Papers, preprints, thesis chapters, and the hardware behind them. Published work links to the journal, with BibTeX; everything else is a pre-written email away.
How mechanical strain reshapes magnetism in thin films, and how to turn that physics into devices: strain-tunable anisotropy, transport under bending, and sensors that read magnetic field and strain at once. I share chapters and the current draft with collaborators and committees.
Shape anisotropy is usually treated as a fixed, geometry-derived constant. Measured field by field it holds only above a crossover field: below it the magnetization is non-uniform and the shape contribution falls short, biasing low-field device models by 8–12%. The decisive evidence needs no fit: devices patterned parallel and orthogonal to the growth easy axis shift in opposite directions.
A physics-guided SRGAN reconstructs high-resolution magnetic field maps from sparse 3×3 sensor arrays, conditioned on the source’s physical parameters to stay physically consistent.
Patterning the ellipse orthogonal to the growth-induced easy axis makes the two anisotropies subtract instead of add, breaking the conventional resolution floor: the effective anisotropy field of a representative device dropped from 6.15 to 2.66 Oe, within 0.2 Oe of a parameter-free model.
Applies multi-objective optimization to magnetic anomaly detection with planar Hall effect sensors, weighing competing detection objectives against one another rather than tuning for a single figure of merit.
A 3×3 array of elliptical PHE sensors plus a neural network reconstructs magnetic source maps. Reaching the same quality without learning would take more than fifty sensors.
Ellipse geometry tunes the anisotropy field from 12 to 120 Oe, a tenfold wider operating range, while every size remains a single magnetic domain that follows the Stoner-Wohlfarth model. My first first-author paper.
A planar Hall effect magnetometer on 25 µm Kapton: equivalent magnetic noise below 200 pT/√Hz flat and below 400 pT/√Hz bent to a 10 mm radius. At publication, the best resolution reported for a flexible magnetic sensor.
Datasets, fabrication notes, teaching material. If I have it and I can share it, I will.