Ferroelectric ferromagnets are exceedingly rare, fundamentally interesting multiferroic materials that could give rise to new technologies in which the low power and high speed of field-effect electronics are combined with the permanence and routability of voltage-controlled ferromagnetism. Furthermore, the properties of the few compounds that simultaneously exhibit these phenomena are insignificant in comparison with those of useful ferroelectrics or ferromagnets: their spontaneous polarizations or magnetizations are smaller by a factor of 1,000 or more. The same holds for magnetic- or electric-field-induced multiferroics. Owing to the weak properties of single-phase multiferroics, composite and multilayer approaches involving strain-coupled piezoelectric and magnetostrictive components are the closest to application today. Recently, however, a new route to ferroelectric ferromagnets was proposed by which magnetically ordered insulators that are neither ferroelectric nor ferromagnetic are transformed into ferroelectric ferromagnets using a single control parameter, strain. The system targeted, EuTiO(3), was predicted to exhibit strong ferromagnetism (spontaneous magnetization, approximately 7 Bohr magnetons per Eu) and strong ferroelectricity (spontaneous polarization, approximately 10 microC cm(-2)) simultaneously under large biaxial compressive strain. These values are orders of magnitude higher than those of any known ferroelectric ferromagnet and rival the best materials that are solely ferroelectric or ferromagnetic. Hindered by the absence of an appropriate substrate to provide the desired compression we turned to tensile strain. Here we show both experimentally and theoretically the emergence of a multiferroic state under biaxial tension with the unexpected benefit that even lower strains are required, thereby allowing thicker high-quality crystalline films. This realization of a strong ferromagnetic ferroelectric points the way to high-temperature manifestations of this spin-lattice coupling mechanism. Our work demonstrates that a single experimental parameter, strain, simultaneously controls multiple order parameters and is a viable alternative tuning parameter to composition for creating multiferroics.

Nature 466, 954-958 (2010) This Letter determined that EuTiO 3 , when appropriately strained, becomes a strong ferroelectric ferromagnet, in agreement with prediction. Strong ferroelectrics are proper ferroelectrics, having polarization as their order parameter, with high paraelectric-to-ferroelectric transition temperatures (T c ). Such ferroelectrics are manifested by a high T c and a high peak at T c in the dielectric constant versus temperature behaviour, signifying that ferroelectricity is driven by the soft mode, which is indicative of proper ferroelectricity. Our measurements of strained EuTiO 3 demonstrate both of these characteristics (shown in Fig. 3 of our Letter), and led us to conclude that strained EuTiO 3 is a strong ferroelectric. In contrast, all well-established prior single-phase ferroelectric ferromagnets are improper or pseudoproper ferroelectrics (that is, with weak ferroelectricity resulting in minuscule P s ). We did not present P s values in our Letter. Second harmonic generation measurements do not provide quantitative values of P s and attempts to determine P s via pyroelectric measurements (Yan, L., Li, J. F. & Viehland, D., personal communication)1 resulted in unphysically high values, presumably owing to electrical leakage. Nonetheless, the magnitude of the P s of our strained EuTiO 3 films can be estimated as follows. In their classic work, Abrahams, Kurtz, and Jamieson 1 established a correlation between P s and T c for displacive ferroelectrics. By studying numerous displacive ferroelectrics they found 5) where T c is the paraelectric-to-ferroelectric transition temperature in K, Dz is the atomic displacement of the 'homopolar' metal atom in Å , and P s is the spontaneous polarization of the ferroelectric in mC cm 22 . Combining these equations to eliminate Dz allows P s to be estimated from T c in displacive ferroelectrics. The huge anomaly of the soft optical phonon near T c that we observe ( Supplementary Fig. 1 of our Letter) shows that strained EuTiO 3 is a displacive ferroelectric, making the aforementioned correlation applicable. Plugging in our measured value of T c (Fig. 3 in our Letter) yields P s 5 29 6 2 mC cm 22 for our strained EuTiO 3 films from this established correlation. This rough estimate is consistent with our first-principles theoretical predictions-P s 5 21 mC cm 22 for EuTiO 3 under 11.1% biaxial tension, corresponding to the strain of our commensurate EuTiO 3 films grown on (110) DyScO 3 . Thus, the data in our Letter shows that appropriately strained EuTiO 3 is a strong ferroelectric ferromagnet.

Lanthanum lutetium oxide thin films were grown on ͑100͒ Si by pulsed laser deposition. Rutherford backscattering spectrometry, atomic force microscopy, x-ray diffraction, and x-ray reflectometry were employed to investigate the samples. The results indicate the growth of stoichiometric and smooth LaLuO 3 films that remain amorphous up to 1000°C. Internal photoemission and photoconductivity measurements show a band gap width of 5.2± 0.1 eV and symmetrical conduction and valence band offsets of 2.1 eV. Capacitance and leakage current measurements reveal C-V curves with a small hysteresis, a dielectric constant of Ϸ32, and low leakage current density levels. © 2006 American Institute of Physics. ͓DOI: 10.1063/1.2393156͔The study of ultrathin gate dielectrics has recently gained great attention due to the technological need to replace SiO 2 films in metal-oxide-semiconductor field-effect transistors ͑MOSFETs͒. 1 The scaling has led to MOSFETs with ultrashort physical gate lengths ͑Ͻ50 nm͒ and insulating SiO 2 -based films with thickness less than 1 nm. At such a thickness, these films suffer from excessively high leakage of charge carriers and poor reliability with respect to dielectric breakdown. Therefore, to overcome these limitations new gate dielectric materials with a higher dielectric constant must be developed to replace the SiO 2 . According to the International Technology Roadmap for Semiconductors, 2 the implementation of high-gate dielectrics with a dielectric constant between 10 and 20 will be required by 2008, which will later be replaced by materials having a larger than 20, in order to meet both low leakage current density and performance requirements.Ternary rare earth oxides ͑e.g., DyScO 3 and GdScO 3 ͒ are emerging as promising candidates for high-applications. As shown by Schlom and Haeni, 3 single crystals of these oxides show values of 20-35 which were also observed for amorphous LaScO 3 , GdScO 3 , and DyScO 3 films deposited on silicon ͑ =22-23͒. 4,5 In addition, these materials fulfill the requirements for large optical band gaps ͑5.6 eV͒ and band offsets ͑2 -2.5 eV͒, 6 while their amorphous phase is stable up to 1000°C ͑for GdScO 3 and DyScO 3 ͒. 4,5 Lanthanum lutetium oxide ͑LaLuO 3 ͒, as a member of this class of ternary oxides, is predicted to have similar properties. 3,7 Experimental data related to high-gate applications of amorphous LaLuO 3 films are, however, still not available. In this letter, we present the results of a systematic study on the microstructural and electrical properties of amorphous LaLuO 3 thin films, deposited on silicon substrates by means of pulsed laser deposition ͑PLD͒.LaLuO 3 films were deposited by PLD using a stoichiometric ceramic target. The target was made by milling a stoichiometric mixture of Lu 2 O 3 ͑Alfa Aesar, 99.99%͒ and La 2 O 3 ͑Alfa Aesar, 99.999%͒ powders with a molar ratio of 1:1. The ground powder was dried and then fired at 1300°C in air for 12 h. After regrinding, the powder was pressed with a uniaxial press ͑3 tons͒. The pellets were the...

Phase-pure, stoichiometric, unstrained, epitaxial ͑001͒-oriented EuTiO 3 thin films have been grown on ͑001͒ SrTiO 3 substrates by reactive molecular-beam epitaxy. Magnetization measurements show antiferromagnetic behavior with T N = 5.5 K, similar to bulk EuTiO 3. Spectroscopic ellipsometry measurements reveal that EuTiO 3 films have a direct optical band gap of 0.93Ϯ 0.07 eV.

No abstract

Crystalline Dy 2 O 3 , Sc 2 O 3 , and amorphous high-k DyScO 3 thin films have been deposited on Si͑100͒ substrates by metallorganic chemical vapor deposition ͑MOCVD͒ using two metal precursors M͑EDMDD͒ 3 ͓MϭDy, Sc; EDMDDϭ6-Ethyl-2,2-Di Methyl -3,5-Decane Dionato͔. The precursors were evaluated in terms of efficiency and growth rate under various conditions, viz. vaporizer and susceptor temperatures, reactor pressure, injection rate, and injection delay between the two precursors. Amorphous DyScO 3 films with nearly correct stoichiometry were deposited within the temperature range of 560-700°C. These amorphous films were smoother than the crystalline binary oxides and reached a density of around 85% of the bulk crystalline density. Amorphous structure and surface smoothness retained up to an annealing temperature as high as 950°C. The thickness of the SiO x interlayer did not vary with deposition temperature, but annealing at temperatures above 900°C increased the interlayer thickness. Electrical properties are promising; the dielectric constant of DyScO 3 ͑k Ϸ 22͒ is much higher than that of the binary oxides Dy 2 O 3 and Sc 2 O 3 ͑k Ϸ 10͒, and the leakage currents are very low compared to SiO 2 .

Dysprosium scandate ͑DyScO 3 ͒ thin films were deposited on Si substrates using metal-organic chemical vapor deposition. Individual source precursors of Dy and Sc were used and deposition temperatures ranged from 480 to 700°C. Films were amorphous with low root mean square roughness ͑ഛ2 Å͒ and were stable up to 1050°C annealing. Electrical characterization yielded C-V curves with negligible hysteresis ͑Ͻ10 mV͒, high dielectric constant ͑ϳ22͒, and low leakage currents. The electrical properties of the DyScO 3 / SiO x / Si stacks were stable up to 800°C for films on native oxide; however, this limit increased to 900°C for films on special chemically grown oxide, suggesting further improvement with proper diffusion barrier. © 2006 American Institute of Physics. ͓DOI: 10.1063/1.2402121͔ Dielectric materials with sufficiently high dielectric constant ͑high k͒ are needed as early as 2007 ͑Ref. 1͒ to replace SiO 2 / SiON in the future complementary metal oxide semiconductor ͑CMOS͒ technology with a subnanometer equivalent oxide thickness. Structural invariance of the thin gate oxide layer within CMOS process thermal budget is preserved with SiO 2 ͑ = 3.9͒ and now with SiON ͑ = 7.0͒. If one considers this as a major criteria the presently favored group IVB oxides, HfO 2 and ZrO 2 , have to be alloyed with high concentrations of Si or Al in order to stabilize the amorphous structure; however, this alloying reduces the dielectric constant ͑k͒ to the values of r ϳ 10-12 ͑Ref. 2͒ for the silicates of Hf and Zr. Hence, this silicate with medium k may facilitate a smooth transition from conventional gate dielectrics to high-k dielectrics ͑k Ͼ 20͒. In order to reach this target another class of dielectrics with even higher dielectric constants and lower leakage currents is needed. Rare-earth based multicomponent oxides in the amorphous state are being considered as the next generation dielectrics after the silicates. 3 Reported work on these types of materials mainly concentrated on the pulsed laser deposition/ molecular-beam or electron beam deposition and these reports suggest a very high potential for these materials in the semiconductor industry. [4][5][6][7] An approach by atomic layer deposition of YScO 3 and metal-organic chemical vapor deposition ͑MOCVD͒ of DyScO 3 films were reported recently. 8,9 Our present effort was to realize films of the rareearth scandate, DyScO 3 , using advanced precursors in an industry friendly high throughput MOCVD process.In this work, DyScO 3 film depositions were carried out in a liquid injection MOCVD reactor ͑Aixtron 2600G3 which can handle five 6 in. wafers simultaneously equipped with a liquid delivery Trijet vaporizer͒. 10 Dy͑EDMDD͒ 3 and Sc͑EDMDD͒ 3 precursors ͑EDMDD = 6-ethyl-2,2-dimethyl-3,5-decane dionato͒ were supplied by Asahi Denka, Japan, and were dissolved in octane with 0.05M concentration. This precursor solution was injected by a TRIJET system with a flow rate in the micro-liter regime. Film depositions were carried out in the temperature range of 480-700°C and the reac...

No abstract