When just looking at peaks does thc co-elute with CBD or just thc isomers with a gc fid?
Gc-fid reads basic stuff easily (major cannabinoids). When you’re starting to deal with some isomers it loses flexibility compared to an HPLC
Yes. GC-FID is the most straightforward for interpreting the data, the simplest.
It give you a “carbon number”, which is quite proportional to the molecular wight of the compounds were are looking at. This makes calibration also more straightforward.
It does however not tell you what the compound is.
Only the retention time (provided you have a reference for matching it) gives you an indication. Against there, it is not a straight an unambiguous identification, since other compound(s) could have the same retention times. But if you now your samples, and using a well set method, this is not an issue. Becomes an issue only with unknown samples.
This is independant on the the chromatographic method (GC vs HPLC) and detection method. This only depends on the type/length of column you use, the eluent, on the temperature (or solvent, pH etc…) program, and flows which are employed. The goal in setting a consistent method is too have a sufficient invidiual separation of the analyte(s) of interest from other componants.
What advantage(s) would HPLC bring in such situation ?
2 Likes
I feel like with GC changing your ramp and gas rate is similar to LC changing your flow rate. I know when working with alcohol isomers that I was able to differentiate with GC, but I had to really juggle the ramp and flow rates. And even then - the difference between integration with FID vs MS was pretty stark. Especially when I had hard time getting peak separation more than say 0.1 or so.
With HPLC, I have always been able to slow the flow rate way down and/or make the column longer to get to that same place a bit easier.
I don’t know if this is the flexibility that @tweedledew is talking about - but for me being able to switch columns out very easily on HPLC has always been a time saver. I know there are parts for GCs that make it easier - but honestly, waiting a couple hours for your GC oven to cool off so you can change something out was always a pain. And I know they have faster cool off times now, but so many people are using instruments from the 90s or even 80s still. I don’t think its quite the same as a two point screw in connection on a HPLC column.
With cannabinoids I picked HPLC for these things because I could really dig into them and make these changes very easily. I had issues getting different lengths of GC column in the past, but those are really supply issues which could happen for anything I suppose.
I love that we are still having this conversation. I also think it is interesting that the international method for this test has landed on an HPLC method. I figure those guys know something about which method had the least amount of variability and the most commercial viability. But I might also be giving the standard writers too much credit. 
7 Likes
I think you can change your solvent to change your curve separation no?
For the standard cannabinoïd tests (flowers, extract, even synthetic soups), one usually does not run into such troubles. Once a consistent method is set, there is no such need for changing it that often. I would also only change the column when it is too degraded, for a newer one. In case one would need to perform more complicated analysis, there is always the possibility to use dual columns, or connect the outlet to another device such a preparatory hplc, a nmr etc…
5 Likes
Are acid forms of hplc more valuable than residuals,terpenes and other capavilities of gcms?
From a breeders perspective you might wanna hunt for the terps and residuals of cannabinoids and trace cbx.
But if your a vape pen producer or maybe a GMP facility for medicine production the terps are unknown ti your production(not sought for) and the most important thing us decarbing ur material…
I ran a hplc for a year as an peon and with no chemical knowledge the doctor there was able to teach me to be selfsustaining from the mobile phase mix to the chromatogram integration.
There was a GCMS there,the sample prep was identical but the knowledge behind to interpret results was deeper and needed much more study that I couldnt perform.
But as I hear cost are significantly different and permits are different cause with HPLC ur workibg with diffwrent solvents and GC acually (beside the sample prep)ddoesnt need solvents to run.
Down time between collon changes for residuals are a drag though cause it needs an obscene time for vacum restoration/cd.
But we now have GC method that is from 1-3% difference in results from an ISO standard HPLC lab and some are acutally 0.05%-0.2% difference. Which in the analytics sphere is very high confidence, considering the complexity of a cannabis sample product of our choices.
So when dialing the meta down have someone with a decade of experience in the analytics sector and you will get excelent results.
GLHF
in my experience it’s NOT waiting that causes pain 
the fans will get the oven and column down pretty fast (mins). it’s the connection to the FID that stays hot…
4 Likes
How does a gc fid loose flexibility to identify isomers over an hplc? Can’t one adjust/ fine tune a column and methodology on a gc fid in a similar fashion to a hplc to dial in isomers and peaks/ resolution? Can these unknown peaks be defined by using a 3rd party gc ms to specify cannabinoids for any questionable co-elution?
there are only so many mobile phases one can use on a GC. Seems like there are fewer stationary phases as well, not sure they’re any more or less expensive, but I would have to agree with @cassin that they’re a hair more difficult to swap out in the GC’s vs HPLC I’ve interacted with.
I don’t know that @tweedledew has gotten to run an HPLC yet, and I haven’t had the opportunity to run cannabis based analytes…but the “flexibility” in mobile phases would be the first place I’d point.
That and the added information that spectral data gives over “yep, that caught fire!”
D10 was misidentified by many labs before someone stood up and said “hey, that can’t possibly be CBC, the absorbance spectra is all wrong”. Which would not have been readily obvious if the detector had been an FID. (except CBC and D10 don’t co-elute in the common GC-FID methods).
3 Likes
Yeah there’s definitely something co-eluting with my cbd, I’ve got a couple sources telling me it’s d10. I think there’s some column options that’ll fix it combined with lower temperature ramp.
I think I’d still opt for the SRI system, simply because I can call them at any time, and they’re all quite smart and helpful. I talked to them about my co-eluting problem and they’re actively trying to get a d10 sample to fix it. So you end up with some real smart folks on your team (that have been making gcs for 40 years) solving your problems for you. Which I don’t think you’d get from most chromatograph suppliers.
5 Likes
They are pretty much the same. Much of your accuracy is going to be contingent on your sample
prep and the error of your pipettes/balances
The advantage of UPLC is you can use columns with a smaller column void volume with smaller particle size with reduced your mobile phase consumption. Of course the backpressure goes up but UPLC can handle it
3 Likes
The Agilent Intuvo 9000 makes GC column swapping super easy. No more cutting columns.
1 Like
Yeah, been jealous of that for some time. I swore off cutting glass columns at one point…pretty sure I’ve made a couple of attempts since then, don’t think I won.
1 Like
Right on. Yea I wish I actually owned some analytical equipment myself. One day hopefully
1 Like
SRI has just released a new GC configuration which is designed especially for measuring cannabinoids in nasty matrixes like MCT, olive oils and butters and gummies. No special sample prep is required.
It works by catching the heavy oils on a short, cheap precolumn and backflushing the heavy oils out to a carbon trap.
Hugh
SRI
3 Likes
D10 THC elutes between D8 and D9 THC on the SRI GC.
See this chromatogram.
3 Likes
So I have a question about this.
We have seen CBC on GCFID in isomerized products…
Could it be that CBC could be missindentified as d10 on GCFID… I also work with smart folksand they tell me thats a HPLC problem…
But we have no way of comfirming cause were on the hunt for d10 thats been confirmed.
But as I undestand CBC can be synthesized via a CBG precursor so a CBD cyclization (acidic reactions) shouldnt yield CBC…
Were actually setting up an LC-MS, to have more reach in the analytics of pesticides and flavonoids and ubknown cannabinoids…
But for a year now the GC-MS/FID was excelent with CBX-TRP even recogbizing mystery peaks with molar masses that we (sadly) have no way of indentifiying…with LOQ 0.005 and LOD 0.03.
For example a molar mass of 314 has at leas 10+reprezentatives and to buy all these standards we would have to contacr no other than his unreachable exclenency Mahmoud ElSohly…
If you have good reference standards, and you think they are co-eluting, just slow it down. You’ll know really quick if that’s really the issue. And you could also change your ramp, since the literature implies their boiling points are different. And you could always increase your number of identifying ions - just cause they both have 231 doesn’t mean there are not others you can pinpoint as special. Look for the others that differentiate if you think you are having an issue with this. You should also be able to tell the difference from how the peak looks when integrated.
Its not an HPLC problem…if you are not using an HPLC. Just cause the chromatography is similar does not mean it is the same.
I feel like with all the resonance structures being generated during this process somehow landing with a tadbit of CBC wouldn’t be impossible. Its just a flip to the left while that bond is broken - and an attachment at another partial negative. I don’t know if it would want to go there - perhaps not. But I’ve seen weirder things happen…if people were all doing this solid state and making sure only one turn was available instead of many… but then people would have better control in general AND less unknown AND more control over the d9THC on the finished product.
I haven’t read about anyone doing it that way though. -shrug-
3 Likes