UCLA

Controlling magnetism on the nanoscale has attracted considerable research interest for the high potential in non-volatile memory and logic applications. Using strain to control magnetization in strain-mediated multiferroic heterostructures is considered the most energy efficient approach, reducing energy dissipation by orders of magnitude. The strain-mediated multiferroic heterostructure has a ferromagnetic element on a ferroelectric substrate. Applying voltage to the ferroelectric substrate induces piezoelectric strain, which manipulates the magnetization of the ferromagnetic element through magnetoelastic effect. Nanomagnets, as information storage bits for non-volatile memory applications, need to be both individually and deterministically controlled. In the present work, two concepts are developed for this aim, one uses an electrode pattern design on a piezoelectric substrate to produce localized strain, and the other consists of architecting the shape of nanomagnets to take advantage of magnetic shape anisotropy. Patterned electrodes are designed and their effect is modeled using finite element simulations. By selectively applying voltage to electrode pairs, various strain configurations are produced between the electrodes, creating localized strain that controls individual nanomagnets. The modeling results were confirmed by experiments that used magnetization characterization techniques including magneto-optical Kerr effect (MOKE) and magnetic force microscopy (MFM). By architecting the geometric shape, “peanut” and “cat-eye” shaped nanomagnets were engineered on piezoelectric substrates. These nanomagnets undergo repeated deterministic 180? magnetization rotations in response to individual electric-field-induced strain pulses. The designs were modeled using micromagnetics simulations. Both concepts provide significant contributions for next generation strain-mediated magnetoelectric memory research. This work opens a broad design space for next generation magnetoelectric spintronic devices.