- Email: [email protected]
Grignard reagent reduced nanocarbon material in oxidative dehydrogenation of n-butane
Journal of Catalysis 360 (2018) 51–56
Contents lists available at ScienceDirect
Journal of Catalysis journal homepage: www.elsevier.com/locate/jcat
Grignard reagent reduced nanocarbon material in oxidative dehydrogenation of n-butane Jiaquan Li a, Peng Yu a, Jingxin Xie a, Yajie Zhang b, Hongyang Liu b,⇑, Dangsheng Su c, Junfeng Rong a,⇑ a
Research Institute of Petroleum Processing, Sinopec, Beijing 100083, China Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, China c Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China b
a r t i c l e
i n f o
Article history: Received 14 December 2017 Revised 18 January 2018 Accepted 19 January 2018
Keywords: Carbon nanotubes Grignard reagent Electrophilic oxygen Reduction Oxidative dehydrogenation
a b s t r a c t A modification route of nanocarbon catalyst based on Grignard reagent reduction of oxidized carbon nanotubes (o-CNTs) has been developed for oxidative dehydrogenation (ODH) of n-butane. The o-CNTs contain considerable amount of electrophilic oxygen species which are responsible for deep oxidation side-reactions and the alkene selectivity in ODH is low. After Grignard reduction, the corresponding electrophilic oxygen groups on the surface of the catalyst were eliminated and the basicity increased. As a result, the side-reactions in ODH were prohibited and the alkene selectivity was significantly improved compared with o-CNTs. The chlorine containing Mg/Cl species were found to have positive effect on the improvement of C4H8 alkene yield. This study provides a method of the preparation of nanocarbon catalyst to achieve higher alkene selectivity for the dehydrogenation reaction. Ó 2018 Elsevier Inc. All rights reserved.
1. Introduction Oxidative dehydrogenation (ODH) of n-butane is a promising process to produce high value butenes and butadiene from lowcost butane. Traditional metal oxides such as supported vanadium and molybdenum oxide catalysts have been widely employed to catalyze this reaction [1,2]. Among the reported metal oxide catalysts, the VAMgAO catalyst presents the highest selectivity to C4 alkenes [1–6]. However, the traditional metal oxide catalysts are unfavourable in side-reaction controlling and suffer serious deactivation by the coke formation during ODH reaction, which restrains their further application. Nanocarbon materials are attracting considerable interest in catalyzing CAH bond activation such as ODH of light alkanes, owing to their competitive catalytic activity and remarkable coke-resistance [7,8]. It has been proven that the electron-rich quinone groups on the surface of nanocarbon are active sites to catalyze ODH reactions through a redox mechanism [8–10], and the electrophilic oxygen groups (peroxide O2 and superoxide O 2 2) attack C@C bonds and are responsible for deep oxidation of the desired alkene products [8]. A variety of heteroatom-doping treatments have been explored on nanocarbons to enhance the catalytic performance focusing on O, N, B and P doping [8,11–13], however,
⇑ Corresponding authors. E-mail addresses: [email protected] (H. Liu), [email protected] (J. Rong). https://doi.org/10.1016/j.jcat.2018.01.021 0021-9517/Ó 2018 Elsevier Inc. All rights reserved.
the employed modification approaches for ODH reaction are still limited and the catalytic influence of other elements has not been investigated. Moreover, further industrial application of nanocarbon catalysts in ODH of n-butane requires higher alkene selectivity which is a priority in developing new modification strategies. Grignard reagent is a reductant presented as RMgX (R is alkyl or aryl and X is halogen). Reduction of graphene oxide with Grignard reagent has been reported to form functional nanocomposites [14,15], but the impact of Grignard reagent reduction on the catalytic performance of nanocarbon materials hasn’t been discussed. Herein, a modification route of nanocarbon materials was developed based on Grignard reagent reduction of oxidized carbon nanotubes (o-CNTs) to control the distribution of the surface oxygen groups and to introduce heteroatoms on carbon nanotubes (CNTs) that have potential effects on the catalyst. We employed C6H5CH2MgCl as the reductant and Cl/Mg precursor to reduce the electrophilic oxygen species on the surface of o-CNT which cause deep oxidation side-reactions, and examined the effect of Cl/Mg dopant on the catalytic behavior in ODH of n-butane. The C4 alkene selectivity of Grignard reduced o-CNT catalyst (Gr-oCNT) was significantly improved, providing new insights into heteroatomdoped carbon-mediated catalyst. The catalytic roles of electrophilic oxygen, Mg and Cl species in ODH reactions were systematically discussed in this work.
52
J. Li et al. / Journal of Catalysis 360 (2018) 51–56
2. Methodology
titration, respectively. The obtained iodometric titration data were validated with similar results from repeated experiments.
2.1. Catalysts preparation
O22 - þ 2KI þ 2H2 SO4 ! O2 þ 2KHSO4 þ I2 þ H2 O
ð1Þ
I2 þ 2Na2 S2 O3 ! 2NaI þ Na2 S4 O6
ð2Þ
cðelectrophilic oxygenÞ ¼ 2 106 V=m
ð3Þ
Commercial available CNTs were purchased from Shandong Dazhan Nano Materials Co., Ltd., and used directly without any purification. The length, inner diameter and purity of CNTs were: 3–12 mm, 12–15 nm and 96%, respectively. 10 g CNTs was oxidized in 125 mL of concentrated HNO3 (65–68%) and 375 mL of concentrated H2SO4 (95–98%) under sonication at 50 °C for 6 h, and the precipitate was filtered out and washed several times with deionized water until the pH of the filtrate reached 7. The precipitate was dried in the atmosphere at 120 °C for 12 h to obtain oxidized carbon nanotubes (o-CNTs). The Grignard reduction of o-CNT was performed as follows. 2 g o-CNTs was added in 20 mL of C6H5CH2MgCl (1.0 M, solution in diethyl ether, Aldrich Chemical Company). After the solution had been stirred under N2 protection for 2 h at 25 °C, the temperature was increased to 65 °C and held for 12 h. Excess C6H5CH2MgCl was washed off by tetrahydrofuran (THF) for several times and the precipitate was filtered out and dried in the atmosphere at 80 °C for 12 h to obtain Gr-oCNT. The samples involved in control experiments include Mg-oCNT, Cl-oCNT, Gr-CNT and Me-oCNT. The preparation procedure was as follows: Mg-oCNT: 2 g o-CNT was added into 20 mL of Dibutylmagnesium (1.0 M, solution in heptane, Aldrich Chemical Company). After the solution had been stirred under N2 protection for 2 h at 25 °C, the temperature was increased to 65 °C and held for 12 h. Excess Dibutylmagnesium was washed off by tetrahydrofuran (THF) for several times and the precipitate was filtered out and dried in the atmosphere at 80 °C for 12 h to obtain Mg-oCNT. Cl-oCNT: 1 g o-CNTs was added into 20 mL of CCl4. After the solution had been stirred under N2 protection for 2 h at 25 °C, the temperature was increased to 80 °C and held for 12 h. The precipitate was filtered out and dried in N2 at 80 °C for 12 h to obtain Cl-oCNT. Gr-CNT was obtained following the same procedure as Gr-oCNT except that o-CNTs was replaced by pristine CNTs. Me-oCNT was obtained following the same procedure as GroCNT except that C6H5CH2MgCl was replaced by CH3MgCl (1.0 M, solution in diethyl ether, J&K Scientific Ltd).
The titration results were validated by the control experiments where no electrophilic oxygen species were detected on pristine CNTs and blank sample (without CNTs). 2.3. Catalyst performance test The catalytic activity of CNT samples in the reaction of oxidative dehydrogenation of butane was measured at 723 K using a 10 mm diameter fixed-bed quartz tube reactor at atmospheric pressure over 300 min. Reactant and product concentrations (weight percent) were measured by online gas chromatography (Agilent model 7890B) equipped with a HayeSep Q column, a HayeSep N column and a molecular-sieve column connected to a thermal conductivity detector, and a HP-PLOT A12O3 column (50 m 0.53 mm 15 lm) connected to a flame ionization detector. Reactant mixtures (Beijing AP BAIF Gases) contained 0.7 wt% butane, 1.4 wt% O2 and 97.9 wt% N2 with a typical gas flow rate of 4500 mL of gas h1 (g of catalyst)1. The selectivity of alkene products is calculated as follows:
selbutene ¼
cbutene 1:034 100% ½cbutane in ½cbutane out
selbutadiene ¼
cbutadiene 1:074 100% ½cbutane in ½cbutane out
where seli (percent) and ci (weight percent) are selectivity and concentration, respectively, of each C4 alkene product in the outlet gas and [cbutane]in and [cbutane]out are the n-butane concentrations in the inlet and outlet gas mixture, respectively. The side-reaction refers to total oxidation of hydrocarbon to form COx. The calculations of side-reaction rates (moles per gram per second) are based on the reaction conditions (gas flow rate and butane concentration) and the catalytic performance. The calculating formula for the side-reaction rate is presented as
v s ¼ 0:189 106 convð1 selÞ 2.2. Iodometric titration method In this work, we used iodometric titration to determine the surface amount of electrophilic oxygen species on CNTs samples [16]. Electrophilic peroxide and superoxide species can oxidize aqueous I— into I2 for subsequent titration with Na2S2O3. Herein, we suppose the electrophilic oxygen species are mostly composed of peroxides and designate peroxide as electrophilic oxygen for convenient calculation. The titration of electrophilic oxygen groups on CNTs samples was performed as following procedure. A 0.3 g CNT sample was added in a KI solution that consisted of 10 mL of KI (100 g/L), 5 mL of H2SO4 (0.1 mol/L), 30 mL of deionized water, and 3 drops of (NH4)6Mo7O24 (30 g/L). The reaction between peroxides on the surface of CNTs and KI is shown in (1). After 30 min under sonication at 25 °C in the dark, I— was oxidized into I2. Then the precipitate was filtered out and washed six times. I2 in the filtrate was titrated with Na2S2O3 (0.002 mol/L) as presented in (2) and the concentration of electrophilic oxygen (mol/g) on CNTs was calculated with (3) where c (mol/g), V (mL) and m (g) represent the electrophilic oxygen concentration, the volume consumption of Na2S2O3 solution and the mass of CNT samples for
where ms (moles per gram per second), conv (percent) and sel (percent) represent the side-reaction rate, the conversion of butane and the C4 alkene selectivity, respectively. A control experiment was conducted under the same reaction condition except there was no catalyst sample added into the reactor. The conversion rate of n-butane was 0%, and neither alkene nor the side reaction product was detected by gas chromatography. 2.4. Characterizations Transmission electron microscopy (TEM) measurements were taken on a FEI Tecnai F20 microscope with an accelerating voltage of 200 kV. Raman spectra were recorded under ambient conditions on a JY LabRAM HR Raman spectrometer with a 325 nm laser beam. The weight concentrations for element C, H, and O were determined on an Elementar Micro Cube elemental analyzer. The weight concentrations for element Mg and Cl were obtained by Rigaku 3013 X-ray Fluorescence Spectrometer (XRF). X-ray photoelectron spectroscopy (XPS) data were obtained with an ESCALab220i-XL electron spectrometer from VG Scientific using 300 W Al Ka radiation. The base pressure was 3 109 mbar.
J. Li et al. / Journal of Catalysis 360 (2018) 51–56
The binding energies were referenced to the C 1s line at 284.6 eV from defect free graphite. Deconvolution of O 1s spectra were performed using mixed Gaussian-Lorentzian component profiles after subtraction of a Shirley background using XPSPEAK41 software. IR spectra was conducted on a Thermo Nicolet iS10 FTIR system. The XRD patterns were obtained on a XRD-6000 with a Cu Ka source at a scan rate of 10° min1.
3. Results and discussion The oxidation process of pristine CNTs in HNO3/H2SO4 (1:3) mixture creates reactive defects for oxygen functional groups to anchor. Then the obtained o-CNT reacted with C6H5CH2MgCl solution to form Gr-oCNT. The Mg/Cl species were incorporated in the catalytic system to form CNT–O–MgCl structure via reactions between C6H5CH2MgCl and oxygen groups on o-CNT such as carboxylic acids and phenols [15]. The morphological and structural properties of o-CNT and Gr-oCNT characterized by transmission electron microscopy (TEM) were shown in Fig. 1A and B suggesting that the reduction treatment has no apparent impact on the structure of o-CNT and no MgO nanoparticles were found on Gr-oCNT. The results were supported by the high-angle annular dark field scanning transmission electron microscopy (STEM) image in Fig. 1C and elemental mapping of a typical area of Gr-oCNT in Fig. S1 (see Supporting Information). Both O and Mg species were well dispersed on Gr-oCNT without visible aggregation of MgO nanoparticles. Furthermore, the powder X-ray diffraction of GroCNT exhibited two peaks attributed to CNTs framework and no peak of crystalline cubic structure of MgO nanoparticles was
53
observed as shown in Fig. 1D, consisting well with the TEM results [17,18]. According to Raman spectroscopy (see Fig. S2 in Supporting Information), The value of ID/IG [19,20] increased from 0.13 to 0.20 after oxidation of pristine CNTs because of the generation of defects by HNO3/H2SO4 etching, while the ID/IG value of Gr-oCNT was 0.19, close to that of o-CNT. Therefore, it can be inferred that the carbon framework was well maintained after Grignard reduction treatment and the Mg species are highly dispersed on o-CNTs. The total concentrations of O, Mg and Cl obtained from element analysis and X-ray fluorescence (XRF) and the surface concentrations of the elements from X-ray photoelectron spectroscopy (XPS) are summarized in Table S1 (see Supporting Information). A comparison of the catalytic performances of pristine CNTs, oCNT and Gr-oCNT in ODH of n-butane executed at 450 °C with an O2/butane ratio of 2.0 is displayed in Fig. 2A. The oxygen content of pristine CNTs is negligible, while after oxidation the total oxygen concentration increases to 13 wt%. As a result, there is a significant increase of butane conversion (from 6.8% to 28.3%) and C4 alkene selectivity (from 7.9% to 23.4%). After the reduction treatment with C6H5CH2MgCl, the alkene selectivity is enhanced to 59% which is among the highest selectivity that has been reported so far on nanocarbon catalyst [8,13,17], and the conversion decreased to 17%. Notably, butadiene selectivity increases from 13% to 21% after Grignard reduction, while the increase of total selectivity of C4H8 (1-butene, cis-butene and trans-butene) was more significant (from 10% to 38%), indicating that Gr-oCNT tends to facilitate the formation of C4H8 butenes instead of butadiene in catalyzing ODH of n-butane. The electrophilic oxygen species (peroxides and superoxides) on o-CNTs account for the deep oxidation of alkene products and
Fig. 1. TEM images of o-CNT (A) and Gr-oCNT (B). (C) HAADF-STEM image of Gr-oCNT. (D) Powder X-ray diffraction patterns of o-CNT and Gr-oCNT.
54
J. Li et al. / Journal of Catalysis 360 (2018) 51–56
Fig. 2. (A) ODH catalytic performance of pristine CNT, o-CNT and Gr-oCNT. (B) Amount of electrophilic oxygen determined by iodometric titration of o-CNT and Gr-oCNT before and after ODH reaction and the relevant side reaction rates. The ODH reaction condition: 723 K, 1 atm, 2:1 O2:butane ratio, 0.2 g catalyst.
carbon skeleton [8,21], resulting in the reduction of C4 alkene selectivity in ODH reactions. In this sense, the promotion of alkene selectivity can be achieved by reduction of oxygen species of high oxidation states to eliminate the electrophilic oxygen on the surface of catalysts. In this work, iodometric titration was utilized to quantitatively determine the amount of surface electrophilic oxygen species on the catalysts. The amount of the electrophilic oxygen on o-CNT and Gr-oCNT before and after ODH of n-butane combined with the corresponding side-reaction rates is presented in Fig. 2B. Compared with o-CNT, the amount of electrophilic oxygen on Gr-oCNT is extremely low, indicating that Grignard reduction treatment eliminates most of the electrophilic oxygen species on o-CNT. After ODH reaction, over half of the electrophilic oxygen groups are removed from the surface of o-CNT, while GroCNT exhibits apparent increase of the amount of electrophilic oxygen which is possibly due to the regeneration of electrophilic oxygen during ODH in the presence of oxygen. The amount of electrophilic oxygen on Gr-oCNT after ODH is approximately half of that of o-CNT, which is in correspondence with the decreased side-reaction rate after Grignard reduction, showing evidence that the electrophilic oxygen species are sites for combustion sidereactions and explains the enhanced selectivity owing to Grignard reduction treatment. In addition, it has been reported that the alkene products of ODH reaction are weekly adsorbed on basic catalysts [3,4,17]. The Grignard reagent is strong Lewis base and reacts with acidic oxygen groups on o-CNT, thereby improving the basicity of o-CNT and facilitating the desorption of alkenes from CNT’s surface. In this sense, higher basicity of the catalyst due to Grignard reduction also attributes to the increased alkene selectivity in this catalytic reaction.
The influence of the reduction treatment on the composition of surface oxygen groups of o-CNTs was analyzed by IR spectra. IR spectra of o-CNT in Fig. 3 demonstrate three signals associated with the typical oxygen containing groups, namely C@O stretching in carboxylic acid and lactone groups at 1696 cm1 [22,23], CAO stretching at 1015 cm1 and CAOH stretching at 1200 cm1 [15]. After Grignard reduction, both C@O and CAOH stretching vibration signals disappear, indicating that C6H5CH2MgCl reacts with CAOH and O@CAO groups on oACNT to form ACAOAMgCl structure during the reduction treatment. The catalytic role of Mg/Cl species in ODH reactions when it is attached on the catalyst was investigated by replacing oACNT with silica gel (SiO2) which contains abundant AOH groups to anchor AMgCl and then conducting Grignard reduction experiment following the same procedure. Catalytic performances of SiO2 and Grignard reduced SiO2 (Gr-SiO2) in ODH of n-butane were presented in Fig. S3 (see Supporting Information). Although both SiO2 and Gr-SiO2 exhibited low conversion of n-butane, a decrease in butane conversion (from 4% to 2%) and a significant increase in alkene selectivity (from 32% to 89%) were observed when Mg/Cl species were anchored on SiO2, which was in correspondence with the results of Gr-oCNTs. In addition, no butadiene was detected in the product gas and the produced alkenes consist entirely of C4H8 alkenes when SiO2 and Gr-SiO2 were employed as catalysts in ODH. Considering that butadiene is primarily generated from second ODH of C4H8 alkenes following similar dehydrogenation pathway to C4H8 formation [8], we can infer that the impact of the formed Mg/Cl species on ODH of butane is more likely to enhance the production of 1-butene, cis-butene and trans-butene than second ODH to produce butadiene, which is also in accordance with former results of catalytic performance of Gr-oCNT. The control experiments were carried out to testify the role of each individual component on Gr-oCNT that may affect the catalytic property. The catalytic performances of a series of CNT samples obtained from different modification procedure are presented in Fig. 4. Oxygen functional groups, Mg and Cl species are considered as three components on Gr-oCNT, with at least one of them absent for each sample to obtain o-CNT, Mg-oCNT, Cl-oCNT, GrCNT and using CH3MgCl as another kind of Grignard reagent to obtain Me-oCNT in comparison with Gr-oCNT. Dibutylmagnesium and CCl4 were utilized as Mg and Cl precursors to form Mg-oCNT and Cl-oCNT samples. Gr-oCNT exhibits a superior alkene selectivity and a moderate butane conversion among all the samples involved. Notably, dibutylmagnesium (C4H9AMgAC4H9) is also a strong reductant that has similar functionality as Grignard reagent
Fig. 3. ATR-IR of pristine CNTs, o-CNT, Gr-oCNT and Gr-oCNT after ODH of butane.
J. Li et al. / Journal of Catalysis 360 (2018) 51–56
Fig. 4. Catalytic performances of o-CNT, o-CNT refluxed with dibutylmagnesium (Mg-oCNT) and CCl4 (Cl-oCNT), CNTs refluxed with C6H5CH2MgCl (Gr-CNT), o-CNT refluxed with C6H5CH2MgCl (Gr-oCNT) and CH3MgCl (Me-oCNT) in ODH of butane. The ODH reaction condition: 723 K, 1 atm, 2:1 O2:butane ratio, 0.2 g catalyst.
to reduce the electrophilic oxygen groups on o-CNT to obtain Mg-oCNT with Mg dopant. XPS results shows that the surface Mg atom concentration of Mg-oCNT is relatively high as 10.3%. According to the iodometric titration results (see Fig. S4 in Supporting Information), the dibutylmagnesium treatment dramatically reduced the amount of the electrophilic oxygen on o-CNT and the concentration of electrophilic oxygen on Mg-oCNT after ODH reaction is even lower than that of Gr-oCNT. However, the total alkene selectivity of Mg-oCNT is merely 26%, much lower than that
55
of Gr-oCNT (59%) and slightly higher than o-CNT (23.4%). It was reported that for the supported vanadium oxide catalyst system, the isolated Mg2+ sites of the support are responsible for deep oxidation of alkenes [24,25]. For nanocarbon catalysts, the presence of metal, such as Fe and Mo, also has no positive contribution to the improvement of alkene selectivity [8,26]. Therefore, we can infer that the sole presence of Mg species on catalysts has little positive effect on the catalytic performance. By refluxing o-CNT in CCl4, a low Cl loading of 0.21% (XPS results) was achieved to give Cl-oCNT and a slight promotion of alkene selectivity of 5% was obtained. Iodometric titration results and deconvolution of O1s XPS show no obvious change of the amount of electrophilic oxygen and C@O concentration (see Figs. S4 and S5 in Supporting Information) on o-CNT after the CCl4 treatment. We infer that the loaded Cl is beneficial to the selectivity improvement. Me-oCNT showed similar catalytic performance to that of Gr-oCNT with a total selectivity of 50%, indicating the indispensable role of Grignard reagent modification. The samples of the control experiments provide a hint that the enhanced reaction selectivity on Gr-oCNT results from joint effects of the reduction of electrophilic oxygen groups and the promoted basicity together with the loading of Cl containing species. On the basis of the above discussion, the overall influence of Grignard reduction treatment on the structure and catalytic performance of o-CNT is depicted in Fig. 5. The Grignard reduction eliminates the undesired electrophilic oxygen species and increases the basicity of the catalyst. The formed Mg/Cl species plays a positive role in catalyzing ODH reactions. As a result, the total alkene selectivity in ODH of n-butane is promoted after the modification treatment.
Fig. 5. Structural and catalytic influence of Grignard reduction on o-CNT in ODH of n-butane.
56
J. Li et al. / Journal of Catalysis 360 (2018) 51–56
4. Conclusions In summary, a modification approach based on Grignard reagent reduction of o-CNT was developed to improve C4 alkene selectivity in ODH of n-butane and provides new insights of heteroatom-doped nanocarbon catalysts. Iodometric titration method was employed in this work to quantitatively determine the amount of electrophilic oxygen species on the surface of CNTs samples. The Grignard reduction of o-CNTs prevents the sidereactions by eliminating the electrophilic oxygen species that are responsible for deep oxidation of alkene products. The enhanced basicity by Grignard reduction facilitates the desorption of alkenes and the generated Mg/Cl species have the functionality of improving C4H8 alkenes yield. The Grignard reduction treatment promotes the alkene selectivity by indispensable contributions of the elimination of electrophilic oxygen, higher basicity of the catalyst and the formation of Cl containing Mg/Cl species. Notes The authors declare no competing financial interests. Acknowledgements This work was financially supported by China Petrochemical Cooperation (No. S213043), the Ministry of Science and Technology (MOST) 2016YFA0204100, the National Natural Science Foundation of China (21573254, 91545110) and Youth Innovation Promotion Association, Chinese Academy of Sciences (CAS). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.jcat.2018.01.021. References [1] R. Grabowski, Kinetics of oxidative dehydrogenation of C(2)-C(3) alkanes on oxide catalysts, Catal. Rev.-Sci. Eng. 48 (2006) 199–268. [2] L.M. Madeira, M.F. Portela, Catalytic oxidative dehydrogenation of n-butane, Catal. Rev.-Sci. Eng. 44 (2002) 247–286. [3] M.A. Chaar, D. Patel, M.C. Kung, H.H. Kung, Selective oxidative dehydrogenation of butane over V-Mg-O catalysts, J. Catal. 105 (1987) 483– 498. [4] T. Blasco, J.M.L. Nieto, A. Dejoz, M.I. Vazquez, Influence of the acid-base character of supported vanadium catalysts on their catalytic properties for the oxidative dehydrogenation of n-butane, J. Catal. 157 (1995) 271–282. [5] C. Tellez, M. Abon, J.A. Dalmon, C. Mirodatos, J. Santamaria, Oxidative dehydrogenation of butane over VMgO catalysts, J. Catal. 195 (2000) 113–124.
[6] X. Liu, L. Duan, W. Yang, X. Zhu, Oxidative dehydrogenation of n-butane to butenes on Mo-doped VMgO catalysts, RSC Adv. 7 (2017) 34131–34137. [7] D.S. Su, S. Perathoner, G. Centi, Nanocarbons for the development of advanced catalysts, Chem. Rev. 113 (2013) 5782–5816. [8] J. Zhang, X. Liu, R. Blume, A. Zhang, R. Schloegl, D.S. Su, Surface-modified carbon nanotubes catalyze oxidative dehydrogenation of n-butane, Science 322 (2008) 73–77. [9] W. Qi, W. Liu, B. Zhang, X. Gu, X. Guo, D. Su, Oxidative dehydrogenation on nanocarbon: identification and quantification of active sites by chemical titration, Angew. Chemie-Int. Ed. 52 (2013) 14224–14228. [10] G. Mestl, N.I. Maksimova, N. Keller, V.V. Roddatis, R. Schlögl, Carbon nanofilaments in heterogeneous catalysis: an industrial application for new carbon materials?, Angew Chemie Int. Ed. 40 (2001) 2066–2068. [11] C. Chen, J. Zhang, B. Zhang, C. Yu, F. Peng, D. Su, Revealing the enhanced catalytic activity of nitrogen-doped carbon nanotubes for oxidative dehydrogenation of propane, Chem. Commun. 49 (2013) 8151–8153. [12] B. Frank, J. Zhang, R. Blume, R. Schloegl, D.S. Su, Heteroatoms increase the selectivity in oxidative dehydrogenation reactions on nanocarbons, Angew. Chemie-Int. Ed. 48 (2009) 6913–6917. [13] Y. Zhang, R. Huang, Z. Feng, H. Liu, C. Shi, J. Rong, B. Zong, D. Su, Phosphate modified carbon nanotubes for oxidative dehydrogenation of n-butane, J. Energy Chem. 25 (2016) 349–353. [14] Y.J. Huang, Y.W. Qin, N. Wang, Y. Zhou, H. Niu, J.Y. Dong, J.P. Hu, Y.X. Wang, Reduction of graphite oxide with a grignard reagent for facile in situ preparation of electrically conductive polyolefin/graphene nanocomposites, Macromol. Chem. Phys. 213 (2012) 720–728. [15] Y. Huang, Y. Qin, Y. Zhou, H. Niu, Z.-Z. Yu, J.-Y. Dong, Polypropylene/graphene oxide nanocomposites prepared by in situ ziegler-natta polymerization, Chem. Mater. 22 (2010) 4096–4102. [16] F.P. Greenspan, D.G. Mackellar, Analysis of aliphatic per acids, Anal. Chem. 20 (1948) 1061–1063. [17] Y.J. Zhang, J. Wang, J.F. Rong, J.Y. Diao, J.Y. Zhang, C.F. Shi, H.Y. Liu, D.S. Su, A facile and efficient method to fabricate highly selective nanocarbon catalysts for oxidative dehydrogenation, Chemsuschem 10 (2017) 353–358. [18] R. Chakravarti, A. Mano, H. Iwai, S.S. Aldeyab, R.P. Kumar, M.L. Kantam, A. Vinu, Functionalization of mesoporous carbon with superbasic MgO nanoparticles for the efficient synthesis of sulfinamides, Chem. – A Eur. J. 17 (2011) 6673– 6682. [19] F. Tuinstra, J.L. Koenig, Raman spectrum of graphite, J. Chem. Phys. 53 (1970) 1126–1130. [20] I.O. Maciel, N. Anderson, M.A. Pimenta, A. Hartschuh, H. Qian, M. Terrones, H. Terrones, J. Campos-Delgado, A.M. Rao, L. Novotny, A. Jorio, Electron and phonon renormalization near charged defects in carbon nanotubes, Nat. Mater. 7 (2008) 878–883. [21] F. Atamny, J. Blocker, A. Dubotzky, H. Kurt, O. Timpe, G. Loose, W. Mahdi, R. Schlogl, Surface-chemistry of carbon – activation of molecular-oxygen, Mol. Phys. 76 (1992) 851–886. [22] J.L. Figueiredo, M.F.R. Pereira, M.M.A. Freitas, J.J.M. Orfao, Modification of the surface chemistry of activated carbons, Carbon 37 (1999) 1379–1389. [23] P.E. Fanning, M.A. Vannice, A drifts study of the formation of surface groups on carbon by oxidation, Carbon 31 (1993) 721–730. [24] A. Dejoz, J.M.L. Nieto, F. Melo, I. Vazquez, Kinetic study of the oxidation of nbutane on vanadium oxide supported on Al/Mg mixed oxide, Ind. Eng. Chem. Res. 36 (1997) 2588–2596. [25] J.M.L. Nieto, A. Dejoz, M.I. Vazquez, Preparation characterization and catalytic properties of vanadium-oxides supported on calcined Mg/Al-hydrotalcite, Appl. Catal. A-General. 132 (1995) 41–59. [26] Z. Bai, P. Li, L. Liu, G. Xiong, Oxidative dehydrogenation of propane over MoOx and POx supported on carbon nanotube catalysts, Chemcatchem 4 (2012) 260–264.
Contents lists available at ScienceDirect
Journal of Catalysis journal homepage: www.elsevier.com/locate/jcat
Grignard reagent reduced nanocarbon material in oxidative dehydrogenation of n-butane Jiaquan Li a, Peng Yu a, Jingxin Xie a, Yajie Zhang b, Hongyang Liu b,⇑, Dangsheng Su c, Junfeng Rong a,⇑ a
Research Institute of Petroleum Processing, Sinopec, Beijing 100083, China Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, China c Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China b
a r t i c l e
i n f o
Article history: Received 14 December 2017 Revised 18 January 2018 Accepted 19 January 2018
Keywords: Carbon nanotubes Grignard reagent Electrophilic oxygen Reduction Oxidative dehydrogenation
a b s t r a c t A modification route of nanocarbon catalyst based on Grignard reagent reduction of oxidized carbon nanotubes (o-CNTs) has been developed for oxidative dehydrogenation (ODH) of n-butane. The o-CNTs contain considerable amount of electrophilic oxygen species which are responsible for deep oxidation side-reactions and the alkene selectivity in ODH is low. After Grignard reduction, the corresponding electrophilic oxygen groups on the surface of the catalyst were eliminated and the basicity increased. As a result, the side-reactions in ODH were prohibited and the alkene selectivity was significantly improved compared with o-CNTs. The chlorine containing Mg/Cl species were found to have positive effect on the improvement of C4H8 alkene yield. This study provides a method of the preparation of nanocarbon catalyst to achieve higher alkene selectivity for the dehydrogenation reaction. Ó 2018 Elsevier Inc. All rights reserved.
1. Introduction Oxidative dehydrogenation (ODH) of n-butane is a promising process to produce high value butenes and butadiene from lowcost butane. Traditional metal oxides such as supported vanadium and molybdenum oxide catalysts have been widely employed to catalyze this reaction [1,2]. Among the reported metal oxide catalysts, the VAMgAO catalyst presents the highest selectivity to C4 alkenes [1–6]. However, the traditional metal oxide catalysts are unfavourable in side-reaction controlling and suffer serious deactivation by the coke formation during ODH reaction, which restrains their further application. Nanocarbon materials are attracting considerable interest in catalyzing CAH bond activation such as ODH of light alkanes, owing to their competitive catalytic activity and remarkable coke-resistance [7,8]. It has been proven that the electron-rich quinone groups on the surface of nanocarbon are active sites to catalyze ODH reactions through a redox mechanism [8–10], and the electrophilic oxygen groups (peroxide O2 and superoxide O 2 2) attack C@C bonds and are responsible for deep oxidation of the desired alkene products [8]. A variety of heteroatom-doping treatments have been explored on nanocarbons to enhance the catalytic performance focusing on O, N, B and P doping [8,11–13], however,
⇑ Corresponding authors. E-mail addresses: [email protected] (H. Liu), [email protected] (J. Rong). https://doi.org/10.1016/j.jcat.2018.01.021 0021-9517/Ó 2018 Elsevier Inc. All rights reserved.
the employed modification approaches for ODH reaction are still limited and the catalytic influence of other elements has not been investigated. Moreover, further industrial application of nanocarbon catalysts in ODH of n-butane requires higher alkene selectivity which is a priority in developing new modification strategies. Grignard reagent is a reductant presented as RMgX (R is alkyl or aryl and X is halogen). Reduction of graphene oxide with Grignard reagent has been reported to form functional nanocomposites [14,15], but the impact of Grignard reagent reduction on the catalytic performance of nanocarbon materials hasn’t been discussed. Herein, a modification route of nanocarbon materials was developed based on Grignard reagent reduction of oxidized carbon nanotubes (o-CNTs) to control the distribution of the surface oxygen groups and to introduce heteroatoms on carbon nanotubes (CNTs) that have potential effects on the catalyst. We employed C6H5CH2MgCl as the reductant and Cl/Mg precursor to reduce the electrophilic oxygen species on the surface of o-CNT which cause deep oxidation side-reactions, and examined the effect of Cl/Mg dopant on the catalytic behavior in ODH of n-butane. The C4 alkene selectivity of Grignard reduced o-CNT catalyst (Gr-oCNT) was significantly improved, providing new insights into heteroatomdoped carbon-mediated catalyst. The catalytic roles of electrophilic oxygen, Mg and Cl species in ODH reactions were systematically discussed in this work.
52
J. Li et al. / Journal of Catalysis 360 (2018) 51–56
2. Methodology
titration, respectively. The obtained iodometric titration data were validated with similar results from repeated experiments.
2.1. Catalysts preparation
O22 - þ 2KI þ 2H2 SO4 ! O2 þ 2KHSO4 þ I2 þ H2 O
ð1Þ
I2 þ 2Na2 S2 O3 ! 2NaI þ Na2 S4 O6
ð2Þ
cðelectrophilic oxygenÞ ¼ 2 106 V=m
ð3Þ
Commercial available CNTs were purchased from Shandong Dazhan Nano Materials Co., Ltd., and used directly without any purification. The length, inner diameter and purity of CNTs were: 3–12 mm, 12–15 nm and 96%, respectively. 10 g CNTs was oxidized in 125 mL of concentrated HNO3 (65–68%) and 375 mL of concentrated H2SO4 (95–98%) under sonication at 50 °C for 6 h, and the precipitate was filtered out and washed several times with deionized water until the pH of the filtrate reached 7. The precipitate was dried in the atmosphere at 120 °C for 12 h to obtain oxidized carbon nanotubes (o-CNTs). The Grignard reduction of o-CNT was performed as follows. 2 g o-CNTs was added in 20 mL of C6H5CH2MgCl (1.0 M, solution in diethyl ether, Aldrich Chemical Company). After the solution had been stirred under N2 protection for 2 h at 25 °C, the temperature was increased to 65 °C and held for 12 h. Excess C6H5CH2MgCl was washed off by tetrahydrofuran (THF) for several times and the precipitate was filtered out and dried in the atmosphere at 80 °C for 12 h to obtain Gr-oCNT. The samples involved in control experiments include Mg-oCNT, Cl-oCNT, Gr-CNT and Me-oCNT. The preparation procedure was as follows: Mg-oCNT: 2 g o-CNT was added into 20 mL of Dibutylmagnesium (1.0 M, solution in heptane, Aldrich Chemical Company). After the solution had been stirred under N2 protection for 2 h at 25 °C, the temperature was increased to 65 °C and held for 12 h. Excess Dibutylmagnesium was washed off by tetrahydrofuran (THF) for several times and the precipitate was filtered out and dried in the atmosphere at 80 °C for 12 h to obtain Mg-oCNT. Cl-oCNT: 1 g o-CNTs was added into 20 mL of CCl4. After the solution had been stirred under N2 protection for 2 h at 25 °C, the temperature was increased to 80 °C and held for 12 h. The precipitate was filtered out and dried in N2 at 80 °C for 12 h to obtain Cl-oCNT. Gr-CNT was obtained following the same procedure as Gr-oCNT except that o-CNTs was replaced by pristine CNTs. Me-oCNT was obtained following the same procedure as GroCNT except that C6H5CH2MgCl was replaced by CH3MgCl (1.0 M, solution in diethyl ether, J&K Scientific Ltd).
The titration results were validated by the control experiments where no electrophilic oxygen species were detected on pristine CNTs and blank sample (without CNTs). 2.3. Catalyst performance test The catalytic activity of CNT samples in the reaction of oxidative dehydrogenation of butane was measured at 723 K using a 10 mm diameter fixed-bed quartz tube reactor at atmospheric pressure over 300 min. Reactant and product concentrations (weight percent) were measured by online gas chromatography (Agilent model 7890B) equipped with a HayeSep Q column, a HayeSep N column and a molecular-sieve column connected to a thermal conductivity detector, and a HP-PLOT A12O3 column (50 m 0.53 mm 15 lm) connected to a flame ionization detector. Reactant mixtures (Beijing AP BAIF Gases) contained 0.7 wt% butane, 1.4 wt% O2 and 97.9 wt% N2 with a typical gas flow rate of 4500 mL of gas h1 (g of catalyst)1. The selectivity of alkene products is calculated as follows:
selbutene ¼
cbutene 1:034 100% ½cbutane in ½cbutane out
selbutadiene ¼
cbutadiene 1:074 100% ½cbutane in ½cbutane out
where seli (percent) and ci (weight percent) are selectivity and concentration, respectively, of each C4 alkene product in the outlet gas and [cbutane]in and [cbutane]out are the n-butane concentrations in the inlet and outlet gas mixture, respectively. The side-reaction refers to total oxidation of hydrocarbon to form COx. The calculations of side-reaction rates (moles per gram per second) are based on the reaction conditions (gas flow rate and butane concentration) and the catalytic performance. The calculating formula for the side-reaction rate is presented as
v s ¼ 0:189 106 convð1 selÞ 2.2. Iodometric titration method In this work, we used iodometric titration to determine the surface amount of electrophilic oxygen species on CNTs samples [16]. Electrophilic peroxide and superoxide species can oxidize aqueous I— into I2 for subsequent titration with Na2S2O3. Herein, we suppose the electrophilic oxygen species are mostly composed of peroxides and designate peroxide as electrophilic oxygen for convenient calculation. The titration of electrophilic oxygen groups on CNTs samples was performed as following procedure. A 0.3 g CNT sample was added in a KI solution that consisted of 10 mL of KI (100 g/L), 5 mL of H2SO4 (0.1 mol/L), 30 mL of deionized water, and 3 drops of (NH4)6Mo7O24 (30 g/L). The reaction between peroxides on the surface of CNTs and KI is shown in (1). After 30 min under sonication at 25 °C in the dark, I— was oxidized into I2. Then the precipitate was filtered out and washed six times. I2 in the filtrate was titrated with Na2S2O3 (0.002 mol/L) as presented in (2) and the concentration of electrophilic oxygen (mol/g) on CNTs was calculated with (3) where c (mol/g), V (mL) and m (g) represent the electrophilic oxygen concentration, the volume consumption of Na2S2O3 solution and the mass of CNT samples for
where ms (moles per gram per second), conv (percent) and sel (percent) represent the side-reaction rate, the conversion of butane and the C4 alkene selectivity, respectively. A control experiment was conducted under the same reaction condition except there was no catalyst sample added into the reactor. The conversion rate of n-butane was 0%, and neither alkene nor the side reaction product was detected by gas chromatography. 2.4. Characterizations Transmission electron microscopy (TEM) measurements were taken on a FEI Tecnai F20 microscope with an accelerating voltage of 200 kV. Raman spectra were recorded under ambient conditions on a JY LabRAM HR Raman spectrometer with a 325 nm laser beam. The weight concentrations for element C, H, and O were determined on an Elementar Micro Cube elemental analyzer. The weight concentrations for element Mg and Cl were obtained by Rigaku 3013 X-ray Fluorescence Spectrometer (XRF). X-ray photoelectron spectroscopy (XPS) data were obtained with an ESCALab220i-XL electron spectrometer from VG Scientific using 300 W Al Ka radiation. The base pressure was 3 109 mbar.
J. Li et al. / Journal of Catalysis 360 (2018) 51–56
The binding energies were referenced to the C 1s line at 284.6 eV from defect free graphite. Deconvolution of O 1s spectra were performed using mixed Gaussian-Lorentzian component profiles after subtraction of a Shirley background using XPSPEAK41 software. IR spectra was conducted on a Thermo Nicolet iS10 FTIR system. The XRD patterns were obtained on a XRD-6000 with a Cu Ka source at a scan rate of 10° min1.
3. Results and discussion The oxidation process of pristine CNTs in HNO3/H2SO4 (1:3) mixture creates reactive defects for oxygen functional groups to anchor. Then the obtained o-CNT reacted with C6H5CH2MgCl solution to form Gr-oCNT. The Mg/Cl species were incorporated in the catalytic system to form CNT–O–MgCl structure via reactions between C6H5CH2MgCl and oxygen groups on o-CNT such as carboxylic acids and phenols [15]. The morphological and structural properties of o-CNT and Gr-oCNT characterized by transmission electron microscopy (TEM) were shown in Fig. 1A and B suggesting that the reduction treatment has no apparent impact on the structure of o-CNT and no MgO nanoparticles were found on Gr-oCNT. The results were supported by the high-angle annular dark field scanning transmission electron microscopy (STEM) image in Fig. 1C and elemental mapping of a typical area of Gr-oCNT in Fig. S1 (see Supporting Information). Both O and Mg species were well dispersed on Gr-oCNT without visible aggregation of MgO nanoparticles. Furthermore, the powder X-ray diffraction of GroCNT exhibited two peaks attributed to CNTs framework and no peak of crystalline cubic structure of MgO nanoparticles was
53
observed as shown in Fig. 1D, consisting well with the TEM results [17,18]. According to Raman spectroscopy (see Fig. S2 in Supporting Information), The value of ID/IG [19,20] increased from 0.13 to 0.20 after oxidation of pristine CNTs because of the generation of defects by HNO3/H2SO4 etching, while the ID/IG value of Gr-oCNT was 0.19, close to that of o-CNT. Therefore, it can be inferred that the carbon framework was well maintained after Grignard reduction treatment and the Mg species are highly dispersed on o-CNTs. The total concentrations of O, Mg and Cl obtained from element analysis and X-ray fluorescence (XRF) and the surface concentrations of the elements from X-ray photoelectron spectroscopy (XPS) are summarized in Table S1 (see Supporting Information). A comparison of the catalytic performances of pristine CNTs, oCNT and Gr-oCNT in ODH of n-butane executed at 450 °C with an O2/butane ratio of 2.0 is displayed in Fig. 2A. The oxygen content of pristine CNTs is negligible, while after oxidation the total oxygen concentration increases to 13 wt%. As a result, there is a significant increase of butane conversion (from 6.8% to 28.3%) and C4 alkene selectivity (from 7.9% to 23.4%). After the reduction treatment with C6H5CH2MgCl, the alkene selectivity is enhanced to 59% which is among the highest selectivity that has been reported so far on nanocarbon catalyst [8,13,17], and the conversion decreased to 17%. Notably, butadiene selectivity increases from 13% to 21% after Grignard reduction, while the increase of total selectivity of C4H8 (1-butene, cis-butene and trans-butene) was more significant (from 10% to 38%), indicating that Gr-oCNT tends to facilitate the formation of C4H8 butenes instead of butadiene in catalyzing ODH of n-butane. The electrophilic oxygen species (peroxides and superoxides) on o-CNTs account for the deep oxidation of alkene products and
Fig. 1. TEM images of o-CNT (A) and Gr-oCNT (B). (C) HAADF-STEM image of Gr-oCNT. (D) Powder X-ray diffraction patterns of o-CNT and Gr-oCNT.
54
J. Li et al. / Journal of Catalysis 360 (2018) 51–56
Fig. 2. (A) ODH catalytic performance of pristine CNT, o-CNT and Gr-oCNT. (B) Amount of electrophilic oxygen determined by iodometric titration of o-CNT and Gr-oCNT before and after ODH reaction and the relevant side reaction rates. The ODH reaction condition: 723 K, 1 atm, 2:1 O2:butane ratio, 0.2 g catalyst.
carbon skeleton [8,21], resulting in the reduction of C4 alkene selectivity in ODH reactions. In this sense, the promotion of alkene selectivity can be achieved by reduction of oxygen species of high oxidation states to eliminate the electrophilic oxygen on the surface of catalysts. In this work, iodometric titration was utilized to quantitatively determine the amount of surface electrophilic oxygen species on the catalysts. The amount of the electrophilic oxygen on o-CNT and Gr-oCNT before and after ODH of n-butane combined with the corresponding side-reaction rates is presented in Fig. 2B. Compared with o-CNT, the amount of electrophilic oxygen on Gr-oCNT is extremely low, indicating that Grignard reduction treatment eliminates most of the electrophilic oxygen species on o-CNT. After ODH reaction, over half of the electrophilic oxygen groups are removed from the surface of o-CNT, while GroCNT exhibits apparent increase of the amount of electrophilic oxygen which is possibly due to the regeneration of electrophilic oxygen during ODH in the presence of oxygen. The amount of electrophilic oxygen on Gr-oCNT after ODH is approximately half of that of o-CNT, which is in correspondence with the decreased side-reaction rate after Grignard reduction, showing evidence that the electrophilic oxygen species are sites for combustion sidereactions and explains the enhanced selectivity owing to Grignard reduction treatment. In addition, it has been reported that the alkene products of ODH reaction are weekly adsorbed on basic catalysts [3,4,17]. The Grignard reagent is strong Lewis base and reacts with acidic oxygen groups on o-CNT, thereby improving the basicity of o-CNT and facilitating the desorption of alkenes from CNT’s surface. In this sense, higher basicity of the catalyst due to Grignard reduction also attributes to the increased alkene selectivity in this catalytic reaction.
The influence of the reduction treatment on the composition of surface oxygen groups of o-CNTs was analyzed by IR spectra. IR spectra of o-CNT in Fig. 3 demonstrate three signals associated with the typical oxygen containing groups, namely C@O stretching in carboxylic acid and lactone groups at 1696 cm1 [22,23], CAO stretching at 1015 cm1 and CAOH stretching at 1200 cm1 [15]. After Grignard reduction, both C@O and CAOH stretching vibration signals disappear, indicating that C6H5CH2MgCl reacts with CAOH and O@CAO groups on oACNT to form ACAOAMgCl structure during the reduction treatment. The catalytic role of Mg/Cl species in ODH reactions when it is attached on the catalyst was investigated by replacing oACNT with silica gel (SiO2) which contains abundant AOH groups to anchor AMgCl and then conducting Grignard reduction experiment following the same procedure. Catalytic performances of SiO2 and Grignard reduced SiO2 (Gr-SiO2) in ODH of n-butane were presented in Fig. S3 (see Supporting Information). Although both SiO2 and Gr-SiO2 exhibited low conversion of n-butane, a decrease in butane conversion (from 4% to 2%) and a significant increase in alkene selectivity (from 32% to 89%) were observed when Mg/Cl species were anchored on SiO2, which was in correspondence with the results of Gr-oCNTs. In addition, no butadiene was detected in the product gas and the produced alkenes consist entirely of C4H8 alkenes when SiO2 and Gr-SiO2 were employed as catalysts in ODH. Considering that butadiene is primarily generated from second ODH of C4H8 alkenes following similar dehydrogenation pathway to C4H8 formation [8], we can infer that the impact of the formed Mg/Cl species on ODH of butane is more likely to enhance the production of 1-butene, cis-butene and trans-butene than second ODH to produce butadiene, which is also in accordance with former results of catalytic performance of Gr-oCNT. The control experiments were carried out to testify the role of each individual component on Gr-oCNT that may affect the catalytic property. The catalytic performances of a series of CNT samples obtained from different modification procedure are presented in Fig. 4. Oxygen functional groups, Mg and Cl species are considered as three components on Gr-oCNT, with at least one of them absent for each sample to obtain o-CNT, Mg-oCNT, Cl-oCNT, GrCNT and using CH3MgCl as another kind of Grignard reagent to obtain Me-oCNT in comparison with Gr-oCNT. Dibutylmagnesium and CCl4 were utilized as Mg and Cl precursors to form Mg-oCNT and Cl-oCNT samples. Gr-oCNT exhibits a superior alkene selectivity and a moderate butane conversion among all the samples involved. Notably, dibutylmagnesium (C4H9AMgAC4H9) is also a strong reductant that has similar functionality as Grignard reagent
Fig. 3. ATR-IR of pristine CNTs, o-CNT, Gr-oCNT and Gr-oCNT after ODH of butane.
J. Li et al. / Journal of Catalysis 360 (2018) 51–56
Fig. 4. Catalytic performances of o-CNT, o-CNT refluxed with dibutylmagnesium (Mg-oCNT) and CCl4 (Cl-oCNT), CNTs refluxed with C6H5CH2MgCl (Gr-CNT), o-CNT refluxed with C6H5CH2MgCl (Gr-oCNT) and CH3MgCl (Me-oCNT) in ODH of butane. The ODH reaction condition: 723 K, 1 atm, 2:1 O2:butane ratio, 0.2 g catalyst.
to reduce the electrophilic oxygen groups on o-CNT to obtain Mg-oCNT with Mg dopant. XPS results shows that the surface Mg atom concentration of Mg-oCNT is relatively high as 10.3%. According to the iodometric titration results (see Fig. S4 in Supporting Information), the dibutylmagnesium treatment dramatically reduced the amount of the electrophilic oxygen on o-CNT and the concentration of electrophilic oxygen on Mg-oCNT after ODH reaction is even lower than that of Gr-oCNT. However, the total alkene selectivity of Mg-oCNT is merely 26%, much lower than that
55
of Gr-oCNT (59%) and slightly higher than o-CNT (23.4%). It was reported that for the supported vanadium oxide catalyst system, the isolated Mg2+ sites of the support are responsible for deep oxidation of alkenes [24,25]. For nanocarbon catalysts, the presence of metal, such as Fe and Mo, also has no positive contribution to the improvement of alkene selectivity [8,26]. Therefore, we can infer that the sole presence of Mg species on catalysts has little positive effect on the catalytic performance. By refluxing o-CNT in CCl4, a low Cl loading of 0.21% (XPS results) was achieved to give Cl-oCNT and a slight promotion of alkene selectivity of 5% was obtained. Iodometric titration results and deconvolution of O1s XPS show no obvious change of the amount of electrophilic oxygen and C@O concentration (see Figs. S4 and S5 in Supporting Information) on o-CNT after the CCl4 treatment. We infer that the loaded Cl is beneficial to the selectivity improvement. Me-oCNT showed similar catalytic performance to that of Gr-oCNT with a total selectivity of 50%, indicating the indispensable role of Grignard reagent modification. The samples of the control experiments provide a hint that the enhanced reaction selectivity on Gr-oCNT results from joint effects of the reduction of electrophilic oxygen groups and the promoted basicity together with the loading of Cl containing species. On the basis of the above discussion, the overall influence of Grignard reduction treatment on the structure and catalytic performance of o-CNT is depicted in Fig. 5. The Grignard reduction eliminates the undesired electrophilic oxygen species and increases the basicity of the catalyst. The formed Mg/Cl species plays a positive role in catalyzing ODH reactions. As a result, the total alkene selectivity in ODH of n-butane is promoted after the modification treatment.
Fig. 5. Structural and catalytic influence of Grignard reduction on o-CNT in ODH of n-butane.
56
J. Li et al. / Journal of Catalysis 360 (2018) 51–56
4. Conclusions In summary, a modification approach based on Grignard reagent reduction of o-CNT was developed to improve C4 alkene selectivity in ODH of n-butane and provides new insights of heteroatom-doped nanocarbon catalysts. Iodometric titration method was employed in this work to quantitatively determine the amount of electrophilic oxygen species on the surface of CNTs samples. The Grignard reduction of o-CNTs prevents the sidereactions by eliminating the electrophilic oxygen species that are responsible for deep oxidation of alkene products. The enhanced basicity by Grignard reduction facilitates the desorption of alkenes and the generated Mg/Cl species have the functionality of improving C4H8 alkenes yield. The Grignard reduction treatment promotes the alkene selectivity by indispensable contributions of the elimination of electrophilic oxygen, higher basicity of the catalyst and the formation of Cl containing Mg/Cl species. Notes The authors declare no competing financial interests. Acknowledgements This work was financially supported by China Petrochemical Cooperation (No. S213043), the Ministry of Science and Technology (MOST) 2016YFA0204100, the National Natural Science Foundation of China (21573254, 91545110) and Youth Innovation Promotion Association, Chinese Academy of Sciences (CAS). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.jcat.2018.01.021. References [1] R. Grabowski, Kinetics of oxidative dehydrogenation of C(2)-C(3) alkanes on oxide catalysts, Catal. Rev.-Sci. Eng. 48 (2006) 199–268. [2] L.M. Madeira, M.F. Portela, Catalytic oxidative dehydrogenation of n-butane, Catal. Rev.-Sci. Eng. 44 (2002) 247–286. [3] M.A. Chaar, D. Patel, M.C. Kung, H.H. Kung, Selective oxidative dehydrogenation of butane over V-Mg-O catalysts, J. Catal. 105 (1987) 483– 498. [4] T. Blasco, J.M.L. Nieto, A. Dejoz, M.I. Vazquez, Influence of the acid-base character of supported vanadium catalysts on their catalytic properties for the oxidative dehydrogenation of n-butane, J. Catal. 157 (1995) 271–282. [5] C. Tellez, M. Abon, J.A. Dalmon, C. Mirodatos, J. Santamaria, Oxidative dehydrogenation of butane over VMgO catalysts, J. Catal. 195 (2000) 113–124.
[6] X. Liu, L. Duan, W. Yang, X. Zhu, Oxidative dehydrogenation of n-butane to butenes on Mo-doped VMgO catalysts, RSC Adv. 7 (2017) 34131–34137. [7] D.S. Su, S. Perathoner, G. Centi, Nanocarbons for the development of advanced catalysts, Chem. Rev. 113 (2013) 5782–5816. [8] J. Zhang, X. Liu, R. Blume, A. Zhang, R. Schloegl, D.S. Su, Surface-modified carbon nanotubes catalyze oxidative dehydrogenation of n-butane, Science 322 (2008) 73–77. [9] W. Qi, W. Liu, B. Zhang, X. Gu, X. Guo, D. Su, Oxidative dehydrogenation on nanocarbon: identification and quantification of active sites by chemical titration, Angew. Chemie-Int. Ed. 52 (2013) 14224–14228. [10] G. Mestl, N.I. Maksimova, N. Keller, V.V. Roddatis, R. Schlögl, Carbon nanofilaments in heterogeneous catalysis: an industrial application for new carbon materials?, Angew Chemie Int. Ed. 40 (2001) 2066–2068. [11] C. Chen, J. Zhang, B. Zhang, C. Yu, F. Peng, D. Su, Revealing the enhanced catalytic activity of nitrogen-doped carbon nanotubes for oxidative dehydrogenation of propane, Chem. Commun. 49 (2013) 8151–8153. [12] B. Frank, J. Zhang, R. Blume, R. Schloegl, D.S. Su, Heteroatoms increase the selectivity in oxidative dehydrogenation reactions on nanocarbons, Angew. Chemie-Int. Ed. 48 (2009) 6913–6917. [13] Y. Zhang, R. Huang, Z. Feng, H. Liu, C. Shi, J. Rong, B. Zong, D. Su, Phosphate modified carbon nanotubes for oxidative dehydrogenation of n-butane, J. Energy Chem. 25 (2016) 349–353. [14] Y.J. Huang, Y.W. Qin, N. Wang, Y. Zhou, H. Niu, J.Y. Dong, J.P. Hu, Y.X. Wang, Reduction of graphite oxide with a grignard reagent for facile in situ preparation of electrically conductive polyolefin/graphene nanocomposites, Macromol. Chem. Phys. 213 (2012) 720–728. [15] Y. Huang, Y. Qin, Y. Zhou, H. Niu, Z.-Z. Yu, J.-Y. Dong, Polypropylene/graphene oxide nanocomposites prepared by in situ ziegler-natta polymerization, Chem. Mater. 22 (2010) 4096–4102. [16] F.P. Greenspan, D.G. Mackellar, Analysis of aliphatic per acids, Anal. Chem. 20 (1948) 1061–1063. [17] Y.J. Zhang, J. Wang, J.F. Rong, J.Y. Diao, J.Y. Zhang, C.F. Shi, H.Y. Liu, D.S. Su, A facile and efficient method to fabricate highly selective nanocarbon catalysts for oxidative dehydrogenation, Chemsuschem 10 (2017) 353–358. [18] R. Chakravarti, A. Mano, H. Iwai, S.S. Aldeyab, R.P. Kumar, M.L. Kantam, A. Vinu, Functionalization of mesoporous carbon with superbasic MgO nanoparticles for the efficient synthesis of sulfinamides, Chem. – A Eur. J. 17 (2011) 6673– 6682. [19] F. Tuinstra, J.L. Koenig, Raman spectrum of graphite, J. Chem. Phys. 53 (1970) 1126–1130. [20] I.O. Maciel, N. Anderson, M.A. Pimenta, A. Hartschuh, H. Qian, M. Terrones, H. Terrones, J. Campos-Delgado, A.M. Rao, L. Novotny, A. Jorio, Electron and phonon renormalization near charged defects in carbon nanotubes, Nat. Mater. 7 (2008) 878–883. [21] F. Atamny, J. Blocker, A. Dubotzky, H. Kurt, O. Timpe, G. Loose, W. Mahdi, R. Schlogl, Surface-chemistry of carbon – activation of molecular-oxygen, Mol. Phys. 76 (1992) 851–886. [22] J.L. Figueiredo, M.F.R. Pereira, M.M.A. Freitas, J.J.M. Orfao, Modification of the surface chemistry of activated carbons, Carbon 37 (1999) 1379–1389. [23] P.E. Fanning, M.A. Vannice, A drifts study of the formation of surface groups on carbon by oxidation, Carbon 31 (1993) 721–730. [24] A. Dejoz, J.M.L. Nieto, F. Melo, I. Vazquez, Kinetic study of the oxidation of nbutane on vanadium oxide supported on Al/Mg mixed oxide, Ind. Eng. Chem. Res. 36 (1997) 2588–2596. [25] J.M.L. Nieto, A. Dejoz, M.I. Vazquez, Preparation characterization and catalytic properties of vanadium-oxides supported on calcined Mg/Al-hydrotalcite, Appl. Catal. A-General. 132 (1995) 41–59. [26] Z. Bai, P. Li, L. Liu, G. Xiong, Oxidative dehydrogenation of propane over MoOx and POx supported on carbon nanotube catalysts, Chemcatchem 4 (2012) 260–264.