Fabrication of La2O3 Uniformly Doped Mo Nanopowders by Solution Combustion Synthesis Followed by Reduction under Hydrogen

This work reports the preparation of La2O3 uniformly doped Mo nanopowders with the particle sizes of 40–70 nm by solution combustion synthesis and subsequent hydrogen reduction (SCSHR). To reach this aim, the foam-like MoO2 precursors (20–40 nm in size) with different amounts of La2O3 were first synthesized by a solution combustion synthesis method. Next, these precursors were used to prepare La2O3 doped Mo nanopowders through hydrogen reduction. Thus, the content of La2O3 used for doping can be accurately controlled via the SCSHR route to obtain the desired loading degree. The successful doping of La2O3 into Mo nanopowders with uniform distribution were proved by X-ray photon spectroscopy and transmission electron microscopy. The preservation of the original morphology and size of the MoO2 precursor by the La2O3 doped Mo nanopowders was attributed to the pseudomorphic transport mechanism occurring at 600 °C. As shown by X-ray diffraction, the formation of Mo2C impurity, which usually occurs in the direct H2 reduction process, can be avoided by using the Ar calcination-H2 reduction process, when residual carbon is removed by the carbothermal reaction during Ar calcination at 500 °C.


Introduction
Molybdenum (Mo) and its alloys have excellent high temperature performance, such as high temperature strength, high creep resistance, low thermal expansion coefficient, and high thermal conductivity [1][2][3]. Thus, they have been widely used as high-temperature components in the aerospace industry, the nuclear industry, metal processing, and other fields [3][4][5]. However, the major drawbacks related to the inherent brittleness and high ductile-brittle transition temperature (DBTT) of Mo alloys not only lead to poor toughness at room temperature, but also to inadequate ductility and strength at high temperature, which greatly limits the widespread application of Mo and its alloys. Therefore, decreasing the DBTT to simultaneously improve both ductility and strength of Mo alloy at room temperature has become a well-known major goal over recent decades [5,6]. As previously reported [2,[7][8][9], the addition of rare-earth oxides, such as La 2 O 3 , Y 2 O 3 , and Gd 2 O 3 into Mo alloys via various doping routes to form rare-earth oxide doped Mo alloys can obviously decrease DBTT. Thus, the rare-earth oxide doped Mo alloys have high toughness and strength at both room and high temperatures. Moreover, the preparation of nano-sized oxide particles with uniform dispersion into the Mo grain and the engineering of Mo grains at nanoscale are key factors for obtaining high performance Mo alloys. Recently, it was clearly demonstrated that Mo alloys can be achieved using high quality La 2 O 3 doped Mo powder synthesized via the liquid-liquid doping route (L-L doping) [2]. Therefore, the rare-earth oxide doped powders composed of nanoparticles with a highly dispersed oxide phase are a crucial factor to obtain high-performance Mo alloys. So far, the conventional doping routes, i.e., solid Mo powders mixed with solid oxide particles (S-S doping) [7,10], solid Mo powders mixed with liquid oxide particles (S-L doping) [11][12][13], and liquid Mo-based compounds mixed with liquid oxide particles (L-L doping) [2,14], have been applied to prepare the designed rare-earth doped Mo powders. Among these strategies, the L-L doping route is the most attractive because it allows mixing of the precursors at the molecular level.
As previously reported [15], the performance of La 2 O 3 doped Mo alloys exceeds those of other doped by rare-earth oxides (i.e., Nd 2 O 3 , Sm 2 O 3 , Gd 2 O 3 , and Y 2 O 3 ). In addition, previous studies about the doping content of La 2 O 3 indicated that the doping content of La 2 O 3 exceeds 2.0 wt.%, which cannot improve the properties of molybdenum alloys [2,16]. Cheng et al. [8] studied the properties of Mo alloy with different doping contents of La 2 O 3 , and the results showed that the Mo alloy with 0.5-1.0 wt.% La 2 O 3 doping simultaneously achieve high strength and great elongation.
Solution combustion synthesis (SCS) is a versatile, energy-efficient, and mass production suitable method that has been used to fabricate several types of nano-sized oxides with different physical and chemical properties [17][18][19]. In addition, SCS is also one of the proper methods to prepare uniformly doped nano-sized oxides, such as Al-doped ZnO nanoparticles [20], Sn-doped α-Fe 2 O 3 nanoparticles [21], and Fe 3+ -W 18 O 49 nanoparticles [22], because precursor mixing at the molecular level ensured that the doped element can be distributed homogeneously throughout the matrix.
In our previous work [23], foam-like MoO 2 nanoparticles, assembled by 20-30 nm nanoparticles, were successfully fabricated by SCS. In this paper, a facile method based on the SCS and subsequent hydrogen reduction route (SCSHR) is proposed to prepare La 2 O 3 uniformly doped Mo nanopowders with various La 2 O 3 loading degrees ranging from 0 to 1.2 wt.%. In fact, the SCSHR route not only ensures the high dispersion of La 2 O 3 as a result of mixing at the molecular level, but also the formation of nanopowders with homogeneous distribution of the oxide nanoparticles. The crystalline phase evolution, morphology, elemental composition, and reduction processes have been studied in detail. Moreover, the formation mechanism of La 2 O 3 doped Mo nanopowders has been discussed.

Synthesis of La 2 O 3 Doped MoO 2 Precursor
MoO 2 precursor with different La 2 O 3 contents were synthesized via SCS using a mixture of AHM, La(NO 3 ) 3 ·6H 2 O, (NH 4 )NO 3 , and glycine as raw materials, as shown in Table 1, with different La 2 O 3 content (0, 0.3, 0.6, 0.9, and 1.2 wt.%, i.e., the mass ratio of La 2 O 3 and Mo), and labeled as P-LM0, P-LM3, P-LM6, P-LM9, and P-LM12, respectively. First, the mixture of metallic salts was dissolved in 100 mL of deionized water under magnetic stirring until a homogeneous solution was obtained. The solution was then poured into a 1000 mL beaker and heated at 160 • C in air using a temperature-controlled electric heating furnace. After heating for about 5 min at 160 • C, as soon as the solvent had evaporated, and a gelatinous mass was formed, an instantaneous combustion reaction occurred, accompanied by expansion of the gelatinous mass and the evolution of gas. The whole combustion reaction process lasted for less than 1 min and a foamy and black MoO 2 precursor doped with different La 2 O 3 contents, i.e., P-LM0, P-LM3, P-LM6, P-LM9, and P-LM12, was synthesized.

Hydrogen Reduction of Mo-Based Precursor
Lanthanum-based MoO 2 precursors (P-LM0, P-LM3, P-LM6, P-LM9, and P-LM12) were reduced by using two different reduction processes, i.e., a direct H 2 reduction process and an Ar calcination-H 2 reduction process. The detailed reduction procedure is shown in Figure 1, and the reduced products are marked as LM0, LM3, LM6, LM9, and LM12, respectively.

Hydrogen Reduction of Mo-Based Precursor
Lanthanum-based MoO2 precursors (P-LM0, P-LM3, P-LM6, P-LM9, and P-LM12) were reduced by using two different reduction processes, i.e., a direct H2 reduction process and an Ar calcination-H2 reduction process. The detailed reduction procedure is shown in Figure 1, and the reduced products are marked as LM0, LM3, LM6, LM9, and LM12, respectively.

Characterization
The crystalline phases of the samples were analyzed using an X-ray diffractometer (Rigaku, D/max-RB12, Tokyo, Japan) with Cu Kα radiation, and a 2θ angle from 10° to 90° with 0.02° increments. The chemical composition at the surface was established using X-ray photoelectron spectroscopy (XPS, Fison VG ESCA210, West Sussex, England). The morphology and size distribution of the nanoparticles were investigated by field emission scanning electron microscopy

Characterization
The crystalline phases of the samples were analyzed using an X-ray diffractometer (Rigaku, D/max-RB12, Tokyo, Japan) with Cu Kα radiation, and a 2θ angle from 10 • to 90 • with 0.02 • increments. The chemical composition at the surface was established using X-ray photoelectron spectroscopy (XPS, Fison VG ESCA210, West Sussex, England). The morphology and size distribution of the nanoparticles were investigated by field emission scanning electron microscopy (FE-SEM, Zeiss, ULTRA 55, Jena, Germany). Transmission electron microscopy (TEM) and high-resolution transmission electron microscopy (HRTEM) were performed on a transmission electron microscope (FEI, Tecnai G2 F20, Hillsboro, OR, USA) with an acceleration voltage of 200 kV. The carbon content in precursors and reduced samples was measured with a carbon/sulfur analyzer (NCS CS-2800, Beijing, China), and their La content was determined by inductively coupled plasma optical emission spectrometry (ICP-OES, Agilent 720, Santa Clara, CA, USA). The La 2 O 3 content of the reduced samples was calculated based on the result of the La content.

Phases and Morphology of Precursor
A schematic diagram of the preparation of La 2 O 3 uniformly doped Mo nanopowders via the SCSHR route is shown in Figure 2. First, MoO 2 precursors with different La 2 O 3 contents were synthesized via SCS using AHM, lanthanum nitrate, ammonium nitrate, and glycine as a molybdenum source, La source, oxidizer, and fuel, respectively. All raw materials were mixed and then heated at 160 • C in air. After about 5 min, an instantaneous combustion reaction occurred, which was accompanied by the evolution of a large amount of gas. The combustion process took less than 1 min and the target precursor samples were synthesized. Next, the obtained precursors were used to prepare La 2 O 3 doped Mo nanopowders by a direct H 2 reduction process and an Ar calcination-H 2 reduction process, respectively. The detailed reduction procedure is shown in Figure 1.

Phases and Morphology of Precursor
A schematic diagram of the preparation of La2O3 uniformly doped Mo nanopowders via the SCSHR route is shown in Figure 2. First, MoO2 precursors with different La2O3 contents were synthesized via SCS using AHM, lanthanum nitrate, ammonium nitrate, and glycine as a molybdenum source, La source, oxidizer, and fuel, respectively. All raw materials were mixed and then heated at 160 °C in air. After about 5 min, an instantaneous combustion reaction occurred, which was accompanied by the evolution of a large amount of gas. The combustion process took less than 1 min and the target precursor samples were synthesized. Next, the obtained precursors were used to prepare La2O3 doped Mo nanopowders by a direct H2 reduction process and an Ar calcination-H2 reduction process, respectively. The detailed reduction procedure is shown in Figure 1.  Figure 3 shows the XRD patterns of the different La2O3 doped precursors, synthesized by SCS. For all precursors, the diffractograms display the main diffraction peaks, which are indexed as monoclinic MoO2 (ICDD No. 78-1069). However, extremely small diffraction peaks corresponding to MoO3 and Mo4O11 phases are also observed in all XRD patterns, indicating that part of the MoO2 on the surface underwent oxidation in air. It was noted that both diffraction peaks of 26 and 29.5 twotheta degree appeared in the P-LM9 precursor sample, indicating that P-LM9 may contain more MoO3 phases than other samples. In addition, lanthanide compounds, such as La2O3 and La(OH)3 diffraction peaks were not observed for P-LM3, P-LM6, P-LM9, and P-LM12 samples, which was due to the low content of the La element. Nano-sized La2O3 has been synthesized by SCS in earlier studies [24][25][26]. Gangwar et al. [24] reported that a pure La2O3 phase can be synthesized by SCS followed by calcination treatment, while La2O3 converts to La(OH)3 upon exposure to the atmosphere due to its hygroscopic nature.  For all precursors, the diffractograms display the main diffraction peaks, which are indexed as monoclinic MoO 2 (ICDD No. 78-1069). However, extremely small diffraction peaks corresponding to MoO 3 and Mo 4 O 11 phases are also observed in all XRD patterns, indicating that part of the MoO 2 on the surface underwent oxidation in air. It was noted that both diffraction peaks of 26 and 29.5 two-theta degree appeared in the P-LM9 precursor sample, indicating that P-LM9 may contain more MoO 3 phases than other samples. In addition, lanthanide compounds, such as La 2 O 3 and La(OH) 3 diffraction peaks were not observed for P-LM3, P-LM6, P-LM9, and P-LM12 samples, which was due to the low content of the La element. Nano-sized La 2 O 3 has been synthesized by SCS in earlier studies [24][25][26]. Gangwar et al. [24] reported that a pure La 2 O 3 phase can be synthesized by SCS followed by calcination treatment, while La 2 O 3 converts to La(OH) 3 upon exposure to the atmosphere due to its hygroscopic nature.  The morphology and microstructure of the precursors were assessed by FE-SEM. Representative images are shown in Figure 4. It can be observed that the P-LM0 precursor sample has a foam-like morphology, as shown in Figure 4a, which is the unique typical morphology obtained by the SCS method. Figure 4b shows that the foam-like P-LM0 is composed of MoO2 nanoparticles with sizes of 20-40 nm as a result of particle agglomeration due to the high surface energy of MoO2 nanoparticles. Similarly, the morphology of La2O3 doped precursors, i.e., P-LM3, P-LM6, P-LM9, and P-LM12, still maintain the foam-like morphology and particle agglomeration. In addition, the particle size of La2O3 doped precursors does not change significantly, indicating that the morphology and particle size of precursors is not affected when doping with low La2O3 amounts, as shown in Figure 4c-f. However, due to the low loading degree of La2O3, the La2O3 nanoparticles are difficult to see in these images. The morphology and microstructure of the precursors were assessed by FE-SEM. Representative images are shown in Figure 4. It can be observed that the P-LM0 precursor sample has a foam-like morphology, as shown in Figure 4a, which is the unique typical morphology obtained by the SCS method. Figure 4b shows that the foam-like P-LM0 is composed of MoO 2 nanoparticles with sizes of 20-40 nm as a result of particle agglomeration due to the high surface energy of MoO 2 nanoparticles. Similarly, the morphology of La 2 O 3 doped precursors, i.e., P-LM3, P-LM6, P-LM9, and P-LM12, still maintain the foam-like morphology and particle agglomeration. In addition, the particle size of La 2 O 3 doped precursors does not change significantly, indicating that the morphology and particle size of precursors is not affected when doping with low La 2 O 3 amounts, as shown in Figure 4c-f. However, due to the low loading degree of La 2 O 3 , the La 2 O 3 nanoparticles are difficult to see in these images.  To clarify whether La2O3 has been successfully doped into the precursor, XPS was performed. Figure 5 shows a survey scan, La 3d spectra, and Mo 3d high-resolution spectra for each precursor. As shown in Figure 5a, no peak of La 3d was observed in P-LM0. However, after La2O3 doping, peaks corresponding to La 3d3/2 and La 3d5/2 appeared whose intensities increase as the amount of  Table 2. As shown, the contents of impurities, such as MoO3 and Mo4O11, were found to be quite high, which far exceeds that indicated by the Rietveld refinement XRD result of a previous study by our research group [27]. This result can To clarify whether La 2 O 3 has been successfully doped into the precursor, XPS was performed. Figure 5 shows a survey scan, La 3d spectra, and Mo 3d high-resolution spectra for each precursor. As shown in Figure 5a, no peak of La 3d was observed in P-LM0. However, after La 2 O 3 doping, peaks corresponding to La 3d 3/2 and La 3d 5/2 appeared whose intensities increase as the amount of  Table 2. As shown, the contents of impurities, such as MoO 3 and Mo 4 O 11 , were found to be quite high, which far exceeds that indicated by the Rietveld refinement XRD result of a previous study by our research group [27]. This result can be related to the XPS technique, which can analyze only the surface with an information depth of up to 10 nm and can be affected by surface oxidation of the MoO 2 nanoparticles.     Table 3 shows the content of C and La elements in each precursor. As observed, each precursor contains traces of carbon. The La content is close to the theoretical content, indicating that the La element is not lost through the SCS method, which ensures the accuracy of La doping.

Preparation of La 2 O 3 Doped Mo Nanopowders
As shown in Figure 1, the P-LM0, P-LM3, P-LM6, P-LM9, and P-LM precursors were reduced by both direct H 2 reduction and Ar calcination-H 2 reduction processes. The corresponding reduction products are referred to as LM0, LM3, LM6, LM9, and LM12 hereafter.
The XRD patterns of the products reduced by direct H 2 reduction and Ar calcination-H 2 reduction process are illustrated in Figures 6 and 7, respectively. As shown in Figure 6, the main diffraction peaks displayed by all diffractograms correspond to the Mo phase while no evidence for MoO 2 peaks is found. This suggests that MoO 2 and impurities (MoO 3 and Mo 4 O 11 ) were completely reduced to the metallic Mo phase. Yet, a small diffraction peak at around 38 • can be observed in each diffraction pattern. It is attributed to the Mo 2 C phase, which may result from the reaction between residual carbon and MoO 2 (or MoO 3 , Mo 4 O 11 ). On the other hand, the XRD patterns of products obtained by the Ar calcination-H 2 reduction process, as shown in Figure 7, shows that zero-valent Mo is the only crystalline phase while Mo 2 C does not appear anymore, clearly suggesting that the formation of Mo 2 C can be avoided via rational preparation of La 2 O 3 doped Mo powders by the Ar calcination-H 2 reduction process. Indeed, the C content was extremely low, i.e., 0.046, 0.043, 0.053, 0.051, and 0.042 wt.% for LM0, LM3, LM6, LM9, and LM12, respectively, as listed in Table 4. Moreover, the C content was much lower than that of samples obtained by direct H 2 reduction. Therefore, in light of these results, it can be stated that the impurity-free La 2 O 3 doped Mo powders can be successfully synthesized by the Ar calcination-H 2 reduction process.   However, as mentioned above, the low amount of La2O3 did not allow its detection by XRD in any of the samples. Regarding Mo, the full width at half maxima (FWHM) of the peak was appropriate to calculate the crystallite size of Mo obtained by the Ar calcination-H2 reduction process   However, as mentioned above, the low amount of La2O3 did not allow its detection by XRD in any of the samples. Regarding Mo, the full width at half maxima (FWHM) of the peak was appropriate to calculate the crystallite size of Mo obtained by the Ar calcination-H2 reduction process  However, as mentioned above, the low amount of La 2 O 3 did not allow its detection by XRD in any of the samples. Regarding Mo, the full width at half maxima (FWHM) of the peak was appropriate to calculate the crystallite size of Mo obtained by the Ar calcination-H 2 reduction process by applying the Scherrer equation. Hence, crystallites of 28.7, 24.3, 23.8, 22.7, and 21.9 nm were obtained for LM0, LM3, LM6, LM9, and LM12, respectively. Figure 8 shows the XPS survey scan and La 3d spectra of La 2 O 3 doped Mo nanopowders obtained by the Ar calcination-H 2 reduction process and Mo 3d high-resolution spectra for the P-LM9 and LM9 sample. As shown in Figure 8a, the increase in La 2 O 3 doping resulted in more intense La 3d peaks, underlying the doping of Mo powders with La 2 O 3 . Figure 8b comparatively displays the Mo 3d high-resolution spectra of P-LM9 and LM9. It can be noticed that a new peak at 280.2 eV appeared in Mo 3d high-resolution spectra of LM9, which is ascribed to metallic Mo. Thus, this is evidence of the formation of a metallic Mo phase in the sample. In addition, well-shaped peaks corresponding to MoO 2 , MoO 3 , and Mo 4 O 11 phases are observed for LM9, suggesting that Mo nanoparticles at the surface were oxidized in air. The Mo 3d high-resolution spectra for the other samples, i.e., LM0, LM3, LM6, and LM12, show similar characteristics to that of LM9.  Figure 8 shows the XPS survey scan and La 3d spectra of La2O3 doped Mo nanopowders obtained by the Ar calcination-H2 reduction process and Mo 3d high-resolution spectra for the P-LM9 and LM9 sample. As shown in Figure 8a, the increase in La2O3 doping resulted in more intense La 3d peaks, underlying the doping of Mo powders with La2O3. Figure 8b comparatively displays the Mo 3d high-resolution spectra of P-LM9 and LM9. It can be noticed that a new peak at 280.2 eV appeared in Mo 3d high-resolution spectra of LM9, which is ascribed to metallic Mo. Thus, this is evidence of the formation of a metallic Mo phase in the sample. In addition, well-shaped peaks corresponding to MoO2, MoO3, and Mo4O11 phases are observed for LM9, suggesting that Mo nanoparticles at the surface were oxidized in air. The Mo 3d high-resolution spectra for the other samples, i.e., LM0, LM3, LM6, and LM12, show similar characteristics to that of LM9.  FE-SEM images of La 2 O 3 doped Mo nanopowders reduced by the Ar calcination-H 2 reduction process are displayed in Figure 9. Aggregates composed of nanoparticles are clearly seen for each product. The foam-like morphology seen in Figure 9d indicates that the samples maintain the original morphology of MoO 2 precursor during the reduction process. The average particle sizes, obtained from the SEM images, increase as the amount of La 2 O 3 decreases (43 nm (LM12) > 49 nm (LM9) > 58 nm (LM6) > 63 nm (LM3) > 68 nm (LM0)), revealing the inhibition of the segregation process induced by La 2 O 3 doping. FE-SEM images of La2O3 doped Mo nanopowders reduced by the Ar calcination-H2 reduction process are displayed in Figure 9. Aggregates composed of nanoparticles are clearly seen for each product. The foam-like morphology seen in Figure 9d indicates that the samples maintain the original morphology of MoO2 precursor during the reduction process. The average particle sizes, obtained from the SEM images, increase as the amount of La2O3 decreases (43 nm (LM12) > 49 nm (LM9) > 58 nm (LM6) > 63 nm (LM3) > 68 nm (LM0)), revealing the inhibition of the segregation process induced by La2O3 doping. TEM micrographs of the LM9 product reduced by the Ar calcination-H2 reduction process are shown in Figure 10. Figure 10a displays aggregates of irregular shapes of Mo nanoparticles. The corresponding selected-area diffraction electron patterns (SAED), shown in the inset, exhibit a

Formation Mechanism of La2O3 Doped Mo Nanopowders
According to theory, SCS involves strong oxidation-reduction reactions. In this reaction system, the La2O3 doped MoO2 precursors were formed by Equation (1) and Equation (2). Furthermore, a large amount of gases generated led to the formation of foam-like morphology [23]. Because the SCS method involves reactions of precursors in solution, the incorporation of the entire amount of doping element is ensured, as well as the uniform distribution of La2O3 into the host material.

Formation Mechanism of La 2 O 3 Doped Mo Nanopowders
According to theory, SCS involves strong oxidation-reduction reactions. In this reaction system, the La 2 O 3 doped MoO 2 precursors were formed by Equation (1) and Equation (2). Furthermore, a large amount of gases generated led to the formation of foam-like morphology [23]. Because the SCS method involves reactions of precursors in solution, the incorporation of the entire amount of doping element is ensured, as well as the uniform distribution of La 2 O 3 into the host material.
Recent studies have shown that the reduction mechanism of MoO 2 to Mo under H 2 depends on the experimental conditions, in particular, the temperature of reduction [28]. The reduction reaction obeys the chemical vapor transport (CVT) mechanism, when the reaction temperature is above 923 K (650 • C). When the reaction temperature is lower than 883 K (610 • C), the mechanism is a pseudomorphic transport mechanism, which results in the nucleation and growth of Mo inside MoO 2 particles. Thus, the Mo products will retain the morphology and size of the initial MoO 2 particles. In light of these results, the preservation of the foam-like morphology and nanostructure of the MoO 2 precursor by the La 2 O 3 doped Mo products obtained at 600 • C is explained.
When the precursors were prepared by the direct H 2 reduction process, the impurities consisting of MoO 3 and Mo 4 O 11 were reduced to MoO 2 based on Equations (3) and (4) [29], respectively. At 600 • C, Equation (5) takes place, forming metallic Mo nanoparticles. Furthermore, an insignificant amount of Mo 2 C impurity formed as a result of the carbothermal reduction in hydrogen of MoO 2 [27,30,31], as shown by Equation (6). La 2 O 3 does not undergo any phase transition due to its high thermodynamic stability.
When the precursors were calcined in Ar at 500 • C, residual carbon was completely consumed via carbothermal reduction, i.e., Equations (7) and (8). Therefore, the La 2 O 3 doped Mo nanopowders without Mo 2 C impurity were formed via the Ar calcination-H 2 reduction process.

Conclusions
Mo nanopowders doped with different amounts of La 2 O 3 , i.e., 0, 0.27, 0.64, 0.88, and 1.16 wt.%, were prepared via the SCSHR route, which allowed a homogeneous distribution of the dopant while preserving the theoretical loading degree. The MoO 2 precursors with different content of La 2 O 3 , synthesized by SCS, displayed similar foam-like morphology and particle sizes of 20-40 nm. Then, La 2 O 3 uniformly doped Mo nanopowders were prepared via different reduction processes, i.e., the direct H 2 reduction and Ar calcination-H 2 reduction. Compared to the direct H 2 reduction processes, the Ar calcination-H 2 reduction process led to La 2 O 3 uniformly doped Mo nanopowders without Mo 2 C impurities. La 2 O 3 uniformly doped Mo nanopowders were composed of nanoparticles with sizes of 40-70 nm, which was in line with the decrease in La 2 O 3 . Hence, La 2 O 3 doped Mo nanopowders retained the original morphology and size of the MoO 2 precursor because the formation of Mo nanoparticles reduced at 600 • C obeys the pseudomorphic transport mechanism. The formation of Mo 2 C impurity was noticed due to the carbothermal reduction occurring when samples were directly reduced under H 2 . On the other hand, when the Ar calcination-H 2 reduction process was applied to form lanthana-based materials, the residual carbon was completely consumed via the carbothermal reduction during the calcination in Ar at 500 • C and thus La 2 O 3 doped Mo nanopowders without Mo 2 C impurity can be successfully prepared. The results obtained herein are valuable to recommend the SCSHR method for preparation of other oxide doped refractory metal nanopowders (e.g., tungsten and niobium).