Contrasting compressibility trends in layered structures Ag(tcm) and Ni(CN)₂

Ag(tcm) (tcm = tricyanomethanide) and Ni(CN)2 are layered structures. Both these materials exhibit area negative thermal expansion (area-NTE) due to 'rippling' of the layers - a displacement pattern that causes the interlayer separation to increase and the effective layer area to decrease (Fig. 1a). We have shown that Ag(tcm) shows negative area compressibility (NAC) under hydrostatic pressure. The latter can be attributed to the rippling phenomenon: when hydrostatic pressure is applied, the effective layer area increases (Fig. 1a). On the contrary, Ni(CN)2 shows shows positive linear compressibility in all directions, albeit much more strongly along the stacking axis than in any direction parallel to the square-grid sheets. This is attributed to transverse phonon modes within the Ni(CN)2 sheets, which can be visualised as 'tilting' of the rigid unit modes (RUMs) (Fig. 1b). A discussion of relating such behaviour to the Grüneisen parameters is undertaken in this study.

Local Structure of Iridate Pyrochlores from Hydrothermal Synthesis

Iridate pyrochlores of general formula M₂Ir₂O₇ have potential applications in catalysis [1]. They also often exhibit unusual magnetic and electronic properties caused by spin-orbit coupling and geometric frustration [2]. A detailed understanding of structure is necessary to enable these properties to be understood and exploited. Because of the propensity of the pyrochlore structure to accommodate structural disorder, we have chosen to utilise the technique of total scattering to examine the structure of M₂Ir₂O₇ (M = Bi, Nd). The sensitivity of our measurements to all the constituent elements is maximised by the combination of both neutron and X-ray total scattering. We find no evidence for magnetic ordering in our samples of Nd₂Ir₂O₇, in contrast to literature reports [3]. By comparing the local structure of our samples with that of one reported to exhibit magnetic ordering, we explore the possibility of a structural origin for the differences in magnetic behaviour. We have found that synthesis method can directly influence the structure of these iridate pyrochlores. Local structural analysis provides evidences of A-site cation deficiency and partial oxidation of Ir(IV) to Ir(V) in samples produced by hydrothermal techniques. Irreversible changes to the lattice parameter upon heating these samples at 400 - 900 ⁰C further support the inference that the cation content is somewhat variable. We report the results of reverse Monte Carlo (RMC) refinements using the program RMCProfile, which is capable of simultaneously fitting to X-ray and neutron data, and therefore provides structural models of the greatest possible accuracy. We also report the results of in situ X-ray total scattering measurements which provide local-scale insight into the interesting thermal behaviour and apparent flexible cation content of these materials.

RMCProfile: A step towards complex modelling

The importance of local structure and disorder in crystalline materials is being recognised more and more as a key property of many functional materials. From negative thermal expansion to improved fuel cell technology and solid state amorphisation to the 'nanoscale' problem, a clear picture of the local atomic structure is essential to understanding these phenomena and solving the associated problems. Total scattering, an extension of the powder diffraction method, is increasingly being used to study crystalline materials. The unique combination of Bragg and diffuse scattering can be used to determine both the average structure and the short-range fluctuations from this average within a single experiment. To maximise the structural information from such data, three-dimensional atomic models consistent with all aspects of the data are required. RMCProfile[1] (see www.rmcprofile.org) expands the reverse Monte Carlo (RMC) modelling technique[2] to take explicit account of the Bragg intensity profile from crystalline materials. Analysis of the RMCProfile-generated atomic models gives more detailed information than is available directly from the data alone. I will give several examples where RMCProfile has been used to successfully study the structure and disorder of crystalline materials to illustrate its potential. As the systems being studied become more complex, the information from many experimental techniques and any prior chemical knowledge needs to be combined into one consistent atomic model. The continued development of RMCProfile and its new capabilities is moving us closer to the complex modelling paradigm[3] required to drive the discovery of new functional materials.