Reaction Behavior of Sm and Valence State Evolution of Sm³⁺ During the Reduction of SmF₃

Abstract

SmF₃ cannot be reduced to metallic samarium by aluminum due to variable valence states of Sm. This study investigates the reduction products of SmF₃ via an aluminothermic reduction. The effect of molar ratios of Al/SmF₃ on the morphology, elemental distribution, crystal structure, and chemical valence of the samples were investigated by scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDS), X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS). The thermodynamic results show that it is feasible for SmF₃ reduction by Al to form SmF₂ in 933~1356 K. SmF_2.413, AlF₃, and Sm(AlF)₅ are obtained under the condition of the molar ratio of Al to SmF₃ at 1:3, 2:3, 3:3, 4:3, and 5:3. The samarium of the reduction products exhibits mixed valence states of Sm³⁺ and Sm²⁺, with the ratio δ of F to Sm determined by a(δ) = −0.1794δ + 5.819 (0 ≤ δ ≤ 0.4615). The presence of adsorbed oxygen in the products facilitates the oxidation process from Sm²⁺ to Sm³⁺. These findings may provide a theoretical basis on the development of valence states for other rare earth elements in aluminothermic reduction.

1. Introduction

Samarium (Sm) and its compounds are widely studied for their unique physicochemical properties and applications in optics, catalysis, and materials science. Among these compounds, samarium (II) fluoride (SmF₂) is notable for its high melting point, thermal stability, and optical performance, making it valuable for high-performance optics and display technologies [1,2,3]. After Ba²⁺ was partially replaced by Sm²⁺ in BaAl₂Si₃O₄N₄, it was found that the photoluminescence spectrum at a low temperature consists of some sharp spectral lines, and BaAl₂Si₃O₄N₄:Sm²⁺ showed excellent thermal stability [4]. In terms of catalysis, Sm: WO₃ nanoparticles were prepared by chemical co-precipitation. The results showed that the photocatalytic performance and photogenerated carrier separation efficiency were improved with increasing the Sm doping concentration from 0 to 7 wt.%, and the superior antibacterial activity against staphylococcus aureus (18 mm) was obtained with the Sm: WO₃ (7 wt.%) sample [5].

The diverse oxidation states of Sm are critical to its functional versatility in materials science and chemistry. While the +3 oxidation state is the most thermodynamically stable and widely observed, the +2 state has gained significant attention due to its distinctive electronic configuration (4f) and relative stability under specific conditions. Zuo et al. [6] systematically investigated the electrochemical behavior of the Sm(III)/Sm(II) redox couple in 2LiF-BeF₂(FLiBe) molten salt in the range of 773–973 K. Their study confirmed the coexistence of Sm(III) and Sm(II) in the molten salt system, with the redox reaction on an inert tungsten electrode exhibiting quasi-reversible kinetics. Furthermore, the growth of single crystals with the composition La_1−y(Sm³⁺_1−xSm²⁺_x)_yF_3−xy (y = 0.04) via the vertical Bridgman method provided direct evidence of the reduction of Sm³⁺ to Sm²⁺ during crystallization. The observed phenomenon was attributed to interactions with carbonaceous species in the melt [7]. In the co-deposition of SmCo in an aqueous solution [8], as well as in the electrodeposition process using 1-Butyl-1-Methylpyrrolidinium dicyanamide ionic liquid [9], NaCl-KCl melts [10], LiCl-KCl Eutectic Molten Salt [11], and LiCl-KCl-SmCl₃ melt [12], only Sm²⁺ was detected. These findings underscore the significance of elucidating the valence state evolution of samarium in diverse chemical environments, as it profoundly influences the structural, electronic, and functional properties of the resulting materials.

However, the reduction of samarium compounds remains a challenging process due to the complex reaction mechanisms and the stability of Sm²⁺. Traditional reduction methods using alkali metals (Li, Na, K) or alkaline earth metals (Ca, Mg) often result in the formation of stable bivalent samarium halides (Sm²⁺) rather than Sm⁰, owing to the tendency of samarium to achieve a stable electron configuration (3d¹⁰4f⁶) close to the semi-full state of the 4f shell [13]. Recent studies have demonstrated that hydrogen reduction can generate non-integer valence states of samarium, such as SmF_2.29 and SmF_x, with the valence state of samarium ranging from +2.0 to +2.5 [14,15,16]. These findings indicate that samarium can exhibit non-integer valence states between +2 and +3, further complicating the reduction process.

Previous studies on the reduction of samarium halides using Li, Na, K, Ca, and Mg have shown that none of these methods can produce Sm. In the preparation of rare earth aluminum alloys, aluminum has been widely used in the production of rare earth aluminum alloys, such as La-Al and Y-Al alloys [17]. While its application in the reduction of SmF₃ to Sm has not been systematically investigated. In this study, the valence state evolution of samarium during the aluminothermic reduction of SmF₃ was investigated. The effect of the Al/SmF₃ molar ratios on the products was examined. The phase structure, chemical valence, and morphology of the reduction products were analyzed. The variation in the valence state of samarium in the reduction products was elucidated. These findings will provide an important reference for understanding the mechanism of samarium valence changes during the aluminothermic reduction process.

2. Materials and Methods

2.1. Raw Materials

The chemical reagents used in this work included SmF₃ powder (purity > 99.99%); it was supplied by Ganzhou Shilei Rare Earth Materials Co., Ltd., Ganzhou, China. Al strip (purity > 99.99%) was supplied by Dongguan Changan Baolu Metal Materials Co., Ltd., Dongguan, China. The argon gas (purity > 99.99%) was supplied by Ganzhou Jianli Gas Co., Ltd., Ganzhou, China. The SmF₃ powder was pretreated before the experiment under vacuum at 423 K for 5 h to remove any moisture. The molar ratios of Al and SmF₃ were set as 1:3, 2:3, 3:3, 4:3, and 5:3.

2.2. Sample Preparation

Before the experiment, Al strips were cut into pieces, mixed with SmF₃ in proportion. The reactants were pressed into chunks using a tablet press (Model FW-4, Tianjin TianGuang Optical Instrument Co., Ltd., Tianjin, China) and loaded into a tungsten crucible. The experiments were conducted using a high-vacuum induction melting furnace (Model VGL-400, Shenyang Oute Vacuum Technology Co., Ltd., Shenyang, China). During the experiment, the reactants were gradually heated to 500~600 °C and maintained at this range for three to five minutes. Subsequently, the temperature rapidly increased until complete melting was achieved, after which heating was discontinued. After undergoing furnace cooling, the reduction products were retrieved and collected.

2.3. Material Characterization

The phase composition of the reduction products was characterized by powder X-ray diffraction (XRD, Empyrean, Netherlands, Cu Kα, 10° ≤ 2θ ≤ 90°), and the XRD data were analyzed using the HighScore software (version 4.9, PANalytical B.V., Almelo, Netherlands). Scanning electron microscopy (SEM, MLA650F, FEI Company, Hillsboro, Oregon, USA) was used to analyze the microstructure of the reduction products, and energy-dispersive spectrometry (EDS, JSM-6480A; JEOL Ltd., Tokyo, Japan) was utilized to analyze the micro-chemical zone of the reduction products. X-ray photoelectron spectroscopy (XPS, 250Xi, Thermo Fisher Scientific, Waltham, MA, USA) was performed to explore the composition detection and valence analysis of the reduction products.

3. Results and Discussion

3.1. Thermodynamic Calculation

The potential reduction reactions of SmF₃ using Al as a reducing agent at various temperatures are presented, along with the corresponding Gibbs free energy values for each reaction in relation to the temperature. The Gibbs free energy change (${\Delta }_{\mathrm{r}}{G}_{\mathrm{m}}^{⊝}\left(T\right)$) for these reactions was calculated using the Gibbs–Helmholtz equation, as shown in Equation (1).

$${\Delta }_{\mathrm{r}}{G}_{\mathrm{m}}^{⊝}\left(T\right)={\Delta }_{\mathrm{r}}{H}_{\mathrm{m}}^{⊝}\left(T\right)={T\Delta }_{\mathrm{r}}{S}_{\mathrm{m}}^{⊝}\left(T\right)$$

(1)

where ${\Delta }_{\mathrm{r}}{H}_{\mathrm{m}}^{⊝}\left(T\right)$is the standard reaction enthalpy, ${\Delta }_{\mathrm{r}}{S}_{\mathrm{m}}^{⊝}\left(T\right)$ is the standard reaction entropy, and T is the absolute temperature in Kelvin. The thermodynamic calculations for the reduction reaction of SmF₃ by Al were performed using data from the handbook [18].

Al(l) + 3SmF₃(s) = AlF₃(l) + 3SmF₂(s),

(2)

$${\Delta G}_1^{⊝}=0.0476T-64.56(\mathrm{kJ}/\mathrm{mol})$$

Al(l) + SmF₃(s) = AlF₃(l) + Sm(s),

(3)

$${\Delta G}_2^{⊝}=0.0271T+143.62(\mathrm{kJ}/\mathrm{mol})$$

Figure 1 illustrates the dependence of Gibbs free energy (${\Delta \mathrm{G}}^{⊝})$ on the temperature (T) for the two chemical reactions (2) and (3). As depicted in Figure 1, it can be observed that the Gibbs free energy of the reaction is positive when SmF₃ reduced by Al to yield metallic Sm, indicating that the reaction is thermodynamically non-spontaneous. In contrast, when the reduction product is SmF₂, the Gibbs free energy of the reaction becomes negative below 1356 K, suggesting that the reaction is thermodynamically feasible. The lower limit of 933 K was chosen based on the melting point of aluminum, ensuring its molten state for the reduction reaction. Therefore, the temperature range of 933~1356 K was selected.

3.2. Phase Analysis

The XRD results of the reduction products corresponding to SmF₃ reduced by Al are shown in Figure 2. It shows a high degree of conformity between the diffraction peaks of the reduction products and those of the Sm₁₃F_32−x standard card (JCPDS No. 00-034-1067), suggesting that Sm₁₃F_32−x (0 ≤ x ≤ 1) is the main reduction product. Furthermore, the XRD pattern of the Sm₁₃F_32−x is entirely different from that of the SmF₃ standard card (JCPDS No. 00-032-0981), confirming that SmF₃ participated in the reaction and was transformed into new phases. Simultaneously, upon comparison with the diffraction peaks of AlF₃ standard card (JCPDS No. 01-083-0719) and Sm(AlF₅) standard card (JCPDS No. 01-081-1722), it is evident that the product’s diffraction peaks closely correspond to the standard peaks. The results indicate the presence of AlF₃ and Sm(AlF₅) in the reduction products. Moreover, variations in the molar ratio of Al/SmF₃ have a minimal impact on the type of product phase. However, according to chemical Equation (2), the reaction products only consist of SmF₂ and AlF₃. Therefore, the formation of Sm(AlF₅) must be attributed to other chemical reactions.

The preparation of Sm(AlF₅) involved the use of SmF₃, Sm, and AlF₃ in a molar ratio of 2:1:3, which were heated to 1023 K and held for 7~10 days under argon protection. The reaction process involved the reduction of Sm from SmF₃, followed by its reaction with AlF₃ to produce SmF₂ [19]. In this reaction process, SmF₂ was obtained through the reduction of SmF₃ with Sm, followed by its reaction with AlF₃. The high temperature (1300 K) accelerates the formation of Sm(AlF₅), which can be expressed by the following equation:

SmF₂(s) + AlF₃(s) = Sm(AlF₅)(s),

(4)

The XRD results indicate that the valence state of Sm in the compound is non-integral. As reported by Greis [15], the partial substitution of Sm²⁺ with Sm³⁺ occurs in Sm₁₃F_32−x, and the chemical formula of samarium fluoride can be expressed as ${\mathrm{Sm}}_{7+\mathrm{x}}^{\mathrm{II}}{\mathrm{Sm}}_{6-\mathrm{x}}^{\mathrm{III}}{\mathrm{F}}_{32-\mathrm{x}}$ or (SmII,SmIII)F_2+δ, indicating that the valence states of samarium ions in the sample are not integers. The XRD results show that the sample does not consist of mechanical mixture of SmF₃ and SmF₂ but rather exhibits an intermediate valence state between 2+ and 3+.

To further determine the value of x in the reduction products,

Table 1. Diffraction data of different Al/SmF₃ molar ratios and standard data of Sm₁₃F_33−x.

Indices of Lattice Planes	d (nm)	d (nm)					Standard Deviation σ× (10−3)	Ratio of F to Sm (δ)	lattice Constant a/nm
		Molar Ratio of Al/SmF3
	Standard Data of Sm13F32−x	1:3	2:3	3:3	4:3	5:3
(131)	0.3340	0.3369	0.3377	0.3364	0.3389	0.3356	1.260	0.41248	0.57450
(312)	0.2898	0.2908	0.2913	0.2916	0.2933	0.2899	1.250	0.41311	0.57448
(134)	0.2055	0.2058	0.2064	0.2070	0.2070	0.2068	0.510	0.41261	0.57449
(523)	0.1747	0.1755	0.1756	0.1755	0.1756	0.1754	0.084	0.41341	0.57448
(262)	0.1671	0.1665	0.1661	0.1683	0.1656	0.1665	1.020	0.41297	0.5748

lists the diffraction data of products with different molar ratios of Al and SmF₃ and Sm₁₃F_32−x standard card. The d values corresponding to each index of lattice planes exhibit a high level of consistency with the standard d value, indicating a significant degree of conformity. The ratio δ of F to Sm has been investigated in the literature [20,21,22], which can be determined through a functional equation; the lattice parameter a(δ) is expressed as:

a(δ) = −0.01794δ + 0.5819, (units in nm)

(5)

where a represents the lattice parameter in the Rhombohedral system with space group R-3. based on the relationship between the chemical formulas SmF_2+δ and Sm₁₃F_32−x. The fluorine-to-samarium ratio can be equated as:

δ = (6 − x)/13

(6)

This equation establishes a quantitative relationship between δ and x, demonstrating how the stoichiometric deviation (δ) varies with the fluorine deficiency (x) in the reduction product. From Equation (6), the range of δ is determined to be 0 ≤ δ ≤ 0.4615.

Based on the data in Table 1, the low standard deviations (σ) indicate high precision in the experimental results. The parameter δ remains nearly constant across different Al/SmF₃ molar ratios, suggesting that the molar ratios of aluminum to SmF₃ does not significantly affect the valence state evolution of samarium in the reduction products. This consistency demonstrates that the valence state of samarium in the reduction products is not influenced by variations in the Al/SmF₃ molar ratios. According to the XRD results, the lattice constant of the reaction product (molar ratio of Al/SmF₃ 1:3) was calculated as 0.5745 nm, and δ was 0.413 by substituting into the function Equation (5). Therefore, the value of x in Sm₁₃F_32−x obtained through Al reduction of SmF₃ was 0.631, indicating that product should be formulated as SmF_2.413. The reduction products consist of a mixture of phases, as confirmed by XRD analysis. The separation of individual components was not attempted in this study, but future research could explore methods to isolate these phases, which may provide further insights into their properties and applications.

3.3. Microstructure of Reduction Products

Figure 3 displays the SEM and elemental scanning images of the reduction products corresponding to an Al to SmF₃ molar ratio of 1:3. As depicted in Figure 3a, the sample’s fracture surface is uneven, with clearly visible fracture cracks. As depicted in Figure 3b–f, the primary constituents of the sample comprise Sm, F, Al, and O. The presence of O can be attributed to its adsorption from the ambient air during sample preparation. The overall distribution of all elements appears to be uniform; however, the localized enrichment of the Al and O elements is observed in Figure 3d and Figure 3f, respectively. The results of the atomic number ratio of the Sm, F, Al, and O elements are nearly 30:7:1:1. Point scanning was conducted by selecting points 1 and 2 in Figure 3a. The results depicted in Figure 3g,h reveal that point 1 predominantly comprised Sm, F, and a minor quantity of O. The mass ratio of Sm to F was determined to be 5.6:1. Hence, the presence of Sm₁₃F_32−x product is plausible at this location. Point 2 is a region with high levels of O and Al, with an O content of 25.50% and an Al of 36.45%, which may exist in the form of Al₂O₃. During the reaction process, Al₂O₃ with a high melting point is enveloped in melt. The mass ratio of Sm and F is almost 3:1, indicating that Sm(AlF₅) may coexist with Al₂O₃.

3.4. Surface Composition and Valence State of Reduction Products

Figure 4 displays the XPS full spectrum scan of the corresponding reduction products obtained at different molar ratios of Al to SmF₃ of 1:3, 2:3, 3:3, 4:3, and 5:3. As depicted in Figure 4a–d, the reduction products exhibit characteristic peaks of Sm 3d, Sm 4d, F1s, O 1s, and Al 2p signals, indicating that the products mainly consist of the Sm, F, O, and Al elements. By comparing the characteristic peak intensities of each element, it can be observed that the characteristic peak intensity of Sm gradually decreases in Figure 4a, while a continuous increase in Al content can be observed in Figure 4b. In Figure 4c, the fluorine (F) content generally increases with the Al/SmF₃ molar ratios. Meanwhile, in Figure 4d, the oxygen (O) content exhibits an overall increasing trend; however, the characteristic peak intensity of O shows little variation among the samples and remains essentially constant. The relative contents of the Sm, F, O, and Al elements in the sample were determined by calculating the characteristic peak area ratio, as presented in

Table 2. Relative content of elements in reduction products with different Al/SmF₃ molar ratios.

Element	Relative Content (%) Molar Ratio of Al/SmF3
	1:3	2:3	3:3	4:3	5:3
Sm	27.20	26.09	24.85	22.80	20.43
Al	5.63	8.22	10.26	11.98	12.31
F	32.97	27.78	29.23	28.85	28.77
O	34.20	37.91	35.66	36.37	38.49

. The proportion of Al content increases from 5.63% to 12.31%, while the relative concentration of Sm gradually decreases with the rise in the Al content, ranging from 27.2% to 20.43%. The content of O remains essentially unaltered at approximately 35%, with any minor fluctuations primarily attributable to the adsorption of atmospheric oxygen by reduction products. The F content exhibits a slight decrease as the Al content increases, but no significant changes are observed.

The XPS high-resolution spectra of the Sm, Al, O, and F elements in the reduction products are presented in Figure 5, Figure 6, Figure 7, Figure 8 and Figure 9. In Figure 5a, the Sm 3d emission exhibits two characteristic peaks of Sm 3d_3/2 and Sm 3d_5/2 due to atomic spin–orbit splitting, with binding energies of 1110.08 eV and 1082.78 eV, respectively, showing a peak distance difference of 27.3 eV that is within the standard range (27.2 eV). After peak splitting, four characteristic peaks are observed. A pair of characteristic peaks at the binding energies of 1082.78 eV and 1073.76 eV match with the states of 3d¹⁰4f⁵ and 3d¹⁰4f⁶, respectively. The binding energies of 1082.78 eV and 1110.08 eV correspond to Sm³⁺ [23], while those at 1073.76 eV and 1100 eV are attributed to Sm²⁺ [24]. Compared to the XPS spectrum of Sm³⁺ in SmF₃, the presence of an additional peak at 1073.76 eV indicates the existence of Sm²⁺ in the reduction products. This confirms that the valence state of samarium in the products includes both +2 and +3 states. The relative contents of Sm³⁺ and Sm²⁺ are 83.53% and 16.47%, respectively, with the predominant form of Sm³⁺.

In Figure 5b, the high-resolution spectrum of O exhibits significant asymmetry. The subsequent peak fitting of the O 1s spectrum reveals two distinct peaks at the binding energies of 527.92 eV and 530.75 eV, respectively. Extensive studies have been conducted on the forms of oxygen [25,26]. Lattice oxygen (O_L) primarily exists as O²⁻ through chemical adsorption, with binding energies typically lower 530 eV. In contrast, adsorbed oxygen (O_A) undergoes physical adsorption via surface adsorption reactions, existing mainly in the intermediate states of${\mathrm{O}}_2^{-}$, ${\mathrm{O}}_2^{2-}$ and O⁻, with binding energies generally falling between 530 and 532 eV. Combined with the previous analysis, it can be inferred that the binding energy of 527.92 eV corresponds to O_L, while the binding energy of 530.75 eV is attributed to O_A. Compared to the XPS spectrum of O in Sm₂O₃, it only exhibits lattice oxygen. The presence of O_A in the reduction products shows surface interactions or partial oxidation during the reduction process. O_A in the reduction products originates from oxygen introduced during sample preparation. In the process of grinding, the high surface energy of fine particle samples renders them susceptible to oxygen adsorption from air, resulting in physical adsorption. The relative contents of O_A and O_L are 95.20% and 4.80%, respectively, as shown in

Table 3. Relative content of Sm and O in products at different Al/SmF₃ ratios.

Molar Ratio of Al/SmF3	Sm (%)		O (%)
	Sm3+	Sm2+	OL	OA
1:3	83.53	16.47	4.80	95.20
2:3	88.92	11.08	7.33	92.67
3:3	81.15	18.85	7.73	92.27
4:3	84.16	15.84	8.28	91.72
5:3	86.65	13.35	9.03	90.97

. Consequently, the oxygen present in the reduction products predominantly exists in the form of O_A. Therefore, the significant presence of Sm³⁺ in the product is attributed to O_A that facilitates the oxidation of Sm²⁺. Faldt’s results confirmed that the presence of oxygen significantly alters the binding energy alignment and peak position of Sm³⁺, leading to the transformation from Sm²⁺ to Sm³⁺ [27].

Figure 5c displays the XPS high-resolution spectrum of Al, which exhibits a well-balanced characteristic peak with a distinct feature at 74.53 eV attributed to Al³⁺. This indicates that Al exists in the form of Al³⁺ [28]. In Figure 5d, combined with the characteristic peak of F 1s at 684.49 eV, the two distinctive peaks emerge after peak splitting, and the F 1s spectrum exhibits dual features of Al-F and Sm-F. The Al-F peak is situated at a binding energy of 686.82 eV [29], whereas the Sm-F peak is observed at 684.08 eV [30]. The relative intensities of these peaks are 38.57% and 61.43%, respectively, corresponding to an approximate ratio of 1:1.6.

In Figure 6a, Figure 7a, Figure 8a and Figure 9a, the binding energies of 1083 eV and 1073 eV correspond to Sm³⁺ [23] and Sm²⁺ [24], respectively, with a Sm³⁺ to Sm²⁺ ratio close to 9:1 (Table 3). The increase in the Al content raises the lattice oxygen content, as shown in Figure 6b, Figure 7b, Figure 8b and Figure 9b and Table 3. The Al 2p peak at approximately 74 eV (Figure 6c, Figure 7c, Figure 8c and Figure 9c) corresponds to Al³⁺, while the F 1s peaks at approximately 684 eV and 686 eV (Figure 6d, Figure 7d, Figure 8d and Figure 9d) are attributed to Sm-F and Al-F bonds, respectively.

4. Conclusions

The reaction behavior of and valence state changes in samarium ions in reduction products were investigated using Al as a reducing agent at different molar ratios of Al/SmF₃. The reduction of SmF₃ to SmF₂ by Al is thermodynamically feasible at 933~1356 K. However, the reduction to metallic Sm is thermodynamically non-spontaneous. The valence state of Sm in the products remains unchanged despite variations in the Al/SmF₃ molar ratios, as confirmed by the nearly constant ratio δ of F to Sm, determined by the functional equation: a(δ) = –0.1794δ + 5.819 (0 ≤ δ ≤ 0.4615). The corresponding products SmF_2.413, AlF₃, and Sm(AlF)₅ are consistently obtained across Al/SmF₃ molar ratios of 1:3, 2:3, 3:3, 4:3, and 5:3. Sm(AlF)₅ is primarily formed from the reaction of AlF₃ with divalent samarium. The XPS analysis shows that samarium in the reduction products predominantly exists in the +3 oxidation state, with a minor presence of the +2 state, in a ratio of approximately 9:1. The high concentration of Sm³⁺ is attributed to the oxidation of abundant adsorbed oxygen in the products. The findings contribute to the understanding of variable valence rare earth elements, and the valence evolution of samarium may play a key role in optimizing alloy preparation processes and achieving desired material properties.