Failing for acetone - An explanation

Some of you might have experienced failure for acetone or other residual solvents.
Sometimes these results just do not make sense. For example, we had laboratories report 5000 ppm acetone (0.5%!!) in terpenes blends made with food grade terpenes. This seemed very unlikely.

Recently our group published a paper that shows that terpenes can degrade into acetone in the analytical equipment under certain conditions. An additional interesting finding was that the presence of cannabinoids can prevent this conversion. Most likely they act as anti-oxidants and prevent this reaction from happening.

Anyway, be aware of this when you fail for acetone and have a high terpene product.

20 Likes

I think that Beta-Pinene also produces acetone over time. Not 100% sure on this, but I had found that in some research several years ago.

2 Likes

Thanks for sharing this. I see you also tested methanol levels with cannabinoids, do you also have the data with terpenes (as done with acetone)?

We did look at the reduction of terpenes in the experiment, but this was not included in the publication as the results for a-Pinene showed an increase, which we did not have a good explanation for. But that result was repeated and we did keep seeing the same thing.

Anyway, attached is the excel file with the data for the terpenes. The vial was heated in the GC oven at the same temp as the autosampler of the headspace injector. This was done under argon for 40 minutes or with air for 40 or 120 minutes. This was compared to an unheated vial. What we saw is that there was less degradation of terpenes in argon atmosphere. Heating time didn’t seem to matter a whole lot for the degradation in air, which probably is because all the oxygen in the sealed vial was already consumed in the first 40 minutes. Terpene Data - Future.xlsx (51.5 KB)

4 Likes

Now of course things like this will depend on the temp of your headspace sampler and the equilibration time, but my conclusion is that GC-Headspace is probably not a good way to analyze terpene content. If you just use liquid extraction and a direct injection you should be able to avoid these reactions.

1 Like

@SWELL

How much did the a-pinene increase (e.g., average % increase)? Is this a-pinene increase statistically significant?

The file is attached in an earlier reply. It is weird, but consistent.
Quantification was with GC-FID and a-Pinene elutes very early.
It is not entirely impossible that the pinene degrades into something with a higher response that overlaps, but that is about all i could come up with.
We did not do a bunch of replicates for the terpene part either because of the long run times and the many variables already investigated. So this limits the statistical interpretation, but the data seems consistent across the various terpenes which I guess you can sort of take as a replicate and show that its not just random variation.

4 Likes

I did a similar experiment myself on my in-house HS-GCMS.

The major acetone producing terpenes are terpinolene, cis-beta ocimene, and a-bisabolol. All the terpenes make acetone but these 3 are off the charts.

5 Likes

I believe this (terpene degradation) can be possible source to the methanol in the Medusa stones head space testing
@Graywolf and @Dukejohnson does this match with your data at all?

1 Like

No, I found methanol in the gas itself. No cannabis was involved at all.

5 Likes

I can verify this with my own experiments as well using HS-GCMS. While acetone is the major product, I’m unsure of the mechanism, terpene degradation produced peaks in the same retention windows and methanol, isopropyl alcohol, ethanol, toluene, and hexane.

What Dukejohnson said. I am not running samples, only reporting results from those on the team that have,

2 Likes

I think that would be terpentine

Anyone seen acetone pop up when testing pure terps samples at confidence?

Like the OP link says, if you had them stored exposed to oxygen your going to get decomp, especially when cannabinoids are removed and especially in terpinolene dom varieties.

1 Like

The opening is talking about acetone concentrations in testing with different atmosphere in the headspace sampling vials

The largest increase in acetone formation was seen for terpinolene which showed an 885% increase in oxygen atmosphere (4603.6 PPM) when compared to analysis under argon (519.9 PPM)

I guess that’s something I can ask them about

5 Likes

Welp, acetone dropped by about a factor of 10 by replacing atmosphere in the headspace sampling vial

3 Likes

Keep in mind that the act of purging will evaporate acetone from the sample.
So if you have acetone and purge it, it will evaporate at least partially during this action.
When we did this experiment, we purged both vials with argon first and then backfilled one with air.
This way it guarantees that what you find was generated by the presence of air and not because the purging removed it. We also ran a blank vial with the same air to exclude that the environmental presence of acetone in the air is what resulted in the detection.

4 Likes

Good stuff, thanks for sharing. Similar to your findings, I’ve observed acetone as high as 9000 ppm when analyzing terpene blends by HS-GC, which drops to 100 ppm under argon instead of air.

2 Likes

We purge the vials with argon right before pipetting the sample into the vial to avoid this. Since argon is denser than air it should stay in there if you move relatively quickly and minimize turbulence.

2 Likes