Surface recombination velocities for 4H-SiC: Temperature dependence and difference in conductivity type at several crystal faces

Abstract

In bipolar SiC devices, which are promising under ultra-high voltage operation, the carrier lifetime is a highly influential parameter for the device performance. Surface recombination is one of the limiting factors for the carrier lifetime, and quantitative values of the surface recombination velocities are required for the design and development of fabrication processes of the devices. In this study, we observe carrier recombination at various temperatures for the Si- and C-faces of n- and p-type 4H-SiC samples and the a- and m-faces of n-type 4H-SiC samples with a treatment of chemical mechanical polishing or reactive ion etching by using the microwave photoconductivity decay method. From the experimental results, we estimate surface recombination velocities and bulk carrier lifetimes of the samples by using an analytical model. As a result, we found the smallest surface recombination velocity of 150 cm/s for the chemical mechanical polished surface of the Si-face of the n-type samples at room temperature. Surface recombination velocities increased with temperature for the chemical mechanical polished surfaces. The surfaces treated with reactive ion etching showed relatively large surface recombination velocities with weak temperature dependence. Based on these results, we discuss the origins of the recombination centers at surfaces of 4H-SiC.

Defect engineering in SiC technology for high-voltage power devices

“…The surface recombination velocity for SiC surfaces after several different processes has been determined. 107,108) However, basic understanding of the temperature dependence of the carrier lifetime is insufficient and thus further investigations are required. In addition, the major recombination path in p-type SiC is also an open question.…”

Suppression of stacking fault expansion in a 4H-SiC epitaxial layer by proton irradiation

“…Here, the typical value for an as-grown n-type SiC epitaxial layer (0.1–1 μs) was adopted for the estimation. Note that the measured lifetime was much smaller than typical value of the bulk carrier lifetime since the surface recombination largely influences on the measured carrier lifetime of the epitaxial layer with the thickness of 10 μm 53 , 55 , 56 . The excess carrier concentration was estimated to be 2 × 10 14 –2 × 10 15 cm −3 , which led to an electronic energy gain (Δ) of around 0.5 ~ 3 mJ/m 2 from the result of calculations reported in Ref.…”

“…The excess carrier concentration in epitaxial layers ( ) caused by UV illumination in the steady state was estimated by the following equation for low excess carrier concentrations, assuming that surface and interface recombination is negligible 53 : where is the bulk carrier lifetime in the epitaxial layer and G is the rate of generation of excess carriers in the epitaxial layer calculated from the photon flux and absorption 54 . The value of is not the same as the carrier lifetime measured by μ-PCD and TR-PL 53 . Here, the typical value for an as-grown n-type SiC epitaxial layer (0.1–1 μs) was adopted for the estimation.…”

Trapping effects and surface/interface recombination of carrier recombination in single- or poly-crystalline metal halide perovskites

“…The excitation sources were pulsed yttrium aluminium garnet lasers with pulse widths of 1 ns and wavelengths of 266 or 355 nm. [28][29][30][31] The repetition frequencies of the lasers were 60-100 Hz, and the intensities of the reflected light from the samples was less than 4% compared with the incident light intensities. For TR-PL, we employed a photomultiplier (Hamamatsu Photonics, H10721) as a detector through a long pass filter at cut-off wavelength of 355 nm to prevent injection of the excitation light.…”

“…In addition, measurements in two (or more) excitation sources with different wavelengths have been performed to observe the contribution of surface recombination. [28][29][30][31] The penetration depth of light depends on wavelength, 20,32,33) and the distributions of the photoexcited carriers will be different between the excitation sources. The different distributions make different contributions of surface recombination to the overall photoexcited carrier recombination.…”