How “hard” (coercive) a ferromagnet can be has been a puzzle for a century. Seven decades ago, William Fuller Brown offered his famous theorem to correlate coercivity with the magnetocrystalline anisotropy fields in ferromagnetic materials. However, the experimental coercivity values have been far below the calculated levels given by the theorem, which is called Brown’s Coercivity Paradox. Researchers have attempted to solve the paradox with sustained efforts; however, the paradox remains unsolved, and coercivity still cannot be predicted and calculated quantitatively by modeling.

Progress has been made in the past 20 years in understanding coercivity mechanisms in nanoscale low-dimensional ferromagnets. In fact, ferromagnetism is a size-dependent physical phenomenon, as revealed by theoretical studies. However, nanoscale ferromagnetic samples with controllable size and shape have been available only in recent times. By adopting newly developed salt-matrix annealing, surfactant-assisted milling, and improved hydrothermal and chemical solution techniques, we used a bottom-up approach to produce nanostructured magnets and have successfully synthesized monodisperse ferromagnetic Fe-Pt, Fe-Co and Sm-Co nanoparticles and Co nanowires with extraordinary properties, which are strongly size- and shape-dependent. A study on size-dependent Curie temperature of the L10 ferromagnetic nanoparticles with sizes down to 2 nm has experimentally proved a finite-size effect. A systematic study of nanowires with extremely high coercivity above their magnetocrystalline anisotropy fields has opened a door to the solution of Brown’s Paradox.

Thermoelectric materials can generate clean electricity by the scavenging of waste heat and realize precise spot cooling by applying current. The performance of thermoelectric materials is determined by the intrinsic chemical bonding mechanism and the extrinsic manipulation of structural defects, both affecting the transport of electrons and phonons. An unconventional chemical bonding mechanism, named metavalent bonding, has been demonstrated unique to the high thermoelectric performance of many chalcogenides. Besides, lattice defects such as impurity clusters, Cottrell atmospheres, and interfacial complexions provide other freedoms to tune the thermoelectric properties in addition to traditional manipulations of the defect's dimensionality and density. In this talk, I will introduce metavalent bonding, experimental methods to characterize bonding and nano-chemistry of defects, as well as the correlations between these features and thermoelectric response.