Nanocrystalline
silicon, amorphous silicon, semiconductor nanowires
This has been a long standing project simulating
nanocrystalline and amorphous silicon thin films. The objective of the research
is to identify, explore, evaluate and model new heterogeneous thin film
materials capable of making a breakthrough in the production of low cost
electricity from sunlight. This project has been part of the NREL national thin
film solar cell team effort.
Our approach has been to use molecular dynamics simulations
using both classical and tight binding methods to obtain insight into mechanisms
of metastability and the structure of nanocrystaline silicon.
Metastability
The Staebler-Wronski effect or
the degradation of thin film silicon solar cells when they are placed in
sunlight is a mystery that has plagued
the research community for more than 20 years. When such solar cells are kept in
sunlight their efficiency drops by 15-20% over a period of several days.
Intensive studies have established that these changes are due to creating of
mid-gap defect states from silicon dangling bonds in the amorphous silicon
films. From an annealed defect
density of 1015 cm-3, the defect density rises to 1016-1017
cm-3 after light-soaking. The metastable changes are reversible and
can be removed by annealing the material at 180-200 C. The metastable defects
are indistinguishable from native dangling bond defects. Significantly the
metastable defects are at least 4 Å distant from hydrogen sites-
indicating an anti-correlation with H.
We proposed a new mechanism
explaining many features of the Staebler-Wronski effect [1,2]. In the first step
sunlight creates an excess density of electrons and holes. The
electrons recombine with holes on weak silicon bonds in the material.
The recombination energy causes the weak silicon bonds to break,
generating silicon dangling bond – floating bond pairs.
Accurate tight-binding molecular dynamics simulations have established
the lowering oif energy barriers for bond-breaking when excited e-h pairs are
present. During the second step, the floating bonds (over-coordinated
silicon atoms) diffuse away from the dangling bonds and move freely
throughout the material since they are a mobile species. In the third step the
floating bonds recombine with themselves or with H in the network, and the
network is left with primarily dangling bonds. An interesting caveat is that
because of charge neutrality charged dangling bond defects are formed in
addition to the neutral species. Many features of the Staebler-Wronski are
explained well by this model (Phys.
Rev. Lett. paper, book
chapter). The model has been featured in news releases (Ames
Lab Inquiry article, Hindu).
Weak Si bond (WB) breaking into a dangling
bond (DB) and a floating bond
(FB) afer e-h recombination
Nanocrystalline silicon was simulated
consisting of nanocrystallites embedded in an amorphous matrix. The presence of
the nanocrystallite improves the ordering of the amorphous matrix which may
account for the higher stability of nanocrystalline silicon towards light
soaking. The nanoscale structure can inhibit the instability and breaking of
weak bonds.
There is a thin disordered region around the
nanocrystal containing an excess density of H where the H density is about twice
the background H density. By performing simulations at various temperatures and
monitoring the motion of H, we found that the excess grain boundary H is
responsible for the anomalous low-temperature H evolution peak that occurs near
400C in addition to the high temperature H evolution peak at 600C.
Powerpoint
presentation on nanocrystalline silicon
Present work is investigating the structure
of semiconductor nanowires, including the structure and orientation of
crystalline core nanowires. There are many similarities to microcrystalline
silicon.
Faculty:
Rana Biswas
Collaborators: Bicai Pan, Y.
Ye
Students: Venkatesh Selvaraj
Funding: NREL,
PRF(ACS)