The TEM images of LCSMO/Mn3O4 composite nanoparticles for samples sintered at temperatures of 700oC, 800oC, and 900oC are depicted in Fig. 1(a), (b), and (c), respectively. All samples contain some agglomerated particles that overlap with each other. Additionally, the particle size distribution of each sample is presented in Fig. 1(d), (e), and (f), with black lines representing the fitting of the log-normal particle size distribution. It shows that the particle size is distributed up to 60 nm for all samples. The average particle size is 28 nm, 30 nm, and 32 nm for the samples sintered at 700oC, 800oC, and 900oC, respectively. The TEM analyses confirmed that increasing sintering temperatures increased particle sizes, as in previous results [15], [32].
The XRD patterns of LCSMO/Mn3O4 composite nanoparticles at different sintering temperatures are presented in Fig. 3. The major peaks of all samples match those of monoclinic La0.85Ca0.15MnO3 (COD no: 1525829 [33]), indicating that Sr has doped into the La/Ca site. The remaining peaks, represented by blue circles, correspond to tetragonal Mn3O4 (COD no: 1514115 [34]), confirming the formation of the LCSMO/Mn3O4 composite phase in all samples. Rietveld refinement analysis confirmed that the fraction of the LCSMO phase is 82.73%, 80.80%, and 76.80% for sintering temperatures of 700oC, 800oC, and 900oC, respectively. Meanwhile, the fraction of the Mn3O4 phase is 17.27%, 19.20%, and 23.10% for the samples sintered at 700oC, 800oC, and 900oC, respectively. This indicates that increasing the sintering temperature results in an increased fraction of the Mn3O4 composite phase. The increase in Mn3O4 composite fraction is likely due to interdiffusion between the LCSMO and Mn3O4 phases. The thermal energy from the sintering temperature might affect the mobility of ions in adjacent phases, facilitating diffusion across phase interface[10].
Figures 4 (a) and (b) illustrate the schematic crystal structures of LCSMO and Mn3O4, respectively. The La, Ca/Sr, Mn, and O atoms are represented by green, blue, purple, and red spheres. The LCSMO sample sintered at 800°C exhibits the largest lattice constants, a and b, along with the smallest lattice constant, c, resulting in the maximum volume of the unit cell. The d-spacing at the highest peak (002) is 0.27304, 0.27314, and 0.27314 nm for sintering temperatures 700, 800, and 900oC, respectively. It is closely matching with the d-spacing estimated from TEM measurement. In the case of Mn-O bond length, notable alterations are observed in the Mn(1)-O(2) and Mn(2)-O(3), along with significant changes in the Mn(1)-O(2)-Mn(1) and Mn(2)-O(3)-Mn(2) bond angles. For the Mn3O4 phase, the sample sintered at 800oC also shows the highest lattice constants a, b, and c, leading to the highest volume of the unit cell. This indicates that the volume of the unit cell in both LCSMO and Mn3O4 phases is highest in the sample sintered at 800oC. The goodness of fit (GoF) value for all samples ranges between 1.50 and 1.90, suggesting good agreement between the fitting model and the data. The structural parameters, including lattice constants, volume of the unit cell, Mn-O bond lengths, and Mn-O-Mn bond angles for LCSMO and Mn3O4, are summarized in Table 1.
Table 1
Rietveld refinement crystallographic parameter of La0.7Ca0.25Sr0.05MnO3/Mn3O4 composite nanoparticles
Crystallographic
parameters
|
Reference
La0.85Ca0.15MnO3
COD: 1525829 [33]
|
La0.7Ca0.25Sr0.05MnO3
|
TSintering= 700oC
|
TSintering= 800oC
|
TSintering= 900oC
|
Crystal structure
|
Monoclinic
|
Monoclinic
|
Monoclinic
|
Monoclinic
|
Space group
|
P 21/c
|
P 21/c
|
P 21/c
|
P 21/c
|
a (Å)
|
7.745
|
7.750
|
7.754
|
7.751
|
b (Å)
|
5.504
|
5.510
|
5.508
|
5.509
|
c (Å)
|
5.474
|
5.461
|
5.463
|
5.462
|
α = γ
|
90
|
90
|
90
|
90
|
β
|
90.090
|
90.51
|
90.30
|
90.50
|
V (Å3)
|
233.35
|
233.14
|
233.31
|
233.18
|
d-spacing at (002)
(nm)
|
0.273675
|
0.27304
|
027314
|
0.27314
|
Fraction (%)
|
|
82.73
|
80.70
|
76.80
|
Bond length (Å)
|
|
|
|
|
Mn(1)-O(1)
|
1.941
|
1.857
|
2.016
|
2.048
|
Mn(1)-O(2)
|
1.912
|
1.920
|
2.083
|
1.834
|
2.013
|
2.048
|
1.837
|
2.084
|
Mn(2)-O(1)
|
1.994
|
2.005
|
1.926
|
1.888
|
Mn(2)-O(3)
|
1.938
|
1.916
|
2.136
|
1.937
|
1.991
|
2.008
|
1.769
|
1.998
|
Bond angle (°)
|
|
|
|
|
Mn(1)-O(1)-Mn(2)
|
159.500
|
161.717
|
159.146
|
159.994
|
Mn(1)-O(2)-Mn(1)
|
162.900
|
166.412
|
163.470
|
163.814
|
Mn(2)-O(3)-Mn(2)
|
162.200
|
162.421
|
166.653
|
160.543
|
Crystallographic parameter
|
Reference
|
|
Mn3O4
|
|
Mn3O4
COD:1514115 [34]
|
TSintering= 700oC
|
TSintering= 800oC
|
TSintering= 900oC
|
Crystal structure
|
Tetragonal
|
Tetragonal
|
Tetragonal
|
Tetragonal
|
Space group
|
I41/amd
|
I41/amd
|
I41/amd
|
I41/amd
|
a (Å)
|
5.765
|
5.756
|
5.757
|
5.756
|
b (Å)
|
5.765
|
5.756
|
5.757
|
5.756
|
c (Å)
|
9.442
|
9.462
|
9.467
|
9.466
|
α = γ = β
|
90
|
90
|
90
|
90
|
V (Å3)
|
313.81
|
313.47
|
313.80
|
313.60
|
Fraction (%)
|
|
17.27
|
19.20
|
23.10
|
Reliability factors
|
|
|
|
|
Rwp
|
|
14.6
|
16.3
|
16
|
χ2
|
|
1.57
|
1.89
|
1.87
|
The crystallite size was determined from the XRD patterns using the Debye-Scherrer formula described in Eq. (1). In this equation, D represents the crystallite size, \(\:K\) is a crystallite shape factor with a value of 0.9, \(\:\beta\:\) is the full width at half maximum (FWHM), and \(\:\lambda\:\) is the X-ray wavelength (\(\:\lambda\:\:\)= 1.5428 Å). The crystallite sizes are 32 nm, 33 nm, and 34 nm for the samples sintered at temperature of 700oC, 800oC, and 900oC, respectively. Table 2 presents the crystallite size estimated from XRD characterization and the particle size obtained from TEM measurements for LCSMO/Mn3O4 composite nanoparticles. It confirms that the higher sintering temperature produced bigger particle and crystallite size.
$$\:D=\frac{K\lambda\:}{\beta\:cos\theta\:}$$
1
Table 2
Crystallite size and particle size of La0.7Ca0.25Sr0.05MnO3/Mn3O4 composite nanoparticles.
|
La0.7Ca0.25Sr0.05MnO3/Mn3O4
|
|
TSintering= 700oC
|
TSintering= 800oC
|
TSintering= 900oC
|
Crystallite size (By XRD)
|
32 nm
|
33 nm
|
34 nm
|
|
Particle size (By TEM)
|
28 nm
|
30 nm
|
32 nm
|
|
Figures 5 (a) and (b) display the temperature dependence of resistivity for LCSMO/Mn3O4 composite nanoparticles, both without and with a 20 kOe applied magnetic field. Particle size and fraction of Mn3O4 are shown in the legend of the figure to illustrate their effects on the transport properties of this system. Ferromagnetic-metallic behavior is observed below 170 K, followed by paramagnetic-insulator behavior above 170 K. The temperature at which this transition occurs is defined as the metal-insulator transition temperature, TMI. The TMI is determined by selecting the temperature at which the resistivity reaches the maximum value. Without the applied magnetic field, TMI is 172 K, 180 K, and 172 K for particle sizes of 28 nm with 17% Mn3O4, 30 nm with 19% Mn3O4, and 32 nm with 23% Mn3O4, respectively. Under the applied magnetic field of 20 kOe, TMI is 184 K, 182 K, and 183 K for particle sizes of 28 nm, 30 nm, and 32 nm, respectively. It is inferred that without the magnetic field, TMI is highest for the 30 nm particle size with 19% Mn3O4, while TMI for the other particle sizes is the same. However, TMI has no difference among all samples under the applied magnetic field in 20 kOe. In addition, the 28 nm particle size with 17% Mn3O4 exhibits the highest resistivity, both with and without an applied magnetic field.
The shift in TMI to the higher temperature region and the change in resistivity suggests that the amount of Mn3O4 fraction disrupts electron transport in this system. The change in the Mn3O4 fraction might alter the ratio of Mn3+ to Mn4+ ions, affecting the double exchange (DE) interaction. This interaction involves electron hopping from a Mn3+ ion to neighboring Mn4+ ion via an O2- ion[35]. The DE interaction relies on the oxygen 2p-orbital and the Mn-3d orbitals, making electron hopping in perovskite manganite sensitive to the conduction bandwidth, W (a.u.)[36] as described in Eq. (2), where \(\:{\theta\:}_{Mn-O-Mn}\) and \(\:{d}_{Mn-O}\) are the bond angle and bond distance of Mn-O-Mn, respectively[36].
$$\:W\propto\:\frac{cos\left[180-\left({\theta\:}_{Mn-O-Mn}\right)\right]}{{d}_{Mn-O}^{3.5}}$$
2
The conduction bandwidth is 0.0918, 0.0938, and 0.0905 (a.u.) for particle sizes of 28 nm with 17% Mn3O4, 30 nm with 19% Mn3O4, and 32 nm with 23% Mn3O4, respectively. This result suggests that the enhancement in the conduction bandwidth causes a reduction in resistivity and TMI shifts towards the higher temperature region. In this regard, the 30 nm sample with 19% Mn3O4 exhibits the lowest resistivity due to the appearance of the highest conduction bandwidth that originates from the shortest Mn-O bond distance compared to other samples and Mn-O-Mn bond angle closer to 180°, facilitating easier electron hopping.
The upper left inset of Fig. 5 (a) and (b) shows an increase in resistivity in the low-temperature region, specifically below 50 K. The temperature at which the resistivity starts to increase in this region is referred to as Tmin. Tmin is about 50 K without an applied magnetic field and about 40 K with an applied magnetic field of 20 kOe. The increase in resistivity is more distinct in smaller particle sizes than in larger ones. The increase in resistivity at low temperatures likely originates from the Kondo-like scattering [37], [38]. However, further investigation about the Kondo effect in this system is needed.
To better comprehend the complex transport mechanism influenced by particle size reduction and variations in the Mn3O4 fraction, the resistivity is fitted using the percolation model. The resistivity is analyzed using Eq. (3) over the temperature range. The detailed derivation of the percolation model has been explained elsewhere [37–42]. In this model, the various contributions to resistivity are represented as follows: \(\:{\rho\:}_{0}\) represents the contribution from grain boundaries, \(\:{\rho\:}_{e}\:\)represents the contribution from electron-electron interaction, \(\:{\rho\:}_{s}\) represents the contribution from Kondo-like spin-dependent scattering, \(\:{\rho\:}_{p}\) represents the contribution from electron-phonon interaction, \(\:{\rho\:}_{2}\) represents the contribution from electron-electron scattering, \(\:{\rho\:}_{\raisebox{1ex}{$9$}\!\left/\:\!\raisebox{-1ex}{$2$}\right.}\)represents the contribution due to the combination of electron-electron, electron-magnon, and electron-phonon scattering, and \(\:{\rho\:}_{\alpha\:}\)represents the contribution from residual resistivity. T denotes temperature, U0 represents the energy difference below \(\:{T}_{C}^{mod}\) or between the ferromagnetic-metallic and paramagnetic-insulator phases, \(\:{T}_{C}^{mod}\) stands for the theoretical Curie temperature, EA denotes activation energy, and kB denotes the Boltzmann constant. The percolation fitting function formula for the temperature dependence of resistivity is applied to the experimental data depicted in Fig. 6. The black solid lines represent the best-fitting results. In contrast, the colored markers represent the experimental data of three samples. Fitting parameters are summarized in Table 3.
$$\:\rho\:\left(T\right)=\:\left[{\rho\:}_{\alpha\:}Texp\left(\frac{{E}_{A}}{{k}_{B}T}\right)\frac{exp\left(\frac{-{U}_{0}\left(1-\frac{T}{{T}_{c}^{mod}}\right)}{{k}_{B}T}\right)}{1+exp\left(\frac{-{U}_{0}\left(1-\frac{T}{{T}_{c}^{mod}}\right)}{{k}_{B}T}\right)}\right]\left[{\rho\:}_{0}+{\rho\:}_{e}{T}^{\raisebox{1ex}{$1$}\!\left/\:\!\raisebox{-1ex}{$2$}\right.}-{\rho\:}_{s}\text{ln}T+{\rho\:}_{p}{T}^{5}+{\rho\:}_{2}{T}^{2}+{\rho\:}_{\raisebox{1ex}{$9$}\!\left/\:\!\raisebox{-1ex}{$2$}\right.}{T}^{\raisebox{1ex}{$9$}\!\left/\:\!\raisebox{-1ex}{$2$}\right.}\left(\frac{1}{1+exp\left(\frac{-{U}_{0}\left(1-\frac{T}{{T}_{c}^{mod}}\right)}{{k}_{B}T}\right)}\right)\right]\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(3\right)$$
3
Table 3
The best fitting parameters obtained from percolation model for La0.7Ca0.25Sr0.05MnO3/Mn3O4 composite nanoparticles.
Fitting parameters | La0.7Ca0.25Sr0.05MnO3/Mn3O4 |
---|
28 nm with 17% of Mn3O4 | 30 nm with 19% of Mn3O4 | 32 nm with 23% of Mn3O4 |
---|
\(\:{\rho\:}_{0}\) (Ωm) | 46.13 | 8.32 | 18.38 |
\(\:{\rho\:}_{e}\) (ΩmK-1/2) | 0.11 | 0.12 | 1.42 |
\(\:{\rho\:}_{s}\) (Ωm) | 5.59 | 1.08 | 1.18 |
\(\:{\rho\:}_{p}\) (×10–10 ΩmK-5) | 42.74 | 3.14 | 54.23 |
\(\:{\rho\:}_{2}\) (×10− 4 ΩmK-2) | 14.80 | 1.79 | 7.59 |
\(\:{\rho\:}_{9/2}\) (× 10− 9 ΩmK-2) | 57.24 | 4.37 | 2.47 |
\(\:{\rho\:}_{\alpha\:}\) (Ωm) | 0.21 | 0.23 | 0.22 |
U0/kB (K) | 1246.3 | 1880.6 | 1165.7 |
EA/kB (K) | 335.9 | 52.4 | 303.4 |
\(\:{T}_{c}^{mod}\) (K) | 270.0 | 275.7 | 272.3 |
R2 (%) | 99.92 | 99.95 | 99.90 |
Among all fitting parameters, the contribution from grain boundaries is the most dominant compared to other parameters. The predominant role of grain boundaries in resistivity arises from electron tunneling between grains, as described by the core-shell model. According to this model, electron motion is significantly affected by the inter-grain distances, where the reduction in the inter-grain distances facilitates an enhancement of electrical conductivity[37–41]. Grain boundary contribution, \(\:{\rho\:}_{0}\) is 46.13, 8.32, and 18.38 Ωm for particle sizes of 28 nm with 17% of Mn3O4, 30 nm with 19% of Mn3O4, and 32 nm with 23% of Mn3O4, respectively. The sample with a particle size of 30 nm with 19% of Mn3O4 has the lowest grain boundary contribution compared to other samples. The lower grain boundary contribution enhances inter-grain electron movement, leading to the lowest electrical resistivity in this sample. The grain boundary contribution has no linear relationship with particle size.
The 30 nm sample with 19% of Mn3O4 has the smallest activation energy (EA/kB), six times smaller than other samples. This lower activation energy means less energy is required for electron to hop between localized states [43]. The smallest activation energy observed in the 30 nm sample with 19% of Mn3O4 compared to the other samples is likely due to the larger conduction bandwidth. The conduction bandwidth significantly influences the double exchange (DE) interaction, with a larger bandwidth strengthening the DE interaction. This stronger DE interaction reduces the energy required for electrons to hop between localized states, resulting in lower activation energy [42]. Thus, the small activation energy in the 30 nm sample with 19% of Mn3O4 aligns with its large conduction bandwidth, where a larger bandwidth leads to lower activation energy.
It shows that the 30 nm sample with 19% of Mn₃O₄ exhibits superior electron conduction compared to the other samples due to the lowest contribution from grain boundaries and the smallest activation energy. The graph for the grain boundary contribution and activation energy (EA/kB) for all particle sizes in La0.7Ca0.25Sr0.05MnO3/Mn3O4 composite nanoparticles shows in Fig. 7 (a) and (b), respectively. In addition, Table 3 shows that the second-highest contribution to resistivity is from Kondo-like scattering, which increases as particle size decreases. This is consistent with the results in the upper left inset of Fig. 5 (a) and (b), where the increase in resistivity at low temperatures is more distinct for smaller particle.
Figure 8 shows the magnetic field dependence of magnetoresistance for the LCSMO/Mn3O4 composite nanoparticles. The magnetoresistance effect is calculated by using Eq. (4), where ρ(H) represents resistivity under an applied magnetic field and ρ(0) represents resistivity without an applied magnetic field.
$$\:MR\left(\%\right)=\frac{\rho\:\left(H\right)-\rho\:\left(0\right)}{\rho\:\left(0\right)}100\%$$
4
It demonstrates that the resistance decreases when applying a magnetic field below 5 kOe, indicating that LFMR is observed in this system only with a low magnetic field application. The magnetoresistance observed when applying a low magnetic field is called the LFMR phenomenon. The LFMR value of the LCSMO/Mn3O4 composite nanoparticles with a magnetic field of 5 kOe at 10 K increased from 25–27% by reducing the particle size from 32 nm to 30 nm. The LFMR value even reach up to 30% at magnetic field of 5 kOe at 5 K for particle size 28 nm. This value is higher than that reported by Vertruyen et al. in 2007 for La0.7Ca0.3MnO3/Mn3O4 composite, which reached only 13% with a magnetic field of 5 kOe at 10 K [10] and comparable with 22‒28% MR ratio with a magnetic field of 5 kOe at 20 K by Vertruyen et al. in 2005 [13]. At 40 K, the LFMR value of the LCSMO/Mn3O4 composite nanoparticles is approximately 23%, still higher than that for pure LCMO, 22% MR Ratio with magnetic field of 3 kOe at 30 K by D. H. Manh. et al. 2010 [14].
If we compared LFMR ratio for our LCSMO/Mn3O4 at 100K for 5 kOe to La0.7Ca0.3MnO3/Mn3O4 composite by Bhame et al. that shows 16% MR ratio at 77 K for 5 kOe[17] and LCMO by Gaur et al. 2010 that MR ratio reach up to 28% at 10 kOe in 80 K[15], our LFMR ratio of LCSMO/Mn3O4 composite nanoparticles is also still comparable that reached up to 20%. Lastly, at 170 K, all samples exhibit an MR ratio of approximately 10%. It can be concluded that reducing particle size into nanoparticle size and adding the insulator composite Mn3O4 to LCSMO to become LCSMO/Mn3O4 composite shows the tendency to improve LFMR. This supports our hypothesis that lowering the sintering temperature produces particle size that reduces to nanoparticle size and adding the insulator composite can enhance the low-field magnetoresistance (LFMR). The MR ratio for La0.7Ca0.25Sr0.05MnO3/Mn3O4 composite nanoparticles and the reference is summarized in Table 4.
For a clear view, we plot the temperature dependence of the MR ratio 5 kOe and MR at 20 kOe for all the particle sizes of LCSMO/Mn3O4 composite nanoparticles in Fig. 9. The temperature dependence of the magnetoresistance ratio with a magnetic field of 5 kOe and 20 kOe for LCSMO/Mn3O4 composite nanoparticles shows that the MR ratio decreases as the temperature increases for all particle sizes. This occurrence is expected due to the thermal energy disrupting the alignment of the magnetic moment, leading to a weakening of the MR ratio. The MR ratio with a magnetic field of 20 kOe is always higher than that of 5 kOe at all temperatures. This is because the higher magnetic field induced more spin alignment, resulting in a more significant decrease in resistivity so that the MR ratio increases[41]. In the case of particle size dependence, there is no significant difference in MR ratio with applied field 5 kOe in the range of particle size 28 until 32 nm below 170 K. The similar behavior also occurred when the applied field is 20 kOe.
Table 4
Comparison of MR ratio for La0.7Ca0.25Sr0.05MnO3/Mn3O4 composite nanoparticles from this study and reference data.
Magnetoresistance ratio |
---|
Present study | Reference data |
28 nm | 30 nm | 32 nm |
30% (5 kOe at 5 K) | 27% (5 kOe at 10 K) | 25% (5 kOe at 10 K) | 13% (5 kOe at 10 K) La0.7Ca0.3MnO3/Mn3O4 [10] |
22‒28% (5 kOe at 20 K) La0.7Ca0.3MnO3[13] |
23% (5 kOe at 40 K) | 24% (5 kOe at 40 K) | 23% (5 kOe at 40 K) | 22% (3 kOe at 30 K) La0.7Ca0.3MnO3 with particle size of 16 nm [14] |
17% (5 kOe at 100 K) | 20% (5 kOe at 100 K) | 15% (5 kOe at 100 K) | 16% (5 kOe at 77 K) La0.7Ca0.3MnO3/Mn3O4 with crystallite size of 35 nm [17] |
28% (10 kOe at 80 K) La2/3Ca1/3MnO3 with particle size of 25 nm[15] |
9% (5 kOe at 170 K) | 9% (5 kOe at 170 K) | 10% (5 kOe at 170 K) | |