Unintentional THC Isomerization in CO2 Distillate

Hello fellow psychonauts and cannabis enthusiasts alike,

Like many other operators on this forum, I am having an issue with CO2 extracted crude isomerizing into other forms of THC during various steps in the process, particularly distillation. We are producing beautiful, clear, potent distillate, but anywhere from 3-15% of is comprised of delta 8, delta 10, delta 10a/6a, or exo-thc, along with other “unknowns”. It is consistent that at least 3-6% of our distillate is made up of these isomers, and we are trying to figure out where and how this is happening.

We see the presence of these isomers in smaller amounts (<2%) even in the crude, and as the crude goes through the winterization process, this number increases just from the heat of the water bath during rotary evaporation. Finally, the number really jumps up from short-path distillation. We were getting distillate with up to 9% Delta 8 THC after some short path runs.

We also run a Root Science wiped film unit. Part of their original SOP includes cooking the oil on a hot plate at 150C for 4 hours to ensure full decarb and devolatilize the oil. We purchased the short path set up to do this process under vacuum and have more control over the devol. Interestingly, the oil run through the short path has produced higher levels of THC isomers than the oil cooked on the hot plate exposed to ambient atmosphere…

From reading various posts on Future4200 including Isomerization of Co2 extracted Crude, Isomerization to delta 8 from heat alone?, and What happened to my THC content? Lab says conversion into delta10. How?, it seems that the acidity of the crude, plus high temperatures, enhanced by reactivity under vacuum are responsible for this uncontrolled isomerization. Although this change can happen with a variety of crude stocks, it seems to me that it is more prevalent with CO2 crude, for whatever reason. Possibly the inclusion of naturally occurring plant fatty acids that are extracted with the non-selective nature of the supercritical CO2 solvent characteristics?

Testing the acidity of crude, or distillate for that matter, is a tricky one. Dissolving it in ethanol and testing the ethanol is not accurate because testing pH in ethanol is itself not a good representation. We have tried dissolving the oil in heptane, washing against water, and testing the water afterwards, noting that the water does indeed become more acidic, although this is a relative measurement at best. Nonetheless, we do believe our CO2 oil is acidic to some extent. We have tried to utilize 3Å molecular sieves in our extractor to catch any water and prevent carbonic acid from forming. We utilize other filter media directly in our extractor to enhance the quality of our crude and pull out some waxes.

If there are other CO2 operators on here, have you encountered this? To what extent?
What is more responsible for isomerization of delta 9 to its analogues: acidity, or heat?
Does vacuum enhance this chemical reaction, or slow it down?
Most importantly, is there a way to treat this distillate and reverse the isomerization?

We are open to engaging with a consultant who has had experience on this topic, especially if they have experience with CO2 extraction.

COA’s and Chromatograms included below for reference:

This is a batch run through the short path:

A deeper look at the same batch:
230206_Distillate Isomers Peak identification:quantification.pdf (317.7 KB)

A few other batches:


not a CO2 operator, but from reading your string, I would look at the residual water content in your input biomass. you could be evolving carboxylic acid through your process, and acid + heat = isomerization. having d8 + exo in your distillate implies that you are creating a tertiary carbocation at C9 upon nucleophilic attack by an acid, meaning that double bond upon ‘stabilization’ can plant itself between C9 and C11 (exo), C7 and C8 (delta 8), and C8 and C9 (delta 9), of which the most thermodynamically stable is actually delta 8.

Given the above is a viable hypothesis, I would then look at variables like residence time in the system, i.e. higher residence time = more saturation of carboxylic acid = more isomerization, or water content of input material = more isomers in finished distillate. The latter would actually be easier to test as you should have water activity or moisture content on your input biomass COA that you could then just extrapolate out or utilize water activity as a function of isomerization in your process. Would be interesting to see if it’s a linear or exponential relationship as well, given that the kinetics for cannabinoid isomerization are exponential.


now, THAT is the right way to ask a damn question around this gin joint…



You are correct in that your isomers are from the acidity present in your crude oil. Now since you mentioned CO2 extraction… Condensation can play a big role in the end product. This can attract water molecules to the oil while the extract is cold coming out of the extractor in some cases. Also, granular “CRC” media can be used in CO2 systems, and they usually have an acidic pH. If the media contained moisture for any reason it would have definitely affect the oil. Neutralizing the oil with Heptane/water washes until the water comes out neutral, will be your best solution to avoid isomers.


interesting that you say this became more of an issue after you started decarbing/devol under vacuum. This may very well be your culprit.

1 Like

I have a lot of experience with both intentionally making and avoiding these isomers. You’re dealing with something that I’ve likely seen and can hopefully help with. Feel free to DM if you’d like to organize a consult to help you avoid these isomers in the future.


Someone much smarter than I weighed in and their opinion was that it is due to acidity, and you saw an increase in isomerization under vacuum because of heat transfer

1 Like

Why is there still thca in your distillate? Are you decarbing your crude?

1 Like

Are you using carbon filtration?

1 Like

I’m sure he/she is decarbing it, doesn’t mean it’s 100 percent complete, nor does his/her process for making distillate also operate at 100 percent efficiency. Ideally, decarboxylation should be taking place during distillation if it hasn’t been already.

1 Like

What you didn’t mention from my quick skim, was your adsorbent media treatment protocol. If you’re using an acidic media, that is A1 recipe for this. I’ve seen this get out of hand into theory so I would start with most likely and work your way down.


I’d bet the acidity is coming from carbonic acid from the extraction process.


CO2 is absolutely excellent for a ton of stuff… But have you given much thought to using a different machine for skinning this cat?

CO2 is second probably only to rosin for ways I wouldn’t want to attempt to make distillate.

If you’re sticking with your CO2 crude, id recommend dissolving at 8:1 methanol/isopropyl to oleo, winterizing, then adding 5:1 hep/hexane to oleo, then adding a large volume of water so the heptane holds your cannabinoids, and most of your water solubles and pH issues should get discarded.


Thanks for your response. Can you elaborate on what you mean by “evolving carboxylic acid through your process”? How does water content factor into that?

We decarb our flower before extraction (unless we are pulling terps from it). We cook our flower at 250F for 2 hours, which should be more than enough but something that I have struggled to understand is how THCa is able to remain in small quantities throughout the process despite being exposed to heat high enough to convert it to THC. We were encountering oil that was decarbing in our wiped film, ruining our vacuum depth, which prompted us to invest in the short path to decarb/devol at the same time.

Interestingly, the chief science officer at Pro verde labs mentioned a paper in which the authors suggested water was necessary to act as an intermediate of some sort in the decarb process, and that decarbing stops or slows down significantly once the water content is at 0%. I havent confirmed this myself though.

We dont use any activated carbon. We use silica 60å, T-5, CRY (Media Bros), and 3å sieves in our extractor to act as a form of CRC + pull out water. I dont believe any of these media are inherently acidic, and we even tested our CRY and T-5 at the lab to confirm they are pH neutral.

This was one of my first places to look at too, but the media seems to be neutral. Nonetheless, I believe the media may be making the oil more prone to oxidation/isomerization…


I just wanted to chime in as a former CO2 guy that this is the most interesting CO2 extraction thread I’ve read on this forum in years. Excellent question, and excellent responses shoutout @chempistry for the chemistry deep dive.


Are the media’s being baked before use?

1 Like

What does your filter stack after your media look like? Is there a reason you’re using your filter media inline rather than in a post processing step? A lot of the common filter medias performed better for me on a regular hochstrom style filter with my crude dissolved in etoh vs running them inline. I also found that t5 will leave residue down to 0.22μm and a lot of my isomers disappeared when I filtered to submicron before solvent recovery and decarb. I think some of those residual media particles don’t play well with leftover moisture when they get heated up.


No, we do not bake the media before use. They remain in 5gal buckets before they are added directly to the columns for each run. Is baking media a standard practice before utilizing them in CRC?