Accepted Manuscript Investigation of temperature dependent electrical characteristics on Au/Ni/β-Ga2O3 Schottky diodes Ang Li, Qian Feng, Jincheng Zhang, Zhuangzhuang Hu, Zhaoqing Feng, Ke Zhang, Chunfu Zhang, Hong Zhou, Yue hao PII:
S0749-6036(18)30725-0
DOI:
10.1016/j.spmi.2018.04.045
Reference:
YSPMI 5659
To appear in:
Superlattices and Microstructures
Received Date: 9 April 2018 Revised Date:
26 April 2018
Accepted Date: 27 April 2018
Please cite this article as: A. Li, Q. Feng, J. Zhang, Z. Hu, Z. Feng, K. Zhang, C. Zhang, H. Zhou, Y. hao, Investigation of temperature dependent electrical characteristics on Au/Ni/β-Ga2O3 Schottky diodes, Superlattices and Microstructures (2018), doi: 10.1016/j.spmi.2018.04.045. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Investigation of temperature dependent electrical characteristics on Au/Ni/β β-Ga2O3 Schottky diodes Ang Li1, Qian Feng1 , Jincheng Zhang1 , Zhuangzhuang Hu1, Zhaoqing Feng1, Ke
1
RI PT
Zhang1, Chunfu Zhang1, Hong Zhou1, Yue hao1, State Key Discipline Laboratory of Wide Bandgap Semiconductor Technology, School
SC
of Microelectronics, Xidian University, Xi’an 710071, China.
M AN U
Abstract Au/Ni/β-Ga2O3 Schottky barrier diodes were fabricated on a 15 µm drift layer mechanically exfoliated from (100)-oriented β-Ga2O3 bulk. The temperature dependent current density-voltage (J-V) characteristics from 300K to 550K were investigated and the barrier height and ideality factor were determined. Compared with the thermionic emission(TE) model, the forward J-V behavior follows thermionic field emission(TFE) model and the reverse J-V characteristics can be explained by Poole-Frenkel model.
AC C
EP
TE D
Index terms: Beta-Gallium Oxide, Schottky barrier diodes, Temperature dependent, Carrier transport mechanism
*
Corresponding Author: Electronic mail:
[email protected] (Qian Feng);
[email protected] (Jincheng Zhang)
ACCEPTED MANUSCRIPT
1. Introduction In recent years, β-Ga2O3 has been getting more and more attention as a promising material in power electronics and optoelectronics applications. Compared to the other wide bandgap semiconductors such as SiC and GaN, β-Ga2O3 has a larger band gap of
RI PT
4.8-4.9eV, a higher theoretical breakdown electrical field of 8MV/cm and a larger Baliga’s figure-of-merit(BOM) of 3400. In addition, the electron mobility of β-Ga2O3 has been estimated to be 300cm2/Vs1)). All the above properties make β-Ga2O3 superior for next generation high voltage and high power devices including Schottky diode(SBD)2)
and
metal-oxide-semiconductor
SC
barrier
field-effect
transistor(MOSFET)3). Moreover, the cost effective, large scale and high-quality
M AN U
Ga2O3 bulk can be grown by melt growth methods, such as Czochiralski, floating zone or edge-defined film-fed growth, which are much appropriate technologies for commercialization of this material4)6), As a unipolar rectifying device, Schottky barrier diodes possess the advantage of fast switching speed and low on-state loss over the bipolar rectifying devices. Thus, many investigations have been carried out
TE D
on the electrical characteristics of Ni, Au, Pt, Cu, Pd and TiN based β-Ga2O3 Schottky diodes2)7)-12). However, much less is known about carrier transport across the Schottky junctions under reverse bias, which is related to the device breakdown voltage, an
EP
important parameter of the SBD. In this paper, the Au/Ni/β-Ga2O3 Schottky barrier diodes were fabricated and the barrier height, ideality factor were determined in the
AC C
temperature range of 300-550K Furthermore, we studied the dependence of leakage current on temperature and electric-field under reverse bias and analyzed the carrier transport mechanism.
2. Experimental Procedure Fig. 1 shows the schematic cross section of the β-Ga2O3 Schottky barrier diode fabricated in this work. The drift layer with the thickness of 15 µm was obtained from Sn-doped (100)-oriented β-Ga2O3 bulk by mechanical exfoliation. Smooth surface was confirmed by atomic force microscope(AFM) image shown in Fig. 2(a) and the root mean square(RMS) was estimated to be 0.664nm. Furthermore, as presented in
ACCEPTED MANUSCRIPT Fig. 2(b), the X-ray diffraction rocking curve peak from (400) plane of the (100)-oriented β-Ga2O3 bulk has a full width at half-maximum of 35.1 arcsec, showing excellent crystal quality. Compared to epitaxial growth13), the defects can be greatly reduced in this way. Si ion implantation was carried out on the back side of the
RI PT
substrate, followed by E-beam evaporation of a Ti/Au (20nm/100nm) metal stack and rapid thermal annealing at 600°C for 60 seconds to form ohmic contact. Then circular Schottky anode electrodes with the diameter of 100 µm were fabricated on the front side by standard photolithography, evaporation of Ni/Au (40um/100um) films and
SC
lift-off.
3. Result and Discussion
M AN U
As shown in Fig. 3, the effective carrier concentration Nd-Na and built-in potential (eVbi) in drift layer were determined to be 2×1017 cm-3 and 1.26 eV, respectively, extracted from the slope and the intercept of the 1/C2 versus V plot of Ga2O3 SBD using the following equation,
(1)
TE D
1 2 = (Vbi − V ) 2 C qε A2 N d
Where A is the Schottky contact area, q is the electron charge and ε is the permittivity of 10ε0 for β-Ga2O3. In addition, the relationship between the Schottky barrier height
EP
eϕb and built-in potential eVbi can be expressed as: eϕb = eVbi + ( Ec − E f ) − e ϕ
(2)
AC C
where Ec and Ef are the conduction band minimum and Fermi level of β-Ga2O3, respectively, and e∆ϕ is the potential barrier lowering caused by the image force. Ec-Ef can be obtained using the following equations: Ec − E f = kT ln(
Nc ) Nd − Na
(3)
Where k is the Boltzmann constant, T is the temperature in K, Nc is the effective density of states of the conduction band, Nd-Na is the effective carrier concentration, m∗ is the effective electron mass of β-Ga2O3, 0.34m0,and h is Planck constant. So the ϕb_cv is found to be 1.33 eV at room temperature.
ACCEPTED MANUSCRIPT The current density-voltage (J-V) curves from 300K to 550K were obtained using Keithley 4200 semiconductor characterization system, as shown in Fig. 4(a) and Fig. 4(b). Fig. 4(c) and its inset show room temperature forward and reverse J-V characteristics in logarithmic scale and forward J-V characteristic in linear scale,
RI PT
respectively. As shown in Fig. 4(c), the current on/off ratio and reverse breakdown voltage Vbr were found to be 1011 and 97V, respectively, at room temperature. The specific on-resistance (Ron), 2.1 mΩ⋅cm2 was determined from the slope of the fitting line to the linear region of J-V curve in the inset of Fig. 4(c). According to thermionic
qV ) nkT
ϕb _ JV
M AN U
J = J S exp(
SC
emission (TE) theory14)-15), the SBD J-V characteristic can be described as follows:
J s = A*T 2 exp(−
kT
)
(4) (5)
Where Js, n, A* and ϕb_JV are saturation current density, ideality factor, Richardson’s constant of 41.1 A cm−2K−2 for β-Ga2O3 at room temperature16) and Schottky barrier height extracted from J-V curves, respectively. Table 1 gives the values of n and ϕb_JV
TE D
obtained from Fig. 4(a) based on equation (4) and (5). It is obvious that both n and ϕb_JV are temperature dependent and show a serious deviation from unity and ϕb_cv at room temperature, respectively. The Richardson plot of ln(Js/T2) versus 1/kT is shown
EP
in Fig. 5 in the temperature range of 300K-550K17)-18). A* and ϕb were extracted to be 7.15 A/cm2K2 and 1.07 eV, respectively. The values of A* is much comparable to the
AC C
theoretical Richardson constant A*=41.1 A/cm2K2, indicating the lower barrier height and the ideality factor much larger than unity cannot be ascribed to the Schottky barrier height(SBH) inhomogeneity. So, there might be other transport mechanisms contributing to carrier transport. Taking the tunneling effect into account and according to the thermionic field emission (TFE) model19)-20), the relationship between J and V can be described by: J = J 0 exp( E0 = E00 coth(
qV ) E0
E00 ) = nkT kT
(6) (7)
ACCEPTED MANUSCRIPT A*T π E00 (ϕbtfe − qV − Vn ) V ϕbtf e − Vn J0 = × exp(− n − ) (8) E00 kT E 0 k cosh( ) kT E00 =
h N d 1/2 ( ) 4π m*ε s
(9)
RI PT
Where J0 is saturation current density for TFE, Vn is the energy difference between Ec and Ef, ϕbtfe is Schottky barrier height for TFE and E00 is characteristic tunneling energy. The values of ϕbtfe and n extracted from J-V-T plots are summarized in Table 1. At room temperature, the barrier height of ϕbtfe=1.32 eV is quite consistent with the
SC
value of ϕb_cv.
M AN U
For three different carrier transport mechanisms (i.e., TE, TFE and field emission (FE) model), the TE is dominant when E00/kT<<1, the TFE when E00/kT≈1 and the FE when E00/kT>>1. The E00/kT determined from Fig. 4(a) are summarized in Table 2. For the values of E00/kT are very close to unity, the TFE may probably be the dominant carrier transport mechanism for Ga2O3 SBD. Since the field-dependent
TE D
effect is obvious in forward characteristics, it will be more serious under reverse bias for the even larger electrical field. The investigation of reverse J-V characteristics was also carried out in the temperature range of 300-550K, as shown in Fig. 4(b). Major reverse current transport mechanisms, i.e., Poole-Frenkel emission21) and Schottky
EP
emission22), were considered. The reverse current based on Poole-Frenkel emission can be described by
AC C
I ∝ E exp(
q kT
qE
πε s
)
(10)
And the reverse current dominated by Schottky emission is given by
Where
E=
E
is
2q ( N d − N a )
εs
I ∝ T 2 exp(
the
(V + Vbi −
maximum
kT ). q
q 2kT
qE
πε s
)
electric
(11)
field,
determined
by
ACCEPTED MANUSCRIPT Fig. 6 depicts the temperature dependent plots of (a) ln (I/E) versus E1/2 for Poole-Frenkel emission and (b) ln (I/T2) versus E1/2 for Schottky emission. All the plots are found to be linear, indicating both Poole-Frenkel and Schottky emissions are present23). In order to confirm which mechanism is dominant, the emission coefficient
q nkT
q
πε s
was calculated, where n=1 for Poole-Frenkel emission and n=2 for
RI PT
S=
Schottky emission21). The calculated and experimental values of S from 300K to 550K are presented in Table 3. For Schottky emission, the experimental values are
SC
almost 4~7 times larger than calculated values at different temperature, while the experimental values are much closer to theoretical values, only 1.5~2.5 times as large
M AN U
as that of ideal ones for Poole-Frenkel emission. Thus, current follows Poole-Frenkel emission under reverse bias in the temperature range of 300-550K.
4. Conclusion
In summary, Ni/β-Ga2O3 Schottky diodes were fabricated on a 15µm Ga2O3 layer mechanically exfoliated from (100)-oriented Ga2O3 bulk and the dependence of J-V
TE D
characteristics on the temperature were investigated. The diodes show satisfactory performances such as high current on-off ratio (1011), low specific on-resistance (2.1 mΩ⋅cm2) and high Schottky barrier height (1.33 eV). From the Richardson plot of
EP
ln(Js/T2) versus 1/kT, A* was calculated to be 7.15 A/cm2K2, much comparable to the theoretical Richardson constant A*=41.1 A/cm2K2, suggesting the deviation of
AC C
ideality factor from unity at room temperature cannot be attributed to the inhomogeneity of SBH. Further analysis indicated the forward current transport was controlled by the TFE mechanism while the Poole-Frenkel emission was the dominant carrier transport mechanism under reverse bias.
Acknowledgments This work was supported by the National Natural Science Foundation of China (NSFC) under Grant Nos.61774116, 61334002 and the 111 Project (B12026).
ACCEPTED MANUSCRIPT
Reference 1) M. Higashiwaki, K. Sasaki, A. Kuramata, T. Masui and S. Yamakoshi, Appl. Phys. Lett. 100 (2012) 013504. 2) M. Higashiwaki, K. Sasaki, K. Goto, K. Nomura, Q. T. Thieu, R. Togashi, H.
RI PT
Murakami, Y. Kumagai, B. Monemar, A. Koukitu, A. Kuramata and S. Yamakoshi, Device Research Conference (DRC), 2015 73rd Annual. IEEE, (2015) 29-30.
3) M. Higashiwaki, K. Sasaki, M. H. Wong, T. Kamimura, D. Krishnamurthy, A.
IEEE International. IEEE, (2013) 28.7. 1-28.7. 4.
SC
Kuramata, T. Masui and S. Yamakoshi, Electron Devices Meeting (IEDM), 2013
M AN U
4) N. Ueda, H. Hosono, R. Waseda and H. Kawazoe, Appl. Phys. Lett. 70 (1997) 3561-3563
5) H. Aida, K. Nishiguchi, H. Takeda, N. Aota, K. Sunakawa and Y. Yaguchi, Jpn. J. Appl. Phys 47 (2008) 8506-8509.
6) A. Kuramata, K. Koshi, S. Watanabe1, Y. Yamaoka, T. Masui1 and S.
TE D
Yamakoshi, Jpn. J. Appl. Phys 55 (2016) 1202A2. 7) A. Jayawardena, A. C. Ahyi and S. Dhar, Semicond. Sci. Technol. 31 (2016) 115002 (6pp).
EP
8) M. Odal, J. Kikawa, A. Takatsuka, R. Tokudal, T. Sasakil, K. Kaneko, S. Fujita, and T. Hitora, Device Research Conference (DRC), 2015 73rd Annual. IEEE,
AC C
(2015)137-138.
9) M. Higashiwaki, K. Konishi, K. Sasaki, K. Goto, K. Nomura, Q. T. Thieu, R Togashi, H. Murakami, Y. Kumagai, B. Monemar, A. Koukitu, A. Kuramata and S. Yamakoshi, Appl. Phys. Lett. 108, 133503 (2016).
10) D. Splith, S. Müller, F. Schmidt, H. v. Wenckstern, J. J. v. Rensburg, W. E. Meye, M. Grundmann, Volume211, Issue1, January 2014 Pages 40-47. 11) M. E. Ingebrigtsen, L. Vines, G. Alfieri, A. Mihaila, U. Badstübner, B. G. Svensson, A. Kuznetsov, Materials Science Forum (Vol. 897, pp. 755-758). 12) M. J. Tadjer, V. D. Wheeler, D. I. Shahin, C. R. E. Jr and F. J. Kub, ECS J. Solid
ACCEPTED MANUSCRIPT State Sci. Technol. 2017 volume 6, issue 4, P165-P168. 13) D. Gogova, M. Schmidbauer and A. Kwasniewski. Cryst. Eng. Comm 17 (35) (2015) 6744-6752. 14) L. Wang, M. I. Nathan, T. Lim, M. A. Khan, and Q. Chen, Appl. Phys.Lett. 68,
RI PT
1267 (1996). 15) Michael Shur, Physics of Semiconductor Devices (Prentice-Hall, Engel-wood Cliffs, NJ, 1990).
16) H. He, R. Orlando, M. A. Blanco, R. Pandey, E. Amzallag, I. Baraille, and M.
SC
Rérat, Phys. Rev. B, vol. 74, no. 19, pp. 195123-1–195123-8, Nov. 2006.
17) Sze S 1981 Physics of Semiconductor Devices 2nd edn (NewYork: Wiley) pp
M AN U
245–311
18) Schroder D 2006 Semiconductor Material and DeviceCharacterization 3rd edn (New Jersey: Wiley-IEEE Press) p160
19) H. Kim, J. Ryou, R. D. Dupuis, S. N. Lee, Y. Park, J. Weon, and T. Y.Seong, Appl. Phys. Lett. 93, 192106 (2008).
2505 (2006).
TE D
20) Y. J. Lin, W. X. Lin, C. T. Lee, and H. C. Chang, Jpn. J. Appl. Phys. Part1, 45,
21) V. Janardhanam, A. Jyothi, K. S. Ahn, and C. J. Choi, Thin Solid Films 546, 63
EP
(2013).
22) J. Lin, S. Banerjee, J. Lee, and C. Teng, IEEE Electron Device Lett. 11, 191
AC C
(1990).
23) D. Tomer, S. Rajput, L. J. Hudy, C. H. Li, and L. Li, Appl. Phys. Lett. 106, 173510 (2015).
ACCEPTED MANUSCRIPT List of figure captions Fig. 1 schematic cross section of the Schottky barrier diode Fig. 2 (a) surface atomic force microscope image and (b) X-ray diffraction rocking
Fig. 3 1/C2-V characteristic curve of Ga2O3 SBD
RI PT
curve peak from (400) plane of the (100)-oriented β-Ga2O3 bulk.
Fig. 4 (a) forward J-V characteristics and (b) reverse J-V characteristics in logarithmic scale in the temperature range of 300-550K and (c) room temperature forward
characteristic in linear scale.
Fig. 5 the Richardson’s plot of ln(Js/T2) versus 1/kT
SC
and reverse J-V curve in logarithmic scale and forward current J-V
M AN U
Fig. 6 Temperature dependent plots of (a) ln (I/E) versus E1/2 for Poole-Frenkel
AC C
EP
TE D
emission and (b) ln (I/T2) versus E1/2 for Schottky emission.
ACCEPTED MANUSCRIPT
Sn-doped Ga 2O3 (15 µm) Ti/Au(20 nm/100 nm) Fig. 1 5
6
6
7 8 9 2.35 nm
Rocking Curve (400)
5
2 3
4
M AN U
1.00
SC
3 4
Intensity (105 a.u.)
0 µm 1 2 0 1
RI PT
Au(100 nm) Ni (40nm)
4 5
0.00
6 7
3
FWHM=35.1 arcsec
2 1
-1.00
0
8 9
14.98
(a)
TE D
-2.22
AC C
EP
Fig. 2
Fig. 3
15.00
15.02 ω (deg.)
(b)
15.04
SC
RI PT
ACCEPTED MANUSCRIPT
(b)
AC C
EP
TE D
M AN U
(a)
(c) Fig. 4
EP
TE D
M AN U
Fig. 5
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
(a)
(b) Fig. 6
ACCEPTED MANUSCRIPT Table 1: Values of n and ϕb based on TE and TFE
300 350 400 450 500 550
TE n 1.428491 1.300885 1.217943 1.176357 1.118972 1.057442
TFE ϕb_JV (eV) 1.027321 1.098278 1.148136 1.161449 1.191984 1.210671
n 1.155382 1.131039 1.12074 1.110743 1.087805 1.05749
ϕ b_JV (eV) 1.32494 1.37358 1.35473 1.37419 1.34452 1.39851
RI PT
Temperature (K)
0.017947 0.019181 0.021020 0.022625 0.022333 0.020719
TE D
300 350 400 450 500 550
E00 (eV)
E00/KT
0.693462 0.635275 0.609158 0.582821 0.517767 0.417692
M AN U
Temperature (K)
SC
Table 2: The experimental E00 and E00/kT for different temperature
Table 3: The calculated and experimental values of S between 300K to 550K Poole-Frenkel emission (V/cm)1/2 Fit
Calculated
Fit
0.009224 0.008021 0.006833 0.00615 0.005591 0.004986
0.02215 0.02103 0.01897 0.01506 0.01335 0.00728
0.004612 0.004011 0.003416 0.003075 0.002795 0.002493
0.02519 0.02407 0.02802 0.02111 0.0164 0.01033
AC C
300 350 400 450 500 550
Calculated
EP
Temperature (K)
Schottky emission (V/cm)1/2
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Highlights 1. Au/Ni/β-Ga2O3 Schottky diodes were fabricated on a 15µm Ga2O3 layer mechanically exfoliated from (100)-oriented Ga2O3 bulk.
RI PT
2. The diodes show satisfactory performances such as high current on-off ratio (1011), low specific on-resistance (2.1 mΩ⋅cm2), high
SC
breakdown voltage (97V) and high Schottky barrier height (1.33 eV) at room temperature.
M AN U
3. By analyzing forward current density-voltage (J-V) characteristics from 300K to 550K, we prove forward current transport was controlled by the TFE mechanism.
TE D
4. By analyzing reverse (J-V) characteristics from 300K to 550K, we prove the Poole-Frenkel emission was the dominant carrier transport
AC C
EP
mechanism under reverse bias.