Ts1500

The crystal plate was inserted in a compact temperature controller (THMSE600 or TS1500, Linkam, Tadworth, UK) which was put under a modified microscope (BX-41, Olympus, Tokyo, Japan) for backscattering measurements. The Brillouin scattering spectrum was measured by using a conventional tandem six-pass Fabry-Perot interferometer (TFP-1 or TFP-2, JRS Co., Zürich, Switzerland) and a diode-pumped solid state laser (Excelsior 532-300, Spectra Physics, Santa Clara, CA, USA). The wavelength of the laser was 532 nm, operated at the power of tens of mW. The unpoled NBT-5%BT is pseudo-cubic without any measurable birefringence. The polarization direction of the incident laser beam was along the pseudo-cubic [100] direction. The free spectral range of the interferometer was set to 75 GHz for probing the acoustic modes. A conventional photon-counting system combined with a multichannel analyzer (1024 channels) was used to detect and average the signal. All the Brillouin spectra were measured with increasing temperature. The details of the experimental setup can be found elsewhere [39 (link),40 (link)].

4H-SiC wafers with low doping (n=2.2 × 10¹⁸ cm⁻³) and chemical mechanical polishing were purchased from Cree. Specimens, 5 × 6 mm², were cut from the wafers and cleaned by sequential ultrasonic baths in acetone, isopropyl alcohol and deionized water to remove grease. After cleaning, a SiC substrate was placed into a heating stage (Linkam, TS1500). This is a small chamber (100 × 100 mm²) that enables the bottom of the substrate to be heated up to 1,000 °C with inert gas supply. The sample was annealed at 650 °C for surface cleaning and it was maintained at the same temperature during laser irradiation in an Ar flow. The laser system used is comprised of a XeCl excimer laser (Coherent, LPX model, λ=308 nm, pulse duration ∼30 ns) and beam delivery optics with a homogenizer optical system. The laser beam is concentrated 5 × by a projection lens through a square pattern mask. Each pulse had a laser fluence of 1,653 mJ cm⁻² and a repetition rate of 1 Hz. For multi-pulse irradiations over 100 pulses, 5 Hz was used to reduce the processing time. Both repetition rates appeared to have the same effect because the heating and cooling produced by the nanosecond pulse are completed within a microsecond. All the laser experiments were performed with the same substrate and experimental conditions except for the number of pulses.

Emitter structures were annealed in a high-temperature heating stage (Linkam, TS1500) at 2 × 10⁻² mbar vacuum pressure using a rough vacuum pump, and in a high-temperature vacuum furnace (RD-G WEBB) at 2 × 10⁻⁶ mbar vacuum pressure. The temperature was ramped at a rate of 10 °C min⁻¹. O₂ concentration in the encapsulated area is measured using an O₂ sensor (Sensor type: SO-B0-010, in the range up to 2000 ppm, purchased from SENSORE Electronic GmbH) and it is assembled on a generic sensor board. Ar gas (99.999% from Linde) at a pressure of 3 bar is used to encapsulate the annealing chamber.

The developed DHT was placed in an air environment for a preliminary determination of the power density of optical excitation at a wavelength that did not lead to a shift of the zero-phonon SiV luminescence line, and, consequently, to heating. The measured power density was used to determine the temperature dependence of the spectral position of the zero-phonon line center in a Linkam TS1500 thermostat, stabilized at a predetermined level with an accuracy of ~ 1 °C. The temperature in the chamber was changed in increments of 10 °C. At each step, the SiV luminescence spectrum was recorded and the spectral position of the zero-phonon line maximum was determined according to the algorithm proposed in our previous work^{1 (link)}.