Oxidative dehydrogenation of propane by molecular chlorine

Oxidative dehydrogenation of propane by molecular chlorine

G Model ARTICLE IN PRESS APCATA-15392; No. of Pages 6 Applied Catalysis A: General xxx (2015) xxx–xxx Contents lists available at ScienceDirect A...

632KB Sizes 2 Downloads 74 Views

Recommend Documents

Oxidative Dehydrogenation of Propane on NixMg1−xAl2O4 and NiCr2O4 Spinels
The NixMg1−xAl2O4, NiCr2O4, and MgCr2O4 spinels have been synthesized, characterized with the XRD and XPS methods, and t

Oxidative dehydrogenation of propane over chromium-loaded calcium-hydroxyapatite
Chromium-loaded hydroxyapatite catalysts Cr(x)/CaHAp (0.1 ≤ x ≤ 3.7 wt.% Cr) were prepared by ion-exchange and character

Oxidative Dehydrogenation of Propane on γ-Al2O3 Supported Vanadium Oxides
γ-Al2O3 supported vanadium oxides have been prepared following the continuous adsorption method. The superficial loading

γ-Al2O3 catalysts for the oxidative dehydrogenation of propane.
Oxidative dehydrogenation of propane over MoO3/γ-Al2O3 catalysts with different Mo loading was studied. The catalysts we

Selective oxidative dehydrogenation of propane at 380°C by Cs2.5Cu0.08H3.34PV3Mo9O40 catalyst precursor
Oxidative dehydrogenation of propane was efficiently catalyzed by the Cs2.5Cu0.08H3.34PV3Mo9O40 catalyst precursor at a

Propane Oxidative Dehydrogenation over Low Temperature Rare Earth Orthovanadate Catalysts Prepared by Peroxyl Method
The high purity rare earth orthovanadates were prepared by the peroxyl method, and found to be a kind of the low tempera

Butadione synthesis by dehydrogenation and oxidative dehydrogenation of 2,3-butanediol
The results of a complex investigation of the process and catalysts for heterogeneous synthesis of butadione (diacetyl)

Thermolytic Molecular Precursor Route to Active and Selective Vanadia–Zirconia Catalysts for the Oxidative Dehydrogenation of Propane
The thermolytic molecular precursor route was employed in attempts to obtain highly dispersed and structurally well-defi

Nanostructured hybrid materials as precursors of mesoporous NiMo-based catalysts for the propane oxidative dehydrogenation
•Hybrid materials are made of Ni and Mo guest ions and a self-assembled copolymer.•The copolymer associates 2 “types of

G Model

ARTICLE IN PRESS

APCATA-15392; No. of Pages 6

Applied Catalysis A: General xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Applied Catalysis A: General journal homepage: www.elsevier.com/locate/apcata

Oxidative dehydrogenation of propane by molecular chlorine N.V. Testova a , A.S. Shalygin a,b , V.V. Kaichev a,b , T.S. Glazneva a,b,∗ , E.A. Paukshtis a,b,c , V.N. Parmon a,b,c a b c

Boreskov Institute of Catalysis, Pr. Akademika Lavrentieva, 5, Novosibirsk 630090, Russia Novosibirsk State University, Pirogova str., 2, Novosibirsk 630090, Russia National Research Tomsk State University, Pr. Lenina, 36, Tomsk 634050, Russia

a r t i c l e

i n f o

Article history: Received 23 March 2015 Received in revised form 7 May 2015 Accepted 16 May 2015 Available online xxx Keywords: Propane oxidative dehydrogenation Ruthenium catalyst Chlorine Ruthenium oxychloride

a b s t r a c t The gas-phase oxidative dehydrogenation of propane by molecular chlorine was studied in a flow reactor with a fixed catalyst bed using a series of ruthenium–titania catalysts at temperatures between 150 and 450 ◦ C. It was found that Ru/TiO2 catalyst prepared by the incipient wetness impregnation of titania with an aqueous hydrochloric acid solution of K4 [Ru2 OCl10 ] revealed the highest catalytic activity. The selectivity toward propylene reached 95% at the propane conversion of 50% and the complete conversion of chlorine at 400 ◦ C. Propane consumption turnover frequency was 45 10−3 s−1 which is comparable with the highest known values obtained at the oxidative dehydrogenation of propane. Diffuse reflectance UV–vis spectroscopy and X-ray photoelectron spectroscopy study showed that ruthenium oxychloride was the active component of the catalyst. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Propylene is one of the most important large-tonnage products of refining of natural gas and petroleum products. It is being isolated from the products of the cracking, pyrolysis, as well as the thermal and catalytic dehydrogenation of propane [1]. The catalytic dehydrogenation of propane is the most direct and selective route for the production of propylene. The dehydrogenation of propane is performed over variety of catalysts, notably, over Cr2 O3 , MoO3 , V2 O5 , TiO2 , and GeO2 [2–4]. The main problem limiting the industrial application of this method is the necessity to use high temperatures in order to increase the olefin equilibrium concentration since the dehydrogenation of propane is an endothermic reaction: C3 H8 → C3 H6 + H2 ,

r Н◦ 298 = 124.27 kJ/mol.

(1)

In this case, the dehydrogenation is usually accompanied by intense deposition of coke due to the high temperatures of the process. Furthermore, one of the undesirable byproducts is methyl acetylene which poisons the polymerization catalysts in the further polypropylene production [5]. To overcome these problems, the oxidative dehydrogenation of propane over various catalysts is being developed [6–8]. The yield of

∗ Corresponding author at: Boreskov Institute of Catalysis, Pr. Akademika Lavrentieva, 5, Novosibirsk 630090, Russia. Tel.: +7 383 3269526. E-mail address: [email protected] (T.S. Glazneva).

propylene in the oxidative dehydrogenation even over the best catalysts known to date does not exceed 35% which is not sufficient to satisfy the economic feasibility. The yield of 40% was achieved over Mg/Dy/Li/Cl/O mixed oxides [9]. The highest selectivity (94–96%) was achieved at 200 ◦ C over PtH3 PMo12 O40 catalyst at the conversion of approximately 20% [10]. The oxidative dehydrogenation by molecular oxygen is the exothermic process: C3 H8 + 1/2O2 → C3 H6 + H2 O,

r Н◦ 298 = −117.57 kJ/mol.

(2)

This process is accompanied by the deep oxidation of propane to carbon dioxide and water: C3 H8 + 5O2 → 3CO2 + 4H2 O,

r Н◦ 298 = −2044.04 kJ/mol.

(3)

The latter complicates the industrial realization of this process due to the necessity of waste-heat utilization. There are attempts to use various organochlorine compounds [11,12], sulfur-containing compounds [13,14], nitrous oxide [15,16], CO2 [17], and halogens [18] as mild oxidants for the dehydrogenation of propane. It is noteworthy that the application of halogens for the dehydrogenation was proposed more than 40 years ago, based on thermodynamic calculations [18]. Pasternak and Vadekar calculated the thermodynamics of the oxidative dehydrogenation of hydrocarbons in the presence of different halogens. Molecular chlorine turned out to be the most advantageous oxidizing agent in accordance with the thermodynamic calculations.

http://dx.doi.org/10.1016/j.apcata.2015.05.018 0926-860X/© 2015 Elsevier B.V. All rights reserved.

Please cite this article in press as: N.V. Testova, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.05.018

G Model

ARTICLE IN PRESS

APCATA-15392; No. of Pages 6

N.V. Testova et al. / Applied Catalysis A: General xxx (2015) xxx–xxx

2

Formed HCl was proposed to be oxidized by oxygen back to Cl2 (the Deacon reaction) [18]. It is assumed that the chlorination of propane leads to the formation of monoclorides of propane: C3 H8 + Cl2 → C3 H7 Cl + НCl,

r Н◦ 298 = −133.36 kJ/mol.

(4)

These monochlorides are easily dechlorinated to form propylene: C3 H7 Cl → C3 H6 + HCl,

r Н◦ 298 = 73.01 kJ/mol.

(5)

The total enthalpy of the reaction: C3 H8 + Cl2 → 2HCl + C3 H6 ,

r Н◦ 298 = −60.34 kJ/mol.

(6)

is −60.34 kJ/mol only, which simplifies the task of waste-heat utilization. For many years, this way of the olefins production was thermodynamically complicated by the necessity to perform the Deacon reaction, for which no effective catalysts were available. Nowadays, a RuO2 /TiO2 catalyst is proposed for the Deacon reaction that allows performing this process at temperatures below 300 ◦ C [19–21]. Catalysts based on ruthenium also show good performance in other reactions involving chlorine or its derivatives, such as the oxychlorination of methane and ethane [22]. Probably, the reactivity of ruthenium compounds in the reactions involving chlorine and organochlorines is due to the high ruthenium affinity to chlorine. Hevia et al. [23] suggested that the active site of the RuO2 /TiO2 catalyst in the Deacon reaction was RuO2−x Clx ruthenium oxychloride. Crihan et al. [20] indicated that RuO2 surface was chlorinated during the catalyzed oxidation of HCl with oxygen by selective replacement of bridging oxygen atoms by bridging chlorine atoms. This view was confirmed by Teschner et al. [24] showed that the surface of RuO2 was extensively chlorinated at typical Deacon reaction conditions. It was shown that the K4 Ru2 OCl10 /TiO2 catalyst was more active in the Deacon reaction in comparison with RuO2 and exceeds all known catalysts for activity and selectivity in the oxychlorination of ethane and methane [22]. The aim of this work was the study of catalysts based on ruthenium oxychloride in the oxidative dehydrogenation of propane by molecular chlorine. 2. Experimental 2.1. Catalyst preparation and characterization Ru/TiO2 -1 catalyst (0.28 wt.% Ru) was prepared by the incipient wetness impregnation of titania (Degussa P-25 with a specific surface area of 57 m2 /g and a pore volume of 0.36 cm3 /g) with an aqueous hydrochloric acid solution of K4 [Ru2 OCl10 ] (Aurat, Russia) with pH = 1 at 90 ◦ C followed by drying at 110 ◦ C and calcination in air at 350 ◦ C for 2 h. Low solubility of K4 [Ru2 OCl10 ] in water complicates the preparation of the catalyst with high ruthenium content (the maximum ruthenium content we were able to achieve by incipient wetness impregnation was 0.4 wt.%). Therefore, we applied another synthesis approaches to obtain Ru/TiO2 catalysts with higher ruthenium content. Ru/TiO2 -2 catalyst (2 wt.% Ru) was prepared by the oxidation of Ru(III) (RuCl3 ) to Ru(IV) by potassium perchlorate (KClO4 ). For this purpose, the titanium oxide was impregnated by the incipient wetness method with a solution of RuCl3 (Aurat, Russia) and dried in air at 60 ◦ C for 2 h. Then, the resulting RuCl3 /TiO2 was impregnated by a solution of KClO4 + KCl (Vekton, Russia). The concentration of KClO4 was 25 g/L and the concentration of KCl was 16 g/L. The resulting sample was washed with deionized water and dried at 60 ◦ C for 2 h. Ru/TiO2 -3 catalyst (5 wt.% Ru) was prepared by the mechanical mixing of K4 [Ru2 OCl10 ] and TiO2 . Necessary amounts of regents were mixed until a homogeneous mass was obtained and then it was triturated in a mortar

for 1 h. The amount of ruthenium in the samples was determined by atomic emission spectroscopy with inductively coupled plasma using an Optima 4300 DV PerkinElmer spectrometer. UV–vis diffuse reflectance (UV–vis DR) spectra were recorded by a Shimadzu UV-2501 PC spectrophotometer equipped with an ISR-240A diffuse reflectance attachment. The spectra were registered with respect to a BaSO4 powder in the range of 11,000 − 53,000 cm−1 . The obtained reflectance spectra were converted into the F(R) Kubelka–Munk function. A crystalline K4 [Ru2 OCl10 ]·H2 O sample was also studied by UV–vis DR spectroscopy for comparison. X-ray photoelectron spectroscopy (XPS) studies were performed on a photoelectron spectrometer (SPECS Surface Nano Analysis GmbH) equipped with a hemispherical PHOIBOS-150 analyzer and an XR-50 X-ray source with a double Mg/Al anode. The XPS spectra were acquired in the fixed pass energy mode using Mg K␣ radiation (h = 1253.6 eV). Relative concentrations of elements were determined from the total intensities of corresponding corelevel spectra using cross-sections according to Scofield [25]. All the spectra were analyzed using the FitXPS software. In short, after subtraction of a Shirley-type background [26], the spectra were fit using Gaussian/Lorentzian line-shapes. The charge effect was corrected by setting the C1s peak (due to adventitious hydrocarbons) at 284.8 eV. For the XPS study, the Ru/TiO2 -1 catalyst was chosen due to the highest catalytic activity. 2.2. Catalyst testing The catalytic experiments were performed in a tubular flow reactor with a fixed catalyst bed (quartz tube of 200 mm long and with an inner diameter of 7 mm) at atmospheric pressure in the temperature range of 150–450 ◦ C. The amount of catalyst placed into the reactor was 1.0 g. Propane and argon (Sibtekhgaz, Russia) with a purity of at least 99.9% were introduced into the reactor using mass-flow controllers. Chlorine produced in an electrolytic cell was supplied with a mixture of argon. All technological tubes and tanks of chlorine were isolated from the light. The quantity of generated chlorine was calculated according to the electrochemical equivalent and the amount of electricity passed through the saturated sodium chloride solution used as an electrolyte. The gas hourly space velocity (GHSV) of propane and chlorine were 200 and 50 (100) h−1 , respectively. The argon was supplied with the GHSV of 400 h−1 . The molar ratio of the reactants was C3 H8 :Cl2 :Ar = 2:(0.5–1):4. The total GHSV of the reactant feed was 650−700 h−1 . Sampling gaseous products was performed directly at the reactor outlet. The reaction products were analyzed by a Tsvet-570 chromatograph equipped with a specially made capillary column (20 m × 0.32 mm) with a DVB-PLOT stationary phase and a flame ionization detector. At the gas-phase chlorination of propane (4), only half of the chlorine atoms leads to the formation of chlorocarbons and the other half provides hydrogen chloride. When performing the reaction in excess of the hydrocarbon with a molar ratio of C3 H8 /Cl2 = 4 (or C3 H8 /Cl2 = 2 in some experiments) at the complete consumption of chlorine and the formation of only monochlorinated hydrocarbon, the conversion of propane will not exceed 25% (or 50%, respectively). The conversion of propane will be lower in case of the formation of deeply chlorinated compounds. 3. Results and discussion 3.1. Catalysts characterization Fig. 1 shows UV–vis DR spectra of K4 [Ru2 OCl10 ] ruthenium oxychloride and the ruthenium catalysts prepared on the basis

Please cite this article in press as: N.V. Testova, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.05.018

G Model

ARTICLE IN PRESS

APCATA-15392; No. of Pages 6

N.V. Testova et al. / Applied Catalysis A: General xxx (2015) xxx–xxx

3

Table 1 Relative concentrations of elements in the subsurface layer of K4 [Ru2 OCl10 ]·H2 O and the fresh and spent Ru–TiO2 -1 catalyst. Sample K4 [Ru2 OCl10 ]·H2 O Ru–TiO2 -1, fresh Ru–TiO2 -1, spent

Data source

[O]/[Ru]

[ORu ]/[Ru]

[Cl]/[Ru]

[K]/[Ru]

[Ru]/[Ti]

XPS Chemical composition XPS XPS

1.1 1.0 20 19

0.55 0.5 – –

2.6 5.0 13 8.0

2.8 2.0 1.5 1.4

– – 0.007 0.010

thereof. The spectrum of ruthenium oxychloride is characterized by bands at 14,000, 17,000, and 20,000 cm−1 (curve 1). The spectrum of the fresh Ru/TiO2 -1 catalyst (curve 2a) contains the bands of K4 [Ru2 OCl10 ] as well as an additional band at 22,600 cm−1 (curve 3a) characteristic for K4 [Ru2 O2 (H2 O)4 Cl8 ] ruthenium complex, probably formed by the partial hydrolysis of K4 [Ru2 OCl10 ] [27,28]. The spectrum of the fresh Ru/TiO2 -2 catalyst (curve 3a) also contains the bands at 14,000, 17,000, and 20,000 cm−1 , however the ratio of intensities of these bands is different from that of the spectrum of K4 [Ru2 OCl10 ], which is probably due to the formation of various ruthenium oxychlorides. The manifestation of the band at 14,000 cm−1 proves the presence of [Ru2 OCl10 ]4− anions. It is well known [29] that the bridging oxygen in this anion has the electrophylic nature and may be responsible for the low temperature oxidation of HCl. The spectrum of the fresh Ru/TiO2 -3 catalyst coincides with that of the source K4 [Ru2 OCl10 ]. An intense band at 31,500 cm−1 observed in the spectra of all the catalysts refers to TiO2 . The UV–vis DR spectra of the spent catalysts are different from the spectra of fresh ones. In the spectrum of Ru/TiO2 -1 (curve 2b), the intensity of the band at 22,600 cm−1 decreased and the intensity of the bands at 14,000 and 20,000 cm−1 increased. For comparison, the UV–vis spectrum of the mechanical mixture of K4 [Ru2 OCl10 ] with TiO2 after the reaction (curve 4b) contains the same bands at 14,000 and 20,000 cm−1 characteristic for K4 [Ru2 OCl10 ]. The spec-

trum of Ru/TiO2 -2 (curve 3b) does not contain any bands besides the edge of 31,500 cm−1 and an increase in absorption in the region of 10,000 cm−1 . Hence, the UV–vis spectroscopy data show the presence of [Ru2 OCl10 ]4− anions in the fresh and spent Ru/TiO2 -1 and Ru/TiO2 -3 catalysts. Crystalline K4 [Ru2 OCl10 ]·H2 O as well as the freshly prepared and spent Ru/TiO2 -1 catalyst were studied by XPS. The atomic ratios of the elements in the subsurface layer of the samples are shown in Table 1. It can be seen that the [K]/[Ru] atomic ratio observed for the crystalline K4 [Ru2 OCl10 ]·H2 O sample slightly differ from the stoichiometric composition. It may be because XPS is a surface-sensitive method, the XPS analysis depth is several nanometers only. At the same time, the [O]/[Ru] atomic ratio on the surface is the same as that in the bulk. The content of chlorine on the surface of K4 [Ru2 OCl10 ]·H2 O is reduced. It seems to be due to the formation of a small amount of RuOx species on the surface of crystalline K4 [Ru2 OCl10 ]·H2 O. Note, that in both cases, oxygen on the surface occupies bridging positions between the two neighboring ruthenium cations. In contrast, the content of chlorine on the surface of the fresh Ru/TiO2 -1 catalyst is overstated. It indicates that the surface of titania is partially covered by chlorine species. The Ru3d5/2 , O1s, and Cl2p3/2 binding energies of the fresh Ru/TiO2 -1 catalyst and reference compounds are shown in Table 2. According to the literature [30,31], the Ru3d5/2 binding energy of RuO2 are in the range of 280.5–281.4 eV, the Ru3d5/2 binding energy of RuO3 and RuO4 are 282.5 and 283.3 eV, respectively. It should be noted that the last Ru–oxide phases were discussed in the literature [30,32,33], and some authors suggested that RuO3 may be present on the surface of RuO2 as an intermediate and cannot be synthesized as a unique compound. (H) Over assigned the Ru3d5/2 binding energy of 282.7 eV to the excitation of the RuO2 plasmon [32,33]. The Ru3d features ascribed to RuO3 and RuO4 in Refs. [30,31] can equally be explained by contaminations. The Ru3d5/2 binding energy of complexes containing chlorine is higher in comparison with oxygen-containing compounds, even at a lower ruthenium charge: the Ru3d5/2 binding energies of RuCl3 and RuCl3 ·H2 O are 281.8 and 282.1 eV, respectively [34,35]. The Ru3d5/2 binding energy of bulk K4 [Ru2 OCl10 ]·H2 O is 282.8 eV; hence, we can speculate that ruthenium is in the Ru4+ state. The O1s spectrum of the bulk K4 [Ru2 OCl10 ]·H2 O sample is presented in Fig. 2. A significant width of the O1s spectrum of Table 2 Ru3d5/2 , O1s, and Cl2p3/2 binding energies for different compounds and catalyst under study. Sample

Fig. 1. UV–vis DR spectra of K4 [Ru2 Ocl10 ]·H2 O (1); Ru/TiO2 -1 catalyst (2a,b); Ru/TiO2 -2 (3a,b); Ru/TiO2 -3 (4a,b); “a” and “b” symbols refer to spectra of the catalyst before and after the reaction test, respectively. The deconvolution of the spectrum to the Gaussian components is shown for 2b curve.

K4 [Ru2 OCl10 ]·H2 O Ru–TiO2 -1 RuO2 RuO4 RuCl3 RuCl3 ·H2 O Metal chlorides KClO3 KClO4 TiO2 a

Binding energy, eV

Reference

Ru3d5/2

O1s

Cl2p3/2

282.8 281.0 280.5–281.4 283.3 281.8 282.1 − − − −

531.7; 533.1 529.9; 531.8 528.4–529.6 – – – – 532.5 532.4 529.9

196.7; 198.0 198.0 – – – – 199.1–200.9 206.7a 209.0a –

This work 30, 31, 37 34 35 38–40 41 This work

The Cl2p binding energy.

Please cite this article in press as: N.V. Testova, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.05.018

G Model APCATA-15392; No. of Pages 6 4

ARTICLE IN PRESS N.V. Testova et al. / Applied Catalysis A: General xxx (2015) xxx–xxx

Fig. 2. O1s core-level spectra of K4 [Ru2 OCl10 ]·H2 O, fresh and spent Ru/TiO2 -1 catalyst.

K4 [Ru2 OCl10 ]·H2 O is due to the presence of two components of approximately equal intensity with the maxima at 531.7 and 533.1 eV. The first peak can be attributed to oxygen anions that are directly bonded to the Ru atoms. The second peak at the higher binding energy corresponds to oxygen in the molecules of water [36,37]. The O1s binding energy of RuO2 is 528.4–529.6 eV which is close to the O1s binding energy of various transition metal oxides [37]. These binding energies are significantly lower than that for ruthenium oxychloride (the first peak in the O1s spectrum). Therefore, it can be concluded that ruthenium oxychloride has a lower partial charge of oxygen (Oı− ) when compared with oxygen in

ruthenium oxides. The O1s spectra of the fresh and spent catalysts consist of an intensive peak at 529.9 eV and a weak peak at 531.8 eV. The first peak certainly corresponds to oxygen in the TiO2 lattice (Table 2) whereas the second peak can be attributed to Ru O Ru species. The last is in good agreement with small values of the [Ru]/[Ti] atomic ratio observed (Table 1). The Cl2p spectrum of bulk K4 [Ru2 OCl10 ]·H2 O contains an intense Cl2p3/2 − Cl2p1/2 doublet with peaks at 198.0 and 199.6 eV, respectively, as well as a significantly less intense doublet with the Cl2p3/2 binding energy equal to 196.7 eV (Fig. 3). As is well known, the Cl2p3/2 binding energy of metal chlorides with the negatively charged chlorine is

Fig. 3. Cl2p core-level spectra of K4 [Ru2 OCl10 ]·H2 O, fresh and spent Ru/TiO2 -1 catalyst.

Please cite this article in press as: N.V. Testova, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.05.018

G Model

ARTICLE IN PRESS

APCATA-15392; No. of Pages 6

N.V. Testova et al. / Applied Catalysis A: General xxx (2015) xxx–xxx

5

Table 3 Temperature dependence of propane conversion and selectivity toward the main products in the propane chlorination over Ru/TiO2 catalysts at the molar ratio of C3 H8 :Cl2 :Ar = 2:1:4 in the initial feed. Sample

Ru content, %

Tr , ◦ C

XC3 H8 , %

SC3 H7 Cl , %

SC3 H6 Cl2 , %

SC3 H6 , %

SC3 H5 Cl ,a %

Ru/TiO2 -1 (impregnation)

0.28

Ru/TiO2 -2 (impregnation)

2.0

Ru/TiO2 -3 (mechanical mixture)

5.0

300 350 400 300 350 400 350 300 350 400

25.3 50.8 47.6 0.2 50.6 3.5 0.3 13.0 22.6 46.3

19.9 12.3 1.1 − 3.1 14.5 71 29.7 20.1 17.9

9.2 − − − − − − 18.0 1.0 0.1

34.5 81.2 94.6 − 88.3 64.3 − 20.6 70.9 78.7

34.2 5.1 3.2 − 7.8 6.8 21 28.9 7.0 2.5

Sb , % 2.2 1.4 1.1 − 0.8 14.4 8 2.8 1 0.8

a According to mass spectrometric analysis, 3 peaks in the chromatogram of the gas sample were attributed to chloropropylenes without specifying the particular structure since their spectra had minor differences. b Selectivity of C1 C2 hydrocarbons and their chlorinated derivatives.

in the range of 199.1–199.5 eV [38–41]. For KClO3 and KClO4 with positively charged chlorine, the Cl2p binding energy is significantly higher and usually observed in the range of 206.3–208.3 eV [41]. In the Cl2p spectra of the fresh and spent catalysts, only one doublet with the Cl2p3/2 binding energy of 198.1 eV is observed. The XPS data for the fresh Ru/TiO2 -1 catalyst differ from that of bulk ruthenium oxychloride (Table 2). The Ru3d5/2 binding energy of the catalyst is 281.0 eV instead of 282.8 eV for K4 [Ru2 OCl10 ]·H2 O. The electronic state of oxygen bonded with ruthenium is the same: the O1s binding energy is 531.7–531.8 eV in both cases. However, the [ORu ]/[Ru] atomic ratio is higher for the catalyst and is equal to 1 instead of 0.5 for bulk ruthenium oxychloride. The Cl2p3/2 binding energy for the fresh and spent catalyst is the same as for bulk ruthenium oxychloride (198.1 eV). Besides, the chlorine content is higher for the supported catalyst. This can be explained by the fact that the impregnation of titanium oxide by the hydrochloric acid solution of K4 [Ru2 OCl10 ]·H2 O leads to the partial chlorination of the titanium oxide surface, and it is difficult to distinguish between the chlorine atoms associated with the support in the form of Ru Cl Ti bridges from the chlorine atoms bonded only with the ruthenium atoms (Ru Cl Ru). Hence, the XPS study of the Ru/TiO2 -1 catalyst showed that the ruthenium is in the Ru4+ state mainly, the oxygen and chlorine atoms possess the same properties as those in the bulk K4 [Ru2 OCl10 ]. Therefore, on the basis of the UV–vis DRS and XPS data, we can assume that the active component of the Ru/TiO2 -1 catalyst is ruthenium oxychloride with the bridging oxygen of the electrophylic nature in [Ru2 OCl10 ]4− anion [29].

be formed by the direct oxidative dehydrogenation with molecular chlorine, like the oxidative dehydrogenation of propane with oxygen [42]. The conversion of propane increases with increase in temperature, and propylene is formed with a high selectivity due to the lack of chlorine in the reaction mixture. The twofold increase of the chlorine content (C3 H8 /Cl2 = 2) in the initial reactant feed leads to increase in the conversion of propane. Table 3 shows the conversion of propane and selectivity toward propylene and chlorinated products in the presence of the Rucontaining catalysts at temperatures above 300 ◦ C and at the molar ratio of the reactants in the initial feed of C3 H8 :Cl2 :Ar = 2:1:4. When comparing the results obtained for the different Rucontaining catalysts, it was found that the selectivity toward propylene at 400 ◦ C over the Ru/TiO2 -1 catalyst (94.6%) prepared by impregnation was higher than that over the Ru/TiO2 -3 catalyst represented the mechanical mixture (78.7%) at similar conversions of propane. For the Ru/TiO2 -2 catalyst, only minor quantities of 1chloropropane were observed at 300 ◦ C at the reactor outlet. The conversion increased significantly at 350 ◦ C and reached 50%. The selectivity toward propylene was 88.3%. A rapid fall in the conversion of propane (3.5%) was observed after raising the reactor temperature up to 400 ◦ C. The subsequent lowering of the temperature down to 350 ◦ C led to a complete deactivation of the sample indicating a change in the catalyst state, probably due to the destruction of ruthenium oxychloride. This is consistent with the UV–vis DR spectroscopy data (Fig. 1) showed that the bands at 14,000, 17,000, and 20,000 cm−1 characteristic for bulk K4 [Ru2 OCl10 ] are absent in the spectrum of the spent catalyst.

3.2. Catalyst performance The catalysts prepared on the basis of ruthenium oxychloride were tested in the reaction of propane with molecular chlorine in the temperature range of 200–450 ◦ C. Fig. 4 shows the temperature dependence of the propane conversion and the selectivity toward propylene oxide, 2-chloropropane, 1-chloropropane, and 1,2-dichloropropane for the Ru/TiO2 -1 catalyst at the molar ratio of C3 H8 /Cl2 = 4. This catalyst showed the highest selectivity toward propylene (96.7% at 400 ◦ C and at the molar ratio of C3 H8 :Cl2 :Ar = 2:0.5:4). It has been found that 1,2-dichloropropane is the main reaction product for the Ru/TiO2 -1 catalyst at low temperatures (up to 300 ◦ C). At high temperatures near 400–450 ◦ C, propylene is formed with a high selectivity. The dehydrochlorination of 1,2-dichloropropane may result in the formation of monochloropropylene or methylacetylene, however the content of these products at high temperatures is negligible. It indicates that besides the reaction path through the chlorination and the subsequent dehydrochlorination over the Ru/TiO2 -1 catalyst, propylene can

Fig. 4. Temperature dependence of propane conversion (1) and selectivity toward propylene (2), 2-chloropropane (3), 1-chloropropane (4), and 1,2-dichloropropane (5) in the propane chlorination. The molar ratio of reagents in the initial feed was C3 H8 :Cl2 :Ar = 2:0.5:4.

Please cite this article in press as: N.V. Testova, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.05.018

G Model APCATA-15392; No. of Pages 6

ARTICLE IN PRESS N.V. Testova et al. / Applied Catalysis A: General xxx (2015) xxx–xxx

6

Hence, the oxidative dehydrogenation of propane using molecular chlorine as an oxidant in the presence of the Ru-based catalyst permits to produce propylene with the selectivity of 95%. Furthermore, this process can be performed, in contrast to the thermal and oxidation processes, at a lower temperature (400 ◦ C vs. 600 ◦ C), which reduces significantly the possibility of cracking of reaction products. The conversion of propane is determined by a total consumption of chlorine and depends on the molar ratio of C3 H8 /Cl2 in the initial reactant feed. The decrease in amount of chlorine in the initial feed leads to the increase in the selectivity toward propylene and to decrease in the content of related chlorinated products. Propane consumption turnover frequency (TOF) calculated for propane conversion of 50% at 400 ◦ C and a flow of 0.2 L/h is equal to 45 × 10−3 s−1 . This TOF value is comparable with that reported for oxidative dehydrogenation of propane over vanadia catalysts [8,43]; however, the conversion of propane in our case is higher. 4. Conclusions A series of ruthenium–titania catalysts was prepared via different synthesis methods and studied in the oxidative dehydrogenation of propane by molecular chlorine. The formation of ruthenium oxychloride was confirmed by XPS and UV–vis DR spectroscopy. The catalyst prepared by the incipient wetness impregnation of titania with an aqueous hydrochloric acid solution of K4 [Ru2 OCl10 ] showed the best catalytic performance. The selectivity toward propylene at the propane conversion of 50% reached 95% at 400 ◦ C and TOF equal to 45 × 10−3 s−1 which is comparable with the highest known values obtained at the oxidative dehydrogenation of propane. Acknowledgments Ministry of Education and Science of the Russian Federation is gratefully acknowledged for supporting the research. The authors thank A.A. Saraev for the XPS measurements. References [1] E. Derouane, V. Parmon, F. Lemos, F. Ribeiro, Sustainable Strategies for the Upgrading of Natural Gas: Fundamentals, Challenges, and Opportunities, Springer, Netherlands, 2003. [2] M. Kocon, P. Michorczyk, J. Ogonowski, Catal. Lett. 101 (2005) 53–57. [3] Y. Baba, A.Y. Jibril, K. Atta, Z.M. Melghit, El-Hadi, H. Ala’a, Al-Muhtaseb, Chem. Eng. J. 193–194 (2012) 391–395. [4] M.D. Putra, S.M. Al-Zahrani, A.E. Abasaeed, J. Energy Chem. 22 (2013) 778–782.

[5] Pat. US 8309776 B2, Method for contaminants removal in the olefin production process, 2012. [6] F. Cavani, F. Trifiro, Catal. Today 24 (1995) 307–313. [7] F. Cavani, N. Ballarini, A. Cericola, Catal. Today 127 (2007) 113–131. [8] C.A. Carrero, R. Schlogl, I.E. Wachs, R. Schomaecker, ACS Catal. 4 (2014) 3357–3380. [9] L. Leveles, S. Fuchs, K. Seshan, J.A. Lercher, L. Lefferts, Appl. Catal. A 227 (2002) 287–297. [10] N.I. Kuznetsova, G.Ya. Popova, L.I. Kuznetsova, V.I. Zaikovskii, S.V. Koscheeva, Yu.A. Chesalova, T.V. Andrushkevich, A.S. Lisitsyna, V.A. Likholobov, S. Han, Appl. Catal. A 477 (2014) 1–7. [11] S. Sugiyama, Yu. Iizuka, E. Nitta, H. Hayashi, J.B. Moffat, J. Catal. 189 (2000) 233–237. [12] S. Sugiyama, N. Fukuda, H. Hayashi, Catal. Lett. 78 (2002) 139–143. [13] P.D. Clark, N.I. Dowling, X. Long, Y. Li, J. Sulfur Chem. 25 (2004) 381–387. [14] Z.A. Premji, J.M.H. Lo, P.D. Clark, J. Phys. Chem. A 118 (2014) 1541–1556. [15] A. Gallardo-Llamas, C. Mirodatos, J. Perez-Ramirez, Ind. Eng. Chem. Res. 44 (2005) 455–462. [16] E.V. Kondratenko, M. Cherian, M. Baerns, Catal. Today 112 (2006) 60–63. [17] V.C. Corberan, Catal. Today 99 (2005) 33–41. [18] I.S. Pasternak, M. Vadekar, Can. J. Chem. Eng. 48 (1970) 212–215. [19] K. Seki, Catal. Surv. Asia 14 (2010) 168–175. [20] D. Crihan, M. Knapp, S. Zweidinger, E. Lundgren, C.J. Weststrate, J.N. Andersen, A.P. Seitsonen, H. Over, Angew. Chem. Int. Ed. 47 (2008) 2131–2134. [21] C. Mondelli, A.P. Amrute, F. Krumeich, T. Schmidt, J. Perez-Ramirez, Chem. Catal. Chem. 3 (2011) 657–660. [22] A.S. Shalygin, E.A. Paukshtis, E.V. Kovalyov, B.S. Bal’zhinimaev, Front. Chem. Sci. Eng. 7 (2013) 279–288. [23] M.A.G. Hevia, A.P. Amrute, T. Schmidt, J. Perez-Ramirez, J. Catal. 276 (2010) 141–151. [24] D. Teschner, R. Farra, L. Yao, R. Schlogl, H. Soerijanto, R. Schomacker, T. Schmidt, L. Szentmiklosi, A.P. Amrute, C. Mondelli, J. Perez-Ramirez, G. Novell-Leruth, N. Lopez, J. Catal. 285 (2012) 273–284. [25] J.H. Scofield, J. Electron Spectrosc. Relat. Phenom. 8 (1976) 129–137. [26] D.A. Shirley, Phys. Rev. B 5 (1972) 4709–4714. [27] A.A. Nesterov, A.V. Bashilov, V.K. Runov, N.M. Kuz’min, Mendeleev Commun. 8 (1998) 95–96. [28] T.M. Buslaeva, S.A. Simanova, Koord. Khim. 26 (2000) 379–387. [29] L.W. Paes, R.B. Faria, J.O. Machuca-Herrera, S.P. Machado, Inorg. Chim. Acta 321 (2001) 22–26. [30] K.S. Kim, N. Winograd, J. Catal. 35 (1974) 66–72. [31] H. Madhavaram, H. Idriss, S. Wendt, Y.D. Kim, M. Knapp, H. Over, J. Assman, E. Loffler, M. Muhler, J. Catal. 202 (2001) 296–307. [32] H. Over, A.P. Seitsonen, E. Lundgren, M. Smedh, J.N. Andersen, Surf. Sci. 504 (2002) L196–L200. [33] H. Over, Chem. Rev. 112 (2012) 3356–3426. [34] B. Folkesson, Acta Chem. Scand. 27 (1973) 287–302. [35] A. Ohyoshi, F. Goetzfried, W. Beck, Chem. Lett. 9 (1980) 1537–1540. [36] S. Zweidinger, D. Crihan, M. Knapp, J.P. Hofmann, A.P. Seitsonen, C.J. Weststrate, E. Lundgren, J.N. Andersen, H. Over, J. Phys. Chem. C 112 (2008) 9966–9969. [37] J.P. Hofmann, S. Zweidinger, M. Knapp, A.P. Seitsonen, K. Schulte, J.N. Andersen, E. Lundgren, H. Over, J. Phys. Chem. C 114 (2010) 10901–10909. [38] S. Ntais, V. Dracopoulos, A. Siokou, J. Mol. Catal. A 220 (2004) 199–205. [39] A. Leon, O. Kircher, M. Fichtner, J. Rothe, D. Schild, J. Phys. Chem. B 110 (2006) 1192–1200. [40] A. Siokou, D. Kefalas, S. Ntais, Surf. Sci. 532–535 (2003) 472–477. [41] W.E. Morgan, J.R. van Wazer, W.J. Stec, J. Am. Chem. Soc. 95 (1973) 751–755. [42] K. Chen, A.T. Bell, E. Iglesia, J. Phys. Chem. B 104 (2000) 1292–1299. [43] A. Dinse, B. Frank, C. Hess, D. Habel, R. Schomäcker, J. Mol. Catal. A 289 (2008) 28–37.

Please cite this article in press as: N.V. Testova, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.05.018