Characteristics of ferroelectric-ferroelastic domains in Néel-type skyrmion host GaV4S8

Abstract

GaV4S8 is a multiferroic semiconductor hosting Néel-type magnetic skyrmions dressed with electric polarization. At Ts = 42 K, the compound undergoes a structural phase transition of weakly first-order, from a non-centrosymmetric cubic phase at high temperatures to a polar rhombohedral structure at low temperatures. Below Ts, ferroelectric domains are formed with the electric polarization pointing along any of the four 〈111〉 axes. Although in this material the size and the shape of the ferroelectric-ferroelastic domains may act as important limiting factors in the formation of the Néel-type skyrmion lattice emerging below TC = 13 K, the characteristics of polar domains in GaV4S8 have not been studied yet. Here, we report on the inspection of the local-scale ferroelectric domain distribution in rhombohedral GaV4S8 using low-temperature piezoresponse force microscopy. We observed mechanically and electrically compatible lamellar domain patterns, where the lamellae are aligned parallel to the (100)-type planes with a typical spacing between 100 nm–1.2 μm. Since the magnetic pattern, imaged by atomic force microscopy using a magnetically coated tip, abruptly changes at the domain boundaries, we expect that the control of ferroelectric domain size in polar skyrmion hosts can be exploited for the spatial confinement and manipulation of Néel-type skyrmions.

Relaxation dynamics of modulated magnetic phases in the skyrmion host GaV4S8 : An ac magnetic susceptibility study

“…These regions are presumably characterized by the coexistence of the magnetic phases [34,37], as has been observed in various skyrmion host compounds in real space via magnetic force microscopy [46] and Lorentz-TEM imaging [4,6,10,47]. In the case of GaV 4 S 8 , the lamellar structure of the ferroelastic domains with a typical thickness in the micrometer range [27] may introduce significant disorder in the vicinity of domain boundaries and may also enhance demagnetizing fields in the sample. However, the relatively wide susceptibility peaks as compared to the values of the critical magnetic fields suggest that disorder is likely to play the key role in the broadening of phase transitions.…”

“…As a result a sizable ferroelectric polarization develops along the rhombohedral axis [25,26]. The depolarization field is reduced by the formation of submicron-sized structural domains of the four possible rhombohedrally distorted variants with the different 111 -type rhombohedral axes [21], assembling into an alternating lamellar domain structure [27].…”

“…Therefore, the structural domain boundaries are necessarily boundaries to the consistent magnetic structures as well [27].…”

Magnetic phases of skyrmion-hostingGaV4S8−ySey(y=0,2Franke1, Huddart2, Hicken3 et al. 2018Phys. Rev. B302323View full textAdd to dashboardCite

“…At room temperature, these materials have a noncentrosymmetric cubic crystal structure with T d point symmetry, but at around 40 K (in both materials) a Jahn-Teller transition changes this to a rhombohedral polar C 3v symmetry by stretching the lattice along one of the four 111 cubic axes [15,17,21]. Below the respective magnetic ordering temperatures (T c ≈ 12.5 K for GaV 4 S 8 [15,22] and T c ≈ 17.5 K for GaV 4 Se 8 [17,23]), applied magnetic fields drive successive transitions between cycloidal (C), SkL, and field-polarized (FP) phases.…”

Superdomain dynamics in ferroelectric-ferroelastic films: Switching, jamming, and relaxation

“…(A good exception however, is [38,39] SrxBa1-xNb2O6 with 0.60<x<0.75 with nano-ferroelastic domains clustered around the fourfold channels in this tungsten bronze, inside 300-nm-diameter ferroelectric domains; and a more recent exception is the spinel GaV4S8, where skyrmion-like ferroelastic domains of diameter 20 nm are found inside ferroelectric domains an order of magnitude larger. [40]) One might ask why, that is, why does the strain energy not often also average to zero in the same way? We believe that the answer is that strain is never screened whereas charge and polarization are for the relevant length scale.…”