Chromatography Forum

Hello everyone,
I'm validating an HPLC method for a pharmaceutical sample and have encountered a strange situation. During one of the precision runs, a small unknown peak appeared at approximately 5.2 minutes. This unknown was not present in the blank or in a few injections I ran previously. I also make all diluents fresh and mobile phases fresh, and I flushed the column before starting the sequence.

At this point I have ruled out carryover and injection error. I am curious if this is maybe a degradation product or possibly related to the column itself. I am curious if anyone else has experienced this seemingly unknown peak during their method validation? What steps did you take to try and isolate the cause of such peaks?

It was interesting to me this topic came up as I was brushing up on some data analysis fundamentals in a generative ai training and thought to myself if there is any advantage to AI-based chromatogram pattern recognition for identifying such issues. Has anyone gone down this route yet?

I would appreciate any suggestions or literature references relating to how to proceed with consideration of an unknown peak, because I want to feel confident I have a robust method before I submit it.

I appreciate your time and support!
henry

Try rinsing your sample vials with a little sample filtrate before filling them.
Some vial vendors do not do as good a job of cleaning them as they would have you believe.

(1) At this time "AI" is based on mass information obtained from the web, which includes more mis-information than factual information so would be unwise to use (AI is currently as error prone as most of the untrained chromatographers on this and other sites). Stupid data in equals stupid data out. AI only reports popular responses or trends which is why it is so important to have actual knowledge to test and check on the responses.
(2) Please consider posting HPLC questions under the HPLC section, not here. This is the Capillary Electrophoresis group.
(3) You provided no HPLC method information of any kind so the only suggestions possible are speculative. "Unknown" peaks require investigation and often indicate use of poor quality method. It could be contamination from sample, sample prep, poor maintenance of HPLC (e.g. change the injector's rotary valve seal or replace the vacuum degasser with a new one). The instrument settings could be inappropriate for the sample or method (i.e. The Reference Wavelength may have been turned 'ON'. *This should always be OFF.

Note: "validated" means nothing but documented. It does not mean the method follows good chromatography practices. Documenting is not the same as understanding. Just look at all the published "validated" methods in the journals and you will conclude that 15-20% are completely irrational and invalid (These statements will be understood by someone with ten plus years of industrial HPLC experience).