|MadSci Network: Chemistry|
It depends... A good look would require initially knowing intensity versus wavelength from the near ultraviolet to the near infrared, and following wavelengths of interest or the whole spectrum over time. As your system is emitting light, optimally you would use a fast scanning spectrofluorimeter (without its excitation source). Can you borrow use of university equipment or secure an instrument manufacturer's cooperation? A quick and dirty look would get by with a photodiode and some simple circuitry to monitor its readout. This would only give you averaged readings and you would need carefully control all variables of sample geometry, time, concentation, mixing rate... temperature, which also effect the reading. If the wavelength composition of emission shifts then you need a sensitivity vs wavelength calibration curve for the photodiode. Slightly better than the worst case is to put a narrow band optical filter in front of your diode to get some idea of intensity vs one narrow wavelength band. In any case, you need a (stirred) thermostatted sample cell. Uncle Al!
Try the links in the MadSci Library for more information on Chemistry.