Effect of phosphate ion on the textural and catalytic activity of titania–silica mixed oxide

Applied Catalysis A: General 220 (2001) 9–20

Effect of phosphate ion on the textural and catalytic activity of titania–silica mixed oxide Santosh Kumar Samantaray, Kulamani Parida∗ Regional Research Laboratory (CSIR), Bhubaneswar 751013, Orissa, India Received 9 January 2001; received in revised form 9 April 2001; accepted 13 April 2001

Abstract Phosphate-modified TiO2 -SiO2 mixed oxide catalysts have been prepared by varying the method of preparation, source and concentration of phosphate ion. The prepared catalysts were compared for their catalytic activity/selectivity in nitration of toluene. The characterisation of the catalysts was performed using X-ray powder diffraction (XRD), Fourier-transform infrared spectroscopy (FT-IR), thermal analysis (TG–DTA), nitrogen adsorption–desorption methods, surface acid strength measured by Hammett indicator method, surface acid sites measured by amine titration method, and phosphate content measured by UV–VIS spectrophotometry. The XRD patterns revealed that phosphate ion stabilises the anatase phase up to 1173 K activation. FT-IR results show that phosphate species strongly bound bidentately, and that both the internal weakly H-bonded hydroxyl groups and free hydroxyl groups are present on TiO2 –SiO2 mixed oxide support. Surface area and surface acidity are found to increase with the increase in phosphate loading up to 7.5 wt.% and thereafter the values decrease drastically. However, average pore radius and total pore volume shows the reverse order. Phosphated samples prepared using H3 PO4 as the source of phosphate ion exhibit higher acidity, and surface area but lower porosity than the samples prepared from (NH4 )3 PO4 , though both the samples contain the same amount of phosphate (7.5 wt.%). Similar results were also observed when varying the method of preparation. TiO2 –SiO2 samples prepared at pH = 3 exhibit higher acidity and surface area but lower porosity than the samples prepared at pH = 7. The acid strength of 7.5P/TiO2 –SiO2 (H) is found to be stronger than that of 100% concentrated H2 SO4 . The material modified with phosphate ion was found to be an efficient and selective catalyst for solvent-free mono-nitration of toluene. Selectivity to the para-product is correlated with the porosity of the material. © 2001 Elsevier Science B.V. All rights reserved. Keywords: TiO2 –SiO2 mixed oxide; Phosphate ion; Mono-nitration; Selectivity

1. Introduction The nitration of aromatic compounds is an important process in both industrial and academic research [1–3]. In particular, the nitration of toluene is useful for producing military explosives such as 2,4,6trinitro-toluene (TNT), pharmaceutical intermediates ∗

Corresponding author. Tel.: +91-674-581636/38/39; fax: +91-674-581637. E-mail address: [email protected] (K. Parida).

such as p-amino benzoic acid (PABA), and colorant intermediates. The noxious homogeneous catalysts in current use, which are sources of pollution, industrial hazard and equipment corrosion [4–6], could be replaced by non-corrosive solid acid catalysts. The use of solid acid catalysts is potentially more attractive because of the ease of removal, recycling and its possible influence on the selectivity. Consequently, in recent years there has been a spate of activity aimed at the development of new nitration methods using solid acid catalysts, and partial success has been achieved.

0926-860X/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 6 - 8 6 0 X ( 0 1 ) 0 0 6 3 8 - X

10

S.K. Samantaray, K. Parida / Applied Catalysis A: General 220 (2001) 9–20

Laszlo and co-workers [7–10] and Smith et al. [11] have independently reported novel methods for the nitration of aromatic compounds. Laszlo and co-workers have developed a reagent known as “claycop”, which is Cu(NO3 )2 supported on acidic montmorillonite clay, that selectively nitrates toluene under Menke conditions (use of acetic anhydride as co-reagent), affording a quantitative yield of mono-nitro toluene with an ortho-, meta-, and para-isomer distribution of 23:1:76. Very high dilution of toluene (1 ml) in CCl4 (2 l) and 120 h reaction time are required to achieve the high conversion and regio-selection for the para-isomer. The disadvantages of this system are the requirement for high dilution, isolation of products, and reutilisation of the catalyst. It is very difficult to remove the excess acetic anhydride during the purification of the products, and the stoichiometric use of copper nitrate makes regeneration of the system rather difficult in industrial applications. Smith et al. discovered that large port mordenite zeolite and benzoyl nitrate in toluene can afford 90% isolable yield for mono-nitro toluene in a 42-h period and with 60% selectivity for para-isomer. Problems associated with benzoyl nitrate include difficulty in handling due to its sensitivity toward decomposition and the tendency toward detonation upon contact with rough surfaces. Although zeolite beta-I and II [12,13] are used due to their shape-selective nature to catalyse the reaction, no major commercial process is yet in operation, due to the rapid ageing of the same. Anions such as SO4 2− , PO4 3− , WO4 2− , CO3 2− , − Cl , F− and OH− have been found to enhance the acidity and certain other physico-chemical properties of the catalyst, like thermal stability and mesoporosity [14–17]. It is well documented that sulphated TiO2 –SiO2 mixed oxides can be used as solid acid catalysts due to their high acid strength [18–23]. More recently, increasing applications of the sulphated TiO2 –SiO2 mixed oxide solid acids are being found in heterogeneous catalysts, such as alcohol dehydration, cumune dealkylation and esterification of acetic acid [18–23]. It is proposed that the active sites were generated by the interaction between hydrous gel and the sulphate ion. Studies on sulphate-modified TiO2 –SiO2 mixed oxides have received much attention because of their versatility in syntheses of organic chemicals and advanced materials. However, no attempt has been made till now to synthesise and characterise the

physico-chemical properties and catalytic activities of phosphated TiO2 –SiO2 mixed oxide. We, for the first time, have prepared and thoroughly characterised the physical as well as the chemical properties of the phosphated TiO2 –SiO2 mixed oxide and tested have for certain selective industrially important chemical reaction. In this paper, we present the results obtained from the studies on the physico-chemical behaviour and catalytic activity of phosphated TiO2 –SiO2 mixed oxide prepared by different methods; the source and concentration of phosphate ion were varied to achieve the mono-nitration of toluene.

2. Experimental 2.1. Materials and methods TiO2 –SiO2 pH = 3 samples: Tetraethylorthosilicate (TEOS, Aldrich, 99%, 48.57 ml) was refluxed at 343 K for 3 h under constant mechanical stirring with a water and isopropyl alcohol mixture of ratio 8. HCl was added to adjust the pH of the initial solution to 3. Tetrabutylorthotitanate (TBOT, Aldrich, 99%, 48.95 ml) dissolved in isopropyl alcohol (1/10 v/v) was added to the above solution at 288 K under vigorous stirring in order to avoid spontaneous precipitation of titania. When the TBOT was added, the reflux temperature was raised up to 353 K. Finally, a colourless homogeneous gel was obtained. This gel was dried at 363 K overnight in an air oven. TiO2 –SiO2 pH = 7 samples: The procedure was similar to the one described for the previous samples. Instead of HCl, NH4 OH (Merck) was employed to adjust the initial pH of the solution to 7. 2.2. Modified catalysts A series of phosphated TiO2 –SiO2 samples were prepared using (NH4 )3 PO4 as a source of phosphate ions by a solid–solid kneading method. For comparison, the other series of phosphated TiO2 –SiO2 samples were prepared by an aqueous impregnation method using dilute H3 PO4 . The suspended mass was evaporated to dryness on a hot plate while stirring. After impregnation with phosphate, the samples were dried in an air oven at 383 K overnight and subsequently

S.K. Samantaray, K. Parida / Applied Catalysis A: General 220 (2001) 9–20

activated at 723, 923, 1023, and 1173 K in a muffle furnace.

11

Prior to adsorption–desorption experiments, the samples were degassed at 393 K at 10−4 Torr for 5 h. 3.6. Acid strength

3. Characterisation 3.1. Determination of phosphate content The concentration of phosphate in the above samples was measured using a UV–VIS spectrophotometer (Varian) following the ascorbic acid reduction method [24].

All the samples dried at 383 K for 6 h in an air oven and cooled in a desiccator and used for determination of acid strength. The acid strength of the catalysts were examined by the colour change method using a Hammett indicator [25], when a powder sample was added to an indicator dissolved in dry benzene. 3.7. Surface acid sites

3.2. X-ray powder diffraction study XRD patterns of the pure and phosphate-modified TiO2 –SiO2 mixed oxide samples were recorded on a Philips X-ray diffractometer using a PW 1830 generator with copper tube, a PW 1820 goniometer fitted with a post-diffracted graphite monochromator and a scintillation detector attached to a PW 1710 diffraction control in the range of 2θ = 10–80◦ at a scanning speed of 2◦ /min. The instrument was operated at 40 kV and 20 mA. 3.3. FT-IR study The IR spectra of pure and phosphate-modified TiO2 –SiO2 mixed oxide samples were recorded at room temperature on 2% sample-KBr pellets (≈50 mg weight) with a Perkin-Elmer (Model Paragon 500) FT-IR spectrometer in the range 4000–400 cm−1 . Before pellets were made, all the samples were degassed at 383 K in vacuum (1 × 10−4 Torr). 3.4. TG–DTA analysis TG–DTA analyses of samples dried at 383 K were carried out in dry air (50 cm3 /min) using a Shimadzu DT-40 thermal analyser in the range of 300–1423 K at a heating rate of 288 K/min. 3.5. Determination of textural properties The specific surface area (BET), pore volumes, average pore radius and pore-size distributions of the catalysts were determined at liquid nitrogen boiling temperature using Quantasorb (Quantachrome, USA).

The acid sites of pure and phosphate-modified TiO2 –SiO2 mixed oxide samples were determined by the amine titration method at H0 = −3.0 (dicinnamalacetone indicator), where H0 is the Hammett acidity function and the pretreatment conditions of the catalysts were identical to that of acid strength. 3.8. Nitration of toluene Solvent-free catalytic nitration of toluene (CDH, AR) was carried out in a double-necked roundbottomed flask at room temperature. In a typical experiment, 0.1 g of the heat-treated sample was taken in the flask and 4.9 mmol of nitric acid (70%) were added. The whole assembly was kept in an ice cold water bath for 5 min. To the above cold solution, 1.5 ml of acetic anhydride (Merck, GR) was added and the mixture was stirred for another 5 min (the amount of acetic anhydride was so chosen that it would allow quantitative conversion to acetyl nitrate and removal of any water present in the nitric acid used). Stirring was followed by drop wise addition of 4.9 mmol of toluene. The reaction mixture was brought back to room temperature and allowed to react for 30 min. The products were separated from catalyst by filtration, followed by washing with CCl4 . The only by-product, acetic acid, was separated from the mixture by a simple distillation technique. The yield and product selectivity of the mono-nitro toluene were analysed by GC using 15% FFAP on 80/100 mesh W (HP) column. Each experiment was repeated thrice to test the reproducibility of the result and the catalyst was recycled without any further treatment to test the reusability.

12

S.K. Samantaray, K. Parida / Applied Catalysis A: General 220 (2001) 9–20

Table 1 Textural parameters of TiO2 –SiO2 and PO4 3− /TiO2 –SiO2 samples activated at 723 Ka Sample no.

Catalysts

Wt.% of PO4 3−

SBET (m2 /g)

Sα (m2 /g)

Average pore radius (Å)

Pore volume

1 2 3 4 5 6 7 8

TiO2 –SiO2 2.5P/TiO2 –SiO2 5.0P/TiO2 –SiO2 7.5P/TiO2 –SiO2 10.0P/TiO2 –SiO2 7.5P/TiO2 –SiO2 (H) 10.0P/TiO2 –SiO2 (H) 7.5P/TiO2 –SiO2 (H∗ )

0.00 2.90 5.52 8.20 11.16 8.12 11.04 8.04

341 354 369 372 261 387 213 265

341 353 367 370 260 386 215 266

12.38 – – 4.15 – 4.02 – 14.24

0.13 – – 0.04 – 0.03 – 0.14

a

H and H∗ indicate H3 PO4 impregnated on TiO2 –SiO2 samples prepared at pH = 3 and 7, respectively.

4. Results and discussion 4.1. Textural properties The phosphate concentrations of the samples activated at 723 K are shown in Table 1. The phosphate concentration in samples activated at 773 K is higher than the corresponding amount added during preparation. This is due to the loss of water (dehydration) and alcohol (adsorbed solvent in course of preparation of catalyst) during activation. Similar types of observation are also reported earlier by Pattnayak and Parida [26].

Figs. 1 and 2 show the TG–DTA curves of the unmodified and phosphate-modified TiO2 –SiO2 gels. The TG curves (Fig. 1) show a notable weight loss of 50% on heating TiO2 –SiO2 mixed oxide up to 1423 K. A simple mass balance shows that the loss of weight corresponds to the quantity of reactants (water, isopropyl, butyl radicals) involved in the synthesis of TiO2 –SiO2 gels. There is no additional weight loss in phosphate-modified TiO2 –SiO2 mixed oxide up to 1423 K, indicating no loss of phosphate up to that temperature. There are three main domains, which can be distinguished from the DTA profiles (Fig. 2).

Fig. 1. Thermogravimetry analysis of (a) unmodified, (b) phosphate-modified TiO2 –SiO2 samples.

S.K. Samantaray, K. Parida / Applied Catalysis A: General 220 (2001) 9–20

13

Fig. 2. Differential thermal analysis of (a) unmodified, (b) phosphate-modified TiO2 –SiO2 samples.

• Between 293 and 523 K, a broad endothermic peak appeared, due to the loss of physisorbed water and the alcohols trapped in the porous texture. • Between 573 and 723 K, a broad exothermic peak due to the combustion of remaining alkoxy group resulting from the incomplete hydrolysis/ polycondensation reaction bound to titanium and silicon. The trapped alcohols with the porous texture may also contribute to this exothermic peak. After thermal treatment at 623 K, the sample became dark brown. • Between 723 and 873 K, the second broad exothermic peak is probably due to a harder oxidation of the carboned residues coming from the combustion [27,28]. Its intensity is lower than that of the first one. An exothermic peak observed prominently in the region near (1200–1350 K) for phosphate-modified TiO2 –SiO2 and, to a lower extent, the one for unmodified TiO2 –SiO2 without any corresponding weight loss are very likely due to solid-state transformations (anatase to rutile phase transformation). Its position at so high a temperature that agrees with the inhibiting effect of phosphate species on the transformation of anatase to rutile already reported [29]. The IR absorption spectra of the unmodified and phosphate-modified TiO2 –SiO2 activated at different

temperatures are shown in Figs. 3 and 4. The band at 935 cm−1 is attributed to the Ti–O–Si bond and the bands at 400–600, 800 and 1064 cm−1 are related to the Ti–O–Ti bond and the Si–O–Si bond, respectively. The bands at 800 and 1064 cm−1 are attributed to the symmetric and asymmetric stretching vibrations of the Si–O–Si groups, respectively [22]. Fig. 3 shows that the Ti–O–Si band intensity varies with varying the method of preparation. The Ti–O–Si band is more intense in the case of the sample prepared at pH = 3 compared to the sample prepared at pH = 7. This band intensity also varies with the source of phosphate ion. The intensity of this band is more in case of phosphated TiO2 –SiO2 (7.5 wt.% PO4 3− ) mixed oxide, when H3 PO4 as the source of phosphate ion, than (NH4 )3 PO4 . At higher activation temperatures (1023 and 1173 K), it is also observed (Fig. 4) that the Ti–O–Si band intensity decreases with increasing Ti–O–Ti and Si–O–Si bands. This is due to a reordering of the solid, which takes place by the breakage of the Ti–O–Si bond and formation of Ti–O–Ti as well as Si–O–Si bonds. This phenomenon has already been studied by Gunji et al. [30]. Fig. 5, the resolved form of IR spectra of phosphated TiO2 –SiO2 mixed oxide, shows a strong sharp absorption band at 1300–1400 cm−1 and broad bands at 1100–1250 cm−1 . The 1300–1400 cm−1 peak is the stretching frequency of P–O bond, whose order

14

S.K. Samantaray, K. Parida / Applied Catalysis A: General 220 (2001) 9–20

Fig. 3. FT-IR spectra of (a) TiO2 –SiO2 , (b) 7.5P/TiO2 –SiO2 (H), (c) 7.5P/TiO2 –SiO2 , (d) 7.5P/TiO2 –SiO2 (H∗ ) samples activated at 723 K.

Fig. 4. FT-IR spectra of 7.5P/TiO2 –SiO2 (H) sample activated at (a) 1023 and (b) 1173 K, respectively.

S.K. Samantaray, K. Parida / Applied Catalysis A: General 220 (2001) 9–20

15

Fig. 5. Resolved FT-IR spectra of 7.5P/TiO2 –SiO2 (H) sample activated at (a) 1023 and (b) 1173 K, respectively.

is close to two (P=O, phosphoryl groups), and the 1100–1250 cm−1 peaks are the characteristic frequencies of PO4 3− . The broad band at 1100–1250 cm−1 resulted from the lowering of the symmetry in the free PO4 3− (Td point group). When PO4 3− is bound to the TiO2 –SiO2 surface, the symmetry can be lowered to either C3V or C2V [31]. Here, the band that split into three peaks (1162, 1193 and 1230 cm−1 ) was assigned to the bidentately bound phosphate ion (C2V point group). These three bands somewhat shifted to lower wave number (Fig. 5b) when the sample is activated at 1173 K (1130, 1162 and 1230 cm−1 ). This agrees with the position of such a fundamental band, as has been directly observed in the case of phosphated zirconia [32] and alumina [33]. The band at 1637 cm−1 attributed to the Si–H2 O absorption is seen on all the spectra of unmodified and phosphate-modified TiO2 –SiO2 mixed oxide samples. The IR spectra of the modified samples show sharp bands in the region 3800–3600 cm−1 which could be

assigned to ν(OH) of free surface hydroxyl groups. From Fig. 4, a very sharp and perfectly symmetric band at 3766 cm−1 is seen if activation is carried out at 1173 K. At 1023 K, the spectrum shows shoulders at 3766 cm−1 (weak) and 3820 cm−1 (intense). The low frequency absorption at 3435 cm−1 is due to ν(OH) of internal weakly H-bonded hydroxyl groups [34], while the higher frequency one at 3820 cm−1 has been assigned to the ν + γ combination of free hydroxyl groups [35]. The intensity of all types of bands for OH groups seems to decrease with phosphate impregnation as well as on activation at higher temperature. Thus the substitution of TiO2 –SiO2 hydroxyl groups with phosphate ion impregnation process is evident, as occurs on ␥-Al2 O3 [36] where less acidic hydroxyls groups are first removed, while hydroxyls of strong acidity are created. The XRD patterns of TiO2 –SiO2 mixed oxide samples activated at 723, 923, 1023, and 1173 K are shown in Fig. 6. It is seen from the XRD patterns

16

S.K. Samantaray, K. Parida / Applied Catalysis A: General 220 (2001) 9–20

Fig. 6. Powder X-ray diffraction patterns of (a) TiO2 –SiO2 , 723 K and 7.5P/TiO2 –SiO2 (H) sample activated at (b) 723 K, (c) 923 K, (d) 1023 K and (e) 1173 K, respectively.

that the TiO2 –SiO2 mixed oxide sample activated at 723 K is amorphous. TiO2 peaks are not detected in TiO2 –SiO2 catalyst though the catalyst contains as high as 50 mol% of TiO2 (TiO2 –SiO2 ). The presence of SiO2 in TiO2 –SiO2 obviously inhibits the growth of TiO2 crystals and results in amorphous TiO2 [37]. However, PO4 3− /TiO2 –SiO2 shows a broad peak characteristic of TiO2 anatase with no evidence of either rutile or silica phases. With increasing the activation temperature from 723 to 1023 K in the case of phosphate-modified TiO2 –SiO2 samples, the intensity of diffraction lines of anatase form of TiO2 gradually increases. But at higher activation temperature (>1173 K), both anatase and rutile forms can be identified. However, from our earlier publication [23], unmodified TiO2 –SiO2 mixed oxides shows both anatase and rutile phase of TiO2 if activation is carried out at higher temperature than 1073 K. Such results show that phosphate could have stabilised the anatase phase up to 1173 K. TG–DTA analyses support such observations. This type of effect is also observed in SO4 2− [38], PO4 3− [39], and WO3 [40] on TiO2 , ␣-FeOOH and ZrO2 , respectively. This result is further supported by IR analysis, i.e. that a structural rearrangement of the gel occurs at higher temperatures (above 1023 K), which gives rise to Ti–O–Ti and Si–O–Si bonds.

Table 1 shows the BET surface area of unmodified and phosphate-modified TiO2 –SiO2 mixed oxide samples prepared by different methods. From the BET surface area measurement, it is found that surface area does not alter much by phosphate impregnation up to 7.5 wt.% (341–372 m2 /g). But at higher phosphate concentration (10 wt.%) the surface area decreases drastically from 372 to 261 m2 /g. It is observed that the sample prepared at pH = 3 has higher surface area (377 m2 /g) than the sample prepared at pH = 7 (265 m2 /g), although both the samples contain the same amount of phosphate ion (7.5 wt.%). The sample containing 7.5 wt.% of PO4 3− (H3 PO4 as the source of PO4 3− ion) has more surface area (387 m2 /g) than the sample containing the same amount of PO4 3− but obtained from a different source of phosphate ion ((NH4 )3 PO4 , 372 m2 /g). The average pore radius is calculated, assuming the pores to be cylindrical, using the formula r = 2V p /S p , where r is the average pore radius, Vp the pore volume, and Sp the specific internal surface area of the pores measured by summing of the pore area over the whole pore system. The reverse trend is also observed for both average pore radius and total pore volume. Therefore, the presence of a low amount of phosphate ion may be responsible for the formation of porous a network. It has been shown [41,42] that in anion modified metal oxides, some of

S.K. Samantaray, K. Parida / Applied Catalysis A: General 220 (2001) 9–20

the hydroxyl bridges originally present in dried unactivated and unphosphated TiO2 –SiO2 mixed oxides are replaced by the phosphate ions. On activation, the formation of oxy-bonds takes place, and results in changes in the Ti–O–Si, Ti–O–Ti and Si–O–Si bond strength due to attachment of the phosphate bridges. Thus, the changes in the bond strength may be responsible for the formation of porous network. However, when phosphate content increases beyond 7.5 wt.%, pore blocking/formation of polyphosphate takes place due to the presence of an excess amount of phosphate ion. N2 adsorption–desorption isotherms of all the samples are of nearly the same type (Fig. 7) and can be assigned as type IV or II in the BDDT classification [43]. From the shape of the curves it can be predicted that sample prepared at pH = 7 consists mainly of mesoporous material. However, the samples prepared at pH = 3 are micro-mesopores. A similar observation can also be drawn from the pore-size distribution curves (Fig. 8) calculated by the BJH equation [44]. Thus due to mesoporosity, samples prepared at pH = 7 show a drastic decrease in surface area. The α s -plots (Fig. 9) of the samples prepared at pH = 3 show a downward deviation. However, the samples

17

Fig. 8. Pore-size distribution curves as a function of pore-radius.

prepared at pH = 7 show an upward deviation. The downward deviation reveals that these samples contain predominantly micropores, along with some mesopores, whereas an upward deviation indicates the presence of mesopores. 4.2. Acid strength Table 2 shows that the Hammett acidity function of 7.5P/TiO2 –SiO2 is ≥−13.16, whereas for TiO2 –SiO2 ,

Fig. 7. N2 adsorption–desorption isotherm of (a) TiO2 –SiO2 , (b) 7.5P/TiO2 –SiO2 (H) and (c) 7.5P/TiO2 –SiO2 (H∗ ) samples activated at 723 K.

Fig. 9. α s -plots of (a) TiO2 –SiO2 , (b) 7.5P/TiO2 –SiO2 (H) and (c) 7.5P/TiO2 –SiO2 (H∗ ) samples activated at 723 K.

18

S.K. Samantaray, K. Parida / Applied Catalysis A: General 220 (2001) 9–20

Table 2 Acid strength and acid sites of the TiO2 –SiO2 and PO4 3− /TiO2 –SiO2 samples activated at 723 K Catalysts

TiO2 –SiO2 2.5P/TiO2 –SiO2 5.0P/TiO2 –SiO2 7.5P/TiO2 –SiO2 10.0P/TiO2 –SiO2 7.5P/TiO2 –SiO2 (H) 7.5P/TiO2 –SiO2 (H∗ )

pKa value of indicators (␮mol/g) −8.2

−12.4

−13.16

−14.52

−3.0

+ + + + + + +

− + + + + + +

− ± ± + ± + ±

− − − − − ± −

210 322 346 372 332 405 380

it is ≥−8.2. These results mean that the weak acid sites of unmodified catalyst are converted into strong acid sites by means of phosphate ion modification. Acids stronger than H0 = −11.93, which corresponds to the acid strength of 100% H2 SO4 , are known as super acids [16]. So, 7.5P/TiO2 –SiO2 acts as a solid super acid catalyst. The increase in acid strength in the modified catalyst is attributed to the double bond nature of P=O, which strengthens the acid sites by the inductive effect [45]. 4.3. Surface acidity The acid sites of unmodified and phosphate-modified TiO2 –SiO2 mixed oxide catalysts after being activated at 723 K for 4 h were determined by titrating the solid suspended in benzene against n-butyl amine using dicinnamalacetone indicator, as reported in Table 2. It is observed that the phosphate-modified catalysts have much more acidity than the unmodified catalyst. When one increases the PO4 3− loading up to 7.5 wt.%, the acidity gradually increases (322–372 ␮mol/g); thereafter it decreases to 322 ␮mol/g on further addition. The initial increase in surface acidity, with an increase in phosphate loading up to 7.5 wt.%, may be due to phosphate monolayer formation. The decrease in the surface acidity at high phosphate concentration (10 wt.%) is probably due to formation of polyphosphate, which decreases the number of Brönsted acid sites and consequently the total number of acid sites [42]. It is also observed that phosphated sample prepared using H3 PO4 exhibit higher acidity (405 ␮mol/g) compared to the samples prepared using (NH4 )3 PO4 (372 ␮mol/g) when the same amount (7.5 wt.%) of phosphate was impregnated. Similarly,

samples prepared at pH = 3 exhibit higher acidity (405 ␮mol/g) than the samples prepared at pH = 7 (380 ␮mol/g). It is reasonable to assume that, during the preparation procedure, the aqueous phosphoric acid protonates all types of hydroxyls (TiO2 –SiO2 ) by an acid–base reaction. However, phosphate of ammonium phosphate by solid–solid kneading method undergoes interaction with all types of basic hydroxyls to a smaller extent, resulting in less acidity compared to phosphate of phosphoric acid by aqueous impregnation method. Similar observations have been reported earlier in the cases of SO4 2− [46] and PO4 3− [47] on alumina. 4.4. Catalytic activity Catalytic outcomes of nitration reaction of toluene are presented in Table 3. It is observed that the yield of mono-nitro toluene in the case of phosphate-modified TiO2 –SiO2 mixed oxide is more than in the case of the unmodified TiO2 –SiO2 catalyst. If one increases the wt.% of PO4 3− up to 7.5, the yield of the reaction does not increase so much. However, with higher phosphate concentration (10 wt.%), the yield decreases drastically. Comparing the TiO2 –SiO2 mixed oxide samples with the same amount of phosphate ion but different sources of the ion (H3 PO4 and (NH4 )3 PO4 ), one sees that the yield of mono-nitro product more or less remained the same, but the P/O ratios are different. A sample prepared from H3 PO4 ion has more para-selectivity (P/O = 2.19) than that of the sample prepared from (NH4 )3 PO4 (P/O = 1.96). The same trend is also observed in case of TiO2 –SiO2 mixed oxide samples prepared by different methods. TiO2 –SiO2 mixed oxide samples

S.K. Samantaray, K. Parida / Applied Catalysis A: General 220 (2001) 9–20

19

Table 3 Catalytic activity of the TiO2 –SiO2 and PO4 3− /TiO2 –SiO2 samples activated at 723 K towards mono-nitration of toluene Catalysts

TiO2 –SiO2 2.5P/TiO2 –SiO2 5.0P/TiO2 –SiO2 7.5P/TiO2 –SiO2 10.0P/TiO2 –SiO2 7.5P/TiO2 –SiO2 (H) 10.0P/TiO2 –SiO2 (H) 7.5P/TiO2 –SiO2 (H∗ )

Yield (%)

58 81 82 83 68 89 59 78

Selectivity (%)

P/O

Ortho (o)

Meta (m)

Para (p)

36 30 32 33 35 31 35 38

1 1 1 2 2 1 2 2

63 69 67 65 63 68 63 60

prepared at pH = 7 (impregnated with 7.5 wt.% of phosphate using H3 PO4 as the source of PO4 3− ion) have less para-selectivity (P/O = 1.57) than the samples prepared at pH = 3 (P/O = 2.19) with the same concentration and source of phosphate ion. The mechanism of nitration reaction involves electrophilic attack on the aromatic rings by the nitronium ion, NO2 + . Brönsted acid sites are responsible for the generation of NO2 + ion from nitric acid, whereas Lewis acid sites are responsible for the selectivity to para-product [12,48]. The earlier work [49] emphasises the activity, yield and selectivity of the catalyst as a function of acidity of the material. However, our observation shows that the selectivity to para-product decreases with increase in porosity of the material, which was also reported earlier by Kwok and Jayasurya [50] on zeolite catalyst (i.e. the higher the porosity, the lower the selectivity to para-product). This observation is reflected on samples prepared using (NH4 )3 PO4 as the source of phosphate ion and TiO2 –SiO2 prepared at pH = 7. In this case, though the porosity increases, the selectivity to para-product is less compared to the samples prepared at pH = 3 and H3 PO4 as the source of phosphate ion, with almost negligible change in acidic properties.

5. Conclusions 1. Modification of the hydrous gel with anions like phosphate can alter the textural as well as the acidity character of the sample. 2. Control of porosity and generation of new strong acid sites, which depend on the method of prepa-

1.75 2.30 2.09 1.96 1.80 2.19 1.80 1.57

ration, source and concentration of phosphate ion affect the catalytic activity of TiO2 –SiO2 mixed oxide. 3. The acidity and porosity of the catalyst control the yield and selectivity to the mono-nitro product.

Acknowledgements The authors are thankful to Dr. V.N. Mishra, Director, Regional Research Laboratory (CSIR), Bhubaneswar, for giving permission to publish this paper and to Dr. S.B. Rao, Head, Inorganic Chemicals Division, for his valuable suggestions and constant encouragement. One of the authors, S.K. Samantaray is obliged to CSIR, New Delhi for a senior research fellowship. References [1] G.A. Olah, R. Malhotra, S.C. Narang, Nitration Method and Mechanism, VCH, New York, 1989, pp. 5–9. [2] S. Borman, Chem. Eng. News (1994) January 17, 18. [3] K. Schofield, Aromatic Nitration, Cambridge University Press, Cambridge, 1980. [4] G. de Almeida, J.L.M. Dufaux, Y. Ben Taarit, C. Naccache, Appl. Catal. A 114 (1994) 141–159. [5] G.A. Olah, Friedel–Crafts Chemistry, Wiley, New York, 1973. [6] H. Miki, US Pat. 4347,393 (1982). [7] L. De Laude, P. Laszlo, K. Smith, Acc. Chem. Res. 26 (1993) 607. [8] P. Laszlo, Acc. Chem. Res. 19 (1986) 121. [9] P. Laszlo, J. Vandormael, Chem. Lett. (1988) 1843. [10] A. Cornelis, L. Delaude, A. Gerstmans, P. Laszlo, Tetrahedron Lett. 29 (1988) 5657. [11] K. Smith, K. Fry, M. Butters, B. Nay, Tetrahedron Lett. 30 (1989) 5333.

20

S.K. Samantaray, K. Parida / Applied Catalysis A: General 220 (2001) 9–20

[12] B.M. Choudary, M. Sateesh, M. Lakshmi Kantam, K. Koteswara Rao, K.V. Ram Prasad, K.V. Raghvan, J.A.R.P. Sarma, Chem. Commun. (2000) 25. [13] K. Smith, A. Musson, G.A. DeBoos, Chem. Commun. (1996) 469. [14] J.M. Camplo, M.S. Chiment, J.M. Martinas, Appl. Catal. 3 (1982) 315. [15] M. Hino, K. Arata, J. Chem. Soc., Chem. Commun. (1980) 851. [16] M. Hino, K. Arata, J. Chem. Soc., Chem. Commun. (1979) 1148. [17] G. Larsen, F. Lotero, M. Nasity, L.M. Petkovu, D.S. Shobe, J. Catal. 164 (1996) 246. [18] J.R. Sohn, H.J. Jang, J. Catal. 136 (1992) 267. [19] J.R. Sohn, H.J. Jang, M.Y. Park, E.H. Park, S.E. Park, J. Mol. Catal. 93 (1994) 149. [20] J. Navarrete, T. Lopez, R. Gomez, Langmuir 12 (1996) 4385. [21] T. Lopez, P. Bosch, F. Tzompantzi, R. Gomez, J. Navarrete, E. Lopez-Salinas, M.E. Llanos, Appl. Catal. A 197 (2000) 107. [22] K.M. Parida, S.K. Samantaray, H.K. Mishra, J. Colloid Interface Sci. 216 (1999) 127. [23] S.K. Samantaray, K.M. Parida, Appl. Catal. A 211 (2001) 175. [24] D.J. Rochford, Aust. J. Mar. Fresh Water Res. 2 (1951) 1. [25] L.P. Hammett, A.J. Deyrup, J. Am. Chem. Soc. 54 (1932) 272. [26] P.K. Pattnayak, K.M. Parida, J. Colloid Interface Sci. 226 (2000) 340. [27] W.T. Minchan, L.G. Menssing, C.G. Pantano, J. Non-Cryst. Solids 108 (1989) 163. [28] A.J. Lecloux, S. Nonet, F. Noville, J.P. Pirard, Ann. Chim. Fr. 16 (1991) 77. [29] J. Criado, C. Real, J. Chem. Soc., Faraday Trans. 1 79 (1983) 2765.

[30] T. Gunji, Y. Nagao, T. Misono, Y. Abe, J. Non-Cryst. Solids 107 (1989) 149. [31] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Co-ordination Compounds, 4th Edition, Wiley, New York, 1986. [32] G. Ramis, G. Busca, V. Lorenzelli, P.F. Rossi, M. Bensitel, O. Saur, J.C. Lavalley, in: Proceedings of the IXth International Congress on Catalysis, Calgary, Canada, 1988, p. 1874. [33] A. Mennour, C. Ecolivet, D. Cornet, J.F. Hemidy, J.C. Lavalley, L. Mariette, P. Engelhard, Mater. Chem. Phys. 19 (1988) 301. [34] S. Kondo, H. Yamagouchi, Y. Kajiyama, T. Ishikawa, J. Chem. Soc., Faraday Trans. 1 80 (1984) 2033. [35] A.A. Tsyganenko, Russ. J. Phys. Chem. 56 (1982) 1428. [36] J.B. Perri, J. Phys. Chem. 69 (1965) 220. [37] T.-C. Liu, T.-J. Cheng, Catal. Today 26 (1995) 71. [38] B.Y. Zhao, X.P. Xu, H.R. Ma, J.M. Gao, D.H. Sun, R.R. Wang, Y.Q. Tang, Acta Physico-Chim. Sinica 9 (1993) 8. [39] K. Kandori, S. Uchida, S. Kataoka, T. Ishikawa, J. Mater. Sci. 27 (1992) 719. [40] D.G. Barton, S.L. Soled, G.D. Meitzner, G.A. Fuentes, E. Iglesia, J. Catal. 181 (1999) 57. [41] B.H. Davis, R.A. Reogh, R. Srinivasan, Catal. Today 20 (1994) 219. [42] A.K. Dalai, R. Sethuraman, S.P.R. Katikaneni, R.O. Idem, Ind. Eng. Chem. Res. 37 (1996) 3869. [43] S. Brunauer, D.W. Demming, L.S. Demming, E. Teller, J. Am. Chem. Soc. 62 (1940) 1723. [44] E.P. Barrett, L.G. Joyner, P.P. Halonda, J. Am. Chem. Soc. 73 (1951) 373. [45] J.R. Sohn, H.J. Kim, J. Catal. 101 (1986) 428. [46] Y. Okamoto, T. Imanaka, J. Phys. Chem. 92 (1988) 7102. [47] J.M. Lewis, R.A. Kydd, J. Catal. 132 (1991) 465. [48] K.M. Parida, P.K. Pattnayak, Catal. Lett. 47 (1997) 255. [49] T. Mishra, K.M. Parida, J. Mol. Catal. 121 (1997) 91. [50] T.J. Kwok, K. Jayasuriya, J. Org. Chem. 59 (1994) 4939.