MIT Department: Electrical Engineering and Computer Science
Undergraduate Institution: University of Central Florida
Faculty Mentor: Jesús A. del Alamo
Research Supervisor: Ethan Lee
I was born in Cienfuegos, a city on the southern coast of Cuba. I am currently attending the University of Central Florida in pursuit of a Bachelor of Science in Electrical Engineering. I am highly interested in the synthesis of nanoscale materials for next generation electronics and energy storage applications. My goals are to one day graduate with a PhD in materials science and to continue in academia by becoming a faculty member. I want to promote the advancement of diversity by bridging the gap for underrepresented students who lack the resources to reach their academic, professional, and personal goals.
2018 Research Abstract
Frequency Dependence of AC On-State TDDB for GaN MIS-HEMTs
Luis D. Hurtado1, Ethan S. Lee2, Jesús A. del Alamo3
1Department of Electrical Engineering, University of Central Florida
2,3Department of Electrical Engineering, Massachusetts Institute of Technology
The Gallium Nitride (GaN) Metal Insulator Semiconductor High Electron Mobility Transistor (MIS-HEMT) is a promising candidate to replace Silicon (Si) based transistors thanks to its high electron mobility, size efficiency, and high frequency operation capability. However, reliability and instability challenges have held back its development and commercial deployment. One key issue, called time-dependent dielectric breakdown (TDDB), stems from the use of a dielectric in the gate stack of the device. During device use, the dielectric experiences an intense electric field. The dielectric accumulates defects as result of the electric field, and the accumulation ultimately leads to a low resistance ohmic contact that shorts the dielectric deeming the device inoperable. Here, we show that the time to breakdown for GaN transistors has a frequency dependency. Our accelerated stress tests demonstrate that AC unipolar at a frequency of 100kHz, 10kHz, and 1kHz increased the time to breakdown by a factor of 2.13, 1.45, and 1.1 respectively. Similarly, AC bipolar stress at 100kHz, 10kHz, and 1kHz increased the breakdown time by a factor of 2.28, 1.68, and 1.44 respectively. The increase in breakdown times shows that extrapolated life time models based on DC measurements may be too conservative. Furthermore, the frequency dependence helps shed more light towards understanding TDDB.