Dimethyl ether synthesis via methanol dehydration: Effect of zeolite structure

Dimethyl ether synthesis via methanol dehydration: Effect of zeolite structure

Accepted Manuscript Title: Dimethyl ether synthesis via methanol dehydration: Effect of zeolite structure Author: Enrico Catizzone Alfredo Aloise Mass...

498KB Sizes 1 Downloads 18 Views

Accepted Manuscript Title: Dimethyl ether synthesis via methanol dehydration: Effect of zeolite structure Author: Enrico Catizzone Alfredo Aloise Massimo Migliori Girolamo Giordano PII: DOI: Reference:

S0926-860X(15)30033-8 http://dx.doi.org/doi:10.1016/j.apcata.2015.06.017 APCATA 15429

To appear in:

Applied Catalysis A: General

Received date: Revised date: Accepted date:

17-3-2015 22-5-2015 15-6-2015

Please cite this article as: Enrico Catizzone, Alfredo Aloise, Massimo Migliori, Girolamo Giordano, Dimethyl ether synthesis via methanol dehydration: Effect of zeolite structure, Applied Catalysis A, General http://dx.doi.org/10.1016/j.apcata.2015.06.017 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

DIMETHYL ETHER SYNTHESIS VIA METHANOL DEHYDRATION: EFFECT OF ZEOLITE STRUCTURE

Enrico Catizzone, Alfredo Aloise, Massimo Migliori*, Girolamo Giordano

Department of Environmental and Chemical Engineering – University of Calabria – Via P. Bucci Cubo 42a – I-87036

Rende (CS), Italy

ABSTRACT In this paper, the effects of either pore size or topology of zeolites were studied in the methanol dehydration to dimethyl ether reaction by comparing catalytic performances of BEA, MFI and FER structures in the temperature range 180°C-300°C. The aim of this study was to investigate how the zeolite catalyst characteristics affect the process performances in terms of methanol conversion and DME selectivity. It was found that the largest-pore 3-D framework zeolite (BEA) was very effective in converting methanol but the channel size and topological connection spaces allowed the fast formation of coke precursors that rapidly

decreases the catalyst performances at higher temperatures. Even if lower in intensity, the same phenomenon was observed for MFI structure (medium pores 3-D framework) confirming stable performances in the temperature range used in the real processes. On the contrary, the 2-D small pores FER zeolite showed a very good selectivity at high temperature also exhibiting a promising conversion rate for an industrial application. Continuous catalytic tests at 300°C, followed by coke deposition analysis, confirmed that BEA rapidly deactivates and exhibited the higher coke formation rate, on the contrary, FER structure exhibited a great stability (conversion and selectivity) as well as a reduced coke formation tendency.

Keywords: Methanol-to-DME; Zeolite structure effect; Coke formation; Catalyst deactivation. Highlights •

The effect of structure on DME production via methanol dehydration was investigated;



Large pores 3-D zeolite (BEA) was very active but poor in stability;



Small pores 2-D zeolite (FER) was less active event though mode stable;



Deactivation is strongly affected by channel size and topology



Coke formation can be modelled as a 1st order system

1. Introduction

During the last decade dimethyl ether (DME) has received an increased attention as sustainable alternative fuel for Diesel engines due to its high cetane number (>55) and to the important reduction of NOx, SOx e PM emissions in exhaust gases [1]. Colourless, non-toxic, non-corrosive, non-carcinogenic and environmentally friendly compound, DME is the simplest ether, with a normal boiling point of -25°C, that can be liquefied above 0.5 MPa at room temperature. Chemical and physical properties of this clean fuel are similar to LPG and this may represents a strategic advantage as the storage and transport facilities of LPG can be easily converted to distribute DME [2]. DME can be industrially synthesized in two way: following the indirect or direct synthesis. Indirect synthesis is a two-steps process: the traditional methanol synthesis from syngas (H2/CO mixture with molar ratio 0.5-2) over Cu/ZnO/Al2O3 (CZA) redox catalyst, in the temperature range 240-280°C and pressure between 3 and 7 MPa [2, 3], followed by methanol dehydration reaction to dimethyl ether over acid catalyst. On the other hand, DME can be also produced by direct synthesis, a one-step process where the two reactions, methanol synthesis (via CO hydrogenation) and the dehydration to DME, take place in the same reactor under process conditions close to those of methanol synthesis [4]. This is a promising process, mainly because of the thermodynamic advantages in case of simultaneous reactions (methanol to DME conversion promotes also the alcohol formation) but also for the lower costs (with respect to the double step synthesis). In addition a major interest to the direct synthesis via carbon dioxide hydrogenation results as a promising option for a fuel effectively reducing the overall CO2 emission [5]. The process (either using CO or CO2) requires two catalytic functions: both the CZA redox catalyst for the methanol synthesis and the acid function for dehydration. Different solutions for the catalyst

optimisation have been proposed, such as bifunctional catalysts [6] or physical mixture of redox and acid catalyst [7], with the second solution being more promising in terms of overall conversion efficiency [8]. Whatever it is the catalyst coupling method, it has been also demonstrated that the properties of the acid catalyst determines the selectivity and durability of bi-functional catalyst as well as the process efficiency overall, since it is controlled by the dehydration step [3]. Despite the importance of the acid catalyst, studies were primarily carried out over γ-Al2O3, exhibiting a high selectivity in the temperature range 200-300°C [9-11]. Moreover, the presence of water as reaction product, significantly deactivates the catalyst, potentially as effect of the active sites blocking by water molecules. In this concern, zeolites (mainly H-MFI) were also investigated as alternative to γ-Al2O3 revealing a better stability to water presence [9] and good conversion and selectivity. In addition these materials offer the possibility to modulate the catalyst properties (acidity, specific surface area ,crystal size and shape selectivity) as a way to improve the process performances [12-14]. Nevertheless, in the temperature range of direct synthesis (260-300°C), the acid catalyst may start to promote other reactions, converting the methanol to a range of different hydrocarbons (Methanol to Hydrocarbon - MTH). In fact, it is well known [15] that the methanol dehydration to DME (Methanol to DME - MTD) is the first step for any MTH process and different industrial processes have been studied leading to olefins (Methanol to Olefins - MTO) or to gasoline (Methanol to gasoline - MTG) [16,17]. In this processes, DME is the first intermediate of methanol conversion even though the direct reaction of the alcohol and the ether to produce hydrocarbons is not the favoured way, except in the induction period. On the contrary, the so called “hydrocarbon pool”, the more recognised

reaction mechanism is based on a catalytic scaffold of larger organic molecules (generally polymethylbenzenes) adsorbed in the zeolite, attracting methanol/dimethyl ether to form alkenes and water in a closed cycle, avoiding which high-energy intermediates [18]. The role of polymethylbenzenes as reaction pool have been widely investigated and the product distribution strongly depends on the catalyst channel size (shape selectivity) and mostly ethylene and propylene are produced when the small pore structures are used, such as SAPO-34 (chabasite framework structure, the most common catalyst for MTO) as only small molecules can easily escape through the narrow pores [19]. Many studies have been proposed to investigate the effect of the zeolite structures in MTO processes and it was found that, apart the zeolite acidity, the channel configuration and dimensions play a relevant role in the product distribution. It was demonstrated that small pores structure, such as CHA, but also larger pores 1-D structures, such MOR, both showed high selectivity for lower olefins, whilst larger pores and interconnected structures, like MFI, FAU and BEA, promoted the formation of bigger alkylaromatics compounds [20]. Despite the fact that typical MTD operating temperature is lower than the MTO process (160-280°C compared to 400°C), selectivity loss are observed also when producing DME and any further reaction has to be limited reducing the coke formation [21]. In fact, coke formation is one of the most relevant aspects in any MTH process, since the potential presence of alkyl-aromatics may easily lead to the formation of solid deposit even in the temperature range of MTD [22]. An interesting debate is still on-going about the preferential location of the coke deposition: in some cases the solid was detected also within the zeolite structures [23], but a great number of researcher agreed that both size and connections of the channels limit the internal solid deposition and coke is most favourably located

outside the pores on the crystal surface [18], according to the hypothesis that the coke precursors can diffuse from the micropores to the external surfaces especially in case of large external surface area and short diffusion path lengths (crystal size and morphology) [24]. In this concern, the same issues may arise in the MTD process, especially in the direct synthesis conditions (high temperature and moderate pressure) and it is interesting to investigate how the zeolite structure can contribute in increase the process selectivity, limiting the coke formation and enhancing the catalyst stability. Those are all are relevant issues for any further application of zeolite structures as acid catalyst in the MTD process. Aiming to contribute in these aspects, in this paper three different zeolite structures FER (8-10 rings, 2-D small pore channels with two channel sizes: 4.2Å x 5.4Å and 3.5Å x 4.8Å), MFI (10 ring, 3-D medium pores structure with two channel sizes: 5,3Å x 5.6Å and 5.5Å x 5.1Å) and BEA (12 rings, 3-D large pore structure, with two channel sizes: 7.6Å x 6.4Å and 5.5Å x 5.5Å) with the same Si/Al ratio were synthesised, characterised and tested in the MTD reaction. The effect of temperature over conversion and selectivity was investigated and Time on Stream (TOS) tests were also performed to evaluate the catalyst stability and coke formation.

2. Experimental 2.1 Samples preparation

As already mentioned, three zeolite structures with different framework characteristics (MFI, BEA, FER) were synthesised, aiming at obtain the same Si/Al ratio in the final catalyst. The MFI was prepared following the synthesis procedure published elsewhere [25] starting from the following synthesis gel (molar composition): 0.10 Na2O – 0.08 TPABr – 1 SiO2 – 0.02 Al2O3 – 20 H2O After the synthesis the samples was filtered with distilled water, dried at 105°C and calcined in air at 550°C to eliminate SDA. The acidic form was obtained after two cycles of ionic exchange with NH4Cl 1 M (1 h each) followed by a calcination at 550°C. The BEA was obtained from the following synthesis gel (molar composition): 0.10 Na2O – 0.2 TEAOH – 1 SiO2 – 0.02 Al2O3 – 10 H2O After 2 hours stirring at room temperature, the gel was left in autoclave at 150 °C for 120 hours. After cooling, the separated solid phase was washed with distilled water and dried at 105°C for 8 hours. The sample was calcined at 550 °C and the same activation procedure (ionic exchange as MFI) was adopted. The FER was prepared from the following gel (molar composition) [26]: 1.5 NH4F – 0.96 HF – 1 TMPP – 1 SiO2 – 0.04 Al(OH)3 – 9.68 H2O The synthesis gel stirred for 90 minutes at room temperature and left in autoclave at 175 °C for 120 hours. After cooling the separated solid phase was washed with distilled water and the acidic form of the catalyst was obtained by a double step calcination: under nitrogen flow and then under air flow (550 °C for 8h each step).

2.2 Characterisation of catalysts In order to verify the obtained catalysts structure, all the samples were characterized by X-Ray powder diffraction spectroscopy (APD 2000 Pro) and the morphology of the crystalline phase was observed on a scanning electron microscope (FEI model Inspect). The Si/Al ratio in the solid was measured by ICP-MS (Perkin-Elmer DRC-e). The specific surface area and the micropores volume of the catalyst were obtained by performing a BET and t-plot analysis of porosimetry data (ASAP 2020 Micromeritics) under nitrogen adsorption at 77 K, after a pre-treatment in vacuum at 200 °C for 12 h. The surface acidity of the samples was determined by NH3-TPD measurements by loading 100 mg of sample in a linear quartz micro-reactor (i.d., 4 mm; l., 200 mm), with a helium carrier flow of 25 STP mL/min. The temperature was increased within the range 100–800°C imposing a heating rate of 10 °C/min. The ammonia desorption was monitored by TCD, calibrated by the peak area of known NH3 pulses. As conditioning step, the sample was pretreated at 300 °C with hydrogen flowing at 100 mL/min for 1 h and then cooled down to 150°C and saturated with a 5 vol.% NH3/He stream flowing at 25 mL/min for 1.5 h. The samples were eventually purged in He atmosphere for about 1 h, until a constant TCD level was obtained [27]. 2.3 Catalytic tests Catalytic test were performed in an experimental apparatus described elsewhere [28]. In order to evaluate the effect of structure on catalytic activity both DME indirect and direct synthesis conditions, methanol conversion tests were carried out in the temperature range 180- 300 °C, loading 150 mg of catalysts (pellets size: 300-500 µm) for each run. A mixture of methanol

(0.056 mol/mol) and nitrogen (60 Nml/min), acting as carrier, was used as a feed, resulting in a Weight Hourly Space Velocity (WHSV) of 2 gMetOH/(gcat h). Before of any test, the catalyst sample was dried under nitrogen flow at 200°C for 3 h and product stream composition was analysed by gas chromatography (Agilent 7890A), equipped with a capillary column (J&W 125-1032) and a FID detector. Experimental results are presented in terms of methanol molar conversion and DME yield defined as it follows:

Methanol conversion =

DME Yield =

Converted methanol (gmol ⋅ min -1 ) Feed methanol (gmol ⋅ min -1 )

2 ⋅ DME outstream (gmol⋅ min -1 ) Converted methanol (gmol ⋅ min -1 )

(1)

(2)

Aiming at investigating the influence of zeolite structure on durability and stability against coke formation, TOS tests were performed at 300°C. Reaction time was modulated according to the catalyst stability: the higher stability against deactivation the longer was the TOS time. In addition TOS at different times were performed in order to monitor the coke deposition over time. The amount of coke deposited on catalyst was measured by thermo-gravimetric analysis (TGA) in air flow (DTG-60 Shimadzu), following the procedure reported by Kim and coworkers [24], by heating the sample from room temperature to 850°C (ramp 5°C/min). The total coke content was estimated as the sample weight loss between 250°C and 850°C, by slightly modifying the temperature range suggested in the literature [29]. In order to verify the absence of any constitution-water loss, also fresh zeolite

was analysed in TGA experiments up to 850°C and no weight loss was observed above 250°C. Any experimental measure was repeated over three independent samples and reported points are average ± standard deviation.

3. Results and discussion 3.1 Textural and chemical properties of catalysts Figure 1 reports the XRD pattern of catalyst after the activation in H-form, revealing the low background as good indicator of high sample crystallinity. Nitrogen adsorption isotherms (as presented in the Supplementary Information Section – Fig. S1) revealed the typical pattern of type I isotherm, confirming the microporous characteristics of the synthesised samples [30]. In addition the adsorption isotherm of MFI sample shows a shoulder at P/P° value around 0.2 and this is a characteristic of uniform ultra-microporous solids with few groups of energetically uniform sites. Textural and chemical characteristics of fresh catalyst pellets are listed in Table 1, revealing the typical sequence of catalysts BET specific surface area, BEA>MFI>FER, according to literature data [31]. In addition it is shown the effect of pore size and topology as the increase in the channels size and interconnections determines an increase of the micropores volume. It is important to notice that the Si/Al “bulk” ratio (i.e. measured on the crystals) is nearly the same for any sample. Catalysts crystals size was measured by SEM and the results reported in Table 1.. Ammonia desorption results are reported in Table 2 and also TPD experiments confirmed a well-accepted trend of increasing acidity (BEA>MFI>FER) and relative values are in the range usually observed for any structure [32, 33]. In

addition, the number of weak and strong sites has been determined by multiplying the overall acid sites number by the fraction active above or below 300°C, respectively. Data show that 3-D structures (BEA and MFI) exhibited a comparable number of acid sites and nearly the same number of strong ones, whilst FER showed a reduced number of acid sites predominantly strong. Curves deconvolution is reported in the Supplementary Information Section Figure S2. 3.2 Catalytic activity The catalytic activity of BEA zeolite in the temperature range from 180°C to 300 °C is reported in Figure 2, together with the calculated theoretical thermodynamic equilibrium value of methanol conversion. The conversion increases as a function of temperature and approaches the equilibrium value around 260°C, up to this temperature also the selectivity toward DME remains above 0.94. For any further increase in temperature a drop of methanol conversion is observed and a significant loss in DME selectivity dropping below 0.5. This effect was attributed to the deposition of heavy compounds on the catalyst surface. This phenomenon is most probably promoted by the large BEA pores that favours the successive reaction of methanol and DME, rapidly leading to olefins and heavier compounds (cyclic and aromatics) as in the MTG process. Figure 3 shows the catalytic test for MFI exhibiting a good performance up to 280°C (when conversion approaches the theoretical thermodynamic equilibrium value) also in terms of selectivity (above 0.95). Also for MFI, at higher temperatures a conversion drop and selectivity loss were observed but, differently from BEA, this may be attributed to temperature-favoured MTH reactions, promoting olefins formation, more than to gasoline-species, as revealed by GC data analysis. By comparing the results with the BEA ones, the

same behaviour was observed even though the deactivation of the MFI catalyst was much slower, according to two co-current factors: (i) the highest acidity of BEA, with respect to MFI, that promotes the MTH reaction, (ii) the different pores size of the two 3-D zeolites (large pores for BEA and medium pores for MFI). When the small pore and 2-D structure of FER is tested (Figure 4), a different behaviour is observed: up to 300°C the catalytic activity remains well below the thermodynamic threshold and no drop in conversion or selectivity was found in the investigated temperature range. In order to compare the effect of structure on the performance in MTD reaction, conversion data with selectivity higher than 0.95 are summarised in Figure 5. It clearly appears that the larger pore structure promotes the methanol conversion, since no significant difference was found between BEA and MFI zeolites and it can be concluded that the difference in pore size of the two 3-D structures does not represent a relevant factor for the reaction, in the investigated conditions. On the contrary the behaviour of the FER was very stable, keeping the highest selectivity up to 300°C, even though the conversion values were lower that the others. This indicates the FER as a reliable options for direct synthesis processes, taking place at temperature above 250°C [3], where the loss in selectivity may represent a key issue. This result is in agreement with some recent interesting findings about the ITQ-6 application (FER is the relative delaminated zeolite precursor) as acid catalyst for direct MTD process [8]. 3.3 Duration test

In order to check the catalyst stability and according to results of catalytic activity, TOS test were performed at the maximum investigated temperature (i.e. 300°C). Data of methanol conversion over time are reported in Figure 6 for the three investigated zeolites. The TOS duration was established for any catalyst according to the system performances: FER revealed a good stability for a long time and the maximum investigation time was fixed at 44h. The conversion measured over time was nearly constant and in agreement with initial value. As for MFI structure, a partial deactivation was observed over time as the methanol conversion cross down the FER one after 15h; in these conditions the formation of olefins was observed and, as a consequence, the maximum TOS was fixed at 24h. On the contrary, BEA exhibited a very fast deactivation (initial conversion about 0.6 according to Figure 2) followed by a partial recovery over time. Nevertheless, the catalyst performance in MTD process was very poor as a significant amount of heavy compounds was formed. Therefore the TOS was limited at 6h. If the selectivity data of Figure 7 are considered, FER structure confirmed a very good stability over 44h TOS, with selectivity always above 0.95. As for the other zeolites, as consequence of the heavy compounds formation, a partial recovery selectivity DME was observed, being the effect more evident for the BEA structure, nearly doubling the selectivity in the investigated TOS. Since coke formation is a complex phenomenon, affected by either acidity or catalyst channels structure, in order to gain information about catalyst deactivation, also data of NH3-TPD of Table 2 have to be taken into account. In fact, according to the well-known difference in structure [32], FER exhibited a lower amount of strong acid sites when compared with both BEA and MFI and this lead to a lower conversion but also an increased stability of the FER catalyst over time. On the contrary, when the

BEA and MFI are considered, a comparable amount of acidic sites was found, therefore the observed differences in terms of conversion and duration can be mainly attributed to the different channels size and configuration of the two zeolites. 3.4 Coke analysis The coke deposition kinetic was also investigated by varying the reaction time up to the maximum TOS as defined above for any investigated catalyst. The coke amount was measured for any exhaust catalyst sample and data are showed in Figure 8. When considering the 3D structures, also in presence of a slightly higher acidity, it clearly appears that the pores of BEA, largest that MFI, promoted a faster and higher coke content also for lower TOS. This indicates that, in presence of an equal concentration of strong acid sites (see Table 2 data) that are responsible for coke formation,[24] the structure topology (3-D for both) is less important than channel size as the larger pore size promotes the coke formation. On the contrary, also considering the lowest acidity, the small pore 2-D structure of FER determines a low and slow deposition of the solid over the catalyst surface, producing after 44 h of TOS, the same amount of coke produced by MFI after 1 h and half of that formed over BEA after 1 h of reaction. In order to gain quantitative information about the coke formation kinetic, experimental data were fitted by using a first order model. This approach was proposed few years ago [34] and successfully applied in similar cases, also including more complicated functionalities such as Temperature and WHSV [35]. In our case the simplest first order kinetic form is used

−t C = C ∞ ⋅ 1 − e τ   

(3)

where C is the specific coke amount [gcoke·100gcat-1], C∞ the plateau value and τ the characteristic time [h]. From this formulation it is easy to estimate the initial coke deposition rate r0 [gcoke·(100gcat·h)-1] as:

r0 =

dC C = ∞ dt t=0 τ

(4)

Experimental data were fitted with a non-linear regression by using a commercial software (CurveExpert Professional 2.0) and parameters are presented in Table 3. It clearly appears that the larger 3-D structure of BEA promotes the coke formation as the estimated coke plateau value C∞ is nearly double than the MFI and much higher than FER. As for the characteristic time τ, it is noteworthy that the difference in channel size of 3-D structures of BEA and MFI did not significantly affect this parameter whilst the small pore 2-D structure of FER results in a much stable catalyst, showing higher τ values of more than 1 order of magnitude. As for the initial coke formation rate parameter r0, values for BEA showed a formation rate much higher than the MFI, confirming the BEA as the structure more sensitive to coke formation. In addition the stable behaviour of FER over time is confirmed by a characteristic time about 2 order of magnitude lower than BEA. All the parameters comparison between different structure confirmed the qualitative behaviour already observed in TOS.

4. Conclusion In this paper the effect of zeolite structure and topology over the performance in the methanol-to-DME reaction was investigated. Three different zeolites structure with the same Si/Al ratio but different in channel topology (3D, 2D) and size (small, medium and large), were compared in terms of catalytic activity and duration over TOS. Catalytic tests revealed that 3-D large pore structure of BEA results on a very reactive catalyst but a significant drop in selectivity was observed at high temperature. When considering the reduced 3D channels size of MFI, even in presence of a comparable acidity, a better stability was observed with a selectivity close to 1 up to 260°C. On the contrary, the narrower channels of the 2-D structure of FER, coupled with a reduced acidity, decreases the catalytic activity but the selectivity was kept near to 1 also when temperature approached 300°C. TOS tests confirmed the good stability over time of FER, either in terms of conversion (very small drop after 44h) or selectivity, also showing that 3-D structures (BEA and MFI) exhibited a less performing behaviour in terms of conversion and selectivity, ending with a much lower stability time. Also the coke formation during TOS was investigated and simplified version of a 1st order kinetic model revealed that 3-D structure of BEA determines the faster initial deposition rate and higher plateau coke content, whilst the small 2-D structure of FER ended in a very stable response with high characteristic coke formation time (2 order of magnitude lower than BEA) and a reduced predicted maximum coke content.

The combination of all these evidences, also considering that temperature represent a key parameter for the DME direct synthesis, suggests that an acid catalyst effective and stable above 240°C, such as FER, can represent a good option for a one step process, to be investigated further in combination with CZA catalyst.

ACKNOWLEDGMENTS The authors gratefully acknowledge Prof. Antonio TAGARELLI, University of Calabria, dr. Francesco FRUSTERI and dr. Giuseppe BONURA, CNR-ITAE “Nicola Giordano” - Messina, for analytical support and fruitful discussions. References [1] C. Arcoumanis, C. Bae, R. Crookes, E. Kinoshita, Fuel 87 (2008) 1014-1030. [2] T. Ogawa, N. Inoue, T. Shikada, Y. Ohno, J. Nat. Gas Chem. 12 (2003) 219-227. [3] J. Kim, M. Park, S. Kim, O. Joo, K. Jung, Appl. Catal. A 264 (2004) 37-41. [4] K. Yoo, J. Kim, M. Park, S. Kim, O. Joo, K. Jung, Appl. Catal. A 330 (2007) 57-62. [5] G. Bonura, M. Cordaro, L. Spadaro, C. Cannilla, F. Arena, F. Frusteri, Appl. Catal. B (2013) 16-24. [6] K. Sun, W. Lu, F. Qiu, S. Liu, X. Xu, Appl. Catal. A 252 (2003) 243-249. [7] T. Takeguchi, K. Yanagisawa, T. Inui, M. Inoue, Appl. Catal. A 192 (2000) 201-209.

[8] A. García-Trenco, A. Martínez, Catal. Today 227 (2014) 144-153. [9] M. Xu, J. Lunsford, D. Goodman, A. Bhattacharyya, Appl. Catal. A 149 (1997) 289-301. [10]

F. Yaripour, F. Baghaei, I. Schmidt, J. Pregaard, Catal. Commun. 6 (2005) 147-152.

[11]

D. Liu, C. Yao, J. Zhang, D. Fang, D. Chen, Fuel 90 (2011) 1738-1742.

[12]

M. Ali, B. Brisdon, W. Thomas, Appl. Catal. A 252 (2003) 149–162.

[13]

S. Hassanpour, F. Yaripour, M. Taghizadeh, Fuel Process. Technol 91 (2010) 1212-1221.

[14]

A. Rownaghi, F. Rezaei, M. Stante, J. Hedlund, Appl. Catal. B 119 (2012) 56-61.

[15]

M. Stocker, Micropor Mesopor. Mater. 29 (1999) 3-48.

[16]

J. Li, Y. Wei, G. Liu, Y. Qi, P. Tian, B. Li, Y. He, Z. Liu, Catal. Today 171 (2011) 221-228.

[17]

U. Olsbye, S. Svelle, M. Bjørgen, P. Beato, T.V. Janssens, F. Joensen, S. Bordiga, K.P. Lillerud, Angew. Chem. Int. Edit.

51 (2012) 5810-5831. [18]

M. Bjørgen, S. Svelle, F. Joensen, J. Nerlov, S. Kolboe, F. Bonino, L. Palumbo, S. Bordiga, U. Olsbye, J. Catal. 249

(2007) 195–207 . [19]

B. Arstad S. Kolboe, J. Am. Chem. Soc. 123 (2001) 8137-8138.

[20]

J.W. Park, G. Seo, Appl. Catal. A 356 (2009) 180-188.

[21]

H. Li, S. He, K. Ma, Q. Wu, Q. Jiao, K. Sun, Appl. Catal. A 450 (2013) 152-159.

[22]

H. Schulz, Catal. Today 154 (2010) 183-194.

[23]

P. Benito, A. Gayubo, A. Aguayo, M. Olazar, J. Bilbao, Ind. Eng. Chem. Res. 35 (1996) 3991-3998.

[24]

J. Kim, M. Choi, R. Ryoo, J. Catal. 269 (2010) 219-228.

[25]

M. Migliori, A. Aloise, G. Giordano, Catal. Today 227 (2014) 138-143.

[26]

A. Chica, U. Diaz, V. Fornés, A. Corma, Catal. Today 147 (2009) 179-185.

[27]

G. Bonura, M. Cordaro, C. Cannilla, A. Mezzapica, L. Spadaro, F. Arena, F. Frusteri, Catal. Today 228 (2014) 51-57

[28]

M. Migliori, A. Aloise, E. Catizzone, G. Giordano, Ind. Eng. Chem. Res. 53 (2014) 14885-14891

[29]

S. Jong., A. Pradhan, J. Wu, T. Tsai, S. Liu, J. Catal. 174 (1998) 210-218.

[30]

G. Leofanti, M. Padovan, G. Tozzola, B. Venturelli, Catal. Today 41 (1998) 207-219.

[31]

J. Groen, L. Peffer, J. Moulijn, J.Pérez-Ramìrez, Micropor. Mesopor. Mater. 69 (2014) 29-34.

[32]

B.Hunger, M. Heuchel, L.A. Clark, R.Q. Snurr, J. Phys. Chem. B 106 (2002) 3882-3889.

[33]

Y. Miyamoto, N. Katada, M. Niwa, Micropor. Mesopor. Mater. 40 (2000) 271-281.

[34]

H. Karge, J.P. Lange, A Gutsze, M. Laniecki, J. Catal. 114 (1988) 144-152.

[35]

H. Hu, F. Cao, W. Ying, Q. Sun, D. Fang, Chem. Eng. J. 160 (2010) 770-778.

TABLES CAPTIONS

Table 1 Table 2

Morphological, chemical and textural properties of the zeolites. Results of NH3 chemisorption measurements of zeolite catalysts.

Table 3

Coke formation: kinetic parameters for 1st order model. Test conditions - Methanol: 5.6 %mol; carrier flow rate: 60 Nml/min; WHSV: 2 gMetOH/(gcat h). Maximum TOS: 6h (a), 24h (b), 44h (c)

FIGURES CAPTIONS Figure 1 XRD pattern of the investigated samples. Figure 2 Methanol conversion (●), equilibrium conversion (▬) and DME yield (▲) over BEA zeolite Test conditions Methanol: 5.6 %mol; carrier flow rate: 60 Nml/min; WHSV: 2 gMetOH/(gcat h)

Figure 3 Methanol conversion (●), equilibrium conversion (▬) and DME yield (▲) over MFI zeolite Test conditions Methanol: 5.6 %mol; carrier flow rate: 60 Nml/min; WHSV: 2 gMetOH/(gcat h) Figure 4 Methanol conversion (●), equilibrium conversion (▬) and DME yield (▲) over FER zeolite. Test conditions Methanol: 5.6 %mol; carrier flow rate: 60 Nml/min; WHSV: 2gMetOH/(gcat h)

■), BEA (▲) and FER (●) structure on MTD process (dashed lines are only reader guideline). Test

Figure 5 Effect of MFI (

conditions - Methanol: 5.6 %mol; carrier flow rate: 60 Nml/min; WHSV: 2 gMetOH/(gcat h)

○), MFI (●) and FER (◊) (. Test conditions -

Figure 6 Catalyst stability over time: methanol conversion at 300 °C over BEA (

Methanol: 5.6 %mol; carrier flow rate: 60 Nml/min; WHSV: 2 gMetOH/(gcat h)

○), MFI (●) and FER (◊) . Test conditions -

Figure 7 Catalyst stability over time: selectivity to DME at 300 °C over BEA (

Methanol: 5.6 %mol; carrier flow rate: 60 Nml/min; WHSV: 2 gMetOH/(gcat h)

○), MFI (●) and FER (◊) as a function of TOS at 300 °C.

Figure 8 Coke amount deposited over BEA (

FIGURES CAPTION ON SUPPLEMENTARY INFORMATIONS Figure S.1

○), MFI (●) and FER (◊)

N2 adsorption isotherms at 77 K of the activated catalyst samples of BEA (

Figure S.2

NH3Temperature desorption curve and deconvolution of the activated catalyst samples.

Si/Al

Microp. Spec. Areaa Microp. Vol.a

BET

SAMPLE [mol/mol] [m2·g-1]

Crystal size

[m2·g-1]

[cm3·g-1]

[µm]

BEA

38

468

353

0.163

0.7

MFI

38

361

186

0.085

4.7 x 5.5

Crystal shape

Spherical Octagonal with twin intergrowths

FER a

34

213

166

0.076

1.7 x 5.4

Laminated

Determined by t-plot method.

NH3-uptake SAMPLE

BEA

Td1 [a]

[µmol/gcat]

[°C]

609

215

Td2 [c] x1 [b]

Weak sites Strong sites x2 [d]

[°C] 0.42

349

0.58

[µmol/gcat] [µmol/gcat] 256

353

MFI

515

223

0.36

396

0.64

185

330

FER

273

203

0.33

412

0.67

90

183

[a] Temperature of maximum desorption of NH3 between 100 and 300°C [b] Fractional population of sites between 100 and 300°C (weak sites) [c] Temperature of maximum desorption of NH3 between 300 and 500°C [d] Fractional population of sites between 300 and 500°C (strong sites)

C∞

τ

r0

r2

[gcoke·100 g-1]

[h]

[gcoke·(100 g·h)-1]

[-]

BEAa

6.50 ± 0.13

0.44 ± 0.07

15.06 ± 2.73

0.996

MFIb

3.51 ± 0.09

0.65 ± 0.11

5.52 ± 1.09

0.987

FERc

1.91 ± 0.18

10.7 ± 3.2

0.19 ± 0.08

0.975

SAMPLE

SUPPLEMENTARY INFORMATION SECTION