Structure and corrosion of iron-based metallic glasses during crystallization
M.J. Duarte1,2,3, S. Klemm1, S. Borodin1, P. Keil1, K. Mayrhofer1, M. Stratmann1, A.H. Romero2, D. Crespo3, J. Serrano4, P. Choi5, S.S.A. Gerstl5, D. Raabe5, F. U. Renner1
1) Department of Interface Chemistry and Surface Engineering 2) CINVESTAV-IPN, Queretaro, Mexico 3) Universitat Politècnica de Catalunya, Castelldefels Spain 4) ICREA- Universitat Politècnica de Catalunya, Castelldefels Spain, 5) Department of Microstructure Physics and Metal Forming
- Figure 1. SAM map of fully crystallized MG Fe50Cr15Mo14C15B6 (800°C, 20 min).
Metallic glasses (MGs), often referred to as amorphous metals, lack long-range order and can be considered as frozen atomic configurations of a liquid phase (1). The combination of metal properties with the formability of usual oxide glasses makes research on MGs very attractive both for fundamental understanding and for the development of tailored technological materials in metallurgy. The absence of microstructural features such as crystal plains, grain and phase boundaries or dislocations, contributes to their special properties (2-4). Moreover, the properties of amorphous alloys also depend on their thermal history. Upon annealing, thermodynamically stable crystalline phases form from the metastable amorphous phase in a devitrification process. The nanocrystalline Metallic glasses (MGs), often referred to as structure that nucleates and evolves after annealing changes the alloy corrosion resistance and other properties (5-7). The control of introducing the respective grain boundaries and segregation profiles by temperature make these alloys also to an interesting model system for fundamental corrosion studies.
Among MGs, amorphous steels, i.e. Fe-based alloys, own valuable properties, including excellent corrosion resistance, high specific strength, and probably the best thermal stability among metallic glasses(8-10). The Fe-based alloy with the nominal Fe50Cr15Mo14C15B6 composition has(11) good mechanical properties and glass forming ability (GFA). In the present work, we investigate the changes in structure, local composition and corrosion behaviour of this alloy upon thermal history and crystallization state. The material was prepared by arc-melting a mixture of the constituents in a purified argon atmosphere. The resultant homogeneous alloys were melted again to prepare melt-spun ribbons of 50 μm thickness and 3 mm wide. For a number of samples crystallization was induced by thermal annealing of ribbons in an argon atmosphere. The samples were annealed at different temperatures (550-800º C), 20 minutes for each set temperature. These temperatures were selected according to the corresponding glass transition temperature Tg (544º C), and crystallization temperatures Tx1 (600º C) and Tx2 (623º C) as a reference as determined by differential scanning calorimetry (DSC).
- Figure 2. With near atomic-scale resolution, metallic glass is exposed (volume: 106 nm X 90 nm X 90 nm containing 23,885,434 atoms) via 3D-APT. Line analysis (a) and selected image (b) of interfaces and precipitates in fully-crystallized MG (800°C, 20 min). Higher concentrations of the minority C and B atoms are indicated with iso-concentration surfaces (yellow, blue) revealing their segregation between the Mo-rich (red) and FeCr-rich (purple) phases.
- Figure 3. Corrosion behaviour of as-produced amorphous MG, a fully-crystallized sample (800°C, 20 min), and an intermediate state (620°C, 20 min). E vs. Ag/AgCl. The inset shows an SEM image of a pitted surface (800°C, 20 min) after corrosion test.
X-ray diffractograms suggest that the main phases formed during crystallization are carbides and borides of the form M23(C,B)6, with M=Fe or Cr, with weaker additional signatures corresponding to a CrMo phase. Nanometer-sized crystallites were formed upon annealing, which renders high-resolution scanning Auger microscopy (SAM) and atom probe tomography (APT) essential techniques to understand the structural and compositional changes. Figure 1 shows a SAM map of the fully-crystalline sample surface, annealed at 800ºC during 20 min, after removing a surface layer of 300 nm by sputtering. The surface shows 10-30 nm large Mo-rich inclusions in a Fe-Cr rich matrix. While SAM shows the chemical lateral composition of the surfaces evaluated in the corrosion tests with a resolution of below 10 nm, APT addresses the internal microstructure with an even better resolution. The combination of SAM and APT can thus link surface properties like corrosion behavior with unprecedented detail in morphology and composition. In agreement with the SAM result, Figure 2a and b show APT data obtained from a fully-crystallized sample (800°C, 20 min). Mo-rich and Fe-Cr-rich areas are clearly visible. Boron and Carbon segregate to FeCr / Mo interfaces.
Linear polarization tests in 0.1M HCl were performed to evaluate the corrosion behaviour. The corrosion tests show that fully amorphous steel presents a higher pitting or breakdown potential and a wider passive range than the partially and fully crystallized ones. The amorphous alloy exhibits hence a better corrosion resistance behaviour. After the corrosion attack the surfaces show pits as can be seen in the inset SEM of Figure 3 for the fully nano-crystalline sample. The breakdown potential decreases considerably with crystallization and this loss in corrosion resistance can be attributed to the nanometer-scale phase separation in the heat treated material. But clearly the phases of different composition, including the grain boundaries, lead to a variation in the stability of the corresponding surface oxide scales as well as to a different proneness towards selective dissolution of the respective parts itself. The exact mechanisms are still evaluated.
Bibliography
1. A. Inoue, Acta Mater. 48 (2000) 279.
2. M.F. Ashby, A.L. Greer, Scripta Mater. 54 (2006) 321.
3. W.L. Johnson, K. Samwer, Phys. Rev. Lett. 95 (2005) 195501.
4. A. Inoue, N. Nishiyama, MRS Bull. 32 (2007) 651.
5. X.J. Gu, S.J. Poon, G.J. Shiflet and M. Widom, APL 92 (2008) 161910.
6. F. Li, B. Shen, A. Makino and A. Inoue, APL 92 (2007) 234101.
7. H.M. Ha, J.R. Miller, J.H. Payer, J. Electrochem. Soc., 156 (2009) C246.
8. V. Ponnambalam, et al., APL 83 (2003) 1131.
9. A. Pardo, M.C. Merino, E. Otero, M.D. Lopey and A. M’hich, J. Non-Cryst. Solids 352 (2006) 3179-3190.
10. P. Pawlik and H.A. Davies, Scripta Mater. 49 (2003) 755-760.
11. V. Ponnambalam, S.J. Poon and G.J. Shiflet, J. Mater. Res. 19 (2004) 1320–1323.
