Download Spectroscopy of Light Emission from a Scanning Tunneling

Spectroscopy of Light Emission from a Scanning
Tunneling Microscope in Air
R. Péchou, R. Coratger, C. Girardin, F. Ajustron, J. Beauvillain
To cite this version:
R. Péchou, R. Coratger, C. Girardin, F. Ajustron, J. Beauvillain. Spectroscopy of Light Emission from a Scanning Tunneling Microscope in Air. Journal de Physique III, EDP Sciences,
1996, 6 (11), pp.1441-1450. <10.1051/jp3:1996195>. <jpa-00249536>
HAL Id: jpa-00249536
https://hal.archives-ouvertes.fr/jpa-00249536
Submitted on 1 Jan 1996
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
J.
Phys.
III
France
6
(1996)
1441-1450
Spectroscopy of Light
Microscope in Air
R.
(*),
P4chou
d'(Iaboration
Centre
B-P.
R.
4347,
(Received
31055
3
PACS.61.16.Ch
PACS.78..66.-w
Emission
Coratger,
des
C.
Mat6riaux
Toulouse
May1996,
NOVEMBER
Cedex,
revised
5
from
Girardin,
et
d'(tudes
Scanning
a
F.
Ajustron
Structurales,
and
29
rue
1996,
PAGE
1441
Tunneling
J.
J.
Beauvillain
Marvig,
France
July1996,
accepted
2
September 1996)
Scanning probe microscopy: scanning tunneling
optical, magnetic force, etc.
Optical properties of specific thin films, surfaces,
superlattices quantum well structures,
structures:
micropartides
scanning
atomic
force,
and
low-dimensional
multilayers
acid
Abstract.
Light emission has been
detected
tip-sample junction of a Scanning Tunat the
neling Microscope (S.T.M.) in air on noble metallic
surfaces.
A spectroscopic
study of emitted
photons for Au-Au and
Ptlr-Au
tunneling junctions is presented.
The general aspect of the
used in the junctions; a study of the
materials
spectra depends on the
spectra as a function of
tunneling current and surface bias voltage reveals similar and reproducible
characteristics.
R4sum4.
Une 4mission de lumiAre a 6t6 d6tect6e
pointe-surface d'un
niveau de la jonction
au
microscope h effet tunnel dans l'air sur des surfaces de mAtaux nobles. Une 6tude spectroscopique
des photons dmis par des jonctions
tunnel
Au-Au et
Ptlr-Au
L'aspect g6n6ral des
est prdsent6e.
mat6riaux
utilis6s
6tude en
fonction
du
tunnel
tension
spectres d6pend des
courant
et de la
une
de polarisation de la jonction r6vAle des
caract6ristiques
similaires
reproductibles.
et
Introduction
Since the pioneering work of
Gimzewski
stimulated
by a
in 1988 [ii, light
emission
et at.
Scanning Tunneling Microscope (S.T.M.) has proven to be a new way of investigating surface
resolution.
In addition, this technique allows
of solids
properties with a high lateral
excitations
studied at a
subnanometric
scale using a high-resolution imaging tool [2-6].
to be
This
phenomenon, previously by Lambe and Mccarthy
detected
polarized planar juncon
theoretical
tions [7], involves
various
interpretations depending on the nature of materials
concerned.
Light emission from noble metallic surfaces has been
interface
attributed
plasmon
to
excited by inelastic
(T.I.P.)
modes
tunneling electrons.
These
localized Tip Induced
Plasmon
modes
expected to radiate in photons if the
translational
of their
invariance
k
momentum
are
II
is broken by surface roughness or by the tip itself [8-10]. In the case of
semiconductor
samples,
electron-hole
recombinations
responsible for photon emission [3, iii.
are
(*)
@
Author
Les
for
#ditions
correspondence
de
Physique
1996
JOURNAL
1442
Using
have
of
this
light
light
drawn
been
emission
as
and
few
up
in air
emission
a
PHYSIQUE
DE
III
N°11
imaging technique, photonic mappings of scanned
interesting properties have been evidenced such as the
surfaces
an
control
[12-14].
frequency
a
spectroscopic study on Au-Au and Ptlr-Au tunneling juncsurface
and tip material,
Au
the wavelength
of the
given
components
a
emitted
light strongly depends on the tip used. However, reproducible phenomena can be
observed
when tunneling
bias voltage are
increased.
The wavelength of the
emitted
current
or
photons is found to be independent of tunneling current but spectra depends on the materials
of tunneling junctions and on applied bias voltage, I-eelectrons.
For each
energy of tunneling
tunneling junction, the average spectrum at fixed bias voltage has been obtained from a set of
This
presents
paper
tions
in
For
air.
recorded
spectra.
1.
Experimental
The
experimental set-up has already
STl/I
operating in the
constant
size"
light
emitted
close
to
normal
to
Set-Up
described
been
tip-sample
of the sample surface.
the
Two
optic
fibers
adapted
was
by five optic fibers with
junction as possible (about
collected
~vas
[12-14].
elsewhere
mode
current
a
I
mm)
and
connected
were
"pocket-
customized
of I
diameter
core
A
photon
for
mm
oriented
placed
each
45°
relative
photomultipliers
to
The
detection.
to
and
as
the
three
spectrometer.
a
Photomultipliers (PM) (Hamamatsu R2949) operate
count
i-ate
of I
count
second
per
at
K in
255
the
in the
185-900
nm
pulse counting mode
range, the complete
with
a
dark
being
range
mappings of scanned
areas.
of an Optic
lfultichannel
Analyser (EG&G, O.M.A,1245) connected
with a C-C-Ddetector
cooled at 173 K and operating in the
The
300-1050
range.
nm
scattering system is a grating blazed at 750 nm. The effective spectral resolution of the specThe
Spectra are integrated over 30 s to increase signal-to-noise ratio.
is of 2 am.
trometer
of our
system (optic fibers, scattering system and C-C-D- detecmeasurement
response
curve
tor) has been taken into account and recorded spectra were corrected at each wavelength by an
function
from the expected
filament
calculated
black body spectrum of a carbon
appropriate
lamp.
simultaneous
This
experimental set-up allows
acquisition of S.T.M, topography, related phoof
the
scanned
mapping
and
emission
spectrum
ton
area.
Samples were
monocrystalline gold films evaporated to a thickness of 50 nm on freshly
sheets
cleaved
heated
mica
at 623 K during 24 hours
at 10~~ Torr [15]. A low deposition
s~~
of
of
allowed
the
gold
layer
along the [I iii direction on the hot
quasi-epitaxy
0,I
rate
nm
diffraction
These samples
substrate
demonstrated
by
X-ray
exhibited
experiments.
mica
as
nm~
high
flat
separated
by
prepared
typically
100
in
size
2
Tips
100
steps.
terraces
x
nm
were
Au or PtIr wires.
by cutting 0.25 mm
diameter
used
draw
to
The
Experimental
2.
EmIssioN
2.I.
recorded
set
photon
up
to
consists
spectrometer
for
1.8 V.
Results
SPECTRA
Au-Au
These
As
A
tunneling
spectra
tunneling junction
bias voltage of1.8 V,
junction being weaker
integrated over 60 s.
and
Au
as
FUNCTION
junctions
are
that
in
of
OF
in
were
Figure
an
TUNNELING
air in
in
shown
spectra
shown
than
Discussions
and
Figure
recorded
2.
Au-Au
The
CURRENT.
the 5 nA
The
I.
in
the
overall
junction,
nA
100
nA
emission
the
spectra
bias
range,
study
same
10
Several
has
rate
been
nA
100
of
given
done
range
a
in
spectra
were
voltage being
at
a
Ptlr-
surface
tunneling
Ptlr-Au
Figure
for
2
have
been
SPECTROSCOPY
N°11
OF
E3fISSION
LIGHT
FROM
A
STM
IN
AIR
1443
ioonA
S
I
°
50 nA
~
~
ii
Z
~
~
Z
IO nA
soo
5 nA
0
600
400
800
1000
WAVELENGTH(nm)
Fig.
Emission
1.
5 nA
100
is
current
spectra
nA
range.
observed.
Bias
of
surface
Au
an
voltage
has
by
scanned
been
to
set
Au-tip
an
V.
1.8
An
at
increase
various
in
tunneling
emission
rate
currents
with
in the
tunneling
iso
I
°°
80 nA
8
~
tl
~/
50
10nA
0
00
(nm)
WAVELENGTH
Fig.
the
Emission
2.
10
An
when
nA
100
increase
the
increased
spectra
nA
in
range.
emission
tunneling current
the tunneling
as
tionately with the
particle responsible
of
Bias
rate
Au
an
voltage
is
has
by
scanned
surface
been
set
clearly evidenced
to
1.8
but
Ptlr-tip
a
at
various
tunneling
currents
in
V.
emission
amplitude of each
wavelengths
remain
unchanged
is simply
photons
Then, the number of emitted
increases
proporof injected electrons,
thereby evidencing that the electron is the
number
for light emission
S-T-MThis
experimental result had already be
in an
is
increased.
current.
The
spectrum
component
JOURNAL
1444
j
PHYSIQUE
DE
III
N°11
2 05 V
j
f
~
l.95
V
u~
l 85 V
I 75
v
65 V
400
1000
800
600
(ml
WAVELENGTH
Fig.
Emission
3.
2.05 V
1.65 V
by
of
spectra
surface
Au
an
Tunneling
range.
by
scanned
been
has
current
set
an
nA.
to 10
Au-tip
at
Cut-offs
various
of each
bias
voltages
spectrum
are
in
the
indicated
arrows.
working principle of a S-T-M- relies on the strong
tip-sample distance dtjp-samp)e. Thus, changes of It in
the 5 nA
100 nA range (at a fixed bias voltage of1.8 V) are expected to cause
important modifications of dtip-sampiei a variation
of about 0.2 nm is expected in the
one-dimension
model
of tunneling
be
considered
fluctuation
around
current
[16]. Such a variation
cannot
as
a
an
mode, the feedback loop of the microscope
constant
current
average dtip-samp)e value: in the
does not allow such
distance.
variations
Emission
wavelengths are thus found
in the tip-sample
Rendell et at, ii?], predicted a dependence of frequencies of
to be independent of drip-sample.
localized
T-I-Pmodes
with (dtjp-samp)e)~R
photons being equal
The
emitted
energy of the
of the
excited
eletromagnetic modes [18], variations
should be
observed
in the
to the
energy
This model
ill-suited
when it
describing the
spectra positions.
to be
to correctly
seems
comes
phenomenon.
mentioned
by Sivel et at. [14]. The
dependence of tunneling current It
SPECTRA
EMISSION
2.2.
Experimental
AS
basic
on
FUNCTION
A
OF
VOLTAGE
BIAS
recorded for Au-Au tunneling junctions in
being set to 10 nA. These spectra are shown
with applied
in Figure 3. The overall
increase
rate of the tunneling junction is found to
bias voltage in this voltage range.
New high-energy
when
in the
components
spectrum
appear
bias voltage is increased:
700 nm-peak can easily be observed
when bias voltage reaches 1.8 V
a
2.2. I.
air in
and
in
the
high
three
to
range.
a
centered
greater
times
2.05
Results.
2.05 V
Several
at
spectra
tunneling
range,
emission
about
660
were
current
be
can
nm
seen
in
the
2.05
V
spectrum.
An
increase
amplitude (the 770 nm-peak and, when bias voltage reaches
bias voltage is also clearly shown:
the 770 nm-peak
becomes
the low-energy
when
bias
voltage
from
increases
1.65 V
components
component
energy
700 nm-peak) with
V, the
1.85
up
shoulder
a
the
1.65 V
V.
blue
than
Cut-offs
shift of
of each
spectrum
high-energy cut-offs
are
indicated
is also
in
evidenced
Figure
when
by arrows.
In this voltage
increasing bias voltage.
3
SPECTROSCOPY
N°11
EMISSION
LIGHT
OF
FROM
STM
A
IN
AIR
1445
300
2I5V
7
3
200
2 05 V
~
~
~/
u~
~
),95 V
'~
loo
1.85 V
175
v
165
v
400
600
800
1000
WAVELENGTH(nm)
Fig.
1.65
by
Emission
4.
V
2.15 V
of
spectra
Au
an
Tunneling
range.
surface
has
current
by
scanned
been
set
Ptlr-tip
a
bias
voltages
spectrum
are
various
at
Cut-offs
20 nA.
to
of each
in
the
indicated
arrows.
PtIr-Au
tunneling
tunneling
current
range,
PtIr-Au
These
junctions.
spectra,
tunneling junctions. The different
obtained in the case
the phenomena
In the
the
remain
centered
2.15
of
case
V
2.15
for
same
about
at
820
several
to
while
set
shown
of
it is
Au-Au
For
located
spectra
a
at
recorded
were
owing
nA
20
Figure 4,
in
components
junctions.
PtIr-Au
nm
junctions,
being
differ
from
in
weaker
the
to
those
air in
the
emission
obtained
1.65 V
rate
for
of
Au-Au
than sharp peaks, but
rather
consist of bumps
tunneling junctions when increasing bias voltage
sample bias voltage of1.65 V, the spectrum is
about 750 nm when sample bias voltage reaches
V.
As the wavelengths of the
different
Discussion.
spectra recomponents of recorded
unchanged when bias voltage is increased, emission frequencies are independent of bias
voltage. The
fundamental
mechanism
of light emission is thus found to be independent of bias
When
bias
voltage. However, new spectral components
in the 1.65 V
2.05 V range.
appear
voltage is increased, the potential barrier and thereby the Injected Electron Energy Distrimodified.
bution (I.E.E.D.) are
The
presence (or absence) of given peaks is therefore closely
2.2.2.
main
linked
the
to
I-E-E-D--
high-energy
Experimental
~~
(discrepancies
el~ip-surface
tral
resolution
cut-offs
observed
spectrometer).
of
our
of
PtIr-Au
are
for
in
good
low
bias
Low-energy
agreement
with
calculated
voltages
cannot
be
cut-offs
are
identical
ones
(l~ut-o~
explained by the
for
a
given
=
spec-
tip-sample
junction.
In
the
shows
at
a
given
for this,
account
between
injected
apex
case
that
and
an
Au-tip
electrons
the
two
and
may
tunneling junctions, the bumpiness observed on recorded
spectra
voltage the I-E-E-D- is different for an Au-tip and for a Ptlr-tip. To
hypotheses can be put forward. The I-E-E-D- is expected to be different
bias
a
Ptlr
lose
low-energy
tail
levels
energy
their
energy
of the
I-E-E-Dcan
one
part
as
of
are
different
through
thereby
in
these
be
two
materials.
Besides,
layer at the PtIr-tip
lengthened. A given tip-sample
contaminant
a
JOURNAL
1446
PHYSIQUE
DE
III
N°11
I
j
a
~
b
~/
#
£
b
400
Fig.
Au-tips
Two
Hz.
0.G25
(b)
Normalized
5.
different
exhibits
a
at
S-T-M-.
main
a
~vas
also
the
bias
peaks
sharp peak at
distinct
configuration
These
a
light
voltage
of
spectra
surface
visible
are
790
at
leads
have
two
one
be
the
2.0
due
to
the
tip
the
on
Spectra
shown
V
4.0
obtained
x
V,
about
790
V
that
fact
and
apex
a
200
a
and
nm
nm~
200
tunneling
current
880
at
5
a
bias
centered
sharp peak
the
same
Differences
scanned
and
with
(a).
two
of
rate
scan
a
while
spectrum
a
range.
sample
observed
given spectrum.
Bemdt
Recorded
shift
in
U-H-V-[18] in an
evidenced
spectra
of
peak
spectral
this
wavelengths
emission
of
experiment is conducted in air: the
presence
modifications
sample surface may cause
important
concerning the I-E-E-D- and Local Densities Of States
our
Voltage
Bias
during
voltage of1.8
recorded
were
surface
a
area
nA
spectrum
on
nm
Au
of15
the
on
Given
a
components
exhibits
from
at
Figure
in
Au-tips
discrete
(b)
by
emitted
of1.8
to
in
observed
Average
different
(nm)
be
the
Spectra
WAVELENGTH
nm.
tunneling barrier particularly
(L.D.O.S.) of the sample surface.
3.
1000
I-E-E-D- and consequently
a given
compared
with
obtained
by
to
spectra
In these
experiments, Au samples were scanned by W-tips.
spectral peak located in the 600-700 nm region. A blue
therefore
results
partially
may
contaminants
in
800
600
at
about
790
at
790
centered
surface
scans
of
V.
The
nm
and
nm.
scanned1N.ith
200
a
first
880
Figure
two
200 nm~ Au
spectrum
nm
5
x
respectively
clearly shows
different
tips
surface
with
(a) clearly
made
while
that
the
second
spectra
two
with
two
evidences
the
same
concerning spectra taken from Figures I
from two
different
and 3 (these
obtained
tunneling junctions made with
spectra have been
the
and under
close tunneling conditions). To average this, a large set of
materials
same
very
experimental spectra is required to study light emission from a given type of tunneling junction.
for different
different
recorded
scanned
with
Au-tips on different Au45 spectra
areas
were
monocrystalline samples at a fixed bias voltage of 1.8 V and a fixed tunneling current of10 nA.
normalized
An ai<erage
obtained by summing up these 45 spectra.
This
spectrum has been
of
shown
A
emitted
light
corresponds
in
Figure
given
6a.
spectrum is
spectrum
to
average
material
are
different.
The
same
remarks
can
be
made
SPECTROSCOPY
N°11
LIGHT
OF
EMISSION
FROM
STM
A
AIR
IN
1447
.5
.25
)
I
)
o.75
i~
o.5
a
0.25
o
500
600
700
80D
been
Normalized
at
a
fixed
obtained
from
voltage
light
by
emitted
V, tunneling
recorded
during scans
of1.8
spectra
45
of
spectra
average
bias
(a)
Au-Au
an
being
current
different
of
l100
(nm)
WAVELENGTH
Fig. 6.
junction
1000
900
set
and
lo
at
Au
surfaces
(b) tunneling
Ptlr-Au
a
nA.
spectra
These
with
different
have
Au
and
Ptlr-tips.
configuration
given tunneling
a
with
a
given tip
tip
geometry,
oxidization
apex
and
state
of a given spectrum do not depend
(I.e.
characteristics
area
only on the sample surface). These tunneling conditions impose a given I-E-E-D- and L-D-D-Sand consequently a given spectrum shape (the fine
observed on
emission
structure
spectra could
be linked to these
peculiar I-E-E-D- and L-D-D-S-)- By summing up all these specific spectra.
contamination
characteristic
these
a
of the
state
large peak
As for
neling
effects
the
in
scanned
tunneling
Au-Au
junctions
toned
are
at
junctions,
fixed
a
over
bias
45
the
and
region than
near-infrared
as
spectra
voltage
of1.8
spectrum
average
the
been
have
V
of two
sum
and
a
can
or
recorded
be
better
three
distinct
for
different
tunneling
described
Ptlr-Au
of10
current
as
components.
nA.
tun-
These
spectrum of Figaverage
First of all, the
Figure
6.
spectra
ure
can
low-energy sides of these average spectra are different. To account for this, the two hypotheses
(modifications of I-E-E-D- due to a
previously
mentioned
layer on the tip apex
contaminant
differences in energy levels) can be proposed.
The second
difference
and
between
these
two
recorded
fib.
have
spectra
been
summed
differences
Two
spectra
average
The
transmission
be
concerns
their
factor
of the
slight difference in I-E-E-Dcut-offs.
in high-energy
may
up
observed
high-energy
obtain
the
normalized
between
the
of
to
cut-offs:
a
potential barrier being
be amplified by the
difference
of
about
0.05
eV is
obseri>ed.
high-energy electrons, a
strong
very
factoi~leading to such offsets
transmission
for
Unfortunately, the average spectrum of light emitted by an AuAu tunneling junction at 2 V
Indeed, a decrease in emission rate is observed when applying higher
is very difficult to obtain.
scanned
light-emitting
than
usual
bias voltages: light
be suppressed
emission
can
a
on
area
by applying
interpretation
can
be
electric
of
attributed
field
of typically 2 V for Au-Au
tunneling junctions [13]. A possible
phenomenon has already been given [14]: this decrease in emission
rate
the
of the
molecules
physisorbed on the gold surface to the high
response
voltage
bias
a
this
to
applied
in
the
tunneling junction.
The
bias
voltage
range
available
for
the
study
JOURNAL
1448
DE
PHYSIQUE
III
N°11
.5
l.25
j
b
e
f
0.75
j~
)
uQ
0.5
0.25
600
500
700
800
900
1000
l100
WAVELENGTH(nm)
Fig. 7.
voltages
Normalized
of
obtained
V
1.8
from
45
average
and 2 V
(a)
spectra
of light
tunneling
during scans
spectra
emitted
(b),
current
recorded
of
by a
being
different
tunneling
Ptlr-Au
set
nA.
10
at
surfaces
Au
junction
These
with
fixed
at
bias
have
spectra
been
Ptlr-tips.
different
junction is thus limited by this 2 V-threshold.
To increase
the
the
has
performed
2V-average
study
been
spectrum,
range
a
same
PtIr-Au
Indeed, the bias voltage threshold leading to a suppression of light
junctions.
on
emission in the case of PtIr-Au
junctions has been found to be greater than the one needed for
Au-Au tunneling junctions (typically 2.3 V in this case).
emitted by PtIr-Au
45 spectra of light
junctions have been recorded at a fixed bias voltage
of
light
emission
from
of 2 V and
presented
using the
and
tunneling
a
in
Fig. 6b)
is
data
same
bias
cut-off
voltage,
The
blue-shifted.
is
the
distribution
presence
[19] with
of
obtain
of10
current
shown
in
of
nA.
Figure
The
7a
(Fig. 7b).
treatment
high-energy cut-offs
voltage is increased,
two
Au-Au
an
voltage
available
these
1.8
the 2
An
offset
spectra,
average
V-normalized
with
of 0.16
their
low-energy tail of the I-E-E-D.
Average spectra given in Figure 7
of emitted
photons being closely
the
different
components
in
individual
eV
sides
is
are
can
be
observed
modified
not
thus
to
has
(previously
obtained
spectrum
average
being roughly
linked
spectra
spectrum
average
V-normalized
between
the
but
dependent
same.
its
on
the
I-E-E-D-.
also
been
the
When
high-energy
applied bias
observed
by
Ito
experimental set-up.
The
authors
reported that polycrystalline
a
U-H-V--S-T-Memitted
characteristic
Au samples
scanned by PtIr tips in an
wavelengths
two
(620 nm and 730 nm) at a fixed bias voltage of 2.5 V and a tunneling current of 5 nA. Here
again, differences in emission wavelengths could partially be due to changes in the I-E-E-Dand for by the
of
caused by the
oxidization
contaminants
in the
state of the tip apex
presence
tunneling barrier when the experiment is conducted in air.
et
at.
different
Conclusion
Spectroscopy of light
different
evidences
spectra
are
found
from a Scanning Tunneling Microscope in air on
near-infrared
wavelengths in the
region. Even if
specific for a given tip-sample configuration, reproducible
emission
emission
to be
Au
the
samples
recorded
phenomena
SPECTROSCOPY
N°11
LIGHT
OF
EMISSION
FROM
A
STM
IN
AIR
1449
wavelengths are found to be independent of tunneling
current
thereby
of
tip-sample
The
wavelengths
of
these
and
distance.
range
different
closely linked to the basic emission
mechanism, are found to
spectral
components,
and high-energy
be independent of bias voltage.
However, new spectral components
appear
cut-offs are
blue-shifted
when the Injected
Electron Energy
Distribution
is modified via applied
PtIr-Au
bias voltage. The results
obtained in the case of Au-Au and
tunneling junctions show
emission
spectra to be dependent upon tip materials and applied bias voltage. Injected Electron
from a S-T-M-.
Energy Distribution
emission
to play a major role in light
seems
can
be
the
in
observed.
nA
5
Emission
nA
100
Acknowledgments
The
authors
like
would
ported by the
thank
to
Ultimatech
G.
Peuch
for
technical
assistance.
This
has
work
been
sup-
program.
References
B., Coombs J-H- and Schlitter R-R-, Photon
emission
Cond.
Matter 72 (1988) 497.
microscope, Z. Phys. B
Photon
induced
emission
by the scanning tunneling microscope,
J-K-,
Gimzewski
[1]
Reihl
with
the
scanning tunneling
[2]
[3]
Berndt
R.,
Microscopy
Renaud
Scanning
687.
Alvarado S-F-, Mapping quantum-well
energy profiles of
scanning-tunneling-microscope-excited
luminescence, Phys.
P.
heterostruc-
III-V
and
by
tures
(1995)
9
Rev.
B
(1991)
44
6340.
[4]
[5]
J.,
Horn
Richter
of
GaAs
20
(1993)
R., Hartnagel H-L-, Sprbssler C-A-, Bischoff M. and Pagnia H., Studies
by scanning tunneling induced photon emission, Mater. Sci. Engin. B
183.
P., Gu M., Syrykh C.,
Dumas
F.,
surfaces
of
observation
Direct
Gimzewski
individual
J-K-,
I., Halimaoui A, and
light emitting
structures
on
Makarenko
nanometer-sized
surfaces, E~trophys. Lett. 23 (1993) 197.
Vazquez de Parga A-L- and Alvarado S-F-, Magnetic
observed
with
scanning-tunneling-microscope-excited
Salvan
porous
silicon
[6]
(1994)
[7]
[8]
J,
[10]
[11]
[12]
[13]
Mccarthy S.L., Light
and
France
Phys.
in
Cobalt
Rev.
Lett.
films
72
37
(1976)
emission
from
inelastic
electron
tunneling, Phys.
Rev.
923.
of tip-induced
Johansson P., Inelastic
tunneling
excitation
surfaces, Phys. Rev. Lett. 67 (1991) 3796.
Persson
B-N-J-, Baratoff A., Theory of photon emission in electron tunneling to metallic
particles, Phys. Rev. Lett. 68 (1992) 3224.
Johansson
P., Monreal R., Apell P., Theory for light emission from a scanning tunneling
microscope, Phys. Rev. B 42 (1990) 9210.
Berndt R, and
Gimzewski
luminescence
from CdS(11§0) studied
with
J-K-, Injection
scanning tunneling microscope, Phys. Rev. B 45 (1992) 14095.
Sivel V., Coratger R., Ajustron F. and
Beauvillain
stimulated
J., Photon
by
emission
scanning tunneling microscopy in air, Phys. Rev. B 45 (1992) 8634.
Sivel V., Coratger R., Ajustron F. and
Beauvillain
J., Control of photon emission by
scanning tunneling microscopy in air, Phys. Rev. B 50 (1994) 5628.
Berndt
plasmon
[9]
dichroism
3726.
Lambe
Lett.
circular
fluorescence,
R.,
Gimzewski
modes
on
J-K-
and
noble-metal
1450
JOURNAL
DE
PHYSIQUE
III
N°11
Sivel V., Coratger R., Ajustron F. and
Beauvillain
J., Interpretation of the control of the
photon emission
stimulated
by STM, Phys. Rev. B 51 (1995) 14598.
template-stripped Au
[15] Hegner M., Wagner P. and Semenza G., Ultralarge atomically flat
surfaces for scanning probe microscopy, S~trf. Sci. 291 (1993) 39.
(Mir Editions, Moscow, 1947).
Mechanics
[16] Landau L, and Lifchitz E., Quantum
tunnel
microstructures
[17] Rendell R.lV, and Scalapino D-J-, Surface plasmons confined by
on
junctions, Phys. Rev. B 24 (1981) 3276.
from the scanning tunneling
emission
microscope, PhD Thesis (Basel
[18] Berndt R., Photon
University, 1992).
spectroscopy of surface
[19] Ito K., Ohyama S., Uehara Y. and Ushioda S., STM light emission
granular Au films, S~trf. Sci. 324 (1995) 282.
micro-structures
on
[14]