Reac Kinet Mech Cat (2018) 123:625–639 https://doi.org/10.1007/s11144-017-1339-z

Kinetics of H2S selective oxidation by oxygen at the carbon nanofibrous catalyst Vasiliy Shinkarev1,2,3 • Gennady Kuvshinov2,4 • Andrey Zagoruiko1,5

Received: 2 November 2017 / Accepted: 17 December 2017 / Published online: 27 December 2017  Akade´miai Kiado´, Budapest, Hungary 2017

Abstract This work is dedicated to construction of kinetic model for the process of H2S selective oxidation into elemental sulfur at carbon nanofibrous (CNF) catalyst. The experiments included the CNF synthesis and kinetic studies. The modified minimization procedure was proposed for kinetic modelling, model discrimination and determination of kinetic parameters. The selected kinetic model provides the qualitatively adequate and quantitatively accurate description of experimental results in a wide range of temperatures (155–250 C), H2S (0.5–2 V%) and O2 (0.25–10 V%) concentrations and reaction mixture humidity (0–35 V% of water vapor). The average value of deviation in the experimental and calculated concentration of key reactants (H2S, O2, SO2) does not exceed 0.04 V%. Such value is comparable with the mean error in maintenance and control of these concentrations in experiments, so the overall model accuracy may be estimated as quite high. The constructed model may be applied for the mathematical modelling, engineering and scale-up of different H2S oxidation processes, based on the CNF catalyst. Keywords Hydrogen sulfide  Sulfur  Oxidation  Selectivity  Carbon nanofibers  Kinetics

Electronic supplementary material The online version of this article (https://doi.org/10.1007/s11144017-1339-z) contains supplementary material, which is available to authorized users. & Andrey Zagoruiko [email protected] 1

Boreskov Institute of Catalysis, Lavrentieva, 5, Novosibirsk, Russia 630090

2

Novosibirsk State Technical University, Marxa, 30, Novosibirsk, Russia 630073

3

OCSiAl Company, Inzhenernaya, 24, Novosibirsk, Russia 630090

4

Sochi State University, Sovetskaya, 26 ‘‘A’’, Sochi, Russia 354000

5

Tomsk Polytechnic University, Lenina, 30, Tomsk, Russia 634050

123

626

Reac Kinet Mech Cat (2018) 123:625–639

Introduction The improvement of the sulfur recovery efficiency in Claus plants is an important task in the area of environmental protection related to processing of natural gas and oil. The main tool to increase sulfur recovery efficiency over the last few decades is the application of the Claus tail gas clean-up processes, including those based on the selective catalytic oxidation of hydrogen sulfide into sulfur by oxygen [1–4]. These processes use the catalysts on the base of various oxides [5, 6], such as iron or vanadia, supported on granular, monolith or fibrous supports, silica or alumina supports [7–13]. Such catalysts are characterized by high H2S oxidation selectivity into sulfur under moderate temperatures (below 250 C) and the molar O2/H2S ratio equal to stoichiometric value or slightly exceeding it (0.5–1). At the same time, at higher O2/H2S values, the sulfur formation selectivity may significantly decrease, especially under increased humidity of the reaction media. This problem may be resolved by application of the carbon-based catalysts. Though the possibility to use the activated carbons as the hydrogen sulfide oxidation catalysts has been known for a long time [14], such catalysts, using the conventional activated carbon, have not found the wide practical application. Most probably, it relates to the developed microporous structure of such carbons, forcing the excessive condensation of forming sulfur in micropores due to capillary effects. This may lead to fast catalyst blocking by liquid sulfur even at temperatures well above the dew point of sulfur vapors. Another problem of traditional activated carbons is their low mechanical strength, insufficient for practical application. One of the most promising directions in the area are the catalysts on the base of carbon nanofibers (CNF) [15, 16]. Such material is formed by the decomposition of hydrocarbons at catalysts containing nickel or iron [15–20]. It consists of densely inter-twinned carbon fibers with the structure, similar to one in graphite. Such catalysts are characterized by developed mesoporous texture and high surface uniformity. They also have a high mechanical strength. As shown earlier [15], such catalysts may be active, stable and quite selective even under significant excess of oxygen (O2/H2S as high as 10–15). The best performance was achieved with the CNF catalysts synthesized over Ni–Cu catalysts [16]. The current study is dedicated to the development of the kinetic model of H2S oxidation at the described CNF material. Such a model is essential for further process engineering and scale-up.

Experimental Catalyst synthesis The CNF synthesis was performed by methane decomposition at the initial catalyst, containing 80 wt% nickel and 10 wt% with Al2O3 as a structural promotor. The initial catalyst was produced by means of an original method. First, the concentrated nitric acid was added under the permanent mixing to the fixed amount

123

Reac Kinet Mech Cat (2018) 123:625–639

627

of pseudo-boehmite until the achievement of acidity modulus equal to 0.1. Then, the batches of Ni(NO3)26H2O and Cu(NO3)23H2O were added and the mixture was stirred until the formation of the uniform paste. This paste was mixed with concentrated water solution of ammonia (* 25% NH3) until the obtaining pH 10.0. The manufactured paste was then placed into a muffle furnace and calcined in the following regime: 150 C/2 h, 250 C/4 h, 550 C/6 h, 750 C/6 h. After calcination at 550 C, the catalyst was partially milled and averaged over the volume. After calcination at 750 C the catalyst was milled in a ceramic mortar and the target fraction of 0.5–1.0 mm was extracted. Al2O3 was required as structural promoter to keep active metal or alloy nanoparticles separated from each other. To synthesize the CNF sample, the prepared initial catalyst (Ni-Cu/Al2O3) was heated in methane flow in the laboratory reactor up to 550 C. The unit methane flow rate was maintained at the level of 120 l/h per 1 g of initial catalyst. The laboratory reactor was made from a quartz cylinder of 30 mm diameter, equipped with soldered inlet and outlet tubes. The synthesis of the sample was stopped when the initial catalyst was still active for methane decomposition, but the conversion of methane was decreased by half. Experimental equipment and methods The scheme of the experimental setup is given in Fig. 1. The laboratory reactor consisted of the vertical tube made of Pyrex glass with internal diameter of 15 mm and height of 100 mm. The catalyst bed was placed inside the reactor on the gasdistributing grid. The catalyst was loaded in the form of thin particles, the preliminary experiments have demonstrated the absence of diffusion limitations. The reactor was situated inside the electric thermostat. The temperature in the reaction zone was measured by a thermocouple located inside the catalyst bed. To provide the efficient mixing of reactants and to improve the temperature uniformity, the catalyst bed was vibro-fluidized by reactor oscillations in axial direction. The initial gaseous flow was prepared by mixing hydrogen sulfide, oxygen and helium in necessary proportions. To provide the required concentration of water

Fig. 1 The scheme of the H2S oxidation installation

123

628

Reac Kinet Mech Cat (2018) 123:625–639

vapors the helium–oxygen mixture was passed through the saturator, operated under fixed temperature. The gas pipeline from saturator to sulfur condenser was heated up to 110 C to avoid condensation of water vapors. The composition of the gaseous mixture at the reactor inlet and outlet was measured by a gas chromatograph Tsvet 500 M, using polysorb separation column and heat conductivity detector. Before the analysis the mixture was passed through calcium chloride desiccator. The experimental program included variation of temperature in the range 150–250 C and variation of concentrations of H2S (0.5–2.0 V%), O2 (0.25–10 V%) and water (0–35 V%). Helium was used as balance in all experiments. Catalyst loading was equal to 0.4 g, the flow rate of the inlet mixture was 1.5 ml/ s. The specific surface of CNF sample was determined from isotherms of lowtemperature adsorption of nitrogen at 77 K; the isotherms were acquired using an adsorption installation NOVA 2200e (Quantachrome) and used for calculating BET surface area, pore volume and pore size. A high-resolution electron microscope JEM-100CX (JEOL) was used for acquiring transmission electron microscopy (TEM) micrographs and a BS-350 (Tesla) microscope for scanning electron microscopy (SEM) micrographs of CNF sample. Processing of the experimental data The following set of catalytic reactions was initially considered for the process of H2S oxidation at CNF: H2 S þ 1=2 O2 ) 1=6 S6 þ H2 O

ð1Þ

H2 S þ 3=2 O2 ) SO2 þ H2 O

ð2Þ

H2 S þ 1=2 SO2 ) 1=4 S6 þ H2 O

ð3Þ

1=6 S6 þ O2 ) SO2

ð4Þ

The experiments were performed in the vibro-fluidized reactor. Therefore, the following CSTR model can be used for description of reaction rates with respect to key processes of consumption of H2S and O2 and formation of SO2:

123

WH2 S ¼

CHin2 S  CHout2 S mcat =F

ð5Þ

WO2 ¼

COin2  COout2 mcat =F

ð6Þ

Reac Kinet Mech Cat (2018) 123:625–639

629

WSO2 ¼

out CSO 2 mcat =F

ð7Þ

Here CH2 S ; CO2 and CSO2 are the volume concentration of reactants (volume fractions), F is the reaction mixture flow rate (st ml/s), mcat is the catalyst loading (g), indexes in and out correspond to reactor inlet and outlet. Sulfur concentration was not measured in experiments, so sulfur was excluded from the list of key reactants. The water formation rate was also excluded from consideration as soon as it is equal to the rate of hydrogen sulfide consumption, thus carrying no additional information. The reaction rates in respect to reactants were represented by formal kinetics equations as a function of outlet mixture composition: Wi ¼

4 X

  mij wj Cout

ð8Þ

1

Here mij is the stoichiometric coefficient for ith reactant in jth reaction, and wj is the rate of jth reaction (j = 1…4 for reactions 1–4). Using the experimentally measured values of outlet concentrations, and various   kinetic hypothesis for wj C out functions, the hypothetic inlet concentrations were calculated for key reactants: mcat ð9Þ Ciin ¼ Ciout þ Wi F The linear minimization criterion z was calculated as the absolute difference between calculated and experimental inlet concentrations, averaged per all experimental points and all key reactants ! y¼n  i¼k 1X 1 X  exp calc  z¼ ð10Þ C  Ciy  k i¼1 n y¼1 iy Here k is the number of key reactants, n is the number of experimental points. For the given task k = 3 (H2S, O2, SO2), n = 112. Use of the criterion of sum of absolute differences (Eq. 10) provides good fit between the experimental and calculated data including the reactants with high concentrations such as H2S and O2 and low concentration such the SO2. In the conventional optimization method based on the least square criterion, the model fits poorly the low concentrations of SO2. Besides, the value of z for the sum of absolute differences (Eq. 10) is directly transferred to average deviation between the concentrations that makes the criterion more physically justified compared to least square criterion. The modified method of steepest descent was used for the search of kinetic constants. This method includes the variation of kinetic constants values by their increase and decrease by the value:

123

630

Reac Kinet Mech Cat (2018) 123:625–639

DKi ¼ Ki s

ð11Þ

Here s is the maximum step coefficient (the initial value was set to 0.1). Such variation is applied to all constants separately, as well as to all their possible combinations. This is required for solution of tasks characterized with strong intercorrelation of constants, when the surface of z values contains the ‘‘ravines’’, located along the diagonals in respect to main axes. The overall amount of possible combinations is defined by the number of kinetic constants to be defined (nk): N ¼ 3nk  1

ð12Þ

Though, the value of N may be rather high, the test calculations showed that this method provides the fastest and most accurate solutions than more simplified methods, based on separate variation of constants. For each set of constants resulting from described variation, the z value is calculated. The set with the minimum z value among all combinations is then established as the basic set for the next step. This procedure is repeated while it provides the improvement (decrease) of z value at each step. At some step, when no further improvement is observed, the step coefficient value s undergoes two-fold decrease and the whole procedure is repeated with this new value and so on. All these steps are stopped when s becomes lower some fixed threshold (we used the value of 10-6).

Results and discussion Characterization of CNF sample The parameters of synthesized CNF are summarized in Table 1. The carbon yield is a weight ratio of a CNF to initial catalyst metal (gCNF/gcat). The initial pre-catalyst was still present in the sample of nanofibrous carbon as contamination. As-prepared samples of nanofibrous carbon were granular materials with a granule size in the range of 1–5 mm. The typical appearance of the granules is shown in Fig. 2a. The study of the materials by scanning electron microscopy Table 1 Description of the synthesized NFC sample Initial pre-catalyst composition, wt%

80 Ni–10 Cu/Al2O3

Carbon deposition temperature, C

550

Carbon yield, gCNF/gcat

230

The amount of metal or alloy in the sample, wt%

0.43

The amount of a structural promoter in the sample, wt%

0.05

BET surface area, m2/g

250

Pore volume, cm3/g

0.59

Average pore diameter, nm

9.5

123

Reac Kinet Mech Cat (2018) 123:625–639

631

Fig. 2 Morphology and structure of CNF sample: a a photography of typical granules; b, c SEM images of the materials; d TEM image of the typical carbon nanofiber of the sample

(SEM) revealed that each granule was formed by a huge number of interlaced nanofibers with typical diameters below 100 nm as shown in the SEM image in Figs. 2b and 2c. The external and internal surface of granules is formed by the surface of nanofibers composing it. Each granule exhibits a high level of porosity due to spaces between nanofibers. Fig. 2c shows that the tips of the nanofibers look brighter than the rest of the nanofiber bodies. The bright spots can be identified as metal particles of an initial catalyst located at the tips of carbon nanofibers. One or several nanofibers associated with the same NiCu alloy particle can be observed in this sample. Fig. 2d shows a high-resolution image of the typical surface of our nanofibrous sample. The graphite layers are preferentially oriented perpendicularly to the surface of nanofibers. The surface structures observed by TEM can be interpreted as binding of adjacent graphite planes. The adjacent layers may bond with each other or linking of the second-neighbor may occur. The edges of graphite planes may also present on the surface. The sample of nanofibrous carbon are predominantly mesoporous material with BET surface area of 250 m2/g. Total pore volume is 0.59 cm3/g and average pore diameter is about 9.5 nm.

123

632

Reac Kinet Mech Cat (2018) 123:625–639

Results of kinetic experiments The results of all performed kinetic experiments are available in the Supplementary Material. The most important experimental trends are presented below in Figs. 3, 4, and 5. For the presentation of results, we used conversions of H2S and oxygen: XH 2 S ¼ 1 

XO ¼ 1 

CHout2 S CHin2 S

ð13Þ

COout2 COin2

ð14Þ

We also used the selectivity of hydrogen sulfide oxidation into elemental sulfur:

Fig. 3 Experimental conversions of hydrogen sulfide (a) and oxygen (b), selectivity of H2S oxidation to sulfur (c) versus temperature in experiments with variation of O2 concentration. Inlet mixture: 1 V% H2S, 0.2–10 V% O2 (see data at the plot for different experiments), 10 V% H2O

123

Reac Kinet Mech Cat (2018) 123:625–639

633

Fig. 4 Experimental conversions of hydrogen sulfide (a) and oxygen (b), selectivity of H2S oxidation to sulfur (c) versus temperature in experiments with variation of H2S concentration. Inlet mixture: 0.5–2 V% H2S (see data at the plot for different experiments), 1 V% O2, 10 V% H2O

S¼1

out CSO 2  CHout2 S

CHin2 S

ð15Þ

Fig. 3 shows the dependence of experimentally measured H2S and O2 conversions, as well as oxidation selectivity upon temperature under variation of inlet oxygen concentration. The inlet concentrations of H2S and water vapor in these experiments were kept at the constant level (1.0 and 10 V%, correspondingly), while O2 concentration was varied in the range from 0.2 to 10 V%. It is seen that conversions of both H2S and O2 are rising with temperature. The increase of oxygen concentration results in increase of H2S conversion and decrease of O2 conversion. The selectivity of oxidation is getting lower with the rise of temperature and oxygen concentration. Fig. 4 shows similar data for experiments with variation of H2S concentration. These experiments were performed under constant O2 and water vapor inlet concentrations (1 and 10 V%, correspondingly) and changing the hydrogen sulfide concentration from 0.5 to 2.0 V%. Again, both H2S and O2 conversions are rising

123

634

Reac Kinet Mech Cat (2018) 123:625–639

Fig. 5 Experimental conversions of hydrogen sulfide (a) and oxygen (b), selectivity of H2S oxidation to sulfur (c) versus temperature in experiments with variation of water vapor concentration. Inlet mixture: 1 V% H2S, 1 V% O2, 0–35 V% H2O (see data at the plot for different experiments)

with temperature. Higher H2S and lower O2 conversion, as well as lower oxidation selectivity, are observed for lower inlet hydrogen sulfide concentration. The influence of humidity of reaction mixture is demonstrated in Fig. 5. Here we used the reaction mixture, containing constant concentrations of H2S and O2 (both 1.0 V%) and variable content of water vapor (0, 10 and 35 V%). It is seen, that water inhibits both the H2S and O2 conversion. At higher temperatures (above 200 C), a definite decrease in selectivity is observed for mixtures with higher humidity. At lower temperatures, the selectivity behavior seems to be more complicated, though the accuracy of selectivity calculation here may be not sufficient due to low conversions, so the final decision on the selectivity trends in this case requires additional studies. In general, we may state that water produces the moderate oxidation selectivity decrease.

123

Reac Kinet Mech Cat (2018) 123:625–639

635

Kinetic model for the process of selective oxidation of H2S Different hypothetic kinetic models were analyzed using the described data treatment approach. Currently, we do not have detailed data on the reaction mechanism, which may provide the formulation of completely physically consistent kinetic equations for reactions 1–4. Most probably, H2S reacts with oxidized catalyst sites with formation of water and adsorbed sulfur. The formed sulfur may either desorb or undergo further oxidation to SO2 in the case of high surface concentration of oxidized sites. The reduced site, formed after desorption of sulfur, may be re-oxidized by oxygen or, less probable, by SO2. It is also reasonable to propose that all reactants (H2S, O2 and H2O) may be reversibly adsorbed at the catalyst surface, thus inhibiting the oxidation reactions. Preliminary data analysis has shown the significant non-linearity of reaction rates with respect to concentrations of reactants, demonstrating the actual complexity of the kinetic model for the considered process. Thus, we used semi-empirical models, constructed on the base of both the observed experimental regularities and some reasonable propositions on the view of kinetic equations. The considered models accounted for various reaction orders in respect to H2S and O2 concentrations, as well as to different types of inhibition factors. The preliminary calculations have shown that the account of the Claus reaction (Eq. 3) provides practically no influence on the quality of kinetic description, therefore the system of considered reactions were limited to reactions 1, 2 and 4. Most probably, this fact is explained by the insignificant rate of re-oxidation of reduced catalyst sites by SO2. After a number of calculations, we selected the following kinetic model:   CH2 S CO0:52 E1   ð16Þ W1 ¼ k1 exp  RT 1 þ k4 eE4 =RT CH2 S þ k5 eE5 =RT CO0:52 ð1 þ k6 eE6 =RT CH2 O Þ   CH0:52 S CO2 E2   W2 ¼ k2 exp  ð17Þ RT 1 þ k7 eE7 =RT CH2 S þ k8 eE8 =RT CO0:52 ð1 þ k6 eE6 =RT CH2 O Þ   1=6 CS6 CO0:52 E3 W4 ¼ k3 exp  RT 1 þ k6 eE6 =RT CH2 O

ð18Þ

The optimal values of parameters for the system 16–18 are given in Table 2. The obtained values of all activation energies are within the physically reasonable range. The rather high value of activation energy E3 for reaction 4 and general view of the Eq. 18, closed to the form of the mass action law, may be an evidence that this reaction occurs (at least partially) via homogeneous non-catalytic mechanism. The results of the kinetic modelling for experiments with variation of H2S and O2 concentrations are given in Figs. 6 and 7. The results are presented in a form of concentration differences DC = Cin - Cout for H2S and O2; and outlet

123

636

Reac Kinet Mech Cat (2018) 123:625–639

Table 2 Parameters of the kinetic model 16–18 Parameter

Measurement unit

Value 8

Parameter

Measurement unit

Value

k1

st ml/(g s)

(3.62 ± 0.25) 9 10

E1

kJ/mol

57.73 ± 3.99

k2

st ml/(g s)

(5.69 ± 0.63) 9 1012

E2

kJ/mol

109.6 ± 18.6

k3

st ml/(g s)

(7.34 ± 0.69) 9 104

E3

kJ/mol

215.6 ± 18.0

k4

n/d

407.5 ± 39.0

E4

kJ/mol

0.24 ± 0.06

k5

n/d

6.34 ± 0.44

E5

kJ/mol

5.04 ± 0.48

k6

n/d

0.661 ± 0.091

E6

kJ/mol

5.06 ± 0.74

k7

n/d

368.2 ± 36.1

E7

kJ/mol

0.47 ± 0.05

k8

n/d

12.47 ± 1.41

E8

kJ/mol

1.89 ± 0.17

Fig. 6 Calculated (lines) and experimental (points) values of DCi: consumption of H2S (a) and O2 (b); formation of SO2 (c) versus outlet oxygen concentration for various temperatures. Initial mixture composition: H2S * 1 V% vol, H2O * 10 V%, range of inlet O2 concentration variation—from 0.25 up to 10% vol. Reaction mixture flow rate—1.5 st ml/s, catalyst loading—0.4 g

123

Reac Kinet Mech Cat (2018) 123:625–639

637

Fig. 7 Calculated (lines) and experimental (points) values of DCi: consumption of H2S (a) and O2 (b); formation of SO2 (c) versus outlet H2S concentration for various temperatures. Initial mixture composition: O2 * 1 V%, H2O * 10 V%, range of inlet H2S concentration variation—from 0.5 up to 2% vol. The reaction mixture flow rate is 1.5 st ml/s, the catalyst loading is 0.4 g

concentrations for SO2. As seen from Eqs. 5–7, these values are proportional to reaction rates WH2 S ; WO2 and WSO2 . It is seen that the proposed model describes quite well the main experimentally observed kinetic regularities, including selectivity decrease under temperature rise and under increase of oxygen concentration, the well-expressed inhibition of reactions 1 and 2 by both hydrogen sulfide and oxygen as well as the less strong inhibition effect of water. The average value of deviation of calculated and experimental concentrations for key reactants (value of criterion z) does not exceed 0.04 V%, this being comparable with the accuracy of concentrations maintenance and analysis in experiments. Therefore, we may summarize that the proposed model provides qualitatively adequate and quantitatively accurate description of experimental data.

123

638

Reac Kinet Mech Cat (2018) 123:625–639

Conclusions The kinetics of the process of hydrogen sulfide selective oxidation on the carbon nano-fiber catalyst was constructed. Kinetic experiments were performed at the synthesized catalyst samples in the vibro-fluidized isothermal reactor. The experimental program included variation of reaction parameters in a wide range: temperature variation included almost all practically important range (155–250 C); variation of H2S and O2 concentrations made possible to study the oxidation reaction under very different conditions: the inlet O2/H2S ratio was varied from 0.24 up to 10 (theoretical stoichiometric value for selective oxidation into sulfur is 0.5). A modified minimization procedure was developed for processing of experimental data. The proposed reaction scheme includes the reactions of H2S oxidation by oxygen into elemental sulfur and into sulfur dioxide, as well as formation of SO2 by oxidation of sulfur vapors. The Claus reaction was excluded from consideration due to its negligible contribution to the process. The proposed model accounts for strong inhibition of H2S oxidation reactions by H2S and O2 as well as for weak inhibition by water vapor. Such inhibition is most probably caused by competitive adsorption of reactants at catalyst surface. High activation energy for the reaction of sulfur oxidation may point to the (at least partial) occurrence of this reaction via the homogeneous non-catalytic way. In general, the model provides a qualitatively adequate description of the experimental array with reproduction of all important regularities with rather high quantitative accuracy. The average value of deviation in the experimental and calculated concentration of key reactants (H2S, O2, SO2) does not exceed 0.04 V%. Such a value is comparable with the mean error in maintenance and control of these concentrations in experiments, so the overall model accuracy may be estimated as quite high. The proposed model may be applied for the mathematical modelling of the different H2S oxidation processes in a wide range of process conditions. Acknowledgements This work was conducted within the frameworks of Russian State budget Project No. 0303-2016-0017 for Boreskov Institute of Catalysis and of Tomsk Polytechnic University Competitiveness Enhancement Program grant.

References 1. Pie´plu A, Saur O, Lavalley JC, Legendre O, Ne´dez C (1998) Claus catalysis and H2S selective oxidation. Catal Rev Sci Eng 40:409–450. https://doi.org/10.1080/01614949808007113 2. Wie¸ckowska J (1995) Catalytic and adsorptive desulphurization of gases. Catal Today 24:405–465. https://doi.org/10.1016/0920-5861(95)00021-7 3. Zagoruiko A, Shinkarev V, Vanag S, Bukhtiyarova G (2010) Catalytic processes and catalysts for production of elemental sulfur from sulfur-containing gases. Catal Ind 4:343–352. https://doi.org/10. 1134/S2070050410040082 4. Zhang X, Tang Y, Siqiu Q, Da J, Hao Z (2015) H2S-selective catalytic oxidation: catalysts and processes. ACS Catal 5:1053–1067. https://doi.org/10.1021/cs501476p

123

Reac Kinet Mech Cat (2018) 123:625–639

639

5. Marshneva VI, Mokrinskii VV (1988) Catalytic activity of metal oxides in hydrogen sulfide oxidation by oxygen and sulfur dioxide. Kinet Catal 29:989–993 6. Davydov AA, Marshneva VI, Shepotko ML (2003) Metal oxides in hydrogen sulfide oxidation by oxygen and sulfur dioxide: I. The comparison study of the catalytic activity. Mechanism of the interactions between H2S and SO2 on some oxides. Appl Catal A 244:93–100. https://doi.org/10. 1016/S0926-860X(02)00573-2 7. Terorde RJAM, Van den Brink PJ, Visser LM, Van Dillen AJ, Geus JW (1993) Selective oxidation of hydrogen sulfide to elemental sulfur using iron oxide catalysts on various supports. Catal Today 17:217–224. https://doi.org/10.1016/0920-5861(93)80026-W 8. Kovalenko ON, Kundo NN, Novopashina VM, Khanaev VM (1998) Investigation of H2S oxidation by oxygen on oxide catalysts in sulfur condensation conditions. React Kinet Catal Lett 64:129–137. https://doi.org/10.1007/BF02475380 9. Shin MY, Nam CM, Park DW, Chung JS (2001) Selective oxidation of H2S to elemental sulfur over VOx/SiO2 and V2O5 catalysts. Appl Catal A 211:213–225. https://doi.org/10.1016/S0926860X(00)00866-8 10. Nguyen P, Edouard D, Nhut J-M, Ledoux MJ, Pham Ch, Pham-Huu C (2007) High thermal conductive b-SiC for selective oxidation of H2S: a new support for exothermal reactions. Appl Catal B 76:300–310. https://doi.org/10.1016/j.apcatb.2007.06.007 11. Mikenin PE, Tsyrul’nikov PG, Kotolevich YuS, Zagoruiko AN (2015) Vanadium oxide catalysts on structured microfiber supports for the selective oxidation of hydrogen sulfide. Catal Ind 2:155–160. https://doi.org/10.1134/S2070050415020075 12. Palma V, Barba D (2014) Low temperature catalytic oxidation of H2S over V2O5/CeO2 catalysts. Int J Hydrogen Energy 39:21524–21530. https://doi.org/10.1016/j.ijhydene.2014.09.120 13. Mikenin P, Zazhigalov S, Elyshev A, Lopatin S, Larina T, Cherepanova S et al (2016) Iron oxide catalyst at the modified glass fiber support for selective oxidation of H2S. Catal Commun 87:36–40. https://doi.org/10.1016/j.catcom.2016.08.038 14. Steijns M, Mars P (1974) The role of sulfur trapped in micropores in the catalytic partial oxidation of hydrogen sulfide with oxygen. J Catal 35:11–17. https://doi.org/10.1016/0021-9517(74)90177-8 15. Shinkarev VV, Glushenkov AM, Kuvshinov DG, Kuvshinov GG (2009) New effective catalysts based on mesoporous nanofibrous carbon for selective oxidation of hydrogen sulfide. Appl Catal B 85:180–191. https://doi.org/10.1016/j.apcatb.2008.07.011 16. Shinkarev VV, Glushenkov AM, Kuvshinov DG, Kuvshinov GG (2010) Nanofibrous carbon with herringbone structure as an effective catalyst of the H2S selective oxidation. Carbon 48:2004–2012. https://doi.org/10.1016/j.carbon.2010.02.008 17. Ba H, Duong-Viet C, Liu Y, Nhut J-M, Granger P, Ledoux MJ et al (2016) Nitrogen-doped carbon nanotube spheres as metal-free catalysts for the partial oxidation of H2S. C R Chim 19:1303–1309. https://doi.org/10.1016/j.crci.2015.09.022 18. Duong-Viet C, Liu Y, Ba H, Truong-Phuoc L, Baaziz W, Nguyen-Dinh L et al (2016) Carbon nanotubes containing oxygenated decorating defects as metal-free catalyst for selective oxidation of H2S. Appl Catal B 191:29–41. https://doi.org/10.1016/j.apcatb.2016.03.018 19. Ismagilov ZR, Shikina NV, Kruchinin VN, Rudina NA, Ushakov VA, Vasenin NT et al (2005) Development of methods of growing carbon nanofibers on silica glass fiber supports. Catal Today 102–103:85–93. https://doi.org/10.1016/j.cattod.2005.02.009 20. Popov MV, Zazhigalov SV, Larina TV, Cherepanova SV, Bannov AG, Lopatin SA et al (2017) Glass fiber supports modified by layers of silica and carbon nanofibers. Catal Sustain Energy 4:1–6. https:// doi.org/10.1515/cse-2017-0001

123