Crystal Structure of the Zirconium Hydride Polymorphs determined by
Neutron and Synchrotron X-ray Powder Diffraction
A. Steuwer1, J. Blomqvist2, T. Maimaitiyili2, O. Zanellato3, C. Curfs4, H.E. du Plessis5, C. Bjerken2
1ESS AB, Stora Algatan 4, 22100 Lund, Sweden and NMMU, Port Elizabeth, 6031 South Africa,
2Malmö University, Malmö, Sweden
3Mateis, INSA, Lyon, France
4ESRF, rue J Horowitz, 38042 Grenoble, France
5SASOL, Sasolburg, South Africa
Abstract: Zirconium alloys, widely used in the nuclear industry, Fig 1, have a strong affinity for hydrogen that leads to
hydrogen pick up during a corrosion reaction when exposed to water. The hydrogen is readily in solution at higher temperature
but precipitates as Zirconium hydrides at ambient temperatures. At least three phases are presumed to exist at ambient
temperature depending on hydrogen concentration and quenching rate. However, some controversy exist regarding the exact
nature, exact structure and stability of the γ-ZrH phase, which is closely related to the -ZrH phase through ordering of the
hydrogen on tetrahedral sites on alternating 110 planes. Aim: Develop a recipe for in-situ preparation of γ-ZrH/D (conflicting
reports exist) with purpose of high resolution neutron powder diffraction studies on deuterated Zr powder samples in order to
re-determine and verify the reported structures that essentially date back to the 1960s [1], and compare those with high
resolution synchrotron X-ray powder diffraction, and supporting results from other techniques.
Fig. 1: Typical fuel assembly, uranium
dioxide, pellets and fuel assembly
(with zircalloy tubing/cladding)
Background: The three crystal structures of ZrH2-x are
a) for x=0 the tetragonal ε-phase (not shown), b)
x=0.34 the cubic, disordered δ-phase, and x=1 the
ordered γ-phase, with hydrogen atoms in tetrahedral
positions on alternating 110 planes as stated in [1,2].
The illustration on the right shows the δ and γ phase
crystal structures visualised with full filling of the
tetrahedral sites. The controversy of the ordered
phases surrounds its RT stability and structure [3].
Figure 2:
Experimental procedure: a) Deuteration: The powder sample
which was used in experiment was prepared in strict order
following literature: First, the sample baked three days at 300 C
(baking at P=5x1010 mbar). Then, the sample exposed to a volume
with a given amount of deuterium, when the volume had reached a
low pressure (deuterium absorbed in the sample), the volume was
refilled. This cycle was repeated until the desired amount of
absorbed deuterium was achieved. During the entire sample
preparation process the powder was handled in a glove box filled
with Argon to avoid any sort of contamination or oxidation.
The final Zr:D ratio (atomic) was almost 1:1, the ideal
stochiometric ratio of the γ-phase. After preparation, the H/D rich
ε-phase (x=0), as well as the more stable δ-phase and remaining
α-phase (pure Zr) are present in the as prepared powder.
b) in-situ experiments: SPODI: the wave length of the neutron
kept as constant (λ=1.548211 Å). The powder was tested both at
room temperature (as prepared) and elevated temperature to
check the stability of the phases, in particular any transition around
180 C and/or 286 C, which has been suggested in the literature.
However, very little γ phase in the annealed sample was detected,
Fig.3 top.
ANSTO: a separate in-situ preparation route (different thermal
cycle) was tried subsequently at the ECHIDNA (HRPD)
diffractometer at ANSTO. A new thermal cycle was found to
successfully produce more traceable amounts of γ-ZrD, see Fig 3,
Additionally, the samples, as prepared and annealed (FRM2) have
been measured on ID31 (λ=0.306592 Å) at the ESRF for
comparison, Fig.3 bottom. The second set of samples (postECHIDNA) are considered for more ID31 measurements.
c) Conclusions: Having finally identified a preparation route for γZrD, we are now preparing to confirm the exact structure with
simultaneous refinement of synchrotron X-ray and neutron
diffraction data, as the obtained data is not in agreement with
published structures, Fig 3.
Fig.3: Preliminary Rietveld fits of the FRM2 (top) and ANSTO
data (middle), and ID31 data (bottom) of the ZrD powders.
References: [1] S. S. Sidhu, N. S. Satyamurty, F. P. Campos, and D. D. Zauberis, Neutron and X-ray Studies on Non-Stoichiometric
Metal Hydrides, in Advances in Chemistry, (American Chemical Society, Washington, DC, 1963), Vol. 39, p. 87. [2] E. Zuzek, J. P.
Abriata, A. San-Martin, and F. D. Manchester, Bulletin of Alloy Phase Diagrams (American Society for Metals, Metals Park, Ohio, 1990),
Vol. 11, No. 4. [3] Steuwer, A; Santisteban, JR; Preuss, M; Peel, MJ; Buslaps, T; Harada, M. Acta Mat. Vol 57, Iss 1, p145-152, 2009
Acknowledgements: FRM2 and ANSTO are gratefully acknowledged for the provision of beam time.
Corresponding author: [email protected] , ESS AB, Stora Algatan, Lund, Sweden, prepared for ECNS 2011, Prague

similar documents