Magnetic properties, critical behavior, and magnetocaloric effect of Nd_1−xSr_xMnO₃ (0.2 ≤ x ≤ 0.5): The role of Sr doping concentration

Perovskite manganites exhibit a diverse range of physical properties, including spin and orbital ordering,¹ piezoelectric properties,² ferroelectric behavior,³ ionic conduction,⁴ superconductivity,⁵ magnetoresistance (MR) effects,⁶ and photovoltaic effects.⁷ These properties are significant for both practical applications and fundamental science. In rare earth manganites RMnO₃ (R representing trivalent rare earth elements), the magnetic and electrical properties are mainly influenced by the transition metal ions and their oxidation states. This includes phenomena such as charge-orbital ordering,⁸ spin-polarized transport,⁹ half-metallic ferromagnetism,¹⁰ and multiferroic behavior.¹¹ Among RMnO₃ compounds, NdMnO₃ stands out as a unique material. Below its Curie temperature (T_C) of 75 K, Nd and Mn magnetic moments exhibit ferromagnetic coupling.^12,13 Both stoichiometric and non-stoichiometric NdMnO₃ show various interesting physical and chemical properties, such as phase separation,^14–20 positive and negative exchange bias,²¹ magnetoelastic effects,²² and negative magnetization with magnetization reversal.^{15–13,21–25} Sudakshina et al. found that the Néel temperature (T_N) of NdMnO₃ is 73.51 K, with negative magnetization observed below 2 K.²³ Research by Hong et al. revealed that the magnetic moment in NdMnO₃ becomes negative below 13 K, with the material exhibiting A-type antiferromagnetic (AFM) ordering and a ferromagnetic (FM) ordering temperature (T_C) of 79 K.²¹ Kumar et al. observed Nd ordering at 15 K and A-type AFM ordering around 75 K in NdMnO_3.²⁴ Additionally, NdMnO₃ displays a significant magnetocaloric effect, with a magnetic entropy change (−ΔS_M) of 4.4 J/kg Sudakshina K near 15 K under an external field of μ₀ΔH = 50 kOe.²⁵ Al-Yahmadi et al. reported a maximum magnetic entropy change of 1.61 J/kg K at 62 K under the same field.²⁶ Saravanan et al. discovered that under an external magnetic field of μ₀ΔH = 50 kOe, the transition from paramagnetic (PM) to antiferromagnetic (AFM) occurs at 56 K, with magnetic entropy changes of 1.85 J/kg K at 60 K, increasing to 2.5 J/kg K at 15 K.²⁷ Kumar and colleagues found that NdMnO₃ undergoes transitions at 15 and 75 K. Under an applied magnetic field of μ₀ΔH = 10 kOe at 15 K, the magnetic entropy change is −2 J/kg K, while under μ₀ΔH = 20 kOe at 73 K, the magnetic entropy change is 0.7 J/kg K.²⁴ 

Most studies on NdMnO₃ have identified two transition temperatures related to magnetic entropy. However, our previous research on the magnetic entropy and critical behavior of NdMnO₃ revealed three distinct maximum magnetic entropy change values, including both positive and negative changes.²⁸ A maximum (−ΔS_M) value of 3.82 J/kg⋅K was observed at μ₀ΔH = 50 kOe and 10–15 K. In the 80–85 K temperature range, a smaller (−ΔS_M) value of 1.22 J/kg K was obtained under an applied magnetic field of μ₀ΔH = 50 kOe. Additionally, at μ₀ΔH = 5 kOe and around 8 K, a maximum negative (−ΔS_M) of −0.557 J/kg K was observed, attributed to the alignment of Mn moments and the ordering of Nd sublattices. Analysis of the critical behavior confirmed the reliability of the KF relation, with an estimated magnetic exchange distance J(r) of approximately 4.563, placing it between the three-dimensional Heisenberg model²⁹ and the mean-field model.³⁰ This behavior can be explained by the antiferromagnetic coupling between the ferromagnetic Mn components at low temperatures around 12 K, which induces the ordering of Nd. The ordered Nd moments cause a spin reorientation of Mn, resulting in magnetization reversal and negative magnetization. At around 53 K, the high-spin state of Mn involved in antiferromagnetic interactions becomes significant, while at around 82 K, the ferromagnetic Mn components exhibit ordering in the a–b plane of NdMnO₃.

For RMnO₃ compounds, doping at the R-site often leads to various intriguing physical properties in manganites.^31–35 Currently, much attention has been focused on A-site substitution, where divalent cations replace trivalent rare-earth ions (R³⁺), generating Mn⁴⁺ ions due to charge compensation. Generally, RMnO₃ exhibits antiferromagnetism and significant Jahn–Teller (J-T) distortion. However, R-site doping can induce double exchange (DE) interactions, reducing the J-T distortion and simultaneously introducing metallic and ferromagnetic characteristics.³⁶ Near the PM insulator to FM metal transition, giant negative magnetoresistance (MR) and magnetic entropy changes are often observed. Doping and temperature variations alter the Mn³⁺/Mn⁴⁺ ion distribution, leading to changes in Mn—O bond lengths and causing structural disorder. This disorder is closely related to the static or dynamic properties of the lattice, orbital degrees of freedom, and charge, which remain topics of ongoing debate. Researchers are particularly interested in understanding which universal principles govern magnetic phase transitions and the mechanisms that normalize interactions near critical points. The critical phenomena of the FM to PM transition in DE systems can be studied using the mean-field method,³⁰ the Ising model,²⁶ and the 3D Heisenberg model.²⁹ Among RMnO₃ oxides, studies on Nd-based compounds are less extensive. The smaller Nd³⁺ ion leads to larger lattice distortions and weaker DE interactions, making Nd-based perovskite manganites particularly intriguing.³⁷ In doped systems like Nd_1−x(Ca, Sr)_xMnO₃, compared to La_1−x(Ca, Sr)_xMnO₃, there is a heightened competition among electron–phonon, electron–electron, and DE interactions. This competition often accompanies pervasive instabilities such as antiferromagnetic superexchange, orbital ordering, and charge ordering.³⁸ Therefore, doping at the Nd site in NdMnO₃ can induce a richer array of magnetic properties, offering fertile ground for further investigation.

In our previous work,²⁸ we investigated the spin reorientation and critical behavior of NdMnO₃. In the current study, we examine the effect of Nd-site doping on its magnetic and magnetocaloric properties and further explore its critical behavior. Sr was chosen as the dopant at the Nd site due to the promising MR and magnetocaloric effects exhibited by Nd_1−xSr_xMnO₃. Al-Yahmadi et al.²⁶ found a maximum magnetic entropy change (−ΔS_M) value of 2.78 J/kg K at 127 K with an applied field of μ₀ΔH = 50 kOe and a relative cooling power (RCP) value of 155.5 J/kg in Nd_0.8Sr_0.2MnO₃. Fkhar et al.³⁹ reported a maximum (−ΔS_M) of 3.12 J/kg K at 257.5 K and μ₀ΔH = 10 kOe in Nd_0.67Sr_0.33MnO₃. Similarly, Xu et al.⁴⁰ observed a maximum (−ΔS_M) of 3.12 J/kg K under μ₀ΔH = 30 kOe at 274 K, with an RCP value of 71 J/kg in Nd_0.55Sr_0.45MnO₃. Ahmed et al.⁴¹ studied the MR of Nd_0.6Sr_0.4MnO₃ and found that MR could reach 12% at 273 K under a magnetic field of 0.6 T. These studies highlight the excellent potential of the Nd_1−xSr_xMnO₃ system for magnetic sensor and refrigeration.

Therefore, we systematically study the doping effect of polycrystalline Nd_1−xSr_xMnO₃ (0.2 ≤ x ≤ 0.5) with NdMnO₃ as the parent compound, investigating the critical behavior through analysis of its magnetic and magnetocaloric effects. Our findings demonstrate that Sr doping in NdMnO₃ (Nd_1−xSr_xMnO₃) induces a ferromagnetic (FM) transition, resulting in significant magnetocaloric effects. A discrepancy between T_CW and T_C can be observed due to the competition between polarons and DE interaction, which is different from previous similar reports on NdMnO₃ by Sr doping.^26,40 The critical behavior analysis via using the self-consistent (modified Arrott plots, MAP) method and the Kouvel–Fisher (KF) relation reveals that Nd_1−xSr_xMnO₃ samples with doping concentrations of x = 0.2, 0.3, and 0.4 adhere to the mean-field model. However, the sample with a doping concentration of x = 0.5 (Nd_0.5Sr_0.5MnO₃) does not conform to any classical model due to the presence of a charge-ordered antiferromagnetic (AFM) phase. For the sample with x = 0.3, (−ΔS_M) reached a maximum of 4.315 J/kg K at μ₀ΔH = 50 kOe, corresponding to a relative cooling power (RCP) of 280.48 J/kg. Remarkably, the x = 0.4 sample displayed (−ΔS_M) of 3.298 J/kg K at μ₀ΔH = 50 kOe near room temperature, with the RCP of 283.64 J/kg. These findings underscore their potential for magnetic refrigeration applications, meanwhile highlight the important role of Sr doping in tuning the magnetic properties, critical behavior, and magnetocaloric effect of NdMnO₃.

Polycrystalline Nd_1−xSr_xMnO₃ (0.2 ≤ x ≤ 0.5) samples were prepared using the high-temperature solid-state reaction method. Initially, stoichiometric amounts of Nd₂O₃, SrCO₃, and MnO₂ powders (Aladdin) with a purity of 99.99% were thoroughly mixed and ground. The mixtures were then calcined at 1100 °C for 24 h. After furnace cooling, the obtained powders were ground again and subjected to a second calcination at 1200 °C for 24 h. The resultant powders were pressed into disc-shaped pellets with a diameter of approximately 1 cm and sintered continuously at 1300 °C for 24 h. All the samples were obtained after furnace cooling.

The crystal structure of the samples was analyzed using x-ray diffraction (XRD) with Cu-K_α radiation (wavelength: 1.5405 Å). The surface morphology of the samples was examined using a scanning electron microscope (SEM), and the elemental composition was analyzed with an energy dispersive spectrometer (EDS). For magnetic analysis, a physical property measurement system (PPMS) incorporating a vibrating sample magnetometer (VSM) was utilized. The relationship between magnetization and temperature (M-T) was examined using different procedures: field-cooled-warming (FCW), field-cooled-cooling (FCC), and zero-field-cooled (ZFC). These procedures were carried out under four distinct magnetic field strengths: 100, 500, 2000, and 10 000 Oe. Magnetization hysteresis loops (M-H) were recorded in the range from −50 to 50 kOe at various temperatures. Additionally, isothermal magnetization curves were measured across different temperature ranges to calculate the corresponding magnetic entropy change for each sample. Electrical transport measurements were performed using the PPMS. The temperature dependence of resistivity (R-T) was measured using the four-probe method under magnetic fields of 0, 1, 2, 4, and 6 T.

The room temperature XRD measurements and Rietveld refinement results for the Nd_1−xSr_xMnO₃ (0.2 ≤ x ≤ 0.5) samples are shown in Figs. 1 and 2(a)–2(h). Figures 2(e)–2(h) display the crystal structures obtained after refinement, viewed from different orientations. The positions of XRD diffraction peaks for the x = 0.2 and x = 0.3 samples are almost consistent with the reference Nd_0.7Sr_0.3MnO₃ (PDF No. 04-008-4443), confirming a single orthorhombic phase with the space group Pnma (No. 62). Similarly, the diffraction peak positions for the x = 0.4 and x = 0.5 samples align with the reference Nd_0.5Sr_0.5MnO₃ (PDF No. 04-011-4383), confirming a single orthorhombic phase with the space group Imma (No. 74). Table I presents the structural parameters obtained from Rietveld refinement. The goodness-of-fit values (χ²) are below 1.5, indicating the reliability of the refinement results.

Figures 3(a)–3(d) show the surface morphology of the Nd_1−xSr_xMnO₃ (0.2 ≤ x ≤ 0.5) perovskites observed using SEM. The agglomerated particles are relatively small, densely distributed, and irregularly oriented. The insets in the top right corners display the average grain sizes, calculated to be 1.90 μm for the x = 0.2 sample, 1.52 μm for the x = 0.3 sample, 1.43 μm for the x = 0.4 sample, and 2.11 μm for the x = 0.5 sample. The left insets in Figs. 3(a)–3(d) reveal a uniform elemental distribution. The EDS spectra of the Nd_1−xSr_xMnO₃ (0.2 ≤ x ≤ 0.5) samples were measured to study their compositions and the compositions are analyzed in Table II. The differences between the experimental and nominal elemental compositions are very small and can be neglected.

In the relation, g is the Landé g-factor. It is noteworthy that transition elements undergo orbital “freezing” when forming metals and compounds, meaning that the orbital magnetic moment is quenched by the crystal field and does not contribute to magnetism. In this case, L = 0, J = S, g_J = 2. Therefore, for Mn³⁺: J = S = 2, g_J = 2, and for Mn⁴⁺: J = S = 1.5, g_J = 2. The calculated values are listed in Table III.

In the undoped sample (x = 0), distinct antiferromagnetic properties are observed.²⁸ Introducing Sr2 + alters the material's characteristics: as x increases, changes occur in the Curie temperature, net magnetization, and spontaneous magnetization, likely influenced by the evolution of double exchange (DE) interactions. With partial substitution of Nd by Sr (0 < x < 0.1), a canted spin structure emerges. As x slightly increases further, the system transitions to a ferromagnetic state while remaining insulating. For compositions where 0.2 ≤ x ≤ 0.48, the material exhibits a ferromagnetic metallic phase below T_C and transforms into a paramagnetic insulating phase above this temperature.⁴³ Optical studies on Nd_0.7Sr_0.3MnO₃ thin films have provided evidence of dynamic Jahn–Teller effects and the presence of magnetic polarons.^44,45 Additionally, chemical potential investigations by Ebata et al. have confirmed the existence of these polarons.⁴⁶ Below T_C, DE interactions dominate, whereas above T_C, polarons become significant, contributing to the complex magnetic behaviors observed.

Figures 4(a)–4(d) depict the magnetization curves of Nd_1−xSr_xMnO₃ (0.2 ≤ x ≤ 0.5) samples in the temperature range of 10–350 K under an external magnetic field of 100 Oe. At low temperatures, the divergence between ZFC and FC magnetization-temperature (M-T) curves suggests a spin glass state^5,28 or other forms of magnetic inhomogeneity. As temperature decreases, a transition from PM to FM state becomes evident. The insets in Figs. 4(a)–4(d) show peaks in dM/dT under the FC mode around 152 K, indicative of ferromagnetic ordering induced by double exchange (Curie temperature, T_C). Notably, Fig. 4(a) reveals a significant deviation of approximately 26 K between T_C derived from the magnetization curve (dM/dT) and the Curie–Weiss temperature (T_CW) obtained from fitting, possibly attributed to polaron effects above T_C. A Griffith-like phase, observed in the x = 0.2 sample,⁵ designates the temperature as the Griffith temperature (T_G). Figure 4(b) displays the magnetization curve for the x = 0.3 sample, showing a PM to FM transition with T_C around 217 K. The inset of Fig. 4(b) confirms a peak in dM/dT under the FC mode at this temperature. The higher doping level (increased Mn⁴⁺ content) promotes double-exchange interactions between Mn³⁺-O-Mn⁴⁺, resulting in elevated T_C and spontaneous magnetization compared to the x = 0.2 sample. Similar magnetic behavior persists, including spin glass characteristics at low temperatures, with T_C differing from T_CW by approximately 9 K. The Griffith-like phase is also present in the x = 0.3 sample. Figure 4(c) exhibits the magnetization curve for the x = 0.4 sample, demonstrating increased spontaneous magnetization and a higher TC. In its inset, dM/dT under the FC mode shows a peak around 285 K, approaching room temperature. According to Al-Yahmadi, substituting 40% of Nd³⁺ with Sr²⁺ establishes strong double-exchange interactions via Mn³⁺—O—Mn⁴⁺ bonds, achieving a nearly equal distribution of Mn³⁺ (δ³⁺) and Mn⁴⁺ (δ⁴⁺), resulting in a high Mn⁴⁺/Mn³⁺ ratio of 97%.²⁶ Consequently, the x = 0.4 sample exhibits the highest spontaneous magnetization among the concentrations studied, with T_C and T_CW nearly identical, highlighting pronounced double-exchange interactions.

With Sr²⁺ doping reaching 50%, a decrease in net magnetization and T_C is observed, attributed to the increase in antiferromagnetic superexchange interactions. Figure 4(d) shows the magnetization curve for the x = 0.5 sample, with the inset displaying dM/dT in the FC mode. A PM to FM transition occurs around 290 K. At the half-doped concentration of x = 0.5, as the temperature decreases from the PM to FM state and further, a charge-exchange-type antiferromagnetic charge-ordered phase emerges, reducing spontaneous magnetization.⁴⁶ 

Figures 5(a)–5(d) present the isothermal magnetization curves for Nd_1−xSr_xMnO₃ (0.2 ≤ x ≤ 0.5) in the external magnetic field range of 0–50 000 Oe. The x = 0.2 sample is measured for magnetization at different temperatures from 100 to 250 K, the x = 0.3 sample from 175 to 300 K, and the x = 0.4 and x = 0.5 samples from 200 to 320 K, to evaluate potential magnetic entropy. Figures 6(a)–6(d) display Arrott diagrams showing the relationship between the square of the magnetization (M²) and H/M, providing insight into the nature of the phase transition. In the M² range, a positive slope indicates a second-order phase transition under the applied magnetic field, according to the Banerjee criteria.⁴⁷ 

Figures 8(a)–8(h) depict the scaled magnetization intensity and magnetic field data separated into two universal curves above and below T_C. The results indicate that the x = 0.5 sample cannot be collapsed into two universal curves, possibly due to the presence of its charge-ordered superexchange antiferromagnetic phase. In contrast, other doping ratios show well-scaled curves. Specifically, the x = 0.2 sample demonstrates better scaling and a clearer collapse into two universal curves using critical exponents calculated via the Kouvel–Fisher relation. For the x = 0.3 and x = 0.4 samples, the critical exponents obtained through the self-consistent method (MAP) provide better scaling of the universal curves. These observations provide initial insights into the more suitable method, and these parameters will be further validated in subsequent steps.

Figure 9 presents the isothermal magnetization curves at temperatures corresponding to the fitted T_C values: 150 K for the x = 0.2 sample in Fig. 9(a), 220 K for the x = 0.3 sample in Fig. 9(b), 280 K for the x = 0.4 sample in Fig. 9(c), and 265 K for the x = 0.5 sample in Fig. 9(d). The insets show slopes obtained from ln–ln scaled critical isotherms, specifically at T_C, as per Eq. (6). Reciprocals of these slopes yield the δ values. The critical exponent δ must satisfy the Widom scaling relation $δ=1+γ/β$⁠.²⁸ Different δ values can be derived using the MAP method and the KF relation, listed in Table IV. Following comparison, the fitting method that aligns more closely with the expected relation is selected, and the parameters calculated via this method are chosen.

For the x = 0.2 sample, the δ value calculated from critical exponents obtained using the KF relation closely matches the δ value fitted from the critical isotherm, consistent with the universal curve. Therefore, the critical parameters derived from the KF relation are chosen. Similarly, for the x = 0.3 and x = 0.4 samples, the δ values calculated from critical exponents obtained using the MAP align better with the δ values fitted from the critical isotherm, consistent with the universal curve. Hence, the critical parameters obtained from the self-consistent method are selected. Regarding the x = 0.5 sample, where the critical exponents are less reliable, a δ value is still fitted from the critical isotherm to maintain data integrity and verify the reliability of the calculated critical exponents. A significant difference of approximately 0.5 is observed compared to the δ values of the other ferromagnetic phases. The KF relation, which produces a smaller difference, is chosen for further verification.

Apart from the x = 0.5 sample, which has coexisting antiferromagnetic phases, the reliable calculation methods for the x = 0.2, 0.3, and 0.4 samples were selected based on the universal curves and δ comparisons. The critical parameters obtained for these samples were then compared with standard models. The results indicate that these parameters are closer to the mean-field model, as shown in Table V.

Figure 10 illustrates the magnetocaloric properties of Nd_1−xSr_xMnO₃ (0.2 ≤ x ≤ 0.5) by depicting the relationship between magnetic entropy change and temperature. The magnetic entropy change (−ΔS_M) represents the temperature variation in a material during magnetization and demagnetization processes under varying magnetic fields. According to the Maxwell relations, the magnetic entropy values (−ΔS_M) are calculated from the M-H curves using

$M i + 1( T i + 1 , H i), M i( T i , H i)$⁠, where are the magnetization rates at different temperatures and at the same magnetic field, respectively.

Figures 10(a)–10(d) illustrate the relationship between magnetic entropy change (−ΔS_M) and temperature under various applied magnetic fields. As expected, the peak values of (−ΔS_M) occur near the transition region (T_C) and increase with stronger magnetic fields. Under an applied field of μ₀ΔH = 50 kOe, the maximum magnetic entropy change (−ΔS_M) varies across the different compositions of Nd_1−xSr_xMnO₃. For the x = 0.2 sample, the maximum (−ΔS_M) reaches 3.108 J/kg K at 160 K. The x = 0.3 sample exhibits the highest (−ΔS_M) value of 4.315 J/kg K at 235 K. As the Sr content increases, the x = 0.4 sample shows a (−ΔS_M) of 3.298 J/kg K at 290 K. The x = 0.5 sample demonstrates the lowest (−ΔS_M) value of 2.433 J/kg K at 280 K among the studied samples. The three-dimensional (3D) magnetic entropy plots are shown in Figs. 11(a)–11(d).

Using the above equation, we can fit the data for magnetic entropy change and RCP, as shown in Figs. 12(a)–12(d). The obtained values of n and δ are listed in Table VI. Comparing n vs δ obtained from the critical parameters, the samples with x = 0.2, 0.3, and 0.4 all match well, while the sample with x = 0.5 shows a significant difference. This discrepancy may be related to its charge-ordered antiferromagnetic phase, consistent with the previous analysis, thus proving the reliability of these results.

The relationship between ΔC_P,H and temperature for μ₀ΔH in the range of 1–50 kOe is shown in Figs. 13(a)–13(d). As the temperature increases, the heat capacity undergoes a regular transition from negative to positive values, corresponding to the Curie temperature T_C transition point. Below T_C, the heat capacity exhibits negative values, while above T_C, the heat capacity exhibits positive values, reflecting a change in the magnetic phase.

The structure, magnetic characteristics, critical behavior, and MCE of Nd_1−xSr_xMnO₃ (0.2 ≤ x ≤ 0.5) were systematically studied using XRD, SEM, and magnetization techniques. All samples exhibit a single orthorhombic phase, but with different space groups: Pnma (No. 62) for x = 0.2 and 0.3 samples, and Imma (No. 74) for x = 0.4 and 0.5 samples. T_C of these samples are 152, 217, 283, and 291 K, respectively, while T_CW are 178, 226, 285, and 285 K, respectively, showing a deviation from T_C. The increase in T_C and spontaneous magnetization is attributed to the development of DE interaction, facilitated by the substitution of Sr²⁺, which introduces mobile Mn⁴⁺ bonds forming Mn³⁺—O—Mn⁴⁺ bonds and influencing Mn positioning. The difference between T_CW and T_C arises from the competition between polarons and double-exchange, with DE dominating below T_C and polarons becoming significant above T_C, resulting in complex magnetic behaviors. Samples with doping concentrations x = 0.2 and 0.3 exhibit a Griffiths-like phase. The slight variation in T_C and reduced spontaneous magnetization observed in the x = 0.5 sample is due to the presence of a charge-ordered superexchange antiferromagnetic interaction. All samples demonstrate significant MCE, with the x = 0.3 sample showing a maximum (−ΔS_M) of 4.315 J/kg K and a corresponding RCP value of 280.48 J/kg under an applied field of μ₀ΔH = 50 kOe. Near room temperature (290 K), the x = 0.4 sample exhibits a (−ΔS_M) of 3.298 J/kg K and an RCP value of 283.64 J/kg under similar conditions. The KF relation is suitable for the samples with x = 0.2 and 0.5, while the MAP method is suitable for the samples with x = 0.3 and 0.4. Except for the x = 0.5 sample, where the presence of the charge-ordered phase complicates critical behavior analysis, the samples with other doping concentrations generally adhere to the mean-field model.

This study on the doping effects of Nd_1−xSr_xMnO₃ provides new insights into the magnetic properties, critical behavior, and MCE of these important manganites. The observed magnetic behaviors and MCE are particularly significant for the potential application of perovskite manganites in magnetic refrigeration. This research is expected to stimulate renewed interest in manganites, prompting further investigation into how doping and structural modifications influence their diverse physical characteristics, including magnetoresistance, MCE, thermoelectric effects, dielectric properties, and optical properties.

This work was supported by the National Natural Science Foundation of China (Nos. 11604067 and U1832143) and the Scientific Research Project of the Department of Education of Zhejiang Province (No. Y202352427). We want to thank colleagues from Beijing Synchrotron Radiation Facility and Shanghai Synchrotron Radiation Facility for their great support.

The authors declare no conflict of interest.

Haiou Wang: Data curation (supporting); Formal analysis (equal); Investigation (lead); Project administration (equal); Writing – original draft (lead); Writing – review – editing (lead). Fuxiao Dong: Data curation (equal); Formal analysis (equal); Investigation (equal). Haochen Wang: Data curation (equal); Writing – original draft (supporting). Bojun Zhao: Data curation (equal); Investigation (supporting). Yan Wang: Data curation (equal); Writing – original draft (supporting). Weishi Tan: Formal analysis (equal); Investigation (supporting).

The data that support the findings of this study are available from the corresponding author upon reasonable request.