DBR‐free semiconductor disc laser on SiC heatspreader emitting 10.1 W at 1007 nm

Abstract

We report a distributed Bragg reflector-free semiconductor disc laser which emits 10 W continuous wave output power at a wavelength of 1007 nm when pumped with 40 W at 808 nm, focused into a 230 μm diameter spot on the gain chip. By introducing a birefringent filter plate in the laser cavity the wavelength could be tuned from 995 to 1020 nm. The laser consisted of a gain chip located at the beam waist of a linear concentric resonator with an output coupling of 2.15%. The gain chip consists of a 1.574-μm-thick resonant periodic gain structure, with ten In 0.13 Ga 0.87 As quantum wells embedded in strain-compensating GaAs 0.94 P 0.06 barrier layers, van der Waals bonded to a silicon carbide intra-cavity heat spreader.Introduction: Semiconductor disc laser (SDLs), also called vertical external cavity surface emitting lasers [1,2], are a flexible semiconductor laser technology which provides high beam quality and high power at a range of wavelengths [3][4][5]. The use of optical pumping allows optimised design of the semiconductor multi-layer stack for optical properties and the external cavity allows the use of intra-cavity elements such as birefringent filters, frequency doubling crystals and semiconductor saturable absorber mirrors [6]. However, the distributed Bragg reflector (DBR) in SDLs adds significantly to the growth time and complexity and limits the thermal performance. In contrast, recently introduced DBR-free SDL technology [7][8][9] offers a potential solution to these limitations. DBR-free SDLs on diamond, emitting 6 W continuous wave (CW) output power at 1055 nm have been demonstrated in [8].Here, we report a DBR-free SDL emitting 10.1 W continuous wave output power at 1007 nm with an incident pump power of 40 W using a silicon carbide (SiC) intra-cavity heat spreader. We demonstrate wavelength tuning, reaching a maximum power of 0.7 W between 995 and 1020 nm by introducing birefringent filter into the cavity.

Thermal Behavior and Power Scaling Potential of Membrane External-Cavity Surface-Emitting Lasers (MECSELs)

“…As it can be seen in Fig. 2, these advantages are remarkable and have pointed out to the possibility of using more affordable heat spreaders with lower thermal conductivity, such as SiC [8], [9], yet enabling power scaling to watt-levels [10], [11].…”

16 W DBR‐free membrane semiconductor disk laser with dual‐SiC heatspreader

“…For lasing wavelengths near 1 µm, the single heatspreader geometry has demonstrated an output power of 6 W when directly bonded to diamond [8]. Recently, using a single SiC heatspreader, Mirkhanov et al [9] reported 6 W of output power at a coolant temperature of 12°C, and 10.1 W at −10°C, while dual-diamond-heatspreader DBR-free SDLs have yielded 3.5 W, thus far limited solely by the available pump power [10]. In this paper, we directly compare the laser performance of DBR-free SDLs using these two geometries.…”

Semiconductor waveguide quantum well lasers

“…The membrane is grown on a GaAs substrate, and between the substrate and the first capping layer is 200 nm of AlAs which acts as an etch stop. The sample is the same geometry as the one used in [17]. In preparing the membrane laser sample, the QW side of a sample chip is adhered to a piece of silicon wafer.…”

“…MQWLs exhibit a number of advantages, some of which are similar to other epitaxially grown semiconductor membranes [15,16,17,18]: (i) the thin growth with released stresses exhibits excellent crystallinity, and hence reduced non-saturable losses; (ii) MQWLs can be grown in about a tenth of the time that is needed to grow a VECSEL or polariton laser, and therefore are an ideal platform to investigate new functionalities; (iii) MQWLs are an ideal geometry in order to extract heat by contact bonding to sapphire or SiC, and can therefore be highly efficient lasers; (iv) MQWL growth is freed from the material (lattice matching) constraints of DBR mirror growth, which can adversely affect the quality of growth (a typically observed characteristic of VECSELs and some polariton lasers); and therefore (v) MQWLs can be designed in a wider selection of design wavelengths as the lack of DBR liberates the development of lasers at wavelengths where appropriate DBR material index contrast is not possible, such as the all-important 1.5 µm telecoms wavelength region; and finally, (vi) because of the high index contrast between the membrane material system and the substrate (silicon carbide, silica etc. ), strong waveguiding is formed which ensures a high overlap of the guided mode with the QW gain region.…”