http://www.diva-portal.org
Postprint
This is the accepted version of a paper published in IEEE Electron Device Letters. This paper has been
peer-reviewed but does not include the final publisher proof-corrections or journal pagination.
Citation for the original published paper (version of record):
Elahipanah, H., Salemi, A., Zetterling, C-M., Östling, M. (2015)
5.8-kV Implantation-Free 4H-SiC BJT With Multiple-Shallow-Trench Junction Termination
Extension.
IEEE Electron Device Letters, 36(2): 168-170
http://dx.doi.org/10.1109/LED.2014.2386317
Access to the published version may require subscription.
N.B. When citing this work, cite the original published paper.
Permanent link to this version:
http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-163477
Abstract—Implantation-free 4H-SiC BJTs with multiple-
shallow-trench junction termination extension (ST-JTE) have
been fabricated. The maximum current gain of 40 at a current
density of 370 A/cm
2
is obtained for the device with active area of
0.065 mm
2
. A maximum open-base breakdown voltage (BV) of
5.85 kV is measured, which is 93% of the theoretical breakdown
voltage. A specific on-resistance (R
ON
) of 28 mΩ.cm
2
was
obtained.
Index Terms—4H-SiC, multiple-shallow-trench JTE,
implantation-free, high-voltage BJT.
I. INTRODUCTION
ILICON carbide bipolar junction transistors (BJTs) in the
3–7-kV range with low on-resistance (R
ON
) are very
attractive devices for industrial applications due to their low
loss and fast switching speed. Besides, gaining lower on-
resistance results in lower loss which is as important as
improving the breakdown voltage to its theoretical value.
Several termination techniques have been developed and
optimized to approach the ideal breakdown voltage and
improve the termination efficiency [1-12]. Among them,
implantation-free devices with low R
ON
have the advantage of
preventing life-time-killing defects and current degradation
caused by ion implantation [10]. These devices are based on
etched junction termination extensions (JTEs) in which the
remaining dose of the JTEs is controlled by the etching depth.
Table I summarized the recent results reported for SiC BJTs.
In this letter, an implantation-free 4H-SiC BJT with
multiple-shallow-trench JTE (ST-JTE) termination is
demonstrated. The effect of material and etching uncertainties
is investigated by numerical simulation and compared with
experimental data. The fabricated ST- JTE device has three
advantages. First, it is completely free of ion implantation and
the following high temperature activation annealing to avoid
any further defects caused by it. Second, compared with multi-
implanted or graded-etched JTEs, the trench structures are
Manuscript received October 29, 2014; revised November 19 and
December 04, 2014; This work was supported by the Swedish Foundation for
Strategic Research.
The authors are with the School of Information and Communication
Technology, KTH Royal Institute of Technology, Kista SE-164 40, Sweden
(e-mail: hosseine@kth.se).
TABLE I. PERFORMANCE COMPARISON BETWEEN THE ST-JTE AND
RECENT
SIC POWER BJTS.
Ref. / Size
*
BV
(kV)
R
ON
(mΩ.cm
2
)
Gain
Termination
Efficiency (%)
[3] / S 3.2 28 28 40
[4] / S 6 28 3 --
[5] / S 23.5 321 63 78
[6] / S 9.2 49 7 --
[7] / L 1.8 4.4 40 --
[8] / L 2.3 4.5 35 85
[9] / L 1.8 2.8 117 --
[10] / S
2.8
4 55
75
[10] / L 6.8 52
[11] / L 10.5 110 75 91
[12] / L 2.7 4 132 > 90
This work / S 5.85 28 40 93
*
"S" FOR SMALL-AREA (< 0.5 MM
2
) LOW-CURRENT (I < 1 A) AND "L"
FOR LARGE
-AREA (> 2.5 MM
2
) HIGH-CURRENT (I > 5 A) BJTS.
Fig. 1. (a) Schematic cross-sectional view of the 4H-SiC with
trenched JTE (ST-JTE), (b) optical microscopic image of the ST-JTE
BJT with seven trenches, and (c) magnified optical microscopic
image of the trench structures. The trenches are formed on the first
JTE with the depth of 300 nm.
5.8-kV Implantation-Free 4H-SiC BJT with
Multiple-Shallow-Trench
Junction Termination Extension
Hossein Elahipanah, Member, IEEE, Arash Salemi, Student Member, IEEE, Carl-Mikael Zetterling,
Senior Member, IEEE, and Mikael Östling, Fellow, IEEE
S
simultaneously formed by the first JTE with no extra mask,
which makes it a simpler and cheaper structure to fabricate;
meanwhile prevents any further misalignment issue. Third, the
maximum BV of 5.85 kV is obtained which is 93% of the
theoretical value [13]. Higher termination efficiency is yet
achievable by increasing the number of trenches and
squeezing the trenches which requires lithography with higher
resolution.
II. D
EVICE STRUCTURE AND FABRICATION
The cross-sectional schematic view of the ST-JTE is shown
in Fig. 1(a). The total area of the device is 0.49 mm
2
. The
width of the emitter and base fingers are 10 and 4 µm,
respectively (see Fig. 1(b)). Trench structures gradually
decreasing in dimensions have been formed on the first JTE
(see Fig. 1(c)). Depending on the design, the depth of the
trenches can be controlled by the etching of the other layers
with no additional mask. Considering the lithography limit to
1 µm, the number, width, and spacing of the trenches are also
optimized to reach the optimized electric field spreading. The
optimized result is obtained for the ST-JTE with seven
trenches as deep as 300 nm; whereas the spacing gradually
increases from 1 µm to 7 µm and the width is gradually
decreasing from 9 µm to 3 µm.
A 100-mm 4H-SiC n
+
-substrate with five different epilayers
on top is used (see Fig. 1(a)). A 2.5-µm layer of SiO
2
is
deposited and patterned as the hard mask. Then, inductively
coupled plasma (ICP) deep etching is used to form the emitter
and mesa. Afterward, reactive ion etching (RIE) is used to
accurately form the JTEs with the depth of D
1
= 300 nm and
D
2
= 160 nm, respectively. Trench structures are
simultaneously formed by the etching step of the first JTE
with the same depth. As a result, the etching uniformity of
±5% is achieved, which significantly improved the yield of
the wafer. Also, extra etchings are performed to investigate
the dose dependency of the breakdown voltage. Subsequently,
sacrificial oxidation was done in O
2
for 1 h. After removal of
the thermally grown oxide, 50 nm SiO
2
was deposited by
PECVD and annealed at 1100 °C in N
2
O environment for 3 h
as the passivation layer. A Ni layer with a thickness of 140-
nm was deposited and patterned on the emitter and back-side
collector. Contacts were annealed in Ar at 950 °C for 1 min. A
110-nm Ni/Ti/Al metal stack with the thickness 0.1/0.15/0.85
was deposited and patterned on the base layer and annealed in
N
2
at 815 °C for 2 min to form the p-contacts [14]. To
improve the current spreading through the contacts, 500 nm of
Al was deposited and formed on the n- and p-contacts. Using
TLM structures, a contact resistivity of 1.4×10
-5
Ω.cm
2
and
1.5×10
-4
Ω.cm
2
are measured for the emitter and base contact,
respectively. Subsequently, a thick oxide was deposited as the
insulator layer and the contact windows were opened. Finally,
2.5-µm Al was sputtered and patterned to form the emitter and
base electrodes.
III. R
ESULTS AND DISCUSSION
The I
C
–V
C
characteristic of the ST-JTE device in forward
and blocking mode is shown in Fig. 2. A maximum current
gain of 40 at I
C
= 0.24 A (J
C
= 370 A/cm
2
) and V
CE
= 4.3 V at
I
C
= 0.1 A are achieved. The current flow in the thick collector
layer is analyzed by 2-D simulation and the estimated
equivalent active area of the device is 0.065 mm
2
. Therefore,
the specific on-resistance of R
ON
= 28 mΩ.cm
2
was calculated
considering the current spreading effect. It should be noted
that the on-resistance of the device is expected to be higher
when scaled up to large-area since the small-area BJT benefits
from significant current spreading in the collector layer. The
forward voltage drop at J
C
= 100 A/cm
2
is 2.6 V. The
maximum open-base breakdown voltage of 5850 V and 4650
V are measured for the ST-JTE and double-JTE, respectively.
These correspond to 93% and 75% of the theoretical value,
respectively, which prove the efficiency of the ST-JTE
structure. Therefore, less termination area is required for the
ST-JTE device leading to reduction of total device cost.
The remaining dose of the dopants in the JTEs is controlled
by the etching process. Fig. 3 illustrates the BV versus the p-
layer dose for ST-JTE compare with conventional double-
JTE. It is apparent that the ST-JTE BJTs have much less
sensitivity to the etching variation. Also, simulation results
show that higher number of trenches results in wider stable
dose range which will result in higher yield of the wafer. The
maximum breakdown voltage variation (ΔBV) of 700 V
and1700 V is measured for the ST-JTE and conventional
double-JTE for
Fig. 2. I-V forward characteristics of the ST-JTE device and
comparison of the breakdown voltages of the ST-JTE and
conventional double-JTE device.
Fig. 3. Experimental and simulated breakdown voltages of the ST-
JTE and double-JTE devices as a function of the remaining dose of
the JTE1. The dose range for the ST-JTE structures are much wider
than that of the double-JTE.
Fig. 4. Simulated electric field distribution along the device at the
cutline of AA′ for the applied reverse voltage of 6 kV.
20% variation in the remaining dose of the p- layer,
respectively. Among the 26 measured ST-JTE devices, 80%
blocks more than 4.5 kV and the yield of high blocking ST-
JTE BJTs (> 5 kV) is about 55%.
The influence of charges at the oxide/SiC interface on the
breakdown voltage of SiC devices has been reported [15]. The
amount of charges in a certain area varies by forming the
trenches. Therefore, taking the presence of these charges into
account, the optimum effective dose of 1.15×10
13
cm
-3
is
found for the first JTE by simulation. Fig. 4 compares the
simulated electric field for the ST-JTEs with different number
of trenches with conventional double-JTE structure at their
optimum dose for the applied reverse voltage of 6 kV. It is
apparent that the electric field is shared by the multiple
trenches leading to alleviated peaks in the electric field.
Therefore, these trenches prevent the field crowding and
shield the high field at the JTE and mesa corners. Thus the
electric field in the ST-JTE is more uniform which results in
higher efficiency of the termination technique. Also, it is
apparent that higher number of trenches with smaller widths
results in better electric field distribution. Considering the
lithography limit to 1 µm, the optimized result is obtained for
the ST-JTE with seven trenches. The smallest dimension
required for realizing the ST-JTE with 14 trenches is 0.5 µm
compare with 1 µm for the device with seven trenches. One
can expect that a finer photolithography improves the
termination efficiency and sensitivity to the etching and
doping variation (remaining dose) of the p-layer. However, it
is recommended not to design too narrow trenches with dense
spacings.
The performance of the ST-JTE BJT compared with that of
the best reported BJTs are summarized in Table I. The
maximum open-base BV of 5850 V and R
ON
of 28 mΩ.cm
2
are obtained, respectively, which is very close to the 4H-SiC
unipolar limit. The termination efficiency of the ST-JTE
device is among the highest values reported to date.
IV. C
ONCLUSION
An implantation-free 4H-SiC BJT with multiple-shallow-
trench JTE structure (ST-JTE) is presented. A specific on-
resistance (R
ON
) of 28 mΩ.cm
2
is obtained for the device with
active area of 0.065 mm
2
. A maximum open-base breakdown
voltage (BV) of 5.85 kV is measured which corresponds to
93% of the ideal value. No extra mask is required to form
these trenches which result in a simpler process without any
further misalignment issue.
REFERENCES
[1] M. Östling, “Silicon carbide based power devices,” in IEDM, San
Francisco, CA, 2010, pp. 13.3.1–13.3.4.
[2] A. Agarwal, M. Das, B. Hull, S. Krishnaswami, J. Palmour, J.
Richmonds, S. H. Ryu, and J. Zhang, “Progress in silicon carbide power
devices,” in Proc. 64th Device Res. Conf., Jun. 2006, pp. 155–158.
[3] C.-F. Huang and J. R. Cooper Jr, “4H-SiC npn bipolar junction
transistors with BV
CEO
> 3,200 V,” in Proc. 14th ISPSD, Jun. 2002, pp.
57–60.
[4] S. Balachandran, C. Li, P. A. Losee, I. B. Bhat, and T. P. Chow, “6 kV
4H-SiC BJTs with specific on-resistance below the unipolar limit using
selectively grown base contact process,” in Proc. 19th ISPSD, Jun. 2007,
pp. 293–296.
[5] H. Miyake, T. Okuda, H. Niwa, T. Kimoto, and J. Suda, “21-kV SiC
BJTs with space-modulated junction termination extension,” IEEE
Electron Device Lett., vol. 33, no. 11, pp. 1598–1600, Nov. 2012.
[6] J. Zhang, H. Zhao, P. Alexandrov, and T. Burke, “Demonstration of first
9.2 kV 4H-SiC bipolar junction transistor,” Electron. Lett., vol. 40, no.
21, pp. 1381–1382, Oct. 2004.
[7] Q. Zhang, A. Burk, F. Husna, R. Callanan, A. Agarwal, J. Palmour, R.
Stahlbush, and C. Scozzie, “4H-SiC bipolar junction transistors: From
research to development - A case study: 1200 V, 20 A, stable SiC BJTs
with high blocking yield,” in Proc. 21st ISPSD, Jun. 2009, pp. 339–342.
[8] M. Domeij, C. Zaring, A. O. Konstantinov, M. Nawaz, J.-O. Svedberg,
K. Gumaelius, I. Keri, A. Lindgren, B. Hammarlund, M. Östling, M.
Reimark, “2.2 kV SiC BJTs with low V
CESAT
fast switching and short-
circuit capability,” Mater. Sci. Forum, vol. 645–648, pp. 1033–1036,
Apr. 2010.
[9] M. Domeij, A. Konstantinov, A. Lindgren, C. Zaring, K. Gumaelius, M.
Reimark, “Large area 1200 V SiC BJTs with β>100 and ρON<3
mΩcm
2
,” Mater. Sci. Forum, vol. 717–720, pp. 1123–1126, May 2012.
[10] R. Ghandi, B. Buono, M. Domeij, C.-M. Zetterling, and M. Östling,
“High-voltage (2.8 kV) implantation-free 4H-SiC BJTs with long-term
stability of the current gain,” IEEE Trans. Electron Devices, vol. 58, no.
8, pp. 2665–2669, Aug. 2011.
[11] S. Sundaresan, S. Jeliazkov, B. Grummel, and R. Singh, “10 kV SiC
BJTs—Static, switching and reliability characteristics,” in Proc. 25th
ISPSD, Jun. 2013, pp. 303–306.
[12] S. G. Sundaresan, S. Jeliazkov, B. Grummel, R. Singh, “Rapidly
maturing SiC junction transistors featuring current gain (β) > 130,
blocking voltages up to 2700 V and stable long-term operation,” Mater.
Sci. Forum, vol. 778–780, pp. 1001–1004, Feb. 2014.
[13] A. Konstantinov, Q. Wahab, N. Nordell, and U. Lindefelt, “Ionization
rates and critical fields in 4H silicon carbide,” Appl. Phys. Lett., vol. 71,
no. 1, pp. 90–92, Jul. 1997.
[14] C.-M. Zetterling, Process Technology for Silicon Carbide Devices,
London, U.K.: IET, Mar 2002.
[15] R. Ghandi, B. Buono, M. Domeij, R. Esteve, A. Schöner, J. Han, S.
Dimitrijev, S. A. Reshanov, C.-M. Zetterling, and M. Östling, “Surface
passivation effects on the performance of 4H-SiC BJTs,” IEEE Trans.
Electron Devices, vol. 58, no. 1, pp. 259–265, Jan. 2011.