Qualification of a fluorescence spectrometer for measuring true fluorescence spectra.

New analytical methods using fluorescence detection are becoming increasingly quantitative and require easy-to-use material standards for fluorometer qualification and method validation. NIST is responding to this need by developing and producing such standards. Reported here is the first step in this process, which is to qualify a research-grade fluorescence spectrometer for measuring true fluorescence spectra of reference material candidates. "True" spectra are defined here as those with fluorescence intensity, either relative or absolute as required, and wavelength both being reported with high accuracy and known precision, after wavelength has been calibrated and corrections for excitation intensity and detection system response have been applied. The uncertainties determined in relative and absolute intensity-corrected fluorescence spectra using both calibrated source (CS)- and calibrated detector (CD)-based methods were compared. The CS-based method gave uncertainties, typically about +/-5% for relative spectral correction, that were about half that of the CD-based method for determining both relative and absolute spectral correction factors. Absolute spectral correction factors can be determined using either method without knowing the optical geometry of the instrument. The absolute spectral correction factors were found to have much larger uncertainties than the corresponding relative correction factors with uncertainties for the CS-based method of +/-10% to +/-15% being typical and +/-20% or more not being uncommon, particularly for excitation and emission wavelengths below 400 nm. Uncertainties arising from detection system nonlinearity and instrument polarization ratios were also explored.