I try to quantify paraquat/diquat in potato using LC/MS similar to the attached method but using different LC column. I use standard in solvent with internal standard (IS) to correct for matrix effect (paraquat has enhancement). The issue is the response of the analytes and IS is not consistent in sample matrix so the number is way off (sometimes is too low, sometimes is too high). I injected the same blank matrix with analytes 5 times and response was not the same. The run is isocratic so the matrix of the first injection may show up in the second or even the third injection. The paper use matrix-matched standard with IS (which is over kill) and it may be necessary to do it. I e-mailed the authors but so far no response. any comments are welcome.
I added the IS at the end to compensate the matrix effect. The method I cited uses IS at the beginning of extraction which will fix lots of things, even if you drop half of the sample extract on the floor. I will try to add IS at the beginning and pretty sure, it will solve the inconsistency issue.
chromatogram showed no clear separation. new solvent system may be tried. if you are using autosampler then response should be close and if manual injection response may vary. Try to mix both STDs and then run the sample to see their separation and RT.
No I use a different HPLC column. I used a mixed phase mode Thermo acclaim Q1 not the HILIC mode. I use this column for glyphosate, 2,4-D and another polar pesticides not retain on C18. This way, one column is for everything, just change the mobile phase. The Q1 is designed for paraquat already (google Acclaim Q1 and paraqut). it gives nice separation (symetrical and basline resolution). Chromatography is not an issue. It is the response of analyte in matrix that is not consistent. IS should take care volume error, extraction error, injector error and matrix effect. Next, I will make a gradient elution to clean the column after each run. After submitting the question, I did the experiment by spiking native and IS just before extraction, but the recovery is still not acceptable. May need to use more cleanup steps and use the grdient the cleanu the column after each injection.
It seems like a good method. I would use class A polypropylene volumetric flasks and centrifuge tubes for the preparation of my standards, ISTD, and samples though. It's unclear if the authors went to this step. They are using a high-end triple quadrupole MS/MS. Unsure if you are, but I think if you are using an ion trap for quantitative work, you will not have as much success, based on our experiences.
everything is plastics not issue here. I use the same instrument 5500 Q-Trap from Sciex
Then its fine error may be due to what you are telling like clean-up,volume, extraction etc.
rana
I finished the work a while back. Please see the list of my publication. Bottom line, need to add IS just before extraction to compensate for the absorption of anlaytes by matrix at low concentration.
@Heydebreck I will check on this but the final decision for what I remember, you have to spike at just before extraction because at the low level, I found that the matrix will absorb a finite amount of analyte and it will be corrected by adding standard at the beginning. Most of the time, I prefer to add IS just before analysis to correct for only matrix suppression and save money. Most of the method on internet (even QuPPe) prefer to add at just before extraction. By adding IS just before extraction will disguise the poor extraction method and if you have poor recovery, you need to fix it and not using IS to compensate.
- Badges
- Science topic