Unmasking the interior magnetic domain structure and evolution in Nd-Fe-B sintered magnets through high-field magnetic imaging of the fractured surface

Conventional magnetic imaging techniques have observed magnetic domains in the polished surface of Nd-Fe-B sintered magnets, but the mechanical processing involved introduces numerous defects that facilitate the nucleation of reversed domains, thereby masking the interior domain structure. Here, we utilize high-field synchrotron x-ray magnetic circular dichroism microscopy to map the elemental and magnetic distributions in the polished and fractured surfaces of a Nd-Fe-B sintered magnet throughout its entire demagnetization process. As the applied field is varied, the domains in the fractured surface behave completely differently from those in the polished surface, thereby unmasking the interior domain structure and behavior. The area-averaged fractured surface coercivity is ${\mu }_0{H}_{\mathrm{c}}^{\mathrm{frac}}=0.85\mathrm{T}$ which is much higher than the area-averaged polished surface coercivity, ${\mu }_0{H}_{\mathrm{c}}^{\mathrm{pol}}=0.5\mathrm{T}$. Most of the local magnetic hysteresis loops are positively or negatively biased from the zero of the field axis. The highest coercivity grains, some of which exceed 2 T, are almost always located in the vicinity of strongly oppositely biased adjacent grains. This indicates that these oppositely biased grains are strongly influencing the magnetostatic field at the sample surface.

Figure 1

SEM images and height maps of the polished and fractured surfaces. (a) SEM image of the polished surface. The scale bar is indicated below the image and the yellow square approximately indicates the area where the height maps and x-ray absorption images (shown later) were taken. (b),(c) Height maps of the polished surface as measured by a confocal microscope and plotted parallel to the magnetic easy axis and a different angle, respectively, to demonstrate the surface morphology. The color scale is the same in (b) and (c). (d)–(f) Same as (a)–(c), respectively, but for the fractured surface. The color scale is the same in (e) and (f), but is different from (b) and (c). Note that the SEM and confocal microscope measurements were made after the x-ray absorption microscopy measurements and the red arrows in (a)–(c) point to a contaminant on the polished surface that was not present during the x-ray absorption microscopy measurements.

Figure 2

Composition mapping of the polished and fractured surfaces. (a) Normalized absorption ratio map, $A(x,y)$, of the polished surface [as defined in Eq. (2)], which is proportional to the quotient of the background-corrected Fe $L_3$ edge and Nd $M_4$ edge absorption maps. (b) Histogram of the image shown in (a). (c) Same as (a), but for the fractured surface. (d) Histogram of the image shown in (c).

Figure 3

Magnetic domain images and composition-dependent hysteresis loops from the polished and fractured surfaces. (a),(b) Polished surface Fe $L_3$ edge XMCD maps taken just above and below its coercivity $H_{\mathrm{c}}^{\mathrm{pol}}$, respectively. (c) Hysteresis loop taken from compositionally different regions of the polished surface, as determined by their $A(x,y)$ value in Fig. 2. The legends indicate the range of $A$ values used to determine each hysteresis loop. (d),(e) Same as (a) and (b), respectively, but for the fractured surface and its coercivity, $H_{\mathrm{c}}^{\mathrm{frac}}$. (f) Same as (c), but the compositionally different regions are determined from their $A(x,y)$ value in Fig. 2. The area shown in (a) and (b) is the same as Fig. 2, and the area shown in (d) and (e) is the same as Fig. 2.

Figure 4

Magnetic domain images from the polished and fractured surfaces in the remanent state. (a),(b) Polished and fractured surface Fe $L_3$ edge XMCD maps, respectively, both taken at ${\mu }_0{H}_{\mathrm{ext}}=0.0\mathrm{T}$ after being saturated at ${\mu }_0{H}_{\mathrm{ext}}=-3.0\mathrm{T}$.

Figure 5

Local coercivities and local magnetic biases of the polished and fractured surfaces. (a),(b) Local coercivity and local magnetic bias, ${\mu }_0{H}_{\mathrm{c}}(x,y)$ and ${\mu }_0{H}_{\mathrm{b}}(x,y)$, of the polished surface, as defined in Eqs. (3) and (4), respectively. The magenta circle in (b) encloses a cluster of grains labeled 1–4. (c) Local hysteresis loops for the grains labeled 1–4 in (b). (d),(e) Same as (a) and (b), respectively, but for the fractured surface. The magenta circle in (e) encloses a cluster of grains labeled 1–3. (f) Local hysteresis loops for the grains labeled 1–3 in (e). The area shown in (a) and (b) is the same as Fig. 2, and the area shown in (d) and (e) is the same as Fig. 2. The color scale is the same in (a) and (d), but is different from (b) and (e).