Skip to content

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

MetadataDetails
Publication Date2022-01-28
JournalIEEE Journal of Quantum Electronics
AuthorsHoy-My Phung, Philipp Tatar-Mathes, Aaron Rogers, Patrik Rajala, Sanna Ranta
InstitutionsTampere University
Citations11
AnalysisFull AI Review Included
  • Core Value Proposition: Membrane External-Cavity Surface-Emitting Lasers (MECSELs) enable superior power scaling through double-side cooling (DSC) and thin gain structures, mitigating thermal rollover limits inherent to traditional VECSELs.
  • Thermal Management Achievement: DSC reduces the maximum temperature rise by a factor of ~2 when using high-conductivity SiC or Diamond heat spreaders, and by a factor of up to ~4 when using thermally inferior Sapphire spreaders.
  • Power Scaling Limits: The maximum pump beam diameter for efficient cooling is limited by the lateral heat flow capability of the spreader material. Sapphire restricts the diameter to approximately 100 ”m, while SiC and Diamond allow more than double this size.
  • Pumping Profile Optimization: Utilizing a 10th order Super-Gaussian pump beam profile significantly reduces peak heating, resulting in a maximum temperature rise near the center that is three times smaller than with a standard Gaussian beam.
  • Thick Membrane Strategy: Double-Side Pumping (DSP) is essential for gain membranes thicker than 1 ”m, providing a crucial benefit by enabling a more homogeneous axial temperature distribution across the structure.
  • Modeling Validation: Thermal simulations using the Finite-Element Method (FEM) were validated against experimental spectral shift measurements, yielding a thermal resistance of ~4.25 K/W for the SiC-cooled 800 nm MECSEL structure.
ParameterValueUnitContext
Gain Membrane Thickness (z0)550nmStandard structure thickness used in simulations
MECSEL Emission Wavelength (lambdaMECSEL)800nmNear-infrared lasing region
Pump Wavelength (lambdaP)532nmOptical pumping source
Quantum Defect (etaQ)0.335-Fraction of pump photon energy converted to heat
Membrane Thermal Conductivity (kGaInP)5.2W/mKCalculated value for GaInP bulk layer approximation
Pump Absorption Coefficient (alpha)5.7 x 104cm-1Measured value for pump light
Diamond Thermal Conductivity (kD)2000W/mKHighest conductivity heat spreader
SiC Thermal Conductivity (kSiC)490W/mKMid-range conductivity heat spreader
Sapphire Thermal Conductivity (kSa)30 to 46W/mKLowest conductivity heat spreader (range)
Bonding Layer Thickness (tB)100nmAssumed thickness for imperfect contact simulation
Bonding Layer Thermal Conductivity (kB)0.4W/mKConservative value for bonding interface
Experimental Thermal Resistance (Rth)~4.25K/WMeasured for SiC-cooled MECSEL (dp = 180 ”m)
Thermal Rollover Limit (Sapphire)~110”mMaximum pump beam diameter for efficient scaling
DSC Temperature Reduction Factor (SiC/Diamond)~2-Ratio of Single-Side Cooling (SSC) to DSC temperature rise
  1. Thermal Modeling: The temperature distribution within the MECSEL structure was simulated using the Finite-Element Method (FEM) via COMSOL Multiphysics, employing axial symmetry (cylindrical coordinates).
  2. Structure Simplification: The complex quantum well (QW) and barrier/cladding layers were approximated as a single bulk layer of GaInP (550 nm thick) with a calculated thermal conductivity of 5.2 W/mK.
  3. Heat Load Calculation: Heat generation Q(r, z) was calculated based on pump absorption via the quantum defect (etaQ). Simulations used either a standard Gaussian beam profile or a 10th order Super-Gaussian profile to model the pump source.
  4. Interface Simulation: A thin, low-conductivity bonding layer (100 nm thick, kB = 0.4 W/mK) was inserted between the gain membrane and the heat spreaders to model the thermal resistance of imperfect contact surfaces.
  5. Boundary Conditions: The outer radial surface of the heat spreaders (radius 0.75 mm) was set to a constant heat sink temperature of 20°C. Thermal convection at the MECSEL sandwich facets was neglected (insulation assumed).
  6. Experimental Validation: The FEM model was validated by measuring the thermal resistance of an operating MECSEL using spectral shift measurements (change in emission wavelength versus dissipated power), confirming a thermal resistance of ~4.25 K/W for the SiC-cooled device.

The enhanced power scaling and thermal stability achieved by DSC MECSELs are critical for applications requiring high continuous-wave (CW) output power and stable beam quality:

  • High-Power Industrial Lasers: Used for materials processing, marking, and cutting, where high CW power (watt-levels and above) is necessary, leveraging the superior heat extraction of SiC and Diamond spreaders.
  • Medical and Surgical Systems: Applications requiring compact, high-power, and potentially tunable sources, such as photocoagulation, dermatology, and advanced surgical tools.
  • Atmospheric and Astronomical Systems: Specifically, for generating high-power, narrow-linewidth beams for sodium guide star applications, which require extremely thermally stable operation.
  • Advanced Sensing and Metrology: The external cavity design allows for narrow linewidth and tunability, beneficial for high-resolution spectroscopy and integrated sensors (e.g., LiDAR).
  • Optical Coherence Tomography (OCT): The development of MECSELs in the near-infrared range (800 nm, 1.5 ”m) provides high-brightness sources suitable for deep-tissue imaging.
View Original Abstract

Membrane external-cavity surface-emitting lasers (MECSELs) have great potential of power scaling owing to the possibility of double-side cooling and a thinner active structure. Here, we systematically investigate the limits of heat transfer capabilities with various heat spreader and pumping parameters. The thermal simulations employ the finite-element method and are validated with experimental results. The simulations reveal that double-side cooling lowers the temperature by about a factor of two compared to single-side cooling when diamond and silicon carbide (SiC) heat spreaders are used. In comparison, the benefit for a thermally worse conductive heat spreader is larger, i.e. a fourfold decrease for sapphire. Furthermore, we investigate the limits of power scaling imposed by the intrinsic lateral heat flow of the heat spreaders that sets how much the pump beam diameter can be enlarged while having efficient cooling. To this end, the simulations for sapphire reveal a limit for the pump beam diameter within the hundred micrometer range, while for SiC and diamond the limit is more than double. Moreover, pumping with a super-Gaussian beam profile could further reduce the temperature rise near the center of the pump area compared with a Gaussian beam. Finally, we investigate the benefits of double-side pumping of thick membrane gain structures, revealing a more homogeneous axial temperature distribution than for single-side pumping. This can be crucial for gain membranes with thicknesses larger than <inline-formula xmlns:mml=“http://www.w3.org/1998/Math/MathML” xmlns:xlink=“http://www.w3.org/1999/xlink”> <tex-math notation=“LaTeX”>$\sim 1,\mu \text{m}$ </tex-math></inline-formula> to fully exploit the power-scaling ability of MECSEL technology.