A detail kinetic study on vapour phase oxidation of ...

1 downloads 0 Views 3MB Size Report
Today 33. (1997) 263–278. [6] T. Blasco, P. Concepcion, P. Grotz, J.M. Lopez Nieto, A. Martinez-Arias, J. Mol. ... Miro, E.A. Lombardo, React. Kinet. Mech. Catal.
Molecular Catalysis 440 (2017) 1–11

Contents lists available at ScienceDirect

Molecular Catalysis journal homepage: www.elsevier.com/locate/mcat

Research Paper

A detail kinetic study on vapour phase oxidation of diphenylmethane over mesoporous V-KIT-6 catalyst D. Santhanaraj ∗,1 , C. Suresh 2 , A. Selvamani, K. Shanthi Department of Chemistry, College of Engineering, Anna University Chennai, 600 025, Tamilnadu, India

a r t i c l e

i n f o

Article history: Received 21 November 2016 Received in revised form 16 July 2017 Accepted 17 July 2017 Keywords: Transition state theory Langmuir–Hinshelwood (LH) model V-KIT-6 catalyst Kinetics Mechanism

a b s t r a c t Mesoporous V-KIT-6 catalysts with different Si/V ratios of 25,50,75 and 100 were synthesized by hydrothermal method under the controlled acidic conditions and characterized by using low angle XRD, ICP-OES, Diffuse reflectance UV–vis, SEM and TEM analysis. The characterization results are confirmed that most of the vanadium was mainly incorporated in the form of an isolated tetrahedral environment with terminal V O bond units. The synthesized catalysts were tested for the vapour phase selective oxidation of diphenylmethane (DPM) using air as an oxidant. A detail kinetic study was carried out on V-KIT-6 (25) catalyst using Langmuir–Hinshelwood (LH) model. The combined studies of Langmuir–Hinshelwood (LH) model and transition state theory (TST) explain that the reaction follows a second-order rate expression with respect to surface coverage of DPM and molecular oxygen. The reaction data derived from the LH modeling is allowed us to calculate the true activation energy and activation entropy of the reaction. The nature of transition state was determined from transition state theory and possible reaction mechanism was proposed for this reaction. © 2017 Published by Elsevier B.V.

1. Introduction Catalytic partial oxidation reactions play an important role in the modern chemical industry for the production of many valuable intermediates in both fine and bulk chemical synthesis. In general, more than 60% of partial oxidized compounds are synthesized by conventional oxidation route using different types of supported metal catalysts [1]. Among the various types of oxidation reactions, the selective partial oxidation of alkyl substituted benzene into their corresponding ketone is a versatile route for the synthesis of several valuable intermediates in fine chemical industries [2]. In this connection, the selective oxidation of diphenylmethane (DPM) to benzophenone (BP) is considered to be an important reaction for the synthesis of many useful commodity chemicals in a large scale production. Literature reports are supporting that benzophenone (BP) is widely used as a component for the synthesis of perfumes and as a starting material for

∗ Corresponding author. E-mail address: [email protected] (D. Santhanaraj). 1 Present address: Department of Chemistry, Loyola College, Chennai 600 034, India. 2 Present address: Electrodics and Electrocatalysis Division, CSIR-Central Electrochemical Research Institute, Karaikudi 630 006, Tamilnadu, India. http://dx.doi.org/10.1016/j.mcat.2017.07.012 2468-8231/© 2017 Published by Elsevier B.V.

the manufacture of dyes, pesticides, drug-related compounds and also as optical filters [3,4]. Although homogeneous catalysts are effectively employed for oxidation reaction, they are not successfully utilized for industrial applications because of its toxic nature, requirement of stoichiometric amounts of catalysts and moreover, it is difficult to separate from the product mixture. In the past decay, there are many attempts have been made to replace the homogeneous catalytic process by suitable heterogeneous catalysts e.g. by incorporation or functionalizing the transition metal ions into the different microporous supports such as zeolites [5], AlPOs [6], and aerogels [7] etc., Nevertheless, microporous molecular sieves have yet another drawback of smaller pore opening, which restricts the approachability (Diffusion limits) of larger reactant molecules into the active sites. In this connection, researchers had taken the greatest efforts to synthesize mesoporous materials such as silica, transitional alumina, and pillared clays. In fact, in 1990, Yanagisawa et al. described the preparation of mesoporous silica with uniform pore size [8]. However, the arrangement of pores in these materials possess irregular in shape and broadly distributed in different sizes. Thus, a gap has been bridged by the discovery of M41S, periodic mesoporous organosilica and SBA-15 families, which have opened up new possibilities for preparing variety of catalysts with several modifications and successfully applied for various types of organic

2

D. Santhanaraj et al. / Molecular Catalysis 440 (2017) 1–11

transformations reactions [9,10] including Suzuki-Miyaura coupling [11], Base catalyzed reaction [12], S-Arylation reaction [13] and oxidative dehydrogenation of propane [14]. A group of mesoporous molecular sieves like MCM type materials containing a high surface area and larger pore diameter attracted much interest for many catalytic reactions, but the poor hydrothermal stability of these materials represents a serious limitation to their practical application especially in high-temperature reactors. Hence, recent reports are emphasizing that KIT-6 material (similar to smaller-pore MCM-48 silica) is a new combined micro- and mesoporous structure which possess higher hydro and thermal stability than MCM type materials [15,16]. Moreover, the unique 3-D mesoporous channels are expected to provide a direct access for guest molecule and thereby facilitating efficient diffusion throughout the pore channel without any pore blocking effect [17,18]. In the recent years, vanadium substituted molecular sieves are extensively used in the oxidative catalytic process [19–22]. Since, the terminal V O bond units are considered as a real active sites for many catalytic oxidation reactions [22,23]. Hence, researchers had taken greatest effort to prepare vanadium containing mesoporous molecular sieves for many bulky molecular transformation [24–27]. In the recent past, there is a quantum of work have been done on the synthesis of vanadium containing mesoporous molecular sieves (MCM-41, MCM-48 and SBA-15) by using different methods such as ion exchange [28], grafting [29], direct hydrothermal synthesis [30], impregnation [31] and chemical vapour deposition [32]. Despite its potential applications, there are only few reports on the direct synthesis of vanadium into the KIT-6 material due to the difficulty in introducing transition metal ions into KIT-6 framework under highly acidic condition. Due to the above mentioned limitations, in the present investigation we introduced vanadium into the KIT-6 support with the different silica to vanadium ratios and are successfully employed for vapour phase oxidation reaction. A detail kinetic analysis is a powerful tool for understanding the kinetic and thermodynamic parameters that can be widely used to insight the reaction mechanism and has been successfully applied for many oxidation reactions [33,34]. However, there is no detail report on kinetic modeling studies using V-KIT-6 as a catalyst for the oxidation reaction. In this regard, we made an attempt to gain a deeper understanding the diphenylmethane oxidation reaction mechanism using a Langmuir-Hinshelwood model combined with transition state theory. The combinations of these two kinetic parameters are allowed to us determine the true activation energy and other thermodynamic parameters. In addition to that, the transition state theory provides the useful information about the nature of the activated complex in a transition state.

2. Experimental 2.1. Synthesis of Si-KIT-6 sample The mesoporous cubic KIT-6 silica material was obtained following the method reported by Ruthstein and Goldfarb [35]. The sample was prepared under acidic condition (1.84 M) in an aqueous solution using TEOS (Teraethylorthosilicate) as the silicon source and a triblock copolymer and n-butanol as a structure-directing agents. Briefly, 4.0 g of pluronic P123 was dissolved in 144 g of deionized (DI) water and 4.67 g of aqueous HCl (35 vol.%) under vigorous stirring. After complete dissolution, 4.8 g of n-butanol was added to the above solution. The mixture was left stirring at 308 K for one hour, after which 8.5 g of TEOS was added at once to the homogeneous clear solution. The stirring was continued for another 24 h at 308 K, followed by aging at 373 K. The resultant white solid product was filtered without washing and dried for 48 h at 368 K. The as-synthesized sample was then calcined in air at 773 K for

6 h with a heating ramp rate of 1 K min−1 to remove the template (P123) and any other organic contamination. 2.2. Synthesis of V-KIT-6 catalysts V-KIT-6 materials were prepared using triblock copolymer and n-butanol as the structure directing agent and TEOS as the silica source. Ammonium metavanadate was used as the metal source. Typically, 4.0 g of triblock copolymer was dissolved in 144 g of deionized water and stirred for 5 h 8.5 g of tetraethylorthosilicate (TEOS) along with 4.8 g of n-butanol (Co-surfactant) and the calculated amount of ammonium metavanadate (NH4 VO3 ) were added directly to the above homogenous solution to prepare different Si/V ratios of 25, 50, 75 and 100. Then, required amount of 0.3 M HCl were used to adjust the pH value of 3.2. The obtained gel was stirred for 24 h and then maintained at 373 K for another 48 h. The green solid product was filtered without washing and dried for 48 h at 368 K. The as-synthesized catalysts were then calcined in air at 773 K for 6 h. 2.3. Characterization techniques Several characterization techniques were employed to understand the properties and structure of catalysts, including ICP-OES, diffuse reflectance UV–vis spectroscopy (DRUV–vis), powder Xray diffraction pattern (XRD), N2 physisorption, 29 Si (Magic Angle Spinning) MAS NMR, scanning electron microscope (SEM) and high resolution transmission electron microscopy (HRTEM), as described below. 2.3.1. Elemental analysis (inductively coupled optical-emission spectroscopy) The amount of vanadium presented in KIT-6 samples was determined by ICP-OES technique (Perkin-Elmer OPTIMA 3000) at SAIF IIT-Madras laboratories. The samples were completely dissolved in a mixture of HF and HNO3 acids prior to the measurements. 2.3.2. X-ray diffraction (XRD) Powder XRD analysis was recorded on a BRUCKER D8 diffractometer using Cu K␣ radiation (␭ = 1.54 Å). The samples were scanned between 0.5 and 70 (2␪) in steps of 0.02 with an analyzing time of five seconds at each data point. 2.3.3. N2 adsorption-desorption analysis Nitrogen adsorption–desorption isotherms were measured with a BEL SORP mini II analyzer at liquid nitrogen temperature. Prior to the experiments, the catalysts were degassed at 573 K for 6 h. Surface area was calculated by the BET method (SBET ) whereas, the pore volume (VP) and pore size distributions were determined from the desorption isotherms by using the Barrett–Joyner–Halenda (BJH) method. 2.3.4. Diffuse reflectance ultra violet spectroscopy DRUV–vis measurements were carried out on a Shimadzu UV–vis spectrophotometer (2450) in the range of 190–800 nm using barium sulfate as the reference compound. 29 Si-MAS

NMR spectroscopy spectroscopy experiments were performed at Indian institute of Science (IISC) Bangalore on a Bruker DSX 300 spectrometer. The NMR spectra were collected at a frequency of 78.2 MHz with a spinning speed of 12 kHz and are cycle delay of 2 ␮s. 2.3.5.

29 Si-MAS

2.3.6. Scanning and transmission electron microscope The morphology of the samples was investigated by using scanning electron microscopy. The sample powder was placed into the

D. Santhanaraj et al. / Molecular Catalysis 440 (2017) 1–11

3

surface of the stub evenly and then coated with gold particles for 2 min using ion sputter coater and viewed on HITACHI ESEM attached with EDAX scanning electron microscope. High resolution transmission (HRTEM) electron microscopic images were recorded on JEOL 200 kV (IIT-Madras) for all the calcined Si-KIT-6 and V-KIT-6 (25) catalysts. 2.4. Catalytic activity test Vapour phase oxidation of diphenylmethane (DPM) was performed on a fixed-bed down flow tubular quartz reactor using CO2 free air as an oxidant, operating at atmospheric pressure in the temperature range of 593–683 K. Prior to the experiment, the reactor tube was loaded with 500 mg of catalyst with excess amount of inert silicon carbide (SiC) material and the temperature of tubular furnace was controlled by thermocouple which was attached inside the reactor tube. The reactant diphenylmethane was fed into the catalytic reactor with the help of syringe pump (Precision flow controller, Coimbatore) at a constant flow rate. The reaction products were analyzed by gas chromatograph (GC-17A, Shimadzu) equipped with a flame ionization detector, using a DB-5 capillary (0.18–0.32 mm I.D. 30 m length) column. The molar concentration of unreacted diphenylmethane and the product yield of benzophenone in the reaction mixture were quantified from the standard calibration curve using phenol as an external standard. The total mole balance of the reaction was always found to be 88% over the period of reaction. For comparison purpose, the reaction was also performed without any catalyst. The spent catalyst was regenerated at a temperature of 773 K for 6 h by continuous flowing of CO2 free dry air. The product yield was calculated as follows: mole of product formed × 100 Product yield (%) = mole of DPM fed

Fig. 1. Low angle X-ray diffraction patterns of Si-KIT-6 and V-KIT-6 catalysts.

Si-KIT-6

V-KIT-6 (25) V-KIT-6 (50)

2.5. Kinetic modeling

V-KIT-6 (75) Microsoft Excel software was used for kinetic modeling. A least square methodology was employed with the help of Solver tool to find the kinetic parameter values and that best fit with the experimental data. A nonlinear least-square regression analysis was used to differentiate between the experimental values and the kinetic model calculated values. The rate equation was expressed on the basis of conventional Langmuir–Hinshelwood (LH) model.

V-KIT-6 (100) Adsorption Desorption

3. Results &discussion 3.1. Characterization of catalysts The amount of vanadium in final calcined V-KIT-6 catalysts was verified by ICP-OES analysis and the obtained results were summarized in Table 1. Interestingly, there was no significant difference of vanadium content (Si/V ratios) in the gel and the final calcined materials, implies that successful incorporation of vanadium into Si-KIT-6 matrix under the controlled acidic conditions. The low angle X-ray diffraction technique was employed to understand the mesoporous nature of calcined pure Si-KIT-6 and V-KIT-6 catalysts with different Si/V ratios and the resultant XRD patterns are depicted in Fig. 1. The ordered three-dimensional (3D) cubic mesoporous features were identified by distinguished Bragg reflection indexed to (211) along with overlapped pattern of (220) plane; these signals represented as the well-ordered mesoporous structure with cubic mesoporous arrangement [36]. However, after modification of Si-KIT-6 material by adding vanadium ion the intensity of XRD signal was gradually decreased with increase in vanadium content from 100 to 25 (Fig. 1) showing a possible

Fig. 2. N2 adsorption-desorption isotherms of Si-KIT-6 and V-KIT-6 catalysts.

destruction of mesoporous arrangements. Furthermore, the d(211) spacing and the unit cell parameter of V-KIT-6 (Table 1) catalysts are relatively larger than that of parent Si-KIT-6 material indicating that the vanadium ions are well bonded into the silicon structure as reported elsewhere [37]. In order to detect the presence of V2 O5 crystals if any formed on catalysts, wide-angle X-ray diffraction analysis were performed in the 2␪ range of 10◦ to 70◦ and compared with XRD patterns of pure crystalline V2 O5 . In all the catalysts, we are not seen any diffraction patterns corresponding to polymeric vanadia crystal confirm that the synthesized catalysts are free from polymeric crystalline vanadia species. The textural properties of V-KIT-6 catalysts were determined from nitrogen sorption measurements and the corresponding values are summarized in Table 2. As shown in Fig. 2 illustrates the typical nitrogen adsorption-desorption isotherms of Si-KIT-6 and V-KIT-6 catalysts. According to the IUPAC nomenclature the isotherms can be classified as type IV [38] and all these sam-

4

D. Santhanaraj et al. / Molecular Catalysis 440 (2017) 1–11

Table 1 Content of vanadium on various mesoporous catalysts. b Relative peak area calculated from 29 Si MAS NMR signals.

a b

Catalysts

Si/V In gel

Vanadium (mol) in gel

Si/Va Calcined

Vanadium (mol)a

(Q3 +Q2 )/Q4b

Reaction Rate (mol h−1 gcat−1 )

V-KIT-6 (25) V-KIT-6 (50) V-KIT-6 (75) V-KIT-6 (100)

25 50 75 100

0.04 0.02 0.013 0.01

26 55 82 117

0.038 0.018 0.012 0.0085

0.73 0.79 0.82 0.89

0.0127 0.0098 0.0079 0.0069

Calculated from ICP-OES analysis. Relative peak area calculated from 29 Si MAS NMR signals.

Table 2 Textural properties of KIT-6 samples. Sample

Surface area (m2 /g)

Pore diameter (nm)

Pore Volume (cm3 /g)

Wall thicknessa (nm)

d(100) spacing (nm)

Unit cell parameter (nm)b

Si-KIT-6 V-KIT-6 (100). V-KIT-6 (75) V-KIT-6 (50) V-KIT-6 (25)

801 788 685 588 555

5.1 4.9 4.8 4.6 4.6

1.31 1.29 1.01 0.99 0.84

16.7 17.9 18.0 18.2 18.9

8.57 8.74 9.49 9.91 10.38

21.79 21.87 22.62 23.02 23.86

Number in parentheses indicate the nominal Si/V ratios. a Unit cell parameter values calculated using a0 = 61/2 d (211) . b Wall thickness = ao − Dp ; where Dp is the pore diameter.

265

1.6 1.4

385

V-KIT-6 (25) V-KIT-6 (50) V-KIT-6 (75) V-KIT-6 (100)

1.2 1.0 0.8 0.6 0.4 0.2 0.0

Fig. 3. DRUV–vis spectra of V-KIT-6 catalysts.

ples were showed a sharp capillary condensation between 0.6–0.8 indicating that the presence of narrow pore size distribution. Besides, incorporation of vanadium into the framework position significantly affects the textural properties of KIT-6 mesoporous materials. As presented in Table 2, the BET surface area, pore diameter and pore volume of all the catalysts decreased notably as the vanadium content increased from 100 to 25 due to the partial destruction of mesoporous structure and these results are inconsistence with XRD analysis. The decreasing behavior of pore diameter and pore volume as a function of vanadium content proven that a large amount of vanadium was located on outside the walls of mesopores channels. The co-ordination environment and oxidation state of vanadium ion in V-KIT-6 samples were studied by DRUV–vis spectroscopy and all the spectra were recorded in a hydrated state with different Si/V ratios are shown in Fig. 3. The existence of two intense UV strong bands at 260 and 370 nm corresponds to V5+ ion in tetrahedral and octahedral coordination environments, respectively. These intense bands are related to the low energy charge transfer bands

which were associated with O-V electron transfer [(␲)t2 (d)e] and [(␲)t1 (d)e] respectively [39]. However after incorporation of vanadium, the intensity of these two bands was gradually increased with decrease in Si/V ratio. The absence of band around 450 nm further proven that V-KIT-6 catalysts typically contain mainly tetrahedral coordinated isolated V5+ ions with terminal monomeric V O bonds. Hence, the UV results are concluded that vanadium oxide species are well dispersed on silicon matrix. 29 Si MAS NMR technique was used for calcined Si-KIT-6 and vanadium incorporated KIT-6 materials and the obtained results are shown in Fig. S2. The NMR spectra consist of two major intense chemical shifts at −110 ppm (Q4 ), −100 ppm (Q3 ) and a very weak signal at −92.1 ppm (Q2 ) corresponding to Si(OSi)4 , Si(SiO)3 OH and Si(OSi)2 (OH)2 units, respectively. As shown in Fig. S2, the intensity of −100 ppm (Q3 ) signal for V-KIT-6 catalysts decreased as the vanadium content increased from 100 to 25 as compared with a pure Si-KIT-6 sample, indicates that there was a possible interaction between surface silanol groups and vanadium ions [40,41]. The successful replacement of surface silanol groups by vanadium ions can be investigated by calculating the Q3 + Q2 /Q4 ratio from the relative intensity of NMR signal [42] and the results are summarized in Table 1. The decreasing behavior of Q3 + Q2 /Q4 ratio with respect to vanadium content suggests that during the incorporation significant quantity of surface silanol groups are replaced by vanadium ions on the external pore walls. The surface morphology of the calcined catalysts was examined by scanning electron microscopy technique and the observed images are shown in Fig. 4. The SEM images of the samples are clearly revealed that the sponge-like structures. However, these materials possess slight deviation of morphology with respect to vanadium content as compared to parent Si-KIT-6 material [43]. This behavior can be explained by the counterion effect where the concentration of Cl− counter ion increases with increase in vanadium content. Thus, the Cl− ion concentration has a lesser binding strength to the surfactant molecule and form less agglomerated particles in the case of V-KIT6 (25) catalyst [44]. Hence, these results are concluding that the concentration of vanadium source during the synthesis can alter the morphological patterns. In addition to that the TEM images are proven that KIT-6 samples are highly ordered lin-

D. Santhanaraj et al. / Molecular Catalysis 440 (2017) 1–11

5

Fig. 4. SEM images of calcined catalysts (A) Low and (B) High magnifications.

ear arrays of pores that are arranged in regular intervals and 3D (dimensional) bi-continuous cubic mesoporous structure (Fig. 5). However, there is no indication for the formation of bulk vanadium oxide species (VOx) on V-KIT-6 catalyst, which also confirms that the vanadium ions are well dispersed into the KIT-6 matrix. All the above characterization results are strongly emphasized that vanadium metal was successfully incorporated in an isolated environment with large quantity. The activity of these catalysts was

tested for vapour phase oxidation reaction in a fixed bed down flow reactor.

3.2. Catalytic measurement 3.2.1. Effect of W/F on vapour phase oxidation of DPM Fig. 6A–D shows the evolution of DPM conversion and BP product yield as a function of W/F at different temperatures 593, 623, 653 and 683 K over V-KIT-6 (25) catalyst.

6

D. Santhanaraj et al. / Molecular Catalysis 440 (2017) 1–11

Fig. 5. TEM images of calcined lower (A) and higher (B) magnifications of Si-KIT-6; lower (C) and higher (D) magnifications of V-KIT-6 (25) samples.

The symbol represents the actual experiment data, whereas the solid line represents the calculated values from Langmuir–Hinshelwood (LH) model. The detail kinetic study will be discussed below (section 3.2.3) in more detail. It can be seen from the figures that the DPM conversion increases with increase in W/F from 8.3 to 41.5 h and the major product was found to be BP which is a rapid oxidation of DPM on the surface of catalyst. Also the study was confirmed that there was no evolution of CO2 gas indicates the total oxidation reaction was almost controlled under the present experimental conditions. These results are inconsistence with Deviaka et al. [45] and they used CeAlPO-5 catalyst using molecular oxygen (Air) as an oxidant. In all the cases, the partial pressure of oxygen from air was calculated based on the mole balance of reaction as shown in Scheme 1.

3.2.2. Effect of time on stream recyclability In order to test the sustainability and reusability of synthesized catalysts, the effect of extended time on stream study was carried out on DPM conversion and BP product selectivity over the V-KIT-6 (Si/V = 25) catalyst under the identical experimental conditions. The catalytic activity measurement was carried out at constant temperature (593 K) with a molar ratio of 1:4 (DPM to molecular oxygen) and a constant W/F of 41.5 h. The obtained results are represented in Fig. 7. Interestingly, the DPM conversion and product selectivity was almost remained the same without any significant loss of activity due to catalyst deactivation over the period of 12 h. The constant profile of reactant conversion and product selectivity indicate that there was an absence of coke deposition or the formed coke

oxidized by catalyst itself during the reaction. The used catalyst after 12 h of the reaction was regenerated in a constant flow of air (75 ml/min) at 753 K for 4 h and tested for its reusability and we observed that the catalyst could be reproducible even after several runs. From the above study we concluded that the catalyst maintains an adequate level of sustainability and stability under the present reaction conditions.

3.2.3. Kinetic modeling Prior to a detail kinetic study on vapour phase oxidation of DPM, it was important to probe the catalyst with internal and external mass transfer limitations. The Koros-Nowak states that the rate of reaction is directly proportional to the concentration of active sites in a catalyst [46]. Madon and Boudart compared a series of supported metal catalysts with different metal loading for mass transfer limitation test, the idea was proposed more than 30 years ago and it has been used successfully in many studies on various types of supported metal catalysts [47]. Recently, Miguel and Resasco [48] reported that alkylation activity increases with increase in acid sites due to the absence of internal mass transfer limitations with respect to different acid site density on Zeolites and they successfully blocked the active sites by simple ion exchange method. However, the criterion has been much less common in transition metal incorporated mesoporous materials such as SBA-15 and MCM-41. In this present investigation, we varied the concentration of active sites on V-KIT-6 catalyst by varying silicon to vanadium ratio from 100 to 25 and used these catalysts to verify any internal mass transfer limitation effect. The final concen-

D. Santhanaraj et al. / Molecular Catalysis 440 (2017) 1–11

7

Fig. 6. Partial pressures of DPM, BP and O2 as a function of W/F at 2 h time on stream on V-KIT-6 (25) catalyst at 593, 623, 653 and 683 K. The points are experimental data, and the lines are fitted data from the Langmuir–Hinshelwood fitting model.

H2

O2

Scheme 1. Vapour phase oxidation of diphenylmethane (DPM).

tration of vanadium content in V-KIT-6 samples were determined by ICP-OES analysis and the obtained values are summarized in Table 1. Fig. 8A shows the specific reaction rate for vapour phase oxidation reaction with respect to Si/V ratios. Interestingly, reaction rates increases linearly as a function of vanadium content indicating that there was no significant internal mass transfer limitation in the mesoporous channels.

In a similar way to test external mass transfer limitations, the reaction was carried out at different space velocities in the reactor while keeping the same W/F. The space velocity was varied by changing flow rate of gas and keeping a constant amount of catalyst. As shown in Fig. 8B, the reaction rates increased from the flow of gas from 7.5 to 42.5 ml/min and then leveled off indicating that there is no external mass transfer limitation under the present

8

D. Santhanaraj et al. / Molecular Catalysis 440 (2017) 1–11

Fig. 7. Effect of time on stream over V-KIT-6 (25) catalyst. Reaction conditions: Temperature: 350 ◦ C; W/F = 6.

experimental conditions. Hence, all the kinetic parameters were studied at 40 ml/min to make a meaningful kinetic modeling. Based on above discussed qualitative analysis, a rate expression was derived from conventional Langmuir–Hinshelwood model, with the following assumptions: (1) Competitive molecular adsorption of Diphenylmethane (DPM), Benzophenone (BP), Water and oxygen over the same sites on the catalyst. (2) Rate inhibition of molecular water is neglected (3) Surface reaction is rate determining step.

with surface coverage for overall reactions can be written as follows: DPM = BP = O2 =

K

DPM DPM + ∗ ↔ DPM∗

KBP

BP + ∗ ↔ BP∗

(1) (2)

KO

O2 + ∗ ↔2 O2 ∗

(3)

KH O 2

H2 O + ∗ ↔ H2 O∗ k

DPM ∗ +O2 ∗ ↔BP ∗ +H2 O + ∗

(4) (5)

k is the reaction rate; KDPM , KO2 , KH2 O and KBP are the adsorption equilibrium constants for corresponding compound, Torr−1 . It is important to mention that Eq. (5) is a rate-determining step which provides the information about the activation barrier. This activation barrier described the energy difference between adsorbed species, DPM* and the transition state DPM. . .O2 (Scheme 2). Based on the assumption the overall reaction rate depends upon the adsorbed concentration of DPM and O2 on the surface of catalyst and the reaction rate can be expressed as: r = kDPM O2

(6)

where  DPM and O2 are surface coverage of adsorbed molecular DPM and O2 , which could be expressed based on the conventional Langmuir-adsorption isotherm equation. The rate equation along

(7)

KBP PBP 1 + KDPM PDPM + KBP PBP + KH2 O PH2 O + KO2 PO2

(8)

KO2 PO2

(9)

1 + KDPM PDPM + KBP PBP + KH2 O PH2 O + KO2 PO2

rDPM = Accordingly, the elementary steps for vapour phase oxidation of diphenylmethane can be described by the following equations, in which * represents an active site.

KDPM PDPM 1 + KDPM PDPM + KBP PBP + KH2 O PH2 O + KO2 PO2



kKDPM PDPM KO2 PO2 1 + KDPM CDPM + KBP CBP + KH20 CH20 + KO2 PO2

2

(10)

The adsorption (K) and rate (k) constants were used as adjustable parameters to optimize the better kinetic fitting for vapour phase oxidation reaction. Fig. 6A–D shows the evolution of partial pressures of reactant (DPM) and product (BP) as a function of W/F for three different temperatures (593, 623, 653 and 683 K) in an isothermal integral reactor and the results are good agreement between the experimental and calculated data for diphenylmethane (DPM) conversion and product yield. This can be clearly seen in Fig. 6A–D, where the solid line calculated from Langmuir-hinshelwood model fit well with experimental data points and the average degree of fitting (R2 ) reached more than 0.96. The resultant adsorption and rate constants at different temperatures are summarized in Table 3. In order to verify Langmuir–Hinshelwood model, the derived values are also used in Eley-Rideal model and we observed that Langmuir–Hinshelwood model are much better fitting than Eley-Rideal model and these results are inconsistence with previously reported by Jingjing Pei et al. [33]. From the kinetic model, it was observed that the rate constant values are increased as a function of temperature, whereas the adsorption constant (K) decreased dramatically. It is noteworthy that, the adsorption constant of DPM is significantly larger than BP, indicating that DPM is strongly adsorbed on the catalyst sites and it was stabilized by radical formation on the surface of the catalyst as shown in reaction mechanisms (Scheme 3). The ther-

D. Santhanaraj et al. / Molecular Catalysis 440 (2017) 1–11

H

TS

O

O

9

H

V

O O

O

O

56 kJ/mol H

O

O

H

V

O O

O

O O O

Reactants

O

H

H

V O

O

O

TS- Transition step-Rate limiting step Products

Scheme 2. Proposed transition state (TS) structures for vapour phase oxidation of Diphenylmethane (DPM).

Table 3 Optimized kinetic and thermodynamic parameter values for the oxidation of DPM obtained from Langmuir–Hinshelwood kinetic model. Temperature (◦ C)

Rate Constant (mol h−1 gcat−1 )

593 623 653 683

0.306 0.486 0.680 0.790

Hadsorption (kJ/mol) Sadsorption (J/mol K) Activation Enthalpy H (kJ/mol) Activation Entropy S (J/mol K)

Adsorption constant (Torr−1 ) KBP

KO2

KH2O

3.037 2.977 2.077 1.397

0.017 0.014 0.011 0.009

0.133 0.061 0.041 0.021

0.0095 0.0018 0.0009 0.0003

DPM

BP

H2 O

O2

−69.29 −124.99 21.91

−22.80 −37.25 –

−47.43 −113.50 –

−58.50 −111.53 –

−309.92







modynamic and kinetic parameters are expressed as a function of temperature according to the conventional Arrhenius k = Ae S

KDPM

−Ea RT

above, the rate-limiting step (5) can be described by the following equations:

−H

and Van’t Hoff equations, K = e R e RT respectively. Remarkably, the resulting enthalpies of adsorption (Heat of Adsorption) of reactant DPM are slightly higher than products, which follows the order of BP > O2 > H2 O. The possible reason might be due to the electron density of methylene group (DPM molecule) can effectively interact with Lewis acid sites of vanadium. Furthermore, the reaction rate ‘k’ obtained from LH model and it can be fitted better with Arrhenius equation. The resultant activation energy from the linear plot (Fig. 9) was calculated as 56 kJ/mol.

3.2.4. Nature of transition state theory (statistical thermodynamics) The transition state theory (TST) explains the reaction rates of elementary chemical reactions. The transition state theory (TST) assumes the occurrence of an activated complex that is in equilibrium (quasi-equilibrium) between the reactants and activated transition state complexes. The resultant activated complex was readily converted into products in a single vibration. As proposed

K#

␯#

DPM ∗ +O2 ∗ ↔ [DPM. . .O2 ] ∗ ↔BP ∗ +H2 O + ∗

(11)

From the above equation, the partial oxidation reaction rate was calculated from transition state theory and the rate will be equal to the concentration of the activated complex multiplied by the vibrational frequency at which a new Carbon Oxygen bond is formed. rateDPM = ␯# K # (DPM. . .O2 ∗)

(12)

where ␯# corresponds to the vibration of Carbon Oxygen bond being formed and the resulting equilibrium constant K# was calculated from statistical mechanics and it can be written as follows. K # = qasym.vib. K #’

(13)

in which qasym.vib. is the partion function of weak anti symmetric vibration mode along the reaction coordinate, leading to formation of new Carbon-Oxygen bond in the transition stat (Scheme 2).  The obtained equilibrium constant K# is known as transition state

10

D. Santhanaraj et al. / Molecular Catalysis 440 (2017) 1–11

H

O

O

H

V

O O

O

O

O2

H

O V O

O

O V

O

H

O

O

O

O

H

O O

H

O

O O

H

O

H

V O

O

O

Scheme 3. Possible reaction mechanism for the formation of BP from diphenylmethane (DPM).

Fig. 8. Initial rates as a function of vanadium concentration (A) and different space velocities (B); Reaction Conditions: W/F = 41.5 h, Reaction time: 2 h.

equilibrium constant. The resultant vibrational frequency in the transition state is completely excited and it can be expressed as qasym.vib. =

1 1 − e−h#/ kB T



kB T h#

(14)

Substitute Eq. (14) in Eq. (13) and it is modified as K# =

kB T #’ K h#

(15)

Similarly, substitute K# equilibrium constant Eq. in (12) and the final transition state equation gives the following expression. rateDPM =

kB T #’ K (DPM. . .O2 ∗) h

where K #’ = e−G/ RT

(16) (17)

Since, G = H − TS, the rate constant expression can be expanded and giving the temperature dependent Eyring Eq. (18). In which H and S are activation enthalpy and entropy respectively. rateDPM =

kB T S −H e R e RT (DPM. . .O2 ∗) h

(18)

Fig. 9. Arrhenius plot for the oxidation of DPM.

It is noteworthy that, LH model equation was used to fit the reaction rate measurements and rate constant includes the true enthalpy and entropy of activation for the C O bond formation with respect to surface adsorbed species. Therefore, the values of H and S in Eq. (18) refer to the differences between the H and S values of the transition state and those of the preferentially adsorbed reactants i.e., two adsorbed molecules on the surface (DPM and O2 ). The goodness of kinetic fitting data based on the second-order Langmuir–Hinshelwood model with respect to adsorbed species on the surface of catalyst are inconsistence with proposed mechanism that involves the hydroperoxo intermediate as shown in Scheme 3. Table 3 summarizes the activation enthalpy

D. Santhanaraj et al. / Molecular Catalysis 440 (2017) 1–11

and entropy of DPM in the transition state. It is clear that the positive value of activation enthalpy system indicates the increase in energy barrier for the rate limiting step. However, the activation entropy is becoming more negative (Entropy loss) due to loss of several vibrational and rotational modes of molecule in the transition state. Based on the complete kinetic analysis we proposed possible reaction mechanism for the oxidation of DPM as shown in Scheme 3. According to Langmuir-Hinshelwood mechanism, both DPM and O2 were preferentially adsorbed on the active sites and the reaction was believed to be proceeding via radical mechanism. In the oxidation reaction cycle, the hydrogen radical from DPM molecule is abstracted by activated oxygen on vanadium center and to form a Metal-alcoholate species. The resultant hydrogen radical shifted from the DPM molecule to oxygen radical represents the rate-determining step. Subsequently the formed transition state complex was decomposed into benzophenone and water on the surface. 4. Conclusions In summary, mesoporous Vanadium incorporated KIT-6 catalysts with different Silicon to vanadium ratios were synthesized by the direct hydrothermal method under the mild acidic condition.DRUV–vis results are confirmed that V-KIT-6 catalysts essentially contain tetrahedral coordinated isolated V5+ species with terminal monomeric V O bonds. In the present study, we have conducted a comprehensive kinetics analysis for vapour phase oxidation reaction over V-KIT-6 (25) catalyst. To gain, insight about the reaction pathway, a detail kinetic study was carried out by using Eley-Rideal and Langmuir–Hinshelwood models and it was found that experimental data values are better fit into Langmuir–Hinshelwood model. According to Langmuir–Hinshelwood model the reaction follows a second-order expression with respect to surface coverage of diphenylmethane (DPM) and molecular oxygen. From the kinetics regime, the heats of adsorption of DPM are significantly higher than those of the reaction products, that is, BP and H2 O. The resultant mechanistic study was proven that DPM preferentially adsorbed on the active sites. Acknowledgment The corresponding author would like to thank the Council of Scientific and Industrial Research (CSIR) of India for providing financial support (Sanction No. 9/468(0402)/2009–EMR-I). References [1] G.J. Hutchings, M.S. Scurrell, Cattech 7 (2003) 90–103. [2] T. Mehler, W. Behnen, J. Wilken, J. Martens, Tetrahedron: Asymmetry 5 (1994) 185–188. [3] G.T. Carroll, N.J. Turro, J.T. Koberstein, J. Colloid Interface Sci. 351 (2010) 556–560. [4] G. Dorman, G.D. Prestwich, Biochemistry 33 (1994) 5661–5673. [5] C. Robert Adams, L. Xu, K. Moller, T. Bein, W. Nicholas Delgass, Catal. Today 33 (1997) 263–278.

11

[6] T. Blasco, P. Concepcion, P. Grotz, J.M. Lopez Nieto, A. Martinez-Arias, J. Mol. Catal. A: Chem. 162 (2000) 267–273. [7] Y. Mizushima, M. Hori, Appl. Catal. A 88 (1992) 137–148. [8] T. Yanagisawa, T. Schimizu, K. Kuroda, C. Kato, Bull. Chem. Soc. Japan 63 (1990) 988–992. [9] N. Pal, A. Bhaumik, RSC Adv. 5 (2015) 24363–24391. [10] S. Rostamnia, E. Doustkhaha, RSC Adv. 4 (2014) 28238–28248. [11] E. Doustkhah, S. Rostamnia, H.G. Hossieni, R. Luque, Chem. Select 2 (2017) 329–334. [12] L.B. Sun, X.Q. Liu, H.C. Zhou, Chem. Soc. Rev. 44 (2015) 5092–5147. [13] S. Rostamnia, E. Doustkhah, R. Bulgar, B. Zeynizadeh, Microporous Mesoporous Mater. 225 (2016) 272–279. [14] Y.M. Liu, Y. Cao, K.K. Zhu, S.R. Yan, W.L. Dai, H.Y. He, K.N. Fan, Chem. Commun. (2002) 2832–2833. [15] X. Liu, B. Tian, C. Yu, F. Gao, S. Xie, B. Tu, R. Che, L. Peng, D. Zhao, Angew. Chem. Int. Ed. 41 (2002) 3876–3878. [16] B. Hu, H. Liu, K. Tao, C. Xiong, S. Zhou, J. Phys. Chem. C 117 (2013) 26385–26395. [17] B. Rabindran Jermy, D.R. Cho, K. Vijayan Bineesh, S.Y. Kim, D.W. Park, Microporous Mesoporous Mater. 115 (2008) 281–292. [18] K. Soni, B.S. Rana, A.K. Sinha, A. Bhaumik, M. Nandi, M. Kumar, G.M. Dhar, Appl. Catal. B: Environ. 90 (2009) 55–63. [19] X. Wang, Z. Guanglin, Z. Chen, W. Jiang, Z. Hongjun, M. Piumetti, B. Bonelli, P. Massiani, Y. Millot, S. Dzwigaj, L. Gaberova, M. Armandi, E. Garrone, Microporous Mesoporous Mater. 223 (15) (2015) 261–267, http://dx.doi.org/ 10.1016/j.micromeso.2015.06.029. [20] M. Piumetti, B. Bonelli, M. Armandi, L. Gaberova, S. Casale, P. Massiani, E. Garrone, Microporous Mesoporous Mater. 133 (2010) 36–44. [21] B. Yaohui, J. Hong, X. Weihong, C. Rizhi, F. Yiqun, S. Irusta, L.M. Comaglia, L.M. Miro, E.A. Lombardo, React. Kinet. Mech. Catal. 116 (2) (2015) 535–547 https://doi.org/10.1007/s11144-015-0898-0. [22] D. Santhanaraj, C. Suresh, M. Gurulakshmi, N.R. Sasirekha, K. Shanthi, React. Kinet. Mech. Catal. 104 (2011) 399–415. [23] D. Santhanaraj, C. Suresh, M. Gurulakshmi, N.R. Sasirekha, K. Shanthi, J. Porous Mater. 19 (2012) 1027–1036. [24] V. Pârvulescu, C. Anastasescu, B.L. Su, J. Mol. Catal. A: Chem. 198 (2003) 249–261. [25] S. Shylesh, A.P. Singh, J. Catal. 228 (2004) 333–346. [26] P. Hurtado, S. Ordonez, H. Sastre, F.V. Dıez, Appl. Catal. B: Environ. 51 (2004) 229–238. [27] M. Albert Vannice, Catal. Today 123 (2007) 18–22. [28] Q. Zhang, Y. Wang, Y. Ohishi, T. Shishido, K. Takehira, J. Catal. 202 (2001) 308–318. [29] G. Du, S. Lim, M. Pinault, C. Wang, F. Fang, L. Pfefferle, G.L. Haller, J. Catal. 253 (2008) 74–90. [30] J. Xu, W. Chu, S. Luo, J. Mol. Catal. A: Chem. 256 (2006) 48–56. [31] V. Fornes, C. Lopez, H.H. Lopez, A. Martinez, Appl. Catal. A: Gen. 249 (2003) 345–354. [32] M. Wark, M. Koch, A. Bruckner, W.J. Grunert, J. Chem. Soc. Faraday Trans. 94 (1998) 2033–2041. [33] J. Pei, X. Han, Y. Lu, Build. Environ. 84 (2015) 134–141. [34] M.M. Barsan, F.C. Thyrion, Catal. Today 81 (2003) 159–170. [35] S. Ruthstein, D. Goldfarb, J. Phys. Chem. C 112 (2008) 7102–7109. [36] X. Liu, B. Tian, C. Yu, F. Gao, S. Xie, B. Tu, R. Che, L. Peng, D. Zhao, Angew. Chem. Int. Ed. 41 (2002) 3876–3878. [37] H.T. Gomes, P. Selvam, S.E. Dapurkar, J.L. Figueiredo, J.L. Faria, Microporous Mesoporous Mater. 86 (2005) 287–294. [38] K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, R.A. Pierotti, J. Rouquerol, T. Siemieniewska, Pure Appl. Chem. 57 (1985) 603–619. [39] P. Selvam, S.E. Dapurkar, J. Catal. 229 (2005) 64–71. [40] S.C. Laha, R. Kumar, Microporous Mesoporous Mater. 53 (2002) 163–177. [41] D.H. Park, C.F. Cheng, H. He, J. Klinowski, J. Mater. Chem. 7 (1997) 159–162. [42] S. Shylesh, A.P. Singh, J. Catal. 233 (2005) 359–371. [43] R. Atchudan, J. Joo, A. Pandurangan, Mater. Res. Bull. 48 (2013) 2205–2212. [44] H.P. Lin, C.P. Kao, C.Y. Mou, S.B. Liu, J. Phys. Chem. B 104 (2000) 7885–7894. [45] S. Devika, M. Palanichamy, V. Murugesan, Chin. J. Catal. 33 (2012) 1086–1094. [46] R.M. Koros, E.J. Nowak, Chem. Eng. Sci. 22 (1967) 470–471. [47] R.J. Madon, M. Boudart, Ind. Eng. Chem. Fundam. 21 (1982) 438–447. [48] M.A. Gonzalez-Borja, D.E. Resasco, AIChE J. 61 (2015) 598–609.