Dear Helen,

welcome,

When exciting a semiconductor by generating excess electrons and hole in it. These excess electron holes will recombine with time. The decay of the excess electrons and holes follow normally an exponential decay and so does the emitted light intensity or photon flux.

The decay time constant is called the minority carrier lifetime Tau. So it can happen that the minority carrier lifetime is injection dependent such that after reaching certain excess carrier concentration the lifetime changes to another value causing the emitted light to decay with other time constant.

The lifetime of the minority carrier itself depends on the recombination centers existing in the material as heavy metal atoms and crystallographic defects.

Best wishes

Dear Helen,

1. If two or more centres (or levels) with different decay rates are involved in the PL, then you will see two or more exponential decays. For example, at liquid helium temperatures, you can observe simultaneously PL from free excitons and localised excitons, with resulting not necessarily mono-exponential decay time. 

2. In some cases (if relaxation of the excited emitters to the ground state may occur through many competing channels) you can see so called stretched exponential decay, which is also not mono-exponential.

3. If pump rate is too high, i.e. if you are working in so called saturation regime, then again you will observe non-exponential decay of PL even from one centre and to one channel. 

Best wishes

Dear Helen,

in case you measure the photoluminescence of a single excited state-that, e.g., can be an exciton (electron-hole pair), but also an elevated electronic state in an heterostructure etc.-you usually measure only photons within a selected energy range. Therefore, your exponential decay represents the decay of that specific excited state. And all possible "decay ways" that empty your investigated excited state influence your exponential decay curve, whether they stemm from a radiative or non-radiative decay path. All of the decay paths will manifest in your exponential decay curve with their individual probability of emptying your excited state, causing an individual "live-time" and therefore different exponents, visible as slopes when plotting the time-resolved photoluminescence data on a logarithmic scale.

If you have more than one exponential decay, that means you have more than the pure, radiative decay by photons (which you are obviously able to detect). That can be, e.g., a non-radiative decay or another radiative decay into other states.

Best wishes

I just read my answer and found it very misleading. Lett me add for clarification: If you see more than one slope in your logarithmic plot of the transient, you have more than one kind of population that decays. Different decay channels of only one kind of population can not produce more than one slope in your plot.