Radiation damage problems in electron microscopy
Domain: Physics
A brief outline is given of the processes of knock-on damage and electronic damage in electron microscopy together with estimates of the damage probabilities. Of the various remedial techniques under investigation, that of low temperature observation seems generally the most promising. Attention is drawn to some of the outstanding problems particularly concerning the nature and behaviour of electronic excitations in molecular crystals and their role in the damage mechanism.
lire la suite replier
Infos
Stats
Publié le : 29/06/2012
Langue : Français
Nombre de pages : 5
Type de la publication : Rapports et thèses
Thème :
Savoirs > Science de la nature
Source : Revue de Physique Appliquée
Tags :
Electron microscope
0 vote(s) 0
17 lecture(s)
0 commentaire(s)
1 téléchargement(s)
Télécharger la publication
Du même auteur :
291
Radiation
damage
problems
in
electron
microscopy
(*)
A.
Howie
(**)
Laboratoire d’Optique
Electronique,
Laboratoire
propre
du
C.N.R.S.,
associé
à
l’Université
Paul-Sabatier,
Toulouse
29,
rue
Jeanne-Marvig,
B.P.
4007, 31055
Toulouse
Cedex,
France
Résumé.
2014
Cet
article
passe
rapidement
en
revue
les
processus
de
création
de
défauts
par
chocs
nucléaires
ou
interactions
électroniques
en
microscopie
électronique
ainsi
que
les
estimations
des
probabilités
de
création
de
défauts.
Parmi
les
différentes
techniques
à
l’étude
tendant
à
diminuer
les
dommages,
les
observations
à
basse
température
semblent
être
d’une
manière
générale
les
plus
prometteuses.
On
attire
l’attention
sur
quelques-uns
des
problèmes
les
plus
importants,
comme
la
nature
et
le
comportement
des
excitations
électroniques
dans
les
cristaux
moléculaires,
et
leur
rôle
dans
le
mécanisme
de
création
de
défauts.
Abstract.
2014
A
brief
outline
is
given
of
the
processes
of
knock-on
damage
and
electronic
damage
in
electron
micro-
scopy
together
with
estimates
of
the
damage
probabilities.
Of
the
various
remedial
techniques
under
investigation,
that
of
low
temperature
observation
seems
generally
the
most
promising.
Attention
is
drawn
to
some
of
the
out-
standing
problems
particularly
concerning
the
nature
and
behaviour
of
electronic
excitations
in
molecular
crystals
and
their
role
in
the
damage
mechanism.
Revue
Phys.
Appl. 15
(1980)
291-295
FÉVRIER
1980,
PAGE
Classification
Physics
Abstracts
61.80
1.
Introduction.
-
Because
of
the
very
extreme
irradiation
conditions
to
which
the
specimen
is
sub-
jected
in
the
electron
microscope,
the
subject
of
radiation
damage
is
often
of
paramount
importance
in
limiting
the
observations
which
can
be
made.
In
recent
years
a
number
of
excellent
reviews
have
been
written,
mainly
in
specialist
publications.
These
cover
on
the
one
hand
the
field
of
knock-on
damage
which
arises
from
the
interaction
between
the
fast
electrons
and
the
atomic
nuclei
[1],
[2],
[3]
and
is
primarily
of
significance
in
high
voltage
electron
microscopy
of
metals
or
semiconductors ;
and
on
the
other
hand
the
question
of
electronic
damage
in
poorly
conducting
materials
such
as
biological
speci-
mens
[4-12],
organic
polymers
[13],
[14]
and
ionic
crystals
[15],
[16]
where
the
damage
occurs
from
the
interaction
of
the
beam
with
the
electrons
in
the
sample.
The
knock-on
phenomenon
is
occasionally
a
nuisance
but
more
generally
forms
the
basis
of
valuable
techniques
for
simulating
reactor
damage
situations
in
a
greatly
compressed
time
scale
and
for
studying
microscopic
diffusion
processes
in
well
controlled
conditions.
The
electronic
damage
is
a
much
more
serious
obstacle
to
electron
microscope
studies
and
drastically
limits
the
resolution
currently
(*) Conférence
présentée
au
Congrès
de
la
Société
Française
de
Physique
(Toulouse).
(**)
Permanent
address :
Cavendish
Laboratory,
Cambridge,
England.
available
in
many
specimens.
Here
we
mainly
deal
with
this
aspect
of
the
subject
since
it
seems
useful
to
draw
the
attention
of
a
wider
audience
to
a
number
of
the
questions
which
it
raises,
particularly
in
solid
state
physics,
in
the
hope
that
their
expertise
might
be
useful
in
assisting
further
progress.
2.
Damage
probabilities.
-
2.1
MINIMUM
ELEC-
TRON
DOSE.
- To
form
an
image
at
resolution
d
when
the
available
contrast
is
C
=
0394I/I,
a
certain
minimum
electron
dose
De
(i.e.
number
of
electrons
incident
on
unit
area
of
the
specimen)
must
generally
be
used.
Assuming
that
a
fraction
f
of
the
electrons
is
effecti-
vely
utilised
for
imaging,
one
requires
that
the
signal
contrast
C
should
exceed
the
noise
to
background
ratio
(IDe
d2)-1/2
by
a
factor
k
usually
taken
to
be
about
5
[17],
[5].
We
thus
obtain
the
expression :
This
shows
for
example
that
to
record
a
satisfac-
tory
image
with
available
contrast
C
=
0.1
at
a
resolution
d
=
0.1
nm
when
f
=
1/2
of
the
electrons
are
utilised,
requires
a
dose
De
of
about
5
x
105
elec-
trons
per
(nm)2
equivalent
to
an
incident
charge
density
of -
7
x
104
Cm-
2.
2.2
KNOCK-ON
DAMAGE
PROBABILITY.
-
When
a
fast
electron
of
kinetic
energy
E
=
(m -
mo)
C2
and
velocity
v
=
fic
collides
with
an
atomic
nucleus
Article published online by
EDP Sciences
and available at
http://dx.doi.org/10.1051/rphysap:01980001502029100
292
of
mass
M »
m,
the
maximum
energy
transferred
occurs
in
conditions
of
head-on
collision
and
is
approximately
given
by
the
expression
[2],
[3] :
When Em
exceeds
a
certain
minimum
value
Ed,
the
threshold
displacement
energy
in
the
specimen
material,
the
atom
can
be
knocked-out
of its
site
with
the
result
that
a
vacancy
-
interstitial
pair
is
pro-
duced.
In
many
metals
and
semiconductors
where
Ed
is
typically
20
eV
or
more,
these
conditions
are
achiev-
ed
at
incident
kinetic
energies
E
which
can
be
several
hundreds
of
keV.
Below
this
threshold
no
damage
is
observed
but
above
threshold
the
damage
even-
tually
becomes
readily
visible
as
the
point
defects
diffuses
in
the
sample
and
agglomerate
to
form
dis-
location
loops,
voids,
etc...
A
great
many
vacancy
-
interstitial
pairs
can
recombine
during
this
process,
so
that
by
the
time
defect
clusters
are
observed,
each
atom
in
the
sample
may
actually
have
been
displaced
from
its
site
many
times
and
the
specimen
structure,
if
at
all
complex,
may
have
been
largely
destroyed.
The
presence
of
point
defects
generated
in
the
knock-
on
damage
process,
can
also
profoundly
affect
speci-
men
properties
as
observed
for
instance
in
the
case
of
in
situ
plastic
deformation
studies
where,
unless
one
can
operate
below
the
threshold
energy,
great
care
is
needed
in
interpreting
the
results
[18].
In
the
case
of
insulating
specimens,
experimental
information
about
knock
on
damage
is
not
available
since
the
effect
is
masked
by
the
electronic
damage
process.
If,
however,
we
assume
that
the
threshold
displacement
energy
Ed
is
as
low
as
5
eV
and
consider
only
light
atoms
Z
such
as
hydrogen
or
carbon,
the
ratio
Em/Ed
becomes
fairly
large
for
kinetic
energies E
of
100
keV
or
more.
The
expression
[2],
[3],
for
the
single
knock-on
cross
section
(in
barns)
then
sim-
plifies
to :
At E
=
100
keV
(fl £r
0.5)
we
then
find
03C3 ~
100
barns
in
H and 03C3 ~
300
barns
in
C.
In
conjunction
with
the
minimum
electron
dose
quoted
above
for
0.1
nm
resolution
we
could
thus
expect
about
0.5
%
of
the
H
atoms
and
perhaps
3’
%
of the
C
atoms
to
be
displac-
ed
in
thin
specimens.
Rather
higher
damage
figures
might
apply
to
thicker
specimens
where
multiple
knock-on
collisions
could
occur
[10].
These
results
suggest
that
although
knock-on
damage
may
not
be
negligible,
it
should
not
present
a
really
serious
problem
in
efficient
high
resolution
imaging
when
only
a
single
good
image
is
required.
2.3
ELECTRONIC
DAMAGE
PROBABILITY.
-The
gene-
ral
significance
of
electronic
damage
in
electron
microscopy
can
be
easily
appreciated
[4-6]
by
refe-
rence
to
the
Bethe
stopping
power
expression
for
the
total
rate
of
energy
loss
per
unit
path
length
of
a
fast
electron
traversing
a
medium
containing n
electrons
per
unit
volume :
The
quantity
7
is
a
mean
excitation
potential
characteristic
of
the
medium
and
equal
to ~
86
eV
in
carbon
where
the
rate
of
energy
loss
of
a
100
keV
electron
is
about
400
eV
per
micron.
In
the
example
taken
previously
of
imaging
at
0.1
nm
resolution,
the
total
amount
of
energy
lost
per
unit
volume
of
a
carbon
sample
would
then
be ~
3
x
1014
erg
cm-3
which,
by
definition,
corresponds
to
an
enormous
radiation
dose
of
3
x
1012
rad.
By
comparison,
reproductive
cell
death
occurs
at
a
dose
of
a
few
hundred
rad
in
animal
cells
and
at
about
104
rad
in
the
most
resistant
organisms
[5].
Although
most
organic
specimens
would
not
exhibit
damage
detec-
table
in
the
electron
microscope
until
doses
in
the
range
of
10’
rad
to
109
rad
are
given,
even
these
figures
are
clearly
quite
insufficient
to
allow
high
resolution
imaging
and
in
many
materials
of
interest
the
best
resolution
that
has
been
achieved
is
not
better
than
2
nm
or
even
5
nm
in
more
sensitive
specimens.
It
should
be
emphasised
qt
this
point
that
there
seems
to
be
general
agreement
[19],
[20],
[13]
that,
although
most
of
the
energy
transferred
by
the
fast
electrons
to
the
specimen
must
ultimately
be
convert-
ed
to
heat,
the
damage
itself
is
not
a
direct
conse-
quence
of
any
heating
effect.
Provided
that
excessive
illumination
intensity
levels
above -
20
Am- 2
are
not
employed,
with
the
specimen
in
good
thermal
contact
with
its
surroundings
and
the
illumination
does
not
fall
on
the
specimen
support
grid,
the
tempe-
rature
rise
of
the
specimen
is
in
general
only
a
few
degrees.
3.
Expérimental
situation.
-
3.1
DAMAGE
CROSS
SECTIONS.
-
In
the
absence
of
a
detailed
theory
of
the
electronic
damage
process,
it
is
necessary
at
present
to
rely
completely
on
empirical
data.
Several
methods
are
available
for
following
the
damage
process
directly
in
the
electron
microscope.
The
destruction
of
the
structure
in
crystalline
specimens
can
be
monitored
either
by
observing
the
disappea-
rance
of
image
contrast
features
such
as
extinction
contours
or
moire
fringes
or
by
the
fading
of
the
intensity
of
the
spots
in
the
diffraction
pattern.
In
the
case
of very
sensitive
samples
the
latter
method
is
easier
since
the
diffraction
pattern
can
be
obtained
by
illuminating
a
large
area
at
rather
low
intensity.
One
can
also
detect
the
mass
loss
from
the
sample
because
of
the
increase
in
transmission
as
it
becomes
293
thinner.
Finally,
if
an
energy
analyser
is
available,
the
changes
in
the
energy
loss
spectrum
can
be
observ-
ed
as
damage
proceeds.
The
disappearance
of
crystallinity
and
the
loss
of
mass
are
essentially
secondary
effects
which
follow
after
the
electronic
excitation
has
occurred
in
sufficient
quantity
to
lead
to
atomic
displacements
in
the
speci-
men.
The
energy
loss
spectrum
is
also
affected
by
these
secondary
effects
but
can
in
addition
provide
useful
information
about
the
initial
changes
in
elec-
tronic
structure.
Detailed
accounts
of
the
results
of
these
observations
on
the
damage
rate
in
a
large
number
of
materials
have
been
given
[5],
[6],
[7],
[9],
[12].
As
one
might
expect,
the
different
kinds
of
measurements
vary
in
their
sensitivity
to
the
damage
effect,
with
the
mass
loss
method
being
perhaps
the
least
sensitive
[9].
There
is
however,
reasonable
agreement
amongst
the
various
methods
in
determin-
ing
the
relative
damage
cross
sections
in
different
materials
which
vary
over
a
wide
range.
3.2
REMEDIAL
TECHNIQUES.
-
A
number
of
expe-
rimental
methods
are
successfully
employed
for
reducing
the
electronic
damage
effect.
Minimal
expo-
sure
techniques
have
been
developed
[21]
so
that,
for
instance,
the
specimen
area
of
interest
is
illumi-
nated
only
during
the
actual
exposure
and
focussing,
etc...
is
done
one
adjacent
regions.
In
the
case
of
periodic
structures,
further
advantage
can
be
taken
[22]
of
image
processing
techniques
which
allow
satis-
factory
images
to
be
obtained
at
electron
doses
far
below
the
value
given
in
eq.
(1).
Low
temperatures
have
also
been
employed
with
the
idea
that,
even
if
the
initial
electronic
excitation
pro-
bability
is
unaffected
by
temperature,
the
subsequent
movement
and
diffusion
of
atoms
(or
any
other
thermally
activated
stage
in
the
subsequent
damage
process)
can
be
inhibited
so
that
the
specimen
struc-
ture
can
be
frozen
in
at
least
for
the
length
of
the
exposure.
The
use
of
liquid
helium
temperatures
has
been
particularly
effective
in
alkali
halides
[15],
[16].
In
organic
specimens
the
improvements
achieved
have
generally
been
less
remarkable
(see
[9])
being
usually
less
than
a
factor
of
5.
Much
more
dramatic
results,
equivalent
to
an
increase
in
damage
dose
by
over
two
orders
of
magnitude
in
some
cases,
have
recently
been
reported
[23],
[45]
for
organic
specimens
studied
in
a
microscope
with
a
superconducting
lens
system
in
which
the
sample
is
held
at
liquid
helium
temperature
and
where
the
stability
allows
very
long
exposure
times
of
100
s
to
be
employed
with
corres-
pondingly
low
incident
beam
current
density
of
~
50 A
m-2
at
the
specimen.
Structures
of
0.45
nm
periodicity
have
thus
been
clearly
imaged
[23]
in
Pt-
stained
bacterial
cell
wall
material
but
could
not
be
imaged
at
normal
exposures
since,
as
the
authors
suggest,
the
beam
current
then
involved
is
sufficient
to
warm
the
specimen
by
a
few
degrees
allowing
atomic
migration
to
occur.
If
these
results
can
be
repeated
with
other
materials,
an
enormous
extension
of
electron
microscopy
in
the
organic
field
may
become
possible
with
the
use
of
high
stability
ultra-
low
temperature
specimen
stages.
Some
reduction
in
damage
has
also
been
achieved
in
a
number
of
cases
(see
[11])
by
enclosing
the
speci-
men
in
a
metal
coating,
embedding
in
ice
or
employ-
ing
a
special
stain.
A
wide
variety
of
effects
may
be
involved
here
and
some
of
the
possible
physical
mechanisms
are
discussed
below.
The
Bethe
stopping
power
expression
suggests
that
the
ionisation
damage
rate
could
be
reduced
by
a
factor
of
about
3
on
raising
the
microscope
voltage
from
100
keV
to
1
MeV
and
this
seems
to
be
in
rough
agreement
with
most
observations.
Unfortunately,
unless
special
photographic
emulsions
are
available
to
compensate
for
the
reduced
sensitivity
to
such
high
energy
electrons,
the
benefit
may
be
lost
because
of
the
increased
exposure
required.
The
image
con-
trast
may
also
be
reduced
if
thicker
specimens
are
not
used.
Considerably
larger
increases
of
lifetime
by
factors
of
30
or
more
at
high
voltages
have
been
observed
in
minerals
[24].
These
figures
relate
to
the
rate
of
formation of
point
defect
clusters
however
which
depends
in
a
complex
and
non-linear
way
on
the *rate
of the
initial
damage
event
[25].
In
crystals,
the
rate
of knock-on
damage
can
depend
on
orientation
because
of
diffraction
channelling
of
the
incident
beam
[26].
Once
again
the
effect
observed
can
be
greater
than
the
factor
of
4
or
so
which
is
the
maximum
one
could
expect
for
the
variation
in
the
rate
of
the
initial
damage
event.
A
similar
explanation
may
apply
to
the
dramatic
increase
in
the
rate
of
defect
cluster
formation
observed
[28],
[29]
in
some
minerals
when
the
Bragg
reflection
condition
is
met.
Even
if
the
initial
event
is
entirely
associated
with
inner
shell
ionisation
in
these
cases
one
would
not
expect
its
rate
to
increase
by
more
than
a
factor
of
perhaps
4
as
a
result
of
Bragg
reflection
of
the
pri-
mary
beam
[30].
The
potential
advantages
and
disadvantages
of
scanning
transmission
electron
microscopy
(STEM)
compared
with
conventional
transmission
micro-
scopy
have
also
been
discussed
[31]
]
with
respect
to
radiation
damage.
At
present
there
is
no
experimental
evidence
of any
particularly
significant
effect
although
there
may
be
some
operational
advantages
in
the
STEM
for
obtaining
minimal
exposure
conditions.
Scanning
diffraction
patterns
can
for
instance
be
conveniently
obtained
from
rather
small
regions
of
polymer
crystals
without
damaging
surounding
areas
[32].
The
temperature
rise
in
the
STEM
is
roughly
similar
to
that
in
the
conventional
instru-
ment
[9].
In
assessing
the
value
of
any
remedial
effect
it
should
be
born
in
mind
that
a
factor
of
4
reduction
in
the
damage
rate
corresponds
to
a
potential
improve-
ment
in
resolution
of 2
(by
eq.
(1)).
In
situations
where
the
resolution
is
actually
limited
by
instrumental
294
performance,
it
has
generally
been
regarded
as
worthwhile
to
expend
very
considerable
sums
of
money
to
achieve
this
kind
of
improvement.
4.
Physical
problems
in
the
electronic
damage
pro-
cess.
-
The
picture
of
the
electronic
damage
based
on
the
Bethe
stopping
power
expression
as
just
outlined,
is
clearly
greatly
oversimplified
and
more
detailed
considerations
reveal
many
problems
whose
solution
would
advance
our
knowledge
about
energy
transfer
and
radiation
processes
in
general
as
well
as
possibly
suggesting
new
methods
for
reducing
the
damage
problem
in
the
electron
microscope.
The
contribution
of
different
processes,
with
energy
loss
AE
=
hco,
to
the
total
stopping
power
is
conve-
niently
given
in
terms
of
the
complex
dielectric
constant
e(q,
w)
of
the
material.
The
probability
per
unit
path
length
p(w)
dw
for
generation
of
an
excitation
in
the
range
03C9 ~
m
+
dw
by
the
fast
electron
is
[33] :
where
6w
=
Iiwjmv2
and
the
momentum
transfer
hq
=
mv(02
+
03B8203C9)1/2.
Information
about
El
and
E2,
the
real
and
maginary
parts
of
e,
can
be
obtained
from
electron
energy
loss
spectra
as
well
as
from
optical
data
where
the
absorption
at
frequency
co
is
proportional
to
E2(0,
03C9).
4. 1
SECONDARY
ELECTRON
EFFECTS.
- The
func-
tion p(03C9)
in
eq.
(5)
can
be
used
to
estimate
the
various
contributions
to
the
total
energy
loss
given
by
the
Bethe
expression
(4).
In
the
case
of
carbon
films
for
instance
it
appears
that
approximately
half
the
contri-
bution
corresponds
to
valence
electron
excitations
with
03C9 ~
30
eV
(plasmons
in
this
case)
and
the
remainder
to
the
generation
of
secondary
electrons
with
energies
greater
than
this.
In
bulk
materials
the
energy
of
these
fast
secondaries
will
mostly
be
convert-
ed
into
further
valence
electron
excitations
but
in
thin
films
they
may
well
escape
with
a
considerable
fraction
of
their
energy
[6],
[14]
which
could
indeed
be
calculated
using
eq.
(5)
together
with
a
knowledge
of
their
directional
distribution.
Although
such
cal-
culations
have
not
been
done
it
is
clear
that
the
Bethe
expression
may
considerably
overestimate
the
energy
deposited
in
the
specimen.
It
should
also
be
pointed
out
that
in
the
case
of
an
insulating
specimen,
the
irradiated
region
will,
as
a
result
of
the
loss
of
these
electrons,
become
charged
up
to
the
point
where
are
all
attracted
back
to
this
area
once
again,
causing
additional
damage.
It
seems
possible
that
in
some
cases
the
beneficial
effect
of
a
conducting
coating
may
be
due
to
the
suppression
of
this
charging
effect
with
the
associated
retum
of
secondaries.
4.2
VALENCE
ELECTRON
EXCITATIONS.
-
The
com-
plex
dielectric
constant
el
+
iG2
provides
a
great
deal
of
information
about
the
valence
electron
excitations
as
well
as
a
comparison
with
optical
data.
The
latter
is
potentially
particularly
valuable
since
one
can
in
principle,
by
detecting
the
damage
generated
by
photons
of
different
energies,
determine
which
excita-
tions
are
critical.
In
the
electron
excitation
case
where
screening
effects
and
collective
oscillations
can
be
of
importance
the
excitation
probability
(eq.
(5))
depends
on
G2/1
e 12
rather
than
on
E2
as
in
the
optical
case.
It
may
be
that
in
some
situations,
for
example
when
organic
molecules
are
supported
on,
or
embedded
in,
a
thin
film
of
some
other
material,
that
the
dielectric
properties
of
this
material
can
be
chosen
to
provide
efficient
screening
and
reduce
the
damage
[34].
In
such
thin
films
one
should
use
an
improved
version
of eq.
(5)
taking
account
of the
effect
of
surface
plasma
modes
[35]
and
their
radiative
properties.
The
related
and
strong
influence
of
support
films
on
molecular
fluorescence
and
energy
transfer
has
already
been
studied
in
some
detail
[46].
In
inorganic
crystals
such
as
alkali
halides
and
silver
halides
where
a
great
deal
of
optical
studies
have
been
made,
the
crucial
role
of
excitons
with
energies
generally
in
the
range
5-10
eV
is
well
esta-
blished
[16],
[36].
These
can
decay
radiatively
(giving
luminescence)
but
can
also
migrate
to
defect
or
impurity
sites
where
the
electron
and
hole
can
be
separated
and
atomic
displacement
follows
by
a
mechanism
which
is
well
understood.
Damage
is
not
observed
in
MgO
for
instance
where
the
atomic
displacement
energy
exceeds
the
energy
of the
exciton.
When
higher
energy
valence
excitations
are
generated
eg.
plasmons,
the
energy
still
seems
to
be
transfered
to
excitons
by
a
process
which
has
an
efficiency
(in
terms
of
energy)
of
about
25 %
in
NaCI
[16].
The
electronic
properties
of
organic
and
inorganic
molecular
crystals
seem
as
yet
rather
poorly
under-
stood.
The
importance
of
excitons
is
recognised,
but
it
is
unclear
whether
the
concepts
of
band
theory
are
very
useful
(see
for
example
[37],
[38]),
so
that,
although
the
conductivity
of
phthalocyanines
and
aromatic
compounds
[39]
might
seem
to
offer
some
explantion
of
their
relatively
good
radiation
resistance,
it
may
be
misleading
to
describe
them
as
organic
semiconduc-
tors.
The
nature
of
the
higher
energy
excitations
in
these
materials
is
also
not
clear.
For
instance,
although
the
energy
loss
spectrum
of
adenine
(and
many
other
organic
compounds)
exhibits
a
broad
peak
at -
20
eV
somewhat
similar
to
the
plasmon
peak
observed
in
amorphous
carbon,
Isaacson
[7],
[9]
points
out
that
el
for
adenine
in
this
frequency
range
is
about
3/4
so
that
the
excitation
is
not
of the
collective
type
(however
see
[40]).
Isaacson
identifies
the
relevant
excitations
at
this
energy
as J -
Q*
transitions
and,
by
reference
to
optical
data
on
damage
processes,
concludes
that
it
is
these
transitions
(and
perhaps
also
inner
shell
ionisations)
rather
than
the
n -
03C0*
transitions
(which
295
provide
the
structure
below
10
eV
in
the
loss
spec-
trum)
which
are
responsible
for
the
damage.
On
the
other
hand,
Ritchie
and
Pope
[41]
suggest
that
the
observations
of
Salih
and
Cosslett
[42]
of
the
impro-
vement
in
the
damage
lifetime
of
coronene
as
a
result
of
applying
a
gold
coating
can
be
quantitatively
explained
by
the
effect
of
the
coating
in
causing
trapping
and
decay
of
excitons
at
the
surface.
In
this
picture,
the
différences
in
the
radiation
sensitivity
of
différent
organic
materials
would
be
primarily
explain-
ed
by
the
differences
in
generation,
diffusion,
trapp-
ing
and
decay
of
excitons.
Since
these
properties
depend
sensitively
on
impurity
and
defect
content
(see
for
example
[43]),
it
would
seem
useful
if
some
further
studies
of
damage
in
the
electron
microscope
could
be
linked
not
only
to
good
energy
loss
data
but
also
to
optical
absorption
and
luminescence
measurements
on
the
same
materials.
Finally
in
materials
such
as
crystalline
polythene,
where
with
careful
techniques
the
development
of
the
damage
can
be
followed
gradually
in
the
electron
micro-
scope
[14],
[44],
it
may
be
possible
not
only
to
observe
the
detailed
structural
changes
which
occur
but
also
to
clarify
the
role
of defects
in
the
process.
The
STEM
would
be
a
convenient
instrument
for
detecting
any
diffusion
of
energy
excitations
from
the
irradiated
area
to
such
defects.
Acknowledgments.
-
1
am
grateful
to
Dr.
B.
Jouf-
frey
for
his
kindness
in
extending
to
me
the
facilities
of
the
Laboratoire
d’Optique
Electronique
at
Tou-
louse
and
for
the
arrangement
of
financial
support
though
the
C.N.R.S.
during
my
stay
there.
A
number
of
colleagues
have
been
most
helpful
in
discussing
some
of
the
problems
of
damage
with
me
and
in
particular
1
would
like
to
thank
A.
Boudet,
P.
W.
Hawkes,
L.
P.
Kubin,
F.
Louchet
and
M.
0.
Ruault
of
Toulouse
together
with
R.
H.
Ritchie
(collaborating
under
a
NATO
Research
grant),
P.
Echenique
and
J.
R.
Fryer.
References
[1]
URBAN,
K.,
In
Microscopie
Electronique
à
Haute
Tension
(Soc.
Française
de
Microsc.
Electron.)
(1975)
159.
[2]
LAIDLER,
J.
J.
and
MASTEL,
B.,
In
Physical
Aspects
of
Electron
Microscopy
and
Microbeam
Analysis
(ed.
Siegel
and
Beaman,
J .
Wiley,
New
York)
(1975)
103.
[3]
MAKIN,
M.
J.,
In
Electron
Microscopy
(Electron
Microscope
Society
of
Canada)
III
(1978)
330.
[4]
STENN,
K.
and
BAHR,
G.
F.,
J.
Ultrastruct.
Res.
31
(1970)
526.
[5]
GLAESER,
R.,
In
Physical
Aspects
of
Electron
Microscopy
and
Microbeam
Analysis
(Ed.
Siegel
and
Beaman,
J.
Wiley,
New
York)
(1975)
205.
[6]
REIMER, L.,
Ibid
(1975)
231.
[7]
ISAACSON,
M. S.,
Ibid
(1975)
247.
[8]
Box,
H. C.,
Ibid
(1975)
279.
[9]
ISAACSON,
M.
S.,
Principles
and
Techniques
of
Electron
Micro-
scopy
(Ed.
Hayat)
7
(1977)
1.
[10]
COSSLETT,
V.
E.,
J.
Microsc. 113
(1978)
113.
[11]
COSSLETT,
V.
E.,
J.
Microsc.
Spectrosc.
Electron.
3
(1978)
551.
[12]
REIMER,
L.
and
SPRUTH,
J.,
J.
Microsc.
Spectrosc.
Electron.
3
(1978),
579.
[13]
GRUBB,
D.
T.
and
KELLER,
A.,
Proc.
5th
Eur.
Cong.
on
Electron
Microscopy
(Inst.
of
Physics,
London)
(1972)
554.
[14]
GRUBB,
D.
T.,
J.
Mater.
Sci.
9
(1974)
1715.
[15]
HOBBS,
L.
W.,
J.
Physique
Colloq.
C
9
34
(1973),
227.
[16]
HOBBS,
L.
W.,
Ultramicroscopy
3
(1978)
381.
[17]
ROSE,
A.,
Adv.
Electron. 1
(1948)
13.
[18]
MARTIN,
J.
L.
and
KUBIN,
L.
P.,
Ultramicroscopy
3
(1978)
215.
[19]
GALE,
B.
and
HALE,
K.
F.,
Br.
J.
Appl.
Phys.
12
(1961)
115.
[20]
REIMER,
L.
and
CHRISTENHUSZ,
R.,
Lab.
Invest. 14
(1965)
1158.
[21]
WILLIAMS,
R.
C.
and
FISHER,
H.
W.,
J.
Mol.
Biol.
52
(1970)
121.
[22]
HENDERSON,
R.
and
UNWIN,
P.
N.
T.,
Nature
257
(1975)
28.
[23]
DIETRICH,
I.,
FORMANEK,
H.,
FOX,
F.,
KNAPEK,
E.
and
WEYL,
R.,
Nature
277
(1979)
380.
[24]
LORIMER,
G.
W.
and
CHAMPNESS,
P.
E.,
In
High
Voltage
Elec-
tron
Microscopy,
Ed.
P.
R.
Swann
and
C.
J.
Hymphreys
(Ac.
Press,
London)
1974,
301.
[25]
BROWN,
L.
M.,
KELLY,
A.
and
MAYER,
R.
M.,
Philos.
Mag. 19
(1969) 721.
[26]
THOMAS,
L.
E.,
Radiat.
Eff.
5
(1970)
183.
[27]
CORDIER,
A.,
DEGEILH,
A.
and
SERIN,
G.,
C.
R.
Hebd.
Séan.
Acad.
Sci.
B
277
(1973)
455.
[28]
MCCONNELL,
J.
D.
C.
(private
communication).
[29]
KUBIN,
L.
P.
(private
communication).
[30]
CHERNS,
D.,
HOWIE,
A.
and
JACOBS,
M.
H.,
Z.
Naturforsch.
28a
(1973)
565.
[31]
MISELL,
D.,
J.
Phys.
Soc.
D10
(1977)
1085.
[32]
SHERMAN,
E.
S.
and
THOMAS,
E.
L.,
J.
Mater.
Sci.14
(1979)
1109.
[33]
PINES,
D.,
Elementary
Excitations
In
Solids
(Benjamin,
New
York)
(1963).
[34]
ECHENIQUE,
P.,
HOWIE,
A.,
RITCHIE,
R.
H.
and
VINAS,
J.
(to
be
published).
[35]
KROGER,
E.,
Z.
Phys.
216
(1968)
115.
[36]
SAIDOH,
M.
and
TOWNSEND,
P.
D.,
Radiat.
Eff.
27
(1975)
1 .
[37]
SHEIN,
L.
B.,
Phys.
Rev.
B
15
(1977)
1024.
[38]
BURSHTEIN,
Z.
and
WILLIAMS,
D.
F.,
Phys.
Rev.
B 15
(1977)
5769.
[39]
AKAMATU,
H.
and
INOKUCHI,
H.,
Proc.
3rd
Carbon
Conf.
(Buffalo)
(1957)
51.
[40]
EMMERSON,
L.
C.,
WILLIAMS,
M.
W.,
TANG
I’LANG,
HAMM,
R.
W.
and
ARAKAWA,
E.
T.,
Radiat.
Res.
63
(1975)
235.
[41]
RITCHIE,
R.
H.
and
POPE,
M.,
Philos.
Mag.
B 40
(1979)
503.
[42]
SALIH,
S.
M.
and
COSSLETT,
V.
E.,
Philos.
Mag.
30
(1974)
225.
[43]
SHIOMI,
N.,
J.
Phys.
Soc.
Japan
21
(1966)
907.
[44]
BOUDET,
A.
and
KUBIN,
L.
P.,
In
Electron
Microscopy
(Elec-
tron
Microscope
Soc.
of Canada)
I
(1978)
498.
[45]
DUBOCHET,
P.,
Chemica
Scripta
(in
press).
[46]
CHANCE,
P.
R.,
PROCK,
A.
and
SILBEY,
R.,
Adv.
Chem.
Phys.
37
(1978)
1.
Chargement...
Télécharger la publication
