Before the introduction of picosecond
laser systems, most Time-Correlated Single Photon Counting spectroscopy
systems used coaxial flashlamps
to act as a light source. A wide range of wavelengths
can be produced with flashlamps. These devices generally produce a pulse
of light about 2 nanoseconds
wide. This is about 1000 times wider than a pulse provided by a laser source, and significantly less powerful (intense).
Flashlamp sources first became available in the 1960's and they were incorporated in most TCSPC systems in the 1970's. Because of their low intensity and slow repetition rate, it was necessary to collect data for long periods of time. If the lamp intensity profile (the shape of an individual pulse) varied during the course of the experiment due to drifting of the instrument, one could have artifacts in the data analysis. Lamp technology has improved steadily and today, lamps can provide repetition rates up to 50kHz and pulse widths around 1 ns. Compared to laser sources, flashlamps are simple to operate and easy to maintain. There have been considerable efforts to obtain even shorter pulses and faster repetition rates. Flash lamps sometimes have a trailing intensity signal as the lamp cools down which can also complicate the data analysis.
The spectral output of the lamp (the wavelengths it can produce) depend both on the type of gas the lamp uses and the pressure inside it. Hydrogen and deuterium lamps can provide light in the ultraviolet (UV) range, but at a low intensity. Nitrogen lamps provide a high intensity at a few select wavelengths in the UV , but not much power at at values between these. Some flashlamps are tuned to the red and near infrared (NIR) wavelengths.
Source: Fluorescence Spectroscopy - Joseph R. Lakowicz
Solid state lasers