Source: https://repository.tudelft. nl/islandora/object/uuid%3Ac1b4471f-ea42 -47cb-a230-5555d268fb4c
Title: Sustainable Production of Cannabinoids with Supercritical Carbon Dioxide Technologies
The goal of this thesis was to develop an alternative extraction method of natural compounds of interest from plant material. In specific, the goal was to avoid using organic solvents as much as possible due to residual solvents problems, low selectivity, high energy consumption, and environmental worries.
The alternative method consists of using supercritical fluid CO2 to extract compounds from plant material. There are numerous advantages to doing SFE with CO2, including CO2 being nonflammable, relatively inert, inexpensive, the ease of removal of the solvent, the plant material being non- hazardous afterwards, the different solubility of compounds depending on the temperature and pressure of the fluid, and low critical temperature allowing for extraction of heat- sensitive materials without damage.
The downsides to using CO2 include it not being a great solvent for larger polar molecules and requiring the stream to always be under high pressure which lead to higher initial investment costs. The higher initial investment costs can be outweighed though by how cheap CO2 is and the fewer steps needed for purification. The focus of the thesis is on the separation of phytocannabinoids (or cannabinoids found in the cannabis plant) from the plant material. There are over 60 different phytocannabinoids with the most commons ones being (-)-D9- tetrahydrocannabinol (D9- THC), cannabidiol (CBD), cannabinol (CBN), cannabichromene CBC), cannabigerol (CBG) and tetrahydrocannabivarin (THCV). This study will focus on D9-THC, CBN, CBD, and CBG. Each of these compounds have their own medicinal effects, from pain relief and nausea relief with D9- THC, a sedative effect with CBN, convulsion, anxiety, and inflammation relief with CBD, and analgesic and anti-inflammatory effects with CBG.
The isolation of these compounds from the plant material is of high interest due to the drawbacks of smoking cannabis and different medicinal effects of each compound.
The production method proposed for cannabinoids with purities higher than 95% involves a pre- treatment step, where the acid forms of the cannabinoids are changed to the neutral ones due to better solubility, extraction using SFE with CO2, winterization of the extract to remove waxes, and then purification through centrifugal partition chromatography (CPC).
The cannabis plant strain used in this thesis is Bedrocan which contains around 18% D9- THC and less than 1% of other cannabinoids, thus the main focus will be on extraction of the D9-THC. CBN can be obtained through specific storage conditions to degrade the D9-THC into CBN. CBD and CBG can be obtained using the same process on different cannabis strains with higher concentrations of other cannabinoids.
CO2 becomes a supercritical fluid at temperatures higher than 31.1 degrees C and pressures higher than 1070 psi. This means that the CO2 can only be described as a fluid as it is indistinguishable between a gas or liquid. This is important because it allows for the tuning of the solvent. By changing the pressure or temperature supercritical CO2 can become more or less liquid-like with increasing or decreasing solvency power. CPC is similar to other chromatography techniques. It uses two immiscible liquid phases and uses a centrifugal field to force the mobile phase through the stationary phase.
Each compound has different interactions with these liquids and thus migrate through the phases at different speeds. Thus, they can be collected at the end of the column in relatively pure amounts. Decarboxylation of D -9-THC is necessary due to the acidic form found in the cannabis plant. Usually this occurs during combustion when smoking the plant, but when it comes to medicinal products it will likely need to be transformed without this step. The usual method for large scale decarboxylation involves organic solvents, basic aqueous solutions, and lots of energy, thus alternatives are preferred. One alternative is to pre-treat the cannabis plant before extraction.
When heating the plant material between 90 and 140 degrees C, the decarboxylation reaction from D9- THCA to its neutral form happens at near 100% selectivity. Since the process happens in a solid-state reaction, which leads to a catalytic process, the process could be estimated with a pseudo first order process. This reaction tends to happen at a lower activation energy than normally assumed possibly due to aliphatic and aromatic acids present as other plant constituents in cannabis. While adding strong acids seem to encourage this reaction and could decrease the activation energy, it causes toxic waste from the process which may be bad for other compounds of interest.
The solubility of D9-THC in supercritical CO2 was found for different temperatures and pressures. Below 1914 psi and 40 degrees C, the solubility could not accurately be recorded due to low solubility. In general, the solubility increases with pressure at all temperatures. At about 2175 psi, the solubility is found to decrease with increasing temperature, and above that pressure the solubility is found to increase with increasing temperature.
Some experimental values for D9-THC in supercritical CO2 from the data collected. At 42 degrees C, changing the pressure from 1914 psi to 3640 psi increased the solubility from by 4 times (0.20 to 0.83). At 54 degrees C, changing the pressure from 2030 psi to 3408 psi increased the solubility by around 6 times (0.33 to 1.99). At 61 degrees C, changing the pressure from 1987 psi to 3190 psi increased the solubility increased the solubility by about 7.3 times (0.32 to 2.33). At 72 degrees C, changing the pressure from 2117 psi to 3190 psi increased the solubility about 3 times (0.98 to 2.95).
At most of the temperatures and pressures evaluated in this study the constants created a good predictability for the solubility. The exception being at above 72 degrees C and low pressures.
The solubility of CBN in supercritical CO2 was found for different temperatures and pressures. In general, the solubility increases with pressure at all temperatures, but not as much as with D9-THC. Interestingly, the highest solubility was found at 53 degrees C.
The article concludes that CBN solubility in supercritical CO2 is different enough from D -9-THC that they could be extracted separately to isolate both compounds. This would include a two step extraction, there the plant material is first extracted at 53 degrees C and 1885 psi for CBN and then 2900 psi at the same temperature for D-9-THC.
The solubility of CBG in supercritical CO2 was found for different temperatures and pressures. In general, the solubility increases with pressure at all temperatures, but by a much less magnitude than the D9-THC. Also, the highest solubility was found at the highest temperature.
The article concludes that the solubility trends for CBG are similar to D9-THC, but the actual values are different enough between the two to extract them separately or through fractionation.
The solubility of CBD in supercritical CO2 was found for different temperatures and pressures. In general, the solubility increases with pressure at all temperatures. The difference in solubility between pressures is similar to CBN. Interestingly, the highest solubility occurs at 53 degrees C, like CBN.
The article concludes that CBD’s solubility trends are more similar to CBN and that they are different enough to D9-THC to be extracted separately.
When comparing all four cannabinoids, the difference in solubility can come from a couple things. This includes their melting point (with solid cannabinoids showing better solubility than liquid ones) and their chemical structures (due to CO2 having a higher affinity for non-polar compounds). Overall, CBN has the highest solubility in supercritical CO2. All of the solubility of the different cannabinoids in supercritical CO2 is on the order of 1-2g per kg of CO2 which place them at high enough for SFE.
An example is described to show how one could extract the majority of D-9-THC without other cannabinoids. In a cannabis plant containing 5% D9-THC and 6% CBD (Bediol strain), a first step extraction at ~1885 psi and 42 degrees C would extract 26 percentage of the THC and all of the CBD.
While the CBD would need to be purified, a large amount of the THC could be collected at very pure amounts using this step extraction method.
It was determined that particle size distribution of the plant material had little influence on extraction yields, and thus weren’t investigated.
The highest total yield (extract weight divided by starting weight) was 23.3 percentage and was found at the highest pressure and lowest temperature, 3335 psi and 40 degrees C respectively. This didn’t vary much from the differences in pressure, with 21 percentage being achieved as low as 2175 psi and is believed to be because the extraction was already being ran to completion. This was at flow rates of CO2 of 6 kg per hour for 3 hours. In terms of THC yield, the best yield was found at lower temperatures (40 degrees C).
In terms of time for extraction (at 2610 psi and 6 kg per hour of CO2), the maximum D9- THC yield was found at around 3.75 hours at 40 degrees C. This yield was 98 percentage. Compared to at 50 degrees C, where the maximum yield was reached at about 1.5 hours, however a maximum yield of 74 percentage is reached. During the extraction time, the D9-THC yield increases linearly in time at the same rate between the two temperatures. In comparison to hexane extraction, the D9-THC yields are about the same (85.3 percentage for CO2 and 85.9 percentage for hexane). The other cannabinoid yields were slightly higher with CO2.
The other cannabinoids were found to have the highest yields at 40 degrees C when varying temperature at 2610 psi. All three other cannabinoid yields decrease with increasing pressure at 40 degrees C, while D9-THC’s yield was stable over pressure ranges. This implies that the two step extraction method at 40 degrees C (first at 2175 psi and then at 2900 psi) could first extract the other cannabinoids and then extract the D9-THC, allowing for a more pure extract of D9-THC. This is consistent with what was stated before.
A winterization step could be avoided to remove waxes by having a two stage separator, where the CO2 to decompressed to a medium pressure to precipitate the waxes, followed by another decompression step to recover the cannabinoids. The exact temperatures and pressures would have to be tuned to the solubility of the cannabinoids in the CO2, but should be feasible. In this thesis, a winterization step was included with hexane. This involves dissolving the extract in hexane and freezing it to precipitate out the waxes.
The extraction curves found in this paper determined that the solvent to feed ratio required for extraction of D9-THC is about 0.7g of D9-THC extract per kg of CO2. This is the same for both 40 and 50 degrees C.
It was found that using CO2 as the stationary phase and a water/ethanol mixture as the mobile phase, that no adequate separation could be achieved. Same with CO2 as the stationary phase and a water/methanol mixture as the mobile phase. There are hopes to use supercritical CO2 as the stationary phase, but no commercial CPC machine can handle the pressures required for such a machine.
With the CO2 SFE process outlined, around 80 percentage of the organic solvents can be recycled and 96 percentage of the CO2 can be recycled. Also, the plant matrix after extraction is clean of organic solvent and can be disposed of much easier than with the hydrocarbon extraction. This favors the CO2 SFE process in relation to the environmental impact of the process.
In conclusion CO2 SFE can be used to extract cannabinoids from cannabis plant material. It is heavily favored economically, environmentally, and regulation wise compared to hydrocarbon extraction. The total amount of process steps is also lower than hydrocarbon extraction. It can produce 85 percentage D9-THC extract after a winterization step, which can be further purified. One method of this is CPC which can produce +99 percentage D9-THC. The cost can be largely reduced by having a lower initial cost of cannabis.