Stuff like this was my job for over 4 decades!
We also assayed methyl paraben by capillary GC after DMF extraction and BSTFA derivatization, but sounds like you do not want to go that route. We've also assayed methyl salicylate by GC as well.
I assume that you're using a detection wavelength of about 280nm, and that you've run the assay out more minutes to demonstrate that your unknown contribution is simply not from a previous injection. And I assume that you have an acidic mobile phase, acidified with some acetic acid, phosphoric acid, or a buffer.
I also assume that you have reduced the injection volume and the co-elution is the same.
So - the easiest thing would be to make standard mix of methyl salicylate and methyl paraben and vary HPLC conditions, start with one parameter at a time, and make note of whether the separation gets better, worse, or stays the same. Maybe the easiest thing to try is to change the temperature of the column by 10C or 20C.
If a temperature change seems to help, investigate mobile phase changes, like more or less organic, changing organic from methanol to ACN or HPLC ethanol (don't know what you're using now), or even adding a wee bit of non-UV THF to the mix, can do wonders.
Flow rate changes won't change things much; if your assay is isocratic, consider doing a gradient.
Trying a different phase column from same or different manufacturer could help.
This is the fun stuff, enjoy the process. This reminds me of a more common assay issue with methyl paraben - close or co-elution with phenoxyethanol - that I solved like 2 decades ago....
Or hire me as consultant!
