The ability to design crystal lattices with specific sublattice geometries has long been sought after especially in the context of harnessing magnetic frustration to elicit emergent physics. One approach which has seen some success recently is to design magnetic sublattices chemically through the use of nonmagnetic linker ligands as scaffolding for the magnetic ions. Here we report the magnetic properties of one such family of materials, the transition metal ($\mathit{TM}$) selenite hydrates with chemical formula $T{M}_3{\left({\mathrm{SeO}}_3\right)}_3{\mathrm{H}}_2\mathrm{O}$. These materials link highly distorted $\mathit{TM}{\mathrm{O}}_6$ octahedra via nonmagnetic [${\mathrm{SeO}}_3{]}^{2+}$ linkers. Studying members with $\mathit{TM}$ = Mn, Co, and Ni we use magnetic susceptibility and neutron powder diffraction to identify antiferromagnetic order in all three compounds with moderate frustration indexes. A comparison of the magnetic structures suggests that while the overall structure of the ${\mathrm{SeO}}_3$ scaffolding remains unchanged the $\mathit{TM}$ effects changes in the magnetic properties through tuning the internal bonding parameters and the single-ion physics. Changing from Mn to Co is found to be particularly consequential manifesting changes in both the ordered-moment direction and in the direction of the ordering vector. Field-dependent measurements of the susceptibility and heat capacity reveal metamagnetic transitions in both ${\mathrm{Mn}}_3{\left({\mathrm{SeO}}_3\right)}_3{\mathrm{H}}_2\mathrm{O}$ and ${\mathrm{Co}}_3{\left({\mathrm{SeO}}_3\right)}_3{\mathrm{H}}_2\mathrm{O}$ indicating nearby magnetic ground states accessible under relatively small applied fields. Density functional theory calculations broadly confirm these results, showing both a sensitivity of the magnetic structure to the $\mathit{TM}$ and its local environment. Although no spin liquid behavior is achieved, these results suggest the fruitfulness of such synthesis philosophies and encourage future work to engender higher frustration in these materials via doping, field, pressure, or larger linker ligands.

• Received 17 October 2019
• Accepted 4 February 2020

DOI:https://doi.org/10.1103/PhysRevMaterials.4.024410