Chromatography Forum

I am testing wastewater samples for the presence of RDX and HMX (explosives) on an Agilent 1200 Series with a Diode Array Detector. When running a standard with 1000 ug/L of these two components, my peak heights are at 3mAU for HMX and 2 mAU for RDX at wavelength 250. At wavelength 254, HMX is about 2 mAU and RDX is 1.25 mAU. Are these peaks too small? What is a rough minimum peak height or area for good quantification of unknowns? Can attenuation help to obtain larger peaks, or do I simply need larger injection volumes? I've noticed my peak shape gets worse with too large of an injection volume. Any other ways to get more response for these low concentrations?

The absolute peak height is relatively unimportant because the display can always be scaled up. The catch is that the noise also get's scaled up and what's important is the signal/noise ratio. Lower limit of quantitation ("LOD" or "LLOD") is commonly accepted to be a peak whose height is 10x the peak-to-peak noise.

You mentioned peak shape issues with larger injection volumes - how much are you currently injecting, and what is your sample diluent? What are your column dimensions and mobile phase?

As Tom pointed out there isn't really a value of peak height that is acceptable.
What you really care about is how well you can quantitate. If you don't have
any rigorous compliance issues then just inject it a few times and look at how reproducible
your integrated areas are.

You can also increase your peak heights by going to lower wavelength. At 220 or 230
I expect your peak heights will go up ten fold or more without much more noise.

Per EPA method 8330, I started out with 100uL injection vol. This is a huge injection and the chromatography was a mess. 5 uL seems to be giving me nice peaks. Sample diluent is ACN. Mobile phase is 50/50 MeOH and H2O. My column dimensions are 4.6 x 150mm; 5um.

Go back to the 100 uL injection, but dilute 1:1 or 1:3 with water. Your peak shapes will be much better. At 1:1 H2O:ACN and 100uL injected, you'll be loading 10x more analyte onto your column than with 5uL and 100% ACN.