Medical Marijuana Musings

At Restek, we’re into medical marijuana, from an analytical standpoint, as cannabis is one of the most complex (and controversial) natural products.  That complexity can make it very challenging to analyze, especially as the suite of analytes expands beyond the usual cannabinoids (e.g. delta-9-THC, cannabidiol, cannabinol) to include other cannabinoids (e.g. cannabigerol, cannabichromene, delta-8-THC, cannabivarin, etc.) and terpenes, sesquiterpenes, oxygenated terpenes, and pesticides.

Except for the pesticides, many of the compounds listed above either have medicinal properties, or contribute to what is called the “entourage effect”.  Simply put, this is where an “extract” of a natural product, which contains many compounds, including what is normally thought of as the active ingredient, is more effective in treatment than the pure active ingredient alone.  Medical marijuana researchers are especially interested in plants that have higher cannabidiol concentrations, as cannabidiol possibly has antioxidant, anticonvulsant, anti-anxiety, and anti-nausea properties, as well as having cytotoxic behavior towards cancer cells.  We recently analyzed 100 seized marijuana samples and found that while most had negligible cannabidiol (CBD) content, four contained CBD that ranged from 0.8 to 4.8%.  More later in another blog on that subject.

We had originally hoped that we could do one extraction procedure (QuEChERS) and determine cannabinoids, pesticides, terpenes, sesquiterpenes, and oxygenated terpenes (using different dilutions and GC and LC runs, of course), but that may be unrealistic, especially for terpenes, which are relatively volatile.  Given that the QuEChERS procedure generates heat when MgSO4 and water are mixed, some of the terpenes are likely lost during extraction, which would lead to a low-biased quantification.  Static or dynamic headspace analysis, or SPME, may be more appropriate for terpene analysis.

Stay tuned for more information on our new analytical efforts in this field and be sure and review the old stuff by going to ChromaBLOGraphy and then using the Search field type in “marijuana” or “cannabis”.

3 Responses to “Medical Marijuana Musings”

  1. matt says:

    Jack, are you having luck using a single QuEChERS extraction procedure to determine both the cannabinoid profile as well as pesticides?

  2. Jack Cochran says:

    Hi Matt:

    Our initial studies looked promising, and I still think it could be done, but I did some “re-extraction” on cannabis previously extracted with the QuEChERS procedure, and was still able to recover some cannabinoids. The percentage was low compared to the first extraction, but we like to be as quantitative as possible. It’s not practical to do a QuEChERS extraction twice on the same material, typically, especially after adding the partitioning salts and centrifuging since the cannabis (or any sample) is now down at the bottom of the centrifuge tube with a hard MgSO4 pellet.

    Another thing that removes some of the attractiveness of trying to do both together is that the sample size for the pesticides work is going to be higher depending on the levels for detection (we likely want to work in the ppb range). Given that many of the cannabinoids are at percent levels, we don’t need near as much sample. This brings us to the point where you can educate me: Absent any FDA guidance on pesticide residue tolerance for medical cannabis, what levels does the industry itself propose as “safe” or “tolerable”?

    I do want to get back to this “one sample preparation” work someday soon if you think it could be viable for labs.

    Thanks so much for reading and your question!


  3. matt says:


    The single sample prep for both would absolutly be in demand for canna labs if one could be developed with reproducable results. I think the problem may be, as you noted, that for pesticide detection the sample size required would be greater than that required for cannabinoid profiling, which would result in a higher concentration of cannabinoids and possibly saturate the detector causing inacurate readings. I guess doing one initial sample prep split into two vials and then diluting only the vial intended for profiling would easily resolve that without too much additional effort.

    As to “Industry Standards” for pesticide residue limits, well, I think that will be hotly debated until government regulation is enacted. We have been researching limits that the OMC in the Netherlands is using and will most likely use those as guidelines absent FDA involvement. I will share them once our research is complete.

Leave a Reply