In the practice of solid state chemistry, structural phase transitions are fairly common events. Nonetheless, their understanding, in terms of both: A rationalization of the observed changes in symmetry pattern and; An understanding of the mechanisms allowing for a particular transformation, are outstanding problems. The thermodynamic classification of phase transitions distinguishes between first and second order transitions, on the basis of the discontinuous behavior of quantities related to first or second derivatives of the free energy, respectively. Small atomic displacements are typically associated with second order phase transitions, and latent heat changes amount to a few calories per gram only. Additionally, the symmetries of the phases surrounding the transition are typically in the relation of a group and a subgroup. Reconstructive phase transitions, on the contrary, involve breaking of (large) parts of the bond scaffolding of the initial structure, and exhibit drastic changes at the transition point, with large latent heat and hysteresis effects. The corresponding atomic displacements can be in the order of the lattice parameters, and no group-subgroup is found, between the symmetry of the phases. These type of transitions have generally a strong first-order character.
Landau theory accounts for continuous, second-order phase transitions. As a phenomenological theory, it does not establish the existence of a phase transition, which remains an experimental fact. It only bridges microscopic characteristics, like space-group symmetries and structural changes, or phonon softening effects, with measurable macroscopic quantities. Therein, distortions are carried by an order parameter, which fully specifies the form of the analytical variational free energy. The latter is continuous and derivable with respect to temperature, pressure and atomic displacement, at the transition point.
First order, non-continuous phase transitions are still within the scope of Landau theory in the mentioned special case of the existence of a group-to-(isotropic) subgroup relationship. In the majority of cases, however, and for the most interesting phase transitions (for basic and applied research), such a relationship is missing, making the choice of an order parameter less straightforward. Most of the allotropic transformations of the elements, many intermetallic systems, and numerous insulating systems belongs to this class. This class also includes most interesting and fundamental electronic effects, like metallization in perovskites or spinel oxides for example.
This very simple fact of a missing symmetry condition has helped reinforcing the opinion of first-order phase transitions being a world apart, and possibly contributed to discouraging a firm theory to develop, able to account for their transformation mechanisms and the change of physical properties across phase transition. The thermodynamic distinction between first and second order phase transitions is too narrow, as, in case of first order phase transitions, it embraces both weakly discontinuous transition and reconstructive ones, where bonds are being strongly modified. Especially, a mean to qualify the distance between two structures (geometric, with respect to symmetry, a.s.o.), is missing. Clearly, a group-subgroup relationship may, and typically does imply shortest shifting distances, as a tiny atomic displacement can already do for a symmetry lowering. Naively, missing such a relation means no constraints, and apparently no means to conclude at a connection of two structures in general, let alone a full mechanistic analysis.
First order phase transitions proceed by nucleation and subsequent growth of the new phase from the initial one. Different from (second-order) continuous phase transitions, they do imply coexistence of the transforming motifs. The discontinuity in some order parameter between the two phases is driven by lowering of the free energy as the new phase forms. However, close to the transition, the original phase remains metastable, and a fluctuation is needed to cause the formation of the new phase to set in. Such a process responds to thermal changes, and depending on the height of the nucleation barrier, its rate may be slower or faster. In the former case, large deviations from equilibrium may be required to achieve transformation to the stable phase, which means that large hysteresis effects will be observed in the course of transformation.
The intent of this work consists in giving a face to the intermediate configurations appearing in first order phase transitions, in solid-solid reconstructive processes. Apart of a mechanistic elucidation, consisting in answering the question “Which atomic displacements bring structural motif A into structural motif B ?”, another purpose of this work is a rather pedagogical one, that is, showing that first-order phase transitions can be understood in detail, not only in principle but in fact. The core of the examples illustrated in this work is concerned with phase transformations where pressure represents the thermodynamic controlling parameter. Pressure is extensively used in chemical synthesis, as a mean to achieve novel properties, optical or mechanical just to mention a few. Additionally, reports on novel high-pressure polymorphs are regularly appearing. In this sense, pressure is a relevant parameter for approaching fundamental questions in solid state chemistry.
| Identifer | oai:union.ndltd.org:DRESDEN/oai:qucosa:de:qucosa:25857 |
| Date | 02 December 2009 |
| Creators | Leoni, Stefano |
| Contributors | Seifert, Gotthard, O'Keeffe, Michael, Grin, Yuri, Technische Universität Dresden |
| Source Sets | Hochschulschriftenserver (HSSS) der SLUB Dresden |
| Language | English |
| Detected Language | English |
| Type | doc-type:doctoralThesis, info:eu-repo/semantics/doctoralThesis, doc-type:Text |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0033 seconds