Forward Active and Blocking Performance of 4H-SiC Bipolar Junction Transistors

2005 ◽  
Vol 888 ◽  
Author(s):  
Santhosh Balachandran ◽  
T. Paul Chow ◽  
Anant Agarwal

ABSTRACTWe evaluate the performance capabilities and limitations of high voltage 4H-SiC based Bipolar Junction Transistors (BJTs). Experimental forward characteristics of a 4kV BJT are studied and simulations are employed to determine the factors behind the higher than expected specific on-resistance (Ron,sp) for the device. The impact of material (minority carrier lifetimes), processing (surface recombination velocity) and design (p contact spacing from the emitter mesa) parameters on the forward active performance of this device are discussed and ways to lower Ron,sp, below the unipolar level, and increase the gain (β) are examined. A correlation between the open base blocking behavior (forward blocking) and the current gain (forward active) for 4H-SiC based high-voltage BJTs with lightly doped collector regions is presented and experimental device characteristics are utilized to verify our numerical analysis.

2006 ◽  
Vol 527-529 ◽  
pp. 1429-1432 ◽  
Author(s):  
S. Balachandran ◽  
T. Paul Chow ◽  
Anant K. Agarwal

We evaluate the performance capabilities and limitations of high voltage 4H-SiC based Bipolar Junction Transistors (BJTs). Experimental forward characteristics of a 4kV BJT are studied and simulations are employed to determine the factors behind the higher than expected specific onresistance (Ron,sp) for the device. The impact of material (minority carrier lifetimes), processing (surface recombination velocity) and design (p contact spacing from the emitter mesa) parameters on the forward active performance of this device are discussed and ways to lower Ron,sp, below the unipolar level, and increase the gain (β) are examined.


2009 ◽  
Vol 615-617 ◽  
pp. 837-840 ◽  
Author(s):  
Yuki Negoro ◽  
Akihiko Horiuchi ◽  
Kensuke Iwanaga ◽  
Seiichi Yokoyama ◽  
Hideki Hashimoto ◽  
...  

Surface passivation of 4H-SiC has been investigated for high current-gain bipolar junction transistors (BJTs). For the characterization of surface passivation, we have introduced the product “sp•Ls” of a surface recombination velocity (sp) and a surface diffusion length (Ls). The sp•Ls value was obtained by analyzing the I-V characteristics of pn diodes. Both BJTs and pn diodes were fabricated with several passivation methods. We have found clear correlation between the sp•Ls value and the current gain of the fabricated BJTs. Optimizing the surface passivation, we realized high performance BJTs with a current gain of 107 and a blocking voltage VCEO of 950 V.


2006 ◽  
Vol 527-529 ◽  
pp. 1441-1444
Author(s):  
Pavel A. Ivanov ◽  
Michael E. Levinshtein ◽  
Anant K. Agarwal ◽  
Sumi Krishnaswami ◽  
John W. Palmour

For 1-kV, 30-A 4H-SiC epitaxial emitter npn bipolar junction transistors, the dependence of the common-emitter current gain β on the collector current IC were measured at elevated temperatures. The collector-emitter voltage was fixed (at 100 V voltage) to provide an active operation mode at all collector currents varying in a wide range from 150 mA to 40 A (current densities 24 - 6350 A/cm2). The maximum room temperature current gain was measured to be βmax = 40 (IC = 7 A) while βmax = 32 (IC = 10 A) at 250oC. The β-IC dependences were simulated using a model which takes into account the main processes affecting the current gain. Minority carrier lifetimes and surface recombination velocity were obtained by means of those considerations.


2011 ◽  
Vol 679-680 ◽  
pp. 649-652 ◽  
Author(s):  
Jang Kwon Lim ◽  
Georg Tolstoy ◽  
Dimosthenis Peftitsis ◽  
Jacek Rabkowski ◽  
Mietek Bakowski ◽  
...  

The 1.2 kV SiC JFET and BJT devices have been investigated and compared with respect to total losses including the gate driver losses in a DC-DC converter configuration. The buried grid, Normally-on JFET devices with threshold voltage of -50 V and -10V are compared to BJT devices with ideal semiconductor and passivating insulator interface and an interface with surface recombination velocity of 4.5•104 cm/s yielding agreement to the reported experimental current gain values. The conduction losses of both types of devices are independent of the switching frequency while the switching losses are proportional to the switching frequency. The driver losses are proportional to the switching frequency in the JFET case but to a large extent independent of the switching frequency in the BJT case. The passivation of the emitter junction modeled here by surface recombination velocity has a significant impact on conduction losses and gate driver losses in the investigated BJT devices.


1970 ◽  
Vol 17 (2) ◽  
pp. 119-124 ◽  
Author(s):  
Karolis GULBINAS ◽  
Vytautas GRIVICKAS ◽  
Haniyeh P. MAHABADI ◽  
Muhammad USMAN ◽  
Anders HALLÉN

n- and p-type 4H-SiC epilayers were grown on heavily doped SiC substrates. The thickness of the p-type layer was 7 µm and the doping level around 1017 cm 3, while the n-type epilayers were 15 µm thick and had a doping concentration of 3 - 5*1015 cm 3. Several different surface treatments were then applied on the epilayers for surface passivation: SiO2 growth, Al2O3 deposited by atomic layer deposition, and Ar-ion implantation. Using collinear pump - probe technique the effective carrier lifetimes were measured from various places and statistical lifetime distributions were obtained. For surface recombination evaluation, two models are presented. One states that surface recombination velocity (SRV) is equal on both the passivation/epi layer interface (S2) and the deeper interface between the epilayer and the SiC substrate i. e. (S1 = S2). The other model is simulated assuming that SRV in the epilayer/substrate (S1) interface is constant while in the passivation layer/epilayer (S2) interface SRV can be varied S2 < S1. Empirical nomograms are presented with various parameters sets to evaluate S2 values. We found that on the investigated 4H-SiC surfaces S2 ranges from 3x104 to 5x104 assuming that the bulk lifetime is 4 (µs. In Ar+ implanted surfaces S2 is between (105 - 106) cm/s.http://dx.doi.org/10.5755/j01.ms.17.2.479


Sign in / Sign up

Export Citation Format

Share Document