Grain Boundaries in Metals

ACCORDING to the transition lattice theory, since in pure metals and single-phase alloys the only difference between the two grains which meet at a boundary is one of direction, the atoms at the boundary take up positions representing a compromise between the two crystal lattice directions. There thus exists a region, a few atomic diameters in thickness, over which a state of disorder exists, the extent of which will be expected to depend upon the relative orientation of the two lattices; there will exist angles for which this disorder is a minimum, an obvious example being the twinning angle. In addition, the direction of the boundary itself relative to the grains would be expected to produce an effect. Lateral misalignment of crystal planes would also be expected to produce `lack of fit' at the grain boundaries, but this effect is most probably smoothed out by the imperfections in the crystal structure. It would be expected, therefore, that any phenomenon depending upon the degree of disorder existing at the boundary would vary in magnitude according to the relative orientations of the grains meeting at the boundary, and with position in the boundary between two given crystal grains if the boundary changed in direction. It is reasonable to suppose that where precipitation of a second phase occurs from the supersaturated solid solution, it occurs more readily in those regions where disorder is greatest, and it would, therefore, be expected that grain boundary precipitation would vary in this manner.