Compositional and Structural - PowerPoint PPT Presentation

1 / 38
About This Presentation
Title:

Compositional and Structural

Description:

Compositional and Structural – PowerPoint PPT presentation

Number of Views:23
Avg rating:3.0/5.0
Slides: 39
Provided by: heikot4
Category:

less

Transcript and Presenter's Notes

Title: Compositional and Structural


1
Compositional and Structural Characterisation of
Indium Nitride using Swift Ions
Heiko Timmers Santosh Shrestha University of
New South Wales, University College, Canberra,
Australia Rakesh Dogra Aidan Byrne Australian
National University, Canberra, Australia Scott
Butcher Marie Wintrebert-Fouquet Patrick P.-T.
Chen Macquarie University, Sydney, Australia
The slides can be found on my home
page www.ph.adfa.edu.au/h-timmers
2
Abstract Despite the sensitivity of indium
nitride to irradiation damage and ion beam
induced dissociation, ion beam techniques can
successfully be applied to characterize this
material. Elastic recoil detection (ERD) analysis
using heavy ion beams enables the direct and
calibration-independent quantification of the
N/In ratio and the impurity content of indium
nitride thin films. Recent work has shown that
the beam-induced dissociation of In and N can
either be avoided or correctly be modelled, so
that precise atomic fractions can be determined
for In, N, O, C and H. Results consistently
indicate that state-of-the art films tend to be
nitrogen-rich in contrast to a common perception
which advocates nitrogen vacancies as the reason
for the native n-type characteristic of indium
nitride. In films grown with RPECVD a correlation
of nitrogen excess and absorption edge is
observed. The magnitude of the nitrogen excess
suggests that at least for films with high
absorption edges most of the nitrogen is
interstitial. The radioisotope probe
Indium-111 has successfully been introduced into
indium nitride grains and indium nitride films by
direct incorporation, by diffusion, and by
ion-implantation. Inside the lattice individual
probe atoms interact with the electric field
gradient at their site. This interaction reflects
the lattice environment in the immediate vicinity
of the probe, extending to typically the three
next atomic layers. The observation of this
interaction with perturbed angular correlation
(PAC) spectroscopy suggests that independent of
growth technique, at the atomic level, indium
nitride is characterized by a large and diverse
number of defects. Rapid-thermal annealing at
600o C, which is near the dissociation
temperature, has been found to reduce this
defectivity for both an RPECVD- and an MBE-grown
film. The results approach those achievable for
gallium nitride, for which a wider range of
annealing temperatures is accessible.
3
Part I n-type conductivity, precise
compositional measurements, nitrogen-excess
Part II implantation damage, radioisotope
studies, annealing behaviour
4
Part I n-type conductivity, precise
compositional measurements, nitrogen-excess
5
(No Transcript)
6
Does n-type InN look like this ?
interstitial H substitutional O
vacant N-site
7
Precise measurements on as-grown InN films
of N/In O-content H-content
8
Composition measurements of InN, an
experimental challenge
Simulation
  • SIMS
  • good for the detection of traces
  • but
  • not quantitative
  • lack of calibration for InN
  • x-ray yields
  • only qualitative
  • lack of calibration for InN
  • Ion beam analysis (RBS, HI ERD)
  • direct
  • quantitative

9
(ERD)
10
(No Transcript)
11
Bulk Molecular Depletion Model
12
MOCVD InN Briot et al.
with 1.3 MeV/u, low-Z 32S projectile ions the
nitrogen effusion during ERD analysis can be
avoided
S ions
Nitrogen (arb. units)
Au ions
number of projectiles (arb. units)
13
Elastic Recoil Detection Analysis (32S
projectiles)
LEXES
N/In

Beware of qualitative, uncalibrated
results !
trend
Samples from Patrick P.-T. Chen Macquarie
University, Sydney
14
includes any surface contamination
N/In
15
Summary of Results for RPECVD films
  • dramatic N-incorporation at lower growth
    temperatures
  • also, increased H,C
  • correlation between N,C
  • and growth temperature absorption edge
  • (Chen, Butcher et al. for details)
  • O-incorporation rather independent of temperature
  • quality, higher-T InN from RPECVD
  • N/In 1.01-1.05
  • O 0.6 atomic-
  • C lt 1 atomic-
  • H lt 0.5 atomic- (includes any surface
    contamination)
  • comparable with MBE material

16
(No Transcript)
17
Depth-profiling of InN films with ERD
MBE InN/AlN (Cornell)
Carbon and Oxygen signals are dominated by
surface results are upper limits
depth
AlN
InN
18
(No Transcript)
19
Outcomes of the Elastic Recoil Detection analysis
  • As-grown InN is always nitrogen-rich
  • The growth techniques introduce significant
    amounts of H, C, and O
  • The incorporation of excess nitrogen into the
  • lattice is unclear, but
  • in addition to substitutional and
    interstitial
  • N atoms, molecular N2 is possible

20
Does n-type InN look like this ?
(would explain n-type characteristic)
21
Part II implantation damage, radioisotope
studies, annealing behaviour
22
Relevance of ion implantation and annealing for
Indium Nitride processing
ion implantation
undesired structural modification
annealing
  • controlled doping
  • controlled structural modification

lateral structure
layered structure
not understood for Indium Nitride
(from Reiner Vianden, University of Bonn)
23
111In/Cd probe
  • standard probe, medical isotope
  • chemically identical to metal atoms
  • occupies In-sites in Indium Nitride

24
Perturbed Angular Correlation Spectroscopy with
the 111In/Cd probe
Detector 2
g1
stop
g2
t
Anisotropy function R(t) number of
g-pairs with time difference
t
start
Detector 1
In
N
25
Implantation Damage and Annealing of MOCVD
Gallium Nitride Perturbed Angular Correlation
Spectroscopy with the 111In/Cd radioisotope probe
(Lorenz, Ruske, Vianden, Appl. Phys. Lett. 80,
2002)
1300?C
as implanted 1013cm-2
annealed
RT
liquid Ga
shock-frozen
annealed
damping frequency decrease
solid Ga
annealed
Time (ns)
The ratio function R(t) reflects the lattice
environment of the 111In/Cd probe
Time (ns)
26
Interpretation
frequency of substitutional probes
Frequency change consistent with
implantation-induced lattice tension
fraction of non-disturbed substitutional probes
60 of probes on Ga-sites in a non-disturbed
lattice
damping of signal diversity of environment
Implantation-induced, non-specific lattice
disturbances
40 of probes in a disturbed environment
27
Perturbed Angular Correlation Spectroscopy of
Indium Nitride
  • only a single study published

Lorenz, Vianden, phys. stat. sol. c (2002)
  • sample contained Indium metal

In
R(t)
In
time ns
InN?
  • frequency assignment tentative
  • and inconsistent

InN?
AlN GaN InN 33 6 153? MHz
28
Frequency measurement for 111In/Cd in Indium
Nitride
Synthesis of bulk Indium Nitride
700 ?C tubular, similar to Schwenzer et al.
(2004)
650 ?C granular
29
650 ?C synthesis
Counts
Commercial InN powder
Confirmation of wurtzite structure with XRD
2-? (?)
Cu Ka
30
Introduction of 111In/Cd by diffusion
  • The previously observed frequency
  • of 153 MHz corresponds to In2O3
  • formed during probe diffusion at 500?C
  • The 111In/Cd frequency in InN is 28 MHz
  • Changes of magnitude, amplitude
  • and damping reflect structural
  • modifications near the In-site

31
Incorporation of 111In/Cd during synthesis
  • The frequency of 28 MHz is confirmed

AlN GaN InN 33 6 28 MHz
damped oscillation
  • This is consistent with the other III-nitrides
  • strong damping of R(t)
  • many and different defects near the In-site
  • not unexpected for a powder
  • possibly static defect configurations

no change
32
Implantation of 111In/Cd into Indium Nitride films
125 keV Implantation 111In radioisotope
probes plus 115InO- parasitic ions
Compositional Analysis with ERD
ERD
ERD
implanted area
  • Implantation fluences of 1014 ions/cm2
    completely remove the nitrogen

33
Implantation of 111In/Cd into Indium Nitride
films at low fluences
160 keV Implantation lt 1013 ions/cm2
125 keV Implantation gt 1013 ions/cm2
InN film from MBE-growth
R(t)
damped InN signal
In-metal
RPECVD InN
after annealing at 400 ?C
R(t)
InN film from RPECVD
Time (ns)
  • results for thin films mirror those for
  • bulk Indium Nitride
  • highly defective lattice
  • no difference between
  • the growth techniques

damped InN signal
after annealing at 400 ?C
Time (ns)
(carried out at the University of Bonn, in
collaboration with R. Vianden)
34
  • Tentative result
  • rapid thermal annealing at 600o C appears to
    significantly cure the lattice
  • pronounced oscillation characteristic of InN ?
  • no/little evidence of metallic In

MBE (Cornell)
RPECVD
no In ?
no In ?
35
Are the defects implantation-induced ? Separation
of implantation damage and radioisotope probes
In-situ production and recoil implantation of the
111In/Cd probe
Bezakova et al. Appl. Phys. Lett. (1999)
  • The undesired flux of ions passes through the
    implanted region at high velocities
  • local damage mainly by the low fluence of 1011
    cm-2 radioisotope probes

36
Furnace annealing of Indium Nitride
MBE-grown film
  • as-implanted signal unchanged
  • defects are native ?
  • with increasing annealing temperature
  • 48 MHz 28 MHz lattice stress is relieved
  • damping is somewhat reduced
  • probe fraction on the In-site increases
  • but, even after 400 ?C, the frequency
    distribution
  • is broad with a width of 30 of 28 MHz
  • many different types of defects
  • near the radioisotope probe on the In-site
  • after 600 ?C In-metal is observed, indicating
    dissociation

37
Conclusions
Does n-type InN look like this ?
  • independent of synthesis and
  • probe introduction,
  • PAC radioisotope studies of InN
  • give a broad and strongly
  • damped frequency spectrum
  • - highly defective material
  • many different types of defects
  • other evidence
  • 1018-1016 cm-3 In-vacancies
  • (Laakso et al. J. Crys. Growth 2004)
  • nitrogen-excess
  • H, C and O contaminations
  • furnace annealing is limited by
  • the low dissociation temperature
  • rapid thermal annealing appears
  • to significantly cure the lattice

,C
38
  • Outlook
  • collect more evidence for nitrogen excess
  • identify form and location of the excess
    nitrogen
  • establish correlations between excess nitrogen
    and other parameters
  • annealing of InN is complex and requires further
    study
Write a Comment
User Comments (0)
About PowerShow.com