Spectroscopic tools for probing the isolated titanium centres in MCM41 mesoporous catalysts

IL NUOVO CIMENTO VOL. 19 D, N, 11 Novembre 1997

Spectroscopic tools for probing the isolated titanium centresin MCM41 mesoporous catalysts (*)

L MARCHESE e), E. GIANOTTI e), T. MASCHMEYER (2)G. MARTRA e), S. COLUCCIA e) and J. M. THOMAS (2)e) Dipartimento di Chimica IFM - via P Giuria. 7, 1-70725 Torino. Italye) Davy Faraday Research Laboratory. The Royal Institution of Great Britain

27 Albemarle Street, London W7X 4BS, UK

(ricevuto il 28 Febbraio 1997; approvato 1'8 Maggio 1997)

Summary. - Catalytically active Ti(IV) centres anchored to the inner walls ofMCM-4l mesoporous silicas have been investigated using a combination of FT-IR,diffuse reflectance UV-visible and luminescence spectroscopy supplemented by theuse of H20 and NH3 as molecular probes of the Ti coordination. The results re-affirmand extend the conclusions reported earlier for the Ti-MCM4l novel epoxidationcatalysts, see MASCHMEYER T., REY F., SANKAR G. and THOMAS J. M., Nature, 378(1995) 159, who, using in situ EXAFS studies, showed that the Ti(IV) centres existlargely as isolated active sites.

PACS 82.65.Jv - Heterogeneous catalysis at surfaces.PACS 81.05.Rm - Porous materials; granular materials.PACS 78.55 - Photoluminescence.PACS 78.30 - Infrared and Raman spectra.PACS 01.30.Cc - Conference proceedings.

1. - Introduction

There is currently great interest in Ti (IV) -based siliceous catalysts for variouskinds of selective oxidation. Starting with the seminal work of Taramasso et al. [1], whoshowed that the so-called titanosilicate TS-1 (a titanium silicalite) was capable of a widerange of low-temperature oxidations with H 20 2 , many others [2-5] have underlined theimportance and value of Ti-based silicate in catalytic contexts.

The synthesis of mesoporous siliceous materials with larger channel apertures, suchas MCM-41 (pores in the 25-100 A range) [6], has expanded considerably the scope forshape-selective Ti-based heterogeneous catalysts for bulky molecules [3,5,7-9]. Forexample the materials prepared by Maschmeyer et al. (anchoring organometalliccomplexes onto the inner walls of MCM-41), exhibit high catalytic performance in the

(*) Paper presented at the "First International Workshop on Reactivity of Oxide Materials.Theory and Experiment", Como, 8, 9 November 1996.

© Societa Italiana di Fisica 1707

1708 L. MARCHESE, E, GIANOTTI, T, MASCHMEYER, ETC,

epoxidation of cyclohexene and in oxidation of bulky cyclic and aromatics compoundsdue to the presence of a high concentration of accessible, well-spaced and structurallywell-defined Ti active sites [7].

These results have prompted us to study the nature and the coordinativeenvironment of Ti active sites in MCM-41 by means of computational and EXAFSapproaches as it has been proposed that the high activity/selectivity performance ofTi-MCM41 is due to isolated tetrahedra titanium centres [10]. It was also inferred thatthe presence of titanium dimers and oligomers should decompose the peroxide (whichis used as oxidizing agent) and that octahedrally coordinated titanium should be inactive for the epoxidation of alkenes in that it lacks free coordination sites [3, 11, 12].

Here we present data obtained by means of FTIR, diffuse reflectance andluminescence UV-Vis spectroscopies, supplemented by the use of NH3 and H20 asmolecular probes, which aid in the clarification of the coordination of titanium sites ofTi-MCM41 prepared by the grafting procedure of Maschmeyer et al. [7].

2. - Experimental

A Ti-MCM41, containing 2% wt Ti, was prepared with the strict exclusion of waterunder an argon atmosphere using conventional Schlenk line techniques and dehydrating the MCM-41 at 250°C under a dynamic vacuum prior to titanium loading.

FT-IR experiments on pelletized samples were recorded with a Bruker IFS88spectrometer at a resolution of 4 em -1, and by means of specially designed cells whichwere permanently connected to a vacuum line (ultimate pressure :::; 10 - 5 torr) to makeadsorption-desorption in situ experiments.

Diffuse reflectance UV-Vis spectra were recorded by means of a Perkin Elmer(Lambda 19) spectrometer equipped with an integrating sphere attachment andphotoluminescence spectra by using SPEX FLUOROLOG-2 1680 spectrometer. Thesamples, in form of powder, were placed in quartz cells which, also in this case, werepermanently connected to a vacuum line for in situ experiments.

All the samples were calcined in 100 Torr O2 at 550°C for 5 to 10 hours in order toeliminate the organic fraction of the Ti-cyclopentadienyl complexes anchored onto thesurface of MCM41; successively, the samples were also evacuated at 550°C.

3. - Result and discussion

3'1. Diffuse reflectance UV data. - It is well established that the diffuse reflectanceUV spectroscopy gives very effective information on the coordination state of Tisites [13-17].

Figure 1 shows diffuse reflectance UV spectra of Ti-MCM41. The sample in vacuo(curve a) presents a band centred at around 230 nm which is gradually replaced, uponwater adsorption, by a new broader band at ca. 240 nm (spectra b-d). The highest dose(curve d) has spectroscopic features which extend down to 320 nm and which cover alsothe region where the original tetrahedral Ti(IV) sites absorbed.

Similar results, obtained by studying Ti-silicalite, were explained by Boccuti etal. [13] as due to tetrahedral Ti(IV) sites, in that case a band at 200-210 nm was found,which undergoes a coordination change to octahedra by insertion of two molecules ofwater as extraligands (a band at 240 nm was observed). The bands were assigned tooxygen to tetrahedral titanium (IV) charge transfer (LMCT) according to the empirical

PROBING THE TITANIUM CENTRES IN MCM41 CATALYSTS 1709

4

MCM41

360320280240

ok---·-----------------·········--·---···········:=:-:=~~~~

200

Wavelength [nrn]

Fig. 1. - Diffuse reflectance UV spectra of Ti-MCM41 activated in vacuo at 550°C (curve a) andupon adsorption of water (curves b-d); the spectrum of a pure MCM-41 is also reported.

optical electronegativity theory [14], which suggests that the (corrected) energy of thefirst Laporte-allowed CT (for TiX4 complexes it is related to Jrtl ---7 de (Jr + a) tz---7 deelectron transfer) is given by equation

(1) V corr (em-I) = 30 OOO[X opt (anion) - Xopt (cation)] .

When reasonable values of optical electronegativities (OH- = 3.45; tetr. Ti(IV) = 1.85and oct. Ti(IV) = 2.05) are substituted in eq. (1), a fairly good agreement with theexperimental data is obtained.

However, an accurate inspection of the 200-210 nm CT band reveals that a strongshoulder at 230 nm is always present in dehydrated Ti-silicalites even in the case ofsamples in which, as suggested by accurate EXAFS, XANES [15, 16], XPS and IH andz9Si NMR [16] investigations, virtually all the titanium exists in tetrahedralco-ordination. Le Noc et al. suggested that the two bands at 200-210 and 230 nm aredue to two different framework sites (Ti(OH)(OSi)3 and Ti(OSi)4, see structures 1 and 3in scheme 1) which have different TiOSi angles. The sites with the larger angles areresponsible for the band at lower wavelengths [16].

Such an assignment seems to contradict the results predicted by eq. (1) and raises aquestion whether the 230 nm band is due to tetrahedral or octahedral Ti(IV) sites.However, it should be considered that the relationship (1) allows to estimate the


Structure 1 Structure 2

Scheme 1

Structure 3

energies for any transition metal ion complex, and it has also to be considered thatvarious factors should be subtracted from the observed CT frequencies to get reliable(corrected v corr) values [14].

In addition, it is reasonable to argue that complexes which have oxygen ligands withdifferent electronegativity, as is the case of the structures depicted in scheme 1, willpresent more than one CT band at different energy. It is indeed well established thatthe oxygen of a siloxane bridge has a lower negative charge than that of a OHgroup [18], and, consequently, it will have an electron transfer at higher energy.

It is likely that tetrahedral Ti centres in Ti-anchored MCM-41 with both siloxy- andhydroxy-ligands (see structures 1 and 2 in scheme 1) are very abundant and this mightwell explain the presence of a very intense band centred at 230 nm rather than at200-210 nm found on Ti-silicalites.

Therefore, it can be argued that the adsorption of two extra ligands (H20), whichexpands the Ti(IV) coordination sphere from tetrahedral to octahedral, occurs onTi-MCM41 catalysts. The mechanism is represented in scheme 2.Altogether these results confirm the XANES and EXAFS studies which suggestedthat a large fraction of titanium in Ti-anchored MCM-41 exists in tetrahedralco-ordination [7].

(A)

H20

Scheme 2

(B)

PROBING THE TITANIUM CENTRES IN MCM41 CATALYSTS

3"2. FT-/R data,

1711

3"2.1. The nature of surface hydroxyls, Figure 2 shows FTIR spectra of atypical pure-silica MCM-41 (a) and the Ti-MCM41 (b) calcined and outgassed at 550°C.The MCM-41 has a narrow band at 3745 cm- I overlapped to a weak and very broadabsorption in the range 3700-3400 cm -I; the assignment of such absorptions isstraightforward because bands in similar positions have been found on variousamorphous silicas [18,19 and references therein], as well as on silicalites [20] and canbe attributed definitely the first to the O-H stretching of free silanols, and the secondto silanols interacting by H-bonds (see scheme in the figure), TGA analysis revealedthat, after heating at 600°C, the MCM-41 has 2,0 OH/nm2

, and it is indeed very likelythat a fraction of hydroxyls are H-bonded,

In Ti-MCM41, besides free silanols (here found at 3743 cm- I) and H-bonded

hydroxyls, a strong shoulder at ca, 3725 cm -I is also present and, even though aclear-cut assignment of the precise nature of such band could not be done, it has to benoted that it is due to the presence of Ti sites onto the surface of the MCM-41. TheTi-MCM41 sample also has a high concentration of hydroxyls (1.8 OH/nm2

) and it islikely that some silanols are close enough to Ti Lewis acid centres, and/or to Ti-OHgroups and this induces a downward shift of the stretching frequency with respect tothat of the free silanols (see the structures depicted in scheme 3), Hydroxyls of thistype should be more acidic than those on the pure MCM-41, and this is, in fact, whatwas found upon adsorption of ammonia (vide infra),

Scheme 3

DFT calculations on H3Si-OH and (H3SiOh-Ti-OH clusters suggest that thestretching vibrations of silanols and Ti-O-H groups are, respectively, at 3722 and3704 cm- I [21], It is remarkable that these values are downward shifted from theexperimental values of the hydroxyls found on the Ti-MCM41 by very similar amounts,23 cm -I for the free silanols and 21 cm -I for the band at 3725 cm -I, so suggesting thatthe band at 3725 cm- I should be assigned to Ti-OH species, However, we believe thatfurther experimental data and computational analysis is necessary to clarify the natureof the hydroxyls at 3725 cm- I more precisely, In particular, as DFT calculations arenormally associated with an error of about ± 5%, the good match between experimentand theory might be fortuitous,

3"2.2, The band at 935 cm -I, A very broad band at 935 cm- I (see fig, 2) is alsoobserved in the IR spectrum of the Ti-MCM41 which is a clear evidence for thepresence of Ti anchored onto the silica surface as similar absorptions have been foundin TS-1 [13,22], Ti-MCM41 [4,8,9], and also in amorphous Ti-silicates [11], In thesematerials, where Ti (IV) ions isomorphously substitute for Si within the siliceous


3743

j 935

T\Si.OH] b980 \ !,~~

3800 3700 3600 3500 3400 1000 950 900 850

Wavenumbers [em-1 ]

Fig, 2, - FTIR spectra of MCM-4I (curve a) and Ti-MCM4I (curve b) obtained after calcining andevacuating the samples at 550°C,

matrix, a band centered mainly at 960 cm-] is constantly found, and, as proposed byBoccuti et al. [13], related to vibrations in structures where Ti0 4 tetrahedra sharecorners with Si04 tetrahedra (scheme 1).

Recent ab initio quantum chemical studies have confirmed the assignment of theabsorption at 960 cm -] as being due to the antisymmetric Ti-O-Si stretchingvibration [23].

Interestingly, the band of the Ti-O-Si connectivity in the case of the Ti-anchoredMCM-4I is found at lower frequency than that of the framework-substituted Ti-silicatesand such an effect might be attributed to the presence of different ligands bound to thetetrahedral Ti sites. In the case of Ti-anchored MCM4I, as suggested by UV-Visreflectance data, it is more likely that one or two ligands are hydroxyls (structures 1and 2 in the scheme 1) rather than Si04 tetrahedra: structures 3 in the scheme 1 shouldbe more abundant in framework-substituted Ti-silicates.

3'2.3. The adsorption of NH 3 . The acidity of hydroxyl groups, both in MCM-4Iand Ti-MCM41, as well as the coordination of titanium centres in Ti-MCM41, wereprobed by using NH3 . On pure MCM-41, NH 3 adsorbs on silanol groups by meansof H-bonds forming very weak complexes similar to the one depicted in the schemeof fig. 3 [18]. The spectroscopic features of these complexes are well known: i) a broadband centred at 3030 cm-] due to the stretching of the H -bonded silanols, ii) bandsat 3405 and 3320 cm-] respectively due to the NH3 asymmetric and symmetricstretching vibrations and iii) a band at 1635 cm-] due to the NH3 asymmetric bendingmode. The ammonia is only weakly adsorbed on the silanols as demonstrated by the


1550

1605

MCM4l

3030I ,~.

340~ 3(20

3391 j 3292

,..-,

~~

'---J

(l) 3745u§

,.Dl-<0rfJ

,.D

~

3600 3200 2800 1750 1500

Wavenumbers [em-I]

Fig. 3. - FTIR spectra of pure MCM-41 and Ti-MCM41 in vacuo at 550°C (dotted lines) and uponadsorption of NH3 (curves 1, 30 Torr). Curves 1 to 16 for Ti-MCM41 and 1 to 13 for MCM-41correspond to decreasing doses of NH3 ; the last spectra were obtained by evacuating the samplesat room temperature for 30 minutes.

fact that all the features of the complex completely disappear, and the narrow bandat 3745 cm- I is restored, upon a simple desorption at room temperature.

However, bands at ca. 3500 (broad) and 1550 (very weak) cm -I are still present oncethe ammonia is desorbed from the surface of the MCM-41 and this clearly suggeststhat new surface species are formed. It is known that distorted surface siloxane bridgescan react with ammonia giving Si-OH and Si-NHz [24,25], and this would explain theappearance of the bands at 3500 and 1550 cm- I which are associated with the newlyformed OH (stretching) and the NHz (bending) groups, respectively.

All the species formed upon adsorption of ammonia on MCM-41 are also formed onTi-MCM41. However, in this latter case, the ammonia is not completely desorbed atroom temperature after a prolonged evacuation, there still being present bands at 3391,


TABLE L - Vibrational modes and wavenumbers (em-I) of the surface complexes formed uponadsorption ofNH3 on MCM-41 and on Ti-MCM4L

v SiO-H , .. NH3

O-H ... NH3

Ti ... NH3

aNHt

MCM-41 Ti-MCM413030 3040

vasym 3405 (8)V sym 3320 3320a asym 1635 1635

vasym 3391v sym 3292a asym 1605

1460

(a) This band cannot be measured with precision because it is ovelapped to the stronger absorption at 3391 em-I

3292, 1605 and ca. 1460 cm- 1. The first three bands can be assigned to NH3adsorbed on

Ti Lewis acid sites as similar complexes have been found also by UV-Vis spectroscopy:the LMCT absorption band at 230 nm of the tetrahedral Ti sites, similarly to the wateradsorption experiment shown in fig. 1, is modified upon NH3 adsorption and this effectis irreversible at RT. Significantly, the stretching and bending modes of NH3 adsorbedon Ti Lewis sites are shifted to lower frequencies than the bands of NH3 adsorbed onsilanols: table I summarizes these results.

The band at 1460 cm-\ not observed on pure MCM-41, can be assigned to the NHtspecies; the presence of these species suggests that on the surface of Ti-MCM41 thereis a fraction of hydroxyls which are sufficiently acidic to protonate some ammoniamolecules. Similar results have been also found for vanadia-based catalysts [26]. Thenature of the hydroxyls responsible of such a reaction is still under investigation.

Finally, the silicon-oxygen-titanium vibration at 935 cm -1 (not shown for the sake ofbrevity) is also deeply modified upon NH3 adsorption in that a blue shift up to ca.990 cm -1 occurs. The phenomenon has been attributed to the fact that when ammoniamolecules are bound to Ti the polarity of the Tio + _0° - -Si bonds is increased and,consequently, more negative charge is accumulated on the [03Si-0]0 - so that theabsorption moves towards the values typical of the stretching of "isolated" Si-Ogroups [15, 19,20,22]. An example of isolated Si-O vibration is that of Si-OH groupswhich in pure MCM-41 absorb at 980 cm- 1 (see fig. 2).

3'3. Photoluminescence data, - The photoluminescence technique has been appliedto the study of anchored titanium oxide catalysts [27-29] and TS-2 [30], particularly byAnpo and co-workers. Photoemission spectra containing features in the range of 400 to600 nm are reported and interpreted as deriving from isolated tetrahedral titaniumcentres. However, only in one paper [30] the corresponding reflectance spectra werereported and there they contain broad bands at A~ 300 nm which suggest that asignificant fraction of Ti sites are present either with coordination ranging from4 to 6 [13, 15, 17] or as TiOz-like phase [16,31].

In order to get unambiguous data for the emitting sites in Ti-containing materials,it is vital to study samples, as in the present case, which have been characterized by in


situ X-ray absorption spectroscopy [7], complemented by reflectance UV-Vis andFTIR spectroscopy (vide supra), and have been shown to contain a high degree oftetrahedral titanium sites.

The photoluminescence spectroscopy allows us to monitor both the ground and theexcited states by collecting the absorption and the emission of light associated withprocesses such as those described in eqs. (2) and (3):

(2)

(3)

with this technique we can probe the presence, if any, of different tetrahedral Ti sites(see scheme 1) in Ti-MCM41.

By exciting the Ti-MCM41 at 298 K with Aex = 250 nm, a complex emission bandwith two main maxima at 434 and 480 nm and two shoulders at ca. 400 and 525-550 nmis obtained (fig. 4A, curve a) and this reveals the presence of at least four differentemitting sites. The overall emission band is greatly quenched upon exposure to O2 at298 K (ca. 80% of the original intensity is lost, cf. curves d-t), meaning that a very largefraction of these sites are located on the MCM-41 silica surface; such an effect is nearlycompletely reversible when the oxygen is desorbed at 298 K (curve g). It might bestressed that no such emission bands are observed in the case of pure siliceousMCM-41 confirming that the emitting sites in Ti-MCM41 are the Ti(IV) centres.

The sites emitting at 434 and 480 nm have very similar excitation spectra (curves aand b) and both present a maxima at 248 nm and this strongly suggests that they stemfrom similar environments. The shift of the maximum in the excitation spectra (ca.250 nm) as compared with the reflectance band (ca. 230 nm) is not surprising. In fact, ithas to be considered that the absorption of tetrahedral Ti(IV) ions is complex [14] andthat the maximum efficiency in exciting the emission does not necessarily coincide withthe maximum absorption due to possible energy transfer processes among levels of thesame sites.

The photoemission of light by Ti sites is a temperature-dependent phenomenon ascan be clearly seen from an inspection of the section B of fig. 4, in which the spectrumrecorded at 77 K (A ex = 250 nm, curves a') is 20-30 times more intense than thatrecorded at 298 K (curve a). The relative intensities of the various components of thetwo spectra is also different being present, in that at low temperature, a very strongband at ca. 500 nm which arises probably from the great enhancement of the shoulderat 525-550 nm. The exposure to O2 at 77 K, 0.15 Torr in this case, nearly completelyquences the emission band (curve d ') confirming that within experimental error all theemitting sites are accessible to the O2 and are, hence, on the surface.

The presence of the larger contribution of sites emitting at 525-550 nm in the lowtemperature spectrum is also confirmed by the appearance in the excitation spectrum(A em = 520 nm, curve c') of a broad shoulder at ca. 270 nm which is not present neitherin the excitation spectra at 298 K nor in that at 77 K collected at Aem = 420 nm (curve b').

1716 L. MARCHESE, E. GIANOTTI, T. MASCHMEYER, ETC.

298K

leem= 434

b\ Ie = 480.. em

;:::i Ccd

L......J

700

B

600500400

17KA

600 700 200 300

Wavelength [nrn]

500400300200

Fig. 4. - Emission and excitation spectra of Ti-MCM4l at 298 (section A) and at 77 K (section B)under high vacuum (ultimate pressure ~ 10- 5 Torr) unless stated otherwise; section A: (a)emission spectrum for Aex = 250 nm, (b) Aem = 434 nm vs. excitation from 200-350 nm, (c) Aem =

480 nm vs. excitation from 200-350 nm, (d), (e) and (f) show the effect of O2 adsorptioncorresponding to 1.5, 20 and 80 Torr O2 respectively on the emission spectra at Aex = 250 nm,(g) shows near reversibility of O2 adsorption when desorbing O2 for 1 h at 298 K (pressure~ 10- 5 Torr); section B: (a') emission spectrum for Aex = 250 nm (spectrum (a) at 298 Kshown, multiplied by 10, for comparison), (b') Aem = 420 nm vs. excitation from 200-350 nm,(c') Aem = 520 nm vs. excitation from 200-350 nm, (d') emission spectrum forAex = 250 nm, corresponding to the adsorption of 0.15 Torr °2 ,

It should be stressed that even though a precise assignment of the variouscomponents cannot be made at this stage, the difference in the relative intensity of thebands at 77 and 298 K confirms that more than one tetrahedral titanium site is presentin Ti-MCM41. In addition, similarly to other high surface area solids [32], energytransfers occur between the sites and such effects are particularly relevant at 298 Kwhere the energy relaxation via non-radiative processes is also very important.

However, owing to the fact that the electron density of the oxygens in TiOH and onTiOSi is certainly different and, additionally, should change slightly when they arelocated in Ti(OH)(OSi)3 or Ti(OH)z(OSi)z surface sites (scheme 1), such groups shouldhave different spectroscopic features. We, therefore, propose that the various bands inthe emission spectra are due to the relaxation of the different oxygen to tetrahedralTi (IV) charge transfer transitions, requiring a minimum of two different tetrahedraltitanium sites.

Finally, it is remarkable that the Ti-MCM41 catalysts have UV-Vis luminescencewhose intensity is strictly related to the dispersion of the Ti(IV) ions onto the innersurface of MCM-41 [33] and such an effect has prompted us to correlate their catalyticactivity/selectivity to their luminescence yield.

PROBING THE TITANIUM CENTRES IN MCM41 CATALYSTS

* * *

1717

Italian ASP and CNR and British EPSRC are gratefully acknowledged for financialsupport. The authors thank Dr. R. OLDROYD for the help in preparing the Ti-MCM41and Prof. A. ZECCHINA for fruitful discussion.

REFERENCES

[1] TARAMASSO M.. PEREGO G. and NOTARI B.. U.S. Pat. 4410501. 1983.[2] TATSUMI T. NAHAMURA M.. NEGISHI S. and H. TOMINAGA. .J. Chem. Soc.. Chem. Commun.

(1990) 476.[3] NOTARI. B., Adv. Catal., 41 (1996) 252; THOMAS J. M., Phi/os. Trans. R. Soc. A. 333 (1990) 173;

see also Nature, 368 (1994) 289.[4] CORMA A., NAVARRO M. T and PEREZ PARIENTE J., .J. Chem. Soc.. Chem. Commun. (1994)

147.[5] TANEV P. T, CHIBWE M. and PINNAVAIA T J., Nature, 368 (1994) 321.[6] KRESGE C. T, LEONOWICZ M. E., ROTH W. J., VARTULI J. C. and BECK J. S., Nature, 359

(1992) 710.[7] MASCHMEYER T, REY F., SANKAR G. and THOMAS J. M., Nature, 378 (1995) 159.[8] BLASCO T, CORMA A., NAVARRO M. T and PEREZ PARIENTE J., .J. Catal 156 (1995) 65.[9] ALBA M. D., LUAN Z. and KLINOWSKI J.,.J. Phys. Chem., 100 (1996) 2178.

[10] SINCLAIR P., SANKAR G., CATLOW C. R. A., THOMAS J. M. and MASCHMEYER T,.J. Phys. Chem.B, 101 (1997) 4232.

[11] a) KLEIN S., THORIMBERT S. and MAIER W. F, .J. Catal., 163 (1993) 476; b) KESHAVARAJA A,RAMASWAMY V., SONI H. S., RAMASWAMY A. V. and RATNASAMY P . .J. Cata I. , 157 (1995)501.

[12] OLDROYD R. D., THOMAS J. M., MASCHMEYER T, MAC FAUL P. A., SNELGROVE D. W., INGOLDK. V. and WAYNER D. M., Angew. Chem., /nt Ed Eng, 35 (1996) 2787.

[13] BOCCUTI M. R., RAo K. M., ZECCHINA A., LEOFANTI G. and PETRINI G., Stud Surf Sci. Catal.,48 (1989) 133.

[14] a) JORGENSEN C. K., Prog /norg Chem., 12 (1970) 101; b) DUFFY J. A., .J Chem. Soc. DaltonTrans. (1983) 1475.

[15] BORDIGA S., COLUCCIA S., LAMBERTI C., MARCHESE L., ZECCHINA A., BOSCHERINI F., BUFFAF., GENONI F., LEOFANTI G., PETRINI G. and VLAIC G., .J. Phys. Chem., 98 (1994) 4125.

[16] LE Noc L., TRONG ON D., SOLOMYKINA S., ECHCHAHED B., BELAND F., CARTIER DIT MOULINC. and BONNEVIOT L., Stud Surf Sci. Cata I. , 101 (1996) 611.

[17] KLEIN S., WECKHUYSEN B. M., MARTENS J. A., MAJER W. F. and JACOBS P. A., .J. Cata I. , 163(1996) 489.

[18] SAUER J., UGLIENGO P., GARRONE E. and SAUNDERS V. R., Chem. Rev., 94 (1994) 2095.[19] MORROW B. A. and McFARLAN A. J., .J. Phys. Chem., 96 (1992) 1395.[20] ZECCHINA A., BORDIGA S., SPOTO G., MARCHESE L., PETRINI G., LEOFANTI G. and PADOVAN M.,

.J. Phys. Chem., 96 (1992) 4991.[21] SINCLAIR P., CATLOW C. R. A. and THOMAS J. M., unpublished results.[22] SCARANO D., ZECCHINA A., BORDIGA S., GEOBALDO F., SPOTO G., PETRINI G, LEOFANTI G.,

PADOVAN M. and TOZZOLA G., .J. Chem. Soc.. Faraday Trans., 89 (1993) 4123.[23] DE MAN A. J. M. and SAUER J., .J. Phys. Chem., 100 (1996) 5025.[24] VIKULOV K., MARTRA G., COLUCCIA S., MICELI D., ARENA F., PARMALIANA A. and PAUKSHTIS

E., Cata l Lett, 37 (1996) 235.[25] MORROW B. A. and CODY 1. A., .J. Phys. Chem., 80 (1976) 1998.


[26] a) RAJADHYAKSHA R A, and KNOZINGER H., Appl Catal" 51 (1989) 81; b) CENT! G"PERATHONER S" TRIFIRO F" ABOUKAIS A" AISSI C. F. and GUELTON M" j Phys. Chern., 96(1992) 2617,

[27] ANPO M" AIKAWA N" KUBOKAWA Y" CHE M" LOUIS C. and K GIAMELLO, j Phys. Chern., 89(1985) 5017,

[28] ICHIHASHI Y" YAMASHITA H. and ANPO M" Proceedings of the XI International ZeoliteConference, August 72-77, 7996, Seoul, Korea, IZC 149.

[29] YAMASHITA H., ICHIHASHI Y., ANPO M., HASHIMOTO M., LOUIS C. and CHE M., j Phys. Chern.,100 (1996) 16041, and references therein.

[30] ZHANG S. G., ICHIHASHI Y, YAMASHITA H., TATSUMI T and ANPO M., Chern. Lett., in press.[31] ZECCHINA A., SPOTO G., BORDIGA S., FERRERO A., PETRINI G., LEOFANTI G. and PADOVAN M.,

in Zeolite Chemistry and Catalysis, edited by P. A. JACOBS et al (Elsevier Science Publ.BV, Amsterdam) 1991.

[32] a) COLUCCIA S. and MARCHESE L., Proceedings of the International Symposium onAcid-Base Catalysis, Sapporo, Nov. 28-Dec 7, 7988, edited by K. TANABE, H. HATTORI, TYAMAGUCHI and T TANAKA (Kodansha, Tokyo) 1989; b) COLUCCIA S. and BORELLO K,Proceedings of the VI I I International Congress on Catalysis, Vol. 3 (Dechema, Frankfurt

am Main) 1984, p. 69.[33] MASCHMEYER T, MARCHESE L., COLUCCIA S., THOMAS J. M., REY F., CORMA A. et ai, in

preparation.

Spectroscopic tools for probing the isolated titanium centres in MCM41 mesoporous catalysts

Documents