Controlling Noncollinear Ferromagnetism in van der Waals Metal-Organic Magnets

Abstract

Van der Waals (vdW) magnets both allow exploration of fundamental 2D physics and offer a route toward exploiting magnetism in next generation information technology, but vdW magnets with complex, noncollinear spin textures are currently rare. We report here the syntheses, crystal structures, magnetic properties and magnetic ground states of four bulk vdW metal-organic magnets (MOMs): FeCl₂(pym), FeCl₂(btd), NiCl₂(pym), and NiCl₂(btd), pym = pyrimidine and btd = 2,1,3-benzothiadiazole. Using a combination of neutron diffraction and bulk magnetometry we show that these materials are noncollinear magnets. Although only NiCl₂(btd) has a ferromagnetic ground state, we demonstrate that low-field hysteretic metamagnetic transitions produce states with net magnetization in zero-field and high coercivities for FeCl₂(pym) and NiCl₂(pym). By combining our bulk magnetic data with diffuse scattering analysis and broken-symmetry density-functional calculations, we probe the magnetic superexchange interactions, which when combined with symmetry analysis allow us to suggest design principles for future noncollinear vdW MOMs. These materials, if delaminated, would prove an interesting new family of 2D magnets.

Figure 1

Crystal structure of MCl₂(pym) viewed along the (a) c-axis and (c) a-axis and MCl₂(btd) viewed along the (b) c-axis and (d) a-axis. The hydrogen atoms are omitted for clarity. (e) Oak Ridge Thermal Ellipsoid Plot (ORTEP) of FeCl₂(pym) showing the coordination environment.

Figure 2

Rietveld refinement of the nuclear structures against powder neutron diffraction data. The measurement temperature for each data set is at FeCl₂(pym), T = 12.5 K; FeCl₂(btd), T = 5 K; NiCl₂(pym), T = 2 K and NiCl₂(btd) T = 2 K. For NiCl₂(pym) the first magnetic Bragg peak (Q = 0.70 Å^–1) was omitted and magnetic Bragg intensity at higher Q was negligible. For NiCl₂(btd) the magnetic Bragg intensity was fixed to values determined from magnetic Rietveld refinement (see Figure 6).

Figure 3

Magnetic susceptibility, χ(T), measurements in zero-field cooled (ZFC) and field cooled (FC) conditions from 2–300 K under a 0.01 T dc field for (a) FeCl₂(pym), (b) FeCl₂(btd), (c) NiCl₂(pym) and (d) NiCl₂(btd).

Figure 4

Isothermal magnetization measurements, M(H), for (a) FeCl₂(pym) at 1.8 K between −2.6 T to 2.6 T, (b) FeCl₂(btd) at 1.8 K between −0.3 T to 0.3T, (c) NiCl₂(pym) at 3 K between −11 T to 11 T and (d) NiCl₂(btd) at 1.8 K between −7 T to 7 T.

Figure 5

Definition of the canting angle γ and angle between the local easy-axes, ϕ and the collinear direction.

Figure 6

Rietveld refinement of the magnetic ground states against temperature subtracted neutron diffraction data. FeCl₂(pym): The model was refined against the I_1.5 K – I_12.5 K data set over 0.36 < Q < 2.37 Å^–1. FeCl₂(btd): The model was refined against the I_1.5 K – I_5 K data set over 0.26 < Q < 1.98 Å^–1. Data at 0.82 < Q < 0.09 Å^–1 were omitted due to incomplete peak subtraction caused by thermal expansion. NiCl₂(pym): The model was refined against the I_2 K – I_30 K data set over 0.29 < Q < 2.61 Å^–1. Data at 1.97 < Q < 2.04 Å^–1 were omitted due to incomplete peak subtraction caused by thermal expansion. NiCl₂(btd): The model was refined against the I_1.5 K – I_30 K data set over 0.59 < Q < 1.61 Å^–1. Data outside this range were omitted due to the absence of magnetic Bragg peaks and the presences of features arising from incomplete subtraction of structural Bragg peaks due to thermal expansion.

Figure 7

Schematic representation of the magnetic ground states of (a) FeCl₂(pym), (b) FeCl₂(btd), (c) NiCl₂(pym) and (d) NiCl₂(btd).

Figure 8

Magnetic diffuse scattering of FeCl₂(btd) fit using an effective field model. Data obtained by temperature subtraction of data measured at 30 K.