Individual assumptions about organic solvents probably vary quite a bit, but normally the term “organic solvents” conjure up putrid visions of poisonous, flammable and toxic vapors, but not all organic solvents fall into that fearful category. We can’t always see or smell them, but they are volatile, and typically measured using a GC/MS with headspace sampler. The testing of cannabis extracts and cannabis-based products for residual solvents seems like a natural extension of the United States Pharmacopeia (USP) monograph 467, which requires that all food or pharmaceutical products that were exposed to solvents during production should be tested for residual solvents. It is just that not all cannabis-complicit states have adopted the existing standard provided by USP monograph 467 which was developed for foods and drugs.
USP <467> provides for solvent classification, where solvents are placed into one of three classes. Class 1 are the truly poisonous, carcinogenic compounds to be avoided at all costs, and then Class 2 which includes methanol with tolerances from 50 to 3,000 PPM. Class 3 solvents include acetone, ethanol, isopropanol, pentane and heptane – all considered low risk when present at less than 5,000 ppm. And butane, propane, isobutene and isopentane were not mentioned in the FDA’s classification because the FDA consider them to be “generally regarded as safe” or GRAS.[1]
So as you now understand, when the test sample is cannabis-based, we are concerned with ensuring that Class 3 solvents including acetone, ethanol, isopropanol, pentane and heptane are under 5,000 ppm, the Class 2 solvent methanol is below 3,000 PPM and we will also look for butane and propane from the GRAS group. But we have a problem. That problem is that during the life cycle of a plant, it produces ethanol, acetone, ispropanol and methanol, creating a background level of inherent residual solvent level that has not always been adequately addressed.[2],[3] The natural production of these Class 2 & 3 solvents by nature is well-established and has recently been a subject at the Association of Commercial Cannabis Laboratories (ACCL) annual meeting,[4] and a talk presented by Justin Fischedick, of Excelsior Analytical Lab, at the AOAC International meeting. Plants make these volatile organics for either signaling, to attract or repel another organism, or because of some metabolic response to wounding or use of nutrients as in fermentation or during osmotic stress.
If the presence of low levels of Class 2 & 3 solvents in cannabis samples is normal as in many other foodstuffs, then how to best accommodate this in our routine testing for residual solvents. Well, this requires that reasonable residual solvent levels be used to deny a product and that they should not just be arbitrary levels across the board. Luckily, the USP <467> sets guidance levels to 3,000 ppm and above for pharmaceutical products,[5] and also recommends only quantifying a solvent if it exceeds the limitation. So what are expected endogenous levels of residual solvents for cannabis flower as well as for cannabis extracts? It should be noted that because residual solvents are volatile, the concentration found in extracts will fluctuate through the life of the product and will not concentrate for other adulterants like pesticides.
The table below presents the ACCL’s Residual Solvent Working Group’s observed typical endogenous levels for three common Class 3 residual solvent levels and one Class 2, ie. methanol, for cannabis flower & cannabis concentrates in comparison to the USP <467> proposed limits and those observed by Digipath Labs in Southern Nevada.
There are other sources of potential contributing residual solvent background numbers as well and that is from both the production lab and the analytical testing lab itself. Both labs use solvents for cleaning benchtops, glassware and lab equipment; not to mention as mobile phases in liquid chromatography. For the production lab, it is important to not purchase low purity solvents; always opt for food or pharmaceutical grade solvents and gases and test products and starting materials before formulating. For the analytical testing lab, the importance of always including the proper blanks for lab air and having adequate ventilation are paramount. Bottom line is that do not be concerned when you see your Certificate of Analysis with higher than expected residual solvent numbers even for your solvent-less extracts. These are important observations to be shared with the cannabis industry and we need to get the word out through education, like this blog, sharing the facts.
[1] Code of Federal Regulations, Title 21, Part 184.1165.
[2] Hood LVS et al. (1973) Headspace volatiles of marijuana. Nature 242:402-403.
[3] Harren FJM, Cristescu SM (2013) Online, real-time detection of volatile emissions from plant tissue. AoB PLANTS 5:plt003;doi:10.1093/asbpla/plt003
[4] Association of Commercial Cannabis Laboratories Annual Meeting. March 22, 2017.
[5] USP 38, Residual Solvents <467>
Written By Cindy Orser, PhD, Chief Science Officer of Digipath Labs