- Email: [email protected]
Influence of Ni-dopant on the properties of synthetic goethite
Journal of Alloys and Compounds 403 (2005) 368–375
Influence of Ni-dopant on the properties of synthetic goethite Stjepko Krehula a , Svetozar Musi´c a,∗ , Stanko Popovi´c b b
a Division of Materials Chemistry, Ruder – Boˇskovi´c Institute, P.O. Box 180, HR-10002 Zagreb, Croatia Department of Physics, Faculty of Science, University of Zagreb, P.O. Box 331, HR-10002 Zagreb, Croatia
Received 6 May 2005; accepted 3 June 2005 Available online 1 August 2005
Abstract The influence of Ni-dopant on the properties of ␣-FeOOH was investigated by XRD, FT-IR, 57 Fe M¨ossbauer spectroscopy and transmission electron microscopy. ␣-FeOOH was synthesized at a highly alkaline pH by precipitation from the FeCl3 solution with the addition of tetramethylammonium hydroxide and autoclaving at 160 ◦ C. The samples doped with Ni2+ ions were precipitated in the same way, but in the presence of varying concentrations of NiCl2 . Solid solutions, having the structure of ␣-FeOOH, were observed in samples with the Ni/Fe ratio up to 0.05. Upon increasing the amount of Ni-dopant the XRD lines were gradually broadened. The sample with the ratio Ni/Fe = 0.10 showed NiFe2 O4 , besides the dominant phase having the structure type of ␣-FeOOH. Shifts of IR bands at 892 and 796 cm−1 were not observed in all samples doped with Ni. For the ratio Ni/Fe = 0.10, the IR bands centered at 631 and 404 cm−1 were significantly broadened. RT M¨ossbauer spectrum of undoped ␣-FeOOH and Ni-doped ␣-FeOOH showed distributions of hyperfine magnetic fields. Bhf decreased from 35.1 T for an undoped ␣-FeOOH to 32.1 T for ␣-FeOOH containing Ni2+ ions (Ni/Fe = 0.05). The saturation of the ␣-FeOOH structure with Ni2+ ions in amounts higher than ∼5 mol% was also observed by M¨ossbauer spectroscopy. The particle size (length) of acicular ␣-FeOOH particles with a maximum in the interval 180–220 nm was slightly decreased with Ni-doping, but the distribution of the length/width ratio showed no change, having a maximum at 4–5. TEM photographs additionally showed small populations of cubic-shaped or pseudocubic particles of ∼10 nm in size for the ratio Ni/Fe = 0.05 and about 10–20 nm in size for the ratio Ni/Fe = 0.10. These particles were assigned to NiFe2 O4 . © 2005 Elsevier B.V. All rights reserved. Keywords: Ni-doping; ␣-FeOOH; XRD; FT-IR; 57 Fe M¨ossbauer; TEM
1. Introduction Goethite (␣-FeOOH) is a well-known polymorph of iron(III)-oxyhydroxide which is naturally present in various soils and sediments. It is also used as a raw material in metallurgy for the production of iron. ␣-FeOOH is usually found in the rust formed by atmospheric or “wet” corrosion of iron (steel). Traditionally, natural ␣-FeOOH was used as a pigment in paintings for many centuries, whereas nowadays it is chiefly the synthetic ␣-FeOOH pigments, with colors ranging from lemon yellow to dark brown. Synthetic ␣-FeOOH particles are also used as starting material in the production of acicular maghemite (␥-Fe2 O3 ) particles via magnetite (Fe3 O4 ) as a transition phase. ∗
Corresponding author. E-mail address: [email protected] (S. Musi´c).
0925-8388/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2005.06.011
Natural ␣-FeOOH often contains different metal cation impurities incorporated into the crystal structure as a consequence of geochemical processes in the history of earth. In some cases, these impurities can be present to a great extent and this can be utilized in metallurgy for the commercial production of impurities in the metallic state. In the laboratory various metal cations can be incorporated into the ␣-FeOOH structure, and many researchers have focused on the influence of the doping of ␣-FeOOH on its chemical, microstructural and physical properties. Stiers and Schwertmann [1] doped ␣-FeOOH with manganese up to ∼15 mol%. It was concluded that originally added Mn2+ ions oxidized to Mn3+ as confirmed by XRD measurements, which showed that the unit-cell size approached to that of groutite (␣-MnOOH). Sileo et al. [2] found that at the molar ratio ≤5.50 (expressed as [Cd] × 100/([Cd]+[Fe])), the Cd2+ ions were all incorporated into the ␣-FeOOH structure. At
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
the molar ratio 7.03, the incorporation of Cd2+ ions into ␣FeOOH was drastically reduced and Cd-substituted hematite (␣-Fe2 O3 ) was additionally formed. Natural Ni-containing ␣-FeOOH samples from Vermelho deposit (Brazilian Amazonia) [3,4] incorporated about 4 mol% of Ni. These samples also contained significant amounts of Al and Cr (1.6 and 0.7 mol%, respectively; average values). Sudakar et al. [5] prepared Ni-containing ␣-FeOOH by oxidation with air of Fe(OH)2 ·xH2 O precipitate at a near neutral pH. At the concentration Ni ≤ 5 at.% the ␣-FeOOH structure as a single phase was found, whereas at >5 at.% the additional presence of an amorphous phase of Ni(OH)2 or NiFe2 O4 was suggested. Incorporation of Cr, Mn and Ni into the ␣-FeOOH structure was investigated by X-ray absorption fine-structure spectroscopy [6]. The isomorphous substitution for Fe3+ by Cr3+ (up to 8 mol%), Mn3+ (up to 15 mol%) and Ni2+ (up to 5 mol%) were found. Balasubramanian et al. [7] used M¨ossbauer spectroscopy, XRD and TEM to investigate Cr-substituted ␣-FeOOH. The size of ␣-FeOOH particles decreased from 200 to 10 nm as the chromium substitution increased from 0 to 10.14 wt.%, while a tendency of decrease in the crystallite size was observed as the Cr concentration increased. Morales et al. [8] investigated the effect of Cr3+ , Cu2+ and Mn2+ ions on the precipitated ␣-FeOOH, with a view to better understanding the corrosion mechanism of weathering steels. Schwertmann et al. [9] prepared Crcontaining ␣-FeOOH particles by the precipitation of Fe3+ or Fe2+ ions and found that ␣-FeOOH could incorporate up to ∼10 mol% Cr. Dos Santos et al. [10] reported that up to 10 mol% of gallium could be incorporated into the ␣-FeOOH structure, and it was concluded that the main effect of Gadoping was a reduced ␣-FeOOH crystallite size. Berry et al. [11] prepared tin-doped ␣-FeOOH by the hydrothermal method. The authors found that originally added Sn2+ ions were oxidized to Sn4+ during the preparation procedure, and because the Fe3+ reduction was not noticed it was concluded that the charge balance in tin-doped goethite was achieved by the formation of cation vacancies. The 119 Sn M¨ossbauer spectra were consistent with octahedrally coordinated tin ions in the ␣-FeOOH structure and sensitive to the changes in the magnetic behavior of ␣-FeOOH resulting from Sn-doping. Researchers have much focused on the incorporation of aluminium ions into the ␣-FeOOH structure, because aluminous goethites are very important constituents of many
369
soils. Taylor and Schwertmann [12] found that the precipitation ions in the presence of aluminium ions at a near neutral pH inhibited the formation of lepidocrocite (␥-FeOOH) in favor of ␣-FeOOH. In acidic conditions the presence of polynuclear aluminium hydroxy complexes retarded the crystallization of any iron(III)-oxyhydroxide polymorph [13]. The dissolution of Al-doped ␣-FeOOH (zero to 10 mol% Al) in 6 M HCl at 24 ◦ C decreased markedly with an increase of incorporated aluminium [14]. Incorporation of Al3+ ions into the ␣-FeOOH structure increased the thermodynamic stability of ␣-FeOOH with respect to hematite (␣-Fe2 O3 ) [15]. Schulze [16] showed that in synthetic Al-substituted ␣FeOOH crystallites the c-dimension was a linear function of Al-substitution between 0 and 33 mol% Al, the a-dimension varied over the same range, whereas the b-dimension was a linear function up to 20 mol% Al. M¨ossbauer spectroscopy found important application in the investigation of Al-doped ␣-FeOOH [17–21]. In our previous work [22] we have focused on the novel method of ␣-FeOOH synthesis in a highly alkaline pH medium, using tetramethylammonium hydroxide to precipitate Fe3+ ions. The same precipitation method has been used in the present work to prepare Ni-doped ␣-FeOOH with the aim to obtain more data about the influence of nickel ions on the properties of precipitated ␣-FeOOH particles. As already pointed out, each substituted metal cation has a specific influence on the properties of ␣-FeOOH. The conditions of synthesis play an important part in the properties of substituted ␣-FeOOH particles.
2. Experimental Analytical reagents FeCl3 ·6H2 O and NiCl2 ·6H2 O supplied by Kemika and 25% (w/w) electronic grade (99.9999%) aqueous solution of tetramethylammonium hydroxide supplied by Alfa Aesar were used. Twice distilled water prepared in own laboratory was used in all experiments. The experimental conditions for the preparation of samples are given in Table 1. A predetermined volume of the TMAH solution was added to the mixed FeCl3 –NiCl2 solutions (a pure FeCl3 solution in the preparation of sample N0). The formed suspensions were vigorously shaken for approximately 10 min, then heated at 160 ◦ C. Autoclaving was performed using a general-purpose bomb by Parr (model 4744), comprising the
Table 1 Experimental conditions for the synthesis of samples N0–N5 Sample
2 M FeCl3 (ml)
[FeCl3 ] (M)
0.1 M NiCl2 (ml)
0.01 M NiCl2 (ml)
[NiCl2 ] (M)
H2 O (ml)
TMAHa (ml)
N0 N1 N2 N3 N4 N5
2 2 2 2 2 2
0.1 0.1 0.1 0.1 0.1 0.1
– – – – 2 4
– 4 8 12 – –
0 0.001 0.002 0.003 0.005 0.010
28 24 20 16 26 24
10 10 10 10 10 10
a
TMAH, tetramethylammonium hydroxide (25%, w/w).
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
370
Table 2 Ni/Fe ratio in the samples as found by PIXE analysis and nominal concentrations of FeCl3 and NiCl2 used in the synthesis of samples N1–N5 Sample
Ni/Fe as found by PIXE
[NiCl2 ] (M)
[FeCl3 ] (M)
[NiCl2 ]/[FeCl3 ]
N1 N2 N3 N4 N5
0.011 0.021 0.032 0.051 0.109
0.001 0.002 0.003 0.005 0.010
0.1 0.1 0.1 0.1 0.1
0.01 0.02 0.03 0.05 0.10
vessel and Teflon cup. After 2 h of heating the precipitates were cooled to room temperature and subsequently washed with twice distilled water using an ultra-speed centrifuge Sorvall RC2-B (up to 20,000 rpm). The precipitates were dried at 60 ◦ C, then analyzed by proton induced X-ray emission – Boˇskovi´c (PIXE) analysis facility on the site of the Ruder Institute, Zagreb. The results of PIXE analysis are given in Table 2. All samples were characterized by X-ray powder diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), 57 Fe M¨ossbauer spectroscopy and transmission electron microscopy (TEM).
XRD patterns were taken at room temperature (RT) using an automatic Philips diffractometer, model MPD 1880 (Cu K␣ radiation, graphite monochromator, proportional counter). FT-IR spectra were recorded at RT using a Perkin-Elmer spectrometer, model 2000. The FT-IR spectrometer was coupled with a personal computer loaded with IRDM (IR data manager) program to process the recorded spectra. The specimens were pressed into small discs using a spectroscopically pure KBr matrix. 57 Fe M¨ ossbauer spectra were recorded in transmission mode using standard instrumental configuration by WISSEL GmbH (Starnberg, Germany). The 57 Co in the rhodium matrix was used as a M¨ossbauer source. The velocity scale and all the data refer to the metallic ␣-Fe absorber at RT. Quantitative analysis of the spectra recorded was made using the MOSSWINN program. TEM observation was made with an electron microscope manufactured by Opton (model EM-10). Before the TEM
Fig. 1. Characteristic parts of XRD patterns of samples N0, N4 and N5, recorded at room temperature.
Fig. 2. Fourier transform infrared spectra of samples N0–N5, recorded at room temperature.
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
Fig. 3.
57 Fe
M¨ossbauer spectra of samples N0–N5, recorded at room temperature.
Fig. 4. Distribution of hyperfine magnetic fields calculated for samples N0–N5.
371
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
372
Table 3 57 Fe M¨ ossbauer parameters calculated for samples N0–N5 Sample
Spectral line
δ (mm s−1 )
Q (mm s−1 )
Bhf (T)
Γ (mm s−1 )
Area (%)
N0 N1 N2 N3 N4 N5
M M M M M M1 M2 M3 Q1 Q2
0.37 0.37 0.37 0.37 0.37 0.37 0.25 0.37 0.35 0.35
−0.26 −0.26 −0.26 −0.26 −0.26 −0.25
35.1 33.8 33.1 32.6 32.1 31.9 45.4 48.9
0.27 0.29 0.28 0.27 0.29 0.31 0.70 0.68 0.45 0.45
100 100 100 100 100 86.3 5.7 5.9 1.1 1.0
0.55 0.95
Errors: δ = ±0.01 mm s−1 , Q = ±0.01 mm s−1 , Bhf = ±0.2 T. Isomer shift is given relative to ␣-Fe.
observation, the powders were dispersed in twice distilled water using an ultrasound bath, then a drop of the dispersion was put on a copper grid previously covered with a thin polymer film.
3. Results and discussion In the present investigation, we have employed a novel procedure in the synthesis of ␣-FeOOH particles, as proposed by Krehula et al. [22]. The principle of this synthesis is based on the precipitation of Fe3+ ions at a very high pH (∼13.5 to ∼13.8) using a strong organic alkali tetramethylammonium hydroxide (TMAH). The main difference between the application of NaOH or KOH on one hand and TMAH on the other lies in the fact that in the case of TMAH the initially formed precipitate at RT is completely dissolved on strong shaking, as visible to the naked eye. After some time of aging there is homogenous reprecipitation of ␣-FeOOH. In the present case, ␣-FeOOH (sample N0) was produced by autoclaving the precipitation system, based on TMAH alkalization, at 160 ◦ C. In the precipitation systems N1–N5, the Ni2+ ions were also present. PIXE analysis of the Ni/Fe ratio (Table 2) showed a good agreement in respect to the ratio of
Fig. 5. Dependence of average hyperfine magnetic field on the Ni/Fe ratio.
nominal concentrations ([NiCl2 ]/[FeCl3 ], thus indicating a quantitative removal of Ni2+ ions from the solution and their incorporation into the solid phase. Fig. 1 shows characteristic XRD patterns of samples N0, N4 and N5. The XRD pattern of sample N0 corresponds to ␣-FeOOH as a single phase. This phase was easily detected using data available in the Powder Diffraction File [23]. ␣FeOOH crystallizes in the orthorhombic space group Pbnm (62) with the unit-cell parameters at RT [23]: a = 0.4608 nm, b = 0.9956 nm, c = 0.3021 nm. Four formula units are present in the unit-cell of ␣-FeOOH. The crystal structure of ␣FeOOH is discussed in a paper by Mathieu and Rousset [24]. Solid solutions having the structure type of ␣-FeOOH as a single phase were observed in samples N1–N4. The XRD pattern of sample N5 additionally contained diffraction lines assigned to NiFe2 O4 [23], besides the dominant phase having the structure type of ␣-FeOOH. A gradual increase in broadening of the diffraction lines of the ␣-FeOOH type phase was observed with an increased Ni2+ concentration. No measurable shifts of XRD lines of the ␣-FeOOH type phase were observed with the increased Ni2+ concentration. ¯ NiFe2 O4 belongs to a face-centered cubic space group Fd 3m (2 2 7), with the unit-cell parameter a = 0.8337 nm. On the basis of XRD measurements it can be inferred that in samples N1–N4 all nickel ions were incorporated into the ␣-FeOOH structure. FT-IR spectra of samples N0–N5 are shown in Fig. 2. The characteristic bands of ␣-FeOOH are well visible. Verdonck et al. [25] applied a method of normal coordinate analysis in the interpretation of the IR spectrum of ␣-FeOOH. For ␣-FeOOH and deuterated ␣-FeOOD the experimental and calculated vibrational frequencies were compared. The experimental IR bands at 630, 495 and 270 cm−1 were rather insensitive to deuteration and on the basis of this finding the authors [25] assigned these bands to Fe–O stretching vibrations. Cambier [26] assigned the bands at 892 and 795 cm−1 to OH bending modes in ␣-FeOOH. The interpretation of bands below 650 cm−1 was similar to that of Verdonck et al. [25], suggesting that an intense IR band around 630 cm−1 was affected by the shape of the ␣-FeOOH particles. FT-IR spectra of samples N0–N4 showed very similar features corresponding to ␣-FeOOH; however, the spectral changes occurred in
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
the FT-IR spectrum of sample N5. The IR band centered at 631 cm−1 and the shoulder at 665 cm−1 were significantly broadened for sample N5 and increased in relative intensity. A shift from 638 to 631 cm−1 was observed. The IR band centered at 404 cm−1 was strongly broadened, whereas the relative intensities of shoulders at 451 and 371 cm−1 were significantly increased in relation to IR bands at 454 and 376 cm−1 recorded for sample N0. Broadening of the bands in the FT-IR spectrum of ␣-FeOOH is the result of the incorporation of Ni2+ ions into the ␣-FeOOH structure. Stiers and Schwertmann [1] measured the shifts of IR bands at 888.1 and 791.7 cm−1 with an increased concentration of Mn3+ ions incorporated into the ␣-FeOOH structure. A significant shift of the IR bands of the same origin, δOH and γ OH , to higher wave numbers was observed for the samples with an increased incorporation of Al3+ ions into the ␣-FeOOH structure [27]. Since in the present case the IR bands of ␣-FeOOH at 892 and 796 cm−1 did not shift with the increase of Ni2+ ions incorporated into the ␣-FeOOH structure (Fig. 2), and because the corresponding shifts of XRD lines were not observed, we have extended the characterization of samples N0–N5 by 57 Fe M¨ossbauer spectroscopy with the aim to investigate possible changes in the M¨ossbauer spectrum of ␣-FeOOH upon incorporation of Ni2+ ions into the ␣-FeOOH structure. The corresponding M¨ossbauer spectra recorded at room temperature are shown in Fig. 3. The M¨ossbauer spectrum of sample N0 can be assigned to ␣FeOOH, which was found as a single phase by previous XRD and FT-IR measurements. Generally, a RT M¨ossbauer spectrum of ␣-FeOOH may vary from the central quadrupole doublet up to a well-developed sextet, with intensity ratios close
373
to 3:2:1:1:2:3. The shape of the M¨ossbauer spectrum of ␣FeOOH depends on factors, such as the degree of crystallinity, particle size distribution and the presence of impurities. The M¨ossbauer spectrum of undoped ␣-FeOOH (sample N0) has shown a sextet, which deviates from the theoretical intensity ratios; this spectrum has shown broadened spectral lines. The calculated distributions of hyperfine magnetic fields for the samples N0–N5 are given in Fig. 4. Bhf decreased from 35.1 T for sample N0 to 32.1 T for sample N4 (Table 3). With a further increase in the Ni2+ substitution there is a further decrease in Bhf . The M¨ossbauer spectrum of sample N5 showed an additional hyperfine magnetic interaction. In accordance with an earlier work [28] about nanocrystalline NiFe2 O4 , the spectrum of sample N5 was fitted for a superposition of additional two sextets corresponding to NiFe2 O4 . Two central doublets Q1 and Q2 have been introduced into this fit, which, perhaps, are due to the presence of a very small fraction of fine oxide particles and/or ferrihydrite. Fig. 5 shows the dependence of an average hyperfine magnetic field on the Ni/Fe ratio. It is noticeable that with the concentrations of Ni2+ ions higher than ∼5 mol% there is a saturation of the ␣-FeOOH structure with Ni2+ ions. Cornell et al. [29] investigated the effect of Ni2+ ions on the transformation of “amorphous” iron(III)-hydroxide into ␣-FeOOH. This transformation proceeded by the dissolution/reprecipitation mechanism, whereas the Ni2+ ions retarded the kinetics of the same process. This effect was found to be greater than the one involving either Mn or Co. A maximum of 5.5 mol% of Ni was incorporated, which is lower than what was found for Co (7 mol%) or Mn (15 mol%) [30]. It is also important to note that dissolution of Ni2+ ions
Fig. 6. (a) Particle size (length) distribution of samples N0, N1, N3 and N5; (b) length to width ratio distribution of particles in samples N0, N1, N3 and N5.
374
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
in addition to the dominant acicular ␣-FeOOH particles. The shape of these very small particles is cubic or pseudocubic, with a size between 10 and 20 nm. It can be concluded that these particles belong to NiFe2 O4 , as already detected by XRD and M¨ossbauer spectroscopy. Inspection of the TEM photograph of sample N4 showed acicular ␣-FeOOH particles and a very small population of fine particles sized ∼10 nm. In line with the previous discussion it can also be concluded that these very fine particles belong to NiFe2 O4 . After this finding, the M¨ossbauer spectrum of sample N4 was additionally fitted for one more sextet (Bhf = 45.8 T) of broad lines and small relative intensity. The addition of this sextet has improved the fitting agreement factor.
4. Conclusion
Fig. 7. TEM photographs of samples: (a) N0 and (b) N5.
from an amorphous precipitate was much slower than dissolution of Fe3+ ions. Sudakar et al. [5] found that the ␣-FeOOH particles precipitated by air oxidation of Fe(OH)2 ·xH2 O in the presence of Ni2+ ions did not change size in comparison with the undoped acicular ␣-FeOOH particles. TEM observation of samples N0–N5 was made to obtain data about a possible influence of Ni-doping on the particle size of ␣FeOOH prepared in a highly alkaline medium. Fig. 6a shows the particle size distribution of acicular ␣-FeOOH particles in samples N0, N1, N3 and N5, based on the TEM observation. The particle size (length) distribution of reference ␣-FeOOH (sample N0) showed a maximum in the range 180–220 nm, whereas for sample N5 this maximum was slightly shifted to 140–180 nm. Fig. 6b shows the distribution of the length/width ratio for the same samples (N0, N1, N3 and N5). The most probable length to width ratio was at 4–5 for all samples, with a slight exception for sample N1. On the basis of these measurements it can be inferred that Ni-doping does not significantly change the size and shape of ␣-FeOOH particles. Bearing this in mind, it can also be inferred that a decrease of Bhf with an increase in Ni-doping concentration, as measured by M¨ossbauer spectroscopy, is dominantly influenced by the formation of solid solutions. However, a small effect of the presence of superparamagnetic particles cannot be excluded. Fig. 7 shows the TEM photographs of samples N0 (Fig. 7a) and N5 (Fig. 7b). In Fig. 7b very small particles are visible
• Synthetic ␣-FeOOH was precipitated at a highly alkaline pH by precipitation from a FeCl3 solution with the addition of tetramethylammonium hydroxide and further autoclaving of the precipitation system at 160 ◦ C. Ni-doped ␣-FeOOH samples were precipitated in the same way, but in the presence of varying concentrations of NiCl2 . The ratio of Ni/Fe in the samples was checked by PIXE analysis. XRD showed that the samples were solid solutions, having the structure type of ␣-FeOOH. The broadening of diffraction lines gradually increased upon increasing the Ni-dopant. The sample having a Ni/Fe ratio of 0.10 also contained NiFe2 O4 , in addition to the dominant phase having the structure type of ␣-FeOOH. • FT-IR spectra showed very similar features up to the ratio Ni/Fe = 0.05. For a higher amount of Ni2+ ions, corresponding to the ratio Ni/Fe = 0.10, the IR band centered at 631 cm−1 with a shoulder at 665 cm−1 was significantly broadened and increased in relative intensity. The IR band centered at 404 cm−1 was strongly broadened, whereas the relative intensities of shoulders at 451 and 371 cm−1 were significantly increased in relation to IR-bands at 454 and 376 cm−1 , recorded for an undoped ␣-FeOOH. Shifts in IR bands at 892 and 796 cm−1 , corresponding to Fe-OH bending vibrations, were not observed for any concentration of Ni-dopant. • 57 Fe M¨ossbauer spectroscopy showed high sensitivity to doping of ␣-FeOOH with Ni2+ ions. The RT M¨ossbauer spectrum of an undoped ␣-FeOOH showed a sextet, which deviated from the theoretical intensity ratios, with broadened spectral lines. The distributions of hyperfine magnetic fields were calculated. Bhf decreased from 35.1 T for the undoped ␣-FeOOH to 32.1 T for a sample with the ratio Ni/Fe = 0.05. M¨ossbauer spectroscopy showed that for the concentrations of Ni2+ ions higher than ∼5 mol% a saturation of the ␣-FeOOH structure with Ni2+ ions takes place. • TEM photographs of samples up to the Ni/Fe ratio 0.10 showed the presence of acicular ␣-FeOOH particles. The particle size (length) of an undoped ␣-FeOOH showed a maximum in the range 180–220 nm. The particle
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
size was slightly decreased with Ni-doping, but the distribution of the length/width ratio did not change, and the maximum probability of this factor was at 4–5. For the ratio Ni/Fe = 0.05 TEM photographs also showed a small population of cubic or pseudocubic-shaped particles ∼10 nm in size, which were assigned to NiFe2 O4 . NiFe2 O4 particles formed for the ratio Ni/Fe = 0.10 were about 10–20 nm in size.
References [1] W. Stiers, U. Schwertmann, Geochim. Cosmochim. Acta 49 (1985) 1909–1911. [2] E.E. Sileo, P.S. Sol´ıs, C.O. Paiva-Santos, Powder Diffraction 18 (2003) 50–55. [3] M.L.M. de Carvalho e Silva, C.S.M. Partiti, J. Enzweiler, S. Petit, S.M. Netto, S.M.B. De Oliveira, Hyp. Interact. 142 (2002) 559–576. [4] M.L. Carvalho e Silva, A.Y. Ramos, H.C.N. Tolentino, J. Enzweiler, S.M. Netto, M. Do Carmo Martins Alves, Am. Mineral. 88 (2003) 876–882. [5] C. Sudakar, G.N. Subbanna, T.R.N. Kutty, J. Mater. Sci. 39 (2004) 4271–4286. [6] B. Singh, D.M. Sherman, J.F.W. Mosselmans, R.J. Gilkes, M.A. Wells, Clay Miner. 37 (2002) 639–649. [7] R. Balasubramanian, D.C. Cook, M. Yamashita, Hyp. Interact. 139/140 (2002) 167–173. [8] A.L. Morales, C.A. Barrero, F. Jaramillo, C. Arroyave, J.-M. Greneche, Hyp. Interact. 148/149 (2003) 135–144. [9] U. Schwertmann, U. Gasser, H. Sticher, Geochim. Cosmochim. Acta 53 (1989) 1293–1297. [10] C.A. Dos Santos, A.M.C. Horbe, C.M.O. Barcellos, J.B. Marimon da Cunha, Solid State Commun. 118 (2001) 449–452.
375
¨ Helgason, A. Boh´orquez, J.F. Marco, J. McManus, [11] F.J. Berry, O. E.A. Moore, S. Mørup, P.G. Wynn, J. Mater. Chem. 10 (2000) 1643–1648. [12] R.M. Taylor, U. Schwertmann, Clays Clay Miner. 26 (1978) 373–383. [13] S.S. Singh, H. Kodama, Clays Clay Miner. 42 (1994) 606–613. [14] U. Schwertmann, Clay Miner. 19 (1984) 9–19. [15] C.J. Yapp, Clays Clay Miner. 31 (1983) 239–240. [16] D.G. Schulze, Clays Clay Miner. 32 (1984) 36–44. [17] E. Murad, U. Schwertmann, Clay Miner. 18 (1983) 301–312. [18] D.D. Amarasiriwardena, E. De Grave, L.H. Bowen, S.B. Weed, Clays Clay Miner. 34 (1986) 250–256. [19] L.H. Bowen, E. De Grave, P.M.A. De Bakker, R.E. Vandenberghe, Hyp. Interact. 54 (1990) 467–472. [20] E. De Grave, P.M.A. Bakker, L.H. Bowen, R.E. Vandenberghe, Z. Pflanzenern¨ahr, Bodenk 155 (1992) 467–472. [21] E. De Grave, G.M. Da Costa, L.H. Bowen, C.A. Barrero, R.E. Vandenberghe, Hyp. Interact. 117 (1998) 245–270. [22] S. Krehula, S. Popovi´c, S. Musi´c, Mater. Lett. 54 (2002) 108– 113. [23] JCPDS PDF cards No. 29-713 and 81-0464 for ␣-FeOOH, cards No. 86-2267, 74-2081 and 10-0325 for NiFe2 O4 ; International Centre for Diffraction Data, Joint Committee on Powder Diffraction Standards, Powder Diffraction File, 1601 Park Lane, Swarthmore, PA 19081, USA. [24] F. Mathieu, A. Rousset, Philos. Mag. A 67 (1993) 533–555. [25] L. Verdonck, S. Hoste, F.F. Roelandt, G.P. Van der Kelen, J. Mol. Struct. 79 (1982) 273–279. [26] P. Cambier, Clay Miner. 21 (1986) 191–200. [27] D.G. Schulze, U. Schwertmann, Clay Miner. 22 (1987) 83–92. [28] M. Goti´c, I. Czak´o-Nagy, S. Popovi´c, S. Musi´c, Philos. Mag. Lett. 78 (1998) 193–201. [29] R.M. Cornell, R. Giovanolli, W. Schneider, J. Chem. Technol. Biotechnol. 53 (1992) 73–79. [30] R. Giovanolli, R.M. Cornell, Z. Pflanzenern¨ahr. Bodenk. 155 (1992) 455–460.
Influence of Ni-dopant on the properties of synthetic goethite Stjepko Krehula a , Svetozar Musi´c a,∗ , Stanko Popovi´c b b
a Division of Materials Chemistry, Ruder – Boˇskovi´c Institute, P.O. Box 180, HR-10002 Zagreb, Croatia Department of Physics, Faculty of Science, University of Zagreb, P.O. Box 331, HR-10002 Zagreb, Croatia
Received 6 May 2005; accepted 3 June 2005 Available online 1 August 2005
Abstract The influence of Ni-dopant on the properties of ␣-FeOOH was investigated by XRD, FT-IR, 57 Fe M¨ossbauer spectroscopy and transmission electron microscopy. ␣-FeOOH was synthesized at a highly alkaline pH by precipitation from the FeCl3 solution with the addition of tetramethylammonium hydroxide and autoclaving at 160 ◦ C. The samples doped with Ni2+ ions were precipitated in the same way, but in the presence of varying concentrations of NiCl2 . Solid solutions, having the structure of ␣-FeOOH, were observed in samples with the Ni/Fe ratio up to 0.05. Upon increasing the amount of Ni-dopant the XRD lines were gradually broadened. The sample with the ratio Ni/Fe = 0.10 showed NiFe2 O4 , besides the dominant phase having the structure type of ␣-FeOOH. Shifts of IR bands at 892 and 796 cm−1 were not observed in all samples doped with Ni. For the ratio Ni/Fe = 0.10, the IR bands centered at 631 and 404 cm−1 were significantly broadened. RT M¨ossbauer spectrum of undoped ␣-FeOOH and Ni-doped ␣-FeOOH showed distributions of hyperfine magnetic fields. Bhf decreased from 35.1 T for an undoped ␣-FeOOH to 32.1 T for ␣-FeOOH containing Ni2+ ions (Ni/Fe = 0.05). The saturation of the ␣-FeOOH structure with Ni2+ ions in amounts higher than ∼5 mol% was also observed by M¨ossbauer spectroscopy. The particle size (length) of acicular ␣-FeOOH particles with a maximum in the interval 180–220 nm was slightly decreased with Ni-doping, but the distribution of the length/width ratio showed no change, having a maximum at 4–5. TEM photographs additionally showed small populations of cubic-shaped or pseudocubic particles of ∼10 nm in size for the ratio Ni/Fe = 0.05 and about 10–20 nm in size for the ratio Ni/Fe = 0.10. These particles were assigned to NiFe2 O4 . © 2005 Elsevier B.V. All rights reserved. Keywords: Ni-doping; ␣-FeOOH; XRD; FT-IR; 57 Fe M¨ossbauer; TEM
1. Introduction Goethite (␣-FeOOH) is a well-known polymorph of iron(III)-oxyhydroxide which is naturally present in various soils and sediments. It is also used as a raw material in metallurgy for the production of iron. ␣-FeOOH is usually found in the rust formed by atmospheric or “wet” corrosion of iron (steel). Traditionally, natural ␣-FeOOH was used as a pigment in paintings for many centuries, whereas nowadays it is chiefly the synthetic ␣-FeOOH pigments, with colors ranging from lemon yellow to dark brown. Synthetic ␣-FeOOH particles are also used as starting material in the production of acicular maghemite (␥-Fe2 O3 ) particles via magnetite (Fe3 O4 ) as a transition phase. ∗
Corresponding author. E-mail address: [email protected] (S. Musi´c).
0925-8388/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2005.06.011
Natural ␣-FeOOH often contains different metal cation impurities incorporated into the crystal structure as a consequence of geochemical processes in the history of earth. In some cases, these impurities can be present to a great extent and this can be utilized in metallurgy for the commercial production of impurities in the metallic state. In the laboratory various metal cations can be incorporated into the ␣-FeOOH structure, and many researchers have focused on the influence of the doping of ␣-FeOOH on its chemical, microstructural and physical properties. Stiers and Schwertmann [1] doped ␣-FeOOH with manganese up to ∼15 mol%. It was concluded that originally added Mn2+ ions oxidized to Mn3+ as confirmed by XRD measurements, which showed that the unit-cell size approached to that of groutite (␣-MnOOH). Sileo et al. [2] found that at the molar ratio ≤5.50 (expressed as [Cd] × 100/([Cd]+[Fe])), the Cd2+ ions were all incorporated into the ␣-FeOOH structure. At
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
the molar ratio 7.03, the incorporation of Cd2+ ions into ␣FeOOH was drastically reduced and Cd-substituted hematite (␣-Fe2 O3 ) was additionally formed. Natural Ni-containing ␣-FeOOH samples from Vermelho deposit (Brazilian Amazonia) [3,4] incorporated about 4 mol% of Ni. These samples also contained significant amounts of Al and Cr (1.6 and 0.7 mol%, respectively; average values). Sudakar et al. [5] prepared Ni-containing ␣-FeOOH by oxidation with air of Fe(OH)2 ·xH2 O precipitate at a near neutral pH. At the concentration Ni ≤ 5 at.% the ␣-FeOOH structure as a single phase was found, whereas at >5 at.% the additional presence of an amorphous phase of Ni(OH)2 or NiFe2 O4 was suggested. Incorporation of Cr, Mn and Ni into the ␣-FeOOH structure was investigated by X-ray absorption fine-structure spectroscopy [6]. The isomorphous substitution for Fe3+ by Cr3+ (up to 8 mol%), Mn3+ (up to 15 mol%) and Ni2+ (up to 5 mol%) were found. Balasubramanian et al. [7] used M¨ossbauer spectroscopy, XRD and TEM to investigate Cr-substituted ␣-FeOOH. The size of ␣-FeOOH particles decreased from 200 to 10 nm as the chromium substitution increased from 0 to 10.14 wt.%, while a tendency of decrease in the crystallite size was observed as the Cr concentration increased. Morales et al. [8] investigated the effect of Cr3+ , Cu2+ and Mn2+ ions on the precipitated ␣-FeOOH, with a view to better understanding the corrosion mechanism of weathering steels. Schwertmann et al. [9] prepared Crcontaining ␣-FeOOH particles by the precipitation of Fe3+ or Fe2+ ions and found that ␣-FeOOH could incorporate up to ∼10 mol% Cr. Dos Santos et al. [10] reported that up to 10 mol% of gallium could be incorporated into the ␣-FeOOH structure, and it was concluded that the main effect of Gadoping was a reduced ␣-FeOOH crystallite size. Berry et al. [11] prepared tin-doped ␣-FeOOH by the hydrothermal method. The authors found that originally added Sn2+ ions were oxidized to Sn4+ during the preparation procedure, and because the Fe3+ reduction was not noticed it was concluded that the charge balance in tin-doped goethite was achieved by the formation of cation vacancies. The 119 Sn M¨ossbauer spectra were consistent with octahedrally coordinated tin ions in the ␣-FeOOH structure and sensitive to the changes in the magnetic behavior of ␣-FeOOH resulting from Sn-doping. Researchers have much focused on the incorporation of aluminium ions into the ␣-FeOOH structure, because aluminous goethites are very important constituents of many
369
soils. Taylor and Schwertmann [12] found that the precipitation ions in the presence of aluminium ions at a near neutral pH inhibited the formation of lepidocrocite (␥-FeOOH) in favor of ␣-FeOOH. In acidic conditions the presence of polynuclear aluminium hydroxy complexes retarded the crystallization of any iron(III)-oxyhydroxide polymorph [13]. The dissolution of Al-doped ␣-FeOOH (zero to 10 mol% Al) in 6 M HCl at 24 ◦ C decreased markedly with an increase of incorporated aluminium [14]. Incorporation of Al3+ ions into the ␣-FeOOH structure increased the thermodynamic stability of ␣-FeOOH with respect to hematite (␣-Fe2 O3 ) [15]. Schulze [16] showed that in synthetic Al-substituted ␣FeOOH crystallites the c-dimension was a linear function of Al-substitution between 0 and 33 mol% Al, the a-dimension varied over the same range, whereas the b-dimension was a linear function up to 20 mol% Al. M¨ossbauer spectroscopy found important application in the investigation of Al-doped ␣-FeOOH [17–21]. In our previous work [22] we have focused on the novel method of ␣-FeOOH synthesis in a highly alkaline pH medium, using tetramethylammonium hydroxide to precipitate Fe3+ ions. The same precipitation method has been used in the present work to prepare Ni-doped ␣-FeOOH with the aim to obtain more data about the influence of nickel ions on the properties of precipitated ␣-FeOOH particles. As already pointed out, each substituted metal cation has a specific influence on the properties of ␣-FeOOH. The conditions of synthesis play an important part in the properties of substituted ␣-FeOOH particles.
2. Experimental Analytical reagents FeCl3 ·6H2 O and NiCl2 ·6H2 O supplied by Kemika and 25% (w/w) electronic grade (99.9999%) aqueous solution of tetramethylammonium hydroxide supplied by Alfa Aesar were used. Twice distilled water prepared in own laboratory was used in all experiments. The experimental conditions for the preparation of samples are given in Table 1. A predetermined volume of the TMAH solution was added to the mixed FeCl3 –NiCl2 solutions (a pure FeCl3 solution in the preparation of sample N0). The formed suspensions were vigorously shaken for approximately 10 min, then heated at 160 ◦ C. Autoclaving was performed using a general-purpose bomb by Parr (model 4744), comprising the
Table 1 Experimental conditions for the synthesis of samples N0–N5 Sample
2 M FeCl3 (ml)
[FeCl3 ] (M)
0.1 M NiCl2 (ml)
0.01 M NiCl2 (ml)
[NiCl2 ] (M)
H2 O (ml)
TMAHa (ml)
N0 N1 N2 N3 N4 N5
2 2 2 2 2 2
0.1 0.1 0.1 0.1 0.1 0.1
– – – – 2 4
– 4 8 12 – –
0 0.001 0.002 0.003 0.005 0.010
28 24 20 16 26 24
10 10 10 10 10 10
a
TMAH, tetramethylammonium hydroxide (25%, w/w).
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
370
Table 2 Ni/Fe ratio in the samples as found by PIXE analysis and nominal concentrations of FeCl3 and NiCl2 used in the synthesis of samples N1–N5 Sample
Ni/Fe as found by PIXE
[NiCl2 ] (M)
[FeCl3 ] (M)
[NiCl2 ]/[FeCl3 ]
N1 N2 N3 N4 N5
0.011 0.021 0.032 0.051 0.109
0.001 0.002 0.003 0.005 0.010
0.1 0.1 0.1 0.1 0.1
0.01 0.02 0.03 0.05 0.10
vessel and Teflon cup. After 2 h of heating the precipitates were cooled to room temperature and subsequently washed with twice distilled water using an ultra-speed centrifuge Sorvall RC2-B (up to 20,000 rpm). The precipitates were dried at 60 ◦ C, then analyzed by proton induced X-ray emission – Boˇskovi´c (PIXE) analysis facility on the site of the Ruder Institute, Zagreb. The results of PIXE analysis are given in Table 2. All samples were characterized by X-ray powder diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), 57 Fe M¨ossbauer spectroscopy and transmission electron microscopy (TEM).
XRD patterns were taken at room temperature (RT) using an automatic Philips diffractometer, model MPD 1880 (Cu K␣ radiation, graphite monochromator, proportional counter). FT-IR spectra were recorded at RT using a Perkin-Elmer spectrometer, model 2000. The FT-IR spectrometer was coupled with a personal computer loaded with IRDM (IR data manager) program to process the recorded spectra. The specimens were pressed into small discs using a spectroscopically pure KBr matrix. 57 Fe M¨ ossbauer spectra were recorded in transmission mode using standard instrumental configuration by WISSEL GmbH (Starnberg, Germany). The 57 Co in the rhodium matrix was used as a M¨ossbauer source. The velocity scale and all the data refer to the metallic ␣-Fe absorber at RT. Quantitative analysis of the spectra recorded was made using the MOSSWINN program. TEM observation was made with an electron microscope manufactured by Opton (model EM-10). Before the TEM
Fig. 1. Characteristic parts of XRD patterns of samples N0, N4 and N5, recorded at room temperature.
Fig. 2. Fourier transform infrared spectra of samples N0–N5, recorded at room temperature.
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
Fig. 3.
57 Fe
M¨ossbauer spectra of samples N0–N5, recorded at room temperature.
Fig. 4. Distribution of hyperfine magnetic fields calculated for samples N0–N5.
371
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
372
Table 3 57 Fe M¨ ossbauer parameters calculated for samples N0–N5 Sample
Spectral line
δ (mm s−1 )
Q (mm s−1 )
Bhf (T)
Γ (mm s−1 )
Area (%)
N0 N1 N2 N3 N4 N5
M M M M M M1 M2 M3 Q1 Q2
0.37 0.37 0.37 0.37 0.37 0.37 0.25 0.37 0.35 0.35
−0.26 −0.26 −0.26 −0.26 −0.26 −0.25
35.1 33.8 33.1 32.6 32.1 31.9 45.4 48.9
0.27 0.29 0.28 0.27 0.29 0.31 0.70 0.68 0.45 0.45
100 100 100 100 100 86.3 5.7 5.9 1.1 1.0
0.55 0.95
Errors: δ = ±0.01 mm s−1 , Q = ±0.01 mm s−1 , Bhf = ±0.2 T. Isomer shift is given relative to ␣-Fe.
observation, the powders were dispersed in twice distilled water using an ultrasound bath, then a drop of the dispersion was put on a copper grid previously covered with a thin polymer film.
3. Results and discussion In the present investigation, we have employed a novel procedure in the synthesis of ␣-FeOOH particles, as proposed by Krehula et al. [22]. The principle of this synthesis is based on the precipitation of Fe3+ ions at a very high pH (∼13.5 to ∼13.8) using a strong organic alkali tetramethylammonium hydroxide (TMAH). The main difference between the application of NaOH or KOH on one hand and TMAH on the other lies in the fact that in the case of TMAH the initially formed precipitate at RT is completely dissolved on strong shaking, as visible to the naked eye. After some time of aging there is homogenous reprecipitation of ␣-FeOOH. In the present case, ␣-FeOOH (sample N0) was produced by autoclaving the precipitation system, based on TMAH alkalization, at 160 ◦ C. In the precipitation systems N1–N5, the Ni2+ ions were also present. PIXE analysis of the Ni/Fe ratio (Table 2) showed a good agreement in respect to the ratio of
Fig. 5. Dependence of average hyperfine magnetic field on the Ni/Fe ratio.
nominal concentrations ([NiCl2 ]/[FeCl3 ], thus indicating a quantitative removal of Ni2+ ions from the solution and their incorporation into the solid phase. Fig. 1 shows characteristic XRD patterns of samples N0, N4 and N5. The XRD pattern of sample N0 corresponds to ␣-FeOOH as a single phase. This phase was easily detected using data available in the Powder Diffraction File [23]. ␣FeOOH crystallizes in the orthorhombic space group Pbnm (62) with the unit-cell parameters at RT [23]: a = 0.4608 nm, b = 0.9956 nm, c = 0.3021 nm. Four formula units are present in the unit-cell of ␣-FeOOH. The crystal structure of ␣FeOOH is discussed in a paper by Mathieu and Rousset [24]. Solid solutions having the structure type of ␣-FeOOH as a single phase were observed in samples N1–N4. The XRD pattern of sample N5 additionally contained diffraction lines assigned to NiFe2 O4 [23], besides the dominant phase having the structure type of ␣-FeOOH. A gradual increase in broadening of the diffraction lines of the ␣-FeOOH type phase was observed with an increased Ni2+ concentration. No measurable shifts of XRD lines of the ␣-FeOOH type phase were observed with the increased Ni2+ concentration. ¯ NiFe2 O4 belongs to a face-centered cubic space group Fd 3m (2 2 7), with the unit-cell parameter a = 0.8337 nm. On the basis of XRD measurements it can be inferred that in samples N1–N4 all nickel ions were incorporated into the ␣-FeOOH structure. FT-IR spectra of samples N0–N5 are shown in Fig. 2. The characteristic bands of ␣-FeOOH are well visible. Verdonck et al. [25] applied a method of normal coordinate analysis in the interpretation of the IR spectrum of ␣-FeOOH. For ␣-FeOOH and deuterated ␣-FeOOD the experimental and calculated vibrational frequencies were compared. The experimental IR bands at 630, 495 and 270 cm−1 were rather insensitive to deuteration and on the basis of this finding the authors [25] assigned these bands to Fe–O stretching vibrations. Cambier [26] assigned the bands at 892 and 795 cm−1 to OH bending modes in ␣-FeOOH. The interpretation of bands below 650 cm−1 was similar to that of Verdonck et al. [25], suggesting that an intense IR band around 630 cm−1 was affected by the shape of the ␣-FeOOH particles. FT-IR spectra of samples N0–N4 showed very similar features corresponding to ␣-FeOOH; however, the spectral changes occurred in
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
the FT-IR spectrum of sample N5. The IR band centered at 631 cm−1 and the shoulder at 665 cm−1 were significantly broadened for sample N5 and increased in relative intensity. A shift from 638 to 631 cm−1 was observed. The IR band centered at 404 cm−1 was strongly broadened, whereas the relative intensities of shoulders at 451 and 371 cm−1 were significantly increased in relation to IR bands at 454 and 376 cm−1 recorded for sample N0. Broadening of the bands in the FT-IR spectrum of ␣-FeOOH is the result of the incorporation of Ni2+ ions into the ␣-FeOOH structure. Stiers and Schwertmann [1] measured the shifts of IR bands at 888.1 and 791.7 cm−1 with an increased concentration of Mn3+ ions incorporated into the ␣-FeOOH structure. A significant shift of the IR bands of the same origin, δOH and γ OH , to higher wave numbers was observed for the samples with an increased incorporation of Al3+ ions into the ␣-FeOOH structure [27]. Since in the present case the IR bands of ␣-FeOOH at 892 and 796 cm−1 did not shift with the increase of Ni2+ ions incorporated into the ␣-FeOOH structure (Fig. 2), and because the corresponding shifts of XRD lines were not observed, we have extended the characterization of samples N0–N5 by 57 Fe M¨ossbauer spectroscopy with the aim to investigate possible changes in the M¨ossbauer spectrum of ␣-FeOOH upon incorporation of Ni2+ ions into the ␣-FeOOH structure. The corresponding M¨ossbauer spectra recorded at room temperature are shown in Fig. 3. The M¨ossbauer spectrum of sample N0 can be assigned to ␣FeOOH, which was found as a single phase by previous XRD and FT-IR measurements. Generally, a RT M¨ossbauer spectrum of ␣-FeOOH may vary from the central quadrupole doublet up to a well-developed sextet, with intensity ratios close
373
to 3:2:1:1:2:3. The shape of the M¨ossbauer spectrum of ␣FeOOH depends on factors, such as the degree of crystallinity, particle size distribution and the presence of impurities. The M¨ossbauer spectrum of undoped ␣-FeOOH (sample N0) has shown a sextet, which deviates from the theoretical intensity ratios; this spectrum has shown broadened spectral lines. The calculated distributions of hyperfine magnetic fields for the samples N0–N5 are given in Fig. 4. Bhf decreased from 35.1 T for sample N0 to 32.1 T for sample N4 (Table 3). With a further increase in the Ni2+ substitution there is a further decrease in Bhf . The M¨ossbauer spectrum of sample N5 showed an additional hyperfine magnetic interaction. In accordance with an earlier work [28] about nanocrystalline NiFe2 O4 , the spectrum of sample N5 was fitted for a superposition of additional two sextets corresponding to NiFe2 O4 . Two central doublets Q1 and Q2 have been introduced into this fit, which, perhaps, are due to the presence of a very small fraction of fine oxide particles and/or ferrihydrite. Fig. 5 shows the dependence of an average hyperfine magnetic field on the Ni/Fe ratio. It is noticeable that with the concentrations of Ni2+ ions higher than ∼5 mol% there is a saturation of the ␣-FeOOH structure with Ni2+ ions. Cornell et al. [29] investigated the effect of Ni2+ ions on the transformation of “amorphous” iron(III)-hydroxide into ␣-FeOOH. This transformation proceeded by the dissolution/reprecipitation mechanism, whereas the Ni2+ ions retarded the kinetics of the same process. This effect was found to be greater than the one involving either Mn or Co. A maximum of 5.5 mol% of Ni was incorporated, which is lower than what was found for Co (7 mol%) or Mn (15 mol%) [30]. It is also important to note that dissolution of Ni2+ ions
Fig. 6. (a) Particle size (length) distribution of samples N0, N1, N3 and N5; (b) length to width ratio distribution of particles in samples N0, N1, N3 and N5.
374
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
in addition to the dominant acicular ␣-FeOOH particles. The shape of these very small particles is cubic or pseudocubic, with a size between 10 and 20 nm. It can be concluded that these particles belong to NiFe2 O4 , as already detected by XRD and M¨ossbauer spectroscopy. Inspection of the TEM photograph of sample N4 showed acicular ␣-FeOOH particles and a very small population of fine particles sized ∼10 nm. In line with the previous discussion it can also be concluded that these very fine particles belong to NiFe2 O4 . After this finding, the M¨ossbauer spectrum of sample N4 was additionally fitted for one more sextet (Bhf = 45.8 T) of broad lines and small relative intensity. The addition of this sextet has improved the fitting agreement factor.
4. Conclusion
Fig. 7. TEM photographs of samples: (a) N0 and (b) N5.
from an amorphous precipitate was much slower than dissolution of Fe3+ ions. Sudakar et al. [5] found that the ␣-FeOOH particles precipitated by air oxidation of Fe(OH)2 ·xH2 O in the presence of Ni2+ ions did not change size in comparison with the undoped acicular ␣-FeOOH particles. TEM observation of samples N0–N5 was made to obtain data about a possible influence of Ni-doping on the particle size of ␣FeOOH prepared in a highly alkaline medium. Fig. 6a shows the particle size distribution of acicular ␣-FeOOH particles in samples N0, N1, N3 and N5, based on the TEM observation. The particle size (length) distribution of reference ␣-FeOOH (sample N0) showed a maximum in the range 180–220 nm, whereas for sample N5 this maximum was slightly shifted to 140–180 nm. Fig. 6b shows the distribution of the length/width ratio for the same samples (N0, N1, N3 and N5). The most probable length to width ratio was at 4–5 for all samples, with a slight exception for sample N1. On the basis of these measurements it can be inferred that Ni-doping does not significantly change the size and shape of ␣-FeOOH particles. Bearing this in mind, it can also be inferred that a decrease of Bhf with an increase in Ni-doping concentration, as measured by M¨ossbauer spectroscopy, is dominantly influenced by the formation of solid solutions. However, a small effect of the presence of superparamagnetic particles cannot be excluded. Fig. 7 shows the TEM photographs of samples N0 (Fig. 7a) and N5 (Fig. 7b). In Fig. 7b very small particles are visible
• Synthetic ␣-FeOOH was precipitated at a highly alkaline pH by precipitation from a FeCl3 solution with the addition of tetramethylammonium hydroxide and further autoclaving of the precipitation system at 160 ◦ C. Ni-doped ␣-FeOOH samples were precipitated in the same way, but in the presence of varying concentrations of NiCl2 . The ratio of Ni/Fe in the samples was checked by PIXE analysis. XRD showed that the samples were solid solutions, having the structure type of ␣-FeOOH. The broadening of diffraction lines gradually increased upon increasing the Ni-dopant. The sample having a Ni/Fe ratio of 0.10 also contained NiFe2 O4 , in addition to the dominant phase having the structure type of ␣-FeOOH. • FT-IR spectra showed very similar features up to the ratio Ni/Fe = 0.05. For a higher amount of Ni2+ ions, corresponding to the ratio Ni/Fe = 0.10, the IR band centered at 631 cm−1 with a shoulder at 665 cm−1 was significantly broadened and increased in relative intensity. The IR band centered at 404 cm−1 was strongly broadened, whereas the relative intensities of shoulders at 451 and 371 cm−1 were significantly increased in relation to IR-bands at 454 and 376 cm−1 , recorded for an undoped ␣-FeOOH. Shifts in IR bands at 892 and 796 cm−1 , corresponding to Fe-OH bending vibrations, were not observed for any concentration of Ni-dopant. • 57 Fe M¨ossbauer spectroscopy showed high sensitivity to doping of ␣-FeOOH with Ni2+ ions. The RT M¨ossbauer spectrum of an undoped ␣-FeOOH showed a sextet, which deviated from the theoretical intensity ratios, with broadened spectral lines. The distributions of hyperfine magnetic fields were calculated. Bhf decreased from 35.1 T for the undoped ␣-FeOOH to 32.1 T for a sample with the ratio Ni/Fe = 0.05. M¨ossbauer spectroscopy showed that for the concentrations of Ni2+ ions higher than ∼5 mol% a saturation of the ␣-FeOOH structure with Ni2+ ions takes place. • TEM photographs of samples up to the Ni/Fe ratio 0.10 showed the presence of acicular ␣-FeOOH particles. The particle size (length) of an undoped ␣-FeOOH showed a maximum in the range 180–220 nm. The particle
S. Krehula et al. / Journal of Alloys and Compounds 403 (2005) 368–375
size was slightly decreased with Ni-doping, but the distribution of the length/width ratio did not change, and the maximum probability of this factor was at 4–5. For the ratio Ni/Fe = 0.05 TEM photographs also showed a small population of cubic or pseudocubic-shaped particles ∼10 nm in size, which were assigned to NiFe2 O4 . NiFe2 O4 particles formed for the ratio Ni/Fe = 0.10 were about 10–20 nm in size.
References [1] W. Stiers, U. Schwertmann, Geochim. Cosmochim. Acta 49 (1985) 1909–1911. [2] E.E. Sileo, P.S. Sol´ıs, C.O. Paiva-Santos, Powder Diffraction 18 (2003) 50–55. [3] M.L.M. de Carvalho e Silva, C.S.M. Partiti, J. Enzweiler, S. Petit, S.M. Netto, S.M.B. De Oliveira, Hyp. Interact. 142 (2002) 559–576. [4] M.L. Carvalho e Silva, A.Y. Ramos, H.C.N. Tolentino, J. Enzweiler, S.M. Netto, M. Do Carmo Martins Alves, Am. Mineral. 88 (2003) 876–882. [5] C. Sudakar, G.N. Subbanna, T.R.N. Kutty, J. Mater. Sci. 39 (2004) 4271–4286. [6] B. Singh, D.M. Sherman, J.F.W. Mosselmans, R.J. Gilkes, M.A. Wells, Clay Miner. 37 (2002) 639–649. [7] R. Balasubramanian, D.C. Cook, M. Yamashita, Hyp. Interact. 139/140 (2002) 167–173. [8] A.L. Morales, C.A. Barrero, F. Jaramillo, C. Arroyave, J.-M. Greneche, Hyp. Interact. 148/149 (2003) 135–144. [9] U. Schwertmann, U. Gasser, H. Sticher, Geochim. Cosmochim. Acta 53 (1989) 1293–1297. [10] C.A. Dos Santos, A.M.C. Horbe, C.M.O. Barcellos, J.B. Marimon da Cunha, Solid State Commun. 118 (2001) 449–452.
375
¨ Helgason, A. Boh´orquez, J.F. Marco, J. McManus, [11] F.J. Berry, O. E.A. Moore, S. Mørup, P.G. Wynn, J. Mater. Chem. 10 (2000) 1643–1648. [12] R.M. Taylor, U. Schwertmann, Clays Clay Miner. 26 (1978) 373–383. [13] S.S. Singh, H. Kodama, Clays Clay Miner. 42 (1994) 606–613. [14] U. Schwertmann, Clay Miner. 19 (1984) 9–19. [15] C.J. Yapp, Clays Clay Miner. 31 (1983) 239–240. [16] D.G. Schulze, Clays Clay Miner. 32 (1984) 36–44. [17] E. Murad, U. Schwertmann, Clay Miner. 18 (1983) 301–312. [18] D.D. Amarasiriwardena, E. De Grave, L.H. Bowen, S.B. Weed, Clays Clay Miner. 34 (1986) 250–256. [19] L.H. Bowen, E. De Grave, P.M.A. De Bakker, R.E. Vandenberghe, Hyp. Interact. 54 (1990) 467–472. [20] E. De Grave, P.M.A. Bakker, L.H. Bowen, R.E. Vandenberghe, Z. Pflanzenern¨ahr, Bodenk 155 (1992) 467–472. [21] E. De Grave, G.M. Da Costa, L.H. Bowen, C.A. Barrero, R.E. Vandenberghe, Hyp. Interact. 117 (1998) 245–270. [22] S. Krehula, S. Popovi´c, S. Musi´c, Mater. Lett. 54 (2002) 108– 113. [23] JCPDS PDF cards No. 29-713 and 81-0464 for ␣-FeOOH, cards No. 86-2267, 74-2081 and 10-0325 for NiFe2 O4 ; International Centre for Diffraction Data, Joint Committee on Powder Diffraction Standards, Powder Diffraction File, 1601 Park Lane, Swarthmore, PA 19081, USA. [24] F. Mathieu, A. Rousset, Philos. Mag. A 67 (1993) 533–555. [25] L. Verdonck, S. Hoste, F.F. Roelandt, G.P. Van der Kelen, J. Mol. Struct. 79 (1982) 273–279. [26] P. Cambier, Clay Miner. 21 (1986) 191–200. [27] D.G. Schulze, U. Schwertmann, Clay Miner. 22 (1987) 83–92. [28] M. Goti´c, I. Czak´o-Nagy, S. Popovi´c, S. Musi´c, Philos. Mag. Lett. 78 (1998) 193–201. [29] R.M. Cornell, R. Giovanolli, W. Schneider, J. Chem. Technol. Biotechnol. 53 (1992) 73–79. [30] R. Giovanolli, R.M. Cornell, Z. Pflanzenern¨ahr. Bodenk. 155 (1992) 455–460.