We have gotten our potency testing down and calculations measuring correctly. Our biggest hurdle now is sample prep. How does everyone sample prep the following?
Crude
Distillate
Biomass
Isolate
Currently, we take 40mg of concentrate into a conical tube and add 40mL of HPLC grade methanol into the tube using a bottle-top dispenser. Then we vortex for 1 min, sonicate at room temp for 15 min, and then vortex again. Then we vial the sample up.
If anyone has any insight, it would be great. Our recovery for these samples is either extremely high (100%+) or extremely low (10% lower than normal).
One thing that stands out is that your preparation workflow seems consistent, so if your recoveries are swinging from significantly above 100% to well below expected values, I’d first investigate sources of variability before changing the extraction procedure itself.
A few things I’d look at are:
Sample homogeneity. Crude and distillate can be highly heterogeneous if not thoroughly mixed before weighing. Small differences in cannabinoid distribution can have a large effect on results.
Weighing accuracy. At around 40 mg, even a small weighing error can translate into a significant potency difference.
Complete dissolution. Ensure the sample is fully dissolved with no visible residue adhering to the tube walls or cap. Sticky concentrates can be difficult to transfer quantitatively.
Volumetric accuracy. Verify the bottle-top dispenser is calibrated. A small dispensing error can have a noticeable impact on calculated potency.
Matrix effects. Crude, distillate, isolate, and biomass often behave differently, so one preparation approach may not be suitable for every matrix. Ideally, each should be validated separately.
Recovery studies. If you haven’t already, perform spike-and-recovery experiments using a certified reference material. This is often the quickest way to determine whether the issue lies in sample preparation, extraction efficiency, or the analytical method.
Method validation. Check repeatability by preparing multiple independent replicates from the same batch. If the precision is poor, the issue is likely occurring during sample preparation rather than instrument analysis.
One question: are the high and low recoveries occurring on the same matrix, or do they differ between crude, distillate, biomass, and isolate? Also, are you using HPLC with UV or a different detector? That information would help narrow down whether the problem is related to sample preparation, calibration, or the analytical system.
100uL of sample and 900uL of more methanol into the auto sample vial. Weigh on scale to get dilution factor and enter into program. For flower do 200, 800ul. For strong isolate go to 50, 950 depending on ur cal curve . What is your method and column?
Thanks for the additional information—that helps narrow things down.
Since your dispenser is verified daily and you’re seeing the same behaviour across crude, distillate and isolate, I’d be inclined to look beyond volumetric error.
A few thoughts:
Have you checked whether the detector is approaching its linear range at the concentrations being injected? Small departures from linearity at the upper end can produce the kind of high/low recoveries you’re describing.
I’d also compare replicate preparations from the same sample versus replicate injections from the same vial. If the injections are consistent but the preparations are not, that points toward sample preparation rather than the instrument.
Another area worth investigating is whether complete dissolution is actually being achieved. Some concentrates can leave trace material on the tube walls or cap, leading to inconsistent recoveries.
Regarding an internal standard, that’s always been one of the trickier aspects of cannabinoid HPLC-UV methods because many compounds either co-elute or absorb strongly at similar wavelengths. If you’re aiming for a true internal standard, I’d look for a compound that:
is absent from cannabis matrices,
is chemically stable,
has similar extraction and chromatographic behaviour,
is well resolved from all cannabinoids,
and has a UV response at your detection wavelength without overlapping any analyte peaks.
Some laboratories avoid using an internal standard altogether for potency testing and instead rely on carefully validated external calibration with rigorous system suitability and recovery studies. If you do decide to use an internal standard, it may require some method redevelopment to ensure adequate chromatographic separation.
Out of curiosity, what column chemistry, mobile phase, and wavelength are you using? Also, are your calibration standards and samples prepared in the same solvent and at similar concentrations? Those details might help pinpoint whether this is a chromatographic issue or a sample preparation issue.
Are these diluted the same way every time? As others have said, you’ll need to dilute your isolate, distillate, and crude much more than the biomass.
There’s a goldilocks zone where linearity is true, outside of that high or low you will have strange results. You can guess and check with serial dilutions until you get close to an expected result. Then do a proper sample prep with that dilution in mind.