Interfacial Cracking Failure Mechanism of Chromium-Coated Zircaloy Cladding for ATF Materials

Abstract

In this study, the interfacial cracking mechanism of Cr-coated Zircaloy cladding tubes was investigated through a combination of tensile testing, advanced microstructural characterization, and first-principle calculations. The results demonstrate the presence of an interlayer (33 nm) between the Cr coating and the Zr substrate, which consists of Cr and O elements. In comparison with the coherent Cr/interlayer interface, the incoherent interlayer/Zr interface is conducive to crack initiation and propagation, resulting from local high strain concentration generated by incompatible deformation between the interlayer and Zr substrate, and O-induced reductions in Cr/Zr interfacial cohesive strength.

1. Introduction

Nuclear energy is gaining significant interest as a form of clean, low-carbon, economical, and sustainable energy. In particular, the development of accident-tolerant fuel (ATF) technology can enhance the fundamental safety of mainstream light-water reactors [1,2]. Compared with the potential CrN, CrAl, TiAlN, and MAX phase ATF coating candidates, Cr-coated zirconium (Zr) alloy cladding exhibits the greatest safety advantages in terms of design basis, expansion, and avoiding serious accidents. It is labelled as the most promising ATF material [3,4].

To date, studies on the effect of Cr coating on Zr alloys have focused on two topics, namely, the oxidation behaviors [5,6,7,8,9,10,11] and mechanical properties [2,12,13,14,15,16] of Zr alloys in the absence or presence of coatings. In these situations, the prerequisite for the improvement of oxidation resistance and mechanical property is the excellent interface structure between Cr coatings and Zr alloy substrates. As reported in many studies [17,18,19,20,21], a Cr-Zr interlayer between the Cr coating and Zr alloy substrate is generally formed due to elemental mutual diffusion driven by the Cr/Zr concentration gradient. Under oxidation and loading conditions, the interlayer plays a crucial role in the formation of cracks. For example, Gong et al. [22] reported that cavities were formed either in the interior of the interlayer or along the interface between the interlayer and Zr alloy substrate in high-temperature steam oxidation environments. Jiang et al. [23] demonstrated that cracks easily initiated from the interior of the Cr-Zr interlayers of pre-oxidized Cr-coated Zr alloy samples during tensile tests and three-point bending tests. For hoop tensile and compressive tests, Li et al. [12] indicated that cracking from the interfaces between Cr coatings and Zr substrates was the dominant failure mode, which was attributed to nano-scaled interface defects. Similar results were reported in Refs. [19,21,24,25]. In brief, previous studies validate the dependence of the crack formation of Cr-coated Zircaloy on the interlayer. However, the question that where nano-scaled cracks preferentially initiate has not been addressed. Also, the underlying deformation cracking mechanism is not well clarified. Different from previous studies concentrating on the effects of nano-scaled Al-enriched loose layers on the cracking failure of ATF coatings [12], the interlayer of commercial Cr-coated Zr alloy cladding tubes was analyzed by advanced characterization techniques in this study. Then, interfacial deformation cracking behavior at the nanoscopic scale of pre-tension samples was investigated, and the corresponding cracking mechanism was elucidated based on density functional theory (DFT) calculations.

2. Experimental Procedures

The material tested in this study consisted of Zr1Nb0.01Cu (wt. %) cladding tubes, whose outer and inner diameters were 9.98 mm and 9.55 mm, respectively. The Cr coating was deposited on the outer surface of the tubes through an unbalanced magnetron sputtering system (25 °C). The detailed coating preparation processes were described in Ref. [12]. Tensile ring-shaped samples with a height of 2 mm were cut from the cladding tubes by electrical discharge machining. The samples were mounted in a special holder to maintain a uniaxial stress state [12] and then subjected to tensile tests on a universal testing machine, as schematically shown in Figure 1a. During tensile tests, a cross-head displacement rate of 0.3 mm min⁻¹ was set [12]. Three tensile tests were conducted for each condition. To reveal the interfacial cracking behavior of Cr-coated Zr alloy, one sample was pre-stretched to a displacement of 0.55 mm and then unloaded. It was worth noting that the sample under a pure tensile load was analyzed, as shown in Figure 1b. Furthermore, the extracted TEM foil samples were situated at the equatorial position as marked by the dotted yellow circle in Figure 1b, where a polar coordinate with an equatorial position (θ = 0°) was defined. The foil samples were extracted by a focus ion beam instrument (FIB, Thermo Fisher Helios G4 UC, Thermo Fisher Scientific Inc., Waltham, MA, USA) and characterized by a transmission electron microscope (TEM, Thermo-Fisher Talos F200X, Thermo Fisher Scientific Inc., Waltham, MA, USA) in bright-field, high-angle annular dark-field (HAADF), and selected-area electron diffraction (SAED) modes. The specific preparation and testing processes referred to previously published Ref. [12]. First-principle DFT calculations were performed using the Vienna Ab Initio Simulation Package (VASP) code [26], in which the projector-augmented wave method was implemented to model electron-ion interactions [27]. Also, the generalized gradient approximation for the exchange correlation energy was used [28]. The convergence criteria for the total energy and atomic force were 10⁻⁵ eV and 0.02 eV/Å, respectively, and the cut-off energy was 400 eV [29,30].

3. Results

Figure 2a,b show the macroscopic appearance of a Cr-coated Zr alloy cladding tube along a cross-section, indicating that a Cr coating with an average thickness of 8 μm is homogenously deposited on the outer surface of the cladding tube. The Cr grains exhibit a random orientation without a texture (Figure 2c). Due to the limited resolution of SEM and EBSD techniques, the interface between the Cr coating and the Zr substrate is not well characterized (Figure 2b). Consequently, advanced STEM-EDS analysis was utilized (Figure 2(d1–d4)). The results demonstrate a three-layer structure consisting of the Cr coating, the interlayer with a thickness of 33 nm (Figure 2(e1–e3)), and the Zr substrate. Here, the interface between the Cr coating and the interlayer is designated as the upper interface, and the interlayer/Zr substrate interface is termed as the lower interface. Differing from previous results that the interlayer consists of Cr and Zr elements [19,21,24,25], only Cr (96.53%) and O (3.47%) elements of the interlayer in weight fractions were detected. The chemical composition of the Cr coating (98.72% Cr and 1.27% O) is similar to that of the interlayer, although there are differences in their O contents. Additionally, the O element is also observed in the Zr substrate, and its content gradually decreases from the lower interface to the interior of the Zr substrate (Figure 2(d3,e3)). Based on the differences in the O element distribution, it is suggested that the entry of O precedes the Cr coating deposition. However, the specific cause and the source of the O element have not been clarified. They may be the result of an incomplete surface polishing or an incomplete vacuum [12].

Figure 3a shows tensile stress–displacement curves of ring-shaped Cr-coated Zr alloy, indicating that the yield strength and tensile strength are 454 MPa and 464 MPa, respectively. Also, the curve of the pre-tensioned sample is given for comparison. Figure 3(b1,b2) present micrographs of the Cr coating and Zr substrate of the undeformed sample, respectively, revealing the elongated appearance of the Cr grains and a dense of dislocations in the Zr substrate. Additionally, bcc-structured Cr coating and hcp-structured Zr substrate are demonstrated by the SAED patterns (Figure 3(b3,b4)). After the pre-tension, the Cr grains in the coating are further elongated (Figure 3(c1)) and the deformation bands along multiple directions in the Zr substrate are generated (Figure 3(c2)).

The deformation behavior of the upper/lower interfaces is our focus and interest. Figure 4 shows the interfacial deformation behaviors of Cr-coated Zr alloy with/without pre-tension. For the sample without pre-tension, the interlayer bonds well with the Cr coating and Zr substrate (Figure 4(a1)), as evidenced by the absence of cracks. Further, HRTEM images demonstrate that the upper interface exhibits a coherent relation resulting from the shared Cr matrix between the Cr coating and the interlayer (Figure 4(a2)), while the lower interface between bcc-structured Cr and hcp-structured Zr is incoherent (Figure 4(b2)). For the pre-tension sample, the upper interface remains well bonded. However, crack initiation and propagation along the lower interface take place (Figure 4(c1)). After comparison with these results, it is confirmed that the incoherent lower interface is more conducive to cracking than coherent upper interfaces.

4. Discussion

To explain the interfacial cracking mechanism along the incoherent lower interface, two aspects have been taken into consideration, i.e., high strain concentration along the lower interface and the reduction of lower interface cohesive strength resulting from the presence of oxygen. For the former, it is established that adjacent lattices of coherent interfaces match well with each other with a small lattice mismatch [31]. Therefore, expectedly, few dislocations are observed in the vicinity of the coherent upper interface (Figure 4(a2)). In light of incoherent lower interfaces, they exhibit a large lattice mismatch, and adjacent lattices along the interfaces still maintain their original lattices and rigidly stack together, resulting in high dislocation density (Figure 4(b2)) [32]. After pre-tension deformation, the upper interface remains coherent to some extent with a few dislocations (Figure 4(c2)), whereas a dense of dislocations near the incoherent lower interface are detected (Figure 4(d2)). Similarly, dislocations in the Cr coating, the interlayer, and the Zr substrate of the undeformed/deformed samples were analyzed and compared, indicating higher dislocation density in the pre-tensioned sample, especially in the Zr substrate (Figure 4(d1)) and the interlayer (Figure 4(d4)). The strain value based on the geometrical phase analysis (GPA) of HAADF scanning electron microscopy (STEM) was analyzed [33], and corresponding 2D strain maps of the in-plane ɛ_XX are shown in Figure 4(a3,b3,c3,d3). For the sample without pre-tension, the strain of the coherent upper interface is small and relatively homogenous (Figure 4(a3)), and the incoherent lower interface exhibits a large strain (Figure 4(b3)). Compared with Figure 4(a3,b3), the pre-tension increases the strain value of both interfaces, especially in incoherent lower interfaces (Figure 4(d3)). This result is consistent with the dislocation density mentioned above. In the process of deformation, the dislocations derived from the interlayer and Zr substrate are hindered and accumulated at the lower interfaces, leading to dislocation pile-ups at the interface with the crack initiation [34].

For the latter, the interface structure between bcc-structured Cr and hcp-structured Zr is constructed based on previous experimental results (Figure 2(e3), Figure 3(b3,b4) and Figure 4(b2)), in which blue, green, and red colors represent Cr, Zr, and O atoms and α, β, and γ indicate the bcc-Cr, hcp-Zr, and their interface, respectively (Figure 5a). The absorption energy ($∆E_a$) is calculated by Equation (1) as [30]:

$$∆E_a=E_{interface}^O-E_{interface}-E_O$$

(1)

where $E_{Interface}^O$ and $E_{interface}$ are the total energy of the interface with and without O, respectively. $E_O$ is the total energy of a single O atom. Note that the negative value of $∆E_a$ corresponds to the strong adsorption effect of O.

The cohesive energy ($∆E_b$) can be obtained by Equation (2) as [29]:

$$∆E_b=E_{\alpha }+E_{\beta }-E_{\alpha +\beta }$$

(2)

where the $E_{\alpha }$ and $E_{\beta }$ are the total energy of α and β, respectively. $E_{\alpha +\beta }$ is the total energy of the interface. α(O) is the doping O atom in the α surface. β(O) is the doping O atom in the β surface. Also, a positive value of $∆E_b$ means a high cohesive strength of the interface.

Figure 5b,c show the bond length and adsorption energy of O in the α, β, and γ systems, revealing that the bond length of Zr-O is greater than that of Cr-O. Additionally, the adsorption energies of O in the α, β, and γ structures are −9.31 eV, −9.92 eV, and −7.84 eV, respectively. These results demonstrate that O is easily adsorbed in the Zr structure, followed by the Cr structure and then the interface. Figure 5d presents the cohesive energy of the Cr/Zr interface in the absence/presence of O. The cohesive energy of the interface without O is 3.53 eV, while it reduces to 2.07 eV and 1.45 eV for O adsorption on Cr and Zr surfaces, respectively, indicating that the doping of O at the interfaces significantly reduces their cohesive strength, which is also in favor of the interfacial cracking in the process of pre-tension deformation.

5. Conclusions

In summary, Cr-coated Zr alloy presents a three-layer structure consisting of the Cr coating, the interlayer, and the Zr substrate. The interlayer, with an average thickness of 33 nm, consists of Cr and O elements. The Cr coating/interlayer interface is coherent, whereas the interlayer/Zr substrate interface is incoherent. During deformation, local strain concentration is easily generated in the incoherent interface, favoring cracking. Meanwhile, DFT calculations demonstrate that the doping of O in the interfaces weakens their cohesive strength, also supporting cracking along incoherent interlayer/Zr interfaces. These results reveal the root cause of the failure of Cr coatings deposited on cladding tubes, providing a potential strategy for optimizing the mechanical properties of Cr-coated Zircaloy cladding for ATF materials. For example, it is possible to change the composition of the interlayer, such as the elimination of O in the interface, to increase the interface strength. Additionally, it is possible to achieve reductions in strain accumulation through the addition of a buffer layer.