Concentration of point defects in 4H SiC characterized by a magnetic measurement B Peng, R X Jia, Y T Wang, L P Dong, J C Hu, and Y M Zhang Citation AIP Advances 6, 095201 (2016); doi 10 1063/1 496254[.]
Concentration of point defects in 4H-SiC characterized by a magnetic measurement B Peng, R X Jia, Y T Wang, L P Dong, J C Hu, and Y M Zhang Citation: AIP Advances 6, 095201 (2016); doi: 10.1063/1.4962545 View online: http://dx.doi.org/10.1063/1.4962545 View Table of Contents: http://aip.scitation.org/toc/adv/6/9 Published by the American Institute of Physics Articles you may be interested in Threading dislocation movement in AlGaN/GaN-on-Si high electron mobility transistors under high temperature reverse bias stressing AIP Advances 6, 095102095102 (2016); 10.1063/1.4962544 AIP ADVANCES 6, 095201 (2016) Concentration of point defects in 4H-SiC characterized by a magnetic measurement B Peng, R X Jia,a Y T Wang, L P Dong, J C Hu, and Y M Zhang School of Microelectronics, Xidian University, Xi’an, 710071, China (Received June 2016; accepted 29 August 2016; published online September 2016) A magnetic method is presented to characterize the concentration of point defects in silicon carbide In this method, the concentration of common charged point defects, which is related to the density of paramagnetic centers, is determined by fitting the paramagnetic component of the specimen to the Brillouin function Several parameters in the Brillouin function can be measured such as: the g-factor can be obtained from electron spin resonance spectroscopy, and the magnetic moment of paramagnetic centers can be obtained from positron lifetime spectroscopy combined with a first-principles calculation To evaluate the characterization method, silicon carbide specimens with different concentrations of point defects are prepared with aluminum ion implantation The fitting results of the densities of paramagnetic centers for the implanted doses of × 1014 cm−2, × 1015 cm−2 and × 1016 cm−2 are 6.52 × 1014/g, 1.14 × 1015/g and 9.45 × 1014/g, respectively The same trends are also observed for the S-parameters in the Doppler broadening spectra It is shown that this method is an accurate and convenient way to obtain the concentration of point defects in 4H-SiC C 2016 Author(s) All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/) [http://dx.doi.org/10.1063/1.4962545] As a wide band gap semiconductor, silicon carbide (SiC) plays an important role in the power electronics industry, owing to its advantages of high electronic breakdown field, high thermal conductivity and stable chemical property In recent years, it has been found that the deep levels in silicon carbide wafers or epilayers prevent the development of high-performance SiC devices,1–5 mainly because of the reduction of carrier lifetimes that lead to a reduction in conductivity.6–10 Since point defects are most common at deep levels, the characterization of point defects in silicon carbide has attracted the interest of many research groups and several methods have been presented in this area, such as electron spin resonance (ESR) spectroscopy,11–14 deep level transient spectroscopy (DLTS)15–17 and positron annihilation spectroscopy (PAS).18–20 Both the intensity of the electron spin resonance spectra and the deep level transient spectra are related to the concentration of defects in the SiC wafers or epilayers However, ESR measurements demand samples which have high defect concentrations, that is, wafers implanted or irradiated with high ion doses, and conventional DLTS measurement are not suitable for samples which have a high resistivity Recently, the magnetic property of silicon carbide has attracted the interest of various research groups, for example, the observed magnetism of neutron irradiated SiC single crystal,21 the room temperature ferromagnetism of 6H-SiC after N implantation22 and the glassy ferromagnetic feature of synthesized Al-doped 4H-SiC samples.23 More recent research results show that the magnetization measured in conventional implanted SiC samples can be decomposed into diamagnetic, paramagnetic and ferromagnetic components.24 The paramagnetic component can be fitted by the Brillouin function ( ) ( ) 2J + 2J + 1 M(x) = N J µ B g coth x − coth x (1) 2J 2J 2J 2J a Electronic mail: rxjia@mail.xidian.edu.cn 2158-3226/2016/6(9)/095201/6 6, 095201-1 © Author(s) 2016 095201-2 Peng et al AIP Advances 6, 095201 (2016) x = g µ B J µ0 H/k BT (2) where M is the magnetization of the paramagnetic component, µB is the Bohr magneton, J is the total angular momentum of the paramagnetic center, g is the Landé factor and N is the density of paramagnetic centers Here, the paramagnetic center can be attributed to point defects, such as Vsi, Vc or VsiVc divacancy, which possess a magnetic moment according to several ESR experiments8,9 and first-principles calculation25,26 results Therefore, a connection between the concentration of point defects and the magnitude of paramagnetism can be proposed This would provide a candidate method to characterize the concentration of point defects in implanted SiC samples by measuring the parameter in the Brillouin function and then calculating the value of the density of paramagnetic centers The specific procedure can be described as follows: first, the g-factor in the Brillouin function can be obtained from an electron spin resonance spectroscopy measurement; then, knowing the type of point defects, the total angular momentum of the paramagnetic center can be obtained from the first-principles calculation Finally, the paramagnetic component can be extracted from the M-H curve of implanted samples However, for practical ion implantation experiments, the high energy and heavy ions will lead to complicated defects or an amorphous region, so a relatively low implantation energy and light ions are needed to induce typical point defects to fit our model In this paper, three aluminum implanted 4H-SiC samples with different doses are used to evaluate our characterization method A one-side polished 4H-SiC (0001) single crystal from TanKeBlue Corporation (Beijing, China) was cut into pieces with dimensions of 10 × 10 × 0.33 mm3 All the samples were implanted with Al ions under a 100 KeV accelerating voltage at room temperature but with different doses of × 1014 cm−2, × 1015 cm−2 and × 1016 cm−2 (referred to as 1E14, 1E15 and 1E16, respectively) During the implantation experiment, samples were perpendicular to the ion beam ESR experiments were performed on a JES-FA200 X-band ESR spectrometer M-H curves of the samples were obtained by a SQUID-VSM magnetometer (Quantum Design, Inc.) Positron annihilation lifetime and Doppler broadening measurements were carried out at the slow positron beam facility in the Institute High Energy Physics, Chinese Academy of Sciences Slow positrons were generated by a 1.85 GBq 22Na radiation source and the positron beam energy range was from 0.18 KeV to 20 KeV The lifetime spectra had a time resolution of 180 ps and contained × 106 events, and the data were fitted by using the program package LT 9.0 The S- and W-parameters were defined as the ratios of the counts in the central low momentum area (510.2–511.8 KeV) and the two flanking high momentum regions (514.83–518.66 keV and 503.34–507.17 KeV) in the DB spectra to the total counts, respectively All calculations were based on density functional theory with the Cambridge Serial Total Energy Package (CASTEP) code The ESR spectra of the 4H-SiC samples before and after ion implantation measured at 130 K are shown in Fig 1(a) with the magnetic field perpendicular to the samples’ surface Compared with the implanted samples, the pristine sample did not present a clear signal in the electron paramagnetic resonance spectrum Meanwhile, the ESR spectrum of the implanted samples exhibited a line shape of a Lorentzian form, which can be attributed to the existence of isolated magnetic FIG (a) ESR spectra of the 4H-SiC samples before and after ion implantation at 130K (b) Temperature dependent ESR spectra of the sample 1E15 095201-3 Peng et al AIP Advances 6, 095201 (2016) moments.11 Fig 1(b) shows the temperature dependent ESR spectra of the implanted sample 1E15 The intensity decreased with an increasing temperature, which proves that the ESR intensity arises from the isolated paramagnetic centers.27 The intensity of the different samples in Fig 1(a) increased with the implantation doses varying from × 1014 cm−2 to × 1015 cm−2, and then reduced with a further increase in ion dosage of × 1016 cm−2, which shows a variation in the trend of the number of isolated magnetic moments The value of the g-factor of the implanted samples were 1.991, 1.995 and 1.996, corresponding to samples 1E16, 1E15 and 1E14, respectively, which indicates a gradual weak spin-orbital coupling of electrons in the localized magnetic moments, which can be attributed to the doping of Al atoms A hyperfine structure was apparent in the ESR spectra at 130 K, which indicates the isolated paramagnetic centers can be attributed to point defects.12 Positron annihilation lifetime spectroscopy measurements were performed to identify the defect type of the implanted samples A lifetime of 135.4 ± 0.3 ps was observed in the pristine sample, with an intensity of 98.36%, which agrees with the previous result obtained for defect-free 4H-SiC.28 The main lifetime value of sample 1E15 was 149 ± 0.4 ps, with an intensity of 97.31%, which is slightly increased from that of the pristine sample Since the lifetime of the carbon vacancy in SiC is approximately equal to the bulk lifetime owing to positrons proposed to be only slightly localized in carbon vacancies,16 and the formation energy of AlC is bigger than that of AlSi, the increase of the lifetime can be attributed to positrons being trapped at silicon vacancies Therefore, VSi was treated as the main type of point defects in our implanted samples The value of Simplanted/Spristine (>1.02) shown in Fig 2(a) at the energy region of 0.18–5 KeV is in good agreement with the ratio of the S parameters for the isolated silicon vacancy in silicon carbide.15,19 The W-S plots for different implanted samples are shown in Fig 2(b) In the W-S representation for different implanted SiC samples, we can obtain approximately straight lines and the slope of the three lines are almost equal The slope of the W-S plot represents the mechanism of positron annihilation after trapping, and the W-S plot has been used to identify the number of defect types which are trapped by positrons in materials.29 This indicates the existence of the same type of open volume defects created by different ion implantation doses in these SiC samples, which corresponds to the silicon vacancy observed from our previous PALS result The magnetic moment of the silicon vacancy is calculated with the CASTEP code A × × supercell sample of SiC with 72 atoms was created to act as the computational model, which is presented in Fig 3(a), with Monkhorst-Pack × × k-points used for integrations of the reduced Brillouin zone Before calculating the properties, structural relaxation was employed The lattice constants and internal coordinates were relaxed with the Perdew-Burke-Ernzerhof (PBE) at the Generalized Gradient Approximation (GGA) exchange-correlation functional The Broyden-Fletcher-Goldfarb-Shanno (BFGS) optimization method was used in this process The change of energy, maximum tolerance of the force, maximum stress, and maximum displacement were set as 0.02 eV/atom, 0.01 eV/Å, 0.03 Gpa and × 10−3 Å, respectively The energy cutoff in our work was set at 400 eV The calculated density of states is shown in Fig 3(b), which indicates that each silicon vacancy yields a magnetic moment of approximately 1.0 µB, which is consistent with previous studies.30 FIG (a) S value as a function of incident positron energy respectively for Al ion implanted sample 1E14, 1E15 and 1E16 (b) W-S plot results of different implanted samples 095201-4 Peng et al AIP Advances 6, 095201 (2016) FIG (a) The supercell structure of Si31C31(VSi)AlSi (b) The calculated total density of states (TDOS) of Si31C31(VSi)AlSi Magnetizations as a function of the magnetic field, performed by SQUID-VSM magnetometer at different temperatures, are shown in Fig The y-axis of the curve was taken to be the magnetization divided by the mass The M-H curves of the pristine sample at K and 300 K, with subtraction of the diamagnetism component of the plastic straw used for mounting samples, are shown in Fig 4(a) The clear diamagnetic property observed at 300 K corresponds to the magnetic property of the intrinsic silicon carbide single crystal It was found that the slopes of the curves were different at K and 300 K owing to the presence of a paramagnetic component However, compared with the diamagnetic background, the paramagnetic component at 300 K had a rather small value As a first-order approximation, we treated the sample at 300 K as totally diamagnetic and then the paramagnetic component at K could be calculated by subtracting the magnetization value measured at 300 K The paramagnetic component of Al implanted samples (1E14, 1E15 and 1E16) measured at K after subtracting the diamagnetic background are shown in Fig 4(b) All curves showed a paramagnetic behavior and a clear deviation from each other FIG (a) M-H curves of the pristine sample at K and 300 K Inset is the paramagnetic component at 5K (b) paramagnetic component of sample 1E14, 1E15 and 1E16 at K and the Brillouin fitting curve 095201-5 Peng et al AIP Advances 6, 095201 (2016) TABLE I The density of paramagnetic centers N and the value of J and g Samples Ja gb N (g −1)c 1E14 1E15 1E16 0.5 0.5 0.5 1.996 1.995 1.991 6.52 × 1014 1.14 × 1015 9.45 × 1014 aJ is the total angular momentum of corresponding point defect is the Lande factor extracted from ESR measurements c N is the calculated density of spins bg Here, we obtained the Landé factor g from the ESR measurement results and determined J = 1/2 from first-principles calculation results By using the Brillouin function to fit the paramagnetic M-H curve, we could calculate the value of N, which also represents the concentration of silicon vacancies The calculated results of N are shown in Table I It was found that the value of N increases at the beginning and then decreases with an increasing implantation dose, and its maximal value emerged at the implantation dose of × 1015 cm−2, which fits well with the intensity of the ESR spectra The hysteresis loops can also be observed in the area of the low magnetic field, as shown in Fig 4(b) According to the previous investigations, the existence of this weak ferromagnetism may arise from the exchange coupling of the J = 0.5 paramagnetic center.31 As the signal of the ferromagnetic component is rather weak, it will not affect the results of our paramagnetic fitting The S-parameter of the Doppler broadening spectra, measured in the slow positron annihilation experiment, was also used to confirm our fitting results of the concentration of point defects The S-parameter of the pristine sample, 1E14, 1E15 and 1E16 as a function of incident positron energy are shown in Fig 2(a) The implanted depth of the slow positron was calculated by the following equation: ( ) 40 1.6 R= E (3) ρ where R is the depth from the sample surface in a unit of nm, ρ is the density of 4H-SiC (3.24 g/cm3) and E is the energy (keV) of the incident slow positron.29 It can be seen that the S-parameter of the implanted samples is larger than that of the pristine sample in the low energy region (0.18–5 KeV), which indicates the effect of ion implantation near the sample surface The S-parameter of sample 1E16 in the high energy (5–15 KeV) region is smaller than that of sample 1E15 but larger than that of sample 1E14, which indicates the change of defect density near the sample surface These results are also in concordance with our previous Brillouin fitting results, which demonstrates the rationality of our characterization method In conclusion, a comprehensive method is presented to characterize the concentration of point defects in ion-implanted silicon carbide by fitting the paramagnetic component of the samples Three aluminum-implanted 4H-SiC samples were used to evaluate our characterization method The g-factor in the Brillouin function can be obtained from electron spin resonance spectroscopy The type of the paramagnetic center can be obtained from positron lifetime spectroscopy, which can be attributed to the contribution of the silicon vacancy The local paramagnetic moment of the silicon vacancy in an Al-doped 4H-SiC crystal was determined by first-principles calculation A superconducting quantum device was used for measuring the M-H curve of the samples With a Brillouin fitting and slow positron annihilation spectroscopy verification, the concentration of point defects in the implanted samples was precisely characterized This method is useful for characterizing the damage level of SiC The decreased tendency of the defect concentration with an increasing ion dose in our implantation experiment can possibly be attributed to the recombination of defects arising from the self-annealing effect ACKNOWLEDGEMENTS Authors would like to acknowledge that this work is supported by the National Natural Science Foundation of China (Grant No 51272202 and No 61234006) 095201-6 Peng et al AIP Advances 6, 095201 (2016) K Kawahara, J Suda, and T Kimoto, J Appl Phys 113, 033705 (2013) F.C Beyer, C.G Hemmingsson, S Leone, Y.C Lin, a Gällström, a Henry, and E Janzén, J Appl Phys 110, 123701 (2011) K Kawahara, G Alfieri, and T Kimoto, J Appl Phys 106, 013719 (2009) K Kawahara, M Krieger, J Suda, and T Kimoto, J Appl Phys 108, 023706 (2010) K Kawahara, J Suda, G Pensl, and T Kimoto, J Appl Phys 108, 033706 (2010) T Miyazawa and H Tsuchida, J Appl Phys 113, 083714 (2013) L Lilja, I.D Booker, J.U Hassan, E Janzen, and J.P Bergman, J Cryst Growth 381, 43 (2013) K Danno, D Nakamura, and T Kimoto, Appl Phys Lett 90, 202109 (2007) L Storasta and H Tsuchida, Appl Phys Lett 90, 062116 (2007) 10 P.B Klein, J Appl Phys 103, 033702 (2008) 11 R.C Barklie, M Collins, B Holm, Y Pacaud, and W Skorupa, J Electron Mater 26, 137 (n.d.) 12 J Isoya, T Umeda, N Mizuochi, N.T Son, E Janzén, and T Ohshima, Phys Status Solidi Basic Res 245, 1298 (2008) 13 N.T Son, P Carlsson, J Ul Hassan, E Janzén, T Umeda, J Isoya, a Gali, M Bockstedte, N Morishita, T Ohshima, and H Itoh, Phys Rev Lett 96, 055501 (2006) 14 M.E Zvanut and V V Konovalov, Appl Phys Lett 80, 410 (2002) 15 A Kawasuso, F Redmann, R Krause-Rehberg, T Frank, M Weidner, G Pensl, P Sperr, and H Itoh, J Appl Phys 90, 3377 (2001) 16 A Kawasuso, F Redmann, R Krause-Rehberg, M Weidner, T Frank, G Pensl, P Sperr, W Triftshäuser, and H Itoh, Appl Phys Lett 79, 3950 (2001) 17 K Kawahara, X.T Trinh, N.T Son, E Janzen, J Suda, and T Kimoto, Appl Phys Lett 102, 112106 (2013) 18 G Brauer, W Anwand, W Skorupa, S Brandstetter, and C Teichert, J Appl Phys 99, 023523 (2006) 19 A Kawasuso, F Redmann, R Krause-Rehberg, M Yoshikawa, K Kojima, and H Itoh, Phys Status Solidi 223, R8 (2001) 20 G Brauer, W Anwand, P Coleman, A Knights, F Plazaola, Y Pacaud, W Skorupa, J Störmer, and P Willutzki, Phys Rev B Condens Matter 54, 3084 (1996) 21 Y Liu, G Wang, S Wang, J Yang, L Chen, X Qin, B Song, B Wang, and X Chen, Phys Rev Lett 106, 087205 (2011) 22 Z.C Lv, X.P Ma, H.W Zheng, R An, C.X Peng, J.D Liu, B.J Ye, C.L Diao, X.Y Liu, and W.F Zhang, Mater Lett 93, 374 (2013) 23 B Song, H Bao, H Li, M Lei, T Peng, J Jian, J Liu, W Wang, W Wang, and X Chen, J Am Chem Soc 131, 1376 (2009) 24 Y Wang, L Li, S Prucnal, X Chen, W Tong, Z Yang, F Munnik, K Potzger, W Skorupa, S Gemming, M Helm, and S Zhou, Phys Rev B 89, 014417 (2014) 25 M Zhao, F Pan, and L Mei, Appl Phys Lett 96, 012508 (2010) 26 X Lin and F Pan, J Supercond Nov Magn 27, 1513 (2014) 27 K.W Lee and C.E Lee, Phys Rev Lett 97, 137206 (2006) 28 Q Xu, T Yoshiie, and M Okada, J Nucl Mater 386-388, 169 (2009) 29 S Jin, P Zhang, E Lu, L Guo, B Wang, and X Cao, Acta Mater 103, 658 (2016) 30 L Yu, H Jin, D Liu, Y Dai, M Guo, B Huang, and Z Zhang, Chem Phys Lett 496, 276 (2010) 31 Y Wang, Y Liu, E Wendler, R Hübner, W Anwand, G Wang, X Chen, W Tong, Z Yang, F Munnik, G Bukalis, X Chen, S Gemming, M Helm, and S Zhou, Phys Rev B 82, 174409 (2015) ... compared with the diamagnetic background, the paramagnetic component at 300 K had a rather small value As a first-order approximation, we treated the sample at 300 K as totally diamagnetic and...AIP ADVANCES 6, 095201 (2016) Concentration of point defects in 4H- SiC characterized by a magnetic measurement B Peng, R X Jia ,a Y T Wang, L P Dong, J C Hu, and Y M Zhang School of Microelectronics,... contribution of the silicon vacancy The local paramagnetic moment of the silicon vacancy in an Al-doped 4H- SiC crystal was determined by first-principles calculation A superconducting quantum device was