Wei
Ren
,
Xi
Lan
,
Lei
Guo
and
Zhancheng
Guo
*
State Key Laboratory of Advanced Metallurgy, University of Science and Technology Beijing, Beijing 100083, China. E-mail: zcguo@ustb.edu.cn
First published on 2nd July 2024
Electrodeposited Invar alloy film is considered an ideal material for fabricating OLED fine metal masks (FMMs) due to its advantages of a simple preparation process, cost-effectiveness, and highly controllable thickness. However, electrodeposited Invar alloy films exhibit a high thermal expansion coefficient (CTE). In this study, the phase transition temperature range of electrodeposited Invar alloy films was obtained and hydrogen annealing experiments were conducted. The results indicate that grain size, carbon (C) and sulfur (S) impurity contents, as well as compositional homogeneity, influence the CTE of electrodeposited Invar alloy films, with grain size dominating. The CTE is inversely proportional to grain size, as the CTE at the grain boundaries of electrodeposited Invar alloy films is significantly higher than at crystallites. After hydrogen annealing at 1073 K, the average grain size is approximately 3.3 μm, with the C content decreasing to 50 ppm and the S content dropping below 10 ppm. The CTE of electrodeposited Invar alloy film is reduced to 1.0 × 10−6 K−1, comparable to that of conventionally manufactured Invar alloy. Consequently, electrodeposited Invar alloy films with near-zero inclusions and extremely low CTE were obtained. This study enables the application of electrodeposited Invar alloy films in FMM fabrication.
Invar alloy films prepared using conventional melt-casting and rolling methods face limitations in reducing material thickness and are significantly affected by high impurity content, thereby notably impacting production yield. In this context, electrodeposition, as a precise alloy preparation technique, has garnered increasing research attention. Compared to traditional metallurgical methods, electrodeposition offers significant advantages in preparing alloy films, particularly in alloy composition modulation, uniformity, and microstructure.11–13 Materials prepared via electrodeposition are theoretically pure, devoid of impurities, and showcase potential advantages in FMM preparation. Moreover, the process of preparing electrodeposited Invar alloy films is simple and cost-effective, making it suitable for large-scale production. This presents the potential feasibility for industrial production, aiming to meet diverse application demands.14–16
Nevertheless, the high CTE of electrodeposited Invar alloy film limits its applications. The high CTE of electrodeposited Invar alloy film can be primarily attributed to two factors. Previous studies have shown that electrodeposited Invar alloy films consist of a metastable BCC phase, equilibrium FCC phase, or mixed BCC + FCC phases.12,17,18 The CTE of the BCC phase is much higher than that of the FCC phase, attributed to the arrangement of atoms in the BCC crystal structure, resulting in relatively large volume changes with temperature variations.12 In addition, the Invar alloy film prepared using the electrodeposition method exhibits a significant nanocrystalline structure with a substantial volume fraction of grain boundaries. Theoretical calculations suggest that the thermal expansion behavior at the grain boundaries is different from that within the grains, with a much higher CTE.19 This phenomenon has been confirmed in nanocrystalline Cu20 and nanocrystalline Ni–P.21 Thus, controlling grain size becomes a feasible method to regulate the thermal expansion behavior of nanomaterials. Compared to the corresponding coarse-grained Invar alloy, the CTE of nanocrystalline Invar alloy (∼9.0 × 10−6 K−1) is relatively high. Previous studies have found that vacuum annealing can reduce the CTE of electrodeposited Invar alloy films. However, sulfide precipitation occurs during vacuum annealing, resulting in a decrease in the purity of the alloy film.22,23
Therefore, to address these issues, we propose a hydrogen annealing strategy for electrodeposited Invar alloy films. This strategy aims to enhance grain growth, diminish internal impurity content, and further lower the CTE of electrodeposited Invar alloy films, which is theoretically feasible. The goal is to prepare Invar alloy films with low CTE and pure, impurity-free properties for FMM. This effort provides a theoretical basis for expanding the industrial applications of electrodeposited Invar alloy. We aim to provide a new perspective and method for the efficient and precise preparation of FMMs in OLED production.
Parameters | Value |
---|---|
NiSO4·6H2O | 150 g L−1 |
NiCl2·6H2O | 30 g L−1 |
FeSO4·7H2O | 106 g L−1 |
C6H6Na6O7·2H2O | 8 g L−1 |
C6H8O6 | 5 g L−1 |
Saccharin sodium | 4 g L−1 |
C12H25SO3Na | 0.3 g L−1 |
Bath pH | 2.5 |
Bath temperature | 323 ± 1 K |
Current density | 5 A dm−2 |
Invar alloy films with a thickness of 25 μm were prepared and their composition was analyzed using inductively coupled plasma mass spectrometry (ICP-MS, ICAPRQ, Thermo, Scientific). Concentrations of C and S were determined using a carbon/sulfur analyzer (EMIA-920V2, Horiba). An electron probe microanalyzer (EPMA-1720H, Shimadzu) was used to analyze the distribution of C and S within the alloy. In situ variable-temperature X-ray diffraction (XRD) analysis of the electrodeposited Invar alloy films was conducted using a variable-temperature X-ray diffraction system (Rigaku, DMAX/2550), enabling the examination of crystal structure, phase composition, and determination of the phase transition temperature. The analysis was performed with a tube voltage of 40 kV, a tube current of 25 mA, and a heating rate of 10 K min−1. High-purity argon gas (99.99%) was used throughout the process to prevent oxidation. For the in situ variable-temperature XRD analysis, a 10 mm diameter sample was placed in an Al2O3 crucible, subjected to a heating rate of 10 K min−1, and scans were obtained over a range of 40–56° 2θ with a step size of 2°. Instrumental line broadening was assessed using a silicon standard sample. Annealing was conducted in a tube furnace under a hydrogen gas atmosphere (flow rate: 200 mL min−1) at temperatures of 773 K, 873 K, 973 K, 1023 K, and 1073 K, respectively, for 1 h, followed by furnace cooling.
Scanning electron microscopy (SEM, Phenom ProX 150) and transmission electron microscopy (TEM, JEM-2010) were used to observe the microstructure and cross-sectional composition distribution before and after annealing. Focused ion beam scanning electron microscopy (FIB-SEM, crossbeam-350, Zeiss) was used to observe the distribution of inclusions within the alloy foil. During the sample preparation process, the automatic cutting mode was used, the slice thickness was set to 30 nm, and the slices were cut layer by layer by the ion beam along the horizontal direction, and SEM imaging was performed alternatively. Finally, the resulting two-dimensional (2D) images were reconstructed in three-dimensional (3D) using Aviso software to visualize the distribution of inclusions inside the alloy. Dimensional changes in the samples were measured relative to temperature using a NETZSCH DIL 402SE expander within the temperature range of 293 K to 393 K, under near-zero load force. During measurements, a heating rate of 5 K min−1 was maintained under the protection of helium gas, and samples measuring 4 mm × 25 mm were used, calibrated with an Al2O3 standard sample. The CTE (α) was obtained by substituting the measurement results between 293 K and 393 K into eqn (1):
![]() | (1) |
Element | Ni | Mn | Al | Ca | Mg | S | C | Si | Fe |
---|---|---|---|---|---|---|---|---|---|
Content (wt%) | 37.32 | <0.005 | <0.005 | <0.005 | <0.005 | 0.0259 | 0.0114 | 0.04 | Bal. |
Fig. 1 displays the EPMA maps for the electrodeposited Invar alloy film, showing a uniform distribution of C within the matrix. This uniform distribution is attributed to the presence of C as an interstitial atom in the Fe–Ni alloy, typically embedded in the alloy lattice in solid solution form. However, localized regions exhibit enrichment of C, coinciding with impurities present in the alloy film. This suggests that certain electrolyte residues adhere to the surface, possibly due to inadequate sample cleaning. Sulfur is also uniformly distributed throughout the alloy film. Considering the very low solubility of sulfur in Fe–Ni alloys, its occurrence in the films is attributed to co-deposition during the electrodeposition process.25
Fig. 2a displays the in situ variable-temperature XRD pattern of the electrodeposited Invar alloy film. The as-deposited sample consists of a mixture of BCC and FCC phases, which is consistent with the research results reported in the literature.12 In contrast to pyrometallurgically produced Invar alloy, which displays only an FCC phase, the electrodeposited alloy exhibits both BCC and FCC phases. According to the theory of preferential growth on surfaces in coatings or films, planes with the highest surface free energy grow fastest to minimize energy. On the crystal surface parallel to the substrate, the plane with the fastest growth rate determines the preferred orientation of the coating.26 Factors such as current intensity influence the growth rate of different crystal planes during electrodeposition, altering the crystal orientation of the film, thus explaining the presence of the BCC phase in the electrodeposited Invar alloy film. The FCC phase diffraction peak of the plated Invar alloy shifts towards higher angles, indicating severe lattice distortion and a decrease in lattice parameters. This slight shift in the diffraction peaks towards higher angles is attributed to internal stress within the coating, which arises from hydrogen permeation and island aggregation during the electrodeposition process.27–29 Concurrently, C atoms, which exist as interstitial atoms, contribute to the peak shift.30 This indicates the presence of residual stress or other elements causing lattice distortion in the as-deposited Invar alloy film.
![]() | ||
Fig. 2 (a) In situ variable-temperature XRD patterns of the electrodeposited Invar alloy films and (b) hydrogen annealing process. |
With increasing annealing temperature, the intensity of the BCC phase diffraction peaks gradually decreases, while the FCC phase diffraction peaks increase. At 773 K, the BCC phase completely disappears, leaving only FCC diffraction peaks, indicating a phase transition occurring between 723 K and 773 K. Furthermore, the FCC phase diffraction peaks gradually shift to lower angles, indicating a reduction in internal stress and a decrease in interstitial carbon atoms.31,32 In the Invar alloy, the BCC phase is metastable, resulting in a high CTE. Therefore, a hydrogen annealing process was designed based on the determined phase transition temperature range, as depicted in Fig. 2b.
![]() | ||
Fig. 3 TEM and SEM images of the electrodeposited Invar alloy films: (a) as-deposited, and (b)–(f) after hydrogen annealing. |
The composition distribution across the cross-section of the electrodeposited Invar alloy film before and after annealing at various temperatures was determined using EDS line scanning, as depicted in Fig. 4. The direction of the line scan is indicated by arrows in Fig. 4a. Significant non-uniformity in the component distribution along the thickness direction was observed in the cross-section of the as-deposited sample. The results presented in Table 2 were obtained through ICP chemical composition analysis, which requires complete dissolution of the sample during detection. This method enables the determination of the total Fe and Ni content in the alloy film. However, the actual composition distribution on the cross-section of the electrodeposited Invar alloy film was found to be non-uniform, with many regions deviating noticeably from the typical Invar composition range. This non-uniformity may affect the thermal expansion property of the alloy film.
![]() | ||
Fig. 4 As-deposited cross-section: (a) SEM image; (b) compositional distribution; (c) SEM image of cross-section after corrosion; (d) compositional distribution after annealing. |
After corrosion, the electrodeposited Invar alloy film exhibited a laminar structure along the thickness direction, characterized by uneven interlayer spacing and layer widths of approximately 50–100 nm, as depicted in Fig. 4c. The gaps between these laminar structures suggest the presence of structural defects and incomplete density in the electrodeposited Invar alloy film. Periodic stirring during electrodeposition resulted in fluctuating hydrogen permeation into the electrodeposited layer, explaining the irregularities in these structures.33 The thickness of the diffusion layer during electrodeposition may also influence the structural density. Studies have demonstrated that a high Fe content can be achieved under high current density conditions. A higher current density is required for the electrodeposition of Invar alloys. However, higher current densities required for the electrodeposition of Invar alloy film may lead to particle aggregation on the coating surface, thereby affecting the coating density. Additionally, large-sized samples may result in uneven current density at the cathode and irregular electrolyte stirring during preparation, leading to non-uniform distribution of components in the deposition layer.34 After annealing, grain growth in the film reached the micrometer level, leading to film densification. Moreover, with increasing annealing temperature the cross-sectional distribution of the alloy film gradually became uniform.
Theoretically, ultra-thin films prepared via electrodeposition are devoid of inclusions, offering a novel approach to fabricating near-zero inclusion Invar alloys. Summarizing findings from other research papers reveals that C influences thermal expansion properties. Metal sulfides have been identified in electrodeposited Ni,36 Fe–Ni alloys,37 and Co–Fe alloys,38 and they precipitate and coarsen during annealing. The presence of inclusions in the Invar alloy films not only affects the yield of FMM but also influences the thermal expansion properties. Fig. 6 illustrates the changes in C and S content in the electrodeposited Invar alloy film before and after hydrogen annealing. After hydrogen annealing, the C and S content in the sample significantly decreases, with the C content reducing to approximately 50 ppm and the S content dropping below 10 ppm. C and S inclusions are effectively eliminated, facilitating the preparation of Invar alloy film with minimal inclusions. C atoms usually exist as interstitial solid solution atoms due to their small atomic size, while S exists mainly as metal sulfides. The equilibrium phase diagram of the Fe–Ni–S system at 973 K–1073 K has been reported in the literature.39 It can be seen that the alloy consists of a Fe–Ni alloy matrix and a Fe–Ni–S solid solution at an Fe content of 37.32%. This is attributed to the strong chemical affinity between S and Fe. During electrodeposition, due to changes in process parameters, S can combine with Fe and Ni to form Fe–Ni–S sulfides. Therefore, during hydrogen annealing, C and S react with hydrogen as follows:
C + 2H2 → CH4 | (2) |
Fe–Ni–S + H2 → Fe–Ni + H2S | (3) |
![]() | ||
Fig. 6 Changes in C and S content in the electrodeposited Invar alloy film before and after hydrogen annealing: (a) C content; (b) S content. |
To confirm the presence of S in the form of metal sulfide within the electrodeposited Invar alloy film, the alloy underwent vacuum heat treatment at 1073 K for 1 h, followed by cooling in the furnace. Fig. 7 presents SEM images of the electrodeposited Invar alloy film after vacuum annealing at 1073 K for 1 h, revealing the presence of granular inclusions with a size of approximately 2 μm. EDS analysis reveals that the inclusions are Fe-rich Fe–Ni–S sulfides.
![]() | ||
Fig. 7 Electrodeposited Invar alloy film 1073 K vacuum annealing for 1 h: (a) SEM image; (b) and (c) EDS point scanning. |
Selected regions of the Invar alloy film were subjected to automated continuous slicing and SEM imaging using FIB, followed by 3D reconstruction using Avizo software, exploiting the contrast between inclusions and the film substrate. The results are presented in Fig. 8. The image clearly illustrates the electrodeposited Invar alloy film demonstrates purity and minimal inclusions, underscoring the capability of electrodeposition techniques to produce Invar alloy films with near-zero inclusions.
Fig. 9 shows the trend of CTE and average grain size with hydrogen annealing temperature. The CTE of the Invar alloy film prepared via electrodeposition is approximately 9.0 × 10−6 K−1 within the temperature range of 293–393 K. As the annealing temperature increases, the CTE decreases and the average grain size increases to the micron level. After hydrogen annealing at 773–1073 K, the average grain from 0.5 μm to 3.3 μm, the CTE decreases to the range of 4.0 × 10−6 K−1 to 1.0 × 10−6 K−1 within a temperature range of 293–393 K. There exists an inverse relationship between CTE and average grain size. After annealing at 1073 K for 1 h, the CTE of the alloy film is approximately 1.0 × 10−6 K−1, which is comparable to the rolled Invar alloy film and meets the requirements for preparing FMMs.
![]() | ||
Fig. 9 Linear expansion curves for: (a) as-deposited; (b) after annealing; electrodeposited Invar alloy film after annealing: (c) CTE; (d) average grain size. |
α = (1 − f)αC + fαGB | (4) |
![]() | (5) |
There is an inverse relationship between grain size and CTE, where larger grain sizes correspond to lower CTEs. The experimental data from Fig. 9 was used to substitute into eqn (5) and to plot the relationship between Δα/αC and 1/d, as shown in Fig. 10. The CTE values of the sample hydrogen annealing at 1073 K was taken as an approximate of αC. The results indicate that the difference in CTEs between grain boundaries and crystallites varies with changes in grain size and decreases as the grain size decreases. Lu and Sui43 observed similar results for nanocrystalline Ni–P (Δα/αC ∼ 60%), and Liu10 observed similar results for nanocrystalline Fe-37 ± 2%Ni samples (Δα/αC ∼ 150%). However, in this study, the observed Δα/αC values for electrodeposited Invar alloy film reached 800%, much higher than values reported by other researchers. A larger value of Δα/αC indicates a greater difference between αGB and αC, implying that the CTE difference between grain boundaries and crystallites is more significant in this study. This suggests that the Invar effect primarily originates from crystallites, rather than at the grain boundaries. Hence, enlarging grain size and reducing grain boundary volume fraction become a new effective way to reduce the CTE of metal materials.
Comparing the results of this study with experimental results from the literature. The literature10 indicates that after vacuum heat treatment at 953 K for 12 h, the grain size of electrodeposited Invar alloy film increased to approximately 10 μm, resulting in a decreased CTE of the alloy to 3.13 × 10−6 K−1. However, after hydrogen annealing at 973 K for 1 h in this study, the grain size of electrodeposited Invar alloy film increased to approximately 2.2 μm, and the CTE of the alloy could be reduced to 2.6 × 10−6 K−1. It is analyzed that the presence of C and S inclusions in electrodeposited Invar alloy film affects its thermal expansion properties.
To facilitate comparison, the electrodeposited Invar alloy film was subjected to vacuum heat treatment at 973 K for 1 h. Subsequently, changes in C and S content, CTE, and grain size of the alloys were observed. The results are shown in Fig. 6 and 9. The CTE of the sample annealed in a vacuum environment is approximately 2.8 × 10−6 K−1, which is higher than that of the sample annealed in a hydrogen atmosphere, as shown in Fig. 9c. The grain size of the sample after vacuum annealing is slightly higher than that of the sample after hydrogen annealing, while the CTE is higher than the latter, as shown in Fig. 9d. This is inconsistent with the variation pattern of CTE and grain size. However, the sample annealed in hydrogen shows a significant reduction in C and S impurities, while the C and S content in the sample annealed in a vacuum remains almost unchanged, as shown in Fig. 6. This indicates that the difference in CTE is attributed to changes in C and S content. Hydrogen effectively removes C and S inclusions within the alloy, further reducing the CTE of the alloy.
Additionally, after annealing, elemental diffusion results in a more uniform cross-sectional composition of the alloy film, approaching the Invar composition range (Fig. 4), which facilitates a further reduction in the CTE of the films.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4tc01964d |
This journal is © The Royal Society of Chemistry 2024 |