You are on page 1of 4

Proceedings of the 2012 24th International Symposium on Power Semiconductor Devices and ICs

3-7 June 2012 - Bruges, Belgium

Study of Electron and Hole Traps in Freewheeling


Diodes for Low Loss and Low Reverse Recovery
Surge Voltage
Satoru Kameyama, Masafumi Hara Tomohiro Kubo, Fumio Hirahara
Electronics Development Division 3 Discrete Semiconductor Division
Toyota Motor Corporation Toshiba Corporation
Toyota, Japan Kawasaki, Japan

Junpei Ebine, Koichi Murakami


Electronics Engineering Division 5
Toyota Technical Development Corporation
Toyota, Japan

Abstract—The purpose of the research described in this paper is with local lifetime control are well-known devices to achieve
to achieve freewheeling diodes (FWDs) with low loss and low low loss [1-3]. Several papers related to local lifetime
reverse recovery surge voltage (Vdsurge). This paper discusses the technologies have indicated that understanding the conditions
relationship between electron and hole traps and the electrical of traps introduced into a device is essential to prevent its
properties of FWDs. Samples with controlled conditions of failure [4-6]. The purpose of the research described in this
electron and hole traps were fabricated by He irradiation and paper is to achieve FWDs with low loss and low reverse
annealing. The trap conditions were evaluated by deep level recovery surge voltage (Vdsurge). This paper discusses the
transient spectroscopy (DLTS) and cathode luminescence (CL). relationship between electron and hole traps and the electrical
The electrical properties of the samples were measured and
properties of FWDs using experiments and simulations. It
simulated to analyze the samples. The analysis clarified that
controlling trap conditions is essential when designing device DC
clarifies that FWDs with low loss and low Vdsurge can be
and AC electrical properties as well as Vdsurge. produced by introducing local lifetime when the trap types,
levels, and concentrations are controllable.
Keywords- Freewheeling diode, lifetime control, electron and
hole traps, reverse recovery surge voltage II. DEVICE STRUCTURE
1.2 kV and 200 A class FWDs introduced with high or low
I. INTRODUCTION concentrations of electron and hole traps were fabricated as
Customer demand for environmentally friendly vehicles shown in Table 1. The process parameters for the sample
such as hybrid electric vehicles (HEVs) and electric vehicles fabrications were selected to control the electron and hole trap
(EVs) is increasing due to their superior energy efficiency. In conditions in the devices. The device structure of the samples is
order to meet this demand, Toyota Motor Corporation described in Fig. 1 [3]. Local lifetime was introduced into the
developed the Toyota Hybrid System (THS). Freewheeling devices by He irradiation and annealing
diodes (FWDs) are one of the key devices in the THS. FWDs
Table 1. Fabricated FWD samples

Intensity at DLTS Intensity at CL


Process parameter
(electron traps) (hole traps)
E1: Ec-0.17 eV E3: Ec-0.43 eV C: Ev +0.35 eV
For electron traps For hole traps
Trap type: OV [7] Trap type: VV [7] Trap type: C iOi [6]
*1 *1
Reference 1.00 1.00

Sample 1 0.44 1.00 0.80 0.75

Sample 2 2.22 1.00 1.15 1.07

*2 *2 *3
Sample 3 1.00 0.22 1.00 1.00 1.00

Sample 4 1.00 1.11 1.09 1.01 1.88


*1: Normalized by process parameters of reference sample
*2: Normalized by DLTS intensities of sample 3
*3: Normalized by CL intensities of sample 3

978-1-4577-1597-6/12/$26.00 ©2012 IEEE 369


Authorized licensed use limited to: Ichiro Omura. Downloaded on January 23,2024 at 08:49:06 UTC from IEEE Xplore. Restrictions apply.
Figure 1. Fabricated FWD structure

Figure 4. Measured CL spectra (samples 3 and 4)


The electron trap conditions were measured by deep level
transient spectroscopy (DLTS) [3], as shown in Figs. 2 and 3.
The hole trap conditions were detected by cathode
luminescence (CL) [4] as described in Fig. 4. The trap levels III. EXPERIMENTAL RESULTS
and intensities are summarized in Table 1. The intensities of The DC and AC electrical properties of the samples in
DLTS and CL spectra represent the concentrations of the traps. Table 1 were measured. Figures 5 and 6 show the measured
E1 and E2 in the DLTS spectra of Figs. 2 and 3 are related to electrical properties of the samples. The trade-off curve
the vacancies in the devices [7]. The C line in the CL spectra of between the forward voltage drop (VF) and reverse recovery
Fig. 4 is generated by the CiOi hole traps in the devices [6]. It is time (trr) of the reference sample, sample 1, and sample 2 are
clear that the electron and hole trap levels and concentrations shown in Fig. 5. The electrical properties of samples 3 and 4
are controlled by the selected process parameters. are on the same trade-off curve as that of samples 1 and 2. On
the other hand, the relationship between the Vdsurge and trr of
samples 1 and 2 in Fig. 6 are different from the trade-off line of
the electrical properties of samples 3 and 4. Figure 6 shows that
a device in which a shorter trr was achieved by controlling the
electron traps (such as sample 1) generates a higher Vdsurge. In
contrast, a device in which a shorter trr was achieved by
controlling the hole traps (such as sample 4) generates a lower
Vdsurge. It is clear that controlling trap conditions is important
when designing device DC and AC electrical properties as well
as Vdsurge. These results show that the targeted trade-off
property between VF, trr, and Vdsurge is achievable when the trap
types, levels, and concentrations are controllable. For example,
point A at the trr of 195 ns in Fig. 6 has a lower Vdsurge than that
of Point B on the same VF – trr trade-off curve.
Figure 2. Measured DLTS spectra of electron traps (samples 1 and 2)

Figure 5. VF – trr measured results


Figure 3. Measured DLTS spectra of electron traps (samples 3 and 4)

370
Authorized licensed use limited to: Ichiro Omura. Downloaded on January 23,2024 at 08:49:06 UTC from IEEE Xplore. Restrictions apply.
Figure 6. trr – Vdsurge measured results Figure 8. Simulated waveforms (models 3 and 4)

IV. SIMULATION RESULTS


To analyze the waveforms in Figs. 7 and 8, the vertical
In order to understand the experimental results, device profiles of the electric field in the devices at the maximum
simulations [8] were carried out. Electron or hole traps were surge voltage in Figs. 7 and 8 were identified as shown in Figs.
introduced into the models as shown in Table 2. The trap levels 9 and 10. It is clear that the introduced traps are related to the
in Table 2 refer to the measured results in Table 1. profiles of the electric fields. The gradient of model 2
The simulated waveforms are described in Figs. 7 and 8. introduced with a higher electron trap level in Fig. 9 is lower
The results are consistent with the experiment results shown in than that of model 1. On the other hand, the gradient of model
Fig. 6. In Fig. 7, the shorter trr of model 2 due to electron trap 3 induced with a higher hole trap level in Fig. 10 is higher than
introduction generates a higher Vdsurge. The shorter trr of model that of model 3.
4 due to the hole trap introduction generates a lower Vdsurge. Equation (1) [7] shows that the gradient of the electric field
Table 2. Simulated FWD models (dE/dx) is related to the effective donor concentration (Neff) due
to the introduced positive or negative trap concentration (NT+).
The relationship of (1) causes the differences in the electric
Energy level [eV] Concentration [a.u.] field profiles shown in Figs. 9 and 10 as well as the differences
Electron traps Hole traps Electron traps Hole traps
in the Vdsurge.
*4
Model 1 1.00
Ec - 0.43 [7] - - dE/dx = q/ε (Neff) = q/ε (ND + NT+). (1)
Model 2 4.00

Model 3 1.00
*4 q: Elementary charge
- Ev + 0.35 [6] -
Model 4 4.00 ε: Dielectric constant of silicon
*4: Normalized by trap concentrations of models 1 and 3

Figure 7. Simulated waveforms (models 1 and 2)


Figure 9. Simulated electric field profiles (models 1 and 2)

371
Authorized licensed use limited to: Ichiro Omura. Downloaded on January 23,2024 at 08:49:06 UTC from IEEE Xplore. Restrictions apply.
Figure 10. Simulated electric field profiles (models 3 and 4)

V. CONCLUSION
Samples with controlled levels of electron and hole traps
were fabricated by He irradiation and annealing. The trap
conditions were identified by DLTS and CL. The electrical
properties of the samples were measured and simulated to
analyze the relationship between the electron and hole traps
and the electrical properties of the samples. The analysis
clarified that controlling the trap conditions of FWDs with
introduced local lifetime is essential when designing DC and
AC electrical properties as well as Vdsurge.

REFERENCES
[1] J. Lutz and U. Scheuermann, “Advanced of the New Controlled Axial
Lifetime Diode”, Proceedings of the 28th PCIM (1994)
[2] K. Nishiwaki, T. Kushida and A. Kawahashi, “A Fast & Soft recovery
Diodes with Ultra Small Qrr (USQ-Diode) Using Local Lifetime Control
by He Ion Irradiation”, Proceedings of ISPSD’01 (2001)
[3] T. Sugiyama, S. Yamazaki, S. Nakagaki and M. Ishiko, “A Study of
Correlation between Traps and Reverse-recovery Characteristics of
FWDs”, Proceedings of ISPSD’04 (2004)
[4] J. Luts, W. Sudkanp and W. Gerlach, “IMPATT Oscillations in Fast
Recovery Diodes due to Temporarily Charged Radiation Induced Deep
Levels”, Solid-St. Electron 42 No.6, 931-938 (1998)
[5] T. Misumi, S. Nakagaki, M. Yamaguchi, K. Sugiyama, F. Hirahara and
K. Nichiwaki, “Analysis of Dynamic Avalanche Phenomenon of PiN
Diode Using He Ion Irradiation”, Proceeedings of ISPSD’06 (2006)
[6] F. Niwa, T. Misumi, S. Yamazaki, T. Sugiyama, T. Kanata and K.
Nishiwaki, “A Study of Correlation between CiOi Defects and Dynamic
Avalanche Phenomenon of PiN Diode Using He Ion Irradiation”,
Proceedings of the PESC (2008)
[7] J. Lutz, H. Schlangenotto, U. Scheuermann and R. D. Doncker,
“Semiconductor Power Devices”, ISBN 978-3-642-11124-2 (2011)
[8] ISE-TCAD Manual ver. 10.0

372
Authorized licensed use limited to: Ichiro Omura. Downloaded on January 23,2024 at 08:49:06 UTC from IEEE Xplore. Restrictions apply.

You might also like