Surface infulence on stability and structure of IIIV nanorods: Firstprinciples studies

1
Surface infulence on stability and structure of
III-V nanorods First-principles studies
Roman Leitsmann and Friedhelm Bechstedt
Institut für Festkörpertheorie- und optik
Friedrich-Schiller-Universität Jena
Sponsored by
2
Motivation

• Application

Eric Mankin mankin_at_usc.edu
Nanowire Memory Cells
In2O3 - Nanowire -gt possible data storage rates
of 40 Gbits/cm2 (USC / NASA Ames Research Center.)
National cancer Institute
Nanowire Sensors
NASA
Logical circuits
Nanowire properties are very sensible to certain
environmental conditions
3-D nanowire-based logic gate (inverter) -gt
push down CMOS gate length to 1 nm
3
Motivation

• Growth

many different growth-techniques
most of them is common

• gaseous environment
• high temperatures

can expect

• surface passivation
• equilibrium rod-structure

4
Motivation

• Growth

InP
depending on growth conditions

• stacking faults
• rotational twin layers
• structural change zb -gt wz

Nanotech. 16, 2903 (2005)
usually growth axis 111/0001
GaAs
hexagon shape of rod cross-section
Appl. Phys. Lett. 79, 3335 (2001)
5
Aim of this study

• Theory

Aim of this study

• ab initio description of III-V nanorods
• investigate the zb -gt wz transition of III-V
  nanorods
• for different rod-structures
• for different environmental conditions
• investigate structural properties like
  bond-length
• extrapolation to thick (gt 5 nm) nanorods

6
Modelling

• Rod types

Theoretical model

• free standing symmetric (C3v) III-V
  semiconductor nanorods
• no influence of the substrate
• total energy considerations
• no kinetic effects
• consider two different rod diameter
• infinitely thick rods are modelled by free
  surfaces

Structural rod models

• hexagonal cross-section
• III-V pair bond parallel to the rod axis in the
  centre of the rod
• consider two polytypes zincblende (zb) and
  wurtzite (wz)
• consider two types of side facets
• type I (112) zb or (1100) wz
• type II (110) zb or (1120) wz

7
Modellig

• Type I

zb

zincblende
wurtzite
Pseudo-hydrogen
dangling bonds
wz
4/3
3/2
dangling bond densities
4v2/(a02 v3)
3v6/(2a02)
8
Modellig

• Type II

zb

Pseudo-hydrogen
zincblende
wurtzite
wz
dangling bonds
1
1
dangling bond densities
2v2/a02
2v2/a02
9
Numerical Details

• DFT-LDA ground state calculations (VASP 4.6.20)
• projector augmented wave (PAW) pseudopotentials
• 200 eV plane wave-cutoff
• preconditioned residuum-minimization method for
  the
• electronic relaxation
• conjugate-gradient-algorithm for the ionic
  relaxation

• Rod-Surface calculations
• vacuum region 2 nm (rod) and 1.5 nm (surf)
• supercell length 12 (rod) 6-8 (surf) atomic
  (double,triple)layers
• BZ-sampling Monkhorst-Pack k-points

10
Modelling

• Formation energies

General procedure

• rod-structure is ideal (bulk-like) - no
  relaxation !!
• passivation with pseudo-hydrogen relaxation of
  H-positions
• calculation of the rod-surface energies

Free rod-surface energy ?
? environmental influence
? diameter dependence
11
Results

• InAs Rod

? Very fast convergence to free surface energies
12
Results

• III-V Rods

other III-V semiconductors
Type I, pass
? same convergence behaviour
13
Results

• Extrapolation

Extrapolation to thick nanorods

• free surface energies
• dangling bond densities of the corresponding
  diameter
• difference between zb and wz (25 meV per pair)
  has to
• be taken into account

gt for very thick rods only zb structure is stable
14
Results

• Phase-diagram

InAs 5 nm Nanorod
clean type II
pass. type II
pass. type I (wz)
15
Results

• Rod structure

Structural properties
(clean rods)

• starting rod-structure is ideal (bulk-like)
• relaxation of c-lattice constant (for diag 2,
  free)
• ?c cideal - c0 0.036 nm 1.7 (zb)
• ?c 0.059 nm 2.8 (wz)
• relaxation of a-lattice constant at c0 (for diag
  2, free)
• ?a aideal - a0 -0.001 nm 0.06 (zb)
• ?a -0.001 nm 0.06 (wz)

gt negative deviation of c/a ratio
16
Summary

• investigated III-V nanorods in wz and zb
  structure
• with ab initio methods
• achieved an extrapolation to thick nanorods
• investigated influence of nanorod-passivation,
• i.e. the influence of the environment
• could explain the occurrence of wz and zb
  nanorods
• found a negative deviation of c/a ratio

17
Thank you for your attention.
Sponsored by
18
Results

• Dangling bonds

• fast convergence
• different for type I
• equal for type II

Type I, wz
Type I, zb
Type II, wz,zb