Modeling the effect of analyte and reference bandwidths on signal and noise magnitudes in spectrophotometric assays

In spectrophotometric assays, it has been well established that the recorded absorption, and therefore the experimentally determined extinction coefficient, decreases as a function of detected bandwidth. This manuscript presents an expression for the extinction coefficient as a function of the critical parameter detected bandwidth per transition linewidth. Calculations for both single channel and multichannel photodetection are presented; the derived expressions are shown to be in good agreement with experimental results. It is important to realize that this systematic bias is present in dilute solutions of low absorptivity, and the experimentally recorded extinction coefficient for a molecular standard such as caffeine can vary approximately 4% or more, depending upon choice of research instrumentation. The magnitude of this bias may be sufficient to effect method robustness, cause interlaboratory discrepancies, and fail system suitability requirements for spectrophotometric assays. The signal to noise ratio, for example as analyzed in HPLC/UV-VIS detected chromatograms, is also a function of the detected bandwidths of both the analyte and reference channels. It is shown here that use of a reference can only increase the baseline noise.