ISSN 0023-1584, Kinetics and Catalysis, 2017, Vol. 58, No. 4, pp. 362–369. © Pleiades Publishing, Ltd., 2017. Original Russian Text © I.G. Tarkhanova, S.V. Verzhichinskaya, A.K. Buryak, V.M. Zelikman, O.I. Vernaya, R.Z. Sakhabutdinov, R.M. Garifullin, T.V. Bukharkina, L.A. Tyurina, 2017, published in Kinetika i Kataliz, 2017, Vol. 58, No. 4, pp. 384–392.
Effect of the Composition of the Immobilized Copper-Containing Ionic Liquid on the Dodecyl Mercaptan Oxidation Kinetics I. G. Tarkhanovaa, *, S. V. Verzhichinskayab, A. K. Buryakc, V. M. Zelikmana, O. I. Vernayaa, R. Z. Sakhabutdinovd, R. M. Garifullind, T. V. Bukharkinab, and L. A. Tyurinaa, e aFaculty
of Chemistry, Moscow State University, Moscow, 119991 Russia Mendeleev University of Chemical Technology of Russia, Moscow, 125047 Russia c Frumkin Institute of Physical Chemistry and Electrochemistry, Russian Academy of Sciences, Moscow, 119071 Russia d Tatar Oil Research and Design Institute, Bugulma, 423230 Tatarstan, Russia e OOO Start-Katalizator, Moscow, 119991 Russia *e-mail:
[email protected] b
Received April 25, 2016
Abstract⎯Basic kinetic parameters of the catalytic oxidation of dodecyl mercaptan in kerosene in the presence of silica-immobilized pyridinium or imidazolium chlorocuprate complexes have been determined. The composition of the copper-containing anions and the oxidation kinetics depend on the nature of the ionic liquid and on the method of its synthesis. The compositions developed in this study are usable in the removal of hydrogen sulfide and light mercaptans from the oil stripping gas. Keywords: dodecyl mercaptan oxidation, kinetics, immobilized catalysts, copper-containing ionic liquids, pyridinium and imidazolium chlorocuprate complexes DOI: 10.1134/S002315841704019X
Interest in non-hydrogenative methods of removing toxic and corrosive sulfur compounds from petroleum fractions has increased because often no cheap hydrogen sources are available or the hydrotreating method is too costly. As for mercaptans, a possible alternative is catalytic oxidative demercaptanization, in which thiols are oxidized with air or oxygen to disulfides and then to thiolsulfonates and sulfonic acids. This approach underlies the Merox technology, which was developed and is widely used by Honeywell UOP. This process and related ones ensure practically complete mercaptan removal from hydrocarbon media; however, their significant drawback is that the reaction over the catalysts it involves takes place only in the presence of an aqueous alkali [1, 2]. This circumstance necessitates performing laborious operations for separation of water–hydrocarbon mixtures and wastewater treatment, thus causing serious environmental problems. The numerous works dealing with non-alkaline oxidative demercaptanization have not resulted in development of efficient wasteless technologies [3–13]. Thus, the problem of creating new catalytic systems for the direct oxidation of hydrogen sulfide and mercaptans remains an environmentally and technologically important one. Earlier, we reported development of catalysts for the oxidation of heavy mercaptans. These catalysts are based on tetraalkylammonium, pyridinium, and imid-
azolium chlorocuprate ionic liquids (ILs) immobilized on mineral acids [14, 15]. Here, we report the kinetics of this model process under the action of the most active systems—imidazolium and pyridinium derivatives—and the catalyst structure effect on the efficiency of the process. A significant element of this study is verification of the applicability of the suggested catalytic compositions to solving the practical problem of removing hydrogen sulfide and light mercaptans from the oil stripping gas. EXPERIMENTAL The catalytic compositions (Fig. 1) were synthesized via procedures reported in our earlier works [14–16]. The support was granular silica gel (Perlkat 97-0 brand, D = 3–4 mm). For the synthesis of immobilized catalysts, silica gel was dried in toluene by azeotropic distillation in a flask fitted with a Dean–Stark apparatus and a backflow condenser. Next, chloropropyltrimethoxysilane was added to the dehydrated silica gel in toluene under stirring (2 : 1, w/w) and the mixture was boiled for 24 h. After the completion of this procedure, modified silica gel was washed with dry toluene and filtered out. The modified support was then placed in a glass tube with a neck and a ground glass joint, and pyridine (catalyst 1) or ethylimidazole (catalysts 2 and 3) was added (1.5 : 1, w/w). The tube was connected to a
362
EFFECT OF THE COMPOSITION OF THE IMMOBILIZED
N +
Si
CunCl2n + m
O O Si Si
+ N
m−
m−
MeO
N
+N N
+ N
363
Si
OMe
MeO O
O O Si Si
CunCl2n + m
Si O
Si Si
Si
OMe
O
O Si Si
2 (n = 2, 3; m = 1, 2) 3 (n = 2, 3, 4; m = 1, 2)
1 (n = 1, 2; m = 1, 2)
Fig. 1. Structures of the heterogeneous catalysts based on copper chloride complexes with silica-immobilized pyridine (1) and imidazole (2, 3) moieties of the ILs.
vacuum system, cooled with liquid nitrogen, and pumped to a residual pressure of 10–3 Torr. This procedure was repeated three times, and the sample was then sealed off and kept in a temperature-controlled oven (110°C for pyridine and 185°C for ethylimidazole) for 18 h. Thereafter, the tube was unsealed and the solid was washed with ethanol and collected on a filter. Copper (II) chloride hydrate crystals were dissolved in rectified ethanol at 40–45°С, and modified silica gel, taken in a tenfold excess over the dissolved copper salt, was introduced into the solution. The mixture was stirred for 1 h and was then decanted, and the solid was vacuum-dried at 90°С. The synthesis of catalyst 3 differed in that, in copper chloride deposition, hydrochloric acid containing an equimolar amount of HCl with respect to copper was added to ethanol. The mixture was stirred and decanted. The precipitate was washed with ethanol two times and was then vacuum-dried at 110°С for 3 h. Textural characteristics of the catalysts were studied by the BEТ and BJH methods on an Autosorb 1 surface area and pore size analyzer (Quantachrome, United States) using samples outgassed at 150°C for 3 h. Specific surface areas and pore sizes were derived from nitrogen adsorption–desorption isotherms at 77 K using the software supplied with the analyzer. The copper content of the synthesized samples was determined by chelatometric titration [14]. The surface-assisted laser desorption/ionization (SALDI) mass spectra of the synthesized compositions were recorded on an Ultraflex mass spectrometer (Bruker, Germany) fitted with a nitrogen laser (wavelength of 337 nm, pulse energy of 110 μJ) and a time-of-flight mass analyzer. The spectra were recorded in the positive- and negative-ion modes. Cluster ions were identified from isotopic distribution data using the IsoPro simulator. The catalytic reaction was carried out in a static semiflow reactor under vigorous stirring at 50–100°С, a catalyst weight of 0.25 to 1.93 g, and a reaction time KINETICS AND CATALYSIS
Vol. 58
No. 4
2017
of 5 min to 1 h. The model mixtures were dodecyl mercaptans solutions in kerosene (bp 125°С) containing 0.02 to 0.078 mol/L mercaptans through which air was bubbled at a rate of 100 mL/min. In the determination of the reaction order with respect to oxygen, the oxygen concentration was varied by diluting the air with nitrogen in prescribed proportions. All experiments were performed at a stirrer rotational speed of at least 900 rpm, at which the conversion of the initial reactant was independent of the stirring vigor. Preliminarily, we experimentally determined the limits of the kinetically controlled region of the process and demonstrated that the kinetics of the reaction remains invariable at air flow rates above 0.01 L/s. The reaction mixture was analyzed by potentiometric titration [15]. Kinetic data were processed using the UNISYS program package, which was developed by Prof. M.G. Makarov, Department of Basic Organic and Petrochemical Synthesis Technology, Mendeleev University of Chemical Technology of Russia. The dependence of the reaction rate on the catalyst particle size was studied in a series of experiments conducted at fixed initial amounts of crushed and spherical catalysts. The particle size of the latter was ~4 times larger than that of the crushed catalyst. The reaction was investigated at 50°С and an initial dodecyl mercaptan concentration of [RSH] = 0.03 mol/L and an oxygen concentration of [O2] = 0.008 mol/L in the gas phase; the copper concentration in the reaction mixture was 10 mol % of the initial mercaptan concentration. The efficiency of catalyst 2 in the purification of the oil stripping gas was studied at the Laboratory of Gas Treatment Technology of Tatar Oil Research and Design Institute in the following way. Air was passed through oil contained in a vessel. Thereafter, the air was passed through a tubular reactor with an inner diameter of 2.5 mm filled with a catalyst (0.063 g). The stripping time was 30 min, and the catalyst temperature was varied from room temperature to 60°C. After being brought into contact with the catalyst, the gas
364
TARKHANOVA et al.
Table 1. Textural characteristics of the initial support and synthesized catalysts SBET, SBJH, VBJH, Dav, Sample nm m2/g m2/g cm3/g Perlkat (initial) 1 3
430 260 190
450* 315 270
0.75* 0.58 0.49
9* 7.5 7.5
S is specific surface area; VBGH and Dav are pore volume and average diameter, respectively. * Data from the certificate of Perlkat.
was successively passed through absorbing solutions containing sodium carbonate (for quantification of hydrogen sulfide) and sodium hydroxide (for quantification of light mercaptans). The amounts of compounds adsorbed by these solutions were determined via the procedure described below. The amount of a sulfur-containing compound reacted was determined from the difference between its concentration in the absorbing solution in the experiment without a catalyst and its concentration in the solution in the experiment involving a catalyst. The residual concentrations of hydrogen sulfide and light mercaptans in oil were determined via an original method developed at Tatar Oil Research and Design Institute [17]. This method includes the following steps: removal of hydrogen sulfide and light mercaptans from the gas phase with an air or inert gas stream (which lowers their partial pressure over the solution and, accordingly, decreases their concentration in oil), their absorption with sodium carbonate and sodium hydroxide solutions, sulfur precipitation from the solutions by adding cadmium acetate, and iodometric back titration of the precipitates. Excess iodine was titrated with a 0.01 N sedum hyposulfite solution. For carrying out an analysis, hydrogen sulfide and light mercaptans were displaced from an oil sample (2–5 g) with an inert gas or air into the absorbing solutions at 50–70°С. RESULTS AND DISCUSSION Structure of Surface Complexes and Textural Characteristics of the Catalysts The synthesis yielded catalysts that were stable in air and contained 2.0–3.2 wt % Cu (Fig. 1).
Figures 2–4 show the SALDI mass spectra of the synthesized catalysts. Clearly, positive ions in the spectrum of 1 are represented by the organic, propylpyridinium cation (121); for 2 and 3, they are represented by the ethylpropylimidazolium cation (137) ethylpropylimidazolium chloride (173). In addition, the mass spectra of 2 and 3 indicate the presence of fragments of ungrafted IL, namely, EtPrImSi(OMe)3 ions (260). Negative ions in the mass spectrum of 1 correspond to the CuCl2 (135) and Cu2Cl3 (233) species; in the spectrum of 2, the negative ions are CuCl2 (135), CuCl3 (170), Cu2Cl3 (233), Cu3Cl4 (333), and Cu3Cl5 (367). The negative-ion spectrum of 3 differs from the spectra of 1 and 2 in that it is made up by a larger number of species: CuCl2 (135), CuCl3 (170), Cu2Cl3 (233), Cu2Cl4 (267), Cu2Cl5 (303), Cu3Cl4 (333), Cu3Cl5 (366), Cu3Cl6 (403), and Cu4Cl7 (503). Thus, the pyridine-based catalyst mainly contains mononuclear and dinuclear chlorocuprates, while the imidazolebased catalysts are dominated by polynuclear chlorocuprates, and the synthesis involving hydrochloric acid (sample 3) favors the formation of larger ionic clusters. The textural characteristics of catalysts 1 and 3, which differ both in the nature of the organic cation and in cluster ion composition, are listed in Table 1. The copper content data for the catalysts are presented in Table 2. Clearly, modification of the surface and increasing the copper content decrease the specific surface area and pore volume of the catalyst. A comparison of the differences between SBJH and SBET for catalysts 1 and 3 suggests that the micropore surface areas for these catalysts are similar. In addition, modification reduces the average pore diameter, but the modifier nature has no significant effect on this characteristic. Dodecanethiol Oxidation Kinetics The catalytic properties of the synthesized catalysts were studied in the following reaction: 2С12H25SH + 1/2O2 = С12H25SSС12H25 + H2O. Table 2 lists specific dodecanethiol oxidation rates (Rsp, mol min–1 g–1) calculated via the formula Rsp = ΔСV/(mτ), where ΔС is the difference between the ini-
Table 2. Copper content and specific dodecanethiol oxidation rate (Rsp) data for the synthesized catalysts* Rsp × 10 4, mol min–1 g–1 Catalyst 1 2 3
first cycle
after 6-h-long operation
0.162 0.203 0.613
0.070 0.081 0.340
Cu content of the catalyst, wt %
mCat, g
[RSH]0, mol/L
2.0 2.2 3.2
1.00 0.78 0.50
0.039 0.036 0.026
* Solution volume, 100 mL; 50°C. KINETICS AND CATALYSIS
Vol. 58
No. 4
2017
EFFECT OF THE COMPOSITION OF THE IMMOBILIZED
365
(а) Intensity, arb. units 232.823
1200 1000 800
134.887
600 400
330.734
120.987
200
213.846 311.752
95.969 169.850
0 100
150
200
250
300
350
400
450
(b) 121.822
261.338
4000 281.250 97.212
164.092
3000 143.511 241.469
2000
360.873 341.943
141.534 183.928
458.562
1000
439.631 196.120
558.325
296.162
82.434 395.749 437.613
537.360 656.146
0 100
200
300
400
500
600
700 m/z
Fig. 2. (a) Negative- and (b) positive-ion SALDI mass spectra of catalyst 1.
tial and final mercaptan concentrations over the reaction time, V is the reaction mixture volume, m is the catalyst weight, and τ is the reaction time. In the first several hours of catalyst operation, Rsp decreases gradually, and in 6 h it reaches its steadystate value, which practically does not change in subsequent catalytic cycles. It is clear from the data presented in Table 2 that the activity of catalyst 2 is only slightly higher than the activity of catalyst 1, possibly to KINETICS AND CATALYSIS
Vol. 58
No. 4
2017
the higher Cu content of 1. Catalyst 3, which contains the largest amount of Cu, shows the highest activity and time-on-stream stability in the model reaction. While its copper content is 1.5 times higher than that of 1 or 2, its activity in the first cycle is 3 times higher and decreases to a lesser extent in 6-h-long operation. The nonlinear dependence of the reaction rate on the metal content of the catalyst surface can be explained in terms of structural features of the copper chloro
366
TARKHANOVA et al.
(а)
Intensity, arb. units 233.071 135.084
2500
331.052
2000 1500
312.065
212.069
1000
310.055 195.091 170.059
500
268.074 293.089
114.030
377.061
0 100
200
300
433.028 477.021
400
572.987
500
600
(b) 137.030
3000
173.039
2000
260.674 111.053 97.024 95.025 62.891
1000
191.104
238.113
80.974 360.592 321.165
393.279 458.484
556.396
0 100
200
300
400
500
m/z
Fig. 3. (a) Negative- and (b) positive-ion SALDI mass spectra of catalyst 2.
complexes and in terms of the mechanism of the catalytic reaction. Indeed, thiol oxidation over coppercontaining metal complex catalysts is a multielectron process proceeding via the formation of dinuclear and polynuclear copper thiolate complexes [18–20]: RSH
[Cu2+ n Cl2n]
H2O2 RSSR
[Cu+nCln(RS)n] + nHCl
O2
At the first stage, the Cu2+ complex interacts with thiol: electron transfer from the thiol group to Cu2+ results in Cu2+ reduction to Cu+ and yields a mixedligand complex between Cu+ and RS radicals, whose recombination yields the corresponding disulfide. Next, the Cu+ complex interacts with oxygen, yielding H2O2 (which subsequently decomposes) and regenerating the initial Cu2+ complex. Obviously, both the two-electron reduction of oxygen and the recombination of thiol radicals in the coordination sphere can take place only on polynuclear species as components of the catalytic complex. As was demonstrated above in the analysis of SALDI mass spectra, these structures are most abundant in catalyst 3. Therefore, assoKINETICS AND CATALYSIS
Vol. 58
No. 4
2017
EFFECT OF THE COMPOSITION OF THE IMMOBILIZED
367
(а)
Intensity, arb. units 366.115 403.131
266.112 170.140
333.084 303.137
2000
503.067
1500 135.078 482.010
1000
500
429.018
466.054 538.026 582.012 618.925
0 100
200
300
400
500
600
(b) 137.030
3000
173.039
2000
260.674 111.053 97.024 95.025 62.891
1000
191.104 238.113
80.974 360.592 321.165
393.279 458.484
556.396
0 100
200
300
400
500
m/z
Fig. 4. (a) Negative- and (b) positive-ion SALDI mass spectra of catalyst 3.
ciation species containing 3 or 4 copper ions are apparently most active in catalysis. This trend—an increase in catalytic activity as a result of the aggregation of transition metal ions in their complexes—is known well both for organic thiols and for sulfide anions [21, 22]. For metal–polymer catalysts involved in the oxidation of inorganic sulfides, this effect is attributed to the formation of extensive electron transport chains in electron transfer from the sulfide anion to oxygen [22]. As would be expected the difference in structural characteristics (nuclrearity of chlorocuprate complexes) between the pyridine- and imidazole-based KINETICS AND CATALYSIS
Vol. 58
No. 4
2017
compositions gives rise to a difference in the kinetics of thiol oxidation in the hydrocarbon medium. Below, we present the rate equations obtained by processing experimental data on dodecanethiol oxidation over the catalysts examined: for catalyst 1,
d[RSH] dτ [O 2 ] [RSH] = k0 [Cu] exp − E , 1 + 60[RSH] 1 + 714[O 2 ] RT −
( )
368
TARKHANOVA et al.
r0 × 103, mol L–1 min–1 1 0.4 2
0.3 0.2 0.1
0
0.02
0.04
0.06
0.08 0.10 [RSH]0, mol/L
Fig. 5. Initial thiol oxidation rate as a function of the initial thiol concentration for the synthesized catalysts: (1) catalyst 1 (mCat = 1.02 g) and (2) catalyst 2 (mCat = 0.78 g). [O2] = 7.9 mmol/L; 50°С.
r0 × 103, mol L–1 min–1 0.4 1 0.3 2 0.2 3 0.1
0
2
4
6
8 10 [O2]0, mmol/L
Fig. 6. Initial thiol oxidation rate as a function of the oxygen concentration in the gas mixture for the synthesized catalysts: (1) catalyst 1 (mCat = 1.02 g, [RSH]0 = 0.038 mol/L); (2) catalyst 2 (mCat = 0.78 g, [RSH]0 = 0.059 mol/L); (3) catalyst 3 (mCat = 0.25 g, [RSH]0 = 0.023 mol/L); 50°С.
where k0 = 3.0 × 106 L2 min–1 mol–2 and Е = 17.5 kJ/mol; for catalysts 2 and 3,
−
( )
d[RSH] [RSH] = k0 [Cu][O 2 ]exp − E , dτ 1 + 60[RSH] RT
where k0 = 8.4 × 106 L2 min–1 mol–2 and Е = 23 kJ/mol (for catalyst 2) and k0 = 8.1 × 107 L2 min–1 mol–2 and Е = 30 kJ/mol (for catalyst 3); [Cu] is the total amount
of copper in the catalyst per unit volume of the reaction mixture (mol/L). Kinetic measurements demonstrated that, for all of the catalysts, the oxidation rate is a homographic function of the thiol concentration and a linear function of the catalyst concentration. Figure 5 shows experimental dependences of the rates of the processes on the on the thiol concentration (points) for both types of catalysts—pyridine- and imidazole-based ones; the lines represent the kinetic data obtained the UNISYS program package. A similar kinetics was earlier observed for cobalt phthalocyanines, which are classical catalysts for the process considered [23, 24]. However, an analysis of the literature demonstrates that other kinetics must be observed for copper catalysts because Cu+ reoxidation is the rate-limiting step of the catalytic cycle in thiol oxidation catalyzed by Cu2+ ions [18, 25]. Therefore, if the reaction is kinetically controlled, the rate equation must involve only the catalyst and oxygen concentrations. As is clear from the above equations, this condition is satisfied only at high thiol concentrations. Another specific feature of the catalysts is that the processes occurring over systems with different organic cations are described by different rate equations. The difference is in the function relating the rate of the process to the oxygen concentration: for the pyridine derivative, this function is homographic, while for the imidazole-based catalysts, the reaction is first-order with respect to the oxidizer throughout the concentration range examined (Fig. 6). We assumed that the change in the order of the reaction with respect to oxygen on passing to catalysts 2 and 3 can be due to the reaction being controlled by internal diffusion because of the limited accessibility of the active sites that are deep in pores. In order to verify this assumption, we carried out experiments on crushed catalysts. The crushing of catalyst 3 increased the thiol conversion rate by more than order of magnitude. In the case of catalyst 1, this effect was noticeably weaker; the rate of the reaction increased by a factor of ~2. This distinction can be explained as follows: an increase in the size of the ionic clusters leads to an increase in the probability of channels in the IL support being blocked. This is confirmed by the fact that the specific surface area and pore volume of catalyst 3 are smaller than those of 1 (Table 1). Obviously, in this case crushing increases the specific surface area and enhances the accessibility of active sites, which, in catalyst 3, are polynuclear copper chloro complexes. Thus, the reaction over catalyst 3 likely proceeds under mixed (kinetic/internal diffusion) control. The rate-limiting step in this case is, possibly, the diffusion of dissolved oxygen in pores, which is consistent with the change in the rate equation. The reaction over catalyst 1 is much less affected by internal diffusion, and, if the reaction proceeds under kinetic or mixed control, a Langmuir–Hinshelwood type mechanism becomes likely. This is indicated not KINETICS AND CATALYSIS
Vol. 58
No. 4
2017
EFFECT OF THE COMPOSITION OF THE IMMOBILIZED
only by the form of the rate equation but also by the activation energy of the process. Such activation energy values are characteristic of heterogeneous catalysis by metal complexes, including supported phthalocyanine catalysts, in thiol oxidation and are substantially smaller the activation energies observed for homogeneous processes [6, 23]. (These works deal with the interaction between a sulfur-containing compound and oxidizer under conditions of their equilibrium adsorption on active sites.) Oxidation of Hydrogen Sulfide and Light Mercaptans in the Oil Stripping Gas Catalyst 2 was additionally tested in the removal of hydrogen sulfide and light mercaptans from an oil stock. The weight fraction of these compounds in the oil was 263 and 7.7 ppm, respectively. We found that the stripping gas that was subjected to catalytic treatment is almost free of hydrogen sulfide and mercaptans, meeting the mandatory requirements imposed on commercial oil. The oil obtained as a result of this treatment also complies with the Russian standards, since the total weight fraction of hydrogen sulfide in the oil subjected to stripping is as small as 40.5 ppm. Note that, at the early stages of stripping, we observed catalyst darkening propagating throughout the bed height, the color of the catalyst changing from yellow to dark brown. By the end of the process, the color had almost completely restored, indicating that the catalyst had been regenerated after the consumption of the entire hydrogen sulfide and light mercaptans. Thus, we synthesized catalysts based on coppercontaining imidazole ILs immobilized on silica gel. These catalytic systems are active in the oxidation of hydrogen sulfide and thiols. They are usable in the oxidation of substrates in both liquids and gaseous oil stocks. ACKNOWLEDGMENTS This study was supported by the Russian Foundation for Basic Research (grant no. 15-03-01995), NIKSA Company, Start-Katalizator Company, and Darville Enterprises Limited. Experiments and were carried out on equipment purchased through the Moscow State University Development Program.
KINETICS AND CATALYSIS
Vol. 58
No. 4
2017
369
REFERENCES 1. Ganguly, S.K., Das, G., Kumar, S., Sain, B., and Garg, M.O., Catal. Today, 2012, vol. 198, no. 1, p. 246. 2. Mazgarov, A.M., Vil’danov, A.F., and Kopylov, Yu.P., Ross. Khim. Zh., 2004, no. 1, p. 67. 3. Das, G., Sain, B., Kumar, S., and Garg, M.O., Catal. Today, 2009, vol. 141, nos. 1–2, p. 152. 4. Reza, E.M., Reza, S.A., and Reza, A., Iran. J. Chem. Chem. Eng., 2013, vol. 32, no. 2, p. 71. 5. Akhmadullina, A.G. and Akhmadullin, R.M., Khim. Tekhnol. Topl. Masel, 2008, no. 6, p. 3. 6. Bukharkina, T.V. and Verzhichinskaya, S.V., Coke Chem., 2005, no. 11, p. 28. 7. Menini, L., Pereira, M.C., Ferreira, A.C., Fabris, J.D., and Gusevskaya, E.V., Appl. Catal., A, 2011, vol. 392, nos. 1–2, p. 151. 8. Javadli, R. and de Klerk, A., Appl. Petrochem. Res., 2012, vol. 1, no. 1, p. 3. 9. US Patent 8460557, 2013. 10. Eur. Patent 2105489, 2009. 11. US Patent 0228528 A1, 2013. 12. US Patent 2828949, 2012. 13. US Patent 8460557, 2013. 14. Tarkhanova, I.G., Gantman, M.G., and Zelikman, V.M., Appl. Catal., A, 2014, vol. 470, no. 1, p. 81. 15. Tarkhanova, I.G. and Konovalov, V.P., Pet. Chem., 2014, vol. 54, no. 3, p. 218. 16. Zelikman, V.M., Tarkhanova, I.G., and Khomyakova, E.V., Kinet. Catal., 2012, vol. 53, no. 2, p. 222. 17. RF Patent 2285917, 2006. 18. Gilbert, B.C., Silvester, S., and Walton, P.H., J. Chem. Soc., Perkin Trans., 1999, vol. 2, p. 1115. 19. Tarkhanova, I.G., Gantman, M.G., Chizhov, A.O., and Smirnov, V.V., React. Kinet. Mech. Catal., 2010, vol. 101, p. 267. 20. Tarkhanova, I.G. and Gantman, M.G., Kinet. Catal., 2011, vol. 52, no. 1, p. 82. 21. Wang, C., Wesener, S.R., Zhang, H., and Cheng, Y.Q., Chem. Biol., 2009, vol. 16, no. 6, p. 585. 22. Pomogailo, A.D., Catalysis by Polymer-Immobilized Metal Complexes, Amsterdam: Gordon & Breach, 1998. 23. Bukharkina, T.V., Mazgarov, A.M., Verzhichinskaya, S.V., and Savkin, K.A., Coke Chem., 2005, no. 2, p. 22. 24. Tyapochkin, E.M. and Kozliak, E.I., J. Mol. Catal. A. Chem., 2005, vol. 242, p. 1. 25. Bagiyan, G.A., Koroleva, I.K., Soroka, N.V., and Ufimtsev, A.V., Kinet. Catal., 2004, vol. 45, no. 3, p. 372.
Translated by D. Zvukov
