ene Chemical Engineering Science, 1968, Vol.23, pp.537-550. Pergamon Press. Printed in Great Britain. Controlling mechanisms in benzene oxidation K, VAIDYANATHAN and L, K, DORAISWAMY Nations Chemical Laboratory, Poona &, India (inst received 27 June 1967; in revised form 10 November 1967) Abstract—The vapour-phuse oxidation of benzene has been the subject of several previous studies, ‘but the controlling mechanisms in this reaction have not been fully elucidated. In the present work, based on a scheme in which benzene is simultaneously and consecutively (through maleic anhydride) ‘oxidised to the final products of combustion (carbox dioxide and water), suitable first-order equations have been set up for the three reactions on a molybdenum-vanadium catalyst. The experimental results indicate that there isa definite shift in the controlling mechanism at about 350°C as evidenced by the change in the activation energy from about 20 kcal/g mole in the lower temperatare range to about 2 Kcall/g mole in the higher temperature range. A detailed analysis of the reaction in the highet temperature region has, however, shown that neither pore diffusion nor external mass transfer is controlling. It hs therefore been postulated that this change is probably due to a change in chemical ‘mechanism and not toa shift from the chemical toa diffusion regime. ares INTRODUCTION SEVERAL studies on the catalytic vapour-phase oxidation of benzene have been reported, but only a few of them relate to the kinetics of the ‘oxidation process. It has been suggested that the reaction proceeds through a number of inter- mediate steps including the formation (succes- sively) of phenol[2,3], hydroquinone(2, 4] and quinone[2, 4,5]. Traces vf formaldehyde and diphenyl have also been reported among the products{6]. However, for the kinetic analysis of the reaction, the following scheme has normally been employed{1, 8, 9]: () benzene +0,—!> maleic, carbon , water | anhydride dioxide (17) 3) @ | carbon , water carbon , water dioxide” (W) dioxide” (W) i) Hammar[7] has studied the oxidation of ben- zene over a number of catalysts consisting of 205 and MoO, on metallic aluminium. He has interpreted the data according to the above scheme. The rate of the first step (r,) is said to be very fast, being limited only by the mass transfer rate. The fact that the activation energy for this step was of the order of 25 kcal/g mole, which is inconsistent with mass transfer control, tas been explained by assuming that the active area of the catalyst available for the reaction is itself temperature dependent. ‘The rate of the total combustion of benzene (74) was found to be almost independent of mass transfer, the rate determining step being either re-adsorption of maleic anhydride or surface reaction. It is difficult to visualise the re-adsorp- tion of maleic anhydride as a possible controlling step since maleic anhydride is not involved in the direct total combustion of benzene to carbon dioxide and water. ‘The three kinetic steps were found by Hammar to be uninhibited first-order reactions with equal apparent activation energies. Ioffe and Lynbarskii(8] also found the oxidation velo- city of benzene to be first-order, but the activa- tion energies of the three steps were found to be unequal. At oxygen concentrations >10 per cent, the benzene oxidation was found to be independent of oxygen concentration in the ‘gas phase[8, 10]. Almost all the results reported in the literature have been presented in terms of first-order Kinetic constants for the three significant 3537steps considered. Among the few attempts to describe a plausible model is that of Hayashi and collaborators[12], who have proposed a steady-state model in which the rate of adsorp- tion of oxygen on the catalyst surface equals the rate of reaction with benzene (not adsorbed) on the surface. A similar model has been proposed by Mars and van Krevelen{13]. On the other hand, Bretton ¢7 al.{14] believe that benzene is adsorbed and reaction occurs when oxygen from the gas phase strikes the adsorbed benzene molecules. ‘The object of the present work is to examine fully the Kinetics of this complex reaction on a selected catalyst, and to elucidate the effect (if any) of mass transfer and pore diffusion. EXPERIMENTAL Apparatus A diagramatic sketch of the reactor system is shown in Fig. 1. The vaporiser (A) consisted Fig. 1, Experimental set-up. 538 K. VAIDYANATHAN apd L. K, DORAISWAMY of a narrow graduated tube with @ bulb; a porcelain bead at the neck of the bulb prevented benzene from being carried away. Metered secondary air was mixed with metered primary air saturated with benzene from the vaporiser in the mixing bulb (F). The preheater (B) was electrically heated to prevent condensation of the organic vapour. The mixed preheated streain was passed down the reactor (C) contain- ing the catalyst and heated by s fluidized bed of bauxite which in turn was heated by electrical resistance wire through a voliage stabilizer and autotransformer, The temperature in the pre- heater and reactor was measured by chromel~ alumel sliding thermocouples. The products were condensed in receivers (1D) and the tail gas dried and passed through carbon dioxide absorbers. Procedure ‘Thiophene-free benzene and air from a com- pressed air cylinder were used. A~VAPORISER B~PREHEATER C-REACTOR D-RECEIVERS FOR SOLID PREDICTS E-GAS ABSORBERS F- MIXING BULBa a ‘Controlling mechanisms in benzene oxidation ‘Air was passed (after drying) through the catalyst bed for 30-60 min prior to the intro- duction of benzene. Readings were taken only after a steady state of operation had been attained. Experiments were carried out over wide ranges of benzene to air ratio, residence time and temperatuse. The life of the catalyst was repeatedly tested and care was taken to ensure that deactivation of the catalyst was absent in all the runs. The different lots of catalyst used for varying the WIF values were taken from the same batch of the catalyst prepared, although results could be reproduced to within 1-2 per cent by using samples from different batches Product analysis Maleic anhydride was estimated by titration against standard alkali[17}. Benzoquinone was analysed by a method analogous to that sug- gested by loffe et al.[18]. This consisted in treating the solution containing benzoquinone with I> in the acid medium. The liberated 1, was then titrated against $,0,-- by an indirect method. Carbon dioxide was estimated by the conventional method of absorption in caustic bulbs. Catalyst A comprehensive review has been presented by Egloff[18] of the various catalysts used in the vapour-phase oxidation of benzene. Vanadium pentoxide in conjunction with molybdenum trioxide appears to be the most widely used catalyst In the present work a number of composi- ns (Table 1) using vanadium pentoxide and ‘Table 1. Details of catalysts employed Totaloxides Ratioof Surface Pore Catalyst content” —-V,0,t0 area (BET) volume No. (MoO, (ang) (aati) 1 4 - - 2 4 4 038 3 4 236 ol 4 25 o - 5 33 5 a molybdenum trioxide on silica gel were evaluated in order to establish the influence of the oxides composition on the kinetic parameters of the reaction. A good catalyst is obviously one which gives high values of kiki +k, and kik, which can be regarded as selectivity criteria for maleic anhydride formation. A typical method of catalyst preparation ( U4 per cent total oxides content with VO: MoO, = 1:1) is given below. 9-255 parts of ammonia meta-vandate (NH,VO,:mol.wt.117) and 8-343 parts of ammonium molybdate ((NH,)jMo;O.. 4H,0:mol.wt. 1235-95) were dissolved in 250 parts of hot 10 per cent oxalic acid solution (10 g of oxalic acid in 100 ml distilled water), To this solution 100 parts of sitica gel (carrier) were added. The mass was evaporated to almost dryness with stirring over a water bath. It was then heated in a furnace at 120-150°C. for one hour, the clusters formed broken up, activated in a mufle for 8hr at 400°C., and then sieved to the required size. In this paper only catalysts 2 and 3 have been considered as they were found to be superior to the other compositions tried (with respect to selectivity and relative cost) and comparable to the reported catalysts. A detailed evaluation of all the five compositions has been presented by Vaidyanathan [27]. The surface areas of catalysts 2 and 3 were determined by the BET method and the pore volumes by the usual displacement technique. The values obtained are recorded in Table 1 KINETIC ANALYSIS In the organisation of the experimental programme for determining the rate constants 6 the three principal reactions involved, the following ranges of the main variables were covered: Concentration of 0-61 x 10-# to 1-86 benzene x 10 atm (4-5-14 mm Hg) Time factor (W/F) Up io 400g/hrgmole ‘Temperature 310°-400°C. The contact time of the reaction has been defined in different ways, viz. as the free space available 539