Oxidative Dehydrogenation of Ethylbenzene over ZSM-5 Type Chromosilicates in the Presence of CO 2

In this work, ZSM-5 type chromosilicate samples as K[Cr]ZSM-5(KCS) and Na[Cr]ZSM-5(NCS) were prepared by hydrothermal method and their catalytic properties were investigated for the oxidative dehydrogenation of ethylbenzene in the presence of CO2 as an oxidant using a fixed-bed stainless steel reactor. The prepared samples were characterized by their morphology (SEM), structural parameters (XRD), and textural parameters (BET). The performance of these catalysts was evaluated in terms of conversion, styrene yield, and selectivity. The KCSBW catalyst (potassium chromosilicate before washing with distilled water) afforded the highest styrene yield, 56.19%, with the selectivity of 96.05% in the presence of CO2 because of the coexistence of potassium ion and Cr2O3 in its structure and their synergistic effect. The influence of the presence of Cr2O3 and sodium or potassium ion on the catalytic activity of the chromosilicate samples in the catalytic EB dehydrogenation process was discussed in detail. Moreover, according to the results, the catalytic activity of the chromosilicate samples (CS) in EB dehydrogenation was increased by decreasing the surface area.


Introduction
Styrene (ST) belongs to the most important monomers in petrochemical industry and is mainly used for the synthesis of many polymers [1]- [4].The commercial production of styrene (ST) is performed by the catalytic dehydrogenation of ethylbenzene (EB), utilizing Fe-K-Cr oxide-based catalysts in the presence of superheated steam as a diluent (reaction I) at high temperatures of 600˚C -700 °C [5].
Recently, the dehydrogenation of EB in the presence of carbon dioxide as a mild oxidant (reaction II) has been extensively investigated [6]- [8].This new strategy is considered an energy-saving and environmentally friendly process.Carbon dioxide, an unwanted greenhouse gas, is depleted by applying in the reaction.According to the literature, the energy required for the new process using CO 2 is 1.9 × 10 8 cal•t −1 , only 13% of which is for the conventional commercial process using steam [1] [9].
According to Mimura et al., there are two possible reaction pathways for the dehydrogenation of EB in the presence of CO 2 [1].The one-step pathway shown in Equation ( 1) includes the direct oxidative dehydrogenation of EB with CO 2 .The two-step pathway shown in Equations ( 2) and (3) includes the reaction coupling of EB simple dehydrogenation with the reverse water-gas shift (RWGS) [10].According to the method reported by Mimura, the dehydrogenation of EB over the catalyst in the presence of CO 2 was estimated to proceed via 45% of a one-step pathway and 55% of a two-step pathway [11].
Various catalysts have been used in the dehydrogenation of EB in the presence of CO 2 , which include ironbased oxides, hydrotalcite-like oxides, zeolites, spinel oxides, active carbon, pillarised clays, active carbon containing promoters (Li, Be, Na, Mg, K, Ca), and carriers for active components (Fe, V, Cr, Cu, Zn, Zr, Ce, La, Ni, Co) [12]- [25].Among these catalysts, vanadia and chromia are believed to be more active in the dehydrogenation of EB with CO 2 [10].Additionally, according to the fact that support plays a decisive role in disposing the dispersion of the active phase, supported metal or metal oxide catalysis has been extensively increasing both in fundamental and industrial applications [26].For an instance, it is believed that the catalysts of iron oxide supported on silicon oxides are active and selective in the above reaction [27].
In this contribution, our main purpose was to prepare chromosilicate samples with different alkaline metals (M[Cr]ZSM-5 (M = K, Na)) and study their catalytic activity in the oxidative dehydrogenation of ethylbenezene in the presence of CO 2 .It is interesting to mention that chromosilicate can be considered the supported chromia in a silicon oxide matrix (as a conventional support).For this purpose, the influence of alkali-metal type (K or Na) of chromosilicate structure as well as the amount of chromium oxide on their catalytic performance in the EB dehydrogenation process was investigated.

Preparing M[Cr]ZSM-5 (M = K, Na)
Na[Cr]ZSM-5(NCS)and K[Cr]ZSM-5 (KCS) were prepared according to the procedure in our previous work [28].Materials that were used for the preparation of the samples were as follows: silicic acid (Merck), chromium (III) nitrate (Merck), potassium (or sodium) carbonate (Merck), and tetrapropylammonium bromide (TPABr, Merck) (as template).Molar ratios of the synthesized samples were Si/Cr = 15, TPABr/Si = 0.17, H 2 O/Si = 35, and K (or Na)/Si = 2.In a typical synthesis, at first the solution of silicic acid and sodium (orpotassium or cesium) carbonate in distilled water was prepared and added slowly to the solution of Cr(NO 3 ) 3 -9H 2 O and sulfuric acid in water with pH = 0.5; then, pH of the mixture adjusted at about 9 using concentrated sulfuric acid.Then TPABr was added; and the obtained gel was maintained at room temperature for about 4 h under vigorous stirring.The resulting gel was transferred into Teflon-lined stainless-steel autoclave.Na[Cr]ZSM-5 or K[Cr]ZSM-5 or Cs[Cr]ZSM-5 was obtained after about 100 h of hydrothermal treatment of the mixtures at 155˚C.After crystallization, the as-synthesized solids were washed several times with distilled water and dried overnight in an oven at 120˚C.Calcination of the samples was made under air for 5 h at 550˚C.At this point, the obtained product (MCS) was washed with distilled water in order to remove chromium species.Thus, KCS samples before and after washing, were labeled as KCS BW and KCS AW , respectively.By the same way, NCS samples before and after washing were labeled as NCS BW and NCS AW , respectively.

Characterization of Catalysts
The X-ray powder diffraction profiles of the samples were examined by means of a STOE X-ray diffractometer system (Version: PKS-2.01)equipped with Cu Kα radiation (λ = 1.540598Å).The applied voltage and current were 40 kV and 40 mA, respectively.The samples were scanned from 2θ of 4˚ -70˚ with the step size of 0.06˚ and counting time of 1.0 sec/step.
Scanning electron microscopy (SEM) and dispersive analysis (EDX) were performed to characterize the morphology and surface composition.The analyses were carried out using Philips XL-30 system equipped with an EDX detector on the samples coated with a thin layer of gold.

Apparatus and Procedure
Figure 1 shows a schematic diagram of the used experimental setup for the dehydrogenation of EB process.The dehydrogenation experiments were carried out on catalysts, pelletized, crushed, and sieved at about 0.5 -1 mm diameter under atmospheric pressure in a continuous-flow type reactor made of a stainless steel tube (4.3 mm i.d. and 500 mm length) placed in an electrically heated furnace.Since the low reactor pressure gradients of the catalyst bed improves the yield and selectivity of ST production, in the industry, it is better to use the catalyst in the form of pellet or extrudate.Thus, our samples were supplied in the form of cylindrical pellets of 1 mm diameter and 0.7 cm length.Prior to the dehydrogenation of EB, the reactor was loaded with 2.0 g of the catalyst with the support of quartz beads.The reaction was performed at 973 K for 4 h.Before EB was introduced into the reactor, the catalyst was pretreated under the flow of N 2 (100 ml•min −1 ) from room temperature to 973 K at the heating rate of 3˚C/min and maintained at this temperature for 30 min.A mixture of EB and CO 2 (CO 2 : EB molar ratio ≤3) was introduced into the reactor using a pump with the feed rate of 36.64 ml•min −1 .The outlet stream from the reactor was passed through a condenser equipped with the ice bath for GC analysis.The obtained products were analyzed by a Varian 3800 CX gas chromatograph apparatus using a flame inductivity detector (FID).Helium was used as the carrier gas, the injection temperature was 180˚C, and the samples were injected into the split injection mode.For each measurement, at least three repeated injections were taken, which obtained reproducible results.ST, toluene, and benzene were the main desired products.The peak area normalization method was used for the quantitative analysis of the products.Benzene was the standard sample (sensitivity factor was set at 1).
( ) where ethylbenzene out or styrene out stands for the amount of ethylbenzene or styrene in the products.Moreover, the styrene yield is calculated using the output data of GC analysis.

XRD Analysis
The XRD patterns of the prepared chromosilicate samples are given in metry in the literature [27], which was characterized by peaks at 2θ = 7.94, 8.86, 23.10, 23.9, and 24.45, representing (011), ( 200), (051), (033), and (313) planes of crystal structure, respectively.Additionally, an extremely small peak was observed at 31.68 due to the crystalline CrO 3 phase in NCS sample which was calcined in air.It is noteworthy that this small peak disappeared after washing the mentioned sample with distilled water.Moreover, there were also some differences in the intensities of some lines between these samples and ZSM-5 zeolite (prepared in our laboratory [29]).These variations confirmed the incorporation of chromium in the structure [28].

BET Surface Analysis
BET surface areas (S BET ) of KCS AW , KCS BW , NCS AW , and NCS BW samples are presented in Table 1.As can be observed, NCS samples, compared with KCS samples, had a relatively higher surface area.Between the NCS or KCS samples, the sample after washing with distilled water (NCS AW or KCS AW ) has higher surface area than the one before washing, which can be related to the removal of significant amounts of Cr 2 O 3 after washing with distilled water.

Morphology Studies
SEM images of KCS and NCS samples, before and also after washing with water, are shown in Figure 3.It can be distinguished that NCS appears in the form of cubic particles; its particle size ranged from 7 to 8 μm, whereas KCS had well-formed and fairly uniform coffin-type crystals of relatively large sizes (about 20 -21 μm).The SEM images of KCS and NCS samples revealed the proper growth of crystal grains.However, a few smaller grains with low crystallinity or even amorphous still existed, which were perhaps due to the incomplete crystallization of some particles or could be assigned to CrO 3 particles.In order to understand the origin of these particles, we decided to perform SEM analysis for the mentioned samples after washing with distilled water.The aim was to understand the nature of these particles.According to Figure 3, it can be inferred that washing the samples with water resulted in removing a large amount of the grains and a small ratio of them was left.So, a large amount of these particles could be related to CrO 3 particles and the remaining particles were the ones which could not completely perform crystallization process.This result is in agreement with EDX analysis (Table 2).

Catalytic Performance of the Prepared Samples
A great deal of research has been conducted for the development of catalysts with improved dehydrogenation activity for the new process, since the commercial K-promoted iron oxide catalysts do not work effectively for EB dehydrogenation in the presence of CO 2 .For this purpose, a variety of metal oxide catalysts, such as iron oxide [30] [31], vanadia [8] [32], chromia [7] [33], ceria [7], and zirconia [34] has been studied by many researchers.
On the other hand, it should be mentioned that silica-and alumina-supported chromium oxides are industrially used for the productions of lower alkenes such as ethene, propene, and isobutene through the dehydrogenation of the corresponding alkanes [35] [36].Carbon dioxide is believed to enhance the dehydrogenation of ethane [8]    [37] [38], propane [4] [24] [39]- [41], or isobutene [34] over supported chromium oxide.The reduction-oxidation property and the appropriate dispersion of chromium species on the support are important in these catalytic reactions [18] [42].Aiming to find alternative catalysts for this reaction, we utilized the mentioned methods for improving the activity of Fe-K commercial catalyst in the dehydrogenation of EB process in the presence of CO 2 .
In this trend, some new catalysts had to be designed on the basis of the commercial catalyst for the purpose of performing the mentioned process with proper efficiency.Chromosilicate (which can be considered the supported chromium oxide in the silicon oxide matrix) can be a good candidate.Furthermore, for the purpose of understanding the roles of chromium oxide and alkali-metal type (K or Na) of chromosilicate structure, in determining the catalytic activity of different types of chromosilicate, various chromosilicate samples were prepared.Table 1 and Figure 4 present the results of the dehydrogenation of EB with CO 2 over the various chromosilicate catalysts and a Fe-K commercial catalyst for comparison.We found that, in the presence of the new catalysts, EB was predominantly converted into ST via an oxidative manner and ST selectivity was improved at the cost of a decrease in conversion [43].According to the results (Table 1, Figure 4), the MCS-based catalysts can be ordered in terms of their selectivity toward ST production as follows: KCS BW ˃ NCS BW ˃ NCS AW ˃ KCS AW This order can be related to the important role of potassium ion as an effective promoter as well as that of Cr 2 O 3 as a proper co-promoter in these catalysts.In the case of KCS BW sample, both potassium ion and Cr 2 O 3 coexisted in this catalyst and the high conversion and selectivity were the result of the mentioned species.By removing the significant amounts of Cr 2 O 3 , via washing, the obtained sample (KCS AW ) had lower selectivity even compared with its sodium analogous (NCS AW ).This result can highlight the importance of the synergic effect of potassium ion and Cr 2 O 3 for improving the catalytic activity of the sample.The presence of sodium instead of potassium as the alkali metal ion in the structure of chromosilicate can slightly enhance the ST selectivity of the CS AW samples (chromosilicate samples after washing with distilled water and removing a significant amount of Cr 2 O 3 ).This trend confirmed that the synergic effect of potassium ion and Cr 2 O 3 for improving the catalytic activity was a little bit more important than that of sodium ion and Cr 2 O 3 .
Moreover, regarding the results of Table 1, it is noteworthy that the catalytic activity of the chromosilicate samples in EB dehydrogenation increased by decreasing the surface area.

Conclusion
The catalytic performance of chromosilicate catalysts as M[Cr]ZSM-5 (MCS) (M = K, Na) was investigated in the oxidative dehydrogenation of EB through CO 2 .The KCS BW catalyst (potassium chromosilicate before washing with distilled water) afforded the highest ST yield, 56.19%, with the selectivity of 96.05% in the presence of CO 2 because of the coexistance of potassium ion and Cr 2 O 3 in its structure.By removing the significant amounts of Cr 2 O 3 , via washing with distilled water, the obtained sample (KCS AW ) had lower selectivity, even in comparison with its sodium analogous (NCS AW ).This result could highlight the importance of the synergic effect of potassium ion and Cr 2 O 3 for improving the catalytic activity of the sample.This trend confirmed that the synergic effect of potassium ion and Cr 2 O 3 for improving the catalytic activity was a little bit more important than that of sodium ion and Cr 2 O 3 .Moreover, regarding the results, it is noteworthy that the catalytic activity of the chromosilicate samples (CS) in EB dehydrogenation is increased by decreasing the surface area.

Figure 1 .
Figure 1.Schematic diagram of the experimental setup used for dehydrogenation of ethylbenzene to styrene.

Figure 4 .
Figure 4.The performance of various chromosilicate samples (M[Cr]ZSM-5) in ethylbenzene dehydrogenation reaction in the presence of CO 2 at 600˚C.

Table 1 .
BET surface area and catalytic activity of the prepared catalysts in EB dehydrogenation at 600˚C and CO 2 flow of 36.64 ml•min −1 .

Table 2 .
EDX information of the samples.