Dear all,
I'm running a reversed-phase HPLC protocol for Irinotecan using standards. I'm getting for each standard very inconsistent peak areas. The peak areas are grouped however in two groups (low/high values). For each group (low or high) the CV% is quite acceptable (5-13%), however the high group average is around 6-8 x times higher the value of the low group average. Blank runs are OK and there is no leakage from the injection port. The same problem persists irrespective of the compound / protocol I'm applying to the column.
I'd like to highlight that I inject manually the column, using a1 20 ul injection loop.
Thank you in advance for your suggestions,
Georgia
This may be caused by the column or mobile phase. If either one is not compatible (not chemically suitable) for the analyte you are looking for, then you'll have bad calibration. I had experience with this in the past. Never got to know the source of the problem. My analyte of interest was never analyzed by anyone before me using LC/MS analysis, so I was doing method development. I tried 3 different column chemistries and a bunch of mobile phases, the peak shape was inconsistent and the calibration wasn't possible.
Additionally, sending it to analysis for another lab could help. also consider the detector itself. Any maintenance needed ? Which detector is it (DAD ?MS ?).
I had similar issues with another type of chromatography, and I had to troubleshoot it and ended up doing an exhaustive check up of each parts. I would recommend thus to follow a systematic protocol, including recommendations from Mukhtar: playing with the different options one per one (column A/no column/columnB)*(mobile phase A, new mobile phase A, mobile phase B, etc.)*injection system*detector. Once you know the source of the issues (or at least, what is NOT the issue) it is a big step forward to find a solution (if there is ^^' ).
Good luck !
As is so often the case with these types of questions, there is no way to provide you with any specific advice unless we first have all of the key analysis details (some of which require being on-site to see your specific instrument and methods). Telling us you use a 20ul loop and manually inject samples is useful, but what about the other 40+ parameters not mentioned. Leaving out this info tells us that you may be using HPLC for the first time so seeking out the advice from someone local to you is going to be the best advice of all.
BTW: Poor RSD in a specific range of values usually hints at problems with: solubility, sample or mobile phase precipitation, poor injector technique, invalid retention (K prime) or resolution (integration or separation invalid), lack of a loading study or column/detector fouling/saturation.
The only types of suggestions we can provide on a forum like this one are general ones. Problems of poor reproducibility are often traced to training issues, not hardware. It is critical that the method being used is selective and suitable for the samples tested. The method must be appropriate for the range of samples tested (concentrations expected) and the type or types of sample matrix present (if applicable). Manual injection also adds additional variables (which should be evaluated with replicates). There really are too many things to list here. As the list is long, here is a link to a free article which may help you troubleshoot the issue (it will provide you with a place to start).
- "HPLC Retention Time Drift, Change, Variability or Poor Reproducibility. Common Reasons for it"; https://hplctips.blogspot.com/2015/11/hplc-retention-time-drift-change.html
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running