stretch)
C – H (stretch)
N – H (stretch) NH+ (stretch)
Alkyl C - H
(stretch)
C – H (stretch)
NH+ (stretch)
Alkyl C - H Aromatic C=C

(stretch) (stretch)
Aromatic C=C
(stretch)
Figure 7. The infrared spectrum for diphenidine hydrochloride.
30
The 1H-NMR for diphenidine (8), shown in Figure 8, shows three identifiable
components of the diarylethylamine structure, and are consistent with the literature
values reported.6
sp2 region
a, i
e
i
f
a
h
e d g
i g, h
f Amine region
a c e
b h
d g
g, h
a c
CD2Cl2
e
b
sp3 region
d
bc
Amine salt peak
f
d
bc
14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
f
Chemical Shift (ppm)
Figure 8. 1H-NMR (400 MHz, CD2Cl2) spectrum of diphenidine hydrochloride.
The broad amine salt peak located at 12.5 ppm, confirms that the piperidine precursor
has been incorporated successfully and with the aromatic region, located between 6.9
and 7.5 ppm having an integral of 10 protons, confirms that diphenidine has been
synthesised. With nitrogen being more electronegative in comparison to carbon,
electron density is drawn towards to the nitrogen consequently increasing the chemical
shift of the surrounding protons. Despite there being scattered data within the aromatic
region making identification of overlapping multiplets difficult, the aromatic protons
of the benzyl bromide precursor will be found with a lower chemical shift, 6.9 ppm,
whilst the benzaldehyde aromatic protons are located at 7.5 ppm. The alkyl protons
located on the ethyl linker appear to be equivalent as they are attached to the same
carbon and exhibit geminal/J2 coupling with each other, but also vicinal/J3 coupling
with the proton attached to the chiral carbon. Being in close proximity to the
electronegative nitrogen, the protons are no longer equivalent and therefore produce
three separate peaks, shown in Figure 8, at 3.3 ppm, 3.9 ppm and 4.1 ppm. The NMR
spectrum also shows two triplets corresponding to the environments within the amine
31
function. With there being 3 different proton environments, a range of J coupling, i.e.
J3 and J4, producing the splitting pattern exhibited.
The 13C-NMR, shown in Figure 9, also coincides with the literature values reported
and reinforces the evidence in the 1H NMR. In comparison with the 1H NMR, there
are now two extra peaks present, from the quaternary carbons on the aromatic rings,
whilst there is no peak present for the N-H group.
γ
β β
α α
CD2Cl2
2
1
Cβ
C1 Cα C2
Cγ
170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

Chemical Shift (ppm)
Figure 9. 13C-NMR (100 MHz, CD2Cl2) spectrum of diphenidine hydrochloride.
The majority of carbons seen in the spectrum are quaternary carbons, which are shown
between 128 ppm and 132 ppm. The clustering of the peaks proves difficult to assign,
however, they do confirm that there are several aromatic systems present. The aliphatic
carbons within the spectrum are more interesting, revealing more insight into the
amine region. The chiral carbon can easily be identified at 73 ppm, again being pushed
downfield as it is located in extremely close proximity to the electronegative nitrogen
atom. The subsequent carbons on the piperidine ring are found in decreasing chemical
shifts the further they are away from the nitrogen atom.
32
5.2. Gas-Chromatography Mass-Spectrometry (GCMS)
Gas-chromatography is an analytical technique that is usually coupled with mass

spectrometry in order to determine the exact composition of a test sample. The
dissolved sample, typically in a volatile solvent such as methanol, is injected into the
GC column and is carried through by an inert gas, for example helium. Theoretically,
depending on the chemical properties of the sample, the compounds elute separately
from the column producing a retention time, then allowing the mass spectrometer to
subject the compounds to ionization energy and to detect their fragments separately.
The use of these two methods in conjunction with each other allows for a finer degree
of sample identification and is now an incredibly useful application in a range of
disciplines.
Diphenidine is identified at a single peak with a retention time of 14.56 minutes as

shown in Figure 10, (a pre-bought standard has been used to produce the spectrum
shown in Figure 10). Compound (8) undergoes fragmentation initially through the α-
cleavage of the ethyl linker, resulting in an iminium ion. This particular fragment has
been identified as the base peak within the mass-spectrum with a m/z = 174. Upon
formation of this iminium species, a secondary fragmentation occurs, cleaving the
Figure 10. A chromatograph showing the retention times of N-benzylpiperidine, eicosane and (8).
33
C=N bond forming a stable aromatic ion, the tropylium cation with a m/z = 91, shown
in Scheme 5.
With the secondary fragmentation producing a highly stable ion (in comparison to the
first iminium species), it is seen in a relatively high abundance in the mass spectrum,
which is shown in Figure 11. Despite this stable fragmentation leading to the tropylium
cation, there are alternative fragments formed, shown in Figure 11, accounting for
peaks at m/z = 181, m/z = 165 and m/z = 55.
However, in several diphenidine samples, particularly in method G, there has been an

impurity seen, having a retention time of 5.70 minutes as shown in Figure 10, which
has been identified as N-benzylpiperidine hydrochloride. The fragmentation that
occurs now produces the tropylium ion, m/z = 91 in the mass spectrum, as the most
stable peak in the fragmentation, followed by the benzylpiperidine ion at the second
highest in abundance, m/z = 174, shown in Figure 12. The suspected fragmentation
pattern of diphenidine is detailed in Scheme 5. It is suspected that this fragment forms
as the major impurity due to the chiral carbon being the most susceptible point of
attack. As nitrogen is more electronegative than carbon, electron density is donated
towards the nitrogen, causing the carbon to be more susceptible to attack, in this
instance, by electron impact (EI) ionisation energy.
Figure 11. The mass-spectrum for diphenidine hydrochloride, that has a retention time of 14.56
minutes and shows distinct peaks at m/z = 174 and m/z = 91.
34
Figure 12. The mass-spectrum for N-benzylpiperidine, now showing m/z = 91 as the base peak.
35
A – Formation of the tropylium ion from diphenidine hydrochloride.
A – Formation of the tropylium ion from diphenidine hydrochloride.
B – other possible fragmentation of diphenidine hydrochloride to form prominent ions present in the mass
spectrum.
B – other possible fragmentation of diphenidine hydrochloride to form prominent ions present in the mass
spectrum.
Scheme 5. The suspected fragmentation pathway for diphenidine hydrochloride. The first
fragmentation, A, shows how the two most abundant peaks from the diphenidine hydrochloride mass-
spectrum is formed. The second fragmentation, B, shows the alternative fragments that are visible
within the spectrum for diphenidine hydrochloride but also N-benzylpiperidine.
36
5.3. Quantitative Gas-Chromatography Validation Method
Instrumentation: GC-MS analysis was performed using an Agilent 6850 GC and

MS5973 mass selective detector (Agilent Technologies, Wokingham, UK). The
mass spectrometer was operated in the electron ionisation mode at 70 eV.
Separation was achieved with a capillary column (HP5 MS, 30 m Å~ 0.25 mm
i.d. 0.25 μm) with helium as the carrier gas at a constant flow rate of
1.0 mL min-1. The oven temperature programme was 120 °C for 3 min, 20°C
min-1 to 170°C, 170°C for 5 min, 10°C min-1 to 210°C, 210°C for 1
min, 20°C min-1 to 250°C, 250°C for 1 min. A 2 μL aliquot of the
samples were injected (manually) with a split ratio of 100:1. The injector
and the GC interface temperatures were both maintained at 280°C
respectively. The MS source and quadrupole temperatures were set at 230°C
and 150°C respectively. Mass spectra were obtained in full scan mode
(50–550 amu). All samples were injected six times.
Sample preparation: 100.0 mg of analytes were weighed accurately into a

100.0 mL clear glass volumetric flask and diluted to volume with methanol
to give a solution containing all components at 1000 μg mL−1. This
solution was then further diluted with methanol and 5 mL of eicosane (1.0
mg mL-1 in methanol) added (in each case) to give calibration standards
(100 mL) containing 50.0 μg mL−1, 100.0 μg mL−1, 150 μg mL−1, 200.0 μg
mL−1 and 250.0 μg mL−1 of each analyte and the internal standard at 25 μg
mL-1.
37
Calibration Curve of N-benzylpiperidine
100
90
80
Peak Area Ratio
y = 0.402x - 12.823
70
R² = 0.9958
60
50
40
30
20
10
0
0 50 100 150 200 250 300
Concentration (Impurity, ug/mL)
Figure 13. A calibration curve for the impurity, N-benzylpiperidine, using standards of 50 ug mL-1, 100
ug mL-1, 150 ug mL-1, 200 ug mL-1 and 250 ug mL-1 and the internal standard, eicosane, at 25 ug mL-1.
Calibration Curve of Diphenidine

120
110
100
Peak Area Ratio
90 y = 0.5057x - 19.394
80 R² = 0.9904
70
60
50
40
30
20
10
0
0 50 100 150 200 250 300
Concentration (Diphenidine, ug/mL)
Figure 14. A calibration curve for diphenidine using standards of 50 ug mL-1, 100 ug mL-1, 150 ug mL-
1
, 200 ug mL-1 and 250 ug mL-1 and the internal standard, eicosane, at 25 ug mL-1.
38
Table 3. Summary of validation data for the quantification of diphenidine hydrochloride (100 µg mL -
1
) and N-benzylpiperidine (100 µg mL-1).
Analyte Diphenidine Impurity

tR(min) 15.49 5.84
RRTa 0.95 0.36
RRFb 1.16 0.86
LODc 6.74 1.07
LOQd 20.42 30.83
Co-efficient (r2) 0.9904 0.9958
Precision (RSD n=6)
50 ug mL-1 7.80 10.26
100 ug mL-1 3.68 3.18
150 ug mL-1 3.47 1.79
200 ug mL-1 3.08 3.50
250 ug mL-1 1.06 3.08
a
Relative retention time.
b
Relative response factor.
c
Limit of detection (based on the standard deviation of the response and the slope).
d
Limit of quantification (base on the standard deviation of the response and the
slope).
39
5.4. Percentage Yield
For batches A1, A2, and A3, the synthesis followed the experimental procedure
outlined in 4.2.1. method A. It was noted that during the addition of the chemicals,
there was little change to the reaction mixture and despite being an exothermic
reaction, proceeded to generate little, if any, heat. Once the final compounds were
yielded, in very low amounts ~0.46 g, 17%, they were analysed via GC-MS and were
shown to contain a significant amount of a single impurity, identified at 5.70 mins
alongside the diphenidine complex, seen at 14.60 mins.
Table 4. All the parameters that were changed during phase 1 and 2, involving the order of addition
and quantities of precursors, stirring time and inert atmospheres.
Batch Parameter change
1. Sigma Aldrich
Order of addition, without argon
A 2. Alfa Aesar
Batch flush Parameter
and piperidine added last
change
3. Acros Organics
1. Sigma Aldrich
Order of
Order of addition,
addition, without
without argon
argon
A B 2. Alfa Aesar
flush,and
flush benzaldehyde added last
piperidine added last
3. Acros Organics
Order of addition, with argon flush,
C Order of addition, without argon
B piperidine added last
flush, benzaldehyde added last
D Order of addition, with argon flush,
C benzaldehyde added last
piperidine added last
20 mmol piperidine, 1 h stir,
E
D benzaldehyde added last, with argon
benzaldehyde added last
flush
E benzaldehyde added last, with argon
F benzaldehyde added last, with argon
flush
flush
2 h stirring, with argon flush,
G
F benzaldehyde added last, with argon
flush
0.5 h stirring, with argon flush,
H 2 h stirring, with argon flush,
G benzaldehyde added last
0.5 h stirring, with argon flush,
H
Table 4. All the parameters that were changed during phase 1 and 2, involving the order of addition
and quantities of precursors, stirring time and inert atmospheres.
40
The results from method A prompted the question whether the physical parameters,
such as order of addition and the presence of moisture/oxygen, severely affected the
yield and contributed to the production of contaminated samples. This led to changing
the physical parameters within the experimental procedure, detailed in Table 4. This
was not only to investigate the change in yield and to determine an optimised method
for the synthesis of diphenidine, but also to investigate whether the isotopic
fractionation would indeed change too.
The first physical parameter to change was the order of addition of the precursors. In
batches A1, A2, and A3, piperidine was added last as this was suspected to initiate the
reaction. However, changing the order of addition and adding the benzaldehyde last
increased the yield by 70% on average, over the six repeats (n=6). During the addition,
a self-induced reflux (reaction is exothermic) was observed along with a colour change
of grey to green-yellow. Upon GC-MS analysis of the samples, there were no
impurities seen; only a single peak was detected for the diphenidine sample at 14.56
mins, and through peak area normalisation calculations, the final compound was
determined to be 100% diphenidine.
An example on how to calculate the purity using peak area normalisation (PAN):
𝑃𝑒𝑎𝑘 𝐴𝑟𝑒𝑎 𝐷𝑖𝑝ℎ𝑒𝑛𝑖𝑑𝑖𝑛𝑒

Peak Area Ratio Diphenidine = 𝑃𝑒𝑎𝑘 𝐴𝑟𝑒𝑎 𝐸𝑖𝑐𝑜𝑠𝑎𝑛𝑒
𝑃𝑒𝑎𝑘 𝐴𝑟𝑒𝑎 𝐼𝑚𝑝𝑢𝑟𝑖𝑡𝑦

Peak Area Ratio Impurity = 𝑃𝑒𝑎𝑘 𝐴𝑟𝑒𝑎 𝐸𝑖𝑐𝑜𝑠𝑎𝑛𝑒
𝑃𝑒𝑎𝑘 𝐴𝑟𝑒𝑎 𝑅𝑎𝑡𝑖𝑜 𝐷𝑖𝑝ℎ𝑒𝑛𝑖𝑑𝑖𝑛𝑒

% Diphenidine = (𝑃𝑒𝑎𝑘 𝐴𝑟𝑒𝑎 𝐷𝑖𝑝ℎ𝑒𝑛𝑖𝑑𝑖𝑛𝑒+𝑃𝑒𝑎𝑘 𝐴𝑟𝑒𝑎 𝑅𝑎𝑡𝑖𝑜 𝐼𝑚𝑝𝑢𝑟𝑖𝑡𝑦) 𝑥 100
𝑃𝑒𝑎𝑘 𝐴𝑟𝑒𝑎 𝑅𝑎𝑡𝑖𝑜 𝐼𝑚𝑝𝑢𝑟𝑖𝑡𝑦

% Impurity = (𝑃𝑒𝑎𝑘 𝐴𝑟𝑒𝑎 𝑅𝑎𝑡𝑖𝑜 𝐷𝑖𝑝ℎ𝑒𝑛𝑖𝑑𝑖𝑛𝑒+𝑃𝑒𝑎𝑘 𝐴𝑟𝑒𝑎 𝑅𝑎𝑡𝑖𝑜 𝐼𝑚𝑝𝑢𝑟𝑖𝑡𝑦) 𝑥 100
Grignard reactions generate better yields when performed under inert conditions and
with thoroughly dried glassware. For methods A and B, the reactions were performed
in an open reaction vessel and with the Barbier-type reaction being similar to that of a
Grignard, it was considered that moisture/oxygen may cause oxidation of the
precursors or intermediate therefore further affect the percentage yield and purity of
diphenidine. Both methods, C and D, were carried out under argon with C being where
41
piperidine was added last whilst D was when benzaldehyde was added last. The
comparison between the two batches wasn’t as substantial as changing the order of
addition. There was only a 6% difference in percentage yield between the two,
however, comparing methods A and C, there’s an increase in percentage yield (by
20%) but more importantly, method C when analysed via GC-MS, showed no
impurities, whereas A showed a single impurity at ~5.70 mins. This may be crucial
when analysing the isotopic fractionation data, see 5.4.2 effect of order of addition in
air. Conversely, methods B and D showed no impurities upon analysis, with both being
100% diphenidine.
As the zinc present is zinc(II), and therefore in a d10 configuration, it can coordinate
with piperidine up to four times in either square planar or distorted tetrahedral
geometry, and thus obey the 18-electron rule. It was suspected that increasing the
number of moles of piperidine in the reaction, would have a considerable impact on
the percentage yield. This hypothesis resulted in methods E and F containing 20 mmol
of piperidine and 30 mmol piperidine respectively. The reactions were performed
using the physical parameters that had produced the best results so far in the
investigation, see 4.2.5. and 4.2.6. method E had a percentage yield of 38% on average
over the replicates whilst method F, had a percentage yield of 64% on average over
the replicates. Despite the obvious difference in yield, the GC-MS data showed that
method E produced pure products in comparison to method F, had a low yet detectable
impurity, at 5.77 mins relating to 1.75% of the overall sample.
F. Gyenes et al investigated the effect of time on the percentage yield of Barbier-type

reactions and concluded that “reaction times greater than 30 min resulted in only a
modest increase in the yield and at 60 min the yield decreased considerably”. 42 It was
determined by prolonging the reaction time that the benzyl bromide is fully consumed
but is also accompanied by the decomposition of the reactive imine intermediate.
Consequently, this resulted in a decreased yield with an increased reaction time.
From these findings and considering what conditions could be used in clandestine
laboratories, the stirring time for the final two batches were changed, from 1 hour to 2
hours and ½ hour respectively for methods G and H. Despite both batches having
consistently low yields, 30% and 28% respectively, there were considerable
42
differences in the texture and appearance of the final compounds. All 6 replicates
making up method G were yellow in colour and grainy, indicating impurities within
the sample. This was confirmed by GC-MS data showing two peaks, again one at 5.75
mins with m/z = 91 and 13.86 mins with m/z = 174. Method H samples were all,
however, white crystalline powders and proved to be pure with single peaks being
obtained at ~14.68 mins from GC-MS data.
From evaluating all GC-MS data, taking into account of the percentage yield data and
peak area normalisation calculations, the most optimised method was found to be
method F, see 4.2.6. Using 30 mmol piperidine rather than 10 mmol or 20 mmol but
also keeping with the 1 hour stirring time, diphenidine hydrochloride was produced in
much higher quantities, with a percentage yield of 64% with only 1.75% being the
impurity, N-benzylpiperidine.
Clandestine laboratories could use any of the methodologies in order to synthesise

diphenidine but would rely on not only the experience of the chemist, quality of the
work environment but also the cost of supplies. The methods mentioned require
specialist apparatus i.e. the use of argon/inert gas, which would mean the use of a
laboratory environment in order to achieve a high yield and purity in which case, it
could be sold as a “premium” product with potentially a higher price. Alternatively, if
the standard method publicised were used to simply product the drug on large scale
for profit, the product is likely to contain a higher percentage of impurities, produced
at a quicker rate with less concern to the final product.
43
5.5. Isotopic Ration Mass Spectrometry
5.5.1. Natural Abundances and Delta
To establish or identify a “chemical fingerprint” for diphenidine hydrochloride, the

ratios of the stable isotopes of both carbon and nitrogen, 13C/12C and 15N/14N can be
measured. On a global scale, the natural isotopic abundances of these elements remain
relatively stable, however, subtle variations do occur during biological, chemical and
physical processes. In the case of diphenidine hydrochloride (8), the variation could
occur during the synthesis and by applying IRMS, it is possible to measure any
potential changes in the isotopic abundance of both carbon and nitrogen. This variation
will be unique to the origin and history of diphenidine, which could help us link the
product to precursor. Isotopic ratios, at natural abundance levels, are measured relative
to international standards which define the measurements scale for particular
isotopes.42
Variations in the natural abundance of stable isotopes are expressed using delta
notation as shown in the following equations:
𝑎𝑏𝑢𝑛𝑑𝑎𝑛𝑐𝑒 𝑜𝑓 𝑡ℎ𝑒 ℎ𝑒𝑎𝑣𝑦 𝑖𝑠𝑜𝑡𝑜𝑝𝑒

Ratio (R) =
𝑎𝑏𝑢𝑛𝑑𝑎𝑛𝑐𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑙𝑖𝑔ℎ𝑡 𝑖𝑠𝑜𝑡𝑜𝑝𝑒
𝑅𝑠𝑎𝑚𝑝𝑙𝑒
δ=( − 1)
𝑅𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑
Delta values are commonly multiplied by 1000 so that they are reported in parts per
thousand (‰) or by 1000,000 to give results in parts per million, ppm.
44
5.5.2. Batches A1, A2 and A3, effect of changing supplier
Based on previous research published involving IRMS analysis on laboratory

synthesised mephedrone samples11, it was theorised that future work could possibly
reach to precursor and/or manufacture differentiation. To apply this to diphenidine
hydrochloride, the precursors were sourced from 3 different manufactures: Sigma-
Aldrich, Alfa-Aesar and Acros Organics, with each batch (A1, A2 and A3) being
synthesised using only the precursors from the same supplier. It is predicted that if all
three suppliers source their precursors from different manufacturers, then upon IRMS
analysis, there would be three separate and distinguishable cluster patterns. The data
from the IRMS analysis is shown in Figure 15.
A1
A2
δ15N
A3
δ13C
Figure 15. A graphical representation of the clustering patterns of the starting material batches (A1,
A2 and A3) after IRMS analysis has been performed. From this, it is apparent that there is a cluster
pattern associated with the Sigma-Aldrich starting materials, batch A1.
δ13C
From looking at the data, there is only one distinguishable pattern of clusters and these
are associated with the Sigma-Aldrich batch, A1 (DSA2-DSA6 on Figure 15). From
the array of clusters, there seems to be little difference in the δ15N values, but a more
comparable difference in the δ13C values.
45
A2 and A3 corresponding to the Alfa-Aesar and Acros Organics batches and have
relatively similar values for both δ15N and δ13C values, all except two outliers, DA01
and DA05 which are both significantly different from the majority.
It is interesting to note that both Acros Organics and Alfa-Aesar are both part of the
Fischer Scientific brand and, therefore, could potentially source their precursors from
the same manufacturer. If the precursors are from the same manufacturer, the relative
isotopic abundances of the elements would essentially be the same, if not, marginally
different. This would account for the similarities in both of the δ13C and δ15N values,
along with the both the carbon and nitrogen percentages. However, Sigma-Aldrich is
a separate brand and could consequently have a different provider of chemicals. This
could potentially explain why batch A1 can be distinguished from A2 and A3, but A2
and A3 cannot be distinguished from each other.
Tables 5, 6 and 7 show the actual values for both δ13C and δ15N along with the
percentages of carbon and nitrogen, for each of the batches. It is a common trend that
the nitrogen percentages are consistently low throughout, ranging from only 4.2 – 4.8
%, despite all carbon percentages remaining comparably high, ranging from 66.6 –
76.4 % across all of the starting material samples. The low nitrogen percentages could
be due to not only the amount of piperidine used, but also the order of addition.
For the reaction to proceed, the organometallic Barbier reagent needs to be present in
abundance and is created through reacting zinc with piperidine, forming a stabilised
complex allowing the piperidine reagent to fully react, leaving behind no unreacted
starting material. During all three batches, A1, A2 and A3, the order of addition meant
piperidine was added last, as it was originally suspected to initiate the reaction. This
suggests that the stabilised complex wasn’t formed, or not in a large enough amount
for the reaction to yield a significant yield, or a clean product, supported by both the
percentage yield results, ~17% on average over all the samples in batch A, and the
GC-MS data, where N-benzylpiperidine was seen in high abundances in addition to
diphenidine.
46
Table 5. The processed data regarding samples DSA02-DSA06, relating to batch A1. All precursors
from the same manufacturer, Sigma-Aldrich.
Sample δ13C %C δ15N %N

DSA02 -24.5 70.6 3.6 4.4
DSA03 -24.5 71.9 3.9 4.5
DSA05 -24.7 72.2 3.7 4.5
DSA06 -24.7 70.5 3.9 4.4
Table 6. The processed data regarding samples DA1-DA6, relating to batch A2. All precursors from
the same manufacture, Alfa-Aesar.

DA1 -25.2 76.4 3.6 4.8
DA3 -25.3 73.8 3.9 4.6
DA5 -25.6 66.6 3.2 4.2
DA6 -25.5 68.7 3.4 4.3
Table 7. The processed data regarding samples DAO1-DAO6, relating to batch A3. All precursors
from the same manufacturer, Acros Organics.

DAO1 -24.4 71.2 1.5 4.2
DAO2 -25.6 72.9 4.3 4.7
DAO3 -25.3 73.1 3.2 4.5
DAO5 -24.4 74.1 1.8 4.4
DAO6 -25.5 72.3 4.3 4.7
47
5.5.3. Methods A and B, effect of order of addition in air
Method A’s replicates were all synthesised without the use of argon, but in a reaction
vessel, open to air. Method A was also prepared using benzyl bromide, benzaldehyde
and piperidine, added individually in that order. Upon the synthesis of method A, it
was noted that the reaction proved to be extremely temperamental, rarely reaching
completion and producing minimal final product, with the average being only ~0.45
g, or 17%. To increase yield and the purity of (8), the order of addition was questioned
first. By keeping the supplier, the same throughout methods B-H (Acros Organics), it
would be possible to draw conclusions accurately if any valuable IRMS was obtained.
BB-BA-P
(28:11:10)
δ15N
B
BB-P-BA
(28:10:11)
δ13C
Figure 16. A graphical representation of the clustering patterns of batches A and B after IRMS
analysis has been performed. The order of addition has been changed but the reaction is still performed
in an open vessel, with the same molar equivalents.
δ13C
Figure 16 shows the effect of changing the order of addition but keeping the reaction
open to air, comparing method A(3) and method B. By adding piperidine, second,
rather than third, theoretically there would be a higher probability of forming the
stabilised Zn/Br intermediate, and consequently a higher yield, which is confirmed by
percentage yield calculations showing an increase of ~70% between methods A and
B. By changing the order of addition, it was expected that the only change would be
48
to the yield, and not to the isotopic ratio. However, changing the order of addition,
from the initial benzyl bromide, benzaldehyde and piperidine to benzyl bromide,
piperidine and then benzaldehyde, has shown to affect the δ15N values, more so than
the δ13C values, shown in tables 7 and 8.
The δ15N are shown to decrease upon changing the order of addition, from 4.3 in batch
A, to as little as 1.0 in batch B, however this decrease has minimal effect on the
percentage of nitrogen incorporated into the final product, which remains stable at ~4.7
% throughout both methods. The decrease in δ15N values is a result that was not
anticipated, however, performing the reaction under argon, methods C and D, will give
more insight into this change. Despite the variance in their δN15 values, their δ13C
values stay relatively constant with a range of only 0.5 across all samples, shown by
samples DAO3 and 6-S2-P2 in Figure 16. From these results alone, it is possible to
distinguish between methods A and B.
Table 8. The processed data regarding samples composing of batch B, whereby the order of addition
has changed, consequently decreasing the δ15N values.

B1 -25.1 72.4 1.2 4.6
B2 -25.1 73.0 1.7 4.6
B3 -25.1 76.0 1.6 4.8
B4 -25.2 72.3 1.5 4.6
B5 -25.2 74.3 1.0 4.7
B6 -25.4 75.2 1.0 4.8
49
5.5.4. Methods C and D, effect of order of addition under argon
There has been an increase in δ13C values methods for methods C and D in comparison to
batches A and D, showing an overall increase of 0.6, which can be seen in Figure 17 and
Tables 9 and 10. Both methods C and D show a condensed region of clustering on the
Y axis but not the X, whereby the majority of samples are located, suggesting
replicable data. This overall pattern follows the initial hypothesis that the δ15N value
would increase upon performing the reaction in an inert atmosphere.
BB-P-BA
(28:10:11)
δ15N
BB-BA-P
(28:11:10)
δ13C
Figure 17. A graphical representation of the clustering patterns of methods A and B after IRMS
analysis has been performed. It is possible to distinguish between both batches due to changes in the
δ15N values.
δ13C
Perhaps the increase observed for both methods, in comparison to A and B, is through
the amount of oxidation occurring upon exposure to air, a variable in this instance is
now controlled that has not previously. Controlling the exposure to air reduces the
amount of oxidation occurring upon the nitrogen element of the piperidine molecule,
meaning that there is a higher percentage remaining in the reaction mixture, in
comparison to that of the exposed and potentially oxidised piperidine in batches A and
B.
50
Through changing the order of addition so that piperidine is added second, there is
additional time for piperidine and benzyl bromide to react, resulting in a higher
probability of the stabilised intermediate forming. In terms of percentage yield, there
is only a 6% difference between C and D and it is also interesting to note that both
batches B and D, whereby the benzaldehyde was added last, show no impurities (N-
benzylpiperidine) in their GC-MS results but are in fact 100% pure samples of
diphenidine. All analytical data concerning the initial batches, A, B, C and D, support
the original suspicion that the order of addition effects the progress of the reaction and
consequently the yield.
Table 9. Processed data regarding samples composing of batch C, whereby the order of addition is the
same as batch A, but is now performed under argon.

C1 -25.7 73.1 1.1 4.6
C2 -25.8 74.6 1.0 4.6
C3 -25.9 73.0 1.1 4.5
C4 -26.0 72.1 1.0 4.4
C5 -26.1 72.5 1.1 4.5
C6 -25.9 73.9 1.1 4.6
Table 10. The processed data regarding samples composing of batch D, whereby the order of addition
is the same as batch B, but is now performed under argon.

D1 -25.9 72.0 4.2 4.5
D2 -25.8 71.4 4.4 4.4
D3 -25.9 70.2 4.4 4.4
D4 -26.0 67.7 4.2 4.2
D5 -25.8 69.1 4.3 4.3
D6 -26.0 69.9 4.3 4.3
51
5.5.5. Batches E and F, effect of mmol piperidine
During the synthesis of method A, it was suggested that piperidine was the component
to initiate the reaction. However, upon performing the synthesis, it was decided that in
order for the reaction to proceed piperidine needs to be added second. In addition to
this, it was hypothesised that zinc would not be fully coordinated by using only 10
mmol of piperidine, therefore methods E and F investigate the effect of increasing this
to initially 20 mmol and then 30 mmol. Although the percentage yield of method E
was low, 38%, in comparison to F, 64%, GC-MS data identified the products to be
clean samples, with F having only a slightly detectable impurity, N-benzylpiperidine
which related to just 1.75% of the sample.
BB-P-BA
F (28:30:11)
BB-P-BA
(28:20:11)
BB-P-BA
(28:10:11)
Figure 18. A graphical representation of the clustering patterns of batches B, E and F after IRMS
analysis has been performed. Upon doubling the equivalents of piperidine, there is a significant
increase in δ15N values followed by a more gradual increasing on increasing 20 mmol to 30 mmol.
Figure 18 shows the IRMS data comparing methods B, E and F. Changing the mmol
of piperidine not only effects the δ15N values but also influences the δ13C values.
Starting with
Figure 18. A batch B,representation
graphical where the mmol of piperidine
of the clustering remains
patterns only
of batches 10and
B, E mmol,
F afterthe δ15N
IRMS
analysis has been performed. Upon doubling the equivalents of piperidine, there is a significant
values are significantly lower than that of E and F, with values as low as 1.0, shown in
increase in δ15N values followed by a more gradual increasing on increasing 20 mmol to 30 mmol.
Table 8. The δ13C values stay relatively constant, with an average of -25.9. However,
upon doubling the equivalent up to 20 mmol, there is a sudden, noticeable increase in
52
the δ15N values, ranging from 4.9 – 5.9, with δ13C values now lower, ranging from -
25.0 to -25.2, shown in Table 11. This evidence supports the hypothesis that more
piperidine is coordinated with zinc, incorporating more nitrogen consequently
increasing not only the δ15N values through the nitrogen element, but also the δ13C
values from the carbon ring. Increasing the equivalents from 20 mmol to 30 mmol
shows a decrease of 0.5 from E to F for δ13C and then an increase of 1.0 for δ15N
values, shown in Table 12 and Figure 17, perhaps now there is steric hindrance from
the piperidine already coordinated, with a lower probability of more coordinating to
create the square planar or distorted tetrahedral geometry.

E1 -25.4 76.5 4.9 4.8
E2 -25.3 73.9 5.9 4.5
E3 -25.4 75.1 5.1 4.7
E4 -25.4 75.1 4.9 4.7
E5 -25.4 75.8 5.4 4.7
E6 -25.5 77.1 5.4 4.8
Table 11. The processed data regarding samples composing of batch E, whereby increasing the
equivalent of piperidine to 20 mmol has increased the δ15N values by 4.1.

F1 -25.0 76.5 5.6 4.8
F2 -25.1 73.9 6.9 4.5
F3 -25.0 75.1 5.7 4.7
F4 -25.1 75.1 5.8 4.7
F5 -25.2 75.8 5.6 4.7
F6 -25.1 77.1 6.5 4.8
Table 12. The processed data regarding samples composing of batch F, whereby increasing the
equivalent of piperidine to 30 mmol has increase the δ15N values by 5.9 from batch B and 2.0 from
batch F.
53
5.5.6. Batches G and H, effect of stirring time
Gyenes et al concluded that in Barbier-type reactions that “reaction times greater than
30 min resulted in only a modest increase in the yield and at 60 min the yield decreased
considerably”.43 By prolonging the reaction time, the reactive imine intermediate
decomposes, resulting in a decreased yield and a higher percentage of impurities, i.e.
side reactions from Wurtz coupling. Both batches G and H, where the reaction times
have changed to 2 hours and ½ respectively. G was composed of yellow, grainy
samples with GC-MS data confirming the impurity, N-benzylpiperidine at 5.75 mins,
supporting the hypothesis of increase impurities and a decreased yield. In the case of
H, the samples were 100% diphenidine and were white crystalline powders, suggesting
that a decreased time gives a purer sample. It was predicted that these results would
affect the IRMS data, by decreasing the δ15N and δ13C values with an increase in time.
Figure 19 shows the effect of stirring time whilst under argon, between batches B, G
and H.
BB-P-BA
(28:10:11)
BB-P-BA
(28:10:11)
BB-P-BA
(28:10:11)
Figure 19. A graphical representation of the clustering patterns of batches B, G and H after IRMS
analysis has been performed. Upon halving the time, there is an increase in δ15N values and upon
halving, there is a significant decrease in δ15N values.
54
In comparison to method B, whose stirring time was just one hour, method H is as
predicted, with higher δ15N values which now range from 1.4 – 1.6. Although still
quite low to values previously seen, for example, in comparison to batch F, it is still
higher than both batches B and G. A shorter reaction time reduces the amount of
oxidation occurring upon the nitrogen element, if any, but also prevents the imine
intermediate being broken down. Comparing this to batch G, the δ15N values have
dropped to as little as 0.5, shown in Tables 13 and 14, suggesting that the majority of
the reactive intermediate is no longer present during the reaction mixture, reflected by
the percentage yield results. With less nitrogen being incorporated throughout the
synthesis due to decomposition of the imine intermediate, a necessity for the Barbier
reaction to proceed, there is less nitrogen available to successfully react. This data
supports the findings from Gyenes and has identified that in this circumstance, it is
possible to distinguish between chosen synthetic pathways in this instance.42
Table 13. The processed data regarding samples composing of batch G, whereby increasing the stirring
time to 2 hours causes a decrease in δ15N values, with the δ13C samples staying relatively constant.

G1 -26.0 72.4 1.5 4.6
G2 -26.1 73.6 1.6 4.7
G3 -26.0 73.8 1.6 4.7
G4 -26.3 73.6 1.6 4.7
G5 -26.1 73.5 1.4 4.7
G6 -26.3 73.9 1.6 4.7
Table 14. The processed data regarding samples composing of batch H, whereby decreasing the stirring
time to 0.5 hour causes an increase in δ15N values, with the δ13C samples staying relatively constant.

H1 -26.0 69.4 0.7 4.3
H2 -26.1 70.7 0.6 4.4
H3 -26.0 72.1 0.7 4.4
Table 13. The processed data regarding samples composing of batch H, whereby decreasing the stirring
-26.1 in δ15N values,
H4hour causes an increase
time to 0.5 with the δ13C samples
71.7 0.5 staying relatively4.4
constant.
H5 -26.1 70.5 0.6 4.3

H6 -26.1 71.5 0.5 4.4
55
6. Conclusion
From evaluating all GC-MS data, taking into account of the percentage yield data and
peak area normalisation calculations, the optimum method was found to be method F.
Using 30 mmol piperidine rather than 10 mmol or 20 mmol but also keeping with the
1 hour stirring time, diphenidine hydrochloride was produced in much higher
quantities, with a percentage yield of 65% with only 1.75% being the impurity, N-
benzylpiperidine. It is possible for clandestine laboratories to use any of the
methodologies mentioned to synthesise diphenidine hydrochloride, however, to
produce a product with few impurities whilst maintaining a profitable yield, the gross
input is consequently higher. A skilled chemist would be required along with a higher
quantity of precursor resulting in a larger initial cost. Alternatively, if the standard
method publicised by Le Gall et al 41 were used to simply produce the drug on large
scale for profit, the product is likely to contain a higher percentage of impurities,
produced at a quicker rate with less concern to the final product. Whichever method
chosen, would depend entirely on the on the supplier, hence why a library of different
physical parameters was created.
From the IRMS data, shown in Figure 15, there is an obvious difference in the δ13C
values between batches A1 compared to A2 and A3, but a minimal difference between
A2 and A3 individually. These two batches correspond to Alfa-Aesar and Acros
Organics, both of which are owned by the Fischer Scientific brand. If the two obtain
their chemicals from the same supplier, this could account for why there is little
difference in their IRMS data. If a seized sample of diphenidine was analysed, it would
be possible to identify between samples synthesised using Sigma Aldrich precursors
but not from that of Alfa or Acros. Through changing the order of addition, and when
the synthesis is performed under argon, there is an increase in δ15N values, in
comparison to when synthesised open to air, where the δ13C values decrease. From
this, it would be possible to distinguish whether a diphenidine sample had been
synthesised in an inert atmosphere or not. An interesting observation is that through
changing the equivalents of piperidine, there is a significant difference on the δ15N
values between batches B and E, relating to 10 mmol and 20 mmol, however, not
56
between 20 mmol and 30 mmol. It would be possible to determine whether the seized
sample was synthesised using 10 mmol or 20 mmol, but not with certainty.
Although some valuable data has been obtained involving IRMS as a form of source
identification, it does have its limitations in comparison to other analytical methods.
Despite giving more detailed results, it does take longer than methods such as GC-MS
and with only a limited library available currently, identification could still be lengthy.
Not only that, it hasn’t been possible to distinguish between all samples in this case
and therefore needs further investigation.
57
7. Future Work
Impurity profiling could be performed whereby the alternative publicised method is

used in order to synthesise diphenidine hydrochloride (8). The investigation should be
conducted again, by another, or group of individuals to identify whether the data is
replicable. In addition, all batches synthesised after batches A were prepared using
precursors from Acros Organics. It would be beneficial to conduct the synthesis using
Sigma-Aldrich and Alfa-Aesar precursors to draw more accurate conclusions between
any significant differences in the IRMS data obtained. Further studies could also focus
on the metabolites of diphenidine or by changing to an aldehyde used to create a new
library of NPSs, for which IRMS analysis could be used to distinguish between them.
58
8. References
1. C. Davidson and F. Schifano, Prog Neuro-Psychopharmacology Biol

Psychiatry, 2016, 64, 267-274.
2. R. P. Archer, R. Treble and K. Williams, Drug Test Anal, 2011, 3, 505-514.
3. F. Schifano, L. Orsolini, G. Duccio Papanti and J. M. Corkery, World
Psychiatry, 2015, 14, 15-26.
4. E. Papaseit, M. Farré, F. Schifano and M. Torrens, Curr Opin Psychiatry, 2014,
27, 243-250.
5. R. L. Luciano and M. A. Perazella, Nature Reviews Nephrology, 2014, 10, 314-
324.
6. J. Wallach, P. V. Kavanagh, G. McLaughlin, N. Morris, J. D. Power, S. P.
Elliott, M. S. Mercier, D. Lodge, H. Morris, N. M. Dempster and S. D. Brandt,
Drug Test Anal, 2015, 7, 358-367.
7. EMCDDA, Europol 2013 Annual Report on the implementation of Council
Decision, 2014.
8. EMCDDA, EU Drug Markets Report: A Strategic Analysis, 2013.
9. EMCDDA, European Drug Report: Trends and Development, 2015.
10. P. Adamowicz and D. Zuba, J. Forensic Sci, 2015, 60, S264-S268.
11. N. NicDaeid, W. Meier-Augensteinn, H. F. Kemp and O. B. Sutcliffe, Anal
Chem, 2012, 84, 8691-8696.
12. S. Gibbons, Clin Toxicol, 2012, 50, 15-24.
13. S. Gibbons and M. Zloh, Bioorg Med Chem Lett, 2010, 20, 4135-4139.
14. J. B. Murray, J psychol, 2002, 136, 319-327.
15. J. P. Smith, O. B. Sutcliffe and C. E. Banks, The Analyst, 2015, 140, 4932-
4948.
16. S. Chiappini, H. Claridge, J. M. Corkery, C. Goodair, B. Loi and F. Schifano,
Human Psychop, 2015, 30, 244-248.
17. F. Schifano, A. Albanese, S. Fergus, J. L. Stair, P. Deluca, O. Corazza, Z.
Davey, J. Corkery, H. Siemann, N. Scherbaum, M. Farre, M. Torrens, Z.
Demetrovics, A. H. Ghodse, L. Di Furia, L. Flesland, M. Mannonen, A.
Majava, S. Pagani, T. Peltoniemi, M. Pasinetti, C. Pezzolesi, A. Skutle, P. Van
Der Kreeft, A. Enea, G. Di Melchiorre, H. Shapiro, E. Sferrazza, C.
Drummond, A. Pisarska, B. Mervo, J. Moskalewicz, L. Floridi and L. S. Y.
Haugen, Psychop 2011, 214, 593-602.
18. K. Kudo, Y. Usumoto, R. Kikura-Hanajiri, N. Sameshima, A. Tsuji and N.
Ikeda, Legal medicine, 2015.
19. K. Hasegawa, A. Wurita, K. Minakata, K. Gonmori, H. Nozawa, I. Yamagishi,
K. Watanabe and O. Suzuki, Forensic Toxicol, 2015, 33, 45-53.
20. J. Hillebrand, D. Olszewski and R. Sedefov, Substance Use and Misuse, 2010,
45, 330-340.
59
21. M. Monaghan, Int J Drug Policy, 2014, 25, 1025-1030.
22. A. B. Petrenko, T. Yamakura, K. Sakimura and H. Baba, Eur J Pharmacol,
2014, 723, 29-37.
23. EMCDDA, 2011 Annual Report: The State of the Drugs Problem in Europe,
2011.
24. J. W. Johnson, N. G. Glasgow and N. V. Povysheva, Curr Opin Pharmacol
2015, 20, 54-63.
25. H. Morris and J. Wallach, Drug Test Anals, 2014, 6, 614-632.
26. V. H. Maddox, E. F. Godefroi and R. F. Parcell, J Med C, 1965, 8, 230-235.
27. J. B. Zawilska, Toxicol Lett, 2014, 230, 402-407.
28. C. L. Stevens, Parke-Davis, USA Patent Office, 1962
29. E. F. Domino, Anesthesiology, 2010, 113, 678-686.
30. EMCDDA, Europol 2010 Monitoring Report: The Implementation of Council
Decision, 2011.
31. EMCDDA, Europol 2013 Joint Report on a New Psychoactive Substance:
Methoxetamine, 2014.
32. M. L. Berger, A. Schweifer, P. Rebernik and F. Hammerschmidt, Bioorg Med
Chem, 2009, 17, 3456-3462.
33. A. Wurita, K. Hasegawa, K. Minakata, K. Watanabe and O. Suzuki, Forensic
Toxicol, 2014, 32, 331-337.
34. N. Gentile, R. T. W. Siegwolf, P. Esseiva, S. Doyle, K. Zollinger and O.
Delémont, Forensic Sci Int, 2015, 251, 139-158.
35. N. M. Beckett, D. I. Grice, J. F. Carter and S. L. Cresswell, Science and Justice,
2015, 55, 57-62.
36. Z. Muccio and G. P. Jackson, Analyst, 2009, 134, 213-222.
37. Z. Muccio and G. P. Jackson, J Forensic Sci, 2011, 56, S203-S209.
38. J. Casale, E. Casale, M. Collins, D. Morello, S. Cathapermal and S. Panicker,
J Forensic Sci, 2006, 51, 603-606.
39. J. R. Ehleringer, D. A. Cooper, M. J. Lott and C. S. Cook, Forensic Sci Int,
1999, 106, 27-35.
40. S. Schneiders, T. Holdermann and R. Dahlenburg, Science and Justice, 2009,
49, 94-101.
41. E. Le Gall, C. Haurena, S. Sengmany, T. Martens, M. Troupel, J Org Chem
2009, 74, 7970-7973.
42. J. Carter, C. Lock, W. Meier-Augenstein, H. Kemp, S. Scheniders, L. Stern, G.
van der Peijl, Good Practice Guide for Isotopic Ratio Mass Spectrometry,
2011, 1, 1
43. F. Gyenes, K. E. Bergmann, J. T. Welch, J Org Chem, 1998, 63, 2824-2828.
60

Thesis - Final Submission Diphedine

Uploaded by

Copyright:

Available Formats

Thesis - Final Submission Diphedine

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Thesis - Final Submission Diphedine

Uploaded by

Copyright:

Available Formats

“Using Isotopic Fractionation to Link Precursor to Product

in the Synthesis of Diphenidine Hydrochloride: A Tool for

Faye Hannah Owen

A thesis submitted in fulfilment of the requirements of the Manchester Metropolitan

The Faculty of Science and Engineering

Signature F. H. Owen Date: 30-Jan-2017

Diphenidine, is an example of a fourth generation NPS that has increased in popularity

From IRMS data it is possible to distinguish between sources of precursors except

The fundamental concept when initially designing an NPS is to expand upon a

2.2. Drug Classification and Legislation

Figure 2. Examples of second (NRG-1, 7) and fourth generation (diphenidine, 8 and

In 2013, the U.K. implemented an arylcyclohexylamine ban, causing recreational

2.3. Clinically Abused Drugs

Many pharmaceuticals, such as anaesthetics and benzodiazepines, are clinically

The most recognised pharmacological feature associated with dissociative agents is

NMDAR activity plays a vital role in physiological processes, for example

1-(1-phenylcyclohexyl)piperidine hydrochloride salt (11), also known as

The original analogue of (11), 1-(1-phenylcyclohexyl)amine (PCA, 12), shown in

Since PCP’s withdrawal as an anaesthetic, it became regarded as a hallucinogenic drug

An example of such a PCP derivative is the ketamine hydrochloride salt (2-(2-

Categorised in 1985 as an essential drug by the World Health Organisation (WHO),

Scheme 2. 1-hydoxycyclopentyl-(o-chlorophenyl)-ketone-N-methylamine (obtained via a Grignard

Ketamine users report feelings of “depersonalisation” once consumed, commonly

In terms of chemical structure, it remains an arylcyclohexylamine class and is

Initially marketed as a “research chemicals” or “not for human consumption”,10

Change in legislation through the restriction and banning of arylcyclohexylamine

Diphenidine hydrochloride was originally synthesised in 1924 via the Bruylants

Similar in structure to that of illicit dissociatives, it is not surprising that

Figure 6. The two enantiomers of diphenidine hydrochloride (8), S and R

The subtle differences in terms of structure that cause diphenidine hydrochloride to be

Through such evidence, it is being recommended that law enforcement agencies

2.6. Isotopic Ratio Mass Spectrometry

The fundamental concept of IRMS as an analytical technique is that it looks beyond

A notable example of such methods was in North-Korea, whereby 20 samples of

Another area of investigation using IRMS is linking precursors to product in the

Diphenidine (isolated as its corresponding hydrochloride salt) is isolated through a

4.2. The synthesis of diphenidine hydrochloride (8)

The specific methods for batches (B – H) are provided in the supplementary

4.2.2. Synthesis of N-benzylpiperidine hydrochloride

Method Physical Parameters

Gas Chromatography Mass Spectrometry spectra were recorded on an Agilent 6890

Sample preparation: 100.00 mg of analytes were weighed accurately into a 100.00 mL

Using a modified version of Le Gall’s experimental method,41 the synthesis is an

Scheme 4. The detailed mechanism for the synthesis of diphenidine hydrochloride.

Table 2. The infrared bands identified in diphenidine hydrochloride.

Peak (cm-1) Band

3030 Alkyl C-H (stretch)

2933 C-H (stretch)

2365 NH+ (stretch)

1604 Aromatic C=C (bend)

1494 Aromatic C-C (stretch)

Alkyl C - H Aromatic C=C

Figure 7. The infrared spectrum for diphenidine hydrochloride.

Figure 8. 1H-NMR (400 MHz, CD2Cl2) spectrum of diphenidine hydrochloride.

170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

Gas-chromatography is an analytical technique that is usually coupled with mass

Diphenidine is identified at a single peak with a retention time of 14.56 minutes as

However, in several diphenidine samples, particularly in method G, there has been an

A – Formation of the tropylium ion from diphenidine hydrochloride.

Instrumentation: GC-MS analysis was performed using an Agilent 6850 GC and