Effects of organic material on magnetoresistance in electron-doped double perovskite

The Curie temperature of electron-doped Sr2FeMoO6 can be optimized significantly due to the band-filling effect, but accompanying an almost absent low-field magnetoresistance (LFMR), which is unfavorable to applications in the magnetoresistive devices operated at room temperature. Our previous works confirmed that, a remarkable enhanced LFMR was observed in Sr2FeMoO6 by modifying the grain boundary with insulating organic small molecules (glycerin, CH2OHCHOHCH2OH). However, in this work, modifying the grain boundary strength of the La0.5Sr1.5FeMoO6 with the insulating organic macromolecules (oleic acid, CH3(CH2)7CH=CH(CH2)7COOH) or small molecules (glycerin), both of them have negligible functions on the magnetoresistance (MR) behavior in La0.5Sr1.5FeMoO6. Contrary to the glycerin-modified Sr2FeMoO6, Sr2FeMoO6/oleic acid composites do not exhibit an obviously increased MR property. Based on the above experimental results and the related works, it is proposed that, maintaining high spin polarization of the carriers at the Fermi level and improving the tunneling process across the grain boundary using the suitable organic materials are decisive factors for optimizing the MR behavior in the similar electron-doped double perovskites.


Introduction
Due to its a half-metallic property with the 100% spin polarization at the Fermi level, a high Curie temperature (T C ) of * 415 K and remarkable large low-filed magnetoresistance (LFMR) behavior, double perovskite Sr 2 FeMoO 6 (SFMO) has been paid a great deal of attention in views of its critical fundamental investigation values and immense potential technological applications for spintronic and magnetoresistive devices operated at room temperature [1].
It is well known that in an ideal SFMO double perovskite structure, the FeO 6 and MoO 6 octahedral arrange alternatively along the three axes of tetragonal structure with the Sr cations occupy the voids between them. Strong antiferromagnetic correlation exists between the localized magnetic moments (Fe 3? : 3d 5 ) and the delocalized electron (Mo 5? : 4d 1 ) in a double-exchange-like type, the magnetic coupling induces a ferrimagnetic state with an ideal saturated magnetization (M S ) of 4 l B per formula unit [2][3][4][5][6][7]. The strength of magnetic coupling in SFMO double perovskite is mainly controlled by the carrier density at the Fermi level, so it indicates that doping electrons in the conduction band is an effective way to enhance T C [8][9][10][11][12][13][14]. This point was confirmed by a substantial enhancement of T C more than 80 K in La x Sr (2-x)-FeMoO 6 [9] and (Ba 0.8 Sr 0.2 ) 2-x La x FeMoO 6 materials [15]. Obviously, the celebrated strategy increases the operating temperature range of the electromagnetic applications in SFMO double perovskite. However, the increased T C in electron-doped double perovskites always accompany with a strong suppression on LFMR effect [9,15,16]. This phenomenon seriously affects functional properties of materials and constrains the technical application. Hence, it is necessary and meaningful to improve the MR effect in electron-doping SFMO system.
As a fact, a remarkable large magnetoresistance (MR) effect can be observed in polycrystalline SFMO ceramics, but it is almost absent in SFMO single crystals or epitaxial films [17][18][19]. This suggests that the LFMR of SFMO is a type of tunneling magnetoresistance; the transport process is related with the spin-dependent scattering occurred at magnetic domain boundaries, so the existence of grain boundary (GB) in SFMO is vital for MR [1]. Previous research proved that LFMR could be improved by enhancing the GB strength in many methods, such as adding the second phase in GB [20,21], slightly oxidizing GB [22,23], reducing the grain size [24,25], and dispersing the grain uniformly [26]. It should be noted that there are three common points in these methods. First, the increment of LFMR value always accompanies with the enhancement of resistivity. Actually, a function of q / expðcs ffiffiffi ffi D p Þ was proposed to express the strength of the GB insulating barriers. From this, the GB strength is directly measured by q [27]. Second, the increment of LFMR value always with the decrement of magnetization. Moreover, the experiment process of above methods is relatively complex. Therefore, it is necessary to establish a facile method which can maintain the magnetization while improving the LFMR value of SFMO. In our previous research, the SFMO ceramic was soaked in organic matter, by the method directly, the resistivity was increased about 500 times and the LFMR value was effectively improved up to -29.5% at 10 K [28]. Based on the achievement, we infer that the method of preparing La 0.5 Sr 1.5 FeMoO 6 (LSFMO)/organic matter composites could be used to eliminate the negative effect resulting from electrons doping in SFMO ceramics and then optimize the LFMR effect while ensuring the magnetization.
The organic matters of oleic acid (CH 3 (CH 2 ) 7-CH=CH(CH 2 ) 7 COOH) and glycerin (CH 2-OHCHOHCH 2 OH) were used as the modifying material in this work. The oleic acid has a long-chain structure with the molecular weight of 282.45 and the glycerin has a short-chain structure with the molecular weight of 92.09. It is obvious that the former has the more excellent insulation.
In this work, three composites of LSFMO/oleic acid, LSFMO/glycerin and SFMO/oleic acid were prepared to investigate the effect of organic materials on the MR behavior in electron-doped double perovskite. The structure, magnetization, electrical resistivity and MR of composites were investigated systematically and comparatively. The main results indicate that neither of the MR behavior in the three experiments has been optimized effectively. It proves that it is critical to select a suitable organic material to modify GB while maintaining the high spin polarization of the carriers at the Fermi level in the electron-doped double perovskite. To modify the GB strength, the as-prepared pure LSFMO powder was post-treated as follows: first, different volume proportions of the oleic acid (CH 3 (CH 2 ) 7 CH=CH(CH 2 ) 7 COOH, V 1 ) and alcohol (V 2 ) were fully stirred to form the mixed solution (V ¼ 0; 0:1; 0:2; 0:3; 0:4; 0:5, V ¼ V 1 V 1 þV 2 ). For each treatment, 20 lL of the mixed solution was pipetted and added into 0.2 g as-prepared LSFMO powder. After six pipetting times, the formed organic/inorganic composite was thoroughly admixed and pressed into disk, then conserved at room temperature (RT). The collected composites were labeled as C1-C6, respectively.

The method of Experiment II
(La 0.5 Sr 1.5 FeMoO 6 /glycerin composite) For contrasting with Experiment I, in this experiment, the modificatory factor oleic acid was substituted by glycerin (CH 2 OHCHOHCH 2 OH). Pure LSFMO powder (in Experiment I) was pressed into disks and the mass of every disk was 0.2 g. To control the participating content of glycerin at GB, the disks were soaked in the isometric mixed reagent of glycerin and alcohol under the protection of N 2 , and kept static at room temperature for 5 and 15 days. Subsequently, the disks were taken out and gently washed three times by alcohol, dried in vacuum condition at RT. The collected disks were labeled as C7 (5 days) and C8 (15 days).

The method of Experiment III (Sr 2 FeMoO 6 /oleic acid composite)
The modified object LSFMO was substituted by pure SFMO. Specifically, the suitable amounts of analytic grade SrCO 3 , Fe 2 O 3 and MoO 3 powders were weighted, ground and sintered at 900°C for 10 h in air. Then the sintered powder was ground again for 3 h and annealed at 1200°C for 12 h in 5% H 2 /95% Ar reducing atmosphere. To modify the GBs, 0.2 g of SFMO powder was mixed homogeneously with 20 lL isopyknic mixed reagent of oleic acid and alcohol solution here. After completely admixing, the composite was pressed into disk and conserved at RT. The collected disk was labeled as C9.
Room-temperature X-ray diffraction (XRD) experiments were performed by X-ray diffraction (XRD) patterns (XRD, Bruker D8 Discover). The microstructure of the samples was examined through high-resolution field emission scanning electron microscope (FESEM, Zeiss SUPRA 40) and transmission electron microscopy (TEM, Philips JEM 2100). The C element's distribution was carried out using energy dispersive X-ray spectroscopy (EDS) coupled with FESEM instrument. The element valence was investigated by X-ray photoelectron spectroscopy (XPS, Escalab250Xi). The magnetization and magnetic transport data of all samples (C1-C9) were carried out by a physical property measurements system (PPMS Quantum Design, 2001NUGC).

Results and discussion
The crystal structure of all prepared polycrystalline samples was detected by the X-ray diffraction (Fig. 1a). This figure confirms the phase purity of all the ceramics and also shows the tetragonal symmetry with space group I4/m. It manifests that the organic molecules have no effect on the structure of original samples, in other words, there was negligible or even no chemical reaction between them. SrMoO 4 as a common second phase exists in SFMO double perovskite can be obtained by oxidation [23,29] or sonication process [30] in experiments. However, such a secondary phase is not shown up in this work which is evidenced by XRD results. The discrepancies between the present study and the previous one [29] may result from the different experimental manipulations, like the prepared method of samples and the content of the solvent in organic matter.
The chemical states of the Fe and Mo cations in SFMO material have a significant influence on the magnetic behaviors and other physical properties. Hence XPS measurements were carried out; the XPS spectra of Fe 2p and Mo 3d for the representative samples of C1, C6, C8, C9 and pure SFMO are shown in Fig. 1b and c, respectively. From Fig. 1b, it can be clearly seen that Fe 2p 3/2 and Fe 2p 1/2 of samples are located at approximately 710.6 eV and 724.3 eV, which indicates that the Fe cations in these samples are mainly in ? 3 oxidation state [31,32]. The double peaks contributed by the Mo 3d 5/3 and Mo 3d 2/3 caused by spin-orbit coupling can be observed in Fig. 1c. The location of the binding energy of the two peaks shows a weak shift, over the range of 231.7-232.4 eV and 234.7-235.3 eV, respectively. All those Mo 3d XPS spectra are in consistent with the reports and that demonstrates the existence of Mo 5? in all measured samples [33][34][35]. The XPS results manifest that the organic materials (oleic acid or glycerin) do not change the chemical states of Fe and Mo cations in LSFMO or SFMO double perovskite. J Mater Sci: Mater Electron (2021) 32:18711-18720 Figure 2 shows the microscopic morphology for C1-C6 composites. The FESEM images (Fig. 2a-f) and TEM image (Fig. 2g) show the grain size of sample is controlled with 10-25 nm in Experiment I. The detailed high-resolution TEM (HRTEM) analysis of the pure LSFMO (inset of Fig. 2g) reveals that particle with the (100) interplanar spacing is measured to be 0.37 nm [36][37][38]. The HRTEM image of a section of C6 in Fig. 2h shows an evidence of the combination of oleic acid molecular and LSFMO particle. Moreover, the corresponding two-dimensional C mappings of C1-C6 are measured with EDS and shown in the insets in Fig. 2a-f. It can be observed that the content of C element increases with the increment of the V values. This further infers that in the current synthetic condition, oleic acid is served Fig. 1 a XRD patterns of all prepared ceramics (C1-C9), b XPS spectra of Fe cation, and c XPS spectra of Mo cation as an insulating barrier that penetrated inside GBs not only adhesive to the surface, otherwise the C content should be independent on the concentration of oleic acid. The phenomenon that LSFMO grains distribute nonuniformity in the measured areas is attributed to the physical prepared process, in which the LSFMO's binding force forming in the annealing time can be destroyed by the subsequent crushing process. Based on the analysis, it seems that the physisorption of oleic acid in LSFMO grain regions might be the dominant states, which is also confirmed by the XRD and XPS analysis.
The magnetic hysteresis loops of LSFMO with different volume proportions of oleic acid/alcohol solution (V ¼ 0; 0:1; 0:2; 0:3; 0:4; 0:5) composites were measured at 50 K and 300 K, shown in Fig. 3a and b, respectively. Evidently, all the samples show the ferromagnetic nature with a well-saturated feature between -2.5 and ? 2.5 T. The mass differences of the pure LSFMO with LSFMO/oleic acid composites are provided in Table 1. It can be observed that the mass of organic matter that entered into LSFMO grain boundaries is too little and this extra mass has negligible effect when we calculate the magnetization values. With the more oleic acid content, the M S values at 50 K are, respectively, 1.46, 1.41, 1.68, 1.33, 1.32 and 1.47 l B /f.u. of the corresponding C1-C6 composite samples and the similar coercive field also can be observed. The analysis suggests the organic material can maintain the magnetization and has no obvious effect on the LSFMO An analogous behavior appears in the M-H curves at 300 K (Fig. 3b). Figure 4 shows the temperature-dependent resistivity curves for C1-C6 composites with 0 T and 1 T extra field. All the ceramics exhibit a semiconductorlike behavior in the range of investigating. From the figure, a systematic enhancement in resistivity happens over the whole temperature range of 10-300 K while increasing the extra amount of oleic acid without reducing the LSFMO content. The measured curves with 0 T and 1 T are almost overlapped in this figure, it suggests extra field has no obvious influence on resistance property. To describe resistivity more directly, the resistivity values (without extra field) of C1-C6 at 50 K and 300 K are shown (Fig. 5). In this picture, values of C6 composite are 138.8 X cm and 27.5 X cm at 50 K and 300 K, respectively, exhibiting * 130 times and * 90 times of C1 (1.1 X cm at   [42]. These reports suggest that the oleic acid can be served as the effective spin transport medium for LFMR behavior. Hence, oleic acid molecule is introduced in La-doped SFMO ceramic to expect to optimize the MR in this work. The MR% versus applied magnetic field (MR%-H) curves for C1-C6 at 50 K and 150 K are depicted in Fig. 6a and b, respectively. From Fig. 6a, the LFMR value at 1 T filed of pure La 0.5 Sr 1.5 FeMoO 6 (C1) is -2.8% and only up to -3.6% of C6 sample which possesses the maximum resistivity. The data verify that oleic acid has a bit positive influence on MR behavior but it is poor when compared with the systems that mentioned before [28,[39][40][41][42]. To clarify the reason of this invalid LFMR in La 0.5 Sr 1.5 FeMoO 6 /oleic acid composite, La 0.5 Sr 1.5-FeMoO 6 /glycerin (Experiment II) and Sr 2 FeMoO 6 / oleic acid (Experiment III) composites are devised and the more in-depth comparisons and discussions are stated in the next content. Figure 7a shows the magnetization versus temperature plots of C1-C6 composite samples that measured from 300 to 450 K. Almost samples undergo a ferromagnetic to paramagnetic transition at the same temperature (* 420 K) which can be obtained from Fig. 7a. There is no obvious change in the magnetic transition temperature of samples after adding the oleic acid molecules. This is yet another confirmation that the organic matter is not beneficial to enhance the ferromagnetic coupling strengths of the LSFMO material.
It is expected that, the insulating big molecules (oleic acid) located at GBs of LSFMO double perovskite can increase the tunneling probability of spin electrons, and thus enhance the LFMR behavior. Unfortunately, the method is proved to be noneffective according to the analysis of Experiment I. However, a remarkable LFMR enhancement in SFMO modified with small molecules (glycerin) was observed in our previous study, in which the LFMR of SFMO/glycerin composite was more than 2 times larger than that of the pure SFMO [28]. Therefore, we designed Experiment II (LSFMO/glycerin) to further study.
To understand the effect of glycerin molecules on LSFMO, the MR% versus applied magnetic field curves are presented in Fig. 7b. The inset of Fig. 7b shows that the LFMR value at 1 T field in 50 K of C1, C7 and C8 samples are -2.8%, -3.9% and -4.1%, respectively. The weakly enhanced LFMR value of LSFMO/glycerin composite has a sharp contrast with SFMO/glycerin [28]. From the experimental design in these two works, glycerin molecule is both introduced into GBs by the same soaking treatments to as the insulating barriers, the unchanged crystal structure of both two systems can be manifested by XRD. However, the data show a remarkable LFMR enhancement in SFMO/glycerin which improves from -12.0% for pure SFMO to -29.5% at 0.5 T field. By comparing, the mainly contrast of LFMR behavior between two systems can be reasonably attributed to the La doping. Once La partially substitute Sr in the SFMO double perovskite, it will lead to two effects: one is band-filling effect, and the other is the increased Fe/Mo anti-site defects concentration, both of them can decrease the amounts of spinpolarized electrons that tunneling across GB and then weaken the LFMR effect. Therefore, we can conclude that maintaining high spin polarization of the carriers at the Fermi level is a crucial factor to obtain an excellent MR behavior in similar electron-doped double perovskite.
To ensure a high spin polarization of carriers at the Fermi level, the SFMO with a high Fe/Mo ordering degree was combined with the oleic acid molecule in Experiment III. It can be directly gotten from Fig. 7c that the resistivity of SFMO can be enhanced by introducing oleic acid in the whole temperature range, similarly with the LSFMO/oleic acid. The MR of the pure SFMO (thin lines) and C9 composite (thick lines) were measured at 10 K, 50 K and 200 K, shown in Fig. 7d. As the picture presents, the LFMR of pure SFMO at 1 T are -16.3%, -12.6%, -4.4%, respectively, and -17.4%, -13.0%, -4.6% for C9 in 10 K, 50 K and 200 K. These data strongly verifies that oleic acid is not beneficial to the LFMR effect of SFMO sample [28].
Organic molecules have been already acknowledged as the suitable spin transport medium in magnetic materials [39][40][41][42]. In these works, the chemical bonding between organic molecules and ferromagnetic grains is the decisive factor for the enhancement of LFMR. Oleic acid is recognized as an efficient barrier when it closely contacts with the Fe 3 O 4 or La 0.7 Sr 0.3 MnO 3 grains with a single  Fig. 7 a Magnetizationtemperature (M-T) curves for C1-C6 measured from 300 to 450 K, b MR% versus applied magnetic field of C1, C7 and C8 measured at 50 K, c resistivity versus temperature curves for pure SFMO and C9 at 10-300 K with zero-field, and d MR% versus applied magnetic field of pure SFMO sample (thin lines) and C9 (thick lines) measured at 10 K, 50 K and 200 K molecular layer. But this phenomenon has not happened in our present work. It may derive from that oleic acid molecule is physically bonded with SFMO grains.
An efficient optimization of LFMR has been observed in SFMO/glycerin [28], which has a sharp contrast with SFMO/oleic acid. The reasons maybe lie in the different molecular structure of the organic matter and the bonding strength between organic molecules and SFMO grain boundary. As for SFMO/ oleic acid composite, due to the different preparation process compared with Fe 3 O 4 or La 0.7 Sr 0.3 MnO 3 grains, the big molecular oleic acid is difficult to form a single layer on the SFMO grain. That contributes a stronger energy barrier between oleic acid and SFMO grain and thus weak the spin injection efficiency, suppress to LFMR effect. But in the SFMO/glycerin system, compared with oleic acid, glycerin has a smaller molecular size and three hydroxyl groups, leads to a relative strong bonding strength with the SFMO interfacial grain. Therefore, glycerin is beneficial to spin injection and the improvement of LFMR effect.
Based on the analysis, we conclude that oleic acid can be recognized as a suitable transport medium when it chemically bonds to ferromagnetic grain. But when it bonds in physically, like the present work, oleic acid cannot optimize efficiently LFMR of the SFMO. In this case, it's available to select the smaller organic molecule with many hydroxyl groups to modify GB to enhance LFMR in the similar double perovskite materials.

Conclusion
La 0.5 Sr 1.5 FeMoO 6 /oleic acid composite was prepared to expect to improve the magnetoresistance of La 0.5-Sr 1.5 FeMoO 6 electron-doped double perovskite in Experiment I, but the result shows that the modifying treatment is noneffective. To make it clear, La 0.5-Sr 1.5 FeMoO 6 /glycerin composite (Experiment II) and Sr 2 FeMoO 6 /oleic acid composite (Experiment III) were designed and the MR behavior were detailedly discussed, respectively. Contrary to the glycerinmodified Sr 2 FeMoO 6 which has made a remarkable enhanced LFMR in our previous works, both of the composites in Experiment II and III don't exhibit the obviously increased MR. Based on the analysis of the experiments and the related works, we conclude the invalid MR effect on La 0.5 Sr 1.5 FeMoO 6 /oleic acid composite results from two respects: the lower spin polarization of carriers at the Fermi level and the weaker spin injection efficiency in GB derived from the oleic acid macromolecule. The observation in this work mainly indicates that maintain a high spin polarization in materials and select the small organic molecule to improve the tunneling process across GB may be the valuable way to enhance the LFMR behavior of La 0.5 Sr 1.5 FeMoO 6 or other similar electron-doped double perovskites. The further work is certainly required and we believe this is an interesting topic for the future work.