Abstract
The depinning of a domain wall in a ferromagnetic metal via macroscopic quantum tunneling is studied based on the Hubbard model. The dynamics of the magnetization verctor is shown to be governed by an effective action of Heisenberg model with a term non-local in time that describes the dissipation due to the conduction electron. Due to the existence of the Fermi surface there exists Ohmic dissipation even at zero temperature,which is crucially different from the case of the insulator. Taking into account the effect of pinning and the external magnetic field the action is rewritten in terms of a collective coordinate, the position of the wall, Q. The tunneling rate for Q is calculated by use of the instanton method. It is found that the reduction of the tunneling rate due to the dissipation is very large for a thin domain wall with thickness of a few times the lattice spacing, but is negligible for a thick domain wall. Dissipation due to eddy current is shown to be negligible for a wall of mesoscopic size.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
This configuration is possible if the anisotropy energy along the easy axis is much larger than the demagnetization energy.
Stamp, P.C.E. (1991) Phys. Rev. Lett. 66, 2802.
Chudnovsky, E.M., Iglesias, O. and Stamp, P.C.E. (1992) B46, 5392.
Stamp, P.C.E., Chudnovsky, E.M. and Barbara, B. (1992) Int. J. Mod. Phys. B6, 1355.
Caldeira, A.O. and Leggett, A.J. (1983) Phys. Rev. Lett. 46, 211;
Caldeira, A.O. and Leggett, A.J. Ann. Phys. 149, 374.
Garg. A. and Kim, G.H. (1989) Phys. Rev. Lett. 63, 2512
Garg, A. and Kim, G.H.(1991) Phys. Rev. B43, 712
H. Simanjuntak,J.(1992) Low Temp. Phys.90 ,405.
Garg, A. (1993) Phys. Rev. Lett. 70, 1541.
Uehara, M., Barbara, B., Dieny, B. and Stamp, P.C.E. (1986) J. Physique 47, 235.
Paulsen, C., Sampaio, L.C., Barbara, B., Fruchard, D., Marchand, A., Tholence, J.L. and Uehara, M. (1991) Phys. Lett. A161, 319;
Paulsen, C, Sampaio, L.C., Barbara, B., T-Tachoueres, R., Fruchart, D., Marchand, A., Tholence, J.L. and Uehara, M.(1992) Europhys. Lett. 19, 643.
Tatara, G. and Fukuyama, H. (1994) Phys. Rev. Lett 72, 772.
The results would be valid even for a thin domain wall with width of a few times lattice constant.
We neglect the effect of magnetic field on electronic states. This is justified as long as UM » γH. In experimental situations with the magnetic field of ≳ 1T and U ≃ 10eV, this condition reduces to M ≳ 10–4 in unit of the Bohr magneton, which is easy to satisfy.
Korenman, V., Murray, J.L. and Prange, R.E. Phys. Rev. (1977) B16, 4032.
These are the results of the Hartree-Fock theory, which describes the essential features of itinerant ferromagnetism but should not be taken literally in comparison with the actual experiments.
RPA summation is needed since we have not taken into account in the Coulomb interaction the process with finite momentum transfer, \( H_{U}^{{(fast)}} \) in the determination of the magnitude of the magnetization, which has been assumed to be uniform.
Herring, C.(Academic, New York 1966) Magnetism, edited by Rado, G.T. and Suhl, Vol. IV.
Awaka, K., Tatara, G. and Fukuyama, H. (1993) Jour. Phys. Soc. Jpn. 62, 1939.
The super-Ohmic contributions, which are of higher orders of (Ω0/?F) in Eq. (20), are smaller than the Ohmic one by a factor of (ΩO/?F)2 ≪ 1 and hence are negligible.
We consider the case of strong anisotropy; K,K? ≫ Ka, where Kd ≡ μo(ħγ)2/a6 is the magnetostatic energy due to demagnetization (γ = e/(2m) is the gyromagnetic ratio and μ0 is the magnetic permeability of a free space). The calculation also applies to a ferromagnet with uniaxial anisotropy -K in z-direction, if one replaces K?→ Kd.
In the absence of the transverse anisotropy, K-, the domain wall cannot tunnel, since without this term Sz is conserved at each site. This fact is expressed in Eq.(26) as the divergence of the domain wall mass as K?→ 0.
The effect of the anisotropy energy on the dissipation due to the itinerant electron is neglected, since the correction would be small by the order of (Ka3/UMo) ≃ 10–3 (for Ka3 ≃10K, M0 ≳ 0.1).
A factor of 1011 (Hz) in the expression of ?0 arises from the magnetostatic energy due to the magnetization of (2ħγS/a3) ≃ 106[A/m] (i.e., S ≃ O(l)).
Callan, C.G., Jr. and Coleman, S. (1977) Phys. Rev. D16, 1762.
The factor of ?5/4 arises from the barrier height and width in the small ? limit as seen by the WKB approximation; Γo ∞(barrier height)1/2 x Q0 ∞? 3/4?1/2.
The integration must be cut off at short time ~ ?0-1, since in the calculation of ?Sdis, we have made use of the bounce solution with zero energy, that is, we have neglected the excited states of variable Q. This approximation would be valid only for small energy transfer ? ≲?0 in the current correlation function . See Yu.Kagan and N. V. Prokof’ev, Zh. Eksp. Teor. Fiz. 90, 2176 (1986) [Sov. Phys. JETP 63, 1276 (1986)].
Schrieffer, J.R. and Wolff, P.A. (1966) Phys. Rev. 149, 491.
The contribution to the local part of the effective action due to the 5 electron renormalizes the magnitude of S and J, but this renormalization can be understood as already included in the values of these quantities.
The numerical factor of Eq. (42) is (1/4) times that of Eq. (39) because no RPA summation is needed in the s-d model.
A rough estimate of η(ch) is as follows. By the definition of η(ch)(≡ ?Sch/(N?)), we can write\( \Delta {S_{ch}} = N{{{\eta ^{(ch)}}} \over {{\lambda ^2}}}\int {dr} \int {dr'{{{{(Q(\tau ) - Q(\tau '))}^2}} \over {{{(\tau - \tau ')}^2}}}} . \) Noting that the integrand is regarded as rate of energy dissipation due to the Joule heat, we have the relation \( {N \over V}{{{\eta ^{(ch)}}} \over {{\lambda ^2}}}{\dot Q^2} \simeq \sigma {E^2}, \)where V = A?L. By use of Eq.(45), we obtain the expression Eq.(52) correct up to a numerical factor.
This is because, in contrast to the case of quantum coherence problem[17], the Ohmic dissipative action is not divergent at long time, and hence the contribution from the Ohmic dissipation is not qualitatively distinct than those from the super Ohmic one and dissipation processes with an excitation gap.
Enz, M. and Schilling, R. (1986) J. Phys. C 19, 1765; L711;
van Hemmen, J.L. and Sütö, A.(1986) Europhys. Lett. 1 481, (1986) Physica 141B, 37.
Hong. K. and Giordano, N. (1994) in this volume Physica B194-196,1009.
Baumberg, J.J., Awschalom, D.D., Samarth, N., Luo, H. and Furdyna, J.K. (1994) Phys. Rev. Lett. 72, 717.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1995 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Tatara, G., Fukuyama, H. (1995). Macroscopic Quantum Tunneling and Dissipation of Domain Walls in Ferromagnetic Metals. In: Gunther, L., Barbara, B. (eds) Quantum Tunneling of Magnetization — QTM ’94. NATO ASI Series, vol 301. Springer, Dordrecht. https://doi.org/10.1007/978-94-011-0403-6_17
Download citation
DOI: https://doi.org/10.1007/978-94-011-0403-6_17
Publisher Name: Springer, Dordrecht
Print ISBN: 978-94-010-4180-5
Online ISBN: 978-94-011-0403-6
eBook Packages: Springer Book Archive