Don’t you think it’s totally flawed that no one calculates percent yield (correctly)??
Mass yields are very misleading imo
I know it’s extraction but I think everyone would benefit from uniform terminology
3 Likes
Nope, I reckon y’all should just follow the cadabinoids bro…
2 Likes
That is what I’m saying
Still saying potency tests are the most important thing when calculating yields.
I just wish we could report %yield as a way to accurately identify the extraction method efficacy
I like testing post-extracted biomass. Seems like the labs are less likely to inflate numbers when they know the biomass is already supposed to be stripped.
5 Likes
…and I’m a broken record on that subject.
thanks for the do wop.
it would be more useful if you included that oh so important potency in your examples…
edit: and now I’m gonna fix that spelling mistake…
1 Like
I’m sorry I didn’t specify that potency tests are needed when calculating yields. I believed that part was not being challenged and you are indeed correct
Just making a comment on the efficacy of extraction methods and wishing that percent yield would be used and reported in extraction discussions as it gives a more accurate representation of the process in some cases
2 Likes
totally agree that measuring ones extraction prowess with raw yield numbers is brain dead.
in fact I’ve been beating folks over the head with the concept since I got here.
damn straight.
3 Likes
I join your crusade but let’s leave religion out😉
Tee shirts are a must
4 Likes
In-house GC-FID is the only way. Latency from testing labs is too high to integrate yield determinations at a pilot or commercial scale.
3 Likes
I have unfortunately found this to be very true as well
1 Like
There are other options…
…and yeah, without in house potency (at a bare min) it’s not only hard to schedule, but also too expensive to achieve.
6 Likes
Now that we’ve got an in-house HPLC I wouldn’t go back to GC-FID as our only real analytics.
A 310MM is absolutely the best tool for “Are there cannabinoids in this? Approximately how many?”
If you’re a cultivator, or even most extractors, it’s likely the best tool for the job.
But to my mind, at about the same price as a new 310MM, a used agilent 1100 with an autosampler is absolutely the way to go. Load up 99 samples at the end of the day and it tests them and spits out the results and puts itself to sleep by morning? Fuck yeah.
If you’re doing real R&D and method development, it’s really hard to beat. I’ve got better things to do than sit in front of the SRI and bang samples into it every 7 minutes.
5 Likes
Ive seen this… before…, somewhere…
Totally agree about the auto sample/inject. It’s also good for assuring uniform sample injections which matter a lot when quantifying purity
Both GC and LC have their specific usefulnesses. I tend to think of the Sri GC as a Toyota truck with no power windows, seats, or video camera- great for the person that knows how to drive and to do things without automation. But it takes time and a lot of precision of sample prep and injection to get accurate quantifiable numbers. I tend to use Gc more qualitatively
I agree that an HpLC is great for cannabinoids and I love it. … but I hate that no one has their equipment and sample prep in a hood or proper ventilation. I see these jammed in a closet sized office with no ventilation more than not. Acetonitrile should be used in a hood. If you can smell it then you are doing harm to your body
3 Likes
But it smells so goooood!
2 Likes
The 1100 is a fantastic instrument. HPLC is wonderful for small molecule drug analysis. I spent a large chunk of my chemist career working HPLC. I left chemistry as a professional after nearly 10 years. Doing software now and I want to create web based dashboards using the report data generated by chromatography software. For chemstation you could export to csv, which is especially handy. If anyone could provide a couple CSVs of report data from samples analysis in chemstation let me know
There is a super handy R package called ChromatographR that will process .UV and .ch files, run various peak finding algorithms, and other wizardry. I’ve started using it to put together an analysis system that is substantially easier than chemstation to get data out of, and to produce nice graphs for in-house COAs which will integrate into our manufacturing platform. I personally find it far more useful than the CSV exports from chemstation, which we only really rely on for doing automated collection and correlation of the results of our calibration runs.
I’m hoping to eventually have R do automated analysis of each sample as it comes off of the HPLC, and build up a sample library of the various compounds and contaminants that can be in our process streams, so it can notify us immediately when something isn’t behaving as expected.
For example: Sample X has Isolate in the name, but a peak identified at 287.5nm suggests that there is a detectable amount of contaminant Y in the sample. Suggest applying process Z to mitigate and enhance product purity.
4 Likes
Ha, I just found a python package that does the same thing!
I believe the R package borrowed heavily from one of the python ones. I prefer R for pure math stuff - mostly because I’ve spent way too much time trying to coax pandas to accept that I only have 64gb of RAM and it needs to just deal with it - though python does have its place.