1. Introduction and methods
Static in-situ annealing of deformed single-crystal halite shows that three distinct temperature dependent stages of dislocation rearrangement result in an overall decrease in the crystallographic variation of the sample.
Substructure dynamics have been investigated in “real-time” by in-situ heating experiments conducted in the SEM. Electron Backscatter Diffraction (EBSD) maps were taken before, during and after each heating stage, to collect detailed information about the crystallographic orientation and misorientation of the sample substructure. Samples were pre-deformed under uniaxial compression at a temperature of ~450 ºC to strains of 0.165, at a strain rate of 6.9*10-6s-1. Samples were then annealed within an SEM in several heating stages at temperatures between 280-470 ºC, with an arbitrary increase in temperature at each heating stage (see fig. 1b). The length of each heating stage varied from 30 minutes to six hours. The setup of the heating stage and an orientation contrast image of the areas of the crystal analysed are shown in fig. 1.
2. Results
Behaviour during annealing of halite can be divided into three distinct phases based on the low-angle boundary (LAB <15º) behaviour and overall changes in the substructure. LABs were divided into five categories based on their morphology, orientation and behaviour. Fig 2 shows a schematic of the boundary types and a table detailing their important features.
Characteristic behaviour of the annealing phases (see fig. 3):
Annealing phase one 280-300 ºC
a) Type 1 and 2 LABs: increase in the misorientation
b) Type 3 and 4 LABs: decrease in the misorientation
c) Type 5: boundary movement with an average velocity of 0.085µm/min
d) subdivision of some subgrains into plateaus of like orientation
Annealing phase two ~300 ºC
a) All LAB types: decrease in the misorientation
b) Type 5: significant decrease in boundary velocity to an average of 0.032µm/min
c) continued subdivision of subgrains and formation of new LABs at plateau borders
Annealing phase three >300 ºC
a) All LAB types: increase in misorientation of remaining LABs
b) Type 5: increase in boundary movement to an average velocity of 0.169µm/min
c) no new plateau formation occurs
3. Substructural evolution during annealing
We suggest that annealing behaviour is both temperature dependent and varies according to which boundary is examined. Fig. 4 shows a diagram of inferred behaviour occurring during annealing. At lower temperatures (T<300 ºC) annihilation of dislocations in the subgrain interior and at the boundary site for LABs aligned with the harder slip system occurs, resulting in a decrease in misorientation. Concurrently, dislocations are added into the boundaries in preferred alignment, which then increase in misorientation. In areas where there are no dislocations of opposite sign, annihilation is not possible and dislocations of like sign begin to align. Consequently, next to these aligned dislocations plateaus of like orientation form. Phase two (T ~300 ºC) marks a switch in behaviour for type 1 and 2 LABs, which begin to decrease as dislocations annihilate at the boundaries. As temperature increases the length scale on which dislocations are attracted to boundaries extends and previously trapped dislocations begin to move. New tilt boundaries form at borders of plateau regions as more dislocations are added. At T>300 ºC changes are dominated by LAB development, where remaining dislocations move towards boundaries as the range of attraction increases further. The cumulation of these processes results in an overall decrease in the crystallographic variation of the sample and thus a significant decrease in the stored energy of the system.