You get good peaks, but the area's for the standards are not good -- not reproducible, and not scaling proportionally with injection? There can be a number of problems with your system -- hardware, derivitaization procedure, and just how "good" your separation really is. I'd start by checking your system -- running a 50:50 mix of eluents, and injecting something simple, caffeine is a favorite, and should work for your column, and see if your system is pumping properly, and your injector is working ok, and your detector is ok, and there is no internal leak.
'Tho it seems to me, that you may be asking a lot from your system, dissolving sample at a widely different pH than running pH. But if that's the method you have, yoiu may have to work with it.