Page 72 - Read Online

Lin et al. Soft Sci 2023;3:14 https://dx.doi.org/10.20517/ss.2023.05 Page 9 of 25

material, weight ratio, and magnetization direction of the magnetic micro/nanoparticles serves as an
effective approach to adjusting the tunable elastic and magnetic properties of the composites [114,115] .

Figure 3A illustrates a soft electronic system based on magnetic composites. The system consists of
magnetic microparticles (neodymium iron boron (NdFeB), average particle radius: ~5.00 μm; remanence:
92.3 emu g ; coercive magnetic field: 6375 Oe) and a porous silicone rubber matrix. The embedded
–1
ferromagnetic particles enable the system to convert mechanical deformations into electricity using coils.
Compared with systems that mix the iron micro/nanoparticles with silicone rubber, the example shown in
Figure 3A can generate electric signals in the absence of external magnetic fields, which simplifies the device
configuration [116,117] .

In the process of heating the mixture of uncured silicone rubber and micro/nanomagnets, an option is to
introduce air bubbles to form porous composites. Such porous structures not only reduce the modulus, but
also provide tiny spaces for the micro/nanomagnets to reorientate and move during impulse magnetization,
eventually forming a chain-like arrangement of the micromagnets associated with the giant magnetoelastic
effect. Compared with the traditional magnetoelastic effect arising from magnetic domain rearrangement
and stress-induced magnetic anisotropy under external magnetic field, this giant magnetoelastic effect is
attributed to the change of micro-magnetic chain structure under mechanical deformation. As shown in the
right frame of Figure 3A, compression changes the chain structure of the magnet, causing a decrease in
surface magnetic flux density. The soft system can withstand a tensile strain of up to 190%, and
demonstrates much lower modulus (at the level of 100 kPa,) compared with conventional magnetic alloy
[118]
(modulus at the level of 10 GPa) . Such features enhance the biomechanical-to-magnetic energy
conversion and open many potential applications in soft electronics, including wearable/implantable energy
harvesters, and stretchable biomedical sensors for continuous monitoring of human pulse waves and heart
rhythms .
[119]
Another advantage of magnetic composites is that the direction of magnetization can be programmed in a
heterogeneous manner, as the composite materials are soft enough to be folded, wrapped, or otherwise
deformed into 3D geometries during magnetization [120,121] . In particular, magnetizing a flexible magnetic film
with a sinusoidal pattern enables high spatial resolution and the capability to decouple different mechanical
[122]
stimuli [Figure 3B] . The soft tactile sensor involves a soft magnetic composite (modulus: ~2 MPa;
thickness: 0.5 mm) formed by mixing polydimethylsiloxane (PDMS) and NdFeB magnetic powders with a
weight ratio of 1:3, a silicone elastomer layer (Ecoflex 00-50, modulus: ~83 kPa), and a commercial Hall
sensor [left frame of Figure 3B]. The soft magnetic composite deforms under external force and causes a
change in magnetic flux density, which can be detected by the Hall sensor embedded in the silicone
elastomer [right frame of Figure 3B]. Here, the sinusoidal pattern (obtained by wrapping the soft magnetic
composite on a cylinder during magnetization) is crucial for decoupling and super-resolution. On the one
hand, the magnetic field distribution caused by the sinusoidal pattern is decoupled into two mutually
orthogonal planes: the magnetic strength B and the magnetic ratio R . The B plane determines the normal
B
force related to the magnetic field rotation angle (or displacement along the z-axis), while the R plane
B
measures the shear force related to the translational movement of the magnetic field (displacement d along
the x-axis). The capability to decouple various mechanical stimuli overcomes the inherent problem of the
strong cross-coupling effects in conventional magnetic tactile sensors [123-125] . On the other hand, extending
the sinusoidal patterns to the form of sensor arrays allows for tactile sensing with super-resolution and
across large areas. Combined with deep learning algorithm, the sensors can achieve a 60-fold improvement
in localization (from 6 to 0.1 mm).