i just use C18 catriage and used two steps of elution.
first step, use 10% methanol to elute
second step use 100% methanol.
discard the eluent from first step since it has a lot large polysacchride and protein.
and inject sample from second step.
since compounds cause hump baseline goes all over the gradient range, I doubt finer gradient from SPE can help.
any suggestion?
First, I defer to anything Uwe Neue would recommend over what I may postulate. He's the expert, I'm just a guy with a lot of time on my hands. You would do well by your analysis to consult him in optimizing your SPE procedure.
My answer should probably be 'maybe', as I'm not an expert at anything, so take everything I say with a giant grain of salt. I think that optimizing your SPE extraction would be easier than adding a cleanup step (and potentially losing analytes of interest), but if you can't get the SPE optimized, I assume you could inject your analytes of interest into a GPC system, note how they separate on your GPC column, and inject a 'blank' bread extract to see how it separates out. If you can separate the components of the bread extract that are causing the hump from your analytes of interest using GPC, you could collect the fraction in which your analytes elute, evaporate the solvent, then reconstitute in your initial mobile phase conditions.
Perhaps I could start the SPE optimization discussion - What type of SPE are you using? Did you try different proportions of H2O:Methanol as an elution solvent? For example, after the 90:10 H2O:MeOH wash, did you try 50:50 MeOH:H2O, 60:40, 70:30, etc? Did you try more than one SPE?