Article pubs.acs.org/IC
Enhancement of Ferroelectricity for Orthorhombic (Tb0.861Mn0.121)MnO3−δ by Copper Doping Jianming Deng,† Muhammad Asim Farid,‡ Meng Zhang,§ Aimei Yang,† Hongxing Zhang,‡ Hao Zhang,‡ Gengfang Tian,∥ Meimei Wu,∥ Laijun Liu,† Junliang Sun,‡ Guobao Li,*,‡ Fuhui Liao,‡ and Jianhua Lin*,‡ ‡
Beijing National Laboratory for Molecular Sciences, State Key Laboratory of Rare Earth Materials Chemistry and Applications, College of Chemistry and Molecular Engineering, Peking University, Beijing 100871, People’s Republic of China † College of Materials Science and Engineering, Guilin University of Technology, Guilin 541004, People’s Republic of Chinaa § Department of Chemistry, School of Science, Beijing Jiaotong University, Beijing 100044, People’s Republic of China ∥ Neutron Scattering Laboratory, Department of Nuclear Physics, China Institute of Atomic Energy, Beijing 102413, People’s Republic of China S Supporting Information *
ABSTRACT: Copper-doped (Tb0.861Mn0.121)MnO3−δ has been synthesized by the conventional solid state reaction method. X-ray, neutron, and electron diffraction data indicate that they crystallize in Pnma space group at room temperature. Two magnetic orderings are found for this series by neutron diffraction. One is the ICAM (incommensurate canted antiferromagnetic) ordering of Mn with a wave vector qMn = (∼0.283, 0, 0) with a ≈ 5.73 Å, b ≈ 5.31 Å, and c ≈ 7.41 Å, and the other is the CAM (canted antiferromagnetic) ordering of both Tb and Mn in the magnetic space group Pn′a21′ with a ≈ 5.73 Å, b ≈ 5.31 Å, and c ≈ 7.41 Å. A dielectric peak around 40 K is found for the samples doped with Cu, which is higher than that for orthorhombic TbMnO3.
1. INTRODUCTION Multiferroic materials LnMnO3 (Ln = Y, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) have attracted considerable interest for their intriguing physics and potential application in memories, sensors, and transducers based on coexistence of magnetic and electric order.1−3 Strong magnetoelectric coupling found in orthorhombic TbMnO3 below 27 K4,5 is very interesting and has inspired a lot of studies. This coupling is attributed to the fact that the ferroelectricity of TbMnO3 is induced by a spiral magnetic order, which is related to the distorted orthorhombic perovskite (ABO3) structure of TbMnO3.6 The corresponding distortion is exhibited by cooperative tilts of the MnO6 octahedra due to the active Jahn−Teller effect of Mn3+ with a 3d4 configuration and enhanced by the small size of the Tb3+ cations.7 The average size of the A (Tb) site can be modified by doping the trivalent cations Ga3+, Y3+, or Nd3+, which in turn can modify the octahedral distortion.8−10 On the other hand, replacing Tb by divalent cations such as Sr2+ or Ca2+ can both change the average size of the A (Tb) site and reduce the Jahn−Teller effect of the B (Mn) site by converting some Mn3+ into Mn4+.11,12 Several tries have also been performed by introducing Sc,13 Cr,14 Fe,15 Co,16 or Ru17 to the Mn site, which may reduce the Jahn−Teller effect of the Mn site. In particular, Mn3+ doped with Cu2+, which is also a Jahn−Teller ion, may enhance the octahedral distortion.18 However, the © 2017 American Chemical Society
ferroelectric ordering temperature indicated by the dielectric peak in the curve of the temperature-dependent dielectric constant is suppressed after these modifications. These results of the ferroelectric ordering temperature in bulk orthorhombic phase constrained the applications of magnetoelectric effects in TbMnO3 to low working temperatures. How to raise the ferroelectric ordering temperature in the bulk TbMnO3 is a challenge. In our previous work, it is found that part of Tb in TbMnO3 can be replaced successfully by Mn to form a solid solution (Tb1−xMny)MnO3.19 Although dielectric constant measurements for (Tb1−xMny)MnO3 exhibit that the ferroelectric ordering temperature is suppressed when a small amount of Mn is doped into the Tb site.20 A new commensurate antiferromagnetic ordering appeared along with the incommensurate magnetic ordering in (Tb1−xMny)MnO3 when a large enough quantity of Mn was doped into the Tb site.21 In this case the dielectric peak was very hard to observe, which indicates that the multiferroicity was inhibited.20 We postulate that the appearance of the commensurate antiferromagnetic ordering breaks the multiferroicity of (Tb1−xMny)MnO3. Is it possible to tune this nonmultiferroic phase to a multiferroic phase again? After careful study a Received: December 12, 2016 Published: February 27, 2017 3475
DOI: 10.1021/acs.inorgchem.6b03024 Inorg. Chem. 2017, 56, 3475−3482
Article
Inorganic Chemistry
at the Tb site (more details are presented in the following discussion). The sample C1 has been suggested to crystallize in the space group Pnma as reported previously.19 Selected area electron diffractions (SAED) of C2−C6 were still obtained and carefully analyzed to make sure the chosen space group Pnma is acceptable because the choice of the space group for a perovskite compound should be very careful.25−27 The SAED patterns of C2−C6 are similar to those shown in Figure 2 for C3, which confirms that the choice of the space group Pnma is acceptable. Therefore, the same structure is suggested to be C1, C2, C3, C4, C5, and C6. The powder X-ray diffraction data for C1−C7 around room temperature can be refined well using the space group Pnma with Rwp < 0.032 and Rp < 0.021. The typical Rietveld plots of X-ray and neutron diffraction data for C5 are shown in Figure 3 (refinement details are listed in the Supporting Information. The data for other samples are also listed in the Supporting Information). Figure 3 shows one reflection around 60° in the neutron diffraction data for C5 marked by an asterisk (*) related to the impurity Mn3−x−yTbxCuyO4 with the structure of Mn3O4 in space group I41/amd. As shown in Figure 4, with the increase of Cu in the sample the lattice parameters b and c first increase to a maximum and then decrease. This is different from that expected by Vegard’s law28 and may be attributed to the fact that the ratio of Cu occupied at Tb and Mn sites changes with the increase of Cu in the samples. For example, one may suggest that when x < 0.04 the Cu may mainly occupy the site of Mn in the samples (maybe a small amount of Cu may occupy the site of Tb with a neglectable effect on lattice parameter). As indicated by the XPS data (see Figure 5), Tb3+, Cu2+, Mn3+, and Mn4+ ions are presented in the samples. To balance the compound (Rietveld refinement of the neutron diffraction data for C1, C3, and C5 indicates that the oxygen sites are almost fully occupied), it is expected that Cu2+ + Mn4+ may replace 2 Mn3+. It is suggested29 that the radius of Cu2+, Mn4+, and Mn3+ with a six-coordinate state is 0.87, 0.67, and 0.785 Å (in high-spin state (HS) or 0.72 Å in low-spin state (LS)). At present, Mn3+ is mainly in the high-spin state (see the Magnetic Property section below). Then one may expect that the lattice parameters a and b should decrease, and c should increase as shown for the solid solution EuMn1−xCuxO330 because Cu2+ + Mn4+ is smaller than 2 Mn3+ (HS) in the six-coordinate state, which disagrees with the results shown in Figure 4. Therefore, it is better to suggest that when x < 0.04, Cu mainly occupies the site of Tb, which is about eight coordinated by oxygen atoms. It is supposed that Cu2+ + Mn4+ may be larger than 2 Mn3+ (HS) in the eight-coordinate state. Then the increase of the lattice parameters with the increase of Cu in the compound can be understood. Of course, the present result does not mean that there is not any Cu in the site of Mn. Further studies are needed to get a clear understanding. As reported for the solid solution (Tb1−xMny)MnO3,19 the lattice parameters a and c decrease with the increase of Mn in the site of Tb while the lattice parameter b increases slightly. The decrease of a and c may be due to the fact that the radius of Mn3+ is smaller than that of Tb3+. The increase of b is therefore attributed to the further structure distortion caused by the change of the average radius of the ions in the site of Tb. In the present case, some of the A-site Mn3+ is replaced by Cu2+ and Mn4+, the increases of the lattice parameters a and c are reasonable because we suggest that Cu2+ + Mn4+ may be larger than 2 Mn3+ (HS) in the eight-
multiferroic phase has been obtained by doping Cu into (Tb1−xMny)MnO3 with the dielectric peaks even moving to high temperature. This interesting result is presented below.
2. EXPERIMENTAL SECTION Samples with a nominal formula (Tb0.84Mn0.16)Mn1−xCuxO3−δ (x = 0.00, 0.025, 0.05, 0.075, 0.10, 0.15, and 0.20 named C1, C2, C3, C4, C5, C6, and C7, respectively) were synthesized from stoichiometric amounts of Tb4O7 (99.95%), CuO (A.R.), and MnCO3 (A.R.). The oven-dried reagents were mixed and homogenized by about 30 min of grinding of mixtures with an agate mortar and a pestle. The mixtures were subjected to 6 h calcination at 800 °C. They were then pressed into pellets to undergo four heat treatments, each for 12 h at 1200 °C, every time followed by a furnace cooling with intermediate grinding. All treatments were done in air. Powder X-ray diffraction (PXRD) data were collected on a PANalytical x’Pert3 Powder with Cu Kα (λ1 = 0.15405 nm and λ2 = 0.15443 nm) radiation (2θ range 5−120° for 10 h; step 0.01313) at 40 kV and 40 mA at room temperature. Powder neutron diffraction patterns for C3 and C5 were collected at temperatures ranging from 1.5 to 300 K on the G4.1 diffractometer installed at the Orphée reactor (Laboratoire Léon Brillouin, CEACNRS Saclay, France). The incident neutron wavelength was 2.427 Å. The X-ray and neutron diffraction data were analyzed using GSAS22,23 and Fullprof24 software. The magnetic properties were investigated with a Cryogenic physical property measurement system (PPMS) from 2 to 300 K. Selected area electron diffractions (SAED) were carried out on a JEM2100F with an accelerating voltage of 200 kV. The X-ray photoelectron spectroscopy (XPS) patterns were acquired with a UK Kratos Axis Ultra spectrometer with an Al Kα X-ray source operated at 15 kV, 15 mA. The chamber pressure was less than 5.0 × 10−9 Torr. Electron binding energies were calibrated against the C 1s emission at Eb = 284.8 eV. The dielectric properties of the samples were measured using a Precision Impedance Analyzer 65120B (Wayne Kerr Electronics).
3. RESULTS AND DISCUSSIONS 3.1. Structural Evolution. Powder X-ray diffraction patterns of C1−C7 collected at room temperature are very similar except small peaks corresponding to impurity found for C6 and C7 as shown in Figure 1. The composition of C1 was
Figure 1. Powder X-ray diffraction patterns of the samples with the nominal formula (Tb0.861Mn0.121)Mn1−xCuxO3−δ: (*) reflections for the impurity.
determined as (Tb0.861Mn0.121)MnO3 using chemical titration and the combined refinement of the X-ray and neutron diffraction data as reported previously.19 Therefore, the nominal formula of C2−C7 are noted as (Tb0.861Mn0.121)Mn1−xCuxO3−δ (x = 0.025, 0.05, 0.075, 0.10, 0.15, and 0.20). It should be mentioned that a small amount of Cu may be present 3476
DOI: 10.1021/acs.inorgchem.6b03024 Inorg. Chem. 2017, 56, 3475−3482
Article
Inorganic Chemistry
Figure 2. Selected area electron diffraction of C3 along [301]̅ , [100], and [101]̅ directions in space group Pnma.
Figure 3. Rietveld plot of the (a) X-ray and (b) neutron diffraction data for C5 around room temperature: (+) observed value, (solid line) calculated value, (marks below the diffraction patterns) calculated reflection positions; the difference curve is shown at the bottom of the figure. One reflection around 60° in the neutron data indicated by an asterisk (*) arises from the impurity Mn3−x−yTbxCuyO4.
Figure 5. XPS spectra of C1, C3, and C5: (a) Tb 3d5/2 and Tb 3d3/2 core levels, (b) Cu 2p3/2 core levels, and (c−e) Mn 2p3/2 and Mn 2p1/2 core levels.
to occupy the site of Mn. When x exceeds a certain value, impurities are found with the lattice parameters of (Tb0.861Mn0.121)Mn1−xCuxO3−δ unchanged as shown in Figure 1. The data seems to agree well with Vegard’s law before and after x = 0.04. However, it is not so sound because the data are less. The maximum value for x in the solid solution is about 0.12 from the present data shown in Figure 4. 3.2
