Fe2o3 Synthesis and Phase Transformation
Fe2o3 Synthesis and Phase Transformation
Fe2o3 Synthesis and Phase Transformation
Ceramics International
journal homepage: www.elsevier.com/locate/ceramint
A R T I C LE I N FO A B S T R A C T
Keywords: It is a general consensus that ε-polymorph of Fe2O3 is not easy to synthesize in the form of bare nano-sized
Transition metal oxides objects owing to its significant thermal instability. Thus, a supporting medium (e.g., silica matrix in most cases)
Powders: Chemical preparation is always necessary for its successful high-temperature synthesis at the sacrifice of its yield. In this study, op-
Magnetic properties timized amounts of protons and nitrate ions were employed in a hydrothermal synthetic route to facilitate the
growth of pure ε-Fe2O3 nanoparticles. The absence of silica matrix ensured a high yield of ε-Fe2O3 product, and
moderate reaction temperature of mere 200 °C was conducive to its scalable energy-efficient production. These ε-
Fe2O3 nanoparticles exhibited large room-temperature coercivity of ca. 16.3 kOe, which was found to be de-
generated to 36 Oe as prolonging the reaction time due to the phase and structure evolution. The preformed ε-
Fe2O3 nanoparticles and post-formed α-Fe2O3 phase would co-align with each other to form nanoflakes for long-
time reaction, being of a character of oriented attachment. The concomitant distorted lattice spacing and dis-
ordered atomic configuration not only contributed to the degeneration of ferromagnetism of ε-Fe2O3 phase, but
also induced unique spin glass behavior.
⁎
Corresponding author.
E-mail address: [email protected] (K. Chen).
https://doi.org/10.1016/j.ceramint.2018.07.162
Received 16 April 2018; Received in revised form 28 June 2018; Accepted 18 July 2018
Available online 19 July 2018
0272-8842/ © 2018 Elsevier Ltd and Techna Group S.r.l. All rights reserved.
J. Ma, K. Chen Ceramics International 44 (2018) 19338–19344
hydrothermal route has found its place in several branches of modern magnetic field down to the investigated temperatures. Then, closed
science and technology in view of its high solvation power, high com- hysteresis loops were measured by applying magnetic fields from 0 to
pressibility and mass transport of solvents [22]. Besides, its mild syn- + 50 kOe, then to −50 kOe, and in turn to + 50 kOe to the end. The
thetic conditions, simple manipulation, and good crystallization of the first 0 → + 50 kOe curve is called the initial magnetization curve.
products confer outstanding merits to this facile route. However, re- The thermo-magnetization [M(T)] curves were measured in zero
garding to the synthesis of metastable phases (e.g., ε-Fe2O3 crystals, field-cooled (ZFC) and field-cooled (FC) procedures. Firstly, the sample
metal alkoxides, amorphous phases) at relatively high temperatures in was cooled from 300 K down to 5 K in a zero magnetic field. Then the
the range, hydrothermal route should be examined very carefully be- magnetization of the sample was measured as heating it from 5 to 300 K
cause it always compromises thermodynamic non-equilibrium states in a given magnetic field. After that, the sample was cooled down again
and commonly yields products with high thermodynamic stability. In to 5 K with the applied field held constant, and the magnetization was
this scenario, the addition of protons and nitrate ions are found to be measured once more during the heating-up process from 5 to 300 K in
critical during hydrothermal procedure to yield pure ε-Fe2O3 phase the same magnetic field.
with no aid of supporting matrix. The merits of this synthetic route lie Collectively, even though the experimental uncertainty in the ab-
in three aspects: (i) no addition of silica matrix ensures a high yield of ε- solute magnetization is ca. 1%, the relative magnetization uncertainty
Fe2O3 phase; (ii) moderate reaction temperature of mere 200 °C not between field points for the same sample is smaller than the data point
only finds a way out of the dilemma between high temperature and size in the figure, and the statistical uncertainties (not shown in this
small size, but also is conducive to its scalable energy-efficient pro- study) are smaller than the point size. Moreover, any effect on magnetic
duction; (iii) by prolonging reaction time, the as-synthesized ε-Fe2O3 properties caused by the shape anisotropy of the sample will not be
nanoparticles can self-assemble into nanoflakes with a character of resolved in our experiments because the measurements were performed
oriented attachment, offering a realistic possibility to tailor ε-Fe2O3 in a on powders with randomly oriented particulates.
suitable shape useful to be integrated in devices.
2.1. Synthesis of ε-Fe2O3 products Protons (provided by (NH4)H2PO4 and (NH4)2HPO4 reagents) and
nitrate ions (provided by Fe(NO3)3·9H2O reagent) are employed in our
The analytically pure reagents Fe(NO3)3·9H2O, (NH4)H2PO4 and hydrothermal route to prolong the reaction process that approaches
(NH4)2HPO4 in experiments were purchased from Sinopharm Chemical final thermodynamic equilibrium states, so that metastable ε-Fe2O3
Reagent Co., Ltd. (China) and used as received without further pur- phase may be readily obtained by intentionally interrupting this reac-
ification. Typically, 3.2 mL of aqueous Fe(NO3)3·9H2O solution tion. Along this line, the reaction is artificially interrupted after 3 h, and
(0.5 mol/L) was mixed with 1.44 mL of isometric (NH4)H2PO4 and the morphology of the product is shown in the inset of Fig. 1a. Clearly,
(NH4)2HPO4 aqueous solution with equimolar concentration of this 3 h-product is composed of numerous nanoparticles of similar sizes,
0.1 mol/L. After ultrasonic agitation for 3 min, distilled water was and no large structures can be observed spreading over the SEM sub-
added to a final volume of 50 mL. Further vigorously stirring for 5 min, strate. Its phase composition can be determined by the XRD pattern as
the mixture was transferred into Teflon-lined stainless-steel autoclaves shown in Fig. 1a, wherein all diffraction peaks can be indexed to the
with a capacity of 100 mL for hydrothermal treatment at 200 °C for 3 h orthorhombic structure of ε-Fe2O3 (Pna21, JCPDS No. 52–1449).
to yield ε-Fe2O3 nanoparticles and 18 h to yield iron-oxide nanoflakes.
After the autoclave had cooled down to room temperature naturally,
the precipitates were centrifuged, washed with distilled water and ab-
solute ethanol, and finally dried in the air at 60 °C for 4 h. Experiments
utilizing the same procedures were repeated five times with consistent
results.
2.2. Characterization
19339
J. Ma, K. Chen Ceramics International 44 (2018) 19338–19344
shown) and typical HRTEM images in Fig. 1b, in which the 2.14 Å
0
fringe spacing corresponds to the (212) inter-planar spacing of ε-Fe2O3.
5K Irrespective of analyzing the relationship between volume distribution
-5
100 K function and number distribution function, the immediately apparent
200 K mismatch of these two average size values can be explained by the fact
-10
300 K
that (i) micro-strains often increase the peak broadening for nano-
particles, leading to smaller particle size according to Scherrer equa-
-15
-60000 -40000 -20000 0 20000 40000 60000 tion; (ii) the variations of the particle shape make the analysis based on
H (Oe) the assumption of spherical geometry less reliable and are probably the
12
origin of the disagreement between the average diameters as de-
(b) termined by XRD and TEM.
One of the most salient magnetic features of ε-Fe2O3 phase is its very
high coercivity (ca. 20 kOe), which can even be used as a sole criterion
to judge whether ε-Fe2O3 phase exists or not in the product. Fig. 2a
shows a set of magnetic hysteresis loops of the as-synthesized ε-Fe2O3
M (emu/g)
Fig. 3. SEM images of control samples after reacting between (a) Fe(NO3)3·9H2O and (NH4)2HPO4, (b) Fe(NO3)3·9H2O and (NH4)H2PO4, (c) FeCl3·6H2O and (NH4)
H2PO4/ (NH4)2HPO4 for 3 h. Their XRD patterns are shown in panel d, e, and f, respectively.
19340
J. Ma, K. Chen Ceramics International 44 (2018) 19338–19344
Fig. 4. (a) XRD pattern, (b) SEM image, (c) TEM image, (d) 3-D topographical MFM image (500 nm × 500 nm × 50 nm), and (e) HRTEM image of iron-oxide flakes.
Typical distorted regions in panel e are separately shown in panel f and g.
(i) symmetric Fe3+ cations with a half-filled d-shell; (ii) defects indu- microspheres consisting of aggregated nanoparticles are prevalent in
cing two possible orientations within a crystallite and thus enhancing (NH4)2HPO4 case due to its much lower H+ concentration. Apart from
the magnetocrystalline anisotropy of ε-Fe2O3 nanoparticles; (iii) single- these substantially morphological difference and size deviation, both
domain character of ε-Fe2O3 nanoparticles, which evokes high mag- the highest H+ concentration in (NH4)H2PO4 case and the lowest H+
netic anisotropy (K ~ 2 ×105 J/m3) and low values of saturation concentration in (NH4)2HPO4 case are not propitious to the formation
magnetization (MS ~ 15 A m2/kg), and hence yielding giant coercivity of ε-Fe2O3 phase. Fig. 3d and e illustrate that α-Fe2O3 is the preferential
according to the formula of HC ∝ K/MS. In any case, this surprisingly and final phase in the resultant 3 h-products, even though the crystal-
large room-temperature coercivity is generally believed to be affected linity is rather poor at high H+ concentration in (NH4)H2PO4 case.
by magnetocrystalline anisotropy as well as particle size and size dis- Thereby, the equimolar hybrid of two etching agents (NH4)H2PO4 and
tribution. (NH4)2HPO4 is fairly crucial in the successful synthesis of ε-Fe2O3 phase
In the preceding analysis, the goal of synthesizing pure ε-Fe2O3 and is thought to be one of the innovation points in our synthetic route.
phase via hydrothermal route at mere 200 °C with no aid of supporting The other innovation point is the utilization of nitrate ions coming from
matrix can be readily achieved by simply adding protons and nitrate Fe(NO3)3·9H2O reagent. If replaced by FeCl3·6H2O with other condi-
ions in reaction. In this case, optimizing the amount of (NH4)H2PO4, tions held constant, the resultant product is of tubular morphology
(NH4)2HPO4 and Fe(NO3)3·9H2O reagents is of utmost importance in mixed with concaved spindles (Fig. 3c). Similar to the other two cases,
determining the success of this synthetic route. In our aforementioned its phase composition is also assigned to hexagonal structured α-Fe2O3
hydrothermal procedure, the molar ratio of (NH4)H2PO4 and (R—3c, JCPDS No. 33–0664). The major role of nitrate ions during the
(NH4)2HPO4 reagents is set to be 1:1 with the same concentration of reaction is their thermal decomposition to gases (including oxygen and
0.1 mol/L. If no (NH4)H2PO4 reagent is added and the volume of nitrogen dioxide), which in turn degenerate the crystallinity of newly-
(NH4)2HPO4 aqueous solution (0.1 mol/L) is doubled with other con- formed iron oxide and prolong the reaction process approaching ther-
ditions held constant, the as-obtained particle is of spherical mor- modynamic equilibrium states.
phology with an average diameter of ca. 200 nm, as shown in Fig. 3a. As stated above, the hydrothermal reaction is intentionally inter-
By contrast, if the volume of (NH4)H2PO4 aqueous solution (0.1 mol/L) rupted at 3 h because such an abortive reaction is conducive to the
is doubled without the addition of (NH4)2HPO4 reagent, smaller-sized formation of metastable ε-Fe2O3 phase. Otherwise, more energetically
(ca. 15 nm) particles are finally obtained (Fig. 3b). According to our stable phases and morphologies will occur in an undisturbed reaction
previous reports [26,27], proton-etching effect is prevalent in these process. Fig. 4a shows an XRD pattern of the as-obtained product after
reactions and dominates the final morphologies and sizes of the pro- 18 h-reaction. As expected, apart from the previously observed ε-Fe2O3
ducts. In this context, scattered finite-sized particles are formed in phase (Pna21, JCPDS No. 52–1449), some reflection peaks can be un-
(NH4)H2PO4 case owing to severe proton-etching effect, whereas ambiguously indexed to hexagonal structure of α-Fe2O3 (R—3c, JCPDS
19341
J. Ma, K. Chen Ceramics International 44 (2018) 19338–19344
6
(a) 5 (b)
4
4 5K
2 100 K
M (emu/g)
200 K
M (emu/g)
3
0 300 K
5K 2
-2
100 K
-4 200 K 1
300 K
-6 0
-40000 -20000 0 20000 40000 0 10000 20000 30000 40000 50000
H (Oe) H (Oe)
3 0.5
(c) (d) 5K 100 K
2
200 K
1
M (emu/g)
M (emu/g)
300 K
0 0.0
-1 5K
100 K
-2 200 K
300 K
-3 -0.5
-40000 -20000 0 20000 40000 -5000 0 5000
H (Oe) H (Oe)
Fig. 5. (a) Magnetic hysteresis loops and (b) initial magnetization curves of iron-oxide flakes measured at different temperatures. FM-component loops deduced from
panel a are plotted from (c) − 50 to + 50 kOe, and (d) − 5 to + 5 kOe.
0.004
(a) 10 Oe 5000 (a) 10 Oe
0.003
4000
FC
1/M (g/emu)
M (emu/g)
0.002 3000
2000
0.001
1000
0.000
ZFC
0
0 50 100 150 200 250 300 0 50 100 150 200 250 300
T (K) T (K)
150
0.06
(b) 100 Oe
(b) 100 Oe
0.05
FC 100
0.04
1/M (g/emu)
M (emu/g)
0.03
50
0.02
0.01
ZFC
0
0.00 0 50 100 150 200 250 300
0 50 100 150 200 250 300
T (K)
T (K)
Fig. 7. The reciprocal of M vs. T curves at (a) 0.01 and (b) 0.1 kOe, with their
Fig. 6. ZFC and FC M(T) curves of iron-oxide flakes measured at (a) 0.01, and
high-temperature zone (marked as rounded rectangles) being plotted as insets.
(b) 0.1 kOe.
19342
J. Ma, K. Chen Ceramics International 44 (2018) 19338–19344
19343
J. Ma, K. Chen Ceramics International 44 (2018) 19338–19344
I. Voicu, M. Balu, I. Morjan, Production of iron-oxide nanoparticles by laser-induced ferrite, Sci. Rep. 6 (2016) 27212.
pyrolysis of gaseous precursors, Appl. Surf. Sci. 154 (2000) 353–359. [25] S. Sakurai, J. Jin, K. Hashimoto, S. Ohkoshi, Reorientation phenomenon in a
[12] Y. Wang, J. Ma, S. Zuo-Jiang, K. Chen, Tailorable magnetic properties of ε-Fe2O3/ magnetic phase of ε-Fe2O3 nanocrystal, J. Phys. Soc. Jpn 74 (2005) 1946–1949.
SiO2 hybrid via alkaline etching, Ceram. Int. 43 (2017) 16482–16487. [26] C. Liu, J. Ma, H. Chen, Periodical structural conversion and its mechanism in he-
[13] M. Popovici, M. Gich, D. Nižňanský, A. Roig, C. Savii, L. Casas, E. Molins, K. Zaveta, matite: from nanospindles, to nanotubes, to nanotires, RSC Adv. 2 (2012)
C. Enache, J. Sort, S. de Brion, G. Chouteau, J. Nogués, Optimized synthesis of the 1009–1013.
elusive ε-Fe2O3 phase via sol-gel chemistry, Chem. Mater. 16 (2004) 5542–5548. [27] J. Ma, X. Zhang, K. Chen, G. Li, X. Han, Morphology-controlled synthesis of he-
[14] R. Zboril, M. Mashlan, D. Petridis, Iron(III) oxides from thermal processes- synth- matite hierarchical structures and their lithium storage performances, J. Mater.
esis, structural and magnetic properties, mössbauer spectroscopy characterization, Chem. A 1 (2013) 5545–5553.
and applications, Chem. Mater. 14 (2002) 969–982. [28] D. Li, M.H. Nielsen, J.R.I. Lee, C. Frandsen, J.F. Banfield, J.J. de Yoreo, Direction-
[15] M. Gich, A. Roig, E. Taboada, E. Molins, C. Bonafos, E. Snoeck, Stabilization of specific interactions control crystal growth by oriented attachment, Science 336
metastable phases in spatially restricted fields: the case of the Fe2O3 polymorphs, (2012) 1014–1018.
Faraday Discuss. 136 (2007) 345–354. [29] H. Xu, W. Wang, W. Zhu, L. Zhou, M. Ruan, Hierarchical-oriented attachment: from
[16] S. Sakurai, A. Namai, K. Hashimoto, S. Ohkoshi, First observation of phase trans- one-dimensional Cu(OH)2 nanowires to two-dimensional CuO nanoleaves, Cryst.
formation of all four Fe2O3 phases (γ → ε → β → α-phase), J. Am. Chem. Soc. 131 Growth Des. 7 (2007) 2720–2724.
(2009) 18299–18303. [30] H. Zeng, J. Li, J.P. Liu, Z.L. Wang, S. Sun, Exchange-coupled nanocomposite mag-
[17] M. Tadic, I. Milosevic, S. Kralj, M. Mitric, D. Makovec, M.L. Saboungi, L. Motte, nets by nanoparticle self-assembly, Nature 420 (2002) 395–398.
Synthesis of metastable hard-magnetic ε-Fe2O3 nanoparticles from silica-coated [31] L. Lu, Z. Ai, J. Li, Z. Zheng, Q. Li, L. Zhang, Synthesis and characterization of Fe-
akaganeite nanorods, Nanoscale 9 (2017) 10579–10584. Fe2O3 core-shell nanowires and nanonecklaces, Cryst. Growth Des. 7 (2007)
[18] J. Jin, S. Ohkoshi, K. Hashimoto, Giant coercive field of nanometer-sized iron oxide, 459–464.
Adv. Mater. 16 (2004) 48–51. [32] N.J.O. Silva, V.S. Amaral, L.D. Carlos, Relevance of magnetic moment distribution
[19] S. Sakurai, J. Shimoyama, K. Hashimoto, S. Ohkoshi, Large coercive field in mag- and scaling law methods to study the magnetic behavior of antiferromagnetic na-
netic-field oriented ε-Fe2O3 nanorods, Chem. Phys. Lett. 458 (2008) 333–336. noparticles: application to ferritin, Phys. Rev. B 71 (2005) 184408.
[20] S. Sakurai, K. Tomita, K. Hashimoto, H. Yashiro, S. Ohkoshi, Preparation of the [33] I.S. Lyubutin, C.R. Lin, Y.V. Korzhetskiy, T.V. Dmitrieva, R.K. Chiang, Mössbauer
nanowire form of ε-Fe2O3 single crystal and a study of the formation process, J. spectroscopy and magnetic properties of hematite/magnetite nanocomposites, J.
Phys. Chem. C 112 (2008) 20212–20216. Appl. Phys. 106 (2009) 034311.
[21] S. Ohkoshi, S. Sakurai, J. Jin, K. Hashimoto, The addition effects of alkaline earth [34] C. Lan, J. Gong, S. Liu, S. Yang, Synthesis and magnetic properties of single-crys-
ions in the chemical synthesis of ε-Fe2O3 nanocrystals that exhibit a huge coercive talline Na2- x Mn8O16 nanorods, Nanoscale Res. Lett. 6 (2011) 133.
field, J. Appl. Phys. 97 (2005) 10K312. [35] Y. Tian, W. Wang, Y. Chai, J. Cong, S. Shen, L. Yan, S. Wang, X. Han, Y. Sun,
[22] A. Rabenau, The role of hydrothermal synthesis in preparative chemistry, Angew. Quantum tunneling of magnetization in a metal-organic framework, Phys. Rev. Lett.
Chem. Int. Ed. Engl. 24 (1985) 1026–1040. 112 (2014) 017202.
[23] E. Taboada, M. Gich, A. Roig, Nanospheres of silica with an ε-Fe2O3 single crystal [36] U.C. Chung, J.L. Mesa, J.L. Pizarro, I. de Meatza, M. Bengoechea, J.R. Fernández,
nucleus, ACS Nano 3 (2009) 3377–3382. M.I. Arriortua, T. Rojo, Li1.43[FeII4.43FeIII
0.57(HPO3)6]·1.5H2O: a phosphite oxoanion-
[24] S. Ohkoshi, A. Namai, T. Yamaoka, M. Yoshikiyo, K. Imoto, T. Nasu, S. Anan, based compound with lithium exchange capability and spin-glass magnetic beha-
Y. Umeta, K. Nakagawa, H. Tokoro, Mesoscopic bar magnet based on ε-Fe2O3 hard vior, Chem. Mater. 23 (2011) 4317–4330.
19344