Chemical potential landscape in band filling and bandwidth-control of manganites: Photoemission spectroscopy measurements

Abstract

We have studied the effects of band filling and bandwidth control on the chemical potential in perovskite manganites R1−xAxMnO3 (R : rare earth, A : alkaline earth) by measurements of corelevel photoemission spectra. A suppression of the doping-dependent chemical potential shift was observed in and around the CE-type charge-ordered composition range, indicating that there is charge self-organization such as stripe formation or its fluctuations. As a function of bandwidth, we observed a downward chemical potential shift with increasing bandwidth due to the reduction of the orthorhombic distortion. After subtracting the latter contribution, we found an upward chemical potential shift in the ferromagnetic metallic region 0.3 < x < 0.5, which we attribute to the enhancement of double-exchange interaction involving the Jahn-Teller-split eg band.PACS numbers: 75.47. Lx, 75.47.Gk, 71.28.+d, Perovskite-type manganites with the formula R 1−x A x MnO 3 , where R is a rare-earth and A is an alkaline-earth metal, exhibit a complex phase diagram as a function of band filling and bandwidth, and competition between these phases leads to remarkable phenomena such as colossal magnetoresistance and the spin, charge and orbital ordering of Mn 3d e g electrons as shown in Fig. 1 [1 , 2, 3, 4]. Key features to understand the complex phase diagram are double-exchange interaction and the instabilities towards spin, charge and orbital ordering. Systems with small and intermediate bandwidths such as Pr 1−x Ca x MnO 3 (PCMO) and Nd 1−x Sr x MnO 3 (NSMO) exhibit the so-called CE-type antiferromagnetic (AF) charge-ordered (CO) phase in the doping range around half-doping x = 0.5, but this tendency disappears in systems with large bandwidths such as La 1−x Sr x MnO 3 (LSMO). This clearly demonstrates the importance of band filling and bandwidth to understand the physical properties of the manganites.The electron chemical potential µ is one of the most fundamental physical quantities of strongly correlated electron systems. The shift of µ as a function of electron density n corresponds to the charge compressibility κ or the charge susceptibility χ c through κ = (1/n 2 )(∂n/∂µ) or χ c = ∂n/∂µ and can be measured through the shifts of photoemission spectra since binding energies in the spectra are experimentally referenced to µ, namely, the Fermi level. Since n is determined by the number of electrons in a unit cell and the unit cell volume and the bandwidth are closely related with each other, it is highly important to study the chemical potential shift ∆µ as a function of both band filling and bandwidth. Recently, suppression of ∆µ as a function of hole doping has been observed in and near the CE-type CO composition range of PCMO and its correlation with the changes of the periodicity of the stripe fluctuations has been pointed out [5]. Pinning of chemical potential by static or dynamic stripe fluctuations in La 2−x Sr x CuO 4 and La 2−x Sr x NiO 4 has also been observed in the composition range where the periodicity of charge order changes ...

In situphotoemission study ofNd1−xSrxMnO3epitaxial thin films

“…The band dispersions obtained by ARPES were almost the same as those of LSMO thin films. These results are almost the same as the previous results of bulk samples 8,14 and surely demonstrate that the electronic structure of the NSMO thin films on STO substrates was almost the same as the bulk samples due to the effects of relaxation. The reduction in the ferromagnetic transition temperature compared to LSMO cannot be ascribed to the decrease in the bandwidth but to the decrease in the coherent spectral weight of the e g band.…”

“…From core-level photoemission studies, we found that the behavior of the chemical-potential shift was not much different from that of the bulk NSMO samples. 8 The band dispersions obtained by angle-resolved photoemission spectroscopy ͑ARPES͒ were almost the same as those of LSMO thin films. 9 These results show that the electronic structure of the NSMO thin films were similar to those of the bulk samples due to the effects of relaxation.…”

“…Mn 2p moves in the same direction as all the other core levels, in contrast to bulk samples. 8 This difference of the behavior of Mn 2p has already been observed in LSMO bulk 12 and thin films on STO substrates, 7 and the origin of this difference is not clear at this moment but probably related to in-plane tensile strain effects from STO substrates. The chemical-potential shift ⌬ can be obtained from the average of the energy shifts of the O 1s and Sr 3d core levels.…”

“…The chemical-potential shift ⌬ can be obtained from the average of the energy shifts of the O 1s and Sr 3d core levels. 8 Figure 2͑b͒ shows ⌬ thus determined plotted together with the bulk NSMO results. 8 The ⌬ of thin films is in good agreement with that of bulk samples.…”

“…8 Figure 2͑b͒ shows ⌬ thus determined plotted together with the bulk NSMO results. 8 The ⌬ of thin films is in good agreement with that of bulk samples. In the bulk samples, there is a suppression of ⌬ in the narrow composition region around x = 0.5, that is, 0.45Յ x Յ 0.55, corresponding to the narrow region of CO in the phase diagram of NSMO.…”