- Email: [email protected]
Magnetic iron nanoparticles prepared by solution combustion synthesis and hydrogen reduction
Chemical Physics Letters 657 (2016) 33–38
Contents lists available at ScienceDirect
Chemical Physics Letters journal homepage: www.elsevier.com/locate/cplett
Research paper
Magnetic iron nanoparticles prepared by solution combustion synthesis and hydrogen reduction Min Huang a, Mingli Qin a,⇑, Zhiqin Cao a,b, Baorui Jia a, Pengqi Chen a, Haoyang Wu a, Xuanli Wang a, Qi Wan a, Xuanhui Qu a a b
Institute for Advanced Materials and Technology, University of Science and Technology Beijing, Beijing 100083, China School of Resources and Environmental Engineering, Pan Zhihua University, Pan Zhihua 617000, China
a r t i c l e
i n f o
Article history: Received 3 December 2015 In final form 19 May 2016 Available online 20 May 2016 Keywords: Magnetic iron nanoparticles Solution combustion synthesis Hydrogen reduction
a b s t r a c t A facile and efficient method has been proposed to prepare iron nanoparticles by combining solution combustion synthesis and hydrogen reduction for the first time. A porous a-Fe2O3 precursor with high specific surface area of 75 m2/g was fabricated by solution combustion synthesis, and then iron nanoparticles with high saturation magnetization of 196.3 emu/g were successfully obtained by hydrogen reduction of the as-synthesized precursor. With the reduction temperature rising from 275 °C to 600 °C, the saturation magnetization of the products increases from 196.3 emu/g to 209.7 emu/g, whilst the coercivity decreases from 611.4 Oe to 98.8 Oe. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction Magnetic nanoparticles have emerged as a promising advanced functional material for a wide field of potential applications such as catalysis, water treatment, environmental remediation, biotechnology/biomedicine, magnetic recording media, electromagnetic wave absorption, magnetic sensors [1–5]. As compared to other magnetic nanoparticles, iron nanoparticles have attracted a great deal of attention due to exceptionally high saturation magnetization, large permeability, superior environmental compatibility and high availability [6–10]. In recent years, multiple techniques have been developed to prepare iron nanoparticles, including thermal decomposition [11,12], sonochemical decomposition [13], vapor phase condensation [14], aqueous reduction of iron salt [15], high-energy milling [16], and hydrogen reduction of nanoscaled iron oxide [17]. Among the above methods, owing to process simplicity, low cost as well as potential high productivity, hydrogen reduction (HR) is utilized to fabricate iron nanoparticles on a commercial basis. Over the past few years, in order to better understand and ameliorate HR, several investigations suggest that the purity and reductive activity of starting materials have great influence on the reduction temperature and microstructure of the asprepared product [18,19]. Thus, the production of iron oxide with
⇑ Corresponding author. E-mail address: [email protected] (M. Qin). http://dx.doi.org/10.1016/j.cplett.2016.05.043 0009-2614/Ó 2016 Elsevier B.V. All rights reserved.
high purity and reactivity is the prerequisite task required for achieving the successful preparation of iron nanoparticles by HR. Solution combustion synthesis (SCS) has been demonstrated to be one of the most appropriate methods to produce oxide nanomaterials [20–23]. It is well known that SCS is essentially exothermal redox reaction between an oxidizer (e.g. metal nitrates) and a fuel (e.g. glycine, urea, citric acid, etc.) in a homogenous aqueous solution within several seconds. Simultaneously, the exothermal reaction provides the energy required for sustaining the combustion reaction without adding external energy. The advantages of SCS thus are as follows: (1) the homogeneous mixing of all the components on the molecular level and the evolution of a large volume of gases result in the formation of nanoscaled products with high specific surface area and resultant high reactive activity; (2) inexpensive starting materials, selfsustained instantaneous reaction features, high yield, as well as simple processing and apparatuses render SCS cost-effective and suitable for mass production. In view of these advantages, a large number of efforts have been made to prepare iron oxides by SCS [24–27]. Despite of such fascinating research results, to our knowledge, no research on hydrogen reduction of iron oxides by SCS has been reported. Hence, in this work, we report a facile and efficient method to prepare iron nanoparticles by combining SCS and HR for the first time. First, by using ferric nitrate as an oxidizer, glycine as a fuel and glucose as an assisted additive and selecting an optimum molar ratio, a porous iron oxide precursor has been fabricated via SCS. Subsequently, iron nanoparticles with high saturation magnetization have been successfully obtained by hydrogen reduction of
34
M. Huang et al. / Chemical Physics Letters 657 (2016) 33–38
the as-synthesized precursor. Moreover, the whole combustion process and hydrogen reduction process have been investigated, and the influence of reduction temperature on magnetic properties of the products has been discussed in detail.
2. Experimental 2.1. Synthesis procedure All the starting materials involving ferric nitrate (Fe(NO3)39H2O, oxidizer and Fe source), glycine (C2H5NO2, fuel) and glucose (C6H12O6H2O, addictive), were commercially purchased and of analytical reagent grade. After a series of optimization experiments, the optimal molar ratios of glycine to ferric nitrate and glucose to ferric nitrate were fixed at 2.8 and 1.5, respectively, and the amount of ferric nitrate was 0.1 M. Just like the typical experimental procedure for SCS, the above starting materials were dissolved in 200 ml deionized water under thoroughly stirring to obtain a rufous homogeneous starting solution at room temperature. Subsequently, the resultant solution was poured into a 2000 ml beaker and heated in a temperaturecontrolled muffle furnace in air. As heating continued, the solution evaporated and gradually developed into a viscous gel. Then, the gel began to swell and a smoldering combustion reaction occurred, coupled with the evolution of a large volume of gases. The whole process of swelling and combustion seemed to undergo a selfpropagating and nonexplosive exothermic reaction and lasted only a few minutes, resulting in the generation of a gray porous foam. Upon further heating, the foam was ignited and the carbon was completely removed, leading to the formation of a loose reddish precursor. Hereafter, hydrogen reduction of the precursor was conducted in a tube furnace. A strict temperature program was followed in all runs, with heating up to the plateau temperature, 150–600 °C, at a constant heating rate of 10 °C/min. The assynthesized precursors were reduced at various temperatures for 2 h in a flowing hydrogen atmosphere at a flow rate of 1 L/min. Finally, when cooled naturally to room temperature, the reduction products were obtained.
2.2. Characterization Thermal behavior of the formed gel sample was investigated by thermal gravimetric analysis (TG) and Differential Scanning Calorimetry (DSC) using a Netzsch 409PC thermal analyzer. The TG and DSC curves were recorded over the temperature range of 25– 600 °C at a heating rate of 10 °C/min in air at a constant flow rate of 20 ml/min. The phase composition of the foam, precursor and reduction products was determined by X-ray Diffraction (XRD), using a TTRAX III diffractometer with monochromic Cu Ka radiation. The average crystallite size was calculated according to the Scherrer’s equation. The presence of Fe3O4 was studied by X-ray photoelectron spectroscopy (XPS), using a ESCALAB 250 Xi spectrometer. The morphology and particle size of the precursor and reduction products were observed by scanning electron microscopy (SEM, JSM-6510) and field emission scanning electron microscopy (FE-SEM,ZEISS ULTRA 55), respectively. The reduction products were further observed and the associated selected-area electron diffraction images were captured by transmission electron microscopy (TEM, Tecnai G2 F30 S-TWIN).The specific surface area (SSA) of the precursor was examined by BET method using an Automated Surface Area & Pore Size Analyzer (QUADRASORB SI-MP, Quantachrome Instruments, Boynton Beach, FL). The magnetic properties were measured by vibrating samples magnetometry (VSM) at room temperature, using a Lake Shore 7307 magnetometer.
3. Results and discussion 3.1. Solution combustion synthesis As described in detail in the experimental section, under continuous heating conditions, the homogeneous starting solution containing ferric nitrate, glycine, and glucose gradually developed into a viscous gel, and the succeeding thermal behavior of the gel was investigated. Fig. 1 depicts the TG and DSC curves of the formed gel. It is clear that the whole process is composed of three major stages with the two dash lines as division marks. In the initial stage (below 136 °C, dehydration), the gel sample undergoes a rapid mass loss of around 40% on the TG curve. On the same temperature interval, a roughly V-shaped endothermic peak at 110 °C is presented on the DSC curve, which can be attributed to the evaporation of remaining water and the desorption of chemically absorbed water in the gel. For the second stage (136–330 °C, smoldering combustion), as compared to the previous stage, the mass decreases by approximately 30% at a relatively gentle rate, accompanied by a wide and weak exothermic peak. The characteristic profile of the DSC curve suggests the low exothermic effect, which is in good agreement with the observed slow smoldering combustion phenomenon. It is well known that, after the removal of the various types of water, the formed gel becomes far more reactive with further heating, and then is closely followed by the vigorous initiation of the combustion reaction. Actually, in this stage the exothermic combustion reaction between ferric nitrate (oxidizer) and glycine (fuel) is conducted [Eq. (1)]. Simultaneously, after absorbing the heat released from combustion reaction, the decomposition of glucose (addictive) occurs [Eq. (2)], resulting in the generation of a gray porous foam [Fig. 2(a)]. Thus, it is the superposition of the exothermic combustion and endothermic decomposition reaction that is responsible for the total low exothermicity and smoldering phenomenon. With respect to the last stage (above 330 °C, decarbonization), as the temperature continues to rise, the sharp exothermic peak, located at approximately 393 °C, corresponds to the complete removal of carbon in the foam [Eq. (3)], leading to the formation of a loose reddish precursor [Fig. 2(b)]. In addition, when the temperature surpasses 450 °C, the mass of the sample keeps almost constant.
6FeðNO3 Þ3 þ 10C2 H5 NO2 ! 3Fe2 O3 þ 20CO2 þ 25H2 O þ 14N2
ð1Þ
C6 H12 O6 H2 O ! 6C þ 6H2 O
ð2Þ
C þ O2 ! CO2
ð3Þ
Fig. 1. TG and DSC curves of the formed gel.
M. Huang et al. / Chemical Physics Letters 657 (2016) 33–38
35
Fig. 2. Original appearance photographs of the foam (a), precursor (b), products reduced at 200 °C (c) and 275 °C (d).
Fig. 3. XRD patterns of the foam and precursor (a), EDS analysis of the foam (b) and SEM image of the precursor (c).
The original appearance photograph of the foam is shown in Fig. 2(a). It can be clearly seen that the foam has exhibited a spongy porous appearance, which is originated from large quantities of gases released slowly from the combustion reaction and decomposition reaction [Eqs. (1) and (2)]. Fig. 3 displays the XRD patterns of the foam and precursor, the EDS analysis of the foam and the SEM image of the precursor. It is found that no obvious Bragg diffraction peaks have been detected in the foam, revealing that it possesses an amorphous structure [Fig. 3(a)]. The corresponding EDS results confirm that the foam consists of C, Fe, and O elements [Fig. 3(b)]. Based on the XRD pattern, the EDS analysis as well as the low exothermicity of the smoldering combustion, it is safe to conclude that the energy liberated from smoldering combustion is not sufficient for the crystallization of iron oxide, rendering the foam a mixture of amorphous carbon and amorphous iron oxide. Furthermore, the precursor is confirmed as a crystalline a-Fe2O3 phase [Fig. 3(a)], which explains the reddish color1 of the precursor [Fig. 2(b)]. This is ascribed to the fact that the higher energy evolved from the oxidation of the car1 For interpretation of color in Figs. 2 and 3, the reader is referred to the web version of this article.
bon promotes the transformation of iron oxide from amorphous state to crystalline state [Eq. (3)]. In addition, it is noticed that the precursor exhibits a porous morphology, which arises from the dispersive effect of gases liberated from Eqs. (1)–(3). The SSA value of the precursor is evaluated to be 75 m2/g, which is undoubtedly explained by the typical porous structure. The crystallite size of the precursor calculated by the Scherrer’s equation is 19 nm. 3.2. Hydrogen reduction To investigate the phase transformation during the reduction process and determine the lowest reduction temperature, the iron oxide precursors were reduced in flowing hydrogen atmosphere at various temperatures for 2 h. The XRD patterns of the products reduced at 150–275 °C are shown in Fig. 4. At 150 °C, no other obvious diffraction peaks are found, except the peaks belonging to a-Fe2O3 phase, which suggests that the temperature is too low to initiate the reduction reaction. The XRD pattern points out that the product reduced at 175 °C is a mixture of Fe3O4 and a-Fe2O3 phases with Fe3O4 as the main phase. This demonstrates that the reduction reaction has already occurred and the onset temperature has been below 175 °C. With the temperature rising to 200 °C, the
36
M. Huang et al. / Chemical Physics Letters 657 (2016) 33–38
Fig. 4. XRD patterns of the products reduced at 150–600 °C.
diffraction peaks from a-Fe2O3 disappear and only diffraction peaks of Fe3O4 exist, implying the formation of a single Fe3O4 intermediate product [Fig. 2(c)]. Subsequently, for the product reduced at 225 °C, the weak diffraction peaks from bcc iron have appeared, and then iron becomes the main phase at 250 °C. When the temperature goes up to 275 °C, it is evident that only diffraction peaks from metal iron are observed in the XRD pattern. This means that the hydrogen reduction of the iron oxide precursor has been completed, and thus the pure iron product is obtained [Fig. 2(d)]. It is worth noting that the reduction temperature is as low as 275 °C, which can be ascribed to the high SSA and resultant high reduction reactivity of the precursor. The low reduction temperature can effectively prevent the agglomeration and coarsening of iron particles. Meanwhile, it is also expected to improve production efficiency and reduce energy cost. The XRD pattern for the product reduced at 200 °C identifies the formation of Fe3O4 as a single intermediate phase. Since the positions of the diffraction peaks in the XRD patterns are nearly identical for Fe3O4 and c-Fe2O3, XPS is employed to illustrate the presence of Fe3O4. Fig. 5 displays the high-resolution XPS spectrum of Fe 2p of the product reduced at 200 °C. The Fe 2p spectrum possesses Fe 2p2/3 and Fe1/2 doublets, located at 710.6 eV and 724.1 eV, respectively, which are in excellent agreement with the literature data for the Fe3O4 standard sample [31]. Furthermore, it was
Fig. 5. High-resolution XPS spectrum of Fe 2p from the product reduced at 200 °C.
previously reported that between the Fe 2p2/3 and Fe1/2 peaks the absence of a shoulder-shaped satellite peak is a typical characteristic for Fe3O4 [31,32]. The absence of the satellite peak has been confirmed in this work. Based on the results of the XRD pattern, the XPS spectra, as well as the typical black color [Fig. 2(c)], it can be concluded that the product reduced at 200 °C is a single Fe3O4 intermediate phase. The microstructure of the product reduced at 275 °C is displayed in Fig. 6. It can be seen that the product consists of irregular sheets with 2 lm in width [Fig. 6(a)]. To acquire more information about the reduction product, the product is subject to TEM analysis. The product presents the two-dimensional sheet structure [Fig. 6(b)], which is in accord with the FESEM image. A closer observation reveals that the sheet is composed of irregular particles with the size of 100 nm [Fig. 6(c)]. The corresponding selected-area electron diffraction results in the inset of Fig. 6(b) further confirm the polycrystalline iron phase, fitting very well with the above XRD results. Besides, the crystallite size is calculated to be 21 nm by the Scherrer’s equation, slightly larger than that of the precursor, implying no obviously coarsening of the crystallite occurs during the reduction process. Such small size can be attributed to the multiple factors of the ultrafine structure of the precursor, the low reduction temperature and the strong dispersive effect of the water vapor evolved from the reduction reaction. 3.3. Magnetic properties The XRD patterns of the products reduced at 275–600 °C are given in Fig. 4. The XRD results illustrate that all the products have the typical bcc iron characteristic. As the reduction temperature rises, the diffraction peaks become sharper and peak intensity increases correspondingly. The average crystallite sizes evaluated by the Scherrer’s equation and SSA values determined by BET method of the products are listed in Table 1. It can be found that, with the rising of the reduction temperature, the crystallite sizes increase, whereas the SSA values decrease. The TEM image of the product reduced at 600 °C is presented in Fig. 6(d). It is noted that the obvious coarsening and aggregation of the iron particles have occurred. Fig. 7 depicts the room-temperature magnetic hysteresis loops of the products reduced at 275–600 °C. The saturation magnetizations are calculated by extrapolating to infinite field, and the main magnetic parameters are listed in Table1. Magnetic measurement results reveal that all the products present typical ferromagnetic characteristics. The iron particles reduced at 275 °C possess a rather high saturation magnetization value of 196.3 emu/g, which is close to 90% of that of the bulk counterpart (218 emu/g). With the reduction temperature rising to 600 °C, the saturation magnetization increases to 209.7 emu/g. It is well know that, for magnetic nanoscaled materials, the saturation magnetization depends strongly on the crystallinity and chemical composition [4,10]. Therefore, the high saturation magnetization of iron nanoparticles at 275 °C herein is attributed to the high crystallinity derived from the moderate temperature condition during the reduction process, whereas the varying saturation magnetization is ascribed to the oxidation of the iron particles. It can be noted from Table 1 that the coercivity of iron particles reduced at 275 °C reaches 611.4 Oe, and the coercivity decreases steeply to 98.8 Oe with the reduction temperature increasing to 600 °C. As reported in previous literatures, for a single-crystal iron particle, the critical size Dc for a single-domain structure increases with increasing the axial ratio n, and the smallest critical size is proximately 20 nm when the axial ratio n equals 1 [28]. By comparing the critical size (20 nm) with the calculated crystallite size (21 nm) of the iron nanoparticles reduced at 275 °C, it is deduced that the crystallite almost exists in a single-domain state. For a
37
M. Huang et al. / Chemical Physics Letters 657 (2016) 33–38
Fig. 6. FSEM image (a), TEM images (b, c) of the product reduced at 275 °C and TEM image (d) of the product reduced at 600 °C.
Table 1 Characteristics of the products reduced at 275–600 °C. Reduction temperature (°C)
Crystallite size (nm)
BET SSA (m2/g)
Mr (emu/g)
Ms (emu/g)
Hc (Oe)
275 400 500 600
21 ± 2 38 ± 3 55 ± 5 62 ± 4
18 ± 1.7 13 ± 1.3 7 ± 0.9 5 ± 0.5
85.3 ± 0.5 52.9 ± 0.3 25.5 ± 0.4 11.9 ± 0.5
196.3 ± 0.7 202.9 ± 0.4 206.5 ± 0.2 209.7 ± 0.5
611.4 ± 3.5 375.6 ± 2.8 174.6 ± 2.3 98.8 ± 1.7
Fig. 7. Room temperature magnetic hysteresis loops of the products reduced at 275–600 °C.
38
M. Huang et al. / Chemical Physics Letters 657 (2016) 33–38
system of single domains, due to the absence of domain walls, magnetization reversal occurs through the coherent rotation of domains, resulting in the high coercivity [29,30]. Meantime, the shape anisotropy arising from the irregular shape can give rise to the high coercivity [30]. Furthermore, as the reduction temperature goes up, the crystallite size becomes significantly larger than the critical size and the crystallite undergoes a transformation from a single-domain state to a multidomain state. Accordingly, Magnetization reversal no longer occurs through the domain rotation and instead commence through the nucleation and motion of the domain walls, rendering the low coercivity. Simultaneously, the decreasing crystallite boundary pinning of the motion of the domain wall can serve to explain the swiftly declining coercivity.
Acknowledgments This work was supported by the National Natural Science Foundation Program of China (51574031) and (51574029), the Natural Science Foundation Program of Beijing (2162027), the Fundamental Research Funds for the Central Universities (06109063) and the National 863 Program (2013AA031101). References [1] [2] [3] [4] [5] [6]
4. Conclusions A facile and efficient method has been proposed to prepare iron nanoparticles by combining solution combustion synthesis and hydrogen reduction. A porous a-Fe2O3 precursor with high specific surface area of 75 m2/g was fabricated by SCS, and then iron nanoparticles were successfully obtained by hydrogen reduction of the as-synthesized precursor. The whole combustion process consists of three major stages, i.e., dehydration, smoldering combustion and decarbonization. The precursor has completed the reduction reaction at the temperature as low as 275 °C. Iron nanoparticles reduced at 275 °C exhibit a high saturation magnetization value of 196.3 emu/g which is close to 90% of that of the bulk counterpart (218 emu/g). The crystallite sizes of the products reduced at 275–600 °C range from 21 nm to 62 nm. The reduction temperature exhibits a significant effect on the magnetic properties of reduction products. With reduction temperature rising from 275 °C to 600 °C, the saturation magnetization of the products increases from 196.3 emu/g to 209.7 emu/g, whilst the coercivity decreases from 611.4 Oe to 98.8 Oe. Finally, it should be emphasized that, due to integration of the common advantages of exceptional simplicity and high efficiency originated from SCS and HR, the method proposed herein will be a promising versatile technique for mass production of highly pure nanoscaled metals and alloys for a variety of applications including catalysis, environmental remediation, data storage, etc.
Conflict of interest The authors declare that there are no conflicts of interest.
[7] [8] [9]
[10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]
[25] [26] [27] [28] [29] [30] [31] [32]
R. Hudson, Y. Feng, R.S. Varma, A. Moores, Green Chem. 16 (2014) 4431. S.C.N. Tang, I.M.C. Lo, Water Res. 47 (2013) 2613. A.H. Lu, E.L. Salabas, F. Schüth, Angew. Chem., Int. Ed. 46 (2007) 1222. M.J. Sharif, M. Yamauchi, S. Toh, S. Matsumura, S.I. Noro, K. Kato, M. Takata, T. Tsukuda, Nanoscale 5 (2013) 1489. N.T.T. Trang, T.T. Thuy, D.M. Mott, M. Koyano, S. Maenosono, Chem. Phys. Lett. 574 (2013) 94. H.M.T. Galvis, J.H. Bitter, C.B. Khare, M. Ruitenbeek, A.I. Dugulan, K.P. de Jong, Science 335 (2012) 835. R.A. Crane, T.B. Scott, J. Hazard. Mater. 211 (2012) 112. Y. Vitta, V. Piscitelli, A. Fernandez, F. Gonzalez-Jimenez, J. Castillo, Chem. Phys. Lett. 512 (2011) 96. S. Cheong, P. Ferguson, K.W. Feindel, I.F. Hermans, P.T. Callaghan, C. Meyer, A. Slocombe, C.H. Su, F.Y. Cheng, C.S. Yeh, B. Ingham, M.F. Toney, R.D. Tilley, Angew. Chem., Int. Ed. 123 (2011) 4292. D.L. Huber, Small 1 (2005) 482. L.M. Lacroix, N.F. Huls, D. Ho, X. Sun, K. Cheng, S. Sun, Nano Lett. 11 (2011) 1641. S. Peng, C. Wang, J. Xie, S. Sun, J. Am. Chem. Soc. 128 (2006) 10676. A. Tiehm, S. Kraßnitzer, Y. Koltypin, A. Gedanken, Ultrason. Sonochem. 16 (2009) 617. J.H. Yu, D.W. Lee, B.K. Kim, T. Jang, J. Magn. Magn. Mater. 304 (2006) 16. J. Fan, Y. Guo, J. Wang, M. Fan, J. Hazard. Mater. 166 (2009) 904. J.E. Muñoz, J. Cervantes, R. Esparza, G. Rosas, J. Nanopart. Res. 9 (2007) 945. H.B. Liu, T.H. Chen, D.Y. Chang, D. Chen, Y. Liu, H.P. He, P. Yuan, R. Frost, Mater. Chem. Phys. 133 (2012) 205. J. Zielin´ski, I. Zglinicka, L. Znak, Z. Kaszkur, Appl. Catal., A: Gen. 381 (2010) 191. W.K. Jozwiak, E. Kaczmarek, T.P. Maniecki, W. Ignaczak, W. Maniukiewicz, Appl. Catal., A: Gen. 326 (2007) 17. F. Li, J. Ran, M. Jaroniec, S.Z. Qiao, Nanoscale 7 (2015) 17590. K. Rajeshwar, N.R. De Tacconi, Chem. Soc. Rev. 38 (2009) 1984. W. Morales, M. Cason, O. Aina, N.R. de Tacconi, K. Rajeshwar, J. Am. Chem. Soc. 130 (2008) 6318. G.S. Maciel, N. Rakov, R.A.S. Zanon, C.E. Fellows, R.B. Guimarães, J.J. Rodrigues, Chem. Phys. Lett. 465 (2008) 258. K.V. Manukyan, Y.S. Chen, S. Rouvimov, P. Li, X. Li, S. Dong, X. Liu, J.K. Furdyna, A. Orlov, G.H. Bernstein, W. Porod, S. Roslyakov, A.S. Mukasyan, Phys. Chem. C 118 (2014) 16264. R. Ianos, A. Tǎculescu, C. Pǎcurariu, I. Lazǎu, J. Am. Ceram. Soc. 95 (2012) 2236. Z. Cao, M. Qin, B. Jia, Y. Gu, P. Chen, A.A. Volinsky, X. Qu, Ceram. Int. 41 (2015) 2806. S.K. Apte, S.D. Naik, R.S. Sonawane, B.B. Kale, J. Am. Ceram. Soc. 90 (2007) 412. M. Sato, Y. Ishii, H. Nakae, J. Appl. Phys. 53 (1982) 6331. T. Iwaki, Y. Kakihara, T. Toda, M. Abdullah, K. Okuyama, J. Appl. Phys. 94 (2003) 6807. D.L. Leslie-Pelecky, R.D. Rieke, Chem. Mater. 8 (1996) 1770. T. Yamashita, P. Hayes, Appl. Surf. Sci. 254 (2008) 2441. M. Descostes, F. Mercier, N. Thromat, C. Beaucaire, M. Gautier-Soyer, Appl. Surf. Sci. 165 (2000) 288.
Contents lists available at ScienceDirect
Chemical Physics Letters journal homepage: www.elsevier.com/locate/cplett
Research paper
Magnetic iron nanoparticles prepared by solution combustion synthesis and hydrogen reduction Min Huang a, Mingli Qin a,⇑, Zhiqin Cao a,b, Baorui Jia a, Pengqi Chen a, Haoyang Wu a, Xuanli Wang a, Qi Wan a, Xuanhui Qu a a b
Institute for Advanced Materials and Technology, University of Science and Technology Beijing, Beijing 100083, China School of Resources and Environmental Engineering, Pan Zhihua University, Pan Zhihua 617000, China
a r t i c l e
i n f o
Article history: Received 3 December 2015 In final form 19 May 2016 Available online 20 May 2016 Keywords: Magnetic iron nanoparticles Solution combustion synthesis Hydrogen reduction
a b s t r a c t A facile and efficient method has been proposed to prepare iron nanoparticles by combining solution combustion synthesis and hydrogen reduction for the first time. A porous a-Fe2O3 precursor with high specific surface area of 75 m2/g was fabricated by solution combustion synthesis, and then iron nanoparticles with high saturation magnetization of 196.3 emu/g were successfully obtained by hydrogen reduction of the as-synthesized precursor. With the reduction temperature rising from 275 °C to 600 °C, the saturation magnetization of the products increases from 196.3 emu/g to 209.7 emu/g, whilst the coercivity decreases from 611.4 Oe to 98.8 Oe. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction Magnetic nanoparticles have emerged as a promising advanced functional material for a wide field of potential applications such as catalysis, water treatment, environmental remediation, biotechnology/biomedicine, magnetic recording media, electromagnetic wave absorption, magnetic sensors [1–5]. As compared to other magnetic nanoparticles, iron nanoparticles have attracted a great deal of attention due to exceptionally high saturation magnetization, large permeability, superior environmental compatibility and high availability [6–10]. In recent years, multiple techniques have been developed to prepare iron nanoparticles, including thermal decomposition [11,12], sonochemical decomposition [13], vapor phase condensation [14], aqueous reduction of iron salt [15], high-energy milling [16], and hydrogen reduction of nanoscaled iron oxide [17]. Among the above methods, owing to process simplicity, low cost as well as potential high productivity, hydrogen reduction (HR) is utilized to fabricate iron nanoparticles on a commercial basis. Over the past few years, in order to better understand and ameliorate HR, several investigations suggest that the purity and reductive activity of starting materials have great influence on the reduction temperature and microstructure of the asprepared product [18,19]. Thus, the production of iron oxide with
⇑ Corresponding author. E-mail address: [email protected] (M. Qin). http://dx.doi.org/10.1016/j.cplett.2016.05.043 0009-2614/Ó 2016 Elsevier B.V. All rights reserved.
high purity and reactivity is the prerequisite task required for achieving the successful preparation of iron nanoparticles by HR. Solution combustion synthesis (SCS) has been demonstrated to be one of the most appropriate methods to produce oxide nanomaterials [20–23]. It is well known that SCS is essentially exothermal redox reaction between an oxidizer (e.g. metal nitrates) and a fuel (e.g. glycine, urea, citric acid, etc.) in a homogenous aqueous solution within several seconds. Simultaneously, the exothermal reaction provides the energy required for sustaining the combustion reaction without adding external energy. The advantages of SCS thus are as follows: (1) the homogeneous mixing of all the components on the molecular level and the evolution of a large volume of gases result in the formation of nanoscaled products with high specific surface area and resultant high reactive activity; (2) inexpensive starting materials, selfsustained instantaneous reaction features, high yield, as well as simple processing and apparatuses render SCS cost-effective and suitable for mass production. In view of these advantages, a large number of efforts have been made to prepare iron oxides by SCS [24–27]. Despite of such fascinating research results, to our knowledge, no research on hydrogen reduction of iron oxides by SCS has been reported. Hence, in this work, we report a facile and efficient method to prepare iron nanoparticles by combining SCS and HR for the first time. First, by using ferric nitrate as an oxidizer, glycine as a fuel and glucose as an assisted additive and selecting an optimum molar ratio, a porous iron oxide precursor has been fabricated via SCS. Subsequently, iron nanoparticles with high saturation magnetization have been successfully obtained by hydrogen reduction of
34
M. Huang et al. / Chemical Physics Letters 657 (2016) 33–38
the as-synthesized precursor. Moreover, the whole combustion process and hydrogen reduction process have been investigated, and the influence of reduction temperature on magnetic properties of the products has been discussed in detail.
2. Experimental 2.1. Synthesis procedure All the starting materials involving ferric nitrate (Fe(NO3)39H2O, oxidizer and Fe source), glycine (C2H5NO2, fuel) and glucose (C6H12O6H2O, addictive), were commercially purchased and of analytical reagent grade. After a series of optimization experiments, the optimal molar ratios of glycine to ferric nitrate and glucose to ferric nitrate were fixed at 2.8 and 1.5, respectively, and the amount of ferric nitrate was 0.1 M. Just like the typical experimental procedure for SCS, the above starting materials were dissolved in 200 ml deionized water under thoroughly stirring to obtain a rufous homogeneous starting solution at room temperature. Subsequently, the resultant solution was poured into a 2000 ml beaker and heated in a temperaturecontrolled muffle furnace in air. As heating continued, the solution evaporated and gradually developed into a viscous gel. Then, the gel began to swell and a smoldering combustion reaction occurred, coupled with the evolution of a large volume of gases. The whole process of swelling and combustion seemed to undergo a selfpropagating and nonexplosive exothermic reaction and lasted only a few minutes, resulting in the generation of a gray porous foam. Upon further heating, the foam was ignited and the carbon was completely removed, leading to the formation of a loose reddish precursor. Hereafter, hydrogen reduction of the precursor was conducted in a tube furnace. A strict temperature program was followed in all runs, with heating up to the plateau temperature, 150–600 °C, at a constant heating rate of 10 °C/min. The assynthesized precursors were reduced at various temperatures for 2 h in a flowing hydrogen atmosphere at a flow rate of 1 L/min. Finally, when cooled naturally to room temperature, the reduction products were obtained.
2.2. Characterization Thermal behavior of the formed gel sample was investigated by thermal gravimetric analysis (TG) and Differential Scanning Calorimetry (DSC) using a Netzsch 409PC thermal analyzer. The TG and DSC curves were recorded over the temperature range of 25– 600 °C at a heating rate of 10 °C/min in air at a constant flow rate of 20 ml/min. The phase composition of the foam, precursor and reduction products was determined by X-ray Diffraction (XRD), using a TTRAX III diffractometer with monochromic Cu Ka radiation. The average crystallite size was calculated according to the Scherrer’s equation. The presence of Fe3O4 was studied by X-ray photoelectron spectroscopy (XPS), using a ESCALAB 250 Xi spectrometer. The morphology and particle size of the precursor and reduction products were observed by scanning electron microscopy (SEM, JSM-6510) and field emission scanning electron microscopy (FE-SEM,ZEISS ULTRA 55), respectively. The reduction products were further observed and the associated selected-area electron diffraction images were captured by transmission electron microscopy (TEM, Tecnai G2 F30 S-TWIN).The specific surface area (SSA) of the precursor was examined by BET method using an Automated Surface Area & Pore Size Analyzer (QUADRASORB SI-MP, Quantachrome Instruments, Boynton Beach, FL). The magnetic properties were measured by vibrating samples magnetometry (VSM) at room temperature, using a Lake Shore 7307 magnetometer.
3. Results and discussion 3.1. Solution combustion synthesis As described in detail in the experimental section, under continuous heating conditions, the homogeneous starting solution containing ferric nitrate, glycine, and glucose gradually developed into a viscous gel, and the succeeding thermal behavior of the gel was investigated. Fig. 1 depicts the TG and DSC curves of the formed gel. It is clear that the whole process is composed of three major stages with the two dash lines as division marks. In the initial stage (below 136 °C, dehydration), the gel sample undergoes a rapid mass loss of around 40% on the TG curve. On the same temperature interval, a roughly V-shaped endothermic peak at 110 °C is presented on the DSC curve, which can be attributed to the evaporation of remaining water and the desorption of chemically absorbed water in the gel. For the second stage (136–330 °C, smoldering combustion), as compared to the previous stage, the mass decreases by approximately 30% at a relatively gentle rate, accompanied by a wide and weak exothermic peak. The characteristic profile of the DSC curve suggests the low exothermic effect, which is in good agreement with the observed slow smoldering combustion phenomenon. It is well known that, after the removal of the various types of water, the formed gel becomes far more reactive with further heating, and then is closely followed by the vigorous initiation of the combustion reaction. Actually, in this stage the exothermic combustion reaction between ferric nitrate (oxidizer) and glycine (fuel) is conducted [Eq. (1)]. Simultaneously, after absorbing the heat released from combustion reaction, the decomposition of glucose (addictive) occurs [Eq. (2)], resulting in the generation of a gray porous foam [Fig. 2(a)]. Thus, it is the superposition of the exothermic combustion and endothermic decomposition reaction that is responsible for the total low exothermicity and smoldering phenomenon. With respect to the last stage (above 330 °C, decarbonization), as the temperature continues to rise, the sharp exothermic peak, located at approximately 393 °C, corresponds to the complete removal of carbon in the foam [Eq. (3)], leading to the formation of a loose reddish precursor [Fig. 2(b)]. In addition, when the temperature surpasses 450 °C, the mass of the sample keeps almost constant.
6FeðNO3 Þ3 þ 10C2 H5 NO2 ! 3Fe2 O3 þ 20CO2 þ 25H2 O þ 14N2
ð1Þ
C6 H12 O6 H2 O ! 6C þ 6H2 O
ð2Þ
C þ O2 ! CO2
ð3Þ
Fig. 1. TG and DSC curves of the formed gel.
M. Huang et al. / Chemical Physics Letters 657 (2016) 33–38
35
Fig. 2. Original appearance photographs of the foam (a), precursor (b), products reduced at 200 °C (c) and 275 °C (d).
Fig. 3. XRD patterns of the foam and precursor (a), EDS analysis of the foam (b) and SEM image of the precursor (c).
The original appearance photograph of the foam is shown in Fig. 2(a). It can be clearly seen that the foam has exhibited a spongy porous appearance, which is originated from large quantities of gases released slowly from the combustion reaction and decomposition reaction [Eqs. (1) and (2)]. Fig. 3 displays the XRD patterns of the foam and precursor, the EDS analysis of the foam and the SEM image of the precursor. It is found that no obvious Bragg diffraction peaks have been detected in the foam, revealing that it possesses an amorphous structure [Fig. 3(a)]. The corresponding EDS results confirm that the foam consists of C, Fe, and O elements [Fig. 3(b)]. Based on the XRD pattern, the EDS analysis as well as the low exothermicity of the smoldering combustion, it is safe to conclude that the energy liberated from smoldering combustion is not sufficient for the crystallization of iron oxide, rendering the foam a mixture of amorphous carbon and amorphous iron oxide. Furthermore, the precursor is confirmed as a crystalline a-Fe2O3 phase [Fig. 3(a)], which explains the reddish color1 of the precursor [Fig. 2(b)]. This is ascribed to the fact that the higher energy evolved from the oxidation of the car1 For interpretation of color in Figs. 2 and 3, the reader is referred to the web version of this article.
bon promotes the transformation of iron oxide from amorphous state to crystalline state [Eq. (3)]. In addition, it is noticed that the precursor exhibits a porous morphology, which arises from the dispersive effect of gases liberated from Eqs. (1)–(3). The SSA value of the precursor is evaluated to be 75 m2/g, which is undoubtedly explained by the typical porous structure. The crystallite size of the precursor calculated by the Scherrer’s equation is 19 nm. 3.2. Hydrogen reduction To investigate the phase transformation during the reduction process and determine the lowest reduction temperature, the iron oxide precursors were reduced in flowing hydrogen atmosphere at various temperatures for 2 h. The XRD patterns of the products reduced at 150–275 °C are shown in Fig. 4. At 150 °C, no other obvious diffraction peaks are found, except the peaks belonging to a-Fe2O3 phase, which suggests that the temperature is too low to initiate the reduction reaction. The XRD pattern points out that the product reduced at 175 °C is a mixture of Fe3O4 and a-Fe2O3 phases with Fe3O4 as the main phase. This demonstrates that the reduction reaction has already occurred and the onset temperature has been below 175 °C. With the temperature rising to 200 °C, the
36
M. Huang et al. / Chemical Physics Letters 657 (2016) 33–38
Fig. 4. XRD patterns of the products reduced at 150–600 °C.
diffraction peaks from a-Fe2O3 disappear and only diffraction peaks of Fe3O4 exist, implying the formation of a single Fe3O4 intermediate product [Fig. 2(c)]. Subsequently, for the product reduced at 225 °C, the weak diffraction peaks from bcc iron have appeared, and then iron becomes the main phase at 250 °C. When the temperature goes up to 275 °C, it is evident that only diffraction peaks from metal iron are observed in the XRD pattern. This means that the hydrogen reduction of the iron oxide precursor has been completed, and thus the pure iron product is obtained [Fig. 2(d)]. It is worth noting that the reduction temperature is as low as 275 °C, which can be ascribed to the high SSA and resultant high reduction reactivity of the precursor. The low reduction temperature can effectively prevent the agglomeration and coarsening of iron particles. Meanwhile, it is also expected to improve production efficiency and reduce energy cost. The XRD pattern for the product reduced at 200 °C identifies the formation of Fe3O4 as a single intermediate phase. Since the positions of the diffraction peaks in the XRD patterns are nearly identical for Fe3O4 and c-Fe2O3, XPS is employed to illustrate the presence of Fe3O4. Fig. 5 displays the high-resolution XPS spectrum of Fe 2p of the product reduced at 200 °C. The Fe 2p spectrum possesses Fe 2p2/3 and Fe1/2 doublets, located at 710.6 eV and 724.1 eV, respectively, which are in excellent agreement with the literature data for the Fe3O4 standard sample [31]. Furthermore, it was
Fig. 5. High-resolution XPS spectrum of Fe 2p from the product reduced at 200 °C.
previously reported that between the Fe 2p2/3 and Fe1/2 peaks the absence of a shoulder-shaped satellite peak is a typical characteristic for Fe3O4 [31,32]. The absence of the satellite peak has been confirmed in this work. Based on the results of the XRD pattern, the XPS spectra, as well as the typical black color [Fig. 2(c)], it can be concluded that the product reduced at 200 °C is a single Fe3O4 intermediate phase. The microstructure of the product reduced at 275 °C is displayed in Fig. 6. It can be seen that the product consists of irregular sheets with 2 lm in width [Fig. 6(a)]. To acquire more information about the reduction product, the product is subject to TEM analysis. The product presents the two-dimensional sheet structure [Fig. 6(b)], which is in accord with the FESEM image. A closer observation reveals that the sheet is composed of irregular particles with the size of 100 nm [Fig. 6(c)]. The corresponding selected-area electron diffraction results in the inset of Fig. 6(b) further confirm the polycrystalline iron phase, fitting very well with the above XRD results. Besides, the crystallite size is calculated to be 21 nm by the Scherrer’s equation, slightly larger than that of the precursor, implying no obviously coarsening of the crystallite occurs during the reduction process. Such small size can be attributed to the multiple factors of the ultrafine structure of the precursor, the low reduction temperature and the strong dispersive effect of the water vapor evolved from the reduction reaction. 3.3. Magnetic properties The XRD patterns of the products reduced at 275–600 °C are given in Fig. 4. The XRD results illustrate that all the products have the typical bcc iron characteristic. As the reduction temperature rises, the diffraction peaks become sharper and peak intensity increases correspondingly. The average crystallite sizes evaluated by the Scherrer’s equation and SSA values determined by BET method of the products are listed in Table 1. It can be found that, with the rising of the reduction temperature, the crystallite sizes increase, whereas the SSA values decrease. The TEM image of the product reduced at 600 °C is presented in Fig. 6(d). It is noted that the obvious coarsening and aggregation of the iron particles have occurred. Fig. 7 depicts the room-temperature magnetic hysteresis loops of the products reduced at 275–600 °C. The saturation magnetizations are calculated by extrapolating to infinite field, and the main magnetic parameters are listed in Table1. Magnetic measurement results reveal that all the products present typical ferromagnetic characteristics. The iron particles reduced at 275 °C possess a rather high saturation magnetization value of 196.3 emu/g, which is close to 90% of that of the bulk counterpart (218 emu/g). With the reduction temperature rising to 600 °C, the saturation magnetization increases to 209.7 emu/g. It is well know that, for magnetic nanoscaled materials, the saturation magnetization depends strongly on the crystallinity and chemical composition [4,10]. Therefore, the high saturation magnetization of iron nanoparticles at 275 °C herein is attributed to the high crystallinity derived from the moderate temperature condition during the reduction process, whereas the varying saturation magnetization is ascribed to the oxidation of the iron particles. It can be noted from Table 1 that the coercivity of iron particles reduced at 275 °C reaches 611.4 Oe, and the coercivity decreases steeply to 98.8 Oe with the reduction temperature increasing to 600 °C. As reported in previous literatures, for a single-crystal iron particle, the critical size Dc for a single-domain structure increases with increasing the axial ratio n, and the smallest critical size is proximately 20 nm when the axial ratio n equals 1 [28]. By comparing the critical size (20 nm) with the calculated crystallite size (21 nm) of the iron nanoparticles reduced at 275 °C, it is deduced that the crystallite almost exists in a single-domain state. For a
37
M. Huang et al. / Chemical Physics Letters 657 (2016) 33–38
Fig. 6. FSEM image (a), TEM images (b, c) of the product reduced at 275 °C and TEM image (d) of the product reduced at 600 °C.
Table 1 Characteristics of the products reduced at 275–600 °C. Reduction temperature (°C)
Crystallite size (nm)
BET SSA (m2/g)
Mr (emu/g)
Ms (emu/g)
Hc (Oe)
275 400 500 600
21 ± 2 38 ± 3 55 ± 5 62 ± 4
18 ± 1.7 13 ± 1.3 7 ± 0.9 5 ± 0.5
85.3 ± 0.5 52.9 ± 0.3 25.5 ± 0.4 11.9 ± 0.5
196.3 ± 0.7 202.9 ± 0.4 206.5 ± 0.2 209.7 ± 0.5
611.4 ± 3.5 375.6 ± 2.8 174.6 ± 2.3 98.8 ± 1.7
Fig. 7. Room temperature magnetic hysteresis loops of the products reduced at 275–600 °C.
38
M. Huang et al. / Chemical Physics Letters 657 (2016) 33–38
system of single domains, due to the absence of domain walls, magnetization reversal occurs through the coherent rotation of domains, resulting in the high coercivity [29,30]. Meantime, the shape anisotropy arising from the irregular shape can give rise to the high coercivity [30]. Furthermore, as the reduction temperature goes up, the crystallite size becomes significantly larger than the critical size and the crystallite undergoes a transformation from a single-domain state to a multidomain state. Accordingly, Magnetization reversal no longer occurs through the domain rotation and instead commence through the nucleation and motion of the domain walls, rendering the low coercivity. Simultaneously, the decreasing crystallite boundary pinning of the motion of the domain wall can serve to explain the swiftly declining coercivity.
Acknowledgments This work was supported by the National Natural Science Foundation Program of China (51574031) and (51574029), the Natural Science Foundation Program of Beijing (2162027), the Fundamental Research Funds for the Central Universities (06109063) and the National 863 Program (2013AA031101). References [1] [2] [3] [4] [5] [6]
4. Conclusions A facile and efficient method has been proposed to prepare iron nanoparticles by combining solution combustion synthesis and hydrogen reduction. A porous a-Fe2O3 precursor with high specific surface area of 75 m2/g was fabricated by SCS, and then iron nanoparticles were successfully obtained by hydrogen reduction of the as-synthesized precursor. The whole combustion process consists of three major stages, i.e., dehydration, smoldering combustion and decarbonization. The precursor has completed the reduction reaction at the temperature as low as 275 °C. Iron nanoparticles reduced at 275 °C exhibit a high saturation magnetization value of 196.3 emu/g which is close to 90% of that of the bulk counterpart (218 emu/g). The crystallite sizes of the products reduced at 275–600 °C range from 21 nm to 62 nm. The reduction temperature exhibits a significant effect on the magnetic properties of reduction products. With reduction temperature rising from 275 °C to 600 °C, the saturation magnetization of the products increases from 196.3 emu/g to 209.7 emu/g, whilst the coercivity decreases from 611.4 Oe to 98.8 Oe. Finally, it should be emphasized that, due to integration of the common advantages of exceptional simplicity and high efficiency originated from SCS and HR, the method proposed herein will be a promising versatile technique for mass production of highly pure nanoscaled metals and alloys for a variety of applications including catalysis, environmental remediation, data storage, etc.
Conflict of interest The authors declare that there are no conflicts of interest.
[7] [8] [9]
[10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]
[25] [26] [27] [28] [29] [30] [31] [32]
R. Hudson, Y. Feng, R.S. Varma, A. Moores, Green Chem. 16 (2014) 4431. S.C.N. Tang, I.M.C. Lo, Water Res. 47 (2013) 2613. A.H. Lu, E.L. Salabas, F. Schüth, Angew. Chem., Int. Ed. 46 (2007) 1222. M.J. Sharif, M. Yamauchi, S. Toh, S. Matsumura, S.I. Noro, K. Kato, M. Takata, T. Tsukuda, Nanoscale 5 (2013) 1489. N.T.T. Trang, T.T. Thuy, D.M. Mott, M. Koyano, S. Maenosono, Chem. Phys. Lett. 574 (2013) 94. H.M.T. Galvis, J.H. Bitter, C.B. Khare, M. Ruitenbeek, A.I. Dugulan, K.P. de Jong, Science 335 (2012) 835. R.A. Crane, T.B. Scott, J. Hazard. Mater. 211 (2012) 112. Y. Vitta, V. Piscitelli, A. Fernandez, F. Gonzalez-Jimenez, J. Castillo, Chem. Phys. Lett. 512 (2011) 96. S. Cheong, P. Ferguson, K.W. Feindel, I.F. Hermans, P.T. Callaghan, C. Meyer, A. Slocombe, C.H. Su, F.Y. Cheng, C.S. Yeh, B. Ingham, M.F. Toney, R.D. Tilley, Angew. Chem., Int. Ed. 123 (2011) 4292. D.L. Huber, Small 1 (2005) 482. L.M. Lacroix, N.F. Huls, D. Ho, X. Sun, K. Cheng, S. Sun, Nano Lett. 11 (2011) 1641. S. Peng, C. Wang, J. Xie, S. Sun, J. Am. Chem. Soc. 128 (2006) 10676. A. Tiehm, S. Kraßnitzer, Y. Koltypin, A. Gedanken, Ultrason. Sonochem. 16 (2009) 617. J.H. Yu, D.W. Lee, B.K. Kim, T. Jang, J. Magn. Magn. Mater. 304 (2006) 16. J. Fan, Y. Guo, J. Wang, M. Fan, J. Hazard. Mater. 166 (2009) 904. J.E. Muñoz, J. Cervantes, R. Esparza, G. Rosas, J. Nanopart. Res. 9 (2007) 945. H.B. Liu, T.H. Chen, D.Y. Chang, D. Chen, Y. Liu, H.P. He, P. Yuan, R. Frost, Mater. Chem. Phys. 133 (2012) 205. J. Zielin´ski, I. Zglinicka, L. Znak, Z. Kaszkur, Appl. Catal., A: Gen. 381 (2010) 191. W.K. Jozwiak, E. Kaczmarek, T.P. Maniecki, W. Ignaczak, W. Maniukiewicz, Appl. Catal., A: Gen. 326 (2007) 17. F. Li, J. Ran, M. Jaroniec, S.Z. Qiao, Nanoscale 7 (2015) 17590. K. Rajeshwar, N.R. De Tacconi, Chem. Soc. Rev. 38 (2009) 1984. W. Morales, M. Cason, O. Aina, N.R. de Tacconi, K. Rajeshwar, J. Am. Chem. Soc. 130 (2008) 6318. G.S. Maciel, N. Rakov, R.A.S. Zanon, C.E. Fellows, R.B. Guimarães, J.J. Rodrigues, Chem. Phys. Lett. 465 (2008) 258. K.V. Manukyan, Y.S. Chen, S. Rouvimov, P. Li, X. Li, S. Dong, X. Liu, J.K. Furdyna, A. Orlov, G.H. Bernstein, W. Porod, S. Roslyakov, A.S. Mukasyan, Phys. Chem. C 118 (2014) 16264. R. Ianos, A. Tǎculescu, C. Pǎcurariu, I. Lazǎu, J. Am. Ceram. Soc. 95 (2012) 2236. Z. Cao, M. Qin, B. Jia, Y. Gu, P. Chen, A.A. Volinsky, X. Qu, Ceram. Int. 41 (2015) 2806. S.K. Apte, S.D. Naik, R.S. Sonawane, B.B. Kale, J. Am. Ceram. Soc. 90 (2007) 412. M. Sato, Y. Ishii, H. Nakae, J. Appl. Phys. 53 (1982) 6331. T. Iwaki, Y. Kakihara, T. Toda, M. Abdullah, K. Okuyama, J. Appl. Phys. 94 (2003) 6807. D.L. Leslie-Pelecky, R.D. Rieke, Chem. Mater. 8 (1996) 1770. T. Yamashita, P. Hayes, Appl. Surf. Sci. 254 (2008) 2441. M. Descostes, F. Mercier, N. Thromat, C. Beaucaire, M. Gautier-Soyer, Appl. Surf. Sci. 165 (2000) 288.