I noticed a relationship between flow rate and peak area on one of my samples. As the flow rate increases the area of the active peak decreases linearly. The gradient was from 10% MeCN:90% H2O (with 0.1% H3PO4 and 0.1% TEA) to 36% MeCN over 40 min with detection at 244nm. (It also occurred in earlier methods using 0.1% HClO4 and 0.1% TEA) It's a pretty drastic change in area too. (Area of 3250000 at flow 0.8mL/min changes to 1540000 at 1.6mL/min)

We originally thought that gradient could have changed the pH and that the peaks absorbance maximum shifted, but the change in MeCN isn't large enough to change the pH enough to shift the maximum. Another idea was adsorption to the column, but if that was true then the data would be the other way around (area would be higher at higher flow, and lower at low flow since it would be stuck to the column).

Does anyone have any ideas?

This is the effect of residency time in the detector flow cell. Within the flow cell the volume of liquid traversed by the light path is fixed. If the flow rate is doubled any given molecule (or the entire peak volume depending on how you want to look at it) will be in the light path for half of the previous amount of time. Since peak area has units of absorbance*time the peak area will be halved.

Thanks. That makes sense, but we never saw it occur with other samples...

Tom's explanation in definitely the right one, and you should observe the same phenomenon with each peak in your chromatogram...