ieee transactions on ultrasonics, ferroelectrics, and frequency control, draft 1 Dielectric Spectroscopy of Nano BaTiO 3 Confined in MCM-41 Mesoporous Molecular Sieve Materials M. Kinka, Juras Banys, W. B¨ ohlmann, E. Bierwirth, M. Hartmann, D. Michel, G. V¨ olkel, and A. P¨ oppl [AU: We need first name for all authors.] ⊲ Abstract—Dielectric properties of barium titanate (BaTiO3) particles, synthesized directly in the pores of MCM-41 materials, have been investigated in the frequency range from 20 Hz to 1 MHz for temperature intervals from 100 K to 500 K. The dielectric spectra of BaTiO3 con- fined in these molecular sieves were compared with the results obtained from the investigation of pure MCM-41 materials. Obtained results confirmed successful incorpora- tion of BaTiO3 into porous matrix, but no phase transition from paraelectric to ferroelectric phase was observed due to the particle size being smaller than the critical size. Also, the overall dielectric response of investigated materials is strongly influenced by adsorbed water molecules. I. Introduction T he tendency of miniaturization in electronics con- stantly produces the need of new materials— composites and, of course, new ways for use of well-known ones. Barium titanate (BaTiO 3 ) is a common ferroelectric material, widely used for producing differently electroni- cal components. Above the Curie temperature of 120 ◦ C, a “bulk” BaTiO 3 crystal is in paraelectric phase and is characterized as a cubic perovskite material. On cooling, it undergoes three displacive-type ferroelectric phase tran- sitions, being in ferroelectric phase at room temperature and having a dielectric constant ε ′ of approximately 3000– 5000 [1]. A major field of current research on ferroelectrics is concerned with the size effect—complete vanishing of ferroelectricity below a certain critical size of the par- ticles. This critical size may amount to a few or some tens of nanometers. It was recently reported to depend on temperature [2]. Although these effects are in general predictable by an extended Landau-Ginzburg theory, it is not yet feasible to produce reliable numerical predictions because we are not dealing with ideal systems but with a complicated interplay of partially delicate and hard-to- control influences, such as sample preparation, free charge Manuscript received July 31, 2005; accepted November 17, 2005. This work was supported in part by the Alexander von Humboldt Stiftung. M. Kinka and J. Banys are with the Physic Faculty, Vilnius Uni- versity, [AU: What city?] Sauletekio 9, 10222 Lithuania (e-mail: ⊲ Robertas.grigalaitis@ff.vu.lt). W. B¨ohlmann, E. Bierwirth, M. Hartmann, D. Michel, G. V¨olkel, and A. P¨oppl are with the Institute of Experimental Physic II, Leipzig University, [AU: What city?] Germany. ⊲ M. Hartmann also is with the University of Kaiserslautern, Kaiser- slautern, Germany. layers on surfaces, shape and size distribution, and others. Therefore, it is desirable to enforce the experimental ba- sis of this rather young branch of science, at best with a maximum set of controlled or controllable parameters. One idea is to enclose the ferroelectric sample into matrices of porous materials to start from a system of known geome- try. By this, geometry-induced influences are reduced, and we are able to concentrate on parameters that are better manageable. Such nanostructures in porous matrices have several advantages compared to systems of quasi-isolated ferroelectric particles, e.g., the stability of their properties due to confinement, which prevents harmful influence of the atmosphere and at the same time allows the prepara- tion of larger quantities of nanomaterials (at least several cubic centimeters). The latter is hardly achievable by con- ventional methods of nanostructure preparation, such as nanolithography and molecular beam epitaxy. Mesoporous MCM-41 materials are very suitable and promising me- dia for such investigations because of their unique struc- tural and chemical properties. This class of silica tube-like materials features extremely large specific surface areas exceeding 1000 m 2 g −1 and possess uniformly sized meso- pores whose diameter can be controlled between 1.5 and 8 nm by the synthesis conditions chosen and template used [3], [4]. The silicon can be partially substituted by various metal ions [5] in order to modify the chemical (hydrophilic) properties of the inner surface of the molecular sieve mate- rials. Recently also other attempts to use porous materials in the fabrication of ferroelectric “nanotubes” have been reported [6]–[8].[AU: References must be cited in or- ⊲ der.] It should be emphasized that nanostructures within porous matrices can be studied by the same experimental techniques used for bulk ferroelectric substances. If ferro- electric materials are imbedded into pores, however, not only the influence of interactions between the nanoparti- cles but also of those with the inner surface of the porous matrices has to be understood. II. Sample Preparation and Measurement Method Samples containing BaTiO 3 were prepared similar to the procedure proposed by Hernandez et al. [9]. Two types of mesoporous MCM-41 materials were synthesized using 0885–3010/$20.00 c  2006 IEEE