The properties of surface oxidation of zirconium by auger electron spectroscopy and secondary ion
mass spectroscopy
by Tair-ji Lee
A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in
Chemical Engineering
Montana State University
© Copyright by Tair-ji Lee (1988)
Abstract:
Zirconium dioxide is one of the most active isosynthesis catalysts. It has also found application as a
construction material for nuclear reactors and high temperature devices.
To develop an improved understanding of the surface oxidation of zirconium, the adsorption of oxygen
on polycrystalline zirconium under ultra high vacuum has been studied using Auger Electron
Spectroscopy (AES) and Secondary Ion Mass Spectroscopy (SIMS). AES measurements have been
conducted for zirconium samples with different extents of oxygen exposure over temperatures between
300 K to 673 K. Auger spectra were also recorded for Zr surfaces with the same extents of oxygen
exposure but accomplished under different oxygen pressures and exposure times. The process of
sputtering an initially 60 L O2-exposed Zr sample was studied employing SIMS. AES spectra were
recorded at sequential interruptions during sputtering.
Oxygen adsorption seems to follow the reaction sequence of first chemisorption, next rapid oxide
nucleation, and finally slow oxide thickening. Diffusion of oxygen onto the bulk of zirconium is
negligible at temperatures below 470 K and becomes significant at temperatures between 470 K and
673 K. Oxygen uptake was found to primarily depend on Oz exposure level and slightly on Oz
exposure pressure (or exposure time). Secondary ion yield of the ZrO+ SIMS signal is closely related
to the oxidation state of the Zr surface. THE PROPERTIES OF SURFACE OXIDATION OF ZIRCONIUM BY
AUGER ELECTRON SPECTROSCOPY
AND
SECONDARY ION MASS SPECTROSCOPY
by
Tair-ji Lee
A thesis submitted in partial fulfillment
of the requirements for the degree
of
Master of Science
in
Chemical Engineering
MONTANA STATE UNIVERSITY
Bozeman, Montana
April, 1988-
ring
J^S/s s
ii
APPROVAL
of a thesis submitted by
Tair-ji Lee
This thesis has been read by each member of the thesis
committee and has been found to be satisfactory regarding
content, English usage, format, citation, bibliographic
style, and consistency, and is ready for submission to the
College of Graduate Studies.
Chairperson, Graduate Committee
Date
Approved for the Major Department
Date
Approved for the College of Graduate Studies
Lfs ~ti ' %
Date
Graduate Dean
iii
STATEMENT OF PERMISSION TO USE
In presenting this
the requirements for
University,
I
a
agree
thesis
in
master's
that
partial fulfillment of
degree
the
at Montana State
Library
shall
make
available to borrowers under
rules
quotations from this
are allowable, without special
permission,
provided
thesis
that
of the Library.
it
accurate
Brief
acknowledgement
of
source is made.
Permission
reproduction
of
professor, or in
when,
in
for
this
his
opinion . of
extensive
thesis
absence,
either,
may
by
quotation
be
the
the
material is for scholarly purposes.
granted
from
by
or
major
Dean of Libraries
proposed
use
of the
Any copying or use of
the material in this thesis for financial gain shall not be
allowed without my written permission.
iv
ACKNOWLEDGEMENTS
The author would
like
to
express
his thanks to the
faculty and staff of the Chemical Engineering Department at
Montana State University for
A special thanks go to
Dr.
for their
excellent
advice
research.
Thanks also goes
their encouragement and help.
Max Deihert and Dr. John Sears
and
guidance throughout this
to Dr. Turgut Sahin who served
on my thesis committee and has been a very helpful friend.
The author wishes to extend his thanks to Mr. and Mrs.
Tim Warner as well as
Kris
and Tim Jr. for their family's
genuine concern and moral supportr to Mrs. Becky Warren and
Mrs. Norma Ritter for their help in tutoring me English.
Next, I would like to thank my father, Huai-Sheng Lee,
and
my
mother,
Chun-Mei
Young,
for
their
moral
and
financial support.
Finally, arid most importantly,
my wife, Lio,
research.
for
her
patience
I
would like to thank
and assistance with this
V
TABLE OF CONTENTS
Pacre
APPROVAL.............................
iii
STATEMENT OF PERMISSION TO USE.......................
ACKNOWLEDGEMENTS.......................................
iv
TABLE OF CONTENTS......................................
v
LIST OF TABLES.............. ............... .........
viii
LIST OF FIGURES......................................
ix
ABSTRACT.... ...................
INTRODUCTION................................. .......
xii
I
BACKGROUND.. ..........................
Zirconium................ .............. .........
3
Auger Electron Spectroscopy and
Auger Transitions of Zirconium.... ........
7
Auger ElectronSpectroscopy....................
Auger Transitions ofZirconium.................
Secondary Ion MassSpectroscopy......................
7
10.
11
vi
TABLE OF CONTENTS (Continued)
Pacre
Oxidation of Zirconium.... . ............... .......
14
Studies by Non-surface-spectroscopic Methods...
Studies by Surface-spectroscopic Methods.......
.14
17
EXPERIMENTAL SYSTEM AND PROCEDURE........... ........22
RESULTS..........
29
AES Results from Sequential Oxygen Exposures......
29
Differentiated and Undifferentiated AES.......
Normalized Peak Height Ratios.................
Energy Shifts........ .................;.......
29
33
35
AES Results from Cyclic Oxygen Exposures..........
45
Auger Spectra above Room Temperature...........
45
Ion Sputtering Monitored with SIMS and AES.........
48
DISCUSSION OF RESULTS................................
52
Influence of Surface Oxidation on Auger Spectra....
54
Adsorption Kinetics of Oxygen on Zirconium........
61
Oz Exposure Pressure Dependence
on Auger Spectra Intensity........................
65
Auger Results above Room Temperature..............
66
Ion Sputtering Monitored with SIMS and AES.... .
67
CONCLUSIONS..................................
71
vii
TABLE OF CONTENTS (Continued)
Pacre
RECOMMENDATIONS...................
REFERENCES....................... .............. .....
APPENDICES.....................
73
.74
79
Appendix A ........................................
80
Appendix B .................
82
viii
LIST OF TABLES
Table
1.
Pacre
Auger Transitions from Oxygen Exposed
Zirconium......................................
H
Dependence of Auger Energies of Zirconium and
Oxygen Transitions on Zirconium Surface
Oxidation......................................
54
3.
Energy Shifts of Auger Peaks............
60
4.
The Energy Shifts of Auger Peaks Determined
by Averaging Peak Maxima and Peak Minima in
dCN(E)*E3/dE and those Determined by the
Maximum Points in N(E) *E Spectra...............
80
Exposure Conditions for the Data in Figure 24__
83
2.
5.
iz
LIST OF FIGURES
Figure
I.
Page
The energy states of electrons in zirconium
metal and the schematic diagram of Zr
(MvNzaNza) Auger transition...........
4
2.
The three possible types of Auger transitions.„.
9
3.
An example of a differentiated Auger
spectrum, taken from slightly oxidized .
zirconium.............................
12
The sequential oxygen exposures and AES
measurements of this study.....................
25
The integrated SIMS spectra of essentially
clean Zr and surface oxidized Zr....... ■.......
28
Undifferentiated "survey" Auger spectra
from a clean Zr sample and a Zr sample
after 100 L Oz exposure at room temperature....
30
Differentiated "survey" Auger spectra
from a clean Zr sample and a Zr sample
after 100 L Oz exposure at room temperature....
30
Multiplexing Auger spectra of Zr(80-180 eV)
from oxygen uptake on zirconium at different
oxygen exposures at room temperature........ . „„
31
Multiplexing Auger spectra of Oxygen(495515 eV) from oxygen uptake on zirconium at
different oxygen exposures at room
temperature..................................
32
The intensity of the ZRsz peak at different
oxygen exposures at room temperature...........
34
The intensity ratio of the ZRize peak from a
Zr sample at different oxygen exposures at
room temperature..............................
36
4.
5.
6.
7.
8.
9.
10.
11.
X
LIST OF FIGURES (Continued)
EiHure
12.
13.
14.
15.
Page
The intensity ratio of the Z R n v peak from a
Zr sample at different oxygen exposures at
room temperature...............................
37
The intensity ratio of the ZRi42 peak from a
Zr sample at different oxygen exposures at
room temperature...............................
38
The intensity ratio of the ZRi4s peak from a
Zr sample at different oxygen exposures at
room temperature...............................
39
The intensity ratio of the ZRivs peak from a
Zr sample at different oxygen exposures at
room temperature................
40
Z
16.
17.
18.
19.
20.
21.
22.
The intensity ratio of the Osoa peak from a
Zr sample at different oxygen exposures at
room temperature. ...........
41
The energy shift of the ZRsz peak from a Zr
sample at different oxygen exposures at
room temperature...............................
43
The energy shift of the Z R n v peak from a
Zr sample at different oxygen exposures at
room temperature...............................
44
The pressure dependence on the Osos peak
intensity ratio using Langmuirs as the
parameter for Oz exposure......................
46
The intensity ratio of the Osos peak from a
Zr sample at different oxygen exposures
above room temperature(linear scale).........
47
The decay of ZrO+ SIHS signal intensity
during sputtering a Zr sample after 60 L
Oz exposure....................................
49
The decay of the Auger intensity ratio of
the Osos peak during sputtering a Zr sample
after 60 L Oz exposure
50
xi
LIST OF FIGURES (Continued)
Figure
23.
24.
25.
26.
-Pacre
The+relationship between the decay of SIMS
ZrO signal and the decay of Auger intensity
ratio of the Osos peak during the process of
sputtering a Zr sample that was initially
exposed to 60 L Oz.............................
51
The variation of the Os os/ZRsz Auger signal
intensities among different experiments.......
53
The intensity ratio of the Osos peak from a Zr
sample at different oxygen exposures at room
temperature (linear scale).....................
56
The energy shift of the ZRsz peak from a Zr
sample at different oxygen exposures at room
temperature (linear scale) ......................
62
xii
ABSTRACT
coi^ium dioxide
is
one
of
the most active
isosynthesis catalysts. It has also found application as a
construction material for
nuclear
reactors and high
temperature devices.
To develop an improved understanding of the surface
oxidation of zirconium, the
adsorption of oxygen on
polycrystalline zirconium under ultra high vacuum has been
studied using Auger
Electron
Spectroscopy (AES) and
Secondary Ion Mass Spectroscopy (SIMS). AES measurements
have been conducted for zirconium samples with different
extents of oxygen exposure over temperatures between 300 K
to 673 K. Auger spectra were also recorded for Zr surfaces
with the same extents of oxygen exposure but accomplished
under different oxygen pressures and exposure times. The
process of sputtering an initially 60 L Oz-exposed Zr
sample was studied employing SIMS.
AES spectra were
recorded at sequential interruptions during sputtering.
Oxygen adsorption seems
to
follow the reaction
sequence of
first
chemisorption,
next
rapid oxide
nucleation, and finally slow oxide thickening. Diffusion
of oxygen onto the bulk of zirconium is negligible at
temperatures below 470 K
and becomes significant at
temperatures between 470 K and 673 K. Oxygen uptake was
found to primarily depend on Oz exposure level and slightly
on Oz exposure pressure (or exposure time). Secondary ion
yield of the ZrO
SIMS signal is closely related to the
oxidation state of the Zr surface.
I
INTRODUCTION
Heterogeneous
catalysis
plays
a
role
of
vital
importance in a wealth of chemical manufacturing processes.
Nevertheless,
there
still
uncoordinated, information
understanding the
exists
relating
mechanisms
quite
to
a
bit
catalysis.
of
Hencer
of heterogeneousIy catalyzed
reactions and how catalytic
activity relates to properties
the catalyst surface has
always been of primary concern
to scientific
investigators
and
practicing technologists
who want to develop and improve catalysts.
decade,
the
advent
of
techniques has provided
execute microscopic
reactions on
systems.
the
The
new
surface—sensitive
improved
application
analysis techniques to the studies
catalysts and
reactions
catalyzed
enhanced
knowledge
of
the
of
the
surface, layers
of
the
analysis
tools for researchers to
investigations
uppermost
During the last
these
structure and
of catalytic
surface—sensitive
of the surface of solid
by
solids has greatly
processes
occurring
at
solid/gas interfaces.
Zirconium dioxide (ZrOz)
catalytic behavior
in
is
known
to exhibit active
dehydrogenation, hydrogenation and
hydrogen exchange reactions Cl].
Studies of the catalytic
2
chemistry of
zirconium
mainly on bulk
dioxide,
zirconium
dioxide
have been performed
on
foil.
layer
The
oxide
stoichiometry
C33
however,
surface
and
has
to
C23.
No major studies
oxide formed on zirconium
been
have
adsorption and exchange of
have proceeded
shown
an
hydrogen
to
activity
have ZrOz
for
the
and oxygen similar to
that of bulk ZrOz E43.
The objective of the
improved
understanding
zirconium
utilizing
present
of
two
the
widely
analysis techniques, Auger
temperature
temperature.
and
at
surface
used
(SIMS).
levels
several
SIMS and AES
chemical state and
is to develop an
oxidation
of
surface-sensitive
Electron Spectroscopy (AES) and
Secondary Ion Mass Spectroscopy
different oxygen exposure
study
surface
Auger spectra at
are investigated at room
temperatures
above
room
are utilized to investigate the
structure
of the oxidized Zr.
This spectral information is
used
to identify the effects
of temperature and extent of
oxygen exposure on the nature
°f the surface oxide formed. ■ They also provide a basis for
identifying the oxidation state of
Zr sample.
Zr on an oxygen exposed
3
BACKGROUND
Zirconium
Bulk zirconium
(atomic
number
40)
metal similar to steel in appearance.
the titanium subgroup IV
of
zirconium consists of five
is a silver-grey
It is an element of
the periodic system.
Natural
stable isotopes, Zr^® (51.46%),
(11-23 %), Zr92 (17.11%), Zr94 (17.4%) and Zr96 (2.8%)
and it
ranks
twelfth
abundance.
The
modifications.
among
pure
The low
hexagonal-close-packed
temperature
structure.
C43.
the
metal
elements in terrestrial
exists
in
two
allotropic
temperature <x—zirconium exhibits a
(h.c.p.)
p-zirconium
has
structure
and
the
high
body-centered-cubic (b.c.c.)
The a-P transition temperature is 1135 K (+5 K)
The
electronic
configuration
ls22s2p63s2p6d104s2p6d25s2
levels of the electrons
in
(or
of
CKr34d25s2).
zirconium
zirconium
is
The energy
are shown in (a) of
Figure I with X-ray designation of each energy level C53.
Zirconium
compounds (e.g.
reacts
HzO,
surface oxide film.
surface of bulk
readily
CO,
This
zirconium
COz,
thin
from
with
...)
oxide
to
oxygen-containing
form a cohesive
film protects the
further oxidation at room
4
INCIDENT
ENERGY
\ Ep>>
5 AugerCMvN23N23
e Mv
VAC
0 eV \
F
n 45+2
N 00
. ?
29
52
.
.\
\
\
VT
180
183
331
mIII
Mn
Jjl.
2223
Lu 1227.
2532
K
17998
-- Ground State
(a)
Figure I.
Final State
(c)
The energy states of electrons in
zirconium metal and the schematic
diagram of Zr(MvNzaNza) Auger
transition.
5
temperature C43.
Zirconium and its
absorbing
and/or
alloys
show
adsorbing
temperatures, even at
very
a
most
low
great affinity for
gases
pressures.
at
elevated
This property
has led to the use of zirconium as a so called "bulk getter
material" in
high-vacuum
technology.
zirconium resembles titanium [63
In
this respect,
which has been thoroughly
studied.
Since zirconium
capture
cross
and
its
sections,
alloys
have
outstanding
small neutron
anti—corrosion
and
mechanical properties, they are widely used as construction
materials for nuclear reactors.
Thus, there have been many
studies of the oxidation of zirconium and its alloys.
Most
of these studies have been concerned with thick oxide films
formed at high
observed
temperatures
that
temperature
is
the
a
[73.
oxidation
process
of
It has generally been
of
zirconium
simultaneous
at
bulk
high
oxygen
absorption and oxide film formation [83.
At low concentrations,
occupying
lattice
oxygen dissolves in zirconium,
interstices
instead
of
forming
stoichiometric oxide compounds.
In the h.c.p. Zr lattice,
the dissolved atoms
two types of interstitial
may
occupy
sites: those which have octahedral symmetry and those which
have
tetrahedral
symmetry.
Between
the
two
possible
interstitial sites, the dissolved oxygen atoms in zirconium
lattices are
primarily
located
in
the larger octahedral
6
sites.
The
occupation
can
be
octahedral sites in an h.c.p.
concentration of 29 atomic
up
to
about
60% of the
zirconium lattice or up to a
percent oxygen, as indicated by
the X—ray diffraction pattern
on a zirconium—oxygen system
C93.
Experimental results
suggest
that
the absorption of
oxygen in bulk zirconium primarily follows a grain-boundary
diffusion at temperatures up to -between 770 and 970 K and a
bulk, or lattice,
diffusion
diffusion is of
minor
below about 770
K
at
higher temperatures.
importance
but
becomes
The
when the temperature is
increasingly important at
higher temperatures C93.
The oxygen diffusion coefficient
in Zr is approximated
D = 0.0661 exp(-44,000/RT) cm2/s
by
for the temperatures in the range of
2
= 16.5 exp(-54,700/RT)
cm /s
range of 920 K to 1770
K
Ritchie et
al.
C103
for
Pemsler
the temperatures in the
(R in cal/gmolK), as reported by
following
the
number of experimental results.
been shown by
560 K to 920 K, and D
C113
analysis
of a large
The oxygen diffusivity has
to
vary
by
a factor of two
depending upon grain orientation.
The rate of diffusion
of
oxygen in zirconium dioxide
is significantly lower than that in the metal [123.
I
Autyer Electron Spectroscopy and
Aucrer Transitions of Zirconium
Auger Electron Spectroscopy
Auger electron
spectroscopy
(AES)
surface-sensitive analytical technique
process.
An
Auger
process,
or
is
a widely used
based
Auger
on the Auger
transition, is an
adjustment to an inner-shell-vacancy
formed
external energetic excitation, which
takes place by having
one electron from
a
less
hole, while a second
electron
into the continuum with
in total energies of
an
the
Auger process (see (b)
&
analysis, the energetic
by an impinging
tightly
bound orbital fill the
(Auger electron) is ejected
energy equal to the difference
initial
and final states of the
(c)
Figure
of
excitation
electron
in an atom by
beam
I)
In an AES
is generally initiated
and
sometimes by a photon
beam.
An Auger transition is
X-ray designated energy
generally
levels
the initial vacancy is in
the
drops to
in
fill
the
hole
as
denoted in terms of
an XYZ transition when
X shell, a Y shell electron
the
X
shell
and the Auger
electron is expelled from the Z shell. This is exemplified
by the MvNzaNz3 Auger transition displayed in Figure I ((b)
& (c)).
Auger transitions
same shells
are
sometimes
that involve electrons from the
not
differentiated
from each
8
other.
In
other
words,
the
omitted in the designation.
levels
in
each shell are
As in the case of MvNz3N23, it
is also denoted as MNN by some researchers.
The kinetic
energy
of
an
Auger electron originated
from an XYZ transition can be estimated by [13]
E(Auger) .= E(X) - E(Y) - E(Z)*
where
E(Auger) : the kinetic energy of the ejected Auger
electron from the Z shell.
E(X)
: the binding energy of an electron in
the initial hole.
E(Y)
: the binding energy of the Y electron
that drops in the X hole.
E(Z)
: the binding energy of the final state.
(i.e. the binding energy of the Z level
in the presence of. a hole in the Y
level).
As shown in Figure 2,
transitions possible.
shell and V stands
there
In this
for
a
CCV
transitions that
involve
Figure, C stands for a core
valence
process is an Auger transition
level electrons.
are three types of Auger
and
one
shell.
Thus, the CCC
that involves only the core
CW
or
processes
two
are the Auger
valence electrons,
respectively, in the Auger processes.
In AES, the
described by an
signal
intensity
of
an
element can be
explicit
function
of
the Auger electron
9
CCV
CCC
—
9—
A
CVV
A
E vac
(Vacuum Level)
(Fermi Level)
♦
E0
----
(Bottom of
Valance B a n d )
■
I
■
4I
I
I
X
Figure 2.
The three possible types of Auger transitions.
C
: core level electrons.
V
: valence level electrons.
CCC : Auger transition involving only core
level electrons.
CCV : Auger transition involving one valence
electron.
C W : Auger transition involving two valence
electrons.
10
escape depth, concentration distribution
a function
of
depth
and
of the element as
electron backscattering factors
[14].
The utility and popularity of AES is based on its high
sensitivity (1% to 5% of
a monolayer can be detected), its
excellent, elemental sensitivity for
He,
the
availability
transmission and the
of
all atoms except H and
energy
good
lateral
analyzers
with
high
resolution afforded by
electron beam excitation.
Auger Transitions of Zirconium
The Auger energies of
the principal Auger transitions
of Zr lie in the ranges of 80 to 180 eV and 1520 to 1950 eV
CIS].
The Auger transitions
80 to 180 eV are
energies.
more
These
of
pronounced than those of the higher
Auger
range are therefore widely
transitions
zirconium in the range of
oxygen
[5]
at
80
zirconium
designated.
is
Auger
shown
to
are
spectrum
in
in the studies of
The Auger transitions of
180 eV and the K W Auger
associated
energy levels, as illustrated in Table
differentiated
the lower energy
investigated
surface oxidation of zirconium.
transition of
the lower energy range of
Figure
I.
of
a
3
with
with the Auger
An example of a
slightly
the
oxidized
Auger peaks
11
Table I. Auger Transitions from Oxygen Exposed Zirconium.
|Ele- I
Auger (#)|
Iment I Transitions j
Symbolic
Name
|
Auger Energy (?)
Clean Zirconium
I
MNN & MMN
I
ZRgz
|
I
MNN & MMN
I
ZRii?
I
120, 116
I Zr IMMN, MMV & MNVI
ZRize
I
126
I
95,
92
I
MNV & MMV
I
ZRl48
I
145, 147
I
MW
j
ZRl75
j
175
o I
KW
I
Osog
I
509.5 (*)
(#) Each of these peaks listed in this table is actually
a combined result of several Auger transitions. For
instance, the primary sources that form ZRaz include
MsMvNi, MvNiNz3, MzMtNi, MtNiNzs and MzMvNi C51.
These transitions are therefore symbolized as MNN &
MMN in this table.
(?) The Auger peak energy in dCN(E)*E3/dE is identified
in this Table by the maximum negative excursion. The
Auger energies of the ZRizs and the Osoa peaks are
taken from the result of this study. The rest of the
Auger energies listed in this Table are taken from
Axelsson et al. Cl?].
(*) measured from a Zr sample with 100 L Oz exposure in
this research.
Secondary Ion Mass Spectroscopy
Secondary
extensively used
analysis.
Ion
as
Mass
a
Spectroscopy
tool
for
(SIMS)
surface
has
been
and bulk solid
The SIMS method possesses the following features
C163:
(I) detection of surface compounds by their fragment
ions.
12
SURVEY
DBW S l 1CN<E)»E3
BES
KINETIC ENERGY, EV
3.
An example of a differentiated. Auger spectrum,
taken from slightly oxidized zirconium.
13
(2) detection of hydrogen and its compounds.
(3) isotope identification with high mass resolution.
(4) significant differences in sensitivity of up to three
orders of magnitude for different surface structures
(e.g. elements, compounds).
(5) capability for quantitative analysis.
(6) information in the monolayer range with detection
limits of less than 10 ^ monolayers.
(7) very small surface disturbance; especially with the
development of static SIMS (SSMIS), using low energy
primary ions, which allows the disturbance of the
specimen surface to be minimized.
This further
enhances its application in surface reaction studies
and monolayer analysis.
In SIMS,
primary
the
ions,
sample
usually
several hundred to
Target
impacts
particles
induced
is
argon
several
are
by
bombarded
ions,
thousand
sputtered
the
primary
due
with
having
a
beam of
energies
of
electron volts (eV)
to
the
ions.
The
cascade of
sputtered
components are neutral atoms and positive or negative ions,
either in their ground or
excited states.
depth of these secondary ions
small number
of
secondary
is
ions
The mean escape
about 6 A.
A relatively
originate
from depth in
excess of 20 A C183.
In the
SIMS
analysis,
the
secondary ions extracted
14
from the target region pass
are separated
according
to a mass analyzer, where they
to
their
mass-to-charge ratios.
The secondary-ion current thus
measured in SIMS depends on
the energy and current density
of the applied primary ion,
the
secondary
ion
yield
concentration of the
per
surface
primary
species,
ion
impact,
the
and the instrument
sensitivity C19ZL
The
mass-to-charge
separated
ions
are
detected by
suitable means and the data are supplied to a recorder or a
computer.
The
data
reflect
chemical bonding in a
specimen
the form and the lateral
information
concerning the
and provide information on
distribution of compounds in thin
surface films.
Oxidation of Zirconium
Studies by Non-surface-spectroscopic Methods
There have
been
a
large
oxidation of zirconium; however,
spectroscopic methods.
from
of
studies of the
very few utilize surface-
This section summarizes the results
non-surface-spectroscopic
surface-spectroscopic
number
methods
methods.
are
Results
reviewed
in
from
the next
section.
The
oxidation
of
absorption of oxygen into
zirconium
the
zirconium dioxide (ZrOz) surface
results
in
both
the
bulk metal and formation of
films.
Due to the very
15
large solubility of oxygen
oxygen in zirconium
and
in zirconium, both diffusion of
in
the zirconium dioxide surface
film are important in the oxidation process [20].
Th-6 initial stage of oxidation
slow
reaction
process,
called
which a thin cohesive
oxide
fil® adheres
to
tightly
which results in
zirconium.
the
oxygen
oxidation is
oxidation,
surface
oxide,
This thin oxide
at room temperature,
during
which the metal
occurs under extended high-
exposure.
associated
oxidation, in
anti—corrosion properties of
oxidizes at accelerated rates,
temperature
forms.
metal
strong
Breakaway
protective
film
the
of bulk zirconium is a
The
with
the
particularly
onset
of
formation
at
edges
and
breakaway
of
a white
corners
of
oxidized specimens [93.
In
the
studies
methods, the kinetics of
using
non—surface—spectroscopic—
oxidation are usually measured by
the mass gain or the thickness of the oxide layer formed as
a function of time.
The
generally
shown
has
been
logarithm of time at
Torr up
parabolic
In the
to
760
in
to
temperatures
higher temperatures, the
pressure.
extent of oxidation of zirconium
kinetics
be
proportional
to
below "670 ""KT203.
depend
the
At
upon the oxygen
"normal pressure" range, approximately I
Torr,
time,
are
two
rate
expressions,
established
pretreatment of specimens [203.
depending
cubic and
on
the
Quite a few varieties of
expressions in the time exponent and activation energy have
16
been reported.
At
intermediate
pressures
Torrr the oxidation has
been
four stages.
represents
Stage
I
observed
rate of weight gain with time.
linear at different rates,
approximately a
factor
10~^ to I
to be divided into
a generally decreasing
Both stage II and III are
but
of
of
in
two
stage
higher
III the rate is
than
in stage I.
stage IV, the rate decreases with time.
low pressures of IO-4
to
IO-6
that the oxidation kinetics
E"ate expression in time
Torr, it has been reported
can
during
parabolic rate expression in
At very
be
described by a linear
initial
exposure and by a
time during extended exposure
[203.
One obvious
feature
zirconium oxidation is
of
that
earlier
a
observations of the
wide discrepancy exists in
the observed rate equations of oxidation among the numerous
studies, although most investigators
the mechanism.
Conflicting
exponents in the rate
[20] .
results
expressions
and
a range of time
have both been reported
Transitions from one rate expression to another have
also been observed in many studies.
rate
agree more or less on
is
markedly
dependent
Moreover the oxidation
upon
the
difference
in
orientation between different grains as reported by Pemsler
[21] .
A wide variety
purity has been
used
of zirconium of different degrees of
for
previous
studies.
Studies by
Kofstad [22] show that impurities have a large effect on Zr
oxidation.
17
Studies by Surface-spectroscopic Methods
Not many
studies
using surface-spectroscopic methods
have been reported for
and most of these
the
have
surface oxidation of zirconium
been
published during the 1980's.
The surface-spectroscopic methods
that
have been utilized
in these studies include AESr XPSr SIMSr LEEDr UPS and work
function.
AES and XPS
were the major instrumental methods
used by most researchers in previous investigations.
Foord
et
al.
E233
absorption of diatomic
between 300 K and 614
gases
entirely
extent
of
adsorption
and
on zirconium at temperatures
spectra.
dissociative
about
the
K by AESr work function measurements
and thermal desorption
that
studied
one
For
Ozr they concluded
adsorption
monolayer
of
occurred
materialr
to
the
with
no
detectable loss by desorption and little diffusion into the
bulk at 300
K.
Saturation
takes place at about 30
that the degree of
increased as the
increased and
number
of
L
of
Oz
exposure. They also showed
attenuation
kinetic
attributed
valence
the chemisorption regime
of
various Zr AES signals
energies
of
this
the
to
electrons
Zr Auger electrons
difference in the
involved
in
the
Auger
transitions.
Tapping C23 studied the
and polycrystalline Zr by XPS
initial oxidation of Zr(OOOl)
and
UPS.
By monitoring the
XPS oxygen(Is) signal as a .function of oxygen exposure, he
IS
concluded
that
oxygen
uptake
saturation at approximately
50
on
to
zirconium
70
L
reached
Oz exposure.
A
reproducible kink at 3.5 L
Oz exposure was observed in the
correlation of the
of
ratios
(O(Is) area
over
exposure.
Tapping
transition
from
nucleation
stage.
estimated the
ZR(3d)
area)
suggested
a
the
XPS signal intensities
to
the
the
extent of oxygen
kink
represented the
chemisorption
region
Based
JXPS
thickness
of
on
the
oxide
film
to
O(Ts)
on
an
oxide
data,
he
zirconium at
saturation coverage as 3 to 4 atomic layers.
Yashonath et al.
ratios
in
C243
transition
studied
metals.
metal Auger intensity
For
the
oxidation
of
zirconium, they claimed that the surface oxide corresponded
to ZrOz above 30 L Oz
exposure.
of zirconium were suggested.
Below 30 L Oz, suboxides
They have estimated the oxide
layers on Zirconium to be about 3 monolayers at very low Oz
exposure and around 6 monolayers for large oxygen exposures
4
O
10
L).
Other investigations have reported that
significantly higher Oz
exposures
are required to achieve
similar thickness of surface oxide on Zr C2, 231.
Sen
et
al.
C251
oxidation of zirconium
XPS and
AES.
They
at
found
chemical shift occurred
pure Zr and Zr after
have
in
investigated
the
surface
room temperature employing both
that
the
an
XPS
approximately
4 eV
Zr(3d) peaks between
high oxygen exposures (around IO6 L).
They suggested that the
change
of
0(2p) band position at
19
different levels of oxygen
the
formation
of
exposures
different
was an indication of
oxide
species.
At oxygen
exposures below 10 L dissociative chemisorption of Oz on Zr
metal occurred,
suboxides.
which
led
to
The chemical shift
of
sub-oxides increased
with
this range.
exposures
At
Oz
the
formation
of various
the XPS energies of the
increasing
Oz
between
exposure within
10
to
25 L , the
suboxides were suggested to convert into 'ZrO', the highest
possible sub-oxide.
With further
the 'ZrO' was converted
to occur via cation
into
ZrOz and oxidation continued
transport
claimed that the presence of
exposures of Oz, some of
to
the surface.
They also
'ZrO' was essentially only at
the surface.
Valyukhov
et
al.
C26,
273
studied
oxidation of polycrystalline zirconium
using XPS and
SIMS.
They
oxygen was accompanied by
proposed to
occur
via
573
initial
K to 873 K
claimed that chemisorption of
the formation of crystallization
centers and growth of ZrOz islands.
one monolayer of ZrOz,
at
the
further
the
After the formation of
oxidation
diffusion
in the bulk was
of adsorbed oxygen
through the ZrOz lattice to the metal-oxide interface.
Hoflund
et
al.
C28,
interaction of polycrystalline
and NzO by XPS and AES.
distinct
have
zirconium
investigated
with
the
Oz, Nz, CO
They found in general surprisingly
low room temperature sticking
observed two
293
states
coefficients (< 0.01).
of
zirconium
They
samples with
20
ererr^ reactivities
whether or not
the
and
Auger
samples
had
spectra, depending upon
been
annealed above the
h-P'C.-b.c.c. phase transition temperature of 1135 K.
further noted that the
They
reduction in chemisorption activity
due to heating above the phase transition temperature could
be restored by argon-ion sputtering.
Krishnan et al. C303
zirconium between 773 K
proposed that the
zirconium
layer.
and
initial
surface
process followed
studied the surface oxidation of
1008
stage
proceeded
by
of
first
nucIeation
Together with the
K
and
results
they claimed that the rate
employing AES.
They
oxidation on a clean
as
a
chemisorption
growth
of
an oxide
of Ritchie et al. [10],
of diffusion of oxygen into the
bulk zirconium was significant and determined the extent of
the
chemisorption
Within
the
regime
at
temperature
chemisorption
regime
temperatures
range
was
found
to
above
investigated,
extend
over
773 K.
the
longer
exposure periods at higher
temperatures.
A surface oxide
was rapidly formed at
temperatures.
The oxide layer
formed was estimated
low
to
be
approximately
I
nm thick at
oxygen exposure around 8 x IO4 L between 773 K and 1008 K.
Axelsson et al.
EELS spectra of clean
They identified the
for these surfaces.
the oxygen
[17]
and
fine
They
chemisorption
have
investigated
oxidized
structures
noted
regime
Zr
the AES and
and of bulk ZrOz.
of the Auger spectra
that the transition from
to
the oxidation regime
21
occurred at around 20 L
Oz..
required to
saturate
completely
oxygen, though
the
Very large oxygen doses were
outermost
the
surface region with
atomic
layers very quickly
became oxide-like.
They observed that the amount of oxygen
uptake depended on
both
pressure.
They further
rate or the transport
rather than the
oxygen
exposure level and oxygen
indicated that the Oz dissociation
of
transport
0
of
atoms
Zr
through the oxide film
atoms through the oxide
film was rate limiting in the surface oxidation process.
22
EXPERIMENTAL SYSTEM AND PROCEDURE
The experiments of this
an ultrahigh-vacuum
research
system.
Auger Microprobe (PHI
595
Physical Electronics Scanning
MULTIPROBE),
Research in Surface Science
at Montana
State
HO
Is-1
University.
turbopump,
sublimation pump.
595
MULTIPROBE
The
analyzer
(CMA).
a
vacuum
system has a
IO-10 Torr and. can be pumped,
a
200
The electron
is
in the Center for
and. Submicron Analysis (CRISS)
normal vacuum pressure of 2 x
by a
were carried, out in
Is”1
pump
and. a
energy analyzer in the PHI
single-stage,
Data
ion
cylindrical
collection,
mirror
reduction
and
interpretation in PHI 595 MULTIPROBE are handled by the PHI
Multiple-technique
Analytical
Computer
System
(MACS), a
utility program developed by Physical Electronics Inc. with
the use of a DEC PDP 11/04 computer.
PHI
595
MULTIPROBE
allows
"survey" and "multiplexing",
for
the CMA to
In
acquire
data.
two
different
scanning
the
modes,
pass energy of
"survey" scheme, the
energy scanning window is set over an electron energy range
large enough to encompass all
For instance, the energy
the Auger peaks of interest.
window
for
"survey" AES in this
study was set from 30 eV to 530 eV, which covered the major
23
transitions of zirconium (80 to
transition of oxygen (495 to
scheme, several
energy
180
515 eV).
windows
only on the Auger energies
eV) and. the K W Auger
of
are
In the multiplexing
established to focus
interest.
energy windows in multiplexing
AES
For example, the
in this study were set
to be 80 eV to 180 eV
for the major Zr transitions and 495
eV to 515 eV for
0
the
K W transition.
requires less data acquisition
of interest for one pass.
since additional
time
Multiplexing AES
time for the Auger energies
It usually has better resolution
is
spent
collecting
data
in the
electron energy regions of interest.
The zirconium foil specimen.of dimensions 10 mm x 5 mm
x 0.03 mm
was
first
etched
solution to remove most of
then solvent-cleaned in
in
a weak hydrofluoric acid
the accumulated oxide layer and
methanol.
inserted into the high-vacuum
beam was rastered over a
2
After
chamber,
mm
by
a
the sample was
3 KeV argon ion
2 mm area to clean the
specimen surface.
The AES measurements were
made
with
a 3 KeV, 167 nA
electron beam rastered over a 0.2 mm by 0.2 mm area central
to the cleaned
"survey"
area
Auger
experimental
of
the
spectra
series
to
were
then used
to
"survey"
collect
collected
establish
zirconium surface., Once the
clean from the
unannealed
prior
cleanness
to
The
each
of the
surface was established to be
spectrum,
data
the
Zr samples.
for
"multiplexing" AES was
the
Auger intensity vs.
24
oxygen exposure.
The exposure pressure of oxygen
system is controlled
by
the
rate
in the PHI 595 vacuum
of oxygen introduction
into the vacuum chamber through a variable leak valve.
variable leak valve
has
at leak rates'up to I
The
control sensitivity and stability
x
IO-10 Torr-Iiter per second.
extent of oxygen exposures
was
expressed
The
in terms of the
product of exposure time and exposure pressure in Langmuirs
(I Langmuir = I L
=
10 6
Torr-sec).
pressures employed in this
Torr to 7 x 10
seconds
to
Torr
1000
research
The oxygen exposure
varied
from 5 x IO-9
and the exposure time ranged from 40
seconds.
A
exposure was achieved either
by
certain
level
of oxygen
a one-time exposure after
cleaning the Zr surface or by applying sequential exposures
to an initially clean
Zr
up to the
time-pressure product (in Langmuirs)
indicated.
particular
For example,
surface with the exposures added
a
achieved by a one time 1000
a set of sequential
4.
In this set of
100
L
Oz exposure was either
sec x IO-7 Torr Oz exposure or
oxygen exposures illustrated in Figure
sequential oxygen exposures, a clean Zr
sample was first exposed to
a
0.3 L Oz exposure, and then
sequentially to 0.7 L, 2 L, 7 L, 20 L and 70 L Oz-exposures
cleaning the sample.
for a 100 L Oz exposed
Therefore, the Auger spectrum
surface was generated at the end of
this sequential exposure set.
metal and after totals of 0.3 L,
Auger spectra for the clean
I
L, 3 L, 10 L, and 30 L
25
of Oz exposure were also obtained in the sequential series
as indicated in Figure 4.
Experimental Procedure
Ar+ sputtering
Auger Spectrum Taken
===> AES <survey)
1==> AES (clean) <multiplexing)
2==> add 0.3 L oxygen
(100 sec x 0.3 E-3 Torr)
3==> AES(0.3 L) (multiplexing)
4==> add additional
0.7 L oxygen
(100 sec x 0.7 E-8 Torr)
5== > AES(1.0 L) (multiplexing)
6 ==>
add additional
2.0 L oxygen
(100 sec x 2.0 E-8 Torr)
7== > AES(3.0 L) (multiplexing)
8 == > add additional
7.0 L oxygen
(100 sec x 7.0 E-8 Torr)
9 == > AES(10. L) (multiplexing)
H
O
Il
"V
0==> clean zirconium
add additional
20. GI L oxygen
(100 sec x 2.0 E-7 Torr)
11 = > AES(30. L) (multiplexing)
12 = > add additional
70.0I L oxygen
(100 sec x 7.0 E-7 Torr)
13 = > AESdOO L) (multiplexing)
Figure 4.
The sequential oxygen exposures and AES
measurements of this study.
To establish the influence of Oz exposure pressure (or
exposure time) on oxygen and zirconium Auger spectra, a set
of one-time oxygen exposures
for
each of the three oxygen
26
exposure levels, 3
several different
L,
10
oxygen
exposure times).
These
oxygen exposures
in
L
100
pressures
L,
were run under
(thus
with different
experiments are denoted as "cyclic
this
t'hs repeated cycles of
and
study
in order to characterize
cleaning and one-time exposures for
a certain oxygen-exposure level.
^
361 K
experiments
to
673
resistively.
K,
The
were
during
made
which
temperature
at temperatures between
the
specimen was heated
measurement
was
made by a
W/W-26% Re thermocouple spot-welded to the back face of the
specimen and, supplementally, by an optical pyrometer.
The SIMS measurements
(L-H) SIMS 100 system that
were
made by a Leybold-Heraeus
included an L-H quadrupole mass
spectrometer, which was attached to the PHI 595 MULTIPROBE.
In the SIMS experiments, a 3 KeV primary argon ion beam was
employed.
The ion beam, with an
uA
2
density of 5 ^ / c m , was rastered over
approximate current
an area of I mm x I
mm.
The
SIMS
100
system
is
integrated secondary ions' scan
capable
over
of
performing
a large range of m/e
(mass—to—charge) ratios or monitoring secondary ions of one
particular m/e ratio vs. time during a sputtering process.
A SIMS scan over an m/e range of I through 200 amu was
initially measured
for
a
zirconium foils of different
Such
integrated
SIMS
clean
zirconium
foil
and for
extents of surface oxidation.
spectra
of
an
oxidized
and
> rr-
an
27
essentially cleaned zirconium foil
5.
The signals occurring between
secondary
ions
of
the
The signals
m/e 90 and 96 are due to
isotopes
appearing between 106 and
ZrO .
are displayed in Figure
112
between
of
are
180
Zr+ .
The signals
due to the isotopes of
and 192 are attributed to
the isotopes of Zrz+ .
can be seen in Figure 5 r the ratio of the intensity
of ZrO+ signalr at m/e
m/e of 90, is quite a
that
has
higher
oxidized zirconium
to
in
to that of Zr+ signal, at
bit larger for the zirconium surface
content.
degree
SIMS signal was used
clean
106,
oxygen
sensitivity to the
study, a
of
of
Because
monitor the process of sputtering
the
zirconium
the
present
specimen
sputtering
and
research.
sputtering
occurred each time the
SIMS
for
For this
surface was initially
SIMS
periodically interrupted to conduct
interruption of
this
surface oxidation, the ZrO+
exposed to 60 L Oz prior to the sputtering process.
the process,
of
During
measurements were
AES measurements.
these
The
Auger measurements
signal intensity at m/e (106)
roughly reached a 30% to 50% decay from its intensity prior
to the previous AES test.
This process continued until the
Auger spectrum
that
clean.
indicated
the
zirconium surface was
28
Positive Secondary Ion Intensity (Arbit. Unit)
Essentially
Cleaned
Zirconium
INTENSITY
Oxidized
Zirconium
INTENSITY X
mass/charge ratio (m/e)
Figure 5
The integrated SIMS spectra of essentially
clean Zr and surface oxidized Zr.
29
RESULTS
AES Results front Sequential Qxvaen Exposures
Typical Auger
spectra
for
clean
zirconium are shown in Figures 6 to 9.
"survey" Auger spectra from 30
both undifferentiated
mode
mode (Figure 7).
top
The
Auger spectral curves
specimen.
clean
to
curves
a
zirconium
specimen.
6) and differentiated
in
100
The bottom curves in
In Figures 6 and I r
530 eV are displayed in
(Figure
from
and oxygen exposed
Figures 6 and 7 are
L Oz exposed zirconium
Figures 6 and 7 are from a
Differentiated
Auger spectra after a sequential
multiplexing
set of Oz exposures on Zr
are presented in Figure 8 for zirconium transitions (80-180
eV) and Figure 9 for oxygen transitions (495-515 eV).
Differentiated and Undifferentiated AES
Auger peaks
become much more
are
evident
removes
the
the
N(E)
function but
pronounced by electronic differentiation,
as shown in Figures 6 and
makes weak
in
features
large
more
7.
The differentiation not only
readily
background
identifiable
consisting
but also
mainly
of
backscattered primary electrons and inelastically scattered
Auger electrons C313 (see Figures 6 and 7).
30
AES SURVZT
Zr(+IOOL O 2 J
Clean Zr
Emmc ZSESGT, ZV
Figure 6. Undifferentiated "survey" Auger spectra from
a clean Zr sample and a Zr sample after 100
L Oz exposure at room temperature.
AES SURVEY
Zr(+IOOL O2)
Clean Zr
E lu m C
ZNO C T . EV
Figure 7. Differentiated "survey" Auger spectra from
a clean Zr sample and a Zr sample after
100 L Oz exposure at room temperature.
31
OXYGEN
EXPOSURE
(L)
d (N (E) *E)/dE
10 0
12B
138
1«
CIKtJIC CHESCT, Vl
Figure 8. Multiplexing Auger spectra of Zr(80-180 eV)
from oxygen uptake on zirconium at different
oxygen exposures at room temperature.
32
O 5O9
OXYGEN
EXPOSURE
(L)
d(N (E) *E)/dE
10 0
30
10
3
I
0.3
clean
CIKETIC EKERCT. EY
Figure 9. Multiplexing Auger spectra of Oxygen(495-515 eV)
from oxygen uptake on zirconium at different
oxygen exposures at room temperature.
33
In
the
differentiated
positions are
maximum
conventionally
negative
Auger spectrum
does
spectrum,
identified
excursion
conventional peak energy
energy in the
Auger
of
by
the
the
the point of
peaks. .
Such
a
designation in the differentiated
not
correspond
undifferentiated
to
the maximum peak
spectrum.
and
in peak energy is usually
not considered significant.
use of differentiated Auger
slope.
The difference
depends on the peak width
in applied AES.
peak
But this difference
The
energies-Is-normally-practiced
In accordance with conventional practice,
the present study is
primarily based on the differentiated
Auger spectra.
Normalized Peak Height Ratios
Auger yield, the factor
peak
amplitude,
is
a
relating surface abundance to
complex
precisely evaluated from
first
changes in the amount of a
quantity
that
principles C323.
can't
be
However,
component on the surface can be
followed by measuring the decay
or growth of an associated
Auger peak height C303.
As indicated in Figure
Zr
does
not
monotonically.
decrease
the
This
intensities is due to
current,
the exposure of oxygen to
intensity
fluctuation
the
experimental factors such
beam
10,
electron
of
in
the
ZRsz peak
Auger
signals'
statistical random variation of
as
changes in incident electron
multiplier
gain
and
other
34
ZR92 PEAK AMPLITUDE (ARBIT. UNIT)
§
«1
O
m
O
O
O
O
O
O
O
Q
O
cn"
O
clean
-I OO
-O 50
0 00
0 50
1.00
I 50
—I
2. QO
LOG ( OXYGEN EXPOSURE (LANGMUIR) )
Figure 10. The intensity of the ZRsz peak at different
oxygen exposures at room temperature.
35
instrumental
factors
variations in
C303.
experimental
peaks are followed in
peak heights taken
this
from
To
compensate
factors,
study
the
Zr
for
such
and 0 Auger
by normalized ratios of
dCN(E)*E3/dE
Auger spectra.
The
normalized peak height ratios are obtained through dividing
the height of the peaks
the ZRg2 peak in the
of
same
interest by the peak height of
spectrum.
The ZRg2 peak height
is used as a normalizing factor because it is a core level,
ccc,
transition.
oxidation only
For
results
a
in
change in line shape since
relatively unaffected by
core
an
level
energy
Auger transition,
shift
with little
the core level wave function is
changes
The core electrons respond
to
in chemical environment.
changes in chemistry as the
result of perturbations in valence shell energies [333.
The normalized peak height
the sequential oxygen exposure
11 through
16
as
a
ratios
tests
function
of
of Auger peaks for
are shown in Figures
the
extent
of oxygen
exposure.
Energy Shifts
The Auger peaks of zirconium between 80 and 180 eV and
the Auger peak of oxygen
between
495
and 515 eV from the
surface oxidized zirconium show energy shifts towards lower
kinetic energies relative to those from clean zirconium, as
indicated by
the
peaks'
positions
transitions and in Figure 9
in
Figure
8
for oxygen transitions.
for Zr
These
36
CM
ZR I26/ZR92 (PEAK HEIGHT RATIO)
o
A
O
CO
O
CO
e
o'
CO
m
o
«
#
CM
O
CM
O
-I
00
-0.50
0.00
0.50
1.00
1.50
2 . 00
LOG ( OXYGEN EXPOSURE (LANGMUIR) )
Figure 11. The intensity ratio of the ZRizs peak from
a Zr sample at different oxygen exposures
at room temperature.
O
CN
*
ZR II?/ZR92 (PEAK HEIGHT RATIO)
O
O
*
*
O
OO
O
X
X
O
CO
O
X
O
O
O
CN
O
O
clean
-I.00
-0 50
0 00
0 50
I.00
I. 50
LOG ( OXYGEN EXPOSURE (LANGMUIR) )
2 . 00
Figure 12. The intensity ratio of the ZRi 17 peak from
a Zr sample at different oxygen exposures
at room temperature.
38
O
m
ZR I42/ZR92 (PEAK HE IGHT RATIO)
m
CM
O
O
O
O
r
^
.
O
O
-n
o
o
i
n
CN
o
o
clean
-I.QO
-0 50
0. 0 0
0.50
1.00
1.50
1
2 00
LOG ( OXYGEN EXPOSURE (LANGMUIR) )
Figure 13. The intensity ratio of the ZRi4z peak from
a Zr sample at different oxygen exposures
at room temperature.
39
O
O
A
A
o
ZR 148/ZR92 (PEAK HEIGHT RATIO)
m
CM
O
O
A
CM
O
m
A
O
O
A
O
m
o
A
O
O
clean
-I
00
-0 50
0.00
0.50
I. 00
1.50
2 QO
LOG ( OXYGEN EXPOSURE (LANGMUIR) )
Figure 14. The intensity ratio of the ZRiua peak from
a Zr sample at different oxygen exposures
at room temperature.
40
ZR175/ZR92 (PEAK HEIGHT RATIO)
0
01
m
o
O
-O
O
in
Q
O
•n
o
O
clean
-i•oo
-b.50
o.oo
o so
1.00
1 .so
LOG ( OXYGEN EXPOSURE (LANGMUIR) )
2
oo
Figure 15. The intensity ratio of the ZRi73 peak from
a Zr sample at different oxygen exposures
at room temperature.
41
O
O
O
O
H
0509/ZR92 (PEAK HEIGHT RATIO)
O-
a
B
o
O
30
O
O
•a
a
o
O
O
O
O
O
o"
clean
I’
r
— 1.00
—0 50
0.00
0.50
1.00
I 50
2 . 00
LOG ( OXYGEN EXPOSURE (LANGMUIR) )
Figure 16. The intensity ratio of the Osog peak from
a Zr sample at different oxygen exposures
at room temperature.
energy shifts are quantified in
average of the shifts of
Auger peaks from
comparison of
maxima
and of the minima of
the
first
derivative
Auger spectra.
energy
shift
determined
in
those determined
Appendix A.
the
this study in terms of the
from
N(E)*E
peaks'
A
this way with
maxima
is given in
The comparison
shows that the average of peak
maximum and peak minimum in
dCN(E)*E3/dE spectra is a good
approximation of the
maximum
energy
of the corresponding
peak in N(E)*E spectra.
For the ZRi4a
(ccv)
transition involving one valence
electron (see Figure
8),
the
zirconium Auger peak
to
occur
which grows
during
oxygen
energy
at
shift
causes a new
a lower kinetic energy
exposure.
Hence, the energy
shift of this peak, involving a valence electron is measured
as the difference
energy-shift
of
peak
the
and
energetic positions between the
the
spectrum from which the
parental
energy-shift
peak
peaks
in
the
form.
same
Only a
change of peak position with little change in line shape is
observed for the core level ccc transitions, ZRs2 and ZRi17
(see Figure 8).
Therefore the energy shifts for core level
transitions are measured
positions of a
clean
as
the
zirconium
differences
and
of the peak
those of a zirconium
after oxygen exposure.
The energy shifts of the
a function
of
oxygen
Figure 17 for the
ZRsz
Auger peaks are displayed as
exposure
peak
and
in
logarithmic
scale in
Figure 18 for the ZRi17
ZR92
.0 00
ENERGY
SHIFT (eV)
J 00
i 00
2 Uti
5 00
6 00
5
fl>
H
S?9it
ft
n
U
0
gs
Q '0 it
nn
ft
1
I n Hl
P
rtUl
It Er
»1 P. HP H- H»
Hi rt
H|
It
O p.
w
ft O
4
UJ Ifl
ro N
3 rO
<t it
P
a
o
?r
Hl
n
n
it 0
u
a 70
ZR92
00.70
07.70
KINETIC
OU 70
I
ENERGY (eV)
04. 70
ZR 1 17
I.PO
H-
SHI FT (eV)
ENERGY
I OO
OO
3 OO
-u
5 . OO
-U
6 Ol
— I
*1
fD
H
CO
PJ Ntj
rt M Cf
rt
I to
O FU rt
a ort
H 'I
an
1O rt io
ro Pr
p, H-
.
IU H1Hi
ft
C
hi rr
h,
n ro o
fl n Hi
O
r
O
0 A
O
X
-< n
Sg
m
x
TJ
Oo
go
m
r_
U) H
5
s
E
%
hi
n
a
2"G
1 15 40
ZR I 17
1 14 40
I 13.40
KINETIC
1 12 40
I I I . 40
E N E R G Y (eV)
I 10.
45
peak.
The energy shifts
not presented here.
The
shifts of the ZRi^a and
in the discussion of
of
the remaining Auger peaks are
reason
the
Osos peaks is presented later
results.
ZRi2e and the ZRiys peaks
not to include the energy
The
are
energy shifts of the
less than the resolution of
the experimental system.
AES Results from Cyclic Oxvaen Exposures
The product of exposure
Langmuir (10 6 Torr.sec)
is
parameter in the study of
The dependence of Oz
on peak
amplitude
pressure and exposure time in
usually
gas adsorption on metal surface.
pressure
attenuation
examined by varying the
used as the exposure
at specific exposure levels
using
this
parameter was
exposure pressures for experiments
of the same Langmuir exposures.
using the peak height ratio
The results are monitored
of the Osog peak and displayed
as a function of exposure time in Figure 19 for 100 L, 10 L
and 3 L of Oz exposure.
Aucrer Spectra above Room Temperature
The Auger
obtained at
intensity
several
for
sequential oxygen exposures
temperatures,
measured
by
the peak
height ratios of the Osog peak, are presented as a function
of oxygen exposure in Figure 20.
46
100 L
100 L
10
L
CL ..
05 -r
100.00
ZOO.00
EXPOSURE
300.OO
* 00.OO
500. 00
1000.00
TIME (sec)
HIGH Oz <-------------------------------- > LOW
Oz
PRESSURE
PRESSURE
Figure 19. The pressure dependence on the Oaos peak
intensity ratio using Langmuirs as the
parameter for Oz exposure.
47
C2 - DATA AT 22SK
C - OATA AT 3S1K
14.on
A - OATA AT ATOK
0 5 0 9 / Z R 9 2 (PE AK H E I G H T RATIO)
u.ou
a .00
ip.tifl
12.on
o - DATA AT STZK
a
A
O
»
0.00
20.00
40.00
60.00
30.00
100.00
120.OO
OXYGEN EXPOSURE (LANGMUIR)
Figure 20
The intensity ratio of the Osog peak from
a Zr sample at different oxygen exposures
above room temperature Clinear scale).
X
48
Ion Sputtering- Monitored with SIMS and AES
The decay of SIMS m/e
of. the Auger Osos peak
106 (ZrO+ ) signal and the decay
during the course of Ar+ sputtering
of a Zr sample, which was
sputtering, are
Figures 21 and
displayed
22,
in
to
versus
the
respectively.
.either of the ZrO+ SIMS
■at any time
exposed
these
signal
two
60 L of Oz prior to
sputtering time in
The signal intensity,
or of the Osos AES signal,
Figures
is represented as the
-
fraction of their
corresponding
decay of the Auger signal
initial intensities.
The
intensity ratio of the Osos peak
(Osos/ZRsz) is plotted against the corresponding fractional
decay of the SIMS m/e = 106 signal in Figure 23 to show the
relative change of the
two different spectral intensities.
In Figure 23, the decay
AES or
SIMS
signal,
of signal intensity, either of the
at
any
fractional decrease in intensity
time,
at
the corresponding initial intensity.
is
expressed by the
that time compared to
|
o
H
S i m s In t e n s i t y (Zr o +)
O
O
O0
«
d
O I
to
d
5
•ta­
ut
O
M
O
O
-1O1OO
2.00
4.00
6.00
SIMS
Figure 21.
-T------ P--t o . 00
12.00
S P U T T E R I N G TIME (MIN.)
8. 00
14. 00
16. 00
T e . oo
The decay of ZrO+ SIMS signal intensity during sputtering
a Zr sample after 60 L Oa exposure.
Note;
The intensity of the ZrO+ SIMS signal at any time is
represented as the fraction of its initial intensity.
Q
M
0509/ZR92 PEAK RATIO
O
O >
O
«3
O
♦
O
to
O
O
O
♦
Ul
O
t
O
M
O
♦
♦
O
O
° 0 .1)0
♦
2! 00
4.00
6.00
SIMS
Figure 22.
6.00
SPUTTERING
10.00
12.00
14.00
16.00
I B . 00
TIME (MIN.)
The decay of Auger intensity ratio of the Osog peak during
sputtering a Zr sample after 60 L O 2 exposure.
Note:
The intensity of the Osog Auger peak at any time is
represented as the fraction of its initial intensity.
51
Q
O
H
a
$#*
O
CO
a
>
IS
<
U
LU
Q
Csl
o>
q:
m
\
CD
O
m O
O IN
Q
Q
O
IN
0. 00
0. 20
0. 4 0
0. 6 0
0. 30
I. oo
SIMS INTENSITY DECAY
Figure 23. The+relationship between the decay of SIMS
ZrO signal and the decay of Auger intensity
ratio of the Osos peak during the process of
sputtering a Zr sample that was initially
exposed to 60 L Oz.
Note:
Signal intensity decay of either AES or SIMS
signal is represented in this Figure as the
fractional decrease at any time during sputtering
compared to the corresponding initial intensities.
52
DISCUSSION OF RESULTS
The main Auger transitions
of 80 to 180 eV and
the range of
energies of
495
the
zirconium in the range
the K W Auger transitions of oxygen in
to
515
eV
observed
oxygen are close to
Table
are
Auger
clearly resolved.
peaks
of
The
zirconium and
the Auger energies reported previously
for these peaks as presented
indicated in
of
2
in
earlier
which
observed for a clean Zr sample
Oz exposure in this study
in Table I.
the
This is
Auger peak energies
and a Zr sample after 100 L
are compared with Auger energies
previously reported.
Reproducibility of experimental data
found to be acceptable
for
that two consecutive Auger
surface appear to
be
caused
by
almost
experimental
spectra are
system
factors
taken.
and
during
It is noticed
taken
exactly
in
absorption
ultrahigh-vacuum
analysis.
spectra
some very minor difference
be
the
in this study is
for the same Zr
the same except for
the fine features which may
of
background
possibly
the
Variations
by
short
in
gases
the
in
variation
time
the
in
period the
Osos/ZRsz Auger
signal intensities are observed for different tests of each
of several oxygen exposure levels as is shown in Figure 24.
53
§
CM-
§
a
0509/ZR92 (PEAK HEIGHT RATIO)
O -
a
I
o
O
a
a
OQ
a
o
o
a
CO
a
a
O
O
o
o
o
clean
Figure
-1.00
-0.50
0. 0 0
0.5 0
1.00
1.50
2 00
LOG ( OXYGEN EXPOSURE (LANGMUIR) )
24.
The variation of the O 5 0 9 / Z R 9 2 Auger signal
intensities among different experiments.
54
The oxygen exposure
Figure 24 are
conditions
listed
variation among the
in
for
Appendix
those data plotted in
B.
In
spite of the
experimental runs, similar development
of intensity ratios vs.
the
extent
of oxygen exposure is
observed.
Table 2. Dependence of Auger Energies of Zirconium and
Oxygen Transitions on Zirconium Surface Oxidation.
Ele- I Transition I
ment j
I clean
I
Auger
metal
Energy(Y)
oxidizedt £)
ZR92
I
ZRi i?
I 118.0
ZRiza
I 126.0
ZRiua
I 148.0
(145, 147)
139.5
ZRl 7 5
I 175.0
(175)
164.0
0 5 0 9
I
93.0
(95, 92)(*)
90.0
(120, 116)
114.5
I I
I
I
Zr I
126.0
I
I
I
I I
I
o I
509.5
(xP) The Auger peak energy in dCN(E)*E3/dE is identified
in this table by the maximum negative excursion.
(5) "oxidized" : Data that were taken in this study after
100 L exposure of Oz on a clean zirconium specimen.
(*) Auger energies in the parentheses are data previously
reported for these transitions, which were cited in
Axelsson et al. [173.
Influence of Surface Oxidation on Auger Spectra
As shown in Figures 8
involves one valence
in peak
amplitude
and 14, the zirconium peak that
electron,
but
changes
the ZRiua peak, attenuates
very
little
in its peak
55
energy as the
hand,
its
oxygen
coverage
energy-shift
amplitude. Figure
13,
increases.
peak,
and
ZRitz,
shifts
grows
towards
energies as the oxygen exposure increases.
in energy shift behavior is due
electrons that form
the
On the other
in
peak
lower kinetic
This difference
to the fact that the Auger
energy—shift
peak originate from
interatomic Auger transitions between oxygen and zirconium.
The
Auger
electrons
that
constitute
primarily originate from metallic
surface.
The
energy-shift
originates from electrons
is.,
Zr
atoms
that
peak
of
have
Zr
the
indication
amplitude
of,
metallic Zr and
of
ZRitz, however, mainly
non-metallic
reacted
the
ZRita
Zr atoms, that
with adsoirbed/absorbed
and
respectively,
the
oxidized
within
Zr,
peak
atoms on or near the
oxygen and have lost their metallic properties.
the peak
ZRita
the
surface
Therefore.,
ZRitz
is an
abundance
of
the depth sensed by
AES.
Except for the peak
the curves of
Auger
height
signal
oxygen exposure fall into
ratio
intensity
of the ZRiza peak,
vs.
the extent of
two classes of similar features.
For the Z R n v , the ZRita and the ZRivs peaks, the intensity
ratios decrease as the exposure of oxygen increases; on the
other hand, the
intensity .ratios
coverage increases for the
ZRitz
increase
and
as the oxygen
the Osos peaks (see
Figures 12 through 16).
As shown in Figure
25,
the
linear plot of intensity
56
S
o_
H
0509/ZR92 (PEAK HEIGHT RATIO)
O
O
H
o>
H
O
CN
r**»
o
m
H
O
to
IO
O
CO S
°o. oo
1
20 OO
OXYGEN
■
40.00
60 00
EXPOSURE
.
80.00
;
100 00
;
120 00
(LANGMUIR)
Figure 25. The intensity ratio of the Osos peak from
a Zr sample at different oxygen exposures
at room temperature (linear scale).
57
ratio
of
indicates
Osos/ZRaz
as
that
saturation
the
a
function
of
of
oxygen
the
Auger
approached between about 10 L and 30 L Oz.
30 L Oz exposure, a
a lower rate
for the
with
ZRi4z
(Figure
13)
still
100 L Oz
the
observed.
for
information
This depth is
the
Osoa
peaks.
This
signal intensity implies that
the oxidation process is
the
However, beyond
oxygen coverage is observed
and
continuing increase in Auger
within
signal is
continuing growth in peak amplitude at
increasing
exposure,
exposure
occurring,
surface
depth
of
at least up to
region of zirconium
the
approximately
atomic layers) for the Osoa peak
and
atomic layers) for the zirconium
peaks
Auger
8
electrons
to 10 A (4 to 5
is about 4 to 5 A (2
(80 to 180 eV), as
estimated by the Auger electron escape depth correlation of
Hangstrum and Rowe
C343.
The continuing change observed
for the ZRi4z peak
suggests
that
more oxygen is adsorbed
and the oxidation of the top two atomic layers of zirconium
surface is not complete at oxygen exposures of up to 100 L.
The ZRizs peak,
unlike
the
other
peaks, shows more
fluctuation in its intensity ratio vs. oxygen exposure (see
Figure 11).
A
careful
N(E)*E Auger spectra
inspection
indicates
ZRizs peak decreases as the
that
of the corresponding
the intensity of the
oxygen coverage increases.
It
is also observed in N(E)AE Auger spectra that a significant
broadening of the
energy shift, which
ZRizs
peak
causes
a
occurs
possibly
due to an
large overlap between ZRizs
58
and the neighboring Z R n 7
peak.
Due to this overlap and
additional broadening, the peak amplitudes of the ZRizs and
the ZRii7 peaks
in
dCN(E)*E3/dE
reflect the true signal
principal Auger peaks
difference
maximum
in
and
the
spectra
intensities
do.
This
amount
minimum
of
as
well as the other
also
explains a larger
energy
averaging
probably do not
shift
between the
of
dCN(E)*E3/dE
scheme
spectra and of N(E)*E spectra for the Z R n 7 peak (see Table
4 in Appendix A).
Energy shifts of
Auger
have been reported for
nickel
C353
and
peaks
during oxygen exposure
many metals, including for example,
molybdenum
C363.
Peak
generally observed to be associated
energy.
of
a
broadening is
with the shift in peak
Such peak shift and peak broadening are indicative
change
in
the
chemical
states
electronic interactions in
the
the surface region of
specimen.
the
energy shifts of a core
to gain information
not
metal
or
a
change
of
and oxygen atoms in
An analysis of the
level Auger transition permits one
only
on
the
oxidation state of
atoms at the surface but also on changes of oxidation state
during the course of a
AES
investigations
vanadium and
the
of
surface
surface
vanadium
Szalkowski and Smorjai
chemical reaction.
[373
such energy-shift analysis.
chemical
oxides
have
They
VOo.sz,
In the
composition
of
VzOs and VOz,
demonstrated the use of
found
a linear energy
shift of about 0.6 eV per vanadium oxidation number for the
59
LaMzaMza peak, a ccc Auger
the
energy
shifts
of
transition.
the
LaMzaV
On the other hand,
transition,
a
ccv
transition, could not he related to the oxidation states of
vanadium.
For this .reason,
only the energy shifts of core
level transitions are analyzed in this 'study.
In the surface oxidation
transitions, ZRsz and ZRn?,
100 L Oz exposure
than
of zirconium, the core Jevel
exhibit
does
less energy shift at
the -ZRi*s peak, a transition
involving a valence electron, as
shown
in Table 3.
These
smaller shifts are consistent with the fact that core level
electrons are less affected
chemical environment.
increases,
the
transitions are
As
energy
in
than
the
valance electrons by the
surface coverage of oxygen
shifts
the
of
direction
these
of
core
level
increasing binding
energy of the electron energy
levels involved in the Auger
transitions.
agrees
This
zirconium is more
tendency
electropositive
with
than
the
fact that
oxygen.
When the
bonding occurs between Zr and adsorbed 0, the electrostatic
shielding influence of the outer electrons of the zirconium
be decreased
and
levels of zirconium
to
thus
cause
shift
in
the inner shell energy
the direction of higher
binding energies.
As indicated in Figures
of the ZRsz and the
very
low
oxygen
ZRii?
exposure
initial exposure region.
17
peaks
and
Most
and 18, the energy shifts
are observed to start at
increase
of
quickly
in the
the energy shifts are
60
observed between I L and 10 L
0z, more energy
shifts
of
occur, but at a smaller
Oz.
the
rate.
Between 10 L and 100 L
ZRsz
The
two core level transitions are over
100 L Oz, ZRsz exhibits
o.O eV shift.
The
a
-4.0
ZRii7
energyshifts of
-I
eV
peak
and the Z R n 7 peaks
eV
these
at 3 L Oz.
At
shift and Z R n 7 has a
exhibits a larger energy
shift than the ZRsz peak, even though both of them are core
level transitions
(see Table 3).
from
similar
However
Appendix
in the energy
from differentiated and
possibly caused by the
and final states
the two peaks exhibit very similar
snergy shifts as measured from
(see Table 4 in
initial
N(E)*E
spectra at 100 L Oz
A).
As already discussed, this
shift
of the ZRii7 peak measured
undifferentiated
peak
Auger spectra is
broadening and overlap of the
ZRizs peak with the Z R n 7 peak.
Auger
peak
I source of Auger I type of I average shift(*)
I transitions
I transi. j from dCN(E)*E3/dE
ZRsz
I
MNN & MMN
I
ccc
I
ZRi i?
I
MNN & MMN
I
ccc
I
I
O
ZRi 4 a
I
MNV & MMV
I
ccv
I
I
O
Table 3. Energy Shifts of Auger Peaks.
05O9
I
KW
I
cw
I
O
M
I
-4.0
io
(*) energy shifts, at 100 L Oz exposure, are based on
kinetic energy in eV.
61
The ZRi4e peak exhibits an
eV at 100 L Oz.
The
-2.0 eV at 100 L Oz,
energy
energy shift of about -7.0
shift of Osos peak is about
compared to its energy when initially
observed after 0.3 L Oz exposure.
Saturation
of
the
energy
shifts
of
core
level
transitions occurs between 10 L and 30 L Oz, as illustrated
in Figure 26 by the ZRsz
of
the
extent
of
peak's energy shift as a function
oxygen
However, similar to the
the Osos, some
intensity
additional
the ZRsz and the ZRii7
This additional
exposure
energy
peaks
energy
for
shifts
a^ter 30 L Oz indicate
on
a
linear scale.
ratios of the ZRi4z and
shifts are observed for
Oz exposure above 30 L .
of
core level transitions
the occurrence of continuing change
in chemical states of the
Zr.
Since the information depth
the ZRsz and the ZRiiv peaks is about two atomic lavers,
the
continuing
incomplete
change
oxidation
in
of
the
zirconium surface at 30 L
indicated by the results
chemical
top
state
two
confirms
atomic
the
layers of
Oz exposure, which has also been
of
intensity ratios as discussed
above.
Adsorption Kinetics of Oxvaen on Zirconium
In their discussion of
on a metal, Fehlner
and
the initial formation of oxide
Mott
[38] postulated the general
reaction sequence in going from a clean surface to an oxide
B5.70
ENERGY (eV)
BB 70
87.70
KINETIC
BB 70
.70
3?
°o'oo
89.70
□
ZR92
4 00
3.00
a
2 00
ENERGY
ZR92
a
.00
SHIFT (eV)
5 00
84.70
6 00
62
20 00
OXYGEN
4.0. 00
60 00
EXPOSURE
30. 00
(LANGMUIR)
100 00
TzS1.oo
Figure 26. The energy shift of the ZRsz peak from a
Zr sample at different oxygen exposures
at room temperature (linear scale).
63
was first chemisorption,
slow oxidationsurface
next
During
chemisorption
saturated.
rapid oxidation and finally
the
initial
sites
During the
are
chemisorption, the
believed
to
become
rapid oxidation, chemisorbed oxygen
nucleates and the oxide starts
to
growing small domains.
In
uppermost monolayer
"oxide"
of
thickening of
oxide
observed this
reaction
the
exist on the surface as
final slow oxidation, the
is
occurs.
probably complete and
Holloway
sequence
in
and
the
Hudson C391
oxidation
of a
nickel single-crystal surface.
The analysis of the energy shifts of the Auger data of
this study and
the
observation
indicate that oxygen uptake on
temperature seems
also
to
of
zirconium
chemisorption process
example,
Foord
et
by
stage
of
the
from a
the
chemisorption
by zirconium
using
kink, a change in
Foord
with
zirconium
reported by Tapping C23.
an
et
the
to
follow
a
researchers, for
the
work function
al. claim that the
reaction
with
Oz
is
a
adsorbates occupying
surface.
oxide
The transition
nucIeation
stage was
In his study on oxygen adsorption
XPS, Tapping
slope,
shown
From
surface
to
reaction sequence
previous
E233L
dissociative chemisorption
sites just below
been
many
al.
the
The initial stage of surface
has
measurement on Zr-Oz system,
initial
a zirconium surface at room
follow
proposed by Fehlner and Mott.
oxidation
of previous investigators
at
3.5
observed a reproducible
L
Oz
in the XPS O(Is)
64
intensity data.
This
transition from
stage.
a
kink
was
suggested to represent a
chemisorption
to
an
oxide nucleation
However, no direct evidence of oxide nucleation was
reported in Tapping's work.
An analysis of the
data of this study
energy
provides
nucleation and oxide
support for the occurrence of
growth.
As
magnitude of energy shift of
measurement of the
change
a
in
surface investigated by AES.
transition between
two
the
chemical states of the
The
energy
exposure
in energy shifts of core
oxygen
L
chemical
interaction between Zr and
this region of
17 and 18, a large change
level Zr Auger transitions occurs
between oxygen exposures of 3
in
and
state
adsorbed
exposure.
occurrence of oxide nucleation and
oxygen exposure since there
state in this region of
levels therefore
atoms between these exposure
As indicated in Figures
change
shift of a ccc
the degree of interaction between
zirconium and adsorbed oxygen
that more
already discussed, the
core level transition is a
oxygen
■represents the change of
levels.
shifts of the Auger uptake
10 L.
This suggests
and more electronic
oxygen atoms occur in
This also suggests the
growth at this level of
exists more change in chemical
oxygen exposure.
concurs with the analysis of Sen
This observation
et al. C253 of XPS Zr(Bd)
data associated with Oz adsorption on Zr.
They showed that
an increasing amount of an intermediate sub-oxide of a high
oxidation state was present between 15
L and 70 L Oz which
65
reached a maximum surface concentration around 25 L Oz.
Oz Exposure Pressure Dependence
on Aucrer Spectra Intensity ____
The dependence of oxygen pressure at specific exposure
levels (in Langmuirs) on
peak amplitude ratio was examined
by varying the
pressures
exposure
Langmuir exposures.
The
for
tests of the same
results, monitored using the peak
height ratio of
Osos
peak
Figure 19 as .a
function
(Osos/ZRsz),
of
indicate that the extent
exposure
are displayed in
time.
These data
of surface oxidation, as measured
by the normalized peak height ratio of Osos peak, increases
slightly as
the
exposure pressure
exposure
decreases
Similar results have
from the N(E)*E
system).
been
Auger
exposure
Their data
time)
at
increases
a
shown
spectra
In their study,
L are achieved with two
1000 sec).
time
given
by
of
or
the oxygen
exposure level.
Axelsson et al. [17]
oxygen (from the Zr-Oz
Oz exposures between IO4 and IO6
exposure time periods (100 sec and
at
low exposure pressure (1000 sec
demonstrate
higher
amounts
of
adsorbed
oxygen.
Since the oxygen uptake . is favored by lower pressures
with longer exposure time, it is expected that the spectral
transition points or the saturation
occur at lower Oz
lower pressures.
exposure
Such
a
in the Zr surface will
levels for experiments made at
dependence
on
pressure can be
66
partially
responsible
for
level reported to reach
the
inconsistent
Oz exposure
the saturation among the published
research papers.
As can be seen
in
Figure
intensity seems less dependent
lower oxygen exposure of 3 L
19,
on
and
oxygen exposure time at
10
these lower oxygen exposure levels,
the oxygen exposed Zr
surface
the rapid oxidation region.
the Osos Auger signal
L than at 100 L.
At
it has been shown that
is in the chemisorption and
On the other hand, the surface
is mainly controlled by the slow oxidation process at 100 L
Oz exposure.
This shows
time dependent when the
nucleated
oxygen.
that the surface reaction is more
zirconium
The
surface is covered with
difference
dependence may therefore indicate
in
exposure-time-
the extent of nucleation
on the Zr surface.
A u c r e r Results above Room Temperature
Dissimilar behavior
during
Oz
exposure
in
to
Auger
peak intensity changes
zirconium
temperatures at or below 470 K
is
and at 673 K.
observed
for
As indicated
in Figure 20, the peak height ratio of Osos obtained at 673
K is apparently lower than those obtained at or below 470 K
at each corresponding
oxygen
for example Foord et al.
in peak intensity
is
exposure.
Previous studies,
[23], show that such a difference
caused
by
a
diffusion of adsorbed
67
oxygen into the bulk of the
variation in
the
zirconium.
magnitude
of
the
.There is a similar
Osos
intensities at
several temperatures of 470
K
Figure 20.
that the diffusion of adsorbed
This
indicates
oxygen into the bulk Zr
at
Further discussion on the
and below as illustrated in
or
below 470 K is very small.
behavior of surface oxidation of
zirconium above room
temperature
the
in
data
reasons.
developed
First,
the
that
some
indicates
incompletely cleaned
above
room
time
intensity
of
Zr
the
prior
to
to
period
Auger
(about
acquisition
important.
Hence,
multiplexing
Auger
when
signal
5
several
of the Osos peak
tested
were
experiments conducted
Secondly,
obtain
for
specimens
the
the
use
spectra
minutes
of
the
requires a
per run) for
A relatively large amount
of adsorbed oxygen can diffuse
data
study
initial
recording each Auger spectrum.
during the
not appropriate from
present
temperature.
multiplexing scheme
fairly long
the
is
into
time
the bulk of the metal
when diffusion becomes
diffusion
obtained
is
is
a
apparent,
the
time-averaged
result and does not represent the true initial condition of
the surface.
Ion Sputtering- Monitored with SIMS and AES
The
SIMS
signal
shows
during sputtering (see Figure
a
rapid
21).
intensity decrease
The rapid decrease in
68
SIMS signal intensity implies that the surface oxide formed
at 60 L Oz
resides
layers, since the
atomic layer.
primarily
SIMS
This
at
signal
the very top few atomic
is
observation
primarily from the top
agrees with the previous
conclusion drawn from the AES results of this study.
In contrast to the
the intensity of the
interruptions
moderate
difference
Osos
during
decrease,
in
behavior
Auger peak, taken at sequential
sputtering,
as
the
of the ZrO+ SIMS signal,
exhibits
indicated
decrease
sputtering between the two
with the fact that the
in
of
relatively
Figure
signal
surface
a
22.
This
intensity during
spectra does not agree
AES signal originates from a larger
information depth, about
3
peak, compared to the SIMS
atomic .layers
signal.
very thick, thicker than the
for
the Osos
If the oxide layer is
information depth of the Osos
Auger signal, the resulting sputter depth profile of either
spectral method will be flat
initially.
constant signal yield, the Osos
when the thickness of
falls below about
would not
sense
3
much
the
signal
signal for the
Both
signals
should
oxide
should
variations if .the
layer
layers,
difference
thickness of oxide layer is
the Auger
signal would start to drop
oxide
atomic
begin
thickness
initial
in
to
on the Zr surface
but
thinner.
exhibit
After the initial
the SIMS signal
intensity until the
For the same reason,
drop
before the SIMS,
between these two levels.
similar
oxide
sputtering profile
thickness
is
only one
I
69
atomic layer or less.
t^ifference
From
depth
intensity of the
of
Osos
this analysis, because of the
surface .sensitivity,
peak
similar decrease during the
signal intensity, if the
should
the
thickness on Zr
the
two
two
spectral methods both relate
sputtering.
oxidized Zr.
three
possible
surface
matches
surface
spectral
classes
signals
The observed faster
ions during
sputtering
of
None of the
oxide layer
the observed behavior of
during
interrupted
decrease of the ZrO+ SIMS
signal during sputtering therefore
of ZrO
a sharper or
sputtering than would the ZrO+
only to the surface abundance of
outcomes of
exhibit
the signal
suggests that the yield
is
closely related to the
oxidation state of Zr surface structure.
The relatively large
when the ZrO+
Figure 23,
SIMS
signal
indicates " that
closely related to
surface.
the
For instance,
about 8% of its initial
sputtering.
this point
the
oxygen.
Auger
is
small.
the
SIMS
oxidation
the
signal that remains
as
can be seen in
signal intensity is
state
of the zirconium
ZrO+ SIMS signal decreases to
intensity after the first minute's
On the other hand, the Osos Auger intensity at
exhibits
initial value.
that
Osos
Zr
an
This
intensity
about
50%
of its
larger Osos signal intensity confirms
surface
Therefore, the
is
still
coexisting
ZrO+ SIMS signal intensity implies
not only related to
of
the
abundance
relatively
replete with
larger decrease of the
that the SIMS signal is
of oxygen but also the
70
oxidation state of the Zr surface investigated.
Further evidence of
the
existence
between the oxidation state of
-yield can
spectra
also
be
taken
found
at
the Zr surface and the SlMS
from
the
sequential
sputtering process.
This is
spectra,
sequential
taken
at
sputtering process, with
of a relationship
nature
of the Auger
interruptions,
during
the
done by comparing these Auger
interruptions
those
during
the
Auger spectra obtained for
,different extents of oxygen exposure to initially sputtered
clean Zr, especially in terms
of energy shifts
and
of Auger line shape, extents
intensity
ratio
of
the Osoa peak.
From this comparison, the Auger spectrum corresponding to a
92% decay in the ZrO+
sputtering) is close
SIMS
to
signal (after about I minute's
those
Auger
spectra taken for a
sputter clean Zr surface after it is exposed to I to 3 L of
oxygen.
This
associated
Oz
with
exposure
a
level
has
chemisorption
been
stage
shown to be
by
previous
researchers C2, 233.
When the SIMS ZrO+ signal shows.a 65%
decay
0.12
initial
(after
about
value,
the
minute's
Auger
spectrum
spectrum of a sputtered clean
oxygen exposure.
the Zr
surface
Zr
sputtering)
resembles
surface
from its
the Auger
after 3 to 10 L
At an oxygen exposure level of 3 to 10 L,
has
been
shown
to
be
undergoing oxide
nucleation.
a
71
CONCLUSIONS
1.
Both, signal intensities
transitions
observed
and energy shifts of the Auger
during
oxygen
uptake
reach approximate saturation at around
within the top two atomic
on zirconium
30 L Oz.
Oxidation
layers of oxidized Zr surface is
not complete at 100 L Oz exposure.
2.
Oxygen
seems
to
adsorption
follow
chemisorption,
on
the
next
zirconium
reaction
rapid
at room
temperature
sequences
oxidation
and
of
first
finally
slow
oxidation, which was postulated by Fehlner and Mott [38].
3.
Oxygen uptake on zirconium
on the exposure
level
but
pressure (exposure time) of
is favored by lower
exposure level.
pressure
is found to depend not only
also
slightly on the exposure
oxygen.
and
This dependence
The adsorption of Oz
longer time at the same
is more dramatic at an Oz
oxygen exposure of 100 L as compared
to
100 L of Oz
is mainly occupied by
exposure,
the
surface
3 L and 10 L.
At
nucleated surface oxide instead of chemisorbed oxygen.
4.
Very little
diffusion of absorbed/adsorbed oxygen into
the bulk of a Zr sample occurs at temperatures below 470 K.
Diffusion
starts
to
become
between 470 K and 673 K.
significant
at
temperature
72
5.
Secondary
ion
yield
from
an
oxidized Zr surface is
observed to be closely related to the level of oxidation of
the surface
structure.
A
chemisorbed oxygen exhibits a
relatively large
Auger
surface
very
signal;
the
that contains mainly
weak SIMS signal but a
SIMS
ZrO+ signal is
larger when the oxygen exists in oxide nuclei.
73
RECOMMENDATIONS
1.
A sputtering study of Zr
after 100 L, .1000 L and 10000
L Oz exposure and of ZrOz (or oxygen-saturated Zr surface),
utilizing SIMS
(static
concurrent AES, would
SIMS
provide
mechanism and the'kinetics
room temperature.
will
Care
of
be
more
more
information about the
surface
should
helpful) and
be
oxidation of Zr at
taken
to ensure the
cleanness of Zr samples used and to avoid possible problems
caused by
the
insulation
property
of
ZrOz
(or oxygen-
saturated Zr).
2.
Extend
spectra
the
to
a
studies
larger
experiments
will
transition
occurs
of
pressure
range
of
Oz
probably . help
among
the
dependence on Auger
exposures.
determine
three
Such
when
reaction
the
sequences
postulated by.Fehlner and Mott.
3.
To have an improved understanding of the behavior of Zr
surface
oxidation
above
room
temperature,
additional
experiments that cover larger and more detailed temperature
ranges are needed.
Similar
sputtering studies on oxygen
exposed Zr at high
temperatures might add valuable insight
to
oxidation
the
surface
temperatures.
process
of
Zr
at
high
74
REFERENCES
75
REFERENCES
I. Yasuko Nakano et al., J . Catal.. 57(1979)1-10.
2. Tapping, R.L., J. Nucl. Mater.. 107(1982)151.
3. Lin, J.M. and Gilbert, R.E., Appl. Surface Sci..
18(1984)315.
4. Blumenthal, W.B.,. "The Chemical Behavior of
Zirconium, "D. VanWostrand Co. , Inc., N'.Y. (1958).
5. Atomic Data. 5(1973)317-469.
6. Shih, H.D., Legg, K.P. and Jona, F., Surface Sci.,
54(1976)355.
7. see, for example, Cox, B., Adv. Corr. Sci. Technol..
5,(1976)173.
8. Kofstad, P., "High Temperature Oxidation of Metals,"
pp. 179-188, Wiley and Sons, New York (1966).
9. Kofstad, P., "High Temperature Oxidation of Metals,"
pp. 156, Wiley and Sons, New York (1966).
;
10. Ritchie, I.G., Atrens, A., J. Nucl. Mater.,
67(1977)254.
11. Pemsler, J.P., J. Electrochem. Soc., 105(1958)315.
12. Kofstad, P., "Nonstoichiometry, Diffusion and
Electrical Conductivity in Binary Metal Oxides,"
pp. 158-159., Wiley-Interscience, Inc., N.Y. (1972).
j
1'I
.
I
76
13. Briggs, D. and Riviere, J.C., in "Practical Surface
Analysis by Auger and X-ray Photoelectron Spectro­
scopy," Briggs, D. and Seah, M.P., Editors, pp.94-95,
. John Wiley & Sons, Ltd, (1983).
14. Joshi, A., Davis, L.E., and Palmberg, P.W., in
"Methods of Surface Analysis," Czanderna, A.W.,
Editor, pp. 181, Elsevier Scientific. Publishing
Company (1975).
15. Davis, L.E., MacDonald, P.W., Palmberg, P.W. ,
Riach, G.E., and Weber, R.E., "Handbook of Auger
Electron Spectroscopy (Physical Electronics
Industries, Eden Prairie, MN)," 2nd ed., pp. 125,
(1976.) .
16. Benninghoven, A., Surface Sci., 35(1973)427.
17. Axelsson, K.O., Keck, K.E., and Kasemo, B., Surface
Sci., 164(1985)109.
18. McHugh, J. A., in "Methods of Surface Analysis,"
Czanderna, A.W., Editor, pp. 228-229, Elsevier
Scientific Publishing Company. (1975).
19. Dawson, P.T. and Walker, P.C., in "Experimental
Methods in Catalytic Research," Vol. 3, pp. 250,
Anderson, R.B. and Dawson, P.T., Editors, New York,
Academic Press (1976).
20. Douglas, D.L., "The Metallurgy of Zirconium," pp. 389405, International Atomic Agency (1971).
21. Pemsler, J.P., J. Electrochem. Soc., 111(1964)381.
22. Kofstad, P., "Effect of Impurities on the Defects in
Oxides and their Relationship to Oxidation of Metals,"
Corrosion, 24(1968)379.
23. Foord, J.S., Goddard, P.J. and Lambert, R.M., Surface
Sci., 94(1980)339.
77
24. Yashonath, S., Sen, P., Hedge, M.S. and Rao, C.N.R.,
J. Chem. Soc. Faraday Trans. I . 79(1983)1229.
25. Sen, P., Sarma, D.D., Budhani, R.C., Chopra, K.L. and
Rao, C.N.R., J. Phvs. F; Met. Phvs.. 14(1984)565.
26. Valyukhov, D.P., Goluhin, M.A., Grebenshchikov, D.M.
and Shestopalova, V.I., Sov. Phys. Solid State, 24(9),
September (1982)1594.
27. Berezinz, N.N., Valyukhov, D.P. and Vorontsov, E.S.,
Russian J. Phvs. Chem., 55(11)(1981)1651.
28. HofIund, G.B. and Cox, F.D., J: Vac. Sci. Technol. Al,
4(1983)1837.
29. HofIund, G.B., Asbury, D.A., Cox, D.F. and Gilbert,
R.E., A p p I . Surface Sci.. 22/23(1985)252.
30. Krishnan, G.N., Wood, B.J. and Cubicciotti, J.
Electrochem. Soc.: Solid State Sci. and Technol,
128(no. I)(1981)191.
31. Joshi, A. , Davis, L.E., and Palmberg, P.W. , in
"Methods of Surface Analysis," Czanderna, A.W.,
Editor, pp. 161, Elsevier Scientific Publishing
Company (1975).
32. Chang, C.C., in "Characterization of Solid Surface,"
Kane, P.F. and Larrabee, Editors, pp. 537, Plenum
Press, New York (1974).
33. Hagstrum, H.D. and Rowe, F.E., in "Experimental
Methods in Catalytic Research," Vol. 3, pp. 88,
Anderson, R.B. and Dawson, P.T., Editors, N.Y.,
Academic Press (1976).
34. Hagstrum, H.D. and Rowe, F.E., in "Experimental
Methods in Catalytic Research," Vol. 3, pp. 57,
Anderson, R.B. and Dawson, P.T., Editors, N.Y.,
Academic Press (1976).
78
35. Horgan, M.A. and Dalins, I., Surface Sci..
36(1973)526.
36. Hass, J.W. , Grant r J.T. and Dooley, G.J. Ill,
J. Appl.Phvs.. 43(1972)1853. 37. Szalkowski, F.J., Aomoriai, G.A., J. Chem. Phvs.,
56(1972)6097.
38.. Fehlner, F.P. and Mott, N.F., Oxidation of Metals.
2(1970)52.
39. Holloway, P.H. and Hudson, J.B., Surface Sci..
43(1974)123.
79
APPENDICES
80
Appendix A
The comparison of Auger energy shift from
dCN(E)*E3/dE with that from N(E)*E spectra
The energy
shifts
of
oxygen
peaks are quantified in this
of the
shifts
of
peak
d[N(E)*E]/dE spectra.
uptake zirconium Auger
study in terms of the average
maxima
Table 4
and
peak
minima
in the
shows the comparison of the
energy shift of the average of the shift of peak maxima and
peak minima
in
dCN(E)*EH/dE
spectra
to those determined
from N(E)^E spectra.
Table 4. The Energy Shifts of Auger Peaks Determined by
Averaging Peak Maxima and Peak Minima in
d[N(E)AE]/dE and those Determined by the Maximum
Points in N(E)AE Spectra.
spectrum
mode
I average shift
I Energy shift by |
from
dCN(E)AE]/dE
I
I N(E)AE maximum j
I
ZRl 4 8
I
0509
I
O
ZRi 17
I
-4.0
I
-5.5
I
-4.0
I
-7.0
j
I
I
i
O
I
ZRd z
I
Pt
I
O
I
(a) data for "oxidized" were taken from 100 L results of
this research.
(b) energy shift is measured in eV relative to the
kinetic energy of the same Auger peak of the clean
zirconium.
81
The difference
between
the
dCN(E)*E3/dE spectra and from
is small.
Z R n 7 peak
two schemes.
exhibits
Possible reason
results
determined from
N(E)*E spectra of this study
a larger difference in the
for this is presented in the
section of discussion of results.
82
Appendix B
The Oxygen Exposure Conditions under which
the Data Compiled in Figure 24 were Taken
All the data plotted in
temperature
starting
with
Figure
a
24 were taken at room
sputter
cleaned
zirconium
specimen without annealing. . The exposure conditions prior
to recording the Auger spectra are listed in Table 5 on the
next page.
For identical
conditions are listed in
0509 signal intensity.
Oz exposure levels, the exposure
descending order of the resulting
83
Table 5. Exposure Conditions for the Data in Figure 24.
Oz Exposure (L)
Exposure Time (sec)
Oz Pressure (Torr)
0.3
100
3 A 10 9
1.0
100
7 A IO-9
3.0
300
I A IO-8
3.0
30
I A IO"7
3.0
100
3 A IO-8
3.0
10.0
Sequential exposure of a + b + c ^
Sequential exposure of a+b+c+d^^
10.0
1000
I A IO-8
10.0
333
3 A IO-8
10.0
100
I A IO"7
10.0
100
I A IO"7
30.0
Sequential exposure of a+b+c+d+e^
30.0
Sequential exposure of g+h<4>
100.0
I A IO'6
100
100.0
Sequential exposure of a+b+c+d+e+f (TjJ)
100.0
Sequential exposure of g+h+i1* 1
(Tj))
where
a
b
C
d
e
f
g
h
i
is
is
is
is
is
is
is
is
is
sec
sec
sec
sec
sec
sec
sec
sec
200 sec
100
100
100
100
100
100
100
100
X
X
X
X
X
X
X
X
X
A
A
A
A
A
A
1.0 A
2.0 A
3.5 A
3.0
7.0
2.0
7.0
2.0
7.0
10_g Torr
i:;
io 7
10
10
10
10
7
7
7
9
Torr
Torr
Torr
Torr
Torr
Torr
Torr
Torr
STATE UNIVERSITY