Session K36: Magnetic Domains and Applied Magnetic Materials

Real Space Observation of Helical Spin Order by Lorentz Microscopy

Helical spin order is widely observed and its average structure has been investigated so far in reciprocal space mainly by neutron diffraction. Here the helical spin order and dynamics are visualized in real space by means of Lorentz electron microscopy. Our system of choice is Fe$_{1-x}$Co$_{x}$Si, which exhibits a helical spin order with a long period ($>$ 30 nm) in a concentration range 0.05 $\le $ $x \quad \le $ 0.8. The N\'{e}el temperature $T_{N}$ and the helix period for $x$ = 0.5 take about 38 K and 90 nm along the [100] direction, respectively. The helical spin order is due to the Dzyalosinsky-Moriya (DM) interaction because of the lack of centrosymmetry of the lattice. The observations were made by means of a Lorentz transmission electron microscope operated at 300 kV. The magnetization distribution was obtained by the transport of intensity equation (TIE) analysis. The real world of the helical spin order proves to be much richer than expected from the averaged structure, as manifested by variegated magnetic defects like atomic boundary and dislocation in the crystal lattice. By applying magnetic fields we can directly observe the deformation processes of the helical spin order, accompanied by nucleation, movement, and annihilation of the magnetic defects.   

Imaging chiral domains and magnetic phase coexistence in holmium metal

In the rare-earth metal holmium, the magnetic moments in the atomic planes form a magnetic spiral structure below T$_{N}$=133 K with an incommensurate repeat unit along the crystal c-axis. This spiral structure forms with different chirality in different regions of the samples, giving rise to chiral domains. The pitch of the spiral evolves with temperature decreasing toward six atomic layers as the crystal is cooled to T=19 K where the magnetic moments cant toward the c-axis, and a commensurate conical magnetic structure results. This magnetic phase transition leads to the formation of a second type of domains where the spiral phase coexists with the conical phase over a temperature range of approximately 1 K. We have combined micro-diffraction with phase-retarding x-ray optics to image both types of magnetic domains simultaneously as a function of temperature. Using circularly polarized x-rays, the satellite peaks associated with the magnetic ordering become sensitive to the chirality of the spirals, and the micro-focused beam provides spatial resolution of a couple microns. We will present images of both chiral domains and spiral/conical domains, which have been shown to occur over length scales of many microns.   

Inverse Transition of Magnetic Domain Patterns in Ultrathin Fe Films on Cu(001): A Quantitative Analysis

The magnetization of ultrathin Fe films on Cu(001) is perpendicular to the film plane. Due to the competition between exchange and dipolar interactions such films break up into domains of opposite magnetization. As the temperature is increased, the magnetic domain pattern undergoes an inverse transition$^1$: a more ordered stripe phase is found at higher temperatures than a less ordered labyrinthine phase. The domain structure has been imaged by means of a SEMPA (Scanning Electron Microscope with Polarization Analysis). We characterize this pattern sequence quantitatively in terms of domain width, correlations, single-configuration energy and density of defects, directly extracted from SEMPA images. \newline\newline $^1$ O. Portmann, A. Vaterlaus, and D. Pescia, Nature {\bf 422}, 701 (2003).   

Magnetic bubble domains in the spin reorientation transition region of Fe/Ni/Cu(001)

Spin reorientation transition (SRT) in Fe/Ni/Cu(001) system was investigated using photoemission electron microscopy (PEEM). In addition to the previously found stripe domains, we also observed bubble domains within a narrow thickness region of the SRT after applying an in-plane magnetic field. Because of the up-down asymmetry of the bubble domains, a small off-normal component of the in-plane magnetic field is necessary to generate the bubble domains. By applying the magnetic field along different directions related to the sample surface, we found that the bubble domains disappear if the field direction is more than $\sim $10 degrees off the sample surface. A temperature dependent study shows that the bubble domains are unstable against annealing and will be converted back to the stripe phase, indicating that the bubble domain phase is a meta-stable phase.   

Regular magnetic domain pattern in TbNi$_{2}$B$_{2}$C due to strong magneto-elastic coupling

$R$Ni$_{2}$B$_{2}$C compounds are of strong interest due to the competition between superconductivity and magnetism. The crystal electric field effects yield strong magnetoelastic effects. The strongest tetragonal-to-orthorhombic distortion is observed for TbNi$_{2}$B$_{2}$C with a relative distortion of 0.6 {\%} in the antiferromagnetic state at 1.5 K. The lowering of the symmetry forces the development of domains of two forms, in which the magnetic moment direction are align along the a or b axis, respectively. For both kind of domains, the lattice distortion is opposite in the ab plane. We investigated the resulting domain using scattering techniques. Temperature-dependent topographic x-ray images show that the domains are very large in (110) directions where as they are very narrow in the direction perpendicular in the ab plane. In X-ray and neutron diffraction sattelite reflections are found, which are produced by the regular pattern of domains with dimensions of 10 nm. A model describing this pattern and consequences for magnetic and superconducting properties will be discussed.   

Disorder Induced Magnetic Memory

To test theories of random microscopic disorder, we first developed and then applied coherent x-raymetrology to a series of thin multilayer films with perpendicular magnetic anisotropy anddegrees of disorder. We used coherent x-rays to generate speckled scattering patterns. The ``random'' arrangement of the speckles is due to the exact configuration of the magneticin the sample. Changes in the domain structure change the speckles, and comparison of thespeckle patterns provides a quantitative determination of how much the domain structure has. We asked (and answered) two important questions. How is the magnetic domain configuration at one point on the major hysteresis loop related to the configurations at the same point on the loop during subsequent cycles: microscopic return point memory(RPM)? We found the RPM is partial and imperfect in the disordered samples. How are the magnetic domains at one point on the major loop related to the domains at the inversion symmetric point on the loop, during the same and during subsequent cycles: microscopic complementary point memory(CPM)? We found the CPM is also partial and imperfect in the disordered samples. In addition, we found that RPM is always larger than CPM. No existing theory was capable of reproducing our results. We developed newthat do fit our observations. Our experimental and theoretical results set new benchmarks forwork.   

Kerr Imaging of Co/Pt multilayers with perpendicular anisotropy

Co/Pt multilayer films are promising materials for applications such as in high density magnetic recording media. For very thin Co layers these structures exhibit perpendicular magnetization. The exchange coupling between adjacent Co layers is ferromagnetic and the coupling strength oscillates as the nonmagnetic Pt layer thickness (t$_{Pt})$ ranges from 3 to 75\r{ }A. In order to understand the magnetization reversal process in these structures, we report on magneto-optical images of a series of [Co(4 {\AA})/Pt(t$_{Pt})$]$_{N}$ multilayers obtained with Kerr microscopy as a function of t$_{Pt}$ and layer repetition N. The images reveal the evolution of the magnetic reversal process that strongly depends on t$_{Pt}$ and therefore on the interlayer coupling. For Co/Pt multilayers with small t$_{Pt}$, e.g., 11 {\AA}, when Co layers are strongly coupled, the whole multilayer switches as a single ferromagnet. As Co layers are farther separated and decoupled, e.g., at t$_{Pt}$ =41 {\AA}, Co layers switch independently and layer by layer switching was observed by Kerr imaging. The response of these distinct magnetic phases to external magnetic fields and their relationship to details of the hysteresis loops will be discussed.   

Sagnac type fiber interferometer for magneto-optic Kerr effect measurement at cryogenic temperatures

We describe a Sagnac type magneto-optical interferometer operating at 1550 nm wavelength in which the fast and slow axis of a single 10-meter long Polarization-Maintaining fiber were used as the Sagnac loop. The last 2 meters of this PM fiber were fed into a cryogenic probe to measure Kerr rotation. This zero-area-Sagnac-loop design is virtually immune to temperature fluctuations and mechanical bending of the fiber, and can work at much lower temperature ($<$ 1 K) compared to apparatus with optical windows. Most important, no perturbing AC magnetic field is needed. Up to now, we have achieved a Kerr rotation sensitivity of $2.5\times 10^{-7}rad/\sqrt {Hz} $ down to liquid Helium temperature, with$4\mu W$of optical power at the detector. Drifts of this apparatus were observed to be less than$0.5\times 10^{-7}rad/Hour$. We studied ferromagnetic transition and magnetic domains of thin (3 to 30 nm) SrRuO$_{3}$ films by cooling them through Tc in zero fields ($<$ 5 mG) and measuring polar Kerr rotations.   

Fluctuation in the depinning of edge magnetic domains due to edge defects in ultrathin films

Magnetic memory devices utilize an optimal reversal field to change the magnetic state of a magnetic memory element. Edge pinned domains (EPD) can lead to the restoration of the original magnetic state. Failure in reversal can occur when the external field is withdrawn before EPD's are completely depinned [1],[2]. We study, via Monte-Carlo simulations, the reversal of edge pinned domains in NiFe ultrathin films with dimensions $0.33 \mu m \times 0.50 \mu m \times 50$ {\AA} and $0.20 \mu m \times 1.00 \mu m \times 50$ {\AA} (aspect ratio 2:3 and 1:5 respectively) at different temperatures. Fluctuations in reversal field were computed relative to that of samples without edge defects, our control. Defects in the edges parallel to the easy axis increase the switching field by an average of 29\% in 1:5 samples and 6.7\% in 2:3 samples. Defects in the edges perpendicular to the easy axis reduce the switching field by an average of 3.6\% in 1:5 samples and 4.2\% in 2:3 samples. The averages are quoted for word ($H_{ext,x}$) and digit ($H_{ext,y}$) line fields that result in a net external field at an angle of $5^{\circ}$ to $55^{\circ}$ to the easy axis (digit line). We show the mechanism for edge pinned domain reversal is by the removal of $180^{\circ}$ domain walls. This removal is by the nucleation of solitons propagating from one end of the sample to the other. Most of the time, the upper and lower walls are removed simultaneously from opposite ends. [1] J. Shi et al, Appl.Phys.Lett. 77, 1692 (2000). [2] A. A. Oriade et al, Jour. App. Phys. 97, 023905 (2005).   

A study of magnetic domain formation and motion in perpendicularly magnetized ultrathin film using the magnetic ac-susceptibility

The Complex, low-frequency ac magnetic susceptibility $\chi $ has been used to study magnetic domain formation and motion in perpendicularly magnetized ultrathin Fe films on a 2ML Ni/W(110) substrate. Analysis of the real and imaginary part of $\chi $ reveals that below a characteristic temperature T$_{sat}$, the stripe-domain density cannot change quickly enough to maintain equilibrium. This is due to the fact that domain wall creation and/or annihilation is itself an activated process, with nucleation energy E$_{n}$, distinct from the pinning of existing domain walls by defects in film structure, with activation energy E$_{a}$. T$_{sat}$, is set by the time scale of the measurements, which is determined by the rate of change of temperature (R). The Magnetic susceptibility was recorded as a function of temperature at different heating rates between 0.03 (K/s) to 1 (K/s). Our results show that below heating rate R$_{c }$(= 0.2 K/s for 1.5ML Fe film) the susceptibility peak temperature (T$_{peak})$ decreases as dT/dR = -200.0 (s). Above R$_{c}$, the susceptibility peak temperature increases as dT/dR = 16.6 (s). Preliminary model calculation show the movement of T$_{peak}$ is due to the change in T$_{sat}$ as the heating rate changes. R$_{c }$is set by the relative values of E$_{n }$and E$_{a}$.   

Imaging Antiferromagnetic Domain Walls with the Hall Effect

We find that the Hall effect in the spin-density-wave state of elemental chromium is sensitive to the underlying magnetic domain structure. A modest hysteresis in the linear Hall coefficient emerges as a function of temperature in the antiferromagnetic phase. By measuring all three components of the resistivity tensor in the plane of the sample we are able to separately consider the effects of domain rearrangement on carrier mobility and density. We also present direct images of hysteretic domain rearrangement, acquired via spatially resolved X-ray charge and magnetic scattering, and we show that these images can be correlated to the transport measurements.   

Anhysteretic magnetization and magnetostriction of thin NiFe films under stress and plastic deformation

The magnetic properties of thin samples of a thin film NiFe sample under tensile stress are investigated. The magnetostriction contribution to dc magnetization under elastic stress and the effect of the plastic strain on the hysteresis loops are discussed. Also, a role of the plastic deformation interrelated with the elastic stress in the magnetization process is established. An experimental system based on a conventional vibrating sample magnetometer equipped with the specially designed loading fixture and optical resonant spectroscopy tension monitoring technique is used to measure anhysteretic permeability and magnetization curve as a function of stress and temperature. This method used to measure anhysteretic permeability and magnetization curve of Ni-Fe as a function of stress and temperature. Anhysteretic permeability was extracted from the anhysteretic $B-H$ curves constructed by degaussing the sample at given longitudinal (parallel to the stresses) dc field. The large positive magnetostriction constant of FeNi samples leads to higher susceptibility and lower coercivity with tensile stress while the large volume magnetostriction results in reduced saturation magnetization. Large stresses imposed on the sample result in plastic strain of the sample which induces increase in dislocation density and domain wall pinning. This causes the gain in hysteresis loss and coercivity to increase at the highest stresses. We also discuss the effect of the Ni composition on results of the measurements.   

Soft magnetic layers for low-field detection in spin-valve and magnetic-tunnel-junction sensors

We have investigated a wide variety of soft magnetic layers as sense layers for magnetic-field sensors. We find that in thin-film form, some of these soft materials can have susceptibilities approaching those of the corresponding bulk material. In general, the highest susceptibilities occur in tri-layer structures with a non-magnetic film separating two soft magnetic films. The alloy Ni$_{77}$Fe$_{14}$Cu$_{5}$Mo$_{4}$ of the mu-metal family is the softest thin-film material we have found, and we can achieve hard-axis susceptibilities of $\sim $10$^{5}$ in tri-layer structures. The hard axis is preferred for magnetic sensors due to its near-linear response. The major impediment we have found to using these very soft layers in low-field sensors is that~the susceptibility decreases by almost two orders of magnitude when the soft structure is incorporated in a standard spin valve or magnetic tunnel junction. In this talk, we will illustrate the problem, show how the structural modifications can minimize the problem, discuss the outlook for the complete elimination of the problem, and assess the prospects for significant improvements in thin-film, low-field magnetic sensors.   

Minimizing 1/f Noise in Magnetic Sensors Using MEMS Flux Concentrators

The 1/f noise of new types of magnetoresistance sensors based on GMR and MTJ limits their sensitivities at low frequencies. Our approach for dealing with this problem is to shift the operating frequency to higher frequencies where the 1/f noise is much lower. The shift is accomplished by placing flux concentrators on MEMS flaps. Springs connecting the flaps are used to establish the proper normal mode. The motion of the MEMS structure, driven to oscillate at 15 kHz by electrostatic comb drives, modulates the field at the position of the sensor. The device was fabricated using SOI wafers, deep reactive ion etching (DRIE), and flip chip bonding. The motion of the permalloy on the MEMS flaps modulates the field by a factor of 2. Driving the motion only requires microwatts of power. Noise measurements indicate that the device is likely to increase the sensitivity of many magnetic sensors at low frequencies by orders of magnitude.   

Enhanced coercivity in melt-spun Sm-Co-Fe-Cu ribbons after low temperature aging

We have studied systematically the influence of the microstructure refinement on the change of coercivity in melt-spun Sm (Co$_{0.45}$Cu$_{0.4}$Fe$_{0.15})_{5}$ alloys after thermal aging. The specimens have been prepared in the form of ribbons with a thickness in the range of 35 - 80 $\mu $m for values of the quenching wheel speed between 5-25 m/s. X-ray diffraction spectra of the as-spun ribbons showed a single phase 1:5 structure. The as-spun ribbons had a coercivity H$_{c}$ = 8 kOe. The as-spun ribbons have been subjected to low temperature aging between 350-400 $^{\circ}$C for different periods of time, in Ar atmosphere. Our results showed that the highest coercivity of 44 kOe was obtained in a sample aged at 400 $^{o}$C for 133 h. Microstructural investigations by transmission electron microscopy reveal that the average grain size of the specimens with the enhanced coercivity is about 500 nm. In order to have an insight into the mechanism of the coercivity enhancement, we are investigating the microchemistry at the grain boundaries by energy dispersive X-ray analysis and measuring the change in the Curie temperature through thermomagnetic measurements.