Systems with a ferroelectric to paraelectric transition in the vicinity of room temperature are useful for devices. Adjusting the ferroelectric transition temperature (T(c)) is traditionally accomplished by chemical substitution-as in Ba(x)Sr(1-x)TiO(3), the material widely investigated for microwave devices in which the dielectric constant (epsilon(r)) at GHz frequencies is tuned by applying a quasi-static electric field. Heterogeneity associated with chemical substitution in such films, however, can broaden this phase transition by hundreds of degrees, which is detrimental to tunability and microwave device performance. An alternative way to adjust T(c) in ferroelectric films is strain. Here we show that epitaxial strain from a newly developed substrate can be harnessed to increase T(c) by hundreds of degrees and produce room-temperature ferroelectricity in strontium titanate, a material that is not normally ferroelectric at any temperature. This strain-induced enhancement in T(c) is the largest ever reported. Spatially resolved images of the local polarization state reveal a uniformity that far exceeds films tailored by chemical substitution. The high epsilon(r) at room temperature in these films (nearly 7,000 at 10 GHz) and its sharp dependence on electric field are promising for device applications.

An overview of the state of art in ferroelectric thin films is presented. First, we review applications: microsystems' applications, applications in high frequency electronics, and memories based on ferroelectric materials. The second section deals with materials, structure ͑domains, in particular͒, and size effects. Properties of thin films that are important for applications are then addressed: polarization reversal and properties related to the reliability of ferroelectric memories, piezoelectric nonlinearity of ferroelectric films which is relevant to microsystems' applications, and permittivity and loss in ferroelectric films-important in all applications and essential in high frequency devices. In the context of properties we also discuss nanoscale probing of ferroelectrics. Finally, we comment on two important emerging topics: multiferroic materials and ferroelectric one-dimensional nanostructures.

Understanding the suppression of ferroelectricity in perovskite thin films is a fundamental issue that has remained unresolved for decades. We report a synchrotron x-ray study of lead titanate as a function of temperature and film thickness for films as thin as a single unit cell. At room temperature, the ferroelectric phase is stable for thicknesses down to 3 unit cells (1.2 nanometers). Our results imply that no thickness limit is imposed on practical devices by an intrinsic ferroelectric size effect.

Predictions and measurements of the effect of biaxial strain on the properties of epitaxial ferroelectric thin films and superlattices are reviewed. Results for single-layer ferroelectric films of biaxially strained SrTiO 3 , BaTiO 3 , and PbTiO 3 as well as PbTiO 3 /SrTiO 3 and BaTiO 3 /SrTiO 3 superlattices are described. Theoretical approaches, including first principles, thermodynamic analysis, and phase-field models, are applied to these biaxially strained materials, the assumptions and limitations of each technique are explained, and the predictions are compared. Measurements of the effect of biaxial strain on the paraelectric-to-ferroelectric transition temperature (T C ) are shown, demonstrating the ability of percent-level strains to shift T C by hundreds of degrees in agreement with the predictions that predated such experiments. Along the way, important experimental techniques for characterizing the properties of strained ferroelectric thin films and superlattices, as well as appropriate substrates on which to grow them, are mentioned. 589 Annu. Rev. Mater. Res. 2007.37:589-626. Downloaded from arjournals.annualreviews.org by Pennsylvania State University on 10/22/07. For personal use only.