Research ArticleChemistry

Propane oxidative dehydrogenation over highly selective hexagonal boron nitride catalysts: The role of oxidative coupling of methyl

See allHide authors and affiliations

Science Advances  15 Mar 2019:
Vol. 5, no. 3, eaav8063
DOI: 10.1126/sciadv.aav8063

Abstract

Hexagonal boron nitride (h-BN) catalyst has recently been reported to be highly selective in oxidative dehydrogenation of propane (ODHP) for olefin production. In addition to propene, ethylene also forms with much higher overall selectivities to C2-products than to C1-products. In this work, we report that the reaction pathways over the h-BN catalyst are different from the V-based catalysts in ODHP. Oxidative coupling reaction of methyl, an intermediate from the cleavage of C─C bond of propane, contributes to the high selectivities to C2-products, leading to more C2-products than C1-products over the h-BN catalyst. This work not only provides insight into the reaction mechanisms involved in ODHP over the boron-based catalysts but also sheds light on the selective oxidation of alkanes such as direct upgrading of methane via oxidative upgrading to ethylene or CHxOy on boron-based catalysts.

INTRODUCTION

Dehydrogenation of propane by either oxidative or nonoxidative approaches is essential to the on-purpose production of propene, a valuable and versatile chemical feedstock. Oxidative dehydrogenation of propane (ODHP) in the presence of O2 faces the challenges of deep oxidation of propane and propene to COx (CO and CO2), often resulting in loss of propene selectivity and yield (1, 2). V-based catalysts have been widely studied in the ODHP with main side products of COx and negligible amount of C2-products (35). However, the boron-based catalysts recently reported for the ODHP show very low selectivity to COx and relatively higher selectivity to C2-products than to C1-products (C2/C1 > 2) (69). For example, at a propane conversion of 4.1%, C2- and C1-selectivities are 8.6 and 3.5% (C2/C1 > 2), respectively, as reported by Grant et al. (9). Huang et al. (7) also reported that the C2/C1 ratio is higher than 2 at a propane conversion of 12% at 520°C over the hexagonal boron nitride (h-BN) nanosheet.

The different C2-selectivities and C2/C1 ratios observed in the ODHP over the h-BN and the V-based catalysts imply that the reaction mechanism is different on these two types of catalysts. It is believed that ODHP over the h-BN catalyst has higher tendency toward the cleavage of only the first C─C bond of propane than that over the V-based catalysts, which results in a higher selectivity of ethylene (10). The relative reactivity of methyl formed via the C─C bond cleavage could also play important roles in affecting the product selectivity on V-based and h-BN catalysts. In the oxidative coupling of the methane reaction (OCM) (1113), methyl is generated via the C-H cleavage of methane, followed by its oxidative reactions, which mainly include three reaction pathways: (i) conversion of methyl to C1-oxygenates (14), (ii) oxidation of methyl to form COx (12), and (iii) coupling of methyl to form the C2-products (15, 16). Therefore, understanding the fate of methyl or elucidation of the reaction pathways of methyl is essential for the in-depth understanding of the reaction mechanisms involved in the ODHP process over the boron-based catalysts.

Here, the reaction pathways of methyl derived from the C─C bond cleavage in ODHP over the h-BN catalyst were systematically studied. Among these reaction pathways, the oxidative coupling of methyl to C2-products was found to result in a higher C2/C1 ratio (>2), contributing to the higher C2-selectivities over the h-BN catalyst in ODHP.

RESULTS AND DISCUSSION

Figure 1 (A and B) shows the influence of reaction temperature on the catalytic performance of ODHP over the h-BN and VOx/γ-Al2O3 catalysts, respectively. On the h-BN catalyst, the main by-product is C2-products and its selectivity increases with increasing propane conversion (Fig. 1A). In addition, the selectivity to C2H4 over the h-BN catalyst is higher than that to C2H6, which can mainly be attributed to the cleavage of propane. On the contrary, the main by-products are COx (CO and CO2), with a negligible amount of C2-products on the VOx/γ-Al2O3 catalyst (Fig. 1B). The C1-products are much more prevalent than the C2-products over the VOx/γ-Al2O3 catalyst, which is typical over the V-based catalysts due to the inevitable over-oxidation of alkanes (C numbers > 1) or olefins in ODHP (17). However, it is interesting that the C2-products (C2H4 and C2H6 shown in fig. S1) are much more prevalent than the C1-products (C2/C1 > 2) over the h-BN catalyst (Fig. 2). This is quite different from that over the VOx/γ-Al2O3 catalyst (6, 9), which showed a C2/C1 ratio of close to 0 (Fig. 2). The observation of a higher C2/C1 ratio (>2) is a general feature for the ODHP over the h-BN and boron-based catalysts (7, 9) at low reaction temperatures, yet its origin remains unaddressed.

Fig. 1 The influence of reaction temperature on catalytic performance of ODHP.

(A) h-BN catalyst; (B) VOx/γ-Al2O3 catalyst.

Fig. 2 Comparison of the C2/C1 ratio on the VOx/γ-Al2O3 and h-BN catalysts at different reaction temperatures.

Because of the complexity of C1-products in ODHP, the C1-products should be carefully examined to draw a definitive conclusion. To exclude the potential contribution of coke formation on the catalyst to the loss of carbon, both the fresh and spent h-BN catalysts were studied using Raman spectroscopy. Except the B-N stretching vibration in h-BN (~1367 cm−1) (18), no C-related peaks (1340 and 1604 cm−1) were found on the spent h-BN catalyst (fig. S2) (19, 20), indicating that there is no coke deposition on the spent h-BN catalyst.

COx and CH4 are the only C1-products detected on the V-based catalysts in ODHP, consistent with the reports elsewhere (2123). On the h-BN catalysts, only COx and CH4 have been reported as the C1-products so far (8, 9, 2426). However, because of the higher tendency toward only the first C─C bond cleavage of propane over the h-BN catalysts compared to the V-based catalysts (10), methyl could thus be produced and the C1-products from the further conversion of methyl via oxidation, such as methanol and formaldehyde (CHxOy), could also be present. To confirm this, mass spectra signals of the products in ODHP over these two catalysts were collected (fig. S3). When the reaction temperature increases from room temperature to 600°C, CHxOy products (methanol and formaldehyde) gradually form and reach a maximum at 600°C over the h-BN catalyst. However, CHxOy products are negligible on the VOx/γ-Al2O3 catalyst over the same range of temperatures studied. The formation of CHxOy on the h-BN catalyst was quantified and summarized in Table 1. Even considering CHxOy in the total C1-products, the C2/C1* ratio is still greater than 2. First, the C─C bond cleavage of propane is expected to form equal molar C1- and C2-products. Assuming that C2-products are not further converted to C1-products (by either continuous C─C bond cleavage or oxidation), the C2/C1 ratio should be at most 2. The fact that the C2/C1* ratio is greater than 2 on the h-BN catalyst (Table 1) suggests that there are other origins for the C2-products formation besides the cleavage of C─C bond in ODHP over the h-BN catalyst.

Table 1 The conversion and selectivity over the h-BN catalyst for ODHP at different reaction temperatures.

C3H8/O2/He/N2 = 1:1:1:8; total flow rate, 33 ml/min. C1* indicates the total C1-products including CHxOy, COx, and CH4. C3 indicates the selectivity to C3H6. C2 indicates the selectivity to C2H4 and C2H6.

View this table:

Ethylene is the main C2 by-product in ODHP over the h-BN catalyst, which is generally considered to originate from the C─C bond cleavage of propane (10). C2-products could also be formed in the presence of methyl via OCM as previously reported on the Mn2O3-Na2WO4/SiO2 catalyst through the methyl oxidative reaction pathway (15). To investigate the possible involvement of OCM in the formation of C2-products on the h-BN catalyst, OCM experiments were conducted over the h-BN and VOx/γ-Al2O3 catalysts at 600°C using a feed gas of V(CH4)/V(O2)/V(N2) = 1/1/1 (fig. S4). As shown in Fig. 3A, only COx products were detected over the VOx/γ-Al2O3 catalyst at a methane conversion of 27.9%. However, both ethylene and ethane (almost 1:1) were also detected in addition to COx products over the h-BN catalyst, and C2-selectivities of 9% were reached at a methane conversion of 25.8% (Fig. 3A), which is similar to the recent report by Wang et al. (27). CHxOy also formed, with a selectivity of 2.1%. Apparently, under the same reaction conditions and at a similar methane conversion, the h-BN catalyst has noticeable OCM activity, while the VOx/γ-Al2O3 catalyst has negligible OCM activity. Noting the similarity of the reaction pathways between OCM and oxidation of methyl, we can infer that the oxidative coupling of methyl to C2-products is involved in ODHP over the h-BN catalyst, which results in a higher C2-product selectivities. Experiments using 13C-labeled propane also suggest that oxidative coupling of methyl is involved in ODHP on h-BN (fig. S5 and table S2). Collectively, the higher C2H4 selectivity and the C2/C1* ratio (>2) on the h-BN catalyst are due to both the C─C bond cleavage of propane to form ethyl followed by an additional C-H cleavage and the oxidative coupling of methyl on this catalyst. When we deducted C2H4 from those formed from the cleavage of C3H8, the selectivities to C2H4 and C2H6 derived from the methyl coupling reaction were similar (Fig. 3B).

Fig. 3 The product distribution of OCM and ODHP over the VOx/γ-Al2O3 and h-BN catalysts.

(A) Conversion and selectivity of OCM over the VOx/γ-Al2O3 and h-BN catalysts. (B) Selectivity to C2-products (C2H4 and C2H6) derived from the methyl coupling reaction in ODHP over the h-BN catalyst at different reaction temperatures.

To estimate the contribution of the oxidative coupling of methyl to the higher C2-product selectivities in ODHP, we first attempted to understand the origins of COx (main products of C1). COx may come from the oxidation of alkanes, olefins, and methyl in ODHP. Therefore, the oxidation reactions of propene, which is more reactive than ethylene, over the h-BN and VOx/γ-Al2O3 catalysts were carried out. As shown in table S1, over-oxidation of propene readily occurs (about 99% COx selectivity) on the VOx/γ-Al2O3 catalyst. However, there is only less than 1% conversion of propene over the h-BN catalyst even at a reaction temperature of 600°C. Because propene is known to be more reactive than ethylene, the fact that there is no over-oxidation of propene suggests that there is no direct over-oxidation of olefins over the h-BN catalyst in ODHP.

However, it is still uncertain whether the over-oxidation of alkanes (C3H8 and C2H6) occurs over the h-BN catalysts until now. Even so, we can conclude that C2/C1 will be less than 2 if the over-oxidation of C2+(olefins/alkanes) does occur. Therefore, it is expected that C2/C1 will be equal or less than 2 in ODHP. When there is no over-oxidation of alkanes or olefins, the amount of C1-products should be equal to that of C2-products, assuming that there is no oxidative coupling of methyl occurring over the h-BN catalyst. In this case, the theoretical value of C1- or C2-selectivities (only cleavage of the first C─C bond) is defined using Eqs. 1 and 2 as belowC1(theoretical value)=(100C3)/3(1)C2(theoretical value)=(100C3)*(2/3)(2)where C1 or C2 represents the total selectivities of one- or two-carbon products, respectively, and C3 presents the selectivity of C3H6. The theoretical values of C1- and C2-selectivities as a function of temperature are shown in Fig. 4A, which are higher and lower than the measured C1- and C2-selectivities, respectively. Therefore, we can assume that the contribution of oxidative coupling of methyl is at least equal to the equation as belowC2(coupling reaction)=(C22*C1*)/3(3)

Fig. 4 The quantitative analysis of product from methyl reactions in ODHP over the h-BN catalyst.

(A) Theoretical and measured selectivities of total C1- and C2-products. (B) Product distributions from methyl reactions in ODHP over the h-BN catalyst at different reaction temperatures.

Therefore, the product distributions derived from methyl in ODHP over the h-BN catalyst at different reaction temperatures are analyzed and shown in Fig. 4B. These products are likely due to the following reaction pathways of methyl in ODHP over the h-BN catalyst: (i) CH4 formation from methyl hydrogenation, (ii) oxidation of methyl to form COx or CHxOy, and (iii) oxidative coupling of methyl to produce the C2-products. The detailed selectivities of C2-products (C2H4 and C2H6) from the methyl coupling reaction are shown in Fig. 3B. The selectivities to C2H4 and C2H6 derived from the methyl coupling reaction in ODHP are almost similar to those derived from the methane coupling reaction (27), indicating that the C2H4 formation is via the CH3* coupling reaction to form C2H6 followed by subsequent dehydrogenation to C2H4. From Fig. 4B, we can see that high temperature favors the formation of COx, CH4, and CHxOy but disfavors C2-products from oxidative coupling (figs. S6 and S7), which is consistent with previous reports (2830). When the reaction temperature further increases to higher than 570°C, the C2/C1 ratio is less than 2, which is possibly due to the inevitable over-oxidation of alkanes or olefins (fig. S8).

On the basis of the above results and discussions, the possible reaction pathways of ODHP over the h-BN catalyst are proposed and shown in Fig. 5. Both C─H and first C─C bonds of propane are cleaved to form CH3CH2*CH2, CH3CH2*, and CH3* radicals over the h-BN surface. The CH3CH2*CH2 radical (6) will continue dehydrogenation to form C3H6 and may also cleave to form CH3CH2* radicals. Meanwhile, C2H4 is generated from CH3CH2* as a main pathway to form the C2-products (mainly from the cleavage of propane). The methyl formed from the first C─C bond cleavage of propane could proceed via four reaction pathways: (i) oxidative coupling to form C2-products, (ii) hydrogenation to form CH4, (iii) selective oxidation to form CHxOy, and (iv) over-oxidation to form COx. Among these reaction pathways, oxidative coupling of methyl makes additional contribution to C2-product selectivity in ODHP, leading to more C2-products than C1-products over the h-BN catalyst.

Fig. 5 The proposed reaction pathways of ODHP over the h-BN catalyst [using a structure model by Shi et al. (31) as an example].

In summary, we report the understanding of reaction pathways in ODHP over the h-BN catalyst, which involve methyl generated from the C─C bond cleavage. Compared with the VOx/γ-Al2O3 catalyst, there are four secondary reaction pathways involving methyl, i.e., methyl hydrogenation to form CH4, methyl oxidation to produce COx and CHxOy, and oxidative coupling of methyl to produce C2-products. Besides the C─C bond cleavage of propane, part of C2-products is formed from the oxidative coupling of methyl in ODHP, which results in a higher C2/C1 ratio (>2) over the h-BN catalyst. This work not only provides further insight into the reaction mechanism of ODHP over the boron-based catalysts but also may broaden its application to other related upgrading reactions on light alkanes, such as selective oxidation of methane to ethylene or CHxOy.

MATERIALS AND METHODS

Materials

h-BN was commercially available from Qinhuangdao Eno High-Tech Material Development Co. Ltd., with a purity of >99%. The sample was used as provided without further chemical or thermal treatment. The h-BN catalysts before and after the reaction were named as h-BN fresh and h-BN spent, respectively.

As a comparison, 5 weight % (wt %) VOx/γ-Al2O3 catalyst was prepared by wet impregnation (23). One gram of γ-Al2O3 support was impregnated with 5.7 ml of 0.1 M NH4VO3 solution, and the resultant slurry was stirred at room temperature for 12 hours. After the impregnation, the catalyst was dried at 90°C overnight and then calcined at 600°C for 4 hours. VOx/γ-Al2O3 (5 wt %) was denoted as VOx/γ-Al2O3.

Catalytic activity measurements

The VOx/γ-Al2O3 and h-BN (80 to 100 mesh) catalysts were tested in a fixed-bed reactor with an inner diameter of 6 mm under atmospheric pressure. The flow rates of pure propane (or propene, methane) and mixture gas (O2:N2:He = 1:8:1) were controlled using two mass flow controllers (Sevenstar) and calibrated to each individual gas to allow a total flow rate of 33 ml/min (C3H8:O2:N2:He = 1:1:8:1). The reaction temperatures were set from 510° to 600°C. The reaction products were analyzed using a gas chromatograph (GC2060, Shanghai Ruimin GC Instruments Inc.) with a flame ionization detector [KB-Al2O3/Na2SO4 column for hydrocarbon (CxHy) measurement] and two thermal conductivity detectors (equipped with TDX-01 column for CO and CO2 measurements and Porapak Q column for methanol and formaldehyde measurements for OCM). Experiments with quartz showed that the reaction conversion was negligible without the catalyst. The carbon balance was checked by comparing the number of moles of carbon in the outlet stream to the number of moles of carbon in the feed. Under the typical evaluating conditions, the carbon balance was within 100 ± 4%. To account for the volume expansion in the reaction, helium was used as the internal standard. C3H8 conversion (X) and product selectivity (S) were calculated using the standard normalization method based on carbon atom balance and were defined as follows.Cbalance=(NiAifi)in(NiAifi)outXsi(%)=NpiApifpiNsiAsifsi+NpiApifpi*100Spj(%)=NpiApifpiNpiApifpi*100GHSV(h1)=Ftotal/Vcatwhere s or p represents the reactant or product; i represents random gas in this reaction system; Ai and fi are the chromatographic peak area and the calibration factor, respectively, of species i; Vcat is the volume of the catalyst loaded in the reactor (ml); Ftotal is the total flow of all inlet gas (ml/min); and N is the number of carbon atoms of the component.

Characterizations

A thermal gravimetric analyzer (TG 209F1, Netzsch, Germany) was used to investigate the carbon deposition of the spent catalysts. The samples were preheated at 80°C and then heated to 970°C at a rate of 10°C/min in air (20 ml/min). Raman spectra were obtained on a Renishaw UV-vis RT 1000 Raman spectrometer with an excitation wavelength of 532 nm and a resolution of 6 cm−1. The sample was pressed into a sheet with a thickness of 2 mm and observed in the Renishaw using a laser scanning confocal microscope.

The reaction products were also identified and quantified by an online mass spectrometer (Pfeiffer OmniStar GSD 320). The m/z (mass/charge ratio) signals of CO, CO2, CH3OH, and CH2O were 28, 44, 31, and 30, respectively. The ODHP at reaction temperatures of 600°C was carried out, and the products were pumped into the mass spectrometer with m/z from 1 to 100.

13C isotopic labeling experiments were performed in a fixed-bed single-pass flow reactor equipped with online mass spectrometry (MS). The details are described as below: A mixture of He and O2 with 9:1 (O2: 1 ml/min) passed the h-BN catalyst at 560°C until the baseline reached flat and then the 13C-labeled propane (Cambridge Isotope Laboratories; 99%; 1 ml each time) was introduced into the flow using an injection syringe at specified temperatures (560°, 540°, 520°, and 510°C). The chemical and isotopic compositions of the reactor effluent were measured by online MS at intervals of 6 s with m/z from 24 to 32.

SUPPLEMENTARY MATERIALS

Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/3/eaav8063/DC1

Supplementary Text

Fig. S1. Product selectivities to C1 (COx and CH4) and C2 (C2H4 and C2H6) in ODHP over the h-BN catalyst at different reaction temperatures.

Fig. S2. Characterizations of the fresh and spent h-BN catalysts.

Fig. S3. Mass spectral signals of the products in ODHP at different reaction temperatures.

Fig. S4. Schematic diagram of OCM process.

Fig. S5. Mass spectral signals of m/z in ODHP with isotopic pulsing experiments over h-BN.

Fig. S6. Distribution of partial products in ODHP over the h-BN catalyst at different reaction temperatures.

Fig. S7. Reaction pathways of methyl over the h-BN catalyst at different reaction temperatures.

Fig. S8. Comparison of C2/C1 ratio on the VOx/γ-Al2O3 and h-BN catalysts at different reaction temperatures.

Table S1. Conversion and selectivity of propene oxidation over the VOx/γ-Al2O3 and h-BN catalysts.

Table S2. 13C-labeled propane and possible products from ODHP.

Reference (32)

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial license, which permits use, distribution, and reproduction in any medium, so long as the resultant use is not for commercial advantage and provided the original work is properly cited.

REFERENCES AND NOTES

Acknowledgments: We thank F. Lin and J. Zheng (Pacific Northwest National Laboratory, USA) for assistance on MS measurements. Funding: This work was supported by the National Natural Science Foundation of China (grants 21576227, 91545114, 91545203, and 21673189) and the Fundamental Research Funds for the Central Universities (grant 20720160032). Author contributions: J.L. and Y.W. conceived the idea for the project. J. Ti. and M.X. conducted the material synthesis. J. Ti. and J. Ta. performed the structural characterizations and catalytic test. J.L., Y.W., and J. Ti. discussed the catalytic results. J. Ti. drafted the manuscript under the guidance of Y.W. All authors discussed and commented on the manuscript. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. Additional data related to this paper may be requested from the authors.
View Abstract

Navigate This Article