Interface States Passivation for p-Channel 4H-SiC MOS Devices

Abstract

Impressive material properties such as wide band gap, high breakdown electric field and thermal conductivity makes silicon carbide (4H-SiC) an attractive semiconductor for high power and high temperature electronics. Applications include, but are not limited to, power and control systems in hybrid and electric vehicles, space crafts and renewable energy sources [1-2]. In addition to discrete power components, an advanced 4H-SiC complementary metal-oxide-semiconductor (CMOS) integrated circuit technology has the potential to revolutionize high temperature electronics, for operational temperatures greater than 200 ºC [3-4], where Si is not efficient or usable. CMOS requires both n-channel (electron) and p-channel (hole) metal-oxide-semiconductor field effect transistors (MOSFETs) and it is critical for both polarities to have adequate performance. In n-channel 4H-SiC devices, nitrogen incorporation at the SiO2/4H-SiC interface reduces interface state density (Dit) energetically located in the upper half of the 4H-SiC band gap and near the conduction band-edge [5]. Research on p-channel is less extensive, but prior research motivates hydrogen [6] and more recent studies suggest nitrogen [7] as effective passivation agents for donor like traps in the lower half of the 4H-SiC band gap. In this work, we investigated several post-oxidation annealing (POA) processes involving nitrogen and hydrogen for Dit reduction in p-type MOS capacitors. For a comparative study, n-type capacitors were also fabricated and characterized. All samples were oxidized at 1150 °C to grow a ~60 nm oxide film and some samples were subjected to POA as summarized in Table 1. The results show that NO and H2 both reduce Dit in the lower half of the 4H-SiC band gap, with NO having a much larger effect. The reduction of positive trapped charge lowers the flat band voltage for p-type 4H-SiC MOS capacitors with different degrees of passivation as shown in Figure 1. The Dit profile near the band-edges, extracted from simultaneous high (100 kHz) - low frequency CV at room temperature, highlights the large reduction associated with NO, as shown in Figure 2. The effect of hydrogen is primarily in the reduction of energetically deep interface traps. To measure the density of the deep traps (Nit, deep), photon assisted CV measurements [8] were carried out, from which Nit, deep was calculated as shown in Figure 3 and summarized in Table 1. Overall, the results confirmed that 2 hour NO annealing at 1175 °C results in the lowest Dit in the lower half of the band gap. Longer NO annealing times or combination of NO with hydrogen resulted in higher Nit, deep as listed in Table 1. Therefore, long channel 4H-SiC p-MOSFETs were fabricated using the 2 hour NO POA process to investigate hole transport and gate control in devices operating at high temperature. The results from the transfer characteristics in Figure 4 indicate that with the increase of temperature, threshold voltage reduces which can be attributed to the reduction of occupied Dit at higher temperatures. A weak temperature dependence of the hole field effect mobility with temperature is present most likely due to surface roughness scattering limitations at higher channel hole concentrations. In this presentation, the field-effect mobility will be compared to Hall measurements for additional insights into surface hole transport. In addition, details of the CV analyses, FET fabrication procedure and high temperature bias instability will be discussed. Acknowledgements: The authors gratefully acknowledge the funding support from US Army Research Laboratory grant ARMY-W911NF-18-2-0160 and National Renewable Energy Laboratory/ US Department of Energy grant NREL-AHL-9-92362-01. References : [1] S. Srdic et al., IEEE Appl. Power Elec. Conf. and Expo.(APEC), 2714 (2016)[2] M. Abbasi et al., in IEEE Trans. Emerg. Sel. Topics Power Electron., 7, 798 (2019)[3]A. M. Francis et al., Proc. IMAPS High Temperature Electron. Conf., 2016, 242, (2016)[4] A. S. Kashyap et al., IEEE (WiPDA), 60 (2013)[5] G. Liu et al. Appl. Phys. Rev. 2, 021307 (2015)[6] M. Okamoto et al. Appl. Phys. Lett., 89, 023502 (2006)[7] K. Tachiki and T. Kimoto IEEE Trans. Elec. Dev., 68, 638 (2021)[8] J. A. Cooper, Jr., phys. stat. sol. 162, 305 (1997) Figure 1