MOF-808: a novel solid-acid catalyst for methanol conversion to environmentally clean fuel dimethyl ether

Abstract In this study, metal–organic frameworks (MOFs) based on zirconium which are called MOF-808 were synthesized by solvothermal method. A series of Zr-MOF were synthesized at different temperatures and reaction times to find the best synthesis samples to produce dimethyl ether from dehydration of methanol. As synthesized MOFs were characterized with x-ray diffraction, standard electron microscopy, and BET, they showed high methanol conversion and selectivity for methanol oxidation. The relationship between crystallinity, reactivity and relation catalysts life was discussed. For the first time, this MOF was used as a solid catalyst in the conversion of methanol to dimethyl ether due to its thermal stability and moderate acidity. The results show that all samples had a crystalline structure and were properly synthesized. The performance of the synthesized catalysts, the selectivity, and the effect of reactor temperature in the conversion of methanol to dimethyl ether in a tubular reactor with weight hourly space velocity = 5 h−1 were investigated. The highest surface area is nearly 1360 m2/g, size dimension of samples is in the range of 200–800 nm, and the highest Langmuir surface area is 1530 m2/g.


Introduction
In recent years, environmental issues such as air pollution and global warming, which are a result of the greenhouse gases, made a strong revision in the usage of fossil fuels. The increasing problems of the energy safety and the future supply of oil made the global community replace the conventional fuels with non-petroleum fuels to increase efficiency and effectiveness in energy consumption (Semelsberger, Borup, and Greene 2006). It has been a decade that dimethyl ether has been generating a broader interest as a promising alternative in the industry as well as a propellant in cosmetics, and in particular in hairsprays and styling foams (Yaripour et al. 2005).
Dimethyl ether (DME, also known as methoxymethane) is the simplest ether organic compound with a chemical formula of CH 3 OCH 3 , and a boiling point of À25 C (Semelsberger, Borup, and Greene 2006;Brunetti et al. 2020;Kampen et al. 2021). DME is a colorless chemical compound, nontoxic, harmless to the environment, and almost an odorless gas at ambient conditions, which is currently being used as a versatile chemically stable propellant for dyeing industry, chemicals, agriculture, and health services (Semelsberger, Borup, and Greene 2006;Brunetti et al. 2020;Kampen et al. 2021). DME could be an ideal choice to replace the common diesel fuels and liquefied petroleum gas (LPG) and this has turned DME to gain considerable attention of whether it can be replaced with diesel fuels and liquefied petroleum gas (LPG) or not (Yaripour et al. 2005;Semelsberger, Borup, and Greene 2006;Kampen et al. 2021). DME is expected as a clean fuel for the following reasons; 1. Unlike other similar ethers, DME could be transported and stored easily and safely without forming flammable peroxides (Azizi et al. 2014). 2. DME has only the C-O and the C-H bonds (there isn't a C-C bond in the structure of DME) and additionally it contains 35% of oxygen, making it an environmentally friendly fuel.Because during its combustion, less amount of carbon monoxides and unreacted hydrocarbons is produced comparing with the natural gas combustion (Azizi et al. 2014). 3. In addition, the similarity of vapor pressure and other properties between DME and LPG has made transportation and storing much easier. Therefore DME could be considered as one of the alternative fuels in the future (Azizi et al. 2014). DME could be produced in two ways (Bakhtyari and Rahimpour 2018). The first is the direct way, where there is only one-stage process with the heterogeneous catalysis as can be seen in Figure 1 (Bakhtyari and Rahimpour 2018).
The second is the indirect way in which acid catalysis is used to decompose methanol (CH 3 OH) which can be seen in Figure 1 (Bakhtyari and Rahimpour 2018).
The most important catalysts in these two processes are alumina (cAl 2 O 3 ) (Yaripour et al. 2005) and acidic zeolites (ZSM-5) (Ortega et al. 2018), and recently, metal-organic frameworks (MOFs) are used in the catalysis for the formation of dimethyl ether (Ortega et al. 2018). MOFs, well known as porous coordination polymers (PCPs), represent a new class of hybrid organic-inorganic materials and have gained special attention over the past two decades (Allendorf and Stavila 2015;Butova et al. 2016).
The PCPs, in comparison with other chemical porous compounds, have got functional flexibility to set their size, shape, dimensions, and functional groups. PCPs can make changes in ligand bridging and metal ions and so they are considered as a multifunctional material with significant physical properties to have luminescence design similar to magnets (Rowsell et al. 2004). The starting materials consist of metal ions (clusters) and organic ligands, and the synthesis process includes different synthesis methods such as solvothermal, physical mixing, and other methods (Rowsell et al. 2004;Furukawa et al. 2013;Tezerjani, Halladj, and Askari 2021;Al-Attri, Halladj, and Askari 2022).
MOF-808 showed high chemical stability (Ly et al. 2018) and excellent catalytic activity under both acidic and neutral pH reaction conditions (Zheng et al. 2018).
In our study, for the first time, MOF-808 was synthesized via solvothermal method with excellent properties as a solid acid catalysis in the dehydration of methanol to produce DME. Due to the excellent properties of MOF-808, especially high thermal, chemical stability, and catalytic activity (Su arez et al. 2018), the performance of these catalysts can be considered challenging to produce DME as a green and environmentally friendly fuel.

Preparation of initial solution
To prepare the initial solution used in the synthesis of MOF-808 crystals, first 2.8 g of zirconium chloride (ZrCl 4 ), 0.84 g of trismic acid (H 3 BTC), 160 ml of DMF, and 200 ml of acidic acid were mixed together. The resulting mixture was stirred at ambient temperature for 1 h using a magnetic stirrer to obtain a uniform solution. After 1 h, the resulting solution was transferred to a glass container and placed in an oven. The heating time in the oven is set to 24, 48, and 72 h and the synthesis temperature to 130 and 150 C. At the end of the heating step, the glass container was placed at room temperature to cool completely (Table 1).

Washing, extraction, and drying
The slurry solution obtained from the crystallization step is biphasic, and a centrifuge is used to separate it from the liquid phase. Fifteen minutes duration is selected for the first step, which is performed for solid separation. The remaining solid is then washed three times with DMF. Then, in order to completely remove the unreacted material from the cavities and the reaction medium, the resulting solid is immersed in 100-200 ml of DMF solution and transferred again to a glass container. Then it is placed in the oven at 120 C for 20 h. The process of washing and extraction with distilled water and acetone is repeated in the same way. In all the steps, the centrifuge speed is set to 6000 rpm. Washed solid was placed in an oven at 200 C for 20 h for drying.

Determining the specifications of the samples
The nature and characteristics of the synthesized samples were analyzed by x-ray diffraction (XRD) (operating at 40 kV and 30 mA with a wavelength of ø¼1.54056 and recorded in step scanning on an Equinox 3000 (INEL, France)), Scanning electron microscopy (SEM) (SEM: SERON, AIs2100) and BET nitrogen adsorption/desorption (N2 adsorption/desorption) at 70 C (BELSOEP-mini II).
To determine the crystallinity using XRD data, the intensity or area below the characteristic peaks is usually used. In this way, regardless of the context of the XRD pattern, the total intensity (or area below) of the main characteristic peaks for each sample is measured relatively to the total intensity (or area below) of the same peaks in the reference sample.
The XRD patterns of the synthesized MOF-808 samples in Figure 2 and Figure S1 show that all the diffraction peaks are sharp and matched well with those of simulated pattern (Li et al. 2015). Also, the XRD pattern of all samples does not exhibit any background showing that no amorphous phase is presented in the samples (Cejka et al. 2007).
The relative crystallinity of the samples from the XRD data was measured based on the reference (Utchariyajit and Wongkasemjit 2010) by the following equation: in which, A is the area under the diffraction peak of each sample at 2Ɵ¼ 4. 3 , 8.3 , 8.7 , 10.1 , 9.34 , 11.0 , 12.52 , and 17.87 and As is the peak area of the sample with the highest amount considered as the standard sample.

Reactor
To test the performance of the catalysts, a fixed bed reactor made of steel with an internal diameter of 5 mm and continuous flow was used. The heat transfer to the reactor is provided by an electric jacket located around the reactor, and the reaction temperature is measured by a thermocouple installed at the point where the catalyst is placed.

2.4.2.
Loading and activating the catalyst Hold 0.3 g of the catalyst at a point in the reactor using two metal grids. Fill the entire volume of the reactor with ceramic spheres that can withstand high temperatures. Then in order to activate the catalyst, nitrogen gas at 500 C temperature, 1 bar pressure, and 100 ml/min flow rate is passed through the reactor bed for 1 h. This is done to remove the gases and the moisture absorbed from the catalyst and to prepare it.

Input feed
After activating the catalyst, the nitrogen gas inlet valve is closed. Feed (pure methanol) enters the reactor. 2.4.4. Operating temperature and pressure One of the most important parameters affecting the conversion rate of methanol and DME production is the operating temperature. In order to study this parameter in detail, its effect was evaluated at three temperatures of 350, 400, and 450 C. All experiments were performed at atmospheric pressure.

Gas chromatography
The output of the reactor enters the condenser and the separated gas is stored in the gas chamber and injected into the gas chromatography (GC) machine. In this device, a plot-Q column is used, in which the gas separation and identification operation is performed. The Agilent 6890N GC device (Santa Clara, CA, USA) with two detectors is used in the process.
2.4.6. Mass space speed per time unit, weight hourly space velocity Mass space velocity per time unit is one of the most important parameters. This parameter shows the ratio of the mass of methanol entering the reactor per time unit to the mass of the loaded catalyst. It has a direct effect on the output of products, and the lower the amount the higher the conversion of methanol and output. This parameter is low when the loaded catalyst amount is high, and/ormethanol inlet flow rate is low resulting in more methanol contact whith the catalyst and accordingly an increase in the conversion rate.
where Q is the input methanol flow rate (1.8 ml/h) and P is the methanol density (0.791 g/ml). By using Equation (3), weight hourly space velocity (WHSV) ¼ 5 h À1 .

XRD, SEM, and BET analysis
As can be seen in Figure 2 and Figure S1, the characteristic peaks of all the synthesized samples corresponded to the sample peaks simulated from the reference, and the XRD pattern of all the synthesized samples was consistent with previous reports (Plessers et al. 2016), indicating that the MOF-808 crystals are synthesized well. Similarly, the main peaks in the MOF-808 simulated model include three large peaks in 2Ɵ ¼ 4.3 , 8.3 , and 8.7 and two smaller peaks in 2Ɵ ¼ 10.1 and 11.0 as shown in Figure 3. The crystallinity is calculated and reported using High score plus Xpert software (Table 1) and according to Equation (1). As can be seen in Table 1, the relative crystallinity also increases with increasing synthesis time and temperature in the MOF-808 sample. The difference in the crystallinity can be attributed to the structural defects of these materials. This means that in many cases the MOF-808 does not have an ideal and complete structure. Loss of the number of coordinated ligands with metal clusters leads to structural defects. Therefore, due to the variation in the number of coordinated ligands, the order of the crystals formed in the long range is reduced and the number of that repeating units is reduced, thus the intensity of X-ray diffraction will also be reduced. As a result, in Figure 2 and Figure S1, synthesized samples of MOF-808 catalysts, by increasing the temperature and synthesis time, the intensity of XRD patterns peaks has been increased.
SEM images were used to study the morphology and crystal size of MOF-808 samples. As can be seen in Figures 3 and 4, all samples have a crystalline structure, which is consistent with XRD analysis. MOF-808 specimens have uniformly separated particles. SEM images (Figures 3  and 4) show that by increasing the synthesis time from 24 to 72 h, the size of particles has been decreased (Al-Attri, Halladj, and Askari 2022).
However, the formed particles cannot be considered completely separate from each other, because the internal growth of octagonal particles with dimensions of 200-800 nm can be seen in the images. The particle size of the samples was calculated using Images software. Figure 5 shows the adsorption and desorption isotherms of MOF-808 samples. The isotherms of all samples are type I, indicating the presence of micropores on the structure of the compounds. The presence of hysteresis, in addition to being a reason for the presence of meso-cavities in the material, can be used to determine the geometry of cavities. The results of this analysis, which are the porosity parameters of MOF-808 samples, such as specific surface areas were determined (Table 2 and Figure 5).
Samples MOF-808-1 and MOF-808-2 with structural defects had a higher specific surface area of BET and total volume than samples MOF-808-3 and MOF-808-4. Changes in the specific surface area of BET and total volume can be attributed to the effect of temperature and time synthesis on the crystals formed, because at low temperatures and short time, MOF-808 crystals have more structural defects. As a result, the specific surface area of BET cavities and the ratio of the outer surface to the micropore surface increase.

Investigation of catalysts performance in MTD process
The synthesized catalysts are loaded into the reactor, and the details are given in Section 2.4. After activating the catalyst, the feed enters the reactor and after the reaction, by analyzing the gas leaving the reactor and the liquid leaving the condenser, the selectivity, yield of dimethyl ether, and the conversion rate of methanol can be obtained as follows (Pimprom et al. 2015;Ahmadova et al. 2018 Many mechanisms investigated methanol dehydration over acidic catalysts (Goda, Abdelhamid, and Said 2020;Goda, Said, and Abdelhamid 2021). In the first stage of the conversion of methanol, molecules of methanol adsorb on the surface of the catalyst. Zr-sites in MOF-808 catalyst can act as acid sites that adsorb methanol molecules. Then, these adsorbed methanol molecules (dissociated) transform to methoxy (CH 3 O) during an oxidation reaction. Finally, these methoxy molecules can be converted to DME by a combination of two methoxy species, or an interaction between a methanol molecule and a methoxy species (Kazemzadeh et al. 2022).

Methanol conversion
In order to calculate the conversion percentage of methanol (according to Equation (4)), the mass of unreacted methanol in the condensed liquid present in the condenser must be calculated. Since the collected liquid contains water and methanol, the weight percentage of water and methanol is obtained by analyzing the liquid and using the calibration curve. By measuring the mass flow rate of the condensed liquid, the mass of methanol does not react and the percentage of methanol conversion is calculated.
In general, DME is produced from a solid-acid catalyst on weakly and moderately acidic sites. As the acidity of the acid catalyst increases, significant amounts of by-products such as hydrocarbons and even coke are formed. As a result, the conversion rate of methanol to DME decreases. Accordingly, modification of the used catalyst acidity is necessary to increase the selectivity and activity of the catalyst. The results of the study of MOF-808 catalysts that had different temperatures and synthesis times show that the synthesis temperature and synthesis times affect the acidity, activity, and longevity of the production catalyst. Figure 6 shows the methanol conversion rate of MOF-808 samples. It can be seen that the methanol conversion rate increases with increasing synthesis temperature and synthesis times.
In the early times after the start of the process, methanol conversion is almost complete and DME production is not highly required. As coke is not formed, the resulting products are easily removed from the pores. Over time, DME production declines due to coke formation and clogging of cavities. As can be seen, the MOF-808-4 had the highest catalytic activity and was inactivated at a lower rate. On the other hand, samples that were synthesized at a lower time and temperature lost their activity at a higher rate. This can be seen in the acidification correction, as the structural defects of MOF-808 catalysts are reduced as the synthesis temperature and time increases. Therefore, the number of metal clusters capable of dehydrating methanol molecules is optimized, thus reducing the acidity of the catalyst. According to research, catalysts with less acidity get less coke and as a result are inactivated in a later time (Askari, Halladj, and Sohrabi 2012). Figure 7 shows the selectivity of the MOF-808-4 catalyst relative to DME during the MTD process. As can be seen, the selectivity at the beginning of the reaction to DME is 98%, which decreases to 97% after 8 h from the start of the  reaction. Therefore, in general, it can be said that selectivity remains constant during the process. Methane and butane are also obtained as by-products during the reaction, accounting for a small percentage. All the synthesized catalysts showed similar behavior in terms of selectivity, which indicates the synthesis temperature and time have no effect on the selectivity of these materials. Due to the similar behavior of the catalysts, the selectivity of only one of the catalysts, MOF-808-4, is plotted.

Investigation of the effect of reactor temperature
One of the most important parameters affecting the conversion rate of methanol is the process temperature. Therefore, the MOF-808-4 sample was tested at 400 and 450 C in addition to 350 C. As shown in Figure 8, when the reactor temperature rises from 350 to 400 C, the conversion rate of methanol increases significantly and the catalyst remains active for a longer time. However, by increasing the temperature from 400 to 450 C, in addition to reducing the conversion rate of methanol, the life of the catalyst also decreases. Therefore, temperatures above and below 400 C reduce the conversion rate of methanol (Schiffino and Merrill 1993;Askari, Halladj, and Sohrabi 2012;Ahmadova et al. 2018;Kampen et al. 2021). This can be attributed to thermodynamic constraints and inactivation of the catalyst, because in the first stage of temperature increase, due to the increase of the reaction rate coefficient, the equilibrium reaction of methanol to DME is accelerated. However, in the second stage of temperature increase, due to the exothermic nature of the methanol to DME conversion reaction (DH¼ À23 Kj/mol), the direction for the reaction changes and moves to the left. In general, investigating the effect of reaction temperature on catalyst inactivation is complex. However, as the reaction temperature increases, the rate of coke formation on the surface and inside the catalyst cavities is expected to increase (Chen, Moljord, and Holmen 2012). Thus, the formed coke blocks the openings of the cavities and reduces the activity of the catalyst.
As can be seen in Table 3, MOF-808 catalyst synthesized in the current work has the specific surface area (BET) of 1350 m 2 /g which is much higher than that of other catalysts. The BET value of other modified and conventional catalysts such as modified Ç-Al2O3 and desilicated ZSM-5 is <500 m 2 /g. So, this can be a positive and effective point for  this catalyst. Besides, the conversion of methanol to DME over MOF-808 is 99% which is higher than these modified conventional catalysts and the selectivity of DME is nearly equal to other catalysts. As a result, MOF-808 can be considered as a proper catalyst for MTD process.

Conclusion
In this article, MOF-808 was synthesized by the solvothermal method to convert methanol to dimethyl ether under methanol dehydration reaction. Samples were synthesized at two temperatures of 130 and 150 C at different times (24, 48, and 72 h). Results of XRD indicated that the MOF-808 crystals are synthesized well, but with increasing temperature and synthesis time, the internal defects decreased and the formed crystals had more completed structures and the specific surface area of BET cavities decreases.
The performance of the synthesized catalysts in a tubular reactor was investigated. The results showed that the process temperature played a significant role in the life and activity of the resulting catalysts. MOF-808 samples synthesized at high temperatures for a long time had the higher conversion rate, but selectivity of all synthesized catalysts showed a similar behavior and remains constant during the process.