Contaminant profiles in extracts

Extraction is at its core an enrichment process: a small amount of concentrated product is obtained from a large amount of plant material, in which the desired substances - cannabinoids, terpenes - are highly enriched. The problem: this enrichment is not selective enough to carry along only the desired molecules. Pesticides often have similar polarity and solubility properties to cannabinoids, which is why they get enriched by the same extraction processes that also concentrate THC and CBD.

How strong this effect can be was shown by a by-now historical but still often-cited report from the Cannabis Safety Institute from 2015: pesticide levels in concentrates were on average around 10 times higher than in the respective source bud, and over 80% of the concentrate samples tested in California at the time contained some detectable residue. These numbers are now a decade old, and since then testing regimes have tightened considerably in many markets - but the underlying chemistry of why contaminants get concentrated during extraction remains valid and unchanged in relevance.

That regulators take this effect seriously is directly visible in the numbers themselves: Oregon sets its pesticide action level for bud at 0.2 ppm, but for extracts at 1 ppm - a permitted value 5 times higher, explicitly because concentration during processing is expected. This is not a loosening of safety standards, but a more realistic calculation: if a bud with 0.2 ppm pesticide is processed into extract, the extract can legitimately show higher values without any more pesticide actually being 'produced' - it is the same pesticide, just concentrated in less material.

For particularly dangerous active substances, however, this calculation does not apply - for those there is no higher tolerance factor, but simply a zero-tolerance ban, regardless of whether it is bud or extract. This brings us to the most concrete example of why some active substances are not just limited, but completely banned.

Myclobutanil (trade name Eagle 20, among others) is a widely used agricultural fungicide against powdery mildew. The problem is not just its basic toxicity, but what happens to it under heat: starting at around 205 °C, myclobutanil decomposes into, among other things, hydrogen cyanide (HCN) and other toxic combustion products. This temperature is routinely and significantly exceeded when smoking, and even more so when dabbing concentrates.

This is exactly why myclobutanil is not simply given a high limit in many markets - including Health Canada and several US states - but is completely banned for cannabis, with zero tolerance for both bud and extracts. This is an instructive example that 'strict limit' and 'complete ban' are two different regulatory responses to different risk levels.

Extracts made with butane, propane, ethanol, or CO2 bring an additional, distinct contaminant issue with them: residual solvents. The permissible limits differ noticeably depending on market and solvent class - as a rough snapshot (these values shift between jurisdictions and change over time): for so-called USP <467> Class 3 solvents like butane or propane, a limit around 5,000 ppm is often set, the state of Washington limits total residual solvent/residual gas to 500 ppm per gram, and individual California product categories range between 1,000 and 5,000 ppm depending on product type.

It's important to correctly categorize the term 'purged': it means that a product has been freed of residual solvent through vacuum and controlled heat to the point where it falls below the respective legal or safety-relevant limit - not that no solvent residue whatsoever remains. 'Purged' is thus a considerably weaker statement than 'solvent-free', even though both terms often sound similar in marketing.

A common but misleading conclusion goes: 'Solventless concentrates like ice-water hash, bubble hash, or rosin do without solvents, so they are automatically cleaner or safer.' For residual solvents that is actually true - this one contaminant panel legitimately does not apply to solventless products, because no solvent is used in the process.

But this does not apply to all other contaminant classes. Washing with ice water or pressing with heat and pressure (rosin) is a mechanical separation - the trichomes are separated from the plant material, but no purification in the sense of removing pesticides, heavy metals, or mycotoxins takes place. Whatever contaminants are present in the source bud essentially carry over unchanged into the finished concentrate - in some cases even with the same concentration effect as with solvent-based extracts, because here too a large amount of plant material yields a small amount of finished product.

In practice, this means: a clean certificate of analysis for a solventless product must still show passed pesticide, heavy-metal, and microbiology panels. The only panel that can legitimately be missing from a solventless product is the residual solvent panel - everything else is just as relevant as for any other concentrate, or even for the bud itself.

So that this doesn't stay abstract, here are two concrete, documented examples from recent times: in June/July 2024, the California regulator DCC recalled 'CUREpen Premium THC Oil' vape cartridges from West Coast Cure/Alkhemist DM at over 200 points of sale - the reason was a detection of chlorfenapyr, a pesticide that is not approved for any food application and is banned for cannabis cultivation in California. In the same year, New Mexico's Cannabis Control Division recalled concentrate products (Live Sugar, Live Diamonds, Shatter) from a specific retailer in Albuquerque due to a banned pesticide.

These recalls are not isolated exceptions from the early days of legalization, but current, officially documented examples - good evidence that contaminant risks in concentrates are not a purely academic topic, even in established, regulated markets with mandatory testing.

An additional, important consideration concerns not the product itself, but the route of consumption: a peer-reviewed study on the transfer of pesticides into smoke found - referring to bud, not specifically to concentrates - that depending on the active substance, about 42-70% of certain pesticides transferred into the smoke when smoked through an unfiltered glass pipe or bong, while a water-pipe filter reduced this share to under 11%. This is not directly transferable to extracts, but it does show a fundamental principle: how a product is consumed affects, in addition to the initial contamination level, how much of a contaminant actually reaches the body.

The practical consequences from all of this are manageable, but effective: