Andrew L. Balk1,2 , Mark E. Nowakowski3, Mark J. Wilson4, David W. Rench4, Peter Schiffer4, David D. Awschalom3, and Nitin Samarth4

1 Center for Nanoscale Science and Technology, NIST, Gaithersburg, MD 20899, USA

2 University of Maryland, Maryland Nanocenter, College Park, MD 20742, USA

3 Center for Spintronics and Quantum Computation, University of California, Santa Barbara, California 93106, USA

4 Department of Physics, The Pennsylvania State University, University Park, Pennsylvania 16802, USA


When a ferromagnet is exposed to a magnetic field, the spins within the ferromagnet act to align with the applied field in a process called magnetization.  Typically, the field required to align a spin with the field is much smaller if there is a region of aligned spins nearby.  This property leads to the formation of patches of flipped spins surrounded by unflipped spins, called domains, which are bordered by transitional interfaces, called domain walls.  During magnetization, the motion of the domain walls is impeded by randomly located pointlike pinning sites, which lead to unrepeatable, unpredictable stick-slip domain wall motion.

However, if a small field is applied, the domain wall reacts to it by flexing in the areas between pinning sites, like a bubble.  Unlike domain wall motion through pinning sites, this flexing motion in between them is energy conservative and repeatable. Direct measurement of flexing is difficult, though, as the corresponding displacements are smaller than the resolutions of most measurement techniques. 

In this work we use a novel electrical technique based on the anomalous Hall effect (AHE) to measure this elastic flexing regime of domain wall motion in (Ga,Mn)As, a ferromagnetic semiconductor.  (Ga,Mn)As has the advantage of having a large AHE, which enables us to probe the domain wall location to nanometer precision and observe domain wall flexing. We use this measurement technique coupled with a feedback loop and trimming electromagnet to first measure the position resolved pinning site strength.  Then we directly measure the flexing regime and observe repeatable, elastic domain wall motion in addition to domain wall depinning events. We analyze this data to determine the density and strength of the pinning sites, which are respectively in the 1014/cm3 and 10s of pN range.  Finally we demonstrate that flexing motion is much faster than larger scale stochastic motion, an observation which may be important for future device and logic applications.