The Impact of Extraction on the Chemical Profile of Cannabis Extracts

The commercial production of cannabis extracts has outpaced the current state of knowledge provided by cannabis research. While there are a myriad of compounds that exist within the cannabis plant that are believed to have therapeutic potential, very little is known about if or how modern extraction processes may affect these compounds.

Speaking at the Analytical Cannabis Science of Cannabis Extraction Online Symposium 2022, Alexander McCorkle, senior research manager at Charlotte’s Web, presented new findings from the team at Charlotte’s Web and researchers at Colorado State University which suggests that extraction methodology does significantly affect the chemical composition of finished cannabis extracts. The group’s research was previously published in Nature’s Scientific Reports.

In addition to these findings, McCorkle also presented new work suggesting that plant maturity and growing region may also have significant effects on a cultivar’s composition.

Extraction method can significantly vary product composition

The ‘Original Formula’ CBD oil produced by Charlotte’s Web has traditionally been done using cannabis plant clones and an isopropyl alcohol (IPA) solvent-based extraction method.

“IPA is sort of a legacy solvent, so we want to see what would be the effects of extracting with ethanol over IPA, beyond the cannabinoid profiles we’ve already looked at,” McCorkle explained.

Additionally, the Charlotte’s Web team had developed a stable feminized hybrid seed, which the researchers also wanted to compare against the original cloned plants.

To study both of these factors, the team harvested material grown from both cloned plants and the feminized seed and compared the chemical composition of the extracts produced from both using IPA solvent, ethanol solvent, and supercritical CO₂ extraction. Each extract was analyzed with a compliment of different approaches, including gas chromatography mass spectrometry (GC–MS) for non-targeted profiling, ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) for the evaluation of specific phytocannabinoids, and inductively coupled plasma mass spectrometry (ICP-MS) for the study of chemical elements.

Using principal component analysis (PCA), the group found that the extracts produced from clones and feminized seeds were very similar. However, there were significant differences in cannabinoid content seen within the three approaches to extraction.

Generally, they found that there was a trend of higher abundance of phytocannabinoids (including CBD and delta-9 THC) in the alcohol extraction methods using ethanol and IPA, compared to the extracts produced by supercritical CO₂ extraction. This trend also held true for the under-researched minor cannabinoids also present in the extracts.

“Depending on the extraction method – and these are all from the same hemp stock – we can have wildly varying amounts of abundance of the contents. Something that was sort of interesting was that cannabicitran was extracted more by alcohol and not very present at all on either of the CO₂ extracts,” McCorkle said.

Cannabicitran (CBT) has also been found in the Rhododendron anthopogonoides plant, which has been used in traditional Chinese and Tibetan medicine as an anti-inflammatory treatment for rheumatoid arthritis and chronic bronchitis.

“There’s not a whole lot of data about cannabicitran in vivo. But part of what we’re interested in doing in this study is trying to understand potentially bioactive compounds that we might not think about very often,” McCorkle said.

Non-cannabinoids are also affected by extraction protocol

Widening the scope away from just cannabinoids, the team also identified several interesting trends in how extraction methods may affect other compounds and nutrients.

As hypothesized, the ethanol solvent tended to extract more polar components compared to the IPA solvent. When comparing the chemical structure of both solvents, IPA features a longer carbon chain with the alcohol group attached at the second carbon, which would explain why it is a less polar solvent than ethanol.

For example, coumaric acid was significantly more abundant in extracts produced by ethanol extraction. Studies have shown that coumaric acid can increase the expression of antioxidant genes and inhibit the generation of reactive oxygen species in vivo. Conversely, shogaol, a potent anti-inflammatory and antioxidant agent, was more abundant in the IPA extracts.

Micronutrients were also found to vary depending on extraction method. Overall, the alcohol extractions were generally more efficient at extracting these nutritional elements; the amount of potassium extracted using the ethanol method approached a level where the extract may be considered a reasonable source of potassium in the diet.

However, more research is still needed to determine whether the variable levels of these compounds might have any significant knock-on effects on the therapeutic potential or biological activity of the final extracts, McCorkle says. Such research may also help to ensure that these extraction methods are not enriching any minor compounds that might present a health risk.

“We found Xylitol enriched in the ethanol extracts. Xylitol is highly toxic in dogs at the right dose. However, at the levels in this extract, [it’s] probably not really a problem. It’s also in strawberries. And those are really good for dogs. But still, it’s important to really look at everything holistically and understand exactly what’s in there,” McCorkle said.

“The type of extraction method you choose matters for more than just cannabinoids and terpenes,” he continued. “Full spectrum extracts should be characterized for both safety and efficacy. There’s a lot more work we can do to really try to understand how those work.”

Plant maturity and growing region also plays an important role

The Charlotte’s Web and Colorado State University researchers also recently conducted experiments designed to assess the effect of plant maturity and growth region on the chemical composition of these plants.

Firstly, they took feminized seed and cannabis clones and cultivated these in a greenhouse under identical conditions, taking flower samples at weeks five, eight, and eleven for analysis. The team used principal component analysis (PCA) in conjunction with GC-MS to study any changes in the plant’s amino acids and sugars, and LC-MS to study larger compounds such as cannabinoids or carbohydrates.

“We saw that the maturity of the plants really drove the chemical composition, a lot more than the propagation method,” McCorkle said. “PCA more or less grouped the samples together by maturity, rather than their propagation method.”

Growth region was also determined to have a significant effect on the plants’ chemical makeup. When comparing otherwise identical crops grown in Oregon, Colorado, and Kentucky, they found notable differences in the amino acids and sugars found in the plants from Oregon compared to the other two locations.

Given the significant effect that growing region and plant maturity can have on a crop, and the transformative effect that different extraction techniques can impose on extracts produced from otherwise identical crops, McCorkle says that it is important for producers and processors to be aware of this. When designing a certain strain or extract product for therapeutic use, it is therefore important to properly assess the chemical composition of any new products.

“Metabolic techniques are a critical and underused tool for the cannabis industry,” McCorkle concluded. “Full spectrum cannabis extracts are complex natural products, and compliance testing only reveals some of their constituents.”

“Chemotype needs to be closely tracked when breeding plants for morphological purposes to avoid changing anything you don't mean to. And lastly, the processing method is critical to the chemical profile of extracts when you’re designing a product for therapeutic purposes.”