Effect of preparation method on the microstructure and optical properties of Y₂O₃–MgO composite ceramics

1. Introduction

Transparent ceramics have been identified as potential candidates for a variety of applications, notably as mid-infrared window materials, due to their excellent mechanical, thermal and optical properties.¹ Nevertheless, the potential for enhancing the operational characteristics of prevalent single-phase infrared (IR) materials, such as Y₂O₃, MgO, ZnS, which are used in extreme temperature conditions and harsh environment, including aerospace applications, is currently limited.^2,3,4 Consequently, yttrium oxide–magnesium oxide (Y₂O₃–MgO) composite ceramics have recently emerged as a new research direction. The Y₂O₃–MgO nanocomposite (Y:Mg) has attracted widespread attention owing to its superior combination of properties. These include high strength at elevated temperatures, excellent abrasion resistance and outstanding resistance to chemical attack (oxidation, corrosion, etc.), as well as a longer wavelength cutoff and lower emissivity compared to other mid-infrared materials.^5,6

It is crucial that both the synthesis and sintering conditions of composite nanopowders are meticulously controlled to fabricate nanocomposites with fine grains and homogeneous microstructures.^7,8 One technique to reduce grain size is to disperse the second phase. In this context, compared with traditional single-case ceramics, the pining effect produced by the second phase in multiphase ceramics can inhibit the grain growth of adjacent phases during the sintering process.⁹ This pinning effect is most effective when the nanophases have comparable volume fractions and are uniformly dispersed within the nanocomposite powder.

Additionally, it has been demonstrated that such homogeneous nanoscales of the composite powders result in the sintering characteristics of the composite powders, leading to the formation of thermally stable microstructures of interconnected and continuous phases in the consolidated bulk material.⁶ The performance of ceramic powder as a starting material is directly correlated with the quality of the final product. To achieve a pure phase and low porosity ceramic sinter, the primary requirement for the initial powder includes: (1) small particle size of the powder, (2) a high dispersion rate with no hard agglomeration and (3) high sintering activity with a uniform particle size distribution.¹⁰

At present, various preparation methods for Y₂O₃–MgO composite powders have been proposed, including the solid-state reaction method, sol–gel,¹¹ sol–gel combustion method,⁵ hydrothermal methods,⁴ citrate-nitrate combustions,¹² chemical co-precipitation,¹³ bioorganic method¹⁴ and glycine–nitrate process. However, there are few reports on the preparation of Y₂O₃–MgO complex phase powders by co-precipitation, so in this experiment, the co-precipitation method will be used to prepare Y₂O₃–MgO nanopowders and compare it with two other milling methods commonly used in the laboratory.

Brard et al. synthesized a homogeneous powder with a grain size of about 10nm by the sol–gel method, and sintered the powder to obtain a fully dense ceramic with an average grain size of 150nm. Alhaji et al. used sucrose as the incendiary agent to prepare Y₂O₃–MgO nanocomposites by burning the sol–gel method, and the TEM results proved that the Y₂O₃–MgO nanopowders grew in three dimensions. Permin et al. developed a self-propagating high-temperature synthesis method of glycine nitrate to produce composite Y₂O₃–MgO nanopowders, and studied the infrared transmittance of Y₂O₃–MgO ceramics related to sintering conditions, and the optimal transmittance was 80.9%@5 $\mu$m  $\mathrm{\text{Hv}}=10.2$GPa.

In this experiment, Y₂O₃–MgO nanopowders are prepared by co-precipitation, sol–gel and glycine nitrate process (GNP). The particle size and microstructure of the powders prepared via three methods are compared following calcination at varying temperatures. Concurrently, a comparative analysis is conducted on the microstructure and transmittance of ceramics sintered from different powders under identical conditions. The calcined powders are hot pressed at 1350°C and 50MPa for 30min.

2. Experimental Section

2.1. Materials

The following materials were utilized in the experimental study.

Magnesium oxide (MgO), with a purity of 99.99%, was obtained from Shanghai Macklin Biochemical, located in Shanghai, China. Yttrium nitrate hexahydrate (Y(NO₃)$\cdot 6$H₂O), with a purity of 99.99%, was sourced from Shanghai Macklin Biochemical, China. Citric acid monohydrate with a purity of 99.9%, was acquired from Shanghai Macklin Biochemical, China. Glycine (NH₂CH₂COOH), with a purity of 99%, was acquired from Shanghai Macklin Biochemical, China. Ammonium bicarbonate (NH₄HCO₃), with a purity of 99%, was purchased from Shanghai Macklin Biochemical, China. Nitric acid (HNO₃), with a purity of 99.9%, was purchased from Shanghai Macklin Biochemical, China. Ethylene glycol which was of analytical reagent grade, was obtained from Shanghai Macklin Biochemical, China.

2.2. Co-precipitation preparation Y₂O₃–MgO nanopowders

For the co-precipitation synthesis of Y₂O₃–MgO nanopowders, a solution containing Y(NO₃)$\cdot 6$H₂O and Mg(NO₃)₂ is mixed in a volume ratio corresponding to a 50:50 ratio of the oxides. A specified quantity of NH₄HCO₃ solution is subsequently added. The mixture is then aged for 12h and centrifuged. The resulting sediment is washed three times by resuspension in deionized water and centrifugation. The washed powder is dried in an oven at 200°C. Following drying, the powders are sieved through a 200-mesh screen and calcined in a tube furnace with a heating rate of 1°C/min, maintaining a temperature of 600°C for 10h.

2.3. GNP preparation Y₂O₃–MgO nanopowders

Y₂O₃–MgO nanopowders are prepared using the GNP (glycine–nitrate process) method by mixing solutions of Y(NO₃) and Mg(NO₃)₂ based on a 50:50 volume ratio of the oxides. Glycine and ethylene glycol solutions are added sequentially in specific proportions. To remove any excess water, the solution is evaporated at 110°C. Once the water has been evaporated, the solution is foamed in an oven preheated to 200°C. After foaming, the material is calcined at a rate of 1°/min to a final temperature of 600°C held for 10h. The reaction between metal nitrate as an oxidant and glycine as fuel results in the completion of the oxide powder preparation, accompanied by the instantaneous release of a large amount of heat.^12,15

2.4. Sol–gel preparation Y₂O₃–MgO nanopowders

For the sol–gel synthesis of Y₂O₃–MgO nanopowders, MgO is first dissolved in HNO₃ to form Mg (NO₃)₂. Y(NO₃) and Mg (NO₃)₂ solutions are then mixed according to a 50:50 volume ratio of the oxides, followed by the sequential addition of a specific proportion of citric acid monohydrate and ethylene glycol solution. The solution is stirred until the water has completely evaporated, after which it is placed into an oven preheated to 200°C and held until white foam forms. The foam is subsequently calcined at 600°C with a rate of 1°/min for a duration of 10h.

2.5. Characterization

The crystalline phases of the powders are characterized by Powder X-ray diffraction (XRD, Bruker AXS AS D2) analysis with Cu K$\alpha$ radiation spanning a 2$\theta$ range of 15–85°. The Scherrer formula is employed as a supplementary tool to estimate the average crystalline size $D_{\mathrm{\text{(XRD)}}}$. The specific surface area of powders is measured using a gas adsorption analyzer (Model Micromeritics ASAP 2010, Norcross, GA) based on the Brunauer–Emmett–Teller (BET) method. The average particle sizes $D_{\mathrm{\text{(BET)}}}$ were calculated from the specific area using the following equation^16,17 :

The theoretical density ($\rho$) of the Y₂O₃–MgO composite with a 50:50 volume percentage is 4.309g/cm³. This value represents the density of the material in its perfectly crystalline and nonporous state. $S_{\mathrm{\text{BET}}}$ (m²/g) is the specific area based on the BET method. Generally, $\varphi =D_{\mathrm{\text{(BET)}}}/D_{\mathrm{\text{(XRD)}}}$ is defined as a factor that reflects the agglomeration extent of the primary crystalline.

The morphologies of calcined Y₂O₃–MgO powders and the microstructures of the sintered ceramics were characterized using a field-emission scanning electron microscope (FESEM, HITACHI, SU8010). The transmission electron microscopy (TEM) of the Y₂O₃–MgO powder was conducted using field-emission transmission electron microscopy (FETEM; FEI Tecnai G2 F20 S-Twin 200KV). The relative densities of Y₂O₃–MgO ceramics were determined using the Archimedes method.

3. Results and Discussion

3.1. Characterization of Y₂O₃–MgO nanopowders

Figure 1 illustrates the XRD patterns of the Y₂O₃–MgO composite powders via various methods. Distinct diffraction peaks corresponding to MgO are observed at 2$\theta$ angles of 37.08°, 43.06° and 62.38°, among others. Similarly, characteristic peaks of Y₂O₃ are identified at 29.34°, 33.8° and 48.64°, and other locations. The XRD pattern can be exclusively indexed to pure Y₂O₃ (PDF No.: 86-1107) and MgO phases (PDF No.: 78-0430), with no detectable secondary phase present. This indicates that no phase transformation or reaction occurred within the Y₂O₃ and MgO grains during the sintering process. Additionally, it is evident from the figure that as the calcination temperature rises, the intensity of the powder’s diffraction peaks increases, signifying improved crystallization.

Solutions prepared with varying concentrations of yttrium nitrate were placed in an oven preheated to 200°C and maintained for 40h. During this period, physical water was released, esterification reactions took place and the nitrates decomposed to form a porous organic foam. The thermal characteristics of the resulting organic foam were then analyzed using a simultaneous thermogravimetric-differential thermal analysis (TG-DTA) instrument. Figure 2 shows the TG-DSC results for Y₂O₃–MgO powder prepared by different methods. The TG data reveal a continuous decrease in the mixture’s weight as the heating temperature increased from room temperature to 600°C. No further weight loss was observed beyond 600°C, indicating that the thermal decomposition process was complete at this temperature. The endothermic peaks in Figs. 2(a)–2(c) are attributed to the vaporization of physical water and water of crystallization.¹⁸ The total weight loss for powder prepared using the co-precipitation method is 54.71%, with the endothermic peak at 201°C in Fig. 2(a) is due to the decomposition of nitrates. The total weight loss for powder prepared using the sol–gel method is 79.29%. while that prepared by the gel-casting process (GNP) method is 31.23%. Several exothermic peaks in Figs. 2(b) and 2(c) correspond to the redox reaction between the esterification reaction products and nitrates, as well as the decomposition of residual organic components, aligning with the combustion temperatures.^18,19

Figure 3 investigates the effects of specific surface area (measured by the BET method) and agglomeration extent on the composite nanopowders synthesized via different methods. Lower $\varphi$ values indicate reduced agglomeration among particles. A $\varphi$ value of 1.0 suggests no agglomeration among crystallites. The calculations show that the powder agglomeration resulting from the sol–gel method is minimal, whereas that from the GNP method is more pronounced. The small specific surface area of the powders prepared by GNP is that in the process of preparing the powder by GNP may be due to the intense combustion reaction during powder preparation, which leads to small particle sizes, good fluffiness and dispersion. However, since the powder is ejected from the crucible during the reaction, this leaves behind powder with inferior fluffiness and dispersion.

Figure 4 presents scanning electron microscopy (SEM) images of Y₂O₃–MgO nanopowders prepared by different methods and calcined for 3h at varying temperatures. The powder prepared by the co-precipitation method presents an irregular spherical shape, and with the increase of calcination temperature, the irregular spherical shape gradually agglomerates into long strips. The powders prepared by sol–gel and GNP methods are spherical with fine grains and uniform shapes, but with the increase of calcination temperature, the powders prepared by GNP methods are more likely to produce hard agglomeration and form large agglomerates. The particle size distributions of local nanoparticles obtained from SEM images are displayed in Fig. 5. As can be seen from the particle size distribution map obtained from the SEM. In this part, it is clear from the figure that the average grain size of the powder grows with the increasing calcination temperature. At the same calcination temperature, the powder prepared by the sol–gel method has the smallest particle size. Since inorganic matter is not fully decomposed at 400°C, and the particle size of the powder at 750°C is larger, powder calcined at 600°C is chosen for subsequent ceramic sintering. At 600°C, the powder’s average particle size prepared by the sol–gel method is approximately 40.59nm, while that prepared by the GNP method is relatively large, about 52.85nm.

Figure 6 demonstrates the different crystallographic planes of Y₂O₃–MgO nanocomposites. The d-spacing values obtained from TEM are consistent with those derived from XRD analysis, showing excellent agreement with the JCPDS reference values. TEM also corroborates the presence of Y₂O₃ and MgO phases in the as-prepared powder. The selected area diffraction (SAD) exhibits a ring-like configuration, indicating that the composite comprises fine grains.²⁰

3.2. Characterization of Y₂O₃–MgO ceramic

Figure 7 illustrates the SEM images of ceramics produced via various methods under hot pressing at 1350°C for 30min. The microstructures exhibit a high degree of homogeneity irrespective of the preparation method used. However, intergranular pores are evident in the Y₂O₃–MgO ceramics prepared via the GNP. The SEM analysis indicates that the bright phase was Y₂O₃, whereas the dark phase was MgO. The integrity of the majority of the MgO grains is preserved, which can be attributed to their significantly higher mechanical characteristics when compared to Y₂O₃. With volume fractions of MgO and Y₂O₃ both at 50%, each phase can function as both the secondary phase and the matrix phase for the other, consequently leading to significant grain growth suppression through grain boundary pinning. This results in the formation of fine-grained Y₂O₃–MgO composite ceramics. Notably, the sol–gel method yields the smallest grain size of approxim ately 0.56$\mu$m, while the GNP method produces the largest grain size, averaging 0.91$\mu$m.

The relative density of ceramics prepared by different methods was measured by Archimedes’ principle. The results showed that ceramics prepared via co-precipitation and sol–gel methods exhibit high and comparable relative densities, reaching 99.86% and 99.93%, respectively. In contrast, those prepared by GNP exhibit a reduced relative density of 90%. Figure 7(b) also reveals a higher incidence of intergranular pores in these ceramics. The small specific surface area of the powder, the presence of agglomeration combined with low sintering activity contribute to the decreased density of the ceramics.

The infrared transmittance of ceramics prepared by different methods without annealing, as shown in Fig. 8, was measured after hot pressing at 1350°C for 30min. The data obtained from Fig. 8 indicates that the transmittance within the IR wavelength range ($\lambda =2$–10$\mu$m) is the highest for Y₂O₃–MgO ceramic prepared by the sol–gel method, peaking at 75.57% at 5.6$\mu$m. The transmittance efficiency is lowest in ceramics prepared by the GNP method, with a value of 4.51% recorded at 6.92$\mu$m. An absorption peak observed at 4.9$\mu$m in the co-precipitation method is associated with carbon monoxide present in the residual pores of the ceramics. The source of carbon in the Y₂O₃–MgO composites includes materials from the hot zone of the press and incomplete washing of the initial precipitate during the washing process.³

Figure 9 depicts the transmittance of Y₂O₃–MgO ceramics following annealing at 1050°C for 40h. After calcination at this temperature, all carbon monoxide is completely oxidized. Upon cooling, the resulting carbon dioxide reacts with either Y₂O₃ or MgO to form corresponding carbonates. This reaction is evidenced by the emergence of two overlapping absorption bands near 7$\mu$m post-annealing, which are assignable to both magnesium carbonate and yttrium carbonate.^21,22 These absorption bands are situated outside the atmosphere transmission window and, therefore, do not impact the functional properties of the composites for most potential applications.

4. Conclusion

This study has systematically investigated the influence of powder preparation methods on the microstructure and optical properties of Y₂O₃–MgO ceramics. The sol–gel method was found to yield powder with the highest specific surface area, which contributed to the highest light transmittance of 85.33% at a wavelength of 5.31$\mu$m. The co-precipitation method produced a composite with a transmittance of 77.21% at 5.44$\mu$m, while the glycine–nitrate process (GNP) resulted in the lowest transmittance. The composites fabricated by the GNP method also exhibited the lowest relative density of 90% and the largest average grain size of 0.91$\mu$m, which is a significant factor affecting the optical transmittance.^23,24 It is widely acknowledged that the optical transmittance of ceramics is primarily influenced by their relative density and grain size. The choice of powder raw materials and preparation methods significantly affects the sintering density. The key differences among the powders in this study were their surface area and particle size. Smaller particle sizes in the ceramic powders lead to a large specific surface area, enhancing the contact area between particles and promoting increased sintering activity. Consequently, the sol–gel method proved to be the most effective for preparing Y₂O₃–MgO composite ceramics with a uniform microstructure and superior optical transmittance.

Acknowledgments

The authors would like to thank the Sichuan Research Center of New Materials for performing Y₂O₃–MgO ceramics and providing technical discussion.