See discussions, stats, andauthor profiles for this publication at: https://www.researchgate.

net/publication/7214941

A 'cold synthesis' of heroin and implications in heroin signature analysis Utility of

trifluoroacetic/acetic anhydride in the acetylation of morphine
Article in Forensic Science International · January 2007
DOI:10.1016/j.forsciint.2006.02.009·Source:PubMed

CITATIONS READS
19 8,229
3 authors:

Luke R. Odell Jana Skope c

Uppsala University AgriFor Scientifi c
123 PUBLICATIONS 2,288 CITATIONS 10 PUBLICATIONS 197 CITATIONS

SEE PROFILE SEE PROFILE

Adam Mccluske y
The University of Newcastle, Australi a
340 PUBLICATIONS 7,957 CITATIONS

SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Arylhydrocarbon receptor pathway View project
A STATISTICAL EVALUATION OF THE CHEMICAL COMPOSITION OF HEROIN SAMPLES WITH THE VIEW TO DISCRIMINATING BETWEEN THEIR POSSIBLE SOURCES A. EKANGAKI, PHD J. SKOPEC, PHD, Z.
SKOPEC, PHD, AND R. J. WELLS, F% View project
Allcontentfollowingthispagewasuploadedby JanaSkopec on20July2018.
Theuser has requestedenhancementofthedownloadedfile.
 Forensic Science International 164 (2006) 221–229
www.elsevier.com/locate/forsciint

A ‘cold synthesis’ of heroin and implications in heroin signature analysis Utility

of trifluoroacetic/acetic anhydride in the acetylation of morphine
Luke R. Odell a, Jana Skopec b,1, Adam McCluskey a,*
a

Chemistry Building, School of Environmental and Life Sciences, The University of Newcastle, University Drive, Callaghan, NSW 2308, Australia
b

AGAL, 1 Suakin Street, Pymble, NSW 2073, Australia

Received 5 October 2005; received in revised form 25 January 2006; accepted 2 February 2006
Available online 23 March 2006

Abstract
Treatment of morphine, at room temperature, with a mixture of trifluoroacetic anhydride (TFAA) and acetic acid (20–30 min) affords good yields of heroin. GC–
MS and HPLC examination shows that heroin produced by this route to be extremely clean, but the product contains slightly less heroin than observed via the more
traditional acetic anhydride (AA) route (76.1% versus 83.55%); and greater quantities of 3-MAM and 6MAM (6.9% versus 0.75% and 7.13% versus 0.63%). The
concentration ratios of the major alkaloid impurities were found to be both production method (TFAA and AA) as well as morphine extraction methodology
dependant. Data contained herein describe the impact of this new production method on current intelligence efforts, largely by-passing existing heroin signature
programs and the UNDCP’s efforts to restrict access to key synthetic precursors. Given the methodology dependency we find that examination of the major alkaloid
ratios is unsuitable for the development of a new heroin signature program.
Further examination of the TFAA methodology allowed the identification of TFAA specific marker compounds, namely bis-trifluoroacetylmorphine (30), 3-
trifluoroacetyl-6-acetylmorphine (31), 3-acetyl-6-trifluoroacetylmorphine (32) and trifluoroacetylcodeine (33). However, the hydrolytic lability of trifluroacetyl esters
requires careful treatment of suspect samples, thus we propose a modification to existing HSP’s in instances were the 6-MAM/WM ratio falls within the average
minimum and maximum values of 6.17 and 17.32. # 2006 Elsevier Ireland Ltd. All rights reserved.
Keywords: Heroin; Signaure program; Acetylation of morphine; Cold synthesis

1. Introduction programs (HSP). In a

Globally the heroinHSP seized samples
seized rises by about 5%are analysed by a
per annum [1]. Withinnumber of different
Australia these figuresanalytical procedures
are even more disturbingwith the aim of
as a result of recordproviding crucial drug
numbers of seizures,trafficking and
222 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

with the mid 1990sdistribution

average of 250 kg perintelligence. HSP
year increasing totypically analyses and
between 450 and 730 kgevaluates the major
per year since 2000synthetic impurities
[2,3]. Indeed, despiteand adulterants. These
education efforts withinapproaches have been
Australia a recentextensively examined
Australian Nationaland reported
Drug Household surveypreviously [1,5–37].
indicates that domesticWorldwide, the
demand is on theimpurity profiling of
increase and that theillicit drugs is being
number of Australianincreasingly utilised as
heroin users has morean intelligence-
than doubled since 1988gathering tool, to
[4]. The semi-syntheticsupport and
nature of heroin, beingcomplement the work
derived from morphineof law enforcement
(Fig. 1) has facilitatedagencies allowing an
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 223

the development andobjective, comparative

implementation of aanalysis of the level of
series of heroincommonality between
signature samples [9–13,32,33].
The chemical
information provided
from an impurity
profile can be used to
provide tactical
intelligence, which
involves establishing
links between different
samples [14,15] and
strategic intelligence,
which refers to the
identification of the
source of a sample
[28–
30], the tracing of
distribution routes and
* Corresponding author. Tel.: +61 249216486; fax: +61 249215472.
E-mail address: Adam.McCluskey@newcastle.edu.au (A. McCluskey).
1 Present address: AgriFor Scientific Pty Ltd, 46 Cook Road, Killara, NSW, Australia.
224 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

identifying new
production processes
[31].
To date HSP analysis
rapidly allows South-
West Asian
(SWA) heroin to be
identified by
significant quantities
of both
0379-0738/$ – see front matter # 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.forsciint.2006.02.009

alkaloids; Mexican
Fig. 1. Acetylation of morphine to yield diacetylmorphine (DAM, heroin).
heroin by its unique
appearance; and South
American heroin by its
high purity and low
Fig. 2. Route specific markers detected in the illicit production of heroin.
acetyl codeine (AC)
papaverine (PAP) and
content. Further,
noscapine (NOS);
quantitative major
South-East Asian
component data have
(SEA) heroin by an
been used successfully
absence of these
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 225

by a number of authors agent available in the

to determine the source clandestine
of heroin manufacture of heroin.
[9,10,12,17,19]. There are currently
Whilst a HSP two other acetylating
typically concentrates agents described in the
on determination of literature: acetyl
geographic origin, it is chloride, which affords
also possible to elicit 1-chloroheroin (3) as a
information relating to route specific marker;
the routes of synthesis and ethylene diacetate,
used in the actual which affords 3-[1-(1-
manufacture of seized carboxymethoxyethyl)
DAM. Standard HSP ]-6-acetylmorphine (4)
analysis is based on as a route specific
acetic anhydride (AA) marker (Fig. 2)
as the principle [5,38,39].
acetylating agent. Over the past 30
However, AA is not years considerable
the sole acetylating work has been done
226 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

developing parameters synthesis of heroin

(Fig. 3, marker [40].
compounds 5–10) for 2. General techniques
the profiling of heroin. and instrumentation
During this period All solvents used
only three acetylating herein were of HPLC
agents have been grade.
reported AA, acetyl HPLC major impurity
chloride and ethylene determination
diacetate, with the The analyses were
former being performed on a Perkin-
employed almost Elmer ISS 200 Series
exclusively. The recent HPLC equipped with
discovery of traces of an LC-235C diode
trifluoroacetic acid in a array detector
seized sample measuring at 240 and
promoted an 225 nm. The
investigation into the chromatographic
use of this and related separation was
reagents in the achieved on an Alltec
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 227

C18 column using a six- comprised 1N sodium

step methanol and hydroxide (32 mL),
phosphate buffer phosphoric acid (11
gradient (see Section mL) and hexylamine
3.1) at a flow rate of (3.5 mL) in millipore
0.76 mL/min. The water (1000 mL).
buffer (pH 2) Pump parameters
228 L.R. Odell et al./Forensic Science International 164 (2006) 221–229
Step Time (min) Methanol (%) Buffer
(%)
1 10 5 95
2 20 30 70
3 6 30 70
4 10 80 20
5 4 80 20
6 10 5 95

Fig. 3. Chemical structures of the major alkaloids found in, and as a result of, the illicit synthesis of heroin.
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 229
Table 1
Comparison of chemical yields and major component analyses for the synthesis of heroin via the TFAA or the AA route
Method Target compounds

M C 3-MAM 6-MAM AC DAM NOS PAP Yield

(%)
TFAAa 1.02 1.13 0.09 0.02 6.9 1.63 7.13 1.56 0.83 0.11 76.1 3.23 <0.01 <0.01 58 5
AAb 0.16 0.29 <0.01 0.75 0.31 0.63 0.28 1.12 0.15 83.55 5.93 <0.1 <0.1 64 12
a
Non-controlled precursor; reaction time 20–30 min; typically 3 equivalents TFAA; room temperature.
b Listed on the International Narcotics Control Board’s ‘RED LIST’: see http://www.incb.org/incb/en/red_list.html; reaction time 2 h; typically 10 equivalents

gas chromatograph
AA; requires heating 80–90 8C.

Each sample under

investigation was equipped with a 5973
weighed (10 mg) into a quadrupole
10 mL volumetric massselective detector
flask. Propiophenone and an Agilent 7683
(1 mL) was added as series autoinjector. The
an internal standard chromatographic
and the mixture diluted separation was
to 10 mL with the achieved on a HP-
addition of the 5MS (5% phenyl
injection solvent (5% methyl siloxane)
methanol, 95% buffer). capillary column
GC–MS analysis (length 30 m, i.d. 0.25
The analyses were mm, film thickness
performed on a 0.25 mm) with helium
Hewlett-Packard 6890 carrier gas at a flow of
1.2 mL/min (pressure
230 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

88.9 kPa). The Each sample under

injection temperature investigation was
was maintained at 260 weighed (50 mg) into a
8C, with 2 mL splitless glass centrifuge tube.
injection. The column Sulphuric acid (1N, 4
was heated at 100 8C mL) and extraction
for 1 min then mixture (60:40 ether–
increased at a rate 6 dichloromethane; 4
8C/min to 240 8C then mL) were added to the
2 8C/min to 280 8C sample. The solution
followed by 6 8C/min was vortexed (15 min)
to 320 8C. and then centrifuged
Acidic and neutral (15 min, 1000 rpm).
trace impurity analysis An aliquot (3 mL) of
All reaction vials, vial the organic phase was
inserts, pipettes and pipetted into a reaction
centrifuge tubes were vial. The solution was
silanised, to replace dried under a stream of
active hydrogen nitrogen at 70 8C. The
groups, prior to use. sample was then
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 231

treated with a transferred to a GC

derivatisation solution vial for analysis.
(250 mL). The GC–MS trace
derivatising solution component analysis
comprised BTFSA (1 was used to determine
mL), TMS (1%) and the relative amounts of
dichloromethane (1 the target compounds
mL). Next in each sample.
benzopinacolone (50 Major component
mL) was added as the analysis
internal standard, at a To 1 mg of each
concentration of sample was added
benzopinacolone/dichl dichloromethane (250
oromethane of 2.5 mL) followed by 50
mg/mL. The reaction mL of internal
vial was capped and standard (prepared as
heated at 70 8C for 30 described above).
min. The solution was Results and
allowed to cool and discussion
232 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

There is literature anhydride followed by

precedent for the use addition of morphine
of trifluoroacetic proved to be an
anhydride as an ester extremely clean,
promoter, with two efficient and elegant
methods commonly route for the
described, differing production of high
only in the order of quality DAM.
reagent addition The TFAA
[41,42]. Dissolution of methodology provides
morphine in TFAA for a four to six-fold
followed by the decrease in reaction
addition of acetic acid time in the absence of
gives rise to very low both heat and a large
yields of DAM, but reagent excess (Table
significant quantities 1). Interestingly,
trifluoroacetyl esters; neither acetic acid nor
conversely dissolution TFAA are restricted
of acetic acid in TFAA precursors, and can
forming the mixed thus be readily
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 233

obtained from significant. Not only

chemical suppliers does it offer the
[43]. From a drug clandestine
laboratories good
Table 2
Alkaloid ratios for the AA and TFAA reactions using the Tas, SWA and SEA
procedures
Methoda Ratiob

yields with far greater

throughput, but also
C/WM NOS/WM PAP/WM 6-
MAM/WM
Tas
AAc <0.04

the potential to
1.48 0.22 0.01 0.02 0.80 0.36
TFAAd 1.14 0.10 <0.01 0.01 0.02 9.86 1.86
SWA

circumvents the
AAe 10.69 0.46 33.1 1.25 4.61 0.13 1.11 0.39
TFAAf 10.92 1.58 24.45 8.63 4.7 1.04 12.35
4.27

current UNDCP
SEAAAe
8.94 0.3 0.21 0.18 0.21 0.23 0.28 0.03
TFAA Basee 9.07 0.09 0.25 0.22 0.06 0.2 9.54 2.7

methods of tracking
TFAA HClg 9.78 0.07 2.12 0.34 0.26 0.1 9.03 1.31
Average
AAh 5.56 4.35 6.68 13.93 0.99 1.91 0.71 0.42
TFAAi

a
heroin production [43].
6.5 4.59 6.58 11.34 1.24 2.11
2.79
10.27

From Table 1, it is
Tas: Tasmanian procedure; SWA: South-West Asian procedure; SEA:
South-East Asian procedure.
b
C = total codeine (C + AC); WM = (%DAM 285.34/

clear that the TFAA

369.42) + (%MAM 285.34/327.38)(%MAM).
n = 5.
n = 8.
n = 3.
n = 5.
n = 4.
n = 11.
reaction gives rise to
anomalous data
n = 20.

enforcement
perspective the outwith that
discovery of such a anticipated in current
simple, rapid and HSP’s. As can be seen
facile method for from Table 1 the
DAM synthesis is product produced via
234 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

the TFAA route that trifluoroacetyl

contains slightly less esters are notoriously
DAM (76.1%) than unstable, readily
that observed via the hydrolysed under
more traditional AA conditions required for
route (83.55%); and analysis (see Section
greater quantities of 3- 6), meaning that their
MAM (6.9% versus detection in
0.75%) and 6-MAM clandestine laboratory
(7.13% versus 0.63%). samples is difficult
Furthermore all TFAA [41,42], this is the
samples contained most probable
traces of morphine and explanation for the
codeine, whereas increased
morphine was only concentrations of 3-
detected in only two MAM, 6-MAM,
and codeine in none of morphine and codeine
the AA samples (data detected.
not shown). It is Current HSP
noteworthy to mention methodologies have
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 235

evolved to deal with 2) were compared by

DAM synthesised via producing a number of
morphine acetylation replicate samples from
by AA at elevated three different
temperature, we thus morphine sources:
felt it crucial that this Tasmanian poppy and
new methodology be typical South-West and
examined and key SouthEast Asian
constituents analysed morphine.
in an attempt to gather
data as to how TFAA
might impinge on
existing HSP
protocols. Accordingly
we examined the
amounts major illicit
heroin components
formed using the two
methods (TFAA and
AA) (Fig. 3 and Table
236 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

Table 2 highlights and codeine ratios

significant differences between Tas, SWA and
in the alkaloid ratios SEA series, can be
observed as a function explained by
of both the reaction differences in the
method (TFAA or AA) morphine extraction
and clandestine processes. The SWA
procedure employed. procedure involves
Variation in the only a crude morphine
noscapine, papaverine

Fig. 4. Trace level impurities targeted by the Australian National Heroin Signature Program.
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 237

extraction and
consequently had the
238 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

highest NOS/WM this series. The

(33.1 and 24.45 versus morphine used in Tas
SEA 0.21, 0.25, 2.12 series had been refined
and TAS <0.01) and using an unknown
PAP/WM (4.61 and chemical process,
4.7 versus SEA 0.21, which removed the
0.06, 0.26 and TAS majority of non-
0.01) ratios. In morphine alkaloids.
contrast, the SEA Our data clearly
employs steps to illustrate that the
remove the majority of amounts of these
noscapine and nonmorphine alkaloids
papaverine, which are related to the
resulted in the smaller morphine extraction
NOS/WM (0.21, 0.25, process and not the
2.12 versus acetylating procedure.
33.1 and 24.45) and Consequently these
PAP/WM (0.21, 0.06, ratios cannot be used
0.26 versus 4.61 and as TFAA specific
4.7) ratios observed in profiling parameters.
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 239

In contrast, the 6- reagent and production

MAM/WM ratio is process independent
significantly higher for making it an ideal
all TFAA produced profiling parameter for
samples (e.g. 9.86 the identification of the
versus 0.8 for Tas use of TFAA in the
series) and remains illicit synthesis of
constant irrespective of heroin.
the morphine sample Given that the major
or clandestine alkaloids are not
procedure employed suitable for
(TFAA reaction 9.86, determination of
12.35 and 9.54, 9.03 TFAA versus AA
for the Tas, SWA, SEA synthesis route we
Base and HCl next turned our
procedures, attention to analysis of
respectively). This the trace impurities
indicates that the 6- (Fig. 4 and Table 3).
MAM/WM ratio is Data presented in
both acetylating
240 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

Table 3 clearly by the N-acetylation

illustrate followed by
demethylation of
Table 3

significant differences
in the quantities of morphine and codeine
morphine related N-oxide, which are
byproducts; 11–13 and transient intermediates
15 are produced when formed during the
AA is employed as the acetylation process
acetylating agent, e.g. [22]. Not surprisingly
13: 38.74, 51.13 and our data confirm the
100 for the Tas, SWA TFAA routes as milder,
and SEA Base resulting in lower
methods, respectively, initial concentrations
as compared with the of these intermediate
TFAA method, e.g. 13: species and hence
7.19, 9.92 and 50.33 lower amounts of 11–
for the Tas, SWA and 13 and 15.
SEA methods,
respectively. These
compounds are formed
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 241

Further examination during the two

of Table 3 reveals a reactions. All three
contrast between the series show that the
relative proportions of AA procedure resulted
thebaine related in the formation of
products formed substantially greater
Analysis of major alkaloids and impurities as a result of the AA and TFAA methods of DAM synthesis
Target Average response and procedure useda
compounds

Tasmanian SWA SEA

AAd TFAAd
AAb TFAAc AAd TFAAd
Morphine related compounds 0.37
11 0.39 0.04 0.07 0.04 17.48 1.97 2.47 0.05 0.58 0.14 0.12
12 14.91 0.98 4.46 1.5 20.58 2.01 10.95 2.54 35.96 6.46 20.25
6.33
13 38.74 1.22 7.19 3.25 51.13 0.85 9.92 2.0 100 5.30 50.33
18.1
14 <0.05 <0.05 2.8 0.54 0.83 0.67 0.04 0.02 0.09
0.04
15 2.56 0.37 0.3 0.16 12.85 1.16 1.82 0.11 18.17 4.70 6.52
2.13
16 <0.05 <0.05 4.19 0.02 0.13 0.1 0.16 0.05 0.17
0.07
Thebaine related compounds 0.4
17 2.13 0.13 0.16 0.06 61.36 4.14 1.76 0.69 0.2 0.06 0.05
18 5.32 0.01 2.1 1.28 100 18.32 20.75 2.45 2.41 0.08 1.45
0.33
19 6.16 0.55 0.47 0.34 97.79 10.56 4.75 0.35 0.84 0.09 0.02
0.01
20 0.22 0.03 0.09 0.03 4.87 0.13 0.34 0.26 0.24 0.07 0.2
0.09
21 0.12 0.02 <0.05 9.11 1.69 0.92 0.31 0.03 0.08 0.02
0.02
22 <0.05 <0.05 3.11 1.25 0.73 0.52 0.01 0.01 <0.05
23 <0.05 <0.05 6.22 0.85 <0.05 <0.05 <0.05
Noscapine and N-norlaudanosine related compounds
24 <0.05 <0.05 0.26 0.05 <0.05 <0.05 <0.05
25 <0.05 <0.05 0.18 0.01 <0.05 <0.05 <0.05
26 <0.05 <0.05 57.21 1.33 3.52 0.02 1.5 0.23 <0.05
27 <0.05 <0.05 1.57 0.01 0.14 0.03 <0.01 <0.05
28 <0.05 <0.05 2.28 0.10 0.06 0.05 <0.01 <0.05
29 0.22 0.15 <0.05 8.73 3.8 5.93 0.38 100 9.9 7.74
0.25
a
Response was calculated against an internal standard (i.e. compound response/internal standard 100), data are then normalised against largest response.
n = 3.
n = 6.
n = 2.
242 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

amounts of impurities methods) are a result

(e.g. 19: AA method: of thebaine being not
6.16, 97.79 and 0.84; reacting under the
TFAA method 0.47, TFAA conditions. The
4.75 and 0.02 for the application of heat to
Tas, SWA and SEA thebaine in the
methods, respectively), presence of TFAA
with the exceptions of results in the rapid
17 in the SEA series decomposition of
(AA: 0.2; TFAA: 0.4)
and impurities 20–23
which were formed in
essentially equal
amounts. Control
experiments, treating
thebaine according to
the TFAA protocol
showed that the
observed differences
(AA versus TFAA
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 243

thebaine. Thus, respectively) (Table 3).

although thebaine The formation of 24
decomposes when and 25 proceeds
heated in TFAA, through the addition of
formation of 17–23 acetic acid across the
(the major AA double bond in 27 and
decomposition 28 [22]. The absence
products) is not the of these compounds in
preferred pathway for samples produced
this reaction [21]. using TFAA can be
Significantly greater explained by the
amounts of noscapine absence of acetic acid
and N-norlaudanosine and the smaller
related compounds quantities of 27 and 28
were produced when present in the mixed
the AA anhydride system. The
methodwasemployed(e formation of 26
.g.26:AA:57.21and1.5; proceeds through the
TFAA:3.52 and <0.05 precursor, noscapine
for SWA and SEA, N-oxide [22]. The
244 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

increased abundance prevalence for attack at

of this compound in the trifluoroacetyl
the AA samples is carbonyl, of the mixed
believed to be a result anhydride formed in
of smaller quantities of situ, increasing with
noscapine N-oxide the reactivity of the
being formed during nucleophile [41].
the TFAA synthesis. Hence the TFAA
Subsequent cleavage procedure favours
of the C1–N bond trifluoroacetylation,
affords 27 and 28. decreasing the amount
The formation of 29 of 29 observed.
involves the Our data clearly
acetylation of the establish that the
secondary amine amounts of these trace
group in N- impurities were
norlaudanosine [22], significantly different
which is a function of between the two
nucleophile strength, reactions. However,
with increased when comparisons are
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 245

made between the Tas, With the currently

SWA and SEA series identified trace organic
these differences were impurities unable to
shown to be highly distinguish between
dependent upon the the two reactions, an
morphine extraction attempt was made to
procedure (e.g. 19: identify TFAA reaction
6.16 versus 0.47, 97.79 specific marker
versus 4.75 and 0.84 compounds. During
versus 0.02 for Tas, this investigation
SWA and SEA, morphine, thebaine,
respectively). oripavine, codeine,
Consequently based on noscapine and
this data set, the trace papaverine were
organic component of treated with AA and
a seized sample cannot TFAA, analysed by
be used as a TFAA GC–MS and their
specific profiling respective products
parameter. compared (Fig. 5).
246 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

The reaction of min (31 m/z = 423)

noscapine, papaverine, and 26 min (32 m/z =
thebaine and 423) in morphine and
oripavinewith TFAA 24 min (33 m/z = 395)
failed to produce any in codeine,
detectable products, respectively. The
which could be used as molecular weights of
TFAA markers. these compounds
However, when suggest the addition of
morphine and codeine two CF3CO groups
were treated with (m/z = 194), one
TFAA and AA CF3CO group and one
significant differences CH3CO group (m/z =
in the reaction 140) to morphine (m/z
products were = 285) and one CF3CO
observed. Fig. 6 shows (m/z = 97) to codeine
the appearance of (m/z = 299),
peaks in the TFAA respectively, assuming
samples at 23.5 min the M-1 is detected in
(30 m/z = 477), 25 the GC–MS. We had
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 247

anticipated that re- the current HSP

examination of our methodology involves
earlier preparations of a 1N H2SO4 extraction
heroin via the TFAA and derivitisation step
route that evidence for prior to analysis, given
the formation of 30–33 the extremely labile
would be forthcoming. nature of
This was not the case. trifluoroacetyl esters it
However, was deemed probable
that the current
methodology
precluded the detection
of these TFAA specific
markers. Consequently
we re-examined the
TFAA heroin synthesis
in the absence of the
Fig. 5. (a) GC–MS analysis of a crude morphine sample after treatment with TFAA. Rt
= 23.5 min = bis-trifluoroacetylmorphine (30); Rt = 26 min =
monotrifluoroacetylmorphine; Rt = 27.7 min = 3-MAM (7); Rt = 29 min = DAM (2); and
extraction and
derivatisation steps, in
Rt = 31 min = IS (benzopinacolone); (b) the same sample after typical pretreatment for GC–
MS analysis; Rt 27.2 min = 6-MAM (8); 27.7 min = 3-MAM (7); Rt = 29 min = DAM
(2); and Rt = 31 min = IS (benzopinacolone).

these instances both 32

248 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

and 33 were detected = 310) and [M-CH2CO

in all SEA and SWA + CF3COOH] (m/z =
samples produced 268) ions.
using TFAA Additionally, the [M-
(Fig. 6). CH2CHO]+ (m/z =
Structural assignment 381) ion, which is
of 30 and 31 was characteristic of an
based on MS aromatic acetate
fragmentation pattern allowed the
analysis. MS analysis assignment of the
shows 30 to have [M]+ acetate group to the 3-
of 423 which suggests position [44]. The
the addition of one spectrum of 30 also
acetyl and one showed significant
trifluoroacetyl ester to peaks at m/z 215 and
morphine. This is 174, which are
supported by a characteristic of
significant [M- morphine skeleton
CH2CO]+ (m/z = 381), fragmentation [45,46].
[M-CF3COOH]+ (m/ z Similarly, 31 was
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 249

determined to have
[M]+ of 395, which is
consistent with the
addition of a
trifluoroacetyl group to
codeine. This was
confirmed by the [M-
CF3COOH]+ (m/ z =
282) base ion. The
spectrum of 31 also
showed characteristic
codeine skeleton
fragmentation peaks at
m/z 229 and 188
supporting the
assignment of 31 as
trifluoroacetylcodeine
[45,46].
250 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

Fig. 6. (a) Chemical structures of TFAA specific markers identified in this work; (b) GC/MS trace of the crude reaction mixture arising from the reaction of morphine and AA; (c) GC/MS
trace of the crude reaction mixturearising from the reaction of morphineand TFAA; (d) GC/MS trace of the crude reaction of codeine and AA; (e) GC/MS trace of the crude mixture arising
from the reaction of codeine and TFAA. (f) GC–MS trace typical of DAM synthesis using TFAA; expanded section shows the presence of 3-acetyl-6-trifluoroacetylmorphine (32) and
trifluoroacetylcodeine (33).

Conclusions quick and currently

The use of TFAA undetectable synthesis
provides a simple, of heroin. To enable
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 251

detection of this sample falls within the

method a number average minimum and
ofprofiling parameters maximum values
have been developed. (6.17–17.32) then a
In Australia, the presumption would be
routine profiling of raised that the sample
heroin using major was made using the
alkaloid analysis TFAA route. Any
involves conversion of samples, which raise
raw alkaloid this presumption,
concentration data into would then be
C/WM, NOS/WM, subjected to an
PAP/WM and additional test to
6MAM/WM ratios confirm the origin of
and its origin assigned the product. A DAM
in accord with the sample with a large 6-
UNDCP data [38]. MAM/WM ratio may
Herein we suggest that also be explained by
if the 6-MAM/WM incomplete acylation
ratio of a seized or poor sample storage
252 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

[25,47,48]. extracted (in a region

Additionally the specific manner) from
presence of 32 or 33 samples seized by the
may be due to the Australian Federal
presence of an Police using the
alternate trifluorinated methods described
reagent (e.g. below. Thebaine,
CF3COCl). oripavine, morphine,
Experimental codeine and
All solvents were papaverine reference
HPLC grade or bulk standards were
solvents re-distilled obtained from the
from glass before use. National Analytical
Morphine used in the Reference Laboratory
Tasmanian series was (NARL), Pymble,
received from Australia.
Tasmanian Alkaloids. Synthetic methods
In all other instances Morphine extraction
morphine utilised in Opium (40 g) was
synthesis was dissolved in boiling
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 253

water (30 mL), to leave behind morphine

which was added base (2.61 g, 6.7%).
Ca(OH)2 (10 g) and the Morphine HCl
heating removed. The Morphine base (2.5 g)
mixture was allowed was dissolved in warm
to cool and stand HCl (1 M, 25 mL) and
overnight. The solution 1 g of charcoal added.
was filtered, the The solution was
residue dissolved in warmed for 10 min
boiling water (10 mL) and then filtered. The
and re-filtered. The filtrate was left to
combined filtrates stand overnight and
were re-heated until filtered to leave behind
the solution steamed, morphine
at this point the pH of hydrochloride (1.14 g,
the solution was 41%).
adjusted to 8–9 with Heroin synthesis
NH4+Cl and left to Trifluoroacetic
stand overnight. The anhydride synthesis.
solution was filtered to Method 1
254 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

Glacial acetic acid carbonate. The

(84 mL, 1.75 mmol) resulting aqueous
was added slowly to solution was extracted
trifluoroacetic with chloroform (3 10
anhydride (311 mL, ml), dried over MgSO4
2.1 mmol) in an ice and evaporated to
bath and stirred for 10 leave heroin base
min. Morphine (200 (157.6 mg, 78%).
mg, 0.7 mmol) was Method 2
added slowly and the Trifluoroacetic
mixture stirred for 20 anhydride (311 mL,
min. The solution was 2.1 mmol) was added
then allowed to warm slowly to morphine
to room temperature (200 mg, 0.7 mmol) in
and diluted with cold an ice bath and stirred
water (1 mL). The for 10 min. Glacial
solution was put back acetic acid (84 mL,
in an ice bath and 1.75 mmol) was added
made basic (pH 8–10) slowly and the mixture
with cold 10% sodium stirred for 20 min. The
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 255

solution was then temperature. The

allowed to come to solution was diluted
room temperature and with cold water (1
evaporated under a mL). The solution was
stream of nitrogen. cooled on ice and
The residue dissolved made basic (pH 8– 10)
in CH2Cl2 and with cold 10% sodium
analysed by GC–MS carbonate. The
(see Section 2). resulting aqueous
Acetic anhydride solution was extracted
synthesis with chloroform (3 10
Acetic anhydride mL), dried over
(600 mL, 5.8 mmol) MgSO4 and evaporated
was added slowly to to leave
morphine (200 mg, 0.7 diacetylmorphine base.
mmol). The resulting South-East Asian
solution was heated procedure
and stirred at 85 8C for Diacetylmorphine
2.25 h and allowed to base (from morphine
cool to room hydrochloride) (60 mg,
256 L.R. Odell et al./Forensic Science International 164 (2006) 221–229

0.16 mmol) was National Analytical

dissolved in acetone (6 Reference Laboratory
mL), and concentrated of the alkaloid
HCl (18 mL) was reference samples used
added and the solution in this work. We also
evaporated under a thank the Australian
stream of nitrogen to Federal Police for
leave behind allowing access to
diacetylmorphine seized heroin samples,
hydrochloride (57 mg, without which this
86%). work would not have
Acknowledgements been possible.
L.O. is grateful for References
financial support from
United Nations Drug Control Program, World Illicit Drugs Report 2000– 01, New
York, 2001.
Australian Bureau of Criminal Intelligence, Australian Illicit Drug Report 1998–99,
2000.

the Australian Forensic

Australian Bureau of Criminal Intelligence, Australian Illicit Drug Report 2000–01,
2002.
Australian Institute of Criminology, 2001. Trends in the Prevalence of

Drug Laboratory
Drugs, Australia 1988–1998, National Drug Strategy Household Survey, Unit Record
Files, 1988–1998, http://www.aic.gov.au/research/drugs/ stats/prevalence/trendsdrugs88-
98.html (3-25-2002).
D.A. Cooper, in: M. Klein, F. Sapienza, H. McClain, I. Khan (Eds.), Clandestinely

(AGAL, Pymble,
Produced Drugs and Precursors, United States Department of Justice Drug Enforcement
Administration, Washington, DC, 1989.
W.R. Heumann, Bull. Narc. 9 (1957) 34.

NSW, Australia). We
M. Chiarotti, N. Fucci, C. Furnari, Comparative analysis of illicit heroin samples,
Forensic Sci. Int. 50 (1991) 47–56.
H. Neumann, Comments on the routine profiling of illicit heroin samples, Forensic Sci.
Int. 44 (1990) 85–87.

are also grateful for the

H. Neumann, Comparison of heroin by capillary gas chromatography in Germany,
Forensic Sci. Int. 69 (1994) 7–16.
P.J. O’Neil, P.B. Baker, T.A. Gough, Illicitly imported heroin products: some physical

donation from the

and chemical features indicative of their origin. Part II, J. Forensic Sci. 30 (1985) 681–691.
B.A. Perillo, R.F.X. Klein, E.S. Franzosa, Recent advances by the US DEA in drug
signature and analysis, Forensic Sci. Int. 69 (1994) 1–6.
P.J. O’Neil, J.E. Pitts, Illicitly imported heroin products (1984–1989): some physical
and chemical features indicative of their origin, J. Pharm. Pharmacol. 44 (1992) 1.
 L.R. Odell et al./Forensic Science International 164 (2006) 221–229 257
H. Huizer, Analytical studies on illicit heroin II. Comparison of samples, J. Forensic Sci. D. Wheeler, T.H. Kinstle, K.L. Rinehart, Mass spectral studies of alkaloids related to
28 (1983) 40–48. morphine, J. Am. Chem. Soc. 89 (1967) 4494.
A. Allen, D. Cooper, J. Moore, D16,17-Dehydroheroinium chloride: synthesis and H. Huizer, Analytical studies on illicit heroin I. The occurrence of
characterization of a novel impurity detected in illicit heroin, J. Org. Chem. 48 (1983) 3951. O3monoacetylmorphine, J. Forensic Sci. 28 (1983) 32.
K. Narayanaswami, H.C. Golani, R.D. Dua, Assay of major and minor constituents of A.R.L. Wijesekera, D. Abeysinghe, K.C. Pathirana, Studies on the degradation of
opium samples and studies of their origin, Forensic Sci. Int. 14 (1979) 181. heroin, Forensic Sci. Int. 67 (1994) 147.
P.A. Hays, C.P. Goddard, M. Japp, I.J. Humphreys, The characterization of illicit heroin
by the analysis of impurities using high-performance liquid chromatography, J. Forensic
Sci. 24 (1984) 561.
K. Narayanaswami, Parameters for determining the origin of illicit heroin samples,
Bull. Narc. 37 (1985) 49.
C. Barnfield, D.L. Byrom, A.V. Kemmenoe, S. Burns, The routine profiling of forensic
heroin samples, Forensic Sci. Int. 39 (1988) 107–117.
A. Johnson, L.A. King, Heroin profiling: predicting the country of origin of seized
heroin, Forensic Sci. Int. 98 (1998) 47.
R.B. Myors, S.V. Skopec, P.T. Crisp, R.J. Wells, Investigation of heroin profiling using
trace organic impurities, Analyst 126 (2001) 679–689.
A. Allen, D. Cooper, J. Moore, C. Teer, Thebaine rearrangements: nonclassical ring
migrations, J. Org. Chem. 49 (1984) 3462.
A. Allen, D. Cooper, J. Moore, M. Gloger, H. Neumann, Illicit heroin manufacturing by
products: capillary gas chromatographic determination and structural elucidation of
narcotine- and norlaudanosine-related compounds, Anal. Chem. 56 (1984) 2940.
F. Bescier, H. Chaudron-Thozet, Chemical profiling of illicit heroin samples, Forensic
Sci. Rev. 11 (1999) 105.
A. Johnson, L.A. King, Heroin profiling: predicting the country of origin of seized
heroin, Forensic Sci. Int. 95 (1998) 47.
J.A. Sibley, Formation of O-6-acetylmorphine in the ‘Homebake’ preparation of heroin,
Forensic Sci. Int. 44 (1996) 110.
R.B. Myors, P.T. Crisp, S.Z. Skopec, R.J. Wells, Investigation of heroin profiling using
trace organic impurities, Analyst 126 (2001) 679.
M. Desage, R. Guilluy, J.L. Brazier, H. Chaudron, J. Girard, H. Cherpin, J. Jumeau, Gas
chromatography with mass spectrometry or isotope-ratio mass spectrometry in studying the
geographical origin of heroin, Anal. Chim. Acta 247 (1991) 249.
F. Becacier, H. Chaudron-Thozet, M. Rousseau-Tsangaris, J. Girard, A. Lamotte,
Comparative chemical analysis of drug samples: general approach and application to
heroin, Forensic Sci. Int. 85 (1997) 113.
S. Klemenc, In common batch searching of illicit heroin samples— evaluation of data
by chemometrics methods, Forensic Sci. Int. 115 (2001) 43.
Viewpublicationstats

K. Janhunen, M.D. Cole, Development of a predictive model for batch membership of

street samples of heroin, Forensic Sci. Int. 102 (1999) 1.
L. Stromberg, L. Lundberg, H. Neumann, B. Bobon, H. Huizer, N.W. van der Stelt,
Heroin impurity profiling. A harmonisation study for retrospective comparisons, Forensic
Sci. Int. 114 (2000) 67.
M. Chiarotti, N. Fucci, Comparative analysis of heroin and cocaine seizures, J. Chrom.
B 733 (1999) 127–136.
P.J. O’Neill, P.B. Baker, T.A. Gough, Illicitly imported heroin products: some physical
and chemical features indicative of their origin, J. Forensic Sci. 29 (1984) 899–902.
B. Remberg, A. Nikiforov, G. Buchbauer, Fifty years of opium characterisation
methods, Bull. Narc. XLVI (1994) 79–108.
N.K. Nair, V. Navaratnam, V. Rajananda, An effective thin-layer chromatographic
system for separating eight opiates and five adulterants, J. Chrom. A 366 (1986) 363.
I.S. Lurie, Application of micellar electrokinetic capillary chromatography to the
analysis of illicit drugs, J. Chrom. A 780 (1997) 265.
P.A. Hays, I.S. Lurie, Quantitative analysis of adulterants in illicit heroin samples via
revered phase HPLC, J. Liq. Chrom. 14 (1991) 3513.
United Nations Drug Control Program, Recommended Methods for Testing Opium,
Morphine and Heroin, United Nations, New York, 1998.
DEA Special Testing and Research Laboratory, Ethylene diacetate—a substitute for
acetic anhydride in heroin production, Microgram 20 (1987) 195.
J. Skopec, S.Z. Skopec, unpublished results.
J.M. Tedder, The use of trifluoroacetic anhydride and related compounds in organic
synthesis, Chem. Rev. 55 (1955) 787.
E.J. Bourne, M. Stacey, J.C. Tatlow, J.M. Tedder, Studies on trifluoroacetic acid Part I.
Trifluoroacetic anhydride as a promotor of ester formation between hydroxy-compounds
and carboxylic acids, J. Chem. Soc. (1949) 2976.
United Nations Office on Drugs and Crime, http://www.unodc.org/unodc/
en/undcp.html.
F.W. McClafferty, F. Turecek, Interpretation of Mass Spectra, fourth ed., University
Science Books, Mill Valley, 1993.
G.A. Cordell, Introduction to Alkaloids, first ed., John Wiley and Sons, Brisbane
Australia, 1981.