**Recent Advances in Corrosion Science**

Special Issue Editor **Jacek Ryl**

MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade • Manchester • Tokyo • Cluj • Tianjin

*Special Issue Editor* Jacek Ryl Gdansk University of Technology Poland

*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland

This is a reprint of articles from the Special Issue published online in the open access journal *Materials* (ISSN 1996-1944) (available at: https://www.mdpi.com/journal/materials/special issues/ Corrosion Science).

For citation purposes, cite each article independently as indicated on the article page online and as indicated below:

LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Article Number*, Page Range.

**ISBN 978-3-03936-176-2 (Hbk) ISBN 978-3-03936-177-9 (PDF)**

c 2020 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.

The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.

## **Contents**



## **About the Special Issue Editor**

**Jacek Ryl** is an Associate Professor in the Faculty of Chemistry, Gdansk University of Technology. He has received a Ph.D. (2010) and habilitation (2018) in Chemical Technology at GUT. His principal area of scientific activity lies in applied electrochemistry and surface physic-chemistry, in particular corrosion science, electrochemical sensors, and materials for energy storage and conversion. He is involved in the development of instantaneous impedance techniques with multisine perturbation, dedicated to the assessment of nonstationary processes. Jacek Ryl is coauthor of over 90 peer-reviewed papers (h-index=21). In 2015–2020, he was the principal investigator of programs Sonata NCN and Iuventus Plus. His most notable awards include a scientific scholarship by the Minister of Science and Higher Education of Poland (2017) and the IV Division (Engineering Sciences) of the Polish Academy of Sciences award for scientific achievements (2019). Currently, he serves on the Editorial Board for Molecules, MDPI.

### *Editorial* **Special Issue: Recent Advances in Corrosion Science**

#### **Jacek Ryl**

Department of Electrochemistry, Corrosion and Materials Engineering, Gdansk University of Technology, Narutowicza 11/12, 80-233 Gdansk, Poland; jacek.ryl@pg.edu.pl

Received: 13 April 2020; Accepted: 17 April 2020; Published: 19 April 2020

The International Union of Pure and Applied Chemistry (IUPAC) and European Federation of Corrosion (EFC) define corrosion as an irreversible interfacial reaction of a material with its environment which results in its consumption or dissolution, often resulting in effects detrimental to the usage of the material considered. Corrosion failure is a significant problem in any given type of industry, leading to substantial economic consequences, but also often influencing human health and the environment negatively, among other unmeasurable factors. The industry estimates indicate that the total direct cost of corrosion ranges between 3% and 5% of GDP [1], while the indirect costs (outages, delays, revenue losses, etc.) while much harder to evaluate, are estimated to be equal to this. These numbers point out that investments in corrosion protection are, by all means, economically justified.

The dynamic development of the global industry and growing demand for new material technologies generates constantly increasing problems regarding premature material degradation and the requirement to determine corrosion mechanisms and to develop new protection/evaluation approaches. This Special Issue, "Recent Advances in Corrosion Science", brings together fourteen articles and one review, providing a snapshot of the recent activity and development in this field.

The corrosion properties of ferrous metals remain the most popular subject of investigation, which naturally found coverage in numerous research articles present within this Special Issue. The primary source of this versatility is achieved by a proper selection of alloying additives and metalworking, which guarantee the demanded mechanical and physicochemical properties. On the other hand, the alteration of metal structure leads to the formation of galvanic microcells, often translating into various forms of local corrosion. The search for alloying additives enhancing the corrosion resistance without sacrificing the desired characteristics continues, intending to reduce alloy corrosion rate and bring measurable economic profits. Within this Special Issue, you will find multiple original research papers strictly devoted to this issue for both ferrous [2–4] and non-ferrous metals [5–8]. The influence of novel microscopy tools, which enable the direct observation of local corrosion processes, cannot be overestimated. For this reason, I would like to recommend a very interesting and important review prepared by Chen et al. [9], referring to the advances in electrochemical atomic force microscopy (EC-AFM), an outstanding tool to perform real-time in situ corrosion studies of galvanic microcells.

Affecting the corrosion process by electrochemical protection (cathodic or anodic), barrier properties obtained with the use of paints or coatings as well as environment modification with dedicated corrosion inhibitors, are the three primary ways to reduce the corrosion rate found in both principle and industrial studies regarding anti-corrosion technologies. All of these research areas are represented within this Special Issue. The works of Xu et al. [10], Tang. et al. [11] and Ryl et al. [12] reveal various aspects concerning the search for efficient organic corrosion inhibitors and the tools used to evaluate protection mechanisms. The studies of Parchoviansky et al. [13] and Winiarski et al. [14] provide an insight on the development of anti-corrosion composite coatings, while an interesting report from Kania and Sipa [15] shows the improved corrosion resistance of anodic zinc coatings, obtained using a new thermal diffusion process.

It is important to emphasize that, nowadays, corrosion issues are not solely connected with the degradation of metals. Modern composite or semiconductor electrode materials are constantly developed to be used in numerous branches of applied electrochemistry, such as energy storage and conversion, electrochemical sensors and electrocatalytic processes. Their stable performance under aggressive environmental factors is often questionable. Thus, the final manuscript of this Special Issue presents work in this new field, which was devoted to high-temperature oxidation and the degradation of boron-doped diamond nanostructures [16].

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Review* **Application of Electrochemical Atomic Force Microscopy (EC-AFM) in the Corrosion Study of Metallic Materials**

#### **Hanbing Chen 1, Zhenbo Qin 2, Meifeng He 1,\*, Yichun Liu <sup>3</sup> and Zhong Wu 2,\***


Received: 27 December 2019; Accepted: 26 January 2020; Published: 3 February 2020

**Abstract:** Electrochemical atomic force microscopy (EC-AFM), a branch of a scanning probe microscopy (SPM), can image substrate topography with high resolution. Since its inception, it was extended to a wide range of research areas through continuous improvement. The presence of an electrolytic cell and a potentiostat makes it possible to observe the topographical changes of the sample surface in real time. EC-AFM is used in in situ corrosion research because the samples are not required to be electrically conductive. It is widely used in passive film properties, surface dissolution, early-stage corrosion initiation, inhibitor efficiency, and many other branches of corrosion science. This review provides the research progress of EC-AFM and summarizes the extensive applications and investigations using EC-AFM in corrosion science.

**Keywords:** EC-AFM; corrosion; metallic materials

#### **1. Introduction**

As is known, all materials have a certain service life and will suffer various forms of direct or indirect damage during use. Although the material will be damaged in a variety of forms, corrosion is the most important and common form, which is a gradual process and cannot be restored. The problem of material corrosion occurs in various fields of the national economy, from daily life to industrial and agricultural production, as well as from advanced science and technology to the development of the national defense industry. Corrosion can lead to huge economic losses and even catastrophic accidents, which not only consume a large amount of resources and energy but also pollute the environment, resulting in a huge loss in the gross domestic product (GDP) [1]. Therefore, it is necessary to clarify the failure mechanism of materials in corrosive environments and take reasonable protective measures to achieve the purpose of preventing or controlling corrosion, to improve the life cycle of materials.

Corrosion is the irreparable damage or deterioration of materials caused by chemical, electrochemical, and physical effects of environmental media. In processes of metallic material corrosion, a chemical or electrochemical multiphase reaction occurs on the surface or interface of the metal, resulting in the conversion of the metal to an oxidized (ionic) state. Corrosion of metallic substance is started via the oxidation of metals.

$$\mathbf{M} \to \mathbf{M}^{n+} + n\mathbf{e}^-,\tag{1}$$

$$\rm{M} + (x + y)\rm{H}\_{2}\rm{O} \rightarrow \rm{MO}\_{x}\rm{(OH)}\_{y} + (2x + y)\rm{H}^{+} + (2x + y)\rm{e}^{-}.\tag{2}$$

In order to maintain the neutral condition, a counter cathodic reaction of the electrolyte occurs [2].

$$2\text{H}\_2\text{O} + 2\text{e}^- \rightarrow \text{H}\_2 + 2\text{OH}^-,\tag{3}$$

$$\text{2H}^+ + \text{2e}^- \rightarrow \text{H}\_2.\tag{4}$$

$$\rm{^1O\_2 + 2H\_2O + 4e^- \to 4OH^-},$$

$$\rm O\_2 + 4H^+ + 4e^- \rightarrow 2H\_2O,\tag{6}$$

$$\text{M}^{n+} + n\text{e}^- \rightarrow \text{M}.\tag{7}$$

The basic process of metallic corrosion in an aqueous solution consists of the anodic dissolution of metals and the cathodic reduction of oxidants. The redox reactions (Equations (1)–(7)) involve the transfer of electrons and ions between the metal and the solution. According to the corrosion kinetics, the anodic oxidation current of the metal degradation is equal to the cathodic reduction current of the oxidant at the corrosion potential. When the metal electrode potential is more positive, the rates of cathodic reactions increase and the rates of anodic reactions decrease accordingly. Conversely, as the metal electrode potential becomes more negative, the effect on the reactions is opposite.

The development of corrosion science is inseparable from the advancement of research methods and instruments. Conventional electrochemical measurements can only obtain macroscopic electrochemical information on the surface of the material. In situ electrochemical scanning probe technology with high spatial resolution facilitated the development of corrosion science, elucidating the microstructure and dynamic properties of materials and interfaces at the molecular/atomic level.

The electrochemical atomic force microscope (EC-AFM) was developed in 1991 on the basis of the atomic force microscope (AFM) [3]. It is well known that AFM is a kind of scanning probe microscope (SPM) and an extension of the scanning tunneling microscope (STM) [4]. The STM was invented by Gerd Binnig and Heinrich Rohrer in the 1980s, and it was able to image individual atoms for the first time. In 1982, Binnig and co-workers indicated that vacuum tunneling with an externally controllable tunnel distance is feasible, even under the conditions of room temperature and non-ultra-high vacuum, which was the first step in the development of scanning tunneling microscopy [5]. The 7 × 7 reconstruction on Si(111) observed in real space with STM was considered the first scientific success [6]. G. Binnig and H. Rohrer went into detail about STM, and they were rewarded with the Nobel Prize in Physics for their work in the field of STM in 1986 [7,8]. At present, STM is still an important method in surface science. However, due to the operation principle of STM being based on the quantum tunneling effect, STM has a severe restriction that requires the samples to be conductive. STM can only directly observe the surface structure of conducting and semiconducting materials. In order to overcome the shortcomings of STM, the AFM was invented by Binnig, Quate, and Gerber in 1986 [4]. The AFM can image the surfaces of a flat sample through the weak interaction force (atomic force) between the probe (mounted to a cantilever) and the sample, which is measured by monitoring the deflection of the microcantilever. Therefore, the AFM is suitable for both conductive and non-conductive samples, and its application field is more extensive. In addition, the AFM is a high-resolution microscope with atomic-scale resolution, which developed into a powerful micro/nanometer-scale surface analysis technique [9,10]. On the basis of different capabilities, a large number of techniques were developed after the invention of AFM, such as Kelvin probe force microscopy (KPFM), magnetic force microscopy (MFM), scanning electrochemical microscopy/atomic force microscopy (SECM–AFM), electrostatic force microscopy (EFM), and EC-AFM. These technologies are not only imaging tools; they can also accurately and quantitatively measure local physical and chemical phenomena.

EC-AFM extends AFM technology to the field of electrochemistry for in situ studying of the solid–liquid interface. The main difference between EC-AFM and AFM in liquids is that the applied potential of the sample with respect to the reference electrode is controlled by a potentiostat [11]. Topographic changes of the sample are achieved by measuring the force between the tip and the substrate in a three-electrode electrochemical cell composed of the counter electrode, reference electrode,

and working electrode (substrate). EC-AFM is applied to in situ electrodeposition [12,13], biological science [14], supercapacitors [15], batteries and electrodes [16], corrosion and protection, and so on. After the invention of EC-AFM, its electrolytic cell, imaging mode, and probe performance were continuously improved, which opens a new prospect for the understanding of the corrosion mechanism [17,18].

In this review, we illustrate the application of EC-AFM to metal corrosion with specific examples, including understanding the properties of passive films, surface dissolution, pitting corrosion, selective corrosion, intergranular corrosion, inhibitor efficiency, coating protection, and other branches of corrosion science. The main purpose of this review is to provide the present situation and trend of corrosion research on various metal materials by EC-AFM and outline future prospects for the technique.

#### **2. Principle and Operation Modes of EC-AFM**

The basic working principle of EC-AFM is similar to that of AFM, and it shares its key components. Primarily, a nanoscale sharp tip located at the end of the cantilever of the probe is used to feel the interaction force between the tip and the sample, a piezo scanner is used for controlling the movement of the tip or sample in the *x*-, *y*-, and *z*-directions, a feedback control system and feedback circuit are used to control the deflection of the cantilever, and a computer system is used to display the results and control the operational parameters [19,20]. In addition, EC-AFM requires an electrochemical cell that can accommodate the working electrode (WE), counter electrode (CE), and reference electrode (RE), as well as a potentiostat for normal potentiostat control in a three-electrode system. Generally, the electrochemical cell is made from chemical-resistant polycarbonate, which can be used with a wide variety of liquids. Eight-degree nose assemblies are recommended for imaging in liquid because the smaller angle considers the different angles at which the laser goes in and out of the fluid, compared to operation in air. The electrochemical cell typically contains retaining clips, an O-ring gasket, and a liquid cell plate. When assembled, the sample itself comprises the bottom of the liquid container. Therefore, the sample must be large enough for the O-ring to seat. The WE should be relatively small, and each point on the WE should be geometrically equivalent to the counter electrode CE, which ensures that the current and potential across the electrode are evenly distributed. Others are based on conventional electrochemical testing requirements. Figure 1 shows the diagram of a typical electrochemical AFM cell with a potentiostat. During the scanning, the sample to which the potential is applied is placed in the electrolytic cell as the working electrode and scanned with the tip mounted on the cantilever. The force between the tip and the sample is measured by monitoring the deflection of the cantilever. By plotting the deflection of the cantilever versus its position on the sample, a topographic image of the sample is obtained [21].

The operation mode of EC-AFM is also similar to that of AFM, and the instrument can be operated either in constant height mode or constant force mode. Based on the interaction force between the tip and the sample, it is generally divided into different operation modes in the imaging process, such as contact mode, non-contact mode, intermittent tapping (tapping mode), torsional resonance mode, and peak force tapping mode. The first is the contact mode, in which the tip of the probe is always in contact with the surface of the sample, and the force between the tip and the sample is a repulsive force. Since the silicon nitride cantilever is soft enough to be deflected and has a high resonance frequency to avoid vibration instability, a silicon nitride probe is often used in contact mode. The contact mode has the advantage of high resolution and high scan speed, but it may cause damage to the surface of the sample. Both the transverse shear force and the capillary force on the surface may have adverse effects [22]. The second type is the non-contact mode, in which the cantilever vibrates above the sample surface, and the distance between the tip and the sample is usually several nanometers [23]. The force between the tip and the sample is van der Waals attraction. In the non-contact mode, there is no damage to the surface of the sample and the lateral force is the smallest, but the resolution is low and the scanning speed is slow. In order to avoid being stuck to the water film on the sample surface, it is often used to scan hydrophobic surfaces and soft samples. However, the development

and improvement of this device in recent years enabled non-contact mode to detect repulsion and obtain atomic resolution images. Miyata used a non-contact-mode AFM to observe the dissolution of calcite in water at high speed [24]. This may well be applicable to the observation of corrosion processes on metal surfaces. The third type is the intermittent tapping (tapping mode), in which the cantilever oscillates at a frequency close to its resonance frequency and the oscillation amplitude is monitored [25]. The cantilever oscillates near its resonant frequency above the surface of the sample, and the probe contacts the sample surface once at the bottom of the oscillation during each vibration period. The difference with the non-contact mode is that the amplitude (amplitude > 20 nm) of the cantilever is larger than that of the non-contact mode (amplitude < 10 nm). It not only reduces damage to the sample and reduces the effects of transverse force, but it also has a higher lateral resolution on most samples. The disadvantage is that the scanning speed is limited.

**Figure 1.** The schematic diagram of electrochemical atomic force microscopy (EC-AFM) electrolytic cell.

In addition, in order to further understand the properties of nanomaterials, a torsional resonance mode (TRmode) AFM was developed, which can measure both vertical and lateral force concurrently [26]. In the torsional resonance mode, the cantilever conducts torsional vibration with the long axis as the center and causes the tip to be in the vibration state. When the probe encounters transverse force on the surface, the system can detect the change in the cantilever torsion vibration and detect the fluctuation in the surface topography of sample [27]. It should be noted that the tapping mode applies a compressive force and the TR mode applies a torsional force; thus, the normal and shear performance is measured in the tapping mode and the TR mode, respectively [28]. Another newly developed operation mode is the peak force tapping mode, in which the tip is oscillated periodically in the vertical direction at a frequency well below cantilever resonance, tapping the sample until the maximum repulsion force reaches the setpoint in each tap [29]. The z piezo is driven with sinusoids rather than a triangular waveform in conventional force–distance (F–D) curves [30]. This mode directly controls the interaction between the tip and the sample, reducing the depth of deformation and the corresponding contact area, minimizing damage to the probe or sample [31,32]. Moreover, the force–distance interactions can be measured directly through peak force tapping mode.

Table 1 summarizes the operation mode of EC-AFM with the necessary parameters added. In the field of corrosion, contact mode and intermittent tapping (tapping mode) are mainly used.


**Table 1.** Comparison of common operation modes of EC-AFM.

#### **3. Application of EC-AFM in Metal Corrosion and Corrosion Protection**

#### *3.1. The Micro-Area Corrosion*

#### 3.1.1. The Corrosion Product Film

EC-AFM can characterize the real-time nanoscale topographic changes of a corrosion product film on a metal surface under different potentials in aqueous environments, which avoids the potential contamination and degradation of the corrosion product film during conventional ex situ characterization and analysis techniques.

A typical example of studying corrosion product films with EC-AFM was presented by Li and co-workers, who studied the surface passivation film and its substructure of carbon steel in a carbonate/bicarbonate solution by EC-AFM in contact mode [33]. They designed a solution container for the EC-AFM, which was large enough to accommodate the three-electrode system and avoid evaporation during testing. The topographic characterization of the sample surface was performed at the passivation potentials of −0.1 V(SCE—saturated calomel electrode), 0.5 V(SCE) and 0.7 V(SCE). The results showed that, at −0.1 V(SCE) and 0.5 V(SCE), as the passivation time increased, the nanoscale features, i.e., the scale-like spots on the surface of the sample, increased in diameter, indicating the growth of the passive film. When the grooves caused by the surface treatment were not observed, a thick passivation film was formed on the surface of the sample. The surface roughness calculated from the topographic profile was small and changed little, indicating that the passive film was relatively uniform. When the potential reached 0.7 V(SCE), the roughness increased and the passive film became non-uniform, as shown in Figure 2.

In addition, based on the morphology, they found that the passive films contained a mixture of Fe3O4, Fe2O3, and FeOOH when passivated at the active–passive transition potential. The inner layer of the passive film was Fe3O4 and the outer layer was Fe2O3/FeOOH. When the film-forming potential changed from −0.1 V(SCE) to 0.5 V(SCE), the chemical composition did not change, but the thickness of the inner layer became thicker, leading to an increase in the thickness of the oxide film from about 4.5 nm to 5.8 nm, and the compactness improved, which made the film more protective. On the contrary, when the potential was at 0.7 V(SCE), although the thickness of the film increased, its chemical composition changed greatly and an amorphous structure appeared, which reduced the corrosion resistance of the film.

Padhy et al. [34] investigated the passive film properties of 304L stainless steel in the nitric acid medium in both ex situ and in situ conditions. The ex situ study on the surface morphology of the passive film showed that the surface morphology of 304L stainless steel depended on the stability of the passive film in nitric acid. The morphology features of the passive film in nitric acid medium increased with time and solution concentration. The passive film in 1 M and 4 M nitric acid solution was stable. However, in 8 M and 11.5 M nitric acid, breakdown of the passive film was observed. In situ surface morphology changes of the passive film were monitored by EC-AFM in 0.1 M, 0.5 M, 0.6 M, and 1 M nitric acid. It was found that, at lower concentrations of 0.1 M and 0.5 M nitric acid, the passive film grew in a platelet-like structure. At nitric acid concentrations of 0.6 M and 1 M, platelet-like structures aggregated, homogenized, and began to deplete from the surface, causing selective dissolution. The results of X-ray photoelectron spectroscopy (XPS) analysis showed that the passive film at lower concentration was composed of a hydroxide and oxide layer. However, at high concentrations, the passive film consisted only of an oxide layer. The passivity of 304L stainless steel under the action of low-concentration nitric acid started from the formation of chromium hydroxide, and the surface was in the form of platelet-like structures. As the concentration increased, the hydroxide layer changed to a homogenous oxide layer. As the concentration continued to increase, the protective oxide film was depleted from the structural heterogeneous zone, leading to the opening of the oxidation boundary. The depletion of the oxide layer caused selective dissolution and local corrosion of 304L stainless steel.

**Figure 2.** The AFM images of the steel specimen polarized at −0.1 V(SCE—saturated calomel electrode) for (**a**) 0.5 min and (**b**) 60 min, at 0.5 V(SCE) for (**c**) 0.5 min and (**d**) 60 min, and at 0.7 V(SCE) for (**e**) 0.5 min and (**f**) 60 min. The black arrow points to scale-like spots. Adapted with permission from Reference [33]; copyright 2017 Elsevier Ltd.

Many researchers reported the corrosion process and mechanism of AA 2024 and identified the corrosion products. Kreta et al. [35] explored the changes on the surface during exposure to the sodium chloride environment. In their experiment, the development of the passive film of AA 2024-T3 was analyzed with different analytical tools during exposure to 0.5 M NaCl. In order to track the morphology changes of the protective hydroxide/oxide film, the in situ EC-AFM measurements were carried out on the site without inclusions in a tapping mode. In situ EC-AFM images were taken from samples exposed to different potentials in various time steps. The parameters such as roughness are not only a function of potential, but also a time-dependent function. It can be observed that the surface roughness gradually increased during immersion in sodium chloride solution, and there was a similar increase in surface roughness after the application of external potential, which was mainly due to the formation of an oxyhydroxide layer. In addition, the deposition of the corrosion products led to the formation of a spiky surface, as shown in Figure 3. The profile lines measured over the corroding intermetallic particle are shown in Figure 3e, which allows observing the formation and roughness changes of the corrosion product film. The further increase in potential was most probably responsible for the significant increase in the thickness of the passivation layer. Under a certain potential, the corrosion potential spikes formed but disappeared after the thicker passivating oxyhydroxide layer was formed.

In terms of biomaterials, the inflammatory state of local body fluids in the body may affect the corrosion of implanted metallic biomaterials such as CoCrMo alloys. In the study of Liu and Gilbert, the surface of CoCrMo alloy was imaged by in situ EC-AFM to evaluate the morphology of the oxide film and surface dissolution under different physiologically possible potentials in phosphate buffer saline (PBS) solution and simulated inflammation (SI) solution (PBS solution with 30 mM H2O2), respectively [36]. EC-AFM images in SI solution showed different topographic changes compared with PBS solution alone, in which the alloy matrix was corroded and the carbide protruded. In contrast, a less compact oxide layer appeared and gradually covered the outermost surface in the case of PBS solution. The authors speculated that the addition of H2O2 changed or removed the passive oxide films and greatly affected anodic dissolution on CoCrMo alloy. In addition, the grain boundaries and carbide boundaries were preferential parts for the dissolution of oxides, which was mainly due to effects of Cr depletion, a heterogeneous structure, and Cr distribution along grain boundaries and carbide boundaries.

In the work of Bearinger and co-workers, the effects of different potentials and hydration on the properties and structure of passive oxide films on the surface of titanium, 6-aluminum, 4-vanadium (Ti-6Al-4V) were investigated by in situ EC-AFM [37]. The AFM imaging of the different surfaces showed that all samples were covered with a protective titanium oxide dome. Due to hydration, the dome area gradually increased, and coalescence occurred with the increase in applied voltage and time. In addition, Bearinger et al. [38] characterized hydration of titanium/titanium oxide surfaces under freely corroding and potentiostatically held conditions using EC-AFM. In contrast to conventional high-vacuum techniques, EC-AFM enables the measurement of morphological surface structures in the in situ hydrated state. The results showed that the titanium surface covered the oxidized dome and grew laterally during hydration. The applied potential altered the growth rate. Under open-circuit potential, growth proceeded approximately six times faster than under −1 V applied voltage. The oxide growth was partly due to the lateral expansion and overgrowth of the dome at the oxide–solution interface. The method was successfully used to study dynamic changes in the surface morphology.

**Figure 3.** The in situ AFM images of AA 2024: (**a**) 226 min, −0.4 V/Ag/AgCl/KClsat, 90 s; (**b**) 328 min, 0 V/Ag/AgCl/KClsat, 10 s. (**c**,**d**) The corroding sites on top of the images (**a**,**b**) as three-dimensional (3D) images. (**e**) The profile lines measured over the corroding intermetallic particle. Reprinted with permission from Reference [35]; copyright 2016 Elsevier Ltd.

Metallic glass has excellent mechanical, magnetic, and chemical properties due to its unique structure. It is noted that the corrosion resistance of metallic glass is not only sensitive to its chemical composition but also inseparable from the structure. The influence of the structure on the corrosion of metallic glass is closely related to the dissolution, passivation, and durability of the passive film. Wang et al. [39] selected Ni50Nb50 metallic glass with polymorphic transition during devitrification to clarify the relationship between the amorphous structure and corrosion, especially the stability of the passive film. By comparing amorphous alloys with its crystalline alloys, the breakdown behavior of passive films and characteristics of the surface film were studied in detail. The surface topography of the samples was characterized by in situ EC-AFM in tapping mode. The results confirmed that the surface of both amorphous and crystalline samples could form metastable pits at nanoscale, but the number of nanometer pits on amorphous samples was much smaller than that on crystalline samples. Moreover, amorphous alloys only had a few deep pits and no shallow pits, as shown in Figure 4a,c, while crystalline alloys had many deep pits and shallow pits, as shown in Figure 4b,d. Although the composition of the passive film on the surface of the two alloys was similar, the amorphous structure could significantly inhibit the formation of pits in corrosive environment. Therefore, amorphous alloys had stronger resistance to pit formation than crystalline alloys.

**Figure 4.** AFM imagines of the surfaces of amorphous and crystallized Ni50Nb50 alloys after polarization at 1.0 V in 1 mol/L HCl solution. (**a**,**b**) Surface morphologies of the amorphous and crystallized samples. Inserts are the corresponding I–t curves during potentiostatic treatments. (**c**,**d**) High-magnification images of selective areas in amorphous and crystallized samples. Adapted with permission from Reference [39]; copyright 2010 Elsevier Ltd.

Phase imaging is an extension of the conventional tapping mode AFM (TM-AFM), through which topographic images and phase images can be obtained simultaneously. The phase images can reflect local nanoscale mechanical properties such as friction, hardness, and viscoelasticity. During the AFM scanning process, different nanometer-scale crystalline phases and their size, shape, and distribution can be identified in the matrix according to the phase change. Zhang et al. [40] studied the formation of the passive film in the Al–Ni–Ce metallic glass by in situ EC-AFM. In situ EC-AFM measurements were performed in open-circuit potential (OCP) conditions and stepped the potential into pitting potential (Epit). They identified α-Al nanophase from the amorphous matrix by TM-AFM. It can be clearly seen from the phase image that the small α-Al islands were randomly distributed on the surface, which had a negative phase shift with respect to the matrix. In the partially magnified phase image, the α-Al island is surrounded by bright ringed patches, as shown in Figure 5.

**Figure 5.** In situ tapping mode (TM) AFM images of the annealed Al88Ni8Ce4 amorphous nanocrystalline sample at open-circuit potential (OCP) held for 3 min in 0.01 mol/L NaCl solution. Topography image (**a**) and phase image (**b**). Three-dimensional (3D) images of phase (**c**) and the (**d**) local enlarged phase image of the marked circle in (**c**). Reprinted with permission from Reference [40]; copyright 2014 Elsevier Ltd.

In addition, the nucleation and growth process of the passive film was observed in real time by in-situ EC-AFM, and it was found that the presence of nanocrystalline α-Al precipitates changed the nucleation mechanism of the passive film in the amorphous matrix from instantaneous nucleation to progressive nucleation. Due to the galvanic coupling between the α-Al nanophase (anodic position) and surrounding amorphous matrix (cathodic position), the formation of the corrosion product Al(OH)3 was in the early stage of formation of the passivation film, which was mixed into the passive film, changing the local structure and composition of the passive film. The density of hydroxide in the passive film was lower than that of the oxidation film formed on the amorphous matrix, which reduced the compactness of the passive film and its stability. Therefore, compared with fully metallic glass, metallic glass containing nanometer α-Al precipitation had lower corrosion resistance.

Most previous research focused on steady-state corrosion reactions of steel, such as the study of pipeline corrosion in bicarbonate solutions or carbonate bicarbonate electrolytes. However, some researchers also studied the early stage of steel corrosion. Li et al. discussed the influence of bicarbonate concentration on the topographic evolution and corrosion mechanism [41]. In their work, the early stage of X100 pipeline steel corrosion in bicarbonate solution with different concentrations was characterized by in situ EC-AFM in the contact mode. The relationship between surface roughness and the corrosion process of pipeline steel was established. The pipeline steel was fixed to the bottom of the homemade solution container as the working electrode, which prevented evaporation of the test solution during testing. The authors found that the early stage of the steel corrosion in 0.01 M NaHCO3 solution could be divided into three stages. When the steel was immersed in the solution, the oxides on the surface of the steel began to dissolve, the OCP dropped, and the surface roughness increased rapidly, representing stage I. When the steel was corroded, the OCP further dropped, and the surface roughness increased gradually, representing stage II. In stage III, as the corrosion of steel reached a steady state, the surface roughness and OCP maintained a stable value, and the formation of corrosion products

reached a dynamic equilibrium state. However, in solutions with increased bicarbonate concentration, such as 0.1 M and 0.5 M NaHCO3, steel could be passivated. In addition, as the passive capacity of bicarbonate solution increased with the increase in concentration, the morphology of the steel surface in 0.5 M NaHCO3 solution was smoother than that in the 0.1 M solution, consistent with the results of power spectral density (PSD) analysis. Compared with the passive film formed at a low concentration (0.1 M), the passive film formed in a solution with a high concentration (0.5 M) could eliminate the large-scale morphological features.

In addition, AFM–SECM is considered to be a particularly attractive research tool in corrosion science [42]. The clear topographical information obtained by EC-AFM and the electrochemical information provided by SECM can be obtained in a single experiment, which makes up for the deficiencies of the two instruments. Izquierdo et al. [43] modified AFM tips to achieve a bifunctional AFM–SECM tip. Conductive microelectrodes were integrated into an AFM probe to obtain an SCEM signal, through which the changes in surface morphology and current density were achieved to observe the progress of corrosion. Local release of Cu2<sup>+</sup> ions was monitored by electrochemical reduction and deposition of metal ions on the AFM–SECM probe. The formation and breakdown of passive layers were characterized from the perspective of surface roughness and current density, reflecting the initial growth of the passive layer, subsequent breakdown, and formation of the pit.

In the research on microzone corrosion carried out by our research group, Ding studied the corrosion behavior and the formation of **a** corrosion product film on nickel–aluminum bronze (NAB) in 3.5 wt% NaCl solution. The in situ AFM in contact mode at open-circuit potential was used to observe the formation of the corrosion product film on different phases [44]. Figure 6 shows the in situ AFM topography of the specimen surface.

**Figure 6.** In situ topography images of nickel–aluminum bronze (NAB) specimen surface after exposure to 3.5 wt% NaCl solution at different times: (**a**) initial; (**b**) 30 min; (**c**) 60 min; (**d**) 120 min; (**e**) 150 min; (**f**) 180 min. Site 1 corresponds to the α phase, site 2 corresponds to the α + κIII eutectoid structure, and site 3 corresponds to the κII phase, while site 4 corresponds to the β' phase. Reprinted with permission from Reference [44]; copyright 2019 MDPI.

The α phase exhibited different corrosion behavior at different locations. Serious corrosion occurred at the lamellar α phase within the α + κIII eutectoid structure, while the α phase far away from the κ phase was not eroded, mainly due to the formation of many micro galvanic cells in the eutectoid. The κII and κIII phases exhibited significant corrosion resistance due to the formation of a stable, dense protective film in a short period of time. It is worth noting that the film on the κII phase

was thicker, related to the κII phase based on Fe3Al intermetallic compounds and containing more iron. Compared with nickel oxide, iron oxide is fluffier and more unstable; thus, the film on the κII phase was thicker. Since the metastable martensite structure struggled to form a protective film, the β' phase suffered the most severe corrosion. When the immersion time reached 150 min, the corrosion product deposition was almost dispersed on the entire surface, and the corrosion depth of each phase was suppressed, as shown in Figure 7. The formation of the corrosion product film tended to be uniform, which prevented the NAB alloy matrix from contacting the corrosive medium and hindered the transport of ions and charges, as well as improved the corrosion resistance of the NAB alloy.

**Figure 7.** In situ line profiles corresponding to sites 1, 2, 3, and 4 marked in Figure 6, respectively: (**a**) α phase; (**b**) α + κIII eutectoid structure; (**c**) κII phase; (**d**) β' phase. Reprinted with permission from Reference [44]; copyright 2019 MDPI.

EC-AFM has great advantages in studying the formation process and microstructure of corrosion product films on a metal surface, especially in characterizing the evolution of early corrosion product films, which is conducive to analyzing the corrosion behavior of metals and promoting the research on the corrosion resistance of metal. The application of EC-AFM in corrosion product films is summarized in Table 2.


**Table 2.** Summary of the application of EC-AFM in corrosion product films.

#### 3.1.2. Pitting Corrosion

Pitting corrosion was extensively studied in the past few decades, as it has greater destructive and hidden dangers. Generally, the pitting mechanism is related to the properties of impurities on the surface or matrix. Pitting corrosion near inclusions, precipitates, or matrix/impurity interfaces is often caused by the inherent potential difference between the matrix and the inclusions or precipitates. Due to the small anode area, the corrosion rate is very fast, which can lead to sudden accidents. When trying to limit or avoid pitting corrosion, EC-AFM is helpful for improving the understanding of random pitting distribution, including nucleation and growth of pitting.

Pitting corrosion of various stainless steels was the focus of research for many years. As the most used material in the orthopedic and orthodontic bracket, AISI 316 L austenitic stainless steel has good mechanical properties. However, because it is susceptible to localized corrosion in chlorine-containing environments, it is often challenged by corrosive environments in the body. Conradi et al. [45] studied localized corrosion of austenitic stainless steel of the type AISI 316 L and duplex 2205 stainless steel (DSS 2205) using in situ EC-AFM in two different solutions, including simulated physiological solution known as Hank's solution (PS) and artificial saliva (AS). The topography and surface roughness of DSS 2205 and AISI 316 L changed with the increase in chloride ion concentration. EC-AFM topographic images showed that the surface of DSS 2205 had prominent square-like corrosion products in both solutions. Compared with AS, the erosion of PS was manifested by the decrease in the time scale of DSS 2205 surface change due to the increase in chloride ion concentration, as shown in Figure 8. Even though the DSS 2205 steel was exposed to anodic potentials in the region of transpassive oxidation, the samples had high pitting corrosion resistance in both solutions. On the contrary, AISI 316 L steel had stable corrosion resistance to AS. However, when AISI 316 L steel was exposed to PS solution, the sample was easily corroded and showed obvious pitting corrosion with the increase in Cl− ion concentration, as shown in Figure 9. The surface of AISI 316 L included ellipse-like deposits. This was due to the change in the chemical composition of the matrix material, which formed different growth patterns on the surface of the sample with the growth of chromium, iron, and nickel oxides. Therefore, DSS 2205 had high corrosion resistance compared to AISI 316 L stainless steel. In medical applications, DSS 2205 will be a promising medical material if the nickel hypersensitivity effect can be reduced in patients receiving treatment.

**Figure 8.** AFM images of DSS 2205 sample: (**a**) in artificial saliva (AS) after being exposed to the potential of 0.8 V for 45 min, (**b**) in AS after additional exposure to the potential of 1 V for 11.6 min, (**c**) in physiological solution (PS) after being exposed to a potential of 0.8 V for 17.2 min, and (**d**) in PS after additional exposure to 1 V for 7.5 min. Adapted with permission from Reference [45]; copyright 2011 Elsevier Ltd.

**Figure 9.** AFM image of AISI 316L sample: (**a**) in AS after exposing the sample to anodic potential of 0.5 V for 10 min, (**b**) in AS after additional exposure of the sample to 0.5 V for 6.6 min, (**c**) in PS immediately after the test cyclic voltammogram in the range of potentials from −0.5 V to 0.8 V, and (**d**) in PS after additional exposure to 0.5 V for 30 s. Adapted with permission from Reference [45]; copyright 2011 Elsevier Ltd.

Martin et al. [46] studied the pitting corrosion of austenitic 304L stainless steel in chloride borate buffer solution using in situ EC-AFM in contact mode. In order to determine whether the pits were randomly distributed at the nanometer scale, the study focused on the location where pits were initiated under controlled potential. In the AFM image, it is clear that the chain of pits (black, right image) was related to the chain of relief islands (white, left image), as shown in Figure 10. However, we can see an intermediate state rather than a typical direct transition, which indicates that the pitting process was a gradual process, as shown in Figure 11. In the final step of surface preparation, the local chemical reactivity of the surface may have led to the formation of a passive film and small oxidized hydroxide aggregate defects. The authors believed that the local chemical defects in the sample preparation process could affect the formation of the passive film, which would eventually cause local differences compared to the rest of the film. These local chemicals or structural defects would reduce the local pitting resistance of the film. In addition, based on the reasonable model that the surface potential under tensile stress would be higher than that under compressive stress, when the strain hardening zone (caused by mechanical polishing) appeared on the surface of the sample, pitting corrosion preferentially initiated in this area.

**Figure 10.** AFM images of the surface at open-circuit potential (**a**) and at pitting potential (**b**). (**c**,**d**) The pit chain (in black, right image) profiles made on images at higher magnification (bottom insets) show that the average relief of the islands was about 5 nm, whereas the average depth of the pits was about 20 nm. Reprinted with permission from Reference [46]; copyright 2008 Elsevier Ltd.

**Figure 11.** AFM images of the surface, taken at (**a**) open-circuit potential, (**b**) corrosion potential after cathodic scan, and (**c**) pitting potential. Reprinted with permission from Reference [46]; copyright 2008 Elsevier Ltd.

It is generally believed that sulfide inclusions (MnS and mixed oxide/sulfide) in a stainless-steel matrix are most likely to cause pitting corrosion. The products produced by the dissolution of sulfide inclusions form a local corrosive environment, which in turn causes more severe pitting corrosion. Wijesinghe et al. [47] discussed the adverse effect of sulfide inclusion on pitting resistance of stainless steel. The results showed that the sulfide inclusions were clustered on the surface of the stainless steel. As the volume composition of the stainless-steel sulfide increased, the number of inclusions per cluster increased. The pitting corrosion was imaged in real time using in situ EC-AFM, and it was found that pits were formed near the sulfide inclusions, consistent with the three main mechanisms proposed previously, i.e., aggressive local chemical reactions based on inclusion dissolution, stressed oxide, and chromium depletion.

In the work of Zhang and co-workers, the pitting corrosion of the solution- and sensitization-treated austenitic stainless steel SUS304 was studied using in situ AFM in 3.5 wt% sodium chloride solution [48]. This study adopted two observation methods. One was in situ continuous observation, in which the corrosion current was continuously applied to the surface of the sample. Another method was in situ interrupted observation, i.e., the observation was carried out at intervals to minimize the impact of probe scanning on corrosion reactions. In situ observation of the corrosion pit of the solution-treated sample showed that the pit became larger as the corrosion time increased from t = 2.46 to 2.70 ks, but no corrosion products were found, as shown in Figure 12. A series of in situ intermittent observations of the sample after solution treatment revealed two large pits (D and E), but the size of the pits did not change with time. In addition, no corrosion products were found on the two large pits, as shown in Figure 13.

**Figure 12.** AFM images of solution-treated SUS304 stainless steel during corrosion in 3.5 wt% sodium chloride solution at 298 K (I = 10 A/m2): (**a**) t = 2.46 ks; (**b**) t = 2.52 ks; (**c**) t = 2.7 ks. Reprinted with permission from Reference [48]; copyright 2005 Elsevier Ltd.

**Figure 13.** AFM images of solution-treated SUS304 stainless steel after corrosion in sodium chloride solution at 298 K (I = 5 A/m2): (**a**) t = 1.5 ks; (**b**) t = 1.5 ks and (**c**) t = 7.2 ks are the magnified images of the framed area of (**a**), showing the presence of the corrosion product. Reprinted with permission from Reference [48]; copyright 2005 Elsevier Ltd.

The authors believed that the corrosion product may have been moved to other locations by the probe or dissolved with the rapid dissolution of the matrix. After the corrosion product was removed, the concentration of local chloride ions and hydrogen ions decreased, and the growth rate of the pits decreased. This means that the corrosion products played an important role in the growth of corrosion pits. When the corrosion product covers the pit, it can lead to an acidic environment and accelerate pitting. On the sensitization-treated sample, the irregular pits were distributed near the grain boundaries. Chromium carbide deposits and pitting occurred in the chromium-depleted area around the carbide, as shown in Figure 14. Cross-sectional profiles of pits along a–b lines in Figure 14a and c–f lines in Figure 14c are shown in Figure 14d. Since no carbide particles were found in the pit according to the cross-sectional profiles, they might have dissolved with pit growth. The dissolution of carbides in pits caused the pit to grow further.

**Figure 14.** AFM images of sensitization-treated SUS304 stainless steel after corrosion in sodium chloride solution (I = 1 A/m2): (**a**) t = 0.6 ks; (**b**) t = 0.9 ks; (**c**) t = 1.2 ks. (**d**) Cross-sectional profiles of pits along a–b and c–f lines. Reprinted with permission from Reference [48]; copyright 2005 Elsevier Ltd.

When we study the corrosion of self-passivation metals covered with a semiconductor protective film, there is no doubt that EC-AFM, which does not need the conductivity of the matrix, is more suitable. Qu et al. [49] investigated the corrosion behavior of pure aluminum using in situ AFM in 0.01 mol/L FeC13 solution. The pitting corrosion process of pure aluminum induced by the potentiodynamic sweep and the repassivation process of active pits were studied. In addition, the effect of mechanical damage on the metal surface caused by AFM tip scratching on the pitting behavior of the sample was emphasized. The topographical images of the sample surface at different immersion times were traced under open-circuit conditions using AFM in contact mode, and the AFM tip scratching process was carried out with a loading force of 800 nN. The results showed that different pitting regions exhibited different pitting activities under the same polarization conditions due to the diversity of physical and electrochemical characterization. The corrosion products contained abundant impure elements such as Fe and Cu. In situ AFM observation of pitting corrosion originated from artificial defects on the aluminum surface showed that physical defects had higher pitting activity and may have been attacked preferentially to pitting corrosion.

In the study of Davoodi and co-workers, the application of the EC-AFM and SECM integrated system for in situ studies of the influence of intermetallic particles on local corrosion of aluminum alloys was introduced [50]. The key to this method is fabricating a dual-mode cantilever/tip that can be used not only as a cantilever for the EC-AFM but also as a microelectrode or nano-electrode tip for the SECM. It can obtain the in situ AFM topographic images and SECM electrochemical current maps simultaneously with micrometer lateral resolution and provide detailed information of localized corrosion related to different kinds of intermetallic particles and the deposition of corrosion products. In their work, when the AA1050 sample was anodically polarized at 300 mV (close to the breakdown potential), the morphology observed using EC-AFM showed many small pits. However, this may have been because many of the active points were close to each other; thus, the local current could not be resolved in the SECM electrochemical current map. During the post-scan, some high-current

sites appeared in the lower part of the SECM image, which may have been related to the presence of pits or trenches on the surface, as shown in Figure 15. Preliminary results indicated that the localized dissolution of aluminum alloy may have occurred below the breakdown potential, but only involved a small number of particles.

**Figure 15.** EC-AFM/scanning electrochemical microscopy (SECM) images of AA1050 in 10 mM NaCl+ 5 mM KI at 300 mV anodic polarization, and tip at +750 mV vs. Ag/AgCl: (**a**) topography; (**b**) electrochemical activity map. Adapted with permission from Reference [50]; copyright 2005 Electrochemical Society.

Metallic zinc is used not only for galvanizing steel but also for various applications such as batteries, die casting, and brass metallurgy. It is necessary to study the local corrosion of pure zinc and zinc alloys. Amin et al. [51] studied the passivation breakdown and pitting corrosion of zinc in 0.5 M sodium hydroxide solution containing different concentrations of ClO3 <sup>−</sup> or ClO4 − anions. The passivation breakdown and pitting sensitivity of the Zn/OH−/ClO3 <sup>−</sup> and Zn/OH−/ClO4 − interfaces were studied by potentiodynamic anodic polarization measurements, and the topography of pitted Zn surfaces was observed by AFM. The experimental results showed that metastable pitting corrosion was mainly caused by Cl<sup>−</sup> ions (generated by reducing ClO3 <sup>−</sup> and ClO4 −). According to the point defect model (PDM), this may have been due to the small size of Cl− ion, which could occupy anion vacancies and cause pitting corrosion. Therefore, Cl− was the invasive ion that caused metastable

pitting, rather than ClO3 <sup>−</sup> and ClO4 <sup>−</sup>. However, even in the absence of Cl<sup>−</sup>, ClO3 <sup>−</sup> and ClO4 − could induce stable pitting, in which ClO3 <sup>−</sup> was more aggressive than ClO4 −, which could be proven by the roughness calculated according to the AFM topography map. Compared with ClO4 −, the Ra value was always higher in the presence of ClO3 −, as shown in Figure 16.

**Figure 16.** AFM images recorded for Zn in 0.5 M NaOH solution containing 0.05 M (**a**,**b**) ClO4 − or (**c**,**d**) ClO3 −. Adapted with permission from Reference [51]; copyright 2013 Springer.

Yong Hwan Kim et al. [52] studied the initial corrosion mechanism of a hot-dip-galvanized surface. The formation of corrosion products, the initiation and growth of pits, and the breakdown of the film were observed using in situ EC-AFM. At the initial stage of passivation, the corrosion product film of ZnO/Zn(OH)2 formed in the dull sector was more unstable than that formed in the bright zone. The authors believed that the relative instability of the passive film in the initial stage was affected by the high-density lath-like structure. The uneven surface structure at the micro-scale, i.e., the lath-like structure, provided a favorable place for the ZnO/Zn(OH)2 formation and pitting.

EC-AFM can provide detailed information of the localized dissolution associated with different kinds of intermetallic particles, as well as the deposition of corrosion products surrounding large particles or covering small pits, including the location of pitting initiation at controlled potentials, and whether pits are randomly distributed at the nanoscale. The application of EC-AFM in pitting corrosion is summarized in Table 3.

#### 3.1.3. Selective Corrosion

Selective corrosion refers to the preferential dissolution of active components in the multicomponent alloy. The relative Volta potential difference is used as the index of selective corrosion caused by microelectronic coupling. Compared with the substrate, the intermetallic phase with positive potential is the cathodic phase. On the contrary, some intermetallics with more negative potential than the matrix show stronger activity during the corrosion process and act as micro-anodes.

Jia et al. [53] investigated the effects of intermetallics on the local corrosion behavior of AZ91 alloy added (La,Ce) mischmetal (MM), including corrosion initiation and propagation. The corrosion morphologies of different intermetallics were imaged using in situ EC-AFM in 0.1 mol/L NaCl solution, and Scanning Kelvin probe force microscopy (SKPFM) was performed to measure the local difference

of the relative Volta potential between the intermetallic phases and the α-Mg matrix. In the experiment, the approximate compositions of several intermetallic phases were identified as Al4(La,Ce), Al8Mn4Ce, and decreased b-Mg17Al12 phases. SKPFM analysis showed that all intermetallics were noble compared with the α-Mg matrix. The observation of in situ EC-AFM on the initial stages of corrosion indicated that the a-Mg matrix surrounded by b-Mg17Al12 or Al8Mn4Ce was prone to pitting corrosion. In detail, the b-Mg17Al12 phase existed in both AZ91 and AZ91 alloys with (La,Ce) MM. Both elongated and dispersed granular b-Mg17Al12 served as effective micro-cathodes, as shown in Figure 17.


**Table 3.** Summary of the application of EC-AFM in pitting corrosion.

**Figure 17.** The polished AZ91 alloy (**a**) surface topography and (**b**) relative Volta potential map of elongated b-Mg17Al12; (**c**) surface topography and (**d**) relative Volta potential map of granular-like b-Mg17Al12. Reprinted with permission from Reference [53]; copyright 2019 Elsevier Ltd.

Al4(La,Ce) intermetallics, whether featuring an acicular shape or rod shape, were believed to produce significant galvanic coupling. However, the α-Mg matrix near the Al4(La,Ce) phase did not have obvious corrosion. Moreover, although the Al8Mn4Ce phase showed a positive potential difference with respect to the matrix, the corrosion resistance of the alloy was not affected by its galvanic coupling, probably due to its lower amount, as shown in Figure 18.

**Figure 18.** Typical in situ surface images for different intermetallics after immersion in 0.1 mol/L NaCl solution for 60 min: (**a**) acicular-like RE phase, (**b**) rod-like RE phase, (**c**) granular RE phase, and (**d**) dispersed granular b-Mg17Al12 in AZ91 alloy with 1.0% (La,Ce) mischmetal (MM) addition. Reprinted with permission from Reference [52]; copyright 2019 Elsevier Ltd.

It can be seen that the intermetallics with the highest Volta potential differences relative to the matrix did not play the role of effective cathodes, just like the Al4(La,Ce) phase, which means that the localization of the cathode reaction was not only dependent on the Volta potential differences between the intermetallics and the matrix. The authors believed that the shape of the coupled anode and cathode had an important effect on the current density, i.e., the geometry of the intermetallic compound had an important influence on the micro-galvanic corrosion.

Zhang et al. [54] studied the anodization process of aluminum 6060 alloy under operating conditions and illustrated the effects of intermetallic particles (IMPs) and anodic Al oxide (AAO) film properties. The in situ EC-AFM measurement was performed continuously to monitor the surface topography changes under anodizing potentials in contact mode, revealing the details of localized dissolution and AAO film formation in Al 6060 samples in 0.2 M Na2SO4 solution. The extruded Al 6060 alloy mainly contained two types of IMPs: AlFeSi primary particles and Mg2Si particles. The Volta potential differences obtained by SKPFM showed that, relative to the aluminum matrix, AlFeSi was cathodic, but Mg2Si was anodic. The in situ EC-AFM measurements showed that AlFeSi particles remained stable, but the local anodic dissolution of Mg2Si particles occurred during anodization, which was consistent with the SKPFM results. As shown in Figure 19, the large particle in region I was likely an α-AlFeSi particle. The particles in regions II, III, and IV were likely anodic Mg2Si particles. The protruded large particle in region I remained stable during anodization, whereas the

small particles in region II in Figure 19a were dissolved upon the anodization. Moreover, upon the anodization at 1 V, some active dissolution started at certain sites (e.g., III), forming small holes, as shown in Figure 19b. During the anodization at 2, 4, and 8 V (Figure 19c–e), the small hole in area III became deeper, as displayed by the profile lines in Figure 19f, and pronounced localized dissolution occurred in area IV. Two small particles remained stable in the dissolved area (area IV in Figure 19d,e). Furthermore, in the area marked as V, localized dissolution at one site resulted in a small hole, and a deposited particle of a few μm in size formed (Figure 19c), but it disappeared at 4 V, exposing a much deeper pit (Figure 19d). In addition, the growth of AAO films occurred with partial anodic dissolution. The thickness of the anodic barrier film increased linearly with the anode potential, but the growth rate decreased due to local anodic dissolution associated with IMPs in the alloy.

**Figure 19.** In situ EC-AFM topography images of the Al 6060 alloy in 0.2 M Na2SO4 solution at (**a**) OCP, (**b**) 1 V, (**c**) 2 V, (**d**) 4 V, and (**e**) 8 V vs. Ag/AgCl. (**f**) Line profiles across the feature III in (**a**–**d**). Adapted with permission from Reference [54]; copyright 2016 Electrochemical Society.

In some cases, the presence of nitrides can adversely affect the corrosion resistance of the material. Bettini et al. [55] studied the effect of nano-sized quenched-in chromium nitride particles on the corrosion behavior of heat-treated 2205 duplex stainless steel (DSS) in an NaCl solution at room temperature and 50 ◦C (slightly higher than the critical pitting corrosion temperature). The relative nobility difference between the precipitated nitrides, austenite, and ferrite in the tested materials was evaluated at room temperature by atomic force microscopy-based Kelvin force microscopy (AFM/KFM). The volt potential mapping at room temperature indicated that the ferritic had a lower relative nobility compared with austenite, and quenched-in Cr2N particles had a higher relative nobility to the surrounding ferritic matrix. The observation results of EC-AFM in 1 M NaCl solution at room temperature showed that the samples after heat treatment showed a wide range of passivation and a very stable surface up to 1.2 VAg/AgCl, where the selective dissolution of ferrite phase occurred but the quenched-in Cr2N particles remained stable, as shown in Figure 20. Figure 20d shows the depth line profile from the image obtained after the polarization at 1.2 VAg/AgCl, presenting a depth of ca. 200 nm between the dissolved ferrite phase and the remaining austenite phase. In addition, the very small nitride particles formed during the fast cooling process did not have enough time for the diffusion of elements and were unlikely to form a considerable composition gradient in the surrounding boundary region, which was one of the reasons for its stability. At temperatures above the critical pitting temperature (CPT = 50 ◦C), rapidly selective dissolution of the austenite phase occurred upon slight anodic polarization, which may have been related to the low content of chromium and molybdenum in the austenite phase. The authors believed that the finely dispersed quenched-in nitrides in the DSS did not cause local corrosion in 1 M NaCl solution. However, the exposure temperature had a great influence on the corrosion resistance of DSS, which changed the selective dissolution behavior of DSS.

**Figure 20.** In situ EC-AFM images of the same area of the 2205 HT in 1 M NaCl at room temperature under electrochemical control at different applied potentials: (**a**) OCP, (**b**) 1.1 VAg/AgCl, and (**c**) 1.2 VAg/AgCl. (**d**) Depth line profile from the arrow in image (**c**) showing particles or particle clusters remaining in the dissolved ferrite area. Reprinted with permission from Reference [55]; copyright 2013 Elsevier Ltd.

Yasakau et al. [56] investigated the mechanism of initial steps of localized corrosion at the cutting edges of adhesively bonded Zn (Z) and Zn–Al–Mg (ZM) galvanized steel substrates. The topography of the initially localized corrosion of Z and ZM samples was measured using in situ AFM equipment under anodic polarization in the corrosive solution. The first corrosion pits at the Z cutting edge

were mainly formed on the zinc layer, and there was no preferential erosion on the adhesive/zinc or zinc/steel, as shown in Figure 21.

**Figure 21.** AFM topography maps made on cutting edge of adhesively bonded Zn (Z) substrate before immersion (**a**) and during 24 min (**b**), 1 h 5 min (**c**), and 2 h (**d**) of immersion in 0.001 M NaCl. (**e**) Evolution of topography across the black line profiles drawn in the same zones on AFM maps. Black arrows indicate local pits on the zinc layer. Reprinted with permission from Reference [56]; copyright 2016 Elsevier Ltd.

However, the types of attacks were different at the ZM cut edge, namely, pitting corrosion of the solid solution and selective dissolution at the eutectic zone. For the local corrosion at the adhesive/zinc interface, the first adhesive disbonding area was located near the cutting edge. The second adhesive disbonding zone was located at the buried deep zinc/adhesive interface, where local corrosion occurred in the solid solution phase and eutectic phase of the sample ZM. Under anodic polarization, the corrosion site was located in the eutectic phase at the interface between the eutectic zone and the solid solution zone, similar to the corrosion at the ZM cutting edge, as shown in Figure 22. In addition, scanning vibrating electrode technology (SVET) and electrochemical impedance spectroscopy (EIS) electrochemical tests showed a decrease in corrosion kinetics at the Z and ZM cutting edges, which was due to the blocking effect of the dense film of corrosion products formed on the zinc and steel surfaces.

**Figure 22.** AFM topography (**a**) and SKPFM (**b**) maps of Zn–Al–Mg (ZM) galvanized coating before immersion and during immersion in 0.005 M NaCl for 4 min (**c**), 9 min (**d**), and 22 min (**e**). (**f**) Evolution of topography across the black line profiles. Reprinted with permission from Reference [56]; copyright 2016 Elsevier Ltd.

Depentori et al. [57] studied the corrosion behavior of neodymium-modified titanium alloy Ti6Al4V2Nd in 1.5 wt% NaCl and compared it with a Ti6Al4V2 matrix using SKPFM and EC-AFM. Neodymium containing intermetallic compounds was precipitated at β-phase boundaries and inside the grains. The Volta potential maps obtained using SKPFM showed that the Volta potential of intermetallic compounds was lower than that of the Ti matrix. Therefore, intermetallic compounds had strong anodic behavior relative to titanium matrix and were the preferred site for local attack in Ti6Al4V2Nd alloy. When the sample was immersed in 1.5 wt% NaCl solution, EC-AFM observation showed that the volume of the local area on Ti6Al4V2Nd increased significantly, which was due to the formation of hydroxide and oxide. In addition, large amounts of debris were observed. The authors suggested this as a sign that the surface was loosely bound to Nd(OH)3. Cyclic voltammetry showed a clear oxidation reaction without a reduction reaction, and the oxidation peak moved to the right as the exposure time increased, which was a clear sign that the electrochemical reaction was irreversible and that the diffusion barrier formed and increased over time. Corresponding to the results of EC-AFM, the diffusion barrier was the Nd(OH)3 layer formed by the corrosion of a large number of precipitates on the surface, which prevented further oxidation.

Davoodi et al. [58] investigated the difference in corrosion behavior between EN AW-3003 (Rolled 3xxx series Al alloys) and a newly developed Al–Mn–Si–Zr fin alloy. The Volta potential of the two alloys determined using SKPFM showed that the intermetallic particles behaved as cathodes relative to the alloy matrix. Compared to EN AW-3003, the Al–Mn–Si–Zr alloy had fewer particles with larger Volta potential difference with respect to the matrix. In situ AFM measurements showed that ring-like corrosion products were deposited on the EN AW-3003 alloy, while only a few corrosion sites and tunnel-like pits were found on Al–Mn–Si–Zr, as shown in Figure 23.

**Figure 23.** In situ AFM images of (**a**) EN AW-3003 and (**b**) Al–Mn–Si–Zr, after two days; (**e**) EN AW-3003 and (**f**) Al–Mn–Si–Zr, after 3.5 days in Sea Water Acetic Acid Test (SWAAT) solution of pH 4. (**c**,**d**) Magnified images of the framed areas of (**a**,**b**), respectively. Adapted with permission from Reference [58]; copyright 2007 Elsevier Ltd.

Topography and electrochemical currents obtained synchronously by the integrated SECM/AFM systems provided information on the pitting precursor and pitting process. Compared with Al–Mn–Si–Zr alloys, EN AW-3003 alloys had more active sites and extensive localized dissolution, resulting in higher material losses. Interestingly, some of the larger micron-sized intermetallic particles initiated localized dissolution at the boundary region of the particle-matrix, while the fine dispersions were not active. The intermetallic particles in the Al–Mn–Si–Zr alloy were few. Although they could induce selective dissolution and form small tunnel-like holes, because of their small weight loss, they were suitable for fin material in heat exchange applications.

The combination of EC-AFM and scanning Kelvin probe force microscopy (SKPFM) can simultaneously obtain topographical changes and Volta potential maps, which helps to better understand selective corrosion behavior and its mechanism. The application of EC-AFM in selective corrosion is summarized in Table 4.


**Table 4.** Summary of the application of EC-AFM in selective corrosion.

#### 3.1.4. Intercrystalline Corrosion

Intercrystalline corrosion is a local corrosion phenomenon that occurs along the grain boundary of a metal material in a corrosive medium and causes the loss of bonding force between grains. When intercrystalline corrosion occurs in the material, there may be no macroscopic change, but the material's strength is almost completely lost, resulting in the sudden destruction of equipment. The main reason for intercrystalline corrosion is the difference in structure and chemical composition between the grains and grain boundaries of the material.

Based on EC-AFM, the contact mode high-speed AFM (HS-AFM) compensates for the shortcomings of AFM with short collection times. The long collection time is a limiting factor for AFM. Contact mode HS-AFM images multiple frames per second, making it orders of magnitude faster than traditional AFM. This enables real-time imaging processes with nano-scale lateral resolution and sub-nanometer-scale height resolution. The increase in speed can not only directly image dynamic nanoscale events, but also macroscopic regions of the sample surface without reducing resolution. It is a valuable imaging tool for in situ observation of nanoscale corrosion initiation events such as metastable pitting, grain boundary dissolution, and short crack formation during stress corrosion cracking.

In the work of Moore and co-workers, local corrosion phenomena such as pitting and intergranular attack (IGA) on thermally sensitized AISI 304 stainless steel were studied using HS-AFM in 1% NaCl [59]. Real-time in situ HS-AFM observations showed that an intergranular pit was formed within 0.5 s during a galvanostatic scan. Intergranular pits were distributed along the grain boundary (GB) in a chain shape, which was caused by the preferential corrosion of GB. Chromium carbide was precipitated along GBs, which resulted in the depletion of the local chromium elements in the area around GBs, greatly reducing the local corrosion resistance of GBs. By using HS-AFM and electrochemical data for the computational model, it was found that the radial diffusion state of the system was reached within 0.01 s, leading to rapid dissolution of materials.

Padhy et al. [60] used EC-AFM to study the surface morphology of austenitic stainless steel in nitric acid medium. In situ EC-AFM results showed that the surface presented a platelet-like structure in low-concentration nitric acid solution (0.1 M, 0.5 M), providing effective protection for the surface. When the concentration of HNO3 was from 0.1 M to 0.6 M, the roughness decreased, which was related to the thinning of the passive film and marked the beginning of corrosion. From the morphology of 0.6 M HNO3, breakdown of the passive film and surface dissolution were observed. As the concentration continued to increase to 1 M, the roughness increased due to the intensification of the surface dissolution and selective dissolution of grain boundaries, as shown in Figure 24.

The early stages of localized corrosion are important as mentioned earlier, because they account for most of the lifetime of intergranular corrosion, stress corrosion cracking (SCC), or pitting, and they are more accessible to take remedial action. Williford et al. [61] used EC-AFM to obtain images of the early stages of intergranular corrosion (IGC) in 304L stainless steel. The observation of EC-AFM showed that IGC was present between the carbides, but not completely around the carbide in the early stage. Later, the grain boundaries became wider, the carbides were shortened, and the IGC completely surrounded the carbides. The matrix between carbides began to corrode, which was best explained by the fact that the carbides were cathodic with respect to the matrix; thus, when the matrix dissolved, they were protected by cathodic protection. In addition, the carbide was connected to adjacent grains through the ligaments of the matrix material, which may have meant that the SCC crack front was bridged by the carbide. The SCC crack front could move forward and surround the carbide, creating an area where the carbide bridged the crack.

**Figure 24.** Surface morphology of 304L stainless steel at 1100 mV in (**a**) 0.1 M HNO3, (**b**) 0.5 M HNO3, (**c**) 0.6 M HNO3, and (**d**) 1 M HNO3. Reprinted with permission from Reference [60]; copyright 2010 Elsevier Ltd.

An AFM with a conductive probe can simultaneously obtain the surface topography and surface potential of the scanned surface. Fu et al. [62] studied the local corrosion of high-chromium cast iron in regions at different distances from interphase boundaries using EC-AFM in the contact mode. According to the measured local potential, a decrease in interface potential between carbide and matrix was observed. Moreover, the corrosion rate of the metal matrix near the primary carbide was significantly higher than that far away from the primary carbide. The morphology map obtained using EC-AFM showed that the area near the primary carbide/matrix interface corroded or dissolved more rapidly than the area away from the interface. Furthermore, the corrosion rate was found to be particularly rapid in areas with sharp edges. The authors suggested that, on the flat side, most of the electrons flowed into the half-space on one side of the matrix. However, at the sharp edge, electrons flowed into a larger space in the matrix, corresponding to a larger total electron flow, which greatly increased the corrosion rate of the sharp edge.

In high-entropy alloys (HEAs), the homogeneous elemental distribution is expected to improve corrosion resistance. In order to develop highly corrosion-resistant HEAs, it is necessary to study the relationship between chemical/microstructure segregation and localized corrosion. In the study of Shi and co-workers, localized corrosion of AlxCoCrFeNi HEAs was studied using in situ EC-AFM in 3.5 wt% NaCl solution. Surface topography changes at micron/submicron scales were monitored under different anodic potentials [63]. In the experiment, with the increase in aluminum content in HEAs, the microstructure changed from a single FCC (face-centered cubic) solid solution to the FCC phase and (ordered/disordered) BCC (face-centered cubic) phase. The EC-AFM image showed that the uniform single-phase Al0.3CoCrFeNi alloy had the best corrosion resistance, and the breakdown of its passive film was in the form of randomly distributed pits, as shown in Figure 25.

**Figure 25.** In situ EC-AFM topography images of the Al0.3CoCrFeNi high-entropy alloy (HEA) in a 3.5 wt% NaCl solution after 10 min of exposure at (**a**) OCP, (**b**) 0.2 V, and (**c**) 0.6 V. Reprinted with permission from Reference [63]; copyright 2018 Elsevier Ltd.

With the increase in aluminum content, the BCC phase appeared in the Al0.5CoCrFeNi alloy, resulting in a heterogeneous microstructure. Pits were formed preferentially along the FCC/BCC phase boundary, leading to the initial breakdown of the passive film, which was manifested as the decline in critical pitting potential value in the potentiodynamic polarization curve, as shown in Figure 26.

Pitting corrosion was not observed in the Al0.7CoCrFeNi alloy with the increase in aluminum content and volume fraction of the BCC phase. In contrast, local corrosion along the dendritic/interdendritic boundaries and selective dissolution of the (Al, Ni)-rich ordered BCC phase occurred, as shown in Figure 27. Therefore, the author believed that, as the microstructure changed from single solid solution to multiphase, the breakdown of the passivation film changed from pitting to phase boundary dissolution, which led to a decrease in corrosion resistance.

Bettini et al. [64] studied the effects of carbides on the corrosion/dissolution behavior of biomedical CoCrMo alloys in PBS solution using EC-AFM at different applied potentials. SKPFM results showed that, compared with the matrix, the Volta potential of carbides was higher. In addition, the Volta potential decreased in the boundary region, which may have been related to local depletion of the main alloy elements. This indicated that the carbide boundary had a greater corrosion tendency and was the preferred site for corrosion/dissolution. In situ EC-AFM measurements of the CoCrMo alloys exposed to PBS showed that, at high anodic potential, a dissolution process at carbide boundaries was observed, and an increase in boundary depth was seen in line profiles across these boundaries, as shown in Figure 28. This was consistent with the SKPFM Volta potential mapping, which showed that some of the boundary areas were weak sites for corrosion/dissolution due to lower relative nobility compared to the matrix.

**Figure 26.** The AFM topography image in air (**a**), and in situ EC-AFM topography images of the Al0.5CoCrFeNi HEA after 10 min of exposure in a 3.5 wt. % NaCl solution at (**b**) OCP, (**c**) 0 V, and (**d**) 0.4 V. (**e**) Three-line profiles across the phase boundaries illustrated in (**a**). (**f**) Vertical profiles along line 1 of the surface in (**a**–**d**). Reprinted with permission from Reference [63]; copyright 2018 Elsevier Ltd.

**Figure 27.** In situ EC-AFM topography images of the Al0.7CoCrFeNi HEA after 10 min of exposure in a 3.5 wt% NaCl solution at (**a**) OCP, (**b**) 0.1 V, and (**c**) 0.1 V, and (**d**) after 15 min of exposure at 0.1 V vs. SCE. (**e**) Three-line profiles across the phase boundaries illustrated in (**a**). (**f**) Vertical profiles along line 1 of the surface in (**a**–**d**). Reprinted with permission from Reference [63]; copyright 2018 Elsevier Ltd.

**Figure 28.** In situ AFM images of the same area of CoCrMo alloy under electrochemical control in PBS solution with pH 7.4, (**a**) after 120 min at OCP, (**c**,**e**) after 10 and 30 min at 0.5 Vsat Ag/AgCl, and (**g**,**i**,**k**) after 10, 30, and 50 min at 0.7 Vsat Ag/AgCl, with marked etching-like dissolution sites on the carbide (where visible). (**b**,**d**,**f**,**h**,**j**,**l**) Depth line profiles at the applied potential and time. Adapted with permission from Reference [64]; copyright 2011 Elsevier Ltd.

Davoodi et al. [65] studied the localized corrosion and preferential dissolution of Al alloys in chlorine solution using an integrated EC-AFM/ SECM system. The integrated EC-AFM/SECM could simultaneously detect topographic changes and electrochemical active sites in the same region and reveals local corrosion processes related to IMPs. The results showed that preferential dissolution occurred in the interfacial region between the alloy matrix and IMPs. The formation of grooves around the larger IMPs indicated that different types of IMPs had different dissolution behaviors. In addition, they found that only a small number of IMPs were involved in the localized dissolution at any given time.

EC-AFM can clearly reveal the formation of trenches and the local dissolution of the grain/phase interface, and it can explore the causes and mechanisms of intercrystalline corrosion in combination with other test tools. Table 5 summarizes the application of EC-AFM in intercrystalline corrosion.


**Table 5.** Summary of the application of EC-AFM in intercrystalline corrosion.

#### *3.2. Metal Protection*

#### 3.2.1. Coating Protection

One of the primary protection techniques for metallic materials is the use of a cover layer on the metal surface to avoid direct contact between the metal and the corrosive medium as much as possible. The coating provides an effective barrier to the substrate, slowing the corrosion of the metal. In general, the metal surface coating can be divided into the metal coating and non-metallic coating. The coating not only slows electron transfer between the anode and cathode; it can also act as a barrier to prevent oxygen from penetrating the cathode reaction. Microcrystalline coating, nanocrystalline coating, gradient coating, composite coating, etc. can effectively improve the overall performance of the material, including corrosion resistance.

Li and co-workers studied the electrochemical mechanism and corrosion protection properties of solvent-borne alkyd composite coating containing 1.0 wt% CeO2 nanoparticles (CeNPs) and 1.0 wt% polyaniline (PANI) for carbon steel in NaCl solution [66]. In their work, the morphology changes of the coatings and redox reactions of PANI at the nanoscale were accurately monitored by linking the volume changes observed using in situ EC-AFM imaging with redox peaks measured using in situ cyclic voltammetry (CV). The results of EC-AFM showed that PANI nanoparticles in the alkyd matrix exhibited contracted morphology in the reduced state, leucoemeraldine base (LB), and expanded morphology in the oxidized state, emeraldine salt (ES). The surface did not change significantly, which indicated that the composite coating was stable in corrosive solutions even under very harsh potential conditions, as shown in Figure 29. OCP and EIS results indicated that the redox reaction of PANI ES/LB forms caused metal passivation, which was an active corrosion protection mechanism.

**Figure 29.** EC-AFM images (3D) of the same scan area of the composite coating in 3.0 wt% NaCl solution obtained after 60 min (**a**) at OCP, (**b**) at −700 mV (vs. Ag/AgCl), and (**c**) at 800 mV (vs. Ag/AgCl). (**d**) Line profiles of CeO2 nanoparticle (CeNP) aggregates when at the two applied potentials drawn by the crossing line. Reprinted with permission from Reference [66]; copyright 2019 Elsevier Ltd.

The evolution of OCP with exposure time under 3.0 wt% NaCl for composite coating and reference coating showed that the potential value of composite coating was higher than that of reference coating during the whole measurement process, indicating that the corrosion resistance of the composite coating was improved. The increase in OCP value was attributed to the fact that CeNPs significantly improved the barrier effect and slowed down the migration of corrosion ions. Therefore, the synergistic effect of PANI and CeNPs greatly improved the barrier performance and the corrosion resistance of alkyd composite coating.

In a series of studies by Li and his colleagues, they also studied the electrochemical activity of 1.0 wt% *p*-toluene sulfonic acid (PTSA)-doped PANI in solvent-borne alkyd composite coating, as well as its self-healing corrosion protection mechanism on carbon steel in 3.0 wt% NaCl solution [67]. Through the CV (cyclic voltammetry) curves and EC-AFM imaging, it was proven that doped PANI at low percentage had extremely high electrochemical activity, showing contraction at the reduction potential and expansion at the oxidation potential, which provided evidence for a reversible redox reaction between the ES and LB forms. The voltage potential diagram obtained using KFM under air conditions indicated that PTSA-doped PANI had sufficiently high electrochemical activity and stable reoxidation ability to keep the passivation region on the metal surface. The PTSA-doped PANI in alkyd composites caused structural mutations due to energy input, which increased the electrochemical activity and provided doped PANI with a good electrochemical connection to the metal surface. It played an important role in improving the self-healing corrosion protection of composite coatings. In addition, the EIS showed increased resistance of the composite coating, which may have been related to the interaction of PANI particles in the alkyd matrix forming a dense and more resistive network.

Moreover, they studied the corrosion protection properties of a waterborne acrylic composite coating with 1.0 wt% acetic acid-stabilized CeNPs on carbon steel in 3.0 wt% NaCl solution [68]. The results of in situ AFM showed that the CeNPs embedded in the composite coating could greatly reduce the nano-pinholes in the waterborne acrylic coating, as well as significantly improve the stability of the coating, which played an important role in improving the barrier property of the coating. In situ EC-AFM indicated that some CeNPs and aggregates were released from the coating surface during exposure, and then some particles and cerium compounds were precipitated, as shown in Figure 30. The presence of CeNPs or aggregates acted as nucleation sites to promote precipitation on the coating surface and inside the coating pinholes, thereby preventing the entry of corrosive ions and the corrosion of the metal matrix.

**Figure 30.** EC-AFM images (3D) of the same area of the composite coating obtained after 14 h of exposure in 3.0 wt% NaCl solution: (**a**) after 60 min at OCP; (**b**) after 75 min at 1.1 V vs. Ag/AgCl. Adapted with permission from Reference [68]; copyright 2015 Electrochemical Society.

Liu et al. [69] studied the pitting behavior of austenitic stainless steel with nanocrystalline (NC) and polycrystalline (PC) microstructures in 3.5 wt% NaCl solution. In situ AFM was used to study the process of passive film formation on the PC alloy and NC alloy under anodic potential. AFM observations showed that the passive film formed rapidly on the PC alloy, and pitting occurred after continuous film formation, which was a slow process of metastable pitting formation and reparation. However, the formation of a passive film on the NC alloy indicated that, due to the accumulation of many small particles on the surface, the oxide particles grew in the original position and eventually became a passive film. The voids and the boundaries of oxide particles may have been inoculation points for metastable pits. Although metastable pits were easy to initiate on the NC coating, the small grain size promoted the diffusion of elements, such that pits could be quickly repaired or healed. Therefore, the pitting mechanism of the NC coating was mainly characterized by rapid metastable pitting initiation and death, and its pitting resistance was higher than that of the PC alloy. In addition, in their other study, the characteristics of both pit initiation and pit growth processes on an austenitic stainless-steel NC coating were monitored using in situ AFM [70]. Pit initiation included the formation of metastable pits and repassivation process. Pitting growth included stable pit growth and material dissolution. The fine grain size promoted the formation and growth of nano-scale oxide particles, which significantly improved the repassivation ability and reduced the probability of stable pit formation. Compared with PC austenitic stainless steel, nanocrystals promoted the formation of metastable pits but reduced the rate of stable nucleation and growth of pits.

In the study of Pan and co-workers, the pitting corrosion of coarse crystal (CC) 304 stainless steel and its NC thin film was studied in 3.5 wt% NaCl solution, especially the influence of nanocrystalline on the pitting process [71]. The whole pit growth process was recorded using in situ AFM and the growth mechanism of stable pits on NC film was understood. The results showed that the initiation site of the pit on the NC film was at the boundary of the oxide particles. As there were lots of boundaries on the surface, metastable pit events on NC films were more likely to occur than those on CC 304 stainless steel, which indicated that nanocrystallization promoted metastable pit processes, as shown in Figure 31.

**Figure 31.** In situ AFM images of nanocrystalline (NC) thin film in the initial pitting stage under anodic polarization in 3.5 wt% NaCl solution: (**a**) 170 s; (**b**) 230 s; (**c**) 340 s. Reprinted with permission from Reference [71]; copyright 2013 Elsevier Ltd.

In addition, the transition from metastable pitting to stable pitting was inhibited due to the excellent repassivation ability of NC films. In the process of NC film deposition, the internal residual stress may have inhibited the formation of lace cover in the process of stable pit growth, and then changed the growth mechanism of the NC film surface stable pit. Therefore, the probability of developing from metastable pitting to stable pitting on NC films was much lower than that on CC 304L stainless steel. Nanocrystallization changed the geometry and growth mechanism of stable pits, slowing down the nucleation and growth process, which improved the pitting corrosion resistance of CC 304L stainless steel. Moreover, they also studied the corrosion behavior of a magnetron-sputtered NC 304L stainless steel coating in 0.05 M H2SO4 + 0.2 M NaCl solution, which was compared with

conventional rolled CC 304L stainless steel [72]. The nanocrystalline structure reduced the adsorption capacity of Cl− on the surface and inhibited the incorporation of Cl− in the passive film. In situ AFM observation showed that the growth rate of the passive film on the NC film was greatly higher than that on the CC 304L stainless steel, as shown in Figure 32. In other words, the nanocrystalline structure improved the growth rate of the passive film and facilitated the healing of the passive film rupture. The composition of the passive film on the NC film determined by XPS had a higher ratio of chromium oxide to iron oxide. The higher content of chromium oxide improved the corrosion resistance of nanocrystalline samples. Since the structure of the passive film was more compact, the ratio of chromium oxide to iron oxide was higher, and the incorporation of Cl− was less, the corrosion resistance of the NC film was greatly improved.

**Figure 32.** In situ AFM images of coarse crystal (CC) 304L stainless steel (**a**,**b**) and a NC thin film (**c**,**d**) (scale 2 μm × 2 μm) in the growth stage of a passive film under anodic polarization in 0.05 M H2SO4 + 0.2 M NaCl solution after passivation of 6 min (**a**,**c**) and 12 min (**b**,**d**). Reprinted with permission from Reference [72]; copyright 2012 Electrochemical Society.

In addition, the passive film growth mechanisms of the NC 304L stainless-steel thin film, deep-rolled bulk nanocrystalline (BN) 304 stainless steel, and CC 304 stainless steel in 0.05 M H2SO4 + 0.2 M NaCl solution were studied using electrochemical measurements and in situ AFM [73]. The growth rate of the passive film on the three materials was in the following order: NC thin film > BN304 stainless steel > CC 304L stainless steel. Nanocrystallization changed the nucleation mechanism of passive films from gradual to instantaneous. The passive film on the CC 304L stainless steel and BN 304L stainless steel had a single-layer structure, while the passive film on the NC film had a multi-layer structure.

EC-AFM can not only detect the state of the coating surface through high-resolution imaging, but also produce coating defects by means of probe scraping to obtain direct information on the coating corrosion resistance. The application of EC-AFM in coating protection is summarized in Table 6.


#### 3.2.2. Corrosion Inhibitor Protection

A corrosion inhibitor is a chemical or a mixture of several chemicals that prevents or slows corrosion when present in a corrosive environment (medium) in the proper concentration and form. The addition of corrosion inhibitors can significantly reduce the corrosion rate of metal materials. At the same time, the original physical and mechanical properties of the metal material can be maintained. The advantages of using corrosion inhibitors to protect metal lie in their low dosage, quick effect, low cost, and convenient use. At present, corrosion inhibitors are widely used in machinery, petrochemical, metallurgy, energy and other industries. In some studies, the authors used in situ EC-AFM under real-time operating conditions to detect changes in the corrosion morphology of the sample after the addition of the corrosion inhibitor, and then investigated the properties of the corrosion inhibitor and speculated on the corrosion inhibition mechanism.

The study of non-toxic corrosion inhibitors is important for replacing classical molecules with sulfur, nitrogen, or aromatic functions. Rocca et al. [74] reported the inhibition conditions and mechanisms of linear sodium heptanoate on copper corrosion. In situ EC-AFM under applied potential allowed observing the morphology of the passive film without damaging the layer. The result showed that, when pH = 5.7 and 11, for 0.08 M NaC7, large non-covering copper heptonate crystals or non-covering copper oxide were formed on the surface, and the corrosion inhibitor efficiency was low in acidic medium. However, when the pH value was 8, for 0.08 M NaC7, a thin layer of heptanoic was formed, which acted as a barrier and effectively protected the metal matrix. The inhibition of sodium heptanoate was related to the formation of a protective layer consisting mainly of copper heptanoate on copper, and the optimum corrosion inhibition conditions were 0.08 M NaC7, pH 8 for copper corrosion.

Bertrand et al. [75] used in situ EC-AFM to study the corrosion behavior of a copper surface immersed in various electrolytes under dynamic potential conditions at room temperature. In sodium sulfate and sulfuric acid solutions, dissolution precipitation occurred on the surface, which changed the topographic characteristics of the metallic surface. On the contrary, when copper was immersed in borate or heptanoate solutions, passivation could be clearly observed. Regardless of the oxidation mode (constant potential or open-circuit condition), a thin passive layer was grown on the surface and was stable over time. In the borate medium, the sediments were composed of Cu2O and CuO oxides, while, in the heptanate electrolyte, a metal soap composed of copper(II) heptanate was detected. In addition, the inhibition mechanism of sodium heptanate was identified, whereby a thin passive layer of copper metal soap was formed on the surface via dissolution precipitation.

The use of polymer corrosion inhibitors attracted attention. On the one hand, they have low cost and good stability. On the other hand, they have multiple adsorption sites that form complexes with metal ions covering the surface and protecting the metal from corrosion. Umoren et al. [76] studied the mechanism of polyacrylic acid (PAA) inhibiting the corrosion of pure cast aluminum in acidic medium and the synergistic effect with iodide ion addition. The in situ AFM morphology of the surface showed that PAA was adsorbed onto the surface of the aluminum and its arrangement was more orderly in the presence of iodide ions, as shown in Figure 33, resulting in higher inhibition efficiency. The authors believed that the alumina film was replaced by the adsorption of KI on aluminum, and then the PAA was adsorbed onto the KI such that the PAA molecules were arranged on the aluminum surface in an orderly manner, which enhanced the inhibition process.

**Figure 33.** In situ AFM image of Al in 0.5 M H2SO4 in the presence of polyacrylic acid (PAA) + KI at different potentials: (**a**) −1.0 V, (**b**) −0.70 V, and (**c**) −0.50 V. (**d**) Magnified images of the framed area of (**c**). Adapted with permission from Reference [76]; copyright 2013 Taylor & Francis.

Zhang et al. [77] investigated the electrochemical corrosion behavior of AISI321 stainless steel in 36% ethylene glycol–water solution. It can be seen from the results of EC-AFM that the passive film became more complete and the number of defects decreased with the increase in polarization time. Moreover, as the passive potential increased, the particle diameter increased and surface defects decreased. When the potential ranged from −0.15 to 0.45 V, an N-type oxide film adhered to the surface of the sample. However, when the potential was between 0.45 and 0.75 V, a P-type oxide film was formed on the surface. The passive film formed at high potential had fewer defects and excellent protective performance. Therefore, AISI321 stainless steel could be passivated in ethylene glycol solution and the passive potential ranged from −0.15 V to 0.75 V. With the increase in passive potential, the protective performance of the passive film was significantly improved.

In the study of Nikhil and co-workers, ethyl-2-cyano-3-(4-(dimethylamino) phenyl) acrylate (ECDPA) and ZnO nanosheet composites were synthesized and used as corrosion inhibitors for copper in 1 M HCl. ECDPA acted as a barrier to acid molecules after adsorbing copper, delaying the corrosion of copper in hydrochloric acid [78]. EC-AFM analysis confirmed that ZnO nanoparticles promoted the adsorption of ECDPA. Adsorption/deposition of a small number of inhibitor molecules was found when ZnO was not added. After the addition of ZnO, the number of inhibitor molecules on the surface was significantly increased, and the size of the inhibitor molecule became larger as the ZnO calcination temperature increased, as shown in Figure 34.

**Figure 34.** EC-AFM two-dimensional (2D) images of (**a**) polished copper, and copper (**b**) corroded, as well as inhibited by (**c**) ethyl-2-cyano-3-(4-(dimethylamino) phenyl) acrylate (ECDPA), (**d**) EZ3 (ECDPA–ZnO at 300 ◦C), and (**e**) EZ5 (ECDPA–ZnO at 500 ◦C) in 10 min of immersion in 1 M HCl at −0.005 V. Reprinted with permission from Reference [78]; copyright 2019 Elsevier Ltd.

The corrosion resistance tests of ECDPA, EZ3 (ECDPA–ZnO at 300 ◦C), and EZ5 (ECDPA–ZnO at 500 ◦C) showed that the composite had better protection performance than ECDPA alone. The maximum inhibition efficiency of ECDPA was approximately 75%, while the composite could be further improved to 78% (EZ3) and 81% (EZ5). The improved corrosion inhibition performance of ECDPA–ZnO may have been related to the inclusion of ZnO nanoparticles, which promoted the adsorption of ECDPA over Cu.

Shaban et al. [79] studied the inhibition of dibenzylsulfoxide (DBSO) and *p*-chlorobenzohydroxamic acid (*p*-Cl-BHA) on copper corrosion in 0.5 M NaCl and 0.1 M Na2SO4, respectively. In situ AFM was used to observe the corrosion and inhibition process of the electrode surface. When DBSO was not added, the surface became rougher due to the dissolution of copper. In the presence of DBSO, the surface was not subject to severe corrosion and a relatively smooth surface was formed, as shown in Figure 35. Therefore, the authors thought that DBSO inhibited the corrosion of copper in sulfate solutions by converting it to a more stable, less soluble sulfide compound.

**Figure 35.** Morphological changes of copper in solution of 0.1 M Na2SO4 (**a**–**d**) and 0.1 M Na2SO4 + 0.5 mM dibenzylsulfoxide (DBSO) (**e**–**h**), at different times: (**a**,**e**) 0 min; (**b**,**f**) 15 min; (**c**,**g**) 30 min; (**d**,**h**) 45 min. Reprinted with permission from Reference [79]; copyright 1998 Springer.

In the absence of *p*-Cl-BHA, the formation and growth of pitting corrosion were detected. The addition of *p*-Cl-BHA significantly impeded the localized corrosion of copper and inhibited the

production of corrosion products that occurred when *p*-Cl-BHA was not added, as shown in Figure 36. Thus, the formation of a stable complex by adsorbing a corrosion inhibitor on a corroded surface effectively hindered further dissolution of the metal.

**Figure 36.** Morphological changes of copper in solution of 0.1 M NaCl (**a**–**d**) and 0.1 M NaCl + *p*-chlorobenzohydroxamic acid (*p*-Cl-BHA) (**e**–**h**), at different times: (**a**,**e**) 0 min; (**b**,**f**) 15 min; (**c**,**g**) 30 min; (**d**,**h**) 45 min. Reprinted with permission from Reference [79]; copyright 1998 Springer.

Cruickshank et al. [80] studied the anodic dissolution of polycrystalline copper in acid medium with or without corrosion inhibitor using EC-AFM. In 0.5 M H2SO4, the preferential corrosion of some grain surfaces occurred and then dissolved along grain boundaries. The addition of benzotriazole (BTA) formed a protective film that effectively inhibited the dissolution of copper. When the anode potential reached 200 mV, the protective film was stable. However, when the potential was as high as 300 mV, the film underwent local breakdown. In addition, Li et al. [81] investigated the effect of BTA on corrosion inhibition of copper using in situ EC-AFM in 0.01 M NaHCO3. The addition of BTA caused the copper surface to form a BTA film, effectively protecting the copper from erosion. The pitting potential of the copper surface covered by the BTA film could be significantly increased by more than 700 mV. Therefore, the author believed that, in fact, pitting was not a concern in the presence of BAT.

Through the morphology testing of EC-AFM, the adsorption–desorption states of corrosion inhibitors on an electrode surface can be intuitively understood, which is very beneficial for the exploration of electrochemical mechanisms of corrosion inhibitors. The application of EC-AFM in corrosion inhibitor protection is summarized in Table 7.


**Table 7.** Summary of the application of EC-AFM in corrosion inhibitor protection.

#### **4. Conclusions and Perspectives**

This review introduced the evolution, principles, and operation modes of EC-AFM, as well as illustrated its application in corrosion science through specific examples.

In summary, EC-AFM can not only perform real-time in situ research on micro-area corrosion (passivation) in the field of corrosion electrochemistry; it also has higher resolution, which provides detailed information on corrosion phenomena, such as activation and passivation at submicroscopic scales, especially in the early stages of corrosion. The surface topography changes of the target sample can be quantitatively analyzed with the information of surface height changes with applied voltage or time, helping us better study the corrosion process and mechanism. In recent years, EC-AFM was continuously improved in electrolytic cells, imaging modes, and probes. Moreover, EC-AFM was combined with SKPFM, SVET, and SECM to study the surface state of metals, greatly expanding the scope of application.

However, EC-AFM also has some problems that affect research, which brings about difficulties obtaining reliable surface topography features, such as drift during the scanning process, alignment of the laser, changes in the refractive index of the electrochemical medium as the process of corrosion goes on, finding accurate resonant frequencies, etc. In addition, the EC-AFM scanning rate is limited, and some rapid interface reactions cannot be monitored in real time.

In the future, the development of EC-AFM will be toward multi-functionality, high sensitivity, high speed, and high efficiency. The continuous improvement of EC-AFM will help carrying out more in-depth research on the in situ dynamic corrosion process, as well as promoting the study of corrosion science.

**Author Contributions:** Conceptualization, Z.W., M.H., and Y.L.; formal analysis, H.C., Z.Q., and Z.W.; writing—original draft, H.C.; writing—review and editing, H.C. and Z.W. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was financed by the National Key Research and Development Program of China (2018YFB0703500) and the National Natural Science Foundation of China (No. 51971155, No. 51771120, and No. 51304136).

**Conflicts of Interest:** The authors declare no conflicts of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **E**ff**ect of Pre-Corrosion Pits on Residual Fatigue Life for 42CrMo Steel**

#### **Dezheng Liu 1,\*, Yan Li 1, Xiangdong Xie 2,\* and Jing Zhao <sup>1</sup>**


Received: 13 June 2019; Accepted: 1 July 2019; Published: 2 July 2019

**Abstract:** The effect of pre-corrosion pits on residual fatigue life for the 42CrMo steel (American grade: AISI 4140) is investigated using the accelerated pre-corrosion specimen in the saline environment. Different pre-corroded times are used for the specimens, and fatigue tests with different loads are then carried out on specimens. The pre-corrosion fatigue life is studied, and the fatigue fracture surfaces are examined by a surface profiler and a scanning electron microscope (SEM) to identify the crack nucleation sites and to determine the size and geometry of corrosion pits. Moreover, the stress intensity factor varying with corrosion pits in different size parameters is analyzed based on finite element (FE) software ABAQUS to derive the regression formula of the stress intensity factor. Subsequently, by integrating the regression formula with the Paris formula, the residual fatigue life is predicted and compared with experimental results, and the relationship of the stress intensity factor, pit depth, and residual fatigue life are given under different corrosion degrees. The fatigue life predicted by the coupled formula agrees well with experiment results. It is observed from the SEM images that higher stress amplitude and longer pre-corroded time can significantly decrease the residual fatigue life of the steel. Additionally, the research work has brought about the discovery that the rate of crack extension accelerates when the crack length increases. The research in this paper also demonstrates that the corrosion pit size can be used as a damage index to assess the residual fatigue life.

**Keywords:** pre-corrosion pits; residual fatigue life; 42CrMo steel; stress intensity factor

#### **1. Introduction**

Most engineering materials are subject to corrosion, and corrosion research has received much attention from different perspectives [1–7]. The 42CrMo (American grade: AISI 4140) is a commonly used ultrahigh-strength steel. Due to its high strength, toughness, and hardenability, 42CrMo steel is widely used in quenched and tempered heavy forgings to build components such as pressure vessels, gears, vehicle axles, and deep oil drilling rod subs [8,9]. However, their corrosion resistance is relatively poor and this material is prone to corrosion. Corrosion pits can weaken the strength of the structure and decrease its fatigue life [10,11], since the fatigue cracks are easy to nucleate at corrosion pits and propagate rapidly under dynamic loads [12]. Thus, to further understand corrosion effects on the strength and residual fatigue life of 42CrMo material, the stress state around corrosion pits should be reasonably estimated.

In recent years, fatigue performance of pre-corroded metallic alloys has been studied extensively. For example, recent works [13–18] have reported the description of the stress concentration arising from corrosion pits based on the finite element (FE) method. Sharland [13] developed a mechanistic model of the propagation stage of an established pit or crevice to simulate the evolution of pit geometry and to describe the corrosion process. However, Wang [14] indicated that a detailed description of both the chemical reactions and the ionic transport was not included in it [13]. Through the use of interaction function to simulate the effect of the corrosion potential and corrosion current density on the corrosion

process, Xu [15] developed an FE model for the simulation of the mechanoelectrochemical effect of pipeline corrosion. Because the fatigue strength of the metallic alloys decreases differently under the different corrosion depths and stress amplitudes [16], Pidaparti [17] established a new FE model to investigate the stress state around corrosion pits. However, the proposed FE model in [17] cannot be used to depict the growth process of corrosion pits with time. Through the use of the damage tolerance method to predict the corrosion fatigue life for 7075-T6 stainless steel, Huang [18] indicted that the nucleation time of pits and the material constant of crack propagation can affect the residual fatigue life of stainless steel significantly. On the basis of [16–18], Zheng [19] proposed that the decrease of fatigue life depends on the pre-corroded time and the size of pre-corrosion pits. Most recent studies on the fatigue life of steel in corrosion environments only consider the factors such as mass loss rate and corrosion pit depth, and rarely involve the stress intensity factor varying with corrosion pits in different size parameters. The stress concentration caused by corrosion pits can dramatically influence the basic macroscopic mechanical properties of a steel [20]. Despite the extensive research on the corrosion of metallic alloys [21–24], there are few detailed studies of the effect of pre-corrosion pits on residual fatigue life for 42CrMo steel.

In this study, the effect of stress distribution around corrosion pits on the stress intensity factor is firstly investigated. The crack nucleation sites and the geometry of corrosion pits are examined by a surface profiler and a scanning electron microscope (SEM), and then the relationship between the fatigue life and the maximum applied stress is obtained by fatigue tests of different pre-corroded specimens. Based on experiment results, an FE model is conducted to investigate stress distribution around corrosion pits, and regression formula of the stress intensity factor varying with corrosion pits in different size parameters is then derived. Subsequently, by coupling the regression formula with the Paris formula [25], the fatigue performance is predicted by the FE method and then compared with experimental tests. Under the different stress amplitudes and pre-corroded times, the FE predictions are finally validated by comparison with the experiment results.

#### **2. Materials and Methods**

#### *2.1. Material and Specimen Preparation*

To better understand the stress intensity factor varying with corrosion pits in different size parameters and to accurately depict the residual fatigue life of 42CrMo steel subject to corrosion, it is necessary to quantitatively investigate the effects of pre-corrosion levels on the fatigue behavior of 42CrMo steel. Experimental specimens were fabricated using the 42CrMo steel (American grade: AISI 4140), which was provided by Baowu Steel Company (Wuhan, China). Specimens were manufactured in the form of flat bare sheets with a thickness of 6 mm by wire cutting machine, the sizes of specimens are provided in Figure 1. The chemical composition was measured using an FLS980-stm Edinburgh fluorescence spectrometer (Edinburgh Instruments Ltd., Livingston, UK). A comparison of the measured chemical composition of 42CrMo specimens and the standard of GB/T3077-2015 (the national standard of the People's Republic of China for alloy structural steel) for 42CrMo steel is presented in Table 1. It can be seen from Table 1 that the chemical composition of the specimens used in this study meets the requirement of the national standard.

**Figure 1.** Schematic diagram of specimen sizes (mm).


**Table 1.** A comparison of the measured chemical composition of 42CrMo specimens and the standard of GB/T3077-2015 for 42CrMo steel (wt.%).

Twelve smooth samples were numbered and divided into four groups. The accelerated neutral salt spray corrosion experiment was established to achieve the pre-corroded steel plates based on the standard of JIS H 8502.5 [26]. A schematic drawing of the sample preparation and the pre-corroded experiment procedure is shown in Figure 2. The sample preparation was carried out in five steps: (1) samples were manufactured in the form of flat bare sheets with the dimensions in Figure 1 by wire cutting machine; (2) the surface of each sample was burnished by the 1000 grit abrasive paper to dispel burrs; (3) samples were rinsed in distilled water; (4) samples were degreased through the use of acetone solvent; (5) all surfaces of samples were rinsed with deionized water and dried by a blower for the pre-corroded experiment.

**Figure 2.** Schematic drawing of the sample preparation and the pre-corroded experiment procedure.

In this study, the pre-corroded experiment procedure can be devised as four main steps: (1) the pure NaCl crystals of 50 g were weighed by an electronic balance and poured into one liter of deionized water and stirred with glass rod at an ambient temperature of (25 ± 1 ◦C) for 5 min; (2) four groups of numbered samples were alternately immersed in 5% NaCl solution at an ambient temperature of (25 ± 1 ◦C) for 0 h, 24 h, 48 h, and 96 h, respectively; (3) the corrosion products were removed by membrane removal solution (prepared with 100 mL HCl, 100 mL deionized water and 0.6 g C6H12N4); (4) all surfaces of samples were rinsed with deionized water and dried by a blower for pit depth measurement.

#### *2.2. Pre-Corrosion Pit Measurements and Fatigue Test*

To achieve the morphology of the corrosion surface and to determine the size and geometry of pre-corrosion pits, a non-contact Dektak150 surface profiler (Veeco Instruments Shanghai Co., Ltd., Shanghai, China) and an S-4800 scanning electron microscope (Hitachi, Tokyo, Japan) were used in this study. A region of 30 mm × 20 mm (along the directions of longitude and transverse, respectively) for each specimen was approximately arranged at the area of pre-estimated fatigue fracture. It is noted that the rate of corrosion changes steadily with the increase of pre-corroded time [27], thus the relationship between the corrosion time and the depth of pre-corroded pits can be expressed by an exponential

function. Based on reference [28], the coefficient of exponential function can be determined by fitting the measured data, and exponential function expression can be described as:

$$
\varphi = 1.235 \, t^{0.775} \tag{1}
$$

where ϕ is the depth of a pre-corroded pit (μm) and *t* is the pre-corroded time (h).

Based on the standard of ASTM E468-90 [29], the fatigue performance of corroded specimens (24 h, 48 h, and 96 h) were evaluated by a hydraulic universal testing machine (Instron-8803, Norwood, MA, USA) under a cyclic load amplitude with a frequency of 10 Hz. The loading operation was force controlled with a proportional error of ±1%. A group of un-corroded specimens were also tested as a control for fatigue studies. The stress ratio was taken as 0.1, and the maximum applied stresses were 100 MPa, 200 MPa, and 300 MPa, respectively. After testing, the fractured specimens were processed by cutting machine and the morphology of the corrosion surface was observed by surface profiler and scanning electron microscope.

The pre-corrosion pit reduced the fatigue life substantially, particularly the interacting pits and sharp pits [30]. In this study, the Paris theory [25] was applied to analyze the relationship between the morphology of fatigue cracks and the stress state around corrosion pits. According to Paris theory [25], the stress intensity factor of cracks can be described as:

$$\mathbb{K}\_I = \sigma \ \sqrt{\pi a} f(\beta) \tag{2}$$

$$
\beta = \frac{a}{b} \tag{3}
$$

where *KI* is the stress intensity factor of the crack, σ is the external stress under the plane stress state, and *a* is the crack length and *b* is the specimen width.

#### *2.3. FE Model*

The FE analysis model of the 42CrMo steel in the form of a corroded flat sheet with a thickness of 6 mm was created using the HyperWorks software (HyperWorks 11.0, Altair Corp., Troy, MI, USA). On the basis of the experiment test, a single semi-elliptical pit model was conducted, and the symmetry plane of the flat sheet and refined mesh around the pit are shown in Figure 3.

**Figure 3.** The FE analysis model.

The CPS8 quadrilateral element type was adopted in the FE model and cracks were prefabricated on the surface through the use of an assigned seam function that was provided by HyperWorks. The sweeping method [31] was used to generate the mesh in the integral region and the singularity of the mesh was controlled by the mesh regeneration technique [32]. The use of the above techniques [31,32] with local refinement and appropriate local mesh density can improve the precision of the calculation. Subsequently, the FE analysis model with a preset corrosion pit was imported into the ABAQUS software (ABAQUS 6.10, Dassault Systemes Simulia Corp., Johnston, RI, USA) to analyze the stress and fatigue performance of the specimens.

#### **3. Results and Discussion**

#### *3.1. Corrosion Surface Characterization*

When the 42CrMo specimens were immersed in NaCl solution, the Cl element in the solution adhered to the surface of specimens and reacted with the Fe2<sup>+</sup> in the metal to form a soluble clathrate, which resulted in the anodic dissolution of the metal surface and the formation of corrosion pits. With the increase of pre-corroded time, more metal dissolved in the anode and the size of corrosion pit became larger and larger [33]. The specimens pre-corroded for 0 h, 24 h, 48 h, and 96 h were examined by an S-4800 scanning electron microscope and the morphologies of corrosion surface characterization are shown in Figure 4.

**Figure 4.** Corrosion surface characterization: (**a**) pre-corroded for 0 h; (**b**) pre-corroded for 24 h; (**c**) pre-corroded for 48 h; and (**d**) pre-corroded for 96 h.

According to Figure 4, corrosion surface characterization can be clearly observed and the obvious corrosion pits after pre-corrosion can be found within the crack source area in each diagram. As shown in Figure 4a, there was no corrosion pit on the surface of specimens without pre-corrosion. With the pre-corrosion treatment, Figure 4b shows that there were local micro-pits and pits distributed densely. With the increase of pre-corroded time, Figure 4c,d shows that corrosion pits became larger since the small pits were interconnected to form larger corrosion pits. A series of electrochemical corrosion processes occurred when the specimen was immersed in the NaCl solution. In this study, the corrosion process can be generally divided into two stages: 1) the initiation of the pits stage and 2) the pit development stage. In the first stage, adsorption and agglomeration of chloride ions took place at certain weak sites and the corrosion pits formed. In the second stage, the pits developed as a result of the anodic dissolution of the metal [34–36]. Thus, the small pits were interconnected to form larger corrosion pits.

To obtain the effect of the stress distribution around corrosion pits on the stress intensity factor and the residual fatigue life of corroded steel, the size of pre-corroded pits should be accurately assessed and the depth of pre-corroded pits can be measured by a surface profiler. The schematic diagram of pre-corrosion pits under the pre-corroded times 24 h, 48 h, and 96 h is shown in Figure 5. From Figure 5, one can see that the depth of the maximum corrosion pit can be regarded as the radius of a semi-circular surface crack. The depth of the maximum corrosion pit was measured by a surface profiler and an SEM.

**Figure 5.** The schematic diagram of pre-corrosion pits under the different pre-corroded times: (**a**) pre-corroded for 24 h; (**b**) pre-corroded for 48 h; and (**c**) pre-corroded for 96 h.

#### *3.2. Fatigue Test Results and Discussion*

Pre-corrosion not only reduces residual fatigue life, but also leads to the change of the stress intensity factor. Under the cyclic load with a frequency of 10 Hz, the local stress concentration around the corrosion pits accelerated the process of fatigue damage and resulted in the reduced fatigue life of the specimen. Three constant stress levels (100 MPa, 200 MPa, and 300 MPa) were carried out to pre-corroded specimens for 0 h, 24 h, 48 h, and 96 h. The fatigue testing results of specimens with different exposure times are shown in Table 2.

Table 2 shows that the fatigue life decreased obviously with the increase of pre-corroded time under the same constant stress level. Furthermore, the fatigue life decreased approximately by more than 12% when the pre-corroded time was doubled, and the fatigue life decreased approximately by more than 11% when the applied stress level was doubled, indicating that both pre-corroded time and applied stress level can significantly affect the fatigue life of specimens, which shows a good agreement with the previous works [37,38]. Although the external load is normally invariant in practical engineering structures and residual fatigue life usually depends on corrosion surface condition and the nominal

stress level, the raised nominal stress caused by the reduction of the cross-sectional area of components due to corrosion still decreases the fatigue life of steel [39]. After fatigue testing, the fracture surface of the corroded specimens under 100 MPa was observed by SEM, and the results are shown in Figure 6.


**Table 2.** The fatigue testing results of specimens with different exposure times.

(**a**) (**b**)

**Figure 6.** The fracture surface of the corroded specimens under 100 MPa: (**a**) pre-corroded for 0 h; (**b**) pre-corroded for 24 h; (**c**) pre-corroded for 48 h; and (**d**) pre-corroded for 96 h.

Figure 6a shows the fracture surface of the non-corroded specimen under the constant stress of 100 MPa, and it can be seen that the crack propagated along grain boundaries. Figure 6b–d reveal that the crack originated at the tip of the corrosion pit and propagated along the grain boundaries, and the white highlight area phenomena was caused by the deposition of corrosive elements. Due to the existence of corrosion defects, stress concentration occurred around the sharp intrusion at the pit bottom and accelerated the process of fatigue damage.

#### *3.3. FE Investigations*

Based on Equations (1)–(3), ABAQUS was applied to compute the stress intensity factor corresponding to different corrosion pit morphology characteristics, which were depicted in the previous section. The same material properties and loading conditions in Section 2.2 were applied to the FE models, and the computed stress distributions around the crack tip are shown in Figure 7.

The ashy area in Figure 7 represents the yielding zone in specimens. The numbers 1, 2, and 3 in Figure 7 denomination represent the constant stress level of 100 MPa, 200 MPa, and 300 MPa, respectively. From Figure 7, it can be seen that, under the same pre-corroded time, the yielding

zone increased with the increase of applied loads through the transverse comparison. Under the same applied loads, it can be seen that the yielding zone increased with the increase of the size of pre-corrosion pits. Moreover, the FEM investigations revealed that the rate of crack extension accelerated when the crack length increased.

**Figure 7.** The stress distributions around the crack tip. (**a1**), (**a2**), and (**a3**): non-corroded specimen under the different constant stress levels; (**b1**), (**b2**), and (**b3**): 24 h pre-corroded specimen under the different constant stress levels; (**c1**), (**c2**), and (**c3**): 48 h pre-corroded specimen under the different constant stress levels; and (**d1**), (**d2**), and (**d3**): 96 h pre-corroded specimen under the different constant stress levels.

#### *3.4. Regression Formula of the Stress Intensity Factor*

By the adoption of Equations (2) and (3), the stress intensity factor around the crack tip in the process of crack propagation for non-corroded specimens can be calculated and compared with the stress intensity factor that was computed by ABAQUS software to validate the accuracy of the FE simulation results. A comparison of the stress intensity factor by theoretical Paris formula and FE simulation is shown in Figure 8.

**Figure 8.** Comparison of the stress intensity factor by theoretical Paris formula and FE simulation.

The maximum percentage error of the calculated stress intensity factor between the theoretical Paris formula and the FEM is 4.7%, which indicates that the theoretical calculation agreed well with the simulated results. Through the integration of the ashy area around the crack tip in Figure 7, the stress intensity factor of the crack tip for pre-corroded specimens can be obtained and the relationship between the stress intensity factor at the crack tip and the depth of the pre-corrosion pits can be established.

Figure 9 shows the relationship between the stress intensity factor at the crack tip and the size of pre-corrosion pits. By using the numerical fitting method, the regression formula to represent the relationship between the stress intensity factor and the depth of the pre-corrosion pits is derived as:

$$\frac{K\_{IS}}{K\_{l0}} = 0.273e^{3.378\varphi} + 0.731\tag{4}$$

where *K*IS is the pre-corroded stress intensity factor, *K*I0 is the non-corroded stress intensity factor, and ϕ is the depth of corrosion pits. The fatigue life can be estimated according to the Paris formula [25]. The Paris formula is shown as follows:

$$\frac{da}{dN} = \mathbb{C}\Lambda K^m \tag{5}$$

where *da*/*dN* is the fatigue crack propagation, Δ*K* is the stress intensity factor at the tip of cracks, and *C* and *m* are the equation coefficients, respectively. For the 42CrMo steel, the coefficients of *C* and *m* are taken as 4.349 <sup>×</sup> <sup>10</sup>−<sup>12</sup> and 3.07, respectively [40]. Through the integration of Equation (5), the fatigue life of 42CrMo steel can be estimated as:

$$N = \int \frac{da}{\mathcal{C}(\Delta K)^{3.0 \mathcal{T}}} \tag{6}$$

By combining Equations (4) and (6), the relationship between the fatigue life of pre-corroded and non-corroded specimens can be written as:

$$N\_{\mathbb{C}} = \left(0.273 \varepsilon^{3.378\varrho} + 0.731\right)^{-3.07} N\_0 \tag{7}$$

where *NC* is the fatigue life of pre-corroded specimens and *N*<sup>0</sup> is the fatigue life of non-corroded specimens.

**Figure 9.** The relationship between the stress intensity factor and the depth of pre-corrosion pits.

Under different stress ranges and corrosion degrees, the residual fatigue life of pre-corrosion specimens is predicted through the use of Equation (7). The comparison of predicted fatigue life and experimental fatigue life is summarized in Table 3.


**Table 3.** Comparison of predicted fatigue life and experimental fatigue life.

Table 2 shows that the fatigue life predicted by the coupled formula agreed well with experiment results, and the percentage errors between predicted values and experimental values were less than 10%. In the case of low corrosion degrees and low stress levels, the percentage errors were less than 5%. With the increase of pre-corroded time and stress levels, the percentage error was increased to approximately 9.3%. The stress intensity factor at the tip of cracks was calculated on the basis of the linear elastic model and the morphology of the pre-corrosion pits simulated by FEM took pre-set defects in the middle of the specimen, while the practical corrosion pits had the characteristics of the uneven distribution and size of the specimen. Thus, the percentage errors between the predicted values and the experimental values were slightly increased with the increase of pre-corroded time and stress levels.

#### **4. Recommendation for Future Research**

In this research, the effect of pre-corrosion pits on the residual fatigue life of the 42CrMo steel was experimentally and numerically studied. It was demonstrated in this paper that the corrosion pit size can be used as a damage index to assess the residual fatigue life. Therefore, the monitoring of corrosion pit size is important not only to know the corrosion status, but also to predict the residual fatigue life of the steel. With the recent development of structural health monitoring technology [41–44], especially the piezoceramic-based active sensing method [45–47], electromechanical impedance (EMI) approach [48–50], and imaging technology [51,52], it is now possible to monitor corrosion pit number

and size [10]. Future work should include the real-time monitoring of the onset and growth of the corrosion pit and the prediction of the residual fatigue life of the 42CrMo steel specimens. Future works should also focus on the quantitative analysis of corrosion pits to improve the predicted accuracy of the residual fatigue life for the corroded 42CrMo steel.

#### **5. Conclusions**

In summary, this study investigated the effect of pre-corrosion pits on the residual fatigue life of the 42CrMo steel. It was found that higher stress amplitude and longer pre-corroded time significantly decreased the residual fatigue life of the steel. The investigation of the residual fatigue life of pre-corrosion specimens was conducted through the use the regression formula and experimental measures, and the fatigue life predicted by the regression formula agreed well with experiment results. Moreover, it was found that the rate of crack extension accelerated when the crack length increased. It is also demonstrated in this paper that the corrosion pit size can be used as a damage variable to assess the residual fatigue life. The recommendation for future research is to monitor corrosion pit characteristics and to predict the residual fatigue life of the 42CrMo steel.

**Author Contributions:** Conceptualization, D.L.; methodology, D.L. and J.Z.; software, D.L. and Y.L.; validation, Y.L.; formal analysis, X.X.; writing—original draft preparation, D.L.; writing—review and editing, X.X.; visualization, Y.L.; supervision, X.X.; project administration, D.L.

**Funding:** This research was funded by Hubei Superior and Distinctive Discipline Group of "Mechatronics and Automobiles" (No. XKQ2019009).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **E**ff**ect of Hot Deformation Process Parameters on Microstructure and Corrosion Behavior of 35CrMoV Steel**

#### **Qiumei Yang 1, Yajun Zhou 1,2,\*, Zheng Li <sup>1</sup> and Daheng Mao <sup>1</sup>**


Received: 16 March 2019; Accepted: 1 May 2019; Published: 6 May 2019

**Abstract:** Hot deformation experiments of as-cast 35CrMoV steel, with strain rates of 0.01 s−<sup>1</sup> and 10 s<sup>−</sup>1, deformation temperatures of 850, 950, and 1050 ◦C, and an extreme deformation reaching 50%, were carried out using a Gleeble-3810 thermal simulator. Electrochemical corrosion experiments were conducted on the deformed specimens. The microstructure was observed by optical microscope (OM), and the corrosion morphology and corrosion products of the specimens were investigated by scanning electron microscopy (SEM), energy disperse spectroscopy (EDS), confocal laser scanning microscopy (CLSM), and X-ray diffraction (XRD) techniques. The results show that the grain size increased gradually with an increase in the deformation temperature at the same strain rate, whereas the corrosion resistance deteriorated. At the same deformation temperature, the grain size becomes smaller as the strain rate increases, which enhances the corrosion resistance. This is mainly attributed to the fine grains, which can form more grain boundaries, increase the grain boundary area, and accelerate the formation of the inner rust layer at the beginning of corrosion. Moreover, fine grains can also refine the rust particles and enhance the bonding strength between the inner rust layer and the matrix. The denseness and stability of the inner rust layer increases as the corrosion process progresses, thereby improving corrosion resistance.

**Keywords:** thermal deformation parameters; 35CrMoV steel; grain size; electrochemical corrosion

#### **1. Introduction**

It is well-known that 35CrMoV steel is a high-quality steel with good hardenability, creep properties, high strength, and a high fatigue limit. It is generally used to manufacture gear, marine power equipment, advanced turbine blowers and impellers, compressor engines, and other important parts which operate under high stress [1]. Under certain conditions, such as high stress and complex marine conditions (especially for complex marine conditions with strong corrosiveness), Cl− has a strong penetrating power that can cause pitting corrosion of carbon steel and easy penetration of the surface defects. Furthermore, the invasion of Cl− can force the surface of the passivation film to rupture, which will aggravate the stress corrosion cracking failure of the steel [2–4]. Therefore, in addition to the mechanical properties, its corrosion performance also seriously threatens the service life, safety, and efficiency of the marine components.

During thermal processing, different thermal deformation parameters have significant effects on the microstructure and properties of the material [5,6]. Numerous studies have been done into the evolution of the microstructure during thermal deformation in recent years [7–9]. Kingkam et al. [10] investigated the effects of the deformation temperature and strain rate on the dynamic recrystallization behavior of high-strength low-alloy steel. It was observed that an increase of deformation temperature led to an increase of grain size and the effects of the microstructure on corrosion behavior were briefly discussed. Xiao et al. [11] observed that the smallest dynamic recrystallization grains were obtained at a higher strain rate and lower deformation temperature during hot deformation. Moreover, high temperatures and low strain rates are beneficial for grain growth. Lin et al. and Xu et al. established the constitutive equation and process diagram of 25Cr3Mo3Nb steel. It is reported that the change of grain size is closely related to the strain rate and, when the strain rate is 1 s−1, a uniform distribution and fine grains can be obtained [12]. Additionally, Huang [13] studied the dynamic recrystallization of 35CrMo steel and found that it was affected considerably by the deformation temperature and strain rate and was difficult to recrystallize at a lower temperature and higher strain rate. Lin et al. [14] studied the microstructure of a Ni-Fe-Cr-based superalloy by thermal compression experiments and the experimental results were consistent with the established dynamic recrystallization model. Quan et al. [15] obtained fine-grained, equiaxed-crystal re-crystallized structures of ASS alloys generated by the thermal deformation activation of dynamic recrystallization. These studies explored the evolution of the microstructure during the thermal deformation process, but rarely involved changes in microstructure that affected the corrosion resistance.

Some studies have exhibited that grain refinement can significantly improve the corrosion resistance of the material [16]. For instance, Argade and Alvarez-Lopez [17,18] showed that grain boundaries can serve as corrosion barriers and considered the role they play in delayed corrosion. Balyanov et al. believed that the rapid passivation of ultrafine Ti makes ultrafine crystals, which have a stronger corrosion resistance than coarse-grained industrial pure Ti [19]. Aung [20] discovered that small grains can reduce the corrosion rate of the AZ31B magnesium alloy. Zhao discussed the effect of the microstructure on the corrosion performance of the AZ91 alloy, and pointed out that the presence of a second phase and surface film can act as a corrosion barrier to prevent corrosion of the matrix [21–23]. Schino explored the inter-granular corrosion rate of AISI 304 austenitic stainless-steel and found that it decreases with a decrease of grain size in an H2SO4-FeSO4 solution. This is due to the increase of grain boundary per unit-volume and the decrease of Cr consumption caused by carbide precipitation, as the inter-granular corrosion rate goes down with the increase of the grain boundary area [24]. The results of Pradhan et al. showed that a fine-grained microstructure has similar or better corrosion resistance to a coarse-grained structure under the optimum distribution of grain boundary characteristics [25]. In addition, it was suggested that the grain boundary is favorable for passivation, high-density dislocations, and grain boundaries, and that grain refinements are propitious to form passivation film as grain refinement is helpful for the rapid formation of a passivation film on the surface, which can reduce the corrosion rate [26]. It was considered that coarse crystal samples are more susceptible to sensitization and inter-granular corrosion than fine-grained samples because their anode and cathode area are lower [27]. It was reported that the degree of sensitization decreases exponentially with the increase of the grain boundary surface-area [28–30], which means that Cr is diffused from the inside to the grain boundary, thereby reducing the growth of carbide precipitates, delaying the sensitization process, and improving corrosion resistance. The above studies show that the grain size, in both magnesium alloy and stainless-steel, affects the corrosion behavior of the material and that grain refinement can improve the corrosion resistance of the material. However, research on the influence of the microstructure evolution of 35CrMoV steel on the corrosion performance in the process of thermal deformation needs further study. Therefore, by adjusting the thermal processing parameters to study the changes of the microstructure in the thermal deformation process, and then analyzing the influence of grain size on corrosion behavior, more appropriate thermal deformation process parameters are obtained, which can provide technical guidance for obtaining large-size forgings with a fine grain size and good performance.

In this study, as-cast 35CrMoV is taken as the research object and thermal compression experiments are carried out at different strain rates and temperatures. The microstructure of 35CrMoV steel after thermal deformation is characterized by optical microscope (OM) and the grain sizes of specimens with different thermal deformation parameters are obtained by a professional micro-image analysis system. The corrosion behaviors of different grain-size specimens obtained under different thermal deformation process parameters are studied in a 3.5 wt% NaCl solution through a three-electrode system. Furthermore, the corrosion performance of the samples is evaluated by electrochemical impedance spectroscopy (EIS) and TAFEL curves, and the corrosion morphology are observed by the energy disperse spectroscopy (EDS), confocal laser scanning microscopy (CLSM), and Hyperfield 3D microscopic systems. Therefore, we can determine the influence of grain size on corrosion resistance and reveal the influence mechanism of the microstructure, obtained under different parameters, on corrosion resistance.

#### **2. Materials and Methods**

#### *2.1. Materials*

Our research material, 35CrMoV steel, is a typical low-alloy steel for industrial use and its chemical composition is shown in Table 1. The as-cast 35CrMoV steel specimens of dimensions Φ 400 × 800 mm and hot compression specimens with size Φ 10 × 15 mm were cut at 1/2R (R is the radius of the as-cast, 200 mm) by a wire cutting machine. Then, the thermal compression test was carried out by using the Gleeble-3810 thermal simulator. The graphite sheet was placed between the two ends of the specimens, as well as the contact surface of the pressure head, before the beginning of the experiment, and lubricant was applied onto the two ends of the specimens so that the specimens could be subjected to uniform deformation during heating and compression. Before the beginning of the experiment, the graphite sheet was placed between the two ends of the specimens and the contact surface of the indenter, and lubricant was applied onto the two ends of the specimens to keep the temperature uniformly distributed during heating and compression. The process of thermal compression is shown in Figure 1. First, it was heated to 1200 ◦C, at a heating rate of 10 ◦C/s−1, for 5 min and then cooled to the deformation temperature (850 ◦C, 950 ◦C, 1150 ◦C), at a cooling rate of 10 ◦C/s−1, with heat preservation for 1 min to eliminate the temperature gradient of the specimens. Then, an extreme deformation, reaching 50%, was carried out at a strain rate of 0.01–10 s−1. Finally, the specimens were quenched in water immediately after compression and the same heat treatment was carried out, the specimens were heated to 850 ◦C for 30 min, then oil-cooled, re-heated to 620 ◦C for 30 min, and finally, were water-cooled.


**Table 1.** Composition of GB 35CrMoV steel (wt%).

**Figure 1.** Schematic representation of hot compression test.

67

#### *2.2. Microscopic Analysis*

The microstructure of the specimens was observed by an Olympus DSX500 optical microscope, and the average grain size was calculated by an artificial cut-off method, according to the standard of GB/T6394-2002. Before the metallographic experiment, the specimens, with a diameter of <sup>10</sup> <sup>×</sup> 5 mm2, were taken along the axial direction of the compressed specimens by a wire cutter, and then they were ground, from coarse to fine sandpaper, on a grinding machine and polished with diamond spray until the surface of the specimens was bright. On this basis, the specimens were corroded in a self-made corrosion solution (2.5 g picric acid + 50 mL distilled water + 2 g dodecyl benzenesulfonic acid, sodium salt), heated, and kept at 60–80 ◦C for 4–8 min. After the corrosion was finished, the specimens were removed from the surface with cotton, washed with distilled water, and dried with hot air, so that the microstructure of the specimens could be observed under an optical microscope.

#### *2.3. Electrochemical Measurements*

Three dynamic potential polarization experiments were performed on the electrochemical workstation (Chi660e, CH Hua, Shanghai, China) using a three-electrode corrosion measurement system. Before the experiment, the specimens were ground to 1500# on the grinder and the polishing cloth was mechanically polished with diamond spray polish until the surface was smooth and free from scratches. Finally, it was washed with distilled water, ethanol, and then dried with hot air. The potentiodynamic polarization was measured by a three-electrode system in a 600 mL corrosion medium cell containing 3.5 wt% NaCl with a saturated calomel (Ag/AgCl) electrode as a reference electrode and platinum-plate as an auxiliary electrode. The surface area of the specimens exposed to the solution was 10 <sup>×</sup> 5 mm2. The specimens were immersed in the corrosive medium for 2400 s to obtain a stable open-circuit potential and, on this basis, electrochemical impedance spectroscopy (EIS) was carried out with a frequency range of 0.01–10<sup>5</sup> Hz and an AC excitation signal amplitude of 10 mV, which can obtain the phase angle diagram of Nyquist and Bode. Then, the appropriate equivalent circuit model was selected and the EIS results were found by equivalent-fitting with the ZSimpWin software. Ultimately, the potentiodynamic polarization curve was measured in the range of −1 to 0 V vs. SCE, at a scanning rate of 0.5 mVS<sup>−</sup>1, and the corrosion potential (Ecorr), corrosion current density (Icorr), anode slope (Ba), and cathodic slope (Bc) were obtained by Tafel fitting.

#### *2.4. Corrosion Surface Analysis*

In order to better observe the corrosion morphology of the specimen surface and analyze the elements of the corrosion products, the surface of the specimens was first dried with a compressed air gun and then was observed by scanning electron microscope (SEM) (EvoMA 10C Zeissi Jena, Oberkochen, Germany). Moreover, the chemical elements of the surface corrosion products were analyzed by energy disperse spectroscopy (EDS) and the surface morphology of corrosion specimens is obtained by using Hyperfield 3D microscopic systems (Keyence VHX-5000, Osaka, Japan) and confocal laser scanning microscopy (CLSM) (Zeiss Axio LSM700, Oberkochen, Germany).

#### **3. Results and Discussion**

#### *3.1. Microstructure*

The microstructure, after hot deformation at different deformation temperatures and strain rates, is shown in Figure 2. The deformation temperatures and strain rates have a significant effect on the microstructure of the 35CrMoV steel. There is a dynamic recrystallization structure in all of the samples and the grain size distribution is not uniform, as displayed in Figure 2. A statistical analysis of the grain size in Figure 2 was carried out and the statistical results are shown in Figure 3. At the strain rate of 0.01 s<sup>−</sup>1, the grain size increased gradually from 12.5 to 15.4 μm when the deformation temperatures increased from 850 to 1050 ◦C (increasing by 20%), which is consistent with the law at the strain rates of 10 and 0.01 s−1. The results show that, at the same strain rate, the grain growth

with the deformation temperature quickly increases, resulting in coarse grains. This can be attributed to higher deformation temperature making the dislocation motion more intense and more favorable to grain boundary migration, which enhances the diffusion and dislocation slip of vacancy atoms, increasing the nucleation of re-crystallization and the dynamic re-crystallization rate. As a result, small grains are continuously swallowed by large grains and the microstructure is coarsened. When the strain rates are increased from 0.01 to 10 s−<sup>1</sup> at the deformation temperature of 850 ◦C, the grain sizes decrease gradually from 12.5 to 10.7 μm, decreasing by 16%. It can be noticed that, as the strain rates are increased from 0.01 to 10 s−<sup>1</sup> at the deformation temperatures of 950 and 1050 ◦C, the grain change is consistent with that at 850 ◦C, which indicates that an increase in strain rate is conducive to grain refinement and acquiring a fine-grained structure at the same deformation temperature. This is mainly due to larger strain rates accelerating the deformation of the samples, and so more strain storage-energy is generated, making the dynamic re-crystallization nucleation rate activate [31]. Thus, there are more phase deformation nuclei and a higher nucleation rate. Moreover, the re-crystallized grain does not have enough time to grow in this case, which reduces the dynamic recovery rate. In addition, the higher strain rate leads to enhanced accumulated strain energy and dislocation density in grains, which makes the dynamic re-crystallization easier to nucleate. It can also be seen from Figure 2 that the number of re-crystallized grains formed along the grain boundary at a strain rate of 10 s−<sup>1</sup> is higher than that at a strain rate of 0.01 s−1. Furthermore, the dynamic recrystallization rate is lower at lower-deformation temperatures, which can be attributed to the decrease of the grain boundary mobility at low deformation temperatures and high strain rates, as well as a tendency toward incomplete dynamic re-crystallization.

**Figure 2.** Microstructure of 35CrMoV steel at different deformation temperatures and strain rates: (**a**) 0.01 s<sup>−</sup>1, 850 ◦C; (**b**) 0.01 s<sup>−</sup>1, 950 ◦C; (**c**) 0.01 s<sup>−</sup>1, 1050 ◦C; (**d**) 10 s<sup>−</sup>1, 850 ◦C; (**e**) 10 s<sup>−</sup>1, 950 ◦C; (**f**) 10 s<sup>−</sup>1, 1050 ◦C.

**Figure 3.** Grain size of 35CrMoV steel at different deformation temperatures and strain rates.

#### *3.2. Electrochemical Analysis*

#### 3.2.1. Potentiodynamic Polarization Curve

In this section, the effects of grain sizes, obtained under different process parameters, on the corrosion resistance of 35CrMoV steel are discussed by potentiodynamic polarization experiments. Moreover, the electrochemical impedance spectroscopy and polarization curves are measured under the steady open-circuit potential (OCP). The change curve of an open-circuit potential, measured in 3.5 wt% NaCl solution, with respect to immersion time, is shown in Figure 4. It can be seen from Figure 4 that the change in the OCP of the studied samples follows a similar trend. At the beginning of the experiment, the potential decreases rapidly and all of the samples slowly reach the OCP when the immersion time is 2400 s. A steady OCP is observed at length.

**Figure 4.** The open-circuit potential (OCP) curves of 35CrMoV steel in 3.5 wt% NaCl solution: (**a**) 0.01 s<sup>−</sup>1, 850–1050 ◦C; (**b**) 10 s<sup>−</sup>1, 850–1050 ◦C.

Figure 5 shows the potentiodynamic polarization curve of 35CrMoV steel immersed in 3.5 wt% NaCl solution after 40 min. It can be easily seen from Figure 5a,b, that the polarization curves obtained at different strain rates and deformation temperatures are similar in morphology and the

Icorr displacement is obvious. There is no passivation and the anode region is an active dissolution of metal, which are the results of the characteristic adsorption of Cl− onto the surface of the specimen, preventing the formation of the passivation film [1]. In addition, the Ecorr shift of the polarization curve and obvious changes in the anode region can also be observed, where the anode region represents the dissolution of the matrix at high potential and the cathode region delegates the cathodic hydrogen evolution reaction associated with water reduction [12].

**Figure 5.** Polarization curves of 35CrMoV steel in 3.5 wt% NaCl solution: (**a**) 0.01 s<sup>−</sup>1, 850–1050 ◦C; and (**b**) 10 s<sup>−</sup>1, 850–1050 ◦C.

Tafel extrapolation is a rapid and effective method used to discuss the corrosion trend and behavior, and the corrosion rate obtained from it is mainly related to the initial surface corrosion [32]. The values of Ecorr, Icorr, Ba, and Bc obtained from the polarization curves using Tafel extrapolation means (as shown in Figure 5) are listed in Table 2. It can be observed in Figure 5 and Table 2 that Icorr increases with an increase of deformation temperature at the same strain rate, and Icorr at the deformation temperature of 850 ◦C is 2.4 μA/cm2, whereas Icorr at the deformation temperature of 1050 ◦C is 5.4 μA/cm2. In comparison, the values Icorr obtained from 850 and 1050 ◦C increased by 2.2 times as much. Furthermore, Icorr increased from 2.3 to 5.3 μA/cm2, which is an increase of 2.3 times, at 10 s<sup>−</sup>1. Therefore, it can be concluded that the higher the deformation temperature of the hot working is, the more easily corrosion occurs. Besides, at the same deformation temperature, Icorr decreases with an increase in strain rate and the Icorr at 850 ◦C and 10 s−<sup>1</sup> is smaller than that at the strain rate of 0.01 s−1. It is well known that the smaller the value of Icorr is, the slower the corrosion reaction rate and the higher the corrosion resistance of the material. Accordingly, this shows that the corrosion resistance of 35CrMoV steel can be improved by obtaining smaller grains at high strain rates and low deformation temperatures.


**Table 2.** Electrochemical test results of 35CrMoV steel.

#### 3.2.2. Electrochemical Impedance Spectroscopy characteristics

Generally, EIS is an effective technique for analyzing and studying the corrosion reaction between the structure of the alloy oxide film and the electrode interface, which can be used to evaluate the corrosion performance of an alloy. Figure 4 shows the curve of the OCP when measured by different specimens. For example, the stable OCP at 850 ◦C and 0.01 s−<sup>1</sup> is about 0.6509 VSCE and the EIS is recorded in the frequency range 0.1–100,000 Hz at the steady potential (0.6509 VSCE). The EIS of 35CrMoV steel is represented by the Nyquist plots and phase angle diagram displayed in Figures 6 and 7. It can be seen from Figure 6 that the EIS of the studied 35CrMoVsteel can be greatly influenced by different deformation temperatures and strain rates. This is because the grain sizes of the specimens, acquired with the different thermal deformation parameters, showed in a comparison that the fine grains can make the grain-boundary area increase and the corrosion rate is accelerated at the early stage of corrosion, resulting in the formation of a dense oxide layer and a delaying and deepening of the corrosion. Consequently, the results show that the grain sizes obtained with the different thermal deformation parameters have a significant effect on corrosion resistance.

**Figure 6.** Nyquist plots of the 35CrMoV steel in 3.5 wt% NaCl solution: (**a**) 0.01 s<sup>−</sup>1, 850–1050 ◦C; (**b**) 10 s<sup>−</sup>1, 850–1050 ◦C.

**Figure 7.** The Bode plots of 35CrMoV steel in 3.5 wt% NaCl solution: (**a**) 0.01 s<sup>−</sup>1, 850–1050 ◦C; (**b**) 10 s<sup>−</sup>1, 850–1050 ◦C.

The radius of the capacitive reactance arc is positively correlated with its corrosion resistance [21,33–35], the larger the arc radius, the lower the corrosion rate of the specimens in the corresponding solution [36–38]. It can be observed from Figure 6 that all samples have similar semi-circular shapes, where the arc lines are similar and their diameters are different, indicating that the corrosion rate of the specimens is different but the corrosion mechanism is the same [20]. In addition, a larger arc radius can be seen at 850 ◦C, with strain rates of 0.01 and 10 s−1, which indicates that its corrosion resistance is better. At the same deformation temperature, the arc radius increases gradually with an increase of strain rate, which implies that the corrosion resistance is better (at the same strain rate), the arc radius decreases with an increase of deformation temperature, which indicates that the corrosion resistance is worse. Moreover, the arc-curve radius also reflects the impedance of the electron transfer process onto the electrode surface. It can be concluded that the hindrance can be enhanced and the corrosion rate can be decreased by a larger arc radius. As for metals, it can be noticed that the large resistance to electron transfer means that the gain and loss of electrons does not occur easily, indicating that the metals are difficult to corrode. Therefore, it is clear from Figure 6 that it is most difficult for corrosion to occur at 10 s−<sup>1</sup> and 850 ◦C, the result with the largest arc radius, which shows that a higher strain rate and lower deformation temperature can obtain smaller grains and better corrosion resistance.

Figure 7 shows the phase angle plots of the 35CrMoV steel with different strain rates and deformation temperatures. The phase angle plots can be divided into the high-, middle-, and low-frequency regions. From Figure 7 it can be seen that the phase angle in the high-frequency region (100–1000 KHz) is small (close to zero), indicating that the impedance of this frequency range is mainly a solution impedance, and that the resistance behavior is independent of the time constant. The phase angle of the intermediate-frequency region (1–1000 Hz) reaches the maximum value, which is a typical characteristic of solubilization. The maximum value of the phase angle of the samples at 0.01 s−1, 850 ◦C and 10 s−1, 850 ◦C moves slightly towards the low-frequency direction. It was found that the capacitance of the double layer is increased [39]. Meanwhile, the presence of a wide angular front in the intermediate- and low-frequency regions implies an interaction between the two relaxation processes, as shown in Figure 7. As per the above analysis, it can be concluded that the two time constants are the reflection of the charge transfer resistance and the corrosion product resistance [40]. In addition, the impedance value, |Z|, of the low-frequency region (0.01–1 Hz) shows the impedance of corrosion reaction, which is one of the parameters used to evaluate the corrosion performance. It indicates that the larger |Z| is, the better the corrosion resistance [41–43]. Additionally, it can be seen from Figure 7 that the impedance value of the specimens decreases with an increase in the deformation temperature, which indicates that the corrosion resistance of the sample surface decreases. Furthermore, at the same deformation temperature, the impedance value increases with the increase of the strain rate. This shows that the corrosion resistance of the sample is stronger. Consequently, the impedance value, |Z|, of the specimen, at 10 s−<sup>1</sup> and 850 ◦C, is observed in Figure 7, which is consistent with the results of the polarization curve, indicating the corrosion resistance is at its best and the corrosion product film on the surface is performing better.

In order to further quantitatively analyze the electrochemical corrosion behavior of the 35CrMoV steel in 3.5 wt%NaCl solution, an equivalent electrical circuit (R(Q (R (QR)) is established to fit the EIS spectra, which can well-express the corrosion mechanism of the metal interface or solution. The equivalent circuit consists of two time constants [44], as shown in Figure 8. The EIS, fitted by the ZsimpWin software (Buokamp, MA, USA), and the fitting results are shown in Table 3. The five different elements used in the equivalent circuit have different physical meanings and can be divided into three different categories. The first part, Rs, is the solution resistance of the electrolyte, which is used to characterize the kinematic velocity of the chloride produced on the surface of the specimens in the electrolyte. In the second part, Qf and Rf are used to imply the capacitance and resistance of the corrosion product respectively, which can reflect the diffusion of ions in the corrosion product layer and the formation of the corrosion product on the substrate surface during electrochemical corrosion. Additionally, Rf is an important parameter used to characterize the protective effect of the oxide layer. It can be found that a larger corrosion product resistance means better protection of the oxide layer and a greater resistance to ion movement. In the third part, Qdl and Rct are used to characterize the

double-layer capacitance and charge transfer resistance of the reaction interface respectively, which are related to the electrochemical corrosion reaction between the sample matrix and the electrolytic solution. Furthermore, Q (in the second and third parts) is a constant phase angle element (CPE), which describes the physical quantity when the parameter of interface capacitance, C, deviates due to dispersion effects. In general, the reactions in the second and third parts occur in the middleand low-frequency regions of EIS, which are related to the corrosion resistance of the specimens. The corrosion resistance of the samples is evaluated by combining the corrosion product resistance Rf with the charge transfer resistance Rct (Rcorr = Rf + Rct) [1,33,35,40]. Generally, the corrosion resistance enhances as Rcorr increases. The polarization resistances (Rcorr) of the two groups of all of the samples studied are shown in Figure 9. It can be seen from Figure 9 that Rcorr decreases with an increase of 35CrMoV steel thermal deformation temperature at the same strain rate, indicating that the corrosion rate is accelerated. This is because an increase in the 35CrMoV hot deformation temperature is conducive to grain growth, resulting in a coarse grain size and lower corrosion resistance. At the same temperature, Rcorr increases with an increase in the strain rate and, more importantly, the highest Rcorr appears at 10 s−<sup>1</sup> and 850 ◦C, indicating that the fine grains obtained at high strain rates and low deformation temperatures can make the oxide layer denser and improve the corrosion resistance.

**Figure 8.** The equivalent electrical circuit of the 35CrMoV steel in 3.5 wt% NaCl solution.

**Figure 9.** The corrosion rate of the 35CrMoV steel in 3.5 wt% NaCl solution.


**Table 3.** Spectra fitting results of experimental 35CrMoV steel in 3.5 wt% NaCl solution.

#### *3.3. Corrosion Products and Corrosion Mechanism Analysis*

In order to further discuss the effect of the grain sizes, obtained under different deformation conditions, on the corrosion resistance of the specimens, the two- and three-dimensional morphology of the specimens surface is observed, on the basis of the potentiodynamic polarization experiment. The two-dimensional corrosion morphology of four typical specimens with deformation temperatures of 850 ◦C and 1050 ◦C, at the strain rates of 0.01 s−<sup>1</sup> and 10 s−<sup>1</sup> respectively, immersed in 3.5 wt% NaCl solution for 1 h and placed for 3 days, is depicted in Figure 10. It can be observed from Figure 10 that the surface of the specimens display local corrosion, where the corrosion holes are in the direction of depth. The black and yellowish-brown areas are corrosion areas, and the bright white areas are un-corroded areas. The corrosion pits on the surface are covered with rust spots and corrosion products, as well as corrosion pits of different sizes. At the same strain rate, with an increase in the 35CrMoV hot deformation temperatures, the local corrosion gradually expands and connects to a larger area of corrosion. Furthermore, the corrosion area becomes smaller and the corrosion degree weakens with an increase of the strain rate at the same deformation temperature. The three-dimensional morphology of the a , b , c , and d (see Figure 10) regions under the CLSM is shown in Figure 11. The size of the etch pits are measured, along the Y direction of the specimens, with the CLSM. The etch pit depths of (a ), (b ), (c ), and (d ) in Figure 10 are 17.17, 49.55, 13.64, and 30.45 μm, respectively. At the same strain rates (0.01 s−<sup>1</sup> and 10 s−1), the surface is locally corroded and there are corrosion pits at the deformation temperatures of 850 ◦C and 1050 ◦C, but the corrosion pit at the deformation temperature of 1050 ◦C is large and deep, which indicates that the corrosion resistance is poor at higher deformation temperatures. This is as the dynamic re-crystallization rate of the sample steel increases at high deformation temperatures, and the re-crystallized large grains continue to swallow small grains and grow. Additionally, at the same deformation temperature, the corrosion surface is smoother and the corrosion pit decreases with an increase of strain rate, which is consistent with the characterization in Figure 10. This is because the structure of the double layer at the interface between the matrix iron and the corrosion product film is prone to preferential adsorption of Cl−, which results in the surface deposition of Cl− fluids, which then combine with cations on the oxidation film to form soluble chloride and corrosion pits. However, the PH value of the corrosion pit decreases with the continuous hydrolysis of chloride, as well as the dissolution of the anode metal of the corrosion pit, and the external Cl− is invaded into the corrosion pit through the corrosion product film, which makes the corrosion proceed further. In this cycle, the deepening and expansion of the corrosion pit depth are the results of the corrosion catalysis of Cl−.

**Figure 10.** Corrosion morphology of the 35CrMoV steel: (**a**) 0.01 s<sup>−</sup>1, 850 ◦C; (**b**) 0.01 s<sup>−</sup>1, 1050 ◦C; (**c**) 10 s<sup>−</sup>1, 850 ◦C; (**d**) 10 s<sup>−</sup>1, 1050 ◦C.

**Figure 11.** Three-dimensional corrosion morphology of the 35CrMoV steel: (**a**- ) 0.01 s<sup>−</sup>1, 850 ◦C; (**b**- ) 0.01 s<sup>−</sup>1, 1050 ◦C; (**c**- ) 10 s<sup>−</sup>1, 850 ◦C; (**d**- ) 10 s<sup>−</sup>1, 1050 ◦C.

Figures 12 and 13 show the surface morphology of the specimens after polarization in 3.5 wt% NaCl solution. It can be observed from Figures (a), (b), and (c) in Figures 12 and 13 that there are cracks and platelet crystal-like corrosion product films on the surface of the specimens. At the same strain rate, the surface corrosion product-coverage area increases and the corrosion product accumulates with an increase of the deformation temperature. However, at the same deformation temperature, the accumulation of corrosion products decreases because of an increase in the strain rate, indicating

that the corrosion degree is weakened. The cracks on the corroded surface are displayed in Figure 12 and in Figure 13a –c , and are due to dehydration after the sample immersion test [45]. The corrosion cracks are formed at the initial stage of corrosion which, with an increase of temperature, become wider, deeper, and looser at the same strain rate, and the corrosion products fall off slightly, indicating that corrosion is deepening. However, at the same deformation temperature, the corrosion cracks on the surface of the specimens become narrower, shallower, denser, and less detached as the strain rate increases, indicating that the corrosion rate is weakened. The platelet-like corrosion products (Figures 12 and 13d –f ) generated on the surface are due to the deepening of the surface corrosion, resulting in rust being generated and the corrosion products accumulating. In other words, with an increase of deformation temperature, the gap in platelet corrosion products becomes wider and looser at the same strain rate, which is more likely to cause oxidative corrosion on the surface, favorable to the further invasion of the corrosive medium which accelerates the corrosion. Furthermore, an increase of the deformation temperature is beneficial to grain growth, which then affects the corrosion rate at the same strain rate. Additionally, the cracks are narrower and the platelet-corrosion product film is densified and uniform at the same deformation temperature, due to strain rate increases. The deformation process becomes rapid, the dynamic recovery rate decreases, and the re-crystallized grains have not yet fully grown, resulting in smaller grains.

**Figure 12.** Scanning electron microscopy (SEM) corrosion morphology of the 35CrMoV steel: (**a**,**a**- ,**d**- ) 0.01 s<sup>−</sup>1, 850 ◦C; (**b**,**b**- ,**e**- ) 0.01 s<sup>−</sup>1, 950 ◦C; and (**c**,**c**- ,**f**- ) 0.01 s<sup>−</sup>1, 1050 ◦C.

The results show that the impedance modulus at the EIS (Figure 6) is the appropriate parameter to characterize the protective performance of the corrosion product film [46–48]. From Figure 6a, it can be seen that no scattering fluctuations were observed in all of specimens at 0.01 Hz, which indicates that the corrosion product film has great stability and can effectively protect the matrix from Cl− in 3.5 wt% NaCl solution, as well as preventing further corrosion [41].

**Figure 13.** SEM corrosion morphology of the 35CrMoV steel: (**a**,**a**- ,**d**- ) 10 s<sup>−</sup>1, 850 ◦C; (**b**,**b**- ,**e**- ) 10 s<sup>−</sup>1, 950 ◦C; and (**c**,**c**- ,**f**- ) 10 s<sup>−</sup>1, 1050 ◦C.

The surface of the specimens was scanned by EDS and two groups of specimens, with different strain rates at deformation temperature 850 ◦C, were selected for analysis. Figure 14 shows the corrosion morphology and EDS analysis of the surface of the 35CrMoV steel at deformation temperature 850 ◦C at 0.01 and 10 s<sup>−</sup>1. Both of the specimens can be observed to have corrosion cracks (Figure 14(a1,b1)) and platelet-like products (Figure 14(a2,b2)) [20], produced on the surface of the 35CrMoV steel. Additionally, there is an obvious accumulation of massive corrosion products, and the thickness of the corrosion layer is uneven. The corrosion cracks (Figure 14(a1,b1)) of the specimens all contain O, Fe, C, and Cr, as revealed by EDS, and the oxygen content on the surface decreases obviously with the increase of the strain rate. Meanwhile, the surface of the platelet-like product (Figure 14(a2,b2)) is mainly composed of O, Fe, C, and Cr, and the oxygen content in 0.01 s−<sup>1</sup> is obviously higher than that in 10 s<sup>−</sup>1, indicating that the corrosion product is mainly oxide. Furthermore, the content of Cr in 10 s−<sup>1</sup> obviously increased and the presence of Cr elements on the surface of the corrosion product film can effectively prevent the deep erosion of Cl−, improve the chemical stability of the corrosion product film, and prevent further corrosion.

**Figure 14.** Corrosion morphology and energy disperse spectroscopy (EDS) analysis of the 35CrMoV steel at the same temperature (850 ◦C) and at different strain rates: (**a**) 0.01 s<sup>−</sup>1; and (**b**) 10 s<sup>−</sup>1.

X-ray diffraction analysis was conducted to determine the types of corrosion products (see Figure 15a) and the matrix (see Figure 15b) after corrosion on the surface of the samples. The energy spectrum analysis results of the 35CrMoV steel matrix and corrosion products after corrosion are displayed in Figure 15. It can be seen from Figure 15 that the surface of the matrix after heat treatment mainly contains Fe, and some studies have shown that its structure is tempered sorbite [49,50]. In summary, the corrosion products on the surface after corrosion are mainly FeOOH. Combined with the EDS analysis of Figure 14, it can be concluded that O, Fe, and C are the main components of the corrosion product. Therefore, it can be inferred that the cathode in the corrosion solution is depolarized by oxygen, due to adequate oxygen supply, and the surface is covered by a thin water film.

**Figure 15.** X-ray diffraction (XRD) analysis of the 35CrMoV steel: (**a**) corrosion products; and (**b**) matrix.

In the early stages of corrosion, Cl− adsorbed on the metal surface play an erosion role and Fe is dissolved from the anode to form Fe2+, combined with Cl<sup>−</sup> to form FeCl2·4H2O [51], and further decomposed to form Fe(OH)2. However, Fe(OH)2 is unstable and decomposed into FeO or oxidized to FeOOH by the O2 dissolved in the water film (see Equations (1) and (2)) [1,52–54]. In addition, some studies have shown that Fe (OH)2 can continue to be oxidized and dehydrated to form Fe2O3 and Fe3O4, which are both dense and difficult to decompose. Accordingly, the reduction in corrosion rate is the result of the Fe2O3 and Fe3O4 hindering the diffusion of oxygen and Cl−, which exist in the inner rust layer [55]. According to the Evans theory, the corrosion products will partially dissolve due to the strong erosive Cl− in the solution. Many cracks will occur in the rust layer, which can provide a gap between the corrosion media (such as O2) and spread, causing the corrosion to continue.

$$\text{Fe}^+ + 2\text{OH}^- \rightarrow \text{Fe}(\text{OH})\_2 \tag{1}$$

$$4\text{Fe(OH)}\_{2} + \text{O}\_{2} \rightarrow 4\text{FeOOH} + 2\text{H}\_{2}\text{O} \tag{2}$$

Grain refinement leads to an increase in grain boundary area per unit volume at lower deformation temperatures and higher strain rates. It is well known that the potential at grain boundaries is lower than inside the grain, which is as the crystal defect density at the grain boundary is large. In addition, the existence of a potential difference (PD) constitutes the grain–grain boundary corrosion micro-battery. As a result, the grain boundary (serving as an anode) has priority in corrosion. In the same corrosion environment, there is a certain potential difference between the grain and the grain boundary. The local anodic corrosion current density experienced by the grain boundary is relatively small when corroded, so it will not intrude very deep holes and cracks, and it will give the rust layer a good compactness. The grain size directly affects the corrosion rate at the beginning of corrosion. In contrast to the coarse-grain rate, the fine-grain corrosion rate is faster, which is mainly due to the fact that grain refinement increases the grain boundary area, and the grain boundary is increased in the microstructure. Besides this, the grain boundary is a high-active zone, which may make it vulnerable to pitting corrosion, leading to surface unevenness and increases in the anode surface area, and causing the electrochemical reaction to proceed rapidly [56]. This would result in a larger corrosion area, heavier corrosion, and faster anodic dissolution, which is conducive to the rapid formation of a protective inner rust layer. Thus, a finer-grain of the specimens is more favorable to the flatness of the rust layer/matrix interface in the specimens and a more-uniform dissolved matrix. At the same time, the rust particles are refined and there is a reduction of the cracks and pores in the rust layer, which can cause the matrix and rust layer to combine firmly. Furthermore, the rust layer becomes thicker, denser, and more stable and the alloying elements on the matrix surface play an important role in the development of corrosion [57,58], resulting in an increase in the self-corrosion potential of the sample while enhancing the corrosion resistance and reducing the dissolution rate of the anode. In addition, the current carrying density of the oxide film was reduced and the dissolution of the oxide film slowed down due to grain refinement [59,60], accordingly improving the corrosion resistance [61].

#### **4. Conclusions**

The microstructure of 35 CrMoV steel under different hot deformation conditions and the short-term corrosion behavior in 3.5 wt% NaCl solution was discussed, and some important conclusions, as follows, were obtained:

(1) Deformation temperatures and strain rates have an important influence on the microscopic structure. At the same strain rate, the grain size increased with an increase in deformation temperature and decreased with an increase in strain rate at the same strain temperature. Amongst all of the hot deformation parameters studied, the grain size at a strain rate of 10 s−<sup>1</sup> and a deformation temperature of 850 ◦C was the smallest (10.7 μm), indicating that a finer-grain structure can be obtained at lower deformation temperatures and higher strain rates.

(2) The difference in grain sizes had a significant effect on the corrosion resistance of 35CrMoV steel. At the same strain rate, the grain size augmented with an increase in temperature and the corrosion resistance decreased with an increase in the grain size, within a certain corrosion time. Furthermore, the corrosion resistance was the worst when the grain size was the largest (15.4 μm). At the same deformation temperature, the higher the strain rate, the smaller the grain size, and therefore, the better the corrosion resistance of the specimens within a certain period of time. The smallest grain size (10.6 μm) was observed at a strain rate of 10 s−<sup>1</sup> and a deformation temperature of 850 ◦C, resulting in the lowest corrosion rate, which indicates low-deformation temperature and high-strain rate can

achieve the effect of refining grains and improving corrosion resistance. However, the long-term corrosion trends of these materials need to be studied further.

(3) The differences in the corrosion resistance of 35CrMoV steel were related to the grain boundary change after grain refinement. Grain refinement increased the grain boundary area and the grain boundary, which accelerated the formation of a protective oxide film at the initial stage of corrosion, improved the stability of the oxide film, and prevented the corrosion from proceeding deeply, thereby improving corrosion resistance.

**Author Contributions:** Conceptualization, Q.Y.; Data curation, Z.L.; Formal analysis, Z.L.; Funding acquisition, Y.Z. and D.M.; Methodology, Q.Y.; Project administration, Y.Z.; Resources, D.M.; Writing—original draft, Q.Y.; Writing—review & editing, Q.Y. and Y.Z.

**Funding:** This research was funded by the national 973 project of China, grant number No.2014CB046702 and the experimental cost were funded by the national 973 project of China.

**Acknowledgments:** The authors would like to acknowledge the financial assistance provided by the Major State Basic Research Development Program of China (No. 2014CB046702).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Corrosion Characteristics of Copper-Added Austempered Gray Cast Iron (AGCI)**

#### **Asiful H. Seikh 1,\*, Amit Sarkar 2, Jitendra Kumar Singh 3,\*, Sohail M. A. Khan Mohammed 4, Nabeel Alharthi 1,5 and Manojit Ghosh <sup>6</sup>**


Received: 16 January 2019; Accepted: 2 February 2019; Published: 6 February 2019

**Abstract:** The aim of this investigation was to assess the corrosion behavior of gray cast iron (GCI) alloyed with copper. Alloyed GCI specimens were austempered isothermally at varying temperatures. After austenitizing at 927 ◦C, the samples were austempered at different temperatures ranging from 260 to 385 ◦C with an interval of 25 ◦C for 60 min. As a result, these samples developed an ausferrite matrix with different percentages of austenite. The resulting microstructures were evaluated and characterized by optical microscope (OM), scanning electron microscope (SEM), and X-ray diffraction (XRD). The corrosion characteristics were determined using potentiodynamic polarization tests and electrochemical impedance spectroscopy (EIS) of these samples. These tests were carried out in a medium of 0.5 M H2SO4 and 3.5% NaCl solution. It was observed from the potentiodynamic polarization results that with increasing austempering temperature, the corrosion rate decreased. All results of the EIS were in accordance with a constant phase element (CPE) model. It was found that with an increase in austempering temperature, the polarization resistance (Rp) increased. The austenite content was also found to influence the corrosion behavior of the austempered gray cast iron (AGCI).

**Keywords:** austempered gray cast iron; austempering temperature; microstructure; potentiodynamic polarization; electrochemical impedance spectroscopy

#### **1. Introduction**

Gray cast iron (GCI) is a potential engineering material, which has a diverse range of applications including use in sophisticated automotive parts [1]. The wide applications of GCI are possible due to its unique properties such as good thermal conductivity, relatively low melting temperature, high damping capacity, and excellent castability [1,2]. The damage of the GCI components at the exterior parts through electrochemical corrosion has been the predominant restricting mechanism against enhancing its life span [3]. The presence of graphitization is a distinguishing feature of the deterioration properties of GCI [3,4]. Attempts have been made to combat the problem of corrosion

with the help of alloying additions with the aim of modifying the microstructure from ferrite to fine pearlite. It is also pertinent to mention that Si plays an important role in controlling the corrosion behavior of GCI: The higher the Si content, the higher the corrosion resistance [5]. Additionally, it is well known that single-phase microstructures like austenite, ferrite, and martensite perform better in corrosive media compared to two-phase mixtures like bainite, pearlite, and tempered martensite [6].

Several researchers [7–10] have tried to assess the effect of heat treatment and composition on the microstructure and corrosion behavior of austempered ductile iron (ADI). Prasanna et al. [7] and Banerjee et al. [8] studied the effect of the austempering treatment on the microstructure and corrosion properties of ductile iron. They found that both mechanical properties and corrosion resistance were enhanced due to the austempering of cast iron. Afolabi et al. [9] observed that the austempering temperature and time influenced the microstructure of the ductile iron, and thus its corrosion behavior was affected by the compositional structures. Hsu and Chen [10] concluded that the enhancement of corrosion resistance in ADI was due to the presence of retained austenite as a result of austempering. Similar studies with GCI are also common where the corrosion resistance and the mechanical properties were improved dramatically by tailoring the heat treatment pattern (tempering, austempering, and quenching [11–14]) and by alloy additions [8]. Further improvement of the mechanical properties, compared to those of conventional GCI, was observed in austempered gray cast iron (AGCI) due to the formation of a matrix of ausferritic structures (ferrite and stabilized austenite) or bainitic ferrite during austempering [12,15]. Thus, the domain of applicability of AGCI is even wider than that of GCI due to its favorable combination of enhanced mechanical properties [11–15] and improved wear characteristics [15–17]. The present literature, however, is lacking in reporting the corrosion behavior of AGCI, although a lot of work can be found on testing the corrosion behavior of ADI [7–10].

The above scientific observations indicated the necessity of the present investigation into the effect of austempering temperatures on the microstructure and corrosion behaviors of copper-alloyed AGCI, in order to establish a correlation between them.

#### **2. Materials and Methods**

#### *2.1. Sample Preparation*

Samples of GCI were prepared from cupola melts in a production foundry. The molten metal, at a temperature of 1420 ◦C was inoculated with 0.25 wt. % of FeSi-based inoculants in the cupola. During tapping, 0.5% Cu pieces of electrolyte grade were added to the metal stream for the sake of alloying. The specimens were cast in the form of standard 30-mm Y-shaped blocks in sand molds as shown in Figure 1a. Corrosion test coupons (Figure 1b) of suitable size (Φ 10 mm × 10 mm) were machined from the as-cast Y blocks. The uniform distribution of fine type-A graphite flakes is promoted by the inoculants during solidification [18]. Cu is soluble in austenite and increases the hardness, strength, corrosion resistance, and transformation time for the austempering process [19,20]. Cu has been accepted as an affordable alloying element for several engineering applications. As a result, the replacement of expensive Ni by Cu may become more prevalent. The final chemical composition (wt. %) of GCI was determined using a spectroscopy spark analyzer as shown in Table 1.

**Figure 1.** (**a**) Dimension (mm) of the Y-block casting and (**b**) schematic of the corrosion test piece (Φ 10 mm × 10 mm).

**Table 1.** Chemical composition of GCI (wt. %).


#### *2.2. Heat Treatment of Samples*

The samples were initially heated to an austenitizing temperature (Tγ = 927 ◦C) and held for 60 min in order to develop a fully austenitic structure (γ). The samples were then rapidly cooled in a molten salt bath comprising 53% KNO3, 40% NaNO2, and 7% NaNO3 at six different austempering temperatures (TA), 260, 285, 310, 335, 360, and 385 ◦C, for 60 min followed by air cooling to complete the phase transformation. Figure 2a schematically represents the entire heat treatment schedule for the austempering process and Figure 2b is the corresponding continuous cooling transformation (CCT) diagram.

**Figure 2.** Schematic diagram of (**a**) heat treatment schedule for austempering and (**b**) CCT diagram for the proposed composition.

#### *2.3. Metallography and X-ray Diffraction (XRD)*

Samples were prepared for metallographic observation using standard polishing techniques. Moreover, the samples were etched using a 2% nital solution for observation under a scanning electron microscope (SEM, JSM 6360, Jeol techniques, Tokyo, Japan). The volume fractions of austenite were calculated by X-ray diffraction (XRD,) analysis as described by Dasgupta et al. [21]. The XRD data were collected using a Rigaku, Ultima III diffractometer (Japan) with a monochromatic copper Fe-Kα radiation (1.54 Å) at 40 kV and 30 mA. Scanning was done at a rate of 1◦/min from 30 to 90◦ to observe the peaks, which were later analyzed using Jade 7 software (7.1.08). The peak positions were analyzed for the (111), (220), and (311) planes of austenite (FCC) and the (110), (200), and (211) planes of ferrite (BCC). The carbon content in austenite (Cγ) at various austempering temperatures was calculated using the following equation:

$$C\_{\gamma} = \frac{a\_{\gamma} - 3.548}{0.044},$$

where *aγ* is the lattice parameter calculated from the angular position of the austenite peak [22].

#### *2.4. Electrochemical (Corrosion) Test*

The heat-treated samples were subjected to electrochemical measurements in a 0.5 M H2SO4 and 3.5% NaCl solution at 25 ◦C (±2 ◦C). The electrochemical studies were performed in a triplicate set of samples to obtain reproducible results. A cell composed of three electrodes was created, including a graphite one, which acts as a counter electrode; a saturated calomel electrode (SCE), which acts as a reference electrode; and the GCI sample which acts as the working electrode (WE), for potentiodynamic polarization and electrochemical impedance spectroscopy (EIS) measurements. The WE area was fixed at 1 cm2. Prior to the tests, the samples were ground and polished using SiC papers of 2500 grit size and rinsed in deionized water followed by immersion in the solution for 30 min in order to stabilize the open circuit potential value. The potentiodynamic polarization tests were carried out from −1.0 to +1.0 V at a scan rate of 1 mV/s, whereas, EIS tests were performed over a frequency ranging from 100 kHz to 0.01 Hz.

#### **3. Results and Discussion**

#### *3.1. Microstructure and XRD Analysis*

Figure 3a,b shows the optical and SEM microstructures of the as-cast gray iron sample, respectively. The matrix of as-cast gray iron is primarily composed of pearlite besides some randomly distributed ferrite. The as-cast specimens were austempered for 60 min and the resulting changes in the microstructure are presented in Figures 4a–f and 5a–f using an optical microscope (OM) and SEM, respectively. The dark, etched needle-like structures represent bainitic ferrite, while the brighter ones represent a mixture of austenite and bainitic ferrite. We can see that the effect of the austempering temperature on the microstructure of austempered irons was significant. It was observed that at lower temperatures (i.e., 260–285 ◦C), very fine needles of bainitic ferrite and austenite were formed and the volume fraction of ferrite was larger. As the austempering temperature increased, the needles of the bainitic ferrite were coarsened along with an increase in the austenite content. Similar observations in ADI were earlier reported by Patutunda et al. [23] and Yang et al. [24].

**Figure 3.** Micrographs of as-cast specimens: (**a**) optical micrograph and (**b**) SEM image.

**Figure 4.** Optical micrographs of samples austempered for 60 min at (**a**) 260 ◦C, (**b**) 285 ◦C, (**c**) 310 ◦C, (**d**) 335 ◦C, (**e**) 360 ◦C, and (**f**) 385 ◦C.

**Figure 5.** SEM images of samples austempered for 60 min at (**a**) 260 ◦C, (**b**) 285 ◦C, (**c**) 310 ◦C, (**d**) 335 ◦C, (**e**) 360 ◦C, and (**f**) 385 ◦C.

Figure 6 presents the quantitative analysis of the XRD pattern. It is evident from the figure that the austempering temperature has a significant effect on the XRD patterns. It was seen that with changing heat treatment temperature, the amount of austenite was changed. The phases detected include ferrite and austenite. Figure 7 shows the volume fraction of austenite and the carbon content of austenite in the AGCI samples as a function of different austempering temperatures. The volume faction of austenite was calculated by Jade7 software built in the XRD. It may be noted from Figure 7 that the austenite content increases with an increase in the austempering temperature. Greater supercooling at a lower austempering temperature resulted in finer ferrite and austenite as also reported by Patutunda et al. [23]. It is well known that the transformation reaction is more likely to be controlled by the nucleation process rather than growth [23,24]. During the process, it is necessary that the carbon must diffuse into austenite through the ferrite zone. At higher austempering temperatures a quite contrasting mechanism prevails due to lower supercooling which makes the nucleation of ferrite slower. This leads to the stabilization of more austenite in addition to incrementing the rate of diffusion of carbon which leads to the formation of coarse ferrite. Thus, the volume fraction of austenite increases with the increase in austempering temperature.

**Figure 6.** XRD phase analysis of austempered gray cast iron (AGCI) for different austempering temperatures held for 60 min (**a**) 260 ◦C, (**b**) 285 ◦C, (**c**) 310 ◦C, (**d**) 335 ◦C, (**e**) 360 ◦C, and (**f**) 385 ◦C.

**Figure 7.** The influence of different austempering temperatures on the volume fraction of austenite and the carbon content of austenite.

#### *3.2. Electrochemical Behavior of As-Cast Gray Iron and AGCI in Different Solutions*

The results of potentiodynamic polarization studies of the as-cast gray iron and the AGCI are shown in Figures 8 and 9. The corrosion of iron in neutral 3.5% NaCl solution occurs according to the following equations,

Anodic reaction:

$$\text{Fe} \rightarrow \text{Fe}^{2+} + 2\text{e}\_{\prime} \tag{2}$$

Cathodic reaction:

$$\text{O}\_2 + 2\text{H}\_2\text{O} + 4\text{e} \to 4\text{OH}^-.\tag{3}$$

When iron is in contact with dilute sulfuric acid (0.5 M H2SO4), an immediate attack on the metal takes place with the formation of hydrogen gas and ferrous ions, as shown in Equations (4) and (5).

Anodic reaction:

$$\text{Fe} \rightarrow \text{Fe}^{2+} + 2\text{e}^-.\tag{4}$$

Cathodic reaction:

$$\text{H}^{\text{+}} + 2\text{e}^{\text{-}} \rightarrow \text{H}\_{2}.\tag{5}$$

The electrochemical parameters are extracted after the extrapolation of the potentiodynamic plots in a Tafel slope. From Figure 7 it is revealed that with increasing austempering temperature, the percentage of austenite increased while the corrosion current density (Icorr) decreased and the corrosion potential (Ecorr) shifted to the cathodic side (Tables 2 and 3). Austenite acts as an anode and ferrite acts as a cathode. The galvanic corrosion is proportional to the cathodic/anodic area. Consequently, with increasing temperature, the austenite percentage increases with a simultaneous decrease in the ferrite percentage. Thus, due to the microstructural homogeneities, distinct localized anodic and cathodic microstructural areas develop, which act as micro-electrochemical cells in the presence of an electrolyte. Thus, the galvanic corrosion decreases in both solutions. From Tables 2 and 3, it can be seen that among the two corrosive mediums, 1 N H2SO4 is more corrosive in all cases.

**Figure 8.** Potentiodynamic polarization curves in 3.5% NaCl solution.

**Figure 9.** Potentiodynamic polarization curves in 0.5 M H2SO4 solution.

**Table 2.** Potentiodynamic polarization results in 3.5% NaCl.


**Table 3.** Potentiodynamic polarization results in1NH2SO4.


Nyquist plots of samples exposed to 3.5% NaCl and 0.5 M H2SO4 solutions are shown in the Figures 10 and 11, respectively. All plots show a depressed semicircle pattern in the whole frequency range, indicating that only one time constant exists between the interface of the solid electrode and the solution. Due to the low impedance value at the lower austempering temperature, the Nyquist plots become suppressed. The corresponding plots are shown in the insets of Figures 10 and 11 for exposure to 3.5% NaCl and 0.5 M H2SO4 solutions, respectively. All the EIS data match well in a constant phase element (CPE) model. In a CPE model, Ru is the solution resistance, Rp is the polarization resistance, and Yo is the admittance. The inserted equivalent circuit shown in Figure 12 was used to fit the EIS

data, and the fitted polarization resistance (Rp) data are shown in Tables 4 and 5 for 3.5% NaCl and 0.5 M H2SO4 solution, respectively. It is known that the diameter of the Nyquist plot represents the Rp. It is also well known that the Rp is inversely proportional to the corrosion rate. With increasing austempering temperature, the diameter of the Nyquist plot increases with a consequent increase in the Rp. It is also seen from Tables 4 and 5 with the error bar (maximum ±7%) that the Rp is higher in 3.5% NaCl solution (Table 4) than it is in 0.5 M H2SO4 solution (Table 5) for all cases.

**Figure 10.** Electrochemical impedance spectroscopy (Nyquist plot) in 3.5% NaCl solution.

**Figure 11.** *Cont*.

**Figure 11.** Electrochemical impedance spectroscopy (Nyquist plot) in1NH2SO4 solution.

**Figure 12.** Equivalent circuit of electrochemical impedance spectroscopy (EIS).


**Table 4.** Electrochemical impedance spectroscopy results in 3.5% NaCl solution.


**Table 5.** Electrochemical impedance spectroscopy results in 0.5 M H2SO4 solution.

#### *3.3. Effect of Austenite Content on Corrosion Behavior*

Figures 13 and 14 show the plots of Icorr obtained from the potentiodynamic polarization diagram against the volume fraction of the austenite for 3.5% NaCl and 1 N H2SO4, respectively. It was observed that with the increasing volume fraction of austenite, the corrosion rate decreased linearly to a sufficient extent in both cases. The linear fit regression value was 0.95 for 3.5% NaCl solution (Figure 13) and 0.84 for 1N H2SO4 solution (Figure 14). A regression value close to 1 means the corrosion rate changes linearly with increasing austenite content.

**Figure 13.** Influence of the volume fraction of the austenite on corrosion rate in 3.5% NaCl.

**Figure 14.** Influence of the volume fraction of the austenite on the corrosion rate in 0.5 M H2SO4.

#### *3.4. Microstructure after Corrosion*

#### 3.4.1. Optical Images after Corrosion in 3.5% NaCl Solution

Figure 15 shows the OM of as-cast gray iron and AGCI samples dipped in 3.5% NaCl solution. The corrosion products consist of compact structures. It is seen that compactness increases with increasing austempering temperature. It was also observed that in optical images of as-cast gray iron (Figure 15a) with a lower austempering temperature (260 ◦C and 285 ◦C), an exfoliation type pattern was present with smaller flake graphite. With increasing austempering temperature (310 ◦C and above), the exfoliation type pattern disappeared with bigger flake graphite. In addition, intergranular type corrosion was observed in the 3.5% NaCl solution. It is also seen that, in as-cast gray iron at a low austempering temperature, more pitting was seen. However, with increasing austempering temperature the pitting density decreased.

**Figure 15.** *Cont*.

**Figure 15.** Optical images of iron samples after a potentiodynamic polarization test in 3.5% NaCl solution. (**a**) As-cast gray iron; and AGCI at (**b**) Tγ = 927 ◦C, TA = 260 ◦C; (**c**) Tγ = 927 ◦C, TA = 285 ◦C; (**d**) Tγ = 927 ◦C, TA = 310 ◦C; (**e**) Tγ = 927 ◦C, TA = 335 ◦C; (**f**) Tγ = 927 ◦C, TA = 360 ◦C; and (**g**) Tγ = 927 ◦C, TA = 385 ◦C.

3.4.2. Optical Images after Corrosion in 0.5 M H2SO4 Solution

Figure 16 shows the optical images (as-cast gray iron and AGCI) of corrosion products in 0.5 M H2SO4 solution. It is seen that pitting formation decreased, with increasing austempering temperature. While with increasing austempering temperature, more metastable pits were formed. So it can be concluded that with increasing austempering temperature, pit formation gradually reduced. At the same temperature, compared to 3.5% NaCl solution, pitting density, radius of pits, and the flake graphite was larger in the case of 0.5 M H2SO4 solution. It is also seen that more pits are formed in 1 N H2SO4 than in 3.5% NaCl solution.

**Figure 16.** Optical images of iron samples after a potentiodynamic polarization test in 1 N H2SO4 solution. (**a**) As-cast gray iron; and AGCI at (**b**) Tγ = 927 ◦C, TA = 260 ◦C; (**c**) Tγ = 927 ◦C, TA = 285 ◦C; (**d**) Tγ = 927 ◦C, TA = 310 ◦C; (**e**) Tγ = 927 ◦C, TA = 335 ◦C; (**f**) Tγ = 927 ◦C, TA = 360 ◦C; and (**g**) Tγ = 927 ◦C, TA = 385 ◦C.

#### **4. Conclusions**

The following conclusions can be drawn from the present investigation:

(a) The microstructure of AGCI consists of special bainitic ferrite (α) and high-carbon austenite (Y) which prevents corrosion. Thus, the corrosion-resistance susceptibility of AGCI is higher than that of as-cast gray iron.

(b) At higher austempering temperatures, the volume fraction of austenite increases with a consequent decrease in the corrosion rate.

(c) In the ausferrite matrix, the corrosion rate depends on the austenite content. An increase in the austenite content results in a decrease in the corrosion rate.

(d) Due to an increase in hydrogen generating reactions,1NH2SO4 is more corrosive than 3.5% NaCl during exposure.

**Author Contributions:** Data curation: A.H.S. and M.G.; formal analysis: A.S., J.K.S., S.M.A.K.M. and M.G.; funding acquisition: A.H.S. and N.A.; methodology: A.H.S. and A.S.; project administration: A.H.S.; Supervision: A.H.S., S.M.A.K.M., and N.A.; validation: A.H.S., Amit Sarkar, and S.M.A.K.M.; writing—original draft: A.H.S., Amit Sarkar, J.K.S., S.M.A.K.M., N.A., and M.G.; Writing—review and editing: A.H.S., A.S., J.K.S., S.M.A.K.M., N.A. and M.G.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors would like to extend their sincere appreciation to the Deanship of Scientific Research at King Saud University for its funding of this research through the Research Group Project No. RG-1439-029.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Non-Isothermal Oxidation Behaviors and Mechanisms of Ti-Al Intermetallic Compounds**

#### **Peixuan Ouyang 1,2,3, Guangbao Mi 1,4,\*, Peijie Li 2, Liangju He 2, Jingxia Cao <sup>1</sup> and Xu Huang <sup>1</sup>**


Received: 25 May 2019; Accepted: 27 June 2019; Published: 30 June 2019

**Abstract:** Non-isothermal oxidation is one of the important issues for the safe application of Ti-Al alloys, so this study aimed to illustrate the non-isothermal oxidation behaviors and the corresponding mechanisms of a TiAl-based alloy in comparison with a Ti3Al-based alloy. The non-isothermal oxidation behaviors of Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys in pure oxygen were comparatively investigated with a thermogravimetry-differential scanning calorimetry (TGA/DSC) simultaneous thermal analyzer heating from room temperature to 1450 ◦C with a heating rate of 40 ◦C/min. When the temperature rose above 1280 ◦C, the oxidation rate of the Ti-46Al-2Cr-5Nb alloy sharply increased and exceeded that of the Ti-24Al-15Nb-1.5Mo alloy owing to the occurrence of internal oxidation. When the temperature was higher than 1350 ◦C, the oxidation rate of the Ti-46Al-2Cr-5Nb alloy decreased obviously due to the generation of an oxygen-barrier β-Al2TiO5-rich layer by a chemical reaction between Al2O3 and TiO2 in the oxide scale. Based on Wagner's theory of internal oxidation, the reason for the occurrence of internal oxidation in the Ti-46Al-2Cr-5Nb alloy is the formation of the α phase in the subsurface, while no internal oxidation occurred in the Ti-24Al-15Nb-1.5Mo alloy due to the existence of the β phase in the subsurface with the enrichment of Nb and Mo.

**Keywords:** titanium aluminides; oxidation; non-isothermal; mechanism; internal oxidation

#### **1. Introduction**

TiAl-based alloys have received considerable attention as high-temperature structural materials for aerospace and automotive applications, since they maintain numerous outstanding properties, such as low density (3.9–4.2 g/cm3), high specific strength, good creep resistance and excellent fireproof performance [1–5]. Nevertheless, the insufficient high-temperature oxidation resistance of the TiAl-based alloys limits their wide application and development [6,7]. Numerous investigations have been carried out to study the high-temperature oxidation behaviors of the TiAl-based alloys, most of which were concerned about the long-term isothermal oxidation at normal service temperature (800–1000 ◦C) [8–10]. Few studies have been focused on the non-isothermal oxidation behaviors of the TiAl-based alloys, which is also deserving of attention since the alloys are often subject to non-isothermal oxidation in service. For example, aero-engine TiAl components are heated rapidly from room temperature to service temperature during the engine startup period. For another instance, aero-engine TiAl blades are likely to suffer from titanium fire, which consists of ignition and propagation combustion processes under the induction of external energies such as high-energy friction, fracture, melt droplets and so on [11], even though the fireproof performance of TiAl-based alloys is much better

than that of conventional titanium alloys. Ignition, which is the precursor process of titanium fire, is essentially a non-isothermal oxidation process [12,13] and the ignition points of titanium alloys are slightly lower than their melting points [11]. Hence, the non-isothermal oxidation of TiAl-based alloys with a rather wide temperature range has a major impact on their safe application in aeroengines.

In addition, the non-isothermal oxidation behaviors of the TiAl-based alloys should be more special than the isothermal oxidation behaviors, since the heating rate is so high that the oxidation may not reach dynamic equilibrium and the temperature range is so wide that phase transition may occur in the alloys. Many studies showed that oxidation temperature and time significantly affect isothermal oxidation behaviors of the TiAl-based alloys [14–17]. Moreover, Vaidya [18] found that during isothermal oxidation, the mass gain of Ti-48Al-2Nb-2Cr alloy exceeds that of Ti-25Al-10Nb-3V-1Mo alloy when the temperature is above 1200 ◦C, which breaks away from the general understanding that the oxidation resistance of the TiAl-based alloys is better than that of the Ti3Al-based alloys. This phenomenon is considered to be attributed to the doubly increased oxygen solubility of γ phase in the TiAl-based alloy at 1200 ◦C [18], while this explanation seems unreasonable since the maximum oxygen solubility of γ phase (~3 at.%) in the TiAl-based alloy is much lower than that of α<sup>2</sup> phase (~13 at.%) in the Ti3Al-based alloys [19]. Hence, in regards to non-isothermal oxidation, it should be of interest to people what behaviors the TiAl-based alloys would exhibit, whether there exists a similar phenomenon that the oxidation rate of the TiAl-based alloys exceeds that of the Ti3Al-based alloys at high temperature, as well as what is the reasonable explanation for this phenomenon. On the basis of the above questions, this paper examines the non-isothermal oxidation behaviors of a TiAl-based alloy in comparison with a Ti3Al-based alloy and illustrates the corresponding oxidation mechanisms. On one hand, this research helps to better understand the ignition mechanisms of TiAl alloys, on the other hand, it might promote the development of new-type TiAl alloys resistant to higher temperature and with less risk to titanium fire by ingredient optimum design.

TGA/DSC simultaneous thermal analysis is a common method to study the non-isothermal oxidation behaviors of metals and their alloys [20–23]. Adopting the TGA/DSC method, G.B.Mi et al. [22] studied the effect of Cr content on the non-isothermal oxidation behaviors of Ti-Cr fire-proof titanium alloys. The results showed that when the Cr content exceeds 10–15 wt.%, the oxidation resistance of Ti-Cr alloys increases with the Cr content due to the precipitation of Cr oxide [22]. Ouyang et al. [23] carried out a research on the non-isothermal oxidation behaviors of a high-temperature near-α titanium alloy (TA29 alloy), meanwhile discussed the effects of lattice transformation and alloying elements on the non-isothermal oxidation behaviors of the TA29 alloy.

In this paper, the non-isothermal oxidation behaviors of a TiAl-based alloy and a Ti3Al-based alloy in pure oxygen were studied with a TGA/DSC synchronous thermal analyzer heating from room temperature to 1450 ◦C with a heating rate of 40 ◦C/min. Combined with microstructural characterization and calculation of oxidation activation energy, the non-isothermal oxidation mechanisms of the TiAl-based alloy were elucidated. Furthermore, the reason for the poorer oxidation resistance of the TiAl-based alloy than that of the Ti3Al-based alloy at high temperature was revealed based on Wagner's theory of internal oxidation.

#### **2. Experimental**

#### *2.1. Specimen Preparation*

The nominal compositions of the TiAl-based and Ti3Al-based alloys studied in this paper are Ti-46Al-2Cr-5Nb (at.%) and Ti-24Al-15Nb-1.5Mo (at.%), respectively. Both the two alloys were prepared by melting, forging, heat treatment and mechanical processing. The microstructure of the as-received Ti-46Al-2Cr-5Nb alloy consists of coarse fully-lamellar γ + α<sup>2</sup> colonies, as shown in Figure 1a. The microstructure of the as-received Ti-24Al-15Nb-1.5Mo alloy is duplex, being composed of equiaxed primary α<sup>2</sup> and/or O phases, B2 transformed structure consisting of flaky secondary α<sup>2</sup> and O phases and residual B2 phase, as shown in Figure 1b. For non-isothermal oxidation experiments, specimens with dimensions of 3 <sup>×</sup> 2 <sup>×</sup> 2 mm<sup>3</sup> were cut from the two kinds of alloy sheets by a computerized numerical control (CNC) dicing saw (SYJ-400, Shenyang kejing instrument company, Shenyang, China. The specimens were ground with sandpapers up to 2000-grit to remove oxide scales and then ultrasonically cleaned with acetone and alcohol.

**Figure 1.** Optical micrographs showing the original microstructure of the (**a**) Ti-46Al-2Cr-5Nb and (**b**) Ti-24Al-15Nb-1.5Mo alloys.

#### *2.2. Non-Isothermal Oxidation Experiment*

Non-isothermal oxidation experiments were carried out in a TGA/DSC simultaneous thermal analyzer (TGA/DSC 1; Mettler Toledo, Zurich, Switzerland). The specimens of both the Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys were heated from room temperature to 1450 ◦C at a heating rate of 40 ◦C/min under a pure oxygen flow of 50 mL/min. The oxygen flow was added to the system once the heating process was started and the flow was cut off when the heating process was completed. After furnace cooling the oxidized specimens were taken out. For each alloy, the mass-gain, mass-gain rate, and heat flux curves were obtained from three repeated experiments with the relative standard deviations less than 10%. Besides, in order to investigate the non-isothermal oxidation behaviors of the two alloys in detail, the specimens were heated to several interested temperatures below 1450 ◦C and then were furnace cooled to obtain the corresponding non-isothermal oxidation products.

#### *2.3. Microstructural Characterization*

Specimens were examined after non-isothermal oxidation using field emission scanning electron microscopy (FE-SEM; Hitachi SU8000, Tokyo, Japan) and X-ray diffraction (XRD; Bruker D8 Advance; Cu Ka, Karlsruhe, Germany) to characterize surface morphologies of the oxide scales and identify oxide phases. For cross-sectional microstructure observation of the oxidized specimens, metallographic specimens were prepared by being placed in the transverse direction, being mounted with cold-setting resin, then being ground with 400–2000 grit SiC sandpapers, being polished with 1.0 μm alumina suspension, being etched with Kroll reagent (92 mL H2O, 3 mL HF, 5 mL HNO3) and finally being coated with a thin layer of carbon. The microstructures and elemental distributions on the cross-section of the oxide scales were characterized using an electron probe microanalyzer (EPMA; Shimadzu EPMA-1720H, Kyoto, Japan).

#### **3. Calculation of Oxidation Activation Energy**

Similar to traditional titanium alloys, the oxidation mass gains of the Ti-Al alloys also come from oxygen dissolution and growth of oxide scales. The two oxidation behaviors are competitive and one of them governs the mass gain at a certain temperature and time [24–26]. The rate-determining steps of the two oxidation behaviors are respectively the diffusion of O atom in the alloy and the diffusion of O2<sup>−</sup> in the oxide scale. Thus, it can be assumed that the mass-gain rate of the Ti-Al alloys during non-isothermal oxidation is limited by one-dimensional diffusion of one species A (A is O atom or O2<sup>−</sup>), so that the mass-gain rate per unit area of the Ti-Al alloys is proportional to the molar diffusion flux of specie A. Furthermore, based on the Fick's diffusion law and assuming that the diffusion of specie A satisfies Arrhenius kinetics, the relationship among the temperature, the mass gain and the oxidation activation energy of the Ti-Al alloys can be derived, as shown in Equation (1). Details for the derived process can be seen in our previous work [23].

$$-\ln\frac{\mathbf{d}(\Delta m)}{\mathbf{d}T} - \ln(\Delta m) = -\ln \mathbf{K}^\* + E/\mathbf{R}T \tag{1}$$

where Δ*m* is the mass gain per unit area, *E* is the oxidation activation energy, R is the molar gas constant and K\* is constant. Let the left side of Equation (1) be equal to *Y*(Δ*m*), thus *Y*(Δ*m*) is linearly related to the reciprocal of temperature and the product of the corresponding positive slope and R is the oxidation activation energy.

#### **4. Results and Discussion**

#### *4.1. Oxidation Mass Gain and Activation Energies*

Figure 2 shows the mass-gain and mass-gain rate curves of the Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys obtained during non-isothermal oxidation. The mass-gain rates of the two alloys were similar when the temperature was lower than 1280 ◦C. However, the mass-gain rate of the Ti-46Al-2Cr-5Nb alloy increased dramatically when the temperature was higher than 1280 ◦C, resulting that the mass gain of the Ti-46Al-2Cr-5Nb alloy significantly exceeded that of the Ti-24Al-15Nb-1.5Mo alloy. This phenomenon is not common but is similar to that reported by Vaidya [18]. Nevertheless, the mass-gain rate of the Ti-46Al-2Cr-5Nb alloy turned to decrease obviously when the temperature was above 1350 ◦C.

**Figure 2.** Comparisons of the mass-gain and mass-gain rate curves between the Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys obtained during non-isothermal oxidation.

According to the evolution of the mass-gain rate, the non-isothermal oxidation process of the Ti-46Al-2Cr-5Nb alloy could be divided into five stages for further study, as shown in Figure 3a. When the temperature was below 870 ◦C (Stage I), the mass gain of the Ti-46Al-2Cr-5Nb alloy was few and could be neglected. When the temperature rose to 870–980 ◦C (Stage II), the mass-gain rate increased slowly and the corresponding mass gain was 0.04 mg/cm2. When the temperature was increased to 980–1280 ◦C (Stage III), the mass-gain rate increased obviously and the corresponding mass gain was 1.05 mg/cm2. When the temperature was raised to 1280–1350 ◦C (Stage IV), the mass-gain rate increased sharply and the corresponding mass gain was 1.68 mg/cm2. Nevertheless, the mass-gain rate decreased significantly when the temperature was above 1350 ◦C (Stage V). It is concluded that the non-isothermal oxidation process of the Ti-46Al-2Cr-5Nb alloy consists of five stages, including nearly non-oxidation (Stage I, <870 ◦C), slow oxidation (Stage II, 870–980 ◦C), accelerated oxidation (Stage III, 980–1280 ◦C), severe oxidation (Stage IV, 1280–1350 ◦C) and decelerated oxidation (Stage V, 1350–1450 ◦C) stages. However, the non-isothermal oxidation process of the Ti-24Al-15Nb-1.5Mo alloy can be only divided into four stages according to the evolution of the mass-gain rate, as shown in Figure 3b, including nearly non-oxidation (Stage I, <800 ◦C), slow oxidation (Stage II, 800–1020 ◦C), accelerated oxidation (Stage III, 1020–1400 ◦C) and severe oxidation (Stage IV, 1400–1450 ◦C) stages. Table 1 gives a summary of the temperature ranges corresponding to the different stages of the non-isothermal oxidation process for the Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys.

**Figure 3.** Stage divisions of the non-isothermal oxidation process for the (**a**) Ti-46Al-2Cr-5Nb and (**b**) Ti-24Al-15Nb-1.5Mo alloys according to the evolution of the mass-gain rates.


**Table 1.** Temperature ranges, oxidation activation energies and matrix phases corresponding to different stages of the non-isothermal oxidation for the Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys.

<sup>1</sup> TiAl here refers to the Ti-46Al-2Cr-5Nb alloy. <sup>2</sup> Ti3Al here refers to the Ti-24Al-15Nb-1.5Mo alloy.

Figure 4 presents the *Y*(Δ*m*)~1/*T* curves and the corresponding linear fitting results of the two alloys. The evolutions of the slops of the *Y*(Δ*m*)~1/*T* curves also clearly demonstrate that the non-isothermal oxidation processes of the Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys are respectively composed of five stages and four stages. It should be noted that since the mass changes of the two alloys at Stage I were mainly induced by the small mass fluctuation of the thermogravimetric analysis equipment, the values of *Y*(Δ*m*) at Stage I fluctuated around a certain value and are not shown in Figure 4. Besides, the *Y*(Δ*m*)~1/*T* curves of both the two alloys at Stages II, III and IV could be positive linearly fitted, so that the oxidation activation energies of the two alloys for the three stages were obtained, as presented in Table 1. It can be seen that the oxidation activation energies of the two alloys at Stage III are similar, while the oxidation activation energies of the two alloys both at Stage II and at Stage IV are quite different, which is related to their oxidation mechanisms and will be discussed in Section 4.3.

**Figure 4.** The *Y*(Δ*m*)~1/*T* curves and the corresponding linear fitting results of the (**a**) Ti-46Al-2Cr-5Nb and (**b**) Ti-24Al-15Nb-1.5Mo alloys.

#### *4.2. Matrix Phases*

Since the non-isothermal oxidation experiments were carried out in a wide temperature range, the Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys were likely to undergo phase transitions during non-isothermal oxidation.

Generally, the phase transition types and temperatures of the TiAl-based alloys are affected not only by the aluminum content (45–48 at.%), but also by kinds of other alloying elements (such as Cr, V, Mn, Nb, Ta, W, Si, C, P and B) and their contents (0.1–8 at.%) [27]. According to the Ti-Al phase diagram [28], the heat-flow and thermomechanical analysis (TMA) derivative curves of Ti-(46~47)Al-2Cr-(2~8)Nb alloy [29], the phase transitions of the Ti-46Al-2Cr-5Nb alloy heating from room temperature to 1450 ◦C are deduced to be γ + α<sup>2</sup> → γ + α and γ + α → α with the respective temperature range of

1175–1215 ◦C (*T*eu) and 1292–1295 ◦C (*T*α). The deduced temperature ranges are in accordance with the two exothermic peaks with the respective temperatures of 1185 and 1226 ◦C in the heat-flow curve of the Ti-46Al-2Cr-5Nb alloy, as shown in Figure 3a. It confirms that the phase transitions of γ + α<sup>2</sup> → γ + α and γ + α → α occurred in the Ti-46Al-2Cr-5Nb alloy during the non-isothermal oxidation. Besides, it can be seen from Figure 3a that there exists another exothermic peak with a temperature of about 1032 ◦C in the heat-flow curve of the Ti-46Al-2Cr-5Nb alloy. It is not yet possible to explain the physical meaning of this peak based on the Ti-Al phase diagram. However, it has been reported that there exists a similar exothermic peak in the heat-flow curve of Ti-46Al-1.9Cr-3Nb alloy with typical duplex structure, which is ascribed to equilibrium transformation or uniformity when heating of the non-equilibrium structure generated by thermal mechanical processing [30]. Therefore, the matrix phases of the Ti-46Al-2Cr-5Nb alloy at different stages of the non-isothermal oxidation process could be concluded, as shown in Table 1.

As for the Ti-24Al-15Nb-1.5Mo alloy, the Nb equivalent is about 20 at.% since the beta phase stability of Mo is 3.6 times of that of Nb [31]. Thus, according to the Ti-25Al-Nb phase diagram [31], the phase transitions of the Ti-24Al-15Nb-1.5Mo alloy heating from room temperature to 1450 ◦C are deduced to be α<sup>2</sup> + O + B2 → α<sup>2</sup> + B2 (1010 ◦C), α<sup>2</sup> + B2 → B2 (1100 ◦C) and B2 → β (1220 ◦C). Since there exist two endothermic peaks with the respective temperatures of near 1000 and 1100 ◦C in the heat-flow curve of the Ti-24Al-15Nb-1.5Mo alloy during non-isothermal oxidation, as shown in Figure 3b, it demonstrates the occurrence of the former two phase transitions. Besides, no peak corresponding to the last phase transition could be found in the heat-flow curve of the Ti-24Al-15Nb-1.5Mo alloy, which is probably due to the small chemical heat for the disordering from B2 phase to β phase. As a result, the matrix phases of the Ti-24Al-15Nb-1.5Mo alloy at different stages of the non-isothermal oxidation process could be also concluded, as presented in Table 1.

#### *4.3. Non-Isothermal Oxidation Mechanisms*

#### 4.3.1. Nearly Non-Oxidation Stage (Stage I)

Similar to the near-α titanium alloy TA29 [23], the oxidation mass gains of the Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys at Stage I were very small and could be neglected (Figure 3), which is ascribed to the fact that a thin titanium oxide film (200 nm) rapidly forms on the alloys at room temperature and passivates the surface [14].

#### 4.3.2. Slow Oxidation Stage (Stage II)

When the temperature rose to Stage II (870–980 ◦C for the Ti-46Al-2Cr-5Nb alloy and 800–1020 ◦C for the Ti-24Al-15Nb-1.5Mo alloy), both the two alloys exhibited slow oxidation behaviors, which is also similar to the near-α titanium alloy TA29 at Stage II with the temperature range of 750–1000 ◦C [23]. As discussed in our previous work [23], the oxidation mechanism of the TA29 alloy at Stage II is oxygen dissolution with the rate-determining step of oxygen diffusion in the alloy. The oxidation activation energies of the TA29 alloy at Stage II (163.9 kJ/mol) [23] is about 20 kJ/mol higher than that of the Ti-24Al-15Nb-1.5Mo alloy (143.2 kJ/mol). The matrixes of the TA29 and Ti-24Al-15Nb-1.5Mo alloys at Stage II are respectively dominated by the α and α<sup>2</sup> phases. Moreover, it has been reported that the activation energy for oxygen diffusion in the α phase is 10–20 kJ/mol higher than that in the α<sup>2</sup> phase [32], which is equivalent to the difference of the oxidation activation energy between the TA29 and Ti-24Al-15Nb-1.5Mo alloys. Therefore, the oxidation mechanism of the Ti-24Al-15Nb-1.5Mo alloy at Stage II is also the oxygen dissolution in the alloy. It indicates that the thin titanium oxide film (200 nm) passivating the surface at Stage I could not prevent oxygen from diffusing into the alloys at Stage II, which might be due to the increase of the solubility of TiO2 in the alloys or the cracking of TiO2 film caused by phase transitions in the oxide film with the temperature increases.

The lower oxidation activation energy of the Ti-24Al-15Nb-1.5Mo alloy in comparison with the TA29 alloy indicates that oxygen atoms diffuse much easier in the former alloy. Since volume diffusion is the main diffusion type in the alloys at such high temperature, the difference of oxygen diffusion ability can be explained by the difference of the lattice structure between the two alloys. The α phase dominating in the TA29 alloy has the close-packed hexagonal structure (hcp, A3), where oxygen atoms tend to occupy the two Ti6 octahedral interstitial sites [33,34], as shown in Figure 5a. The α<sup>2</sup> phase dominating in the Ti-24Al-15Nb-1.5Mo alloy has the ordered close-packed hexagonal structure (DO19), where oxygen atoms prefer to occupy the Ti6 octahedral interstitial site instead of the Al2Ti4 site [35,36], as shown in Figure 5b. Though the number of the Ti6 interstitial site in the lattice of the α<sup>2</sup> phase is lower than that in the lattice of the α phase, which leads to the lower oxygen solubility in the α<sup>2</sup> phase than in the α phase, the covalence of Ti-Al bond in the lattice of the α<sup>2</sup> phase makes electrons aggregate between Ti and Al atoms, resulting in the weaker Ti-O bond strength in the lattice of the α<sup>2</sup> phase compared with that in the lattice of the α phase [32]. Therefore, the diffusion resistance of oxygen atoms in the lattice of the α<sup>2</sup> phase is smaller than that in the lattice of the α phase. That is to say, the required activation energy for oxygen diffusion in the α<sup>2</sup> phase is smaller than in the α phase.

**Figure 5.** Octahedral interstitial sites for oxygen atoms in the lattices of the (**a**) α, (**b**) α<sup>2</sup> and (**c**) γ phases.

As for the Ti-46Al-2Cr-5Nb alloy, the oxidation activation energy at Stage II (217.8 kJ/mol) is higher than that of the Ti-24Al-15Nb-1.5Mo alloy (143.2 kJ/mol). The γ phase dominating in the Ti-46Al-2Cr-5Nb alloy at Stage II has the ordered face-centered cubic structure (L10), where oxygen atoms can only occupy the interstitial octahedral sites surrounded by both titanium and aluminum atoms (Al4Ti2 and Al2Ti4), as shown in Figure 5c. Since the covalence of Ti-Al bond makes electrons aggregate between Ti and Al atoms, the Ti-O bond strength in the Al4Ti2 and Al2Ti4 octahedrons is higher than that in the Ti6 octahedron. Hence, the oxygen diffusion ability in the lattice of the γ phase is much weaker than that in the lattice of the α<sup>2</sup> phase, leading to the higher oxidation activation energy of the Ti-46Al-2Cr-5Nb alloy than that of the Ti-24Al-15Nb-1.5Mo alloy. The above analysis manifests that the oxidation mechanism of the Ti-46Al-2Cr-5Nb alloy at Stage II is the same with the Ti-24Al-15Nb-1.5Mo alloy, namely, oxygen dissolution in the alloy. Besides, since the oxygen solubility in the γ phase is much lower than that in the α<sup>2</sup> phase, the temperature range corresponding to Stage II for the Ti-46Al-2Cr-5Nb alloy (870–980 ◦C) is narrower than that for the Ti-24Al-15Nb-1.5Mo alloy (800–1020 ◦C).

#### 4.3.3. Accelerated Oxidation Stage (Stage III)

When the temperature was raised to Stage III (980–1280 ◦C for the Ti-46Al-2Cr-5Nb alloy and 1020–1400 ◦C for the Ti-24Al-15Nb-1.5Mo alloy), the two alloys exhibited accelerated oxidation behaviors and the oxidation activation energies are respectively 249.8 and 244.8 kJ/mol, which is close to the diffusion activation energies of O2- and Ti4<sup>+</sup> in TiO2 (234 kJ/mol [37] and 257 kJ/mol [38], respectively). It indicates that the oxidation mechanisms of the two alloys at this stage are mainly the growth of oxide scales dominated by TiO2. Since the diffusion rates of O2<sup>−</sup> and Ti4<sup>+</sup> in TiO2 are much

lower than that of oxygen atom in the alloy, the rate-determining step at this stage is the diffusion of O2<sup>−</sup> and Ti4<sup>+</sup> in the oxide scale so that the oxidation activation energy is independent with the matrix phases of the alloys.

Figure 6 demonstrates the surface morphologies and XRD pattern of the oxide scale formed during heating the Ti-46Al-2Cr-5Nb alloy to the end of Stage III. As shown in the XRD pattern (Figure 6d), there are high contents of γ and α<sup>2</sup> phases besides rutile TiO2 and corundum α-Al2O3. It manifests that the oxide is composed of rutile TiO2 and corundum α-Al2O3 (hereafter referred to as TiO2 and Al2O3), and the thickness of the oxide scale is less than the detection depth of X-ray (~20 μm). The oxide scale is a multiple-layer structure (Figure 6a), the outer layer of which consists of coarse TiO2 crystals (Figure 6b) and is prone to peel off, exposing the inner layer of fine TiO2 and Al2O3 crystals (Figure 6c).

Figure 7 presents the cross-sectional morphology and the corresponding elemental distribution maps of the oxide scale on the Ti-46Al-2Cr-5Nb alloy. The oxide scale is identified to be in the order of the TiO2 layer/Al2O3-rich layer/TiO2 + Al2O3 mixed layer from the outside to the inside. The total thickness of the oxide scale (~13 μm) is indeed less than the detection depth of X-ray, which is consistent with the result of the XRD pattern. An Al-depleted layer that deemed to be α<sup>2</sup> phase [16] is generated in the subsurface of the alloy. Besides, there is a transverse crack between the intermediate and inner layers (Figure 7a). However, the crack is considered to be produced during preparation of the metallographic specimen instead of during oxidation and during cooling, since no such crack is found in the oxide scale formed during heating the alloy to the end of Stage IV and the alloy exhibited much severer oxidation behavior at Stage IV than at Stage III. Thus, the crack is not taken into account when discussing the formation process of the oxide scale at Stage III.

**Figure 6.** (**a**–**c**) Surface morphologies and (**d**) X-ray diffraction (XRD) pattern of the oxide formed during heating the Ti-46Al-2Cr-5Nb alloy to the end of Stage III.

**Figure 7.** (**a**) Cross-sectional morphology and (**b**–**f**) the corresponding elemental distribution maps of the oxide scale formed during heating the Ti-46Al-2Cr-5Nb alloy to the end of Stage III.

The oxide scale formed during heating the Ti-46Al-2Cr-5Nb alloy from room temperature to the end of Stage III (Figures 6 and 7) is considered to be mainly generated at Stage III since the Ti-46Al-2Cr-5Nb alloy exhibits nearly non-oxidation behavior at Stage I and the oxidation mechanism at Stage II is oxygen dissolution in the alloy. The three-layer oxide scale structure formed at Stage III (980–1280 ◦C) is similar to the common oxide scale structure formed during isothermal oxidation at 800–1000 ◦C [14]. Hence, the growth mechanism of the oxide scale at Stage III could be deduced from the isothermal oxidation mechanism at 800–1000 ◦C. Firstly, an Al2O3 film is rapidly generated on the oxygen-saturated Ti-46Al-2Cr-5Nb alloy [39]. Since the Al2O3 film is grown by the inward diffusion of O2<sup>−</sup> along the Al2O3 grain boundaries and the outward diffusion of Al3<sup>+</sup> along the Al2O3 lattice, the Al2O3 grain boundaries suffer compressive stress, resulting in curling deformation or even rupture of the Al2O3 film [40]. Thus, Ti4<sup>+</sup> and O2- would diffuse in opposite directions along the cracks of the Al2O3 film [40]. As a result, an outer TiO2 layer and an inner TiO2 + Al2O3 mixed layer are respectively formed on the outside and the inside of the ruptured Al2O3 film, as shown in Figure 7. The growth process of the oxide scale at Stage III demonstrates that the growth rate of the oxide scale is mainly controlled by the diffusion of Ti4<sup>+</sup> and O2<sup>−</sup> in TiO2, which is consistent with the result of the oxidation activation energy of the alloy at Stage III. Therefore, it is further confirmed that the oxidation mechanism of the Ti-46Al-2Cr-5Nb alloy at Stage III is the growth of the oxide scale dominated by TiO2.

As for the Ti-24Al-15Nb-1.5Mo alloy, the oxide scale formed during heating the alloy to the end of Stage III has a three-layer structure, which is in the order of the TiO2 + Al2O3 mixed layer/TiO2-rich layer/TiO2(Nb, Mo) + Al2O3 mixed layer from the outside to the inside, as shown in Figure 8. Consistent with the oxidation activation energy, the structure of the oxide scale also indicates that the oxidation mechanism of the Ti-24Al-15Nb-1.5Mo alloy is mainly the growth of the oxide scale dominated by TiO2. Besides, a thin β-Ti layer enriched in Al, Nb and Mo elements was formed in the subsurface of the alloy. The microstructure between the β-Ti layer and the matrix is composed of α-Ti(O) grains and a small amount of β-Ti phase enriched in Al, Nb, and Mo in the grain boundaries. As mentioned in Section 4.3, the matrix is dominated by β phase at the end of Stage III. Thus, it is referred that the microstructure between the subsurface of the alloy and the matrix is stabilized to α phase by the inward-diffusing oxygen, while the microstructure of the subsurface still maintains β phase due to the enrichment of Nb and Mo.

**Figure 8.** (**a**) Cross-sectional morphology and (**b**–**f**) the corresponding elemental distribution maps of the oxide scale formed during heating the Ti-24Al-15Nb-1.5Mo alloy to the end of Stage III.

#### 4.3.4. Severe Oxidation Stage (Stage IV)

The two alloys exhibited severe oxidation behaviors when the temperature was raised to Stage IV, while the corresponding temperature range of the Ti-46Al-2Cr-5Nb alloy (1400–1450 ◦C) is much higher than that of the Ti-24Al-15Nb-1.5Mo alloy (1280–1350 ◦C). The oxidation activation energies of the Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys are respectively 985.0 and 608.6 kJ/mol, which is significantly higher than those at Stage III (249.8 and 244.8 kJ/mol). Thus, the oxidation mechanisms at Stage IV should be somewhat different from that at Stage III.

Figure 9 shows the surface morphologies and XRD pattern of the oxide scale formed during heating the Ti-46Al-2Cr-5Nb alloy to the end of Stage IV. As shown in the XRD pattern (Figure 9d), the oxide scale is mainly composed of a large amount of TiO2 as well as a small amount of Al2O3 and β-Al2TiO5. Besides, the content of the γ phase is much lower than that detected after heating to the end of Stage III (Figure 6d). It manifests that the thickness of the oxide scale increased significantly at Stage IV compared with that at Stage III, which is consistent with the drastic increase of the oxidation rate at Stage IV (Figure 3). As shown in Figure 9a,b, the outer layer of the oxide scale is still a spalling-prone TiO2 layer, but it is relatively thin and dense compared with that formed at Stage III (Figure 6a,b). Below the outer layer, a few irregular sintering structures of β-Al2TiO5 coexist with relatively regular crystals of TiO2 and Al2O3, as shown in Figure 9c. It can be seen from the TiO2-Al2O3 phase diagram [41] that the reaction of TiO2 + Al2O3 → β-Al2TiO5 occurs when the temperature is higher than 1200 ◦C. Therefore, the irregular sintering structures of β-Al2TiO5 was produced by the reaction between TiO2 and Al2O3 in the oxide scale at Stage IV (1280–1350 ◦C).

Figure 10 presents the cross-sectional morphology and the corresponding elemental distribution maps of the oxide scale. The oxide scale is identified to be in the order of the TiO2 layer/TiO2 + Al2O3 mixed layer/Al2O3-rich layer/TiO2-rich layer/TiO2(Nb, Cr) + Al2O3 mixed layer from the outside to the inside. In addition, an Al-depleted layer enriched in Nb and Cr elements is generated in the subsurface of the alloy. Moreover, Al2O3 oxides are dispersed as islands in the subsurface, indicating the occurrence of internal oxidation at Stage IV. Since internal oxidation deteriorates the oxidation resistance of TiAl-based alloys [14], the severe oxidation behavior of the Ti-46Al-2Cr-5Nb alloy at this stage resulted from the internal oxidation of Al, which is probably the cause for the much higher oxidation energy of the Ti-46Al-2Cr-5Nb alloy at Stage IV than that at Stage III.

**Figure 9.** (**a**–**c**) Surface morphologies and (**d**) XRD pattern of the oxide scale formed during heating the Ti-46Al-2Cr-5Nb alloy to the end of Stage IV.

**Figure 10.** (**a**) Cross-sectional morphology and (**b***–***f**) the corresponding elemental distribution maps of the oxide scale formed during heating the Ti-46Al-2Cr-5Nb alloy to the end of Stage IV.

As for the Ti-24Al-15Nb-1.5Mo alloy, the surface morphologies and XRD pattern of the oxide scale formed during heating the alloy to the end of Stage IV are shown in Figure 11. The XRD result shows that the oxide scale is mainly composed of TiO2 accompanied with a small amount of β-Al2TiO5 (Figure 11d). The outer layer of the oxide scale is crimping (Figure 11a), consisting of coarse rod-like TiO2 particles and ridged structures composed of fine β-Al2TiO5 particles (Figure 11c). The outer layer is prone to exfoliation (Figure 11a), revealing the layer of fine TiO2 particles (Figure 11b).

**Figure 11.** (**a**–**c**) Surface morphologies and (**d**) XRD pattern of the oxide formed during continuously heating the Ti-24Al-15Nb-1.5Mo alloy to the end of Stage IV.

Figure 12 presents the cross-sectional morphology and the corresponding elemental distribution maps of the oxide scale on the Ti-24Al-15Nb-1.5Mo alloy. The structure of the oxide scale is in the order of the TiO2 + Al2TiO5 layer/TiO2 layer/coarse Al2O3 layer/porous TiO2(Nb, Mo) layer from the outside to the inside. The porous TiO2(Nb, Mo) inner layer and the coarse Al2O3 immediate layer indicate that the Al2O3 particles formed in the inner layer dissolved in the surrounding TiO2(Nb, Mo), then migrated outward and re-precipitated in the immediate layer at Stage IV, the reason for which is that the decreasing oxygen pressure around the inner layer with the thickening of the oxide scale increases the solubility of Al2O3 in TiO2 and reduces the stability of Al2O3 [17]. It is reported that the dissolution, migration and re-precipitation of Al2O3 destroys the oxygen-blocking ability of the original Al2O3 barrier layer, leading to breakaway oxidation [17,42]. However, there is a transverse crack below the oxide scale and whether the crack contributed to the severe oxidation should be discussed. According to the crimping structure and the locally spalling morphology of the oxide scale (Figure 11a), the crack is considered to be induced during cooling after oxidation. The reason is that the oxide scale and the matrix were respectively subjected to compressive and tensile stresses during cooling since the thermal expansion coefficient of the oxide scale is generally smaller than that of the metal, and curling deformation was prone to occur in the oxide scale for stress release. Thus, the severe oxidation behavior of the Ti-24Al-15Nb-1.5Mo alloy at Stage IV is independent of the formation of the transverse crack. Consequently, the severe oxidation at this stage is due to the dissolution, migration and re-precipitation of Al2O3, which might lead to the higher oxidation energy at Stage IV than that at Stage III.

In addition, a thin α-Ti(O) layer was formed at the interface between the oxide scale and the alloy, beneath which is a 10-μm-thick layer composed of β-Ti rich in Nb, Mo, Al elements (referred to as *i* in Figure 12a) and several fine α-Ti grains with an orientation nearly perpendicular to the matrix surface (referred to as *ii* in Figure 12a). The microstructure between the β-Ti layer and the matrix is still composed of α-Ti(O) grains (referred to as *iii* in Figure 12a) and a small amount of β-Ti phase in the grain boundaries (referred to as *iv* in Figure 12a). The chemical compositions of these structures are shown in Table 2. The microstructures demonstrate that more oxygen diffused into the subsurface of the alloy at Stage IV in comparison with at Stage III, resulting in the transition from β-Ti to α-Ti(O) in the near subsurface layer.

**Figure 12.** (**a**) Cross-sectional morphology and (**b**–**f**) the corresponding elemental distribution maps of the oxide scale formed during heating the Ti-24Al-15Nb-1.5Mo alloy to the end of Stage IV.


**Table 2.** Chemical compositions of the microstructures in the Ti-24Al-15Nb-1.5Mo alloy examined by an electron probe microanalyzer (EPMA) after heating the alloy to the end of Stage IV.

#### 4.3.5. Decelerated Oxidation Stage (Stage V)

When the temperature increased to Stage V (1350–1450 ◦C), the oxidation rate of the Ti-46Al-2Cr-5Nb alloy decreased remarkably. Figure 13 presents the surface morphologies and XRD pattern of the oxide scale formed during heating the Ti-46Al-2Cr-5Nb alloy to the end of Stage V. There is no obvious exfoliation morphology on the surface of oxide scale (Figure 13a). Further, the scale becomes denser (Figure 13b) and the crystals in the surface of the scale become larger (Figure 13c) compared with those formed at Stage IV. As shown in the XRD pattern (Figure 13d), the oxide scale consists of a large quantity of β-Al2TiO5 as well as some TiO2 and Al2O3, indicating that the sintering reaction between TiO2 and Al2O3 in the oxide scale aggravated at Stage V.

**Figure 13.** (**a**–**c**) Surface morphologies and (**d**) XRD pattern of the oxide scale formed during heating the Ti-46Al-2Cr-5Nb alloy to the end of Stage V.

Figure 14 shows the cross-sectional morphology and the corresponding elemental distribution maps of the oxide scale. The oxide scale from the outside to the inside is identified to be in the order of the TiO2 layer/Al2O3-rich layer/TiO2-rich layer/Al2TiO5 layer/mixed layer of TiO2 and fine Al2O3 flakes. It is interesting that the TiO2 (referred to as *i* in Figure 14a) and the fine Al2O3 flakes (referred to as *ii* in Figure 14a) distribute alternately in the inner layer, the compositions of which are presented in Table 3. White coarse particles (referred to as *iii* in Figure 14a) with an orientation perpendicular to the oxide/substrate interface were generated in the subsurface of the substrate. The detected composition as shown in Table 3 manifests that the white particles might be an alloy or an intermetallic of Nb. Moreover, there are a lot of fine Al2O3 flakes distributing around the white coarse particles. Besides, it should be mentioned that there is a transverse crack in the oxide scale, which is not found in the oxide scale formed during heating the alloy to the end of Stage IV. This crack is considered to be produced during cooling after oxidation or during the preparation of metallographic specimens instead of during oxidation, since the oxidation rate at Stage V was significantly reduced in comparison with at Stage IV.

Through comparing the cross-sectional structures of the oxide scales respectively formed by heating the Ti-46Al-2Cr-5Nb alloy to the end of Stage IV and Stage V (Figures 10 and 14), the growth process of the oxide scale at Stage V can be inferred as follows: (1) oxygen from the atmosphere diffused into the alloy and the Al-depleted zones were preferentially oxidized nearby the Al2O3 flakes which were generated due to the internal oxidation at Stage IV, thus forming the inner oxide layer structure where TiO2 and Al2O3 flakes distributed alternately; (2) due to the limited solubility of the alloying elements (such as Nb, Cr and Al) in TiO2, the alloying elements diffused towards the underlying substrate during the growth of the oxide scale, resulting in the formation of massive white coarse particles of Nb alloy or intermetallic in the newly-formed subsurface of the substrate; (3) the continuous inward diffusion of oxygen further induced the occurrence of internal oxidation in the subsurface, leading to the formation of fine Al2O3 flakes around the white coarse particles and even

the underlying substrate. However, at the same time of the oxide growth, the sintering reaction between TiO2 and Al2O3 in the oxide scale (mainly in the TiO2 + Al2O3 mixed layer) aggravated and an β-Al2TiO5-rich layer was formed in the oxide scale (Figure 14a). Since Al2TiO5 has higher oxygen resistance than TiO2 [5,43], the formation of the β-Al2TiO5-rich layer could effectively slow down the inward diffusion of oxygen and the oxidation rate decreased. Hence, the decelerated oxidation behavior of the Ti-46Al-2Cr-5Nb alloy at Stage V is due to the generation of an oxygen-barrier β-Al2TiO5-rich layer in the oxide scale by the reaction between TiO2 and Al2O3 in large scales.

**Figure 14.** (**a**) Cross-sectional morphology and (**b**–**f**) the corresponding elemental distribution maps of the oxide scale formed during heating the Ti-46Al-2Cr-5Nb alloy to the end of Stage V.

**Table 3.** Chemical compositions of the microstructures in the inner layer of the oxide and in the



#### *4.4. Reasons for the Occurrence of Internal Oxidation in the Ti-46Al-2Cr-5Nb Alloy*

In comparison with the Ti-24Al-15Nb-1.5Mo alloy, the Ti-46Al-2Cr-5Nb alloy suffered catastrophic oxidation at the temperature range of 1280–1350 ◦C due to the occurrence of internal oxidation. In order to improve the non-isothermal oxidation resistance of the Ti-46Al-2Cr-5Nb alloy, it is essential to shed light on the reasons for the occurrence of internal oxidation in the Ti-46Al-2Cr-5Nb alloy instead of in the Ti-24Al-15Nb-1.5Mo alloy. In accordance to the occurrence conditions of internal oxidation [44], it seems that the Ti-24Al-15Nb-1.5Mo alloy mainly dominated by α<sup>2</sup> phase should be more prone to suffer internal oxidation than the Ti-46Al-2Cr-5Nb alloy mainly dominated by γ phase, since both the oxygen solubility and the oxygen diffusion rate in α<sup>2</sup> phase are higher than those in the γ phase

and the Al content in the α<sup>2</sup> phase is lower than that in the γ phase. However, this is not the case. Therefore, the actual phases in the Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys at the temperature range of 1280–1350 ◦C should be taken into account.

The initial temperature for the occurrence of internal oxidation in the Ti-46Al-2Cr-5Nb alloy (1280 ◦C) is close to the temperature of γ + α → α phase transition (about 1226 ◦C, see Figure 3a). Hence, before the occurrence of internal oxidation (T < 1280 ◦C), the substrate of the Ti-46Al-2Cr-5Nb alloy was mainly dominated by γ phase and the subsurface of the alloy was dominated by an Al-depleted layer of α<sup>2</sup> phase [16], as illustrated in Figure 15a. However, when the temperature rose above the initial temperature for the occurrence of internal oxidation (T > 1280 ◦C), the main phases both in the substrate and the subsurface Al-depleted layer of the Ti-46Al-2Cr-5Nb alloy transformed to the α phase, as illustrated in Figure 15b. As for the Ti-24Al-15Nb-1.5Mo alloy at the temperature range for the occurrence of internal oxidation in the Ti-46Al-2Cr-5Nb alloy (1280–1350 ◦C), the substrate was mainly dominated by β phase and the subsurface was also dominated by β phase due to the enrichment of Mo and Nb elements, even though the microstructure between the subsurface and the substrate was composed of α-Ti(O) grains and a small amount of β-Ti phase, as illustrated in Figure 15c. It is obviously seen from Figure 15 that the phases in the subsurface where internal oxidation occurs are different. Hence, in order to reveal the effect of the phases in the subsurface on the internal oxidation tendency, the four phases involving in the Ti-Al alloys (γ, α2, α and β) were studied.

**Figure 15.** Schematic diagrams of the phase structures for (**a**) the Ti-46Al-2Cr-5Nb alloy when T < 1280 ◦C, (**b**) the Ti-46Al-2Cr-5Nb alloy when T > 1280 ◦C and (**c**) the Ti-24Al-15Nb-1.5Mo alloy at the temperature range of 1280–1350 ◦C.

The tendency of internal oxidation for the different phases in the Ti-46Al-2Cr-5Nb alloy at 1280 ◦C can be identified based on Wagner's theory [44]. The critical criterion for the transition from internal oxidation to external oxidation is:

$$N\_{\rm Al} > N\_{\rm crit(Al)} = \left(\frac{\pi \text{g}}{2} \frac{V\_{\rm m}}{V\_{\rm ox}} \frac{D\_{\rm O}}{D\_{\rm Al}} N\_{\rm O}\right)^{0.5} \tag{2}$$

where *N*crit(Al) is the critical Al content required for the transition from internal oxidation to external oxidation, g\* is a constant factor in the case of an index of 0.3, *V*<sup>m</sup> and *V*ox are respectively the molar volumes of the Ti-46Al-2Cr-5Nb alloy and Al2O3 oxide (in cm3/mol, Vox = 25.5 cm3/mol), *D*<sup>O</sup> and *D*Al are respectively the diffusion coefficients of O and Al atoms in the lattice of the alloy, *N*<sup>O</sup> and *N*Al are respectively the oxygen content and Al content in the subsurface of the alloy. The composition in the subsurface of the Ti-46Al-2Cr-5Nb alloy after heating to 1280 ◦C was detected to be Ti-29.8Al-3.6O-3.0Cr-5.4Nb (in at.%) by EPMA, thus the values for *N*<sup>O</sup> and *N*Al are respectively 0.036 and 0.298. The values for the other parameters involved in Equation (2) are given in Table 4. It should be noted that the specific value for the oxygen diffusion coefficient in the γ phase has not been reported, but it is readily inferred that the oxygen diffusion coefficient in the γ phase is lower than that in the α<sup>2</sup> phase from their lattice structure differences, as mentioned in Section 4.3.2.


**Table 4.** Critical Al contents required for the transition from internal oxidation to external oxidation for the different phases in the subsurface of the Ti-46Al-2Cr-5Nb alloy at 1280 ◦C.

Table 4 presents the calculation results of the critical Al contents required for the transition from internal oxidation to external oxidation for the different phases in the subsurface of the Ti-46Al-2Cr-5Nb alloy at 1280 ◦C. The critical Al contents required for the α<sup>2</sup> and γ phases are respectively 0.085 and less than 0.05, which is much lower than the actual Al content in the subsurface of the substrate (*N*Al = 0.298). It indicates that if the subsurface of the Ti-46Al-2Cr-5Nb alloy is dominated by γ phase and/or α<sup>2</sup> phase, an external oxide scale would form on the alloy and internal oxidation could not occur. However, the critical Al contents required for the α and β phases are respectively 2.08 and 1.62, which is impractical since the contents are more than 1. This situation is caused by the fact that Wagner's theory of internal oxidation is based on a large number of idealized conditions that often could not be met in the actual system [44]. Nevertheless, Wagner's critical criterion can still reflect the tendency of internal oxidation for the different phases in the subsurface of the Ti-46Al-2Cr-5Nb alloy. The higher the required critical Al content, the more easily internal oxidation will occur. Therefore, the tendency of internal oxidation for the phases is in the order of α > β > α<sup>2</sup> > γ. Thus, it is easy to understand why internal oxidation occurred in the Ti-46Al-2Cr-5Nb alloy when the temperature exceeded 1280 ◦C, since the main phase in the subsurface of the Ti-46Al-2Cr-5Nb alloy changed from α<sup>2</sup> phase to α phase when the temperature was higher than 1280 ◦C. In addition, the main phase in the substrate of the Ti-24Al-15Nb-1.5Mo alloy was β phase at the temperature range of 1280–1350 ◦C due to the enrichment of large amounts of β-stabilizing elements (Nb and Mo), so that no internal oxidation occurred in this alloy. Consequently, it is concluded that the formation of α phase in the subsurface is the basic reason for the occurrence of internal oxidation in the Ti-46Al-2Cr-5Nb alloy. The tendency of internal oxidation in the TiAl-based alloys could be reduced through avoiding the formation of α phase in the subsurface by optimizing alloy ingredients, such as increasing the Al content or adding β-stabilizing elements. However, it should be noted that the optimization of alloy ingredients is complicated and should be further investigated since not only the tendency of internal oxidation, but also other properties of the alloys such as the mechanical properties, should be taken into consideration in practical engineering.

#### **5. Conclusions**

1. The non-isothermal oxidation behaviors of the Ti-46Al-2Cr-5Nb and Ti-24Al-15Nb-1.5Mo alloys are similar when the temperature is below 1280 ◦C, while the Ti-46Al-2Cr-5Nb alloy exhibits poorer oxidation resistance than the Ti-24Al-15Nb-1.5Mo alloy when the temperature exceeds 1280 ◦C, even though the oxidation rate of the Ti-46Al-2Cr-5Nb alloy decreases significantly when the temperature is above 1350 ◦C.


**Author Contributions:** Conceptualization, G.M. and P.L.; methodology, P.O.; investigation, P.O.; writing—original draft preparation, P.O.; writing—review and editing, G.M. and P.L.; supervision, X.H. and J.C.; project administration, L.H.; funding acquisition, G.M. and P.L.

**Funding:** This research was funded by the National Natural Science Foundation of China (Grant No. 51471155) and the Aviation Innovation Foundation of China (Grant No. 2014E62149R).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
