Decarboxylation Efficiency

As does mine.

1 Like

Here is the link to a paper by Steep Hill Lab in California, one of the most respected.

Its not crystal clear what they are saying, but the bottom line is they say the theorectical conversion rate is .877 but then on top of that add a .75 rate for the side reactions and the end rate is .66.
My data says .68.
If you can achieve a higher conversion rate than that in your lab or process, I would like to explore the details. It may be that by adding a catalyst or some other method you can increase the conversion rate, but smoking a joint just does the same thing that happens in the GC. There are no catalysts, just heat. So maybe its more accurate to say that the heat based conversion of THCA to d9THC is .68.

1 Like

Any insight in the decarboxila of cbd?
Trying to figure out the best way to decrb thc-a and not the cbd-a that is in a mixture of 2T:30C
Thank you

That’s not really a peer reviewed paper. It’s more of a blog post. I have an article here that discusses this exact topic. And it does claim that people doing GC analysis see a 65% effiency, however recent studies with HPLC show decarboxylation effiencies that are much higher. “More recently, Perrotin-Brunel
et al.11 studied the kinetics and molecular modeling of
the decarboxylation of THCA-A using HPLC. The proposed pseudo-first-order, acid catalyzed keto–enol mechanism for the decarboxylation process was found to
be >95% efficient.” These conclusions are more in keeping with what I have seen in my own studies.

Here is the link to the article:DecarboxylationstudyofAcidiccannabinoids.pdf (1000.3 KB)


we do a cannabinoid mass balance for our entire process, from incoming flower to final distillate, via an in house HPLC, and there are several spot checks along the way as well. i promise you, the decarb step is damn near quantitative.

im torn - i would love to share data but i feel like youre just trying to sell your GCs…

For thc ? Or cbd?
Most papers I read say correct for thc
Cbd seems to have a lot more side reactions

1 Like

Yeah I’ve noticed the same thing with my data. Even mass losses that I see are way above this value.

1 Like

Truthfully his data says more about the non viability of gc for testing cannabinoids. It’s just not the right tool for the job.


no, same for cbd, always near quantitative, youre doing something wrong if you lose that much during decarb!

ethanol some cbd buds, and then reduce your solvent. get a potency on the oil. eg. 60% cbda, 10% cbd, and its mass is 500g.
thats 300g cbda, 50g cbd.
theoretical cbd on this is 50 + (300*.87) = 50 + 261 = 311g total cbd after decarb.

now decarb it. get another potency. youll have ~311g of cbd. this stuff doesnt just disappear.

last time i saw someone waving around that link from steep hill, they were selling $300 “decarboxylators” to unsuspecting DIY stoners at an expo. they show you the “paper” and then demonstrate that their product (practically a fucking coffee pot) decarbs near quantitatively with potency numbers from before and after they decarb it.

is it a coincidence that the only other person to push this .68 number is also trying to sell a “solution” to this “problem”?

Have you tried working under nitrogen? It helps dry out residuals and minimizes degradation pretty well. I’ve cooked oil over 130 at atmospheric with no degradation under n2

1 Like

If I get any degradation it s in the 0.5%
Not worth the hassle
My interest has peaked figuring out how to decarb the thc-a portion in cbd-a crude for this remediation tech proposed
Reason I am trying to find Alice :rofl:
Down the rabbit hole


How much THC?

According to this, Cbda decarb rate can be influenced by the amount of THC present. Not sure why, and they don’t have much data.

1 Like

I read the paper you suggested and I understand they were able to get a near perfect conversion of THCA to d9THC using a vacuum oven and excluding O2 and light. I also gathered that the instrumentation they used was powerful and expensive.
My purpose here is not to argue that GC is a superior measurement technique to HPLC, just to try to explain that it works about as well ( for a fraction of the cost ) and delivers a result which mirrors most people’s experience of decarbing. I have also tried to correct some misunderstandings regarding GC vs HPLC measurement.
There is also one other confusing number and that is when you says 95% efficiency you are really saying the conversion is THCA times .877 then times .95 = .833. But when I and Steep HIll say the conversion is .68 we are really saying THCA times .877 times .75 =.68. So there is no disagreement that the real conversion rate is not the widely used .877 number. We just don’t know for sure what the real number is. If you smoke a joint, the number is probably very close to what happens in the GC.
If you have the best equipment and process then you get the higher conversion rate you report.


once again no. it is .877 in any normal scenario.

how hot does your GC get? i dont doubt you can degrade a sample of d9 in a GC (which is part of what you’re saying) but if you decarb thca in a “normal” scenario (ex. boil a crude oil on a hot plate) you will see near perfect conversion.

you are proposing only 75% efficiency?? to what? d8? past that? cbn? this is demonstrably false!
it is simply foolish to make such a claim. many of us do this every single day, with analytics. i dont lose 1/4 of my thc on the hot plate. no one does.

@1234user said it best:

Maybe you’re right. I will re-do my tests.
However, I don’t think Steep Hill used a GC for their analysis.
And I don’t quite see the connection between the disagreement on the conversion rate and whether a GC is a poor choice for cannabis analysis. Surely not everyone can afford the equipment used in the paper.

The GC injector is normally set to 200C, and the sample passes through this temperature for only a second or two. The way I do my test is to start with a Restek THCA standard at .1%. I derivitize that standard first to make sure its 100% THCA and has not already partially decarbed.
Then I take two equal aliquots of THCA and inject one into the GC and get a peak ( the GC will decarb it to d9THC ). I take the other aliquot and externally decarb it. To do that I first evaporate the methanol it is dissolved in and then heat the residue at some temperature and time. The paper gives some guidance as have many other papers which generally suggest that 150-160 is best, but too long results in losses. I then re-constitute the decarb to the original volume with clean methanol and inject into the GC. If CBN is produced by too much decarb heat, I will see a CBN peak. If I derivitize the decarb I can tell if there is any un-decarbed THCA remaining.

In the past when I have followed this procedure the peak resulting from the decarbed aliquot has been the same size as the non-decarbed aliquot suggesting that the GC decarb is the same as the external decarb.

If you could suggest any improvement to my method I would be grateful as this is a subject which comes up regularly in my work.

200c is certainly hot enough to cause degradation. 150c is also hot enough for degradation to occur.

im not exactly sure about improvement on this particular method. however, to get a clear picture of this set of reactions, a GC is not the tool. it can be calibrated, however, in my opinion, to be of great use for this industry.

decarb samples. track via hplc. that way you will be able to see at what point you begin to make d8 and/or cbn. abide by that data rather than what a gc can tell you.

The GC can easily see the d8 and CBN peaks.
In the past I have seen the CBN peak increase with over-decarbing, but I have not see any increase in d8.

if youre shooting isolate and getting .68, you cooked your sample. i would again recommend playing with an hplc, some form of thca, and a hot plate. time and temp dependent on amount, form of thca, and method of heat. we do this daily. if i decarbed the oil and lost 1/4 of it my boss would be asking an awful lot of questions!

1 Like

Thanks for putting up with my questions.
When you start your process with the THCA oil, do you actually measure it with your HPLC to know that the oil is 100% THCA and not a mix of THCA and d9THC. When you do measure it does the result from the HPLC actually give you exactly 100% or is it 90-110% THCA with no d9THC detected.