awn... . . , ‘
J . , . . (”WWW .

“:31: a.
2. ‘ . .

H

I
i

.
. «HS...
,flafiub.....

gm

 

m!. n no.

A . . 9

1.2.1.1 ..
.: .56:

‘ 'l‘l? .

. x.
571$“; nwhmqavfim: .31..
2.3.1542, :1. VA... a). basal...“
a: {#3 . 2:. A

.uv .

. "Era-(3.x
3‘ It III}. 5.9:
I’I .1

«suffix-iv .iOI.

{3.3.1.3}. 36‘..- 3.3“:

Ilv..1.:£.1ivfst§sei:£.izn nu
, ‘xi‘lbn. .-

u..$.J viii...
I (in?! ‘ . 3‘03:
7 E!!!
%.\n!v.¢li '3‘ o
’3‘}. {xi .3 LII...
awn‘vsuu? as #. | . I
2.. v31: 1- .
.

2...: . .a-iai.sc.rl.lelv8lt 9|
{I- 32.35;?! 10....CGIIQEIII: ‘3
. 1:... it! ulvsvlllninnnf
0503i

.3 I.

9.99.

tit.

ti. ll¥‘|.fi

~33... OAS! its:

III.’I!I..5W91«II;-
\l (3!...

3
I$X.|€\l

.3th

 

6.. V

3: . ‘56! 1... 2|.

5.. I’Li.1?1ln.~'vxsl|
(0..

‘ ,I
"novels! 5!!"

c 1 {r , Sui-tn
. .tt .llsstt...

I
11.... 6 d

 

 

 

F LIBRARY l
Michigan State
Universny

 

This is to certify that the
thesis entitled

OPTIMIZATION OF A MICROWAVE-ASSISTED
EXTRACTION PROCEDURE FOR THE EXTRACTION OF
ORGANIC IMPURITIES FROM SEIZED MDMA TABLETS

presented by

PATRICIA JEAN JOINER

has been accepted towards fulfillment
of the requirements for the

MS. degree in Forensic Science

 

 

MM

 

Major ProfessWs'Signature

September 3, 2009

 

Date

MSU is an Affirmative Action/Equal Opportunity Employer

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 Kthroleoc8-PrelelRC/DateDuo.indd

 

 

OPTIMIZATION OF A MICROWAVE-ASSISTED EXTRACTION PROCEDURE
FOR THE EXTRACTION OF ORGANIC IMPURITIES
FROM SEIZED MDMA TABLETS
By

Patricia Jean Joiner

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Forensic Science

2009

ABSTRACT
OPTIMIZATION OF A MICROWAVE-ASSISTED EXTRACTION PROCEDURE
FOR THE EXTRACTION OF ORGANIC IMPURITIES
FROM SEIZED MDMA TABLETS
By
Patricia Jean Joiner

The controlled substance 3,4-methylenedioxymethamphetamine (MDMA), also
known as ecstasy, is often ingested in tablet form. Because of the lack of quality control
in the synthesis of MDMA, impurities from the starting materials as well as intermediates
and by-products of reactions during the synthesis are often present in the tablets. The
organic impurities are extracted using a liquid-liquid extraction (LLE) procedure and are
analyzed by gas-chromatography-mass spectrometry (GC-MS) to Obtain the organic
impurity profile. Due to the limitations using LLE, alternative extraction procedures are
desired for the extraction of impurities.

In this work, a microwave-assisted extraction (MAE) procedure in conjunction
with a headspace solid-phase microextraction (HS-SPME) procedure is optimized for the
extraction Of organic impurities from MDMA tablets. The extraction buffer, pH, and
concentration were optimized for use in the MAE. Using a full factorial design, the MAE
parameters of ramp time, extraction time, and extraction temperature were determined to
be significant and were then optimized using a circumscribed central composite (CCC)
design. The HS-SPME parameters of extraction time and extraction temperature were
optimized empirically. The Optimized MAE/I-IS-SPME procedure was compared to a
HS-SPME procedure and a LLE procedure based on the literature [1] showing that

MAE/HS-SPME and HS-SPME are possible alternatives to LLE.

[1] Van Deursen et al. Sci Justice 2006; 46: 135-152.

ACKNOWLEDGEMENTS

I would first like to thank my advisor Dr. Ruth Smith for all of her guidance and
patience throughout the project start to finish. Without her help and support, this work
would not have reached its potential.

I thank my committee members Mr. Frank Schehr and Dr. Christina DeJong for
their input and help in preparing my thesis.

I would like to acknowledge the Michigan State Police for the supply of MDMA
tablets. The tablets were essential in the practical application of this work which could
not have been completed otherwise.

I would also like to thank the Forensic Sciences Foundation Lucas Grant and the
Michigan State University Graduate School for partial funding of this project. The funds
were essential in the completion of this work.

I thank my fellow forensic chemistry students who were always available for a
second opinion and for support whenever I needed it. Their thoughts and ideas helped to
remind me of the big picture and the overall goal of the work.

Finally, I would like to thank my family for their continued support. I especially

want to thank Adam Giebink for his love and support even when things were tough.

iii

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................ vii
LIST OF FIGURES ........................................................................................................... ix
KEY TO ABBREVIATIONS ............................................................................................ xi
Chapter 1 Introduction ........................................................................................................ 1
1.1. MDMA Background ................................................................................................ 1
1.1.1. MDMA History ................................................................................................. 1
1.1.2. MDMA Use ...................................................................................................... 2
1.1.3. MDMA Chemistry and Pharmacology ............................................................. 3

1.2. MDMA Production .................................................................................................. 4
1.2.1. MDMA Synthetic Routes ................................................................................. 4
1.2.2. Tablet Production .............................................................................................. 5

1.3. MDMA Profiling ..................................................................................................... 6
1.4. Alternative Extraction Procedures ......................................................................... 10
1.4.1. Headspace Solid-Phase Microextraction ........................................................ 10
1.4.2. Microwave-Assisted Extraction ...................................................................... 13

1.5. Research Objectives ............................................................................................... l4

1 .6. References .............................................................................................................. l 6
Chapter 2 Theory .............................................................................................................. 20
2.1. Microwave-Assisted Extraction ............................................................................. 20
2.2. Headspace Solid-Phase Microextraction ............................................................... 22
2.3. Gas Chromatography-Mass Spectrometry ............................................................. 26
2.3.]. Gas Chromatography ...................................................................................... 27
2.3.2. Mass Spectrometry .......................................................................................... 30

2.4. Experimental Design .............................................................................................. 33
2.4.1. Screening Design ............................................................................................ 34
2.4.1.1. Full Factorial Design ................................................................................ 34
2.4.1.2. Analysis of Full Factorial Design ............................................................ 36

2.4.2. Optimization Design ....................................................................................... 39
2.4.2.1. Circumscribed Central Composite Design ............................................... 39
2.4.2.2. Analysis of Circumscribed Central Composite Design ........................... 41

2.5. Retention Time Alignment .................................................................................... 43
2.6. Pearson Product Moment Correlation Coefficients ............................................... 44
2.7. References .............................................................................................................. 45
Chapter 3 Materials and Methods ..................................................................................... 47
3.1. Sample Preparation ................................................................................................ 47
3.1.1. Simulated Sample ........................................................................................... 47
3.1.2. MDMA Exhibits ............................................................................................. 47

3.2. Optimization of Microwave—Assisted Extraction Procedure ................................. 48

3.2.1. Optimization of Extraction Buffer .................................................................. 48
3.2.2. Determination of Significant Parameters... ..................................................... 51
3.2.3. Optimization of Significant Parameters .......................................................... 53
3.3. Optimization of Headspace Solid-Phase Microextraction Procedure .................... 54
3.4. Liquid-Liquid Extraction Procedure ...................................................................... 55
3.5. Gas Chromatography—Mass Spectrometry ............................................................. 55
3.6. Comparison of Extraction Procedures ................................................................... 56
3.7. References .............................................................................................................. 58
Chapter 4 Results and Discussion ..................................................................................... 59
4.1. Sample Preparation ................................................................................................ 59
4.1.1. Simulated MDMA Sample .................................................. . ........................... 59
4.1.2. MDMA Exhibits ............................................................................................. 59
4.2. Optimization Of Microwave-Assisted Extraction Procedure ................................. 60
4.2.1. Selection of Extraction Buffer ........................................................................ 60
4.2.1.1. Investigation of Phosphate Buffers .......................................................... 60
4.2.1.2. Investigation of Tris Buffers .................................................................... 66
4.2.1.3. Investigation of Carbonate Buffer ........................................................... 68
4.2.2. Determination of Significant Parameters ........................................................ 69
4.2.3. Optimization of Significant Parameters .......................................................... 71
4.3. Optimization of Headspace Solid-Phase Microextraction Procedure .................... 74
4.3.1. Optimization of Extraction Time .................................................................... 75
4.3.2. Optimization of Extraction Temperature ........................................................ 78
4.4. Comparison of MAE/HS-SPME, HS-SPME, and LLE ......................................... 80
4.4.1. Simulated Sample ........................................................................................... 80
4.4.2. MDMA Exhibits ............................................................................................. 81
4.4.2.1. MDMA Exhibit MSU900-01 ................................................................... 83
4.4.2.2. MDMA Exhibit T-17 ............................................................................... 92
4.4.2.3. MDMA Exhibit T-27 ............................................................................... 99
4.4.3. Summary ....................................................................................................... 105
4.5. Comparison of MDMA Exhibits ......................................................................... 108
4.5.]. Comparison Of MDMA Exhibits MSU900-01, T-17, and T-27 ................... 108
4.5.2. Comparison of MDMA Exhibits T-17 and CJ-FSOS .................................... 110
4.6. Summary .............................................................................................................. 112
4.7. References ............................................................................................................ 1 14
Chapter 5 Conclusions and Future Work ........................................................................ 116
5.1. Conclusions .......................................................................................................... 1 16
5.2. Future Work ......................................................................................................... 120
5.3. References ............................................................................................................ 122

Appendix A: Synthesis Schemes of 3,4-MethylenedioxyphenyI-Z-Propanone (MDP2P)
......................................................................................................................................... 123

Appendix B: Synthesis Schemes of 3,4-Methylenedioxymethamphetamine (MDMA)

from MDPZP ................................................................................................................... 124
Appendix C: Experimental Runs for Full Factorial Screening Design ........................... 125
Appendix D: Experimental Runs for CCC Optimization Design ................................... 126

Appendix E: MANOVA Table for Simulated Sample Components from Full Factorial
Design ............................................................................................................................. 127

Appendix F: Full List of Compounds Extracted from MDMA Exhibits MSU900-01, T-
17, and T-27 using MAE/HS-SPME, HS-SPME, and LLE ........................................... 129

Appendix G: Full List of Compounds Extracted from MDMA Exhibits T-l 7 and CJ-FSOS
using HS-SPME (unidentified impurity numbers do not correspond to Appendix F) 146

vi

LIST OF TABLES

Table 2.1: Example set up of full factorial design with three parameters ........................ 35
Table 2.2: Calculations for the sum of squares ................................................................. 37
Table 2.3: Calculations for the degrees of freedom .......................................................... 38
Table 2.4: Calculations for the mean squares and F-values .............................................. 38
Table 2.5: Example set up of CCC design with two parameters ...................................... 40
Table 3.1: High and low parameters for DJ” factorial screening design ........................... 51
Table 3.2: GC-MS parameters for HS-SPME analysis and LLE analysis ........................ 56

Table 4.1: Physical characteristics of MDMA exhibits (averages based on ten tablets) .. 60

Table 4.2: Average abundance and RSD values of simulated sample components in
phosphate buffers (average peak areas based on three replicates) ............................ 61

Table 4.3: Average abundance and RSD values of simulated sample components in tris
buffers (average peak area based on three replicates) .............................................. 67

Table 4.4: Average abundance and RSD values of simulated tablet components in
carbonate buffer (average peak area based on three replicates) ............................... 69

Table 4.5: Optimum microwave parameters from the CCC design .................................. 73

Table 4.6: Relative standard deviations for simulated sample components extracted by the
three procedures ........................................................................................................ 81

Table 4.7: Number of impurities and components extracted from MDMA exhibit
MSU900-01 ............................................................................................................... 83

Table 4.8: Average PPMC coefficients and standard deviations of MDMA exhibit
MSU900-01 associated with each extraction procedure ........................................... 91

Table 4.9: Number of impurities and components extracted from MDMA exhibit T-17
................................................................................................................................... 93

Table 4.10: Average PPMC coefficients and standard deviations of MDMA exhibit T-17
associated with each extraction procedure ................................................... . ............ 98

vii

Table 4.11: Number of impurities and components extracted from MDMA exhibit T-27

................................................................................................................................... 99
Table 4.12: Average PPMC coefficients and standard deviations of exhibit T-27
associated with each extraction procedure .............................................................. 104
Table 4.13: Summary of number of impurities and components extracted from each
MDMA exhibit by each extraction procedure; also shown are average PPMC
coefficients and standard deviations for each exhibit extracted by each procedure
................................................................................................................................. 106
Appendix C: Experimental Runs for Full Factorial Screening Design ........................... 125
Appendix D: Experimental Runs for CCC Optimization Design ................................... 126

Appendix E: MANOVA Table for Simulated Sample Components from Full Factorial
Design ............................................................................................................................. 127

Appendix F: Full List of Compounds Extracted from MDMA Exhibits MSU900-01, T-
17, and T-27 using MAE/HS-SPME, HS-SPME, and lLLE ........................................... 129

Appendix G: Full List of Compounds Extracted from MDMA Exhibits T-l 7 and C] -F SOS
_ using HS-SPME .............................................................................................................. 146

viii

LIST OF FIGURES

Figure 1.1: Chemical structure of MDMA with the methylenedioxy substitution on the
aromatic ring outlined and the (It-carbon labeled ........................................................ 3

Figure 2.1: Schematic of HS-SPME with the arrows representing the movement of the

analytes ..................................................................................................................... 22
Figure 2.2: Schematic of a gas chromatograph ................................................................. 27
Figure 2.3: Polydimethyl siloxane phase in CC columns ................................................. 29
Figure 2.4: Schematic of a mass spectrometer .................................................................. 30
Figure 2.5: Schematic of an ion trap mass analyzer ......................................................... 31
Figure 2.6: Schematic of set up of experiments for CCC design ...................................... 40
Figure 2.7: Graph of desirability function for maximization at different values of 3

(adapted from references 13 and 14) ........................................................................ 42
Figure 3.1: Representative tablets from each exhibit: a) exhibit MSU900-01 (pink, purple,

and green); b) exhibit T-17 (blue); c) exhibit T-27 (pink) ........................................ 48
Figure 3.2: Schematic of assembled microwave vessel with sample ............................... 49

Figure 3.3: Schematic of assembled microwave vessel with quartz insert and sample.... 52

Figure 4.1: Chromatograms of the simulated sample in l M phosphate buffer at a) pH 6
and b) pH 8; an * indicates that a peak was present in the blank ............................. 63

Figure 4.2: Estimated response surface for methamphetamine from CCC design with a
plus sign (+) indicating the response at the optimum settings for the MAE
parameters ................................................................................................................. 74

Figure 4.3: Chromatograms of HS-SPME extractions at a) 60 minutes, b) 40 minutes, and
c) 10 minutes; an asterisk (*) indicates that a peak was present in the blank ........... 77

Figure 4.4: Chromatograms of HS-SPME extractions at a) 80 °C and b) 40 °C; an asterisk
(*) indicates that the peak was present in the blank .................................................. 79

Figure 4.5: Chromatogram of MDMA exhibit MSU900-01 extracted by MAE/I-IS-SPME;
an asterisk (*) indicates that the peak was present in the blank ................................ 84

ix

Figure 4.6: Chromatogram of MDMA exhibit MSU900-01 extracted by HS-SPME; an
asterisk (*) indicates that the peak was present in the blank .................................... 85

Figure 4.7 : Chromatogram of MDMA exhibit MSU900-01 extracted by LLE ............... 86

Figure 4.8: Chromatogram of MDMA exhibit T-l 7 extracted by MAE/HS-SPME; an
asterisk (*) indicates that the peak was present in the blank .................................... 94

Figure 4.9: Chromatogram of MDMA exhibit T-17 extracted by HS-SPME; an asterisk
(*) indicates that the peak was present in the blank .................................................. 95

Figure 4.10: Chromatogram of MDMA exhibit T-17 extracted by LLE .......................... 96

Figure 4.11: Chromatogram of MDMA exhibit T-27 extracted by MAE/HS-SPME; an
asterisk (*) indicates that the peak was present in the blank .................................. 100

Figure 4.12: Chromatogram of MDMA exhibit T-27 extracted by HS-SPME; an asterisk

(*) indicates that the peak was present in the blank ................................................ 101
Figure 4.13: Chromatogram of MDMA exhibit T-27 extracted by LLE ........................ 102
Figure 4.14: Chromatograms of MDMA exhibits a) T-l7 and b) CJ-FSOS extracted by

HS-SPME; an asterisk (*) indicates that the peak was present in the blank ........... 11 1
Appendix A: Synthesis Schemes of 3,4-Methylenedioxyphenyl-Z-Propanone (MDP2P)

................................................................................................................................. 123
Appendix B: Synthesis Schemes of 3,4-Methylenedioxymethamphetamine (MDMA)

from MDP2P .......................................................................................................... 124

amu

ANOVA

BZP

CAR

CCC

CSA

DEA

DVB

DVB/CAR/PDMS

EM

FDA

FTIR

GC

GC-MS

HPLC

HS-SPME

ICP-MS

LLE

MAE

MAE/I-IS-SPME

MANOVA

KEY TO ABBREVIATIONS

Atomic mass unit

Analysis of variance
N-benzylpiperazine

CarboxenTM

Circumscribed central composite
Controlled Substances Act

Drug Enforcement Administration

Divinylbenzene

Divinylbenzene/CarboxenTM/polydimethy|siloxane
Electron multiplier

Food and Drug Administration

Fourier transform infrared spectroscopy

Gas chromatography (gas chromatograph)

Gas chromatography-mass spectrometry

High performance liquid chromatography
Headspace solid-phase microextraction
Inductively-coupled plasma mass spectrometry
Liquid-liquid extraction

Microwave-assisted extraction

Microwave-assisted extraction/headspace solid-phase
microextraction

Multivariate analysis of variance

xi

MDA

MDEA

MDMA
MDP2P
MDPZP-oxime
MDP2-propanol
MS

m/z
N-formyl-MDA
N-formyI-MDMA
PDMS
PDMS/DVB
PPMC

RSD

SPME

TFMPP

3,4-Methylenedioxyamphetamine
3,4-Methylenedioxyethylamphetamine
3,4-Methylenedioxymethamphetamine
3,4-Methylenedioxyphenyl-2-propanone
3,4-Methylenedioxyphenyl-Z-propanone oxime
3,4-Methylenedioxyphenyl-Z-propanol

Mass spectrometry (mass spectrometer)
Mass to charge ratio
N-fonnyl-methylenedioxyamphetamine
N-formyl-methylenedioxymethamphetamine
Polydimethylsiloxane
Polydimethylsiloxane/divinylbenzene
Pearson product moment correlation
Relative standard deviation

Solid-phase microextraction

1-(3-Triflouromethyl)pheny|piperaz ine

xii

Chapter 1 Introduction
1.1. MDMA Background

The controlled substance 3,4-methy1enedioxymethamphetamine (MDMA), also
known as the club drug “ecstasy,” is a dangerous and illegal substance. MDMA is a
synthetic amphetamine-type stimulant often ingested in tablet form. This psychedelic
drug has many effects on the body including euphoria and distortions in perceptions [l ].
1.1.1. MDMA History

MDMA was synthesized in the early 19005 by the German pharmaceutical
company Merck as an intermediate product in an attempt by the company to synthesize a
drug to stop bleeding [2]. The compound was patented in 1914 by Merck under German
Patent number 274350 [3]. In the 19505, the US Army was reported to have conducted
studies with the drug, testing its toxicity in several animals including mice, and rats [4,5].

A major resurgence of the drug occurred in the mid-19705. Dr. Alexander
Shulgin synthesized MDMA for experimentation. He ingested the drug and took careful
notes of its effects including euphoria [6]. Around the same time, psychiatrists were
utilizing the compound with patients to enhance communication; however, MDMA had
never been tested by the Food and Drug Administration (FDA) for this use [7]. In the
19805 the drug first became available on the streets. After being emergency scheduled in
Schedule I of the Controlled Substances Act (CSA) in 1985, MDMA was permanently
placed in Schedule I by the Drug Enforcement Administration (DEA) in 1988. As a
Schedule I controlled substance, MDMA was thought to have a high potential for abuse

and no proven medical uses [8,9].

l - 1 .2. MDMA Use

The term “ecstasy” encompasses a broad range of synthetic tablets that are
ingested orally, often in a club or rave setting [10,11]. Other names for these tablets
include Adam, XTC, Beans, E, X, Hug Drug, and Disco Biscuit [12]. Most ecstasy
tablets contain MDMA, though not all do. Ecstasy mimic tablets often contain a mixture
ofN-benzylpiperazine (BZP), a Schedule 1 controlled substance under the CSA, and 1-(3-
triflouromethyl)phenylpiperazine (TFMPP) which is not currently controlled [13].
Ecstasy tablets which include MDMA often also contain methamphetamine and 3,4-
methylenedioxyamphetamine (MDA) in addition to other substances such as ketamine,
caffeine, and diazepam [10]. The tablets can vary in size, shape, color, and markings
with masses ranging from approximately 0.2 g to 0.3 g. The percentage of MDMA in

tablets ranges from approximately 30-50% of the total mass [10].

According to an ongoing study by the University of Michigan, MDMA use has
th th th

fluctuated over the last 13 years [14]. The study surveyed 8 grade, 10 grade, and 12
grade students concerning their use and attitudes towards illicit drugs. The number of

students who reported having used MDMA at least once in their life peaked in 2001 with

12% of 12th grade students reporting MDMA use. Since 2001, the number of students
reporting MDMA use has decreased to approximately 6% of 12th grade students.

However, the study also reported the perceived risk associated with using MDMA once

or twice. The percentage of students who associate a risk with using MDMA has slowly

decreased since 2005 with less than half of the 8th grade and 10th grade students

responding that risk is involved. Because of the trend of attitudes towards MDMA, there

is a concern that a resurgence Of the drug’s popularity may occur.

1.1.3. MDMA Chemistry and Pharmacology

MDMA is a member of the phenethylamine class of compounds. It is structurally
similar to amphetamine and methamphetamine with the addition of a methylenedioxy
group on the aromatic ring (Figure 1.1). The methylenedioxy substitution on the
aromatic ring gives MDMA its hallucinogenic properties while the methyl group on the

a-carbon gives MDMA its stimulant properties [15].

 

 

 

C
I /O
NH
/ \9. ........ .

Figure 1.1: Chemical structure of MDMA with the methylenedioxy substitution on the
aromatic ring outlined and the a—carbon labeled

When ingested orally, MDMA enters the blood stream and reaches peak levels
about two hours later [11]. In the body, MDMA may remain unchanged or be
metabolized to MDA as it is excreted. Because the half life of MDMA is eight hours,
effects of the MDMA can last for several days afier the drug has been ingested [11].
Effects of MDMA on the body can include euphoria, distortions in perception, increased
energy, and a feeling of closeness with other people [1,11]. MDMA can also cause
increased heart rate, nausea, blurred vision, faintness, and hyperthermia [1].

When MDMA reaches the brain, it has the most effect on serotonin neurons [16].
The MDMA acts as a substrate for the serotonin transporter on the pre-synaptic cell.
Once in the nerve terminal, MDMA displaces serotonin from vesicles which increases the
amount of serotonin released into the synapse between neurons. The areas of the brain

most affected by the increase in serotonin levels include the prefrontal cortex, which is

involved in decision making; the thalamus, which is involved in sensory processing; and
the amygdala, which is involved with fear and anxiety reactions [16].

1.2. MDMA Production

1.2.1. MDMA Synthetic Routes

Several routes are available for synthesizing MDMA with routes using 3,4-
methylenedioxyphenyI—2-propanone (MDP2P) as the starting material being the most
common [17]. Because MDP2P is now regulated by the DEA, it too must be synthesized
in the clandestine laboratory. Two of the more common routes for MDP2P synthesis are
shown in Appendix A [18]. In the first synthesis, safrole is extracted from sassafras oil, a
naturally occurring substance in eastern North America and eastern Asia [19]. Safrole is
then converted to isosafrole through isomerization using potassium hydroxide and
ethanol. lsosafrole can also be obtained from industrial sources thus bypassing the
safrole extraction and isomerization steps [18]. lsosafrole glycol is produced by the
oxidation of isosafrole using formic acid and hydrogen peroxide. Finally MDP2P is
generated by dehydration of isosafrole glycol using sulfuric acid [18].

The second synthesis of MDP2P begins in the same way with safrole being
converted to isosafrole. However, using sulfuric acid and sulfanilic acid, isosafrole is
oxidized to form piperonal. Piperonal is then converted to B-nitroisosafrole via the
Knoevenagel-Walter condensation using nitroethane. After the formation of an
intermediate oxime, l-(3,4-methylenedioxyphenyl)-2-propanone oxime (MDPZP oxime),
through oxide-reduction, MDP2P is formed by hydrolysis using acetic acid [18].

After MDPZP is synthesized, the MDMA synthesis begins. While there are other

synthesis methods that can be used, the two most common are reductive amination and

the Leukart synthesis [17,18]. Appendix B shows a schematic of the synthesis of MDMA
from MDP2P by both of these routes.

For the reductive amination route, the MDPZP is reacted with methylamine to
form an imine intermediate: 1,2-(methylenedioxy)-4-(2-N-methyl-iminopropyl) benzene.
This compound is then reduced to MDMA using one of several reducing agents such as
sodium cyanoborodhydride or sodium borohydride [18,20]. During the reduction of the
imine, 3,4-methylenedioxyphenyI-Z-propanol (MDPZ-pronanol) is formed from a side
reaction which lowers the yield of MDMA. To limit the production of MDPZ-propanol
and increase the production of MDMA, a laboratory may use the “cold method.” When
the cold method is used, the temperature of the mixture during synthesis is cooled to -20
°C which increases the selectivity of the reducing agent to form MDMA [21].

Two variations of the Leukart synthesis are commonly used. The first involves
the reductive amination of MDPZP to N-formyl-methylenedioxymethamphetamine (N-
formyl-MDMA) using methylformamide [22]. Through hydrolysis, MDMA is then
formed [20]. Alternatively, MDP2P can be converted to N-formyl-
methylenedioxyamphetamine (N-formyI-MDA) using formamide through reductive
amination [22]. Then, through reduction with lithium aluminum hydride, MDMA is
formed [20].

1.2.2. Tablet Production

Using one of the aforementioned synthetic routes, MDMA is synthesized in
clandestine laboratories [17,23]. The impurities from chemicals used as starting
materials as well as by-products and intermediates of reactions during the synthesis are

present in the MDMA powder, though often in low concentrations [17]. The synthesized

MDMA is then mixed with additives including adulterants and diluents. Adulterants,
such as caffeine, are compounds that are added to the drug to enhance its effects. Other
controlled substances such as methamphetamine and amphetamine may also be added to
further enhance the effects of the MDMA. Diluents, such as lactose, are added to dilute
the MDMA so that more tablets can be produced from a single batch of MDMA.
Additionally, color dyes are also used to dilute the MDMA and to give the tablets their
color.

The mixture of MDMA and additives is then pressed into tablets using various
tablet presses that give the tablets different shapes and logos (or imprints). In a
clandestine laboratory, a single batch of MDMA can be divided and pressed into tablets
with different colors and logos. Therefore, exhibits that look physically different may in
fact contain MDMA that was synthesized in the same batch and therefore have the same
chemical properties [10,17]. The percentage of MDMA present in the final tablets varies
by batch and manufacturer [10]. Tablets currently being seized by law enforcement
contain about 30% MDMA (by mass) as well as methamphetamine and caffeine. Also,
some tablets being seized contain a mixture of MDMA and BZP [24].
1.3. MDMA Profiling

The goal of profiling MDMA tablets is to link tablets from different exhibits to a
common batch of MDMA or link tablets to a common production source in an effort to
determine drug trafficking routes. Historically, physical characteristics of tablets have
been used in MDMA tablet profiling [10]. However, the physical characteristics alone
may not be sufficient to compare tablets from different exhibits because clandestine

laboratories can manufacture tablets that appear physically different [10,23].

The limitations of profiling MDMA tablets based only on the physical appearance
of the tablets were demonstrated by Cheng et al. [10]. Using over 123,000 MDMA
tablets seized from 613 cases in Hong Kong, the group recorded physical characteristics
and studied the chemical composition of tablets using Fourier transform infrared
spectroscopy (FTIR), high performance liquid chromatography (HPLC), and gas
chromatography-mass spectrometry (GC-MS). Many of the physically similar tablets
were likely not from a common production source because different impurities and
additives were present. On the other hand, some tablets that had different physical
characteristics were found to possess similar impurity profiles (similar impurities present
at similar concentrations), indicating that the tablets may have originated from the same
production source. This study demonstrated the downfalls of tablet profiling based on
physical characteristics alone and highlighted the need for profiling tablets based on the
chemical composition.

The profiling Of tablets based on chemical composition can be completed at
different levels. The overall chemical components of the tablet can be used in
compositional profiling or the organic impurities present in the tablet can be used in
organic impurity profiling [25]. To study the organic impurities in MDMA tablets, the
tablets are typically ground and dissolved in an aqueous solvent and then extracted into
an organic medium. The extraction is referred to as liquid-liquid extraction (LLE). After
LLE, the extract is analyzed, often by GC-MS, with the resulting chromatogram referred
to as the organic impurity profile of the tablet. By studying and comparing the impurity
profiles of the tablets, a more complete comparison of tablets from different exhibits can

be obtained. The identity of impurities present can be used to determine the synthetic

route used to manufacture the MDMA. In addition, tablets from different exhibits can
possibly be linked based on the impurities.

Van Deursen et al. developed a LLE method for the extraction of organic
impurities from seized MDMA tablets [26]. In the developed procedure, whole MDMA
tablets were ground and dissolved in 0.33 M phosphate buffer at pH 7. After several
agitation steps, including vortexing, centrifugation, and sonication, 400 uL of toluene
were added to the solution. After a few final agitation steps, such as rotative shaking and
centrifugation, the toluene layer was extracted and analyzed by GC-MS. Good
repeatability (average relative standard deviation, RSD, without outliers was 6%) over six
extractions was achieved based on the peak areas of 22 selected impurities in the
chromatograms (for example MDP2P, MDP2-propanol, and N-fonnyl-MDMA). Also,
good reproducibility among days over a two week time span was reported with RSD
values of 7% and 8% (without outliers) using two separate MDMA exhibits. Based on
the impurities detected and identified in the MDMA exhibits, the group determined that
the MDMA present in many tablets seized in the Netherlands was manufactured using the
reductive amination route. The group also noted that differences between the tablets
were likely due to different reducing reagents used in the manufacturing process.

In addition to between exhibit comparisons, the impurities present in MDMA
tablets can be utilized to determine the method used to synthesize the MDMA. Palhol et
al. identified 29 impurities in 52 MDMA samples seized in France using LLE followed
by GC-MS [17]. Impurities from starting materials and side reactions were present in the
final tablets with different synthetic routes yielding different impurities. For example, the

Leukart reaction gave N-fomiyl-MDMA, an impurity not observed in other routes. Also,

the bromopropane route was the only route to show brominated impurities. The most
common synthetic route in Europe was determined to be reductive amination from
MDP2P based on the impurities in the analyzed tablets. The group also compared tablets
from different exhibits based on the impurities present and determined that similar levels
of the same impurities indicated that the MDMA was produced in a common batch. The
same impurities present at different levels indicated that, while the same synthetic route
was used, the MDMA originated from different batches from the same clandestine
laboratory. Different impurities present indicated that the samples were unrelated.
(However, a clandestine laboratory may produce batches of MDMA using different
synthetic routes yielding different impurities.) Therefore, a more definitive comparison
of tablets is possible when organic impurities contained in the tablets are considered,
rather than only the physical characteristics.

Using organic impurity profiles obtained by the LLE procedure developed by van
Deursen et. al [26], Weyermann et al. created a standardized procedure that could be
used across many laboratories. With participating laboratories using the same procedure,
the results can be pooled to create a database. Throughout the work, the differences in
eight impurities among 26 MDMA exhibits were studied [23]. Some of these impurities
included: MDP2P, MDP2-propanol, and N-formyl-MDMA. These eight organic
impurities were selected to compare tablets based on good reproducibility between
replicate analyses and large variability among samples from different exhibits. The
correlation among the samples was assessed using Pearson product moment correlation
(PPMC) coefficients based on the peak areas of the eight impurities. Using this method,

successful discrimination of many of the exhibits was achieved (although it is not clear

how many exhibits were discriminated). However, because only eight impurities were
considered in the comparison, potentially discriminatory information found elsewhere in
the chromatogram was overlooked.
1.4. Alternative Extraction Procedures

Despite the successful use of LLE for organic impurity profiling there are many
limitations of LLE. For example, LLE can efficiently extract components such as
methamphetamine, MDMA, and caffeine. The efficient extraction of these components
is not desirable because the compounds can result in large, broad peaks in the
chromatogram that can potentially mask impurities present at lower concentrations.
Also, a relatively large sample mass is required. For the van Deursen method, a full
tablet is used for each extraction. Therefore, when exhibits contain a small number of
tablets, possibly only one tablet, re-testing cannot be performed [26]. Because organic
solvents are used for LLE, costs increase and organic waste is generated. Therefore,
alternative extraction methods to LLE are desirable to overcome these limitations.
1.4.1. Headspace Solid-Phase Microextraction

Solid-phase microextraction (SPME) is an extraction procedure in which a
polymer coated fiber is introduced to the sample either directly in contact with the sample
(immersion SPME) or in the headspace of the sample (HS-SPME). Solid-phase
microextraction allows for the pre-concentration of analytes onto the fiber thus allowing
for the detection of components present at low concentrations. Headspace solid-phase
microextraction has been used as an alternative method of extraction to LLE for profiling

synthetic illicit drugs such as methamphetamine.

10

Kuwayama et al. Optimized HS-SPME parameters such as sample mass,
extraction time, extraction temperature, and fiber type for the extraction of organic

impurities from methamphetamine [27]. In the developed procedure, t he optimized
extraction procedure included exposing a divinylbenzene/Carboxenm/

polydimethylsiloxane (DVB/CAR/PDMS) fiber to the headspace of 50 mg solid sample
for 30 minutes at 85°C. Because the HS-SPME pre-concentrated the analytes on the
fiber, less sample mass was required for this extraction compared to LLE. The group
applied statistical procedures such as Euclidian distance, cosine distance, and correlation
coefficients to compare the samples based on the abundance of eight impurities. It was
determined that a logarithmic conversion followed by a cosine distance calculation was
best for discriminating tablets from different batches and classifying tablets from the
same batch. The group demonstrated that HS-SPME was a quick and simple extraction
method for organic impurities while minimizing the extraction of methamphetamine.

Koester et al. also used HS-SPME to obtain impurity profiles Of
methamphetamine [28]. The HS-SPME profiles were compared to profiles obtained
using other extraction procedures including LLE, acid dissolution, base dissolution, and
solvent dissolution. The HS-SPME procedure, which involved sampling the headspace
of the solid sample, extracted 30 impurities while the LLB procedure extracted only eight
impurities. Methamphetamine was present in the HS-SPME profiles; however, it was
present at lower concentrations and did not dominate the profile, as observed in the LLE
profiles.

The application of SPME to MDMA was described by Kongshaug et al. [29].

The MDMA sample was dissolved in 0.1 M acetate buffer at pH 5 which was chosen to

11

avoid the over-extraction of MDMA. A polydimethylsiloxane/divinylbenzene
(PDMS/DVB) fiber was exposed to the sample for 30 minutes at 90 °C. The two
different SPME sampling modes (immersion and headspace) were investigated. The HS-
SPME sampling mode was preferred due to the increased lifetime of the fiber. In
immersion SPME, the fiber degraded more quickly because it was in direct contact with
the liquid sample. The PDMS/DVB fiber extracted more impurities (though it is unclear
how many more) than the PDMS fiber because the PDMS/DVB fiber extracted analytes
with a wider range of polarities. To assess the precision of the extraction, the RSD values
of peak areas was calculated to be 2-13%, showing that the HS-SPME procedure was
precise. The HS-SPME procedure extracted a similar number of impurities to LLE with
similar chromatography but without the need for organic solvents.

Bonadio et al. optimized a HS-SPME procedure for the extraction of impurities
from a ground MDMA tablet [30]. The developed method involved pre-heating the vial
with 40 mg of ground MDMA sample for 15 minutes at 80 °C and then exposing a
PDMS/DVB fiber to the headspace for 15 minutes at 80 oC. The group chose the 10

impurities with the most repeatable peak areas and applied data pre-treatments such as a
normalization using the 4th square root and the logarithm. Principal components analysis

was then used to identify clusters of similar MDMA samples. The group also compared
their developed method to a LLE procedure [31]. The LLE procedure extracted 15 more
impurities than the HS-SPME procedure, but the HS-SPME sample preparation was

simpler because fewer steps were required.

12

1.4.2. Microwave-Assisted Extraction

Microwave-assisted extraction (MAE) uses microwave energy to heat a solution
under pressure. The pressure allows the heating of the solvent to temperatures higher
than its atmospheric boiling point. The higher MAE temperatures are achieved in a few
minutes which is faster than heating a sample by conventional heating. Theoretically,
MAE offers a highly efficient extraction in part because the sample is heated more evenly
due to the heating mechanisms (discussed in chapter 2).

So far, MAE has had limited use in drug analysis applications. The procedure has
been applied to the extraction of the active ingredients in pharmaceuticals by Hoang et al.
[32]. Using MAE, the group achieved 97-102% extraction efficiency when the amount
extracted was compared to the amount given on the label of the pharmaceutical. The
extraction efficiency compares to conventional extraction procedures; however, the MAE
took only seven minutes compared to the conventional 30 minutes. The RSD values
between replicate extractions were 1.4% showing that the MAE procedure had good
repeatability that compared well with the conventional extraction.

The use of MAE in combination with SPME has been-reported in the literature as
a method that Offers a highly efficient extraction (MAE) while maintaining the desired
selectivity (SPME). Bieri et al. used focused MAE (microwave-assisted extraction at
atmospheric pressure) followed by immersion SPME to extract cocaine from coca leaves,
and the fiber extract was analyzed by GC-MS [33]. The MAE involved placing 100 mg
Of coca leaves in 5 mL methanol and exposing it to microwave energy for 30 seconds.
The effects of pH, extraction time, and extraction temperature on immersion SPME were

also studied. The group determined that the Optimum SPME procedure involved diluting

l3

the microwave extract (50:1) in 50mM phosphate buffer at pH 8.1. The PDMS fiber was
immersed in a sample for 15 minutes at 25 °C. By using MAE with HS-SPME, the total
extraction time was reduced by 29 minutes. The MAE procedure allowed a quick and
efficient extraction while the addition of the SPME step allowed for more selective
extraction of the cocaine from coca leaves.

Carro et al. applied a combined MAE/HS-SPME procedure to the extraction of
polybrominated compounds from aquaculture samples [34]. Using a central composite
experimental design, MAE parameters, such as solvent type, solvent volume, extraction
temperature, and extraction time were studied and optimized. Following MAE, a clean-
up step and HS-SPME step were performed. Because of the pre-concentration ability of
the fiber in HS-SPME, lower detection limits were achieved when the HS-SPME step
was included compared to analyses where HS-SPME was not included. For example, for
the compound heptachlor, the limit of detection for MAE/HS-SPME was 80 pg/g of
sample whereas, with MAE alone, the limit of detection was 690 pg/g of sample. This
study demonstrated the possibility of MAE used in conjunction with HS-SPME to
achieve lower limits of detection.

1.5. Research Objectives

The objective of this research is to develop a MAE/HS—SPME procedure for the
extraction of organic impurities from seized MDMA tablets. Microwave-assisted
extraction allows for an efficient extraction; however, because of the high efficiency, a
more selective extraction procedure is desired to follow the MAE. Headspace solid-

phase microextraction was chosen because it allows for the selective extraction of the

14

organic impurities while minimizing the extraction of the more concentrated components
such as methamphetamine, MDMA, and caffeine.

This work was completed in two parts. The MAE/HS-SPME procedure was
optimized first. This involved determining an appropriate extraction buffer and HS-
SPME parameters and then optimizing the MAE parameters. The pH, concentration, and
type Of the extraction buffer for MAE were optimized to achieve a precise extraction that
limited sample carry-over between extractions. The HS-SPME extraction time and
extraction temperature were optimized empirically to allow for the extraction Of
impurities without overloading the fiber. The MAE parameters ramp time, extraction
time, and extraction temperature were optimized using experimental design procedures.
A full factorial design was used to determine the significant parameters and a
circumscribed central composite (CCC) design was used to Optimize the significant
parameters.

After the method was optimized, three MDMA exhibits were utilized to compare
the optimized MAE/HS-SPME procedure to a HS-SPME procedure and a LLE procedure
based on the literature [26]. The goal was to determine which extraction procedure

extracted the most impurities and would be the most useful in crime laboratories.

15

1.6. References

[1] National Institute on Drug Abuse. NIDA InfoFacts: MDMA. 2008. Available at
http://www.drugabuse.gov/infofacts/ecstasy.html. (Accessed 22 December 2008).

[2] Freudenmann R, Oxler F, Bemschneider-Reif S. The Origin of MDMA (Ecstasy)
Revisited: The True Story Reconstructed from the Original Documents. Addiction 2006;
9: 1241-1245.

[3] E Merck, assignee. Verfahren zur Darstellung von Alkyloxyaryl-, Dialkyloxyaryl-
und Alkylendioxyarylaminopropanen bzw. Deren am Stickstoff monoalkylierten
Derivaten. German Patent 274350. 16 May 1914.

[4] Parrot AC. Human Psychopharrnacology of Ecstasy (MDMA): A Review of 15 Years
of Empirical Research. Hum Psychopharmacol Clin Exp 2001; 16: 557-577.

[5] Hardman HF, Haavik CO, Seevers MH. Relationship of the Structure of Mescaline
and Seven Analogs to Toxicity and Behavior in Five Species of Laboratory Animals.
Toxicol Appl Pharmacol 1973; 25: 299-309.

[6] Shulgin A, Shulgin A. PiHKAL: A Chemical Love Story. Transform Press,
Berkeley, CA; 1991.

[7] National Institute on Drug Abuse. Research Report Series: MDMA (Ecstasy) Abuse.
2006. Available at http://www.drugabuse.gov/ResearchReports/MDMA/default.html.
(Accessed 22 December 2008).

[8] Drug Enforcement Administration. Drug Scheduling. Available at
http://www.usdoj.gov/dea/pubs/scheduling.html. (Accessed 5 January 2009).

[9] Drug Enforcement Administration. Irma Perez Story. Available at
http://www.usdoj.gov/dea/pubs/states/newsrel/perez_story.html. (Accessed 05 January
2009).

[10] Cheng WC, Poon NL, Chan MF. Chemical Profiling of 3,4-
Methylenedioxymethamphetamine (MDMA) Tablets Seized in Hong Kong. J Forensic
Sci 2003; 48: 1249-1259.

[11] McKim WA. Drugs and Behavior: An Introduction to Behavioral Pharmacology.
Prentice Hall, Upper Saddle River, NJ; 2003.

[12] Drug Enforcement Administration. MDMA (Ecstasy). 2006. Available at
http://www.usdoj.gov/dea/concern/mdma.html. (Accessed 7 April 2008).

16

[13] Microgram Bulletin. Ecstasy Mimic Tablets (actually containing BZP,
l-(3-Trifluoromethyl)phenylpiperazine (TFMPP), Caffeine , and 1,4-Dibenzylpiperazine)
in Olathe, Kansas. Vol 42, No 5, May 2009. Available at
http://www.usdoj.gov/dea/programs/forensicsci/microgram/mgOS09/mg0509.pdf
(Accessed 2 Aug 2009).

[14] Johnston LD, O'Malley PM, Bachman JG, Schulenberg JE. "Various Stimulant
Drugs Show Gradual Decline among Teens in 2008, Most Illicit Drugs Hold Steady."
University of Michigan News Service: Ann Arbor, MI. Available at
http://www.monitoringthefutureorg. (Accessed 5 January 2009).

[15] Smith F P ed., Siegel JA series ed. Handbook of Forensic Drug Analysis. Elsevier
Academic Press, Burlington, MA; 2005.

[16] Parrott AC. Recreational Ecstasy/MDMA, the Serotonin Syndrome, and Sertonergic
Neurotoxicity. Pharrn Biochem Behav 2002; 71: 837-844.

[17] Palhol F, Boyer S, Naulet N, Chabrillat M. Impurity Profiling of Seized MDMA
Tablets by Capillary Gas Chromatography. Anal Bioanal Chem 2002; 374: 274-281.

[18] Gimeno P, Besacier F, Bottex M, Dujourdy L, Chaudron-Thozet H. A Study of
Impurities in Intermediates and 3,4-Methylenedioxymethamphetamine (MDMA)
Samples Produced via Reductive Amination Routes. Forensic Sci Int 2005; 155: 141 -
157.

[19] Nie ZL, Wen J, Sun H. Phylogeny and Biogeography of Sassafras (Lauraceae)
Disjunct Between Eastern Asia and Eastern North America. Plant Systematics and Evol
2007; 267: 191-203.

[20] Swist M, Wilamowski J, Parczewski A. Determination of Synthesis Method of
Ecstasy based on the Basic Impurities. Forensic Sci Int 2005; 152: 175-184.

[21] Swist M, Wilamowski J, Zuba D, Kochana J, Parczewski. Determination of
Synthesis Route of 1-(3,4-Methylenedioxyphenyl)-2-propanone (MDP2P) based on
Impurity Profiles of MDMA. Forensic Sci Int 2005; 149: 181-192.

[22] Li J]. Name Reactions: A Collection of Detailed Reaction Mechanisms. Springer
Berlin Heielberg, New York; 2006.

[23] Weyermann C, Marquis R, Delaporte C, Esseiva P, Lock E, Aalberg L, Bozenko Jr.

J S, Dieckmann S, Dujourdy L, Zrcek F. Drug Intelligence based on MDMA Tablets
Data: 1 Organic Impurities Profiling. Forensic Sci Int 2007; 177: 11-16.

17

[24] Microgram Bulletin. Ecstasy Combination Tablets (containing MDMA and
Methamphetamine) in Haltom City and Fort Worth Texas. Vol 42, No 4, April 2009.
Available at
http://www.usdoj.gov/dea/programs/forensicsci/microgram/mg0409/mg0409.pdf.
(Accessed 2 Aug 2009).

[25] Teng SF, Wu SC, Tsay WI, Liu C. The Composition of MDMA Tablets Seized in
Taiwan. Huaxue 2005; 63: 468-480.

[26] van Deursen MM, Lock ERA, Poortman-van der Meer AJ. Organic Impurity
Profiling of 3,4-Methylenedioxymethamphetamine (MDMA) Tablets Seized in the
Netherlands. Sci Justice 2006; 46: 135-152.

[27] Kuwayama K, Tsujikawa K, Miyaguchi H, Kanamori T, Iwata Y, Inoue H, Saitoh S,
Kishi T. Identification of Impurities and the Statistical Classification of
Methamphetamine using Headspace Solid Phase Microextraction and Gas
Chromatography-Mass Spectrometry. Forensic Sci Int 2006; 160: 44-52.

[28] Koester CJ, Andresen BD, Grant PM. Optimum Methamphetamine Profiling with
Sample Preparation by Solid-Phase Microextraction. J Forensic Sci 2002; 47: 1002-1007.

[29] Kongshaug KE, Pedersen-Bjergaard S, Rasmussen KE, Krogh M. Solid-Phase
Microextraction/Capillary Gas Chromatography for the Profiling of Confiscated Ecstacy
and Amphetamine. Chromatographia 1. 999; 50: 247-252.

[30] Bonadio F, Margot F, Delemont O, Esseiva P. Optimization of HS-SPME/GC-MS
Analysis and Its Use in the Profiling of Illicit Ecstasy Tablets (Part 1). Forensic Sci Int
2009; 187: 73-80.

[31] Bonadio F, Margot P, Delemont O, Esseiva P. Headspace solid-phase
Microextraction (HS-SPME) and liquid-liquid extraction (LLE): Comparison of the
Performance in Classification of Ecstasy Tablets (Part 2). Forensic Sci Int 2008; 182: 52-
56.

[32] Hoang TH, Sharma R, Susanto D, Di Maso M, Dwong E. Microwave-Assisted
Extraction of Active Pharmaceutical Ingredient from Solid Dosage Forms. J Chromatogr
2007; 1156: 149-153.

[33] Bieri S, Ilias Y, Bicchi C, Veuthey J, Christen P. Focused Microwave-Assisted
Extraction Combined with Solid-Phase Microextraction and Gas Chromatography-Mass

Spectrometry for the Selective Analysis of Cocaine from Coca Leaves. J Chromatogr
2006; 1112: 129-132.

18

[34] Carro AM, Lorenzo RA, Fernandez F, Phan-Tan-Luu R, Cela R. Microwave-
Assisted Extraction Followed by Headspace Solid-Phase Microextraction and Gas
Chromatography with Mass Spectrometry Detection (MAE-HSSPME-GC-MS/MS) for
Determination of Polybrominated Compounds in Aquaculture Samples. Anal Bioanal
Chem 2007; 388: 1021-1029.

19

Chapter 2 Theory
2.1. Microwave-Assisted Extraction

Microwave-assisted extraction (MAE) utilizes microwave energy to extract a
sample from a matrix into solution. The solution is heated under pressure which allows
for the solvent to be heated above its atmospheric pressure boiling point resulting in
efficient extractions [1].

The heating of the sample takes place through two mechanisms which occur at the
same time: diopolar rotation and ionic conduction [2,3]. The dipolar rotation
phenomenon is due to solvent’s molecular dipoles aligning with the electrical field. As
the electric field oscillates, the dipoles are forced into motion to stay in alignment, thus
creating friction and heating the solvent. Ionic conductiOn is caused by the
electrophoretic movement of the ions as a result of the applied electric field. Ions with a
small charge will generally move more slowly than ions with a higher charge. Also, as
the mass of the ion increases, the movement of the ion decreases. As the solvent resists
this movement, the friction created heats the solvent [2,3]. Due to these heating
mechanisms, the solvent is theoretically heated more uniformly than with conventional
heating methods such as a hot plate.

Several considerations must be taken into account when choosing a solvent for
MAE. The first is the polarity of the solvent. As mentioned earlier, the dipole of the
solvent molecules aligns with the electric field. A more polar molecule with a larger
dipole is more vigorously realigned with the electric field creating more heat than a non-
polar molecule with a smaller dipole. Therefore, polar solvents such as alcohols and

water will absorb more microwave energy than non-polar solvents such as hexane [1].

20

The dielectric loss coefficient (3”) indicates the ability of a material to absorb
microwave energy and convert it to heat. Molecules with a larger dielectric loss
coefficient are able to absorb the microwave energy more effectively therefore achieving
more optimal heating than molecules with a lower dielectric loss coefficient. The
dielectric constant (e’) is the ability of the material to be polarized by the electric field
[1]. The ratio of the dielectric loss coefficient and the dielectric constant, known as the
loss tangent or tangent delta (tan 8), describes the material’s ability to convert

electromagnetic energy to heat [1]. With a dipole moment of 1.87, a dielectric constant

of 78.3, and a tan 5 of 1570 x 104, water is a good solvent for microwave chemistry

because it adequately absorbs the microwave energy and converts it to heat. In contrast,
non-polar solvents which are not heated efficiently by microwave energy have lower
dipole moments and dielectric constants. For example, hexane has a dipole moment of
less than 0.1 and a dielectric constant of 1.88 [1].

Commercially available microwave lab stations are generally used to perform
microwave chemistry due to the dangers involved with the higher pressures, sometimes
as much as 65 bar, that can be achieved. After the sample and solvent have been placed
in the microwave vessels, the vessel is assembled and placed on a rotor inside the unit to
allow rotation of the vessels during the extraction. This overcomes the limitation of a
non-uniform electrical field by altering the location of the microwave vessels throughout
the extraction. The temperature is monitored during the extraction and microwave
energy can be supplied at varying levels to ensure that the system remains at the desired
temperature. After the extraction, the samples are cooled and then analyzed directly or

subsequently extracted by additional procedures.

21

2.2. Headspace Solid-Phase Microextraction

Headspace solid-phase microextraction (HS-SPME) is an analytical extraction
technique that utilizes a polymer coated fiber to extract analytes from the sample
headspace. A solid sample or aqueous sample is placed in a vial with a septum cap. The
fiber is placed in the headspace above the sample for a specified extraction time at a
specified extraction temperature. When sampling above an aqueous solution, the
analytes move from the solution into the headspace. Then, from the headspace, the

analytes absorb onto the fiber. Figure 2.1 shows a schematic of HS-SPME.

   

Fiber holder ——>

Fiber—

Path ofanalyte
movement

 

Sample solution ——

 

Figure 2.1: Schematic of HS-SPME with the arrows representing the movement of the analytes

At the end of the extraction, the fiber is retracted, removed from the sample, and
typically placed in the heated inlet of a gas chromatograph for analysis. In the GC inlet,
the analytes are thermally desorbed from the fiber and carried onto the column in the
flow of the mobile phase for subsequent separation [4].

Headspace solid-phase microextraction is an equilibrium based technique. During

the process, analytes form an equilibrium among three phases: the fiber coating and the

22

aqueous phase, the headspace and the aqueous phase, and the fiber coating and the
headspace. The concentration of analytes in the different phases at equilibrium is given
by Equation 2.1 [5]

Col/S = Cth + CSVS + Cfo (2.1)

where, C0 is the initial concentration Of the analyte in the aqueous sample, variables C

and Vare the concentrations and volumes, respectively, of the analyte in the different
phases that are represented by the following subscripts: h corresponds to the headspace, 5
corresponds to the sample matrix, and f corresponds to the fiber.

The equilibrium formed by the analyte is based on the partition coefficients of the

analyte between the phases. The partition coefficient between the phases is the ratio of

the concentration of the analyte in each Of those phases. The partition coefficient Kfl, for

an analyte between the fiber and the headspace is given by Equation 2.2

C
f
K =—
fh Ch

(2.2)

where, Cfis the concentration of the analyte on the fiber and C], is the concentration of

the analyte in the headspace. Similarly, the partition coefficient Khs of the analyte

between the headspace and the sample matrix is given by Equation 2.3

K — (2-3)
’15 CS

and the partition coefficient Kf, of the analyte between the sample matrix. and the fiber is
given by Equation 2.4

C
_ J:
K“ — C, (2.4)

23

Based on the partition coefficients, the mass Of analyte that is absorbed on the

fiber (nf) is summarized by Equation 2.5 [4].

n _ KfSI/J‘Vsco (25)
f Kfsz + Khth + V5

 

In the denominator, the terms Kfsz and [(th describe the analyte on the fiber

and in the headspace. Based on Equation 2.5, several alterations can be made to the

system to increase the mass of the analyte that is absorbed by the fiber [4]. One way is to

increase the concentration C0 of the analyte in the sample. If more of the analyte is

present, more is available to be extracted by the fiber.

Another way to increase the mass of an analyte extracted is to change the partition

coefficient between the fiber and the sample, Kfs. This can be accomplished by changing

the extraction temperature. For analytes with a high affinity for the fiber, when the
extraction temperature is increased, the partition coefficient of analyte between the
headspace and the solvent is increased. Therefore, a higher mass of the analyte is
extracted by the fiber. [4,5]. On the other hand, if an analyte has a lower affinity for the
fiber, a small mass of the analyte will absorb onto the fiber. When the temperature is
increased, the molecule is more likely to desorb from the fiber causing a lower mass of

the analyte to be extracted by the fiber.

The partition coefficient between the headspace and the solvent (Khs) can be

affected by pH. At lower pH values, molecules are protonated and, as the pH increases,

the molecules become deprotonated. The deprotonated molecules are more volatile (have

24

lower boiling points) and have a higher affinity for the headspace than the protonated
form. Therefore, a larger mass of the deprotonated analyte is extracted [5].

Agitating the solution can also affect the mass of analyte that absorbs onto the
fiber. As the analytes in the sample move from the solution into the headspace, stirring
the sample helps to ensure that additional analytes in solution are able to transfer into the
headspace. Therefore, the time required for equilibrium to be obtained between the
solution and the headspace is decreased. This is especially useful for analytes with lower
volatilities where only a small concentration of the analyte is in the headspace at
equilibrium. Once the analytes are in the headspace they transfer to the fiber quickly
because gasses have higher diffusion coefficients. [4].

The characteristics of the fiber can also be altered to affect the extraction of the

analyte. The volume of the fiber, V}; can be increased by lengthening the fiber or

increasing the thickness of the fiber’s coating. However, a larger fiber volume results in
a longer equilibration time because more time is required for the analyte to adsorb into
the fiber pores which are not as accessible in a thicker fiber [5]. When a thicker fiber
coating is used, the molecules take longer to desorb from the fiber, lengthening the
analysis time.

The type of fiber coating can affect the mass of the analyte extracted by changing

the partition coefficient between the fiber and the sample, Kfs [4,5]. Different coatings
have different chemical properties to extract different classes of analytes. Several types
of fibers are commercially available. Polydimethylsiloxane (PDMS) is a common base

used for the fibers. This coating is a non-polar coating that is used for extracting non-

polar compounds and semi-polar compounds such as aromatics and esters. A

25

divinylbenzene (DVB) coating is often used in conjunction with PDMS coatings to
broaden the range of analyte polarities that are extracted. The DVB coating is used to
extract moderately polar compounds such as amines. Analytes adsorb, or are retained, in
the pores of the DVB coating. This increases the sensitivity of the fiber to analytes

present at trace levels. However, one drawback to the DVB coating is that it is fragile

and can be stripped off the fiber easily. A CarboxenTM (CAR) coating is also used with a

PDMS coating. CarboxenTM is made up of pores of various sizes and therefore allows

for the adsorption and extraction of highly volatile compounds [4]. A combination fiber
of the various coatings mentioned (PDMS, CAR, and DVB) is available and expands the
range of analytes that can be extracted in a single extraction.

Because HS-SPME is a non-exhaustive extraction technique, the sample is not
used in its entirety allowing for re-testing if necessary. In addition, because of the ability
to pre-concentrate the analytes on the fiber, HS-SPME provides good sensitivity for the
extraction of compounds present in the sample at trace levels, for example trace level
impurities in MDMA tablets.

2.3. Gas Chromatography-Mass Spectrometry

Gas chromatography-mass spectrometry (GC-MS) is a common analytical
technique used to separate a mixture into its components and to detect those components.
Gas chromatography is a separation technique that is based on the interaction between the
mixture’s components and the mobile and stationary phases. After separation by the GC,
the components enter the detector which, in the case of GC-MS, is a mass spectrometer.
In the mass spectrometer, the analytes are ionized and fragmented, and the ions are

separated according to the mass to charge (m/z) ratios.

26

2.3.1. Gas Chromatography
A schematic of a gas chromatograph (GC) is shown in Figure 2.2 with the major

components labeled.

FIOW I Injector
controller It

CD & fl Detector

' Q

I
Column

 

 

 

 

 

 

 

-Column
oven

 

 

 

 

 

 

Carrier gas
cylinder

Figure 2.2: Schematic of a gas chromatograph

An inert carrier gas, or mobile phase, is required for CC analysis to move the
sample through the system. Helium, hydrogen, and nitrogen are all common carrier
gasses [6]. The flow rate and pressure of the carrier gas are regulated using gauges and
controllers. For many GC-MS applications, flow rates are typically 1 mL/min.

The sample is introduced to the GC through the inlet. In order for a sample to be
analyzed by GC, it must be volatile and thermally stable. The inlet is kept at high
temperatures, generally at least 50 °C above the boiling point of the analytes to ensure
that the sample volatilizes [6]. When an injection is performed, it is important that it be
completed quickly to allow the sample to move onto the column in a tight band. If a
sample volume is too large or the injection is made slowly, the sample spreads resulting
in band broadening which causes peaks to become broader, affecting peak shape and

resolution in the final chromatogram.

27

Sample injections can occur in the form of a liquid sample using a standard GC
syringe (e.g. sample dissolved in solvent), a gas sample using a gas-tight syringe (e.g.
from direct headspace sampling), or desorbed from a SPME fiber. Common inlets for
liquid and gas injections contain a septum and a glass liner. The septum seals the inlet to
prevent air from entering the instrument. The glass liner provides an inert surface in

which to inject and volatilize the sample without retaining the sample. For SPME fiber
analysis, a Merlin MicrosealTM is used in place of the septum and a narrower glass inlet

liner is used. The Merlin plays a similar role to the septum. The glass liner is narrower
to focus analytes onto the column that desorb slowly from the fiber.

Once vaporized in the inlet, the sample is carried onto the column which is
housed in an oven. In the column, the sample components interact with the stationary
phase which slows the sample so that it does not travel through the column as quickly.
Different components of the sample interact with the column stationary phase to different
extents. Some components will have strong interactions with the column and will be
slowed more than components that spend little time interacting with the stationary phase.
This causes the separation of the sample into its individual components.

Several different stationary phases are available for columns. Polydimethyl
siloxanes are a common group of stationary phases with the general form shown in
Figure 2.3. In the polydimethyl siloxane coating, the R-groups are all methyl (-CH3)
groups making the stationary phase non-polar. For other stationary phases, the R-groups
can be changed to a different group, such as a phenyl (-C6H5) group, to make the

stationary phase more polar [6].

28

 

 

 

:0

Q

O
”—21—”

O

c_n

JD

 

 

L. _ n
Figure 2.3: Polydimethyl siloxane phase in CC columns

The column is contained within an oven which heats the column at a specific rate
as determined by the user. Temperature programming can increase the speed of analysis
by increasing the temperature of the oven. When the analysis time is shorter, the sample
spends less time on the column, and therefore band broadening is decreased. Often, the
oven temperature will start low (e.g. 40-60 °C). After a given amount of time, the oven
will start to heat the column at a given rate, for example 10 °C/minute, until reaching the
desired final temperature (e.g. 280 °C). The different temperatures allow for the
separation of the components based on their boiling points. Components with lower
boiling points interact less with the column at low temperatures and elute from the
column first. Components with higher boiling points interact more with the column and
do not reach the detector until the higher oven temperatures are reached [6].

The separated sample components are then carried into the detector. The result of
the GC analysis is a chromatogram that plots the abundance of molecules detected
against the time at which they were detected, or the retention time. Each set of analytes
(in sufficient concentration) that go through the detector appears in the output
chromatogram as a peak. The retention time of the analyte will change when the
temperature program or type of column are changed because the analyte will interact

differently with the column.

29

2.3.2. Mass Spectrometry

A schematic of the sections of the mass spectrometer (MS) are shown in Figure
2.4. As shown in the figure, the system is under vacuum which typically operates at
pressures of 30-40 mTorr. This greatly lengthens the mean free path of the sample
molecules. In other words, the length of time between analyte collisions is much longer
leading to fewer collisions between ions. When the ions collide, they are neutralized and
therefore are not detected. Also, the vacuum reduces the possibility of contamination
from the environment and protects surfaces from water vapor that would otherwise cause

corrosion [6].

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Pump

 

 

 

:-
1

1 Vacuum
I

I

I

I-

Figure 2.4: Schematic of a mass spectrometer

The GC column feeds into the MS through the transfer line. The transfer line is
kept at high temperatures (e.g. 300 °C) to ensure that the separated components are not
lost through condensation. The end of the column is located at the ion source to allow for
the ionizatiOn of the sample.

One of the most common forms of ionization is electron ionization. This
ionization source contains a heated filament that releases electrons at a particular energy,
often 70 eV. The sample is introduced into the path of the electrons where ionization

occurs. Typical bond energies in a molecule can range from 10 to 20 eV. Therefore, the

30

70 eV electrons supply ample energy to the molecule to ionize and fragment the
molecule. Positive ions are formed during ionization due to the loss of an electron.
Compared to the formation of positive ions, the formation of negative ions is inefficient
and therefore few negative ions are produced. Ions of multiple charges can be formed
during the ionization process; however, during electron ionization, singly charged ions
are the most common. The fragments are useful in determining the structure of the
sample molecule because the fragmentation pattern of a molecule is consistent under the
same conditions [7].

A series of negatively charged focusing lenses attract the positive ions and focus
them into a thin beam for transfer into the mass analyzer. Ion trap mass analyzers contain
two end caps, which have openings to allow for the entry and exit of the ions and a ring

electrode for RF voltage oscillation. Figure 2.5 shows a cross-section of an ion trap mass

analyzer.
Ring Electrode
End Cap End Cap
\1
From Ion :> To
Source Detector

 

 

 

Ring Electrode Ions

Figure 2.5: Schematic of an ion trap mass analyzer

The ions of different masses are held in the middle of the trap and move around in
a figure-eight trajectory because of the voltage applied to the trap [7]. Inside the trap,

helium is present to reduce the energy of the ions through collisions. This ensures that

31

the ions stay in a tight group in the center of the trap. At a particular RF voltage applied
to the ring electrode, ions of a certain mass will become destabilized and leave the trap
through an aperture in the exit end cap. As the voltage is cycled, all masses within a
given range (for example 50-650 atomic mass units, amu) will be detected once in the
cycle (if the mass is present) [7].

As the ions leave the mass analyzer they can be detected by a continuous dynode
electron multiplier (EM) which converts the ions into a signal of electrons. The EM has a
curved conical appearance and can have a voltage of -10 kV at the opening which
changes to +10 kV at the end of the detector when operating in positive ion mode [6]. As
the ions hit the surface of the EM, electrons are ejected. This starts a cascade of
electrons. The ejected electrons are attracted to another region of the EM with a more
positive voltage. The electrons strike the next section causing more electrons to be
ejected. At the end of the multiplier, the electrons ejected constitute the current that is
sent to the amplifier system and then to the data system.

The process of ionization, mass analysis, and detection can occur several times a
second and is performed throughout the GC analysis. Therefore, every completely
separated component of the sample will be detected by this process independently of the
other components in the order they elute from the column.

The output of the mass spectrometer is a mass spectrum which contains the mass
to charge ratios of the ions and the abundance at which they were detected. In a GC-MS
analysis, every time point in the GC chromatogram has a corresponding mass spectrum.

This allows for the definitive identification of compounds. The retention time of the

32

analyte from the GC and the fragmentation pattern from the MS is unique to that
particular molecule.
2.4. Experimental Design

Experimental designs are used in research for many different applications. These
statistical designs allow the experimenter to learn more about a system or procedure in
fewer experiments than with a one-at-a-time experimental set-up, thus saving time and
money. Experimental designs can be used to identify experimental parameters that affect
the outcome, optimize important parameters in a process, or improve the robustness of a
procedure [8,9]. Often, a screening design will be used to determine parameters that have
a significant effect on the outcome followed by an optimization design to determine the
optimum settings of the significant parameters.

In the process of setting up an experimental design, the parameters, or effects, to
be studied are selected and the levels of the parameters are set. Levels are the number of
values at which the parameter will be studied. For example, in a two-level design, high
and low values of a parameter are studied. Also, the responses, or outcomes, are chosen
based on what system is being studied.

Next the set of experiments are planned. Often the set of experiments to be
performed are randomized to reduce experimental bias. If the set of experiments are to
be completed on different days or are to utilize different batches of materials, the design
can be divided into smaller sections called blocks. Dividing the experiments into blocks
takes the differences between days or batches of material into account in the data

analysis, often giving a more accurate view of the effects on the response. Some designs

33

employ a randomized block design in which each level of each parameter occurs once
and only once in each block [10].

The confounding of the effects of parameters and interactions between parameters
must been taken into account. If the parameters or interactions are confounded, it means
that the data analysis cannot separate the effects of the confounded parameters or
interactions, and they appear as one effect on the response [11]. The resolution of the
design is related to the confounding. In a resolution 111 design, the effects of the main
parameters are not confounded with other main parameters. However, one or more of the
interactions may be confounded with a main parameter or with a two-factor interaction.
In a resolution IV design, main parameters are not confounded with other main
parameters or two-factor interactions. However, two-factor interactions may be
confounded with one another. In a resolution V design, there is no confounding among
the main parameters or the two factor-interactions.

2.4.1. Screening Design

Screening designs are employed by experimenters to examine parameters to
determine which have a significant effect on the outcome. There are several types of
screening designs including full factorial, fractional factorial, and Plackett-Burrnan
designs [8]. Full factorial designs allow for the determination of the effects of the
parameters and the interactions between the parameters on the experimental response.
2.4.1.1. Full Factorial Design

For full factorial designs, the number of experiments to be performed is based on

Equation 2.6

E=KN (2.6)

34

where, E is the number of experiments, K is the number of levels, and N is the number of
parameters. For a two-level design with four factors, 16 experiments must be performed.
Fractional factorial designs, in which only a subset of the experiments in the full factorial
design is performed, can be used when many parameters are being studied. However, not
as much information about each parameter and interaction is learned from the fractional
factorial design when compared to the information gained from a fiill factorial design.

In fiIll factorial designs every combination of parameters at the various levels are
studied. For example, for a two-level, three-parameter study, Table 2.1 lists an example
set up of experiments. In the table, +1 indicates the high level and -1 indicates the low
level. The levels are then translated into experimental values. For example, parameter A
may be extraction temperature with a high value of 120 °C and a low value of 80 °C,
parameter B may be extraction time with a high value of 20 minutes and a low value of
10 minutes, and parameter C is the concentration of the sample in solution with a high
value of 2 M and a low value of 1 M. So, for experiment 1, an extraction temperature of
120 °C, an extraction time of 20 minutes, and a 2 M solution are used to complete the
experiment.

Table 2.1: Example set up of full factorial design with three parameters
Experiment Parameter A Parameter B Parameter C

 

+1 +1 +1
+1 +1 -1
+1 -1 +1
-1 +1 +1

+1 -1 -1
-1 +1 -1
-1 -1 +1
-I -I -I

OONONMAWNi—t

35

2.4.1.2. Analysis of Full Factorial Design

After the design has been set up and the experiments have been completed, the
experimental responses are used to build mathematical models using linear regression
techniques. The model allows for the estimation of values for other settings of
parameters not tested and is used to identify which parameters have a significant effect on
the response [8].

Often, multivariate analysis of variance (MANOVA) is used to determine which
parameters have a significant effect on the response when more than two variables are
studied. The MANOVA calculations are a statistical technique that is used to separate
variation due to random error (or uncontrolled factors) from variation due to changing a
controlled factor. The calculations then determine if the change in the outcome due to the
change in the control factor is significant.

The MANOVA calculations begin with the calculation of the sum of squares (SS)
for each response for the parameters, the interactions between the parameters, the error
(or residuals), and the total for the design based on the equations given in Table 2.2 [12].
Only two parameters are shown in the table; however, the calculations are the same if

more parameters and interactions are involved in the design.

36

Table 2.2: Calculations for the sum of squares

 

 

 

Source of Variation Sum of Squares
55 = Exit + nyz __ (2)02
Parameter A A n n N

 

 

55,, =(2y31 + 23132) _ (2y)Z

 

 

 

 

 

Parameter B n n N
_ 23’3131 2313231 23512132 23512232
SSAB — + + +
I . "A131 "A231 "A132 "A232
nteraction AB (2 ) 2
- A}; - 55,, — 553
Error SSError = SSTotal _ (SSA + 558 + 55.48)
(2302
T0131 SSTotal = (Z yZ) _ N

 

In the table, variable y corresponds to the observed responses of the design. For

the parameters, the subscripts A1 and A2, for example, correspond to the low level and
high level of the parameter studied. Therefore y,“ corresponds to the responses from the

design when parameter A was at its low level. The variable n corresponds to the number
of experiments at the particular level and the variable N corresponds to the total number
of experiments performed. Similar designations are used for parameter B. For the
interactions, y is again the observed response. The subscript A18], for example,
corresponds to the response when both A and B were at the low levels.

Next, the degrees of freedom, df, are calculated according to the equations given

in Table 2.3. Degrees of freedom are the number of variables that are available to fit the

37

model. In the equations, the variable, T, is the number of levels of each parameter with

the subscript corresponding to the parameter.

Table 2.3: Calculations for the degrees of freedom
Source of Variation

 

Degrees of freedom

 

ParameterA dfi4 = TA — 1
Parameter B df B = 73 — 1
Interaction AB de3 = (TA - 1) (TB — 1)
Error derror = df Total " (de + de + 51/23)
Total dfTotal = Number of Experiments — 1_

 

The mean squares (MS) and F-values are then calculated for each parameter,
interaction, and error (or residuals) based on the equations given in Table 2.4.

Table 2.4: Calculations for the mean squares and F-values
Source of Variation

 

 

 

 

Mean Square F-value
MS _ SSA F _ MSA
ParameterA A '" dfi4 '4 _ MSErmr
553 M53
M5 = — F = ——
Parameter B B de B M SError
SSAB MSAB
InteractionAB M543 .—.j d f], B FAB = MSError
SSError
EITOI' MSError _ df Error

 

The significance of the parameter’s or interaction’s effect can be determined by
comparing the calculated F-value to a critical F-value fiom a statistical table for the

required confidence level. If the calculated F value is smaller than the critical F value,

38

then the effect is not significant at that confidence level. If the calculated F value is
larger than the critical F value, the effect is significant at that confidence level [12].
2.4.2. Optimization Design

After determining the significant parameters in a procedure, the optimal setting of
the parameters can be determined using an optimization design. A circumscribed central
composite (CCC) design allows for the determination of second order interactions
(squared terms) in addition to the interactions between parameters.
2.4.2.1. Circumscribed Central Composite Design

A CCC design contains a factorial design in conjunction with a star design and
center points [9]. The factorial design is the same as discussed earlier. The star design
involves setting experimental points at i (1 according to Equation 2.7

a = [2N]“4 (2.7)

2 U4

where, N is the number of parameters. For two parameters, a = [2 ] = 1.41. For three

parameters, (1 equals 1.68. These experiments allow for the determination of the squared
terms in the model. The center points involve experimental parameters in the middle of
the design and the center point is often replicated to determine the error in the system
[11].

Figure 2.6 shows a schematic of the correlation between the types of experimental
points (factorial, star, and center) for a CCC design with two parameters being tested. In
the figure, the small circles represent the factorial points, the stars represent the star
points, and the diamond represents the center point. At each point, the first number in the
ordered pair corresponds to the setting of one parameter while the second number in the

ordered pair corresponds to the setting of the second parameter. The schematic is drawn

39

for a design with two parameters, but the same principal applies when three parameters

are being studied.

(-1,1) ,

 

 

(-1.41,0) (1.41,0)

 

 

 

 

('la'l)

 

(0,-1 .41)
Figure 2.6: Schematic of set up of experiments for CCC design

These theoretical values (:1, i a, and 0) are then converted to experimental
values to perform the set of experiments. For example, if extraction temperature with
high and low levels of 120 °C and 80 °C, respectively, is considered, the factorial points
can be converted to 80 °C (-1) and 120 °C (+1). The star points are converted to 72 °C (-
1.41) and 128 °C (+1.41) and the center point is converted to 100 °C (0). Table 2.4
shows an example set of experiments for 3 CCC design. Only one center point is shown

in this set up, however, the center point is usually replicated at least five times [9].

Table 2.5: Example set up of CCC design with two parameters

 

Experiment Parameter A Parameter B Type of Point

 

1 +1 +1 Factorial
2 +1 -1 Factorial
3 -1 +1 Factorial
4 -l -l Factorial
5 +1 .41 0 Star

6 0 +1 .41 Star

7 -l .41 0 Star

8 0 -1 .41 Star

9 0 0 Center

 

40

2.4.2.2. Analysis of Circumscribed Central Composite Design

After the experiments are completed, the responses are determined and a
mathematical model is built for each response using linear regression analysis. Next,
each response is optimized individually based on whether the response is to be
maximized or minimized. The experimental data and model are used to determine the
optimum settings for the parameters that result in the maximum (or minimum) for each
response [13].

A desirability function combines the separate responses into a single function
[14]. This function is then used to optimize the parameters based on all responses, not
just each individually. Equation 2.8 is used to optimize the desirability of responses that

are to be maximized [13,14].

0 y < low
y — low 5 .
dmax = (——_) r _<_ S 2.8
high _ low low y high ( )
1 y > high

As the response is maximized, the desirability, d, approaches one. The variable,
y, is the predicted response from the model. The variable, high, is the value of y above
which the desirability is at its maximum or one. The variable, low, is the lowest
acceptable value of y, and any value lower would yield an unacceptable desirability of
zero [13,14].

The variable, 5, is the shape of the desirability function. When 5 is set at one, the
desirability function is linear. If s is less than one, the desirability is almost equally
acceptable over the range. If s is greater than one, only the values closest to the

maximum (or minimum) are acceptable, thus limiting the range of acceptable values

41

[13,14]. Figure 2.7 shows the shape of the desirability function when different values of

s are used.

 

s=0.l

05 S: 1.0

Desirability (d)

s= 10.0

 

 

 

 

 

low high

Predicted Response

Figure 2.7: Graph of desirability function for maximization at different values of 5
(adapted from references 13 and 14)

Equation 2.9 is used to optimize the desirability of responses that are to be

minimized [13,14].

1 y < low
y - high 5 .
d... = (___), s _<.
low - high low y high (2.9)
0 y > high

Overall, as the response is minimized, the desirability d approaches one. This
equation is the opposite of the equation for dmax. The variable y is the predicted response
based on the model. The variable low is the value of y below which the desirability is at
its maximum or one. The variable high is the highest acceptable value for y, and any
value above it would result in an unacceptable desirability of zero.

To determine the overall desirability D of the system, the set of parameters that

yields the highest value for D according to Equation 2.10 is determined [13,14]

. _1_
D = (c111 x 352 x ...x d1.” )(21) (24°)

42

where, (11, etc. is the desirability of the individual responses and I is the impact of each

response. The impact of each response is set on a scale of one to five according the
user’s determination of the importance of the response. Five indicates high importance
and one indicates low importance [13,14]. The resulting optimum set of parameters
theoretically gives the optimum results based on the range of parameters studied in the
design.
2.5. Retention Time Alignment

Chromatograms of replicates of the same sample can be retention time aligned to
overcome instrumental drift between analyses. The Line Up software (Infometrix,
Bothell, WA) utilizes a correlation optimized warping algorithm to align the
chromatograms. For this algorithm, a target chromatogram is chosen, and the
chromatograms to be aligned are compared to the target one at a time. The alignment
starts at the end of the chromatogram and then moves towards the beginning. The
algorithm divides the chromatogram into sections. The user defined “slack” parameter
sets the number of data points to be included in each of the sections. The user defined
“warp” parameter specifies how many data points the section can be stretched or
compressed when matching the chromatogram to the target. The closeness of the match
is determined by calculating the Pearson product moment correlation (PPMC) coefficient
(discussed in section 2.6.) between the target and the aligned chromatogram for each
section. When the highest PPMC coefficient is determined for the section, the algorithm
moves to the next section of the chromatogram, again maximizing the PPMC coefficient.
This process continues through the length of the chromatogram, resulting in the aligned

chromatograms. The process is repeated for all other chromatograms in the sample set.

43

2.6. Pearson Product Moment Correlation Coefficients
Pearson product moment correlation (PPMC) coefficients are used to determine

the similarity between two variables according to Equation 2.11.

r = Z{(x-f)(y-7)} (2.11)
$1206 - EJZIIZO' - 7J2]

 

 

For chromatographic applications, the x variables correspond to one
chromatogram and the y variable corresponds to the second chromatogram. The variable
r can have a value between +1 and -1. Correlation values greater than zero indicate a
positive correlation, with values from 0.8-] indicating strong positive correlation. Values
less than zero indicate a negative correlation, while a correlation value of zero indicates

no correlation [15].

44

2.7. References

[1] Kingston HM ed, Haswell SJ ed. Microwave-Enhanced Chemistry: Fundamentals,
Sample Preparation and Applications. American Chemical Society, Washington, DC;
1997.

[2] Kaufrnann B, Christen P. Recent Extraction Techniques for Natural Products:
Microwave-Assisted Extraction and Pressurised Solvent Extraction. Phytochemical

Analysis 2002; 13: 105-113.

[3] Eskilsson CS, Bjorklund E. Analytical-Scale Microwave-Assisted Extraction. J
Chromatog A 2000; 902: 227-250.

[4] Wercinski SA ed. Solid Phase Microextraction: A Practical Guide. Marcel Dekker,
Inc., New York; 1999.

[5] Pawliszyn J. Solid Phase Microextraction: Theory and Practice. Wiley-VCH, New
York; 1997.

[6] Skoog DA, Holler FJ, Nieman TA. Principles of Instrumental Analysis. Thomson
Learning, Inc., United States; 1998.

[7] De Hoffmann E, Stroobant V. Mass Spectrometry: Principles and Applications. John
Wiley and Sons, Inc., England; 2007.

[8] Araujo PW, Brereton RG. Experimental Design 1. Screening. Trends Anal Chem
1996; 15: 26-31.

[9] Araujo PW, Brereton RG. Experimental Design 11. Optimization. Trends Anal Chem
1996; 15: 63-70.

[10] Peterson RG. Design and Analysis of Experiments. Marcel Dekker, New York City;
1985.

[11] Antony J. Design of Experiments for Engineers and Scientists. Butterworth-
Heinemann, Amerdam; 2003.

[12] Frigon NL, Mathews D. Practical Guide to Experimental Design. John Wiley and
Sons, Inc., New York; 1997.

[13] Statgraphics. Design of Experiments -— Multiple Response Optimization. StatPoint,
Inc., Hemdon, VA; 2005.

[14] Derringer G, Suich R. Simultaneous Optimization of Several Response Variables. J
Qual Tech 1980; 12: 214-219.

45

[15] Devore JL. Probability and Statistics for Engineering and the Sciences, 4th Ed.
Duxbury Press, Belmont, CA; 1995.

46

Chapter 3 Materials and Methods
3.1. Sample Preparation
3.1.1. Simulated Sample

A simulated 3,4-methylenedioxymethamphetamine (MDMA) sample of known
components was prepared for use in the microwave optimization studies. Benzylamine
hydrochloride (0.5-2% of sample), 2-phenethylamine hydrochloride (0.5-2% of sample)
methamphetamine hydrochloride (0.5-2% of sample), MDMA (0.1% of sample), and
ephedrine (0.5-2% of sample) were purchased from Sigma-Aldrich (St. Louis, MO) and
used as received. Caffeine (92-98% of sample; Eastman, Rochester, NY) was included as
an adulterant. All components were homogenized with a mortar and pestle. Due to cost,
MDMA (0.1% of sample) was only used in the simulated tablet for the selection of the
buffer for microwave-assisted extraction (MAE).
3.1.2. MDMA Exhibits

Three exhibits of MDMA tablets were received from the Michigan State Police
Forensic Science Division. An exhibit is a set of tablets with similar physical properties
that is obtained by the police at one time from one location. For each exhibit, the
physical characteristics of the tablets were recorded, and several tablets from each exhibit
were homogenized with a mortar and pestle for use. Photographs of representative

tablets from each exhibit are shown in Figure 3.1.

47

 

Figure 3.1: Representative tablets from each exhibit: :1) exhibit MSU900-01 (pink, purple, and green);
b) exhibit T-17 (blue); c) exhibit T-27 (pink)

3.2. Optimization of Microwave-Assisted Extraction Procedure
3.2.1. Optimization of Extraction Buffer

An Ethos EX Microwave Lab Station (Milestone Inc.; Shelton, CT) was used for
all microwave-assisted extractions. When the MDMA sample is introduced to the buffer,
the sample has the potential to alter the pH of the solution. Buffers were used as the
extraction solvent instead of water alone due to the ability of the buffer to maintain the
pH of the solution. Based on a review of the literature, three different buffers at three
different pH values and concentrations were investigated [1-5].

Preliminary studies were performed with a 0.05 M carbonate buffer, pH 10, which
was prepared using sodium bicarbonate (Sigma) and 2 M sodium hydroxide (Spectrum,
New Brunswick, NJ). Phosphate buffers at concentrations of l M, 0.5 M, and 0.1 M

were prepared using potassium phosphate-monobasic, KH2P04 (Mallinckrodt, Paris, KY)
and sodium phosphate-dibasic, NazHPO4-7H20 (Jade Scientific, Canton, MI). For each

concentration, buffers at three different pH values (6, 7, and 8) were prepared using 2 M
sodium hydroxide (Spectrum) to adjust the pH. Tris buffers at concentrations of 1 M, 0.5

AL I ' A,

M, and 0.1 M were prepared using trisGIydrux, ...... J ) (Mallinckrodt).

48

For each concentration, buffers at three different pH values (7, 8, and 9) were prepared
using concentrated hydrochloric acid (EM; Gibbstown, NJ) to adjust the pH.

Extractions were performed using each buffer at each pH and concentration. A 75
mg mass of the simulated sample was transferred to a TeflonTM microwave vessel
(Milestone Inc.) and 10 mL of the appropriate buffer was added. The vessel was then
assembled according to the manufacturer’s recommendations and a fiber optic
temperature probe was inserted into the reference vessel to accurately monitor the
temperature during the extraction. Figure 3.2 shows a schematic of the assembled

reference vessel. Spring

  

Shield

Th ermowell

 

Sample/buffer Vessel
solution

Figure 3.2: Schematic of assembled microwave vessel with sample
After the assembled microwave vessels were placed in the microwave unit, the
instrument was programmed to heat for 15 minutes (ramp time) to 100 °C (extraction
temperature) and hold at 100 °C for 15 minutes (extraction time). During the extraction,
the vessels were rotated in the microwave to allow for even heating of all samples. At the

end of the extraction, the vessels were allowed to cool to 50 °C before being opened.

49

Then, 5 mL of the extract were transferred to a 10 mL amber glass vial (Supelco,
Bellefonte, PA) containing a stir bar.

The solution was further extracted using a headspace solid-phase microextraction
(HS-SPME) procedure previously developed in our laboratory. The vial was pre-heated
at 70 °C for five minutes, with stirring. A StableFlex divinylbenzene/Carboxenm/
polydimethylsiloxane (DVB/CAR/PDMS) fiber (Supelco, Bellefonte, PA) was exposed
to the headspace for 20 minutes, with stirring. Finally, the fiber was retracted and
analyzed by gas chromatography-mass spectrometry (GC-MS) using a Thermo Focus gas
chromatograph with a Polaris Q mass spectrometer (Thermo Fisher Scientific Inc.,
Waltham, MA).

The TeflonTM microwave vessel wasc leaned by rinsing with distilled water,
acetone, and methanol. Next, 10 mL of fresh distilled water were added to the vessel
which was then assembled. The vessel was then cleaned in the microwave with a 10
minute ramp to 160 °C and a 20 minute hold at 160 °C. After cooling, the water in the
vessels was discarded, and the vessels were rinsed with fresh distilled water. The
cleanliness of the vessels was assessed by performing blank extractions which were ,
performed exactly as described for sample extraction, but with no simulated sample
present.

Triplicate extractions of the simulated sample followed by the blank extractions
were performed for each of the buffers. The optimum buffer was chosen based on the
number of simulated sample components extracted, the precision of the extraction, and

the level of simulated sample component carryover between extractions.

50

3.2.2. Determination of Significant Parameters

A full factorial experimental design was performed to screen for significant
parameters in the MAE procedure. Ramp time, extraction time, and extraction
temperature were studied, and the high and low values for each parameter were

determined based on practical limitations (Table 3.1).

Table 3.1: High and low parameters for full factorial screeninLdesign

 

Parameter High Value Low Value
Ramp Time (min) 20 10
Extraction Time (min) 20 10
Extraction Temperature (°C) 120 80

 

The experimental design was generated using Statgraphics Centurion software
(Version XV, Statpoint, Inc., Hemdon, VA). For the screening design, a block
randomized set of 16 experiments in four blocks was generated. Each block contained
four extractions: two center point extractions and two other extractions. Center point
extraction parameters included the middle point of the high and low values for each
parameter, in this case a 15 minute ramp time, 15 minute extraction time, and 100 °C
extraction temperature. The two other extractions tested the various combinations of the
high and low values for the parameters.

Preliminary extractions indicated that methamphetamine carryover was present in
the microwave vessel between extractions. This was potentially due to adsorption of the
sample onto the TeflonTM microwave vessel. To overcome the carryover problem, quartz
inserts were used which theoretically minimize carryover because the sample would not
adsorb onto the quartz. For this study, 50 mg of the simulated sample were placed in 5.5

mL of 1 M phosphate buffer (pH 8) in the quartz insert. The insert was then placed in the

51

TeflonTM microwave vessel with 10 mL buffer in the vessel. The buffer in the vessel

outside of the insert was required for accurate temperature monitoring by the fiber optic
probe in the therrnowell. The vessel was assembled according to the manufacturer’s
recommendation (Figure 3.3) and extracted using the ramp time, extraction time, and

extraction temperature specified in the full factorial design given in Appendix C.

Spring

  
  
  

Buffer

 

Sample/buffer
solution

Quartz insert Vessel
Figure 3.3: Schematic of assembled microwave vessel with quartz insert and sample

After MAE, 5 mL of the extract was transferred to an amber glass vial for
subsequent extraction by HS-SPME using a similar procedure as previously described
(section 3.2.1) except with a 40 minute extraction time rather than 20 minutes. All
extracts were analyzed by GC-MS.

Integrated peak areas of each component of the simulated sample were used as the
responses for the appropriate extraction. Using Statgraphics software, statistically
significant extraction parameters for each tablet component were determined based on
multivariate analysis of variance (MANOVA).

The quartz inserts were cleaned between each sample extraction by rinsing with

distilled water, acetone, and methanol. Next 10 mL distilled water were added to the

52

insert which was then placed in the vessel with 15 mL distilled water. The vessels with
the quartz inserts were then cleaned in the microwave with a 10 minute ramp to 130 °C
and a 20 minute hold at 130 °C. After the vessels cooled, the water was discarded, and
the inserts were rinsed with fresh distilled water. -

3.2.3. Optimization of Significant Parameters

After determining that ramp time, extraction time, and extraction temperature
were significant parameters for MAE, a circumscribed central composite (CCC)
optimization design was performed to determine the optimum settings for these
parameters. Statgraphics software was used to generate the CCC design with 23
randomized experiments (Appendix D). The values for the factorial points of the CCC
design were the same as for the full factorial design discussed in section 3.2.2. For the
star points, ramp time and extraction time were studied with at 23 minutes and 7 minutes
while extraction temperature was studied at 134 °C and 67 °C. Extractions were
performed exactly as described in section 3.2.2 except using the ramp rate, extraction
time, and extraction temperature specified in the CCC design. The inserts were cleaned
between each extraction using the procedure given in section 3.2.2.

As before, integrated peak areas of each simulated sample component were used
as the responses for the appropriate extraction. Using Statgraphics software,
mathematical models were developed for each sample component based on the peak area
for each extraction. The responses were then optimized based on whether the goal was to
maximize or minimize the peak area. A desirability function was used to combine the
desired responses (peak areas) into a single fiinction to optimize the extraction

parameters.

53

3.3. Optimization of Headspace Solid-Phase Microextraction Procedure

After selecting the optimum buffer for use in MAE, it was necessary to determine
the optimum HS-SPME extraction time and extraction temperature for future MAE/HS-
SPME extractions. For the optimization, a homogenized sample from one MDMA
exhibit (MSU900—01) was extracted using HS-SPME with no prior microwave extraction.
All HS-SPME extractions used a 23-gauge StableFlex divinylbenzene/
CarboxenTM/polydimethylsiloxane (DVB/CAR/PDMS) fiber (Supelco; Bellefonte, PA)
that was conditioned daily before use as recommended by the manufacturer. After
conditioning, the fiber was analyzed by GC-MS to ensure that the fiber was clean.

For the extraction, 50 mg the homogenized MDMA exhibit was placed in 5 mL of
l M phosphate buffer at pH 8 in a 10 mL amber glass vial containing a stir bar. The vial
was pre-heated by suspending it in a water bath at the specified extraction temperature for
five minutes with stirring. Extraction time and extraction temperature were studied
empirically; that is, one parameter was changed while the other was held constant.
Extraction times of 10-60 minutes in 10 minute increments were studied while holding
the extraction temperature at 70 °C. Then, extraction temperatures of 40-80 °C were
studied using an extraction time of 40 minutes. At the end of the extraction time, the
fiber was retracted and analyzed by GC-MS. The optimum HS-SPME extraction time
and temperature were determined based on the combination that offered a compromise
between the number and abundance of the impurities extracted and acceptable

chromatography.

54

3.4. Liquid-Liquid Extraction Procedure

The procedures for liquid-liquid extraction (LLE) were adapted from the method
developed by van Deursen et al. [1]. Phosphate buffer (0.33 M at pH 7) was prepared
using potassium phosphate-monobasic (Mallinckrodt) and sodium phosphate-dibasic
(Jade Scientific) with 2 M sodium hydroxide (Spectrum) used to adjust the pH.

For the extraction, 200 mg of either the simulated sample or the MDMA exhibits
were placed in 4 mL of the phosphate buffer. The sample was vortexed for 10 seconds
followed by sonication for 10 minutes and centrifugation for eight minutes. After adding
400 11L of toluene (Mallinckrodt) with eicosane (Aldrich) as an internal standard (0.020
mg/mL), the sample was gently agitated, then sonicated for 10 minutes, and centrifuged
for five minutes. The toluene layer was transferred to a GC vial insert (Restek, West
Chester, PA). Manual injections were made using 1 uL of sample with 0.5 uL air.

3.5. Gas Chromatography-Mass Spectrometry
A Thermo Focus gas chromatograph with a Polaris Q ion trap mass analyzer

(Thermo Fisher Scientific Inc.) was used for all analyses. The GC was equipped with a
inTM-Sms column (30 m, 0.25 mm id, 0.25 pm df; Restek). The mass spectrometer was

operated in full scan mode from 50-650 m/z with the electron ionization source operating
at 70 eV.

For HS-SPME extractions, a Merlin MicrosealTM septum replacement (Merlin
Instrument Company, Half Moon Bay, CA) was used instead of a traditional septum. A
narrow splitless inlet liner with an internal diameter of 0.8 mm was used for HS-SPME
extractions to better focus the components that desorb from the fiber onto the head of the

column. The GC-MS parameters used for HS-SPME experiments are given in Table 3.2.

55

Blank MAE/HS-SPME extractions were analyzed using the same GC-MS parameters for
samples with a minor change to the GC temperature program: the hold at the end of the
program was reduced to one minute for time efficiency.

For LLE, a BTO 17 mm CenterGuide septum (Restek) and a traditional
split/splitless liner were used. The GC-MS parameters used for LLE injections were

based on those reported by van Duersen et al. [1] and are given in Table 3.2.

Table 3.2: GC-MS parameters for HS-SPME analysis and LLE analysis

 

 

 

 

MAE/HS-SPME and HS—SPME LLE
Injection Port 260°C; splitless 1 minute, then 100:1 split 250°C; 50:1 split
Carrier Gas Helium, 1 mL/min Helium, 0.5 mL/min
Oven Program
Initial 60°C for 2 minutes 90°C for 1 minute
Ramp 8°C/minute 8°C/minute
Final 300°C for 15 minutes 300°C for 10 minutes
MS Transfer Line 275 °C 275 °C
[on Source 225 °C 225 °C
MS Solvent Delay 2 minutes 4 minutes

 

Using instrument software (Xcaliber 1.4, Thermo Fisher Scientific Inc.), peak
areas of the simulated sample components were integrated and used in subsequent data
analysis.

3.6. Comparison of Extraction Procedures

Replicate extractions of the simulated sample were performed to compare the
precision of the optimized MAE/HS-SPME procedure (four replicates) and the optimized
HS-SPME procedure (five replicates) with the LLE procedure based on the literature [1]
(three replicates). For each extraction procedure, the relative standard deviation (RSD) of
the integrated peak area for each component in the simulated sample was calculated.

The three homogenized MDMA exhibits were then extracted in triplicate by each

extraction procedure. The extraction procedures were evaluated based on the number of

56

impurities extracted, the overall chromatography, and the precision of the extraction for
each exhibit. Rather than calculate RSD values for the individual impurities in each
exhibit to assess precision, Pearson product moment correlation (PPMC) coefficients
were calculated between each pair-wise combination of replicates to assess the
correlation among replicate chromatograms. Prior to calculating PPMC coefficients,
chromatograms were retention time aligned using a commercially available retention time
alignment algorithm (Line Up, Infometrix, Bothell, WA).

Using the HS-SPME chromatograms, the three MDMA exhibits were compared
based on the identity of the impurities extracted by the three extraction procedures. The
origin of the impurities as well as the synthetic route used to synthesize the MDMA in the

exhibit was investigated.

57

3.7. References

[1] van Deursen MM, Lock ERA, Poortman-van der Meer AJ. Organic Impurity Profiling
of 3,4-Methylenedioxymethamphetamine (MDMA) Tablets Seized in the Netherlands.
Sci Justice 2006; 46: 135-152.

[2] Andersson K, Jalava K, Lock E, Huizer H, Kaa E, Lopes A, Poortman-van der Meer
A, Cole MD, Dahlen J, Sippola E. Development of a Harmonised Method for the
Profiling of Amphetamines IV. Optimisation of Sample Preparation. Forensic Sci Int
2007; 169: 64-76.

[3] Kuwayma K, Tsujikawa K, Miyaguchi H, Kanamori T, Iwata Y, Inoue H, Saitoh S,
Kishi T. Identification of Impurities and the Statistical Classification of
Methamphetamine using Headspace Solid Phase Microextraction and Gas
Chromatography-Mass Spectrometry. Forensic Sci Int 2006; 160: 44-52.

[4] Palhol F, Boyer S, Naulet N, Chabrillat M. Impurity Profiling of Seized MDMA
Tablets by Capillary Gas Chromatography. Anal Bioanal Chem 2002; 374: 274-281.

[5] Swist M, Wilamowski J, Zuba D, Kochana J, Parczewski. Determination of Synthesis
Route of l-(3,4-methylenedioxypheny1)-2-propanone (MDP2P) based on Impurity
Profiles of MDMA. Forensic Sci Int 2005; 149: 181-192.

58

Chapter 4 Results and Discussion
4.1. Sample Preparation
4.1.1. Simulated MDMA Sample

The components in the simulated sample, which was used to optimize the
microwave-assisted extraction (MAE) parameters, were chosen to provide a wide span of
retention times and peak abundances in the resulting impurity profiles. Benzylamine and
phenethylamine were chosen because of their structural similarity to methamphetamine,
MDMA, and impurities commonly observed in MDMA tablets. Methamphetamine and
caffeine were included because they are common adulterants added to the synthesized
MDMA before it is pressed into tablet form [1]. Ephedrine was included because it is a
common starting material for the synthesis of methamphetamine [2]. MDMA was
included to make the simulated tablet as realistic as possible; however, it was not
included in the data analysis due to the low quantity present in the sample.
4.1.2. MDMA Exhibits

The physical characteristics of the three MDMA exhibits used throughout this
study are given in Table 4.1. The average diameter, height, and mass were calculated

based on ten tablets selected from each of the exhibits.

59

Table 4.1: Physical characteristics of MDMA exhibits (averages based on ten tablets)

 

 

Exhibit Number,“ Tablet Tablet Tablet Average Mirage Average
Identity Tablets In Color Logo Sha Diameter Herght Mass
Exhilit pa (mm) (mm) (g)
P' k/ / Circular,
MSU900-OI 100 m green Alligator beveled 8.0 5.0 0.2705
purple
edge
Circular,
T- l 7 20 Blue Horses hoe beveled 8.0 4.0 0.2423
edge
Circular,
T-27 20 Pink Heart beveled 8.0 4.8 0.2693
edge

 

4.2. Optimization of Microwave-Assisted Extraction Procedure
4.2.1. Selection of Extraction Buffer

The goal of the buffer study was to determine the optimum buffer for use with
MAE. It was desired that the buffer extract all components of the simulated sample with
acceptable precision and no carryover between extractions. The simulated sample was
extracted in triplicate using the different buffers at different pH values and different
concentrations. For each set of triplicates, the percent relative standard deviations (RSD)
of the peak areas of the simulated sample components were calculated. Blank extractions
were performed and analyzed following each sample extraction and cleaning procedure
to determine if sample carryover was detected.
4.2.1.1. Investigation of Phosphate Buffers

Phosphate buffers at pH 6, 7, and 8 were studied to correspond to the buffering

range of phosphate which has a pKa of 7.2 [3]. Buffers of lower pH were not studied

because basic impurities, which are commonly found in MDMA tablets, are extracted
more efficiently at higher pH values [4]. Buffer concentrations of l M, 0.5 M, and 0.1 M

were studied based on a review of the literature [2,4,5]. The average peak areas and RSD

60

values for the simulated sample in phosphate buffers analyzed in triplicate are

summarized in Table 4.2.

Table 4.2: Average abundance and RSD values of simulated sample components in phosphate buffers

(average peak areas based on three replicates)

 

 

 

 

 

mod mo+mE ._ wwfim meoN; new wo+m2 .m of 0:6me
r. t a... 2.. 8s rimmed S: aseafim
2mm oo+mmm._ my.— oo+mo~._ o0.— oo+mwm._ o2 0E850fi80£02 w
mosm B+m2§m ow. ,_ _ ho+mwm6 mm .m mo+mwm._ ad 0£ES>505£A
mm .2 mo+mwoN .mdm wo+mmm.m om; mo+mmm.m md ofifissficom
omen wo+mw_ ._ 3.2 memm; mmdm wo+m§4 0.3 0:0me
2 .mm co+mma6 3.3m co+mw9w wmsm B+m§ ._ od. 0560,15
RH wo+mwm4u Sis wo+mvwd mms wo+m3\m od— ofififlofifimfiofi N.
a. a. ... a. a. * wd ogflxfivcuzm
fl .2 no+mwwd amas— ho+m_mm.v _~.wm B+mmo.m md opium—Ecum—
ESm 21mg)” 8.3 hc+mmmé 3.6m h¢+mmafi 93 «0&0me
.. ... .. a .. .. 0.2 £6210
mm SN 3+5 m ._ em .2 bo+mmmg Saw timovd od— 0£ESo€Em£02 o
.. .. a .. .. .. as baggage;
a. * _.. a * *. md oEEm—chom
NP— ao a0.— ao a0.— a0
am: ..\e 0w”.— u><m amy— ..\e on“. ”can amy— .x. 0m”.— H><m “vi bran—E I: Lou—.m—
E —.e .2 m6 2 _

 

 

 

 

 

* denotes that the impurity was not detected

** denotes that the impurity co-eluted with a siloxane peak from the fiber

61

As the concentration of the buffer increased, the average peak area of the
simulated sample components generally increased indicating a higher concentration of
sample was extracted. As the concentration of the buffer increased, the stability of the
solution increased which makes the solution more thermally stable. A more stable
solution would experience less degradation when exposed to microwave energy than a
less stable solution. Therefore, phosphate buffers of higher concentration are more
desirable.

At lower pH values, not all components of the simulated sample were extracted.

Also, as the pH of the buffer increased, the abundance of the peaks generally increased.

This is due to the pKa of the components. In a solution with the pH less than the pKa of

the component, the salt (or protonated) form of the component dominates the equilibrium
between the salt and the free base form. The protonated form is less volatile than the fi'ee
base form which indicates that the protonated form has a lower affinity for the headspace
than the free base form during headspace-solid phase microextraction (HS-SPME) [6,7].
Because the components of the sample generally move from the sample solution to the
headspace and then from the headspace to the fiber [7], more of the analyte in the
headspace implies that more of the analyte is available for HS-SPME. To illustrate the
effect of pH on the extraction of the simulated sample, Figure 4.1 shows chromatograms

of the simulated sample extracted from 1 M phosphate buffers at pH 6 and pH 8.

62

 

a) 2.00E8 ; *

Abundance

*-
methamphetamine

*

E caffeine

 

 

 

 

 

 

0.00E8 __1__._AL --,A.L.__. . -
2 7 12 17 22 27

Retention Time (min)

 

 

 

 

 

   

 

 

 

 

 

 

 

b) 2001-:8
t o
"=- .“5’
i s
>‘ B
8 3 .2 a. *
5 -D g g
'3 5’ e a a
5 2‘ ° .5 .s
.0 3 E .5 a1: 80:)
< a o a
.E’ «a 0
Q.* G.)

 

 

 

 

12 17 22 27
Retention Time (min)

\1 .

2

Figure 4.1: Chromatograms of the simulated sample in 1 M phosphate buffer at a) pH 6 and b) pH 8;
an * indicates that a peak was present in the blank

At pH 6, benzylamine, phenethylamine, and ephedrine were not extracted.
Methamphetamine and ephedrine both have a pKa of 10.0, but ephedrine was not

detected at pH 6 while methamphetamine was detected. This was due to the lower
boiling point of methamphetamine making it more volatile than ephedrine. The higher
methamphetamine volatility gives methamphetamine a higher affinity for the headspace
than ephedrine, therefore allowing methamphetamine to be extracted at higher

abundances.

63

For HS-SPME, acceptable RSD values usually fall between 1% and 10% [7]. The
RSD value for caffeine was high (greater than 27%) for all concentrations at pH 6, and
the RSD value for methamphetamine was high (greater than 10%) at 0.5 M and 0.1 M.
This may be due to the sample not being completely homogenized when the aliquots
were taken for extraction. Another contributing factor may be slight fluctuations in the
temperature of the water bath (plus or minus 2-3 °C) during HS-SPME. These factors
combined can decrease the precision observed because the extraction conditions are not
the same between replicate extractions.

At all concentrations of phosphate buffers at pH 6, there was no sample carryover
in the microwave vessel after the cleaning procedure. Despite the adequate cleaning of
the microwave vessels, the phosphate buffers of pH 6 were not chosen as the optimum
buffer because not all components of the simulated sample were extracted. Also, the
components that were extracted (methamphetamine and caffeine) were not extracted
precisely, as shown by the high RSD values.

For pH 7 phosphate buffers, phenethylamine was not extracted at any buffer
concentration. Overall, high RSD values (greater than 10%) were observed for
benzylamine, ephedrine, and caffeine (in l M and 0.5 M buffers at pH 7). Again, this
could have been due the lack of homogeneity in the sample or changes in the extraction
temperature (plus or minus 2-3 °C). At all concentrations, there was no sample carryover
in the microwave vessel after cleaning. However, phosphate buffers of pH 7 were not
chosen as the optimum buffer because not all components were extracted and

components that were extracted did not show good precision.

64

All simulated sample components were extracted using pH 8 phosphate buffers.
At pH 8, using the 0.5 M and 0.1 M concentration buffers, ephedrine co-eluted with a
siloxane from the fiber making accurate identification and peak area integration difficult.
Therefore the RSD value for ephedrine was not calculated for these concentrations. The
co-elution and higher fiber bleed may have been due to a new fiber being used. After
several extractions, there was no filrther co-elution between ephedrine and siloxane using
this fiber. At 0.5 M and 0.1 M, RSD values were high for benzylamine (greater than
15%) and phenethylamine (greater than 11%) again due to slight changes in the sample
and extraction temperature between replicate extractions. The RSD values were
acceptable (less than 10%) for all impurities at the 1 M concentration pH 8. The
concentration of the buffer can influence the partition coefficient between the sample and
the headspace affecting the equilibration between the two phases. The 1 M buffer is
more stable than the 0.5 M and 0.1 M buffers. Therefore, the simulated sample may be
more stable in the 1 M buffer during MAE, allowing for a more precise extraction.

The 0.5 M and 0.1 M buffers (pH 8) showed methamphetamine and benzylamine
carryover in the microwave vessel between extractions, and hence, these buffers were not
chosen as the optimum buffer. There was no sample carryover in the microwave vessels
between extractions with at 1 M concentration. Because the simulated sample is more
soluble in the higher concentration buffer, more of the sample goes into solution and less
is left in the microwave vessel following the extraction. The 1 M phosphate buffer at pH
8 was chosen as the optimum phosphate buffer for MAE because all components were
extracted with acceptable RSD values and there was no carry-over of the sample in the

microwave vessel between extractions.

65

4.2.1.2. Investigation of Tris Buffers

Tris buffers were studied at pH 7, 8, and 9 to correspond to the buffering range of

tris(hydroxymethylaminomethane) which has a pKa of 8.1 [3]. Buffers of higher pH

were avoided because methamphetamine and MDMA are extracted more efficiently at
higher pH values [4]. Based on a review of the literature, buffer concentrations of 1 M,
0.5 M, and 0.1 M were studied [5]. The average peak area and RSD values for the

simulated sample analyzed in triplicate in the tris buffers are summarized in Table 4.3.

66

Table 4.3: Average abundance and RSD values of simulated sample components in tris buffers

(average peak area based on three replicates)

 

 

 

 

 

 

om.m ho+mhwd m— .Vm Rim—mmd w~ .NN wO+mKoA 0.: oiotao
a... .I. I. .1. .I. .1. CA: 055915
_m.N ofmoo; 2 .N RTE _ ._ mm; o¢+m2 ._ 0.8 0£ES0£QENEOE o
wad wo+WVN._ VN; g w¢+mhv4 2 .3 wo+mmwg wd oEEflxfiococm
wc.v wceummmae vofi wo+m~mm.v S .v wgmmmd m6 oEEflchom
3.3» ho+mo¢d chflu wO+mGo._ VPMM ho+m©od 0.: 05.0th
a... 1. I. a... .I. .1. od— ofiufionmm
NNA: w$mmo._ mm.m wo+mNm.N cow wO+mmn._ 0.2 0EESOLQEMEOE w
... a. a. ... ... .._ w d OEEm—buvcocm
wmdN ho+mmbé wed 59.9: .w mvfim hgmhmg md DEEEEOm
mvd RIM—NOV .2.... 50+me .N .11. ho+mmofi 0.: 050taU
a. a a. a. ... ... S: 0E€0fim
thm ho+m3¥ .2.... ©O+m _ m .m .12.. wo+mm_ .m od— 03580:?«502 h
a. * a * * a. w .a uEEscEBr—vzm
odd 21mg.— ..cI. hgmmlwg .13.. bng m6 BEEN—Sum
N 0.— N O N 2 N 0 N 0.- fl 0
Dmm .x. on”. ”can 55¢ ..\a on“ H><m aid 3 0w“... ”.5”— av—n brine: in a eh: m
E m .e 2 m6 2 —

 

 

 

 

 

* denotes that the impurity was not detected

** denotes that the impurity co-eluted with a siloxane peak from the fiber

***denotes that impurity was present but the buffer was only analyzed once; therefore,

no RSD value was calculated for the particular concentration/pH combination

67

For tris buffers of pH 7, phenethylamine and ephedrine were not extracted at any
buffer cOncentration, and replicates of 1 M and 0.5 M tris buffers at pH 7 were not
completed. Therefore, RSD values for these two buffers were not calculated. The pH 7
tris buffers showed no carryover in the microwave vessel from the simulated sample
between extractions. However, because not all impurities were extracted, tris buffers of
pH 7 were not chosen as the optimum buffer for MAE.

At all concentrations of tris buffer at pH 8, phenethylamine was not extracted, and
ephedrine was extracted but co-eluted with a siloxane peak from the fiber. At all
concentrations there was no sample carryover in the microwave vessel between
extractions. However, because not all impurities were extracted, tris buffers of pH 8
were not chosen as the optimum buffer for MAE.

For tris buffers of pH 9 all impurities were extracted; however, ephedrine co-
eluted with a siloxane peak making peak area integration difficult. Therefore, no RSD
value was calculated for ephedrine. However, methamphetamine carryover in the
microwave vessel was observed at all concentrations and benzylamine carryover was
observed in the 1 M buffer. Therefore, tris buffers of pH 9 were not chosen as the
optimum buffer for MAE.
4.2.1.3. Investigation of Carbonate Buffer

A 0.05 M carbonate buffer at pH 10 was chosen based on a review of the
literature [8,9] for preliminary work at the start of this project. The average peak areas
(n=3) and RSD values were determined for the components of the simulated sample in

the carbonate buffer and are summarized in Table 4.4.

68

Table 4.4: Average abundance and RSD values of simulated tablet components in carbonate buffer
(average peak area based on three replicates)

 

 

 

 

 

0.05 M
Bufl’e r pH Impurity pKa Ave rage % R SD
Pe ak Are a
Benzylamine 9.3 4.05E+08 5.89
Phenethylamine 9.8 2.64E+08 8.72
10 Methamphetamine 10.0 1 .92E+09 l .71
Ephedrine 10. 0 1 .60E+08 8.26
Caffeine 14.0 1.55E+08 7.74

 

All impurities were extracted in the carbonate buffer with good precision.
However, because methamphetamine carryover was observed in the microwave vessels
between extractions, carbonate buffer was not chosen as the optimum buffer for MAE.

Thus, the optimum buffer selected for all subsequent extractions was a 1 M
phosphate buffer at pH 8. The simulated sample components were extracted with good
precision from this buffer, and no sample carryover was seen in the microwave vessel
between extractions.

4.2.2. Determination of Significant Parameters

Using the simulated sample, a filll factorial screening design was performed to
determine if the microwave parameters of ramp time, extraction time, and extraction
temperature were significant in the extraction of organic impurities. The full factorial
design was chosen because the parameters and the interactions between the parameters
could be studied and the significance of each determined.

The high values for ramp time and extraction time (20 minutes) were chosen for
time efficiency as longer extractions would become impractical. The low value for ramp
time (10 minutes) w as chosen to allow the microwave sufficient time to reach the
extraction temperature. The low value for the extraction time (10 minutes) was selected

to allow time for the sample to dissolve into solution. The high extraction temperature

69

(120 °C) was selected to avoid possible thermal decomposition of the sample while the
low extraction temperature (80 °C) was chosen to ensure that the sample completely
dissolved into solution.

After the set of experiments was completed (see Appendix C), the peak areas of
the simulated sample components were integrated. Using Statgraphics software,
multivariate analysis of variance (MANOVA) was performed to determine which of the
microwave parameters had a significant effect on the extraction of the simulated sample
components. Ephedrine was not included in the analysis because it co-eluted with a
siloxane peak from another new fiber that made accurate peak area determination
difficult.

Because of the blocked experimental design, the interactions between the
parameters were confounded with the block effects, or the day to day differences. This
means that the effects of the interactions and the effects of the blocks could not be
differentiated from one another. Because this design was an initial screening used only to
determine which of the parameters had an effect on the extraction, the confounding with
the block effects was not problematic.

Using MANOVA, the sum of squares for each of the main effects, the interactions
plus blocks, and blocks alone were calculated, and the degrees of freedom were
determined. The mean square value of each parameter and interaction was calculated
followed by the determination of the F-ratio for each parameter and interaction. The
significance of this value was determined by comparing the calculated F-value to a
critical F-value at the 95% confidence level. Complete MANOVA tables for each of the

simulated sample components can be found in Appendix E.

70

For benzylamine, the interaction between extraction time and extraction
temperature had a significant effect on the peak area. Because the interaction between
extraction time and extraction temperature was significant, both parameters were
important and investigated further in the subsequent optimization design. For
methamphetamine, three parameters and interactions significantly affected the peak area:
extraction temperature, the interaction between ramp time and extraction time, and the
interaction between extraction time and extraction temperature. For caffeine and
phenethylamine, no parameter or interaction of parameters had a significant effect on the
peak areas. Because all three parameters (ramp time, extraction time, and extraction
temperature) had a significant effect on the peak area of one or more of the components,
all three parameters were included in the optimization design.

4.2.3. Optimization of Significant Parameters

To optimize the microwave ramp time, extraction time, and extraction
temperature, the parameters were studied in a circumscribed central composite design
using the simulated sample. The circumscribed central composite design was chosen
because it allowed for the second order interactions to be determined and more complete
mathematical models to be built [10]. After the set of experiments was completed (see
Appendix D), the peak areas of the simulated sample components were integrated.
Again, ephedrine was not included in the data analysis because of co-elution with a
siloxane peak from the fiber.

Using Statgraphics software, the first step in determining the optimum parameters

was to model the data collected during the experiments. Linear regression analysis was

71

used to fit a second-order mathematical model for each component’s peak area. This
resulted in four models, one for each sample component.

Next, the peak area of each impurity was optimized individually based on whether
the goal was to maximize or minimize the area. The peak areas of methamphetamine and
caffeine were minimized. In MDMA tablets, these compounds are adulterants, not
impurities. Therefore, these peaks should be minimized to avoid over-extraction and
broad peaks that can mask impurities present at low concentrations. Meanwhile,
benzylamine and phenethylamine were maximized because they represent impurities in
the MDMA tablets, and the goal is to maximize the extraction of impurities.

A desirability function was created for each simulated sample component
individually. Then, the desirability of each individual component was combined to
determine the optimum settings for the MAE. This allowed for the determination of the
extraction parameters that allowed the methamphetamine and caffeine peaks to be
minimized and the benzylamine and phenethylamine peaks to be maximized. The
variable, 5, or the shape of the desirability function, was set to one (linear) for all
responses (or peak areas).

The impact, or importance, was set for each component on a scale of 1-5. The
impact of methamphetamine was set to 5 because it was important to minimize the
carryover of methamphetamine between samples. Also, it was important to minimize the
possibility of a broad methamphetamine peak masking impurities present at lower
concentrations. Minimizing caffeine was not of great importance because there was no
carryover of caffeine between experiments, and the peak is not so broad as to mask other

impurities present at similar retention times. Therefore, the impact was set to 2.

72

Benzylamine and phenethylamine maximization was important in maximizing the levels
of impurities present in the sample, and the impact was set to 5. The optimum MAE

parameters determined from the desirability function are shown in Table 4.5.

Table 4.5: Optimum microwave parameters from the CCC design

 

 

Parameter Optimum
Ramp Time (min) 23
Extraction Time (min) 23

Extraction Temperature (°C) 100

 

To visualize the estimated peak area at various settings, including the optimum
settings, of the parameters, estimated response surface graphs were drawn for the
components. To determine the estimated response, the values of the parameters were
entered into the mathematical model constructed for the particular component. The plot
was then constructed from the responses of several different sets of conditions. A
separate plot was created for each component.

For example, Figure 4.2 shows the estimated response surface for
methamphetamine. For this plot, the extraction temperature was held at 100 °C in the
equation for methamphetamine, and the values for the parameters of ramp time and
extraction time were varied. Although the peak areas are shown as discrete lines, the

peak area is a continuum. The peak area at the optimum setting is marked with a plus

sign (+).

73

 

 

 

 

 

 

25 L“ Methamphetamine
A r Estimated Peak Area
v F
0 20 f “'— 2.28E8
'5 _——— _______.____ ---2.36E8
5 .:—--——--—--....._~ 3 --2.44E8
-.: -__ ““" “- .\ \ / - - 2.52E8
8 IS “ ~ ~ )
5 _ e \ / - - 2.60E8
>< - ‘ a .
11.1 _ t I / / / / - - - 2.68138
. 1 I - . .
10 _1 I. i l i / I /. ./i a 1 . . . . 1 2.76E8
IO 15 20 25

Ramp Time (min)

Figure 4.2: Estimated response surface for methamphetamine from CCC design with a plus sign (+)
indicating the response at the optimum settings for the MAE parameters

At the optimum settings of the parameters (23 minute ramp time and 23 minute
extraction time), the predicted peak area for methamphetamine is between 2.20E8 to
2.28E8 which are the lowest values calculated for the estimated peak area. This result is
expected since the desired outcome of methamphetamine is minimization. From this
plot, it can be seen that a 10 minute ramp time and 20 minute extraction time gives
similar results to a 23 minute ramp time and 23 minute extraction time. However, these
values for the parameters were not chosen because the peak areas of the other
components are at their maximum or minimum at a ramp time and extraction time closer
to 23 minutes. The Optimum parameters represent a compromise of parameter settings
based on all the responses.

4.3. Optimization of Headspace Solid-Phase Microextraction Procedure

Using 1 M phosphate buffer at pH 8 and the homogenized batch of exhibit
MSU900-01, the optimum HS-SPME extraction time and extraction temperature were
investigated. The samples for this study were not microwave extracted prior to HS-

SPME to determine the effect of HS-SPME on the extraction of the MDMA exhibit. For

74

the HS-SPME optimization, extraction times of 10-60 minutes in 10 minute increments
and extraction temperatures of 40-80 °C in increments of 10 °C were studied. The
extraction time range was chosen for time efficiency while the extraction temperature
range was chosen for practical limitations of the water bath. The number of impurities
extracted and the peak shape were evaluated to determine the optimum HS-SPME
extraction time and extraction temperature.

4.3.1. Optimization of Extraction Time

When the extraction temperature was held at 70 °C, the longer extraction times
(40, 50, and 60 minutes) extracted five more impurities from the MDMA exhibit than the
shorter extraction times (10, 20, and 30 minutes). In general, impurities were extracted at
higher abundances at longer extraction times because of the pre-concentration of the
impurities on the fiber. For impurities with a higher affinity for the fiber than the
headspace, the longer extraction time allowed more time for the impurities to absorb or
adsorb onto the fiber. However, if the impurities have a lower affinity for the fiber, it is
possible that the longer extraction time would give more time for the impurities to desorb
from the fiber. This did not appear to be the case for the impurities extracted from the
exhibit because the longer extraction times had more abundant peaks across the retention
time range of the chromatogram.

Even though more impurities and components were extracted in higher
abundances at longer extraction times, the high abundance is not always desirable.
Figure 4.3 shows a comparison of chromatograms for the 60, 40, and 10 minute
extractions, all with an extraction temperature of 70 °C. At longer extraction times (60

and 40 minutes, Figure 4.53 and b, respectively), the higher abundances of MDMA and

75

caffeine led to poor peak shape and broadened peaks with 60 minutes showing the worst
peak shape. The broad peaks are caused by the high concentration of the sample
overloading the gas chromatograph (GC) column resulting in band broadening.
Therefore, the longest extraction times, 50 and 60 minutes, were not investigated further
as there is a greater chance that trace level impurities present in the retention time range

of MDMA and caffeine would. be masked by the broad peaks.

76

 

a) 1.50E8

*
MDMA

y)?»
R
W
* MDPZ-Propanol

Diethyl Phthalate
Caffeine

Methamphetamine

   

Abundance

 

 

 

 

r
F

 

 

 

 

 

 

 

 

 

2 7 12 I7 22 27
Retention Time (min)
b) 1.50E8 * <
i .E 2 a
s . E 52‘
3 .e i n- z
E °i E 2 flag
2 I 3 2+: 2
3 /' \ 2 a .C.‘ 5
f3 / \ E is
__/ \T\ %*U
, 12.2 12.3 * g fll
000130 . . NEP_LIJ WW. II I 1 , L

 

 

 

2 7 12 17 22 27

Retention Time (min)

 

   
   

 

 

 

 

 

 

C) 1.50E8 *
.E
E < E
s ‘5 E t ..
s E 2 :3
t: E D. a. 5
.8 g g&
< :11! —>'~
§\ s
u '5
000130 1 L xi Ill ALL Ii_._ i I l l 1

Retention Time (m in)

Figure 4.3: Chromatograms of HS-SPME extractions at a) 60 minutes, b) 40 minutes, and
c) 10 minutes; an asterisk (‘) indicates that a peak was present in the blank

77

At the shorter extraction times (10 minutes, Figure 4.3c), many impurities were
extracted at lower abundances. Piperonal, which co-eluted with an unidentified
compound, was not detected above the baseline noise at the 10 minute extraction time.
The detection of piperonal is important for determining the synthetic route of 3,4-
methylenedioxyphenyl-2-propanone (MDP2P) which is a common starting material in
the synthesis of MDMA. Forty minutes (Figure 4.3b) was chosen as the optimum
extraction time because it offered a compromise between higher impurity abundance and
improved peak shape.

4.3.2. Optimization of Extraction Temperature

When the extraction time was held at 40 minutes, the higher extraction
temperatures (70 and 80 °C) extracted impurities at higher abundances than the lower
extraction temperatures (40, 50, and 60 °C). At higher extraction temperatures, more of a
compound volatilizes than at lower extraction temperature because the compounds have a
higher partition coefficient between the headspace and the sample. Thus, more of the
sample is in the headspace to absorb onto the fiber [7,11]. By lowering the temperature,
less of the compound volatilizes; and therefore, less of the sample is available in the
headspace to absorb or adsorb onto the fiber resulting in less abundant peaks in the
chromatogram.

However, at higher extraction temperatures, higher volatility compounds may
actually desorb from the fiber during the extraction time due to the lower partition
coefficient [7,11] For example, Figure 4.4 shows a comparison of extractions at 80 °C

and at 40 °C. An unidentified peak with a retention time of 2.17 minutes (labeled “a”)

78

was extracted at a higher abundance with an extraction temperature of 40 °C than with an

extraction temperature of 80 °C.

 

 

 

 

  

 
 
 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a) 1.50E8 . * __
1; o < O
' E .5 2 5
E s 9 8T
8. ‘5 “r
s '55 €- 3:
s l _ E /9 s 3::
'U /\ \. H D— / '—' D
5 r" I g 8'. it a"? E
.2 r \K D\ i U
, \ 2 3.
122 12.3 /§ :1:
O
f_,__ L r 42
2 7 12 17 22 27
Retention Time (min)
b) 1.5038 i *
.2
8 g a
: w <
co 0 5
'2 *fi- 1:. E ‘é'
3 5 E 2 n.-
< s D N s
2 5\ E g
E m
a t * U
l l l l *1
0.0050314 _ 1111 l. i -I L IA I l 1 A; l

 

 

 

 

 

2 7 12 17 22 27
Retention Time (min)

Figure 4.4: Chromatograms of HS-SPME extractions at a) 80 °C and b) 40 °C;
an asterisk (*) indicates that the peak was present in the blank

Even though impurities were extracted at higher abundances at 80 °C, the broad
MDMA peak may have masked additional impurities present at trace levels that elute at a
similar retention time to MDMA. Therefore, 80 °C was not chosen as the optimum
extraction temperature. At lower extraction temperatures, for example 40 °C, the broad

peaks were not observed; however, some impurities such as piperonal and isosafrole were

79

not extracted above the baseline noise due to their lower volatility. The detection of these
impurities is important because their presence indicates the possible synthetic route used
to synthesize the MDP2P [12]. Ultimately, an extraction temperature of 70 °C was
chosen as the optimum temperature for the extraction of organic impurities because
impurities were extracted at higher abundance than the lower temperatures while
maintaining better chromatography than the 80 °C extraction. Therefore, the optimum
HS-SPME conditions were a 40 minute extraction at 70 °C.

Thus, based on the results of the optimization studies, the optimum MAE/HS-
SPME procedure involves a microwave extraction with a 23 minute ramp time to 100 °C
and a 23 minute extraction at 100 °C. The subsequent HS-SPME parameters include a
five minute pre-heat at 70 °C and a 40 minute extraction at 70 °C.

4.4. Comparison of MAE/HS-SPME, HS-SPME, and LLE
4.4.1. Simulated Sample

To determine the precision of each of the three extraction procedures (MAE/HS-
SPME, HS-SPME, and liquid-liquid extraction, LLE), replicate extractions of the
simulated sample were performed by each of the procedures (four replicates for
MAE/HS-SPME, five replicates for HS-SPME, and three replicates for LLE). For each
set of replicates, the peak areas of each simulated sample component were integrated and

RSD values were calculated (Table 4.6).

80

Table 4.6: Relative standard deviations for simulated sample components extracted by the three
procedures

 

 

 

o/oRSD
Component MAE/HS-SPME HS-SPME LLE
(n=4) (n=5) (n=3)
benzylamine 4.61 5.74 **
phenethylamine 28.32 3.46 **
methamphetamine 9.81 5.50 9.37
ephedrine * 4.88 **
caffeine 8.73 6.35 7.61

 

* Co-elution with siloxane from fiber; ** Component not detected

For SPME, acceptable RSD values range from 1-10%. The MAE/HS-SPME
procedure showed good precision with RSD values less than 10% for benzylamine,
methamphetamine, and caffeine. Ephedrine co-eluted with a siloxane peak, therefore
making accurate peak area integration difficult. The co—elution may have been due to a
new fiber being used. Phenethylamine showed poor precision resulting in a RSD value
greater than 25%. Phenethylamine may not be thermally stable at the higher
temperatures reached by the microwave (100 °C), and therefore is not extracted precisely.

The HS-SPME procedure showed a precise extraction of the simulated sample
with all components being extracted with RSD values less than 10%. The LLE procedure
only extracted two of the simulated sample components and did so with RSD values less
than 10%. Despite similar precision in extraction for the three procedures, more of the
simulated sample components were extracted using MAE/HS-SPME and HS-SPME than
with LLE.
4.4.2. MDMA Exhibits

Homogenized batches of the three MDMA exhibits were extracted in triplicate by
each of the three extraction procedures (MAE/HS-SPME, PIS-SPME, and LLE). For

each exhibit, the ability of the three extraction procedures to extract impurities was

81

assessed based on the number and identity of impurities. To compare the precision of the
extraction procedures, the triplicate chromatograms were retention time aligned and
Pearson product moment correlation (PPMC) coefficients were calculated between
replicate extractions for each exhibit.

A fit” list of all compounds extracted from the three MDMA exhibits is given in
Appendix F. The compounds extracted from the exhibits have been divided into two
groups: impurities and other components. Impurities are chemical compounds which
originate from the reactions involved in synthesizing the MDMA. Other components are
additives, adulterants, and main active ingredients. Components include caffeine, fatty
acids, methamphetamine, MDMA, lidocaine, and phthalates (not present in the blank).

The GC-MS parameters for the LLE analysis are based on the literature [4]. This
set of parameters differs from the parameters optimized for MAE/HS-SPME and HS-
SPME analysis in two ways that affect the retention time of impurities and components:
the carrier gas flow rate and the initial starting temperature of the GC oven. The flow
rate for LLE analysis is 0.5 mL/min while for MAE/HS-SPME and HS-SPME it is 1
mL/min. The initial GC oven temperature for LLE is 90 °C while for MAE/HS-SPME it
is 60 °C. Therefore, a difference in retention times of approximately four minutes is
expected when comparing LLE to MAE/HS-SPME or HS-SPME. Also, retention time
differences of about 0.1 minutes are observed when comparing MAE/HS-SPME and HS-
SPME chromatograms due to slight changes in the instrument as a result of routine

maintenance, for example cutting the end of the column.

82

4.4.2.1. MDMA Exhibit MSU900-01

In total, 62 different compounds were extracted from exhibit MSU900-01 by the
three extraction procedures and are summarized in Table 4.7.

Table 4.7: Number of impurities and components extracted from M DMA exhibit MSU900—01

 

 

Extraction Number of Impurities Number of Other Components
MAE/HS-SPME 42 6
HS-SPME 46 5
LLE 8 6

 

Figures 4.5, 4.6, and 4.7 show the chromatograms of exhibit MSU900-Ol
extracted by MAE/HS-SPME, HS-SPME, and LLE, respectively. The chromatograms
have been truncated to show only the region from 2-30 minutes as this is the region of
interest. Peaks in the chromatograms are labeled with the identity of the component or

impurity. Peaks labeled with an asterisk (*) are present in the blank, some of which are

siloxanes from the SPME fiber.

83

 

 

PIOV stunned

41AAL.‘114LA

vwow pammsqns [£11121
=mw=>0pn~\>

911193593
4

 

 

*-

VEICIIAI*\ _

mow 9112mm 14119106
IOWdOld'ZdClI/‘l

 

 

* dZdGW

oo esos ——-—91
113 1/1

 

 

 

 

 

 

 

 

 

1.00E8

ounumoqdweqiaw t ,
/ *
\o_ ouonlofixotpaualfiqtaw/
F‘ .1
/ V?
/ ~ 4
9101593

o
. LL]
o
aouepunqv q
o

w]
J
J

27

22

17

12

Figure 4.5: Chromatogram of MDMA exhibit MSU900-01 extracted by MAE/HS-SPME;

an asterisk (*) indicates that the peak was present in the blank

84

Retention Time (min)

 

 

 

 

vwcm pammsqnswma -
ammpnNfi
auiageg) * :'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.00E8

*-
* I
* VEI(IV\I\ _
VWGW otemqtqdlfiqmrbg
* 2
[ouedoId-zclqw / dZdCIIAI
alogesosl—et
Z 4
[womdtd/ J
ouiwmaqdmeqtow 4"]
/ *
j V? auanlotfixotpaualfimow/ _
M .2
9101.198 J
:3
Lu
O
oouepunqv Q
c

Figure 4.6: Chromatogram of MDMA exhibit MSU900-01 extracted by HS-SPME;
an asterisk (*) indicates that the peak was present in the blank

85

27

22

17

12

Retention Time (min)

 

 

 

PW owns J

ppv named .
Rah-3

euro 9
. J} C) H /

/
vwaw pammsqns Mung

 

QWEXO dZdCU/Vfl

GIBIWIWJIIQDQECI—ar
vwow j

IOWdOJcI-ZJCIW
chICIW /

 

 

 

>—

aunnmothueqtow _a

t

 

 

1.00E8

0.00E0

oouepuan

Figure 4.7: Chromatogram of MDMA exhibit MSU900-01 extracted by LLE

86

27

22

17

12

Retention Time (min)

More impurities were extracted from exhibit MSU900-Ol using MAE/HS-SPME
and HS—SPME than LLE. Headspace solid-phase microextraction pre-concentrates the
impurities on the SPME fiber allowing the impurities to be extracted and detected above
the background noise. However, there is no pre-concentration in LLE [4] and
compounds present in the sample at trace levels may not be detected above the
background noise in the chromatogram.

The impurity 3,4-methylenedioxytoluene was extracted using MAE/HS-SPME
and HS-SPME, but not using LLE. This impurity is a by-product formed during the
synthesis of MDP2P from safrole. This lack of detection was a limitation of the GC
temperature program used for LLE analysis. The initial temperature of the GC program
for LLE analysis was 90 °C with a four minute solvent delay to prevent saturation of the
detector. After four minutes, the mass spectrometer turned on at which time the oven
temperature was approximately 114 °C. Therefore, compounds with boiling points lower
than 114 °C were not observed in the chromatogram. In contrast, for MAE/l-IS-SPME
and HS-SPME, the initial oven temperature was 60 °C and the mass spectrometer delay
was only two minutes, which allowed the sample to desorb from the fiber. Therefore,
more impurities were observed at the beginning of the MAE/HS-SPME and HS—SPME
chromatograms.

The impurities safrole, piperonal, and isosafrole were also only detected by
MAE/I-lS-SPME and HS-SPME; however, this is due to the pre-concentration of the
impurities on the SPME fiber. The presence of these impurities is indicative of the

synthetic route used to synthesize MDP2P. Safi'ole is a common starting material for

87

MDP2P and is often extracted from sassafras oil, a naturally occurring substance [12].
Piperonal and isosafrole are intermediates in the synthesis of MDP2P [l 3].

The impurities MDP2P and 3,4-methylenedioxyphenyl-2-propanol (MDP2-
propanol) were extracted by all three procedures but at higher concentrations by
MAE/HS-SPME and PIS-SPME. The impurity MDP2-propanol is a product of a side
reaction in the synthesis of MDMA from MDP2P using reductive amination [8]. During
the reductive amination, MDP2P is converted to an imine intermediate by amination
reaction with methylamine. The intermediate is then reduced to MDMA during which a
side reaction occurs converting the MDP2P to the alcohol form (MDP2-propanol).

The impurity 3,4-methylenedioxyethylamphetamine (MDEA), which is
chemically similar to MDMA, was extracted by MAE/HS-SPME and HS-SPME at
similar levels but not extracted by LLE due to its low concentration in the exhibit. While

the origin of this impurity has not been tested or confirmed, it is possible that it originates

from the amination of MDP2P by ethylamine (CH3CH2NH2) which is an impurity

present in methylamine (CH3NH2) [12].

Some tablet impurities, such as l-(3,4-methylenedioxyphenyl)-2-propanone
oxime (MDP2P oxime), were extracted by LLE but were not extracted by MAE/HS-
SPME and HS-SPME. This oxime impurity originates from the synthesis of MDP2P
from safrole through the B-nitroisosafrole route (see synthesis route 2 in Appendix A)
[12]. Liquid-liquid extraction has the advantage of extracting components of low
volatility that are not extracted by MAE/HS-SPME or HS—SPME thus giving
complementary information to the SPME profiles. Therefore, more information about

the sample can be gained if both HS-SPME (or MAE/HS-SPME) and LLE are used.

88

The impurity N-methyl-(l,2-methylenedioxy)-4-(l-ethyl-2-aminopropyl) benzene
(ethyl substituted MDMA) was extracted by all three procedures. While the exact origin
of this impurity is not known, it is similar in structure to N-ethyl,N-methyl(l,2-
methylenedioxy)-4-(2-aminopropyl)benzene which is a by-product of the reductive
amination of MDP2P by ethylamine which is an impurity in methylamine [12].
Therefore, the impurity detected in the chromatogram may be an indicator of the
reductive amination route of MDMA synthesis.

A similar number of tablet components were extracted from exhibit MSU900-Ol
by all three extraction procedures. Although these components are not organic impurities
which are the focus of this work, their extraction and detection in the tablet is important
to provide additional information on the tablet production and manufacturing process.

Methamphetamine, MDMA, diethyl phthalate, and caffeine were extracted by all
three procedures. Methamphetamine and caffeine are adulterants which are added to the
synthesized MDMA before it is pressed into tablets to enhance the effects of the MDMA
[1]. Diethyl phthalate is a plasticizer which is used as a binder in the tableting process
[9]. Methamphetamine and diethyl phthalate were extracted in higher abundances by
MAE/HS-SPME and l-IS-SPME than by LLE because of the pre-concentration of the
compounds on the SPME fiber.

Caffeine was extracted at higher concentrations by LLE than by MAE/HS-SPME
and HS-SPME. In the LLB chromatogram, the broad caffeine peak had a baseline width
of 0.8 minutes whereas in the MAE/HS-SPME chromatogram, the caffeine peak had a
baseline width of 0.2 minutes. In LLE, other extracted compounds that have a similar

retention time to caffeine would likely be masked by the broad caffeine peak. In contrast,

89

 

using MAE/HS-SPME and HS-SPME, these compounds are less likely to be masked
because the caffeine peak is narrower. The higher concentration of caffeine in LLE is
due to its volatility. With a sublimation point of 180 °C [14], caffeine has a low volatility
that limits its extraction by MAE/HS-SPME and HS-SPME, but does not affect its
extraction by LLE. Even though caffeine has a high sublimation point, caffeine is still
extracted by HS-SPME due to the high concentration in the sample. According to

Equation 4.1 (which was previously discussed in Chapter 2), as the concentration of the
compound in the initial solution, Co, increases, a higher mass of the compound, nf, will be
extracted by the fiber.
n _ 19, Vii/SC.
f Kfsvf + 10.th + V. (4")

 

Lidocaine, a local anesthetic that can be added to the. MDMA before it is pressed
into tablets, was detected in the MAE/HS-SPME and HS-SPME chromatograms. The
MAE/HS-SPME and HS-SPME allow for components and impurities present at trace
levels to be pre-concentrated on the fiber, thus allowing the compounds to be extracted
and detected above the baseline. However, in LLE, trace level impurities may be
difficult to detect above the background noise since there is no pre-concentration.

Fatty acids such as palmitic acid and stearic acid are used as lubricants in the
tableting process [9]. Palmitic acid was extracted by both MAE/HS-SPME and LLE,
with LLE extracting the component at higher concentrations because of the low volatility
of the component. However, the fatty acid was not extracted using SPME alone.
Theoretically, a highly efficient extraction is achieved by MAE allowing for all
components to be extracted into solution. In contrast, during HS-SPME alone, the entire

sample did not completely dissolve into solution. Analytes transfer more readily fi‘om

9O

solution into the headspace than from a solid into the headspace. Therefore, following
MAE, the entire sample was in solution and more of the sample transferred into the
headspace for extraction.

Meanwhile, stearic acid was only extracted by LLE. Because of the higher
boiling point of the stearic acid compared to palmitic acid, the HS-SPME equilibrium
would favor the solution and little of the component would be present in the headspace to
be extracted by the fiber. The mass extracted by the fiber, if any, was too small to be
detected above the background. Therefore stearic acid was not observed in the MAE/HS-
SPME or HS-SPME chromatograms. However, because LLE is not dependent on the
volatility of the compound, stearic acid was extracted by LLE.

To determine the precision of the extraction procedures, the chromatograms were
retention time aligned, and, using the entire chromatogram, the average PPMC
coefficients of each set of triplicates and their standard deviations were calculated and are

shown in Table 4.8.

Table 4.8: Average PPM C coefficients and standard deviations of MDMA exhibit MSU900-01
associated with each ex traction procedure

 

 

Extraction Average PPMC Standard Deviation
(n=3)
MAE/HS-SPME 0.9501 0.0264
HS-SPME 0.9271 0.0399
LLE 0.9330 0.0190

 

A PPMC coefficient between 0.8-1.0 indicates a strong positive correlation [15].
Ideally, PPMC coefficients of replicates should be at least 0.99, indicating close to
perfect correlation between the replicates. However, variability in the sample, sample
preparation, and extraction procedure among replicate extractions can cause lower PPMC

coefficients. For example, the sample may not have been completely homogenized prior

91

to analysis. Also, slight fluctuations in the temperature of the water bath used for HS-
SPME (plus or minus 2-3 °C) could have caused the variation between replicate
extractions. Retention time misalignments also could have contributed to the lower
PPMC coefficients. For example, in the chromatogram of this exhibit, several small
peaks were present at the beginning of the chromatogram (first 10 minutes) that were
only slightly higher than the baseline noise. These peaks were not well aligned by the
alignment algorithm which may have contributed to the lower PPMC coefficients.
Despite these factors, the PPMC coefficients of the replicates were greater than 0.92
indicating a strong correlation among replicates and a precise extraction.

The PPMC coefficient of the MAE/HS-SPME replicates is higher than HS-
SPME. Because the sample was entirely in solution following MAE, the MAE/HS-
SPME extraction was more precise than with HS-SPME alone. The standard deviation
shows the range of differences in the PPMC coefficients among the replicates. The
standard deviation of the HS-SPME procedure was higher than the standard deviation for
the MAE/HS-SPME replicates showing that the replicates of the MAE/HS-SPME
procedure were more similar than the replicates of the HS-SPME procedure. The LLE
procedure showed a similar PPMC coefficient to HS-SPME; however, for LLE, the
standard deviation was approximately half that for HS-SPME. Again, this is due to
misalignments in the HS-SPME chromatograms, mainly in the early eluting peaks.
4.4.2.2. MDMA Exhibit T-17

In total, 66 different impurities and components were extracted from exhibit T-l 7

by the different procedures and are summarized in Table 4.9.

92

Table 4.9: Number of impurities and components extracted from MDMA exhibit T-17

 

 

Extraction Number of Murities Number of Other Components
MAE/HS-SPME 40 5
HS-SPME 35 5
LLE 23 7

 

Figures 4.8, 4.9, and 4.10 show the chromatograms of exhibit T-l7 extracted by
MAE/HS-SPME, HS-SPME, and LLE, respectively. The chromatograms have been
truncated to show only the region from 2-30 minutes as this is the region of interest.
Peaks in the chromatograms are labeled with the identity of the component or impurity.

Peaks labeled with an asterisk (*) are present in the blank, some of which are siloxanes

 

from the SPME fiber.

93

 

vwcrw pemmsqns MIMI
amevom

N
outages inf—é

 

VEICIW 911211311qu -x-

 

y

 

 

 

VIAIGW j

———\.5

/ * fl
[ouedOJJ-ZJCUAI (IZJCIW
empeqda
*

1___%

 

 

OIQdI / .
[BU 'd oroges/a

 

ounnmoqdmeqiow

ouonloMxotpeuq/(qiow

I... J

 

 

 

 

 

 

1.00E8

oouepunqv

0.00E0

Figure 4.8: Chromatogram of MDMA exhibit T-l7 extracted by MAE/HS-SPME;
an asterisk (‘) indicates that the peak was present in the blank

94

27

22

17

12

Retention Time (min)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.00E8

d;
r
.4
1
i.
vwow pelnltlsqnsvima —
ameoopn\> ,
gurgling) * g
I * g
VEICIW 91919‘I1‘1d\ * _» _
\\
VIAIGW fl:
/ * fl
IOWdO-Id'ZcICIW chICIW g
ouppang _.
~x- //)
019d! 4
[nu "1 91015934
Q
outwmoqdmeqnw
ouonmmxotpouopfiqiow \
N
c.2-
J
:3
LL}
o
aouepunqv o_
c

Figure 4.9: Chromatogram of MDMA exhibit T-l7 extracted by HS-SPME;
an asterisk (*) indicates that the peak was present in the blank

95

27

22

17

12

Retention Time (min)

 

 

Llc

_‘—

P19V011991S 4)

prov 0311111ch
91119001311

 

euioyeg)

VEICIW

 

 

 

9‘“le dZdCIW\

9112mm \

 

 

 

\_
VWCIW J3,

 

 

Iomdmd-zaow

dedow/

'4
ampaqda/ ,

1

_

ounumoqdmeqtow ___é.

 

 

1.00E8

oouepunqv

0.00E0

Figure 4.10: Chromatogram of MDMA exhibit T-l7 extracted by LLE

96

27

22

17

12

Retention Time (min)

More impurities were extracted using the MAE/HS-SPME and HS-SPME
procedures than using the LLE procedure due to pre-concentration of the impurities on
the SPME fiber. There were five more impurities extracted by MAE/HS-SPME than by
HS-SPME alone due to the theoretically higher extraction efficiency of the microwave
which extracts the entire sample into solution. In HS-SPME alone, some of the sample
remains in the solid state. Compounds in solution move more easily into the headspace
than compounds that are in the solid form. Therefore, with the entire sample in solution
more impurities were extracted by MAE/HS-SPME.

The impurities 3,4-methylenedioxytoluene, safrole, and piperonal were extracted
by both MAE/HS-SPME and HS-SPME. Ephedrine, which is a starting material in the
synthesis of methamphetamine [2], was extracted by all three procedures. Ephedrine can
also be an adulterant added to the MDMA after the MDMA was synthesized. However
due to the low abundance of ephedrine in the LLE chromatogram, it is more likely that
the ephedrine was present in the tablets at low levels suggesting it was an impurity from
the synthesis of methamphetamine.

The impurities MDP2P and MDP2-propanol were extracted by all three
procedures. The MDP2P was extracted at trace levels in the LLE chromatogram due to
its low concentration in the sample and the lack of pre-concentration in LLE. 1n the
chromatograms of all three procedures, MDEA was the most abundant peak. The
impurity N-methyl-(l,2-methylenedioxy)-4-(l-ethyl-2-aminopropyl) benzene (ethyl
substituted MDMA) was extracted by MAE/HS-SPME and HS-SPME but was not
extracted by LLE. Once again, MDP2P oxime was only extracted by LLE due to its

volatility.

97

More tablet components were extracted by LLE than by the other two procedures
because LLE is not dependent on the volatility of the compound. For example, the fatty
acids palmitic acid and stearic acid were only extracted by LLE. However, palmitic acid
was extracted by MAE/HS-SPME from exhibit MSU900-01. Palmitic acid and stearic
acid were present in lower concentrations in exhibit T-I7 compared to exhibit MSU900-
01 as is evident from the less abundant peaks. When a lower concentration of a
compound is present in the starting solution, there is less available for extraction by the
fiber based on Equation 4.1. Therefore, the fatty acids were not detected above the
baseline in MAE/HS-SPME or HS-SPME for exhibit T-l7 while palmitic acid was
detected from exhibit MSU900-01 by MAE/HS-SPME.

Methamphetamine, MDMA, caffeine, and lidocaine were extracted by all three
procedures. A phthalate peak was detected by all three procedures but at higher
concentrations in the MAE/HS-SPME and HS-SPME chromatograms due to its pre-
concentration on the SPME fiber.

The average PPMC coefficients and the standard deviations of the triplicate

chromatograms for each extraction procedure are shown in Table 4.10.

Table 4.10: Average PPMC coefficients and standard deviations of MDMA exhibit T-l7 associated
with each extraction procedure

 

 

Extraction Aver:§:;)PMC Standard Deviation
MAE/HS-SPME 0.9943 0.0016
HS-SPME 0.9975 0.0008
LLE 0.9812 0.0050

 

The three extraction procedures have similar PPMC coefficients of 0.98 or higher
indicating a very strong correlation between the replicates. There were fewer

misalignments for exhibit T-17 than for exhibit MSU900-01 resulting in the higher

98

PPMC coefficients for exhibit T-17. The low standard deviations show that the replicate

extractions were very similar indicating good precision.

4.4.2.3. NIDMA Exhibit T-27

In total, 75 different impurities and components were extracted from exhibit T-27

and are summarized in Table 4.11.

Table 4.11: Number of impurities and components extracted from MDMA exhibit T-27

 

 

Extraction Number of Impurities Number of Other Components
MAE/HS-SPME 50 6
HS-SPME 46 5
LLE I4 6

 

Figures 4.1], 4.12, and 4.13 show the chromatograms of exhibit T-27 extracted by
MAE/HS-SPME, HS-SPME, and LLE, respectively. The chromatograms have been
truncated to show only the region from 2-30 minutes as this is the region of interest.
Peaks in the chromatograms are labeled with the identity of the component or impurity.
Peaks labeled with an asterisk (*) are present in the blank, some of which are siloxanes

from the SPME fiber.

99

 

sptoV [(1195 poiemiesu {1 5

\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

vwow pemmsqns 14913
9111935121) ‘3
' 3
vaaw \
919mm Wilam\*j
quw j
- 3
IOWdOJd zaow dZdCIW
QIOIJBSOSI
[anomdtd \
9101;93 ———>1_
outwmaqdweqiow

 

 

3|:
a.“ LML MALL LL

 

 

1.00E8

aouepunqv

0.00E0 7r

Figure 4.11: Chromatogram of MDMA exhibit T-27 extracted by MAE/HS-SPM E;

an asterisk (*) indicates that the peak was present in the blank

100

27

22

17

12

Retention Time (min)

 

1x1-
27

1

1x1

prov [(1112 :1 peimnmsun ‘_
vwow 991111118an [AIME—93

j
outageg ‘5’

J

VEICIW\ *—%
metequlfimaid \j

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.00E8

VIAIGW
* j»
Iomdmd-zacm 'dzdow / 910995081
* nauoxodtd
J
ounnmothueqiow J
-)t- f
/ m.
«j : -§
2 \ ._: ausmofixotpeuolfiqiaw
9101598 T 4
O
1.1.}
O
oouepunqv 3

Figure 4.12: Chromatogram of MDMA exhibit T-27 extracted by HS-SPM E;
an asterisk (‘) indicates that the peak was present in the blank

10]

22

17

12

Retention Time (min)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.00E8

/
spiov [(1112 :1 potexmesun
Z
autoyef) /{_€
vwcrw palnlnsqns Mina

aurixo dzctowxsi

awteqmd Ifiqwtafa é
vwow
IOWdOJcI-ZdCiW-j
dZdCIW

ounuetoqdmeqiow .4
O
LL]
0
oouepunqv o_
0

Figure 4.13: Chromatogram of MDMA exhibit T-27 extracted by LLE

102

27

22

17

12

Retention Time (min)

More impurities were extracted by the MAE/HS-SPME and HS-SPME
procedures than by the LLB procedure. The MAE/HS-SPME procedure extracted six
impurities that eluted during the first ten minutes of the GC-MS analysis that were not
present in the HS-SPME extract analysis. Higher molecular weight compounds may be
degrading during the MAE resulting in the formation of lower molecular weight
compounds with lower boiling points. Because these compounds have not yet been
identified, this hypothesis cannot be confirmed.

The impurity 3,4-methylenedioxytoluene was only detected in the HS-SPME
chromatogram. The impurity may have been present in the MAE/HS-SPME extract, but
other unidentified peaks were present in the retention time range may have masked the
impurity since it is only present at trace levels. The impurities safrole, piperonal,
isosafrole, and MDEA were extracted by both MAE/HS-SPME and HS-SPME but not
detected in the LLE chromatograms because of their low concentration. The impurities
MDP2P and MDP2-propanol were extracted by all three procedures, but at lower
abundances by LLE than the other two extractions due to the lack of pre-concentration.
The impurity N-methyl-(1,2-methylenedioxy)-4-(l-ethyl-2-aminopropyl) benzene (ethyl
substituted MDMA) was extracted by all three procedures at similar abundances. The
impurity MDP2P-oxime was extracted by LLE but not by MAE/HS-SPME or HS-SPME
due to its low volatility

The components methamphetamine, MDMA, diethyl phthalate, caffeine, and fatty
acids were extracted by all three procedures. Unsaturated fatty acids were extracted in
higher concentrations by LLE resulting in broad peaks. Unsaturated fatty acids are best

separated by polar GC columns while saturated fatty acids can be separated by non-polar

103

 

columns [16]. The column used for this research, an inTM-Sms, is a non-polar column
composed of a 5% diphenyl and 95% dimethylpolysiloxane stationary phase. Therefore,
it is possible that the fatty acids present in exhibit T-27 are unsaturated fatty acids
resulting in the poor chromatography. The fatty acids were extracted by MAE/HS-SPME
and HS-SPME but in low concentrations due to the low volatility of the components.

The average PPMC coefficients and the standard deviation associated with the
triplicates of each extraction procedure are shown in Table 4.12. As mentioned earlier,
ideally, PPMC coefficients of replicates should be at least 0.99. The values for these
replicates show strong correlation, but not as strong as expected for replicates. This is
again possibly due to slight changes in the extraction procedure such as the small
fluctuations in the temperature of the water bath. The HS-SPME procedure shows the
lowest PPMC coefficient, but still indicates a strong correlation. The higher PPMC
coefficients for MAE/HS-SPME and LLE indicate these two procedures are more precise
than the HS-SPME procedure. The standard deviations of the MAE/HS-SPME and LLE
procedures are lower than the standard deviation for the HS-SPME procedure. This
indicates that the replicates for MAE/HS-SPME and LLE are more similar than replicates

of the HS-SPME procedure.

Table 4.12: Average PPMC coefficients and standard deviations of exhibit T-27 associated with each
extraction procedure

 

 

Extraction Aver:§:;PMC Standard Deviation
MAE/HS—SPME 0.9798 0.0044
HS-SPME 0.9388 0.0221
LLE 0.9776 0.0093

 

104

4.4.3. Summary

Table 4.13 summarizes the number of impurities and other components extracted
from each exhibit by each extraction procedure. As evident in the table, the number of
impurities extracted fi'om each exhibit by MAE/HS-SPME and HS-SPME is greater than
the number of impurities extracted by LLE. The number of other components extracted
from each of the exhibits by each of the procedures is similar. The average PPMC
coefficients and standard deviations are also shown in the table. The PPMC coefficients
of the replicates indicate a strong positive correlation among replicates of each MDMA
exhibit extracted by each procedure. However, as mentioned previously, these PPMC
coefficients are not as strong as expected for replicates (expect 0.99 or greater) because

of small changes in the extraction procedure between replicates.

105

 

 

 

 

 

386 2.36 c E mag
$8.0 336 m S» mac/Emmi
good wonod 0 cm mgmmfig
cmood m _ mod N. mm m5.—
woood mnoad m mm gammm
o _ ood 336 m ow mgmmmzhfiz m
oiod 33.0 o w m5;
386 :Nod m we muzmmm:
$36 536 o NV mZLméE
.8535: Ann—c 8: 2.9580 moat—EE— =ocogxm .325—

Eéaam 02mm owes: 552.355 3.3.5.2

 

Table 4.13: Summary of number of impurities and components extracted from each MDMA exhibit by
each extraction procedure; also sh own are average PPMC coefficients and standard deviations for each

exhibit extracted by each procedure

Despite extracting more impurities than LLE and HS-SPME with good precision,
MAE/HS-SPME has several downsides that make its use for the organic impurity
profiling of MDMA tablets unrealistic at this point. The MAE/HS-SPME procedure
requires the use of an additional instrument which increases the cost of analysis. In
addition, extra time is required to complete the analysis, and because vessels and inserts
are reused, there is the possibility of sample carryover between extractions.

The HS-SPME procedure extracts more impurities than the LLB procedure with
similar precision. During this procedure, the sample is left unattended for 40 minutes
allowing the scientist to continue with other work. Because the vials used for HS-SPME
can be discarded, the possibility of carryover is overcome using HS-SPME alone,
assuming the fiber is properly cleaned after each experiment. However, special
equipment must be purchased to use HS-SPME which increases the cost of analysis. For
example, the fibers and fiber holder as well as different GC inlet parts (e.g. liner, Merlin
MicrosealTM, etc.) must be purchased.

The LLE procedure extracts the fewest impurities from the MDMA exhibits of the
three procedures studied. Because the LLE procedure uses equipment commonly found
in the laboratory (e.g. sonicator, centrifuge, and vortex), the cost of analysis is relatively
low. However, the procedure requires more hands-on time during the extraction because
no single step in the extraction procedure is longer than 10 minutes. Therefore, sufficient
time for other work is not available.

If the costs of the fibers and other equipment can be absorbed, HS-SPME is the
best choice for extracting organic impurities from seized MDMA exhibits among the

three procedures discussed in this work. The HS-SPME procedure offers a compromise

107

between the other two procedures: it extracts more impurities than LLE, but does so at a
lower cost and shorter analysis time than MAE/HS-SPME. As mentioned earlier, LLE
extracts components that were not extracted by HS-SPME. Therefore, if sufficient time,
sample, and equipment are available, more information about the MDMA sample can be
obtained by performing both HS-SPME and LLE with different aliquots of the same
exhibit.
4.5. Comparison of MDMA Exhibits

Because HS-SPME is the best choice for extracting impurities from MDMA
tablets, the HS-SPME chromatograms of each of the MDMA exhibits were used to
compare the exhibits to one another and to determine the possible synthetic route used to
manufacture the MDMA in each exhibit.
4.5.1. Comparison of MDMA Exhibits MSU900-01, T-l7, and T-27

The impurities 3,4-methylenedioxytoluene, safrole, and piperonal were extracted
from all three exhibits. These impurities are present from the synthesis of MDP2P which
was also extracted from all three exhibits. lsosafrole, an intermediate in the synthesis of
MDP2P from safrole, was extracted from exhibits MSU900-01 and T-27 but not from
exhibit T-17. The presence of safrole and piperonal in all three exhibits suggests that the
MDP2P in all exhibits was synthesized from safrole. However, the oxidation of
isosafrole may have been more efficient for exhibit T-l7 (see reaction schemes in
Appendix A).

The utility of LLE in addition to HS-SPME is illustrated by the presence of the
impurity MDP2P oxime in the LLE chromatograms of all three exhibits. This impurity

indicates that the MDP2P was synthesized from the reduction of B-nitroisosafrole. Based

108

on the presence of safrole, piperonal, and MDP2P oxime, it is likely that the MDP2P
oxime in all three exhibits was synthesized through Route 2 shown in Appendix A [12].

The presence of N-methyl-(l,2-methylenedioxy)-4-(1-ethyl-2-aminopropyl)
benzene in all three exhibits suggests that the reductive amination route may have been
used to convert MDP2P to MDMA. This impurity is structurally similar to N-ethyl,N-
methyl(l,2-methylenedioxy)-4-(2-aminopropyl)benzene which is a by-product of the
reductive amination of MDP2P by ethylamine, an impurity in methylamine [12]. Also,
MDP2-propanol is formed during the reduction of MDP2P to MDMA. However,
because the impurities N-methyl-(l,2-methylenedioxy)-4-(l-ethyl-2-aminopropyl)
benzene and MDP2-propanol are not limited to the reductive amination route, the
hypothesis that the reductive amination route was used to manufacture the MDMA in the
three exhibits cannot be proven.

There are many similar active ingredients and other tablet components including
adulterants and additives. Methamphetamine, MDMA, and caffeine were present in all
three exhibits. Exhibits MSU900-01 and T-l7 both contained lidocaine which was not
present in exhibit T-27. Diethyl phthalate was extracted from exhibits MSU900-01 and
T-27; however, a different, unidentified, phthalate was extracted from exhibit T-17. Fatty
acids were only extracted by HS-SPME from exhibit T-27.

The exhibits can be discriminated from one another based on the presence and the
levels of the impurities present in the tablets. In addition to the lack of isosafrole, the
presence of ephedrine and the high abundance of MDEA in exhibit T-l7 discriminate
exhibit T-17 from the other two exhibits, MSU900-01 and T-27. Exhibit T-27 contained

three major unidentified impurities (numbers 75, 81, and 87 in Appendix P) which were

109

 

not present in either of the other two exhibits thus discriminating exhibit T-27 from
exhibits MSU900-01 and T-l7. Even though the MDP2P and MDMA present in the
three exhibits could have been synthesized by the same synthetic route, the three exhibits
were likely produced by different laboratories because many tablet impurities and
components were present in different concentrations [9].
4.5.2. Comparison of MDMA Exhibits T-17 and CJ-F805

A fourth MDMA exhibit was available for comparison: CJ-FSOS. Exhibit C]-
F805 was obtained from the Michigan State Police Forensic Science Division Laboratory
in Northville, MI, in March of 2007. This exhibit had similar physical characteristics to
exhibit T-l7, which was obtained from the Michigan State Police Forensic Sciences
Division Laboratory in Bridgeport, MI, in January of 2009. Only one tablet was
available for analysis from exhibit CJ-FSOS. This tablet was ground with a mortar and
pestle, and three 50 mg aliquots were extracted by the same HS-SPME procedure as used
for the other MDMA exhibits. The resulting impurity profiles of exhibit CJ-FSOS were
compared to profiles obtained from exhibit T-l7. Figure 4.14 shows a comparison of the
chromatograms of the two exhibits extracted by HS-SPME, and Appendix G lists all

impurities and components extracted from these two exhibits.

110

 

(.1)

1...!

 

 

L

l

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

r 7 v
2

l L

_ 1

I 1

m .

.L I
<EDE vflfiumaam 35m J. P. <EQE 08333.5 TEE r

05835 05335
26th * oEofimU
Ia ., , ¥
<mQZ nan—msfimI/I * i 7 .m <mn=2 3m.a5.E/a

1 al. ( l/ I

/jfl. m g

<20: 1“ F. <35: * \v 1

\ a. N“ n a \
Esaoaéoz aegis/WEE .m .05 5-292 :22an hum—2
“Etna—Em m 05.6215
.2. . mm. M a.

o_ofim\L

 

 

 

oEESofiEanE

econ—ouhonocoifioE

 

 

8.2

~\_/
81

Ilm
A“
V IN 7

s
i
/ x

\

 

 

L

 

 

 

a) 1.00E8

oofivcan<

 

 

 

oEESonn—Emfioi

 

0520356053;on

/

 

 

 

 

 

0.00EO
1.00E8

b)

OU§U==£<

27

22

17

12

Retention Time (min)

Figure 4.14: Chromatograms of MDMA exhibits a) T-l7 and b) CJ-FSOS extracted by HS-SPM E;

an asterisk (*) indicates that the peak was present in the blank

Ill

As evident from the chromatograms, these two exhibits appear chemically similar.
Most major peaks and identified peaks are present in both samples. However, the two
exhibits differ by a few minor peaks. For example, safrole was extracted from exhibit T-
17 but was not detected in exhibit CJ-FSOS. Safrole was present in exhibit T-17 at trace
levels; therefore, if safrole was present in exhibit CJ-F SOS at lower levels than in exhibit
T-17, it would not have been detected above the baseline noise. Because these two
samples have similar overall profiles that only differ slightly in impurities extracted and
relative abundances, it is possible that these samples were produced by the same
laboratory using the same synthetic route but potentially with the MDMA produced in
different batches, which is consistent with the physical characteristics [9]. However,
since only one tablet was available from exhibit CJ-FSOS, a definitive conclusion cannot
be made.
4.6. Summary

A MAE/HS-SPME procedure was optimized for the extraction of organic
impurities from seized MDMA tablets. Using a 1 M phosphate buffer at pH 8, the MAE
part of the procedure included a 23 minute ramp time to 100 °C and a 23 minute
extraction time at 100 °C. The HS-SPME part of the procedure included a 5 minute pre-
heat at 70 °C and a 40 minute extraction at 70 °C. This combination of procedures was
compared to a LLE procedure available in the literature [4] and to the HS-SPME
procedure alone using a simulated MDMA sample and seized MDMA tablets.

While MAE/HS-SPME extracted more impurities overall than HS-SPME alone or
LLE, HS-SPME alone was determined to be the most practical procedure for the

extraction of organic impurities of the extraction procedures studied. The limitations of

112

 

MAE/HS-SPME, which included a longer extraction time, higher costs, and
contamination problems, made the MAE part of the extraction impractical. The
shortcomings of LLE, which included fewer impurities extracted, made LLE a less
desirable extraction procedure than HS-SPME.

Three MDMA exhibits were differentiated based on their chemical composition
using HS-SPME alone. However, it was ultimately determined that HS-SPME and LLE
gave complimentary information to one another, and therefore, if time and money allow,
both procedures should be performed. Using both HS-SPME and LLE, the possible
synthetic routes used to manufacture the MDP2P and the MDMA were determined for

the three MDMA exhibits.

113

 

4.7. References

[l] Byrska B, Zuba D. Profiling of 3,4-Methylendioxymethamphetamine by Means of
High-Performance Liquid Chromatography. Anal Bioanal Chem 2008; 390: 715-722.

[2] Kuwayma K, Tsujikawa K, Miyaguchi H, Kanamori T, Iwata Y, Inoue H, Saitoh S,
Kishi T. Identification of Impurities and the Statistical Classification of
Methamphetamine using Headspace Solid Phase Microextraction and Gas
Chromatography-Mass Spectrometry. Forensic Sci Int 2006; 160: 44-52.

[3] C. Thompson. 2004 http://www.umt.edu/medchem/teaching/LectureS- ‘ '
Pharmaceutics%20(buffer-partition).pdf <Accessed 8 June 2009>.

[4] van Deursen MM, Lock ERA, Poortman-van der Meer AJ. Organic impurity profiling ;;
of 3,4-methylenedioxymethamphetamine (MDMA) tablets seized in the Netherlands. Sci
Justice 2006; 46: 135-152. 3

 

[5] Andersson K, Jalava K, Lock E, Huizer H, Kaa E, Lopes A, Poortman-van der Meer
A, Cole MD, Dahlen J, Sippola E. Development of a Harmonised Method for the
Profiling of Amphetamines IV. Optimisation of Sample Preparation. Forensic Sci Int
2007; 169: 64-76.

[6] Coumbaros JC, Kirkbride KP, Klass G. Application of Solid-Phase Microextraction to
the Profiling of an Illict Drug: Manufacturing Impurities in Illicit 4-
Methoxyamphetamine. J Forensic Sci 1999; 44: 1237-1242.

[7] Pawliszyn J. Solid Phase Microextraction: Theory and Practice. Wiley-VCH, New
York; 1997.

[8] Swist M, Wilamowski J, Zuba D, Kochana J, Parczewski. Determination of Synthesis
Route of 1-(3,4-methylenedioxyphenyl)-2-propanone (MDP2P) based on Impurity
Profiles of MDMA. Forensic Sci Int 2005; 149: 181-192.

[9] Palhol F, Boyer S, Naulet N, Chabrillat M. Impurity profiling of seized MDMA
tablets by capillary gas chromatography. Anal Bioanal Chem 2002; 374: 274-281.

[10] Araujo PW, Brereton RG. Experimental Design 11. Optimization. Trends Anal Chem
1996; 15: 63-70.

[1 l] Wercinski SA ed. Solid Phase Microextraction: A Practical Guide. Marcel Dekker,
Inc., New York; 1999.

[12] Gimeno P, Besacier F, Bottex M, Dujourdy L, Chaudron-Thozet H. A Study of

Impurities in Intermediates and 3,4-Methylenedioxymethamphetamine (MDMA)
Samples Produced via Reductive Amination Routes. Forensic Sci Int 2005; 155: 141-157.

114

[13] Gimeno P, Besacier F, Chaudron-Thozet H, Girard J, Lamotte A. A Contribution to
the Chemical Profiling of 3,4-Methylenedioxymethamphetamine (MDMA) Tablets.
Forensic Sci Int 2002; 127: 1-44.

[14] Moffat AC, Jackson JV, Moss MS, Widdop B. Clarke’s Isolation and Identification
of Drugs. Pharmaceutical Society of Great Britain, London; 1986.

[15] Devore JL. Probability and. Statistics for Engineering and the Sciences, 4th Ed.
Duxbury Press, Belmont, CA; 1995.

[I6] Baer I, Margot P. Analysis of Fatty Acids in Ecstasy Tablets. Forensic Sci Int 2009;
doi 10.1016/ i.forsciint.2009.03.015.

 

115

Chapter 5 Conclusions and Future Work
5.1. Conclusions

A microwave-assisted extraction (MAE) procedure was developed and optimized
for the extraction of organic impurities from seized MDMA tablets. Three different types
of buffers at three different pH values and concentrations were studied to determine the
optimum buffer for use with the microwave extraction. Using a simulated sample
containing benzylamine, phenethylamine, methamphetamine, ephedrine, and caffeine, the
optimum buffer for use with the microwave was determined to be 1 M phosphate buffer
at pH 8. This buffer allowed for the precise extraction of all components with no sample
carryover in the vessels between extractions. Buffers of lower pH did not extract all
components of the simulated sample or extracted components with poor precision.
Buffers of higher pH extracted all components; however, carryover of the sample in the
microwave vessel between sample extractions limited the use of high pH buffers.

A full factorial experimental design in four blocks was used to determine the
microwave parameters that had an effect on the extraction of impurities from MDMA
samples. Using the simulated sample, the parameters of ramp time, extraction time, and
extraction temperature were studied. Ramp time and extraction time were studied at
times of 10 and 20 minutes with a center point of 15 minutes, while the extraction
temperature was studied at 80 °C and 120 °C with a center point of 100 °C. From this
experimental design, all three parameters were found to have a significant effect on the
extraction of the simulated sample components.

The three parameters of ramp time, extraction time, and extraction temperature

were then optimized using the simulated sample in a circumscribed central composite

116

 

 

(CCC) experimental design. In the CCC design the range of values for the ramp time and
extraction time was 7-23 minutes and the range of values for the extraction temperature
was 66-134 °C. Using a desirability function, the optimum parameters determined were
determined to be a 23 minute ramp time to 100 °C and a 23 minute extraction time at 100
°C. These parameters allowed for the minimization of the methamphetamine and
caffeine peaks while allowing for the maximization of benzylamine and phenethylamine
peaks.

Because of the efficient extraction of the microwave, a second, more selective
extraction technique, headspace solid-phase microextraction (HS-SPME), was utilized to
selectively extract the organic impurities from the sample. The HS-SPME parameters of
extraction time and extraction temperature were optimized empirically. The extraction
time was studied over a range of 10-60 minutes holding the extraction temperature at 70
°C. The extraction temperature was studied over a range of 40-80 °C holding the
extraction time at 40 minutes. The shorter extraction times (10-30 minutes) and lower
extraction temperatures (40-60 °C) did not extract as many impurities as the longer
extraction times (40-60 minutes) and higher extraction temperatures (70-80 °C).
However, the longest extraction times (50-60 °C) and highest extraction temperature (80
°C) resulted in the extraction of components, such as MDMA, in high concentrations
resulting in broad peaks that could potentially mask impurities present at lower
concentrations. An extraction time of 40 minutes and an extraction temperature of 70 °C
were chosen as the optimum HS-SPME parameters because these parameters offered

high impurity abundance without sacrificing chromatography.

ll7

 

 

Finally, the developed MAE/HS-SPME and HS-SPME procedures were
compared to a liquid-liquid extraction (LLE) procedure from the literature [1] using three
seized MDMA exhibits: The combination of MAE/HS-SPME allowed for the extraction
of more impurities and components than HS-SPME and LLE. However, MAE/HS-
SPME had limitations that restricted its use for organic impurity profiling such as
increased extraction time and increased cost. Also, a problem of sample carryover in the
microwave vessels and inserts between extractions limited the use of MAE for the
extraction of impurities from MDMA exhibits. The LLE procedure required the least
amount of new equipment therefore offering the lowest cost. However, more analyst
involvement was required to complete the steps of the extraction and fewer components
were extracted and detected. Because of these limitations, HS-SPME was determined to
be the most practical extraction procedure of the three procedures studied for the
extraction of organic impurities from MDMA exhibits. The HS-SPME procedure offered
the extraction of more impurities than LLE and did not have the limitation of sample
carryover between extractions like MAE. Also, the cost of HS-SPME analysis was less
than the MAE/HS-SPME analysis because the microwave was not needed.

All three extraction procedures allowed for the extraction of impurities and
components from the MDMA exhibits that helped to determine the synthetic route used
to manufacture the MDMA in the exhibits. Using the chromatograms of the MDMA
exhibits extracted by HS-SPME and LLE, the possible synthetic route for the synthesis of
3,4-methylenedioxyphenyl-2-propanone (MDP2P) and MDMA were determined. The
MDP2P present in the exhibits could have been synthesized from safrole through the B-

nitroisosafrole route (Route 11 shown in Appendix A). The MDMA may have been

118

cal-9‘

 

synthesized using the reductive amination route (shown in Appendix B); however, few
impurities were identified to establish this link and therefore the determination of the use
of this route is a preliminary hypothesis. Based on the chromatograms of the MDMA
exhibits obtained by HS-SPME, the three MDMA exhibits were successfully
differentiated from one another based on chemical composition.

This work shows that the developed HS-SPME procedure can be utilized for
organic impurity profiling of MDMA exhibits. Because of the pre-concentration of the
impurities on the fiber, many impurities are extracted from MDMA samples, even those
present at trace levels. Therefore, the synthetic route used to manufacture the MDMA
could be determined with more certainty and more points of comparison among exhibits
are available, thus aiding law enforcement in the connection of tablets from different
MDMA exhibits.

Organic impurity profiling is more often performed in research laboratories than
in local state and city crime laboratories. Ofien, local law enforcement laboratories are
only interested in the active ingredients present in the MDMA tablets, such as MDMA
and methamphetamine. Therefore, the MAE/HS-SPME and HS-SPME procedures
optimized during this work (which were optimized for impurity extraction) are more
likely to be utilized by research laboratories. However, HS-SPME may still be applicable
to local crime laboratories. Because of the simple sample preparation (the sample is
ground and placed in 5 mL of buffer) and the possibility of automation, the HS-SPME
procedure could be optimized for the extraction of active ingredients and thus be relevant

to local crime laboratories.

119

 

5.2. Future Work

Even though the MAE/HS-SPME procedure developed in this work has downfalls
that hinder its use in a crime laboratory at this time, further work could be performed to
successfully allow its use. Because one of the main downfalls of the use of the
microwave was the sample carryover in the vessels between extractions, studies could be
performed to determine a more efficient procedure to clean the microwave vessels and
the quartz inserts. This would reduce the overall time required for MAE as well as
overcome the contamination problem.

The application of Pearson product moment correlation (PPMC) coefficients can
be expanded from comparing replicates of the same exhibit to comparing chromatograms
from different exhibits. Other statistical procedures, such as hierarchical cluster analysis
and principal components analysis, could also be applied to the data to determine the
level of similarity or association among exhibits. By applying statistical procedures, the
determination of the similarity of MDMA exhibits would be objective instead of
subjective as was the case in this work, thus minimizing the possibility of experimenter
bias.

As evident from this work, many unidentified impurities were extracted from the
MDMA exhibits. Further work could be completed to determine the identity of these
compounds. Tandem mass spectrometry is a technique that allows for the selection of
target ions and fiirther fragmentation of these ions in order to determine their structure.
Ion trap mass spectrometers, like the one used in this work, have the ability to perform

tandem mass spectrometry. By identifying more of the impurities extracted from MDMA

120

 

exhibits, more clues to the synthetic route used and more points of comparison among
exhibits would be available.

Much of the work in identifying impurities in MDMA tablets has focused on the
organic impurities. The study of the inorganic impurities present in the exhibits could
also be useful for comparing tablets. Inductively-coupled plasma mass spectrometry
(lCP-MS) can be used to identify trace metals present in the tablets. Once again, more
points of comparison among exhibits increases the ability to link tablets from different
MDMA exhibits to a common source or, alternatively, increase the certainty with which
tablets are determined to be unrelated.

The extraction of organic impurities can assist law enforcement in determining
the synthetic route used to manufacture the MDMA as well as link tablets from different
exhibits to a common production source. Headspace solid-phase microextraction has
proved useful in the extraction of the impurities. While some work still needs to be
completed before the technique is fully applicable to crime laboratories, the extraction

procedure shows promise in its eventual use.

121

l...‘" ‘TI- _. _

 

5.3. References

[1] van Deursen MM, Lock ERA, Poortman-van der Meer AJ. Organic Impurity Profiling
of 3,4-Methylenedioxymethamphetamine (MDMA) Tablets Seized in the Netherlands.
Sci Justice 2006; 46: 135-152.

 

122

Appendix A: Synthesis Schemes of 3,4-Methylenedioxyphenyl-2-Propanone (MDP2P)

“(12>

 

Safrole
KOH
EtOH
0
Route 1 "C02” lsosafrole C6H7NO3$ Route 2
H202 H2804
OH %
O
OH O O
lsosafrole glycol Piperonal
H2804 C2H5N02
O\C ::> mi:
Mcthylcnodioxyphcnyl-Z- B—Nitroisosal‘rolc
propanonc(MDP2P)
CH3COOH

m0
N >
HO/ 0

3,4-Mcthylcnodioxyphcnyl-2-
propanonc oxime

CH3COOH
\n/Uo
. >
O
Mcthylcncdioxyphcnyl-Z-

propanonc (MDP2 P)

123

 

 

Appendix B: Synthesis Schemes of 3,4-Methylenedioxymethamphetamine (MDMA)

from MDP2P

22824582 I

£2: 12%Eaeégoaaoaa
4292 "ECO/m 4303;560:2362
O O

A. a... \QQLA

O
385:3 23:3
0

 

vzmsz ”E
zomzmaz .=
3:2 ._
2.3892323591505.?3 $ng ozoEQEQN
52% 04-9560835/oev- N d .Mcmfinqzxofiocflmnsz
o Iz\
A u: .5 .2 g wazmmnv
0
E533 95 0:13—

124

Appendix C: Experimental Runs for Full Factorial Screening Design

 

 

 

 

 

 

Block Ramp Time Extraction Time Extraction Temperature
(minl (min) (°C)
1 15 15 100
1 10 20 120
1 20 10 80
1 15 15 100
2 10 10 120
2 20 20 80
2 15 15 100 f 1
2 15 15 100 ; '
3 20 10 120 '-
3 10 20 80
3 15 15 100
3 15 15 100
4 20 20 120
4 10 10 80 g.
4 15 15 100
4 15 15 100

 

125

Appendix D: Experimental Runs for CCC Optimization Design

 

Ramp Time Extraction Time Extraction Temperature

 

 

Run (min) (min) (0 C) Type of Point
1 15 15 100 center point
2 10 10 8O factorial
3 20 20 80 factorial
4 15 15 66 star
5 15 15 134 star
6 15 15 100 center point
7 20 I 0 80 factorial
8 15 15 100 center point
9 1 O 20 80 factorial
10 15 15 100 center point
1 1 15 15 100 center point
12 15 15 100 center point
13 23 15 100 star
14 7 15 100 star
15 20 20 120 factorial
16 10 10 120 factorial
17 15 15 100 center point
18 15 15 100 center point
1 9 20 10 120 factorial

20 15 15 100 center point
21 10 20 120 factorial
22 15 23 100 star

23 15 7 100 star

 

126

Appendix E: MANOVA Table for Simulated Sample Components from Full Factorial

 

 

 

 

 

 

 

 

 

 

Design
Benzylamine
Source Sum of Squares Mean Square F-Ratio
A:Ramp Rate 9.63E+l4 1 9.63E+14 4.65
B:Extraction Time 1.21E+l4 1 1.21E+14 0.59
CzExtraction Temperature 9.50E+14 1 9.50E+14 4.58
AB+block 1.25E+14 1 1.25E+14 0.61
AC+block 1.SZE+14 1 1.52E+14 0.73
BC+block 1.63E+] 5 1 1.63E+15 7.89
blocks 1.89E+15 3 6.31E+14 3.04
Total error 1.24E+15 6 2.07E+14
Total (corr.) 5.55E+15 15
Caffeine
Source Sum of Squares Mean Square F-Ratio
A:Ramp Rate 2.03E+14 1 2.03E+14 0.06
BzExtraction Time 7.21 E+14 1 7.21E+14 0.21
CzExtraction Temperature 3.67E+14 1 3.67E+14 0.11
AB+block 2.33E+15 1 2.33E+15 0.69
AC+block 1.51E+15 l 1.51E+15 0.45
BC+block 1.44E+15 1 1.44E+15 0.43
blocks 1.87E+16 3 6.24E+15 1.84
Total error 2.03E+16 6 3.39E+15
Total (corr.) 4.45E+16 15
Methamphetamine
Source Sum of Squares Mean Square F-Ratio
A:Ramp Rate 4.84E+15 l 4.84E+15 2.85
BzExtraction Time 3.55E+15 1 3.55E+15 2.09
CzExtraction Temperature 8.27E+16 l 8.27E+16 48.6
AB+block 1.21E+16 1 1.21E+16 7.11
AC+block 7.64E+15 1 7.64E+15 4.49
BC+block 1.64E+16 1 1.64E+16 9.64
blocks 2.11E+l6 3 7.02E+15 4.13
Total error 1 .02E+16 6 l .70E+15
Total (corr.) 1 .60E+17 15

 

127

 

’ "' ION“.

 

 

Phenethylamine

 

 

Source Sum of Squares Mean Square F-Ratio
A:Ramp Rate 6.90E+13 1 6.90E+13 0.07
B:Extraction Time 4.08E+11 l 4.08E+11 0
CzExtraction Temperature 2.29E+14 l 2.29E+14 0.22
AB+block 1.41E+15 1 1.41E+15 1.35
AC+block 1.01E+l4 1 1.01E+14 0.1
BC+block 1.84E+14 1 1.84E+14 0.18
blocks 1.96E+15 3 6.53E+14 0.62
Total error 6.27E+15 6 1.04E+15
Total (corr.) 9.56E+15 15

 

128

 

'W‘d! r - ._ n
I‘ "

Appendix F: Full List of Compounds Extracted from MDMA Exhibits MSU900-01, T-

17, and T-27 using MAE/HS-SPME, HS-SPME, and LLE

 

www

3mm

5N9
5 TH
_ c.8252

we .me .8 .8

o Basses:

 

N56

5N4.
N TH
5.8352

3. .aoN .mS

a 35525

 

Nvd

><><

5N9
5 TH
3.08sz

mm. .m_N.:N

a 85.52::

 

BN4.
h _..._.
3.25392

_m_ .2: .3. .R. .mm.

e 85.52::

 

O‘
V

00
V

RNA.
5 T...
_ c-8252

hm Ko .3 .3 .mm

w. 852.85:

 

O5
N

hmnh
5 TH
SioooDmE

he .mm .55 .mm .3

4 85:52.5

 

OO
xxxxxxf'zxfljxxx
v

BN4.
5 TH
5-25sz

mm .3 .R. Km Kc

m 3528...:

 

x
.mN
.mN

5N4.
5 TH
3.8252

mm_ .5 .mm_ .mm_ H».— .E

N 35202::

 

X><><><><XXXXXXXXXXXXXXXXXXXKX

x
AEEV .5—
HA:—

xxxxxlfixxxlfixmxxxrflx><><><l\,><><><><><

x
2:5

.5— BEA—mam: .5— mEmmimEm<E

x
oNN
x
2:5

5N4.
5 TH
8-25me

.EExH

om\mm .3. .3. .8 .3 .K

a}:

_ 85.52::

.0353 3563:...—

 

129

I-

II..I.L|‘~. 4 1

 

 

 

 

 

 

 

 

 

 

x x x Emir—i
x .33 9.3 :9 3 we .3 2 85.52::
s. 33 :3 3.83sz
x a; m; 3;.
x 33 3.» :3 3 .3 .3 2 Engage:
x mm.” a; 3.8332
x x x hN-H
x 33 8.” :9 E .3 .3 we .3: 2 35.52::
s. x x 3.23sz
x a.” x 32
x x x :é o: .3 .3 .3 .3 .3 3 33.525
x x x 3.8892
x m; x 29
x t.» R.» :9 3 .3 .02 32 ”5205235350243
x cm.» 3.” 3.83m:
x x NE 29
x x x :4. 3 .3 .3 2 85202::
x x x .3382
x x x hNuF
x 2:. :3 3 E R .3 .w: 2 Engage:
x 8.3 x 3.833:
x 8.0 36 R;
x 8.0 o; :3 3 .3 .3 .m: .E : 35.52::
x 36 MS 3.83%:
x x 8.9 39
x x x :3 m2 .3 .3 .3 o. BEEES
x x x 3.8332
3.5 5. 3:5 3.5
5.. 5. “Emma: .5. ”.2.—35:2 .325 NE 552 .23.??—

 

I30

 

 

 

 

 

 

 

 

 

 

 

511%.
x RR. .3. R9
x R... an: :9 B .R. .8. .3. .228...
x RR. .3. 8.8.82
x x x hN-P
x 2.: 2.: :9 3. .R .8. .2. R 85.52::
x x x .38sz
x 3.: mm. . R9
x 9.... E... :9 .9 .R .35.... .5 .8. 23mm
x 2. . . E. _ 8.83m:
x x x hN-F
x .32 3.: :9 R .3 .3 .m: .9: R 35.52::
x x x 8.83m:
x Rd. .2: R9
x x x :9 a: .R .5 .8. R 39:202....
x 3.: .2: 3.85m:
x x x bNaP
x 8... 8.... :9 8. .8 .mm .9; R 85.82::
x x x 8.83m:
x 8.9 2:: R9
x x x :9 .2. 4.2 .Rm. a. 85.52::
x x x 3-83.0.2
x x 2; R9
x x x :9 an .R .. : am. .2 35.82::
x x x 5.83%
S. a... $5 R9
a... $5 2... :9 o: .3 .mm 2.532.323:
3.. .3 a; 8-8sz
3.5 .5. 3.5 .55.
m1... .5. 55.5%. .5. m2.m.m=\m<.z .32.... 3:. 5:5... .2328...

 

131

 

 

 

 

 

 

 

 

 

 

 

x 3...: R: R9
x 3.: x :9 3.: .39 .R .8. .3 on 85.52::
x .6.: R: 3.8332
x x .6.: R9
:9 R: .2. .3 .:. R couscous:
x x x 3.83m:
x 3.: x R9
x x x :9 mm .3 .3 .3 R 85.52....
x x x 3.833.).
X X X thP
.9... 8.: m. .m. :9 3: .w: .R a: .3: R 85.52::
x x x 3.833.).
X X X FNIH.
S... x x :9 mm .3 .3 .8 cm 89:202....
x x x 3.833.).
X X X thP
8... 3.: 3.: :9 3 .3... .R .3 .3 2.62....
x x x 3.8ng
x 3.: 3.: R9
x x x :9 3.. .R .3. .8. .3. .8. 22.88.
x 3.: 3.: .983sz
x 3.: x R9
x x x :9 3.. .o: .3. .3 .2. R 8.5.525
x 3.: 3.: 5838.2
X X X hNul—L
x x x :9 .m. .3... .2. .3 3 3552::
x 3.: x 5833.).
2:5 .5. .55. 3.5 .55.... as has... .2332...

HQ..—

.5— BEA—mum: .5— m2mw-w—¢m—<E

132

 

 

 

 

 

 

 

 

 

 

R3. 8.... 93.... R9

83. N93. 3.... :9 o: .39 .3: .99 .33. -N-.9wwmmmwwuwmwmmmmét3

3R3. 3.3.. .93. 3.333sz
x x 3. R9
x x x :9 .: .3... .33. .3: .99. 33 BEEfiS
x 33.... 3...... 3.333232
3. x 33.... R9
x x x :9 99 .3: .33.. .33. .33 93 couscous:
x x x 3.333sz
x 33.3. x R9
x x x 9 .9 m: .3. .2359 .9. . .33. 33 8:53.32:
3. x x 3.333sz
x 3.: 33.... R9
x 3.3.... mm... 9 .9 39 .33. .99 .33 .3 33 Season;
x x 3. 3.33323:
x x x R9
3. x x 9 .9 .3 .3: .3: 33 Snags:
x 33.. 33.... 3.333232
3. 3.... x R9
3. x x :9 3: .33. .99 .33. .3 33 85.52::
x x x 3.833.).
x 33.3. 33.... R9
3. x x :9 39 .99 .:.. .3 .33. N3. BEBE:
x x 3. 3.333232
3. 33.: 33.: R9
3. x x 9 .9 33. .99 .3. . .3 .3 couscous:
x x x 3.3333:

.55. 5. 3.5 Ea.
9...... 5. ”.233: 5. mzmm-m=.m<2 .32.... :5 5.2.3... .23....65

 

133

 

 

 

. 9N-..

 

 

 

 

 

 

 

 

 

3 3 3 9 .9 33:3. .2 3.33 .33 3.. 333.232....
33... . 33.3. 33.3. 333333.).
3 93.3. 3.3. R9
3 3 3 :9 33 .3. .333 .33. .33 3. 333352....
3 3 3 3.333332
X X X hNH—t
.3... 3 3 :9 33.99 .33 .33. .33. .3 3.. 333.2323:
3 3 3 $3333.32
3.9.... 33.3. 3.3. R9
3.... 33.3. 33.3. :9 33. .33. .33. .33 3...£5333§3Mwfimww=23532i3
3.... 93.3. 33.3. 3.8323:
X X X BNIP
3 3 3 9 .9 33. .33 3.. 3333:3335
3 .33. 3 3.33323...
3 33.3. 3 R9
3 3 3 :9 3333:3313. .3 3339.35.33:
3 3 3 3.333233.
3 33.3. 3.3. R9
3 3 3 :9 39 .99 .33. .33 .3 3.. 32....53..5
.3 3 3 3-83:3:
33.3. 3 3 R9
3 3 3 9 .9 33. .33 .33. .33 .3 33 32.3.5335
3 3 3 $333232
33.... 93.... 33.... R9 22.333.33.523
33.3. 33.3. 3.3. :9 39 .39 .33. .99 .33. .33. .3....33...-3-. 3:2... 333.32.... 3.332-... 3
33.3. 3.3.... 33.3. 3.333232
~55 .5. 2...: ~55
3.... 5. 3.2.3-3.. 5. 3233-33.22 33.33... 3.... 33.5... 3.3.3.2....

 

 

I34

 

 

 

 

 

 

 

 

 

 

3. .3. 3 3 R9
3 3 3 9.9 33:33. .99 .33 .3 33.3.3323:
3 3 3 3.333232
33.3. 3 3 R9
3 3 3 :9 33. .33. .99 .39 .33 33 323.232....
3 3 3 3.33323...
3 33.3. 3...3. R9
3 . .3. 9...3. 33.3. 9.9 99 .39 .39 33.3.3.3. 3...... www.mmwao. 3.332-... 3
3 33.3. 9...3. 3133333.).
33... 33.3. 33.3. 93.9
3 3 3 :9 39. .99. .3... 3.3.3.3.... .3523.
33... 33.3. 33.3. 3.33323...
3 33.3. 3 R9
3 3 3 9.9 ..3. ..3. .99 .33. .33 3.. 335.332....
3 3 3 33333.32
3 3 33.3. 93.9
3 3 3 9.9 .3.. .33. ...3. 3.. 32....53..5
3 3 3 3.333232
3 3 3. .3. R9
3 3 3 :9 9.. .33. .33. .33 .9... 9.. 33.33523:
3 33.3. 3.3. 3-33323...
X X X hN-F
33... 3 3 9.9 33 .33 .39 .33 ..3 3.. 33333333...
3 3 3 3.333232
X X X bN-F
33... 93.3. 33.3. :9 33 ...3. ..3. .33 .3... 3.3.3.3....
3 3 3 3.33323...
2...... .5. 2...... .:..5
3.... 9.. 3.2.3.3.. 9.. n..3....3-3.....3..>. .:....33. 3.... 3.3.5... .2233...

 

I35

 

 

m..m.
2.2
N...“—

mm...
3. TH
.:.:omDmS.

2. .mm. .8. .:.... .3

38.393.2sz 0:...8 0:03.323.
3-3333333...3333.3...33.-...3...

 

3%.:

XX

9»:

XX

mm...
3. TH
3180sz

.a .2. .3. do.

33 323.232....

 

.53...

X

:0:

on:

R...
h T...
. 31.8sz

E.— .E. .m: .9... d..—

33 33.....333.3:

 

X

wad.

NNF
h 7...
3:89.52

No .mm .0» .3

mm 3:35.325

 

XXXXXXX

0:.—

mm...
3. T...
5.25sz

a: .Nom .3.

33 333.232....

 

XXXXX

mmd.

XXXXXXXXX

8.:

.:.:

mm...
3. 7..
5.25sz

mm. .53. .8. .mm .wom

33 323.232....

 

2.6.

.36.

um...
h T...
3.25sz

.w. .3. .mo. .2.

..3 335.5335

 

2.6.

mm:—

mm:—

3...:—

R...
n T...
.:.oooDm—z

2.1213 .8. .3.

33 323.232....

 

XXXXXXXXX

x
9...: .5.
@43—

X

X
93.3.
3...:

X

x
2.6.
2...:

by— m—Smmum: .5— mEmm-mEm~<—Z

RA.
5 T...
8.63sz

33......

v0. Kc. .3. .oc.

E...

33 32.35233.

3.....3... .333.3.>3...

 

136

 

 

 

9.92

9h-..
9. T...
5.95952

mm. .09.. .3

33 33.33523:

 

vod—

9.~-...
9. T...
3.80392

3 4m. .3 do.

9.: BE.§E=D

 

9h-..
5 T...
3.2.0me

on .3. .2.. .2:

33 323.232....

 

XXXXXXXXXX

$9..

No.9.—
3.:

mm...
9. T9.
.:-oooDmS.

a... .93 .3.

mo coarse—3...:

 

XXXX

3.9.—

mm...
b T...
3-95sz

N9. .3 .3.. .wom do.

we 35.52::

 

5.9..

9.N-._.
9. T...
.oéooDmE

mo. .3.. .m: .m: .om.

33 32.35.33:

 

XXXXXXXXXXXXXXXXX

XXXX

93.9..

XXXXXX

3.9..

9.N-...
n _ -9.
3.959sz

.:N .3. .2: .mo.

33 323.232....

 

3.:
X

2.2

9.N-._.
9. T...
.:.ooaDmS.

mm. .m... .m9.. ..mm .3

.3 32.333335

 

x
o_.m_
x
AEEM Py—
mun—1—

X
X
X
X
X
X

2...:

XXXXX

X
3...:

.59 mgr—mug.— .5— mEmmnmEm<2

9b-...
9. T...
.oéooDm—z

33...—um

gm .3 69.1.2

39...

cc 3:35.223

3.35... 333.333...

 

137

 

 

.3....

mod.

9.~.....
9. T...
3.53%.).

w... .2..

99 3333523:

 

XXXXXX

5.».

9.~.......
9. T...
3.25sz

.o .59. .Nm. .3.

39 83:32::

 

cm...

Nod.

hm-..
9. T...
3.825.).

mm. 6w. .3. .mm.

39 33:52::

 

mm.».

X

9.~-...
9..-...
5.0.532).

”9.. .wcm .8. 6m. .2.

39 85352::

 

$3.”.

9.N-...
9..-...
5.825.).

won .mm. .N9. .wm do.

39 3333523:

 

XX

93....

mm...
b T...
3.82%.).

m... .M9.. .2. .o.~

99 83352.5

 

XXXXXXXXXXXX

No...

X

hm-..
9. T...
5.25sz

ma .3 .95. 69..

.9 85.52::

 

X

Na...

XXXX

ofiw.

rm...
9. T...
3.83%.).

9.9. .mm. .wom .8. .mm

39 32.33532:

 

XXX

x
.:.—5 .5.
m1...

mm.w.
x
x
.:.—5

.5. mEmmnmE .5. HEmmumEm<E

9.~-...
9. T...
35035.).

.3.—.39..

.oN do. 50.

ES

33 83:52::

93:52 .2233...

 

138

 

 

 

 

 

 

 

 

 

 

33.3. 33.3. 3.3. 93....
33.3. 393. 33.3. 9.-. 33. .3. 2.3.35
39.3. 39.3. 33.3. 3.83:3:
3 33.3. 39.3. 93.3
3 3 3 9.-. .3. .33. .33. 33 33.-353...:
3 3 3 3.83:3:
X X X FNIP
33.3. 3 3 9.-.. 33 .33. .33 3.33. 3.3 33.-.3523:
3 3 3 3.83:3:
3 33.3. 33.3. 93..-
3 3 3 9.3 3 .3.. .33. 33 323.3523:
3 3 3 3.83:3...
X X X hN-H.
3 3 33.3. 9.-. 333 .39 .3 33 33.-.3532:
3 3 3 3.8323...
33.3. 33.3. 3.3. 93-3
3 3 3 9.-. 33.99 .3.: .33 .3 .3 33.-.3523:
3 3 3 533333....
3 3 3 93.3
3 3 3 9.-. 3.3 .39. .99 .333. 33 33.-.3523:
3 3. .3. 33.3. 3.83:3:
3 333. 3.3. 93-..
3 3 3 9.-. 33. .3 .93. ..3. 39 33.-3.23.3.5
3 3 $333232
X X X FNIH.
3 3 3 9.-. 33. .33 .33 .39. .3 39 83.2.2.5
33.... 3 3 $8323.).
3...: .5. .:.E. .:.E.
3...... .5. 3.2.3.3.. 3... 3.2.3-3353. 33......— NE 3.3.3... .2332...

 

139

 

.u'

 

 

 

 

 

 

 

 

 

 

3 3 3 9.N-...
3 3 3 9.7... 3.3. .39. .9.9. .3.: Na 32.52::
3 03.3 3 8.25me
3 3 3 9?...
3 3.36m modm 9..-... man .33. .9.9.N .2.. .a uoEEQvED
3 3 3 3.8..sz
3333. RM “WNW mm 33. .99 .33 .33. 2335.. 222.325.33.353
mod. Snow 33.8 .o-ocaDmS. .YT..mxo_cu:o_.£.oE N 3.352: 2
3 3 3 9.N-...
33.3. 3 3 9.7... 3. .mo. .3 .2. .3 9.. 3.2.5223
3 3 3 3.8352
3 33.8 whom 9.3-3
3 3 3 9..-... a: .mo. .3 on .3.-.35323
3 3 3 5.23sz
3 3 3 R-..-
w..o. wmdN atom 9..-... .3 .Q. .33 .33 3.2382.
3 3 Wow 3.3.3 3.233sz
3 aflom 3:3 mm...-
3 odom mm.o~ 9..-... 9.3. .mm. .33. .33 33 .32....5233
3 3 3 .o-oooDmS.
33.3. o _ .om om.o~ 9.~-...
3 3 3 5.... N3. .2.. 5 .6.-552::
3 3 3 .o-oooDmS.
3 3 3 9.-...
33. 3 3 9..-... 3. .mm .3 63. :2... 33 3.2.5223
3 3 3 .o-ooaDmS.
.:.E. .5. .:..5 ~55. .32.... 3.5 .355... .3.—23:6...—
HA-. .5. mini-m: .5. mEmm-mEmS). . . . . .

 

140

 

 

3.NN

codm
3.3

9.-...
9. .-...
3.23sz

9.3. .33 .3 .3.

9.. 83:32::

 

9.N.N~

9.....NN

93¢.
9. T...
.o-oooDmS.

3 .39. 69.133. .3...

33 35.32::

 

9.3.3

9.o.NN

3m...
9. .-...
333sz

39.. .3

3 35.52::

 

hN-H
9. .-...
.963sz

mom .3 .9.N~ .N3.

8 85.52::

 

X><><><><><><

N23

933..
9. T...
.o-oooDm—z

39. .mo. .8. d3.

3 33:52::

 

XXXXXKXXXXXXXXXXXX

9....N

93-...
9. T..-
.c-oooDm—z

9.9. .8. .9... .3..

a 35.52::

 

93.9..

X

8.3

hN-H
n T...
.c-oooDwS.

3N. do .39. .93 .mm

23 3.83 3.3.2.8....

 

3.9..
2.:

><><><><><

93-...
9. T...
353sz

33 .3. .33 .mm .93

28 2.3.5....

 

3
3
3
3:... h.“
mun—1—

36m

>6

3
3:5

.5— HSEm-m: Pm HELm-mEm<—>.

9.N-...
9. T...
3.33me

.333”.—

m: 63 .03.

n9...

3 35:82.5

33...—93 .:.—23.5....

141

 

I-I'I‘ Ilia! 3|...

 

XXXXXXXXXXXXXXXXX

X

3.3.

mm...
om...
x
x
x
.:..5 .5.
3....

No.3
M9..mm

XXXX

owdm

XXXXXXX

NNNM
3
9......

.5. mini-m: .5— 52.533336:

3..m~

3..mm

3.3

8.3

2.3

2.3

XXXXXX

2.3-N

X
X
X
X
X
X
X
X

.2....

9.m-...

9. .-...
.063sz
93-...

9. T...
.o-oooDm—z
9.N-....

9. .-...
35335.).
9.N-...

9. .-...
.o-oooDmS.
3H.-

3 .-...
.o-oooDmS.
RN.-

9- .-...
.o-oooDmS.
9.N-...

9. T...
.o-oooDmE
9-N-...

h T...
.o-oooDmS.
mm...

9. .-...
.o-oooDmS.

.:....3m

mm .39 .3 .3.

N...~...3m.~3.

3v. .3

.3 .mm. .mm .3 63.

.3m .3. .30.

3 .3m. ..m~ do.

3. do .3 .93 .mm

gm .93. .39. .mm .9.3

co. .3 .9... .3

N9...

8. 85.5.35

8. 8.3.8.3....

we. 3.3.2025.

8. 33:52::

8. 8.3.5.3....

8. 33:52....

22. 3...... 3.3.3.325

23 2.3.3

8. 8.3.5.35

33...»... .2233...

142

 

 

 

 

 

 

 

 

 

 

x x x 9.N-...
x x 2.6m 9..-... o9-N .9.9. .mm. m: 3.....32..5
x x x .o.255m.>.
x x x 9.N-...
x x x 9..-... mm .5 .9b. .5... 3. 32:82.5
NNMN x x 3.2552).
x x x 9.N-...
.2mm x x 9..-... no. .o9.~ .9.9. .mm. m: 35.32....
x x x 5.2559).
x x x 9.N-...
x x x 9..-... R .5... .2.. N: 3.....3225
oon x x .o.o.55m.>.
x 3.0... owém mm...
x x mhém 9..-... 9.5. .5. Z. 3.....32c5
x No.8 whom 3.25522
x x x 9.N-...
x x 2-... mm .N9. .5. 9.. 3.....32.5
x mmdm mmém 5.2552).
x x x 9.N-...
x .52 .w.mm 9..-... gm 5.. .5 .5... 5... 35.32:...
x x x .225sz
x oohm o9..m~ 2-...
x x x 9..-... 5 .3. .Na. 2.. 32.5.2.5
x wmdm 52mm 3.2552).
x x x hm-..
x x x 9..-... mo. .N9. .3.. 9b. 3.....32c5
x w..mm x .225sz
.:.E. .5. .:.E. .:.E.
Q...— ...~. HST—Wm: .5. mEmm-mEm—<E .325. N9... 5.2.0... 3.5.2.5....

 

143

 

wvwm

><><><><><><><><

mafia

X

><><><><><

modm
3.3

X

3.:

><><><><><

2.3

K

2.:

oofim

5N4.

5. TH
_o.oooDm2
RF
N. TH
3.8352
NNL.

N. TH
3.25sz
5N9
5 TH
5.25me

3. £2 6% sme—

mam 53

m2 .2: Wm. .vom .8—

K. .NE .9:

vm— 3:55ch

mfi BEEoEcD

mm. vuczcoEcD

a. voctcogcb

 

mN.mN

XXXX

many.
s Th
3.23sz

no. .91

ca 85202::

 

amfim
wmsm
amsm

um?
5 TH
215252

of .03 6:

a: 35202::

 

><><><><><><

_N.mm

XXXXXXXXXX

mmfim

x

BN4.
N. Th
5.8252

3.— dc— .hf .3:

2 _ Butane:

 

omflm

BNKN

RNA.
5 TH
_o-oooDm—2

mm: .2: .8.

E 35282:

 

XXXXXX

x
3.5 .5.
34

x
x
2.3
3:5

.5— m—meum: .5— mEmmumEm<2

X
X
X
X

x
x
x
x
2:5

smur
N. TH
5.8252

.325— .

~N~.o¢_.vm~.m¢~

a}:

E 352025

£23.: .2235..—

144

I I! u...

 

 

 

x x x R#

MEN x x 2-9 as .5 .2: .EN 92 85202::
x x x 3.2332
x x x. 2.9
x x x :9 a: in 53 R. BEEQES
x 3% ~58 3.8252

€25 .5— AEEV :55
mi E mgr—Wm: b. mzmm-mEm<2 .325 NE $.52 33:38..

 

145

Appendix G: Full List of Compounds Extracted from MDMA Exhibits T-17 and CJ-FSOS
using HS-SPME (unidentified impurity numbers do not correspond to Appendix F)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Provisional Identity m/z T47, CJ-FSPS
RT (mmLRT (mm)
Unidentified 1 137, 153, 82, 77, 91, 155 x 2.51
Unidentified 2 83, 55, 77, 67 2.79 2.87
Unidentified 3 133, 77, 78, 104, 151 4.36 x
Unidentified 4 117, 115, 77/91/118 6.65 x
Unidentified 5 94, 66, 65 x 6.65
Unidentified 6 118, 77, 91, 133 7.04 7.15
3,4-Methy1enedioxytoluene 135, 136, 77, 78 8.17 8.26
Unidentified 7 146, 105, 79, 77, 91 8.39 8.49
Unidentified 8 ' 91, 92, 65 8.54 8.62
Unidentified 9 105, 77, 51, 78, 106 9.32 9.41
Methamphetamine 58, 91, 150 9.64 9.73
Unidentified 10 146, 58, 105, 91 9.95 10.02
Unidentified 11 135, 108, 91, 82, 69, 58 x 10.42
Unidentified 12 147, 148, 91, 89, 77 10.58 10.66
Unidentified 13 107, 135, 77, 97 x 11.18
Safrole 162, 131, 103, 77, 104, 78 11.43 x
Unidentified 14 135, 166, 77, 136 12.15 12.24
Piperonal 149, 150, 121, 65 12.22 12.30
Ephedrine 58, 56, 77, 71, 91 12.91 13.00
Unidentified 15 135, 136, 77, 178, 179 13.03 13.12

 

 

146

 

T-l 7 CJ-F805

 

 

Provnsnonal Identity m/z RT (min) RT (min)
Unidentified 16 91, 105, 77, 133, 79 13.67 x
Unidentified 17 91 , 93, 77, 105, 79 14.20 14.30

 

3,4-Methylenedioxypheny1-2-propanone

14.72 4,7
(Mopzp) 135, 77. 178,79, 136 1 6

 

3,4-Methylenedioxyphenyl-Z-propanol

(MDP2_pmpanol) 135, 136, 77, 180, 78,79 15.12 15.20

 

3,4-Methy1enedioxymethamphetamine

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(MDMA) 58, 135, 136, 194 15.58 15.64

Unidentified 18 149, 65, 121, 194, 58 15.87 x
3’4‘Methylene‘zfiggi‘famphetami“e 72, 70, 77 (208) 16.47 16.50 ‘

Unidentified 19 208, 72, 105, 77 17.02 17.09
Unidentified 20 208, 207, 222, 72, 91 x 17.23
Unidentified 21 231, 208, 175, 133/246 x 17.62
Unidentified 22 195, 208, 167, 165 x 17.73
Unidentified 23 190, 147, 148, 208, 188 17.82 x
Unidentified 24 121, 208, 107, 163 x 17.82
Unidentified 25 135, 190, 107/208, 147, 148 x 17.92
Unidentified 26 107, 121, 149, 208 x 18.10
Unidentified 27 58, 100, 208, 135, 77 18.20 x
Unidentified 28 100, 58, 72, 135, 208 18.36 x
Unidentified 29 107, 149, 121, 208, 100 x 18.43
Unidentified 30 191, 192, 57 x 18.93
Unidentified 31 120, 91 , 214, 77, 93/121 x 19.23

Caffeine 194, 193 19.74 19.86

147

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Provisional Identity m/z T47, CJ-FSPS
RT (mm) RT (mInL
Unidentified 32 58, 165, 152, 167 20.26 x
Lidocaine 86, 58, 72, 91 20.38 20.47
Nmejg‘yyllggxgz 5132:3311?!- 162,58, 77, 135 20.77 20.85
Unidentified 33 147, 277, 189, 292 20.84 x
Unidentified 34 146, 147, 105, 77 21.17 x
Unidentified 35 227, 143, 242, 228 x 22.23
Unidentified 36 148, 168, 176, 73, 91 22.27 x
Unidentified 37 217, 232, 215, 202, 231 x 22.65
Unidentified 38 99, 117, 55, 100 23.22 x
Unidentified 39 148, 190, 149 23.39 x
Unidentified 40 91, 148, 281 24.10 x
Unidentified 41 149, 91, 119, 284 25.71 25.80
Unidentified 42 191, 150, 164, 192 x 26.27
Unidentified 43 149, 167, 160, 176, 281 27.32 x
Unidentified 44 135, 192, 77 x 27.97

 

148