2
Multiferroic Materials

Wanjun Peng, Ziyao Zhou, and Ming Liu

Xi'an Jiaotong University, Electronic Materials Research Laboratory, Key Laboratory of the Ministry of Education and International Center for Dielectric Research, 28 Xianning West Road, Xi'an, 710049, China

2.1 Introduction

Ferromagnets, generated by the spontaneous, uniform orientation of atomic or molecular magnetic moments, have been investigated for more than 2500 years and just unfolding [1]. Ferroelectricity, named so based on the likeness to ferromagnetism, was discovered merely a century ago. Multiferroic materials [2] with the coexistence of at least two ferroic orders have recently aroused ever-growing attention because of their potential for significantly broadening applications. On the one hand, they combine the performances of two materials that were formerly separate from different fields. On the other hand, the coupling interaction between the various states can produce added functionalities not present in either state alone, such as the magnetoelectric () effect discovered more than a century ago [3,4].

The ME response can be divided into two categories. One is the direct ME effect, which produces an electric polarization P by applying a magnetic field H:

where E denotes the electric field and a(aE) is the ME (ME voltage) coefficient.

The other is the simultaneously converse ME effect, that is, the emergence of magnetization M upon applying an electric field E:

In a multiferroic material, where ferroelectric, ferromagnetic, as well as a strong enough ME coupling coexist ideally, the electric (magnetic) polarization-magnetic (electric) field curves (P-H or M-E curves) would show a hysteresis response, as schematically demonstrated in Figure 2.1, which resemble the celebrated ferroelectric or ferromagnetic hysteresis loops.

Figure 2.1 Schematic illustration of an ideal case of the multiferroics [1] . There is a P-H or M-E hysteresis loop similar to the celebrated ferroelectric or ferromagnetic hysteresis loops.

From the viewpoint of physical architectures, multiferroic materials can be classified into two types: single phase and composite. The intrinsic ME coupling exists in some natural monophasic substances and has been found in more than 10 compounds so far such as BiFeO3 () and rare earth manganates. However, the applications of most single-phase compounds are strictly limited due to the low Curie temperatures (below room temperature) and a weak inherent ME coupling (especially above room temperature).

Alternatively, multiferroic ME composites [5-7] combining ferroelectric and ferromagnetic phases have been gradually coming into view and have become "hot." In ME composites, remarkable ME coupling can be produced because of the cross-interaction between the phases although neither of the constituent phases has ME effect, which was first proposed by Van Suchtelen as a product tensor property [7] . Generally, the composite ME coupling is the product effect of the magnetostrictive effect (magnetic/mechanical effect) in the ferromagnetic phase and the piezoelectric effect (mechanical/electrical force) in the ferroelectric one, namely [8],

Overall, ME composites have much larger ME effect at room temperature than single-phase compounds, which makes the practical application of multiferroic materials a significant step forward. Thus, various ME composites have been investigated recently, from bulk ME composites to thin films.

However, the research history of ME materials is not smooth sailing. In 1894, Curie pointed out through symmetry analysis that there might be intrinsic ME coupling effect in some crystals. In 1961, American scientists first reported the essential ME effects observed in Cr2O3 at low temperatures, which led to a small climax in the early studies of ME effects in the 1970s. At the same time, the concept and materials of ME composite appeared for the first time. However, due to the lack of practical applications, the limitation of low-temperature conditions, and the complexity of the coupling mechanism involved, all related studies entered a low glacial period of nearly 30 years. In the recent 10 years, with the tremendous progress of material preparation technology, characterization means, and theoretical calculation, as well as the urgent need for new information functional devices in the modern information society, research on multiferroic materials and methods has witnessed unprecedented rapid development [9].

2.2 Single-Phase Multiferroics

According to the mechanism of formation, we can divide many magnetoelectric multiferroic materials into four main categories [10]:

1. (1) Perovskite-type compounds such as BFO, BiMnO3 (), and PbFe1/2Nb1/2O3. In these materials, the only pair of 6s electrons of Bi ions at A site of perovskite provides a ferroelectric order, while the 3d transition metal electrons at B site provide a spin order. In general, the ferroelectric properties of these materials are first-order order parameters, which make the electric field actively interact with the spin.
2. (2) Structural dislocation materials, commonly known as rare earth manganites such as RMnO3 (R = Sc, Y, etc.) and RMn2O5 (R = Y, Tb, etc.). In this kind of material, the ferroelectric polarization temperature is usually higher than room temperature, but the ME coupling can be formed at antiferromagnetic Neil temperature (~70 K).
3. (3) Magnetic charge ordered materials, represented by LuFe2O4. The ionic non-centrosymmetry in such materials leads to ferroelectric polarization.
4. (4) Multiferroic bulk materials, represented by TbMnO3 and DyMnO3. Their ferroelectricity is based on the long-range order of spin in magnetic materials.

Undoubtedly, the most widely studied single-phase ferroelectric material is ABO3-type perovskite oxide. In ferroelectric materials with this structure, most of the ferroelectricity originates from B-position ions located in the center of the oxygen octahedron, which deviates from the center of the oxygen octahedron below Curie temperature, reducing the symmetry of crystal structure and separating the positive and negative charge centers to form electric dipole moments. Generally, ABO3-type perovskite ferroelectrics have no electron occupation in the d orbit of B ions and behave as d0 states. On the contrary, it is impossible to produce any magnetic order because of the absence of local magnetic moments resulting from vacancies in electrons in the d orbit, which indicates that the mechanisms of conventional ferroelectric and magnetic orders are mutually exclusive at the atomic scale. Therefore, an additional driving force that satisfies both the structural symmetry condition of ferroelectric crystals and the electronic shell structure condition of magnetic crystals is vital.

From past research, there is no doubt that the most striking single-phase multiferroic is BFO, in bulk, nanoparticles, or in thin films [11]. BFO is the only material with ferroelectric Curie temperature and antiferromagnetic Nile transition temperature much higher than room temperature (which can realize ferroelectricity at room temperature and coexist with antiferromagnetism), and also with strong ME coupling characteristics, which thus can achieve field-controlled magnetization. The research upsurge originated from the study of BFO thin films epitaxially grown on (001) SrTiO3 () single crystal substrates by pulsed laser deposition () reported by Ramesh's group [12] in 2003, as shown in Figure 2.2. In this study, for the first time, they observed remarkable ferroelectric properties with full electrochemical strength Ps = 50-60 µC cm-2 and magnetization Ms = 150 emu cm-3. It is understood that the ferroelectric order and magnetic order in BFO originate from the contributions of different ions, namely, Bi ions and Fe ions, which leads to weak intrinsic coupling. For example, the ME coefficient of La-doped BFO thin films calculated by Jang et al. [13] is about 10 mV cm-1. Subsequently, Zhao et al. [14] succeeded in observing one-to-one correlations and coupling relationships between ferroelectric and antiferromagnetic orders in BFO by piezoelectric force microscopy () and X-ray magnetic circular dichroism-X-ray photoemission electric microscopy (), as shown in Figure 2.3. On this basis, ME coupling is only found at specific polarization reversal (e.g. 109° and 71°) [15]. Eom and coworkers [16] realized selective polarization reversal on nanoscale by scanning electric probe field, which laid the foundation for electrical field control of the antiferromagnetic direction in rhombic phase BFO. In addition, since the tetragonal BFO epitaxial crystals were synthesized, the ME effect has attracted widespread attention. Yang and...