Pr/sup 3+/ Co-Doped 2.8-/spl mu/m Erbium Fiber Laser

Abstract

Tel. +41-31-631 8937,Fax+41-31-631 3 7 6 5With rising power in cw and pulsed operation erbium 3-pm lasers work their way into a variety of applications in surgery Despite the unfavorable ratio of upper-level to lower-level lifetime cw laser opcrattori is achieved in fluoride host materials solely due to the small branching ratio on the laser transitioii 11). CW output powers exceed 1 W in LiYF4 121 and 150 mW in a ZBLAN fiber [3]. Owing to geometr). and erbium concentration the population mechanisms are different m clystals and fibers [4]. Thcsc mechanisms determine different concepts of energy recycling in the mentioned matcrials.In highly doped crystals ground-state absorption (GSA) and interionic processes play a dominant role Excited-state absorption (ESA) is weak, because the excitation density is small compared with the %round-state density [4], Interionic upconversion from the lower laser level recycles energy into the upper lascr Icvel. Energy lost by a second upconversion from the upper laser level is automatically recycled via cross relaxation. The concentration dependence of the interionic processes offers highest efficiency in Er'+( I5 %)'LiYF4 [2]. The experimentally obtained slope cfiaeitcy of 40 % [4] is clearly hi&er than the Stokes limit of 35 % under 973-nm pumping The theoretical limit including energy recycling is 56 % [4]. I n fibers, in contrary, a low dopant concentration and a strong pump confinement in combmation with long Iifetmies of the laser levels lead to ground-state bleaching and a strong excitation of the laser levels. ESA dominates GSA, whereas interionic processes have only a small influence [41. Groundstate bleaching quires the depletion of the lower laser level by pump ESA and repopulation of the ground state via fluorcscences. However, two pump photons are absorbed for the creation of one laser photon, In addition, part of thc ESA excitation is not automaticallv recycled, but leads to competitive lasing at 850 nni and the saturation of 3-pm output power [j] Cascade lasing at 1.7 pm [3] avoids competitive lasing, recyclcs the energy into the upper laser level, and ensures that each pump photon is converted into one laser photon. Thus, the theoretical limit ofthe slope efficiency under 791-nm excitation is enhanced to 27 % 141 which IS only slightly smaller than the Stokes limit of 29 %. Experimentally a slope efficiency of 23 "/. IS obtained [3]. The recycling concept ofcascade lasing may also be applicable to other fiber lasers.;:iw, P 2 LiYF4 71-pm (circles) transition pumped of the at cascade 973 nm laser and the u1 a 0Obtained slope efficiencies versus absorbed I I ZBLAN fiber (squares) pumped at 791 nm 2 40 2oo 4oo 6w pump power are 40 %and 23 %, respcctivelv Absorbed Pump Power [mW] The erbium 2.8-pm ZBLAN fiber laser is a promising candidate for surgical applications. Owing to pump excited-state absorption (ESA) from the lower laser level and subsequent competitive lasing at 850 nm [I] the 2.8-pm laser usually saturates at some 15 mW output power. Recently a cascade-las...

The route toward a diode-pumped 1-W erbium 3-/spl mu/m fiber laser

“…Only lifetime quenching by co-doping depletes the ESA level and repopulates the ground state in one step. But this approach is not sufficient to prevent ESA from having a crucial influence [9]. The proper combination of the possible solutions answers both questions simultaneously and leads to the desired doubleclad diode-pumped high-power device.…”

“…Since double-clad ZBLAN fibers are rather expensive, it is apparent that these ideas are best investigated in computer simulations which conducted experiments so well in recent years [4], [5], [7], [9]. In addition, the understanding of the population dynamics is deepened, system parameters can be optimized, and possible pitfalls can be identified that have to be investigated spectroscopically before performing the main experiment.…”

“…Thermal redistribution within each multiplet is assumed to take place instantaneously. For consistency with [4], [7], and [9], let the 2.7-m laser originate from the second Stark level of the multiplet. The 850-nm laser is assumed to originate from the first Stark level of the multiplet.…”

“…In addition, it increases energy diffusion within the lower laser level and improves lifetime quenching by Pr co-doping [12]. The quenching process weakens ground-state bleaching and population of the ESA level [9]. Detuning the pump wavelength from the GSA peak is substituted by doubleclad pumping which automatically reduces GSA, ground-state bleaching, excited-state population, and ESA.…”

“…The requirement of high pump intensity, however, excludes the cascade regime from diode pumping. Third, the unexpectedly low performance of a different approach, the quenching of the lower laser level by Pr co-doping [8], could be explained by insufficient quenching and the successful competition of ESA with subsequent 850-nm lasing [9]. So far, it has been possible only with ESA to overcome the bottleneck of ground-state bleaching and long excited-state lifetimes.…”