Abstract
This study investigates the cyclic oxidation behaviour of a thermal barrier coating (TBC) system applied to Inconel 718 alloy substrates. The TBC comprises a NiCr bond coat and a YSZ top coat, applied using the air plasma spray method. Cyclic oxidation studies involve subjecting the coated samples to repeated heating and cooling cycles, with mass gain measurements taken after each cycle. Vibrational modes and surface morphology are analysed before and after oxidation. Results reveal a gradual increase in mass gain up to the 4th cycle, attributed to interface bonding and diffusion-limited growth. Thermal cycling induces stresses, leading to micro cracking and accelerated oxidation. Oxidation kinetics follows both parabolic and linear laws, with rates decreasing over cycles. Kl values consistently exceed Kp, indicating faster linear growth. The study enhances understanding of TBC behaviour under cyclic oxidation conditions, critical for improving durability in practical applications.
Keywords
Introduction
Improving the corrosion resistance of superalloy materials utilised in high-temperature environments is essential to safeguard turbine components, such as compressors and combustion chambers, from erosion during operation. Superalloy materials often face various surface-related challenges. Nickel-based superalloys are renowned for their exceptional strength, toughness, and resistance to deterioration at elevated temperatures, making them well-suited for applications in nuclear power plants, aircraft, defense systems, turbines, and other industries subjected to harsh conditions.1–3 These properties are attributed to their unique microstructure, which consists of a solid solution of gamma (γ) phase and gamma prime (γ′) phase. The γ′ phase is responsible for the high-temperature strength and creep resistance, while the γ phase ensures toughness and ductility. 4 Iron and cobalt compositions are commonly incorporated into superalloys to stabilise their structure by forming gamma (γ) and gamma prime (γ′) phases, with nickel playing a significant role in enhancing this property by promoting the formation and stability of these phases, as well as increasing the solubility of other alloying elements. 5
Inconel 718 (EN8), a well-known superalloy, offers a wide range of material compositions and mechanical strengths, with modifications tailored for specific applications. At elevated operating temperatures, nickel and chromium synergistically shield materials from surface damage, including corrosion, oxidation, and erosion.6–8 Prolonged exposure to high temperatures can result in oxidation of high temperature materials like EN8 superalloys, which may lead to failures and significant consequences. Such oxidation can cause metallic components to reach the end of their service life prematurely. Thermal barrier coatings (TBC) are preferred for protecting metallic parts in high temperature applications because of their excellent thermal insulation properties.9–11 Bond coats are applied to the substrate to improve adhesion and reduce thermal expansion mismatch with the top coat. The top surface layer is made of a ceramic material known for its low thermal conductivity and strong resistance to high temperatures.12–14
Additionally, TBC provide resistance to oxidation and corrosion. TBCs are typically added to superalloys in gas turbine engines with the aim of reducing turbine inlet temperatures and improving the efficiency and longevity of the hot section components.15,16 Various processes are utilised in the production of bond coats, including high-velocity oxy-fuel (HVOF), atmospheric plasma spray (APS), low-pressure plasma spray (LPPS), etc. However, only a limited number of deposition methods, such as APS are available for top coats with high melting points.17,18 Yttria-stabilised zirconia (YSZ) is commonly used in TBC systems due to its low thermal conductivity and high coefficient of thermal expansion, which closely matches that of metals. However, when subjected to temperatures exceeding 1200 °C, YSZ undergoes a phase transformation from the tetragonal phase to a combination of tetragonal and monoclinic phases, leading to an increase in sintering effects. This phase transformation is associated with a volume expansion of approximately 4%, which can lead to material degradation and a decrease in material integrity. The monoclinic phase is less stable and can lead to cracking and spallation of the coating, reducing its effectiveness as a thermal barrier. Therefore, maintaining the tetragonal phase is crucial for the long-term stability and performance of YSZ coatings at high temperatures.19,20
Oxidation represents a critical concern within TBC when subjected to service conditions. These coatings are comprised of top layers designed to shield the underlying substrate from harsh environmental factors, including high temperatures. However, despite their protective role, the top coatings of TBC exhibit characteristics such as an ionic-permeable structure and inherent defects like porosity and voids, originating from the coating production process.21–23 Consequently, oxygen ingress occurs through these pathways, infiltrating the coating system. Upon reaching the underlying bond coat layer, the oxygen initiates oxidation reactions, ultimately compromising the integrity and performance of the TBC system. Thus, understanding and mitigating oxidation mechanisms are crucial for enhancing the durability and longevity of TBCs in practical applications.24,25 There are few research related oxidation behaviour of YSZ based TBC on EN8 superalloy as follows.
Bahamirian et al. 26 devised a layer-by-layer TBC comprising CoNiCrAlY/YSZ/nanostructured ZGYbY on IN738LC superalloy. The layer-by-layer composition of TBCs, like CoNiCrAlY/YSZ/nanostructured ZGYbY, impacts their oxidation resistance by improving mechanical properties, increasing oxidation life, and restricting crack propagation. The use of an intermediate YSZ layer between the bond and top coats enhances mechanical properties and delays TBC destruction. The employment of ZGYbY nanostructured coating significantly improves oxidation resistance compared to YSZ coatings. Additionally, the formation of non-transformable ZrO2 phases and the reduction of thermal conductivity in ZGYbY coatings contribute to their enhanced oxidation resistance. Tong Li et al. developed 27 a TB Coating over a Ni-based alloy, comprising an 8 wt% Y2O3 stabilised ZrO2 (8YSZ) topcoat and either a NiCoCrAlYHfZr or a NiCoCrAlY bond coat. The cyclic oxidation behaviours of the three coatings were evaluated at 1100 °C. The results demonstrated that a uniform thermally grown oxide (TGO) layer with fine inclusions rich in reactive elements was formed in the YSZ-NiCoCrAlYHfZr TBC system, mitigating the adverse effects resulting from the localised enrichment of Hf-rich precipitates. Kadir Mert Doleker and co-worker 28 explores the resilience of NiCr coatings against oxidation in high-temperature settings, highlighting their applicability in TBC at 1000°C. Through the use of APS technique, NiCr powders are successfully deposited on nickel-based superalloy Inconel 718 substrates, followed by YSZ top coatings. The research findings demonstrate the ability of the combined NiCr and YSZ coatings to withstand oxidation, suggesting their potential for cost-effective applications in extreme temperature environments. The same research group also conducted oxidation study on double-layered YSZ/La2Zr2O7 over superalloy and found that double layered coating improves the oxidation stability up to 1150 °C. 29 Pavithran et al. developed 30 YSZ-based TBC over additive manufactured Inconel 718 alloy and investigate its oxidation behaviour. Result revels that YSZ coating applied to the Inconel 718 alloy significantly enhanced its thermal stability while with compared to the uncoated Inconel 718 alloy. Ghosh et al. 31 carried out comparative analysis on oxidation behaviour of micrometre size and nanostructured YSZ coating on Inconel 718 alloy. Results reveals that the nanostructured YSZ coating on the Inconel substrate exhibits notably better resistance to oxidation compared to the micrometre-sized YSZ-coated specimen. Based on the available literature, it is evident that both YSZ and hybrid YSZ coatings enhance the oxidation resistance of Inconel 718 alloy up to 1000°C for continuous durations of up to 24 h. However, practical applications often involve subjecting components in gas turbine engines to cyclic temperature conditions. These components experience fluctuating temperatures during engine operation, including start-up, shutdown, and operational cycles. However, still there is a need for an in-depth understanding of the cyclic oxidation behaviour of YSZ-based TBC. Therefore, this research aims to investigate the cyclic oxidation behaviour of TBCs and comprehend their cyclic failure.
Materials and method
In this research, Inconel 718 alloy substrates measuring 25 × 25 × 5 mm are utilised for coating purposes. Prior to coating, the substrates undergo mirror polishing using varying grades of SiC emery paper. The air plasma spray method is employed to apply a TBC, with NiCr (80/20) serving as the bond coat. The YSZ coating is developed with a spray distance of 200 mm, a current of 500 A, and a powder feed rate of 40 g/min. The resulting coating thickness is 50 µm, with 250 µm for the bond coat and top coat. Cyclic oxidation studies are conducted by subjecting the coated materials to repeated heating and cooling cycles, ranging from 1000°C to room temperature, at 6-h intervals. Mass gain is measured after each cycle, and the process is halted upon detection of coating cracking or detachment. Holmarc made Confocal laser Raman spectrometer worked with 532 nm green laser is employed to identify vibrational modes in coated substrates before and after oxidation studies. Surface morphology analysis of the oxidised samples is performed using SEVO-18 CAREL ZEISS scanning electron microscopy, while phase analysis is conducted using an X-ray diffractometer (BRUKER) machine with 2 theta scan range of 20° to 80°. Surface roughness of the developed coating was analysed using the Zeta 20, a state-of-the-art 3D optical profilometer. This advanced instrument enables precise measurements, capturing of intricate surface features with high resolution. Moreover, the mechanical properties, including hardness and elastic force, were evaluated using the SHIMADZU EZ-XS texture analyser.
Result and discussion
The surface roughness was assessed in three separate places, yielding an average value of 104 µm, as illustrated in Figure 1(a). Complementing these numerical findings, 3D surface image showcased in Figure 1(b) provides a visual representation of the surface roughness variation. Notably, the images revealed a spectrum of roughness levels, spanning from 58.2 µm in the green region to 137.2 µm in the yellow region, underscoring the heterogeneity of the coating's surface topography. In texture analysis, three-point bending method is adopted with a controlled speed of 5 mm/s. The obtained results indicated an average hardness force of 549.48 N, signifying the coating's resistance to indentation and deformation. Furthermore, the elastic forces, calculated as 2.53 N/mm², provide valuable information regarding the material ability to return to its original shape after stress application.

(a) Surface roughness and (b) 3D surface profile of coated surface.
Figure 2 depicts the mass gain in developed coating with respect to cycles. Herein mass gain increases slowly upto 4th cycle and further there is no mass gain for 5th and 6th cycle. The gradual increase in mass gain observed up to the 4th cycle in oxidation studies can be attributed to interface bonding between coating and substrate material at high-temperature environment. Initially, when a material is exposed to elevated temperatures at oxidising atmosphere, it typically undergoes initial oxidation, resulting in the formation of a thin, protective oxide layer on coated surface. This initial oxide layer acts as a barrier, retarding the ingress of oxygen and slowing down the oxidation rate. 32 During initial stage the mass gain is ∼0.08 g that form a barrier against oxidation. As the number of cycles increases, the formed oxide layer tends to thicken, making it more effective at inhibiting further oxidation. Moreover, the growth of oxide layers is often governed by diffusion processes, where oxygen atoms need to diffuse through the formed oxide layer at initial stage to reach the substrate material. This diffusion-limited growth can become a rate-limiting step as the oxide layer thickness, leading to a gradual increase in mass gain. Furthermore, there may be a saturation effect at initial cycle where the available active sites on coating surface become progressively occupied by oxide as the oxidation process continues. As these sites become saturated, the rate of oxidation may slow down.33,34

Oxidation behaviour of developed coating (a) with respect to cycles, (b) oxidation kinetics.
The thermal cycling inherent could also contribute to the observed behaviour. Repeated heating and cooling cycles might induce the thermal stresses in developed coating, which has potential to form micro cracking in the oxide layer which can be visualised from SEM morphology depicted in Figure 3. It can also be attributed to thermal expansion and contraction mismatches between the oxide layer and the underlying material. Additionally, oxidation of the substrate material and the diffusion of oxygen into the coating cause oxidation-induced swelling, exerting mechanical pressure on the YSZ coating and contributing to pore formation as the material tries to accommodate the increased volume. These cracks might temporarily expose to coated surface that leads to accelerated oxidation rates. During testing condition an oxide layer forms on the YSZ surface during the first cycle, the formed oxide layer predominantly consist of zirconium dioxide (ZrO2) and yttrium oxide (Y2O3). This layer acts as a diffusion barrier, slowing the ingress of oxygen from the atmosphere into the substrate. 35 YSZ exceptional ionic conductivity enables oxygen ions to migrate through the coating to maintain the protective oxide layer on the Inconel surface. Oxygen ions diffuse through the YSZ layer to the metal surface, where they participate in oxidation reactions, forming a continuous and protective oxide layer. This layer acts as a barrier, preventing further oxidation of the underlying metal and enhancing its resistance to degradation. The continuous supply of oxygen ions ensures the durability and longevity of the protective oxide layer, maintaining the structural integrity and oxidation resistance of the coated component. Moreover, YSZ's ability to stabilise the zirconia crystal structure prevents phase transformations, enhancing the protective layer's durability. The thermal expansion characteristics of YSZ closely match those of Inconel substrates, minimising the risk of delamination during thermal cycling. Nevertheless, the effects of thermal cycling should not be underestimated, as they can stress the YSZ coating and impact its integrity. Additionally, the presence of yttrium as a reactive element in YSZ allows it to diffuse into the substrate, forming complex oxides at the substrate-coating interface, enhancing adhesion and stability.

Surface morphology of YSZ-coated samples at (a) Cycle 1, (b) Cycle 3, (c) Cycle 4, (d) Cycle 5.
The mass loss was not observed after 4th cycle, it might be owing to formation of a stable and protective oxide layer after initial cycles of oxidation. This oxide layer acts as a barrier, preventing further oxidation of the underlying substrate (Inconel). Once this protective layer is formed, it can limit or prevent additional mass gain. YSZ is an excellent oxidation resistance due to the formation of a stable yttria-rich oxide layer at high temperatures. After several cycles, the oxidation behaviour of the YSZ coating may stabilise, leading to no significant increase in mass. After repeated cycles, the system might reach an oxidation equilibrium where the rate of oxidation is balanced by other processes such as diffusion or the formation of stable oxide layers. At this equilibrium, there may be no net increase in mass despite continued exposure to oxidising conditions.2,36 The phenomenon of increasing oxidation kinetics with time at constant cycle temperature can be observed in Figure 3(a). As materials exposed to oxidising environments, they form oxide layers over the surfaces. Initially, these oxide layers act as barriers, retarding the rate of further oxidation by slowing down the diffusion. The oxide layer typically thickens over time, making it more challenging for oxygen to penetrate and react with the YSZ. Examining Figure 3(a)–(b), it becomes apparent that until the 4th cycle, oxidation abides by the parabolic law, indicating that corrosion primarily results from diffusion growth, involving the migration of outer cations and inner anions. Subsequently, the oxidation process shifts to adhere to a linear law. The kinetics of oxidation where the rate of increase in oxide layer thickness follows a parabolic relationship with time. The linear law is an alternative model that assumes a linear relationship between oxide layer thickness and time.
Both Kp and Kl values decreases as the number of oxidation cycles increases (Table 1). This suggests that the rate of oxide layer growth decreases over time. It could be due the formation of protective oxide layers, depletion of reactants, or changes in the diffusion properties. Changes in diffusion properties, influenced by environmental conditions, can alter the rate at which reactants diffuse to reaction sites, impacting overall reaction kinetics and subsequently decreasing Kp and Kl values. Kl values are consistently larger than Kp values. This indicates that the linear model predicts a faster rate of oxide layer growth compared to the parabolic model. The relative change in Kp and Kl values over cycles can provide insights into the dominant oxidation kinetics. The decrease in Kp is more significant compared to Kl which suggests that the parabolic relationship between oxide growth and time becomes less applicable as the oxidation progresses. The values of Kp decrease from 2.289 × 10−10 to 9.481 × 10−11 over the first six cycles. This decline suggests a reduction in the rate of oxide layer growth, indicating that the oxidation process becomes progressively slower with each cycle. For instance, the rate of oxide layer thickening may decrease by approximately 58.5% from the first to the sixth cycle. The values of Kl also exhibit a similar decreasing trend from 1.127 × 10−7 to 2.962 × 10−8 over the same six cycles. This decline indicates a diminishing rate of oxide layer growth when modelled linearly, with a reduction of approximately 73.7% observed from the first to the sixth cycle. Throughout all cycles, Kl consistently surpasses Kp highlighting the tendency of the linear model to predict a faster oxide layer growth rate compared to the parabolic model.34,37
The phenomenon of increased pore formation with more oxidation cycles in YSZ-coated Inconel is primarily due to several interrelated factors. Firstly, the cyclic heating and cooling during oxidation induce thermal stresses within the coating and substrate, leading to crack formation and subsequent pore development. It can also be attributed to thermal expansion and contraction mismatches between the oxide layer and the underlying material. Additionally, oxidation of the substrate material and the diffusion of oxygen into the coating cause oxidation-induced swelling, exerting mechanical pressure on the YSZ coating and contributing to pore formation as the material tries to accommodate the increased volume. Moreover, the diffusion of gases such as oxygen through the coating can create pressure gradients and promote chemical reactions that degrade the coating, further facilitating pore formation. The cyclic nature of the oxidation process accelerates material degradation mechanisms like grain boundary oxidation and interfacial reactions, exacerbating pore formation over successive cycles. Lastly, as the YSZ coating and Inconel substrate undergo chemical and structural changes due to oxidation and degradation, their integrity weakens, making them more susceptible to pore formation.
X-ray mapping of sample during the 5th cycle is illustrated in Figure 4(a)–(d), revealing significant insights into its composition. The mapping confirms the presence of key elements such as Y and Zr, indicative of YSZ coating. Notably, detection of oxygen, accounting for approximately 35% of composition, unequivocally verifies oxidation of coated samples. Oxygen is visualised in a light green hue, while yttria and zirconium are, respectively, represented by grey and red, facilitating clear identification and characterisation of constituent elements.

(a)–(d) X-ray mapping of sample at 5th cycle: (a) selected region, (b) EDS, (c) mapping, (d) element percentage.
Figure 5(a)–(c) depicts the oxidised samples at various cycle times, revealing a noticeable increase in the oxide layer thickness with each successive cycle. During oxidation, oxygen from the surrounding environment reacts with the substrate material (Inconel) and diffuses through the YSZ coating, resulting in formation of oxides layer. These oxides accumulate gradually over subsequent cycles, thereby augmenting the thickness of the oxide layer. It is noteworthy that the oxidation rate initially increases up to the 4th cycle, after which it stabilises due to the formation of a passivation layer. The cyclic nature of the oxidation process, characterised by alternating heating and cooling cycles, plays a significant role in influencing oxide layer growth. Thermal cycling enhances the diffusion of oxygen and other reactive elements, thereby promoting accelerated oxide layer growth compared to continuous oxidation process.

Cross-section of samples at different cycle: (a) 1st cycle, (b) 3rd cycle, (c) 5th cycle.
The XRD pattern of YSZ coating is taken after 5th cycle and presented in the left panel of Figure 6(a), which is matched with that of the standard ICSD card of Zr0.8Y0.2O1.9 which matches with the literature. 38 Furthermore, the XRD results revealed the formation of a crystalline cubic (Fm–3 m) Zr0.8Y0.2O1.9 phase even after the 5th cycle. The diffraction data shows the presence of only the cubic/tetragonal phase of zirconia, with no monoclinic peaks except for a possible trace at 28° for the 5th cycle sample (Figure 6(a)). The peak overlap for these phases prevented a conclusive determination of whether the samples were cubic or tetragonal. However, no splitting of peaks at (2 2 0) or (4 0 0) was observed, which would be expected if a ≠ c. Also, the peaks that show YSZ crystal phase are wide. These wide peaks may indicate the formation of nanosized particles on the substrate or an amorphous characteristic of the coating. The absence of any strong monoclinic peaks in the XRD data that report mixtures of cubic/tetragonal and monoclinic phases are the most common products. The peak intensities from XRD indicate some preferred orientation for cubic and tetragonal ZrO2 having the highest intensity peak for (1 1 1) planes, and the second most intense is (2 2 0) from the literature, 38 but after the 5th cycle, the planes (1 1 1) and (3 1 1) have a more prominent intensity. The (2 0 0) planes intensity decreased as compared to (3 1 1) plane of YSZ. By comparing the relative intensities of (1 1 1) at 2θ ∼ 30.9°, (220) at 51°, and (3 1 1) at 61° with the samples measured in literature, 39 it appears that growth of the (1 1 1) orientation is suppressed and growth of (2 2 0) and (3 1 1) is enhanced for YSZ after the 5th cycle. The lattice parameter of cubical structure is calculated as follows, a = 5.04164 Å and cell volume is calculated as 128.1489 Å3.

(a) XRD of oxidised at sample 5th cycle, (b) Raman spectrum of oxidised sample at different cycles.
Figure 6(b) shows the Raman spectra obtained for YSZ coating. The Raman data exhibit six distinct peaks of tetragonal ZrO2, which supports XRD data. The Raman data shows the more prominent peaks of tetragonal structure rather than cubic structure or the cubic Raman signal may be coincided with the tetragonal signals. A slight difference in the frequencies of these peaks is observed as the YSZ cycle increases. t/c-ZrO2 consists of two sets of zirconium-oxygen bonds. Raman active modes and the tetragonal symmetry of zirconia has two molecules per unit cell and six Raman active modes. The same kind of results have been reported by other authors, as most of the bands were observed near 250, 450, and 620 cm−1 and used as an indication of the tetragonal structure. The most pronounced peak for all samples is seen at 460 cm−1. It originates from F2g-symmetry Brillouin zone (BZ) centre phonons. The broadening of F2g peaks is explained by contributions from non-BZ centre phonons due to the disorder-induced relaxation of the selection rules. A1g and E2g peaks are also present between 100 and 450 cm−1. All samples have peaks near 170, 250, 330, 390, 450, and 568 cm−1, but the parent compound possesses an extra peak at 516 cm−1, which may be due to more distortion in ZrO2. 40 It can be confirmed that the results from Raman spectroscopy and XRD are in good consistent. The mixed crystal system of tetragonal/cubic with no trace of monoclinic phase is confirmed.
Oxidation kinetics of coated substrate.
Conclusion
The study focused on cyclic oxidation behaviour of YSZ-based TBCs on Inconel 718 alloy substrates revealing the gradual formation of a protective oxide layer during initial cycles. Thermal stresses induced crack formation and pore development impacting the integrity of the coating over successive cycles. Composition and phase analysis confirmed stable crystalline phases in the coating even after multiple oxidation cycles, indicating its resilience. Oxidation kinetics exhibited a transition from parabolic to linear growth behaviour, suggesting a reduction in oxidation rate over time. These insights underscore the importance of understanding oxidation mechanisms to enhance TBC durability in high-temperature environments, prompting further research into optimised coating designs and deposition methods.
Footnotes
Authors’ contribution
R. Prabhu, V. Kavimani, and Jhelai Sahadevan contributed to data curation, methodology, resources; V. Kavimani and R. Prabhu contributed to formal analysis, writing & review; V. Kavimani contributed to project administration, writing – original draft, editing, supervision; P.M. Gopal contributed to editing, review, investigation, software, validation, visualisation. All authors have read and agreed to the published version of the manuscript.
Data availability
The data used to support the findings of this study are included within the article. Further data or information is available from the corresponding author upon request.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethical approval
Not applicable.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
