AnsweredAssumed Answered

Cannabis testing lab - chemstation issues with concentrate baseline & calcs

Question asked by jamartellini on Jul 25, 2019
Latest reply on Aug 14, 2019 by kbusch

Recent run with a previously tested R&D concentrate sample, sandwiched between 2 blanks

We have agilent 1100 series (HPLC with vacuum degasser, binary pump, autosampler, thermostated column compartment and MWD).  UV & Vis were on, along with the pump, 30 mins before the run.  Baseline was perfectly balanced at zero for several minutes before starting.


No air bubbles (we have filters), the blank and sample are both HPLC grade methanol.  We syringe filter the samples. 

The detector autobalance is always on for the prerun.

Has anyone seen this happen before, and know why?  We didn't change the established protocol, and this particular sample had been tested numerous times previously for R&D (I weighed and prepped it fresh before the run).


We are testing Cannabis concentrate, and trying to calculate %THC using the ESTD% method.  Our protocol is set to use 0.1 g sample, dilute at 1:100, and add the exact test weight in mg to the sequence column (we don't add the multiplier or dilution, because they cancel each other out 0.01 & 1:100).  Our peak areas have been huge for just concentrate this past week.  Flower and edibles test and calculate perfectly normal using the same saved method, as long as they are not in the same batch being run with the concentrate (when I ran flower samples with the concentrate batch, the calculations and peak areas were also crazy). 

The chemstation is all of a sudden calculating THC content well over 100% only for concentrates, and we can't figure out why/how.  All blanks are clear, and all internal standards look perfectly normal.  We expanded our calibrations, to try and account for more peak area, but the calculations are still crazy!  Why is it working and calculating flower perfectly normal on one run, and concentrate doesn't make sense on another?  They both use the same prep (flower test weight is 0.25 g, concentrate is 0.1 g), diluents, and calculations.  Both are diluted 1:100, and we enter the test weight, dilution factor and multiplier (1/test weight in mg) into the sequence table. 


Pic 1 below: Our normal looking cannabis flower (THC-A is in flower, since it is not converted to THC-A until it is decarbed with heat)

Pic 2 below: Baseline normal.  We tried running with a lower test weight, and it helped the peak area, but calculations are still wrong.  Concentrate has been decarbed, so only has THC.

Pic 3 below: Normal test weight, with different dilution.  The baseline dipped below zero, during the run. 


Troubleshooting done: Ran concentrate at differet dilutions and test weights.  Made new mobile phase, flushed system, washed system, recalibrated with expanded calibration level, tested all tips, pipettes, and diluents for any THC contamination, did intensity/cell/dark current tests on detector (all passed).


Any advice would be greatly appreciated!!  We just opened for business a month ago, and I have only been training on this software for a few weeks, but concentrates will be a significant portion of our business.  I did peptide purification, with reverse phase HPLC during my PhD, so I am studying as much as I can in the manuals, but this problem has so far perplexed everyone I have turned to, and I have clients waiting for results.  If you have the golden ticket, drinks are on me!!

Cannabis flower - looks normal (same prep, method, and calcs)Measuring a lower test weight, helped with peak area, but calculations still unusualDifferent dilutions of prepped concentrate, with normal test weight, have same calculation issue