Plasma formation in diode pumped alkali lasers sustained in Cs

Abstract

In diode pumped alkali lasers (DPALs), lasing action occurs on the resonant lines of alkali atoms following pumping by broadband semiconductor lasers. The goal is to convert the efficient but usually poor optical quality of inexpensive diode lasers into the high optical quality of atomic vapor lasers. Resonant excitation of alkali vapor leads to plasma formation through the excitation transfer from the 2P states to upper lying states, which then are photoionized by the pump and intracavity radiation. A first principles global model was developed to investigate the operation of the He/Cs DPAL system and the consequences of plasma formation on the efficiency of the laser. Over a range of pump powers, cell temperatures, excitation frequency, and mole fraction of the collision mixing agent (N2 or C2H6), we found that sufficient plasma formation can occur that the Cs vapor is depleted. Although N2 is not a favored collisional mixing agent due to large rates of quenching of the 2P states, we found a range of pump parameters where laser oscillation may occur. The poor performance of N2 buffered systems may be explained in part by plasma formation. We found that during the operation of the DPAL system with N2 as the collisional mixing agent, plasma formation is in excess of 1014–1015 cm−3, which can degrade laser output intensity by both depletion of the neutral vapor and electron collisional mixing of the laser levels.

Implementing and constraining higher fidelity kinetics for DPAL models

“…For the usual extent of population inversion, electrons experience net energy gain from collisions with potassium. Inversion between D1 and D2 even gives a noticeable contribution, a fact previously recognized by Markosyan and Kushner [5], and Zatsarinny et al [4]. Even if all the alkali were ionized, energy transfer via elastic collisions with helium still dominate electron-electron energy transfer.…”

“…Wallerestein et al [3] considered an even broader set of processes involving the bound electrons, and emphasized building intuition for the hierarchy of processes based on their timescales, and likely availability of reactants. However, Zatsarinny et al [4], and Markosyan et al [5,6], did consider a non-Maxwell-Boltzmann (non-MB) electron energy distribution through the use of finely-grained lookup tables [7]. We also relax the Maxwell-Boltzmann assumption, but model, and evolve the electron distribution a bit differently, and focus on experimental comparisons.…”