Paste-Injection of Low-Density Barium Hexaferrite Magnets with Soft Magnetic Iron Phase

Abstract

Permanent magnets of varying shapes and sizes are increasingly produced. For hexaferrite magnets, it is challenging to incorporate polymers and a soft magnetic phase in the form of paste before injection molding or extrusion free-forming. In this study, hard magnetic barium hexaferrite/soft magnetic iron composites with a density of 2.28–2.34 g/cm³ are obtained after paste-injection molding and subsequent sintering at 1150 °C for 5 h. Variations of the binder (143.5–287.0 mg poly(vinyl alcohol), PVA) and the plasticizer (75–150 mg poly(ethylene glycol), PEG-400) in the ceramic–polymer paste give rise to comparable remanent magnetization (33.10–33.63 emu/g) and coercivity (3854–3857 Oe). Unlike all-ferrite systems, the presence of a soft magnetic metal phase is not detrimental to the coercivity. However, the remanent and saturation magnetizations are not substantially increased. The addition of 1% and 5% of iron oxide in the ceramic–polymer paste gives rise to hard/soft composites with lower densities of 2.11 and 2.14 g/cm³. The coercivity is increased to 3942–3945 Oe; however, the maximum energy product is reduced.

1. Introduction

Permanent magnets are traditionally manufactured by powder compaction and injection molding. Casting is also used in the case of alnico magnets. The advent of new technology has increased the variety of shapes and sizes of magnets used in machines and devices. To keep up with such demands, additive manufacturing has been adopted for producing permanent magnets with complex and diverse geometries [1,2,3,4,5]. White et al. demonstrated the laser-engineered net shaping of alnico magnets without any molds and binders [6]. Near-net-shape alnico magnets were also produced by powder bed fusion by either laser melting [6] or electron beam melting [7]. To obtain higher maximum energy products, Nd-Fe-B typed magnets were manufactured by powder bed fusion, vat photo-polymerization, binder jetting, and fuse filament fabrication [1,2,3,4,5]. Palmero et al. also investigated the fuse filament fabrication of rare-earth-free magnets by filling manganese- aluminum in polyethylene (PE) [8] and acryl butadiene styrene (ABS) [9]. Magnetic polymer composites drawn into filaments can be extruded using commercial 3D printers.

For hard ferrites, strontium hexaferrite (SrFe₁₂O₁₉) and barium hexaferrite (BaFe₁₂O₁₉) with the M-type hexagonal structure have been produced as ceramic magnets with varying dimensions. These hexagonal ferrites are low-cost alternatives to rare-earth magnets [10,11]. Examples of ferrite polymer composites for fuse filament fabrications are 76 wt.% BaFe₁₂O₁₉ in ABS [12] and 90 wt.% SrFe₁₂O₁₉ in ethylene ethyl acrylate (EEA) [13]. In addition to fuse filament fabrication, extrusion free-forming is a promising route for printing bulk ferrites with diverse geometries. As with other extrusion techniques, a large fabrication volume is attainable at a relatively low cost. A large volume of ceramic paste could be injected into molds or extruded through a nozzle of commercial 3D printers. BaFe₁₂O₁₉ magnets can be synthesized from BaCO₃ and Fe₂O₃ precursors according to the following reaction:

BaCO₃ + 6Fe₂O₃ → BaFe₁₂O₁₉ + CO₂

(1)

After sintering at 1200 °C, Peng et al. obtained 95% of the theoretical saturation magnetization of the BaFe₁₂O₁₉ [14]. Wei et al. also successfully used extrusion free-forming to produce BaFe₁₂O₁₉ and SrFe₁₂O₁₉ magnets with coercivities of 4–6 kOe and saturation magnetizations close to the theoretical values [15]. The 3D-printed SrFe₁₂O₁₉ magnets exhibited the maximum energy product exceeding 2.5 MGOe. Alternatively, Yang et al. fabricated complex-shaped SrFe₁₂O₁₉ magnets from 3D gel printing and found the maximum magnetic energy products as high as 3.31 MGOe after sintering at 1300 °C [16]. From the paste-injection molding, a coercivity of up to 4 kOe and a saturation magnetization of around 73 emu/g were obtained in low-density BaFe₁₂O₁₉ magnets [17].

In addition to sintered and polymer-bonded magnets, hexagonal ferrites have been produced into composites to enhance the coercivity and remanent magnetization contributing to the maximum energy product. Yang et al. used mixed NdFeB and SrFe₁₂O₁₉ pastes to increase the maximum energy product of 3D-printed magnets [18]. As early as 1991, Kneller and Hawing described the exchange-spring magnetic coupling between the hard and soft magnetic layers [19]. The combination of high coercivity from the former with high magnetization from the latter has also been investigated in core–shell structures and particulate composites. These hard/soft magnetic composites have attracted significant interest as permanent magnets. Composites of SrFe₁₂O₁₉ and magnetite exhibited a coercivity of 3.3–3.7 kOe and saturation magnetization around 70 emu/g [20,21]. Other spinel ferrites, including nickel and zinc ferrites, were incorporated into hard ferrites using various methods [22,23,24,25]. Magnetic properties are dependent on the phase composition and particle size. The exchange coupling is enhanced between hard hexagonal ferrite and soft spinel ferrite prepared by one-pot chemical synthesis, compared to the physical mixing of powders. Cobalt ferrite was used to produce hard/soft magnetic composites with either SrFe₁₂O₁₉ [26,27] or BaFe₁₂O₁₉ [28,29,30,31,32]. The soft ferrite phase may increase the magnetization, but reduce the coercivity [20,21,22,27,28,29]. Nevertheless, the composites of modest maximum energy products still attract interest in microwave absorbing and antimicrobial properties [27,28,29,33,34].

Ferromagnetic cobalt (Co) and iron (Fe) also exhibited exchange coupling with hard ferrites, providing alternatives to the all-ferrite systems [35,36,37,38]. SrFe₁₂O₁₉/Co and SrFe₁₂O₁₉/Fe-Co were synthesized in the form of core–shell structures [36,37]; and the Fe phase can also be formed by the iron oxide precursor [38]. In this study, four formulations of aqueous ceramic pastes are used to synthesize soft magnetic Fe in hard magnetic BaFe₁₂O₁₉; and the magnets obtained from paste-injection molding are compared. These ceramic pastes can possibly be printed without relying on molds; this is advantageous for producing permanent magnets with varying shapes and sizes in future applications.

2. Materials and Methods

2.1. Manufacturing of Hard/Soft Magnetic Composites

BaCO₃ (>99.0 purity, HiMedia) and Fe₂O₃ (99.0% purity, SIAL) powders were used as starting materials in a stoichiometric ratio of 1:6. The powders were ground in a mortar and then sieved through a 100-mesh. Ceramic–polymer pastes were prepared using different formulations listed in

Table 1. Formulations of the ceramic–polymer pastes, and the density averaged from 3 samples of sintered hard/soft magnetic composites.

Formulation/Samples	BaCO3/Fe2O3 (g)	Additional Fe2O3 (g)	PVA (mg)	PEG-400 (mg)	EFKA®4620 (mg)	DI Water (mL)	Density of Magnets (g/cm3)
I/S18–S20	15	-	287.0	150	40	4.6	2.34 ± 0.03
II/S21–S23	15	-	143.5	75	40	4.6	2.28 ± 0.05
III/S24–S26	14.85	0.15	574	300	40	4.6	2.15 ± 0.02
IV/S27–S29	14.25	0.75	574	300	40	4.6	2.11 ± 0.06

and shown in Figure 1. For the first formulation (S18–S20), the BaCO₃/Fe₂O₃ powders were mixed with the binder (287 mg of poly(vinyl alcohol), PVA). The plasticizer (150 mg of poly(ethylene glycol), PEG-400), the dispersing agent (40 mg of EFKA^® FA4620), and DI water (4.6 mL) were then added. The PVA and PEG-400 in the aqueous-based paste were reduced by half for formulation II (S21–S23). To increase a soft magnetic phase, formulations III and IV were modified by adding Fe₂O₃ in a percentage of 1% and 5%, respectively. The amounts of PVA and PEG-400 were increased to 574 and 300 mg, accordingly. Each homogeneous paste was injected into 0.4 × 0.4 × 1.8 cm³ molds using syringes. The magnets were left to dry for 3 h at room temperature, and subsequently sintered at 1150 °C for 5 h. The sintering is universally aimed to enhance the BaFe₁₂O₁₉ phase; however, the grain growth by excessive sintering may lead to multidomains, which tend to reduce the coercivity.

2.2. Characterization of Hard/Soft Magnetic Composites

The mass and size of each magnet were measured to determine its density in g/cm³. The mixed phases were identified by an X-ray diffractometer (XRD; Rigaku, SmartLab, Austin, TX, USA) using 1.54060 Å CuKα1 radiation. A scanning electron microscope (SEM; FEI, Quanta 250, Hillsboro, OR, USA) with an energy-dispersive X-ray spectrometer (EDS; Oxford, X-max50, UK) examined the morphology and elemental compositions on the surface of magnets.

Using a vibrating sample magnetometer (VSM; in-house developed and calibrated with Lakeshore 730,908), the in-plane magnetic field (H) from an electromagnet was scanned between −17.5 kOe and 17.5 kOe. The resulting magnetic flux was then detected by a coil around a sample of a vibrating magnet. Magnetic properties were then deduced from the hysteresis loop by plotting the magnetization (M) against the external magnetic field. The coercivity and the remanent magnetization are the x-intercept and y-intercept, respectively. The maximum energy product ((BH)_max) was determined from the second quadrant of each hysteresis loop. Using the law of approach to saturation in a regime close to the maximum applied field detailed in [39,40], the saturation magnetization (M_s) was approximated from the plot between M and 1/H² according to the following equation:

$$M=M_s \left[1-\frac{b}{H^2}\right]$$

(2)

where b is a constant.

3. Results and Discussion

3.1. Density and Morphology of Hard/Soft Magnetic Composites

The hexaferrite powder compaction gives rise to a density close to a theoretical value of around 5 g/cm³ [33]. By contrast, magnets from the paste injection without high-pressure compaction in Table 1 have much lower densities of 2.11–2.34 g/cm³. The standard deviation of the density averaged from three replicates is 0.06 g/cm³ or less; indicating control over the density of magnets synthesized from the ceramic–polymer paste. The reproducibility of the process confirms that the injection or extrusion free-forming can be practically applied using these paste formulations.

All the SEM images in Figure 2 reveal that the magnets are similarly composed of hexagonal platelets and short prisms of hundreds of nanometers in cross-section. This size is comparable to the range of high coercivity. These particles tend to form clusters by ferrimagnetic interaction through the Van der Waals force [23]. They are not closely packed and the voids observed in Figure 2 are consistent with a lower density than compact magnets [33]. The lowest density is obtained in Figure 2d with many voids among large clusters.

3.2. Phase and Elemental Composition of Hard/Soft Magnetic Composites

Elemental composition mapping is exemplified by the spectra in Figure 3. The Fe and Ba are evenly distributed on the surface of each magnet. By contrast, O is depleted in some areas, appearing as a dark background. After subtracting the amount of C from the residual peak, the composition of Fe:Ba is close to the stoichiometric ratio of BaFe₁₂O₁₉ with a considerable variation in O from 15.05 to 25.09. For samples S26 and S29 prepared with additional Fe₂O₃, the Fe:Ba value is slightly increased to 11.85:1 and 11.49:1, respectively. Although the elemental compositions are consistent with BaFe₁₂O₁₉, the EDS technique does not identify the phase and the ferrite types cannot be distinguished. The quantitative analysis from EDS is also susceptible to errors because data are only acquired from a small area of each sample; in addition, the peaks associated with carbon, oxygen, and other light elements overlap. The XRD is used to complement EDS and characterize the phase, as with previous reports on hard/soft ferrite composites [22,23,24].

The M-type BaFe₁₂O₁₉ phase comparable to the literature [41] is confirmed in every sample by XRD spectra in Figure 4. The diffracted peaks from the crystallographic planes of (110), (107), (114), (203), (205), (217), (2011), (220), (2014), and (317) are at 30.31°, 32.20°, 34.11°, 37.08°, 55.06°, 56.60°, 63.06°, 67.36°, and 72.59°, respectively (JCPDS: 00-043-0002). Interestingly, the most intense peak in each spectrum is characterized as the metallic Fe at 43.50° from the (110) plane (JCPDS: 01-071-4650). The minor Fe peak at 63.20°, corresponding to the (200) plane, is also present near a minor BaFe₁₂O₁₉ peak. The Fe peak intensities of three replicates (S18–S20, S21–S23, S24–S26, and S27–S29) vary substantially within each formulation. Nevertheless, the crystallite size, determined from the width of the major Fe peak using the Scherrer equation [42], is relatively constant. The average Fe crystallite sizes obtained from different formulations are comparable at 24 nm. This size is smaller than a twice domain wall thickness of hexaferrites; enabling the exchange coupling between hard and soft magnetic phases [34].

3.3. Magnetic Properties of Hard/Soft Magnetic Composites

All the hysteresis loops in Figure 5 are smooth without kink or shift, signifying the exchange coupling in hard/soft magnetic composites [37]. For composites with incomplete exchange coupling [33,34], the stepped loops or kinks in the second quadrant were observed because of separate switching of the hard and soft phases. Because the increase in magnetization does not saturate under 17,500 Oe, the saturation magnetization is deduced from the plots between M and 1/H² following the law of approach to saturation. From Figure 6, a coefficient of determination (R²) of 0.9969–0.9985 indicates a good fitting of hysteresis data; and the y-intercept gives rise to a saturation magnetization over 67 emu/g.

All the M_r/M_s ratios in

Table 2. Coercivity (H_c), remanent magnetization (M_r), saturation magnetization (M_s), and maximum energy product ((BH)_max) of the hard/soft magnetic composites averaged from 3 replicates.

Samples	Synthesis Composition	Magnetic Properties
		Hc (Oe)	Mr (emu/g)	Ms (emu/g)	Mr/Ms	(BH)max (MGOe)
S18–S20	BaCO3 + Fe2O3 (287 mg PVA, 150 mg PEG)	3857 ± 14	33.10 ± 0.32	67.21 ± 0.54	0.492	0.215 ± 0.008
S21–S23	BaCO3 + Fe2O3 (143.5 mg PVA, 75 mg PEG)	3854 ± 30	33.63 ± 0.54	68.34 ± 1.23	0.492	0.210 ± 0.007
S24–S26	BaCO3 + Fe2O3:Fe2O3 (99:1) (574 mg PVA, 300 mg PEG)	3942 ± 13	34.85 ± 0.31	70.85 ± 0.40	0.492	0.202 ± 0.004
S27–S29	BaCO3 + Fe2O3:Fe2O3 (95:5) (574 mg PVA, 300 mg PEG)	3945 ± 10	33.59 ± 0.59	68.57 ± 1.01	0.490	0.182 ± 0.005

of approximately 0.5 correspond to isotropic magnets with the assembly of non-interacting particles [27]. A value higher than 0.5 also indicates strong exchange coupling between the hard and soft phases [22]. Different amounts of the PVA binder and PEG-400 plasticizer in the paste insignificantly influence the magnetic properties of samples S18–S20 and S21–S23. Table 2 shows the average coercivity, remanent magnetization, and maximum energy product of approximately 3850 Oe, 33 emu/g, and 0.2 MGOe, respectively. By adding Fe₂O₃ in the synthesis, the saturation magnetizations and coercivities of samples S26 and S29 in Figure 5b are slightly higher than those in Figure 5a. However, the maximum energy products are reduced. Since all the samples are composed of identical phases, different magnetic properties are attributable to the particle clustering and voids of magnets from the paste-injection molding. Sample S27–29, with the lowest density of 2.11 g/cm³, has the lowest maximum energy product of 0.182 MGOe; however, it has the highest coercivity of 3945 Oe.

Compared to the BaFe₁₂O₁₉ from the paste-injection molding [17], the inclusion of the soft phase in this study does not reduce the coercivity and M_r/M_s ratio of the hard phase. This finding is similar to other hard ferrite/soft metal core–shell structures [36,37]. Although the hard magnetic properties are superior to some BaFe₁₂O₁₉/CoFe₂O₄ and BaFe₁₂O₁₉/Fe₃O₄ composites [28,29,38], the coercivities are still less than 4 kOe; and the magnetizations are not substantially raised by the Fe addition. It follows that the maximum energy products in this study are much lower than the value obtained from the paste extrusion of BaFe₁₂O₁₉ [15].

4. Conclusions

Facile paste-injection molding produced hard/soft magnetic composites with reproducible density and magnetic properties. The paste was prepared from BaCO₃ and Fe₂O₃ with a 287 mg PVA binder, 150 mg PEG-400 plasticizer, 40 mg EFKA^® FA4620 dispersing agent, and 4.6 mL DI water. After sintering at 1150 °C for 5 h, the low-density magnets were composed of BaFe₁₂O₁₉ and Fe phases. The reduction in PVA and PEG-400 by half did not significantly alter the saturation magnetization over 67 emu/g and the coercivity of around 3850 Oe. Interestingly, the coercivity was increased to 3942 and 3945 Oe by adding 1 and 5% Fe₂O₃ to the ceramic–polymer paste. Using additional Fe₂O₃, the saturation magnetization was slightly increased; however, the maximum energy product was reduced. Although the maximum energy product was still lower than common permanent magnets, the saturation magnetization and coercivity were comparable to those of other hard/soft magnetic composites; moreover, the finding regarding the paste formulation is useful for future development.