Negative Thermal Expansion Properties of Sm 0.85 Sr 0.15 MnO 3- δ

: A novel negative thermal expansion (NTE) material composed of Sm 0.85 Sr 0.15 MnO 3- δ was synthesized using the solid-state method. By allowing Sr 2+ to partially replace Sm 3+ in SmMnO 3 , the ceramic material Sm 0.85 Sr 0.15 MnO 3- δ exhibits NTE properties between 360K and 873K, and its average negative thermal expansion coefﬁcient was − 10.08 × 10 − 6 /K. The structure of Sm 0.85 Sr 0.15 MnO 3- δ is orthogonal, the space group is pbnm, the morphology is regular, and the grain size is uniform. The results of X-ray diffraction and XPS (X-ray photoelectron spectroscopy) suggest that the NTE phenomenon is related to the electron transfer of Mn ions. With the increase in temperature, Mn 4+ is rapidly transformed into Mn 3+ , accompanied by Mn 4+ O 6 octahedron distortion and oxygen defects. It was found that the sample volume continually decreased at the same time.


Introduction
We know that most instruments are composed of various materials, but with increases in temperature, different thermal expansion coefficients of various constituent materials may lead to thermal mismatches, and small cracks in the equipment can lead to performance failures and even instrument damage. NTE materials have attracted considerable research attention in the production of composites with accurately controllable positive, negative, or zero coefficients of thermal expansion [1][2][3][4][5][6][7][8][9][10][11].
Kurimamachiya-chouses conducted research on Sr 2+ partly substituting Gd 3+ in GdMnO 3 . They pointed out that Gd 1-x Sr x MnO 3-δ had excellent NTE properties [18]. L. J. Fu reported the NTE material of Er 0.7 Sr 0.3 NiO 3-δ with Sr 2+ partly substituting Er 3+ in ErNiO 3 [19]. These studies suggest that the substituting method is an effective way to prepare new kinds of NTE materials with excellent properties [16,[18][19][20]. In the present study, we conducted research on Sr 2+ partly substituting Sm 3+ in SmMnO 3 . The thermal properties are discussed.

Experimental Procedures
The sample was prepared according to the conventional solid-state method. Analyticgrade Sm 2 O 3 (purity 99.5%), SrO (purity 99.5%), and MnO 2 powder were used as raw materials. Using MnO 2 as the raw material, Mn 2 O 3 powder was prepared by burning in a 923 K furnace for 10 h. Sm 2 O 3 , SrO, and Mn 2 O 3 powders were mixed according to the mole ratio of Sm:Sr:Mn = 0.85:0.15:1. The mixtures were ground using an agate mortar for 1 h and then ground with ethanol for 2 h. The obtained mixtures were then dried for 1 h at 353 K in a baking oven. Afterward, the mixtures were pressed into cylindrical-shape compacts (Ø10 × 5 mm) using a powder pellet machine (769YP-15A, 200 MPa). The compacts were initially sintered in a pipe furnace (AY-BF-555-180) at 1273 K for 10 h in air and subsequently sintered at 1623 K for 10 h. The sample was allowed to cool in the furnace naturally.
The linear thermal expansion coefficient was measured using a dilatometer Linseis L76 (heating and cooling rates of 5 K/min). The XRD measurement was carried out using Bruker D8 Advance with CuKα radiation. The XRD pattern of the sample was analyzed using X'Pert HighScore Plus software. The lattice constants a, b, and c and the unit cell volume of the sample were calculated using powderX software and the least square method. The surface morphology of the sample was observed using the FEI Quanta 250 scanning electron microscopy (SEM), and the EDS energy spectrum was obtained using an Appllo XP. The TGA and DSC were tested using a LabsysTM thermal analyzer. The XPS (X-ray photoelectron spectroscopy) was performed using a Thermo Scientific K-Alpha instrument for the valence analysis of the Mn element. The BET tests were performed to determine the size and volume of the holes using an ASAP2460 device. Figure 1a is the XRD pattern of the sample at RT. Comparing the XRD pattern with the JCPDS cards for SmMnO 3 (00-025-0747), Eu 0.9 Sr 0.1 MnO 3 (No. 00-051-0252), and Eu 0.8 Sr 0.2 MnO 3 (00-051-0251), we found that the diffraction peaks were similar to those of the JCPDS cards, except for some shifts, which suggests that the as-prepared sample had similar structure to that of SmMnO 3 , Eu 0.9 Sr 0.1 MnO 3 , and Eu 0.8 Sr 0.2 MnO 3 . It can be confirmed that the ceramic Sm 0.85 Sr 0.15 MnO 3-δ crystallizes in an orthorhombic structure. As the ionic radius of Sr 2+ (ionic radius 1.18 Å) is bigger than that of Sm 3+ (ionic radius 0.958 Å), the difference in the ionic radius may cause lattice distortion. As Sr 2+ partly substitutes for Sm 3+ , the diffraction peaks also shift. raw materials. Using MnO2 as the raw material, Mn2O3 powder was prepared by bu in a 923 K furnace for 10 h.

Phase Analysis
Sm2O3, SrO, and Mn2O3 powders were mixed according to the mole ratio of S Mn = 0.85:0.15:1. The mixtures were ground using an agate mortar for 1 h and then gr with ethanol for 2 h. The obtained mixtures were then dried for 1 h at 353 K in a b oven. Afterward, the mixtures were pressed into cylindrical-shape compacts (Ø 10 × 5 using a powder pellet machine (769YP-15A,200 MPa). The compacts were in sintered in a pipe furnace (AY-BF-555-180) at 1273 K for 10 h in air and subsequ sintered at 1623 K for 10 h. The sample was allowed to cool in the furnace naturally The linear thermal expansion coefficient was measured using a dilatometer L L76 (heating and cooling rates of 5 K/min). The XRD measurement was carried out Bruker D8 Advance with CuKα radiation. The XRD pattern of the sample was ana using X'Pert HighScore Plus software. The lattice constants a, b, and c and the un volume of the sample were calculated using powderX software and the least s method. The surface morphology of the sample was observed using the FEI Quan scanning electron microscopy (SEM), and the EDS energy spectrum was obtained an Appllo XP. The TGA and DSC were tested using a LabsysTM thermal analyze XPS (X-ray photoelectron spectroscopy) was performed using a Thermo Scienti Alpha instrument for the valence analysis of the Mn element. The BET tests performed to determine the size and volume of the holes using an ASAP2460 devic Figure 1a is the XRD pattern of the sample at RT. Comparing the XRD pattern the JCPDS cards for SmMnO3 (00-025-0747), Eu0.9Sr0.1MnO3 (No.00-051-0252), Eu0.8Sr0.2MnO3 (00-051-0251), we found that the diffraction peaks were similar to th the JCPDS cards, except for some shifts, which suggests that the as-prepared sampl similar structure to that of SmMnO3, Eu0.9Sr0.1MnO3, and Eu0.8Sr0.2MnO3. It ca confirmed that the ceramic Sm0.85Sr0.15MnO3-δ crystallizes in an orthorhombic structu the ionic radius of Sr 2+ (ionic radius 1.18Å ) is bigger than that of Sm 3+ (ionic radius 0.9 the difference in the ionic radius may cause lattice distortion. As Sr 2+ partly substitut Sm 3+ , the diffraction peaks also shift.    Figure 2a shows the SEM image of the sample. We found that the ceramic sample was composed of homogenous spherical or elliptic spherical particles with some obvious agglomerations. There were pores and microcracks in the sintered body. The size of the particles was uniform, with an average grain size of about 1~2 µm. The EDS analysis of the sample revealed the primary elements of Sm, Sr, Mn, and O, and their atomic ratio (Sm:Sr:Mn:O) was about 0.85:0.15:1:3 (seeing Table 1). Combined with the XRD analysis, we identified the composition of the samples as being Sm 0.85 Sr 0.15 MnO 3 . J. Compos. Sci. 2022, 6, x FOR PEER REVIEW 3 of 11 Figure 2a shows the SEM image of the sample. We found that the ceramic sample was composed of homogenous spherical or elliptic spherical particles with some obvious agglomerations. There were pores and microcracks in the sintered body. The size of the particles was uniform, with an average grain size of about 1~2 μm. The EDS analysis of the sample revealed the primary elements of Sm, Sr, Mn, and O, and their atomic ratio (Sm:Sr:Mn:O) was about 0.85:0.15:1:3 (seeing Table 1). Combined with the XRD analysis, we identified the composition of the samples as being Sm0.85 Sr0.15 MnO3.

Thermal Expansion Property
Figure 3a-c show the relative length (dL/L) with the temperature increases of SmMnO3, SrMnO3, and Sm0.85Sr0.15MnO3-δ, respectively. SmMnO3 ( Figure 3a) and SrMnO3 ( Figure 3b) showed positive thermal expansion. Calculating according to the curve, the expansion coefficients were 5.24 × 10 −6 /K and 12.7 × 10 −6 /, respectively. When the temperature was below 360 K, the ceramic Sm0.85Sr0.15MnO3-δ showed a positive thermal expansion of 0.46875 × 10 −6 /K. As the temperature increased, the ceramic Sm0.85Sr0.15MnO3δ showed an NTE property in the range of 360 to 873 K. The average linear expansion coefficient was −10.08 × 10 −6 /K.    (Figure 3b) showed positive thermal expansion. Calculating according to the curve, the expansion coefficients were 5.24 × 10 −6 /K and 12.7 × 10 −6 /K, respectively. When the temperature was below 360 K, the ceramic Sm 0.85 Sr 0.15 MnO 3-δ showed a positive thermal expansion of 0.46875 × 10 −6 /K. As the temperature increased, the ceramic Sm 0.85 Sr 0.15 MnO 3-δ showed an NTE property in the range of 360 to 873 K. The average linear expansion coefficient was −10.08 × 10 −6 /K. Figure 4 shows the high-temperature XRD patterns of ceramic Sm 0.85 Sr 0.15 MnO 3-δ from RT to 873 K. As the temperature increased, the diffraction peaks of Sm 0.85 Sr 0.15 MnO 3-δ moved slightly to small angles, except three diffraction peaks (31.54 • , 33.79 • , and 52.65 • ) that moved to a large angle. Figure 5 shows the variation in the Sm 0.85 Sr 0.15 MnO 3-δ lattice parameters and volume with temperature increases, which was calculated using the powderX software. In a, c in Figure 5, the increase occurred gradually, while in b in Figure 5, it decreased as the temperature increased gradually. We believe that the thermal expansion of Sm 0.85 Sr 0.15 MnO 3-δ was due to anisotropy. We can see that from RT to 360 K, Sm 0.85 Sr 0.15 MnO 3-δ showed a positive expansion property. As the temperature increased to 360~873 K, Sm 0.85 Sr 0.15 MnO 3-δ showed an NTE property with the average linear expansion coefficient of −3.33 × 10 −6 /K. However, the original calculation of the negative thermal expansion coefficient of Sm 0.85 Sr 0.15 MnO 3-δ in this temperature range was −10.08 × 10 −6 /K, according to Figure 3. As can be seen from Figure 2 above, there were pores and microcracks in the crystal. Therefore, we believe that when the temperature rises, the crystal squeezes the open space, namely, these pores and microcracks, which is another reason for the negative thermal expansion.  Figure 4 shows the high-temperature XRD patterns of ceramic Sm0.85Sr0.15MnO3-δ from RT to 873K. As the temperature increased, the diffraction peaks of Sm0.85Sr0.15MnO3-δ moved slightly to small angles, except three diffraction peaks (31.54°, 33.79°, and 52.65°) that moved to a large angle. 20 Figure 4 shows the high-temperature XRD patterns of ceramic Sm0.85Sr0.15MnO3-δ from RT to 873K. As the temperature increased, the diffraction peaks of Sm0.85Sr0.15MnO3-δ moved slightly to small angles, except three diffraction peaks (31.54°, 33.79°, and 52.65°) that moved to a large angle. 20    according to Figure 3. As can be seen from Figure 2 above, there were pores a microcracks in the crystal. Therefore, we believe that when the temperature rises, crystal squeezes the open space, namely, these pores and microcracks, which is anot reason for the negative thermal expansion.  Table 2 shows the pore size, pore volume, and BET surface area of the sample. T specific surface of the material itself is large, and the general level of adsorption is go When the pore structure of carbon materials is more complex, it is easy to have a flexi hole, and the pore size becomes larger after gas adsorption. With the doping of S oxygen defects are caused, and the gas is adsorbed in the pores. With the further dop of Sr 2+ , the adsorption oxygen saturation does not change. With the increase temperature, the gas is sintered out, the b-axis shrinks at the same time, and the pore s becomes smaller, resulting in the negative expansion property.

Sample
Pore Size (nm) Pore Volume (cm 3 /g) BET Surface Area (m 2 Sm0.85Sr0.15MnO3-δ 15.7842 0.002563 0.6351 Figure 6a is the XPS spectrum of Sm0.85Sr0.15MnO3-δ; the characteristic peaks of Sm, Mn, and O are shown in the figure, respectively. The surface of the sample was free fr any pollutants, and element C was used for the calibration of the XPS atlas.   Table 2 shows the pore size, pore volume, and BET surface area of the sample. The specific surface of the material itself is large, and the general level of adsorption is good. When the pore structure of carbon materials is more complex, it is easy to have a flexible hole, and the pore size becomes larger after gas adsorption. With the doping of Sr 2+ , oxygen defects are caused, and the gas is adsorbed in the pores. With the further doping of Sr 2+ , the adsorption oxygen saturation does not change. With the increase in temperature, the gas is sintered out, the b-axis shrinks at the same time, and the pore size becomes smaller, resulting in the negative expansion property. Table 2. Pore size, pore volume, and BET surface area of the sample.

Sample
Pore Size (nm) Pore Volume (cm 3  show the XPS spectra of Mn. In the XPS spectrum, the sample had a bimodal structure, which indicates that the Mn elements on the sample surface existed in two forms: Mn 3+ and Mn 4+ , which led to the oxygen vacancy. The presence of the oxygen vacancy facilitated the movement of electrons between Mn 4+ and Mn 3+ . The oxygen vacancy also led to the shortening of the bond length of the Mn-O bond, which led to lattice distortion and generated internal stress; this reduced the bond angle of Mn-O-Mn and increased the double-exchange effect.

Discussion
SmMnO3 is a typical manganite perovskite structure. The structure of SmMnO3 is shown in Figure 7. As for the MnO6 octahedron in SmMnO3, the distortion was caused by a change in the length of the Mn-O bond.

Discussion
SmMnO 3 is a typical manganite perovskite structure. The structure of SmMnO 3 is shown in Figure 7. As for the MnO 6 octahedron in SmMnO 3 , the distortion was caused by a change in the length of the Mn-O bond.

Discussion
SmMnO3 is a typical manganite perovskite structure. The structure of SmMnO3 is shown in Figure 7. As for the MnO6 octahedron in SmMnO3, the distortion was caused by a change in the length of the Mn-O bond.  There are three kinds of common modes for this change [22][23][24][25][26] as follows. (1) The surface tension contract model Q 1 , as shown in Figure 8a. Six oxygen atoms of the unit cell move close to or far away from the manganese atom at the same time, making the Mn-O bond length decrease or increase significantly. This model can increase the energy of the system, which is not conducive to the system energy being able to decrease and makes the system extremely unstable in turn. (2) The plane distortion model Q 2 , as shown in Figure 8b. In a unit, two oxygen atoms in the horizontal plane leave a manganese atom, while the other two oxygen atoms become close to the manganese atom. The location of the two oxygen atoms in the vertical plane remains unchanged. (3) The expansion mode, or inspiratory mode Q 3 , which is shown in Figure 8c. In a MnO 6 octahedron, the two oxygen atoms in the vertical plane leave manganese atoms, while the four oxygen atoms in the horizontal plane become close to the manganese atom simultaneously. For a MnO 6 octahedron, the Q 1 and Q 2 models normally exist. Since the Q 1 model is unstable, the distortion of the MnO 6 octahedron is mainly the Q 2 model, also called the plane distortion model. There are three kinds of common modes for this change [22][23][24][25][26] as follows. (1) The surface tension contract model Q1, as shown in Figure 8a. Six oxygen atoms of the unit cell move close to or far away from the manganese atom at the same time, making the Mn-O bond length decrease or increase significantly. This model can increase the energy of the system, which is not conducive to the system energy being able to decrease and makes the system extremely unstable in turn. (2) The plane distortion model Q2, as shown in Figure  8b. In a unit, two oxygen atoms in the horizontal plane leave a manganese atom, while the other two oxygen atoms become close to the manganese atom. The location of the two oxygen atoms in the vertical plane remains unchanged. (3) The expansion mode, or inspiratory mode Q3, which is shown in Figure 8c. In a MnO6 octahedron, the two oxygen atoms in the vertical plane leave manganese atoms, while the four oxygen atoms in the horizontal plane become close to the manganese atom simultaneously. For a MnO6 octahedron, the Q1 and Q2 models normally exist. Since the Q1 model is unstable, the distortion of the MnO6 octahedron is mainly the Q2 model, also called the plane distortion model. We used MnO2, Sm2O3, and SrO as the raw materials to prepare Sm0.85Sr0.15MnO3-δ. In the reaction process, there was a reciprocal transformation between Mn 3+ and Mn 4+ .
When Sr 2+ substitutes the Sm 3+ in SmMnO3, Sr 2+ will occupy the position of Sm 3+ . To maintain the valence balance, electron transfer occurs in the Mn 3+ converting into Mn 4+ in Sm0.85Sr0.15MnO3-δ. Additionally, the p electron of O 2-will migrate to the orbit of the nearby Mn 4+ , and the d electron of Mn 3+ will migrate to the orbit of the nearby Mn 3+ . Thus, this mechanism results in the electronic conduction and position exchanges of Mn 4+ and Mn 3+ ions. The system energy remains unchanged throughout. This process is known as the double exchange [27]. The structure of Mn 3+ -O 2− -Mn 4+ forms in the process. However, according to the theory of Zener [28], the route of electron transfer between two Mn 3+ changes between Mn 3+ and Mn 4+ . In order to keep the electron transfer between two Mn 3+ , the magnetic moment between Mn 3+ and Mn 4+ ions should be parallel to each other. In this situation, it is favorable for there to be more electron transfer between Mn 3+ and Mn 4+ ions.
According to the analysis of the variable-temperature XRD data, we considered that the thermal property of Sm0.85Sr0.15MnO3-δ might be related to the interaction of the lattice vibration and electron transfer between Mn 3+ and Mn 4+ . As the temperature rose, the lattice vibrated dramatically and Mn 4+ converted into Mn 3+ . Moreover, the electron transfer rate increased between the Mn 3+ and Mn 4+ ions. The number of Mn 3+ ions that can cause the Jahn-Teller [29] effect increased. The oxygen ions in the Mn 3+ O6 octahedron became slant, or even produced oxygen defects, making the unit cell volumes decrease. From RT to 360 K, the unit cell volume increased. The reason is that the contribution of the lattice vibration to the thermal expansion exceeded that of the MnO6 octahedral distortion and oxygen defects. As the temperature increased, Sm0.85Sr0.15MnO3-δ showed a low positive thermal expansion property, and above 360 K, the unit cell volume decreased. With more Mn 4+ We used MnO 2 , Sm 2 O 3 , and SrO as the raw materials to prepare Sm 0.85 Sr 0.15 MnO 3-δ . In the reaction process, there was a reciprocal transformation between Mn 3+ and Mn 4+ .
When Sr 2+ substitutes the Sm 3+ in SmMnO 3 , Sr 2+ will occupy the position of Sm 3+ . To maintain the valence balance, electron transfer occurs in the Mn 3+ converting into Mn 4+ in Sm 0.85 Sr 0.15 MnO 3-δ . Additionally, the p electron of O 2− will migrate to the orbit of the nearby Mn 4+ , and the d electron of Mn 3+ will migrate to the orbit of the nearby Mn 3+ . Thus, this mechanism results in the electronic conduction and position exchanges of Mn 4+ and Mn 3+ ions. The system energy remains unchanged throughout. This process is known as the double exchange [27]. The structure of Mn 3+ -O 2− -Mn 4+ forms in the process. However, according to the theory of Zener [28], the route of electron transfer between two Mn 3+ changes between Mn 3+ and Mn 4+ . In order to keep the electron transfer between two Mn 3+ , the magnetic moment between Mn 3+ and Mn 4+ ions should be parallel to each other. In this situation, it is favorable for there to be more electron transfer between Mn 3+ and Mn 4+ ions.
According to the analysis of the variable-temperature XRD data, we considered that the thermal property of Sm 0.85 Sr 0.15 MnO 3-δ might be related to the interaction of the lattice vibration and electron transfer between Mn 3+ and Mn 4+ . As the temperature rose, the lattice vibrated dramatically and Mn 4+ converted into Mn 3+ . Moreover, the electron transfer rate increased between the Mn 3+ and Mn 4+ ions. The number of Mn 3+ ions that can cause the Jahn-Teller [29] effect increased. The oxygen ions in the Mn 3+ O 6 octahedron became slant, or even produced oxygen defects, making the unit cell volumes decrease. From RT to 360 K, the unit cell volume increased. The reason is that the contribution of the lattice vibration to the thermal expansion exceeded that of the MnO 6 octahedral distortion and oxygen defects. As the temperature increased, Sm 0.85 Sr 0.15 MnO 3-δ showed a low positive thermal expansion property, and above 360 K, the unit cell volume decreased. With more Mn 4+ ions converting into Mn 3+ , the Mn 3+ O 6 octahedral distortion was enhanced and oxygen defects occurred. These contributed more to the thermal expansion than the lattice vibration. Therefore, Sm 0.85 Sr 0.15 MnO 3-δ shows a negative thermal expansion property between 360 K and 873 K.
The DSC and TGA results of ceramic Sm 0.85 Sr 0.15 MnO 3-δ also support the above statements. Figure 9a presents the DSC curve of ceramic Sm 0.85 Sr 0.15 MnO 3-δ . In the curve, Sm 0.85 Sr 0.15 MnO 3-δ has an endothermic peak at about 360 K. This shows that more Mn 4+ ions were converted to Mn 3+ with the increase in temperature. Thus, Mn 3+ O 6 octahedral distortion was enhanced and oxygen defects occurred. The unit cell volume began to decrease, which is consistent with the results calculated by the high-temperature XRD (seeing b in Figure 5). As electron transfer occurred between the Mn 3+ and Mn 4+ ions, the amount of Mn 4+ decreased, and oxygen ions in the Mn 3+ O 6 octahedron became slant or even produced oxygen defects. The TGA results of Sm 0.85 Sr 0.15 MnO 3-δ confirm the existence of oxygen defects. In Figure 8b, the weight of the Sm 0.85 Sr 0.15 MnO 3-δ sample decreased when the temperature increased from RT to 873 K. In addition, the variabletemperature XRD (seeing Figure 5) showed that there was no phase transition with the increase in temperature. As electron transfer occurred between the Mn 3+ and Mn 4+ ions, Mn 4+ O 6 converted into Mn 3+ O 6 and oxygen defects appeared. Therefore, we consider that the loss of the weight can be ascribed to the oxygen defects.
J. Compos. Sci. 2022, 6, x FOR PEER REVIEW 9 ions converting into Mn 3+ , the Mn 3+ O6 octahedral distortion was enhanced and oxy defects occurred. These contributed more to the thermal expansion than the la vibration. Therefore, Sm0.85Sr0.15MnO3-δ shows a negative thermal expansion prop between 360K and 873K. The DSC and TGA results of ceramic Sm0.85Sr0.15MnO3-δ also support the ab statements. Figure 9a presents the DSC curve of ceramic Sm0.85Sr0.15MnO3-δ. In the cu Sm0.85Sr0.15MnO3-δ has an endothermic peak at about 360 K. This shows that more Mn 4+ were converted to Mn 3+ with the increase in temperature. Thus, Mn 3+ O6 octahe distortion was enhanced and oxygen defects occurred. The unit cell volume bega decrease, which is consistent with the results calculated by the high-temperature X (seeing b in Figure 5). As electron transfer occurred between the Mn 3+ and Mn 4+ ions amount of Mn 4+ decreased, and oxygen ions in the Mn 3+ O6 octahedron became slan even produced oxygen defects. The TGA results of Sm0.85Sr0.15MnO3-δ confirm the existe of oxygen defects. In Figure 8b, the weight of the Sm0.85Sr0.15MnO3-δ sample decrea when the temperature increased from RT to 873K. In addition, the variable-tempera XRD (seeing Figure 5) showed that there was no phase transition with the increas temperature. As electron transfer occurred between the Mn 3+ and Mn 4+ ions, Mn converted into Mn 3+ O6 and oxygen defects appeared. Therefore, we consider that the of the weight can be ascribed to the oxygen defects. Moreover, the non-stoichiometric ratio of Sm0.85Sr0.15MnO3-δ caused the mole r mismatch of the Sm, Mn, and O atoms. Some lattice vacancies and interstitials existe the crystal lattice, making the lattice distortion continuous. In the structure analysis crystal distortion was found to have a direct impact on the bond length and angle of MnO6 octahedron. As for ABO3, when we conducted the substitution in the A posi with a different ionic radius, especially in the non-stoichiometric ratio mangan perovskite, the size mismatch effects of the A position ion together with lattice space interstitial caused a difference in the crystal structure. These eventually led to a g change in the lattice parameters and unit cell size [30][31][32].

Conclusions
(1) A novel negative thermal expansion material composed of Sm0.85Sr0.15MnO3-δ synthesized using the solid-state method with an NTE coefficient of −10.08 × 10 from 360 to 873 K. (2) The particles were homogenous spherical or elliptic-spherical particles wit uniform particle size of about 1~2 μm.  Moreover, the non-stoichiometric ratio of Sm 0.85 Sr 0.15 MnO 3-δ caused the mole ratio mismatch of the Sm, Mn, and O atoms. Some lattice vacancies and interstitials existed in the crystal lattice, making the lattice distortion continuous. In the structure analysis, the crystal distortion was found to have a direct impact on the bond length and angle of the MnO 6 octahedron. As for ABO 3 , when we conducted the substitution in the A position with a different ionic radius, especially in the non-stoichiometric ratio manganese perovskite, the size mismatch effects of the A position ion together with lattice space and interstitial caused a difference in the crystal structure. These eventually led to a great change in the lattice parameters and unit cell size [30][31][32].

Conclusions
(1) A novel negative thermal expansion material composed of Sm 0.85 Sr 0.15 MnO 3-δ was synthesized using the solid-state method with an NTE coefficient of −10.08 × 10 −6 /K from 360 to 873 K. (2) The particles were homogenous spherical or elliptic-spherical particles with a uniform particle size of about 1~2 µm.