[BNC-all] Defense exam: Sambit Palit, May 30th, 10:00AM, BRK 1001

Sambit Palit spalit at purdue.edu
Thu May 29 11:39:42 EDT 2014


Dear all,

I invite you for some free food, snacks and coffee on the occasion of my 
final defense examination.
The details are as given below.

Regards,
Sambit


Date: Friday, May 30th
Time: 10:00 AM
Location: Birck 1001

Title: Charging and Breakdown in Amorphous Dielectrics: Phenomenological 
Modeling Approach and Applications

Abstract:

Amorphous dielectrics of different thicknesses (nm to mm) are used in 
various applications. Low temperature processing/deposition of amorphous 
thin-film dielectrics often result in defect-states or electronic traps. 
These traps are responsible for increased leakage currents and bulk 
charge trapping in many associated applications. Additional defects may 
be  generated during regular usage, leading to electrical breakdown. 
Increased leakage currents, charge trapping and defect 
generation/breakdown are important and pervasive reliability concerns in 
amorphous dielectrics.

We first explore the issue of charge accumulation and leakage in 
amorphous dielectrics. Historically, charge transport in amorphous 
dielectrics has been presumed, depending on  the dielectric thickness, 
to be either bulk dominated (Frenkel-Poole (FP) emission) or contact 
dominated (Fowler-Nordheim tunneling). We develop a comprehensive 
dielectric charging modeling framework which solves for the transient 
and steady state charge accumulation and leakage currents in an 
amorphous dielectric, and show that for intermediate thickness 
dielectrics, the conventional assumption of FP dominated current 
transport is incorrect, and may lead to false extraction of dielectric 
parameters. We propose an improved dielectric characterization 
methodology based on an analytical approximation of our model. Coupled 
with ab-initio computed defect levels, the dielectric charging model 
explains measured leakage currents more accurately with lesser empiricism.

We study RF-MEMS capacitive switches as one of the target applications 
of intermediate thickness amorphous dielectrics. To achieve faster 
analysis and design of RF-MEMS switches in particular, and 
electro-mechanical actuators in general, we propose a set of fundamental 
scaling relationships which are independent of specific physical 
dimensions and material properties; the scaling relationships provide an 
intrinsic classification of all electro-mechanical actuators. However, 
RF-MEMS capacitive switches are plagued by the reliability issue of 
temporal shifts of actuation voltages due to dielectric charge 
accumulation, often resulting in failure due to membrane stiction. Using 
the dielectric charging model, we show that in spite of unpredictable 
roughness of deposited dielectrics, there are predictable shifts in 
actuation voltages due to dielectric charging in RF-MEMS switches.  We 
also propose a novel non-obtrusive, non-contact, fully electronic 
resonance based technique to characterize charging driven actuation 
shifts in RF-MEMS switches which  overcomes limitations in 
conventionally used methods.

Finally, we look into the issue of defect generation and breakdown in 
thick polymer dielectrics. Polymer materials often face premature 
electrical breakdown due to high electric fields and frequencies, and 
exposure to ambient humidity conditions. Using a field-driven correlated 
defect generation model, coupled with a model for temperature rise due 
to dielectric heating at AC stresses, we explain measured trends in 
time-to-breakdown and breakdown electric fields in polymer materials. 
Using dielectric heating we are able to explain the observed lifetime 
and dielectric strength reduction with increasing dielectric 
thicknesses. Performing lifetime measurements after exposure to 
controlled humidity conditions, we find that moisture ingress into a 
polymer material reduces activation barriers for chain breakage and 
increases dielectric heating.

Overall, this thesis develops a comprehensive framework of dielectric 
charging, leakage and degradation of insulators of different thicknesses 
that have broad applications in multiple technologies.


More information about the BNC-all mailing list