서지주요정보
Study on magnetic domain wall roughness in ferrimagnetic CoGd thin film = 준강자성체 CoGd 박막에서 자구벽의 요철도(凹凸度) 연구
서명 / 저자 Study on magnetic domain wall roughness in ferrimagnetic CoGd thin film = 준강자성체 CoGd 박막에서 자구벽의 요철도(凹凸度) 연구 / Kyoung-Hoon Kim.
발행사항 [대전 : 한국과학기술원, 2020].
Online Access 원문보기 원문인쇄

소장정보

등록번호

8035760

소장위치/청구기호

학술문화관(문화관)B1층 보존서고

MPH 20001

휴대폰 전송

도서상태

이용가능(대출불가)

사유안내

반납예정일

리뷰정보

초록정보

Magnetic domain wall (DW) motion has long been a subject of research in spintronics field because of its fundamental interest as well as technological applications. When the magnetic DW is driven by a weak magnetic field at finite temperature, the motion is generally occurred by successive thermal activation process, which we call DW creep motion. In this creep regime, the DW can be considered as an elastic string which has finite roughness due to the quenched disorders. As the roughness governs the scaling law of DW speed, it is highly important to determine the roughness in disordered media. In this study, we investigate the roughness of magnetic DW in two different magnetic systems: ferromagnetic Co/Pt and ferrimagnetic GdCo, both are made by sputtering method. By developing a program that can automatically analyze the DW roughness from the obtained DW image, we analyze the DW roughness of Co/Pt and GdCo thin films. The results show that the wandering exponent is similar in two different magnetic systems. This suggests that the DW roughness does not much depend on the magnetic properties, but depend on the fabrication conditions.

자구벽은 오랫동안 스핀트로닉스 분야에서 기술적인 적용뿐만 아니라 기초학문적인 흥미로 인해 많은 자성과학자들의 연구주제가 되었다. 자구벽이 절대 0도 이상에서 약한 자기장으로 움직이게 될 때, 그 운동은 연속적인 열적 활성화 과정을 통해 발생되는데, 이러한 조건으로 운동하는 영역을 크립 영역이라고 부른다. 이 영역에서, 자구벽은 마치 어떠한 요철도를 가진 고무줄로 생각할 수 있다. 이 요철도가 자구벽의 운동 속도를 결정하기 때문에 무질서한 매질에서 요철도를 결정하는 것이 매우 중요하다. 이 연구에서 강자성체와 준강자성체라는 두 가지 다른 자성계에서 각각 자구벽의 요철도를 조사하였다. 주어진 자구벽 이미지로부터 자구벽의 요철도를 자동적으로 계산하는 프로그램을 만들어 스퍼터링 방법으로 증착된 Co/Pt(강자성체)와 $Co_xGd_{100-x}$(준강자성체) 박막의 자구벽 요철도를 분석하였다. 그 결과 강자성체와 준강자성체에서 에러범위 내에서 동일한 자구벽의 요철도를 얻었다. 이는 자구벽의 요철도가 시료의 거칠기나 결함에 기인하고, 따라서 자성의 특성보다 시료를 만드는 방법에 의존함을 의미한다.

서지기타정보

서지기타정보
청구기호 {MPH 20001
형태사항 iii, 59 p. : 삽화 ; 30 cm
언어 영어
일반주기 저자명의 한글표기 : 김경훈
지도교수의 영문표기 : Kab-Jin Kim
지도교수의 한글표기 : 김갑진
학위논문 학위논문(석사) - 한국과학기술원 : 물리학과,
서지주기 Including references.
QR CODE

책소개

전체보기

목차

전체보기

이 주제의 인기대출도서

Diamagnetism (a) Microscopic spin structure of diamagnet. (b) Reaction of diamagnet in response to the magnetic field.

Paramagnetism (a) Microscopic spin structure of paramagnet. (b) Reaction of paramagnet in response to the magnetic field.

Ferromagnetism (a) Microscopic spin structure of ferromagnet. (b) Reaction of ferromagnet in response to the Magnetic Field. (c) The hysteresis loop ofthe ferromagnetic object. To switch the magnetization, the magnetic field whose magnitudeis Hc (the coercive field) at least should be applied to this system [6].

Ferrimagnetism and Antiferromagnetism (a) Microscopic spin structure of ferrimagnet. (b) Reaction of ferrimagnet in response to the magnetic field. For ferrimagnet, it is similar to antiferromagnet microscopically (atomically antiparallel), but it resembles the ferromagnet macroscopically (net magnetization). (c) Microscopic spin structure of antiferromagnet. Neighboring spins aligns in antiparall

The Free Energy - The Length of DW segments Graph. From L=0toL=Lc,the curveis

Magnetic field-driven DW velocity [7]. AtT=0, the DW cannot move until the driving field reaches the threshold field (Hc). On the other hand, there is thermal fluctuation at the finite temperature. Due to this, DW can move even below the critical field.

Where does the wandering exponent come from? [16] (a) The Position of DW (b) The Evolution of DW (c) W(L=const,t) -t graph(Left) and log(w) - log(t) graph (Right) (d) The graph is log(w(:L.t=const.) - log(l).

Pinning Potential of the Surface the DW Creeps for Random Bond Disorder [7]. Here, theyellow elastic curve represents a DW. Potential landscape shows the pinning potential for Random Bond Disorder [8].

Pinning Potential for Random Field Disorder [4]. Here, the magnetic atoms (Mn) are located on the defect spots. The magnetic atoms set apart from another magnetic atom, so the interaction between magnetic atom too weak to work dominantly. In this case, the total field applied to magnetic atom is main boost to move DW [8].

The MOKE Microscope used in this research.

Schematic Diagram for MOKE Microscopy

Image Data to be obtained from this MOKE Microscope. The directions of the magnetization of bright zone and dark one are opposite. The border of these 2 region should be a Domain Wall.

Graph of Magnetic Field Generated by Current - Current Flowing through the Electromagnet. Above graph is obtained by measuring the magnetic field when changing the current flowing through the electromagnet

The Specification of the sensor of the CCD Camera used in this experiment

Rotated Direction of Polarization. After reflecting, the direction of polarization is rotated and dependent on that of magnetization of the sample.

Sample used in this paper (a) Co/Pt sample (b) CoxGdi-x sample (x= 73, 75, 79, 81). The substrate layer looks even thinner than the real. Also, for the convenience, the thickness of the Co/Pt samples are exaggerated.

Co/Pt Hysteresis Loops (a) MOKE hysteresis loop (Raw Data) (b) normalized hysteresis loop (obtained from (a))

Hysteresis Loop for CoxGd100-x. Left Column : Intensity of MOKE Signal [Arb. Unit] VS. Magnetic Field [G] Right Column : Relative MOKE Signal [Arb. unit] VS. Magnetic Field [G]

Coercive Fields for Each Sample

Brightness Profile. The graph on the rightis a brightness profile of the column in the red box on Fig. 3-3 (left). The i-th value is calculated by average with the original brightness 10 points including the i-th point.

Differentiated Brightness Profile with taken absolute value. The i-th value is gotten by {{(+1)-th value} - {i-th value}]. The brightness near the pixel containing DW segments is rapidly changed. So, the pixel having the maximum value on this graph can be regarded as DW. After the procedures so far is applied to all columns, then, we will geta DW profile.

Domain Wall Profile. The raw image data on the left (Fig. 3-3) is converted with the graph on the right.

In (Average Correlated Function) Vs. In (L) Graph

Slope: Wandering Exponent(7) : the Slope of "ln (AFC) - In (L)" graph. Exactly, the slope is 27. When we apply linear fitting to the above graph, the first 10 points and points after extremum are excluded.

The Experimental Environment (Driven Current) for Each Sample

Wandering Exponent 27, for Co/Pt. The left one is S. Lemerle's measurement as reference, and the right one is measured and analyzed by our LABVIEW-based program.

Snap shots (a) C073Gd27 (b) CozsGd25 (c) CozgGd21 (d) cost

Wandering Exponent Distribution for the different composition. (x-axis : measurement number, y-axis : exponent 23) (a) Co73Gd27 (27=1.37+0.12) (b) CozsGd25 (23=1.21 士0.10) (c) CozgGd21 (27 =1.36士0.14) (d) CosiGd19 (23 =1.25+0.11)

Wandering Exponent Values for Sample. The red box is Ferromagnet, and the green one is Ferrimagnet. The wandering exponents tolerably change within error ranges even if the composition between Co and Gd is different.

Surface of sample (Atomic Force Microscopy Image) (a) Co/Pt (b) Cos1Gd19 (c) CozgGd21 (d) Com7Gd (compensated) (e) ConsGd25 (f) Co73Gd27

Average Grain Diameters and Wandering Exponents for Samples