Oxidation-controlled magnetism and Verwey transition in Fe/Fe3O4 lamellae

rw gu g P rials, Adv d Faculty of Applied Sciences, Ton Duc Thang University e Department of Physics, University of South Florida, Ta f Institute of Experimental Physics, Slovak Academy of S al University, Hanoi. rg/licenses/by/4.0/). den crystal lattice bic inverse spinel n accompanied by erties. It has been ear 125 K occurs stoichiometry [9]. ons from the stoi- It is not observ- To the best of our knowledge, only Yamamuro et al. reported on the structural and magnetic properties for Fe/Fe3O4 lamellae-like structures, but not on their Verwey transition [4]. Furthermore, the temperature dependence of magnetization in these materials was not fully understood. Authors [13e15] reported that the Bloch law (Ms(T)z T3/2) could be applied well for magnetic nanoparticles * Corresponding author. Institute of Materials Science, Vietnam Academy of Science and Technology, 18 Hoang Quoc Viet, Ha noi, Viet Nam. ** Corresponding author. E-mail addresses: manhdh.ims@gmail.com (H.M. Do), skorvi@saske.sk (I. Skorvanek). Contents lists available at ScienceDirect Journal of Science: Advance journal homepage: www.el Journal of Science: Advanced Materials and Devices 5 (2020) 263e269Peer review under responsibility of Vietnam National University, Hanoi.compounds containing Fe3O4 is the presence of Verwey phase able in Fe3O4 nanoparticles with average size below 6 nm [10e12].through controlled oxidation. © 2020 The Authors. Publishing services by Elsevier B.V. on behalf of Vietnam Nation This is an open access article under the CC BY license ( 1. Introduction Heterogeneous magnetic systems composed of ferromagnetic and ferrimagnetic phases, such as Fe/Fe3O4, have shown interesting magnetic properties (exchange coupling, exchange bias, etc.) [1e6] as well as their potential applications (enhanced specific absorp- tion, microwave absorption, efficient removal of dyes from waste- water, etc.) [3,7,8]. One of the interesting phenomena related to transition (TV), which is widely known as a sud change from a monoclinic structure to the cu structure [9e11]. This structural change is ofte other anomalies in electrical and magnetic prop reported that the first-order phase transition n only in bulk and single-crystal Fe3O4 of perfect The TV tends to be suppressed because deviati chiometry due to lattice defects, surface effect Verwey phase transition Modified Bloch law temperature dependence of high-field magnetization of the lamellae is analyzed by the modified Bloch law. Our study demonstrates the possibility of tuning the magnetism in iron/iron oxide nanosystemsa r t i c l e i n f o Article history: Received 21 February 2020 Received in revised form 24 March 2020 Accepted 3 April 2020 Available online 10 April 2020 Keywords: Lamellar structure Fe/Fe3O4 nanocomposites Oxidation-controlled magnetismhttps://doi.org/10.1016/j.jsamd.2020.04.001 2468-2179/© 2020 The Authors. Publishing services b ( Ho Chi Minh City, Viet Nam mpa, FL 33620, USA ciences, Watsonova 47, 040 01 Kosice, Slovakia a b s t r a c t The structural and magnetic properties of Fe/Fe3O4 nanocomposites, synthesized by combined high energy ball milling and controlled oxidation, have been studied. An X-ray diffraction analysis of the crystal structure of the nanocomposites confirmed the coexistence of Fe and Fe3O4 phases. An increase of the oxygen concentration during oxidation process led to the formation of a higher fraction of the Fe3O4 phase with good crystallinity and stoichiometry. The morphology of the nanocomposites revealed a lamella-like structure with a thickness of about 30 nm. The saturation magnetization decreased when the phase fraction of Fe3O4 increased. The coercivity was enhanced at low temperatures (
100 K) but decreased at high temperatures, due to thermal fluctuation effects on the anisotropy in the Fe3O4 phase. Interestingly, the lamellae exhibited a sharp Verwey transition near 120 K, which is often suppressed or absent in nanostructured Fe3O4 due to the poorly crystalline, off-stoichiometric characteristic. Theb Graduate University of Science and Technolog c Laboratory of Magnetism and Magnetic Mate anced Institute of Materials Science, Ton Duc Thang University, Ho Chi Minh City, Viet Nama Institute of Materials Science, Vietnam Academy of Science and Technology, 18 Hoang Quoc Viet, Ha Noi, Viet Nam y, Vietnam Academy of Science and Technology, 18 Hoang Quoc Viet, Hanoi, Viet NamOriginal Article Oxidation-controlled magnetism and Ve lamellae Hung Manh Do a, b, *, Thi Hong Le a, b, Xuan Phuc N Thi Hong Ngo a, Trung Hieu Nguyen a, Thanh Phon Jozef Kov
ac f, Ivan Skorvanek f, **y Elsevier B.V. on behalf of Vietnamey transition in Fe/Fe3O4 yen a, Hong Nam Pham a, ham c, d, Manh Huong Phan e, d Materials and Devices sevier .com/locate/ jsamdNational University, Hanoi. This is an open access article under the CC BY license (aexp)) [21,22]. Extra peaks of g-Fe2O3 were not seen in our XRD patterns. Besides, the aexp value was estimated to be ~8.396 Å for both M10 and M30 (see Table 1), which is equivalent to that of bulk magnetite (8.396 Å) [21]. Moreover, the positions and the relative intensity ratios of the diffraction peaks of the present samples match well those of the published results for magnetite [25]. Fig. 1. XRD patterns of the Fe/Fe3O4 samples. The experimental and calculated patterns are dots and full line, respectively. Table 1 Average crystalline size (D), experimental lattice constant (aexp) and weight fraction of Fe and Fe3O4 phases. Sample D (nm) aexp (Å) % weight Fe Fe3O4 Fe Fe3O4 Fe Fe3O4 M0 25.5 e 2.863 100 e M10 62.4 67.5 2.866 8.396 62 38 ced Materials and Devices 5 (2020) 263e269with high TC up to room temperature, whereas authors [16e18] noticed some deviation from Bloch law in such nanoparticles. In this paper, we present the results of a detailed study of the structural and magnetic properties of Fe/Fe3O4 lamellae prepared by high-energy ball milling and subsequently controlled oxidation. Following the structural and morphological characterizations, we have investigated effects of varying Fe3O4 phase fraction on the magnetic properties of the Fe/Fe3O4 lamellae. The occurrence of a sharp Verwey transition at ~120 K as well as the temperature dependence of high-field magnetization are analysed and discussed. 2. Experimental 2.1. Synthesis of Fe3O4 lamellae Commercial Fe powder with particle size less than 100 mm (Merck) was milled in a planetary ball mill (Fritsch P 6) in air using hardened steel containers and balls. The milling parameters were used as follows: A ball-to-powder weight ratio of 15:1, a milling speed of 450 rpm, and a milling time of 2 h (see Ref. [19] for more details). The as-milled powder was used for controlled oxidation under (oxygen and nitrogen) gas flow with oxygen of 10 and 30% volume at 773 K (500 C) for 1 h and then cooled down to room temperature. The gas sources of oxygen and nitrogenwere obtained from Singapore Oxygen Air Liquid Pte Ltd and the desired con- centrations were controlled by the flow-through mixing through mass-flow-controllers (Aalborg, GFC17) [20]. We labeled M0 for Fe milled for 2 h, and M10 and M30 for the subsequently oxidized samples corresponding to the used oxygen concentrations. 2.2. Characterization of the samples The structural characterization was performed by SIEMENS D5005 diffractometer using Cu-Ka radiation at l ¼ 0.154 nm. The diffraction patterns were collected with 2q in the range of 20oe70. A careful refinement for X ray diffraction (XRD) patterns for sam- ples have done before evaluating the average crystal size, lattice constant, and the proportion of the Fe and the Fe3O4 phases by using a commercial X'Pert HighScore Plus software [15]. The morphology was studied using scanning electron microscopy (SEM) and bright field transmission electron microscopy (BFTEM) techniques with a Hitachi S-4800 equipment. The temperature and field dependence of magnetization was measured using the superconducting quantum interference device (SQUID) magne- tometer MPMS-XL-5 (Quantum Design) in the temperature range of 5e300 K and in magnetic fields of up to 50 kOe. 3. Results and discussion 3.1. Structural and morphological analysis The XRD patterns of the samples are shown in Fig. 1. Two peaks at the 2q values of 44.7 and 65.0 were indexed to body-centered cubic (bcc) Fe. We did not see any signature of secondary phases indicating that sample M0 was single phase. All new peaks of samples M10 and M30 were indexed to magnetite (Fe3O4) and maghemite (g-Fe2O3) patterns. They were strong and sharp, indi- cating that the both samples are well crystallized. It is generally known that the basic structure of geFe2O3 is close to that of Fe3O4, resulting in appearance of the XRD peaks at the similar positions [21,22]. There are several approaches that have been used to distinguish these phases such as neutron diffraction [23] and XRD methods (extra peaks due to (321) and (221) lattice planes corre- H.M. Do et al. / Journal of Science: Advan264spond to the g-Fe2O3 [24], or the experimental lattice constants M30 45.5 52.4 2.868 8.396 57.8 42.2 Therefore, we can assume that the new peaks belong to the major crystalline phase of Fe3O4 with good crystallinity and stoichiom- etry. This hypothesis has been further supported by the M(T) data, as shown in Fig. 5, in which a sharp change in the magnetization near the TV ~ 120 K appears to observe. Values of average crystallite size D were evaluated from all peaks of Fe3O4 and Fe and are summarized in Table 1. It can be seen in Table 1 that D of the Fe phase in M10 and M30 increased about two times as compared to M0 due to the annealing at 773 K for 1 h. The estimated fractions of Fe and Fe3O4 are ~62% and ~38% for M10, and ~57.8% and ~42.2% for M30, respectively. This finding also pointed out that the presence of the higher oxygen content helped accelerate oxidation of Fe, leading to an increasing growth rate of the Fe3O4 phase. This observation is consistent with reports of Chin et al., in which the growth rate of Fe3O4 nanosheets was enhanced by flowing O2 gas into the solution during oxidation [26]. The effect of the Fe3O4 fraction on the magnetic properties of the Fe/Fe3O4 nanocomposites will be discussed below. Fig. 2 shows the representative SEM and BFTEM images of M10 andM30. These images revealed a high density of lamella structure. The edges of the lamellae are irregular and the lengths of the lamellae are in the range of a micrometer, with thicknesses of about 30 nm. Since the lamellae are relatively thin (in the nanometer range), they are almost transparent under high-energy electron beam (see Fig. 2b and d). It was also found that increasing the amount of oxygen in the oxidation process did not change the morphology of the samples. It can be noted that completely different morphologies of the samples indicated in our previous work [19] and that of this work are due to the different fabrication 3.2. Magnetic properties Magnetization versus field (MH) measurements were per- formed on all samples at various temperatures from 5 to 300 K and shown in Fig. 3. The M(H) curves shown in Fig. 3 are smooth and shoulderless, irrespective of the presence of two magnetic phases (Fe and Fe3O4). The existence of shoulders in the M(H) curve in a dual-phase magnetic system is often derived from the decoupling of cooperative magnetic moments between two components’ magnetic phases [4,27]. The presently observed smooth and shoulderless M(H) curves would thus be a signature of relatively good exchange-coupling between two magnetic phases (Fe and Fe3O4) in our samples. The magnetic parameters of the samples at 300 K were deduced from this figure and summarized in Table 2. As can be seen from this table, MS decreased from 220 emu/g for M0 to 177 emu/g and 172 emu/g for M10 and M30. The results are fully consistent with increasing Fe3O4 content due to the controlled oxidation. It should be noted that theMS of bulk Fe3O4 at room temperature (92 emu/g) [28] is lower than that of Fe (220 emu/g) (see Table 2). Therefore, the larger values ofMS obtained for M10 (177 emu/g) and M30 (172 emu/g) relative to bulk Fe3O4 also affirm the presence of both Fe and Fe3O4 phases in these samples, which is consistent with the XRD analysis (see Fig. 1 and Table 1). Fig. 4 shows a comparison of the coercivity as a function of temperature for the three samples. The coercivity at low temper- atures (
100 K) of M10 and M30 has increased three times as compared to M0. For these samples, we assume that a thin layer of Fe3O4 could be formed on to surface of Fe due to oxidation (a H.M. Do et al. / Journal of Science: Advanced Materials and Devices 5 (2020) 263e269 265conditions. The first is the morphology of the 10 h as-milled Fe, while the second is that of the Fe/Fe3O4 lamellae (Fe powder was only milled for 2 h and then annealed at 500 C for 1 h). Besides, the size of lamellae is larger than their crystal size because a lamellar contains many different crystals.Fig. 2. SEM and BFTEM morphologies ofthickness of ~30 nm for the Fe/Fe3O4 lamellae as shown in Fig. 2). The interfacial coupling between Fe and Fe3O4 could be a main driving force for the enhanced coercivity below 100 K. The authors [6] stated that Fe3O4 shell shows typically a soft magnetic behavior. This is different from the g-Fe2O3 shell which exhibits strong spinsamples M10 (a, c) and M30 (b, d). ced Materials and Devices 5 (2020) 263e269H.M. Do et al. / Journal of Science: Advan266disorder and hence an enhanced magnetic anisotropy [29]. How- ever, at room temperature, the coercivity decreased and began to converge (13 Oe, 20 Oe and 21 Oe for M0, M10 and M30, respec- tively) due to the thermal fluctuations in the shell overcoming the anisotropy of Fe3O4 [27]. The phase fraction of Fe and Fe3O4 in M10 and M30 could be determined by comparing the room temperature MS values, normalized to the total mass of the sample, using the following equation: Msðaþ bÞ¼ aMFes þ bMFe3O4s (1) Fig. 3. M(H) hysteresis loops of M0, M10 and M30 samples taken at indicated tem- peratures. The insets show a magnified low-field region of the M(H) loops.where a and b are % wt. of Fe and Fe3O4 phases and (a þ b) ¼ 100%, MFes andM Fe3O4 s are the saturation magnetization of Fe and Fe3O4. From the values of MS at 300 K, the weight fractions of the Fe3O4 phase was determined to be 32.3% wt. and 36.9% wt. (see Table 2). These results are relatively consistent with that derived from the XRD data. Fig. 4. Temperature dependent coercivity for samples M10 and M30. along their respective easy axes, when the magnetocrystalline anisotropy of Fe3O4 becomes dominant at low field, derived from the ferrimagnetic shell region. The temperature dependence of magnetization in heteroge- neous magnetic systems, including core/shell nanoparticles of Fe3O4/g-Fe2O3 and Fe3O4/a-Fe2O3 [31,32], has been reported, indicating the important roles of surface and interface spins. As discussed in the previous section, the magnetic phases in M10 and M30 are expected to be exchange coupled. Therefore, the change of high field magnetization in the temperature range of 5e300 K for these samples was also investigated. The tempera- ture dependence of magnetization well below the Curie tem- perature for most ferromagnetic materials is well described by Bloch law [14]: MSðTÞ¼MSð0Þ h 1BT3=2 i ðfor T << TCÞ (2) ced Materials and Devices 5 (2020) 263e269 267H.M. Do et al. / Journal of Science: AdvanTemperature dependence of field-cooled magnetization, MFC(T) in an applied field of 100 Oe is shown in Fig. 5 for M10 andM30. The MFC(T) curves show a sharp change of magnetization at ~120 K that coincides well with TV of Fe3O4 [9,30]. So far, it is well known that TV disappears due to loss of crystal quality (imperfect stoichiometry or/and poor crystallinity) in bulk, single crystals of Fe3O4 [9e11]. This is the first time that the Verwey transition has been observed in a Fe/Fe3O4 nanocomposite system with a lamellar structure. It also implies that the Fe3O4 phase has a good crystallinity and stoichiometry, which is consistent with both the XRD and SEM analyses. M30 sample with the higher phase fraction of Fe3O4 exhibits a sharper change in magnetization at ~120 K than that of M10 (see Fig. 5). We can define “jump” at low field DM ¼ Mtop e Mbottom at 120 K from the M(T) curve. The DM values are 0.8 emu/g and 1.2 emu/g coresponding to 14% and 23% of the overall magnetization for FC conditions of M10 and M30. Ong et al. [6] believed that the low-field jumps are due to a sudden reorientation of moments where B is a constant, which is related to the exchange integral Je (B ~ 1/Je 3/2) and MS(0) is the saturation magnetization at zero temperature. However, theoretical calculations and experimental results [16e18] have reported that the temperature dependence of magnetization for small ferromagnetic clusters and nanoparticles is given by a modified Bloch law: MSðTÞ¼MSð0Þ½1BTa (3) Fig. 5. Magnetization changes as a function of temperature under field-cooled con- dition for samples M10 and M30. Field applied during the measurement was 100 Oe. Table 2 The saturation magnetization (MS) and coercivity (HC) of the samples at 300 K. The fraction of Fe and Fe3O4 phases calculated from Eq. (1) in the text. Sample MS (emu/g) HC (Oe) % wt. Fe Fe3O4 M0 220 13 100 e M10 177 20 67.7 32.3 M30 172 21 63.1 36.9Fig. 6. Temperature dependence of saturation magnetization for M10 and M30. Solid curves are the best-fit curves to the experiment data with error bars of near 1% using Eq. (3) described in the text. ced MSince the magnetization tends to saturate at ~50 kOe, we can consider that MS(T) ¼ M(T, H ¼ 50 kOe). Then, the temperature dependence of MS data was fitted by using Eq. (3). The best-fit curves were obtained, as presented in Fig. 6 (solid curve). It points out that MS(T) is indeed well described by a Ta law in the whole experimental temperature range (5 K < T < 300 K). The a values were 1.74 and 1.72 for M10 and M30, respectively. These results are different from those for single-phase ferromagnetic nanoparticle systems with high TC [14,15] in which the Bloch law could be applied well up to room temperature. However, authors [31] also reported some deviation from the original Bloch's law for Fe3O4/g-Fe2O3 ferrimagnetic/ferrimagnetic core/shell nano- particles, where the surface and interface have contributed to the magnetization effectively. The deviation from the Bloch T3/2 law of MS(T) lamellae is smaller for the Fe/Fe3O4 ferromagnetic/ferrimagnetic lamellae than for the Fe3O4/g-Fe2O3 core/shell nanoparticles. This result is reasonable because the lamellae have a magnetic volume larger than that of Fe3O4/g-Fe2O3 core/shell nanoparticles leading to the smaller contributions of surface and interface to MS and its tem- perature dependence. 4. Conclusion In summary, the Fe/Fe3O4 lamellae with a thickness of about 30 nm have been successfully synthesized by high energy ball milling and subsequently oxidized via controlling oxygen con- centration at 500 C for 1 h. The Fe3O4 phase in this structure has good crystallinity and stoichiometry and its content in- creases with oxygen concentration during oxidation. The magnetic properties of the Fe/Fe3O4 lamellae were systemati- cally studied. The Verwey transition in these Fe/Fe3O4 lamellae was observed for the first time. The saturation magnetization decreased as the phase fraction of Fe3O4 increased, whereas the Verwey transition became sharper. The temperature depen- dence of the saturation magnetization in the Fe/Fe3O4 lamellae could be predicted by the modified Bloch law i