Atomistic weak interaction criterion for the specificity of liquid metal embrittlement

www.nature.com/scientificreports OPEN Atomistic weak interaction criterion for the specificity of liquid metal embrittlement Masatake Yamaguchi1,2,3*, Tomohito Tsuru3,4, Mitsuhiro Itakura5 & Eiji Abe2,6 Liquid metal embrittlement (LME) occurs in some solid–liquid metal elements’ couples (e.g., Fe-Zn and Al-Ga), called specificity. Although some material parameters like solubility and bonding energy were suggested as controlling factors, none could be attributed satisfactorily. Here we have unveiled the primary factor that governs the specificity of LME. From first-principles calculations compared with a systematic surveillance test result, we found that the grain-boundary (GB) adsorption energy shows near-zero values in all embrittling couples; the interaction between solid and liquid metal atoms is weak when an atom from the liquid state penetrates the grain boundary of the solid. Furthermore, we found that the calculated surface adsorption energy that promotes bond-breaking does not correlate to the specificity. Therefore, we consider that the penetration of a liquid metal atom surrounded by weakly interacting solid metal atoms is necessary before the bond-breaking assisted by surface adsorption occurs at a microcrack tip. This mechanism is also applicable for transgranular cracking along low-energy boundaries and crystal planes. While liquid metal atoms penetrate and diffuse into solid GB macroscopically before cracking, liquid metal’s surface adsorption stronger than GB adsorption should promote the bond-breaking of solid metal. In conclusion, the atomistic penetration precedes the surface-adsorption-assisted bond-breaking and controls the specificity of LME. Solid metals are susceptible to brittle fracture when they come in contact with certain liquid metals (e.g., Fe–Zn and Al–Ga). This phenomenon is called liquid metal embrittlement, discovered in 18741. One of the most severe embrittling cases is polycrystalline aluminum alloys embrittled by liquid gallium; liquid gallium penetrates aluminum GBs macroscopically at high s peed2. Zinc atoms’ penetration, diffusion, and subsequent liquid zinc penetration into steel GBs have been observed because molten zinc embrittlement has attracted much attention by increasing steels’ s trength3,4, especially in the automotive industry. Luo et al. have clarified that bismuth forms bilayer interfacial phase along nickel G Bs5. In these cases, the subsequent embrittlement occurs due to the change in GB strength induced by liquid metal atoms existing in the GB before cracking. On the other hand, embrittlement occurs even when exposed to the liquid metal without detectable atomic diffusion or liquid penetration into GBs, lath boundaries in martensitic steel, and crystal lattice planes. Although liquid mercury atoms do not seem to penetrate and diffuse into high-strength steel’s GBs and lath boundaries, embrittlement does occur along those b oundaries6. Even single crystal aluminum exhibits embrittlement by liquid gallium7 after significant plastic deformation. The extremest example is the mercury embrittlement of single and polycrystal zinc; fracture occurs within seconds after coating mercury on stress-loaded zinc8. In this way, LME occurs even if liquid metal atoms do not exist along the cracking path before fracture. One of the curious phenomena that many researchers have tried to solve is why embrittlement occurs or does not occur depending on a given solid–liquid metal elements’ couple under a specific test condition. Such selectivity is called the specificity of L ME9, which has been known for over a century, yet it is not well understood. Some material parameters suggested as controlling factors are solubility, interatomic bonding energy, and intermetallics formation. However, none of them could be attributed to the specificity of LME s atisfactorily10,11. We considered that the crucial atomistic mechanism of LME must be hidden behind the mystery of the LME specificity. 1 Center for Computational Science & e‑Systems, Japan Atomic Energy Agency, 2‑4 Shirakata, Tokai‑mura, Naka‑gun, Ibaraki 319‑1195, Japan. 2Department of Materials Science and Engineering, The University of Tokyo, 7‑3‑1 Hongo, Bunkyo‑ku, Tokyo 113‑8656, Japan. 3Elements Strategy Initiative for Structural Materials, Kyoto University, Yoshida‑honmachi, Sakyo‑ku, Kyoto 606‑8501, Japan. 4Nuclear Science and Engineering Center, Japan Atomic Energy Agency, 2‑4 Shirakata, Tokai‑mura, Naka‑gun, Ibaraki 319‑1195, Japan. 5Center for Computational Science & e‑Systems, Japan Atomic Energy Agency, 178‑4 Wakashiba, Kashiwa, Chiba 277‑0871, Japan. 6Research Center for Structural Materials, National Institute of Materials Science, 1‑2‑1 Sengen, Tsukuba, Ibaraki 305‑0047, Japan. *email: Scientific Reports | (2022) 12:10886 | https://doi.org/10.1038/s41598-022-10593-2 1 Vol.:(0123456789) www.nature.com/scientificreports/ a c Fe 4.0 Bi, 271, n 2.0 Al Tl, 304, n Pb, 328, n 3.0 Na, 98, n Te, 450, n Hg, -39, n Sn, 232, e 1.0 Cd, 321, n Ga, 30, e -1.0 Se, 221, e -1.5 Embriˆlement Non-embriˆlement Li, 181, n 0.0 -1.0 -0.5 1.0 1.5 Hg, -39, n 0.5 Cd, 321, n Se, 221, e 0.0 Zn, 420, n -0.5 Embriˆlement Non-embriˆlement Li, 181, e -1.0 0.5 Tl, 304, n Na ,98, n Ga, 30, n Zn, 420, e 0.0 Bi, 271,n Sn, 232, n 1.0 In, 157, n La‰ce dissoluon energy (eV/atom) La‰ce dissoluon energy (eV/atom) 2.0 Pb, 328, n In, 157, n 1.5 2.5 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 Grain-boundary adsorpon energy (eV/atom) Grain-boudary adsorpon energy (eV/atom) d b Fe 0.0 Hg Cd Zn Tl Na Li In Pb Sn Ga Bi Te Al Se Zn 0.0 Cd Hg In Ga Tl Pb Sn Bi Li Na Se -0.2 -0.4 -1.0 Surface adsorpon energy (eV/atom) Surface adsorpon energy (eV/atom) -0.5 -1.5 Embriˆlement -2.0 Non-embriˆlement -2.5 -0.6 -0.8 -1.0 Embriˆlement -1.2 Non-embriˆlement -1.4 -1.6 -3.0 Figure 1.  Calculated atomistic energies of liquid metal elements compared with the specificity of LME. (a) GB adsorption energies on FeΣ3(111) symmetrical tilt GB and lattice dissolution energies in Fe of liquid metal elements. (b) Surface adsorption energies of liquid metal elements on Fe(111) surface, which is a fracture surface of FeΣ3(111) GB. Inserted figures show computational cells of the Fe-Pb case. (c) GB adsorption energies on AlΣ5(012) symmetrical tilt GB and lattice dissolution energies in Al of liquid metal elements. (d) Surface adsorption energies of liquid metal elements on Al(012) surface, which is a fracture surface of AlΣ5(012) GB. Negative energy indicates attractive interaction between atoms. Embrittlement or Non-embrittlement refers to Rostoker et al.’s specificity data of LME10, as shown in Table 1. The number in the label box is the melting temperature (degrees of Celsius). The green box indicates that a binary intermetallic compound exists (e), whereas the other box does not (n)22. Test temperature (℃) 30 50 125 180 210 250 260 (...truncated)