Synthesis and structure determination of ferromagnetic semiconductors LaAMnSnO₆ (A = Sr, Ba)

Abstract

LaAMnSnO₆ (A = Sr, Ba) have been synthesized by high temperature solid-state reactions under dynamic 1% H₂/Ar flow. Rietveld refinements on room temperature powder X-ray diffraction data indicate that LaSrMnSnO₆ crystallizes in the GdFeO₃-structure, with space groupPnma and, combined with transmission electron microscopy, LaBaMnSnO₆ in Imma. Both space groups are common in disordered double-perovskites. The Mn³⁺ and Sn⁴⁺ ions whose valence states were confirmed by X-ray absorption spectroscopy, are completely disordered over the B-sites and the BO₆ octahedra are slightly distorted. LaAMnSnO₆ are ferromagnetic semiconductors with a T_C = 83 K for the Sr- and 66 K for the Ba-compound. The title compounds, together with the previously reported LaCaMnSnO₆ provide an interesting example of progression from Pnma to Imma as the tolerance factor increases. An analysis of the relationship between space group and tolerance factor for the series LaAMnMO₆ (A = Ca, Sr, Ba; M = Sn, Ru) provides a better understanding of the symmetry determination for double perovskites.

---

Introduction

The perovskite structure is one of the most extensively studied structures in geology, materials science, solid-state physics and chemistry. The arrangement of atoms is thermodynamically extremely stable, hence the structure has considerable compositional and structural flexibility. Thus a very large number of double perovskites (AA′BB′O₆) have been investigated in the past decades. Either A- or B-site cation ordering can be realized and several comprehensive reviews have been published on the subject.^1–4 As a general rule, large size and/or charge differences of A- or B-cations facilitate the structural ordering. Typically, octahedral tilting distortions occur in the majority of the double perovskites when the A-site cation is too small for the large cubo-octahedral cavities; this distortion will generally affect the physical properties. Howard et al. and Woodward et al. proposed the group theoretical analysis and structural symmetry deduction of the superstructures arising from both tilting distortions and B-cation ordering.^2,5–7 A general structural feature of perovskites is the high degree of pseudo-symmetry, which always renders the establishment of the precise structure of a given perovskite a challenge.

Among the double perovskites, there are only a few examples, which demonstrate an unusual layer-by-layer ordering of the B-cations: Ln₂CuSnO₆ (Ln = La, Pr, Nd, Sm), La_2−xSr_xCuSnO₆ and La₂CuZrO₆.^8–10 Such unique layered ordering is believed to be driven by a strong Jahn–Teller deformation of the CuO₆ octahedra, which results in four short in-plane Cu–O distances (1.985–2.003 Å in La₂CuSnO₆⁸) and two long apical ones (2.367–2.395 Å) directed nearly perpendicular to the CuO₂ and SnO₂ planes. Taking into account the effect of the Jahn–Teller active B-cations, previous efforts have been devoted to the Mn³⁺-based double perovskites, including LnCaMnSnO₆ (Ln = La, Pr, Nd, Sm–Dy)¹¹ to form B-cation layer-ordered double perovskites. A cooperative Jahn–Teller deformation of MnO₆ in LaCaMnSnO₆ was indeed observed, but it was much weaker than in La₂CuSnO₆; additionally, an alternation of short and long Mn–O bonds is found within the ac-plane.¹¹

In an effort to prepare new B-cation, layer-ordered double perovskites, two new double perovskites LaAMnSnO₆ (A = Sr, Ba) have been prepared with comparably large A-cations. However, no cation ordering of Mn³⁺ and Sn⁴⁺ was found, probably, because of the small charge and size differences. The structures were carefully examined by powder X-ray diffraction, electron diffraction and high resolution transmission electron microscopy. LaSrMnSnO₆ crystallizes in the orthorhombic Pnma GdFeO₃-structure type with a unit cell of ∼√2a_p × 2a_p × √2a_p, while LaBaMnSnO₆ belongs to another space group, Imma, with the same unit cell set. Therefore, LaAMnSnO₆ with A = Ca, Sr, Ba represents an interesting series of compounds with a symmetry progression from Pnma of LaCaMnSnO₆ to Imma of LaBaMnSnO₆ with variation of the A-cation size.

Also discussed is the structural similarity with a related series of Mn³⁺-based double perovskites, LaAMnRuO₆ (A = Ca, Sr, Ba), where there is still controversy on the exact symmetry of LaSrMnRuO₆ and LaBaMnRuO₆.^12–16 The LaAMnSnO₆ phases are ferromagnetic semiconductors with T_C = 83 K and 66 K for the Sr and Ba compounds, respectively.

Experimental section

Polycrystalline samples of LaAMnSnO₆ (A = Sr, Ba) were prepared by high temperature solid-state reactions.

Stoichiometric amounts of La₂O₃ (obtained by the dehydration of La(NO₃)₃·6H₂O at 800 °C), SrCO₃ (Aldrich, 99.995%) or BaCO₃ (Aldrich, 99.98%), Mn₂O₃ (Aldrich, 99.999%) and SnO₂ (Aldrich, 99.99%) were ground with an agate mortar, pressed into pellets and treated for 10 hours at 900 °C. The resulting powders were reground, re-pressed and heated for 120 hours with three intermediate regrinding steps at certain conditions. Specifically, LaSrMnSnO₆ can be obtained by heating the pellet at 1400 °C in air, but with ∼10% admixture of La₂Sn₂O₇, which suggests that the main phase is somehow off-stoichiometric. Therefore, a pure sample for the Rietveld refinement and property measurements was prepared under a dynamic 1% H₂/Ar flow at 1250–1275 °C. Similarly, LaBaMnSnO₆ can not be synthesized at 1400 °C in air. However, by heating the pellet between 1125 and 1150 °C under 1% H₂/Ar flow, a single phase of LaBaMnSnO₆ can be obtained, which, however, decomposed at higher temperatures (∼1175 °C), probably due to the reduction of Sn⁴⁺ to Sn. It should be noted that the 1% H₂/Ar gas should be dried by going through a gas drier containing CaSO₄ particles, before flowing through the sample. LaAMnSnO₆ (A = Sr, Ba) are black powders.

The purity of both products was confirmed by powder X-ray diffraction (PXD). No visible impurity reflection peaks can be seen (Fig. 1). The PXD data were recorded on a Bruker D8-Advance diffractometer (in Bragg–Brentano geometry with Cu Kα radiation λ = 1.5406 Å, SOL-X solid-state detector, 40 kV and 40 mA, step scan 10–120°/0.02°/14 s). Rietveld refinements were performed with the TOPAS software package.¹⁷ The samples for electron microscopy investigation were prepared by dispersing the powder in ethanol and depositing it on a holey carbon grid. Energy dispersive X-ray (EDX) analysis was performed in a JEOL 5510 scanning electron microscope equipped with the Oxford INCA system. EDX measurements on LaBaMnSnO₆ on 112 positions showed the cation ratio to be La_1.0(2)Ba_1.0(2)Mn_0.9(3)Sn_1.1(3)O_x, confirming the stoichiometry of the compound within experimental error. Electron diffraction (ED) patterns were obtained on a Philips CM20 transmission electron microscope, and high resolution transmission electron microscopy (HRTEM) images on a Tecnai G2. The Mn and Sn X-ray absorption near edge spectroscopy (XAS) were collected simultaneously in both the transmission and fluorescence mode on powder samples on beam line X-19A at the Brookhaven National Synchrotron Light Source. The dcmagnetic susceptibility measurements were carried out on powder samples with a Quantum Design MPMS-XL superconducting quantum interference device (SQUID) magnetometer. Powder samples were placed in a gelatin capsule fastened in a plastic straw for immersion into the SQUID. Typical zero field cooling (ZFC) and field cooling (FC) magnetizations in the temperature range 5–300 K were performed under various external fields from 100 Oe to 1 T. When higher fields were needed after each ZFC–FC cycle, the sample was pulled out from the low temperature chamber to room temperature atmosphere to avoid the magnetic hysteresis effect. The isothermal magnetizations at 5 K were measured up to 5 T. Resistivity measurements were performed with a standard four-probe technique in the SQUID with a Keithley (model 181) equipment. Thermogravimetric analysis (TGA) was carried out on an SDT-Q600 instrument with a heating rate of 10 °C min⁻¹ from 30 to 1100 °C under O₂ atmosphere.

Fig. 1 Rietveld refinement of the powder X-ray diffraction patterns for LaSrMnSnO₆ (a) and LaBaMnSnO₆ (b). The symbol ○ represents the observed pattern and the solid line is the calculated pattern; the vertical marks below the diffraction patterns are the expected reflection positions, and the difference curve is also shown below the diffraction pattern. The structure views along the [010] and [101] directions are shown for both compounds.

Results and discussion

Fig. 1 gives the Rietveld refinements of the PXD for both LaSrMnSnO₆ and LaBaMnSnO₆, the cell lattice sets are the same ∼√2a_p × 2a_p × √2a_p for both, while the space groups (SG) are different: Pnma (no. 62) for the former one and Imma (no. 74) for the latter. Both SG are common for disordered double-perovskites. Details of the space group determination are discussed in the sections below. The corresponding refined cell parameters and selected bond distances are listed in Table 1. Atomic coordinates and thermal displacement factors are given in Tables 2 and 3. At first, the occupancies for all atoms were freely refined, and converged to be very close to 0.5 for La, Sr, Ba, Mn, and Sn and 1 for O. Therefore, in the final cycle of the Rietveld refinement, the occupancy factors were fixed (see Tables 2 and 3), which reflects a completely disordered structural model for both the A- and B-sites. It should also be noted that the cooperative Jahn–Teller distortion of the BO₆ octahedra observed in LaCaMnSnO₆ is not seen in the title compounds. In contrast, the MnO₆/SnO₆ octahedra in both compounds are barely distorted (Table 1), although half of them are occupied by the Jahn–Teller active Mn³⁺ cation. It should be noted that the barely distorted octahedra are from the average structure obtained with PXD in our study. However, experimental techniques sensitive to local structure, such as pair distribution function technique, would be needed to determine if the coordination around Mn³⁺ is truly barely distorted on a local scale. Table 1 presents the results of the title compounds along with a series of double-perovskites LaAMnRuO₆ (A = Ca, Sr, Ba) and LaCaMnSnO₆ previously reported in the literature, which are similar to the title compounds. As mentioned above, there is still uncertainty about the exact symmetry of LaSrMnRuO₆ and LaBaMnRuO₆. As summarized in Table 1, there have been differing conclusions. The structure of LaSrMnRuO₆ was first reported in 2000 to be pseudo-cubic by Ramesha et al.,¹⁸ then in 2004 as orthorhombic Pnma by Dass et al.,¹² and in 2006 as orthorhombic Imma by Bune et al.¹³ The report of Imma symmetry was based on Rietveld refinements of synchrotron PXD data, where a small amount of Pnma phase was also observed along with the major phase by ED. In 2008, LaSrMnRuO₆ was again assigned to Pnma based on neutron diffraction data.¹⁴LaBaMnSnO₆ was first reported with the ideal cubic structure (Pm3̄m) in 2000 by Horikubi and Kamegashira,¹⁵ then as orthorhombic Pnma by Granado et al.¹⁶ Thus it appears that LaSrMnRuO₆ can form either in Pnma or Imma, while LaBaMnRuO₆ has either Pnma or Pm3̄m symmetry. Here, our structural studies of LaSrMnSnO₆ and LaBaMnSnO₆, together with the previous analysis on LaAMnRuO₆, are instructive on space group determination for double-perovskites.

Table 1 Crystallographic and magnetic parameters for LaAMnBO₆ (A = Ca, Sr, Ba and B = Ru, Sn)

	LaCaMnRuO6	LaSrMnRuO6		LaBaMnRuO6		LaCaMnSnO6	LaSrMnSnO6	LaBaMnSnO6
Ref.	16	12,14	13	16	15	11	This work	This work
a t 1 Factor = 〈A–O〉/√2〈M–O〉. b t 2 Factor = (〈RA〉 + RO)/√2(〈RB〉 + RO); RO2^− = 1.40 Å, RLa3^+ = 1.16 Å for CN = 8, 1.27 Å for CN = 10, and 1.36 Å for CN = 12; RCa2^+ = 1.12 Å, RSr2^+ = 1.26 Å for CN = 8 and 1.36 Å for CN = 10; RBa2^+ = 1.57 Å for CN = 11 and 1.61 Å for CN = 12; RMn3^+ = 0.645 Å, RRu4^+ = 0.62 Å, RSn4^+ = 0.69 Å for CN = 6 (CN stands for coordination number). The order of t2 factors: LaCaMnSnO6 < LaCaMnRuO6 < LaSrMnSnO6 < LaSrMnRuO6 < LaBaMnSnO6 < LaBaMnRuO6.
S.G.	Pnma	Pnma	Imma	Pnma	Pm3̄m	Pnma	Pnma	Imma
Tilt sys.	a^−b ^+a^−	a^−b ^+a^−	a^0b^−b ^−	a^−b ^+a^−	a^0a^0a^0	a^−b ^+a^−	a^−b ^+a^−	a^−b^0a ^−
a/Å	5.5099(2)	5.5257(1)	7.79296(9)	5.5975(5)	3.96309(8)	5.6701(1)	5.6400(5)	5.729(3)
b/Å	7.7553(3)	7.8126(2)	5.51255(6)	7.9163(8)		7.9022(1)	7.9773(7)	8.1065(7)
c/Å	5.4819(3)	5.5664(1)	5.55603(6)	5.6197(8)		5.55446(8)	5.6435(6)	5.730(3)
V/Å^3	234.25(3)	240.300(7)	238.682(2)	249.02(5)	62.2	248.873(5)	253.91(4)	266.1(2)
T C (K)	∼200	∼220, ∼250	∼225	∼200	∼200	∼65	83	66
µ eff (µB)	1.77(2)	1.83(5)	2	1.96(4)		5.0	5.12	5.31
A–O/Å	2.38	2.46(1)	2.52	2.56/2.77	2.80 × 12	2.30(1)	2.39(2)	2.7(1)
	2.53	2.67(1)	2.77 × 2	2.87		2.52(1)	2.702(7)	2.871(9) × 2
	2.40 × 2	2.52(1) × 2	2.58 × 4	2.72 × 4		2.41(2) × 2	2.55(1) × 2	2.79(4) × 4
	2.66 × 2	2.67(1) × 2	2.95 × 4	2.85 × 2		2.70(2) × 2	2.77(1) × 2	2.95(5) × 4
	2.73 × 2	2.80(2) × 2		2.91 × 2		2.76(2) × 2	2.78(1) × 2
〈A–O〉/Å	2.56	2.64	2.74	2.79	2.80	2.57	2.66	2.85
M–O1/Å	1.9847(5) × 2	1.9854(4) × 2	1.968(2) × 2	2.0006(7) × 2	1.98 × 6	2.059(3) × 2	2.046(4) × 2	2.04(1) × 2
	1.980(1) × 2	1.979(2) × 2	1.9716(9) × 4	1.9846(4) × 2		1.99(1) × 2	2.00(1) × 2	2.029(5) × 4
	1.997(1) × 2	1.982(2) × 2		1.9899(3) × 2		2.102(9) × 2	2.04(1) × 2
〈M–O〉/Å	1.99	1.98	1.97	1.99	1.98	2.05	2.03	2.03
t 1 Factor^a	0.91	0.94	0.98	0.99	1.00	0.89	0.93	0.99
t 2 Factor^b	0.89	0.91	0.94	0.98	1.00	0.87	0.89	0.96
M–O1–M/°	155.3	159.3	163.5	163.2	180	147.3	154.3	169.3
M–O2–M/°	155.5	163.9	165.9	175.4	180	152.3	161.3	173.7

---

Table 2 Atomic coordinates and atomic displacement parameters for LaSrMnSnO₆ from PXD Rietveld refinement using space groupPnma (no. 62). Rp = 4.3%, Rwp = 6.2% and gof = 2.01

Atom	Position	x	y	z	U eq/Å^2	Occupancy
La/Sr	4c	0.0218(1)	1/4	0.9976(8)	0.011(6)	0.5/0.5
Mn/Sn	4b	0	0	1/2	0.008(6)	0.5/0.5
O1	4c	0.494(1)	1/4	0.080(3)	0.007(1)	1
O2	8d	0.278(2)	0.472(1)	0.718(2)	0.007(1)	1

---

Table 3 Atomic coordinates and atomic displacement parameters for LaBaMnSnO₆ from PXD Rietveld refinement using space groupImma (no. 74). Rp = 4.7%, Rwp = 6.4% and gof = 2.02

Atom	Position	x	y	z	U eq/Å^2	Occupancy
La/Ba	4e	0	1/4	0.500(3)	0.0139(6)	0.5/0.5
Mn/Sn	4a	0	0	0	0.0028(5)	0.5/0.5
O1	4e	0	1/4	0.03(2)	0.034(2)	1
O2	8g	1/4	0.514(8)	1/4	0.034(2)	1

---

The major issue in determining the structure of a perovskite-based material is to choose the right symmetry, namely the space group and unit cell. BO₆ octahedral tilting distortions (in terms of off-symmetric cation or oxygen positions) lead to various pseudo-cubic symmetries. One governing parameter is the tolerance factor, which can be either defined as t₁ = 〈A–O〉/√2〈M–O〉, or t₂ = (〈r_A〉 + r_O)/√2(〈r_B〉 + r_O) (Table 1). According to group theoretical analysis,² the space groupsPnma and Imma are characterised by (a⁻b⁺a⁻) and (a⁻b⁰a⁻) tilt systems, respectively. Thus the Pnma and Imma structure models differ only in the details of the octahedral tilting, and it is reasonable to expect a smooth progression from Pnma to Imma as the tolerance factor increases, specifically reflected by the removal of the in-phase tilt along the b-direction.

As shown in Table 1, there is good agreement that LaCaMnRuO₆ and LaCaMnSnO₆ phases crystallize in the Pnma structure with the lowest tolerance factors. Due to the small size of the Ca²⁺ cation, the MnO₆/SnO₆ octahedra are forced to tilt in order to optimize the Ca²⁺ environment by going from 12- to 8-coordinated. It can be visualized by the deviation from 180° of the M–O–M angle. Apparently, the lower tolerance factors correspond to the lower M–O–M angles as shown in Table 1. LaSrMnSnO₆ also has a low t value. The PXD data obtained for LaSrMnSnO₆ were fit with both Pnma and Imma models. However, the results are unambiguous; the overall fit is much better with the Pnma model. Furthermore, there are several Pnma peaks observed (indicated by arrows in Fig. 1a) that would be systematically absent if the structure possessed Imma symmetry. Both in-phase and out-of-phase tilting can be clearly seen in Fig. 1a. The M–O1–M and M–O2–M angles are 154.3° and 161.3°, respectively, which indicate moderate tilting distortions. It is clearly not possible to assign this material to the (a⁻b⁰a⁻) system (Imma).

For LaBaMnSnO₆, the PXD pattern was refined with Pnma, Imma and Pm3̄m, with acceptable convergence factors Rwp = 6.6%, 6.4% and 6.9%, respectively. The small differences between the Rwp values are not sufficient to determine the correct symmetry. Therefore, all three possibilities were tested with electron diffraction. The ED patterns are shown in Fig. 2. All reflections can be indexed with a cell ∼√2a_p × 2a_p × √2a_p, this is the indexation scheme used for Fig. 2. Pm3̄m with ∼a_p × a_p × a_p can be eliminated as a possibility, since the weak reflection indexed on the [100] pattern in Fig. 2 as 011 has a d-value of approximately 4.7 Å. In fact, the 0kl : k + l = 2n + 1 reflections in the [100] zone correspond to superstructure reflections in the 〈110〉_p zone that are typical for perovskites in which an anti-phase tilt component is present.¹⁹ The hkl reflection condition derived from all tilt series is hkl : h + k + l = 2n. This eliminates also the possibility of Pnma, leaving Imma. On the other hand, all 〈110〉_p ED patterns observed show a super-reflection, while for Imma it should be present in [100] (=[110]_p), but not in [001] (=[1̄10]_p). However, HRTEM images (Fig. 3) show that nanoscale orientation variants (twin related variants) occur of [100] and [001], as proven by the Fourier transforms. Any selected area electron diffraction will enclose an area containing both zones together, and on the ED pattern this gives the overlap of all reflections of type (eee); the absence of ½(ooe) in [001] will be obscured by the presence of ½(eoo) at the same position in [100] (where o indicates an odd index, and e an even index). This explains the seemingly contradicting occurrence of superstructure reflections in all 〈110〉_p ED patterns, and leaves Imma as a space group in agreement with all experimental TEM data. Indeed, the Rietveld refinement with Imma gives the best fit (R_wp = 6.4%). The structural views along the [010] and [101] directions are shown in Fig. 1b, which are consistent with the proposed tilting system (a⁻b⁰a⁻). The M–O1–M (169.3°) and M–O2–M (173.7°) bond angles are much closer to 180° than those in LaSrMnSnO₆ (Table 1) which indicates that LaBaMnSnO₆ is very close to the undistorted cubic perovskite structure (Pm3̄m), and that is exactly the reason why the PXD pattern is simple, and weak reflections, that violate the cubic symmetry can only be seen by ED.

Fig. 2 ED patterns of the main zones of LaBaMnSnO₆.

Fig. 3 HRTEM image of LaBaMnSnO₆ of an area showing [001] domains next to [100] domains.

The tolerance factor (t₂ = 0.91 for Pnma and 0.94 for Imma) for LaSrMnRuO₆, the most studied compound in this series and for which the correct space group is still in question, is between that of LaSrMnSnO₆ (Pnma, t₂ = 0.89) and LaBaMnSnO₆ (Imma, t₂ = 0.96). The observed tilt angles of LaSrMnSnO₆ (M–O1–M = 154.3° and M–O2–M = 161.3°) are larger than those of LaSrMnRuO₆ (M–O1–M = 159.3° and M–O2–M = 163.9° for a Pnma model; M–O1–M = 163.5° and M–O2–M = 165.9° for an Imma model). Bune et al.¹³ and Woodward et al.,¹⁴ although both carried out careful work to determine the symmetry of LaSrMnRuO₆ arrived at: Imma and Pnma, respectively. The possible reason for the different conclusion is that the materials in the above two studies were different, due to the slightly different preparation conditions. The structural tolerance factor of LaSrMnRuO₆ could be coincidently at this boundary between Pnma and Imma. For LaBaMnRuO₆ with the larger tolerance factor the symmetry is therefore expected to be Imma or higher.

Mn and Sn XAS measurements can be used to probe the respective oxidation states in these compounds, as illustrated in Fig. 4. In Fig. 4a the Mn–K edges of LaAMnSnO₆ (A = Sr, Ba) are compared to those of a series of Mn-standards. The chemical shift of the Mn absorption edge to higher energy, with increasing formal oxidation state can clearly be seen by comparing the Mn(II)–Sr₂ReMnO₆, Mn(III)–LaMnO₃ and Mn(IV)–CaMnO₃ curves at the absorption coefficient (µ) value of 1.0.^20,21 As is conventional, the absorption-edge-step in µ has been normalized to unity in the range of 70–200 eV above the edge. The close coincidence of the Mn–K spectra of LaAMnSnO₆ compounds to that of LaMnO₃ in the µ ≈ 1 range, clearly indicates the Mn(III) state for these compounds.

Fig. 4 (a) Mn–K edge XAS spectra for the compounds LaAMnSnO₆ (A = Sr, Ba) and the standards MnO, Sr₂ReMnO₆, LaMnO₃, and CaMnO₃. (b) Sn–L₁XAS spectra for LaBaMnSnO₆ and the standards, Sn, SnO, SnO₂, and LaCaCrSnO₆.

In Fig. 4b the Sn–L₁ edge of LaBaMnSnO₆ is compared to those of a series of Sn-standards. As a p-block element, the frontier bonding orbitals of Sn are the 5s/5p states. The Sn–L₁ near edge involves transitions into the p-symmetry states above the Fermi energy, and consequently the structure and chemical shift of the Sn-L₁ edge is particularly valence-state sensitive. This is similar to Sb-based compounds studied by our group in the past.²² The Sn-L₁ edge chemical shifts between the Sn(0)–Sn, Sn(II)–SnO, and Sn(IV)–SnO₂spectra are clear in Fig. 4b. The chemical shift of LaBaMnSnO₆ clearly identifies it as a Sn(IV) compound. The spectrum of another perovskite based Sn(IV) compound, studied in our lab, LaCaCrSnO₆ is shown to emphasize the typical spectral shape for Sn(IV) in this structure.²³ The close similarity of the Mn edges between the A = Ba and Sr (discussed above) supports the analogous assignment of Sn(IV) in LaBaMnSnO₆ more than reasonable.

There is no detectable oxygen deficiency in either compound, although they were synthesized under a reducing gas flow. TGA measurements for LaAMnSnO₆ (A = Sr, Ba) under O₂ atmosphere up to 1100 °C (Fig. S1 in the ESI†) show no weight increase, but rather a small decrease (∼0.3 wt%), probably due to a slight decomposition. PXD patterns were recorded after the TGA measurements (Fig. S2 in the ESI†), which indicate that the structures were retained.

The zero-field cooled and field cooled (ZFC–FC) temperature-dependent magnetic susceptibilities for LaSrMnSnO₆ were measured in the range of 5–300 K at various external fields (Fig. 5a). The reciprocal susceptibility χ⁻¹(T) above 220 K follows the Curie–Weiss law reasonably with θ = 179 K, C = 3.28 cm³ K mol⁻¹ (inset of Fig. 5a). The experimental effective magnetic moment µ_eff is estimated to be 5.12 µ_B, which is in agreement with the spin only value calculated as g√S(S + 1) µ_B = 4.9 µ_B, for S = 2 and g = 2 for Mn³⁺. Irrespective of the random distribution of Mn and Sn cations over the B-site, the large positive value of θ suggests strong ferromagnetic (FM) interactions dominating between Mn³⁺ cations, which is consistent with the monotonous increase of χT above 60 K (Fig. 5b). With decreasing temperature, the susceptibility increases sharply, starting at ∼100 K, which evidences FM ordering of the sample; the ZFC–FC curves diverge at lower temperatures, in addition, the ZFC shows a decreasing tendency, suggesting a spin or cluster glassy component.^24–26 However, this glassy component almost disappears when the external magnetic field (H) is higher than 1 kOe (Fig. 5a). The FM ordering temperature could be estimated from the differential of the χT curve (inset of Fig. 5b), which shows a pronounced and negative peak at ∼83 K. LaBaMnSnO₆ shows similar magnetic properties. The ZFC–FC curves between 100 Oe and 1 T are shown in Fig. 5c. The reciprocal susceptibility χ⁻¹(T) above 150 K follows the Curie–Weiss law and yields C = 3.52 cm³ K mol⁻¹ and θ = 82.6 K, as shown in Fig. 5c. T_C for LaBaMnSnO₆ is estimated to be ∼66 K by the differential of the χT curve (inset of Fig. 5d). LaCaMnSnO₆ also shows a T_C transition at ∼65 K.¹¹ The T_C progression is similar to that in the series of LaAMnRuO₆ (A = Ca, Sr, Ba), where the Sr-compound also exhibits the highest T_C ≈ 250 K while T_C for the Ca- and Ba-analogous are both ∼200 K (Table 1).¹⁶ The M–O–M bond angle is the largest for the Ba compounds in both the Sn and Ru series (Table 1), and would be expected to maximize super-exchange, and lead to the highest T_C; it appears, however, that super-exchange alone is not sufficient to understand the magnetic properties of these systems.

Fig. 5 (a) Temperature dependence of the magnetic susceptibility: ZFC–FC plots for LaSrMnSnO₆ under different external magnetic fields (100, 500 and 1000 Oe). The inset is the Curie–Weiss fit with the data above 220 K; (b) χTvs.T curve at 1 kOe FC condition for LaSrMnSnO₆. The inset is the d(χT)/dTvs.T curve; (c) ZFC–FC curves for LaBaMnSnO₆ under different external fields (100 Oe, 500 Oe, 1 kOe, 5 kOe and 1 T). The inset is the Curie–Weiss fit; (d) χTvs.T curve at 1 kOe FC for LaBaMnSnO₆. The inset is the d(χT)/dTvs.T curve.

In Fig. 6 the isothermal magnetization curves at 5 K for LaSrMnSnO₆ show a rapid saturation at ∼3.77 µ_B, slightly lower than the theoretically expected value (gSµ_B = 4 µ_B) for spin-only moments, which is coincident with the strong FM interactions. The FM coupling in LaBaMnSnO₆ is relatively weaker and the saturation magnetization at 5 T is ∼2.03 µ_B (50% of that expected). A higher H is necessary to obtain full saturation. As shown in the insets of Fig. 6, LaAMnSnO₆ (A = Sr, Ba) curves show a small magnetic remnant and coercive fields of 0.53 µ_B, 0.07 µ_B and 130 Oe, 163 Oe, respectively. Whether the remnant fields are from FM or a magnetic glass component, both compounds show typical soft ferromagnetic behavior.

Fig. 6 Field dependence of the magnetization plots for LaAMnSnO₆ (A = Sr, Ba) at 5 K. The insets are the low field enlargements.

In Fig. 7, the temperature variations of the resistivity for both compounds show a semiconducting behavior above 200 K. The estimated activation energies (E_a) ∼0.274 eV for LaSrMnSnO₆ and 0.284 eV for LaBaMnSnO₆ are typical for semiconductors. The ZFC–FC curves are identical. No magnetoresistivity was observed in the paramagnetic region. Below 200 K, the signal is out of the range of the measuring device.

Fig. 7 Temperature dependence of resistivity for LaAMnSnO₆ (A = Sr, Ba) with and without an external magnetic field (1 T).

Conclusions

In conclusion, two new Mn³⁺-based double perovskites LaAMnSnO₆ (A = Sr, Ba) have been prepared under reducing atmosphere. The structures were carefully characterized by PXD and ED to be in Pnma (a⁻b⁺a⁻) and Imma (a⁻b⁰a⁻), respectively, which are typical space groups for disordered perovskites. The BO₆ octahedra in both title compounds are barely distorted, although half of them are occupied by the Jahn–Teller active Mn³⁺ cation. LaAMnSnO₆ (A = Sr, Ba) are ferromagnetic semiconductors with a T_C = 83 K and 66 K, respectively. The main achievement of this work is the unambiguous determination of the space group symmetry for the series LaAMnSnO₆ (A = Ca, Sr, Ba), which shows an interesting progression from Pnma to Imma with increasing tolerance factor, t, specifically reflected by the removal of the in-phase tilt of the BO₆ octahedra along the b-direction. Moreover, the analysis and comparison of the relationship between space group and tolerance factor provides a better understanding of the symmetry observed in the analogous series, LaAMnRuO₆ (A = Sr, Ba).

Acknowledgements

This work was partially supported by NSF-DMR 0541911 and 0966829 grants (MG, TY). J.H. acknowledges financial support from the European Union under the Framework 6 program under a contract for an Integrated Infrastructure Initiative. Reference 026019 ESTEEM.

Notes and references

1. M. T. Anderson, K. B. Greenwood, G. A. Taylor and K. R. Poeppelmeier, Prog. Solid State Chem., 1993, 22, 197–233.
2. C. J. Howard, B. J. Kennedy and P. M. Woodward, Acta Crystallogr., Sect. B: Struct. Sci., 2003, 59, 463–471.
3. P. K. Davies, H. Wu, A. Y. Borisevich, I. E. Molodetsky and L. Farber, Annu. Rev. Mater. Res., 2009, 38, 369–401.
4. G. King and P. M. Woodward, J. Mater. Chem., 2010, 20, 5785–5796.
5. P. M. Woodward, Acta Crystallogr., Sect. B: Struct. Sci., 1997, 53, 32–43.
6. P. M. Woodward, Acta Crystallogr., Sect. B: Struct. Sci., 1997, 53, 44–66.
7. M. W. Lufaso and P. M. Woodward, Acta Crystallogr., Sect. B: Struct. Sci., 2001, 57, 725–738.
8. M. T. Anderson and K. R. Poeppelmeier, Chem. Mater., 1991, 3, 476–482.
9. M. T. Anderson, K. R. Poeppelmeier, S. A. Gramsch and J. K. Burdett, J. Solid State Chem., 1993, 102, 164–172.
10. M. Ducau, K. S. Suh, J. Senegas and J. Darriet, Mater. Res. Bull., 1992, 27, 1115–1123.
11. A. M. Abakumov, M. D. Rossell, S. A. Seryakov, M. G. Rozova, M. M. Markina, G. Van Tendeloo and E. V. Antipov, J. Mater. Chem., 2005, 15, 4899–4905.
12. R. I. Dass, J. Q. Yan and J. B. Goodenough, Phys. Rev. B: Condens. Matter Mater. Phys., 2004, 69, 094416.
13. R. O. Bune, M. X. Lobanov, U. Popov, M. Greenblatt, C. E. Botez, P. W. Stephens, M. Croft, J. Hadermann and G. Van Tendeloo, Chem. Mater., 2006, 18, 2611–2617.
14. P. M. Woodward, J. Goldberger, M. W. Stoltzfus, H. W. Eng, R. A. Ricciardo, P. N. Santhosh, P. Karen and A. R. Moodenbaugh, J. Am. Ceram. Soc., 2008, 91, 1796–1806.
15. T. Horikubi and N. Kamegashira, Mater. Chem. Phys., 2000, 65, 316–319.
16. E. Granado, Q. Huang, J. W. Lynn, J. Gopalakrishnan and K. Ramesha, Phys. Rev. B: Condens. Matter Mater. Phys., 2004, 70, 214416.
17. TOPAS V2.1: General Profile and Structure Analysis Software for Powder Diffraction Data, Bruker AXS, Karlsruhe, Germany.
18. K. Ramesha, V. Thangadurai, D. Sutar, S. V. Subramanyam, G. N. Subbanna and J. Gopalakrishnan, Mater. Res. Bull., 2000, 35, 559–565.
19. A. M. Glazer, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1975, 31, 756–762.
20. M. Croft, D. Sills, M. Greenblatt, C. Lee, S. W. Cheong, K. V. Ramanujachary and D. Tran, Phys. Rev. B: Condens. Matter, 1997, 55, 8726–8732.
21. G. Popov, M. Greenblatt and M. Croft, Phys. Rev. B: Condens. Matter Mater. Phys., 2003, 67, 024406.
22. T. Mandal, A. Abakumov, M. Lovanov, V. Poltavets, M. Croft, J. Stalick and M. Greenblatt, Chem. Mater., 2008, 20, 4653–4660.
23. T. Mandal, V. Poltavets and M. Greenblatt, J. Solid State Chem., 2008, 181, 2325–2331.
24. J. W. Cai, C. Wang, B. G. Shen, J. G. Zhao and W. S. Zhan, Appl. Phys. Lett., 1997, 71, 1727–1729.
25. D. D. Stauffer and C. Leighton, Phys. Rev. B: Condens. Matter Mater. Phys., 2004, 70, 214414.
26. S. Karmakar, S. Taran, B. K. Chaudhuri, H. Sakata, C. P. Sun, C. L. Huang and H. D. Yang, Phys. Rev. B: Condens. Matter Mater. Phys., 2006, 74, 104407.

---

Footnote

† Electronic supplementary information (ESI) available: EDX, TGA, and PXD patterns after TGA measurements. See DOI: 10.1039/c0jm02614j

---