Reliability issues investigation in Silicon Germanium HBT Yogesh

Reliability issues investigation in Silicon Germanium HBT
Yogesh KumarVerma, Dr. R.K. Chauhan
Department of Electronics and Communication Engineering
Madan Mohan Malaviya Engineering College
Gorakhpur-273010, INDIA
[email protected], [email protected]
Abstract-:
This paper covers and contrasts the
various reliability issues of SiGe HBTs. Self heating
and elevated junction temperature are focused as
emerging major reliability issues which effects both
reliability and performance of the device. As a
foundation for study, an analytical thermal model is
developed. Based on the model different parameters
of 200 GHz SiGe HBT are measured using
SILVACO-ATLAS Device Simulator.
Keywords-: SiGe HBT, reliability, scaling.
I. Introduction
SiGe HBT are vertical transport devices.
Commercially there exists four generations of SiGe
HBT technology. These generations are classified as
first generation, second generation, third generation,
and fourth generation. First generation comprises of
peak unity gain cut-off frequency of 50GHz,
followed by 100-120GHz of second generation SiGe
HBT technology. Third generation has peak unitygain cut off frequency of 200GHz range. Fourth
generation also called as research level has record
peak unity-gain cutoff frequency of 350GHz. SiGe
HBT technology exists in BiCMOS implementation
(SiGeHBT+SiCMOS).This BICMOS implementation
of SiGe HBT provides the optimum advantages of its
potential and makes it an excellent contender in radar
systems, communications, THz sensing, Imaging and
communications, analog applications, electronic
warfare and the most important domain of β€œmixedsignal” ICs.
II. Reliability issues of SiGe HBT
The reliability of a system is defined as the ability of
the system to perform successfully in both expected
and unexpected circumstances. There are different
reliability issues for different devices. Consider a
capacitor, its main reliability issues are operating
temperature, ripple current, inrush current and
operating voltage. In the same fashion for SiGe HBT,
temperature solely is the key concern.
According to J. D. Cressler reliability issues of SiGe
HBTs are low temperature conditions, high
temperature
conditions
and
radiation
rich
environment. He proposed that SiGe HBT can
successfully operate under all these three extreme
environment conditions.
J. D. Cressler also proposed that reliability issues for
mixed circuit applications of SiGe HBTs are thermal
effects; impact ionization induced bias point
instabilities, and operating voltages.
According to J. Kuchenbecker, reliability issues of
SiGe HBTs may be two factors. First is electrical
degradation due to DC life tests and second due to the
influence of hot electron stress on electrical behavior
of devices. He provided the curve between
normalized DC current Gain and stress time (hours)
that clearly indicated that no major problem occurs
during life tests. He also proposed hot carrier stress
experiments which provide a degradation mechanism
located at the surface of device.
According to Zhang Wan-rong, the major reliability
issues for a single MBE SiGe HBT are reduction in
DC current gain and increase in turn-on voltage. AC
current gain degrades slowly. He also proposed that
introduction of Ge into base potentially maintain the
junction reliability.
Greg Freeman proposed that increase in collector
concentration is the most important reliability
concern. As increase in collector concentration
affects the current density and avalanche current. He
proposed that reduction in strip widths of SiGe HBTs
will reduce electro migration and self heating effects.
He also proposed that the ratio of emitter polysilicon
to single-crystal interface needs to be minimized to
reduce base current and emitter resistance shifts. He
proposed that reduction in base resistance is needed
to maintain a high pinch in voltage. At high voltage
designing, avalanche introduces a degradation
mechanism which cannot be ignored.
SiGe HBT has ability to operate in extreme
environment. Extreme environment is the important
profitable corner of the market for electronics.
Extreme environment includes very low operating
temperatures, very high operating temperatures, and
radiation rich environment. Very low temperatures
mean temperatures below 70K or even 4.2K. Very
high temperatures mean temperatures up to 473K or
573K. Radiation rich environment may be space, a
high vibration environment, a high pressure
environment, a low pressure environment, or a
caustic or chemically corrosive environment (inside
the human body). The unique band gap engineered
feature of SiGe HBT offers great potential to make it
operate successfully without need of any
modification. This ultimately makes SiGe HBT
economically strong at both IC and system level. This
unique feature of SiGe HBT to operate successfully
outside the domain of conventional commercial
specifications maintains its reliability.
III. Self Heating
Temperature is responsible for most device failure
mechanisms. Temperature variation in SiGe HBT
may be internally in device or due to extreme
environment conditions. Temperature variation may
be due to rise in junction temperature Ξ”Tj which is an
important reliability concern. This basically occurs
due to self heating. Since increased temperature
causes most of the device degradation problems, so
its suppression is mandatory.
π›₯𝑇𝑗 = π‘…π‘‘β„Ž βˆ— 𝑃𝑑𝑖𝑠𝑠
(1)
π‘…π‘‘β„Ž
is thermal resistance and 𝑃𝑑𝑖𝑠𝑠 is the
dissipated power. A straightforward procedure to
suppress π›₯𝑇𝑗 is the reduction of both π‘…π‘‘β„Ž and
𝑃𝑑𝑖𝑠𝑠 . Rth and Ξ”Tj both vary with emitter strip length
and emitter strip width.
IV. Results and Discussion
(i) Variation in π‘…π‘‘β„Ž with 𝐿𝐸 and π‘ŠπΈ
For constant power density and constant emitter
width, on decreasing emitter strip length 𝐿𝐸 , there is
sharp increase inπ‘…π‘‘β„Ž . Table 1 represents the π‘…π‘‘β„Ž
(K/W)
values
at
different
𝐿𝐸
(µm).
Table-1: π‘…π‘‘β„Ž (K/W) values at different 𝐿𝐸 (µm)
𝐿𝐸 (µm)
π‘…π‘‘β„Ž (K/W)
10
2500
6.44
3500
5
4500
2.44
6500
1
11000
0.67
14500
Emitter length Vs Rth for 200 GHz SiGeHBT
16000
14000
12000
10000
8000
6000
4000
2000
0
0
2
4
6
8
10
12
Figure 1: π‘…π‘‘β„Ž (along Y-axis) variation with 𝐿𝐸 (along X-axis)
From figure 1 it is clear that on decreasing
emitter strip length 𝐿𝐸 , there is sharp increase in
π‘…π‘‘β„Ž . On the other hand for constant power
density and constant emitter length 𝐿𝐸 , on
decreasing emitter strip width π‘ŠπΈ , there is
increase in π‘…π‘‘β„Ž
but variation is not as
noticeable as in case of 𝐿𝐸 scaling, due to
relatively small change in cross sectional area,
shown in figure 2.
Table-2: π‘…π‘‘β„Ž (K/W) values at different π‘ŠπΈ (µm)
π‘ŠπΈ (µm)
π‘…π‘‘β„Ž (K/W)
0.27
7241.96
0.25
7298.36
0.2
7439.36
0.15
7580.36
0.1
7721.36
0.03
7918.77
WE(µm) Vs Rth(K/W)
8000
7900
7800
7700
7600
7500
7400
7300
7200
0
0.05
0.1
0.15
0.2
0.25
0.3
Figure 2: π‘…π‘‘β„Ž (along Y-axis) variation with π‘ŠπΈ
(ii) Variation in π›₯𝑇𝑗 with 𝐿𝐸 and π‘ŠπΈ
there is smaller temperature rise π›₯𝑇𝑗 as shown in
figure 3.
For constant power density and constant emitter
width, on decreasing emitter strip length, 𝐿𝐸 (µm)
Table-3: π›₯𝑇𝑗 values at different 𝐿𝐸 (µm)
𝐿𝐸 (µm)
π›₯𝑇𝑗
10
29.17
6.44
26.67
5
25
2.44
19.17
1
13.33
0.67
8.33
LE(µm) Vs Tj
35
30
25
20
15
10
5
0
0
2
4
6
8
Figure 3: 𝑇𝑗 (along Y-axis) variation with 𝐿𝐸
On the other hand for constant power density
and constant emitter strip length, on decreasing
emitter strip widthπ‘ŠπΈ , 𝑇𝑗 suppression is more
pronounced as shown in figure 4.
Table-4: 𝑇𝑗 values at different π‘ŠπΈ
π‘ŠπΈ (µm)
𝑇𝑗 (K)
0.27
38.33
0.25
36.31
0.2
31.28
0.15
26.24
0.1
21.21
0.03
14.16
10
12
WE(µm) Vs Tj(K)
45
40
35
30
25
20
15
10
5
0
0
0.05
0.1
0.15
0.2
0.25
0.3
Figure 4: 𝑇𝑗 (along Y-axis) variation with π‘ŠπΈ
Lateral scaling imposes positive impacts on
devices in terms of self-heating management.
Vertical scaling is expected to have a much less
impact on π‘…π‘‘β„Ž than lateral scaling. It is because
no significant impact exists on heat dissipation
path due to vertical scaling. But Jc increment for
vertically scaled devices, results in raised
dissipated power which causes increase in 𝑇𝑗 ,
although π‘…π‘‘β„Ž remains unchanged.
VI. Conclusion
In
this
paper the major
reliability
concerns
of SiGe HBTs are observed including
junction temperature. It has been observed that
π‘…π‘‘β„Ž variation with 𝐿𝐸 for constant π‘ŠπΈ and constant
power density is more strongly dominating than
π‘…π‘‘β„Ž
variation with π‘ŠπΈ for constant 𝐿𝐸 and
constant power density. 𝑇𝑗 Suppression is more
pronounced with varying π‘ŠπΈ for constant 𝐿𝐸 and
constant power density as compared with variation
of 𝑇𝑗 with 𝐿𝐸 for constant π‘ŠπΈ and constant power
density.
VI. References
[1]. John D. Cressler β€œOn the potential of SiGe
HBTs for Extreme Environment Electronics”,
Proceedings of the IEEE, VOL93, No.9, pages
pp 1559-1582, SEPTEMBER 2005.
[2]. Peter Ashburn, β€œSiGe Heterojunction
Bipolar Transistors”, John Wiley & Sons
Publication, 2003.
[3]. ATLAS User’s Manual Device Simulation
Software, SILVACO International, 2004.
[4]. Frank Larin, β€œRadiation Effects in
Semiconductor Devices”, John Wiley & Sons
Publication.
[5]. Ankit Kashyap and R.K. Chauhan , β€œProfile
Design Optimization of SiGe Heterojunction
Bipolar
Transistors
for
High
Speed
Applications”, Journal of Computational and
Therotical Nanoscience, Volume 5, 2238-2242,
2008.
[6]. Ankit Kashyap and R.K. Chauhan, β€œEffect
of the Ge Profile design on the performance of
an
n-p-n SiGe HBT-based analog circuit”,
Microelectronics journal, MEJ: 2554, (2008).
[10]. John D. Cressler, β€œSilicon-Germanium as
an Enabling Technology for Extreme
Environment Electronics”, IEEE Transactions
on Device and Materials Reliability, VOL. 10,
NO. 4, pages pp 437-pp 448, December 2010.
[11]. Yogesh Kumar Verma and Dr.
R.K.Chauhan, "Study of Radiation and
Temperature Effects on Devices with Si and
SiGe Base" ISSN: 2249-1945, GJCAT, Vol2
(1),2012,pp917-pp920.
Authors
Er. Yogesh Kumar Verma born on 25 June 1988
received his B.Tech. Degree in Electronics and
Communication Engineering from Sharda Group
of Institutions, Mathura in 2009 and M.Tech.
Degree in Electronics and Communication
Engineering with specialization in Digital
Systems from Madan Mohan Malaviya
Engineering College, Gorakhpur in 2012. His
research interest includes Device modeling in
general and Modeling and Simulation of HBT
based devices in particular. His area of interest
also includes analysis of Mixed Signal Circuits
and Integrated Circuits. He is currently working
as Assistant Professor in Electronics and
Communication Engineering Department in
Noida Institute of Engineering and Technology,
Greater Noida.
Contact: 07503697453
Email id: [email protected]
R. K. Chauhan was born in Dehradoon, India
in 1967. He received the B.Tech. degree in
Electronics & Communication Engineering,
from G.B.P.U.A.T Pantnagar, in 1989 and M.E.
in Control & Instrumentation, from MNNITAllahabad in 1993 and Ph.D in Electronics
Engineering, from IT-BHU, Varanasi, INDIA in
2002. He joined the department of ECE, Madan
Mohan
Malviya
Engineering
College,
Gorakhpur, India as a lecturer, in 1993, as an
Assistant Professor since 2002 and thereafter as
an Associate Professor since Jan, 2006 to till
date in the same institute. He also worked as a
Professor in Department of ECE, Faculty of
Technology, Addis Ababa University, Ethiopia
between 2003 to 2005. He is reviewer of
Microelectronics Journal, CSP etc. His research
interests include device modeling and simulation
of MOS, CMOS and HBT based circuits. He
was selected as one of top 100 Engineers of
2010 by International Biographical Centre,
Cambridge, England.
E-mail: [email protected]
Ph: +91-9235500556