Phase Transition in Iron Thin Films Containing Coherent Twin Boundaries: A Molecular Dynamics Approach

Abstract

Using molecular dynamics (MD) simulation, the austenitic and martensitic phase transitions in pure iron (Fe) thin films containing coherent twin boundaries (TBs) have been studied. Twelve thin films with various crystalline structures, thicknesses and TB fractions were investigated to study the roles of the free surface and TB in the phase transition. In the austenitic phase transition, the new phase nucleates mainly at the (112)bcc TB in the thicker films. The (111¯)bcc free surface only attends to the nucleation, when the film is extremely thin. The austenitic transition temperature shows weak dependence on the film thickness in thicker films, while an obvious transition temperature decrease is found in a thinner film. TB fraction has only slight influence on the austenitic temperature. In the martensitic phase transition, both the (1¯10)fcc free surface and (111)fcc TB attribute to the new body-center-cubic (bcc) phase nucleation. The martensitic transition temperature increases with decreased film thickness and TB fraction does not influent the transition temperature. In addition, the transition pathways were analyzed. The austenitic transition obeys the Burgers pathway while both the Kurdjumov–Sachs (K–S) and Nishiyama–Wassermann (N–W) relationship are observed in the martensitic phase transition. This work may help to understand the mechanism of phase transition in the Fe nanoscaled system containing a pre-existing defect.

Shear Deformation Helps Phase Transition in Pure Iron Thin Films with “Inactive” Surfaces: A Molecular Dynamics Study

“…Thus, it could be concluded that the Meyer-Entel potential might be the only one in the EAM class, which can describe both the austenitic and martensitic phase transitions. Here, we should mention that the equilibrium transition temperature of 550 ± 50 K predicted by the Meyer-Entel potential [20] differs from the experimental value of 1184 K. This is due to the fact that this potential does not include the magnetic entropy, which stabilizes the bcc phase at low temperatures [10,28]. However, it has been confirmed that the transition mechanism will not be affected by this inexact transition temperature [2][3][4]19,21,28].…”

“…Firstly, it is not unexpected that new phase nucleates at the intersection site between two types of defects, in our case the free surface and the phase boundary. This area provides the largest energetic and structural fluctuations for the new phase nucleation [10]. Here, two values should be given: The phase boundary energy with the N-W geometry amounts to 3.84 J/m 2 (0.96 J/m 2 ) with (without) considering the misfit dislocations [31] and the surface energy of the (111) fcc surface is 1.40 J/m 2 [19].…”

“…Here we should note that the film thickness (surface-volume ratio) has a significant influence on the transition behavior. It has been confirmed that a higher surface-volume ratio is conducive to the phase transition [3,10,19]. The effect of film thickness on the transition will be discussed in Section 3.4.…”

“…The fcc and hcp are closely related due to their close packed nature and they differ only in the stacking sequence. In previous simulation works [2,3,10,19], the mixture of the fcc and hcp phases is denoted as the close packed (cp) phase. The shear deformation acts exactly as the necessary shear to shift the fcc atom layers to build SFs.…”

“…They also found that dislocation structure is changed during the reverse martensitic phase transition. Recently, Wang et al [10] investigated the phase transition in Fe thin films containing coherent twin boundaries and discussed the dependence of austenitic and martensitic transition temperatures on the film thickness and twin boundary (TB) fraction.…”