Etching methods for revealing nanoscale precipitates and carbides in Ni-based superalloys
Seung Gyu Hong, Cho Hyeon Lee, Seong Hyeon Yang, Minyu Kang, Dawon Kang, Hyeong Jin Park, Dae Won Yun, Nokeun Park, Hyun-Uk Hong, Jae Bok Seol
TL;DR
This paper introduces tailored etching methods to reliably reveal nanoscale precipitates and carbides in Ni-based superalloys for accurate microstructural analysis.
Contribution
The study presents composition-specific etching protocols for different Ni-based superalloys to achieve stable and phase-selective SEM imaging.
Findings
Nitric-acid etchant effectively reveals γ′ precipitates in Haynes® 282 and Model alloy 1 but not in Model alloy 2.
An HF-containing etchant provides stable γ′ contrast in Model alloy 2 with finer precipitates.
Chloride-based etchants are necessary to reveal carbides in Nb/Ta-containing model alloys.
Abstract
Gamma prime (γ′) precipitates and grain boundary (GB) carbides govern the high-temperature performance of Ni-based superalloys, and their reliable quantification is essential for microstructural evaluation and alloy development. However, conventional etching procedures are often transferred between alloys without considering composition-dependent changes in γ′ size and fraction or carbide population, which can cause unstable contrast, γ-matrix damage, and unreliable image-based interpretation. Here, we establish composition-tailored etching conditions for Haynes® 282 (γ′ ~23 nm) and two model alloys with modified Al–Ti and Nb–Ta contents, and evaluate their suitability for phase-selective Scanning electron microscopy (SEM) imaging. After identical mechanical preparation, γ′ precipitates in Haynes® 282 and Model alloy 1 are clearly revealed using a nitric-acid etchant, whereas the same…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8- —National Research Foundation of Korea
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsHigh Temperature Alloys and Creep · Advanced Materials Characterization Techniques · Intermetallics and Advanced Alloy Properties
Introduction
Nickel-based superalloys are widely used in applications that require excellent mechanical properties and oxidation resistance at elevated temperatures, such as gas turbines, aircraft engines, and nuclear power systems (Pollock and Tin 2006, Shin et al. 2019, FA et al. 2014, Ramakrishnan and Dinda 2019, Boehlert and Longanbach 2011). The high-temperature properties of these alloys are primarily governed by γ′ precipitates that are distributed within the γ matrix (Tiwari et al. 2024, Xu et al. 2024, Shaikh et al. 2021, Ishtiaq et al. 2025). In particular, γ′ precipitates make significant contributions to strength, creep resistance, and thermal stability, and their size, volume fraction, and spatial distribution are key factors controlling the overall performance of the alloy (Tai et al. 2024, Huang et al. 2025, Ali et al. 2020). Therefore, precise observation and reliable characterization of γ′ precipitates are essential for understanding and optimizing the properties of Ni-based superalloys. However, because γ′ precipitates typically exist on a fine scale of several tens of nanometers, appropriate chemical or electrochemical etching is required to achieve sufficient contrast for microscopic observation. Conventional etching methods often suffer from limitations such as insufficient contrast between γ′ precipitates and the γ matrix, or damage and distortion of the γ matrix due to excessive etching (Salehi et al. 2012). In addition, depending on the extent of etching, precipitates from both surface and subsurface layers may become visible simultaneously, which can lead to misleading interpretation of the γ′ distribution in image-based analyses (Szczotok and Reichel 2020, Smith et al. 2018). For these reasons, the development of improved etching methods is necessary to enable more precise and reliable microstructural observation of γ′ precipitates.
Another microstructural feature that strongly affects the high-temperature properties of Ni-based superalloys is the carbides precipitated along GB (Kim et al. 2025, Ghorbani et al. 2025, Wang et al. 2025). These carbides include coarse MC carbides and fine M₂₃C₆ carbides, both of which play crucial roles in controlling GB strength, creep rupture life, and the susceptibility to intergranular damage (Dong et al. 2012, Hu et al. 2012). Changes in alloy composition modify the amount, morphology, and chemical stability of these carbides, and thus require corresponding adjustments in etching conditions to reveal them effectively (Li et al. 2014, Szczotok et al. 2006). Nevertheless, conventional etching practices have not sufficiently accounted for this compositional dependence, which has limited the clear and consistent observation of GB carbides across different alloy variants (Fang et al. 2024). Therefore, developing etching methods that can distinctly reveal GB carbides while accommodating compositional variations is important for both academic studies and practical alloy evaluation.
In standard metallographic practices and handbooks, etchant selection for Ni-based superalloys is typically categorized based on the target phase to be revealed. (International 1999) For revealing γ′ precipitates, chemical etchants such as Glyceregia or electrolytic etching using phosphoric-acid-based solutions are commonly used, as they preferentially dissolve the γ matrix and thereby expose γ′ morphology (International 1999). Conversely, grain boundaries and carbides, including MC and M₂₃C₆, are typically characterized using reagents such as Adler’s reagent or electrolytic oxalic acid, which selectively target interfacial regions or specific carbide phases (International 1999). While these standardized recipes provide a useful general baseline and have been successfully applied to a range of conventional alloys, they are largely formulated as qualitative guidelines. Crucially, they offer limited insight into how essential parameters—such as etching time, applied voltage, or reagent concentration—should be optimized for compositional variations that alter γ′ size and fraction, carbide distribution, or γ-matrix passivation behavior. Consequently, applying these standard protocols to compositionally modified or advanced Ni-based superalloys can lead to inconsistent phase contrast, excessive matrix dissolution, or etching artifacts, highlighting the critical need for composition-dependent etching strategies.
Several studies have investigated etching procedures for nickel-based alloys. Khodabakhshi et al. examined service-exposed IN738 turbine blades and investigated how different etchant types and etching conditions affect the visibility of microstructural features, particularly γ′ precipitates and related phases, in the root and airfoil regions (Khodabakhshi et al. 2018). More recently, Fang et al. applied various chemical and electrolytic etchants to Ni600 and Ni625 alloys and compared how effectively these etchants reveal GB carbides, Ti-containing phases, and chromium-depleted regions around Cr₂₃C₆ carbides (Fang et al. 2024). These studies provide important practical guidelines for selecting etchants for specific commercial alloys and target phases. Nevertheless, most existing studies have been limited to individual alloy systems or have focused on either γ′ precipitates or GB carbides, rather than treating both phases within a transferable framework that simultaneously addresses phase-selective contrast, γ matrix damage tolerance, and reproducibility across compositionally related Ni-based superalloys. In particular, when the alloy composition is modified, such as through changes in the total Al and Ti content or the addition of strong carbide-forming elements like Nb and Ta, there has been insufficient systematic discussion on how etching conditions effective for a base alloy should be adjusted for compositionally modified γ′-strengthened Ni-based superalloys. Based on previously reported studies, this study proposes composition-dependent etching procedures for Haynes^®^ 282 and its modified variants, separately optimized for γ′ precipitates and GB carbides. The emphasis of this study is on establishing composition-dependent, analysis-ready etching windows that ensure stable phase contrast, minimal γ matrix damage, and high reproducibility across repeated observations.
In this study, we establish composition-dependent, phase-selective etching procedures for Haynes^®^ 282 and two compositionally modified model alloys, separately optimized for γ′ precipitates and GB carbides. Rather than providing a single “best” etchant, we define practical etching conditions that yield (i) stable phase contrast, (ii) minimal γ matrix damage, and (iii) consistent interpretability across repeated observations, enabling robust microstructural assessment. The proposed protocols therefore serve as a transferable metallographic framework that supports not only qualitative phase identification but also reliable downstream image-based analysis (e.g., precipitate size statistics and GB carbide morphology/continuity) in γ′-strengthened Ni-based superalloys.
Method
Materials and heat treatment
Haynes^®^ 282 is a Ni-based superalloy that utilizes γ′ precipitation hardening as its primary strengthening mechanism, providing both excellent high-temperature creep resistance and weldability (Joseph et al. 2017, Muhammad and Park 2025). In this study, Model alloy 1 and Model alloy 2 were designed based on the composition of Haynes^®^ 282. The contents of several alloying elements were adjusted to control the size, fraction, and stability of γ′ precipitates, with the aim of producing distinct microstructural features suitable for systematic evaluation of etching behavior and phase-selective observation. The chemical compositions of Haynes^®^ 282, Model alloy 1, Model alloy 2 are summarized in Table 1.
Table 1. Chemical compositions of Haynes^®^ 282 model alloy 1 and model alloy 2wt%AlTiCoCrMoNbTaNiHaynes^®^ 2821.532.1910.219.58.5< 0.1< 0.01Bal.Model alloy 11.53.8121011Bal.Model alloy 21110.5184.511Bal.
All three alloys were produced by vacuum induction melting (VIM) at a 5 kg scale. The cast ingots were hot rolled with a 60% thickness reduction at 1150 °C, followed by homogenization heat treatment at 1150 °C for 1 h. After homogenization, the specimens were solution-treated at 1150 °C for 1 h and subsequently aged at 700 °C for 16 h to obtain the final microstructures used for metallographic examination.
Metallographic specimen Preparation
For metallographic preparation, samples from Haynes^®^ 282, Model alloy 1 and Model alloy 2 were mounted using a carbon phenolic resin (metallography consumables parts phenolic resin; MTDI, Korea) to ensure stability during grinding and polishing. Mounting was performed using a hot mounting press (ETOS-100; MTDI, Korea) at 180 °C under a pressure of 170 kgf/cm² for 6 min, followed by cooling for 5 min.
Grinding was carried out using a polishing machine (FOBOS-200 S; MTDI, Korea) and SiC abrasive papers of various grit sizes. All grinding steps were performed under a load of 22 N using an automatic polishing head (FOBOS-300; MTDI, Korea). The polishing head and the disc were rotated in opposite directions at 150 rpm and 140 rpm, respectively. Standard SiC abrasive papers with grits of #400, #600, #800, and #1200 were sequentially used, with each grinding step performed for 2 min to obtain a flat and uniform surface. Subsequently, automatic fine polishing with diamond suspensions was conducted to further improve the surface quality. The same equipment configuration as in the grinding process was used, with the applied load set to 17 N while maintaining the same rotational conditions. A polishing cloth (DiaMat, Final-POL) compatible with diamond polishing suspensions (3 μm and 1 μm; Struers, Denmark) was employed. Fine polishing with the 3 μm suspension was carried out for 5 min, followed by 1 μm polishing for 10 min. As the final step, an oxide polishing suspension (OPS) was used for 15 min under a load of 17 N, again with the same rotational conditions as in the previous stage. The detailed grinding and polishing conditions are summarized in Table 2.
Table 2. Grinding and Polishing conditions for Haynes^®^ 282, model alloy 1 and model alloy 2GrindingGritPlaten RPMSample RPMLoadTimeModeFluid400150 rpm140 rpm22 N2 minContraWater600150 rpm140 rpm22 N2 minContraWater800150 rpm140 rpm22 N2 minContraWater1200150 rpm140 rpm22 N2 minContraWater Polishing
Cloth
Platen RPM
Sample RPM
Load
Time
Mode
Fluid DiaMat150 rpm140 rpm17 N5 minContraWaterFinal-POL150 rpm140 rpm17 N10 minContraWaterChem-Pol150 rpm140 rpm17 N15 minContraWater
In this study, composition-dependent etching procedures were applied separately for γ′ precipitates and GB carbides. The detailed compositions and etching times of each etchant are described in Sect. 3.1 and 3.2 and summarized in Tables 3 and 4. To ensure that the selected conditions are not case-specific, the etching time for each solution was calibrated using sequential time-variation tests on specimens prepared under identical grinding/polishing conditions. In this study, an “etching window” is defined as the time range in which the target phase (γ′ or carbide) is revealed with high phase contrast while avoiding artifacts that compromise interpretation, such as excessive γ matrix roughening, pitting, or edge blurring. Under-etching typically results in insufficient phase contrast and discontinuous feature delineation, whereas over-etching leads to pronounced surface relief and preferential attack of the γ matrix, which can obscure true phase boundaries. The optimal etching time reported in Tables 3 and 4 was therefore selected within this window by maximizing the visibility of the target phase while maintaining a smooth and stable γ matrix background.
Scanning electron microscopy
After etching, microstructural characterization of γ′ precipitates and carbides was performed using field-emission scanning electron microscopes (SU-8700, Hitachi, Japan; JSM-7610 F, JEOL, Japan). All observations were conducted at an accelerating voltage of 5 kV, a working distance of 8.1 mm, and in secondary electron (SE) mode. γ′ precipitates were examined at magnifications between 60,000 and 100,000, while carbides were analyzed at magnifications between 4,000 and 8,000, depending on particle size and distribution. GB carbides were further characterized by energy-dispersive X-ray spectroscopy (EDS; Ultim Max 65, Oxford Instruments, UK) attached to the scanning electron microscopy (SEM), using an accelerating voltage of 15–20 kV. *The use of a low accelerating voltage (5 kV) for SEM imaging was selected to enhance surface sensitivity and minimize *electron penetration depth, thereby maximizing topographic and phase contrast after chemical etching while suppressing subsurface signal contributions. In contrast, a higher accelerating voltage (15–20 kV) was employed for EDS analysis to ensure sufficient excitation of characteristic X-rays from Nb, Ta, and Cr containing carbides and to improve the reliability of compositional analysis. Representative images were recorded for each etching condition to evaluate phase contrast, morphology, and the presence of etching-induced artifacts (e.g., γ matrix pitting or excessive relief). To verify compatibility with quantitative image analysis, the SEM images were additionally processed using ImageJ with a consistent segmentation workflow, confirming that γ′ precipitates and GB carbides can be clearly separated from the γ matrix. To improve the reproducibility of image-based precipitate quantification under fixed etching and imaging conditions, a minimum of ten SEM fields were analyzed for each alloy under identical imaging and processing parameters, and field-to-field variation of the measured precipitate area fraction and equivalent size was evaluated. This validation step ensures that the proposed etching conditions provide not only qualitative visibility but also a stable basis for image-based quantification. Prior to quantitative image analysis, SEM micrographs were screened using predefined analysis-ready criteria, including surface pitting, surface relief, precipitate–matrix contrast, and segmentation readiness. Only images meeting all criteria were used for subsequent quantitative analysis, with detailed criteria provided in the Additional file 1.
Results and discussion
Etching behavior of γ′ precipitates in Haynes® 282 and model alloys
Haynes^®^ 282 was selected as a benchmark alloy because γ′ revelation using nitric-acid-based etchants has been widely reported and provides a reliable reference for composition-dependent protocol development. After mechanical polishing, etching was performed on Haynes^®^ 282, Model alloy 1 and Model alloy 2. The compositions and etching times of the etchants used to reveal γ′ precipitates are summarized in Table 3. Etchant 1 is a nitric-acid solution composed of 10 mL HNO₃ and 10 mL distilled water. Etchant 2 is a mixed-acid solution prepared by combining 33 mL distilled water, 33 mL acetic acid, 33 mL HNO₃, and 1 mL HF. The optimal etching times summarized in Table 3 were determined through systematic time-variation tests, identifying the specific durations that maximized the contrast between precipitates and the γ matrix without inducing excessive surface pitting. Notably, the width and position of the etching window depend on alloy chemistry through changes in γ′ size/fraction and surface passivation behavior. Accordingly, the time-variation approach provides a practical route to translate an etchant that works for a base alloy into a composition-adjusted condition that remains phase-selective for modified alloys.
Table 3. Chemical compositions and etching times of γ′ revealing solutions used for Haynes^®^ 282 and the model alloysNameTargetEtching solution compositionTime (s)Etchant 1γ′HNO_3_ (10 ml) + H_2_O (10 ml)16Etchant 2γ′H_2_O (33 ml) + Acetic acid (33 ml) + HNO3 (33 ml) + HF (1 ml)240
The γ′ precipitate morphologies in Haynes^®^ 282 and Model alloy 1 etched with Etchant 1 are presented in Fig. 1. In Haynes^®^ 282 and Model alloy 1, a dispersion of nanoscale γ′ precipitates is clearly resolved within the γ matrix. The precipitates appear as dark, nearly spherical particles with clear interfaces, while the surrounding γ matrix exhibits only shallow surface relief without severe pitting or step-like attack. This indicates that, for Haynes^®^ 282 and Model alloy 1, the nitric-acid etchant produces a sufficient difference in dissolution rate between the γ matrix and the γ′ phase and provides stable phase-selective contrast suitable for microstructural analysis. Given the robust γ/γ′ contrast obtained for Haynes^®^ 282 and Model alloy 1 under Etchant 1, further under-/over-etching analysis was not pursued for these alloys.
Fig. 1SEM images of γ′ precipitates in (a, c) Haynes^®^ 282 and (b, d) Model alloy 1 using etchant 1. Panels (a, b) show relatively low magnification images illustrating the overall dispersion of γ′ precipitates, while panels (c, d) present the corresponding high magnification images of the regions indicated by the red boxes
The etched surface of Model alloy 2 after using Etchant 1 under the same polishing and etching conditions is presented in Fig. 2. In this case, discrete γ′ precipitates cannot be distinguished from the γ matrix. The surface exhibits no discrete precipitate-like features at the expected size scale. This result demonstrates that the nitric-acid condition that is effective for Haynes^®^ 282 and Model alloy 1 is inadequate for revealing the γ′ precipitates in Model alloy 2. Systematic time-window evaluation was focused on Model alloy 2, where Etchant 1 failed to produce analysis-ready contrast. To determine an analysis-ready etching window for Model alloy 2, the HF-containing mixed-acid etchant (Etchant 2) was applied in a stepwise manner with 1-min increments. Representative SEM images demonstrating the etching response of Model alloy 2 at 1 min intervals, including the under-etching, optimal etching, and over-etching regimes, are provided in additional file 2.
Fig. 2SEM images of γ′ precipitates in (a, b) Model alloy 2 using etchant 1. Panel (a) shows relatively low magnification images illustrating the etched surface, while panel (b) presents the corresponding high magnification images of the regions indicated by the red boxes
The γ′ microstructure of Model alloy 2 after using Etchant 2 under identical mechanical preparation is presented in Fig. 3. Under this mixed-acid condition, fine dark γ′ precipitates are distinctly distributed within the brighter γ matrix. The γ′ particles are uniformly distributed and their very small size can be recognized across the observed region, while the γ matrix surface remains sufficiently smooth to avoid misinterpretation caused by excessive relief. These observations confirm that a stronger mixed-acid etchant is required to obtain clear γ′ contrast in Model alloy 2. The different etching responses of the three alloys can be rationalized by considering the alloy chemistry and the resulting γ′ characteristics.
Fig. 3SEM images of γ′ precipitates in (a, b) Model alloy 2 using etchant 2. Panel (a) shows relatively low magnification images illustrating the overall dispersion of γ′ precipitates, while panel (b) presents the corresponding high magnification images of the regions indicated by the red boxes
Haynes^®^ 282 and Model alloy 1 contain higher combined aluminum and titanium contents than Model alloy 2, which promotes the formation of relatively coarser γ′ precipitates with a higher volume fraction (Chu et al. 2023, Baldan 2002). These coarser and more abundant γ′ particles generate a contrast against the γ matrix, so that even a relatively mild nitric-acid solution can selectively attack the γ matrix and outline the γ′ phase. In Model alloy 2, the lower aluminium and titanium contents lead to a reduced γ′ volume fraction and much finer precipitates, which weakens the intrinsic contrast between the two phases. At the same time, the strong oxidizing action of the nitric-acid solution stabilizes a chromium-rich passive film on the surface, further masking the γ/γ′ interface (Salehi et al. 2012). The addition of hydrofluoric acid in Etchant 2 promotes dissolution of this passive layer and activates the surface, allowing the fine γ′ precipitates in Model alloy 2 to be selectively revealed (Strehblow et al. 1979). Overall, these results highlight that γ′ etching conditions must be tailored to the alloy composition, especially when aluminum and titanium contents are reduced and the γ′ precipitates become extremely fine.
The effect of HF addition on the etching behavior can be rationalized by considering the interplay between surface passivation and the chemical stability of oxide films. In Ni-based superalloys, the γ matrix, enriched in Cr and other oxide-forming elements, readily develops a thin but chemically stable Cr-rich passive oxide layer in acidic environments (especially in nitric acid) (Morshed-Behbahani and Nasiri 2025). This passive film suppresses the uniform anodic dissolution of the γ matrix, thereby limiting etching sensitivity and phase contrast, particularly when nanoscale γ′ precipitates are present. The introduction of HF alters this behavior by destabilizing the passive oxide film through fluoride complexation and enhanced oxide solubility (Guo et al. 2022). Fluoride ions are known to form soluble complexes with metal cations within surface oxides, effectively promoting the localized breakdown of passivation and facilitating controlled matrix dissolution (Dai et al. 2021). Importantly, the γ′ phase, which is comparatively enriched in Al and Ti and exhibits higher intrinsic chemical stability under these conditions, remains less susceptible to HF-assisted dissolution. Consequently, HF-containing etchants selectively increase the electrochemical activity of the γ matrix relative to the γ′ phase. This mechanism amplifies the phase contrast and enables the clear delineation of nanoscale γ′ precipitates without inducing excessive matrix attack, consistent with the experimental observations in Model alloy 2.
Etching behavior of GB carbides in Haynes® 282 and model alloys
GB carbides are another important microstructural constituent that strongly influences the creep strength and long-term stability of Ni-based superalloys (Dong et al. 2012, Hu et al. 2012). In Haynes^®^ 282, Model alloy 1, and Model alloy 2, carbides precipitate preferentially along GB and can be revealed by appropriate chemical etching. The etchants and etching times used for observing GB carbides are summarized in Table 4. Etchant 1 is the nitric-acid solution already used for γ′ observation, whereas Etchant 3 is a chloride-containing CuCl₂–HCl–ethanol solution.
Table 4. Chemical compositions and etching times of GB carbides revealing solutions used for Haynes^®^ 282 and the model alloysNameTargetEtching solution compositionTime (s)Etchant 1CarbideHNO_3_ (10 ml) + H_2_O (10 ml)16Etchant 3CarbideCuCl_2_ (2.5 g) + HCl (50 ml) + Ethanol (50 ml)1
Representative etched GB of Haynes^®^ 282 obtained with Etchant 1 is presented in Fig. 4. In Haynes^®^ 282, the GB network is clearly delineated as nearly continuous chains of fine carbides. The carbides appear as small dark particles decorating the GB network and the triple junctions, while the surrounding γ matrix exhibits only moderate surface relief without severe pitting. Considering that the Nb and Ta contents of Haynes^®^ 282 are very low, these GB carbides are interpreted to be predominantly Cr-rich M₂₃C₆, with only a minor contribution from MC carbides (Boehlert and Longanbach 2011, Joseph et al. 2017, Pike 2008). The nitric-acid etchant therefore provides sufficient contrast to trace the GB morphology in Haynes^®^ 282.
Fig. 4GB carbides in Haynes^®^ 282 etched using Etchant 1. Fine Cr-rich M₂₃C₆ carbides decorate GB and triple junctions
In contrast, the etched GB for Model alloy 1 and Model alloy 2 using Etchant 1 under the same polishing and etching conditions is presented in Fig. 5. In both alloys, the GB is only weakly visible and large portions of the GB network remain indistinct. Distinct carbide precipitates decorating the GB cannot be clearly observed, and the overall surface contrast remains flat and featureless. These results indicate that the nitric-acid condition that is effective for Haynes^®^ 282 is inadequate for revealing GB carbides in the two model alloys. Further increasing the time of the nitric-acid-based etchant did not lead to improved delineation of GB carbides in either Model alloy 1 or Model alloy 2, indicating that insufficient selectivity rather than etchant aggressiveness was the limitation.
Fig. 5SEM images of GB regions in (a) Model alloy 1 and (b) Model alloy 2 etched using Etchant 1. The GB and carbides are not clearly revealed, indicating that the nitric-acid etching condition is insufficient for resolving carbide morphologies in these alloys
To obtain clear GB contrast in Model alloy 1, Etchant 3, a CuCl₂–HCl–ethanol solution, was used after identical mechanical preparation. The resulting GB microstructure, together with SEM images, EDS elemental mapping, and line-profile analyses, is presented in Fig. 6. Under this condition, GB carbides are distinctly exposed. Coarse blocky particles along GB are identified as MC carbides enriched in Nb and Ta, whereas finer particles decorating the GB are identified as Cr-rich M₂₃C₆ carbides. These identifications are supported by the elemental mapping and the concentration profiles measured across representative particles in Fig. 6. The chloride-based etchant produces strong contrast between the carbides and the γ matrix while maintaining a relatively smooth γ matrix surface, which enables the simultaneous observation of MC and M₂₃C₆ morphologies along the GB network. This result indicates that the selection of an appropriate etchant must be adjusted according to alloy composition to achieve reliable GB carbides revelation.
Fig. 6GB carbides in Model alloy 1 using Etchant 3. **a **SEM image and EDS mapping showing coarse MC and fine M₂₃C₆ carbides along the GB. b, **c **EDS line profiles identifying (b) Cr-rich M₂₃C₆ and (c) Nb/Ta/Ti-rich MC carbides
The GB microstructure of Model alloy 2 etched with Etchant 3 is presented in Fig. 7. GB is clearly traced by a mixture of coarse MC carbides and finer M₂₃C₆ carbides, and the interfaces between carbides and γ matrix are sharply defined. As in Model alloy 1, MC carbides are enriched in Nb and Ta, whereas M₂₃C₆ carbides are enriched in Cr, as confirmed by the corresponding EDS mapping and line profiles. This etching condition prevents deep groove formation along the GB and preserves a surface state suitable for detailed microstructural evaluation.
Fig. 7GB carbides in Model alloy 2 using Etchant 3. **a **SEM image and EDS mapping revealing the distribution of MC and M₂₃C₆ carbides. b, **c **EDS line profiles confirming the presence of (b) Cr-rich M₂₃C₆ and (c) Nb/Ta/Ti-rich MC carbides
While this study emphasizes phase-selective etching to obtain clear, analysis-ready contrast for specific constituents (e.g., nanoscale γ′ or specific carbide types), practical metallography often employs objective-dependent etching, in which a single etchant intentionally reveals multiple phases (International 1999). Indeed, the chloride-based Etchant 3 used in this study provides concurrent contrast for both grain-boundary carbides and γ′ precipitates, which is advantageous for rapid microstructural screening and for correlating precipitate distributions with grain boundary networks. However, multi-phase revelation typically involves a trade-off between contrast hierarchy and quantitative fidelity, as conditions optimized for one constituent may increase matrix attack or complicate reproducible quantification of another (Salehi et al. 2012). In this context, our composition-dependent strategy does not reject multi-phase etching, but rather provides a framework to rationally tune the effective etching window according to alloy chemistry, whether the objective is selective enhancement of a single phase or balanced, simultaneous revelation of multiple constituents.
To determine the optimal etching condition for GB carbides, the CuCl₂-containing etchant was applied in a stepwise manner with 0.5 s intervals. Based on this systematic evaluation, an etching time of 1 s was selected as the optimal condition and is summarized as Etchant 3 in Table 4. Representative SEM images illustrating under-etching, optimal etching, and over-etching regimes are provided in the Additional file 3.
The different etching responses of the three alloys can be explained by considering the alloy composition and the resulting carbide populations. In Haynes^®^ 282, GB is decorated mainly by fine Cr-rich M₂₃C₆ carbides, for which a relatively mild nitric-acid solution is sufficient to delineate the GB network. In contrast, in Model alloy 1 and Model alloy 2, the addition of Nb and Ta promotes the more frequent precipitation of Nb/Ta/Ti-rich MC carbides together with M₂₃C₆ along GB. This change in carbide population alters the local corrosion response and reduces the effectiveness of the nitric-acid etchant.
In addition, the γ matrix compositions of the two model alloys differ significantly, particularly in terms of Co content as well as Cr and Mo concentrations. These compositional differences influence the dissolution behavior of the γ matrix in chloride-containing CuCl₂–HCl–ethanol solutions, resulting in a narrow and composition-sensitive etching window for reliable carbide exposure. Accordingly, the use of chloride-based etchants with carefully controlled, alloy-specific conditions is required to simultaneously distinguish MC and M₂₃C₆ carbides while maintaining an acceptably smooth γ matrix surface. Overall, these results highlight that GB carbide etching conditions must be tailored to alloy composition and carbide type, rather than being directly transferred from one Ni-based superalloy to another.
The effectiveness of chloride-containing etchants for revealing MC carbides can be rationalized by considering the interfacial electrochemical behavior of Nb- and Ta-rich carbide phases. MC carbides exhibit high thermodynamic and chemical stability in oxidizing acidic environments, which limits their selective delineation when nitrate-based or purely oxidizing etchants are employed (Morshed-Behbahani and Nasiri 2025, Yue et al. 2025). Under such conditions, both the carbides and the surrounding γ matrix tend to remain passivated, resulting in weak interfacial contrast. Chloride ions are well known for their ability to locally disrupt passive films and promote localized dissolution processes, such as pitting (Zhang et al. 2018, Parangusan et al. 2021). In the context of Nb/Ta-rich MC carbides, chloride ions facilitate interfacial attack primarily by destabilizing the passive films at the carbide–matrix interface, thereby promoting localized depassivation and selective dissolution at these chemically and electrochemically heterogeneous regions (Zhang et al. 2018, Alexiadis and Fuchs 2025). This chloride-assisted depassivation preferentially initiates at chemically and structurally heterogeneous regions, such as the MC–γ interface, where local electrochemical potential differences are enhanced (Zhang et al. 2018, Parangusan et al. 2021). Consequently, chloride-containing etchants promote selective dissolution at the carbide–matrix interface without inducing extensive general corrosion of the surrounding γ matrix. This mechanism enhances the delineation and contrast of MC carbides and provides a fundamental explanation for the superior performance and reproducibility of chloride-based etchants observed in this study.
Quantitative analysis of γ′ precipitates from SEM images
The collected SEM images were processed using ImageJ software for binarization, and representative processed images are presented in Fig. 8. These high-contrast images, obtained through the optimized etching procedures established in Sect. 3.1 and 3.2, served as a robust basis for quantitative microstructural analysis. Based on grayscale thresholding, the γ′ precipitates were segmented from the γ matrix, allowing statistical measurement of the average size and areal fraction of the precipitates. The areal fraction measured from 2D sections was used as an estimate of the γ′ volume fraction under standard stereological assumptions. Particle size was evaluated as an equivalent circular diameter from segmented areas, and at least 10 fields were analyzed per alloy.
Fig. 8. Processed images for quantitative analysis of γ′ precipitates in (a) Haynes^®^ 282, (b) Model alloy 1 and (c) Model alloy 2. Dark particles correspond to γ′ precipitates and the bright background to the γ matrix
The quantitative results are summarized in Table 5. The reference alloy, Haynes^®^ 282, exhibited a volume fraction of approximately 12.29% and an average particle size of 23.4 nm. These values are in good agreement with the γ′ fraction and size ranges typically reported in the literature for standard heat-treated Haynes^®^ 282 (Khodabakhshi et al. 2018, Joseph et al. 2017), thereby validating the reliability of the etching and image processing protocols employed in this study.
Table 5. The quantitative results of Haynes^®^ 282, model alloy 1 and model alloy 2Haynes^®^ 282γ′ volume fraction (%)γ′ average size (nm)12.29 ± 0.6223.4 ± 3.7Model alloy 119.02 ± 0.5834.1 ± 4.6Model alloy 27.99 ± 0.4717.2 ± 2.3
Significant microstructural variations were observed in the model alloys, reflecting the intended compositional modifications. Model alloy 1, designed with higher Al and Ti contents, showed the highest volume fraction of 19.02% and a coarser average particle size of 34.1 nm. This increase is attributed to the higher concentration of γ′-forming elements, which enhances γ matrix supersaturation and the driving force for precipitation during aging. Conversely, Model alloy 2, with reduced Al and Ti contents, exhibited the lowest volume fraction of 7.99% and a fine particle size of 17.2 nm. These results confirm that the proposed etching methodology is sensitive enough to accurately detect and quantify changes in the γ′ precipitate fraction resulting from compositional adjustments.
Implications for microstructural evaluation of Ni-based superalloys
The present results highlight that chemical etching should be regarded not merely as a sample preparation step, but as a critical variable that directly governs the reliability of SEM-based microstructural evaluation in γ′-strengthened Ni-based superalloys. In particular, when precipitate populations approach the nanoscale and multiple carbide types coexist along GB, inappropriate etching conditions can lead to misleading contrast, artificial surface relief, or incomplete phase exposure, thereby compromising both qualitative interpretation and subsequent quantitative analysis.
For γ′ precipitates, the findings demonstrate that etchant suitability is strongly coupled to the intrinsic precipitate population, including size, volume fraction, and interparticle spacing. When γ′ precipitates are relatively coarse and abundant, as in Haynes^®^ 282 and Model alloy 1, mild nitric-acid etching is sufficient to generate stable phase-selective contrast. However, as the γ′ population becomes finer and more densely distributed, as in Model alloy 2, the same etching condition fails to resolve individual precipitates and instead produces ambiguous surface contrast. This indicates that γ′ visibility in SEM images is not solely determined by microscope resolution, but is critically constrained by the interaction between precipitate characteristics and surface activation during etching.
A similar methodological implication applies to GB carbides. In alloys where GB are decorated predominantly by fine Cr-rich M₂₃C₆ carbides, mild etching conditions can adequately delineate the GB network. In contrast, the coexistence of Nb/Ta/Ti-rich MC carbides with M₂₃C₆ introduces a heterogeneous corrosion response along GB, necessitating etchants that provide stronger chemical selectivity. Importantly, the present results show that the effectiveness of such etchants is further modulated by γ matrix composition, which influences passivation and dissolution behavior and results in a narrow, composition-sensitive etching window. As a consequence, reliable GB characterization cannot be achieved through universal etching recipes, but requires deliberate matching between etchant chemistry and the expected carbide population.
From a practical standpoint, these observations have direct consequences for quantitative microstructural analysis. SEM images acquired under inadequately matched etching conditions may artificially exaggerate precipitate size, obscure particle continuity along GB, or introduce contrast artifacts that bias image segmentation. In contrast, the composition-tailored etching strategies established in this study produce stable, analysis-ready microstructures that can be robustly processed using standard image analysis tools, such as ImageJ, for precipitate size statistics and GB decoration assessment. Therefore, the methodology presented here provides a transferable framework for ensuring that SEM-based microstructural evaluation reflects intrinsic alloy characteristics rather than artifacts introduced during sample preparation.
Finally, the reproducibility and analysis-readiness of the proposed etching protocols were assessed to support their practical transferability. It should be noted that the absolute etching time can vary with subtle differences in surface state (final polishing quality), solution freshness, and ambient temperature. Therefore, while Tables 3 and 4 provide practical starting points, we recommend confirming the etching window by brief time-variation tests when the preparation route or laboratory conditions differ. Importantly, the composition-dependent trends and the decision logic presented here remain valid, because they originate from phase chemistry (γ′ size/fraction and passivation behavior) and carbide-forming tendency (Nb/Ta-assisted MC formation). This practical calibration guideline is intended to maximize inter-laboratory reproducibility of metallographic interpretation.
The reproducibility and reliability of the proposed etching protocols were evaluated through repeated metallographic preparation and etching under identical mechanical polishing and chemical etching conditions. For each alloy, the optimized etching conditions consistently produced stable and phase-selective contrast, allowing γ′ precipitates and GB carbides to be clearly distinguished from the γ matrix across different regions of the specimens.
Notably, the etching responses were robust within a practical processing window, such that small variations in etching time did not result in abrupt loss of contrast or excessive surface damage. This robustness is particularly important for routine microstructural evaluation, where minor experimental fluctuations are often unavoidable.
Furthermore, the resulting SEM images exhibited sufficiently uniform contrast and limited surface relief to enable reliable binarization and quantitative image analysis using standard image-processing tools. These observations indicate that the established etching strategies provide not only visually clear microstructures but also reproducible and analysis-ready datasets, supporting their use in comparative studies and systematic characterization of γ′ precipitates and GB carbides in Ni-based superalloys.
Although this study focuses on conventionally processed polycrystalline Ni-based superalloys, the underlying chemical principles governing the proposed etching protocols are expected to remain relevant to other microstructural morphologies, such as directionally solidified (DS) and single-crystal (SX) superalloys. However, distinctive microstructural characteristics—including the absence of transverse grain boundaries in SX alloys and the strong crystallographic alignment in DS materials—may influence both the local etching response and the interpretation of contrast. In such cases, etching outcomes may be significantly modulated by crystallographic anisotropy rather than the grain-boundary-assisted attack typically observed in polycrystalline aggregates (Liu et al. 2022). Accordingly, while the etching chemistry established here provides a robust baseline, etching parameters and contrast interpretation should be adapted with careful consideration of microstructural morphology and orientation effects when applying these protocols to DS or SX superalloys (Szczotok and Reichel 2020).
In this study, microstructural analysis was primarily conducted using secondary electron (SE) imaging of chemically etched surfaces. SE imaging provides high spatial resolution and strong sensitivity to etching-induced surface topography, which is particularly advantageous for resolving and quantitatively analyzing nanoscale γ′ precipitates(Egerton and Zhu 2022, Wang et al. 2024).
By contrast, backscattered electron (BSE) imaging generates contrast mainly from atomic number differences and is less sensitive to surface relief (Wang et al. 2024, Timischl and Inoue 2018). While BSE imaging is well suited for robust phase identification of compositionally distinct constituents such as Nb- or Ta-rich carbides, SE imaging was intentionally selected here to maximize the morphological definition required for γ′ precipitate size statistics. This distinction clarifies that the reliance on SE imaging in this study reflects the primary objective of evaluating etching-induced delineation and analysis-readiness, rather than phase identification alone.
It is also worth noting that electropolishing followed by backscattered electron (BSE) imaging is a well-established approach for revealing γ/γ′ microstructures, as it leverages atomic number (Z) contrast to distinguish phases while minimizing surface relief artifacts (Tucker et al. 2026, Murr et al. 2011). While such techniques are highly effective for flat-surface phase discrimination, the chemical etching strategies proposed in this study should be viewed as complementary rather than exclusive. They offer a more accessible and flexible alternative for laboratories where electropolishing facilities or optimized BSE conditions may not be readily available. Furthermore, chemical etching provides the distinct advantage of simultaneously delineating the topographic morphology of grain-boundary carbides, which is essential for assessing particle continuity and morphology but can be less distinct on perfectly flat, electropolished surfaces.
Conclusions
In this study, composition-dependent etching procedures were developed to enable clear SEM observation of γ′ precipitates and GB carbides in a reference Ni-based superalloy, Haynes^®^ 282, and two compositionally modified model alloys.
After identical mechanical polishing, γ′ precipitates in Haynes^®^ 282 and Model alloy 1 were successfully revealed using a simple nitric-acid solution, whereas the same condition failed to generate sufficient γ/γ′ contrast in Model alloy 2. This difference is attributed to the lower Al and Ti contents in Model alloy 2, which lead to a reduced γ′ volume fraction and much finer precipitates. By introducing a stronger mixed-acid etchant containing acetic acid, HNO₃, and a small amount of HF, nanoscale γ′ precipitates in Model alloy 2 were clearly exposed while avoiding severe surface roughening. These results demonstrate that γ′-selective etching must be tailored to alloy chemistry, particularly to changes in Al and Ti that control γ′ size and fraction.
GB carbides also exhibited clear composition-dependent etching behavior. In Haynes^®^ 282, GB decorated mainly by fine Cr-rich M₂₃C₆ carbides, with only sparse MC carbides, was distinctly delineated using the same nitric-acid etchant that was effective for γ′ observation. In contrast, Model alloy 1 and Model alloy 2, which contain higher Nb and Ta and therefore form more abundant Nb/Ta-rich MC carbides together with M₂₃C₆ along GB, required chloride-containing CuCl₂–HCl–ethanol etchants to obtain stable boundary contrast. Under optimized conditions, both MC and M₂₃C₆ carbides were clearly revealed along the GB network.
Overall, the etching strategies established in this study provide a practical framework for phase-selective SEM observation of γ′ precipitates and GB carbides in Ni-based superalloys. The resulting microstructures exhibit sufficient contrast and surface quality to serve as a reliable basis for both qualitative assessment and subsequent quantitative image analysis. When combined with complementary high-resolution information obtainable from TEM, the proposed SEM-oriented etching procedures can support more accurate microstructural characterization and can be adapted to other Ni-based superalloys by considering their Al–Ti levels and carbide-forming elements such as Nb and Ta, as well as the γ matrix composition.
Supplementary Information
Supplementary Material 1.
Supplementary Material 2.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1M. Ishtiaq, J.E. Jung, H.J. Bae et al., Observation of interface disruption and Lomer-Cottrell locks in a crept L 12-strengthened Ni-based Superalloy. Mater. Sci. Eng. A 148570 (2025). 10.1016/j.msea.2025.148570