Modeling and optimal design of silicon superjunctions considering charge imbalance

• Published in: Semiconductor science and technology Vol. 40; no. 5; pp. 55014 - 55022
• Main Authors: Bhargav, P N S, Gurugubelli, Vijaya Kumar
• Format: Journal Article
• Language: English
• Published: IOP Publishing 30.05.2025
• Subjects: breakdown voltage; charge imbalance; design formulas; optimum design; specific on-resistance; superjunction; trade-off
• ISSN: 0268-1242; 1361-6641
• DOI: 10.1088/1361-6641/add36a

Charge imbalance ( C . I . ) in superjunction (SJ) power devices adversely affects their breakdown voltage V Br if the specific on-resistance R on , sp is not compromised and vice versa. The literature dealing with C . I . is void of an approach that evaluates the ionization integral by substituting an accurate model for the electric field along the critical path. Our work addresses this gap, thereby providing a robust analytical model that relates the pillar width, length and doping to V Br and C . I . Further, we propose an optimum design methodology that reveals the following insights. The optimum pillar length depends only on V Br but not on pillar width or C . I . Whereas, the optimum doping depends on all the above. Unlike in a balanced SJ, where R on , sp can be lowered by decreasing the pillar width down to the physical limits of fabrication, we demonstrate that in the presence of C . I . , R on , sp increases rapidly below a certain width which is physically realizable, e.g. as high as 2.7 µ m for a 600 V SJ device with 10% C . I . This optimum width is found to increase with V Br and C . I . In addition, we provide closed-form models for the optimum pillar width, length and doping, using which a simple, non-iterative method to design an optimum SJ that minimizes R on , sp for a target V Br and a given C . I . margin is proposed. The developed models are validated by 2D numerical simulations and compared against available experimental data.