Thermal Conductivity Measurement Setup for Low-Temperature Characterization of Laser Materials

Cryogenic lasers substantially increase the efficiency and power of existing room-temperature lasers.

Operating lasers at cryogenic temperatures gained maturity only in the mid-90s due mostly to the progress in transparent, laser-grade ceramics and semiconductor pump sources. The main limiting factors for power scaling of room-temperature solid-state lasers are thermal effects such as thermal lensing, induced polarization losses, and fracture.

(a) A schematic of the thermal conductivity measuring setup attached to the cryofridge assembly. Wires have been omitted for clarity. (b) A picture of the finished setup fixtures with a sample in place. (c) The entire thermal conductivity measuring station.
These detrimental thermal effects can be substantially suppressed by cryogenic cooling. The cryogenic temperatures modify the spectroscopic characteristics of the laser media. A possible concern is that the bandwidth of both the emission and absorption of the cryogenic laser host may be reduced down to less than 1 nm, introducing more stringent requirements on line width and emission peak stability of pumps. Fortunately, the recent progress in semiconductor pump sources, such as development by industry of high-power spectrally narrowed diode lasers, allows for the building of kW level cryogenic solid-state lasers.

Although the cryogenic concept is mostly applicable to so-called low quantum defect lasers where the heat deposition is kept at the very minimum, this approach was used to develop a diode-pumped, cryogenically cooled 2.7- μm erbium (Er3+):yttria (Y2O3) ceramic laser, operating on quasi-three-level transitions. The approach demonstrated 14W of continuous wave (CW) optical power, and nearly diffraction-limited output; this output was strictly pump-power-limited. Nearly quantum defect (QD)-limited 27.5% optical-to-optical slope efficiency was largely achieved due to implementation of a sufficiently narrowband highpower pump source — a surface-emitting distributed feedback (SE-DFB) laser.

All of the above justifies the necessity of accurate thermal characterization of laser host materials at a wide temperature range from liquid helium to room temperatures.

The figure shows the overall schematic of the thermal conductivity measuring setup, which was designed around two copper plate fixtures that can be attached to the superstructure of a CTI Cryodyne cryogenic refrigerator. The upper copper plate attaches directly to the cryofridge assembly, while the lower plate connects via four 6-32 nylon screws attached from the bottom. The sample to be measured gets sandwiched between the two copper plates, with thin indium layers (~0.2 mm thick) ensuring good thermal contact between the sample and the copper. Accurate temperature readings are obtained using silicon diode sensors that have been indium-soldered to the upper and lower copper plates near the sample contact area. The lower copper plate has a thick-film chip resistor soldered to the bottom, which can generate up to 100W of heating power. A programmable power supply was used as the electrical source for the lower heater. Not shown in the figure is an aluminum intermediate heat shield that attaches to the upper copper plate and enshrouds the entire setup. This shield is necessary to mitigate any radiative heat transfer that might occur between the sample and the dewar tail of the cryofridge.

Thermal conductivity value is not acquired until the temperature at each point of the sample is constant. A LabVIEW front panel was created for the thermal conductivity measuring program, which includes an example data set further illustrating the flow of an experiment. First, the cryofridge is driven to the temperature at which the measurement will be made using an independent controller, and the temperature sensor readings above and below the sample are allowed to reach steadystate. Then, a power value is selected for the lower heater in order to achieve a small (5-6 K) temperature gradient across the sample. The sample dimensions are then input into the LabVIEW program in order to properly calculate thermal conductivity.

The measured temperature range was extended lower in these measurements because of the overall reduced thermal conductivity (lower heat flux) intrinsic to these samples compared to YAG. Immediately apparent in the results for this series of samples is the trend that increasing the doping content leads to lower thermal conductivity. This phenomenon is attributed to an increase in phonon scattering, which inherently follows from decreasing purity in the higher-doped samples. While there are no direct comparisons for this series of samples, the values are comparable to similar rare-earth doped ceramics.

This work was done by Zachary D. Fleischman and Tigran Sanamyan of the Army Research Laboratory. ARL-0179