Abstract
Silicon carbide (SiC) is a promising material for power devices because of its wide bandgap and high thermal conductivity. However, carbon remaining in the metal–SiC interface (carbon cluster) adversely affects the ohmic contacts. As a result, the ohmic contact characteristic worsens and the on-state power loss of the device increases. In addition, the carbon cluster in the metal–SiC interface degrades the device’s high-temperature thermal stability. In this paper, Co/Si/Co/Si is used as an ohmic contact material to reduce the carbon cluster and improve the thermal stability. The specific contact resistivity was measured, and the effect of the Co:Si composition ratio and the annealing condition for the ohmic contact was analyzed. The thermal stability was investigated by measuring the specific contact resistivity through a thermal duration test. The measurement results were analyzed using the X-ray diffraction method.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
Silicon carbide (SiC) has a bandgap three times higher than Si and good thermal conductivity [1]. These material characteristics make SiC suitable for power devices [2]. In this study, 4H–SiC was used, which has a large bandgap, and good electron and hole mobility [3]. 4H–SiC has suitable material characteristics as a power device material; however, it has limitations in terms of forming ohmic contacts. Metal silicide is formed at the metal and 4H–SiC junction, and the remaining carbon collects on the surface (carbon cluster). Carbon clusters degrade the ohmic contact characteristics and lowers the thermal stability of the device [4]. Therefore, it is necessary to reduce the carbon cluster to lower the specific contact resistivity and improve the thermal stability of the device [5].
In this study, CoSi2 is used as barrier material to form ohmic contacts with a low specific contact resistivity and improve the thermal stability [6]. CoSi2 is a material used for ohmic contact of various semiconductors such as SiGe [7]. Especially, CoSi2 has a low specific contact resistivity with 4H–SiC, and it is stable at high temperatures [8, 9]. However, when silicide is formed after Co is deposited on 4H–SiC, it is difficult to suppress the occurrence of carbon clusters. This is because 4H–SiC supplies Si necessary for silicide formation, and C generates carbon clusters. Therefore, this study suppressed the occurrence of carbon clusters using Co/Si/Co/Si as a barrier layer [10]. Si in the Co/Si/Co/Si multilayer reacts with Co to form a silicide. Therefore, the silicide is formed without breaking the Si–C bond, and the carbon cluster can be reduced. This investigation analyzed the effect of the Co:Si composition ratio, and the annealing conditions on ohmic contact quality and thermal stability in Co/Si/Co/Si. Through our research, it will be possible to obtain the most optimal ohmic contact formation condition using Co/Si/Co/Si multilayer.
This study is organized as follows. Section 2 describes the experimental conditions and the ohmic contact fabrication. Section 3 discusses the results, and analyzes the ohmic contact characteristics and thermal stability according to the Co:Si composition ratio and the annealing conditions. In addition, X-ray diffraction was applied to analyze the results. Finally, Sect. 4 presents the conclusion of this paper.
2 Experiments
2.1 Contact fabrication
The wafer for ohmic contact formation was 4° off-axis and has a Si-face (0001) 4H–SiC. The wafer consists of a substrate, a buffer layer, and an epitaxy layer. The n-type substrate has a concentration of 5 × 18 cm−3, and the n-type buffer has a thickness and concentration of 0.5 μm and 1 × 18 cm−3, respectively. The epitaxy layer has a thickness and concentration of 21 μm and 3.8 × 15 cm−3, respectively.
High-concentration n- and p-type ion implantations were performed on 4H–SiC for ohmic contact. High doping forms the source/drain of MOSFET. As n- and p-type dopants, nitrogen and aluminum, which are commonly used in power devices, were used. The doses of nitrogen and aluminum were 1 × 15 cm−2, and the implant conditions were 200 keV, 250 °C and 20 s [11, 12]. Post-implantation annealing is required to activate the implanted dopant. However, Si evaporation and surface roughness occur during post-implantation annealing at 1700 °C. To prevent this, a capping layer made out of graphite was used [13]. PR was coated on the 4H–SiC surface and burned at 600 °C. After forming the graphite capping layer, the wafer was annealed at 1700 °C for 30 min. After activating the dopant, the graphite capping layer was removed using ashing and wet cleaning.
After ion implantation, 6000 Å of SiO2 was deposited using STS PECVD. PR patterning was performed through HMDS coating, PR coating, soft baking, aligning, hard baking, and developing the PR. Afterwards, SiO2 was etched to create a circular transmission line measurement (CTLM) pattern. Co/Si/Co/Si multilayers were deposited to fit the CTLM pattern. Co/Si/Co/Si multilayer was deposited by sputtering method using KVS-T8860 under condition of 100 W and 1 mtorr. The multilayer deposited on SiO2 was removed by performing the lift off method.
After deposition, two-step annealing was performed to form Co silicide, as a barrier material [14]. Al was deposited to produce the final sample for measure resistance. The multilayer deposition thickness and annealing conditions are described in Sect. 2.2, which discusses the experimental conditions. The wafer preparation and ohmic contact fabrication processes are illustrated in Figs. 1 and 2, respectively.
The ion implant implantation process was performed at the Korea Electrotechnology Research Institute (KERI) using E220. The high-temperature annealing was performed at the National Institute for Nanomaterials Technology (NINT) using ALPHA-805D. DC measurement was performed by FPP-5000 4-POINT PROBE. Metallization is just formed by depositing Al. X-ray diffraction was performed at Yonsei Center for Research Facilities (YCRF) using SmartLab.
2.2 Experimental conditions
2.2.1 Co:Si composition ratio variation
The Co:Si composition ratios of the Co/Si/Co/Si multilayers were split into 1:2, 1:2.5, and 1:3 to form CoSi2, a thermodynamically stable phase. The overall thickness of the Co/Si/Co/Si multilayer was deposited at 84 nm. The Co/Si/Co/Si multilayers with thicknesses of 14/28/14/28 nm, 12/30/12/30 nm, and 10.5/31.5/10.5/31.5 nm were deposited for 1:2, 1:2.5, and 1:3 ratios, respectively.
2.2.2 Annealing temperature variation
Co/Si/Co/Si multilayers have a two-step annealing procedure to form the Co silicide. The experiments were conducted to investigate the effect of first-step annealing on the ohmic contacts. The first annealing temperature was split into 500 °C, 600 °C, 700 °C, 750 °C, 800 °C, 850 °C, and 900 °C.
3 Measurement results
3.1 Co:Si composition ratio variation
3.1.1 Ohmic contact characteristic
Figure 3 displays the ohmic contact characteristics for n- and p-type 4H–SiC with Co:Si composition ratios of 1:2, 1:2.5, and 1:3. The annealing condition was 600 °C for 4 min and 950 °C for 1 min. The reason for setting the second annealing temperature to 950 °C is because it follows the temperature of [15]. The measurement results show the ohmic contact characteristics under all of the conditions. When the Co:Si composition ratio is 1:2.5, it has the smallest resistance. The specific contact resistivity according to the Co:Si composition ratio is listed in Table 1. The reason for these results can be found in the formation of the Co silicide, because CoSi2 has a low resistivity [16] and is mostly formed when the Co:Si composition ratio is 1:2.5. A detailed analysis is covered in Sect. 3.3.
3.1.2 Thermal stability
Figure 4 displays the specific contact resistivity as the temperature varies from 25 to 400 °C. In Fig. 6, the specific contact resistivity decreases rapidly between 25 and 200 °C, but becomes constant after 200 °C. Between 25 and 200 °C, more free carriers are produced as the temperature increases, thus lowering the specific contact resistivity. However, after 200 °C, the change in the specific contact resistivity decreases because the mobility and temperature dependency of the ionization are opposed to each other [17]. At high temperatures, the free carrier density increases with the temperature; however, the mobility decreases with phonon scattering. As a result, the specific contact resistivity becomes constant. In the n-type, the specific contact resistivity decreases by 93.4%, 91.3%, and 93.2%, when the Co:Si composition ratios are 1:2, 1:2.5, and 1:3, respectively, according to the temperature change (25 to 400 °C). For the p-type, the specific contact resistivity decreases by 94.0%, 92.6%, and 93.7% when the Co:Si composition ratios are 1:2, 1:2.5, 1:3, respectively. As a result, the variation of the specific contact resistivity was the smallest for the Co:Si composition ratio of 1:2.5 for both n- and p-types. The measurement results are presented in Tables 2 and 3.
Figure 5 depicts the specific contact resistivity for 20 h of exposure at 400 °C. In the n-type, the specific contact resistivity variation of the Co:Si composition ratios of 1:2, 1:2.5, and 1:3 are 7.47%, 3.54%, and 5.23%, respectively. In the p-type, the specific contact resistivity of the Co:Si composition ratios of 1:2, 1:2.5, and 1:3 were changed by 6.11%, 2.85%, and 5.37%, respectively. In the case of the Co:Si composition ratio of 1:2.5 for both n- and p-types, the variation of the specific contact resistivity was smallest. This is because CoSi2 had the most formation, as it is the most thermodynamically stable phase [18]. The specific contact resistivity variations are presented in Table 4.
3.2 Annealing temperature variation
3.2.1 Ohmic contact characteristics
Figure 6 shows the ohmic contact characteristics of n- and p-type 4H–SiC according to the annealing conditions. The Co:Si composition ratio is 1:2.5. When the first-step annealing temperature is 500 °C, 600 °C, 700 °C, 750 °C, and 800 °C, it shows ohmic contact characteristics. In addition, it can be seen that as the temperature increases up to 750 °C, and the amount of current flowing at the same voltage increases. This means that the higher the first-step annealing temperature (i.e., up to 750 °C), the smaller the resistance. When the temperature reaches 800 °C, a small amount of carbon cluster is formed, and an ohmic contact with poor characteristics is formed. When the temperature is above 850 °C, the formation of carbon clusters increases and the ohmic contacts are not properly formed. The specific contact resistivity according to the first annealing temperature is shown in Table 5. When the first-step annealing temperature is 750 °C, the smallest specific contact resistivity is shown for the n- and p-types.
3.2.2 Thermal stability
Figure 7 shows the specific contact resistivity as the temperature varies from 25 to 400 °C. When the first-step annealing temperature was 500 °C, 600 °C, 700 °C, 750 °C, and 800 °C, ohmic contact was formed and the specific contact resistivity could be measured. However, at 850 °C and 900 °C, ohmic contact was not formed; therefore, the specific contact resistivity was not measured. As the first-step annealing temperature increases (up to 750 °C), the specific contact resistivity variation decreases. However, at 800 °C, the thermal stability decreases because the formation of the carbon cluster begins [19]. The specific contact resistivity with the temperature is presented in Tables 6 and 7. When the first-step annealing temperature is 750 °C, it can be seen that the specific contact resistivity varies when the temperature is the smallest.
Figure 8 depicts the specific contact resistivity during 20 h of exposure at 400 °C. When the first-step annealing temperature was 850 °C and 900 °C, ohmic contact was not formed; therefore, the specific contact resistivity could not be measured. As described in Table 8, the higher the first-step annealing temperature (up to 750 °C), the smaller the variation of the specific contact resistivity. However, the variation of the specific contact resistivity increases when the first-step annealing temperature is 800 °C. As a result of 20 h exposure at 400 °C, it is most thermally stable when the first-step annealing temperature is 750 °C.
3.3 X-ray diffraction analysis
3.3.1 Co:Si composition ratio variation
Figure 9 shows the X-ray diffraction (XRD) results when the Co:Si composition ratio is (a) 1:2, (b) 1:2.5, and (c) 1:3 after the first-step annealing process. More CoSi2 was formed when the Co:Si composition ratio was 1:2.5 and 1:3 in comparison to 1:2. This is because sufficient Si is supplied to form the Co silicide by the Co/Si/Co/Si multilayer. Figure 10 displays the XRD results according to the Co:Si composition ratio after the second-step annealing process. When the Co:Si composition ratio was (a) 1:2, the CoSi2 could not be formed due to the lack of Si. When the Co:Si composition ratio was (b) 1:2.5, sufficient Si was supplied and CoSi2 was formed. In the case of (c) 1:3, because a large amount of Si was supplied, Si remaining after the forming of CoSi2. As a result, when the Co:Si composition ratio is 1:2.5, CoSi2 was formed while suppressing the carbon cluster most efficiently. Therefore, the specific contact resistivity was the smallest and the thermal stability was the best.
3.3.2 Annealing temperature variation
Figure 11 shows the XRD results when the Co:Si composition ratio is 1:2.5 after the first-step annealing condition is (a) 500 °C, (b) 750 °C, and (c) 900 °C for 4 min. As the temperature of the first-step annealing increases, the formation of CoSi increases. However, the carbon cluster is formed when the temperature is more than 800 °C. Figure 12 shows the XRD results after two-step annealing when the Co:Si composition ratio is 1:2.5. The first-step annealing conditions were (a) 500 °C, (b) 750 °C, and (c) 900 °C for 4 min. The second-step annealing conditions were all at 950 °C for 1 min. The formation of CoSi2 increases when the first-step annealing temperature is up to 750 °C. However, after 800 °C, the carbon cluster is formed. The higher the first-step annealing temperature is the more energy is supplied to break the Co–Co and Si–Si bonds. As a result, the formation of CoSi increases, which causes the formation of CoSi2 during second-step annealing. However, when the temperature of the first-step annealing increases above 800 °C, the Si–C bond is broken. As a result, the unreacted carbon collects on the surface and degrades the ohmic contact quality.
4 Conclusion
In this paper, a Co/Si/Co/Si multilayer was used as a barrier material to form the 4H–SiC ohmic contact using Co silicide. To form Co silicide efficiently, experiments were carried out with varying Co:Si composition ratios and annealing conditions. This study aimed to form CoSi2 so it has a low resistivity and is thermodynamically stable. When the ratio of Si to Co was low, there was not enough Si to form CoSi2. In contrast, when the Si ratio is high, the Si remains and the contact characteristics are poor. There is an increase of CoSi2 formation at higher first annealing temperatures. However, if the temperature is too high, the SiC bond is broken and a carbon cluster is formed, thus resulting in poor contact characteristics. As a result, when the Co:Si composition ratio is 1:2.5 and the annealing condition is 750 °C for 4 min and 950 °C for 1 min, then CoSi2 is best formed. In this condition, it has the lowest specific contact resistivity and is thermally stable. As such, contact using the Co/Si/Co/Si multilayer has a low specific contact resistivity and is thermally stable at high temperatures. This makes it suitable for contact barrier materials for power devices.
References
B.J. Baliga, Modern Power Devices (Wiley, New York, 1987)
W.J. Choyke, Silicon Carbide Recent Major Advances (Springer, Berlin, 2004)
B.J. Baliga, Fundamentals of Power Semiconductor Devices (Springer, New York, 2008)
M.G. Rastegaeva, A.N. Andreev, A.A. Petrov, A.I. Babanin, M.A. Yagovkina, I.P. Nikitina, The influence of temperature treatment of the formation of Ni-based schottky diodes and ohmic contacts to n-6H-SiC. Mater. Sci. Eng. 46, 254–258 (1997)
D. Bae, G. Ahn, C. Jeong, K. Kim, Ni/W/Ni ohmic contacts fot both n- and p-type 4H–SiC. Electr. Eng. 100, 2431–2437 (2018)
N. Lundberg, M. Ostling, CoSi2 ohmic contacts to n-type 6H–SiC. Solid State Electron. 38, 2023–2028 (1995)
O. Nur, M. Willander, CoSi2/Si1−xGex/Si(001) heterostructures formed through different reaction routes: silicidation-induced strain relaxation, defect formation, and interlayer diffusion. J. Appl. Phys. 78, 7063 (1995)
Q Xiang, B Yu, H Wang, M Lin, Dig, High performance sub-50nmCMOS with advanced gate stack, in 2001 Symposium on VLSI Technology. Digest of Technical Papers (IEEE, Kyoto, 2001), pp. 23–24
J.A. Edmond, J. Ryu, J.T. Glass, R.F.J. Davis, Electrical contacts to beta silicon carbide thin films. Electrochem. Soc. 135, 359–362 (1988)
N.I. Cho, K.H. Jung, Y. Choi, Improved ohmic contact to the n-type 4H-SiC semiconductor using cobalt silicides. Semicond. Sci. Technol. 19, 306–310 (2004)
G.K. Reeves, Specific contact resistance using a circular transmission line model. Solid State Electron. 23, 487–490 (1980)
A. Frazzetto, F. Giannazzo, R. Lo Nigro, V. Raineri, F. Roccaforte, Structural and transport properties in alloyed Ti Al Ohmic contacts formed on p-type Al-implanted 4H-SiC annealed at high temperature. J. Phys. D 44, 255302 (2011)
Y. Negoro, T. Kimoto, H. Matsunami, Electronic behaviors of high dose phosphorus-ion implanted 4H-SiC (0001). J. Appl. Phys. 96, 224–228 (2004)
J.P. Gambino, E.G. Colgan, Silicides and ohmic contacts. Mater. Chem. Phys. 52, 99–146 (1998)
K. Smedfors, C.M. Zetterling, M. Ostling, Sputtered ohmic cobalt silicide contacts to 4H–SiC. Mater. Sci. Forum 821, 440–443 (2015)
S.J. Yang, C.K. Kim, I.H. Noh, S.W. Jang, K.H. Jung, N.I. Cho, Study of Co- and Ni-based ohmic contacts to n-type 4H–SiC. Diam. Relat. Mater. 13, 1149–1153 (2004)
F. Zhao, M.M. Islam, C.F. Huang, Study of SiO2 encapsulation for aluminum and phosphorus implant activation in 4H–SiC. Mater. Lett. 64, 2593–2596 (2010)
M.C. Poon, C.H. Ho, F. Deng, S.S. Lau, H. Wong, Thermal stability of cobalt and nickel silicides. Microelectron. Reliab. 38, 1495–1498 (1998)
S.W. Park, Y.I. Kim, J.S. Kwak, H.K. Baik, Investigation of Co/SiC Interface Reaction. J. Electron. Mater. 26, 172–177 (1997)
Acknowledgements
This research was supported by the Ministry of Trade, Industry, and Energy (MOTIE) (Ministry of Science and ICT (MSIT), Korea, under the Information Technology Research Center (ITRC) support program (IITP-2020-2018-0-01421). This study was also supervised by the Institute for Information & Communications Technology Promotion (IITP).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Kim, T., Kim, J. & Kim, K. Processing and characterization of co silicide ohmic contacts to 4H–SiC. J Mater Sci: Mater Electron 31, 16299–16307 (2020). https://doi.org/10.1007/s10854-020-04178-w
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10854-020-04178-w