The present invention relates to a catalyst based on zeolites, a process for their making and their use in the production of olefins from methanol.

Light olefins, in particular propylene, are important starting materials for many major processes in chemical industry. For example, propylene is an important starting material in the production of methionine and its hydroxyl-analogue. Therefore, the demand for light olefins continues to rise every year. The primary production of light olefins is based on either steam cracking or recovery from refinery processes. However, the existing steam cracking and refinery processes can barely meet the increasing demand for light olefins. Therefore, the prices for light olefins, in particular for propylene, have continued to rise and there is no trend reversal in sight. Thus, there is a need for alternative processes for the provision of light olefins. One of these alternative processes is the methanol-to-olefin reaction, also known in its short form as MTO reaction, which is performed in the presence of acidic zeolite based catalysts. When propylene is the major product of this reaction, it also referred to as the methanol-to-]Dropylene reaction (MTP). The MTO reaction has attracted extensive attention as an alternative route for the production of light olefins, because the starting material methanol is easily to get from the huge reserves of natural gas. Since its discovery in the late 1970s, a variety of zeolite catalysts for the MTO reaction as well as modification procedures for these catalysts have been investigated in order to obtain MTO catalysts which the highest possible selectivity for the formation of propylene. Extensive studies were done on the reaction paths in the MTO reaction over zeolite catalysts. They showed that there are two competing reaction mechanisms, depending on the specific type of zeolite catalyst used: An olefin- cracking-mechanism, which gives olefins such as propylene and ethylene, and a side-chain- mechanism, which gives unsaturated hydrocarbons with five and more carbon atoms, especially aromatics. Accordingly, it is preferred that the MTO reaction proceeds via the olefin-cracking- mechanism and thus, the occurrence of the side-chain-mechanism must be avoided or at least significantly reduced. Another important aspect is the deactivation of the catalyst during the MTO process. For example, the silicoaluminophosphate SAPO-34 with a pore size of ca. 0.43 nm is an excellent catalyst for the selective production of olefins, ethylene and propylene, in the MTO reaction. However, this catalyst is rapidly deactivated by coke accumulation in the internal narrow channels of the SAPO-34 crystals. By comparison, the zeolite H-ZSM-5 appears to be stable in the MTO reaction because its pores are of medium size (ca. 0.55 nm), which therefore are less susceptible to any blockage caused by coke accumulation. The disadvantage of the H-ZSM-5 zeolite however, is that its strong acid sites lead to the formation of large amounts of by-products, especially aromatics, which are formed through oligomerization reactions. Therefore, several modifications of acidic zeolites, such as H-ZSM-5 and H-MCM-22, were examined to provide catalysts, which are selective for the formation of light olefins and at the same time resistant towards deactivation in the MTO reaction. The published international patent application WO 2009/092781 A2 discloses zeolite catalysts, which were modified with phosphor and an alkaline earth or rare earth metal, and their use in the catalyzed MTO reaction. These catalysts are prepared by simple treatment of H-ZSM-5 catalysts having a known Si to Al ratio with phosphoric acid, followed by treatment with an aqueous solution of a calcium or lanthanum salt. Compared to the starting H-ZSM-5 zeolite, the modified zeolitic catalysts give an improved yield for olefins, in particular propylene, in the MTO reaction. It is believed that these results are due to a deactivation of strong acid sites in the zeolite because the incorporated calcium or lanthanum cations are less acidic than the original acid sites of the zeolites. However, the approach of the publication WO 2009/092781 A2 is not based on a systematic basis. Due to the presence of water, which is always formed as a by-product in the MTO reaction, a hydration shell is formed around the multivalent cations of calcium and lanthanum. Several side reactions can proceed in this hydration shell which then significantly affect the selectivity for the olefin formation and it is not foreseeable what side reactions may proceed and if so, to what extent. Therefore, it appears that the results presented in the publication WO 2009/092781 A2 are not reproducible and rather, appear to be based on fortune and selection.

Liu et al. studied the effect of the modification of a high silica H-ZSM-5 catalyst with phosphor on the propylene selectivity in the MTO reaction (Catalysis Communication 2009, 10, 1506-1509). The examined zeolite was a H-ZSM-5 catalyst with a Si to Al ratio of 220 which was modified by impregnation methods using salts of various main group metals or transition metals and phosphoric acid. The best propylene selectivity was observed when the employed zeolite was loaded with 0.1 percent by weight of phosphoric acid, based on the total weight of the catalyst. However, the authors of this paper did not further investigate the optimum loading of a zeolite with phosphorous. Therefore, it appears that the results of this paper are not based on a target-oriented and thus systematic approach how to modify a zeolite with respect to an optimized selectivity for the formation of propylene in the MTP reaction. Rather, it appears that the results of this paper are just the outcome of a mere screening for promoters which are beneficial for an improvement in propylene selectivity. Accordingly, this paper does not contain a real teaching how to modify an existing zeolitic catalyst with respect to an improved propylene selectivity.

Also investigated was the effect of phosphate modification on the zeolite framework, the pore system and the acid site densities of a H-ZSM-5 (M. Dyballa et al., Chemie Ingenieur Technik 2013, 85, No. 1 1 , 1719-1725). According to this paper the best selectivity for the formation of olefins was achieved with a H-ZSM-5 having a Si to Al ratio of ca. 21 which was impregnated with 5 percent by weight of phosphate. However, neither this specific catalyst nor the other catalysts of this paper appear to be particularly selective for the formation of a specific olefin, because this paper is completely silent with respect to the yield for individual olefins, especially with respect to propylene. Accordingly, this catalyst does not appear to be suitable for the use in an industrial MTO process, in particular not for an MTP process. D. V. Vu et al. also examined the effect of phosphate modification on the propylene selectivity of H- ZSM-5 catalysts (Journal of the Japan Petroleum Institute 2010, 53(4), 232-238). The best selectivity for the formation of propylene was obtained with a H-ZSM-5 catalyst having a Si to Al ratio of 155, which was loaded with 3 per cent by weight of phosphor, based on the total weight of the catalyst. However, the reaction conditions used for testing the thus modified zeolite-based catalysts in the MTO reaction make the test results rather questionable: the modified zeolite-based catalysts were always tested at a W/F ratio (mass of the catalyst [kg] divided by the feed rate of methanol [mol/h]) of 0.065 kg ^■ h ^■ mol \ which corresponds to a weight hourly space velocity (WHSV) of less than 0.5 h ^" . By comparison, in industrial applications the WHSV typically ranges from 1 to 6 h \ where a WHSV of 1 h ^" favors the propylene selectivity and a WHSV of 6 h ^~ favors the methanol conversion but at the same time leads to a deactivation of the catalyst. Due to the extremely low WHSV used in this paper the number of molecules in the pores of the zeolites was always very low, which favors the formation of propylene, independently from the specific catalyst used. Therefore, the reason for the high propylene selectivity observed in this paper is not the specific zeolite-based catalyst used but the very low WHSV. Accordingly, the good results in the paper do not allow the conclusion that the catalysts of this paper did have a high selectivity for the formation of propylene. Rather, it appears that the catalysts of this paper are not suitable for use in an MTP reaction on industrial scale.

Summarizing, a variety of different modified zeolite catalysts are known as well as a variety of processes for modifying zeolite catalysts. However, the known processes for the post-synthesis modification of zeolites only lead to random and thus not reproducible results with respect to the selectivity for the formation of propylene from methanol in catalyzed MTP reactions. Accordingly, so far it was not possible to modify zeolite catalysts with respect to an increase in propylene selectivity in MTP reactions in a target-oriented and accurate way by means of specific post-synthesis treatments.

It is therefore an objective of the present invention to provide a process for the preparation of zeolite- based catalysts for the methanol-to-olefin reaction, in particular the methanol-to-propylene reaction, which provides these catalysts with an optimized Br0nsted acid site density of the highest reproducibility.

It was found that this objective is solved by adjusting the density of the Br0nsted acid sites of a zeolite to a value of from ca. 0.08 mmol/g to ca. 0.4 mmol/g, determined by means of quantitative H MAS NMR spectroscopy.

By comparison, temperature-programmed desorption of ammonia (NH3-TPD) and infrared (IR) spectroscopy, in particular Fourier transform infrared (FTIR) spectroscopy, have been so far the most widely used methods for characterizing the acidity in zeolites. However, these methods do not allow an accurate determination of specific acid sites. The temperature-programmed desorption of ammonia typically involves saturating the surface of the zeolites with ammonia under specified adsorption conditions, followed by a linear ramping of the sample temperature in a flowing inert gas stream. The ammonia concentration in the effluent gas stream is followed by absorption/titration or mass spectrometry. Alternatively, the experiment is carried out with a microbalance, in particular a quartz crystal microbalance, with a zeolite having been loaded with ammonia and changes in sample mass are followed continuously. The amount of ammonia, which desorbs above a specific temperature, is taken as the acid-site concentration, and the peak desorption temperatures are typically used to calculate the heights of adsorption. However, the investigation of the acid site of zeolite by means of temperature-programmed desorption in general has several flaws. The major problem with temperature-programmed desorption of ammonia is that the small ammonia molecules will load all available acid sites, regardless whether they are Lewis or Br0nsted acid sites. For example, ammonia may be also physically adsorbed in addition to the ammonia which is chemically adsorbed at Br0nsted acid sites as ammonium ions. Therefore, the temperature-programmed desorption of ammonia cannot distinguish between the Lewis and Br0nsted acid sites. Consequently, this method always gives too high values for the Br0ndsted acid site density. Another problem with temperature-programmed desorption of ammonia is that the quantification is rather complicated, and in general it is too high, which also contributes to the high inaccuracy of this method. Yet another problem with temperature-programmed desorption of ammonia is that the adsorption of ammonia of ammonia at acid sites and the desorption of ammonia from these sites are equilibrium processes which take a long time. In order to get reliable values for the heights of adsorption, the ammonia concentration in the effluent gas stream may only be measured after eastablishment of the equilibrium. However, since the establishment of equilibrium in temperature-programmed desorption of ammonia takes long, it is typically not awaited and thus the results obtained by means of NH3-TPD are too high. Alternatively, temperature-programmed desorption can also be performed with pyridine instead of ammonia. Compared to ammonia, pyridine is larger and therefore, it is rather questionable whether this relatively large molecule can reach all Br0nsted acid sites. As is already the case with the NH3-TPD, the quantification for the temperature- programmed desorption of pyridine is rather complicated and generally gives too high values. Infrared spectroscopy of zeolites acidity is based on the change of O-H stretching frequencies observed upon loading a zeolite with a small molecule. In practice, zeolites are loaded through titration with pyridine. However, like the temperature-programmed desorption with pyridine this method is rather problematic because it is not clear whether the relatively large pyridine molecules really reach all (available) Br0nsted acid sites. Another major problem with infrared spectroscopy is again the calibration, which strongly depends on the specific acid sites of the zeolite in question. Depending on the specific acid sites which are the basis for the analysis by infrared spectroscopy the results vary by a factor of up to ca. 70 %. This is due to the fact that the absorption coefficient is a function of the loading of the acid sites with a probe molecule and of the wave number of the used light. The reason for this is that the analysis by infrared spectroscopy is based on the Lambert-Beer law. According to the Lambert-Beer law the absorbance Ελ of a material for a specific light of the wavelength λ is described by the equation ΕΛ = ε * c * d, wherein c is the molar concentration of the absorbing substance in a fluid (mol/l), d is the thickness of the irradiated body, and ελ is the decadic extinction or spectral absorption coefficient for the wavelength λ, which is a specific physical quantity for every individual absorbing substance. Therefore, the specific extinction coefficient had to be determined for every single type of zeolite in order to obtain accurate and reproducible results for the determination of the Bransted acid site density by means of infrared spectroscopy. However, the extinction coefficients of zeolites cannot be accurately determined and therefore, infrared spectroscopy is not suitable for determination of the Br0nsted acid site density. Regarding the different extinction coefficients of zeolites for the quantification by infrared spectroscopy references is made to Karge, H. G., Weitkamp, J., Molecular Sieves 4, Science and Technology, Characterization I, Springer, 2004. The extinction coefficients of specific zeolites listed in the tables of this reference strongly vary and thus, illustrate how vague and inaccurate the quantification by infrared spectroscopy is. A further major problem when trying to apply any kind of formalism to acidity in zeolites is that zeolites are not molecules with only one single type of acidic proton. Rather, zeolites contain a variety of different types of proton donor sites within a continuous framework. Accordingly, there is not a range of proton affinities for a given zeolite, but these proton affinities also change when they are loaded with a probe molecule. For example, the commonly used zeolite H-ZSM-5 contains at least three different types of hydroxyl groups. The first hydroxyl group is the acidic (AI-(OH)-Si) hydroxyl group which is observed at a wavenumber of 3,605 cm ^-1 . However, the intensity of this band decreases with increasing content of alumina in the H-ZSM-5. Further, this band disappears when the zeolite is transformed to its ammonium form. A second hydroxyl group is observed at a wavenumber of 3,740 cm ^-1 in H-ZSM-5 and in almost all other silicas. This band is considered to be associated with isolated Si-OH at the exterior of the crystallites. Most H-ZSM-5 samples also show a third broad band in the IR spectrum centered at approximately 3,500 cm ¹ , which appears to be related to defects in the crystalline structure, possibly nested silanols which would be needed for charge compensation at a cation vacancy. Accordingly, infrared spectroscopy is only suitable for recognition of specific frequencies resulting from the loading of zeolite with a probe molecule. However, this method gives only comparative data for Bransted and Lewis acid sites. Summarizing, infrared spectroscopy is not suitable for quantitative measurement and in particular not for quantitative determination of the density of a type of acid site, and especially not the density of the Bransted acid sites.

Obviously, so far experts in the field of zeolite-based catalyst for the MTP have not thought of any correlation between the selectivity for the formation of propylene and specific values of the Bransted acid side density. It is believed that the lack of knowledge of this correlation is due to the fact that there was so far no reliable method which facilitates the precise and reproducible adjustment of the Br0nsted acid site density in zeolite catalysts. Accordingly, all statements in the literature with respect to the influence of post-synthesis modification of zeolite-based catalysts on propylene selectivity in MTP reactions are not reliable, when they are based on infrared spectroscopy or tern peratu re- programmed desorption of ammonia.

By comparison, H MAS NMR spectroscopy is a powerful tool for the direct investigation of hydroxyl groups. In general, protonated zeolites, i.e. zeolites which are present in their H-form, often also referred to as their protonated form, contain different hydroxyl groups. Br0nsted protons are responsible for the acidity of the zeolite in question, and their number can be decreased by a dealumination treatment. Silanol groups are present on the external surface, in the mesopores and at defect sites, where the zeolitic framework is interrupted. Alumina species with hydroxyl groups on their surface are present in the pores. Their presence may be due to that during the synthesis not all the aluminum atoms originally present in the synthesis gel are introduced into the zeolite lattice or, after dealumination, only some of the dealuminated aluminum atoms are completely extracted from zeolite crystals. All of these hydroxyl groups can be investigated by means of H MAS NMR spectroscopy. Like in IR or FTIR spectroscopy and Nh -TPD, the H MAS NMR spectroscopy involves the saturation of hydroxyl groups of the zeolite with ammonia. However in contrast to IR or FTIR spectroscopy, H MAS NMR spectroscopy does not require the knowledge of the specific coefficient of a zeolite, which is a function of the loading with ammonia and of the used wavelength. Further, H MAS NMR spectroscopy allows the distinction between the different acid sites of a zeolite: the ammonia molecules absorbed at the Br0nsted acid sites are converted to ammonium ions, which are well defined and highly symmetrical compounds. The presence of this ion type at the Br0nsted acid sites leads to a decrease in chemical anisotropy, which results in spectra with very narrow and sharp signals. As a result the signals of the hydroxyl groups, which are loaded with ammonia, do not overlap in the thus obtained spectra or at least they only overlap to an insignificant degree and this only in rare cases. Accordingly, in contrast to the aforementioned analysis methods the quantitative H MAS NMR spectroscopy allows to make a distinction between the different hydroxyl groups of a zeolite. For example, the signals of the acidic bridge-hydroxyl-groups of a ZSM-22 zeolite are observed at 6.2 and 4.1 ppm and the signals of the SiOH groups are observed at 1.8 and 2.2 ppm. After loading of the dehydrated zeolite with ammonia the signals of the corresponding ammonium ions are observed at 6.8 ppm. By comparison, the signals of the SIOH groups are not influenced by ammonia due to their lower acid strength.

In addition, modern H MAS NMR spectroscopy also allows the quantitative measurement of spectra. More specifically, modern H MAS NMR spectroscopy allows to determine the direct dependency of the signal intensity on the individual spin number. This allows the accurate determination of the concentration or density of specific hydroxyl groups in a reliable and reproducible way. In addition, solid state NMR spectroscopy has also proven to be particularly suitable for zeolites because of the rigid and crystalline nature of zeolites.

The specific Bransted acid site density can already be set in the direct synthesis of the zeolite or by modifying a zeolite in a post-synthesis treatment. Said post-synthesis treatment of a zeolite can be i) the loading of a zeolite with phosphate and/or ii) the partial exchange of ions of the zeolite. It is also possible to combine the direct synthesis with any one of the post-synthesis treatments i) to ii) or to combine the post-synthesis treatments i) and ii) with each other. One object of the present invention is therefore a process for the preparation of a zeolite-based catalyst for the preparation of propylene from methanol, comprising the step of setting the Bransted acid site density of a zeolite to a value of from ca. 0.08 mmol/g to ca. 0.4 mmol/g, wherein said Br0nsted acid site density is set in the direct synthesis of a zeolite, wherein the direct synthesis of the zeolite comprises the steps of a1 ) synthesizing a zeolite with a framework Si to Al ratio of from ca. 55: 1 to ca. 210: 1 and a2) converting said zeolite with a solution of a diluted mineral acid or an ammonium solution into the H-form of said zeolite, followed by drying and/or calcining the thus obtained zeolite, and/or by modifying a zeolite in a post-synthesis treatment selected from the group consisting of i) loading the zeolite with phosphate, and/or ii) partially exchanging ions of the zeolite, wherein the Br0nsted acid site density is determined by means of quantitative H MAS NMR spectroscopy, and wherein the step a2) and/or the post-synthesis treatment is/are repeated until the zeolite-based catalyst has a Br0nsted acid site density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g.

In context of the present invention the term Br0nsted acid is used according to the Bransted definition of an acid as a species with a tendency to give up a proton (J. Bransted, J. Reac. Tran. Chim. 1923, 42, 718). The Bransted acidity is a quantifiable concept and the use a pK _a scales for aqueous solutions or proton affinities for gas-phase acid/base reactions facilitates the determination of the strength of an acid or base. However, the real problem when trying to determine the Bransted acidity is the lack of an acceptable scale of solid acidity comparable to pK _a scales for aqueous solutions or proton affinities for gas-phase acid/base reactions. Further, NH3-TPD and FTIR spectroscopy are not able to either distinguish between Br0nsted and Lewis acid sites or to give accurate quantitative statements with respect to the density of each of these acid sites.

In catalyzed MTP reactions Br0nsted acid sites transfer protons of strongly acidic hydroxyl groups to methanol molecules, and the protonated methanol molecules further react under elimination of water to propylene and other light olefins. Consequently, the catalytic performance of zeolites in catalyzed MTP reactions strongly depends on the acid strength and the accessibility of the Br0nsted acid sites. In general, zeolites in their protonated form contain different kinds of hydroxyl groups, which can in principle be distinguished in acidic and non-acidic sites. Accordingly, a clear distinction between acidic and non-acidic sites is only possible when a complete proton transfer from the catalyst surface to the adsorbate molecule is facilitated and this requires the use of strongly basic probe molecules. By comparison, the use of weak basic molecules is more suitable for the discrimination between Br0nsted acid sites with different acid strengths. The adsorption of ammonia as probe molecules is a suitable method for the quantification of the number of strong Br0nsted acid sites. Ammonium ions are formed upon their adsorption at accessible and strong hydroxyl groups, and a narrow H MAS NMR signal is observed at δ-m = 6.0 - 7.0 ppm. Due to this shift, the respective signal does not overlap with the signals of other hydroxyl groups of zeolites, for example of the SiOH hydroxyl group which is observed at a shift of ca. 1.8 to 2.3 ppm. Further, this method allows a determination of the number of accessible Bransted acid sites with high accuracy, because the resulting signal for the ammonium ions has a narrow line width and an intensity, which is four times larger than the H MAS NMR signal of the acidic hydroxyl groups in the unloaded material. For quantitative studies, the measuring time of the NMR experiments has to be large compared to the spin-lattice relaxation times of the different hydroxyl groups, which typically are in the range of from 1 to 10 seconds.

Accordingly, the term quantitative H MAS NMR spectroscopy is used in context of the present invention to denote any H MAS NMR spectroscopy which is performed with a strongly basic probe molecule for a duration which is longer than the spin-lattice relaxation times of the different hydroxyl groups in the examined zeolite. Preferably, a quantitative H MAS spectroscopy according to the present invention is performed with ammonia as probe molecule for a time of more than 10 seconds. The terms NMR, H NMR and MAS NMR spectroscopy (magic angle spinning, MAS) either alone or in combination with each other are used as known to the person skilled in the art.

The determination of the Br0nsted acid site density is typically performed by the following procedure: The MAS NMR spectra are measure during the rotation of the zeolite sample in the solid state around the so-called magic angle, the value of which is approximately 54.7356°, with a Bruker Avance III 400 WB-spectrometer. The spectra are analyzed with the Bruker software TopSpin, WINFIT and WINNMR. All measure catalysts or zeolites are measured in a rotor made of zirconium oxide with a diameter of 4 mm and at a rotation speed of 8 kHz. In case the H-form of the zeolite-based catalyst was prepared by means of an ammonium exchange, a complete deammoniation is performed at a vacuum pump set, and the dehydration of the H-form of the zeolite is performed. For this purpose, it is suitable to perform for example an evacuation of the catalyst material at a temperature of ca. 450°C for ca. 12 hours. Then, the acidic zeolite-based catalyst is saturated with ammonia at room temperature, for example at an ammonia pressure of ca. 100 mbar for ca. 20 minutes. Subsequently, the physisorbed ammonia is desorbed by means of evacuating the ammonia-loaded zeolite-based catalyst at a temperature of ca. 180°C for ca. two hours. The determination of the density of the Br0nsted acid sites of the zeolite-based catalyst is done by means of H MAS NMR spectroscopy of the catalyst samples loaded with ammonia and the quantitative examination of the integral intensities of the H MAS NMR signals of the ammonium ions at a chemical shift of 6.0 to 7.0 ppm by means of comparison with the signal intensities of zeolite H, Na-Y with a Bransted acid density of 1.77 mmol/g as an intensity standard. Here, the four hydrogen atoms in the ammonium ions correspond to one Br0nsted acid site of the dehydrated zeolite-based catalyst.

In context of the process according to the present invention the term ca. 0.08 mmol/g is used to denote deviations of +/- 0.02 mmol/g from the explicitly mentioned value. Thus, the term ca. 0.08 mmol/g includes all real number values of from 0.06 mmol/g to 0.1 mmol/g, in particular the values 0.06, 0.07, 0.08, 0.09 and 0.10 mmol/g. In context of the process according to the present invention the term ca.0.4 mmol/g is used to denote deviations of +/- 0.04 mmol/g from the explicitly mentioned value. Thus, the term ca.0.4 mmol/g includes all real number values of from 0.36 mmol/g to 0.44 mmol/g, in particular the values 0.36, 0.37, 0.38, 0.39, 0.40, 0.41 , 0.42, 0.43 and 0.44 mmol/g.

According to the process of the present invention the zeolite-based catalyst can already be provided with a Br0nsted acid site density of from ca.0.08 mmol/g to ca.0.4 mmol/g in the direct synthesis of said zeolite. For this purpose, zeolites with 10-ring channels and a framework Si to Al ratio of from ca.55:1 to ca.210:1 are prepared in a direct synthesis (step a1). The framework Si to Al ratio can be determined by use of wet chemical methods, for example by means of inductively coupled jDlasma optical emission spectrometry, also known as ICP-OES, and the incorporation of aluminum in the zeolite framework can be followed and checked by means of ²⁷ AI MAS spectrometry. The thus obtained crude zeolite is then typically subjected to a calcination. This heat treatment removes the water still being present in the pores of the zeolite and the residues of the preparation procedure, in particular the template compounds which facilitate the formation of the zeolite structure. Thus, the term synthesizing a zeolite and the term direct synthesis of a zeolite comprise all preparation steps which are necessary in order to provide a zeolite with or without a calcination step of the crude zeolite. After its synthesis, the zeolite is converted into its H-form by treatment with a solution of a diluted mineral acid or an ammonium solution to provide the zeolite with a Bransted acid site density of from ca.0.08 mmol/g to ca.0.4 mmol/g, followed by drying and/or calcining (step b1). Finally, the Bransted acid site density of the zeolite is determined by means of quantitative H MAS NMR spectroscopy. If the zeolite does not have the required Br0nsted acid site density, the step b1) will be repeated until the zeolite has a Br0nsted acid site density of from ca.0.08 mmol/g to ca.0.4 mmol/g.

In context of the process according to the present invention the term ca.55:1 is used to denote all deviations of +/- 10% of the explicitly mentioned value. Thus, the term ca.55:1 includes all integral and real number values of from 50:1 to 60:1, in particular the values 50:1, 51:1, 52:1, 53:1, 54:1, 55:1, 56:1, 57:1, 58:1, 59:1 and 60:1. In context of the process according to the present invention the term ca.210:1 is used to denote all deviations of +/- 10% of the explicitly mentioned value. Thus, the term ca.210:1 includes all integral and real number values of from 189:1 to 231:1, in particular, the values 189:1, 190:1, 191:1, 192:1, 193:1, 194:1, 195:1, 196:1, 197:1, 198:1, 199:1,200:1,201:1, 202:1, 203:1, 204:1, 205:1, 206:1, 207:1, 208:1, 209:1, 210:1, 211:1, 212:1, 213:1, 214:1, 215:1, 216:1, 217:1, 218:1, 219:1, 220:1, 221:1, 222:1, 223:1, 224:1, 225:1, 226:1, 227:1, 228:1, 229:1, 230:1 and 231:1.

The synthesis of the zeolite is in principle not limited to any specific procedure and thus can be performed as known to the person skilled in the art, for example as described in Verified Syntheses of Zeolitic Materials, H. Robson, 2 ^nd edition, Elsevier, Amsterdam 2001. The Bransted acid density of a given zeolite can also be set to the required value by loading said zeolite with a phosphate. In principle, the treatment of a zeolite with a phosphate containing solution can be done with any phosphate containing compound, for example with a solution containing phosphoric acid, diammonium hydrogen phosphate, ammonium dihydrogen phosphate or a metal salt of dihydrogen phosphate, hydrogen phosphate or phosphate. Preferably, the loading of a zeolite with phosphate is done using ammonium dihydrogen phosphate. The loading of the zeolite with phosphate is performed until the content of phosphate loaded onto the zeolite has a value of ca. 10 percent by weight at the most, based on the total weight of the catalyst, and determined by means of inductively coupled jDlasma atomic emission spectroscopy (ICP-AES), which is also known as inductively coupled jDlasma optical emission spectroscopy (ICP-OES). The term ca. 10 percent by weight is herein used to denote all deviations of +/-1 percent from the explicitly mentioned value and thus, includes all integral and real number values from 9 to 1 1 percent by weight, in particular the values 9, 10 and 11 percent by weight. The modified zeolite is subsequently dried and/or calcined. Afterwards, the thus obtained zeolite is checked whether its Br0nsted acid site density has the required value of from ca. 0.08 mmol/g to ca. 0.4 mmol/g. If the Br0nsted acid site density of the zeolite does not have the required value, it will be necessary to wash the zeolite, preferably with demineralized water, until the zeolite has a Bransted acid density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g.

Thus, in an embodiment of the process according to the present invention, the loading of a zeolite with phosphate comprises the following steps of: a3) treating a zeolite with a phosphate comprising solution, b3) if necessary, washing the zeolite obtained in step a3), and c3) drying and/or calcining the zeolite, wherein the sequence of steps a3) to c3) is performed at least once to provide a zeolite-based catalyst with a Bransted acid density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g.

In the context of the process according to the present invention the treating of a zeolite with a phosphate comprising solution in step a3) comprises i) impregnating a zeolite with a phosphate comprising solution, preferably an excess of a phosphate comprising solution, and ii) a thermal treatment of the impregnated zeolite. The sequence of these sub-steps leads to a loading of the zeolite with phosphate, followed by a reaction of the phosphate with the acid sites or the structural features of the zeolites to give a permanent loading of the zeolite with phosphor or a phosphate. Accordingly, the second sub-step leads to a permanent binding of the phosphor or phosphate to the zeolite. Therefore, the thermal treatment of the impregnated zeolite is preferably performed as a calcination. In any case the Br0nsted acid site density of the post synthesis-modified zeolite is always determined after a complete run of the sequence of steps a3) to c3). Therefore, it is also possible to examine the impact of a definite amount of a phosphate on the Br0nsted acid site density of a zeolite: first, the Bransted acid site density of a directly synthesized zeolite is determined by means of quantitative H MAS NMR spectroscopy. The thus obtained zeolite is subjected to step a3) with a definite amount of a specific phosphate, followed by steps b3) and c3). Then, the Br0nsted acid site density of the post synthesis-modified zeolite is determined again by means of quantitative H MAS NMR spectroscopy and the correlation between the definite amount of the phosphate used is correlated with the determined values of the Br0nsted acid site densities, e.g. as a graph in a diagram. This diagram allows to determine the amount of the specific phosphate which is necessary to provide the directly synthesized or other post synthesis-modified zeolites with the required or other desired values of the Br0nsted acid site density. Therefore, in a preferred embodiment of the process according to the present invention the loading of a zeolite with phosphate therefore further comprises the additional steps of determining the Br0nsted acid site density of the chosen zeolite prior to step a3), correlating the Br0nsted acid site density measured prior to step a3) and after the complete run of steps a3) to c3) with the amount of the phosphate used, and

determining the necessary amount of the specific phosphate in order to provide the zeolite with a Bransted acid site density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g.

In context of the process according to the present invention the loading of a zeolite with a phosphate and in particular, the treating of a zeolite with a phosphate comprising solution is not limited to any specific treatment. Rather, the term treating a zeolite with a phosphate comprising solution is used to denote all techniques through which a zeolite is contacted with a liquid containing the phosphate to be loaded onto the zeolite. Preferably, the phosphate is loaded onto the zeolite by means of impregnation. In context of the process according to the present invention impregnation can be performed through any of the following procedures: i) Impregnation by soaking or with an excess of the solution, ii) dry or pore volume impregnation, iii) incipient wetness impregnation, iv) deposition by selective reaction with the surface of the zeolite, v) impregnation by percolation, vi) co- impregnation and vii) successive impregnation (these procedures are known to the person skilled in the art of zeolites and are further described in the Manual of Methods and Procedures for Catalyst Characterization, prepared for pulication by J. Haber, J. H. Block and B. Delmon, Pure & Appl. Chem., 1995, Vol. 67, Nos. 8/9, 1257-1306).

Further, it is also possible to provide a zeolite in its cationic form or in its H-form with the required Bransted acid site density through partial exchange of ions of the zeolite. When the zeolite is present in its cationic form, for example in its sodium form, said zeolite is treated with a solution of a diluted mineral acid or with an ammonium solution, and when the zeolite is present in its H-form, it is treated with a solution of alkali metal ions. Afterwards, the thus treated zeolite is dried and/or calcined.

Accordingly, in an alternative embodiment of the process according to the present invention, the partial ion exchange of the zeolite comprises the following steps of: a4) treating a zeolite present in its cationic form with a diluted mineral acid or an ammonium solution, or a4') treating a zeolite present in its H-form with an alkali metal ion containing solution, b4) if necessary, washing the zeolite obtained in step a4) or a4'), and c4) finally drying and/or calcining the zeolite, wherein the sequence of steps a4) or a4') to c4) is performed at least once to provide a zeolite-based catalyst with a Br0nsted acid density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g.

The process according to the present invention is not subject to any limitations regarding the use of an alkali metal ion containing solution in step a4'). However, in order to facilitate the exchange of protons of the zeolite it is advantageous to use small alkali metal ions. Therefore, the alkali metal in step a4') is preferably lithium, sodium and/or potassium, particularly preferred is sodium.

In any case the Br0nsted acid site density of the post synthesis-modified zeolite is always determined after a complete run of the sequence of steps a4) or a4') to c4). Therefore, it is also possible to examine the impact of a definite amount of a mineral acid or ammonium compound (in case of step a4)) or of an alkali metal ion (in case of step a4')) on the Br0nsted acid site density of a zeolite: first, the Br0nsted acid site density of a directly synthesized zeolite is determined by means of quantitative H MAS NMR spectroscopy. Then said zeolite is subjected to a partial ion exchange by treatment with a definite amount of specific mineral acid or ammonium compound in step a4) or by treatment with a definite amount of an alkali metal ion in step a4'), followed by steps b4) and c4). Then, the Br0nsted acid site density of the post synthesis-modified zeolite is determined again by means of quantitative H MAS NMR spectroscopy. The definite amount of the mineral acid or ammonium compound (in case of step a4)) or of an alkali metal ion (in case of step a4')) used is correlated with the determined values of the Br0nsted acid site densities measured, e.g. as a graph in a diagram. This diagram allows to determine the amount of the mineral acid or ammonium compound (in case of step a4)) or of an alkali metal ion (in case of step a4')) which is necessary to provide the directly synthesized or other post synthesis-modified zeolites with the required or other desired values of the Br0nsted acid site density. Therefore, in a preferred embodiment of the process of the present invention the partial ion exchange of the zeolite further comprises the additional steps of determining the Br0nsted acid site density of the chosen zeolite prior to step a4) or a4'), - correlating the Br0nsted acid site density measured prior to step a4) or a4') and after the complete run of steps a4) or a4') to c4) with the amount of the mineral acid or ammonium compound (in case of step a4)) or of an alkali metal ion (in case of step a4')) used, and determining the necessary amount of the specific mineral acid or ammonium compound (in case of step a4)) or of an alkali metal ion (in case of step a4')) in order to provide the zeolite with a Br0nsted acid site density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g.

It was further found that the Br0nsted acid site density of a zeolite can also be set to the desired value of from ca. 0.08 mmol/g to ca. 0.4 mmol/g through a combination of the direct synthesis of a zeolite with at least one post-synthesis modification of a zeolite selected from the group consisting of i) partially dealuminating the zeolite, ii) loading the zeolite with phosphate, iii) partially exchanging ions of the zeolite, and a combination of at least two of said post-synthesis modifications i) to iii). This combination has the advantage that it allows to provide zeolite catalysts with a Bransted acid site density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g, where the direct synthesis of a zeolite or a post- synthesis treatment of zeolite alone is not sufficient to provide the zeolite with the desired or required Br0nsted acid site density. Any of the aforementioned combinations is typically performed in that either the direct synthesis or a post-synthesis treatment is performed first, followed by a specific post- synthesis treatment and optionally a further post-synthesis treatment either identical to or different from the treatment performed before and after drying and/or calcination the Bransted acid site density is determined by means of quantitative 1 H MAS NMR spectroscopy.

Therefore, in a preferred embodiment of the process according to the present invention the direct synthesis of a zeolite is combined with at least one of the post-synthesis treatments selected from the group consisting of i) partially dealuminating the zeolite, ii) loading the zeolite with phosphate, and iii) partially exchanging ions of the zeolite.

In an alternative preferred embodiment of the process according to the present invention at least two of the post-synthesis treatments selected from the group consisting of i) partially dealuminating the zeolite, ii) loading the zeolite with phosphate, and iii) partially exchanging ions of the zeolite are combined.

Preferably, the direct synthesis of the zeolite is combined with the loading of said zeolite with phosphate, optionally followed by washing, or the direct synthesis of the zeolite is combined with the loading of said zeolite with phosphate and the thus obtained zeolite is subjected to a partial ion exchange, optionally followed by washing. From a practical point of view it is preferred to sequentially perform the combination of the direct synthesis of the zeolite with the loading of said zeolite with phosphate, optionally followed by washing, or to sequentially perform the combination of the direct synthesis of the zeolite with the loading of said zeolite with phosphate and subjecting the thus obtained zeolite to a partial ion exchange, optionally followed by washing, wherein the second treatment is always started after completion of the first treatment. However, from a theoretical point of view it is also conceivable to perform the individual treatments at the same time.

In principle, the process according to the present invention is not limited regarding the structure of the zeolite which is subjected to the specific treatment through which the specific Bransted acid density is adjusted. However, the process according to the present invention is particularly suitable for zeolites with 10-ring channels. In addition, zeolites having 10-ring channels also proved to be advantageous catalysts in the MTP reaction.

Therefore, in one embodiment of the process according to the present invention the zeolite used has 10-ring channels.

Zeolites with 10-ring channels, which proved to be suitable for the process according to the present invention, are the ZSM-5, ZSM-1 1 and the ZSM-22 zeolites.

A disadvantage of ZSM-22 zeolites is that their use as catalysts in MTP reactions is limited to specific process parameters, in particular to a specific weight hourly space velocity (WHSV). The WHSV is one of the economically most relevant parameters of catalyzed processes on industrial scale because it represents the quotient of the mass flow of the reactants divided through the mass of the catalysts in the reactor. A high WHSV is therefore preferred in order to keep the reactor volume as small as possible. However, it was observed that an economically advantageous WHSV of 4 h ^~ leads to an immediate deactivation of ZSM-22 zeolites in catalyzed MTP reactions. By comparison, ZSM-22 zeolite catalysts work without any problems at a relatively low WHSV of ca. 0.5 h ^~ . In general, the lower the WHSV is, the more propylene is formed by ZSM-22 zeolites. It is believed that this is due to the specific structure of the ZSM-22 zeolites, which is characterized by pores with elongated channels and by comparison, ZSM-22 zeolites has fewer pore openings than zeolites of the ZSM-5 or ZSM-1 1 type. The lower the WHSV is, the lower the amount of molecules in the channels of the ZSM-22 zeolites. It is believed that this favors the formation of propylene and at the same time it reduces the formation of aromatics, which is believed to result from further proton transfer reactions. By comparison, it is believed that the larger amount of molecules at a higher WHSV does not only favor the formation of higher olefins and aromatics over the light olefins but also a coking. Since zeolites of the ZSM-22 type have fewer openings for the channels than ZSM-5 or ZSM-1 1 zeolites, they are more susceptible for any blocking of the pores and/or channels due to the depositions resulting from the coking. Therefore, ZSM-22 zeolites are less suitable for industrial applications and in particular for use as catalysts in MTP reactions than ZSM-5 or ZSM-1 1. Accordingly, ZSM-5 or ZSM-1 1 zeolites are advantageous for MTO and MTP reactions. Accordingly, in a preferred embodiment of the process according to the present invention the zeolite has a structure of the ZSM-5 or ZSM-1 1 type.

Further preferred are zeolites of the ZSM-5 type which always show a good catalytic performance in the MTP reaction.

It was found that independently from the chosen type of modification of the zeolite, a Bransted acid site density can always or at least in most cases be set to a value of from ca. 0.08 mmol/g to ca. 0.4 mmol/g, when the process according to the present invention starts from a zeolite with a framework Si to Al ratio of less than 55: 1 .

Therefore, in one embodiment of the process according to the present invention the zeolite, which is subjected to at least one of the post-synthesis treatments, has a framework Si to Al ratio of less than ca. 55:1.

As already discussed above in detail, a zeolite which is particularly suitable for use in the MTO or MTP reaction has a Bransted acid site density of ca. 0.08 mmol/g to ca. 0.4 mmol/g. Since the process for the preparation of zeolite-based catalysts according to the present invention allows the accurate determination of the Bransted acid site density, the present invention therefore, also provides a method for selecting zeolite-based catalysts for their use in the methanol-to-olefin process, wherein the zeolite-based catalysts suitable for use in the methanol-to-olefin process have a Br0nsted acid site density of ca. 0.08 mmol/g to ca. 0.4 mmol/g, wherein the Bransted acid site density is determined by means of quantitative H MAS NMR spectroscopy. The preparation process and the selection method according to the present invention give access to zeolite-based catalysts which are particularly suitable for the MTP reaction because they have a Br0nsted acid site density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g and 10-ring channels. Therefore, the present invention also provides for zeolite-based catalysts obtainable or obtained by a process according to the present invention and for zeolite-based catalysts selected through the method according to the present invention.

Another object of the present invention is therefore a catalyst on the basis of a zeolite having 10-ring channels, characterized in that said catalyst has a Br0nsted acid site density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g, which is determined by means of quantitative H MAS NMR spectroscopy.

In one embodiment of the catalyst according to the present invention, the zeolite has a structure of the ZSM-5, ZSM-1 1 of ZSM-22 type.

Preferably, the zeolite has a structure of the ZSM-5 or ZSM-1 1 type, particularly preferably the ZSM- 22 type. In another embodiment of the catalyst according to the present invention, the zeolite-based catalyst contains from ca. 0.5 to ca. 10 percent by weight of a phosphate, based on the total weight of the catalyst.

In yet another embodiment of the catalyst according to the present invention, from ca. 5 to ca. 20% of the hydrogen atoms in the framework of the zeolite are replaced with an alkali metal.

In principle, the catalysts according to the present invention are not subject of any limitation regarding a specific alkali metal. However, alkali metals whose corresponding ions are small, are preferred because the protons of zeolites are easier replaced with smaller alkali metal ions, such as lithium, sodium or potassium, than with bigger alkali metal ions such as rubidium or cesium. Therefore, the alkali metal ion is preferably lithium, sodium and/or potassium, particularly preferably sodium. Without limitations the catalysts which are obtainable or obtained by a process according to the present invention are suitable for use in the preparation of olefins from methanol containing feed streams, preferably in the MTO reaction. In order to obtain the highest possible yield for propylene, the feed stream preferably contains as much methanol as possible, particularly preferred is a feed stream essentially consisting of methanol. It is also possible to recycle olefins with chain lengths of more than 3 carbon atoms into a process for the preparation of olefins from methanol comprising feed streams to increase the yield of propylene. It is believed that the recycled olefins with more than three carbon atoms are broken as a result of the proton transfers in the MTO reaction and the thus obtained fragments are reacted to give further propylene. Also the catalysts according to the present invention can be used in the preparation of olefins from methanol and particularly in the MTO reaction.

A further object of the present invention is therefore the use of a catalyst according to the present invention or a catalyst obtainable or obtained by a process according to the present invention in the preparation of olefins from a methanol comprising feed stream, preferably in the methanol-to-olefin process (MTO process). The most common olefins obtained in the MTO process are ethylene, propylene and butylene.

The catalysts which are obtainable or obtained by a process according to the present invention give the best results in the catalyzed preparation of propylene from methanol. Also the catalysts according to the present invention give the best results in the catalyzed preparation of propylene from methanol.

Thus, in one embodiment of the use according to the present invention propylene is prepared from a methanol comprising feed stream. The present invention is further described by the following items:

1. Process for the preparation of a zeolite-based catalyst for the preparation of propylene from methanol, comprising the step of setting the Br0nsted acid site density of a zeolite to a value of from ca. 0.08 mmol/g to ca. 0.4 mmol/g, wherein said Br0nsted acid site density is set in the direct synthesis of a zeolite, wherein the direct synthesis of the zeolite comprises the steps of a1 ) synthesizing a zeolite with a framework Si to Al ratio of from ca. 55: 1 to ca. 210: 1 and a2) converting said zeolite with a solution of a diluted mineral acid or an ammonium solution into the H-form of said zeolite, followed by drying and/or calcining the thus obtained zeolite, and/or by modifying a zeolite in a post-synthesis treatment selected from the group consisting of i) loading the zeolite with phosphate, and/or ii) partially exchanging ions of the zeolite, wherein the Br0nsted acid site density is determined by means of quantitative H MAS NMR spectroscopy, and wherein the step a2) and/or the post-synthesis treatment is/are repeated until the zeolite-based catalyst has a Br0nsted acid site density of from ca. 0.08 mmol/g to ca.

0.4 mmol/g.

Process according to item 1 , wherein the loading of a zeolite with phosphate comprises the following steps of: a3) treating a zeolite with a phosphate comprising solution, b3) if necessary, washing the zeolite obtained in step a3), and c3) drying and/or calcining the zeolite, wherein the sequence of steps a3) to c3) is performed at least once to provide a zeolite-based catalyst with a Br0nsted acid density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g.

Process according to item 2, wherein the loading of a zeolite with phosphate therefore further comprises the additional steps of

- determining the Br0nsted acid site density of the chosen zeolite prior to step a3),

- correlating the Br0nsted acid site density measured prior to step a3) and after the complete run of steps a3) to c3) with the amount of the phosphate used, and

- determining the necessary amount of the specific phosphate in order to provide the zeolite with a Br0nsted acid site density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g.

Process according to item 1 , wherein the partial ion exchange of the zeolite comprises the following steps of: a4) treating a zeolite present in its cationic form with a diluted mineral acid or an ammonium solution, or a4') treating a zeolite present in its H-form with an alkali metal ion containing solution, b4) if necessary, washing the zeolite obtained in step a4) or a4'), and c4) finally drying and/or calcining the zeolite, wherein the sequence of steps a4) or a4') to c4) is performed at least once to provide a zeolite- based catalyst with a Br0nsted acid density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g.

5. Process according to item 4, wherein the partial ion exchange of the zeolite further comprises the additional steps of

- determining the Br0nsted acid site density of the chosen zeolite prior to step a4) or a4'),

- correlating the Br0nsted acid site density measured prior to step a4) or a4') and after the complete run of steps a4) or a4') to c4) with the amount of the mineral acid or ammonium compound (in case of step a4)) or of an alkali metal ion (in case of step a4')) used, and

- determining the necessary amount of the specific mineral acid or ammonium compound (in case of step a4)) or of an alkali metal ion (in case of step a4')) in order to provide the zeolite with a Br0nsted acid site density of from ca. 0.08 mmol/g to ca. 0.4 mmol/g.

6. Process according to any one of items 1 to 5, wherein the direct synthesis of a zeolite is combined with loading the zeolite with phosphate, and/or partially exchanging ions of the zeolite.

7. Process according to any one of items 1 to 6, wherein the zeolite used has 10-ring channels. 8. Process according to item 7, wherein the zeolite has a structure of the ZSM-5 or ZSM-1 1 type. 9. Process according to any one of items 1 to 8, wherein the zeolite, which is subjected to at least one the post-synthesis treatments, has a framework Si to Al ratio of less than ca. 55:1 . 10. Catalyst on the basis of a zeolite having 10-ring channels, characterized in that said catalyst has a Bransted acid site density of from ca. 0.8 mmol/g to ca. 0.4 mmol/g, which is determined by means of quantitative H MAS NMR spectroscopy. 1 1. Catalyst according to item 10, wherein the zeolite has a structure of the ZSM-5, ZSM-1 1 or ZSM-22 type.

12. Catalyst according to item 10 or 1 1 , wherein the zeolite-based catalyst contains from ca. 0.5 to ca. 10 percent by weight of a phosphate, based on the total weight of the catalyst.

13. Catalyst according to any one of items 10 to 12, wherein from ca. 5 to ca. 20% of the hydrogen atoms in the framework of the zeolite are replaced with an alkali metal. 14. Use of a catalyst according to any one of items 10 to 13 or of a catalyst obtainable or obtained by a process according to any one of claims 1 to 9 in the preparation of olefins from a methanol comprising feed stream.

15. Use according to item 14, wherein propylene is prepared from a methanol comprising feed stream.

Figures: Figure 1 illustrates the change of the acidic bridge-hydroxyl-groups of a zeolite in H MAS NMR spectrum upon loading a zeolite with ammonia.

Examples:

I. Comparative examples

1. Directly synthesized zeolite

The directly synthesized zeolite is an H-ZSM-5 zeolite in powder form, commercially available as HZX from Tricat Inc., a member of the Eurecat Group, Houston Texas. According to the information from Tricat Inc. said zeolite has a framework S1O2 to AI2O3 ratio of 45 to 50, which approximately corresponds to a framework Si to Al ratio of ca. 22.2 to 25. Analysis of the Si to Al ratio of the obtained HZX powder by ICP-OES showed that this zeolite has a Si to Al ratio of ca. 20: 1.

2. Directly synthesized zeolite

An H-ZSM-5 zeolite with a framework Si to Al ratio of 1000: 1 was prepared according to the procedure of Ghamami, K., Sand, L.B. (Zeolites 1983, Vol. 3, 2 ^nd issue, p. 155). The reactant materials for the synthesis were: (1 ) a microfine precipitated silica (Philadelphia Quarts QUSO F20) S1O2 * 0.4 H2O, (2) a colloidal silica sol (DuPont LUDOX ^® AS-40) S1O2 * 5 H2O, (3) dried aluminum hydroxide gel (Reheis F-2000), AI2O3 * 5 H2O, a 29% ammonium hydroxide solution, a 25% solution of tetrapropylammonium hydroxide, tetrapropylammonium bromide and water. As ammonium hydroxide solution is volatile and as tetrapropylammonium hydroxide absorbs carbon dioxide from the air, measures were taken to minimize the errors which arise from these sources. The autoclave was loaded to 75% capacity. The Bransted acid site density of the thus obtained zeolite was lower than 0.03 mmol/g, determined by quantitative H MAS NMR spectroscopy. 3. Template free synthesized zeolite

An H-ZSM-5 zeolite with a framework Si to Al ratio of 20: 1 was prepared according to the following procedure: A solution of colloidal silica in water was added under stirring to a solution of sodium aluminate in water. After homogenization a solution of sodium hydroxide was added to this solution and the resulting mixture was stirred for 30 minutes. Dynamic crystallization was performed in a steel autoclave equipped with a teflon inlet at 160°C for 120 hours. Then the reaction mixture was quenched, separated by means of a Nutsche filter and washed with sufficient amount of water, followed by calcination. Finally, the Bransted acid site density was determined by quantitative H MAS NMR spectroscopy to be 0.79 mmol/g.

4. Dealuminated zeolite

An H-ZSM-5 zeolite with a framework Si to Al ratio of 20: 1 was prepared, placed in 75 cm ³ of concentrated nitric acid per gram of zeolite and stirred at 80°C for 4 hours. Subsequently, the thus modified zeolite was washed with water until nitrate was no longer detected, and then calcined. Finally, the Br0nsted acid site density of the thus obtained zeolite was determined by quantitative H MAS NMR spectroscopy to be 0.73 mmol/g.

5. Phosphate modified zeolite

Phosphate modification of zeolites was generally done using the incipient wetness method. Impregnation with incipient wetness method is done according to this procedure: First, the volume of the impregnation solution is determined by spraying water over a zeolite, which was dried before at 80°C for 30 min. The test impregnation solution is drawn into the interior of the zeolite by capillary forces. By test impregnation with water the volume of the impregnation solution was determined to be 0.22 cm ³ per gram of zeolite. Because of similar pore volumes for the zeolites to be impregnated the required volume for the impregnation is the same for each zeolite.

An H-ZSM-5 zeolite with a framework Si to Al ratio of 20: 1 was prepared and loaded with 2.7 % of a phosphate by impregnating the zeolite one time loading with a solution of 5% ammonium dihydrogen phosphate in water using the incipient wetness method described above, followed by washing and drying. The thus obtained modified zeolite was calcined and its Bransted acid site density was 0.5 mmol/g, determined by quantitative H MAS NMR spectroscopy. 6. Alkali ion exchanged zeolites

A directly synthesized ZSM-5 zeolite was partly ion exchanged with lithium ions. In the first case (comparative example 6a) 94% of the protons of said zeolite were replaced with lithium cations by treatment of the ZSM-5 zeolite with a lithium chloride solution having a concentration of 7.1 M using the incipient wetness method (see above). Subsequently, the treated zeolite is dried and calcined in air. An excess of lithium chloride is removed by suspending the treated zeolites for 10 minutes in 2 cm ³ per gram of zeolite, followed by filtration and washing with 500 cm ³ of water. The thus obtained modified zeolite has a Br0nsted acid site density of 0.05 mmol/g, determined by means of quantitative H MAS NMR spectroscopy. In the second case (comparative example 6b) 46% of the protons of the zeolite were exchanged by treatment of the ZSM-5 zeolite with a lithium chloride solution having a concentration of 4.8 M using the incipient wetness method to give a modified zeolite with a Br0nsted acid site density of 0.42 mmol/g, determined by means of quantitative H MAS NMR spectroscopy.

7. Catalyzed MTP reactions

The zeolite-based catalysts of the comparative example 1 to 6 were tested in the MTP reaction after a time on stream of three hours, at a WSHV = 4 h ^~ and a temperature of 450°C. The yields for the specific olefins obtained with the different catalysts are summarized in Table 2.

Table 1 : Summary of the results of the MTP reaction, * = dimethylether is the only product.

II. Examples according to the invention

1. Directly synthesized zeolite An H-ZSM-5 zeolite with a framework Si to Al ratio of 130:1 was prepared according to the following general procedure: tetraorthosilicate, tetrapropylammonium hydroxide and water were mixed and stirred for 14 hours. To this mixture was added a solution of aluminum isopropoxide, tetrapropylammonium hydroxide, sodium hydroxide and water. The resulting mixture was stirred for 30 minutes. Crystallization was performed in an autoclave, made of steel and equipped with a teflon inlet, at 160°C for 24 hours. Thereafter, the reaction mixture was quenched, the zeolite was separated from the reaction mixture by filtration and washed with a sufficient amount of water, followed by calcination, ion exchange with ammonium nitrate and drying and/or calcination. The Br0nsted acid site density of this zeolite was 0.13 mmol/g, determined by quantitative H NMR MAS spectrometry.

2. Phosphate modified zeolite

An H-ZSM-5 zeolite with a framework Si to Al ratio of 130: 1 , specifically the catalyst of example 1 , was loaded with 0.5% of a phosphate (the phosphate was applied by loading with a solution of 5% of ammonium dihydrogen phosphate, followed by washing with water) using the incipient wetness method described above. The Br0nsted acid site density of this zeolite was 0.10 mmol/g, determined by quantitative H MAS NMR spectroscopy.

3. Phosphate modified and partially ion exchanged zeolite An H-ZSM-5 zeolite with a framework Si to Al ratio of 20:1 , specifically the HZX power of comparative example 2, was loaded with 3.2% of a phosphate, based on the total weight of the catalyst (the phosphate was applied by loading with a solution of 5% of ammonium dihydrogen phosphate, followed by washing with water) using the incipient wetness method described above. Afterwards, 19% of the protons of the zeolite were exchanged with sodium ions (performed by washing with a solution of 0.05 M sodium chloride) using the incipient wetness method described above. The Bransted acid site density of this zeolite was 0.15 mmol/g, determined by quantitative 1 H MAS NMR spectroscopy.

4. Phosphate modified and partially, ion exchanged zeolite

An H-ZSM-5 zeolite with a framework Si to Al ratio of 20:1 , specifically the HZX power of comparative example 2, was loaded with 5% of a phosphate, based on the total weight of the catalyst (the phosphate was applied by loading with a solution of 5% of ammonium dihydrogen phosphate, followed by washing with water) using the incipient wetness method as in comparative example 5. Subsequently, 8% of the protons of the zeolite were exchanged with sodium ions (performed by washing with a solution of 0.05 M sodium chloride) using the incipient wetness method as in comparative examples 6a and 6b. The Bransted acid site density of this zeolite was 0.21 mmol/g, determined by quantitative H MAS NMR spectroscopy. 5. Catalyzed MTP reaction

The zeolite-based catalysts of the examples 1 to 4 according the invention were tested in a catalyzed MTP reaction. The catalytic performance after a time on stream (TOS) of 3 hours, at a WHSV of 4 h ^~ and a temperature of 450°C. The yields for the specific olefins obtained with the different catalysts are summarized in Table 2.

Table 2: Summary of the results of the MTP reaction

In comparison to the catalytic performance of the zeolite-based catalysts in the comparative examples, the zeolite-based catalysts according to the present invention lead to significantly improved yields for propylene and in addition, to a significantly lower amount of by-products.