Cannabis oil: Chemical evaluation of an upcoming cannabis- based medicine

Full text

Cannabinoids 2013;1(1):1-11
© International Association for Cannabinoid Medicines 1
Original article
Cannabis Oil: chemical evaluation of an upcoming
cannabis-based medicine
Luigi L Romano, Arno Hazekamp
Department of Pharmacy, University of Siena, Italy
Plant Metabolomics group, Institute of Biology, Leiden University, The Netherlands
Abstract
Concentrated cannabis extracts, also known as Cannabis oils because of their sticky and viscous
appearance, are becoming increasingly popular among self-medicating patients as a claimed cure
for cancer. In general, preparation methods for Cannabis oils are relatively simple and do not re-
quire particular instruments. The most well-known example of such a product is called ‘Simpson
oil’. The purpose of the extraction, often followed by a solvent evaporation step, is to make canna-
binoids and other beneficial components such as terpenes available in a highly concentrated form.
Although various preparation methods have been recommended for Cannabis oils, so far no stud-
ies have reported on the chemical composition of such products.
Recognizing the need for more information on quality and safety issues regarding Cannabis oils,
an analytical study was performed to compare several generally used preparation methods on the
basis of content of cannabinoids, terpenes, and residual solvent components. Solvents used include
ethanol, naphtha, petroleum ether, and olive oil. The obtained results are not intended to support or
deny the therapeutic properties of these products, but may be useful for better understanding the
experiences of self-medicating patients through chemical analysis of this popular medicine.
Keywords: cannabis oil, Rick Simpson oil, cancer, cannabinoids, terpenes
This article can be downloaded, printed and distributed freely for any non-commercial purposes, provided the original work is prop-
erly cited (see copyright info below). Available online at www.cannabis-med.org
Author's address: Arno Hazekamp, ahazekamp@rocketmail.com
Introduction
Cannabinoids exert palliative effects in cancer patients
by reducing nausea, vomiting and pain, and by stimu-
lating appetite [1]. In addition, preclinical evidence has
shown cannabinoids to be capable, under some condi-
tions, of inhibiting the development of cancer cells by
various mechanisms of action, including apoptosis,
inhibition of angiogenesis, and arresting the cell cycle
[2,3]. As a result of such exciting findings, a growing
number of videos and reports have appeared on the
internet arguing that cannabis can cure cancer. But
although research is on-going around the world, there
is currently no solid clinical evidence to prove that
cannabinoids - whether natural or synthetic - can effec-
tively treat cancer in humans. It is therefore important
to be cautious when extrapolating preclinical results to
patients.
Anecdotal reports on cannabis use have been historical-
ly helpful to provide hints on the biological processes
controlled by the endocannabinoid system, and on the
potential therapeutic benefits of cannabinoids. The
antiemetic [4], appetite-enhancing [5], analgesic [6],
and muscle-relaxant effects [7] and the therapeutic use
of cannabinoids in Tourette’s syndrome [8] were all
discovered or rediscovered in this manner. But alt-
hough it is possible - and even desirable - that cannabis
preparations exert an antineoplastic activity in, at least
some, cancer patients, the current anecdotal evidence
reported on this issue is still poor, and, unfortunately,
remains far from supporting that cannabinoids are
efficacious anticancer drugs for large patient popula-

Original Article
2 Cannabinoids  Vol 7, Issue 1  May 5, 2013
tions [9]. It should be noted, however, that the potential
effects of terpenes on cancer, either alone or in combi-
nation with cannabinoids, have not yet been addressed
in laboratory studies. Indeed, the synergistic effect
between cannabinoids and terpenes is often claimed to
be the major difference between ‘holistic’ herbal prepa-
rations of cannabis, and products based on single can-
nabinoids [10]. Moreover, self-medicating patients
often use extraction methods and/or administration
forms that are quite different from conditions used in
(pre)clinical studies, possibly resulting in different
serum profiles of cannabinoids and their metabolites
[11] and, consequently, in different therapeutic effects.
Because of this gap between clinical research and real
experiences, the curative potential of whole cannabis
preparations for the treatment of different cancer types
remains unclear.
In recent years an increasing number of patients have
been using concentrated extracts of herbal cannabis,
which, because of its sticky and viscous appearance,
has become known as “Cannabis oil”. Among the self-
medicating population, it is firmly believed that these
products are capable of curing cancer, a claim that is
backed up by numerous anecdotal patient stories. Can-
nabis oil is a concentrated extract obtained by solvent
extraction of the buds or leaves of the cannabis plant.
Various non-polar solvents have been recommended
for this purpose, including petroleumether, naphtha,
alcohol and olive oil. The purpose of the extraction,
often followed by a solvent evaporation step, is to
make cannabinoids and other beneficial components
such as terpenes available in a highly concentrated
form. In general, preparation methods for Cannabis oil
are relatively simple and do not require particular in-
struments. For this reason, people who have access to
cannabis, either home grown or obtained from licensed
pharmacies, dispensaries, coffee shops or the black
market, may prepare it at home by themselves.
In particular, the captivating story of a former patient
called Rick Simpson, a Canadian who claims to have
cured his skin cancer through repeated topical applica-
tion of Cannabis oil produced according to his own
recipe, has received increasing attention. His detailed
story is described on his website [12] and in a docu-
mentary film called “run from the cure” [13] where
various cancer patients describe the therapeutic effects
of ‘Simpson’ oil on their medical condition. In both the
website and documentary, it is explained in detail how
to prepare and administer the product. The method
suggests the use of naphtha or petroleum ether as a
solvent for the extraction, without specifying a particu-
lar quality or source. Both solvents are a mixture of
petroleum hydrocarbons (PHCs), often available in a
wide range of qualities. In general, petroleum ether and
naphtha refer to very similar products, even though
different names may be used around the world; e.g. in
some countries naphtha is equivalent to diesel or kero-
sene fuel. As a result, extensive discussions on solvent
choice can be found on web-forums. Following the
success of Simpson oil, a number of related recipes
have sprung up, emphasizing small but significant
changes to the original recipe. Examples include focus-
ing on extraction with safer solvents such as ethanol
[14], or preventing exposure to organic solvents alto-
gether, by using olive oil [15].
Since cancer is a devastating disease that affects a large
proportion of the world population, it causes some
patients to seek alternative treatments outside the realm
of modern medicine. With a growing interest in Can-
nabis oils for self-medication it is important not to
overlook the importance of quality control and stand-
ardization. In this regard it should be noted that none of
the production methods for Cannabis oil have been
validated in published literature, and no reports have
been made on the chemical composition of these prod-
ucts either. As a result, although many believe Canna-
bis oil may cure cancer, no one seems to know what is
actually in it. Instead, the positive effects of Cannabis
oil are based almost exclusively on case-reports by
people who have used it. This paper evaluates the ef-
fects of preparation methods, and particularly the sol-
vents used, on the final composition of the different
Cannabis oils. The obtained results are not intended to
support or deny the therapeutic properties of these
products, but may be useful for better understanding
the experiences of self-medicating patients through
chemical analysis of this popular medicine.
Materials and Methods
Plant material
Cannabis plant material used in this study was of the
variety ‘Bedrocan’ (19% THC w/w) and was obtained
from Bedrocan BV (Veendam, The Netherlands) where
it was cultivated under standardized conditions accord-
ing to the requirements of Good Agricultural Practice
(GAP). Only female flower tops were used (‘Cannabis
Flos’). After harvest, the plant material was air-dried in
the dark under constant temperature and humidity for 1
week. Dried flowers were manicured to remove leaves
and stems, and finally cut in smaller pieces. The same
cannabis material is officially dispensed through Dutch
pharmacies under the medicinal cannabis program of
the Netherlands, supervised by the Office of Medicinal
Cannabis (OMC). The plant material was homogenized
by grinding, and stored at -20°C until used.
Chemicals and solvents
Ethanol (HPLC grade), methanol (HPLC grade), acetic
acid (analytical grade) and activated charcoal (analyti-
cal grade) were purchased from Sigma-Aldrich (Stein-
heim, Germany). Petroleum ether (boiling point 40-
65°C; analytical grade) was purchased from Boom BV
(Meppel, The Netherlands). Naphtha (light hydrotreat-
ed petroleum distillate; Coleman® fuel) was purchased
from the Coleman Company (Wichita, USA). Olive oil
(extra virgin quality) was purchased from a local gro-
cery store. Deuterated chloroform (CDCl3) was from
Eurisotop (Gif-sur-Yvette, France). Pure ethanolic
standards for delta-9-tetrahydrocannabinol (THC) and

Grotenhermen
Cannabinoids  Vol 5, No 1  January 23, 2010 3
delta-9-tetrahydrocannabinolic acid (THCA) were
produced as previously described [16,17]. Cellulose filter paper for filtration of extracts was from Whatman
Ltd. (Maidstone, UK).
Table 1: Detailed description of the five different protocols used for preparation of Cannabis oils.
Preparation step
1) NAPHTHA
2) PETROLEUM
ETHER
3) ETHANOL
4) OLIVE OIL I
5) OLIVE OIL II
CANNABIS (g)
5g
5g
5g
5g
10g
SOLVENT (mL)
Naphtha
(200 mL)
Petroleum ether
(200 mL)
Ethanol
(200 mL)
Olive oil (20 mL) +
water (70 mL)
Olive oil
(100 mL)
EXTRACTION/
FILTRATION
Extraction #1:
5 g cannabis + 100 mL
naphtha, agitate 20
min. (a)
•
Filtration with filter
paper
•
Extraction #2:
Same cannabis + 100
mL naphtha, agitate 20
min. (a)
•
Filtration with filter
paper
•
Combine extracts
Extraction #1:
5 g cannabis + 100 mL
petr. ether, agitate 20
min. (a)
•
Filtration with filter
paper
•
Extraction #2:
Same cannabis + 100
mlLpetr. ether, agitate
20 min. (a)
•
Filtration with filter
paper
•
Combine extracts
Extraction #1:
5 g cannabis + 100 mL
ethanol, agitate 20
min. (a)
•
Filtration with filter
paper
•
Extraction #2:
Same cannabis + 100
mL ethanol, agitate 20
min. (a)
•
Filtration with filter
paper
•
Combine extracts
5g cannabis + 20 mL
olive oil + 50 mL water.
Heat in water bath
~98°C for 60 min.
•
Before filtration, let it
stand to cool off.
•
Filtrate by pressing (b)
•
Rinse the plant
material with 20 mL of
hot water
•
Filtrate by pressing (b)
•
Combine extracts
10 g cannabis + 100
mL olive oil. Heat in
water bath ~98°C for
120 min.
•
Before filtration, let it
stand to cool off.
•
Filtrate by pressing (b)
EXTRACT CLEAN-UP
N/A
N/A
(optional): Filter
extract over a column
filled with activated
charcoal
N/A
N/A
EVAPORATION/
SEPARATION
Evaporate solvent in
water bath ~98°C
under stream of
nitrogen gas
Evaporate solvent in
water bath ~98°C
under stream of
nitrogen gas
Evaporate solvent in
water bath ~98°C
under stream of
nitrogen gas
Let the solution stand
to separate water and
oil. Put it in the freezer
(-20°C) overnight
N/A
RECONSTITUTION
Reconstitute residue
with EtOH to 100 mL
Reconstitute residue
with EtOH to 100 mL
Reconstitute residue
with EtOH to 100 mL
Collect upper layer
(oil) by pouring it off
the frozen water layer
Collect the oil
EXTRACT
CONCENTRATION
(cannabis/solvent)
5 g/100 mL
5 g/100 mL
5 g/100 mL
5 g/20 mL
10 g/100 mL
DILUTION FACTOR FOR
ANALYSIS
20x
20x
20x
100x
40x
FINAL
CONCENTRATION
(cannabis/solvent)
2.5 mg/mL
2.5 mg/mL
2.5 mg/mL
2.5 mg/mL
2.5 mg/mL
a): agitate by using a shaking platform @ 120 rpm
b): separate oil from plant material by using a French coffee press
Effects of preheating
Preheating of cannabis samples has been recommended
as a way to potentiate the final extract, i.e. to decar-
boxylate the acidic cannabinoids naturally present in
cannabis plant material, such as THCA and CBDA,
and turn them into their more potent counterparts such
as THC and CBD [18,19]. Therefore, we tested two
decarboxylation methods by heating cannabis plant
material (1 g in an open glass vial) under two condi-
tions: I) in a water bath at a low boil (temp. 98-100°C)
for 5 min, and II) in an oven heated at 145°C for 30
min. Unheated samples were used as a control for these
experiments. All experiments were done in duplicate.
Subsequently, samples were extracted as previously
described [20,21] and analyzed by HPLC and GC.
Preparation of concentrated extracts
Five different extraction protocols for the preparation
of concentrates were assessed. Details are described in
table 1. These included a naphtha (1) and a petroleum
ether extraction (2) according to the procedure of Rick
Simpson [12,13]; an ethanol extraction based on an

Original Article
4 Cannabinoids  Vol 7, Issue 1  May 5, 2013
authoritative Dutch website on Cannabis oil [14]; and
two olive oil extractions using different heating dura-
tions (4, 5) based on popular Youtube videos [15].
Chemically, naphtha and petroleum ether are very
similar solvents, and sometimes hard to distinguish
because of the many different qualities available. In the
context of this study we selected an industrial quality
naphtha that was sold as camping fuel (Coleman®) and
contains added chemicals for improving stability, while
the petroleum ether used was of laboratory quality, and
represents a more pure and better characterized prod-
uct. Both solvents may be purchased by inexperienced
patients under the name naphtha or petroleum ether.
All preparation methods consisted of only a few simple
steps, typically involving one or two extraction steps,
separating plant material from solvent, and finally (in
case of organic solvents) an evaporation step to pro-
duce a concentrate. For the ethanol extraction (3) we
also tested the effect of filtration over activated char-
coal, intended to remove chlorophyll which is strongly
extracted by ethanol and may add an unpleasant
‘green’ flavour to the extract. Because the different
extraction methods used different solvent-to-plant
ratios, all extracts were finally diluted in ethanol to
obtain a solvent-to-plant ratio of 2.5 mg/mL in order to
allow direct chromatographic comparison of canna-
binoid and terpene contents by high performance liquid
chromatography (HPLC) and gas chromatography
(GC).
GC/FID analysis
Because of the heat applied during injection and sepa-
ration, GC is not able to show the presence of acidic
cannabinoids without sample derivatization. As a re-
sult, GC reveals the total cannabinoid content (acidic +
neutral cannabinoids) after decarboxylation, only.
However, terpenes can be efficiently analyzed by GC.
Therefore, an Agilent GC 6890 series (Agilent Tech-
nologies Inc., Santa Clara, CA, USA) equipped with a
7683 autosampler and flame ionization detector (FID)
was used for the analysis of cannabis terpenes as previ-
ously described [20,21]. The instrument was equipped
with a DB5 capillary column (30 m length, 0.25 mm
internal diameter, film thickness 0.25 μm; J&W Scien-
tific Inc., Folsom, CA, USA). The injector temperature
was 230°C, with an injection volume of 4 μL, a split
ratio of 1:120 and a carrier gas (N2) flow rate of 1.2
mL/min. The temperature gradient started at 60°C and
increased at a rate of 3°C/min until 240°C which was
held for 5 min resulting in a total run time of 65 min.
The FID temperature was set to 250°C. The GC was
controlled by Agilent GC Chemstation software ver-
sion B.04.01
HPLC analysis
Cannabinoid profiles were studied in more detail by
HPLC, which enables the differentiationof acidic can-
nabinoids (THCA, CBDA etc.) and their neutral ana-
logues (THC, CBD etc.). Analyses were carried out
using an Agilent (Agilent Technologies Inc., Santa
Clara, CA, USA) 1200 series HPLC system, consisting
of a G1310A pump, an G1322A solvent degasser, and
a G1329A autosampler. Full spectra were recorded in
the range of 200-400 nm using a G1315D photodiode-
array (PDA) detector. Chromatographic separation was
achieved using a Phenomenex C18 column (type
Kinetex, 2.6 μm, 3 x 100 mm). Equipment control, data
acquisition and integration were performed with Ag-
ilent Chemstation software. The mobile phase consist-
ed of methanol and water, acidified with 25 mM formic
acid. Initial setting was 75% methanol (v/v), which was
linearly increased to 100% methanol over 10 min.
After maintaining this condition for 1 min, the column
was re-equilibrated under initial conditions for 4 min,
resulting in a total runtime of 15 min. The flow-rate
was set to 0.5 mL/min, the injection volume was 2 μL,
and the detection wavelength was 228 nm. All experi-
ments were carried out at a column temperature of 40
ºC.
NMR analysis
Proton Nuclear Magnetic Resonance (1H-NMR) analy-
sis for detection of solvent residues was performed by
dissolving sample aliquots in deuterated chloroform.
Spectra were recorded on a Bruker DPX 300MHz
spectrometer, as previously described [17].
Results and Discussion
Effects of preheating
In the cannabis plant, cannabinoids are biosynthesized
as their acidic forms, characterized by the presence of a
carboxyl group attached to the phenolic ring. Acidic
cannabinoids can be rapidly converted into their ‘neu-
tral’ analogues under the influence of heat or extended
storage [18], which causes loss of the relatively unsta-
ble carboxyl group in the form of carbon dioxide (de-
carboxylation). Preparation of cannabis oil, mainly
intended for oral use, usually involves temperatures
that are relatively low compared to other forms of ad-
ministration where heating of the material is typically
performed at much higher temperatures (e.g. smoking,
vaporizing or baking). For a more thorough decarboxy-
lation, preheating of herbal cannabis before preparation
of cannabis oil has been suggested, for example by
placing the cannabis in an oven.
Besides cannabinoids, the cannabis plant contains a
range of terpenes, which are the volatile compounds
that give cannabis its distinct smell and may act syner-
gistically with cannabinoids [10]. Although preheating
the plant material may release more of the known ac-
tive (neutral) cannabinoids, it may simultaneously also
cause loss by degradation or evaporation of compo-
nents such as terpenes. Our tests were intended to bet-
ter clarify the balance between desired decarboxylation
and unwanted degradation. Unheated cannabis material
was analyzed as a control.
Figure 1A shows the cannabinoid profile of the decar-
boxylated samples, obtained by HPLC analysis. The

Romano & Hazekamp
Cannabinoids  Vol 7, Issue 1  May 5, 2013 5
mild water bath treatment did not lead to significant
changes in the acidic-to-neutral cannabinoid ratio. In
contrast, the oven treatment resulted in a complete
decarboxylation of the major cannabinoids detected.
THCA, cannabigerolic acid (CBGA) and canna-
bichromenic acid (CBCA) had all fully converted into
THC, cannabigerol (CBG) and cannabichromene
(CBC), respectively. Further conversion of THC into
its’ main degradation product cannabinol (CBN) only
took place to a small degree during the oven treatment.
Figures 1B and 1C show the terpene profile acquired in
our decarboxylated samples using GC. Compared to
the untreated control, monoterpenes (the most volatile
class of terpenes) were reduced to about half of their
original levels even after exposing the plant material to
boiling water for just 5 min. After the more intense
oven treatment, only small traces of the monoterpenes
terpineol, myrcene and terpinolene could still be de-
tected. As may be expected, the less volatile sesquiter-
penes were more resistant to the mild treatment with
the water bath. However, most of them were lost in the
oven treatment, and only traces of gamma-cadinene
and eudesma-3,7(11)-diene remained.
These data indicate that significant decarboxylation of
the major cannabinoid acids occurs only by exposure to
higher temperatures for extended time (oven at 145°C
for 30 min), which is in agreement with previous stud-
ies [18,22]. However, under these conditions all major
terpenes present were affected by significant evapora-
tion. Although milder decarboxylation using a boiling
water bath may be efficient when applied for longer
time [22], the terpene profile already changes signifi-
cantly after only 5 min of treatment. For this reason, all
further experiments were carried out without applica-
tion of a preheating step.
Analysis of the extracts: cannabinoid and terpene con-
tent
Analysis by HPLC to reveal the ratio between acidic
and neutral cannabinoids in the different extracts was
limited to the main cannabinoids THCA and THC.
Results are shown in Figure 2A. Most extracts con-
tained only a small proportion of THC (5-10% of total
THCA + THC content), as a result of the relatively low
heat of max. 100°C applied during the evaporation
(protocol 1-3) or extraction (protocol 4-5) step. A nota-
ble exception was the naphtha extract, which was
found to contain 33% of total THCA + THC content
present in the form of THC. This is remarkable because
the extract prepared with petroleum ether did not show
the same composition, even though both solvents are
chemically quite similar. Perhaps added chemicals (e.g.
for stability) in the naphtha used in this study may be
responsible for the observed difference.
Analysis of the extracts by GC indicated that the major
components present in the cannabis material used were
the monoterpenes beta-pinene, myrcene, beta-
phellandrene, cis-ocimene, terpinolene and terpineol,
and the sesquiterpenes beta-caryophyllene, humulene,
delta-guaiene, gamma-cadinene, eudesma-3,7(11)-
diene and elemene. This is in agreement with previous
reports on cannabis variety ‘Bedrocan’ [20,21].
The extraction solvents showed comparable efficiency
for extracting terpenes, with the notable exception of
naphtha (Figure 2B and 2C). While this solvent gener-
ally extracted terpenes less efficiently than the other
solvents, several terpenes could not be detected at all in
the naphtha extract. It is not known whether (i) these
components were not extracted from the plant material,
(ii) were degraded or evaporated during the extraction
protocol, or (iii) GC retention times for these compo-
nents were changed as a result of interaction with sol-
vent components. Interestingly, the use of petroleum
ether (chemically very similar to naphtha) did not show
the same absence of components.
The use of olive oil as extraction solvent was found to
be most beneficial based on the fact that it extracted
higher amounts of terpenes than the other sol-
vents/methods, especially when using an extended
heating time (120 min; protocol 5). This may be ex-
plained by the highly non-polar but also non-volatile
character of olive oil, resulting in a good solubilization
of terpenes while limiting their loss by evaporation.
Treatment of the ethanolic extract with activated char-
coal, intended to remove chlorophyll, resulted in a
considerable reduction of cannabinoid content (~50%)
as well as all other sample components (data not
shown). For this reason, the use of charcoal should not
be recommended and was not further evaluated in our
study.
Residual solvent testing
Naphtha and petroleum ether are mixtures of various
hydrocarbon solvents with a range of boiling points,
typically between 30 - 200°C. All the solvent compo-
nents should be considered harmful and flammable,
and some of them, such as hexane and benzene, may be
neurotoxic. Both naphtha and petroleum ether are con-
sidered potential cancer hazards according to their
respective Material Safety Data Sheets (MSDS) pro-
vided by manufacturers. Moreover, products sold as
naphtha may contain added impurities (e.g. to increase
stability) which may have harmful properties of their
own [23]. For these reasons, the naphtha and petroleum
ether extracts were analyzed for residual solvent con-
tent.
Analysis by GC as well as NMR revealed significant
residues of petroleum hydrocarbons (PHCs) in the
naphtha and petroleum ether extracts. As may be ex-
pected, mainly PHCs with a higher boiling point (as
indicated by longer GC retention times) were detected,
as they are more resistant to the evaporation procedure
used (Figure 3A). In the naphtha extract, based on GC
peak areas, the content of naphtha residue was roughly
similar to the total content of terpenes remaining in the
extract (Figure 3B).
Reconfirmation using an actual patient sample
In order to confirm our experimental results, we also
analyzed a sample provided by a patient in the Nether-
lands who produced his own cannabis oil using

Original Article
6 Cannabinoids  Vol 7, Issue 1  May 5, 2013
Figure 1: (A) Effect of (pre-)heating on the cannabinoid (HPLC analysis), (B) monoterpene and (C) sesquiterpene
composition (GC analysis) of herbal cannabis material. (THCA: tetrahydrocannabinolic acid; THC: tetrahydrocannabinol;
CBN: cannabinol; CBGA: cannabigerolic acid; CBG: cannabigerol; CBCA: cannabichromenic acid; CBC: cannabichromene)

Romano & Hazekamp
Cannabinoids  Vol 7, Issue 1  May 5, 2013 7
Figure 2: (A) Effect of five different preparation methods on the cannabinoid (HPLC analysis), monoterpene and
sesquiterpene composition (GC analysis) of concentrated cannabis extracts.

Original Article
8 Cannabinoids  Vol 7, Issue 1  May 5, 2013
Figure 3a: Residual naphtha solvent components present in the naphtha extract as indicated by GC analysis. Dotted lines are
added for easier comparison. All chromatograms are shown at the same vertical scaling.

Romano & Hazekamp
Cannabinoids  Vol 7, Issue 1  May 5, 2013 9
Figure 3b: GC analysis showing the same ethanol and naphtha extracts as above (Fig. 3a), but now using a larger time scale
to compare total peak area of naphtha components to the sesquiterpenes present in these samples.
Bedrocan® cannabis and following the Simpson meth-
od as described in the internet. The patient was a 50
year old male suffering from cancer of the (left) tonsil
and the tongue. The analytical results (data not shown)
were equivalent to our lab experiments described
above, confirming the residual presence of PHCs at
significant concentrations in a product that is intended
for self-medication of cancer.
Conclusions
Concentrated cannabis extracts, also known as Canna-
bis oils, are increasingly mentioned by self-medicating
patients as a cure for cancer. Despite this growing
popularity, so far no studies have been reported on the
chemical composition or on the different preparation
methods of such products. Recognizing the need for
more information on quality and safety issues regard-

Original Article
10 Cannabinoids  Vol 7, Issue 1  May 5, 2013
ing Cannabis oils, the small study presented here com-
pared on the basis of cannabinoid, terpene, and residual
solvent content a few generally used recipes for prepa-
ration of Cannabis oils,.
Based on the results of our preheating experiments,
comparing a mild water bath treatment to more intense
heating in an oven, it can be concluded that it is not
feasible to perform decarboxylation of cannabinoids,
without significant loss of terpene components. This is
particularly important because of the fact that users of
Cannabis oils often claim the holistic nature of canna-
bis components to be responsible for its therapeutic
effects. Retaining the full spectrum of terpenes present
in fresh cannabis material should therefore be a major
focus during optimal Cannabis oil production.
When comparing five methods of Cannabis oil prepara-
tion, some interesting differences were observed be-
tween the resulting extracts. Specifically the prepara-
tion method described by Rick Simpson has attracted
quite a following of self-medicating patients. This
method favours the use of naphtha as solvent for can-
nabinoid extraction, without specifying issues regard-
ing quality or safety. According to the Simpson web-
site: “All these solvents […] are poisonous in nature,
but if you follow these instructions solvent residue in
the finished oil is not a concern. […] Even if there was
a trace amount of solvent residue remaining, the oil
itself would act upon it to neutralize any harmful poi-
sonous effect.” [13]. In other words, the curative prop-
erties are considered to be strong enough to counteract
any and all potential negative effects caused by residu-
al solvents. Chemical analysis of our laboratory sam-
ples, as well as a sample obtained from a patient,
showed that the heavy fraction (components with high
boiling point) of naphtha indeed remains in the extract
despite the recommended evaporation step. Based on
GC-FID peak areas, the total content of PCHs roughly
equalled the total content of terpenes present in the
extract. The potential harmful effects of these solvent
residues have been discussed above.
It should be noted that as a result of sample viscosity,
the more concentrated an extract becomes, the more
difficult it will be to remove the residual solvent from
it. In such a case, applying more heat will increase
evaporation, but simultaneously more terpene compo-
nents will be lost as well. Especially under conditions
where Cannabis oil is prepared by simple household
methods, there will always be a trade-off between re-
sidual solvents and terpene content. For this reason, the
use of non-toxic solvents should always be advised, so
that potential residues are not harmful to health.
As extraction solvents for the production of Cannabis
oils, ethanol and olive oil were shown to perform much
better, extracting all terpenes and cannabinoids tested
very efficiently. Additionally, these solvents are not
harmful. Unfortunately, pure ethanol efficiently ex-
tracts chlorophyll from cannabis, which will give the
final extract a distinct green colour, and often unpleas-
ant taste. Removing chlorophyll by filtering the ethanol
extract over activated charcoal was found to be very
effective, but it also removed a large proportion of
cannabinoids and terpenes, and is therefore not ad-
vised. Additionally, in most countries consumption-
grade ethanol is an expensive solvent, as a result of
added tax on alcohol products.
Of the solvents tested, this leaves olive oil as the most
optimal choice for preparation of Cannabis oils for
self-medication. Olive oil is cheap, not flammable or
toxic, and the oil needs to be heated up only to the
boiling point of water (by placing a glass container
with the product in a pan of boiling water) so no over-
heating of the oil may occur. After cooling down and
filtering the oil, e.g. by using a French coffee press, the
product is immediately ready for consumption. As a
trade-off, however, olive oil extract cannot be concen-
trated by evaporation, which means patients will need
to consume a larger volume of it in order to get the
same therapeutic effects. In a follow-up study on the
use of Cannabis oils, there should be more focus on the
characteristics and motivations of those who use it for
self-medication.
Acknowledgements
LR was funded for this research by the Puglia (Italy)
regional government with a “Ritorno al futuro” grant.
AH is the head of Research and Development at Bed-
rocan BV, the Netherlands.
References
1. Guzmán M. Cannabinoids: potential anticancer
agents. Nat. Rev. Cancer. 2003; 3(10): 745-755.
2. Calvaruso G, Pellerito O, Notaro A, Giuliano M.
Cannabinoid-associated cell death mechanisms in
tumor models (review). Int. J. Oncol. 2012;
41(2): 407-413.
3. Velasco G, Sánchez C, Guzmán M. Towards the
use of cannabinoids as antitumour agents. Nat.
Rev. Cancer 2012; 12(6): 436-444.
4. Dansak DA. Cannabis as an antiemetic and appe-
tite stimulant in cancer patients. McFarland &
Co.: Jefferson, NC, 1997; pp 69-83.
5. Plasse TF, Gorter RW, Krasnow SH, Lane M,
Shepard KV, Wadleigh RG. Recent clinical expe-
rience with dronabinol. Pharmacol. Biochem.
Behav. 1991; 40(3): 695-700.
6. Noyes R, Baram DA. Cannabis analgesia. Com-
prehensive Psychiatry 1974; 15(6): 531-535.
7. Clifford DB. Tetrahydrocannabinol for tremor in
multiple sclerosis. Ann. Neurol 1983; 13(6): 669-
671.
8. Müller-Vahl KR, Kolbe H, Schneider U, Emrich
HM. Cannabis in movement disorders. Forsch.
Komplementärmed. 1999; 6 (Suppl 3): 23-27.
9. Guzman M 2012. Do cannabinoids cure cancer?
IACM website: http://www.cannabis-
med.org/index.php?tpl=faq&red=faqlist&id=274
&lng=en
10. Russo EB. Taming THC: potential cannabis syn-
ergy and phytocannabinoid-terpenoid entourage
effects. Br. J. Pharmacol. 2011; 163: 1344-1364.

Romano & Hazekamp
Cannabinoids  Vol 7, Issue 1  May 5, 2013 11
11. Grotenhermen F. Pharmacokinetics and pharma-
codynamics of cannabinoids. Clin. Pharmacokin.
2003; 42(4): 327-360.
12. Simpson R. 2008. Documentary film: Run from
the cure. Available on youtube.
http://www.youtube.com/watch?v=0psJhQHk_GI
13. Simpson R. 2013: Rick Simpson official website:
http://phoenixtears.ca/
14. Bruining W. 2013. Wernard Bruining. Mediwiet
[Medi-weed] official website:
http://www.mediwiet.nl
15. Dr. Diane. 2013. Youtube instruction video:
http://www.youtube.com/watch?v=vs66uyiH968
16. Hazekamp A, Simons R, Peltenburg-Looman A,
Sengers M, Van Zweden R, Verpoorte R. Prepar-
ative isolation of cannabinoids from Cannabis sa-
tiva by centrifugal partition chromatography. J.
Liq. Chrom. Rel. Technol. 2004; 27: 2421-2439.
17. Hazekamp A, Choi YH, Verpoorte R. Quantita-
tive analysis of cannabinoids from Cannabis sati-
va using 1H-NMR. Chem. Pharm. Bull. 2004;
52(6): 718-721.
18. Veress T, Szanto JI, Leisztner L. Determination
of cannabinoid acids by high-performance liquid
chromatography of their neutral derivatives
formed by thermal decarboxylation in an open re-
actor. J. Chromatogr. 1990; 520: 339-347.
19. Hazekamp A, Bastola, K Rashidi H, Bender J,
Verpoorte R. Cannabis tea revisited: A systematic
evaluation of the cannabinoid composition of
cannabis tea. J. Ethnopharm. 2007; 113: 85-90.
20. Fischedick JT, Hazekamp A, Erkelens T, Choi
YH, Verpoorte R. Metabolic fingerprinting of
Cannabis sativa L., cannabinoids and terpenoids
for chemotaxonomic and drug standardization
purposes. Phytochem. 2010; 71(17-18): 2058-
2073.
21. Hazekamp A, Fischedick JT. Cannabis - from
cultivar to chemovar. Towards a better definition
of cannabis potency. Drug Testing and Analysis
2012; 4: 660-667.
22. Fischedick JT, Glas R, Hazekamp A, Verpoorte
R. A Qualitative and Quantitative HPTLC Densi-
tometry Method for the Analysis of Cannabinoids
in Cannabis sativa L. Phytochem. Anal. 2009; 20;
421-426.
23. IARC 1989: WHO International Agency for
Research on Cancer. IARC monographs on the
evaluation of carcinogenic risks to humans. Vol-
ume 47.
http://monographs.iarc.fr/ENG/Monographs/vol4
7/mono47.pdf