High-temperature crystal chemistry of sodium zirconium phosphate (NZP)

  • PDF / 625,977 Bytes
  • 9 Pages / 595.44 x 841.68 pts Page_size
  • 96 Downloads / 253 Views

DOWNLOAD

REPORT


D. K. Agrawal and H. A. McKinstry Materials Research Laboratory, The Pennsylvania State University, University Park, Pennsylvania 16802

Anthony J. Perrotta Alcoa Technical Center, Alcoa Center, Pennsylvania 15069 (Received 18 July 1986; accepted 26 March 1987) High-temperature crystal structures of NZP (Na! + x Zr 2 P 3 _ x Si^O,2) have been determined by x-ray measurements made on single crystals. Thermal expansion of NZP (x = 0.11) parallel to the hexagonal c axis is positive (about 22.4X 10~ 6 "C" 1 between 25° and 700 °C), whereas expansion perpendicular to c is slightly negative (about 5.4X 10~ 6 °C~'), resulting in an average volume thermal expansion o f l l . 8 X l O ~ 6 ° C ~ 1 . A proposed structural model to interpret this anisotropic thermal expansion of NZP is tested to prove the model's validity. In this model the rotation of the phosphate tetrahedron is coupled to the rotation of the zirconium octahedron. The observed thermal expansions of sodium, zirconium, and phosphorus cation coordination polyhedra are 10.8, 0.00, and — 0.23 (all X 10~ 6 "C^ 1 ), respectively. The large thermal expansion of the sodium site is offset by rotations in the Zr-P polyhedral framework, thus yielding the low net expansion of NZP.

I. INTRODUCTION The sodium zirconium phosphate NaZr 2 P 3 O ]2 (NZP) structure, first determined 20 years ago by Hagman and Kierkegaard,1 has several interesting features: (i) high ionic conductivity,2 (ii) low thermal expansion,3"5 and (iii) enormously varied ionic substitution in the structure.6 The structural framework is such that a wide range of elements may be substituted at each of the different lattice positions (Li, Cs, Ca, Ba, etc., for Na; Ti, Ge, Hf for Zr, and Si, S for P). Sljukic et al.1 observed that substitution of alkali for Na caused the cell parameters to change in an unusual way. Replacing Na with a large cation caused the c dimension of the rhombohedral cell to increase but anomalously the a dimension decreased. Alamo8 noticed that the polyhedra are coupled and was able to postulate a model that could account for this anomaly. The details of this model have been described in the recent publication by Lenain et al.9 The stability of this structure for a wide range of chemical elements makes this a candidate for a universal sponge for nuclear waste host material.10 Boilot et al.11 noticed that the thermal expansion behavior of Na 1 + J C Zr 2 P 3 _ J C Si x Oi 2 (NASICON) was similarly anomalous. Heating an alkali NZP material caused the c dimension to increase and the a dimension to decrease. Lenain et al.12 examined all the alkali metal ions and found that Li had a positive expansion in both directions while K, J. Mater. Res. 2 (3), May/Jun 1987

http://journals.cambridge.org

Rb, Cs all behaved in a fashion similar to NZP. Limaye et al.13 found that CaZP was similar but in Sr and Ba analogues the signs were reversed, i.e., the a axis expanded and the c axis contracted. Since all members of the family should, in principle, form crystalline solutions, a wide range of anisot