Name
Characterization of the charge state of the silicon vacancy in 4H-SiC using low-energy muon spin spectroscopy
Description

The silicon vacancy (VSi) in 4H silicon carbide (4H-SiC) is a well-known single photon emitter and spin center, and a strong contender for applications in a variety of quantum technologies, including quantum computation, communication networks, and quantum sensing [1, 2]. Schottky barrier diodes (SBDs) based on SiC have been studied with photoluminescence (PL), and the results show an increase in the emission intensity of the VSi in the space charge region (SCR), suggesting the possibility to manipulate the charge state occupation by controlling the band-bending [3]. Furthermore, transitions between the bright and so-called dark states (q = 0, −2, −3) are suspected to occur when an electrical bias is applied, however, the dark states cannot be observed with PL. Here, low-energy muon-spin spectroscopy (LE-μSR) measurements were performed on an SiC-SBD, to directly probe the SCR with nanometer depth-resolution. In the LE-μSR experiment, a beam of positive muons (μ+) is implanted in the sample to obtain information about the magnetic and electronic environment of the material. The final muonium state depends on the defects present and the doping type and concentration in the material [4,5]. In the case of the VSi, the complex formed with the μ+ serves as a probe of the charge occupation of the defect [6]. The SBD sample was prepared on an n-type SiC epilayer purchased from CREE/Wolfspeed with 10 μm thickness and doping concentration ND =1 × 1017 cm−3, which was irradiated with protons to a fluence of 1× 1014 cm−2. After irradiation, a 20 nm Al layer was deposited on SiC. The sample was annealed at 425 ◦C for 30 minutes, to obtain a detectable concentration of VSi (∼3.0 × 1015 cm−3) but reduce the VC concentration (∼3.7 × 1015 cm−3) below the sensitivity level of LE-μSR for this defect [6,7]. The concentration of the defects was obtained using deep-level transient spectroscopy (DLTS). As shown in Fig. 1 a), by using μ+ implantation energies E in the range of 2–20 keV, it is possible to probe the Al layer and the SiC epilayer up to a depth of ∼130 nm. Fig. 1 b) shows a full energy scan, where at 2 keV to 5 keV the Al layer is probed, and the diamagnetic fraction FD is high due to Mu+ final state in the metal. Between E = 6 and 20 keV, the SiC epilayer is probed where Mu− is expected to form for such a doping concentration. In addition to the conventional LE-μSR setup, an external electric bias was applied to manipulate the band-bending and width of the SCR. The measurements with applied electric field were performed at 10 K, in freeze out regime of carriers, and 100 K at which ionization of the donors in SiC is about 50%. At 100 K (Fig. 2 a)) the applied reverse bias (−0.6 V) causes a drop in FD in the probed region, suggesting an increase in the SCR width, as expected. Similarly, a forward bias (0.4 V) increases FD, due to Mu−formation as free electrons are pulled towards the interface and the SCR becomes narrower. At this temperature, the changes in the VSi charge state are not directly observed due to the presence of free electrons, but these results highlight that band bending in the SBD can be induced by biasing. The measurements at 10 K (Fig. 2 b)) show an effect only when an applied forward bias is applied, with a reduction of FD, at the expense of neutral Mu0 formation. At a bias of 0 V and −0.6 V, the VSi charge state could. therefore, be q = −1 or q = −2, but no clear distinction is possible from the muon signal. Interestingly, at a bias of 0.4 V, indeed a charge transition to q = −3 seems to be possible, and only the filling of this state allows for an electron pick-up to form the Mu0 state.

Speakers
Maria Mendes Martins - APS - ETH Zurich
Date
Friday, October 4, 2024
Time
11:30 AM - 11:50 AM
Location Name
Room 305
Track
Point Defects