Are you not filtering these before the final dilution?
AmenâŚ
why bother? 50 mg conc. fully dissolves into 10 mL methanol by sonication and vortex, then centrifuged. Any particulate matter that isnât soluble in methanol crashes out.
pipetteing seems like something that would be a no brainer, but it really is an acquired skill, and yes absolutely, you can have variance from operator to operator, sample to sample, just because they pipette with different techniques, or lack of techniques.
4 Likes
Prep is as followed:
50 mg into 10 mL methanol; sonicate at 50C for 20 mins. Vortex at 2500 rpm for 10 mins. Centrfiuge at 3000 rpm for a few mins. Dilute 50x (20.4 uL + 1.000 mL into methanol. Vortex in vial for 5-10 secs to mix.
Cal range - 2 - 100 ppm
No we can change the method. The method is fine. Works great if I have to find the value within 20%, 15% and probably 10%. To me, personally, 76%, 80%, 85% is all close. But the client doesnât see it this way. And I can understand that.
1 Like
Itâs not the act of pipetting thatâs the issue. We can verify the validity of everybodyâs pipetting skill by weighing low, mid and high volumes and running stats on multiple data points.
But pipetting water or methanol at certain volumes is not the same as pipetting a 50mg conc + methanol. The residue of that water or methanol on the outside of the tip can easily be wiped off before dispensing. And thatâs what I believe I am missing.
There definitely are many ways and alternatives to fix this problem, but they donât decrease the cost or time necessarily. There needs to be multiple readings in order to get a pretty good grasp on the real value. And I dontâ want to double my work.
1 Like
Out of curiosity, do you flush or attempt to flush the pipette tip with solvent solution after dispensing?
Wondering if that is something I should take care to do
Absolutely do. I have always done it. I do believe some sample residue will be stuck on the inside of the tip and it needs to be washed out.
this is probably the real answer. never thought about it. That would get us a true dilution factor. but a pain, if we did it for all conc. samples
It just takes the guessing/anxiety out for me. At the cost of weighing shit constantly lol. But I saw dramatic improvements after implementing.
But just be aware youâre just translating those potential errors to a scale as well. Just make sure itâs calâd to not end up in the same situation as now.
Is there a reason you centrifuge your samples after vortexing? That would definitely send your heavier compounds to the bottom of the sample tube.
Also wow thatâs a small cal curve range. How many points are you including?
of course. balance is verified from 0.0100 - 1.0000 g with 4 diffferent weights
1 Like
Centrifuge helps crash out wax material that canât dissolve into the methanol. Distillate fully dissolves. As long as my cannabinoids arenât crashing out, which I donât think they are.
Cal curve uses 5 points. Detector saturation and peak distortion begins around 200 ppm for certain analytes
Okay I see your reasoning. I donât think the cannabinoids would crash out but since theyâre heavier than pure methanol, I would think putting it in the centrifuge will drag them downward and defeat all the work you did to vortex them, and might be why itâs more concentrated at the bottom.
You can use a spin filter attached to a syringe and that will help filter out the waxes in your solution. Pull up ~2 mL with a syringe, twist on the filter, and push it into a microcentrifuge tube, then perform your dilution from the little part you pulled out of the sample.
1 Like
Excellent help everyone. Thank you all very much. We will begin running some internal tests on these ideas to check the variability. I do believe this to be the main source of our wide variabilty.
What the range for your cal curve? Have you looked at response factors for your standard levels at higher concentration?
Iâd be interested in knowing what some people are expanding their ranges to based on only a good R^2 value.
This is one reason i prefer microliter syringes instead of autopipettes.
At the last lab I worked at, we were testing THC and CBD from 5 - 500 PPM. It was an 8-point curve (500, 250, 125, 100, 75, 50, 20, 5), and I could usually get a 0.999 R^2 value.
The instrument was a Shimadzu LC-2030, using the method prepared by Shimadzu with their NexLeaf column. Itâs been a few months but I remember the 500 ppm having a response factor around ~800000.
Our third party lab used an Agilent and their curve ran from 0.75 - 200 ppm (Iâm a bit fuzzy on their low end but I know it went up to 200 on the high end), and my numbers were usually within 2-3% of their results, and around 5% RPD.