Synthesis of α-Fe2O3 and Fe-Mn Oxide Foams with Highly Tunable Magnetic Properties by the Replication Method from Polyurethane Templates

Open cell foams consisting of Fe and Fe-Mn oxides are prepared from metallic Fe and Mn powder precursors by the replication method using porous polyurethane (PU) templates. First, reticulated PU templates are coated by slurry impregnation. The templates are then thermally removed at 260 °C and the debinded powders are sintered at 1000 °C under N2 atmosphere. The morphology, structure, and magnetic properties are studied by scanning electron microscopy, X-ray diffraction and vibrating sample magnetometry, respectively. The obtained Fe and Fe-Mn oxide foams possess both high surface area and homogeneous open-cell structure. Hematite (α-Fe2O3) foams are obtained from the metallic iron slurry independently of the N2 flow. In contrast, the microstructure of the FeMn-based oxide foams can be tailored by adjusting the N2 flow. While the main phases for a N2 flow rate of 180 L/h are α-Fe2O3 and FeMnO3, the predominant phase for high N2 flow rates (e.g., 650 L/h) is Fe2MnO4. Accordingly, a linear magnetization versus field behavior is observed for the hematite foams, while clear hysteresis loops are obtained for the Fe2MnO4 foams. Actually, the saturation magnetization of the foams containing Mn increases from 5 emu/g to 52 emu/g when the N2 flow rate (i.e., the amount of Fe2MnO4) is increased. The obtained foams are appealing for a wide range of applications, such as electromagnetic absorbers, catalysts supports, thermal and acoustic insulation systems or wirelessly magnetically-guided porous objects in fluids.


Introduction
Reticulated sponges made of polyurethane (PU) have been commercially available since the 1950s [1]. The first attempts to transfer porous templates into ceramic (i.e., oxide) foams by the powder slurry replication method date from early 1960s [2]. At present, the method has become widely available for many applications involving porous oxide materials: electromagnetic wave absorbers, gas sensors, catalysts, oil-water separators and lithium-ion batteries [3][4][5][6][7][8]. As a consequence, the polymeric sponge replication process has consolidated as a promising technique to create cellular oxide structures with 3D interconnected pores, characterized by high strength and high corrosion resistance in acid and alkaline media [9][10][11]. In the 1970s, metallic foams also started to be prepared by this method and they found applications such as battery electrodes, catalysts or filters [12][13][14]. Over the past ten years, the replication method has been extended to manufacture porous steels [15][16][17][18] and porous Cu-based [19] and Ti-based alloys [20,21].
While many studies have focused on metallic foams [15][16][17][18][19][20][21], fewer efforts have been devoted to synthesizing inorganic oxide foams through the polymer sponge replication process [22]. So far, to the best of our knowledge, the possibility to use this technique to produce oxide foams with magnetic properties has not been explored.
Manganese ferrite (MnFe 2 O 4 ) is a well-known ceramic compound with electrically insulating and soft ferrimagnetic properties at room temperature. It has been widely used in the electronics industry to fabricate magnetic cores for read/write heads for high-speed tape or hard disk recording [23,24]. More recently, MnFe 2 O 4 has emerged as a promising material to be used as a catalyst [25,26], in hydrogen production technologies [24] or for oil-water separation [7]. Diverse preparation techniques, such as high energy ball milling [24,27], co-precipitation [28,29] or sol-gel routes [30], have been used to obtain MnFe 2 O 4 powders and nanoparticles. However, the preparation of porous MnFe 2 O 4 remains challenging and rather elusive.
Meanwhile, magnetic foams made of Fe 2 O 3 , Fe 3 O 4 , Co or Ni are appealing since they combine the intrinsic properties of magnetic materials with the aforementioned advantages of the porous structures, constituting magnetic materials with high surface area, low density and high strength-to-weight ratio. For instance, ultralight Fe 2 O 3 /C foams produced using polyelectrolyte-grafted PU sponges [7] exhibit one of the highest oil-absorption capacities among the reported counterparts. Magnetic Fe 3 O 4 nanoparticles/PU composites produced by in-situ blending methods have also been reported as good candidates for wastewater treatments, acting as carriers for immobilized microorganisms [30].
Therefore, the aim of the present work is to use the replication method to produce Fe and Fe-Mn oxides open-cell porous structures. The magnetic behavior of the foams is tuned, first, by the composition of the metallic slurry precursor and, secondly, by adjusting the N 2 flow. The influence of the N 2 flow and the Mn content, as well as the synergistic effect of porosity and magnetism are investigated.

Materials and Methods
Commercial Fe (97%) and Mn (99%) powders were used to coat a reticulated PU sponge by the impregnation method. First, the powders were mixed and mechanically milled in a planetary mill device (Fritsch Pulverisette 5, Fritsch, Idar-Oberstein, Germany) to reduce the powder size and to obtain the targeted composition, Fe or Fe-30Mn (Fe:Mn ratio of 70:30 nominal wt %). The raw powders were milled under Ar atmosphere with a ball-to-powder weight ratio of 10:1 at 300 rpm for 15 h. The particle size of the as-milled powders, calculated from a Scanning Electron Microscopy (SEM) image using the ImageJ software, was 10.8 ± 1.4 mm. To prepare the slurry, the milled powders were mixed with poly(ethylene glycol) (PEG) and distilled water. PEG acted as a binder and was used to control the slurry viscosity and to favor the adhesion of the powder particles to the sponge before sintering. The composition of the slurry is listed in Table 1. Subsequently, 1 cm 3 of a commercial reticulated PU sponge, acting as organic template, was immersed into the slurry for 5 min to allow complete impregnation. The sponge was removed from the suspension and was squeezed to ensure that only a thin layer of slurry covered the skeleton of the PU template without blocking the pores. Then, the impregnated template was dried at room temperature for at least 12 h before sintering. Finally, the template was thermally removed at 260 • C and the debinded powders were sintered in a tubular furnace (Carbolite MTF 9/15, Parsons Lane, UK) at 1000 • C for 2 h under N 2 flow ranging from 180 to 650 L/h. SEM observations were done on a Zeiss Merlin microscope (Carl Zeiss Microscopy, Jena, Germany) equipped with an energy dispersive X-ray (EDX) spectroscopy detector (Oxford Instruments, Abingdon, UK) for compositional analyses. X-ray diffraction (XRD) was carried out on a Philips X'Pert diffractometer (Philips, Amsterdam, The Netherlands) using Cu K α radiation. The measurements were performed in the angular range 2θ = 30-100 • with a step size of 0.026 • . Furthermore, Rietveld refinement of the XRD patterns using the X'Pert HighScore Plus software (PANalytical, Almelo, The Netherlands) was carried out to determine the cell parameters and the percentage of the constituent phases.
Hysteresis loops were recorded at room temperature using a vibrating sample magnetometer (VSM) from MicroSense (LOT-QuantumDesign, Darmstadt, Germany), with a maximum applied magnetic field of 20 kOe.  (Figure 1b-d) containing slurries. At a first glance, a rather homogeneous pore distribution can be observed, with pore sizes of the order of 400 µm. Fully-compact pore walls were observed in the foams produced from the Fe-and FeMn-containing slurries obtained at the lowest N 2 flow (Figure 1a,b); however, at larger N 2 flows, the pore walls exhibit a nanoporous morphology although the distribution and the size of the big pores was not compromised. The composition of the foams obtained at different nitrogen flow rates, measured by energy-dispersive X-ray (EDX) analyses, is listed in Table 2  To shed light on the microstructure of the open-cell foams X-ray diffraction was carried out ( Figure 2). As suggested by EDX analyses, the foams produced from the Fe-containing slurry consist of α-Fe2O3. In contrast, the XRD of the foams produced from the FeMn-containing slurry indicate a mixture of α-Fe2O3, FeMnO3 and Fe2MnO4 phases. Phase contributions and cell parameters estimated by Rietveld refinement are listed in Table 3. Sintering the FeMn-containing foams at higher N2 flow results in larger amounts of Fe2MnO4. As can be observed in Table 3, the amount of Fe2MnO4 increases from 8.5 to 74% when increasing the N2 flow from 180 to 650 L/h. Accordingly, the Fe2O3 phase percentage decreases from 60 to 26%. FeMnO3 only forms at a N2 flow of 180 L/h.   Table 3. Sintering the FeMn-containing foams at higher N 2 flow results in larger amounts of Fe 2 MnO 4 . As can be observed in Table 3, the amount of Fe 2 MnO 4 increases from 8.5 to 74% when increasing the N 2 flow from 180 to 650 L/h. Accordingly, the Fe 2 O 3 phase percentage decreases from 60 to 26%. FeMnO 3 only forms at a N 2 flow of 180 L/h. To shed light on the microstructure of the open-cell foams X-ray diffraction was carried out ( Figure 2). As suggested by EDX analyses, the foams produced from the Fe-containing slurry consist of α-Fe2O3. In contrast, the XRD of the foams produced from the FeMn-containing slurry indicate a mixture of α-Fe2O3, FeMnO3 and Fe2MnO4 phases. Phase contributions and cell parameters estimated by Rietveld refinement are listed in Table 3. Sintering the FeMn-containing foams at higher N2 flow results in larger amounts of Fe2MnO4. As can be observed in Table 3, the amount of Fe2MnO4 increases from 8.5 to 74% when increasing the N2 flow from 180 to 650 L/h. Accordingly, the Fe2O3 phase percentage decreases from 60 to 26%. FeMnO3 only forms at a N2 flow of 180 L/h.    Figure 3 shows the hysteresis loops of the open-cell porous foams produced from the Fe-and FeMn-containing slurries at different nitrogen flow rates. A linear magnetization versus field behavior (reaching 0.6 emu/g for an applied field of 20 kOe) was observed in the Fe-O foams, in agreement with the reported behavior for antiferromagnetic α-Fe 2 O 3 . Clear hysteresis loops are observed in the FeMn-containing foams as a result of the ferrimagnetic character of the Fe 2 MnO 4 phase, which exhibits a theoretical saturation magnetization, M s , of 77 emu/g at room temperature [31]. The FeMnO 3 phase is known to exhibit a weakly ferrimagnetic response with M s~0 .23 emu/g [32]. Accordingly, the saturation magnetization of the FeMn-containing foams ranges from 5 emu/g to 52 emu/g depending on the N 2 flow rate, i.e., the relative amount of FeMnO 3 Figure 3 shows the hysteresis loops of the open-cell porous foams produced from the Fe-and FeMn-containing slurries at different nitrogen flow rates. A linear magnetization versus field behavior (reaching 0.6 emu/g for an applied field of 20 kOe) was observed in the Fe-O foams, in agreement with the reported behavior for antiferromagnetic α-Fe2O3. Clear hysteresis loops are observed in the FeMn-containing foams as a result of the ferrimagnetic character of the Fe2MnO4 phase, which exhibits a theoretical saturation magnetization, Ms, of 77 emu/g at room temperature [31]. The FeMnO3 phase is known to exhibit a weakly ferrimagnetic response with Ms ~ 0.23 emu/g [32]. Accordingly, the saturation magnetization of the FeMn-containing foams ranges from 5 emu/g to 52 emu/g depending on the N2 flow rate, i.e., the relative amount of FeMnO3 and Fe2MnO4 phases. For instance, an Ms value of 37.7 emu/g would be expected for the foam produced under a N2 flow of 350 L/h as it contains 49% of Fe2MnO4. The observed Ms in this case (40 emu/g) is in quite good agreement with the expected value. The foam produced under a N2 flow of 650 L/h has Ms = 52 emu/g which is also in good agreement with the theoretically calculated value (57 emu/g).

Conclusions
Highly porous foams with homogeneous open-cell structure have been obtained by the replication process using polyurethane templates. The foams produced from the Fe-containing slurry consist of α-Fe 2 O 3 , independently of the N 2 flow. In contrast, the phase composition of the foams produced from the FeMn-containing slurry can be tuned by adjusting the N 2 flow. The main phases of the Fe-Mn-O foams sintered under a N 2 flow of 180 L/h are α-Fe 2 O 3 and FeMnO 3 with minor content of Fe 2 MnO 4 . Increasing the N 2 flow to 350 L/h results in a mixture of α-Fe 2 O 3 and Fe 2 MnO 4 . The amount of manganese ferrite is further increased at a flow rate of 650 L/h. Hence, addition of Mn as well as the adjustment of the N 2 flow allows tailoring the magnetic response of the foams from practically non-magnetic to soft ferrimagnetic. The magnetic behavior reported in this work for the Fe30Mn foams obtained at a N 2 flow rate of 350 and 650 L/h could be of interest for a wide range of applications, such as electromagnetic absorbers, catalysts supports, thermal and acoustic insulation systems or wirelessly magnetically-guided porous objects in fluids.