Unconventional magnetic materials including non-collinear antiferromagnets (nc-AFM), p-wave magnets and altermagnets, are an emerging frontier for quantum spintronics and hybrid quantum devices. Critical to the application of these materials is control over the magnetic domain state, as their unique, symmetry-driven properties vanish in a multi-domain limit. However, the mechanisms governing domain formation in materials with compensated local moments remain poorly understood. In this work, we examine the ferrimagnetic to nc-AFM phase transition of Mn3NiN using scanning nitrogen-vacancy centre magnetometry. We provide nanoscale mapping of the magnetic domain evolution on cooling and correlate the local stray fields with global magnetometry and anomalous Hall effect measurements. We observe the formation of a disordered, dendritic domain structure whose roughness is quantified using its fractal dimension. The fractal dimension steadily increases on cooling through the transition, saturating at a value of ∼1.55 in the non-collinear phase, but the domain area distribution does not show any significant changes. We show this behaviour cannot be explained by the balance of demagnetisation energy and domain wall energy, and conclude elastic contributions and defects are a critical factor to explain the domain size.
Figures
16 panels with data across 5 figures. Each panel page shows the plot, its columns and its files; each data.csv begins with a header naming the paper, the panel, the source, the license and the provenance route.
Fig. 1
panel (b): X-ray reflectivity scan of the Mn$_3$NiN film against $2\theta$ (deg), on a log scale. The measured points are shown; the printed fit curve is not. Reflectivity is in detector counts, where the printed axis carries no numbers. data.csv
panel (d): Stray field map of the Mn$_3$NiN film surface at 260 K: out-of-plane field ($\mu$T) against position $x$, $y$ ($\mu$m). Axes in micrometres replace the printed scale bar. data.csv
Fig. 2
panel (a): Scanning nitrogen-vacancy magnetometry stray field map of Mn$_3$NiN at 230 K on cooling: field ($\mu$T) against position $x$, $y$ ($\mu$m). Axes in micrometres replace the printed scale bar. data.csv
panel (b): Scanning nitrogen-vacancy magnetometry stray field map of Mn$_3$NiN at 210 K on cooling: field ($\mu$T) against position $x$, $y$ ($\mu$m). Axes in micrometres replace the printed scale bar. data.csv
panel (c): Scanning nitrogen-vacancy magnetometry stray field map of Mn$_3$NiN at 190 K on cooling: field ($\mu$T) against position $x$, $y$ ($\mu$m). Axes in micrometres replace the printed scale bar. data.csv
Fig. 3
panel (a): Mean stray field per $\mu$m$^2$ of each field map ($\mu$T/$\mu$m$^2$) against temperature (K), red stars, one per image. The printed error bars and the field-cooled magnetisation curve on the right-hand axis are not shown. data.csv
panel (b): Anomalous Hall resistivity $\rho_{xy}$ ($\mu\Omega$cm) against temperature (K), measured on cooling in 150 mT. The temperature axis is labelled here; the print draws this panel on a wider scale (180 to 290 K) than the labels under (c). data.csv
panel (c): Mean absolute stray field per $\mu$m$^2$ of each field map ($\mu$T/$\mu$m$^2$) against temperature (K), one square per image. The printed error bars are not shown. data.csv
Fig. 4
panel (a): Stray field map at 190 K used for the domain analysis: field ($\mu$T) against position $x$, $y$ ($\mu$m). The print shows it without axes or colour bar; the colour limits are those of Fig. 2. data.csv
panel (b): The 190 K field map of (a) binarised with Otsu's threshold: red above the threshold, blue below, against position $x$, $y$ ($\mu$m). Computed without preprocessing, since the paper states none; it differs from the printed map on about 11% of pixels. data.csv
panel (c): Regions of the 190 K map of (b) identified as positive domains, each connected region coloured by its label; other pixels are blank. Computed from the deposited maps with Otsu's threshold and no minimum domain size, which the paper does not state; it differs from the print. data.csv
panel (d): Regions of the 190 K map of (b) identified as negative domains, each connected region coloured by its label; other pixels are blank. Computed from the deposited maps with Otsu's threshold and no minimum domain size, which the paper does not state; it differs from the print. data.csv
panel (e): Perimeter ($\mu$m) against area ($\mu$m$^2$) of each domain in the three 190 K maps, on log axes, with a linear fit per map. No offsets are added, where the print shifts the maps apart. Domains come from Otsu's threshold after Gaussian smoothing ($\sigma$ = 1 px), keeping domains of at least 10 px; the paper does not state these steps, which were inferred to match the print. data.csv
panel (f): Fractal dimension $D_f$ of each field map against temperature (K), with the standard error of the fitted slope as error bars. The printed dashed guide to the eye is not shown. Domains come from Otsu's threshold after Gaussian smoothing ($\sigma$ = 1 px), keeping domains of at least 10 px; the paper does not state these steps, which were inferred to match the print. data.csv
panel (g): Fractal dimension $D_f$ of each field map against its mean absolute field per $\mu$m$^2$ ($\mu$T/$\mu$m$^2$, as in Fig. 3(c)), with the standard error of the fitted slope as error bars. Domains come from Otsu's threshold after Gaussian smoothing ($\sigma$ = 1 px), keeping domains of at least 10 px; the paper does not state these steps, which were inferred to match the print. data.csv
Fig. 5
panel (1): Domain area ($\mu$m$^2$) of every domain in the field maps at each temperature (K), drawn as points over a quartile box where the print draws a violin. Computed from the deposited maps with Otsu's threshold and no minimum domain size, which the paper does not state; it differs from the print. data.csv
Cite
Freya Johnson, Jan Zemen, Helena Knowles, Lesley F Cohen. Uncovering domain morphology in an unconventional magnet with scanning diamond quantum magnetometry. Mater. Quantum. Technol. 6, 025201 (2026). https://doi.org/10.1088/2633-4356/ae64a0
When you use hosted data, cite the original paper and give the panel's URL so a reader can find the exact values you used.