Colloids and Surfaces A: Physicochem. Eng. Aspects 234 (2004) 17–25
Synthesis and textural-structural characterization of magnesia, magnesia–titania and magnesia–zirconia catalysts Mar´ıa A. Aramend´ıa a , Victoriano Boráu a , César Jiménez a , Alberto Marinas a , José M. Marinas a , José A. Nav´ıo b , José R. Ruiz a , Francisco J. Urbano a,∗ a
Department of Organic Chemistry, University of Córdoba, Edificio Marie Curie, Campus de Rabanales, Edificio C-3, Córdoba E-14014, Spain b Instituto de Ciencia de Materiales, Centro de Investigaciones Cient´ıficas “Isla de la Cartuja”, Centro Mixto CSIC-Universidad de Sevilla, Avda. Américo Vespucio s/n. 41092, Sevilla, Spain Received 22 July 2003; accepted 18 November 2003
Abstract Several pure magnesium oxides together with some mixed systems (magnesia–titania and magnesia–zirconia) were synthesised, using the sol–gel technique. Chemical, textural and structural characterisation of the solids was carried out through a wide range of physical techniques. In all the cases, the presence of titanium or zirconium led to an increase in surface area as compared to pure magnesia. X-ray diffraction (XRD) revealed that no mixed phases were observed, but MgO (periclase), TiO2 (anatase) and ZrO2 (tetragonal) phases. As for mixed oxides, surface and bulk composition is quite similar in Mg–Zr system, whereas titanium tends to accumulate in the surface of Mg–Ti solids. The surface of the catalysts appears clearly carbonated as revealed by the XPS and diffuse reflectance infrared (DRIFT) spectra, revealing the basic character of the catalysts prepared. © 2003 Elsevier B.V. All rights reserved. Keywords: Magnesia; Magnesia–titania; Magnesia–zirconia; Textural characterization; Structural characterization
1. Introduction MgO, CaO and BaO were once regarded as catalytically inert materials, but are currently known to be highly active catalysts for certain base-catalysed reactions, if properly activated. Currently, all these materials are typically basic solid catalysts, particularly MgO, which can be considered as a reference material among solid base catalysts [1]. MgO is mainly obtained by thermal treatment of magnesium hydroxide or carbonate, and, more recently, by the sol–gel method. Textural and acid–base properties depend, to a great extent, on the synthesis conditions (pH, gelifying agent, sequence of the addition of reagents, calcination temperature, etc.) [2,3]. The modification of the acid–base properties of the MgO is also usually carried out by mixing it with other oxides, metallic ions or noble metals, which has revealed as a very effective way of tailoring the activity towards many organic processes. ∗
Coresponding author. Tel.: +34-957-218638; fax: +34-957-212066. E-mail address:
[email protected] (F.J. Urbano).
0927-7757/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfa.2003.11.026
Mixed oxides containing MgO were thoroughly studied recently. The mixtures are often prepared in the hope of finding synergetic effects, i.e. to produce a material with properties surpassing those of a linear combination of its constituents. Among these mixed oxides are the MgO–ZrO2 and MgO–TiO2 systems. Zirconia is known to exist in monoclinic, tetragonal and cubic phases, and as an amorphous solid. Of these modifications, the monoclinic form is thermodynamically stable under ambient conditions, while tetragonal and cubic phases are metastable. In some cases, such as zirconia ceramic materials, the synthesis parameters could be controlled to stabilise the high temperature phases, tetragonal and cubic, at ambient conditions, in order to give effect to their potential applications. These phases are stabilised by making solid solutions with some additives such as MgO [4,5]. In addition to its ceramic applications, ZrO2 is a promising catalytic material [6], since, apart from its high thermal stability, it exhibits both acidic and basic properties, which could be tailored through the addition of MgO. Again, the synthetic procedure is very important to obtain mixed solids with the desired textural and acid–base properties [7].
18
M.A. Aramend´ıa et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 234 (2004) 17–25
As far as TiO2 is concerned, this material has many applications, such as its use as pigment in paintings, in the production of electrochemistry electrodes, capacitors, solar cells, etc. Titania is also of interest in catalysis, where it can be used as the support or the catalyst. In the latter case, one of the most relevant uses is in the photocatalytic degradation of organic pollutants in water [8]. Titania can be found in rutile, anatase or brookite phases. Among them, anatase is the preferred for photocatalytic applications. The properties of titania can be modified when it is mixed with another oxide, such as SiO2 or ZrO2 . However, little research has been carried out dealing with the synthesis, characterisation and application of MgO–TiO2 mixed oxides, specially from the catalytic point of view. Bokhimi and co-workers [9,10] studied MgO–TiO2 mixed oxides prepared by the sol–gel technique. Such authors reported that, in addition to the stable phases of pure magnesia and titania (periclase, anatase and rutile), three intermediate compounds were detected in concentrations that depended on the MgO/TiO2 weight ratio and the annealing temperature. These phases were nanocrystalline, which generated samples with large surface areas that could be of interest in catalysis. Our research group has large experience on the application of pure magnesia, zirconia and, more recently, titania as catalysts, photocatalysts and catalytic supports for different organic processes such as reduction of unsaturated carbonyl compounds [11], hydrodechlorination [12] or the Meerwein–Ponndorf–Verley reaction [13]. One step forward would mean the search for more versatile catalysts by obtaining binary magnesia–titania and magnesia–zirconia mixed oxides to explore its catalytic properties. In this sense, in this work, several mixed MgO–TiO2 and MgO–ZrO2 oxides together with pure oxides (used as references) were synthesised and characterised from a textural and structural point of view, by a wide range of physical techniques. In a further paper, in preparation, surface chemical characterization will be carried out.
2. Materials and methods
99 ml min−1 ) until pH = 10 was reached and the corresponding gel was formed. This magnesium gel was aged for 72 h, vacuum-filtered, air-dried (24 h) and placed in an oven at 383 K for 24 h. Calcination of gels was carried out in a furnace at 873 K for 3 h, either in the air or in oxygen flow. The solids obtained were labelled Mg-AIR and Mg-OX. Three magnesia–titania mixed oxides were synthesised by dissolving the corresponding amount of magnesium nitrate (31.8, 50.9 and 57.3 g) and titanium oxide (5, 2 and 1 g) in water (Milli-Q; 542, 868 and 976 ml). The magnesium–titanium gels were treated as described above: precipitated with NaOH 1 M, aged, vacuum-filtered, air-dried, oven dried (383 K) and calcined in oxygen flow. The solids obtained were labelled MgTi5-OX, MgTi19-OX and MgTi31-OX. The magnesia–zirconia mixed oxide was synthesised by dissolving 50.9 g of magnesium nitrate and 5.2 g of zircomium oxychloride in 1 l of water (Milli-Q). The magnesium–zirconium gel was treated as described above: precipitated with NaOH 1 M, aged, vacuum-filtered, air-dried, oven dried (383 K) and calcined in oxygen flow. The solid obtained was named MgZr27-OX. 2.2. Thermal study of the precursors 2.2.1. Thermogravimetric analysis (TGA)–differential thermal analysis (DTA) analysis Thermogravimetric analysis and differential thermal analysis were recorded on a Setaram Setsys 12 instrument. Temperature was ramped from 323 up to 1373 K at 10 K min−1 . Experiments were carried out in an argon atmosphere (40 ml min−1 ). 2.2.2. Temperature-programmed-mass spectrometry (TP-MS) experiments TP-MS experiments were carried out in an on-line device consisting of a flow reactor coupled to a VG Sensorlab quadrupole mass spectrometer, operating in the multiple ion monitoring (MIM) mode. Fifty milligram of catalyst were placed in a tubular reactor (20 cm length and 1 cm i.d.). Experiments were carried out in an Ar flow rate of 50 ml min−1 and temperature was raised at a rate of 10 K min−1 .
2.1. Catalyst synthesis 2.3. Structural and textural characterization Catalysts were synthesised using the sol–gel technique, starting from Mg(NO3 )2 ·6H2 O (Merck, art. 5853), ZrOCl2 ·6H2 O (Merck, art. 1.08917) and TiO2 (Fluka, art. 89490). The nomenclature of the oxides includes the symbol of the metal or metals followed, in the case of mixed oxides, by the atomic Mg/Zr or Mg/Ti ratio, as determined by ICP. Finally “AIR” or “OX” indicate that calcination was carried out either in the air or in the oxygen flow. Pure MgO catalysts were synthesised by dissolving 88 g of magnesium nitrate in 1.5 l of water (Milli-Q). Then, the mixture was stirred and NaOH (1 M) was added (at
XPS data were recorded on 4 mm × 4 mm pellets, 0.5 mm thick, prepared by pressing at 10 t cm−2 the powdered materials which were outgassed at 150 ◦ C in the prechamber of the instrument up to a pressure < 2.67×10−6 Pa to remove chemisorbed water from their surfaces. The Leibold-Heraeus LHS10 spectrometer main chamber, working at a pressure < 2.67×10−7 Pa and a constant pass energy of 50 eV, was equipped with an EA-200 MCD hemispherical electron analyser with a dual X-ray source working with Al K␣ (hν = 1486.6 eV) at 20 mA, using C(1s) and Ti(2p3/2 ) as energy references ( 284.6 and 458.5 eV, respectively).
M.A. Aramend´ıa et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 234 (2004) 17–25 1
0 -5 -10 -15 -20 -25 -30 -35 -40 -45 -50 -55 (a) -60
-2 -3 -4 -5
-1
loss of weight (%)
0 -1
DTG (% min )
-6 645
-7
-7 M.S. signal (a.u.)
-8
H2O
-9 CO2 -10 400
600
800
1000
1200
1400
Temperature (K)
(b)
Fig. 1. Thermal study of precursor of Mg-AIR and Mg-OX. (a) TGA-DTG profile and (b) TP-MS profile. CO2 (m/z = 44) and H2 O (m/z = 18) were monitored.
796
635 713
other precursor, are not shown since it did not undergo any loss of weight on calcination. This is the reason why as titanium content increases loss of weight decreases. Therefore, such values can be ranked in the order MgTi31-gel (28.6%) 467
Elemental analysis of mixed oxides was carried out on a Perkin-Elmer 1000 ICP spectrophotometer after dissolution of the samples in nitric acid. Mg band, ZrOCl2 ·8H2 O and TiF6 (NH4 )2 were used as standards. Selected wavelengths were λ = 279.079, 343.823 and 334.941 nm for Mg, Zr and Ti, respectively. The textural properties of solids (specific surface area, pore volume and mean pore radius) were determined from nitrogen adsorption–desorption isotherms at liquid nitrogen temperature by using a Micromeritics ASAP-2000 instrument. Surface areas were calculated by the BET method, while pore distributions were determined by the BJH method. Prior to measurements, all samples were degassed at 383 K to 0.1 Pa. X-ray analysis of solids was carried out using a Siemens D-5000 diffractometer provided with an automatic control and data acquisition system (DACO-MP). The patterns were run with nickel-filtered copper radiation (λ = 1.5406 A) at 40 kV and 30 mA; the diffraction angle 2θ was scanned at a rate of 2◦ min−1 . FT-Raman spectra were obtained on a Perkin-Elmer 2000 NIR FT-Raman system with a diode pumped Nd:YAG laser (9394.69 cm−1 ). It was operated at a resolution of 4 cm−1 throughout the 3600–200 cm−1 range to gather 64 scans. Diffuse reflectance infrared (DRIFT) experiments were conducted on a Bomen MB-100 instrument with an “environmental chamber” (Spectra-Tech). It was operated at a resolution of 8 cm−1 throughout the 4000–400 cm−1 range to gather 256 scans. Diffuse reflectance UV-Vis spectra were performed on a Cary 1E (Varian) instrument, using polytetraethylene (density = 1 g cm−3 and thickness = 6 mm) as reference material.
19
0
-7
Mg-gel H2O
-8
-15 -9 CO2
-30
3. Results and discussion MgTi5-gel H2O
-8
-15 CO2
-30
MgTi19-gel
0
-9 -10 -7
H2O
-8
CO2
-9
-15
M.S. signal (a.u.)
Prior to calcinations, a thermal study of precursors was carried out by TGA-DTG and TP-MS. Solids consisting in pure magnesia (Mg-OX and Mg-AIR) were obtained by calcination of the same gel, labelled Mg-gel. Results found for thermal study of such a gel are shown in Fig. 1. As can be seen from the TGA-DTG profiles, there is a loss of weight centred at 645 K, mainly due to the loss of water. Such a loss means 32.3% of the initial weight. The difference between the obtained percentage of loss (32.3) and the theoretical value corresponding to the transformation of Mg(OH)2 into MgO (30.9%), could be ascribed to the presence of surface carbonates, as revealed by the CO2 released and detected by the mass spectrometer (m/z = 44). As far as magnesium–titanium precursor gels are concerned, their corresponding TGA and TP-MS profiles are shown in Fig. 2. Those obtained for Mg-gel are also included so that they can be better compared. The ones for TiO2 , the
loss of weight (%)
0
3.1. Thermal study of precursors
-10 -7
-30 0
MgTi31-gel
-10 -7
H2O
-8
-15 -9
CO2 -30
-10 400
600
800
1000
1200
1400
Temperature (K) Fig. 2. Thermal study of precursors of magnesium–titanium systems by TGA and TP-MS. CO2 (m/z = 44) and H2 O (m/z = 18) were monitored.
20
M.A. Aramend´ıa et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 234 (2004) 17–25
-35
1 0 -1 -2 -3 -4 -5 -6 -7 -8
0
1
0 -5 -10
Mg-gel
-15 -20 -25 645
-5
0 473
-10
MgZr27-gel
-15
-1 -2
-20
-3
-25
-4
-30 644
-35
Catalyst
Mg-AIR Mg-OX MgTi5-OX MgTi19-OX MgTi31-OX MgZr27-OX
DTG (% min-1)
loss of weight (%)
-30
Table 1 Chemical composition of the catalysts synthesised as determined by XPS
739
-5
Zr-gel
Heat flow (a.u.)
-10 -15 -20 -25 485
-30
120 100 80 60 40 20 0 -20
743
600
800
1200
> MgTi19-gel (26.8%) > MgTi5-gel (17.3%). Such losses are due do the Mg(OH)2 to MgO transformation (loss of H2 O) together with the CO2 loss from the surface carbonates, as stated above. Finally, regarding magnesium–zirconium gel, its thermal analysis profiles (Fig. 3) have some characteristics in common with both Mg-gel (loss of weight centred at 644 K) and a Zr-gel (loss at 473 K). Both cases correspond to endothermal processes associated to loss of water (m/z = 18) and CO2 (m/z = 44). Moreover, this figure shows that for the Zr-gel a highly exothermic peak appears at 743 K that is associated to the crystallisation of the amorphous gel to the monoclinic or tetragonal phases (insert in Fig. 3). This feature is absent in the MgZr-gel, indicating that either the hydrous oxide is already crystalline or more likely that the final oxide is amorphous [14,15]. Once the thermal study of precursors had been completed, calcination temperature was chosen. Therefore, we selected 873 K, since most of the loss of weight had already been completed at such temperature. Moreover, our previous experience in the synthesis of magnesium oxides allowed us to ensure that the resulting solids would still have a high surface area and a high enough acid–base sites density [2,3].
1.1 1.1 2.2 1.0 0.3 0.8
Mg-AIR
1305.6
532.0 533.8 532.4 534.2 530.9 532.6 530.4 532.2 532.3 534.1 532.2 534.1
MgTi5-OX MgTi19-OX
MgZr27-OX
Fig. 3. Thermal study of precursor of MgZr27-OX by TGA-DTG. For a better comparison, those for the corresponding single oxides are also included.
Na (1s)
0.0 0.0 5.9 2.6 2.1 0.0
O (1s)
-0.8
1400
Ti (2p)
0.0 0.0 0.0 0.0 0.0 1.1
Mg (1s)
MgTi31-OX
1000
Zr (3d)
59.5 60.3 57.4 57.1 56.8 57.9
284.6a
-0.6
Temperature (K)
O (1s)
30.5 29.8 24.5 29.9 33.0 32.4
C (1s)
-0.4
400 600 800 100012001400 Temperature (K)
-35 400
-0.2
Mg (1s)
8.9 8.8 10.0 9.4 7.9 7.7
Catalyst
Mg-OX 0.0
C (1s)
Table 2 Binding energy values (in eV) for all the catalysts, obtained by XPS
-5
0
Element concentration (%)
a
292.8 284.6a 292.2 284.6a 291.1 284.6a 292.5 284.6a 295.8 284.6a 296.1
1306.0 1304.6 1303.9 1305.7 1305.7
Zr (3d)
Ti (2p)
Na (1s) 1074.3 1074.8
458.5a 464.4 458.5a 464.1 460.2 465.6
1072.1 1072.3 1074.3
183.4 185.4
Internal reference.
3.2. Chemical composition: XPS and ICP One of the first questions to be addressed once the catalysts had been obtained was not only the global composition, in percentage of each element, for all our systems but also whether such a composition was the same in the surface of the particle as in the bulk. Therefore, a study of surface (XPS) and total (ICP) composition was made. Results are shown in Tables 1–3. 3.2.1. Pure magnesium oxides As far as magnesium oxides are concerned, experimental surface Mg/O ratio (about 0.5, Table 1) is closer to that expected from Mg(OH)2 than to MgO. Moreover, both solids are similar as far as carbon and sodium content are concerned. From XPS results, presented in Table 1, a tentative surface composition of solids, assuming a formula Mgx (OH)y Oz (CO3 )t Nal , could be proposed. Thus, x, t and l are obtained directly from such table, while y and z can Table 3 Comparison of surface (XPS) and total (ICP) chemical composition of mixed oxides Catalyst
Theoretical
XPS
ICP
Mg/Ti atomic ratio
MgTi5-OX MgTi19-OX MgTi31-OX
1.0 4.0 9.1
4.2 11.4 15.8
5.0 18.7 30.5
Mg/Zr atomic ratio
MgZr27-OX
3.3
29.1
27.4
M.A. Aramend´ıa et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 234 (2004) 17–25
be obtained from the oxygen balance and electroneutrality principle. Oxygen(%) = y + z + 3t 2x + l = y + 2z + 2t
(oxygen balance) (electroneutrality principle).
In summary, the surface composition of Mg-AIR and Mg-OX could be expressed by the formulae: Mg − AIR → Mg30.55 O11.49 (OH)21.46 (CO3 )8.86 Na1.06 Mg − OX → Mg29.79 O9.20 (OH)24.64 (CO3 )8.82 Na1.10 Regarding binding energies (Table 2), there are two different oxygen signals at about 532 and 534 eV. Binding energy for O(1s) at 530.4 eV is usually assigned to MgO, whereas it appears at 532.2 eV in hydroxylated species [16]. However, Sugiyama et al. [17] attributed bands at 532 and 533.8 eV to carbonates and hydroxycarbonates, respectively. Since, in our catalysts, the presence of carbon as carbonate is confirmed by the existence of a C(1s) signal at 292 eV [17,18], our O(1s) signals could be due to carbonate and/or hydroxides (532 eV) and hydroxycarbonates (533.8 eV), in agreement to an experimental surface Mg/O ratio close to 0.5. 3.2.2. Magnesium–titanium oxides There is a tendency of titanium to accumulate in the catalyst surface, as shown by the fact that Mg/Ti ratio, determined by XPS, is always lower than that obtained by ICP (see Table 3). Moreover, as titanium content increases there is also a remarkable increase in sodium content in the catalyst surface (see Table 1). This may indicate that some kind of sodium titanate has been formed, as reported by other authors [19,20]. Tentative surface compositions were calculated for magnesium–titanium systems: MgTi5 − OX → Mg24.54 (OH)0.29 (CO3 )9.96 Na2.17 O27.28 Ti5.88
21
Values obtained for Ti(2p) binding energy suggest that oxidation state for such a species is +4 [21,22]. These results were taken into account on calculating the above written formulae. Two signals corresponding to O(1s) binding energy appear at higher values than expected for either pure magnesia or titania, suggesting the presence of hydroxyl groups and carbonates (C(1s) signal at about 290 eV). 3.2.3. Magnesium–zirconium oxide Mg/Zr ratio is quite similar in the surface (XPS) and the bulk of the catalyst (ICP). Moreover, such a value is very high, probably due to the high pH value (pH = 10) used in the synthetic procedure. The selection of such a value was motivated by the solubility product of Mg(OH)2 . Therefore, calculations allowed us to conclude that at pH = 10 only 0.52% of Mg2+ species remain in solution, whereas, for instance, at pH = 9, such a value would go up to 52%. Nevertheless, at pH = 10 part of precipitated zirconium will redissolve, which could account for the high Mg/Zr ratio in the resulting catalyst (Table 3). Regarding XPS results, a tentative surface composition of the solid can be proposed, in a similar way as for the other systems. It could be expressed by the formula: MgZr27 − OX → Mg32.36 Na0.85 Zr 1.11 O19.81 (OH)14.91 (CO3 )7.74 As already observed in the analysis of Mg-OX by XPS, there are two signal corresponding to O(1s). Signal at 532.2 eV could be assigned to carbonate or hydroxide species whereas that appearing at 534.1 eV might be attributed to hydroxycarbonates. As for Zr(3d) signals they appear at 183.4 and 185.4 eV and correspond to Zr4+ . This result was taken into account on proposing the corresponding surface formula. 3.3. Textural and structural characterization
MgTi19 − OX
3.3.1. Textural characterization: nitrogen adsorption–desorption isotherms Textural properties of the solids are summarised in Table 4 including the BET surface area, pore diameter and pore volume, including the distribution of such pore volume into
→ Mg29.90 (OH)5.40 (CO3 )9.40 Na1.00 O23.50 Ti2.60 MgTi31 − OX → Mg33.00 (OH)7.30 (CO3 )7.90 Na0.30 O25.80 Ti2.10
Table 4 Specific surface area (SBET ), pore volume (Vp ) and pore mean diameter (dp ) obtained for all the solids Catalysts
Mg-AIR Mg-OX MgTi5-OX MgTi19-OX MgTi31-OX MgZr27-OX
SBET (m2 g−1 )
53 47 52 67 78 74
dp (A)
404 346 281 340 336 284
Vp % Micropore
% Mesopore
% Macropore
Total (ml g−1 )
0.0 0.0 0.1 0.0 0.0 0.0
32.8 34.9 72.5 63.3 69.8 73.1
67.2 65.1 27.4 36.7 30.2 26.9
0.57 0.42 0.46 0.68 0.81 0.64
TiO2 Fluka (precursor of magnesium–titanium oxides) is not included, since data could be subject to big mistakes as a result of the low surface (9 m2 g−1 ) of the catalyst.
M.A. Aramend´ıa et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 234 (2004) 17–25
3.3.2. X-ray diffraction (XRD) patterns The X-ray diffraction patterns for Mg-AIR and Mg-OX (not shown) exhibit three the characteristic peaks for periclase variety (2θ = 37.2, 43.1 and 62.5, respectively) [24,25]. Regarding magnesium–titanium oxides (Fig. 4), these systems are constituted by TiO2 (mainly anatase) and MgO as periclase. No mixed phases such as MgTi2 O5 , MgTiO3 or Mg2 TiO4 [9,10] are observed. On the other hand, although the XPS results indicated the presence of sodium titanate, no peaks related to this phase are observed in the XRD pattern. However, we can not discard the presence of amorphous sodium titanate, since it has been reported to form after alkali treating and heating [20]. As far as MgZr27-OX is concerned, its diffraction pattern (Fig. 5), exhibit bands corresponding to MgO (periclase) and ZrO2 (tetragonal phase), though again no mixed phases are observed. 3.3.3. Raman spectroscopy FT-Raman spectra do not provide much information as far as pure magnesia are concerned due to fluorescence problems that forced us to work at low laser power in order to avoid the saturation of the detector. Only two bands at 282 and 446 cm−1 were observed in the gel, disappearing on calcination. Similar results were obtained by Mestl et al. for Mg(OH)2 [26].
• TiO2(anatase) ∇ MgO (periclase) MgTi31-OX
Intensity ( a.u.)
micropores (<2 nm), mesopores (2–50 nm) and macropores (>50 nm). All solids present type II isotherms, associated to macroporous solids, according to Brunauer, Deming, Deming and Teller (BDDT) classification. Hysteresis loops are H3 type (De Boer classification [23]), characteristic of slit shaped, non-uniform pores. Both Mg-AIR and Mg-OX solids have similar surface areas and pore distribution (Table 4). For the magnesium– titanium oxides the specific surface areas are in all cases higher than that of Mg-OX, despite the low surface area of pure titania precursor (Fluka TiO2 , SBET = 9 m2 g−1 ). Moreover, as titanium content decreases BET area increases. These results could be explained in terms of a retardation of crystallisation of magnesium oxide in the presence of titania. However, TGA-DTA profiles cannot cast further light on this, since no peaks due to crystallisation (the so-called glow exotherm) were observed. Unpublished results of our research group show that the presence of TiO2 in zirconium–titanium systems synthesised in a similar way as magnesium–titanium ones, retards the appearance of the glow exotherm due to ZrO2 crystallisation. Parallel to the surface area increase there is a change in the pore size distribution, from predominantly macroporous (Mg-OX) to mainly mesoporous (Mg–Ti systems), (Table 4). The same results can be observed for the magnesium–zirconium oxide in comparison to the Mg-OX solid: an increase in the specific surface area and change from macroporous to mesoporous.
MgTi19-OX
MgTi5-OX
0
20
40
60
80
2 theta
Fig. 4. X-ray diffraction patterns corresponding to magnesium–titanium systems.
As far as the magnesium–titanium systems are concerned, since titanium was added as titania, the spectra of the gels (Fig. 6) show already bands corresponding to anatase with Raman shifts at 199, 398, 516 and 640 cm−1 [27–29]. Moreover, the same bands that appeared in Mg-gel can be observed, their intensity increasing with the magnesium content. Calcination of the gels results in the disappearance of such bands. Finally, for the magnesium–zirconium gel, bands at 281 and 445 cm−1 attributed to Mg-gel can also be detected. However, after calcination, the spectrum of MgZr27-OX (not represented) does not allow us to distinguish characteristic
ZrO2 (tetragonal)
intensity (a.u.)
22
MgZr27-OX
Mg-OX (periclase)
0
10
20
30
40
50
60
70
80
2 theta
Fig. 5. X-ray diffraction pattern of MgZr27-OX compared with that of Mg-OX and a zirconium oxide.
M.A. Aramend´ıa et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 234 (2004) 17–25
23 1378
•TiO2 (anatase)
Mg
1455
Mg 3739
MgTi31-gel
Mg-OX 3612
MgTi31-OX 1517 1381
Absorbance (a.u.)
Intensity (a.u.)
3743
MgTi19-gel
3585
MgTi19-OX 1385 3742 3608
3743
1536
MgTi5-OX Ti-OX
1505 3565
1397 1165
MgTi5-gel
4000
3000
2000
1000
-1
1000
750
500
Wavenumbers (cm )
250
Fig. 8. DRIFT spectra of magnesium–titanium oxides and the corresponding single oxides.
Raman Shift (cm-1)
Fig. 6. Raman spectra of the precursor gels of magnesium–titanium oxides.
bands of zirconia in the region 200–800 cm−1 probably due to the need of operating at low laser power because of the presence of magnesia.
1516
1378
3739 3611
3.3.4. DRIFT experiments Fig. 7 shows the DRIFT spectra for both Mg-AIR and Mg-OX systems. The broad and complex absorption at 2900–3700 cm−1 corresponds to the stretching vibrations of different types of hydroxyl groups. It is generally considered that “free” OH groups produces a sharp band at higher wavenumbers, (above 3700 cm−1 ). The OH groups, which interact by hydrogen bonding with their surrounding neighbours, are represented by the band below 3700 cm−1 , which is assigned to perturbed OH hydroxyls groups [30]. On the other hand, Mg-O stretching appears at about 1400–1441 cm−1 . However, carbonate and bicarbonate
Absorbance (a.u.)
MgZr27-OX
Mg-AIR
Mg-OX
4000
3000
2000
1000
Wavenumbers (cm-1)
Fig. 7. DRIFT spectra of Mg-OX, Mg-AIR and MgZr27-OX.
bands also appears in the region below 2000 cm−1 and, as stated above, the XPS measurements proved that our samples were carbonated. Unidentate carbonate exhibits a symmetric O–C–O stretching at 1360–1400 cm−1 and an asymmetric O–C–O stretching at 1510–1560 cm−1 . Bidentate carbonate shows a symmetric O–C–O stretching at 1320–1340 cm−1 and an asymmetric O–C–O stretching at 1610–1630 cm−1 . Bicarbonates show a C–OH bending mode at 1220 cm−1 as well as a symmetric and an asymmetric O–C–O stretching at 1480 and 1650 cm−1 , respectively [31]. According to this statement and Fig. 7, the main carbonate species in our solids are unidentate carbonates that exhibit bands at 1378 and 1516 cm−1 corresponding to the symmetric and asymmetric O–C–O stretching bands. For the magnesium–titanium systems, DRIFT spectra Mg–Ti systems are shown in Fig. 8, together with those of Fluka TiO2 calcined at 873 K (labelled Ti-OX) and Mg-OX, for comparison. In this figure, we can observe how the pure TiO2 does not present any absorption bands due to hydroxyl or carbonate species. Therefore, since from XRD data we concluded that no mixed phases were formed, any feature that appears in the magnesium–titanium systems are to be due to the magnesium atoms. Thus, we can observe both “free” (3743 cm−1 ) and bridging OH (3500–3600 cm−1 ) hydroxyls. On the other hand, there are also bands due to unidentate carbonate species, although their intensity is much lower than those corresponding to pure Mg-OX. This seems to indicate that magnesium–titanium mixed oxides are less basics than pure magnesium oxides (Mg-OX or Mg-AIR). Finally, the DRIFT spectrum for MgZr27-OX (Fig. 7) is very similar to that of magnesia. Free and hydrogen-bridged OH hydroxyls can be observed in the OH stretching region as well as unidentate carbonate species in the region below 2000 cm−1 . In this sense, we must note that the carbonate
24
M.A. Aramend´ıa et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 234 (2004) 17–25 0.08 TiO2-Fluka 0.07 0.06
MgTi19-OX
F (R) (a.u.)
0.05 0.04 0.03 MgZr27-OX 0.02 0.01 0.00 Mg-OX 200
300
400
500
600
Wavelength (nm)
Fig. 9. Diffuse reflectance UV-Vis spectra for some of the systems.
bands are also very strong in the MgZr27-OX solid indicating a strong surface basic properties. 3.3.5. UV-Vis spectroscopy Nowadays, numerous physical techniques can be used to study the catalysts. In particular, the absorption bands occurring in the visible and near-UV regions are used in heterogeneous catalysis as a technique to obtain information on the electronic structure of heterogeneous catalysts [32,33] as well as a way to evaluate the capability of a solid material to absorb either in the near-UV or in the visible regions (optical properties of materials); the later could be interesting to establish the potential use of a photocatalyst into the visible region for solar chemistry reactions [34–36]. Diffuse reflectance UV-Vis spectra of some of the studied systems are shown in Fig. 9, where Kubelka and Munk function is plotted as a function of wavelength. As can be seen, pure TiO2 and MgTi19-OX samples have similar band-gap since they have similar absorption wavelength (around 400 nm). This circumstance, together with the fact that titania is mainly present in MgTi19-OX as anatase (as shown by XRD and Raman) enables its potential use as photocatalyst. The same can be said of MgTi5-OX and MgTi31-OX (not represented in Fig. 9). The only difference is that their intensities increase with the content in titania. As for MgZr27-OX, due to the presence of zirconia, it could show a certain photocatalytic activity but probably much lower than that of titania containing systems. 4. Summary Different magnesium-containing solids were synthesised by the sol–gel technique. A thermal study of precursor by TGA-DTA and TP-MS led us to choose 873 K as the calcinations temperature, as most of the loss of weight is already completed at such a temperature.
A comparative study of surface (XPS) and bulk (ICP) composition showed the tendency of titanium to accumulate in the surface of MgO–TiO2 systems, whereas, as far as MgZr27-OX, Mg/Zr ratio is quite similar in both cases. Regarding magnesium–titanium systems, the content in sodium increases with the content in titanium, indicating that a kind of sodium titanate could be formed. However, this is not detected by XRD and if formed it must be amorphous. Mg-AIR and Mg-OX are macroporous solids whereas MgTi5-OX, MgTi19-OX, MgTi31-OX and MgZr27-OX are mesoporous. Surface areas of mixed oxides are, in all cases, higher than those of pure magnesia, suggesting the possibility of a retardation in crystallisation of MgO, in the presence of zirconium and titanium. XRD and Raman analysis showed that no mixed phases are present but MgO (periclase), TiO2 (anatase) and tetragonal zirconia. The synthetic procedure, in which titanium was added as TiO2 , favoured the maintenance of TiO2 as anatase in Mg–Ti systems after calcination. This circumstance, together with the substancial increase in surface area (as compared to Fluka TiO2 ) and UV-Vis results, make us think that, in the absence of other influencing factors, Mg–Ti solids could potentially be used as photocatalysts. Acknowledgements The authors gratefully acknowledge the financial support from Junta de Andaluc´ıa and Ministerio de Ciencia y Tecnolog´ıa in the framework of Project BQU2001-2605 (co-financed with FEDER funds). References [1] K.Tanabe, M. Misono, Y. Ono, Y. Hattori, in: New Solids Acids and Bases, Studies in Surface Science Catalysis, Vol. 51 (1989). [2] M.A. Aramendia, V. Borau, C. Jimenez, J.M. Marinas, A. Porras, F.J. Urbano, J. Catal. 161 (1996) 829. [3] M.A. Aramendia, V. Borau, C. Jimenez, J.M. Marinas, A. Porras, F.J. Urbano, J. Mater. Chem. 6 (1996) 1943. [4] E.N.S. Muccillo, R. Muccillo, N.H. Saito, Mater. Lett. 25 (1995) 165. [5] R. Muccillo, E.N.S. Muccillo, N.H. Saito, Mater. Lett. 34 (1998) 128. [6] T. Yamaguchi, Catal. Today 20 (1994) 199. [7] T. Settu, Ceram. Int. 26 (2000) 517. [8] J.M. Herrmann, Catal. Today 53 (1999) 115. [9] X. Bokhimi, J.L. Boldu, E. Munoz, O. Novaro, T. Lopez, J. Hernandez, R. Gomez, A. Garcia-Ruiz, Chem. Mater. 11 (1999) 2716. [10] T. Lopez, J. Hernandez, R. Gomez, X. Bokhimi, J.L. Boldu, E. Munoz, O. Novaro, A. Garcia-Ruiz, Langmuir 15 (1999) 5689. [11] M.A. Aramendia, V. Borau, C. Jimenez, J.M. Marinas, A. Porras, F.J. Urbano, Appl. Catal. A 172 (1998) 31. [12] M.A. Aramendia, V. Borau, I.M. Garcia, C. Jimenez, A. Marinas, J.M. Marinas, F.J. Urbano, J. Catal. 187 (1999) 392. [13] M.A. Aramendia, V. Borau, C. Jimenez, J.M. Marinas, J.R. Ruiz, F.J. Urbano, Appl. Catal. A: 244 (2003) 207. [14] G.K. Chuah, S. Jaenicke, B.K. Pong, J. Catal. 175 (1998) 80. [15] G.K. Chuah, Catal. Today 49 (1999) 131. [16] Y.H. Hu, E. Ruckenstein, Catal. Lett. 43 (1997) 7.
M.A. Aramend´ıa et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 234 (2004) 17–25 [17] S. Sugiyama, T. Ookubo, K. Shimodan, H. Hayashi, J.B. Muffat, Bull. Chem. Soc. Jpn. 67 (1994) 3339. [18] V.R. Choudhary, V.H. Rane, R.V. Gadre, J. Catal. 145 (1994) 300. [19] M.F. Chen, X.J. Yang, Y. Liu, S.L. Zhu, Z.D. Cui, H.C. Man, Surf. Coat. Technol. 173 (2003) 229. [20] F. Liang, L. Zhou, K. Wang, Surf. Coat. Technol. 165 (2003) 133. [21] J.P. Nogier, A.M. De Kersabiec, J. Fraissard, Appl. Catal. A 185 (1999) 109. [22] J. Huuhtanen, M. Sanati, A. Andersson, T.A. Lars, Appl. Catal. A 97 (1993) 197. [23] Ceramics De Boer, in: B.H. Everett, F.S. Stone (Eds.), The Structure and Properties of Porous Materials, Butterworths, London, 1958. [24] M.A. Aramendia, J.A. Benitez, V. Borau, C. Jimenez, J.M. Marinas, J.R. Ruiz, F.J. Urbano, Colloids Surf. A 168 (2000) 27. [25] S. Ardizzone, C.L. Bianchi, B. Vercelli, Colloids Surf. A 144 (1998) 9. [26] G. Mestl, M.P. Rosynek, J.H. Lunsford, J. Phys. Chem. B 101 (1997) 9321.
25
[27] P.K. Dutta, P.K. Gallagher, J. Twu, Chem. Mater. 5 (1993) 1739. [28] K. Lagarec, S. Desgreniers, Solid. State. Commun. 94 (1995) 519. [29] G.A. Tompsett, G.A. Bowmaker, R.P. Cooney, J.B. Metson, K.A. Rodgers, J.M. Seakins, J. Raman Spectrosc. 26 (1995) 57. [30] W. Ignaczak, W.K. Jozwiak, E. Szubiakiewicz, T. Paryjczak, Pol. J. Chem. 73 (1999) 645. [31] V.K. Diez, C.R. Apesteguia, J.I. Di Cosimo, Catal. Today 63 (2000) 53. [32] E.Garbowski, E. Praliaud, Electronic spectroscopy, in: B. Imelick, J.C. Vedrine (Eds.), Catalyst Characterisation. Physical Techniques for Solid Materials, Plenumm Press, New York, 1994, Chapter 4. [33] J.M. Thomas, W.J. Thomas, Principles and Practice of Heterogeneous Catalysis, VCH, New York, 1996. [34] S. Sakthibel, B. Neppolian, M.V. Shankar, B. Arabindoo, M. Palanichamy, V. Murugesan, Sol. Energy Mater. Sol. Cells 77 (2003) 65. [35] C. Hu, Y. Tang, J.C. Yu, P.K. Wong, Appl. Catal. B 40 (2003) 131. [36] M. Iwasaki, M. Hara, H. Kawada, H. Tada, S. Ito, J. Colloid Interface Sci. 224 (2000) 202.