Download Radiation Effects of SiGe and III-V Based Material

Radiation Effects of SiGe and III-V Based Material HBTs
Technology: Comparison and trends
C P Jain1, K.B Joshi2
1
Department of Electronics, Banasthali University, Banasthali, Rajasthan, India-304022
2
Department of Physics, M.L.Sukhadia University, Udaipur, India-313001
1
[email protected]
Abstract
Silicon Germanium (SiGe) & III-V based material Hetero-junction Bipolar Transistor (HBT)
technology has been developed as a significant technology for both wired and wireless
application for its excellent performance in analog and RF application and compatibility with
CMOS process.
The study of radiation effect of SiGe & III-V HBT technology is very important to spacecraft
designers to introduce the latest technology into the system. In this paper, the Radiation hardens
characteristics of SiGe HBTs as well as III-V HBTs are measured in rad/s. The characteristics
are compared in with and without radiation for different doses of alpha particle and the
influences on above device are investigated and correlation between base and collector current in
SiGe and III-V HBTs are established to reduce radiation impact.
Keywords— Radiation, SiGe HBT, GaAs HBT, InP HBT, Rad Hard Bipolar devices
1. Introduction
In this paper, we mainly focus on to make device and circuit resistive or harden against
radiation environment and for bipolar process; for years, Si BJT has been widely used all over in
the semiconductor and integrated circuit technologies [1],[2],[3]. A massive research that has
been done into silicon technology over the many years has resulted in incredibly cheap, reliable,
and simple manufacturing and design processes [3]. The mysterious world of the III-V transistor
and other band gap-engineered technologies occupies only a small slice of the overall
semiconductor pie due to the expensive cost and low-yield typical of these advanced platforms
[2]. These technologies are used only for those applications where we require the highest speed,
lowest-noise, and highest gain [2].
2
Presently at Department of Physics & Astronomical Science, Central University of Himachal Pradesh,
Post Box 21, Dist- Kangra, HP- 176215
1
Clearly, a space qualified IC technology must show sufficient radiation tolerance to support high
speed circuit applications as well possess total dose tolerance. SiGe & III-V HBT technology has
generated considerable interest in the space community due to its robustness to total ionizing
dose radiation (TID) without any additional hardening[4]. But, recently it has found that high
speed SiGe HBT digital logic circuits were vulnerable to SEU [5]. Hence it is important to study
the radiation effect on HBTs. To understand these effects in HBTs [6], we use calibrated two
dimensional (2-D) device simulations to assess the characteristics of SiGe HBTs.
2. Simulation Details
Radiation dose is defined as the amount of energy deposited into unit mass of the material of
interest. The units of radiation dose are 1 Gy (Gray) = 1 Joule/Kg = 100 rads. Since this energy
loss is material dependent, the type of the material is appended onto the unit. A dose is always
referenced to a mass of material. The term dose rate is used to indicate the dose per unit time
(e.g. Gy/min or Rad/min) that a device will experience at a position from a radiation source.
In Technology Computer Aided Device (TCAD) tool, Devices simulation is divided into two
different models that are calculated simultaneously at each DC bias point or transient time step.
l. Optical ray trace using real component of refractive index to calculate the optical intensity at
each grid point.
2. Absorption or photo-generation model using the imaginary component of refractive index to
calculate a new carrier concentration at each grid point.
Here we define the photo-generation rate in a C-INTERPRETER function written into a text file
that can be supplied to the program. The file returns a time and position dependent photogeneration rate to the program. This returned value is multiplied at every node point defined
during mesh initialization.
The X. ORIGIN, Y. ORIGIN and X. END, Y. END parameters set the coordinates of the starting
and the last point of the line segment. The default values correspond to the top left and bottom
left corners of the device considered.
The Standard beam input syntax allows specification of plane waves with Gaussian or flat-top
(top-hat) irradiance profiles.
In this simulation, we have assumed that the radiation on the device has created a uniform Photogeneration rate of different dose rate vary from 1E10 to 25E30 on devices.
2
3.
Device Characterization
3.1 Bipolar Technology
In this simulation, the Ic verse Vce characteristic is plotted for different values of base
current of the silicon BJT using Gummel model.
To construct the NPN transistor structure, in the initial stage, we specify the mesh, region and
electrode statement in TCAD ATLUS. In doping state statement, we added analytical doping
profile statement. The structure has a heavy n+ emitter 1.0e18 peak base concentration, a buried
collector layer and heavy p+ extrinsic base contact.
In contact statement, we define the emitter contact as an N-type polysilicon emitter. Surface
recombination is also specified at the polysilicon/silicon interface. Next, the material statement is
used to specify the electron and hole SRH lifetime. The model bipolar statement is used to
specify that the default of bipolar model to be set. The default set of bipolar model includes:
concentration dependent mobility, field dependent mobility, band gap narrowing, concentration
dependent lifetime and auger recombination.
Figure 1 Cross-section of Bipolar Junction Transistor
From figure 1 it is seen that, as we increase the radiation dose, the drain current increases. These
trapped charges produce new interface states at the frontier SiO2-Si interface that also decreases
minority carrier lifetime and increases the junction leakage current. The impact of oxide trapped
charges on bipolar devices is much smaller than for MOSFETs because the oxide is not an active
part and the surface doping is much greater for bipolar transistors than for MOSFETs. Both bulk
and interface defects decrease the overall gain and increases the leakage current. The major
effects of radiation on bipolar transistors are a gain loss and an increase of leakage current. The
degradations are higher for power bipolar transistors, especially at low currents. The damages
3
that produced from a high dose rate exposure of a bipolar transistor are smaller to the ones that
can be found with a low dose rate [3].
In this simulation, the syntax for the Ic/Vce simulation is described in considering above NPN
structure. The key point is the use of current boundaries conditions on the base contact as defined
by contact name = base current. This allows constant base current to be forced while Vce is
ramped. The family of Ic/Vce curve can be overlaid in plot.
Figure 2 Overlay results of Id/Vd for dose 7e12
Figure 3 Results of Id/Vd for dose 7e12
Figure 4 Result for dose2E20
Figure 5 Result for dose2.7E25
3.2 The SiGe HBT Technology
The study of radiation effects on SiGe technology can be divided into two main topics. Total
ionizing dose (TID) and single event effects (SEE). TID effects are those caused by long-term
exposure to radiation that slowly ionizes the oxides within a device and usually appears as
increased base leakage current [4].
SiGe HBTs have posed a desirable side benefit of possessing an inherent hardness to ionizing
radiation, and have been shown to be TID tolerant to multi-rad dose levels without any additional
hardening. This unique feature of SiGe HBTs is because of the presence of Ge in the base region
4
as the major source of degradation in Si transistors is the ionization damage in the emitter-base
(EB) spacer oxide. The radiation hardness of SiGe HBTs, is a result of their natural structural
properties that is the heavily doped extrinsic base region, very thin EB spacer, and very small
active volume are the main technological features of SiGe HBTs that promisingly help the device
to be became TID tolerant. Yet, minor radiation induced degradation can still be observed in the
performance of SiGe HBTs. In forward-mode operation, ionization damage in SiGe HBTs will
cause a insignificant increase in the base current at low injection that will consequently degrade
the current gain. Although SiGe HBTs are TID tolerant in regard to single-event effects and
these devices lack immunity and require mitigation techniques that can be accomplished at the
circuit level or at the device level or at both levels [5], [6].
Figure 7 to figure 9 shows the Gummel plot characteristics of a SiGe HBT (which has not been
radiation hardened in any way) both before and after exposure.
The majority of this is similar to Silicon bipolar Gummel plot in BJT section, this description
will focus on the specific SiGe syntax.
In region statement, we define the SiGe region in the base of an NPN bipolar transistor. The
composition fraction of the area inside of x.min, x.max, y.min, and y.max is defined by the
x.composition parameter. Outside this box the Ge composition fraction rolls of linearity to zero
over the distance specified by the grad.* parameters.
The material parameters for the SiGe are defined by material material = SiGe… statement. Here
the life time is set differently for the SiGe and Si region. In this model, we specify common
models for Si and SiGe regions, but this is not essential. The material parameter of the model
statement can be used to set separate model for each material.
The electrical part of this file is same as the silicon file above.
Figure 6 Cross section of SiGe HBT Transistor
5
Figure 7 Result for dose 1E18
Figure 8 Result for dose 2E20
Figure 9 Result for dose 10E20
3.3 GaAs /AlGaAs HBT Technology
In this simulation we take the simulation of an AlGaAs/GaAs HBT to extract Ic and Ib verse
Vbe and plot the bipolar gain. The majority is similar to the Silicon bipolar Gunmmel plot. Here
we focus on specific AlGa and AlGaAs syntax.
In the region statement we used the graded AlGaAs composition fraction. The parameter x.comp
defines the Al content inside the box defined by x.min, x.max, y.min, y.max statement. Outside
the box the Al content fails linearly to zero in a distance grad.*. Since devices constructed in
ATLAS syntax must be in rectangular shape, a silicon dioxide region is used to fill the space to
the left of the emitter.
The low field motilities for the AlGaAs region is set in the material statement. Keep in mind that
the concentration dependent mobility model ‘conmob’ can be applied to GaAS so two model
statement are used to defined the model used in all region and then just the GaAs region.
6
The model statements are used to specify the following set of models: field dependent mobility,
SRH and optical recombination. Optical recombination (also known as band to band
recombination) is important effect in III-V devices. The parameter ‘optr’ sets this model on.
The electrical simulation is same as in silicon BJT.
Figure 10 Cross-section of GaAs HBT Transistor
Figure 11 Characteristic without radiation
From figure 11 to figure 15, it is seen that to illustrate GaAs field effect transistor (FET) does not
have any oxide that can trap charges. Little threshold voltage shift resulting from charge trapping
is experienced. Radiation creates defect centers in the emitter and base materials. These two
effects increase the base current and lead to a small decrease of current gain [7].
7
Figure 12 Result for dose 1E18
Figure 13 Result for dose 2E20
Figure 14Rresult for dose 2E15
Figure 15 Result for dose 2E20
3.4 InP/InGaAs HBT technology
An HBT structure based on the InGaAs-InP material created. The structure is then passed to
ATLUS for electrical stimulation. The input file consists of the following two main stages.
The HBT geometry material region, doping profile, and electrodes are defined in the first stage
of the input construct in DEVEDIT. The p-n-n device is based on a lattice matched InGaAs-InP
material system. It consists of highly doped InGaAs cap region followed by another cap region
made of InP. The next region, also made of InP, constitute the actual emitter. As usual in HBTs
the emitter (InP) has wider energy band gap than the base and collector (InGaAs). The base is
followed by the n-subcollector and n+ collector regions. The substrate is made of undoped InP.
In this structure each region was uniformly doped. The Ga composition fraction was also
specified. Finally the mesh was generated automatically by specified basic mesh constraints and
refining it along x- and /or y- direction in the important areas of the device [8].
8
Figure 16 Cross section of InP HBT Transistor
The first active statement in the ATLAS portion of the input file is material parameter and
models definition. Here the energy band gap, densities of states, and other fundaments material
parameter for InGaA are defined based on the composition specified. the respective default value
are used for InP. The band alignment is defined using the align parameter on the material
statement. Material and model parameters can be defined in ATLAS on material by material or
region by region basis. The latter possibilities are used here to define carrier lifetime, low field
motilities, and saturated velocities taking into consideration the doping levels in the respective
regions.
The same set of physical models is applied here to all the region/materials. Schockley-Read- Hall
recombination, electric field dependent motilities with GaAs- like velocity-field characteristic,
and Fermi- Dirac statistics. To reflect the different properties of the material with regard to
critical field, ecrtn defines in separate model statement for InP and InGaAs material.
The simulation is first performed to obtain the Gummel plot by biasing simultaneously the base
and collector with respect to the emitter up to 12 V.
Figure 17 Characteristic without radiation
9
The simulated ID - VG and IB-VG characteristics of an InP HBT transistor with different levels of
hot carrier damage .As expected the threshold voltage increases as a result of oxide interface
charge. As a result changes in drain current and base current by using different radiation doses
applies.
Figure 18 Result for dose 1E15
Figure 19 Result for dose 10E20
Figure 20 Results of InP HBT Id/Vd for dose 2*10E6
Figure 21 Results of InP HBT Id/Vd for dose 1*1E14
4. Conclusion
In this paper we have given an overview of radiation effects in Silicon-Germanium hetero
junction bipolar transistors (SiGe HBT), BJT and GaAs/AlGaAs HBT & InP/InGaAS HBTs. We
start by reviewing HBTs and study the impact of ionizing radiation with and without radiation.
This effects are pronounced in SiGe HBT circuits ensure adequate tolerance for many orbital
missions. SiGe HBT, GaAs HBT & InP HBT technology thus offers many interesting
possibilities for electronic systems.
10
Acknowledgement
The author wishes to acknowledge Prof. A. Shastri for guiding and providing research
facilities at Banasthali University. The author (CPJ) is grateful to Department of Science and
Technology, New Delhi for supported this work through SERC grant.
References
[1]
J. P. Colinge, “Hardening Integrated Circuits against Radiation Effects,” IEEE NSREC
1997 Short Course Tutorial.
[2]
P.Buchman, “Total dose hardness assurance for microcircuits for space environment,”
IEEE Trans. Nuc.Sci., vol.NS-33, p.1352, Dec 1986.
[3]
A.H. Johnston, B.G. Rax and C.I. Lee, “Enhanced damage in linear bipolar integrated
circuits at low dose rate,” IEEE Transactions on Nuclear Science, vol. 42, no. 6, pp.
1650-1659, 1995.
[4]
L. Larson, M. Case, S. Rosenbaum, D. Rensch, P. MacDonald, M. Matloubian, M. Chen,
D. Harame, J. Malinowski, B. Meyerson, M. Gilbert, and S. Maas, “Si/SiGe HBT
technology for low-cost monolithic microwave integrated circuits,” in Tech. Dig. IEEE
Int. Solid-State Circuits Conf., 1996, pp. 80–81.
[5]
P.W. Marshall, M.A. Carts, A. Campbell, D. McMorrow, S. Buchner, R. Stewart, B.
Randall, B. Gilbert, and R.A. Reed, "Single event effects in circuit-hardened SiGe HBT
logic at gigabit per second data rates," IEEE Trans. Nucl. Sci., vol. 47, pp. 2669-2674,
2009.
[6]
J. D. Cressler, D. L. Harame, J. H. Comfort, J. M. C. Stork, B. S. Meyerson, and T. E.
Tice, “Silicon-germanium heterojunction bipolar technology: The next leap in silicon?,”
Tech. Dig. IEEE Int. Solid-State Circuits Conf., 1994, pp. 24–27.
[7]
M.E. Kim et al., “GaAs Heterojunction Bipolar Transistor Device and IC Technology for
High performance Analog and Microwave Application,”IEEE Trans. Microwave Theory
Tech., vol.37, pp. 1286-1303, 1989.
[8]
S. Blayac, M. Riet, J. L. Benchimol, F. Alexandre, P. Berdaguer, M. Kahn, A. Pinquier,
E. Dutisseuil, J. Moulu, A. E. Kasbari, A. Konczykowska, and J. Godin, “MSI
InP/InGaAs DHBT technology: Beyond 40 Gbit/s circuits,” in Proc. 14th Indium
11
Phosphide Related Materials Conf., vol. B1-4, Stockholm, Sweden, May 2002, pp. 51–
54.
12