

Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Community
Ask the community for help and clear up your study doubts
Discover the best universities in your country according to Docsity users
Free resources
Download our free guides on studying techniques, anxiety management strategies, and thesis advice from Docsity tutors
This document reports on a study investigating magnetization reversal phenomena in a single submicron wire using the giant magnetoresistance (gmr) effect. The researchers measured the electric resistance in external magnetic fields and observed the critical field for domain wall nucleation. They found that the process occurs through a thermally activated process down to 5 k.
Typology: Papers
1 / 3
This page cannot be seen from the preview
Don't miss anything!
Journal of Magnetism and Magnetic Materials 198}199 (1999) 58} 60
! Institute for Chemical Research, Kyoto Uni v ersity, Uji 611-0011, Japan " Faculty of Science and Technology, Keio Uni v ersity, Yokohama 223-8522, Japan
Abstract
Magnetization reversal phenomena in a submicron magnetic wire with a trilayer structure were investigated by measuring the electric resistance in external magnetic "elds. The critical "eld for domain wall nucleation was observed to be slightly di!erent for each "eld sweep. The magnetoresistance measurements were repeated 200 times with a sweep rate of 6 Oe/s. The obtained distribution of critical "elds had two peaks and both moved to higher "elds as temperature was decreased. This result suggests that domain wall nucleation occurs through a thermally activated process down to 5 K. ( 1999 Elsevier Science B.V. All rights reserved.
Keywords: Magnetization reversal; Submicron magnetic wire; Magnetic domain wall; Giant magnetoresistance
Magnetism in mesoscopic systems has become an im- portant topic from both scienti"c and technological points of view. However in general, experimental studies of magnetic nanostructures have been made on samples consisting of a large number of nearly identical particles, since most measurement techniques are not sensitive enough to measure the magnetization of a single particle. Most of the single-particle or single-wire properties were inevitably hidden behind the distribution of size or shape. Recently the techniques of magnetic force micros- copy (MFM) [1], electron holography [2], and micro- superconducting quantum interference device (SQUID) magnetometery [3}5] have made it possible to study individual magnetic particles. While these experiments give us crucial information, they have considerable tech- nical di$culties. As reported in a previous paper [6], the authors demonstrated that very small magnetization changes in a single NiFe(200 As )/Cu(100 As )NiFe(50 As ) tri- layer wire with 0.5 lm width can be detected by using the giant magnetoresistance (GMR) e!ect. In ferromagnetic wires, it is expected that magneti- zation reversal takes place by nucleation and propaga-
tion of the magnetic domain wall which lies in a plane perpendicular to the wire axis. The process of magneti- zation reversal attracts interest especially at low temper- atures where the macroscopic quantum tunnelling (MQT) process may be dominant. The MQT of a domain wall in a ferromagnetic metal wire has been recently investigated from both theoretical [7] and experimental points of view [8]. To determine the mechanism of mag- netization reversal, the temperature dependence of the critical "eld should be very useful. In this paper, we present the results of measurements on the nucleation "eld of the magnetic domain wall in a single submicron wire based on a non-coupled- type GMR e!ect at several temperatures. The samples were prepared by lift-o! techniques on electron-beam evaporated NiFe(400 As )/Cu(200 As)/NiFe(50 As ) trilayer "lms. Due to the large Cu-layer thickness, the inter- layer exchange coupling between NiFe layers is negli- gible. The magnetoresistance (MR) measurement was performed at 5, 20 and 50 K. The external mag- netic "eld H%95 was applied along the wire axis. The resistivity was determined using a four-point DC tech- nique. As seen in an image of scanning electron micros- copy in Ref. [6], the sample has four current}voltage terminals where the voltage is probed over a distance of 20 lm. Furthermore, the sample has an arti"cial neck
0304-8853/99/$ } see front matter ( 1999 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 8 8 5 3 ( 9 8 ) 0 0 6 2 0 - 9
Fig. 1. Resistance as a function of the external magnetic "eld at 5 K. The magnetic "eld was applied along the wire axis. The magnetic domain structures inferred from the resistance measurement are schematically shown.
(0.35 lm width) introduced at 13 distance from one volt- age probe. Fig. 1 shows a typical resistance change of the trilayer system as a function of the applied "eld at 5 K. Prior to the measurement, a magnetic "eld of 1000 Oe was ap- plied in order to achieve magnetization alignment in one direction. Then the resistance was measured as the "eld was swept continuously towards the counter direction at a rate of 6 Oe/s. The result shown in Fig. 1 essenti- ally con"rms the previous results found for a NiFe(200 As )/Cu(100 As)/NiFe(50 As) trilayer wire with 0.5 lm width [6]. The ratio of the resistance changes at "rst and second leap is 1 : 2. This means that one-third of the total magnetization of the thin 50-As-thick NiFe layer changes its direction at the "rst leap in Fig. 1, since the GMR changes are proportional to the fraction of the layer that switches. The ratio of one-third corresponds to the ratio of length between one voltage probe and the neck to the overall length of the wire between the voltage probes. Therefore, in this case, a magnetic domain wall nucleates in the shorter part of the wire and propagates to the neck in the "eld up to 60 Oe and then the wall is pinned up to 86 Oe at the neck. An abrupt resistance decrease at 100 Oe corresponds to the magnetization reversal of the 400-As-thick NiFe layer. The domain wall in the thinner layer can nucleate more easily than in the thicker layer, since wall energy in the thinner layer is smaller. The feature of magnetization reversal is reproducible but the numerical values of critical "elds have certain distributions. Therefore, the same measurement as de- scribed in a previous paragraph was carried out 200 times at 5, 20, and 50 K. We focus on the magnetization reversal of the 400 As-thick NiFe layer, since the magnet- ization reversal "eld of the 50 As -thick NiFe layer is too small to obtain a quantitatively con"dent result. All results can be categorized into the three types as shown in Fig. 2. Type (A), (B) and (C) corresponding to the following magnetization reversal process, respectively. (A) The magnetic domain wall nucleates in either the shorter or longer part of the wire and passes through the arti"cial neck and propagates to the end of the wire. (B)
Fig. 2. Observed three types of resistance change corresponding to magnetization reversal of 400-As thick NiFe layer at 5 K.
Fig. 3. Histogram of the nucleation "eld of the magnetic domain wall in the 400-As thick NiFe layer at 50 K.
The magnetic domain wall nucleates in the shorter part and is trapped in the arti"cial neck. (C) The magnetic domain wall nucleates in the longer part and is trapped in the arti"cial neck. The ratio of occurrence of these three types at 5 K is (A) : (B) : (C)"135 : 28 : 37. The ratio of type (A) decreases with decreasing temperature. Fig. 3 shows a histogram of the observed "eld for all three cases as indicated by H# in Fig. 2. The H# corresponds to the nucleation "eld of the magnetic domain wall. Fig. 3 suggests that the observed values are in a rather narrow "eld range but the distribution has two peaks and these two peaks were also recognized at 5 and 20 K. The reason why the two peaks exist is not clear at this stage. In the present sample, the 20 lm wire is not magnetically isolated but is connected with probes for MR measure- ments which are also composed of the same magnetic layer. Therefore, it is hard to identify the initial structure of domain nucleation. It is considered that domain wall nucleation starts on the outside of the wire and there exist e!ectively two potential barriers. Although the rea- son for two potential barriers is not understood at this stage, we can compare their temperature dependence. In Fig. 4, the "elds of the two peaks, which are indicated by (L) and (R) in Fig. 3, are plotted as a function of temper- ature. The peak "elds increase with decreasing temper- ature down to 5 K, suggesting that the nucleation takes place by a thermally activated process. The study at lower temperatures, using a dilution refrigerator, is now in progress to investigate the e!ect of MQT.
K. Shigeto et al. / Journal of Magnetism and Magnetic Materials 198 } 199 (1999) 58 } 60 59