Wechselwirkung magnetischer Wände mit versetzungen und anderen kristallfehlern in nickel

Abstract

Magnetic domain wall motion and interaction with dislocations and other crystal defects were observed by transmission electron microscopy in Ni single crystal foils (99,995%), which were electrochemically thinned by a jet method. The specimens were mounted in a tiltable straining stage with a magnetization device with beam alignment. The investigated transparent regions had a thickness of ≈ 200 nm. The predictions of two models for domain pinning were investigated. 1. 1) Conservative force model by H. Träuble. Statistically oriented dislocations build up local variations of strain which generate a conservative force gradient in magnetostrictive crystals. In this gradient the walls are assumed to move independently from each other. If a magnetic field is applied they are expected to move at first continuously and reversibly along a distance exceeding the wall width before proceeding by an irreversible jump at a critical field. 2. 2) Wandwölbungs model by H. Markert. This model was introduced to account for magnetically enhanced mechanical creep in strained ferromagnetics and for ultrasonic effects on the hysteresis curve of Ni. On applying a field, the wall is expected to move at first continuously piling up mobile dislocations which entangle, become fixed, and stop the wall. On further increasing the field, the wall bows between the pinning centers. At a critical field the wall breaks away at some point and proceeds in a catastrophic process by an irreversible jump. Furthermore induction experiments apparently indicate a strong coupling between walls and dislocations. Contrary to this, electron microscope experiments reveal neither continuous reversible wall motion prior to jump nor piling up of dislocations by a wall with wall pinning. Displacements of dislocations by a moving wall did occur, however, only on straining the crystal to its elastic limit. Thus interactions of domain walls and dislocations were very weak. Pinning of walls due to dislocations never occurred. Wall pinning took place at grain boundaries due to inhomogeneous strain, at slip traces due to decreased wall area, i.e., wall energy, at inclusions or holes, and most frequently in regions where no inhomogenities were visible. Walls did only move by jumps and usually jumped back to their original or almost original position on decreasing the field, if low fields had been applied. On decreasing higher applied fields the walls frequently surpassed their original position probably due to demagnetizing fields of switched remote domains.