In a laser gain medium, amplification is associated with the population of an excited state, from which stimulated emission can occur. Even without stimulated emission, the lifetime of this upper-level population is finite due to spontaneous emission and possibly due to additional quenching effects. Typically, the upper-state population decays exponentially with a certain decay time (the upper-state lifetime), assuming the absence of pumping and stimulated emission. More precisely, the decay time is the time after which this population has decayed to 1/e (≈ 37%) of the initial value.
Spontaneous emission leads to fluorescence, the lifetime of which (fluorescence lifetime) is of course identical to the upper-state lifetime.
The decay rate is enhanced if there are additional transitions to other lower-lying energy levels. In addition, the upper-state lifetime can be reduced by quenching processes, involving e.g. the deexcitation at impurities or crystal defects, or energy transfers between different laser ions. The decay of the upper-state population is then not necessarily of exponential nature; non-exponential decay is frequently observed (particularly for measurements with sufficiently short pump pulses). For example, some quenching processes lead to a fast decay, as long as the upper-level population is high, but have little influence on later stages of the decay.
For some gain media such as Cr:forsterite and Cr:YAG, the upper-state lifetime is strongly temperature-dependent. The reason can be phonon-assisted non-radiative relaxation, which becomes stronger at higher temperatures.
An effective upper-state lifetime can be defined under lasing conditions, which includes the effect of stimulated emission. For a four-level laser medium, the effective upper-state lifetime is reduced e.g. by a factor of 2 when the laser is pumped twice above the threshold pump power.
Note that the lower laser level can also have a finite lifetime, the so-called lower-state lifetime.
The upper-state lifetimes of different kinds of laser gain media differ considerably:
- Assuming the existence of dipole-allowed transitions, excited levels of atoms or ions typically have lifetimes of the order of nanoseconds.
- The lifetime of carriers in the conduction band of a direct bandgap semiconductor (as used for semiconductor lasers) is also typically a few nanoseconds.
- Rare-earth-doped gain media typically operate on weakly allowed transitions, leading to much longer lifetimes between a few microseconds (e.g. for titanium–sapphire) and ≈ 8–10 milliseconds (e.g. for erbium-doped fiber amplifiers). Their upper laser levels are called metastable.
Note that the threshold powers for different gain media vary much less than the upper-state lifetimes do, since long upper-state lifetimes imply low emission cross sections, and the threshold power depends on the σ−τ product (see below).
Importance for Lasers
A long upper-state lifetime in a laser gain medium means that a significant population inversion can be maintained with a relatively low pump power. The gain efficiency and thus also the threshold pump power of a laser also depend on the emission cross section (apart from other factors); the threshold pump power is inversely proportional to the product of upper-state lifetime and emission cross section (called the σ−τ product), at least for four-level lasers.
A long upper-state lifetime is desirable for continuously pumped Q-switched lasers, because it makes it possible to store large amounts of energy.
Measurement of the Upper-state Lifetime
The upper-state lifetime can be measured e.g. by populating the upper laser level with a short laser pulse and monitoring the decay of the fluorescence. (Chopping a continuous-wave beam is also an option, but the switching is then much slower, and the level of population achievable may be smaller.) Note that in highly doped media the measured upper-state lifetime may be increased by reabsorption of the fluorescence, particularly if radiation trapping due to total internal reflection at the surfaces of the medium enhances this effect [1, 2]. Reabsorption effects can be suppressed by using a lightly doped powder of the substance immersed in a liquid with comparable refractive index, or by dominantly recording fluorescence from some edge of a sample, using a pinhole (pinhole method). Note that the fluorescence decay becomes non-exponential in situations with significant reabsorption. Non-exponential decay can also result from various other conditions.
|||D. S. Sumida and T. Y. Fan, “Effect of radiation trapping on fluorescence lifetime and emission cross section measurements in solid-state laser media”, Opt. Lett. 19 (17), 1343 (1994)|
|||H. Kühn et al., “Model for the calculation of radiation trapping and description of the pinhole method”, Opt. Lett. 32 (13), 1908 (2007)|
|||I. G. Kisialiou, “Free of reabsorption upper-state lifetime measurements by the method of transient gratings”, Appl. Opt. 51 (22), 5458 (2012)|
If you like this article, share it with your friends and colleagues, e.g. via social media: