Marijuana and the Cannabinoids

62 ElSohly et al.
6.7. Application of the Marijuana Fingerprint System to Analysis
of Hashish Samples
The fingerprints of hashish exhibits are expected to differ greatly from those of
marijuana because hashish is a product of Cannabis processed to concentrate the
cannabinoids, primarily THC. For this study the GC/MS data of the hashish samples
were obtained using the same fingerprint template developed for marijuana, not a new
set of chromatographic peaks specific to the typical hashish profile.
Five countries were represented in the 68 hashish exhibits provided for the study,
but only three broad regions: South America (Colombia), the Middle East (Lebanon),
and Southwest Asia (Afghanistan, India, and Pakistan). A model based on the five
countries produced correct classifications at rates of 67% Colombia, 100% Lebanon,
50% Afghanistan, 67% India, and 73% Pakistan. Because it was noted that the
misclassified Afghan, Indian, and Pakistani exhibits all fell in the other Asian classes,
those countries were combined, and a second model was created with South America,
Middle East, and Southwest Asia as the classes. In the second model, Southwest Asia
had 98% correct hits, whereas Colombia and Lebanon were 67 and 100%, respectively,
leading us to postulate that the manufacturing methods particular to a region
may induce distinct differences in the chemical profiles of hashish. The anomalies in
the Colombian samples were attributed to the small number of available exhibits.
Although the Cannabis fingerprint system as designed for marijuana reliably
determined the origins of hashish samples, a fingerprint based on the actual peaks
found in hashish chromatograms would undoubtedly improve the accuracy. Additionally,
a study of a marijuana profile compared with the profile of hashish made from
that same marijuana could offer insight into the design of a hashish database.
6.8. Examination of Chemical Profiles for Distinguishing Peaks
Characteristic of Specific Regions
To determine if certain chemical “marker” compounds could be present in marijuana
plants from one region, but absent in plants from another region, data were
again crunched, and Pirouette offered some likely candidates to test this so-called
silver bullet theory.
Three sesquiterpenes—peak 70, peak 92, and peak 63—were predominantly found
in domestic fingerprints. Peak 70 was present in 54% of the domestic specimens and
absent in the foreign ones, peak 92 in 90% of the domestic and 13% of the foreign, and
peak 63 in 93% domestic and 14% foreign specimens. Peak 92 was identified as cisnerolidol,
but the others were only tentatively identified because reference standards
for those compounds could not be obtained. Mass spectral evidence suggested that
peak 63 was γ-elmene and peak 70 either α- or γ-gurjunene.
Peak 130, identified as butylated hydroxytoluene, was detected only in foreign
specimens, particularly Jamaican, but never in domestic ones. A sesquiterpene, peak
86, possibly γ-cadinene or β-farnesene, was totally absent from Colombian, Jamaican,
and Mexican fingerprints but was detected in more than 50% of the Thai and some
domestic profiles. Peak 100, a sesquiterpene identified as guaiol, was detected in only
a few Californian, Hawaiian, and Mexican specimens.