Deep-level defects in silicon carbide (SiC) are critical to the control of the performance of SiC electron devices. In this paper, deep-level defects in aluminum ion-implanted 4H-SiC after high-temperature annealing were studied using electron paramagnetic resonance (EPR) spectroscopy at temperatures of 77 K and 123 K under different illumination conditions. Results showed that the main defect in aluminum ion-implanted 4H-SiC was the positively charged carbon vacancy (VC+), and the higher the doping concentration was, the higher was the concentration of VC+. It was found that the type of material defect was independent of the doping concentration, although more VC+ defects were detected during photoexcitation and at lower temperatures. These results should be helpful in the fundamental research of p-type 4H-SiC fabrication in accordance with functional device development.

  • EPR results show that the main defect type in p-type 4H-SiC after high temperature annealing is VC+.

  • The type of p-type 4H-SiC defect is independent of the Al doping concentration.

  • The concentration of VC+ defect in p-type 4H-SiC increases with the increase of aluminum ion doping concentration.

Silicon carbide (SiC) has a large band gap, high electron mobility, and good thermal conductivity, all important for applications in high power, high frequency, and high-temperature-resistant electronic devices.1–4 For SiC devices, as for other semiconductor devices, the quality and crystal defects of the material are critical to their fabrication and performance. However, defects such as vacancies, interstitials, or more complex defects can be introduced during crystal growth, material processing or lifetime engineering, such as ion implantation for selective doping or electron irradiation for carrier lifetime control.5 Such defects often appear deep in the band gaps of semiconductors, and, hence, have a strong influence on the electrical and optical properties of the material. Ion implantation is widely used, particularly, in the fabrication of functional semiconductor devices such as PN diodes. However, due to the collision cascade induced by the energized ions, ion implantation will always produce deep-level defects in SiC samples, which will have large impacts on the performance of the device, including compensation ratio and carrier lifetime.6–8 Therefore, it is important to study such defects in silicon carbide in order to better understand and possibly control them.

Compared to other polytypes of silicon carbide, 4H-SiC is widely used, owing to its wide band gap.9–12 However, the origins of deep-level intrinsic defects in 4H-SiC are still not known due to the large differences among samples. Many defects in p-type 4H-SiC have been found, such as Z1/2, EH6/7, EI5, EI6, etc.13,14 Among them, structural defects of EI5 and EI6, which originate from VC+ at quasicubic (k) and hexagonal (h) sites, respectively, have been confirmed by means of electron paramagnetic resonance.14–17 The current electron paramagnetic resonance (EPR) study results showed that, for EI5, there were two space groups at C3v and C1h with temperature change, while EI6 had only one space group at C3v.14 However, no accurate conclusion could be drawn about the structure of the other defects.16 

According to the test principle of EPR, an alternating electric field with a frequency of γ perpendicular to the external magnetic field was applied to the test sample, and = gβH was satisfied, magnetic field resonance absorption will occur. Here h is Planck's constant, g is a dimensionless vector, called the g factor of electrons, and β is the Bohr magneton of electrons. Then the spin electron g factor of the non-even electron is calculated according to the following formula.18 

g=βH=0.714484ν/MHzH/G
(1)

Since the g factor can reflect the interaction between the electron spin motion and the orbital motion in the paramagnetic molecule, the g factor can be used to characterize the environmental and electronic states around the atom of the element being tested. So a change in g factor indicates a certain change in the detected defect structure, such as a different symmetry structure.

It is well known that when performing an EPR test, from theory the integral area under the absorption curve is proportional to the number of unpaired electrons in the sample, and the relative intensity approximates the peak-to-peak amplitude of a reciprocal curve measured under specific conditions.19 In this paper, the relative concentration of defects is calculated by the following formula:19 

ϑ=Y×(ΔHpp)2
(2)

where ϑ represents the relative concentration of the defect, Y represents the sum of the absolute values of the peak and the valley, and ∆Hpp represents the width of the magnetic field between the peak and valley of the EPR, that is, the line width.

This paper focuses on the defects of p-type 4H-SiC after high-temperature annealing. Compared with other studies, our annealing temperature was higher but in the range of temperatures typically used in 4H-SiC device processing after ion implantation.20 The effects of implanted ion concentration, EPR detection temperature, and illumination on the concentration of the defects were studied.

In this paper, a 4H-SiC sample was investigated, wherein the substrate and the epitaxial layer were both n-type silicon carbide, and the epitaxial layer had a thickness of 5.3 μm and a doping concentration of approximately 6 × 1015 cm−3 (Table 1). A p-type doping layer (thickness of approximately 200–250 nm) was prepared by different aluminum ion implantations at room temperature for each quarter of the wafer. Doping concentrations for the different quarters of S1-S3 were 5 × 1016 cm−3, 5 × 1017 cm−3, and 1 × 1018 cm−3. After implantation, the sample was annealed at 1700 °C for 30 min in an Ar atmosphere at 20 mbar. A carbon capping layer on top of the samples was removed after the annealing. In order to consider the influence of both the sample substrate and the epitaxial layer on the EPR experiments, a 4H-SiC substrate, with an epitaxial layer of 5.3 μm thickness, was also characterized and compared with the nitrogen-doped and aluminum-doped layers over the same epitaxial layer with a doping concentration of approximately 5 × 1019 cm−3.

Table 1.

Details of samples used in Raman characterization.

LabelEpitaxyImplantationAnnealing
Carrier concentration (cm−3)Thickness (μm)ElementSurface concentration (cm−3)Temperature (°C)Time (min)
Gepi ≈6.0 × 1015 ≈5.3 –  1700 30 
GN ≈6.0 × 1015 ≈5.3 <5.0 × 1019 1700 30 
GAl ≈6.0 × 1015 ≈5.3 Al 5.0 × 1019 1700 30 
S1 ≈6.0 × 1015 ≈5.3 Al 5.0 × 1016 1700 30 
S2 ≈6.0 × 1015 ≈5.3 Al 5.0 × 1017 1700 30 
S3 ≈6.0 × 1015 ≈5.3 Al 1.0 × 1018 1700 30 
LabelEpitaxyImplantationAnnealing
Carrier concentration (cm−3)Thickness (μm)ElementSurface concentration (cm−3)Temperature (°C)Time (min)
Gepi ≈6.0 × 1015 ≈5.3 –  1700 30 
GN ≈6.0 × 1015 ≈5.3 <5.0 × 1019 1700 30 
GAl ≈6.0 × 1015 ≈5.3 Al 5.0 × 1019 1700 30 
S1 ≈6.0 × 1015 ≈5.3 Al 5.0 × 1016 1700 30 
S2 ≈6.0 × 1015 ≈5.3 Al 5.0 × 1017 1700 30 
S3 ≈6.0 × 1015 ≈5.3 Al 1.0 × 1018 1700 30 

Gepi: no ion implantation was performed in this area, and thus no surface concentration is shown here.

The EPR instrument used in this experiment was a German Bruker SOE-072 A300 electronic paramagnetic resonance spectrometer in the X-band (∼9.5GHz) range, and the magnetic field was perpendicular to the sample surface. For illumination, a Xenon lamp (1.5 eV) was used as a light source to test carbon vacancies at wavelength 808 nm with an illumination time of 10 min. Under different temperatures and illuminations (by light excitation, VC+ can be activated and the intensity of the EPR signal will respond21), the defects in 4H-SiC following Al ion implantation and high-temperature annealing were studied.

EPR experiments were conducted over the same epitaxial (Gepi), nitrogen-doped (GN), and aluminum-doped (GAl) layers with a doping concentration of approximately 5 × 1019 cm−3, as shown in Fig. 1. Table 2 gives the g factor and relative concentration of the defect. The defect type for both Gepi and GN was VC, while the defect for GAl was VC+. It can be seen that GN had the lowest carbon vacancy defect concentration, while GAl had the highest carbon vacancy. This was because nitrogen ion implantation had a certain compensating effect on the carbon vacancies in the Gepi, so that the carbon vacancy defect concentration would drop.22 On the other hand, the aluminum ion atom radius was relatively larger, producing more point defects. The aluminum ions were also more likely to assume the positions of the silicon atoms, so that the compensation of the carbon vacancies was not as obvious as that for the nitrogen ions.22 By comparison of the three areas on the same substrate, it was shown that EPR well-characterized the carbon vacancy statuses for the different doping setups over the 4H-SiC substrate.

Fig. 1.

EPR spectra of 4H-SiC samples measured at 123 K and without illumination.

Fig. 1.

EPR spectra of 4H-SiC samples measured at 123 K and without illumination.

Close modal
Table 2.

g factor and relative concentrations of defects corresponding to EPR spectra for different experimental conditions.

SamplesDoping concentration (cm−3)g factorRelative concentration of defects
Gepi  2.00381 1,056,782.7 
GN <5 × 1019 2.00361 608,525.3 
GAl 5 × 1019 2.00364 2,346,948.7 
SamplesDoping concentration (cm−3)g factorRelative concentration of defects
Gepi  2.00381 1,056,782.7 
GN <5 × 1019 2.00361 608,525.3 
GAl 5 × 1019 2.00364 2,346,948.7 

Fig. 2 shows the results of EPR measurements of samples S1–S3 at different temperatures and under different illumination conditions. Among them, Fig. 2(a) shows the EPR spectrum measured at 123 K and dark conditions (without an external light source). When the microwave frequency was about 9.5 GHz and the magnetic field was 351.4 mT, the EPR spectra of the three test samples S1–S3 showed clear signals. The position of the central magnetic field in the experimental results was nearly the same as that reported by Wu,18 under similar measurement conditions. Here, the EPR result also indicated that the central magnetic field position was independent of the Al doping concentration (for the investigated range) and remained stable at 351.4 mT. Fig. 2(b) shows the results of an EPR experiment that was performed after 10 min of illumination and in a light environment at 123 K. There was a clear signal on the EPR spectrum at 353.4 mT. There was also virtually no change in the position of the central magnetic field with and without illumination, indicating that the position of the central magnetic field was also independent of illumination, related only to the EPR instrument used and the microwave frequency chosen. It can be seen from Fig. 2(c) and (d) that the peak-to-valley asymmetry appeared in the EPR line at 77 K, indicating that the 4H-SiC material was anisotropic.19 

Fig. 2.

EPR results for samples S1–S3 at different temperatures under different illumination conditions. (a) EPR spectra of 4H-SiC sample measured at 123 K without illumination; (b) EPR spectra measured at 123 K with 10 min of illumination (1.5 eV); (c) EPR spectra measured at 77 K without illumination, and (d) EPR spectra measured at 77 K with 10 min of illumination (1.5 eV).

Fig. 2.

EPR results for samples S1–S3 at different temperatures under different illumination conditions. (a) EPR spectra of 4H-SiC sample measured at 123 K without illumination; (b) EPR spectra measured at 123 K with 10 min of illumination (1.5 eV); (c) EPR spectra measured at 77 K without illumination, and (d) EPR spectra measured at 77 K with 10 min of illumination (1.5 eV).

Close modal

3.2.1. g factor analysis

Table 3 summarizes the g factor corresponding to the EPR spectra obtained under different experimental conditions. When the measurement was performed without illumination and at a temperature of 123 K, the g factor was consistent with the experimental results of Son et al., which determined that the g factor of the carbon vacancy VC+ was 2.00322.23 The defects represented by the EPR signals of the three different samples were considered as VC+. The g factor was sensitive to temperature and illumination conditions. As shown in Fig. 3, the illumination excitation at a temperature of 123 K caused a decrease in the g factor, which was consistent with the trend visible from the results of Ping.19 As shown in Fig. 4, when the experimental conditions were in darkness, the g factor also increased with increasing temperature from 77 K to 123 K. However, the g factor and EPR signal are both nearly the same for the situations of with and without illumination at 77 K. These results showed that there would be a critical temperature between 77 K and 123 K for EPR measurements, so that the detectable carbon vacancy defects structures would change beyond this critical temperature, such as changes in structural symmetry.

Table 3.

g factor corresponding to EPR spectra for different experimental conditions.

SamplesDoping concentration (cm−3)g factor
123 K in darkness123 K illumination (1.5 eV)77 K in darkness77 K illumination (1.5 eV)
S1 5 × 1016 2.00384 1.99111 1.99383 1.99383 
S2 5 × 1017 2.00381 1.99199 1.99448 1.99162 
S3 1 × 1018 2.00400 1.99199 1.99332 1.99277 
SamplesDoping concentration (cm−3)g factor
123 K in darkness123 K illumination (1.5 eV)77 K in darkness77 K illumination (1.5 eV)
S1 5 × 1016 2.00384 1.99111 1.99383 1.99383 
S2 5 × 1017 2.00381 1.99199 1.99448 1.99162 
S3 1 × 1018 2.00400 1.99199 1.99332 1.99277 
Fig. 3.

Change in the g factor with illumination for 10 min at a temperature of 123 K.

Fig. 3.

Change in the g factor with illumination for 10 min at a temperature of 123 K.

Close modal
Fig. 4.

Change in the g factor with temperature and without illumination.

Fig. 4.

Change in the g factor with temperature and without illumination.

Close modal

3.2.2. Relative concentrations of defects

Fig. 5 shows the relative concentrations of defects measured at 123 K under different illumination conditions. It can be seen that, as the doping concentration of aluminum increased, the relative concentration of defects increased significantly, indicating that the concentration of VC+ was sensitive to and grew with ion doping concentrations ranging from 5 × 1016 cm−3 to 1 × 1018 cm−3. When the concentration of the implanted aluminum ions increased, the degree of damage to the sample increased, thereby causing the carbon atoms to leave their original positions and form more carbon vacancies.

Fig. 5.

Relative concentration of defects at 123 K (a) without, and (b) with illumination for 10 min for different Al doping concentrations.

Fig. 5.

Relative concentration of defects at 123 K (a) without, and (b) with illumination for 10 min for different Al doping concentrations.

Close modal

Fig. 6 shows the variation in relative concentrations of defects at different temperatures and illuminations for 10 min. It can be seen that, after illumination at a temperature of 123 K, the relative concentration of defects increased significantly by a factor of around 500. It was believed that light excitation provided energy to the defect, so that more VC+ was available to absorb energy and be detected.19 Results also showed that light excitation converted the carbon vacancy defect from VC2+ and VC, which cannot be detected by EPR, to VC+, which can be detected by EPR, so that the concentration of defects increased significantly.21 In addition, it could be seen that, as the temperature dropped from 123 K to 77 K without illumination, the relative concentration of defects increased dramatically by a factor of about 5000 compared to the measurement at 123 K without illumination. This indicated that, when the temperature changed from 123 K to 77 K, interference from phonon vibration was significantly reduced, so that more VC+ defects were detected at 77 K. It also showed that the effect of temperature on the detected defect concentration was stronger than that of illumination. In future studies, the hyperfine constants of the EPR spectrum can be increased to more accurately determine the structure of the defects, including symmetry.14 

Fig. 6.

Change in relative concentrations of defects with ion doping concentration at different temperatures and illuminations.

Fig. 6.

Change in relative concentrations of defects with ion doping concentration at different temperatures and illuminations.

Close modal

In this paper, Al-implanted 4H-SiC samples of differing doping concentrations were studied by EPR after high-temperature annealing at different temperatures and under different illumination conditions. Experimental results showed that the main defect type in p-type 4H-SiC (Al-implantation) after high-temperature annealing is VC+, and the anisotropic structure of 4H-SiC becomes evident from EPR results at 77 K. The type of material defect is independent of Al doping concentration. The concentration of VC+ defects in 4H-SiC increases with increasing aluminum ion doping concentration, with more VC+ defects being detected during photoexcitation and at lower temperatures.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

The study is supported by the National Natural Science Foundation of China (No. 51575389, 51761135106), the National Key Research and Development Program of China (2016YFB1102203), the State Key Laboratory of Precision Measurement Technology and Instruments (Pilt1705), and the '111' Project by the State Administration of Foreign Experts Affairs and the Ministry of Education of China (Grant No. B07014). The authors would like to thank Mr. Te Wu and Mr. Kang Li for their help for the EPR measurements.

1.
Roccaforte
F
,
Fiorenza
P
,
Greco
G
, et al 
Emerging trends in wide band gap semiconductors (SiC and GaN) technology for power devices
.
Microelectron. Eng.
2018
;
187
:
66
77
.
2.
Weitzel
CE
,
Palmour
JW
,
Carter
CH
, et al 
Silicon carbide high-power devices
.
IEEE Trans. Electron Devices
1996
;
43
:
1732
1741
.
3.
Wei
R
,
Song
S
,
Yang
K
, et al 
Thermal conductivity of 4H-SiC single crystals
.
J. Appl. Phys.
2013
;
113
(
5
):
53503
.
4.
Sugawara
Y.
Recent progress in SiC power device developments and application studies
.
ISPSD’03. 2003 IEEE 15th International Symposium on Power Semiconductor Devices and ICs. 2003. Proceedings
.
2003
.
10
18
.
5.
Miyazawa
T
,
Tsuchida
H
.
Point defect reduction and carrier lifetime improvement of Si- and C-face 4H-SiC epilayers
.
J. Appl. Phys.
2013
;
113
(
8
):
83714
.
6.
Storasta
L
,
Tsuchida
H
.
Reduction of traps and improvement of carrier lifetime in 4H-SiC epilayers by ion implantation
.
Appl. Phys. Lett.
2007
;
90
(
6
):
62116
.
7.
Miyazawa
T
,
Ito
M
,
Tsuchida
H
.
Evaluation of long carrier lifetimes in thick 4H silicon carbide epitaxial layers
.
Appl. Phys. Lett.
2010
;
97
(
20
):
202106
.
8.
Kawahara
K
,
Suda
J
,
Pensl
G
, et al 
Reduction of deep levels generated by ion implantation into n- and p-type 4H–SiC
.
J. Appl. Phys.
2010
;
108
(
3
):
33706
.
9.
Wenzien
B
,
Käckell
P
,
Bechstedt
F
.
Quasiparticle band structure of silicon carbide polytypes
.
Phys. Rev. B
1995
;
52
(
15
):
10897
.
10.
Roccaforte
F
,
Fiorenza
P
,
Greco
G
, et al 
Challenges for energy efficient wide band gap semiconductor power devices
.
Phys. Status Solidi A
2014
;
211
(
9
):
2063
2071
.
11.
Izhevskyi
VA
,
Genova
LA
,
Bressiani
JC
,
Bressiani
AHA
.
Silicon carbide. Structure, properties and processing
.
Cerâmica
2000
;
46
(
297
):
4
13
.
12.
Jayant Baliga
B.
Wide Bandgap Semiconductor Power Devices: Materials, Physics, Design, and Applications
.
Woodhead Publishing
;
2018
.
13.
Hiyoshi
T
,
Kimoto
T
.
Elimination of the major deep levels in n-and p-type 4H-SiC by two-step thermal treatment
.
Appl. Phys. Express
2009
;
2
(
9
):
91101
.
14.
Umeda
T
,
Isoya
J
,
Morishita
N
, et al 
EPR and theoretical studies of positively charged carbon vacancy in 4H−SiC
.
Phys. Rev. B
2004
;
70
(
23
):
235212
.
15.
Umeda
T
,
Isoya
J
,
Morishita
N
, et al 
EPR identification of two types of carbon vacancies in 4H−SiC
.
Phys. Rev. B
2004
;
69
(
12
):
121201
.
16.
Zvanut
ME
.
Electron paramagnetic resonance of electronic-grade SiC substrates
.
J. Phys. Condens. Matter.
2004
;
16
(
46
):
R1341
.
17.
Isoya
J
,
Umeda
T
,
Morishita
N
, et al 
EPR identification of intrinsic defects in Si
.
Phys. Status Solidi B
2008
;
245
(
7
):
1298
1314
.
18.
Wu
Z
.
Study on Ion Irradiation Effects of Single Crystal 6H-SiC and 4H-SiC
.
Harbin, China
:
Harbin Institute of Technology
;
2018
(in Chinese).
19.
Cheng
P
,
Zhang
Y
,
Zhang
Y
, et al 
ESR characters of intrinsic defects in epitaxial semi-insulating 4H-SiC illuminated by Xe light
.
J. Semicond.
2009
;
30
(
12
):
123002
.
20.
Kimoto
T
,
Cooper
JA
.
Fundamentals of Silicon Carbide Technology: Growth, Characterization, Devices and Applications
.
Singapore, Piscataway, New Jersey
:
Wiley IEEE Press
;
2014
.
21.
Kawahara
K
,
Xuan
TT
,
Son
NT
, et al 
Quantitative comparison between Z1/2 center and carbon vacancy in 4H-SiC
.
J. Appl. Phys.
2014
;
115
(
14
):
143705
.
22.
Erlekampf
J
,
Kallinger
,
Weiße
J
, et al 
Deeper insight into lifetime-engineering in 4H-SiC by ion implantation
.
J. Appl. Phys.
2019
;
126
:
045701
.
23.
Son
NT
,
Hai
PN
,
Janzén
E
.
Carbon vacancy-related defect in 4H and 6H SiC
.
Phys. Rev. B
2001
;
63
(
20
):
201201
.

Xiuhong Wang, State Key Laboratory of Precision Measuring Technology and Instruments, School of Precision Instruments and Opto-Electronics Engineering, Tianjin University, China. Ms. Wang is studying for Master degree. Her research interests include: silicon carbide material and deep level defect.

Zongwei Xu, Associate Professor, State Key Laboratory of Precision Measuring Technology and Instruments, School of Precision Instrument and Opto-Electronics Engineering, Tianjin University, China. His research interests include: defect engineering in wide band gap semiconductor, Micro/ nanofabrication using focused ion beam, Raman and fluorescence spectrum, etc.

Mathias Rommel, Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) and Fraunhofer Institute for Integrated Systems and Device Technology IISB, Germany. His research interests include: focused ion beam (FIB), nanoimprint lithography, electrical scanning probe microscopy, and deep level transient spectroscopy (DLTS).

Bing Dong, State Key Laboratory of Precision Measuring Technology and Instruments, School of Precision Instrument and Opto-Electronics Engineering, Tianjin University, China. Mr. Dong is studying for Master degree.

Le Song, Associate Professor, State Key Laboratory of Precision Measuring Technology and Instruments, School of Precision Instrument and Opto-Electronics Engineering, Tianjin University. Prof. Song received his PhD degree in 2008 and is currently working at Tianjin University. His research interests include: Bionic measurement and control method and system implementation, Precision measurement in advanced manufacturing.

Clarence Augustine TH Tee, PhD (Cambridge, 2000), BEng (Hons, 1st Class, 1997), PEPC (Electrical), CEng (UK), ACPE (Electrical), FIET (UK), SMIEEE (USA), Fellow CCT (UK), FellowCTES (UK), Fulbright Fellow, is a Professor (Dept. of Electrical Engineering) and Director (Industrial Liaison Div.), Faculty of Engineering, University of Malaya, Malaysia. His current research interests are in interdisciplinary research involving Engineering, Life Sciences/Medicine and Deep Learning, Micro-Nanotechnology/Nanoscience and Photonics (incl. Plasmonics/Surface Optics).

Fengzhou Fang, Professor, State Key Laboratory of Precision Measuring Technology and Instruments, School of Precision Instrument and Opto-Electronics Engineering, Tianjin University. His research interests are in the areas of micro/nano manufacturing, optical free form manufacturing, bio-medical manufacturing, ultra-precision machining and metrology.