The role of interfaces and domain structures in the performance of thin-film devices has continues to be a source of confusion. Devices such as non-volatile ferroelectric memories, ferroelectric-gate field effect transistors, electrooptic modulators, and pyroelectric sensors depend critically on the management of domain structures, interfaces, and space charge. Recent developments in x-ray scattering techniques and technical advances in sources and optics have enabled a new range of experiments to probe these issues. Taking advantage of these developments, Paul Evans, Max Lagally, and coworkers are conducting two sets of complementary x-ray scattering experiments: microdiffraction experiments to study of domain nucleation and growth at length scales as short as 200 nm, and soft x-ray scattering experiments to gain information at lateral length scales of 5 to 500 nm.

In comparison with other techniques presently used to study the domain structure and evolution in ferroelectric thin films, x-ray microdiffraction has several advantages. X-ray diffraction is a structural tool and thus, in contrast to scanned-probe techniques, does not require that large electric fields be applied to the sample, thereby avoiding complications associated with interpreting the piezoelectric response of the film. The contrast leading to x-ray scattering in these experiments is a consequence of the dependence of x-ray absorption on the details of the local structure and electronic state of solids. In particular, individual ferroelectric domains can be distinguished because the x-ray structure function depends on the domain orientation. Experiments since the beginning of the seed project in January have imaged ferroelectric polarization switching in lead zirconium titanate thin films and connected the onset of polarization fatigue with a local structural relaxation. Since microdiffraction requires a high-brilliance hard x-ray source, these experiments are performed at the APS.

Evans and coworkers are also developing a complementary area-integrated technique to obtain information at smaller length scales. This approach again takes advantage of the sensitivity of x-rays to the ferroelectric polarization of the sample at short length scales. The strongest absorption signatures of these local effects are in general near L-edge transitions, which for transition metal ions are found in the soft x-ray energy range. Non-resonant scattering can be performed as a control and will allow purely structural roughness to be separated from the "charge" roughness, implicated in models of fatigue, for which, at present, there is no other probe. These scattering experiments are being performed at the Univ. of Wisconsin Synchrotron Radiation Center. These principles can be extended to organic materials using resonant features at the carbon K x-ray absorption making possible the distinction of local structure based on purely chemical contrast.