- Email: [email protected]
Kinetic study of goethite dehydration and the effect of aluminium substitution on the dehydrate
Thermochimica Acta 545 (2012) 20–25
Contents lists available at SciVerse ScienceDirect
Thermochimica Acta journal homepage: www.elsevier.com/locate/tca
Kinetic study of goethite dehydration and the effect of aluminium substitution on the dehydrate Haibo Liu a,b , Tianhu Chen a,∗ , Qiaoqin Xie a , Xuehua Zou a , Chengsong Qing a , Ray L. Frost b,∗∗ a b
School of Resource and Environmental Engineering, Hefei University of Technology, China School of Chemistry, Physics and Mechanical Engineering, Science and Engineering Faculty, Queensland University of Technology, Australia
a r t i c l e
i n f o
Article history: Received 30 April 2012 Received in revised form 18 June 2012 Accepted 19 June 2012 Available online 11 July 2012 Keywords: Goethite Al-substituted goethite TG DTG Adsorption activation energy
a b s t r a c t Goethite and Al-substituted goethite were synthesized and were characterized using XRD and XRF. The kinetic study of goethite dehydrate was investigated by TG and DTG at different heating rates (2, 5, 10, 15, 20 ◦ C/min) and the effect of Al substitution for Fe on dehydrate was studied. The results showed that two types of absorbed water with the same Ed values of 3.4, 6.2 kJ/mol were confirmed on goethite and Alsubstituted goethite. Three types of hydroxyl units were proved, one being on the surface and the other two being in the structure of goethite. The substitution of Al for Fe in the structure of goethite decreases the desorption rate of hydroxyl, increases the dehydroxylation temperature, broadens the desorption peaks in DTG curves, and improves the Ed values from 19.4, 20.4, 26.1 kJ/mol to 21.6, 30, 33.6 kJ/mol when Al substitution comes to 9.1%. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Goethite (␣-FeOOH) occurs in soils, rocks and throughout the various compartment of the global ecosystem and is frequently used as an important raw material to produce magnetic iron oxide and pigments [1–5]. The structure of goethite is orthorhombic and each iron atom has six octahedrally distributed oxygen and hydroxyl neighbours in an almost perfect hexagonal close-packing with the 6-folded coordinated Fe atoms occupying the octahedral position [6]. The Fe atoms are arranged in double row to form what can be described as double chains of octahedra running the length of the c-axis, while the bonding between the double chains consists of relatively weak hydrogen bonding directed through apical oxygen ions along the a-axis [7]. In this case, non-stoichiometric hydroxyl units incorporated into the goethite structure during crystal growth were found and researched widely [8–11]. Excess OH or non-stoichiometric OH occurs in goethite and mostly strongly affects the unit cell a dimension [9,10,12] and reduce goethite dehydroxylation temperature [9,13]. Besides, goethite formed from Fe2+ systems usually contains more excess OH and has a lower dehydroxylation temperature than goethite synthesized from Fe3+ systems.
∗ Corresponding author. Tel.: +86 05512903990. ∗∗ Corresponding author. Tel.: +61 7 3138 2407; fax: +61 7 3138 1804. E-mail addresses: [email protected], [email protected] (T. Chen), [email protected] (R.L. Frost). 0040-6031/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tca.2012.06.024
Furthermore, the 6-folded coordinated Fe can be replaced by Al, Cr, Co, etc., among which the substitution of Al for Fe was well demonstrated and has been proved to occur universally in natural goethite [14–21]. Al substitution amount for Fe differs from different natural goethite ranging from zero to 33 mol% [2,22,23]. And the effect of Al substitution on the crystal structure of goethite has been researched for decades [24–26]. The dehydroxylation of goethite can be influenced by Al substitution, particle size and structural defects. The report [22] showed that the dehydroxylation temperature and specific surface area of goethite increased and unit cell dimensions of goethite decreased as Al-substitution increased. These researches [9,13,22,27] showed a larger amount of non-stoichiometric hydroxyl units is associated with a higher aluminum substitution. A shift to a higher wavenumber of bending and hydroxyl stretching vibrations is attributed to the effects of aluminum substitution associated with non-stoichiometric hydroxyl units on the a–b plane relative to the b–c plane of goethite (Table 1). Frost et al. reported [28] two endotherms found at 75 and 225 ◦ C for synthetic goethite and attributed to the loss of water and the dehydroxylation of the goethite, respectively, and four peaks were found in the DTA curve of goethite. In addition, evidence for the existence of two types of surface hydroxyl groups has also been reported [29,30]. However, no report about the kinetic study of dehydration and dehydroxylation for goethite and the effect of Al substitution was forthcoming. Therefore, kinetic study about dehydration and dehydroxylation for goethite the effect of Al substitution on that is investigated and a basic message about the types of hydroxyl in goethite is provided by the characterization of TG and DTG in this paper.
H. Liu et al. / Thermochimica Acta 545 (2012) 20–25
21
6000
Table 1 d-Spacings of reflections for goethite and substituted goethite. 021
111
ICSD SG SAG-3.6 SAG-6.9 SAG-9.1 SAG-9.9 SAG-11.5
4.957 4.978 4.968 4.959 4.956 4.953 4.948
4.158 4.159 4.190 4.174 4.174 4.177 4.175
3.371 3.388 3.382 3.371 3.371 3.369 3.371
2.680 2.696 2.691 2.683 2.682 2.680 2.680
2.568 2.585 2.580 2.573 2.571 2.570 2.568
2.438 2.452 2.449 2.441 2.440 2.439 2.438
2. Experimental 2.1. Synthesis of goethite and Al-substituted goethite 2.1.1. Preparation of goethite 100 g of Fe(NO3 )3 ·9H2 O and 400 mL deionized water were placed in a 1000 mL beaker. Fe(NO3 )3 ·9H2 O was dissolved by stirring continuously. After dissolution, KOH with a concentration of 5 mol/L and a concentration of 0.1 mol/L were used to regulate the pH at 13.9 ± 0.1 pH units. After finishing the above, the beaker was sealed with preservative film to prevent evaporating and then put into thermotank controlled at 70 ◦ C. After 6 days, the beaker was taken out to removal of redundant KOH by centrifugation several times till the pH came to neutral. After centrifugation, the deposits were dried at 105 ◦ C, cooled to room temperature and ground to obtain powder for further characterization. The sample is labelled as synthetic goethite (SG).
2000 6000
2.2. Characterization 2.2.1. X-ray diffraction The synthetic goethite and Al-substituted goethite were prepared as pressed powders and mounted in stainless steel sample holders. The powder X-ray diffraction (XRD) patterns were recorded on a Philips PANalytical X’Pert PRO diffractometer using Cu K␣ radiation operating at 40 kV and 40 mA. XRD diffraction patterns were taken in the range of 10–70◦ at a scan speed of 2◦ min−1 with 0.5◦ divergence slit size. Phase identification was carried out by comparison with those included in the Inorganic Crystal Structure Database (ICSD). 2.2.2. Thermogravimetry Thermogravimetric analysis of the synthetic goethite and synthetic Al-substituted goethite were obtained by using TA Instruments Inc. Q500 TGA operating at different heating rate (2, 5, 10, 15, 20 ◦ C/min) from 38 ± 1 to 1000 ◦ C in a high-purity nitrogen atmosphere with a flowing rate for balance (40 cm3 /min) and a flowing rate for sample (60 cm3 /min). Approximately 38 ± 0.5 mg of finely ground dried sample was heated in an open platinum crucible.
SAG-9.9
4000 2000 6000
SAG-9.1
4000 2000 6000
SAG-6.9
4000 2000 6000
SAG-3.6
4000 2000 6000
SG
4000 2000 1500 1000 500 0 10
2.1.2. Preparation of Al-substituted goethite 9.378 g of Al(NO3 )3 ·9H2 O and 90.905 g Fe(NO3 )3 ·9H2 O were placed in a 1000 mL beaker and then 400 mL deionized water were put into the beaker. The Al(NO3 )3 ·9H2 O and Fe(NO3 )3 ·9H2 O were dissolved by stirring continuously. After dissolution, KOH with a concentration of 5 mol/L and a concentration of 0.1 mol/L were used to regulate the pH at 13.9 ± 0.1 pH units. The following steps are same as above. The sample is got with an Al substitution of 9.1 mol% in fact (10 mol% Al substitution in theory, Al/(Al + Fe)) and labelled as synthetic Al-substituted goethite (SAG-9.1). Alsubstituted goethite with different Al substitution (3.6, 6.9, 9.9, 11.5 mol%) is got by changing the mass of Al(NO3 )3 ·9H2 O. The Al substitution amount is calculated based on the results of the chemical composition measured on a Shimadzu XRF-1800 with Rh radiation.
SAG-11.5
4000
20
30
130 021 111
130
110
120
120
110
020
020
Intensity/Cps
Sample
ICSD(96-900-2159)
40
50
60
70
2θ/(º) Fig. 1. XRD patterns of synthetic goethite and Al-substituted goethite.
DTG component analysis was undertaken using the Jandel “Peakfit” software package (Jandel Scientific, CA, USA) 3. Results and discussion 3.1. XRD Fig. 1 shows the XRD patterns of synthetic goethite, Alsubstituted goethite with different Al substitution and goethite from ICSD (96-900-2159). These reflections ((0 2 0), (1 1 0), (1 2 0), (1 3 0), (0 2 1), (1 1 1), etc.) are observed and indentified as goethite compared with the ICSD (96-900-2159). In addition, the variation of d-spacings of goethite reflections after the occurrence of Al substituting Fe in the structure of goethite is presented (Table 2). The d-spacings are obtained from the XRD patterns using the software of X’Pert HighScore Plus. As is shown in Fig. 2, all d spacings of goethite reflection derived from ICSD (96-900-2159) are lower than that of the synthetic goethite in the experiment, which should be ascribed to the different preparation methods. What is more important, all d-spacings decrease slightly after the addition of Al(NO3 )3 ·9H2 O during the preparation of goethite, which is attributed to the smaller Al3+ ion radius than that of Fe3+ . This is good agreement with the report [4]. Schulze [4] has reported that diffraction peaks became broad and shifted to high diffraction angle (namely smaller d-spacings) with the increase of Al substitution in the structure of goethite. Moreover, all d-spacings of selected reflections decrease with an increase of Al substitution except (1 1 0). As shown in Fig. 2, linear relationship between unit cell dimension (UCD) and unit cell volume (UVC) and Al substitution is negatively related. All the data is got from XRD patterns by analysis of the Rietveld refinement method. It is worth to note that the correlation coefficient between a/b dimension and Al
22
H. Liu et al. / Thermochimica Acta 545 (2012) 20–25
Table 2 Component analysis of the mass loss of the TG and DTG for goethite and Al-substituted goethite. Sample
Heating rate (◦ C/min)
Step 1 dehydration
Step 2 dehydration
◦
Mass loss (%)
T ( C)
SG
2 5 10 15 20
1 1 1 0.8 1
38 38 38 38 38
± ± ± ± ±
SAG-9.1
2 5 10 15 20
1.2 1.3 1.1 1.2 1.3
38 38 38 38 38
± ± ± ± ±
Step 3 dehydroxylation ◦
◦
◦
Step 4 Total mass dehydroxylation loss (%) Mass loss (%)
Tm1 ( C)
Tm2 ( C)
Mass loss (%)
Tm3 ( C)
Tm4 ( C)
Tm5 ( C)
1 1 1 1 1
1.3 1.3 1.3 1.3 1.3
53.9 56.1 61.7 67.3 67.3
80.9 85.7 90.5 96.2 101
10.6 10.5 10.5 10.7 10.6
177.5 183.5 195.5 203 206
212.4 223.6 241.2 250 250
242.7 258 272.4 279.7 286.1
1.6 1.6 1.5 1.6 1.5
14.5 14.4 14.3 14.4 14.4
1 1 1 1 1
1.3 1.4 1.2 1.2 1.4
53.7 57.7 66.5 66.5 66.5
84.9 93.8 102.6 105 105
10.1 10.1 10.1 10.2 10.1
196.3 216.4 223.6 225.2 230
256.4 270 281.3 291.7 299
288.5 306.9 320.5 327.7 338.1
1.1 1.3 1.3 1.1 1.3
13.7 14.1 13.7 13.7 14.1
3.2. TG and DTG of goethite The TG and DTG curves of synthetic goethite with different heating rate (2, 5, 10, 15, 20 ◦ C/min) are shown in Fig. 3. The results of the component analysis of the TG and DTG curves are provided in Table 2. Four mass loss steps are observed in all TG curves and five peaks after fitting are found in all DTG curves. An average value of 14.4% total mass loss is observed over the ambient to 1000 ◦ C range, as is displayed by the TG curves. An about 1% mass loss is found which is assigned to the superficial adsorbed water and can be desorbed at 38 ± 1 ◦ C because all samples experience isothermal for 10 min at 38 ± 1 ◦ C before heating. A steady 1.3% mass loss are observed, which are attributed to the absorbed water. Combining with the DTG curves, the second step should be divided into two dehydration steps. Namely, there are three types of adsorbed water on the surface of goethite.
A third mass loss with an average value of 10.6% is observed at slightly high temperature which is attributed to the dehydroxylation. According to the DTG curves, three peaks can be found in the third step which should be assigned to different dehydroxylation. That is to say, there are three types of hydroxyl in the structure or on the surface of goethite. However, the theoretical mass loss for the conversion of goethite to hematite according to the equation (2FeOOH → Fe2 O3 + H2 O) should be precisely 10.1%. Therefore, the mass loss of dehydroxylation for the synthetic goethite is higher than the theoretical value. In addition, the evidence for the existence of two types of surface hydroxyl group has been reported by Parfitt [29], Russel [30] and Rochester [31]. Therefore, the third peak should be ascribed to desorption of hydroxylation which has a relative strong interaction with broken bond on the surface of goethite. Another evidence for this speculation is the less peak area than that of the latter two dehydroxylations. Moreover, excess OH occurring in goethite has been reported [10,12]. The last about 1.6% mass loss is observed which is contributed to desorption of the remanent hydroxyl. Comparison with the five DTG curves, every DTG curve has five peaks after fitting and these peaks temperature increases with an increasing heating rate. Therefore, it is suggested that the more rapid the heating rate, the higher peak temperature. Obviously, five sub-steps including dehydration and dehydroxylation occur 3.024
9.96 9.94
2
2
R =0.992
9.92 9.90 9.88 0
2
4
6
8
R =0.994
3.016
b/A
a/A
◦
Mass loss (%)
substitution is considerably higher that of c dimension. a-Axis and ˚ respectively, b-axis dimension decreases by 0.066 A˚ and 0.016 A, when Al substitution increases from 0 to 11.5 mol%. It indicates that a-axis is more sensitive than other axes which agrees well with the report [8]. This phenomenon also proves the occurrence of Al substitution for Fe in the structure of goethite used in the experiments.
10
Al substitution/mol%
3.008
3.000
12
4.63
0
2
4
6
8
10
Al substitution/mol%
12
139.2 2
R =0.739
4.62
2
R =0.953
138.4
UVC/A3
c/A
◦
4.61 4.60
137.6 136.8
4.59 0
2
4
6
8
10
Al substitution/mol%
12
0
2
4
6
8
10
Al substitution/mol%
Fig. 2. Linear relationship between UCD and UCV and Al substitution.
12
H. Liu et al. / Thermochimica Acta 545 (2012) 20–25
23
Fig. 3. TG and DTG of (a–e) goethite with an elevated heating rate 2, 5, 10, 15, 20 ◦ C/min.
in goethite regardless of the heating rate. As is well-known, much more energy would be needed to overcome energy barrier to accomplish the step with high peak temperature. Therefore, based on the research [32], the following formula is used to calculate desorption activation energy of every step including dehydration and dehydroxylation, 2 ln Tm − ln ˇ = Ed /(RTm ) + ln(Ed /AR), where Tm , ˇ, Ed , A, R represent peak temperature, heating rate, adsorption activation energy, pre-exponential factor, constant, respectively. The linear relationship between 2 ln Tm − ln ˇ and 1/Tm /103 is obtained, where 2 ln Tm − ln ˇ is labelled as y-axis and 1/Tm /103 is labelled as x-axis, as is displayed in Fig. 4(a). Then the Ed and A calculated by the slope and intercept are shown in Table 3. The 3.4, 6.2 kJ/mol of Ed
corresponding to adsorbed water is substantially lower than 19.4, 20.4, 26.1 kJ/mol corresponding to the hydroxyl in the structure or on the surface of goethite. 3.3. TG and DTG of substituted-goethite The TG and DTG curves of synthetic Al-substituted goethite with different heating rate (2, 5, 10, 15, 20 ◦ C/min) are shown in Fig. 5. The results of the component analysis of the DTG curves are provided in Table 2. Four mass loss steps are observed in all TG curves and five peaks after fitting are found in all DTG curves, which is the same as that of SG.
Table 3 The results of Ed (kJ/mol) and A for SG and SAG-9. Sample
Ed (kJ/mol)A
Tm1
Tm2
Tm3
Tm4
Tm5
SG
Ed A
3.4 740
6.2 282
19.4 9 × 104
20.4 1.2 × 104
26.1 4.5 × 104
SAG-9.1
Ed A
3.4 271
6.2 355
21.6 194
30 98
33.6 71
24
H. Liu et al. / Thermochimica Acta 545 (2012) 20–25
Fig. 4. The relationship between 2 ln Tm − ln ˇ and 1/Tm /103 for SG (a) and SAG-9.1 (b).
The attribution of the four steps according to TG curves and five peaks based on DTG curves for SAG-9 are same as that of SG. However, these temperatures of desorption peaks increase when Al substitutes for Fe in the structure of goethite compared with
that of SG, especially these temperatures of dehydroxylation. These temperatures corresponding to dehydroxylation for SAG-9 increase by about 25, 40, 50 ◦ C, respectively, in comparison to that of SG. This is good agreement with the report [22,28]. The research of Ray
Fig. 5. TG and DTG of (a–e) Al-substituted goethite with an elevated heating rate 2, 5, 10, 15, 20 ◦ C/min.
H. Liu et al. / Thermochimica Acta 545 (2012) 20–25
showed that the temperature of mass loss steps increases regularly with the increase in %Al substitution. A conclusion from the report of Ruan indicated that goethite synthesized from ferrous system altered to hematite with DTGA maximum increasing from 236 ◦ C to 273 ◦ C for 0–30.1 mol% Al-substitution. The reason evidently should be ascribed to the smaller atomic radius of Al than that of Fe, which results in a stronger coordination as OH− is linked via the bond of Al O to Al3+ . To investigate the effect of Al substitution on desorption activate energy of dehydration and dehydroxylation, the above formula is used to calculate the Ed . The linear relationship between 2 ln Tm − ln ˇ and 1/Tm /103 is illustrated in Fig. 4(b) and calculated results are presented in Table 3. The both have a positive relativity for dehydration of SAG-9 same to that of SG. As is shown in Table 3, the same Ed and the different A are observed for the two types of goethite. However, the Ed for dehydroxylation of SAG-9 increases from 19.4, 20.4, 26.1 kJ/mol to 21.6, 30, 33.6 kJ/mol when compared with that of SG. Such an effect is in harmony with the results of changes in the temperature of desorption peaks as is observed from DTG curves. In a contrary, the values of A for SAG9 decrease dramatically compared with that of SG. The A partly denotes the desorption rate. Therefore, the change of A values is positively associated with the changes of DTG curves. As is indicated in Fig. 4, all adsorption peaks become broad and shift to high temperature. 4. Conclusions Three types of adsorbed water and three types of hydroxyl units are observed using TG and DTG for the synthetic goethite and Alsubstituted goethite. The desorption of the first adsorbed water occurs easily in spite of the temperature of 38 ± 1 ◦ C. The values of Ed for the other two types of adsorbed water are 3.4, 6.2 kJ/mol for goethite and aluminous goethite. However, the Ed values of three types of hydroxyl for Al-substituted goethite have an increase from 19.4, 20.4, 26.1 kJ/mol to 21.6, 30, 33.6 kJ/mol when compared with that of goethite. The first hydroxyl should be ascribed to be linked to the surface broken bond and the latter two hydroxyls are assigned to locate in the structure of goethite. The A values of dehydroxylation for aluminous goethite decrease dramatically compared with that of goethite. The substitution of Al for Fe in the structure makes desorption peaks of hydroxyl become broad and shift to high temperature due to the slow desorption rate. The results provide a baseline kinetic information on the dehydration and dehydroxylation of goethite and the effect of Al substitution on that. The way can be extrapolated to the relative study of minerals on the types of hydroxyl. Acknowledgements This study was financially supported by Natural Science Foundation of China (no. 41172048, no. 41072035, no. 41130206) and PhD Programs Foundation of Ministry of Education of China (no. 20110111110003). The authors appreciate the financial support and the School of Chemistry, Physics and Mechanical Engineering, Science and Engineering Faculty, Queensland University of Technology, for providing the infrastructure to this research. References [1] D. Walter, G. Buxbaum, W. Laqua, The mechanism of the thermal transformation from goethite to hematite, J. Therm. Anal. Calorim. 63 (2001) 733.
25
[2] R.M. Cornell, U. Schwertmann, The Iron Oxides: Structure, Properties, Reactions, Occurrences and Uses, Second Editor, Weinheim, 2003, p. 509. [3] C. Zee, D.R. Roberts, D.G. Rancourt, C. Slomp, C.P. Slomp, Nanogoethite is the dominant reactive oxyhydroxide phase in lake and marine sediments, Geology 31 (2003) 993–996. [4] S.K. Ghose, G.A. Waychunas, et al., Hydrated goethite (␣-FeOOH) interface structure: ordered water and surface functional groups, Geochim. Cosmochim. Acta 74 (7) (2010) 1943–1953. [5] V. Morozov, S. Vasilev, Effect of isomorphic substitutions on the Mössbauer and magnetic parameters of goethite, Eurasian Soil Sci. 43 (7) (2010) 795– 801. [6] G. Busca, N. Cotena, et al., Infrared spectroscopic study of micronised geothite, Mater. Chem. 3 (4) (1978) 271–283. [7] F.J. Ewing, The crystal structure of diaspore, J. Chem. Phys. 3 (1935) 203– 207. [8] H.D. Ruan, R.J. Gilkes, Dehydroxylation of aluminous goethite: unit cell dimensions, crystal size and surface area, Clays Clay Miner. 43 (1995) 196– 211. [9] D.G. Schulze, U. Schwertmann, The influence of aluminium on iron oxides. X. properties of Al-substituted goethites, Clay Miner. 19 (1984) 521–539. [10] D.G. Schulze, U. Schwertmann, The influence of aluminium on iron oxides. XIII. Properties of goethites synthesised in 0–3 M KOH at 25 ◦ C, Clay Miner. 22 (1987) 83–92. [11] D.G. Lewis, U. Schwertmann, The effect of [OH] on the goethite produced from ferrihydrite under alkaline conditions, J. Colloid Interface Sci. 78 (1980) 543–553. [12] D.G. Schulze, The influence of aluminium on iron oxides. VIII. Unit-cell dimensions of Al-substituted goethites and estimation of Al from them, Clays Clay Miner. 32 (1984) 36–44. [13] C.J. Goss, The kinetics and reaction mechanism of the goethite to hematite transformation, Mineral. Mag. 51 (1987) 437–451. [14] K. Norrish, R.M. Taylor, The isomorphous replacement of iron by aluminium in soil goethites, J. Soil Sci. 12 (1961) 294–306. [15] C. Janot, H. Gibert, X. de Gramont, R. Biais, Étude des substitutions Al–Fe dans des roches latéritiques, Bull. Soc. France Mineral. Cristallogr. 94 (1971) 367–380. [16] B.G. Davey, J.D. Russell, M.J. Wilson, Iron oxide and clay minerals and their relation to colours of red and yellow Podzolic soils near Sydney, Australia, Geoderma 14 (1975) 125–138. [17] D. Nahon, C. Janot, A.M. Karpoff, H. Paquet, Y. Tardy, Mineralogy, petrography and structures ofiron crusts (ferricretes) developed on sandstones in the western part of Senegal, Geoderma 19 (1977) 263–277. [18] J.M. Bigham, D.C. Golden, L.H. Bowen, S.W. Buol, S.B. Weed, Iron oxide mineralogy of well-drained ultisols and oxisols. I. Characterization of iron oxides in soil clays by Mössbauer spectroscopy, X-ray diffractometry, and selected chemical techniques, Soil Sci. Soc. Am. J. 42 (1978) 816–825. [19] E. Mendelovici, Sh. Yariv, R. Villalba, Aluminum-beating goethite in Venezuelan laterites, Clays Clay Miner. 27 (1979) 368–372. [20] J. Torrent, U. Schwertmann, D.G. Schulze, Iron oxide mineralogy of some soils of two river terrace sequences in Spain, Geoderma 23 (1980) 191–208. [21] R.W. Fitzpatrick, U. Schwertmann, Al-substituted goethite—an indicator of pedogenic and other weathering environments in South Africa, Geoderma 27 (1981) 335–347. [22] H.D. Ruan, R.J. Gilkes, Dehydroxylation of aluminous goethite; unit cell dimensions, crystal size and surface area, Clays Clay Miner. 43 (2) (1995) 196–211. [23] K. Norrish, R.M. Taylor, The isomorphous replacement of iron by aluminum in soil goethites, J. Soil Sci. 12 (1961) 294–306. [24] K. Jónás, K. Solymár, Preparation X-ray, derivatographic and infrared study of aluminium-substituted goethites, Acta Chim. Acad. Sci. Hung. 66 (1970) 383–394. [25] M.B. Fey, J.B. Dixon, Synthesis and properties of poorly crystalline hydrated aluminous goethites, Clays Clay Miner. 29 (1981) 91–100. [26] R.C. Mackenzie, G. Berggren, Oxides and hydroxides of higher valence elements, in: R.C. Mackenzie (Ed.), Differential Thermal Analysis, Academic Press, London, 1970, pp. 271–302. [27] H.D. Ruan, R.L. Frost, J.T. Kloprogge, L. Duong, Infrared spectroscopy of goethite dehydroxylation. II. Effect of aluminium substitution on the behaviour of hydroxyl units, Spectrochim. Acta A 58 (2002) 479–491. [28] R.L. Frost, Z. Ding, H.D. Ruan, Thermal analysis of goethite, relevance to Australian indigenous art, J. Therm. Anal. Calorim. 71 (2003) 783–797. [29] R.L. Parfitt, J.D. Russell, et al., Confirmation of the surface structures of goethite (␣-FeOOH) and phosphated goethite by infrared spectroscopy, J. Chem. Soc., Faraday Trans. 1: Phys. Chem. Condens. Phases 72 (1976) 1082–1087. [30] J.D. Russell, R.L. Parfitt, et al., Surface structures of gibbsite goethite and phosphated goethite, Nature 248 (5445) (1974) 220–221. [31] C.H. Rochester, S.A. Topham, Infrared study of surface hydroxyl groups on goethite, J. Chem. Soc., Faraday Trans. 1: Phys. Chem. Condens. Phases 75 (1979) 591–602. [32] R.J. Cvetanovic, Y. Amenomiya, Application of a temperature-programmed desorption techniques to catalyst studies, Adv. Catal. 17 (1967) 103–149.
Contents lists available at SciVerse ScienceDirect
Thermochimica Acta journal homepage: www.elsevier.com/locate/tca
Kinetic study of goethite dehydration and the effect of aluminium substitution on the dehydrate Haibo Liu a,b , Tianhu Chen a,∗ , Qiaoqin Xie a , Xuehua Zou a , Chengsong Qing a , Ray L. Frost b,∗∗ a b
School of Resource and Environmental Engineering, Hefei University of Technology, China School of Chemistry, Physics and Mechanical Engineering, Science and Engineering Faculty, Queensland University of Technology, Australia
a r t i c l e
i n f o
Article history: Received 30 April 2012 Received in revised form 18 June 2012 Accepted 19 June 2012 Available online 11 July 2012 Keywords: Goethite Al-substituted goethite TG DTG Adsorption activation energy
a b s t r a c t Goethite and Al-substituted goethite were synthesized and were characterized using XRD and XRF. The kinetic study of goethite dehydrate was investigated by TG and DTG at different heating rates (2, 5, 10, 15, 20 ◦ C/min) and the effect of Al substitution for Fe on dehydrate was studied. The results showed that two types of absorbed water with the same Ed values of 3.4, 6.2 kJ/mol were confirmed on goethite and Alsubstituted goethite. Three types of hydroxyl units were proved, one being on the surface and the other two being in the structure of goethite. The substitution of Al for Fe in the structure of goethite decreases the desorption rate of hydroxyl, increases the dehydroxylation temperature, broadens the desorption peaks in DTG curves, and improves the Ed values from 19.4, 20.4, 26.1 kJ/mol to 21.6, 30, 33.6 kJ/mol when Al substitution comes to 9.1%. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Goethite (␣-FeOOH) occurs in soils, rocks and throughout the various compartment of the global ecosystem and is frequently used as an important raw material to produce magnetic iron oxide and pigments [1–5]. The structure of goethite is orthorhombic and each iron atom has six octahedrally distributed oxygen and hydroxyl neighbours in an almost perfect hexagonal close-packing with the 6-folded coordinated Fe atoms occupying the octahedral position [6]. The Fe atoms are arranged in double row to form what can be described as double chains of octahedra running the length of the c-axis, while the bonding between the double chains consists of relatively weak hydrogen bonding directed through apical oxygen ions along the a-axis [7]. In this case, non-stoichiometric hydroxyl units incorporated into the goethite structure during crystal growth were found and researched widely [8–11]. Excess OH or non-stoichiometric OH occurs in goethite and mostly strongly affects the unit cell a dimension [9,10,12] and reduce goethite dehydroxylation temperature [9,13]. Besides, goethite formed from Fe2+ systems usually contains more excess OH and has a lower dehydroxylation temperature than goethite synthesized from Fe3+ systems.
∗ Corresponding author. Tel.: +86 05512903990. ∗∗ Corresponding author. Tel.: +61 7 3138 2407; fax: +61 7 3138 1804. E-mail addresses: [email protected], [email protected] (T. Chen), [email protected] (R.L. Frost). 0040-6031/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tca.2012.06.024
Furthermore, the 6-folded coordinated Fe can be replaced by Al, Cr, Co, etc., among which the substitution of Al for Fe was well demonstrated and has been proved to occur universally in natural goethite [14–21]. Al substitution amount for Fe differs from different natural goethite ranging from zero to 33 mol% [2,22,23]. And the effect of Al substitution on the crystal structure of goethite has been researched for decades [24–26]. The dehydroxylation of goethite can be influenced by Al substitution, particle size and structural defects. The report [22] showed that the dehydroxylation temperature and specific surface area of goethite increased and unit cell dimensions of goethite decreased as Al-substitution increased. These researches [9,13,22,27] showed a larger amount of non-stoichiometric hydroxyl units is associated with a higher aluminum substitution. A shift to a higher wavenumber of bending and hydroxyl stretching vibrations is attributed to the effects of aluminum substitution associated with non-stoichiometric hydroxyl units on the a–b plane relative to the b–c plane of goethite (Table 1). Frost et al. reported [28] two endotherms found at 75 and 225 ◦ C for synthetic goethite and attributed to the loss of water and the dehydroxylation of the goethite, respectively, and four peaks were found in the DTA curve of goethite. In addition, evidence for the existence of two types of surface hydroxyl groups has also been reported [29,30]. However, no report about the kinetic study of dehydration and dehydroxylation for goethite and the effect of Al substitution was forthcoming. Therefore, kinetic study about dehydration and dehydroxylation for goethite the effect of Al substitution on that is investigated and a basic message about the types of hydroxyl in goethite is provided by the characterization of TG and DTG in this paper.
H. Liu et al. / Thermochimica Acta 545 (2012) 20–25
21
6000
Table 1 d-Spacings of reflections for goethite and substituted goethite. 021
111
ICSD SG SAG-3.6 SAG-6.9 SAG-9.1 SAG-9.9 SAG-11.5
4.957 4.978 4.968 4.959 4.956 4.953 4.948
4.158 4.159 4.190 4.174 4.174 4.177 4.175
3.371 3.388 3.382 3.371 3.371 3.369 3.371
2.680 2.696 2.691 2.683 2.682 2.680 2.680
2.568 2.585 2.580 2.573 2.571 2.570 2.568
2.438 2.452 2.449 2.441 2.440 2.439 2.438
2. Experimental 2.1. Synthesis of goethite and Al-substituted goethite 2.1.1. Preparation of goethite 100 g of Fe(NO3 )3 ·9H2 O and 400 mL deionized water were placed in a 1000 mL beaker. Fe(NO3 )3 ·9H2 O was dissolved by stirring continuously. After dissolution, KOH with a concentration of 5 mol/L and a concentration of 0.1 mol/L were used to regulate the pH at 13.9 ± 0.1 pH units. After finishing the above, the beaker was sealed with preservative film to prevent evaporating and then put into thermotank controlled at 70 ◦ C. After 6 days, the beaker was taken out to removal of redundant KOH by centrifugation several times till the pH came to neutral. After centrifugation, the deposits were dried at 105 ◦ C, cooled to room temperature and ground to obtain powder for further characterization. The sample is labelled as synthetic goethite (SG).
2000 6000
2.2. Characterization 2.2.1. X-ray diffraction The synthetic goethite and Al-substituted goethite were prepared as pressed powders and mounted in stainless steel sample holders. The powder X-ray diffraction (XRD) patterns were recorded on a Philips PANalytical X’Pert PRO diffractometer using Cu K␣ radiation operating at 40 kV and 40 mA. XRD diffraction patterns were taken in the range of 10–70◦ at a scan speed of 2◦ min−1 with 0.5◦ divergence slit size. Phase identification was carried out by comparison with those included in the Inorganic Crystal Structure Database (ICSD). 2.2.2. Thermogravimetry Thermogravimetric analysis of the synthetic goethite and synthetic Al-substituted goethite were obtained by using TA Instruments Inc. Q500 TGA operating at different heating rate (2, 5, 10, 15, 20 ◦ C/min) from 38 ± 1 to 1000 ◦ C in a high-purity nitrogen atmosphere with a flowing rate for balance (40 cm3 /min) and a flowing rate for sample (60 cm3 /min). Approximately 38 ± 0.5 mg of finely ground dried sample was heated in an open platinum crucible.
SAG-9.9
4000 2000 6000
SAG-9.1
4000 2000 6000
SAG-6.9
4000 2000 6000
SAG-3.6
4000 2000 6000
SG
4000 2000 1500 1000 500 0 10
2.1.2. Preparation of Al-substituted goethite 9.378 g of Al(NO3 )3 ·9H2 O and 90.905 g Fe(NO3 )3 ·9H2 O were placed in a 1000 mL beaker and then 400 mL deionized water were put into the beaker. The Al(NO3 )3 ·9H2 O and Fe(NO3 )3 ·9H2 O were dissolved by stirring continuously. After dissolution, KOH with a concentration of 5 mol/L and a concentration of 0.1 mol/L were used to regulate the pH at 13.9 ± 0.1 pH units. The following steps are same as above. The sample is got with an Al substitution of 9.1 mol% in fact (10 mol% Al substitution in theory, Al/(Al + Fe)) and labelled as synthetic Al-substituted goethite (SAG-9.1). Alsubstituted goethite with different Al substitution (3.6, 6.9, 9.9, 11.5 mol%) is got by changing the mass of Al(NO3 )3 ·9H2 O. The Al substitution amount is calculated based on the results of the chemical composition measured on a Shimadzu XRF-1800 with Rh radiation.
SAG-11.5
4000
20
30
130 021 111
130
110
120
120
110
020
020
Intensity/Cps
Sample
ICSD(96-900-2159)
40
50
60
70
2θ/(º) Fig. 1. XRD patterns of synthetic goethite and Al-substituted goethite.
DTG component analysis was undertaken using the Jandel “Peakfit” software package (Jandel Scientific, CA, USA) 3. Results and discussion 3.1. XRD Fig. 1 shows the XRD patterns of synthetic goethite, Alsubstituted goethite with different Al substitution and goethite from ICSD (96-900-2159). These reflections ((0 2 0), (1 1 0), (1 2 0), (1 3 0), (0 2 1), (1 1 1), etc.) are observed and indentified as goethite compared with the ICSD (96-900-2159). In addition, the variation of d-spacings of goethite reflections after the occurrence of Al substituting Fe in the structure of goethite is presented (Table 2). The d-spacings are obtained from the XRD patterns using the software of X’Pert HighScore Plus. As is shown in Fig. 2, all d spacings of goethite reflection derived from ICSD (96-900-2159) are lower than that of the synthetic goethite in the experiment, which should be ascribed to the different preparation methods. What is more important, all d-spacings decrease slightly after the addition of Al(NO3 )3 ·9H2 O during the preparation of goethite, which is attributed to the smaller Al3+ ion radius than that of Fe3+ . This is good agreement with the report [4]. Schulze [4] has reported that diffraction peaks became broad and shifted to high diffraction angle (namely smaller d-spacings) with the increase of Al substitution in the structure of goethite. Moreover, all d-spacings of selected reflections decrease with an increase of Al substitution except (1 1 0). As shown in Fig. 2, linear relationship between unit cell dimension (UCD) and unit cell volume (UVC) and Al substitution is negatively related. All the data is got from XRD patterns by analysis of the Rietveld refinement method. It is worth to note that the correlation coefficient between a/b dimension and Al
22
H. Liu et al. / Thermochimica Acta 545 (2012) 20–25
Table 2 Component analysis of the mass loss of the TG and DTG for goethite and Al-substituted goethite. Sample
Heating rate (◦ C/min)
Step 1 dehydration
Step 2 dehydration
◦
Mass loss (%)
T ( C)
SG
2 5 10 15 20
1 1 1 0.8 1
38 38 38 38 38
± ± ± ± ±
SAG-9.1
2 5 10 15 20
1.2 1.3 1.1 1.2 1.3
38 38 38 38 38
± ± ± ± ±
Step 3 dehydroxylation ◦
◦
◦
Step 4 Total mass dehydroxylation loss (%) Mass loss (%)
Tm1 ( C)
Tm2 ( C)
Mass loss (%)
Tm3 ( C)
Tm4 ( C)
Tm5 ( C)
1 1 1 1 1
1.3 1.3 1.3 1.3 1.3
53.9 56.1 61.7 67.3 67.3
80.9 85.7 90.5 96.2 101
10.6 10.5 10.5 10.7 10.6
177.5 183.5 195.5 203 206
212.4 223.6 241.2 250 250
242.7 258 272.4 279.7 286.1
1.6 1.6 1.5 1.6 1.5
14.5 14.4 14.3 14.4 14.4
1 1 1 1 1
1.3 1.4 1.2 1.2 1.4
53.7 57.7 66.5 66.5 66.5
84.9 93.8 102.6 105 105
10.1 10.1 10.1 10.2 10.1
196.3 216.4 223.6 225.2 230
256.4 270 281.3 291.7 299
288.5 306.9 320.5 327.7 338.1
1.1 1.3 1.3 1.1 1.3
13.7 14.1 13.7 13.7 14.1
3.2. TG and DTG of goethite The TG and DTG curves of synthetic goethite with different heating rate (2, 5, 10, 15, 20 ◦ C/min) are shown in Fig. 3. The results of the component analysis of the TG and DTG curves are provided in Table 2. Four mass loss steps are observed in all TG curves and five peaks after fitting are found in all DTG curves. An average value of 14.4% total mass loss is observed over the ambient to 1000 ◦ C range, as is displayed by the TG curves. An about 1% mass loss is found which is assigned to the superficial adsorbed water and can be desorbed at 38 ± 1 ◦ C because all samples experience isothermal for 10 min at 38 ± 1 ◦ C before heating. A steady 1.3% mass loss are observed, which are attributed to the absorbed water. Combining with the DTG curves, the second step should be divided into two dehydration steps. Namely, there are three types of adsorbed water on the surface of goethite.
A third mass loss with an average value of 10.6% is observed at slightly high temperature which is attributed to the dehydroxylation. According to the DTG curves, three peaks can be found in the third step which should be assigned to different dehydroxylation. That is to say, there are three types of hydroxyl in the structure or on the surface of goethite. However, the theoretical mass loss for the conversion of goethite to hematite according to the equation (2FeOOH → Fe2 O3 + H2 O) should be precisely 10.1%. Therefore, the mass loss of dehydroxylation for the synthetic goethite is higher than the theoretical value. In addition, the evidence for the existence of two types of surface hydroxyl group has been reported by Parfitt [29], Russel [30] and Rochester [31]. Therefore, the third peak should be ascribed to desorption of hydroxylation which has a relative strong interaction with broken bond on the surface of goethite. Another evidence for this speculation is the less peak area than that of the latter two dehydroxylations. Moreover, excess OH occurring in goethite has been reported [10,12]. The last about 1.6% mass loss is observed which is contributed to desorption of the remanent hydroxyl. Comparison with the five DTG curves, every DTG curve has five peaks after fitting and these peaks temperature increases with an increasing heating rate. Therefore, it is suggested that the more rapid the heating rate, the higher peak temperature. Obviously, five sub-steps including dehydration and dehydroxylation occur 3.024
9.96 9.94
2
2
R =0.992
9.92 9.90 9.88 0
2
4
6
8
R =0.994
3.016
b/A
a/A
◦
Mass loss (%)
substitution is considerably higher that of c dimension. a-Axis and ˚ respectively, b-axis dimension decreases by 0.066 A˚ and 0.016 A, when Al substitution increases from 0 to 11.5 mol%. It indicates that a-axis is more sensitive than other axes which agrees well with the report [8]. This phenomenon also proves the occurrence of Al substitution for Fe in the structure of goethite used in the experiments.
10
Al substitution/mol%
3.008
3.000
12
4.63
0
2
4
6
8
10
Al substitution/mol%
12
139.2 2
R =0.739
4.62
2
R =0.953
138.4
UVC/A3
c/A
◦
4.61 4.60
137.6 136.8
4.59 0
2
4
6
8
10
Al substitution/mol%
12
0
2
4
6
8
10
Al substitution/mol%
Fig. 2. Linear relationship between UCD and UCV and Al substitution.
12
H. Liu et al. / Thermochimica Acta 545 (2012) 20–25
23
Fig. 3. TG and DTG of (a–e) goethite with an elevated heating rate 2, 5, 10, 15, 20 ◦ C/min.
in goethite regardless of the heating rate. As is well-known, much more energy would be needed to overcome energy barrier to accomplish the step with high peak temperature. Therefore, based on the research [32], the following formula is used to calculate desorption activation energy of every step including dehydration and dehydroxylation, 2 ln Tm − ln ˇ = Ed /(RTm ) + ln(Ed /AR), where Tm , ˇ, Ed , A, R represent peak temperature, heating rate, adsorption activation energy, pre-exponential factor, constant, respectively. The linear relationship between 2 ln Tm − ln ˇ and 1/Tm /103 is obtained, where 2 ln Tm − ln ˇ is labelled as y-axis and 1/Tm /103 is labelled as x-axis, as is displayed in Fig. 4(a). Then the Ed and A calculated by the slope and intercept are shown in Table 3. The 3.4, 6.2 kJ/mol of Ed
corresponding to adsorbed water is substantially lower than 19.4, 20.4, 26.1 kJ/mol corresponding to the hydroxyl in the structure or on the surface of goethite. 3.3. TG and DTG of substituted-goethite The TG and DTG curves of synthetic Al-substituted goethite with different heating rate (2, 5, 10, 15, 20 ◦ C/min) are shown in Fig. 5. The results of the component analysis of the DTG curves are provided in Table 2. Four mass loss steps are observed in all TG curves and five peaks after fitting are found in all DTG curves, which is the same as that of SG.
Table 3 The results of Ed (kJ/mol) and A for SG and SAG-9. Sample
Ed (kJ/mol)A
Tm1
Tm2
Tm3
Tm4
Tm5
SG
Ed A
3.4 740
6.2 282
19.4 9 × 104
20.4 1.2 × 104
26.1 4.5 × 104
SAG-9.1
Ed A
3.4 271
6.2 355
21.6 194
30 98
33.6 71
24
H. Liu et al. / Thermochimica Acta 545 (2012) 20–25
Fig. 4. The relationship between 2 ln Tm − ln ˇ and 1/Tm /103 for SG (a) and SAG-9.1 (b).
The attribution of the four steps according to TG curves and five peaks based on DTG curves for SAG-9 are same as that of SG. However, these temperatures of desorption peaks increase when Al substitutes for Fe in the structure of goethite compared with
that of SG, especially these temperatures of dehydroxylation. These temperatures corresponding to dehydroxylation for SAG-9 increase by about 25, 40, 50 ◦ C, respectively, in comparison to that of SG. This is good agreement with the report [22,28]. The research of Ray
Fig. 5. TG and DTG of (a–e) Al-substituted goethite with an elevated heating rate 2, 5, 10, 15, 20 ◦ C/min.
H. Liu et al. / Thermochimica Acta 545 (2012) 20–25
showed that the temperature of mass loss steps increases regularly with the increase in %Al substitution. A conclusion from the report of Ruan indicated that goethite synthesized from ferrous system altered to hematite with DTGA maximum increasing from 236 ◦ C to 273 ◦ C for 0–30.1 mol% Al-substitution. The reason evidently should be ascribed to the smaller atomic radius of Al than that of Fe, which results in a stronger coordination as OH− is linked via the bond of Al O to Al3+ . To investigate the effect of Al substitution on desorption activate energy of dehydration and dehydroxylation, the above formula is used to calculate the Ed . The linear relationship between 2 ln Tm − ln ˇ and 1/Tm /103 is illustrated in Fig. 4(b) and calculated results are presented in Table 3. The both have a positive relativity for dehydration of SAG-9 same to that of SG. As is shown in Table 3, the same Ed and the different A are observed for the two types of goethite. However, the Ed for dehydroxylation of SAG-9 increases from 19.4, 20.4, 26.1 kJ/mol to 21.6, 30, 33.6 kJ/mol when compared with that of SG. Such an effect is in harmony with the results of changes in the temperature of desorption peaks as is observed from DTG curves. In a contrary, the values of A for SAG9 decrease dramatically compared with that of SG. The A partly denotes the desorption rate. Therefore, the change of A values is positively associated with the changes of DTG curves. As is indicated in Fig. 4, all adsorption peaks become broad and shift to high temperature. 4. Conclusions Three types of adsorbed water and three types of hydroxyl units are observed using TG and DTG for the synthetic goethite and Alsubstituted goethite. The desorption of the first adsorbed water occurs easily in spite of the temperature of 38 ± 1 ◦ C. The values of Ed for the other two types of adsorbed water are 3.4, 6.2 kJ/mol for goethite and aluminous goethite. However, the Ed values of three types of hydroxyl for Al-substituted goethite have an increase from 19.4, 20.4, 26.1 kJ/mol to 21.6, 30, 33.6 kJ/mol when compared with that of goethite. The first hydroxyl should be ascribed to be linked to the surface broken bond and the latter two hydroxyls are assigned to locate in the structure of goethite. The A values of dehydroxylation for aluminous goethite decrease dramatically compared with that of goethite. The substitution of Al for Fe in the structure makes desorption peaks of hydroxyl become broad and shift to high temperature due to the slow desorption rate. The results provide a baseline kinetic information on the dehydration and dehydroxylation of goethite and the effect of Al substitution on that. The way can be extrapolated to the relative study of minerals on the types of hydroxyl. Acknowledgements This study was financially supported by Natural Science Foundation of China (no. 41172048, no. 41072035, no. 41130206) and PhD Programs Foundation of Ministry of Education of China (no. 20110111110003). The authors appreciate the financial support and the School of Chemistry, Physics and Mechanical Engineering, Science and Engineering Faculty, Queensland University of Technology, for providing the infrastructure to this research. References [1] D. Walter, G. Buxbaum, W. Laqua, The mechanism of the thermal transformation from goethite to hematite, J. Therm. Anal. Calorim. 63 (2001) 733.
25
[2] R.M. Cornell, U. Schwertmann, The Iron Oxides: Structure, Properties, Reactions, Occurrences and Uses, Second Editor, Weinheim, 2003, p. 509. [3] C. Zee, D.R. Roberts, D.G. Rancourt, C. Slomp, C.P. Slomp, Nanogoethite is the dominant reactive oxyhydroxide phase in lake and marine sediments, Geology 31 (2003) 993–996. [4] S.K. Ghose, G.A. Waychunas, et al., Hydrated goethite (␣-FeOOH) interface structure: ordered water and surface functional groups, Geochim. Cosmochim. Acta 74 (7) (2010) 1943–1953. [5] V. Morozov, S. Vasilev, Effect of isomorphic substitutions on the Mössbauer and magnetic parameters of goethite, Eurasian Soil Sci. 43 (7) (2010) 795– 801. [6] G. Busca, N. Cotena, et al., Infrared spectroscopic study of micronised geothite, Mater. Chem. 3 (4) (1978) 271–283. [7] F.J. Ewing, The crystal structure of diaspore, J. Chem. Phys. 3 (1935) 203– 207. [8] H.D. Ruan, R.J. Gilkes, Dehydroxylation of aluminous goethite: unit cell dimensions, crystal size and surface area, Clays Clay Miner. 43 (1995) 196– 211. [9] D.G. Schulze, U. Schwertmann, The influence of aluminium on iron oxides. X. properties of Al-substituted goethites, Clay Miner. 19 (1984) 521–539. [10] D.G. Schulze, U. Schwertmann, The influence of aluminium on iron oxides. XIII. Properties of goethites synthesised in 0–3 M KOH at 25 ◦ C, Clay Miner. 22 (1987) 83–92. [11] D.G. Lewis, U. Schwertmann, The effect of [OH] on the goethite produced from ferrihydrite under alkaline conditions, J. Colloid Interface Sci. 78 (1980) 543–553. [12] D.G. Schulze, The influence of aluminium on iron oxides. VIII. Unit-cell dimensions of Al-substituted goethites and estimation of Al from them, Clays Clay Miner. 32 (1984) 36–44. [13] C.J. Goss, The kinetics and reaction mechanism of the goethite to hematite transformation, Mineral. Mag. 51 (1987) 437–451. [14] K. Norrish, R.M. Taylor, The isomorphous replacement of iron by aluminium in soil goethites, J. Soil Sci. 12 (1961) 294–306. [15] C. Janot, H. Gibert, X. de Gramont, R. Biais, Étude des substitutions Al–Fe dans des roches latéritiques, Bull. Soc. France Mineral. Cristallogr. 94 (1971) 367–380. [16] B.G. Davey, J.D. Russell, M.J. Wilson, Iron oxide and clay minerals and their relation to colours of red and yellow Podzolic soils near Sydney, Australia, Geoderma 14 (1975) 125–138. [17] D. Nahon, C. Janot, A.M. Karpoff, H. Paquet, Y. Tardy, Mineralogy, petrography and structures ofiron crusts (ferricretes) developed on sandstones in the western part of Senegal, Geoderma 19 (1977) 263–277. [18] J.M. Bigham, D.C. Golden, L.H. Bowen, S.W. Buol, S.B. Weed, Iron oxide mineralogy of well-drained ultisols and oxisols. I. Characterization of iron oxides in soil clays by Mössbauer spectroscopy, X-ray diffractometry, and selected chemical techniques, Soil Sci. Soc. Am. J. 42 (1978) 816–825. [19] E. Mendelovici, Sh. Yariv, R. Villalba, Aluminum-beating goethite in Venezuelan laterites, Clays Clay Miner. 27 (1979) 368–372. [20] J. Torrent, U. Schwertmann, D.G. Schulze, Iron oxide mineralogy of some soils of two river terrace sequences in Spain, Geoderma 23 (1980) 191–208. [21] R.W. Fitzpatrick, U. Schwertmann, Al-substituted goethite—an indicator of pedogenic and other weathering environments in South Africa, Geoderma 27 (1981) 335–347. [22] H.D. Ruan, R.J. Gilkes, Dehydroxylation of aluminous goethite; unit cell dimensions, crystal size and surface area, Clays Clay Miner. 43 (2) (1995) 196–211. [23] K. Norrish, R.M. Taylor, The isomorphous replacement of iron by aluminum in soil goethites, J. Soil Sci. 12 (1961) 294–306. [24] K. Jónás, K. Solymár, Preparation X-ray, derivatographic and infrared study of aluminium-substituted goethites, Acta Chim. Acad. Sci. Hung. 66 (1970) 383–394. [25] M.B. Fey, J.B. Dixon, Synthesis and properties of poorly crystalline hydrated aluminous goethites, Clays Clay Miner. 29 (1981) 91–100. [26] R.C. Mackenzie, G. Berggren, Oxides and hydroxides of higher valence elements, in: R.C. Mackenzie (Ed.), Differential Thermal Analysis, Academic Press, London, 1970, pp. 271–302. [27] H.D. Ruan, R.L. Frost, J.T. Kloprogge, L. Duong, Infrared spectroscopy of goethite dehydroxylation. II. Effect of aluminium substitution on the behaviour of hydroxyl units, Spectrochim. Acta A 58 (2002) 479–491. [28] R.L. Frost, Z. Ding, H.D. Ruan, Thermal analysis of goethite, relevance to Australian indigenous art, J. Therm. Anal. Calorim. 71 (2003) 783–797. [29] R.L. Parfitt, J.D. Russell, et al., Confirmation of the surface structures of goethite (␣-FeOOH) and phosphated goethite by infrared spectroscopy, J. Chem. Soc., Faraday Trans. 1: Phys. Chem. Condens. Phases 72 (1976) 1082–1087. [30] J.D. Russell, R.L. Parfitt, et al., Surface structures of gibbsite goethite and phosphated goethite, Nature 248 (5445) (1974) 220–221. [31] C.H. Rochester, S.A. Topham, Infrared study of surface hydroxyl groups on goethite, J. Chem. Soc., Faraday Trans. 1: Phys. Chem. Condens. Phases 75 (1979) 591–602. [32] R.J. Cvetanovic, Y. Amenomiya, Application of a temperature-programmed desorption techniques to catalyst studies, Adv. Catal. 17 (1967) 103–149.