During the past few years, olefin metathesis reactions have opened up many new industrial routes to important petrochemicals, polymers, and specialty chemicals [1, 2]. Olefin metathesis is a powerful carbon–carbon bond rearrangement reaction in which the unsaturated carbon–carbon bonds can be redistributed in the presence of a catalyst. This route has been proved effective to produce value-added olefins according to market demand. Typical examples are the self-metathesis reaction of propene [3] and its reverse reaction (metathesis of ethene and 2-butene to propene) [4-6]. Besides, much attention has been paid to the development of metathesis catalysts from fundamental research to commercial application [7].
The increasing demand for propene has led to propene synthesis routes being focused on during the last few years [8-10]. To date, several olefin metathesis reaction routes using C4 olefins as raw materials have been developed, including self-metathesis of 1-butene [11-13], cross-metathesis of ethene and 2-butene [14-17] or 1-butene [18], and metathesis of 2-butene [19]. In the context of declining international oil supply and the emergence of shale gas, the price of propene is volatile and sometimes falls below than that of ethene. On basis of surplus C4 olefins, self-metathesis of 1-butene to ethene and 3-hexene provides a promising route to produce value-added olefins. Here, 3-hexene could be further isomerized to 1-hexene, which is an important comonomer for the production of linear low-density polyethylene and high-density polyethylene. However, few articles on 1-butene self-metathesis have been reported, and only a few patents have been disclosed [20-22].
In this work, we present a novel route to produce ethene and hexene by the self-metathesis reaction of 1-butene. In the complicated reaction network, metathesis of 1-butene and 2-butene to propene is demonstrated to be the main side reaction. Isomerization of the 1-butene double bond is the main step determining target olefin selectivity. Consequently, the major objective of the present study is to find a catalyst suitable for the 1-butene self-metathesis reaction. The factors that influence the product distribution in the metathesis reaction are considered.
Na-type mordenite zeolite and Al2O3 composite supports were prepared by extruding a mixture of Al2O3 (prepared from pseudo-boehmite) and the Na-type mordenite zeolite (Si/Al = 20, Fushun Petroleum Company, China) into strips with a diameter of about 2 mm. The H-type mordenite and Al2O3 composite supports were ion-exchanged with 0.8 mol/L NH4NO3 solution at 80 ℃ three times to replace Na+ with NH4+, and then washed and dried. The H-type composite supports were obtained by calcination at 520 ℃ in air for 3 h, ground and then passed through 20–40 mesh filters. The Na-type mordenite and Al2O3 composite supports are denoted as NaMOR-xAl2O3, while the H-type mordenite and Al2O3 composite supports are labeled as HMOR–xAl2O3, where x represents the weight percent of alumina. Al2O3-x and SiO2 supports were provided by the Aluminum Corporation of China Limited, Shandong Branch. Catalysts labeled as xMo/Al2O3-A were prepared by incipient wetness impregnation using an aqueous solution of (NH4)6Mo7O24·4H2O and left at room temperature for 24 h; here, x indicates the weight percent of Mo in the catalysts. The catalysts were dried at 120 ℃ for 2 h and finally calcined at 600 ℃ for 2 h.
Nitrogen (N2) physisorption experiments were performed at −196 ℃ on a Micromeritics ASAP 2020 instrument. Prior to analysis, samples were outgassed at 350 ℃ under a vacuum of 10−3 Pa for 8 h. Specific surface area was calculated by the standard multipoint Brunauer-Emmett-Teller (BET) method with a P/P0 range of 0.05–0.30. Pore volume was determined by N2 adsorption at P/P0 = 0.995.
Powder X-ray diffraction (XRD) patterns were collected on a PANalytical X'pert PRO diffractometer using Cu Kα radiation at 40 kV and 40 mA. The chemical composition of samples was determined on a Philips Magix 601X X-ray fluorescence (XRF) spectrometer.
Temperature-programmed desorption of ammonia (NH3-TPD) experiments were carried out in a quartz microreactor (i.d. 4 mm), which was connected to an online gas chromatograph (Shimadzu GC-8A) equipped with a thermal conductivity detector (TCD). The TCD response was calibrated by doses containing known amounts of NH3. Typically, the sample (140 mg) was first pretreated at 605 ℃ for 0.5 h in helium with a flow rate of 25 mL/min, then cooled to 150 ℃ and saturated with NH3. Subsequently, NH3 was purged in a helium flow for a certain period at 150 ℃ until a stable baseline was obtained. The reactor temperature was raised from 150 to 605 ℃ at a heating rate of 19.8 ℃/min.
Infrared (IR) spectra were recorded with a Bruker Vertex 70 IR spectrometer at 4-cm−1 optical resolution. Prior to measurements, the samples were pressed into self-supporting wafers with a diameter of 1.3 cm and pre-treated in an IR cell at 450 ℃ for 1 h under 10−2 Pa. The cell was cooled to room temperature to record the background spectra. Each sample was saturated with pyridine (Py) and then the excess Py was removed under vacuum at 150 ℃ for 0.5 h. After cooling to room temperature again, the sample was subjected to IR measurements to quantify the amount of acid sites.
Temperature-programmed reduction of H2 (H2-TPR) experiments were carried out using a custom-made setup. Prior to measurements, each sample (100 mg) was pretreated at 550 ℃ for 1 h under Ar flow to remove water and other contaminants and then cooled to 200 ℃. The reduction step was carried out in a mixture flow of 10% H2/90% Ar from 200 to 900 ℃ at a heating rate of 15 ℃/min. The H2 signal was monitored online by a TCD detector.
Ultraviolet-visible (UV-Vis) spectra were recorded in a diffuse reflectance mode with a JASCO V-550 spectrometer at room temperature in the range of 200–600 nm at a rate of 40 nm/min.
High-resolution transmission electron microscopy (HRTEM) images were recorded on a JEM-2100 transmission electron microscope (JEOL, Japan) with an accelerating voltage of 200 kV.
The self-metathesis of 1-butene was carried out in a fixed bed microreactor with a length of 320 mm and diameter of 7 mm. The catalyst (1 g) was located in the middle part of the fixed-bed reactor. The catalyst was activated in situ at 500 ℃ for 2 h in N2 flow before the reaction. The reaction products were analyzed by an online Agilent 7890B gas chromatograph (GC) equipped with a flame ionization detector and Al2O3 capillary column.
The GC results indicated that cis-, trans-2-pentene and cis-, trans-3-hexene were the major products with C5 and C6 structures, respectively. To simplify treatment, we denoted products with the same carbon number as Cn=, where n represents carbon number.
Conversion (X(1-C4=)) and selectivity (S(Cn=)) were calculated using the following equations:
For simplicity, [1-C4=]F and [1-C4=]p represent the weight percent of the component in the feed gas and product, respectively; [Cn=]p is the weight percent of product with carbon number n.
The Gibbs free energy change is an important parameter in thermodynamic analysis. Thermodynamic data were obtained from the relevant thermodynamic manual. The correlation for the Gibbs energy of formation was a series expansion in temperature [23],
ΔfG is the Gibbs energy of formation of an ideal gas, kJ/mol; a, b, and c are regression coefficients for chemical compounds; and T is temperature, K. The Gibbs energy change for the reaction ΔrG was calculated by Eq. (2).
Scheme 1 illustrates the main possible reaction pathways in the 1-butene metathesis reaction network. Ethene, hexene, propene, and pentene are the main metathesis products. To improve the selectivity for Metathesis Ⅰ, 1-butene isomerization and the subsequent Metathesis Ⅱ should be suppressed. We first analyzed the reaction network from the viewpoint of thermodynamics. ΔrG values can be used to infer whether the reaction was thermodynamically favored at a given temperature. It is generally accepted that a reaction occurs spontaneously when ΔrG < 0. Reactions are possibly favorable when 0 < ΔrG < 50 kJ/mol [23]. As shown in Fig. 1, the ΔrG values for Metathesis Ⅰ and Ⅳ were around 4–6 kJ/mol and increased with the reaction temperature, indicating that low temperature was favorable for the target Metathesis Ⅰ reaction. ΔrG for Metathesis Ⅱ was below 0 and was as low as −6 kJ/mol for 1-butene isomerization. This means the double bond isomerization and side Metathesis Ⅱ reaction occurs more easily than the Metathesis Ⅰ reaction. Thus, the metathesis catalyst should be carefully optimized to selectively promote the Metathesis Ⅰ reaction.
Mo/HMOR-Al2O3 [24], Mo/Al2O3 [25], and Mo/SiO2 [7] have been proved good metathesis catalyst candidates in different metathesis catalytic reactions. Here, they were evaluated in the 1-butene self-metathesis reaction. The detailed catalytic results under different reaction conditions are summarized in Table 1. Dramatic differences between feed conversion and product selectivity were observed for the different supported catalyst systems. Over the 6Mo/HMOR-50Al2O3 catalyst, high 1-butene conversion was observed. However, propene selectivity was much higher than that of ethene, indicating that Metathesis Ⅱ dominated in reaction network; i.e., 1-butene first isomerized to 2-butene, followed by the cross-metathesis of 1-butene and 2-butene. This behavior was in agreement with our previous report in which Mo/HMOR-Al2O3 was the effective metathesis catalyst for the conversion of 1-butene to propene [24]. Product selectivity for 2-butene was also high, which meant 6Mo/HMOR-50Al2O3 was an effective double-bond isomerization catalyst. To suppress the formation of 2-butene, the 6Mo/NaMOR-50Al2O3 catalysts was used. As listed in Table 1, product selectivity for ethene and hexene improved, whereas 1-butene conversion decreased markedly compared with the case for Mo/HMOR-Al2O3. Low support acidity suppressed both the isomerization and metathesis activity of the system.
When 6Mo/Al2O3-A was used, high ethene selectivity was observed on the premise of high 1-butene conversion. The molar selectivity for ethene and hexene exceeded 85%, which indicated that the self-metathesis of 1-butene predominated in the metathesis reaction network. In contrast, 6Mo/SiO2 showed poor catalytic performance at 100 ℃ and 1-butene conversion increased with the reaction temperature. However, ethene selectivity decreased sharply. The selectivity for 2-butene exceeded 90%. In other words, 6Mo/SiO2 was a good double-bond isomerization catalyst, especially at high reaction temperature.
In general, 6Mo/Al2O3-A was a good candidate for the 1-butene self-metathesis reaction among the above evaluated catalysts. This catalyst could effectively promote the target reaction at low reaction temperature. Over 6Mo/Al2O3-A, metathesis of 1-butene and 2-butene to propene/pentene was the main side metathesis reaction. The double-bond isomerization reaction of butene, the key intermediate step determining the target olefin selectivity, was sensitive to catalyst acidity and reaction temperature. Low acidity and temperature could suppress 2-butene formation and the following Metathesis Ⅱ side reaction. Thus, Mo/Al2O3 with proper acidity was proved to be a good candidate catalyst for 1-butene self-metathesis.
It is generally accepted that metathesis activity is closely related to the metal loading. Optimal Mo loading on the Al2O3-A support was therefore investigated. Table 2 lists the physicochemical properties of xMo/Al2O3-A samples. The results of N2 physisorption experiments revealed that the BET surface area of the Al2O3-A support was 326 m2/g and decreased upon Mo loading [26]. Quantitative analysis of NH3-TPD profiles showed that the acid density of Al2O3-A decreased from 0.335 to 0.295 mmol/g after 2% Mo loading. However, the total number of acid sites increased with the Mo content when the Mo loading was higher than 4%. For 10Mo/Al2O3-A, the total acid density was higher than that of the Al2O3-A support. As shown in Fig. 2, the desorption peak intensity at around 240 ℃ was the biggest among all the samples, indicating that Mo loading induced the formation of new weak acid sites. Mo species preferentially interacted with the strong acid sites on the Al2O3 support and a high Mo content led to the formation of new acid sites. Similar phenomena were also observed by Miranda et al. [27]. Crystalline MoO3 was not observed in the XRD patterns of the xMo/Al2O3-A catalysts (Fig. 3).
It is well accepted that the olefin metathesis reaction follows the metal carbene mechanism over heterogeneous catalysts. The redox properties of the supported metal oxides are directly associated with carbene formation. Interactions between Mo species and alumina were investigated by H2-TPR measurements, as shown in Fig. 4. The H2-TPR profiles of all the samples contained two main reduction peaks. The peak at low temperature was consistent with the reduction of octahedral Mo species to tetrahedral Mo species dispersed on the Al2O3-A support [28]. The peak signal at high temperature was attributed to both the further reduction of the excess octahedral Mo species and the partial reduction of tetrahedral Mo species strongly interacting with Al2O3 [29]. As illustrated in Fig. 4, the ratio of the reduction peak area at 475 ℃ rose with Mo loading, indicating that the quantity of octahedral Mo species increased with Mo loading. Quantitative analysis results after fitting are listed in Table 2. Meanwhile, the reduction peak at higher temperature shifted to lower temperature with increasing Mo loading, indicating that the interaction between Mo species and the support weakened and aggregated species formed.
The effects of Mo content of the xMo/Al2O3-A catalysts on the self-metathesis of 1-butene were investigated. As displayed in Fig. 5(a), the highest butene conversion was observed over the 10Mo/Al2O3-A catalyst. Similar activity was observed when the Mo loading was in the range of 6%–8%. When the Mo content was 2%, butene conversion was low (15%). It should be noted that all the catalysts showed good catalytic stability during the evaluation period.
In the case of product selectivity, the situation was quite complicated. Propene/pentene selectivity increased with Mo loading and followed the trend of butene conversion. Combined with the NH3-TPD results, it was deduced that high acid density promoted the double-bond isomerization and following Metathesis Ⅱ reaction. According to the thermodynamic results, ΔrG of Metathesis Ⅱ was much lower than that of Metathesis Ⅰ. Once 2-butene was produced, Metathesis Ⅱ occurred preferentially, which led to the formation of propene and pentene. When the Mo loading was in the range of 2%–6%, product selectivity for ethene was around 40%. Compared with ethene, selectivity for hexene was a little lower because of the oligomerization side reaction. Taking into account both conversion and selectivity, the optimal Mo loading of xMo/Al2O3-A was around 6%.
The metathesis catalytic activity of 6Mo/Al2O3-A was closely linked with the reaction parameters. The effects of reaction temperature, pressure, and feed space velocity on product distribution were investigated in detail. It is generally accepted that Mo-based heterogeneous metathesis catalysts are activated at medium to high temperature (150–300 ℃) [30]. For the 1-butene self-metathesis reaction, the situation was quite different. As shown in Fig. 6, 1-butene conversion increased with the reaction temperature. It should be pointed out that propene/pentene selectivity also increased with the reaction temperature although ethene/hexene were the main products. At 60 ℃, the selectivity for ethene and hexene was high (90%). However, 1-butene conversion was low (20%). The optimal reaction temperature was between 80 and 100 ℃.
Although 1-butene self-metathesis is an equal molecular reaction, reaction pressure also influenced the catalyst activity because high pressure could increase the collision probability of reactant molecules. As illustrated in Fig. 7, 1-butene conversion increased from 37% to 47% with rising reaction pressure. High reaction pressure also promoted the Metathesis Ⅱ side reaction. Propene selectivity slightly increased with pressure. Suitable reaction pressure should be chosen to guarantee the high yields of ethene and hexene. Fig. 8 presents the influence of 1-butene WHSV on the metathesis activity of 6Mo/Al2O3-A. It was obvious that 1-butene conversion decreased with rising WHSV. However, the selectivity for Metathesis Ⅰ increased with WHSV. Thus, a long residence time on the catalyst could promote double-bond isomerization and the subsequent Metathesis Ⅱ reaction. In contrast, high WHSV of 1-butene was beneficial for Metathesis Ⅰ (ethene/hexene).
To achieve high yields of ethene and hexene, both 1-butene conversion and product selectivity should be considered. On the basis of the above discussion, the relative optimal reaction conditions for the 6Mo/Al2O3-A catalyst in the 1-butene self-metathesis reaction are 80–100 ℃, 0.5–1.0 MPa and 3–4.5 h−1. Under the optimal conditions, 1-butene conversion could be as high as 47% on the premise of 85% product selectivity for ethene and hexene.
It is well known that the metathesis performance of Mo-based catalysts is influenced by the nature of the support [31, 32]. On the basis of the above results, Al2O3 was chosen as the support. The textural and acid properties of Al2O3 may influence the state of Mo species and consequently metathesis activity. To investigate the relationship between reaction performance and support nature, a series of Al2O3 supports with different textural properties were selected. To distinguish the different alumina supports, Al2O3 from the above discussion is designated as Al2O3-A; the other three supports are labeled as Al2O3-B, -C, and -D.
Textural and physicochemical properties of the Al2O3-x supports and 6Mo/Al2O3-x catalysts are listed in Table 3. The BET surface areas of the supports ranged from 137 to 326 m2/g. Introduction of Mo species simultaneously decreased the surface area and pore volume. The XRF results indicated the Mo loading on each support was around 5%.
Quantitative analysis of the NH3-TPD profiles indicated that the acid density of Al2O3-x followed a similar trend to that of BET surface area. Py-IR spectra were acquired to obtain more information about the types of acid sites. As shown in Fig. 9, Al2O3-A only contained Lewis acid sites (1450 cm−1). Mo species interacted with the Lewis acid sites, which weakened the intensity of the peak at 1450 cm−1. As mentioned in Section 3.3, Mo loading may introduce new acid sites on Mo/Al2O3 catalysts. The Py-IR results indicated that no new Br nsted acid sites were formed by Mo loading. The number of Lewis acid sites on Al2O3-D was the lowest among the four supports.
UV-Vis diffuse reflectance spectra can reveal the surrounding geometries of Mo species. Fig. 10 shows the UV-Vis spectra of 6Mo/Al2O3-x catalysts. Two bands centered at around 220 and 270 nm were observed. The band at 220–250 nm is commonly attributed to tetrahedral molybdate, while that at 260–320 nm is assigned to the Mo–O–Mo bridge of octahedral molybdate and polymerized molybdate species [28, 33]. The content of octahedral molybdate in 6Mo/Al2O3-D was the highest among that of the four samples because of its lowest surface area. This was in accordance with the quantitative analysis results of H2-TPR (Table 3).
The dispersion of Mo species on alumina was further characterized using HRTEM images. As illustrated in Fig. 11(a), a uniform distribution of Mo nanoparticles over the alumina support was observed. Little difference was observed between the 6Mo/Al2O3-B and -C samples. In the case of 6Mo/Al2O3-D, the dispersion of Mo species was poor and some particles aggregated together into clusters, as indicated in Fig. 11(d). These observations are in accordance with the UV-Vis and H2-TPR results. In general, a high BET surface area of the alumina support is beneficial for the dispersion of Mo species.
Metathesis activities of the 6Mo/Al2O3-x catalysts were investigated and are listed in Table 4. The highest 1-butene conversion was observed on 6Mo/Al2O3-A and it gradually decreased following the order of 6Mo/Al2O3-B, 6Mo/Al2O3-C, and 6Mo/Al2O3-D. Combined with the above characterization results, it was deduced that effective dispersion of Mo species and high acid density were beneficial for the conversion of butene. Although the highest 1-butene conversion was obtained on 6Mo/Al2O3-A, its ethene/hexene selectivity was lower than those of 6Mo/Al2O3-B and 6Mo/Al2O3-C. Thus, a large number of acid sites may promote the double-bond isomerization and subsequent metathesis Ⅱ reaction. To suppress the side metathesis reaction, suitable acid density was necessary. In the case of 6Mo/Al2O3-D, poor catalytic stability was observed. The high ratio of octahedral Mo species led to the rapid deactivation of the catalyst.
In general, the 1-butene double-bond isomerization reaction was an important step that determined the product selectivity. Suppression of isomerization was critical to guarantee the selectivity for metathesis Ⅰ. For the Mo/Al2O3 catalyst, low reaction temperature and high WHSV were beneficial to increase the ethene/hexene selectivity. The optimal Mo loading was between 4% and 6%. Al2O3 with a large surface area contributed to the effective dispersion of Mo species on the support. To guarantee the good catalytic performance of Mo/Al2O3 catalysts, the acid density and state of Mo species should be balanced.
The catalytic performance of a series of Mo-based catalysts in the self-metathesis reaction of 1-butene under various reaction conditions was investigated. The 6Mo/Al2O3 catalyst exhibited the best catalytic performance of the examined systems at around 80 ℃ and 1.0 MPa with a 1-butene WHSV of 3.0 h−1. Under these conditions, 1-butene conversion was as high as 47% on the premise of ethene selectivity of 42%. In the complicated reaction network, ΔrG of Metathesis Ⅰ (1-butene → ethene + 3-hexene) was much higher than that of Metathesis Ⅱ (1-butene + 2-butene → propene + 2-pentene). Double-bond isomerization of 1-butene was proved as the step determining the target olefin selectivity. An alumina support with large surface area was beneficial for the effective dispersion of Mo species and formation of tetrahedral Mo species. Suitable acidity and Mo dispersion were vital factors that promoted the 1-butene self-metathesis reaction. A high ratio of octahedral Mo species may lead to the rapid deactivation of the Mo/Al2O3 catalyst.
2018,
Vol. 39 Issue (1): 37-46 
PDF
2018,
Vol. 39 Issue (1): 37-46 PDF