Project topics and materials
Get Complete Project Material File(s) Now! »
Chapter 2 Defects in semiconductors
Introduction
Defects are always present in crystalline solids. They have been shown to signi cantly a ect the performance of silicon carbide based devices. This chapter examines defects found in semiconductors. Several ways in which they can be classi ed depending on the material in which they exist are listed. Only n-type 4H -SiC is discussed in this thesis. Detailed descriptions of the most prominent defects found in n-4H -SiC are given as well as some background theory on the di usion process. In this work only zero dimensional defects were investigated, and they will be referred to as point defects. Point defects form the medium for di usion in semiconductors.
Point defects
Examples of point defects (Figure 2.1) are: Vacancies: arise when a component is missing from a regular lattice point. In ionic compounds vacancies can pair up on cation and anion sub-lattices to form Schottky defects (Feng, 2017).
Interstitials: formed when host atoms are located at an interstitial point instead of a regular lattice position. In ionic compounds when an ion occupies an interstitial position other than its own lattice site, a Frenkel defect is formed (Jimenez and Tomm, 2016).
Impurities: these are foreign atoms found isolated inside the lattice (Enderlein and Horing, 1997).
Antisites: formed when certain concentrations of components exchange posi-tions in sub-lattices (Queisser et al., 2003).
Thermodynamically, defects can be classi ed based on the Gibbs potential ( G)(Paul and Divinski, 2017)
where H is the change in enthalpy, S is entropy and T is temperature.
i)Equilibrium defects
They decrease the free energy causing the material to be more thermodynami-cally stable. This is because their production results in an increase in con gura-tional enthalpy. They usually arise in the lattice during material growth (Smith and Wagner, 1987).
ii) Non-equilibrium defects
They increase the free energy causing the material to be more unstable. The increase in enthalpy dominates the increase in entropy. In the presence of suf-cient thermal energy the system will try to minimize these defects. Examples are irradiation induced or damage induced defects (Ryoken et al., 2005).
Shallow and deep level defects
Defects can be classi ed as shallow or deep levels. Shallow levels arise due to small perturbations of the lattice potential by an electrostatic potential generally created by a substitutional doping atom, able to bind an electron (if created by a donor atom) or a hole (if created by an acceptor atom), the same way an electron is bound by an H+ ion, i.e. the trapped carrier sees a long-range 1=r potential.The properties of a shallow level can be approximated by the Bohr model of the hydrogen atom (Poole and Charles, 2004). They can be intentionally introduced in a semiconductor to control the Fermi level in a semiconductor.
Deep levels are defects with properties deviating from those predicted by the Bohr model of the hydrogen atom. The perturbations to the lattice potential are much larger in the trap’s vicinity. They have a steep potential and their electron wave functions are localized at the defect site.
Capture and emission
The occupancy of deep levels is changed through carrier transitions between bands Schroder (2006) as shown in Figure 2.2. The possible transitions are
- Carrier regeneration
- Electron trapping
- Hole trapping
- Recombination
On Figure 2.2 a recombination event occurs in the order (1) then (3). A regeneration happens if (4) occurs after (2). If (1) is followed by (2) or (3) is followed by (4) then the event is classi ed as trapping.
The Z1=2 defects in n-4H -SiC
The prominent defect levels observed in as-grown 4H -SiC in the temperature range (20 { 380 K) we studied are known as the Z1=2 centers observable in conventional DLTS scans at an energy of approximately 0.56 { 0.71 eV below the conduction band minimum. These defects limit the carrier lifetime in high purity epitaxial layers (Son et al., 2012). Several di erent viewpoints exist in literature pertaining to their identity. Eberlein et al. (2003) carried out theoretical calculations and they showed that the Z1=2 was thermally stable and bistable. They modeled the defect as a pi-bonded di-carbon interstitial complex next to a nitrogen atom. Contrary to this, Trinh et al. (2013) theoretically linked the defects to a carbon vacancy.
The Z1=2 has been shown by Son et al. (2012) to be related to a carbon vacancy and made up of two negative U centres with inverted ordering. Negative U behaviour is when one defect traps two electrons simultaneously with the second one being more strongly bound than the rst (Thang, 2015).
Doyle et al. (1998) showed that in VPE grown material, the Z1 center originated from a defect of an intrinsic nature. Using secondary ion mass spectroscopy (SIMS) they proved that the defect was not linked to transition metals. Further, Zhang et al. (2003) showed that the concentration of intrinsic defects in the semiconductor was a factor of growth temperature. Pintilie et al. (2002) investigated the formation of the Z1 defect by varying nitrogen dopant concentrations and C/Si ratios during CVD. The formation of the defect was also observed to be altered by the C/Si ratio. They, therefore, correlated it to the incorporation of nitrogen doping and suggested that it was a nitrogen complex with carbon interstitials or silicon vacancies.
Asghar et al. (2007) used DLTS to characterize defects in hot layer CVD grown epi-layers and they observed the Z1 at Ec 0:61 eV and an additional overlapping level introduced by the radiation which they attributed to the Z2 overlapping with the pre-existing Z1 peak. They reported that alpha particle irradiation transforms the Z1 into the Z1=2 and they related the Z2 to the silicon vacancy.
The temperature range that was studied does not include the EH6=7 but it is also an important defect in 4H -SiC.
Irradiation induced defects in 4H -SiC
The non-ionizing energy loss (NIEL) hypothesis has been successful in estimating energy displacements that account for damage in lattice structures. Conventionally where is the atomic density of the target material, T is the kinetic energy of the recoil material, Tmax the maximum energy that can be transferred to a recoil nucleus by an incident particle, Ed is the threshold displacement energy, d =dT is the di erential interaction cross section, and G(T ) the Lindhard energy partition function (Lindhard et al., 1963).
4H -SiC has been shown to be stable under harsh radiation environments (Nappi and Seguinot, 2004). Its threshold displacement energy is remarkably high, being 35 eV in Si and 20 eV in C (Jiang et al., 2016). These values are almost triple those of gallium arsenide which is also considered a radiation hard material (Claeys and Simoen, 2002). Threshold energy refers to the minimum energy a lattice atom receives before being displaced to an interstitial position in an ion-solid interaction at cryogenic temperatures (Crawford and Slifkin, 2013). In order to apply silicon carbide to radiation hard devices, an in-depth knowledge of its response in such environments is required for re ned control purposes.
Several studies have been carried out to create harsh radiation conditions from heavy ions on 4H -SiC under controlled environments, and probe the ensuing material per-turbations. In an electrical study, Kalinina et al. (2002) irradiated 4H -SiC with 245 MeV Kr ions, and showed that it increased the concentration of Z1 levels. There-after Kalinina et al. (2015) also irradiated 4H -SiC UV detectors with 167 MeV Xe ions and showed that the service life and radiation endurance of the devices increased at higher temperatures (23 { 180 C).
Storasta et al. (2004) irradiated 4H -SiC epi-layers with low energy electrons (80{ 250 keV) and observed a linear increase in defect concentration with increasing radiation dose. They, therefore, speculated that no vacancies or di-vacancies were induced by the radiation. (Hemmingsson et al., 1997) studied electron irradiated epitaxial layers and showed that the capture cross section of the induced defects is temperature dependent. Doyle et al. (1998) showed that the Ec 0:32 eV, Ec 0:62 eV and Ec 0:68 eV levels introduced by electron irradiation annealed out at room temperature in VPE grown material.
Paradzah et al. (2015) irradiated Ni/4H -SiC Schottky diodes with 5.4 MeV alpha particles and revealed that the alpha particle irradiated devices had degraded diode characteristics. Omotoso et al. (2015) irradiated 4H -SiC Schottky diodes with high energy electrons and showed that it induced the Ec
Ec 0:71 eV levels. They also showed that after irradiation, the Richardson constant signi cantly increased, implying that it exacerbates inhomogeneity of the Schottky barrier.
Castelletto et al. (2014) realized an ultra-bright, photo-stable single photon source in 4H -SiC through optimized electron irradiation and annealing. They attributed the source to the carbon-antisite vacancy pair created by electron irradiation
Declaration of Authorship
Abstract
Acknowledgements
List of Figures
List of Tables
Abbreviations
Physical Constants
Symbols
1 Introduction
1.1 Motivation
1.2 Objectives
1.2.1 Layout of the dissertation
1.3.1 Properties and applications
2 Defects in semiconductors
2.1 Introduction
2.2 Point defects
2.2.1 Shallow and deep level defects
2.2.2 Capture and emission
2.3 The Z1=2 defects in n-4H-SiC
2.4 Irradiation induced defects in 4H-SiC
2.5 Defects induced by metalization processes in n-4H-SiC .
2.6 Point defects and diusion .
2.7 Defect annealing
2.8 Raman spectroscopy on 4H-SiC
2.8.1 Confocal Raman microscopy
3 Electrical measurements
3.1 Introduction
3.2 Depletion region
3.3 Capacitance-voltage measurements
3.4 Deep level transient spectroscopy (DLTS)
3.5 Laplace transform DLTS
3.6 Defect depth proling
3.7 Current-voltage measurements
4 Experimental techniques
4.1 Introduction
4.2 Device fabrication
4.3 Swift heavy ion irradiation
4.4 Annealing of samples
4.5 Characterization Techniques
5 Results and discussion
5.1 Introduction
5.2 Native defects
5.3 Sputter deposition
5.4 Annealing
5.5 Swift heavy ion irradiation
6 Conclusions
6.1 Introduction
6.2 Future work
Bibliography
Appendix
GET THE COMPLETE PROJECT
