Below is a cache of http://www.mne.psu.edu/motta/JNM_Oxide_04.pdf. It's a snapshot of the page taken as our search engine crawled the Web.
The web site itself may have changed. You can check the current page or check for previous versions at the Internet Archive. Yahoo! is not affiliated with the authors of this page or responsible for its content.
Structure of zirconium alloy oxides formed in pure water studied with synchrotron radiation and
Structure of zirconium alloy oxides formed in pure
water studied with synchrotron radiation and
optical microscopy: relation to corrosion rate
Aylin Yilmazbayhan
a
, Arthur T. Motta
a,*
, Robert J. Comstock
b
,
George P. Sabol
c
, Barry Lai
d
, Zhonghou Cai
d
a
Department of Mechanical and Nuclear Engineering, The Pennsylvania State University, University Park, PA 16802, USA
b
Science and Technology Department, Westinghouse Electric Co., 1340 Beulah Rd, Pittsburgh, PA, USA
c
Consultant to Westinghouse Electric Co., 1340 Beulah Rd, Pittsburgh, PA, USA
d
Advanced Photon Source, XFD 401 B3194, Argonne National Laboratory, 9700 South Cass Ave., Argonne, IL 60439, USA
Received 4 April 2003; accepted 22 August 2003
Abstract
A detailed study was undertaken of oxides formed in 360 C water on four Zr-based alloys (Zircaloy-4, ZIRLOe,
1
Zr2.5%Nb and Zr2.5%Nb0.5%Cu) in an eort to relate oxide structure to corrosion performance. Micro-beam X-
ray diraction was used along with transmitted light optical microscopy to obtain information about the structure of
these oxides as a function of distance from the oxidemetal interface. Optical microscopy revealed a layered oxide
structure in which the average layer thickness was inversely proportional to the post-transition corrosion rate. The
detailed diraction studies showed an oxide that contained both tetragonal and monoclinic ZrO
2
, with a higher fraction
of tetragonal oxide near the oxidemetal interface, in a region roughly corresponding to one oxide layer. Evidence was
seen also of a cyclic variation of the tetragonal and monoclinic oxide across the oxide thickness with a period of the
layer thickness. The results also indicate that the nal grain size of the tetragonal phase is smaller than that of the
monoclinic phase and the monoclinic grain size is smaller in Zircaloy-4 and ZIRLO than in the other two alloys. These
results are discussed in terms of a model of oxide growth based on the periodic breakdown and reconstitution of a
protective layer.
2003 Elsevier B.V. All rights reserved.
1. Introduction
For the last decade, the corrosion behavior of Zr-
based alloys has been at the forefront of LWR fuel
technology. The corrosion resistance of fuel cladding
and structural components often limits economic
improvements in fuel utilization, such as those associ-
ated with higher heat uxes, uid temperatures and core
residence times. This challenge to fuel performance has
been addressed in the past through optimization of the
chemistry and microstructure of the existing commercial
alloys, Zircaloy-2 and -4 and the Zr1.0% and 2.5%Nb
alloys, and introduction of alloys containing both Sn
and Nb [1,2]. This empirical development has resulted in
the correlation of alloy microstructure with corrosion
behavior and in practical thermo-mechanical processing
schemes for achieving optimum corrosion response. It is
now well established that corrosion resistance of Zirca-
loy-4 in pressurized water reactors (PWR) is improved
when the size of the second-phase particles (SPPs) is
greater than about one-tenth micron [3] and the tin
content is in the low range of the specication [4,5]. In
*
Corresponding author. Tel.: +1-814 865 0036; fax: +1-814
865 8499.
E-mail address:
atm2@psu.edu
(A.T. Motta).
1
ZIRLO is a trademark of Westinghouse Electric Co.
0022-3115/$ - see front matter
2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.jnucmat.2003.08.038
www.elsevier.com/locate/jnucmat
Journal of Nuclear Materials 324 (2004) 622 contrast, small SPP size (less than one-tenth micron) is
needed for maximum corrosion resistance of the 1.0%
and 2.5%Nb alloys [6,7] and also for ZIRLO [1]. Small
SPP size is also needed in the Zircaloys for resistance to
nodular corrosion in boiling water reactors (BWRs).
The mechanisms for such corrosion improvements are
not well understood.
As a result of the already extensive optimization of
the existing alloys, it is believed that further signicant
improvement in corrosion resistance requires a more
mechanistic understanding of the corrosion process to
allow selective emphasis on those characteristics favor-
able to superior corrosion behavior. Therefore, a fun-
damental understanding of the structural development
and properties of the oxide layer is needed. Eorts to
provide such fundamental data have been ongoing for
some time [810]. Unfortunately, many of the studies
focus on a limited set of oxide features, or are limited by
alloy availability. In addition, the resolution limits of the
experimental techniques are often insucient for detec-
tion of signicant eects.
The commonly accepted macroscopic description of
the corrosion kinetics of zirconium-based alloys in
aqueous media divides the total process into two regimes
[1113]. An initial pre-transition region that is approxi-
mately parabolic with respect to time is followed by a
post-transition region of more accelerated kinetics with
an approximately linear dependence on time. The onset
of the accelerated corrosion regime is called the Ôoxide
transition and is characterized by either the exposure
time or by the oxide thickness at which the change in
kinetics occurs. However, in reviewing the corrosion
data with more scrutiny, it is observed that this simple
description of the kinetics is only an approximation. The
kinetics in pre-transition are not parabolic, but display a
cubic dependence on time [14,15]. More importantly, the
post-transition regime is composed of several periods of
corrosion that mimic pre-transition kinetics in a cyclical
sequence. On individual samples, the cyclic nature of the
kinetics can be easily distinguished through several
repetitions. The cyclic nature of post-transition corro-
sion has been observed and noted for some time [16,17]
and observations [8,9,18,19] of stratication in the oxide
lms have been correlated with the cyclic kinetics.
Acknowledging that oxide growth results from oxy-
gen migration through the oxide, the rate of oxidation
may be controlled by either the ionic or electronic con-
ductivity of the oxide layer. For pure zirconium, evi-
dence [13,20] indicates that electronic conductivity is
rate controlling at PWR fuel cladding temperatures,
290400 C. For the oxide formed on Zircaloy-2, on the
other hand, electric potential measurements across the
oxide indicate corrosion rate control by ionic transport
processes. For the Zr2.5%Nb alloy the relative impor-
tance of ionic and electronic processes depends upon the
metallurgical structure of the alloy. For material with a
ne distribution of b-Nb precipitates (quenched and
aged) for which the corrosion resistance is very good,
ionic transport is rate controlling at the temperatures of
interest. It has also been suggested [20,21] that relative
changes in the ionic and electronic conductivity caused
by irradiation, such as fast neutron damage of the oxide
and radiolysis of the coolant, may be responsible for in-
reactor accelerated corrosion. Regardless of the role of
electronic conduction in the corrosion kinetics, the ex-
tent of corrosion results from the mass transfer of oxy-
gen through the oxide layer and subsequent conversion
to more oxide at the metal/oxide interface. Furthermore,
for out-reactor post-transition corrosion of alloys of
commercial signicance, it is generally believed that
ionic transport of oxygen through the oxide layer con-
trols corrosion kinetics.
From studies of the nature of the oxide lms, a sig-
nicant number of oxide features have been identied.
The oxides are a mixture of the stable monoclinic ZrO
2
phase and a tetragonal ZrO
2
phase, the latter stabilized
by local conditions in the oxide, such as stresses, small
grain size and dissolved alloying elements [22]. The grain
morphology consists of a mixture of equiaxed and
columnar grains [2326] and it has been proposed that
the columnar grains are protective [27,28]. These oxides
are known to develop cracking and porosity during their
growth, which various researchers have linked to the
transition in kinetics [9,18,29]. The SPPs present in the
base metallic alloy are incorporated unoxidized into
the oxide layer and only undergo oxidation after
some residence time in the oxide [30,31], which some
researchers have associated with the onset of oxide
transition [26]. There is a clear degradation of corro-
sion behavior when the oxides are formed in lithiated
water. It has been proposed [27] that the eect of Li is to
destroy the inner barrier layer by preventing develop-
ment of the columnar structure. Instead, equiaxed grains
are formed that are less protective.
Thus, there is considerable data on oxide observa-
tions.
However,
conicting
data
prevent
further
extrapolation of t