10 pages

A Kinetic Study of the Thermal and Photochemical Partial Oxidation of Cyclohexane with Molecular Oxygen in Zeolite Y

of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
A Kinetic Study of the Thermal and Photochemical Partial Oxidation of Cyclohexane with Molecular Oxygen in Zeolite Y
  Journal of Catalysis  204,  440–449 (2001)doi:10.1006/jcat.2001.3403, available online at on A Kinetic Study of the Thermal and Photochemical Partial Oxidationof Cyclohexane with Molecular Oxygen in Zeolite Y R. G. Larsen, A. C. Saladino, T. A. Hunt, J. E. Mann, M. Xu, V. H. Grassian, 1 and S. C. Larsen 1 Department of Chemistry, University of Iowa, Iowa City, Iowa 52242 Received May 22, 2001; revised August 29, 2001; accepted August 29, 2001 The kinetics of the thermal and photochemical oxidation of cy-clohexane in zeolite Y were investigated using  ex situ  GC productanalysis and  in situ  FTIR and solid state NMR spectroscopies. Theresults show that cyclohexyl hydroperoxide, cyclohexanone, andcyclohexanol (and water) are formed during the thermal and pho-tochemical oxidation of cyclohexane in BaY. The overall percentconversion of cyclohexane decreases dramatically at cyclohexaneloadings of greater than 3 cyclohexane molecules per supercage.Pronounced deuterium kinetic isotope effects were observed forboth the thermal and photochemical cyclohexane oxidation reac-tions, indicating that a proton transfer step is a rate-limiting step inthe reaction mechanism. For the thermal oxidation of cyclohexaneinBaYandNaY,activationenergiesof62( ± 9)and85( ± 3)kJ/mol,respectively, were measured.  c  2001 Elsevier Science INTRODUCTION Thepartialoxidationofhydrocarbonsissignificanttothechemical industry because these oxidation reactions areused to convert petroleum hydrocarbon feedstocks intochemicals important in the polymer and petrochemical in-dustries. Liquid phase air oxidations are generally pre-ferred by the chemical industry because of the mild reac-tion conditions (1). However, conversions of the oxidationprocesses are typically very low in order to maintain highselectivity. This is necessary because the desired partial ox-idation products can easily be further oxidized under typ-ical reaction conditions. Current liquid phase methods forthe autooxidation of cyclohexane using soluble cobalt ormanganese catalysts only exhibit acceptable product selec-tivity ( > 80%) when cyclohexane conversion is very low( < 5%) (2, 3). Although more efficient processes areknown, economic and safety considerations strongly favoroxidants, such as air or oxygen for use in large-scale pro-duction, and have lead to the large-scale adoption of a costeffective, yet seemingly inefficient production process.The inefficiency associated with low conversion and theneed to separate catalyst from products has motivated thesearch for solid catalysts that are active for the oxidation 1 To whom correspondence should be addressed. of cyclohexane (4–10). Frei and coworkers demonstratedthat the oxidation of cyclohexane to cyclohexanone withmolecular oxygen occurs with very high selectivity undermild thermal and photochemical conditions in cation-exchanged zeolites, such as NaY (10–12). Figure 1 shows aY zeolite with one molecule of adsorbed cyclohexane andillustrates the relative sizes of cyclohexane and zeolite Ypores. An additional benefit of this approach is that theoxidation of cyclohexane by molecular oxygen in NaYis desirable from an environmental perspective; the highselectivity minimizes waste, and the process utilizes theclean and inexpensive oxidant, molecular oxygen.Frei and coworkers proposed a reaction mechanism fortheoxidationofcyclohexaneinNaYthatinvolvedachargetransfer complex, [(cyclohexane) + O − 2  ], as follows (10,12–15):. [1]The hypothesis is that the charge transfer complex is sta-bilized by the exchangeable cation in the zeolite and thatthis stabilization allows access to the charge transfer stateby visible light irradiation or by thermal activation. In thenext step of the reaction, a proton from the cyclohexanecation radical is abstracted by O − 2  to form HO 2  and a cyclo-hexyl radical. The HO 2  radical attacks the cyclohexyl rad-ical to form cyclohexyl hydroperoxide under both photo-chemical and thermal conditions. Using Fourier transforminfrared (FTIR) spectroscopy, the formation of cyclohexylhydroperoxide, which reacts thermally to form cyclohex-anone and water, was observed (10). Complete selectivityinthephotooxidationofcyclohexanetocyclohexanonewasreported at conversions as high as 40%, based on  in situ FTIR measurements.Since the stabilization of the charge transfer state iscrucial to this reaction, the cation identity is thought to 0021-9517/01 $35.00c  2001 Elsevier ScienceAll rights reserved 440  CYCLOHEXANE OXIDATION IN ZEOLITE Y  441 FIG.1.  A zeolite Y supercage with one molecule of adsorbed cyclo-hexane. play a defining role in the zeolite reactivity. Vanoppen andcoworkersstudiedthermalcyclohexaneoxidationreactionsin cation-exchanged Y zeolites, CaY, SrY, BaY, and NaY(16, 17). They observed the following trend in reactivityfor cyclohexane oxidation to cyclohexanone in gas and liq-uid phase reactions: CaY > SrY > BaY > NaY. The initialstudy by Vanoppen  et al  . (16), in which gas phase cyclo-hexane was absorbed into cation-exchanged zeolites in thepresence of oxygen, concluded that, based on the cationdependence, the results were consistent with the mech-anism proposed by Frei involving the formation of analkane–oxygen charge transfer complex. However, in alater study that focused on cyclohexane oxidation usingliquid cyclohexane, it was found by analysis of the re-action kinetics that a classical autoxidation process wasmost compatible with the data (17). It was suggested bythe authors that the reactivity in the presence of a liq-uid phase was distinct from that of the gas phase system.Further, it was suggested that this difference could be at-tributed to the lower mobility of the reactant moleculeswithin the zeolite pores at the high loadings characteris-tic of the liquid phase system (17). The results of thesestudies raise questions about the mechanism of the cy-clohexane oxidation reaction at low loadings in the zeo-lite, typical of gas phase reaction conditions, and abouthow the reaction mechanism changes as a function of loading.The objective of the current study was to examine the ki-neticsofthethermalandphotochemicaloxidationofcyclo-hexane in BaY and NaY using  ex situ  GC product analysisand  in situ  spectroscopic methods, such as Fourier trans-form infrared (FTIR) and solid state nuclear magnetic res-onance (NMR), at various loadings. Although there arenow several studies that support the hypothesis that theelectric field at cation sites in zeolites facilitates the exci-tation of a O 2 · hydrocarbon charge transfer, the evidencealso suggests that the overall forward reaction is not thesolegoverningfactor(12,16–21).Forexample,ifthechargetransfer state were the sole governing factor, the reactionthreshold would be expected to be linearly related to theionization potential (IP) of the hydrocarbon. However, anonlinear relationship between the reaction threshold andthe IP was observed by Frei and coworkers and has beenattributed to differences in reaction quantum efficiencies(14). These changes in quantum efficiencies could be duein part to the kinetic competition between an electrontransfer back reaction and the proton transfer between thecation radical and superoxide ion in the charge transferstate.Specifically, the impact of the proton transfer rate onthe oxidation process and the effect of cyclohexane load-ing were examined in this study. Kinetic isotope studieswith deuterated cyclohexane were undertaken to deter-mine whether the proton transfer step enters into the over-allforwardrate.Forsolutionphasephotoinitiatedelectron-transfer reactions involving the formation of alkylbenzenecation radicals, it has recently been shown that a deuteriumisotope effect results when the rate of deprotonation iscompetitive with the rate of back electron transfer (22–24).Kinetic isotope effects between 1.5 and 5.6 have been ob-served,dependingonthealkylbenzeneanddonormoleculeused (22–24).Using mixtures of normal and perdeuterated reactantmolecules, kinetic studies of the thermal (35, 45, 55, 65,and 75 ◦ C) and photochemical oxidation of cyclohexane inBaY were conducted with  ex situ  product analysis usingGC and GC/MS. In every case examined, the measuredratio of normal to deuterated reaction products, in the ini-tial rate regime, showed a pronounced isotope effect. Theeffect of cyclohexane loading was also examined to estab-lishalinkbetweenearlierFTIRstudiesofFreiandcowork-ers (10) and the liquid phase oxidation study of Vanoppenandcoworkers(17).Theactivationenergiesforthethermaloxidation of cyclohexane in BaY and NaY were measured.  Insitu NMRandFTIRspectroscopieswereusedtomonitorcyclohexane oxidation in the zeolite pores. EXPERIMENTAL Zeolite Sample Preparation BaY was prepared from NaY (Aldrich) by standard ion-exchangeproceduresat90 ◦ Cusinganaqueous0.5MBaCl 2 solution. The elemental composition of Al, Si, and Ba wasdeterminedbyinductivelycoupledplasma/atomicemissionspectroscopy(ICP/AES)usingaPerkin–ElmerPlasma400.The Si/Al and Ba 2 + /Al ratios for BaY were 2.4 and 0.33,respectively.  442  LARSEN ET AL. Ex Situ Product Analysis with GC and GC/MS Zeolites were activated in vials by heating on a vacuumrack to 300 ◦ C overnight to remove adsorbed water. Theactivated samples contained in vials were then placed in aglove bag filled with oxygen. The sample vials were cappedand removed from the glove bag, and the desired amountof cyclohexane was injected into the vial through the sep-tum in the cap. Control experiments, which used no zeolitein the sample vial, were also conducted. The samples wereeither heated in a water bath or forced air convection ovenfor thermal reactions or irradiated with a 500-W mercurylamp (Oriel Corp.) for photochemical reactions. A broad-band long-pass filter was placed in front of the lamp forvisible light excitation (Oriel Corp. filter 59472, % T   = 0 at400 nm). After the reactions were completed, acetonitrilewas added to the sample through the septum. The samplewasstirredfor90min.andcentrifugedfor2min.at10 , 000 g .The supernatant was then analyzed by GC or GC/MS withan FID detector and a 5% phenyl/95% methylpolysiloxanecapillary column. When available, standards of the prod-ucts were injected separately to determine retention timesand response factors.The calibration for quantitative analysis of the data wasdone in two ways. First, standard solutions of cyclohexane,cyclohexanone, and cyclohexanol in acetonitrile were pre-pared and a calibration curve was constructed. In addition,a second calibration curve was constructed by addingspecific amounts of cyclohexane, cyclohexanone, or cyclo-hexanol to activated BaY. Then the hydrocarbons wereextracted from BaY with acetonitrile in order to calibratethe amount of each compound that was extracted from thezeolite.Thecalibrationwithzeoliteindicatedthat ∼ 90%of theadsorbedcyclohexane,cylohexanone,andcylcohexanolwasextractedfromBaY.Cyclohexylhydroperoxidewasas-sumedtohavethesamecalibrationcurveastheoxygenatedproducts, cyclohexanol, and cyclohexanone. Using thesetwo calibration curves, the mass balance was determinedto range from approximately 70 to 85% in the experimentsreportedhere.Oneexplanationforthemassbalanceisthatcyclohexane is lost either through leakage from the septumof the sample vial or by volatilization during the extractionprocess. Initial rates were calculated using kinetic dataobtained at cyclohexane conversions of less than 15%.  In Situ Product Analysis with  FTIR  and Solid State  NMR  Spectroscopies In situ  FTIR spectra were recorded with a Mattsoninfrared spectrometer equipped with a narrowband MCTdetector. Each spectrum was taken by averaging 500 scansat an instrument resolution of 4 cm − 1 . The infrared samplecell used in this study has been described previously (25).Briefly, ∼ 50–100mgzeolitewascoatedontoaphoto-etchedtungsten(3 × 2cm 2 )gridheldfromawaterslurry.Thetung-sten grid coated with the zeolite was mounted onto nickel jaws that are attached to a copper feed-through. The sam-ple can be heated to 900 ◦ C. The temperature of the samplewas measured with a thermocouple wire spotwelded to thecenter of the grid. The entire assembly was mounted insideof the IR cell, a 2 34  stainless steel cube with BaF 2  windows.The IR cell was then evacuated by a turbomolecular pumpto a pressure of 1 × 10 − 7 Torr. Zeolites were heated undervacuum at 300 ◦ C overnight to remove adsorbed water.Cyclohexane was loaded into the zeolite by adsorptionunder an equilibrium vapor pressure of the liquid at roomtemperature. The excess hydrocarbon was pumped out for5 min and oxygen was then added at a pressure of approxi-mately600Torr.A500-Wmercurylamp (OrielCorp.)witha water filter was used as the light source for photolysis. Abroadband long-pass filter was placed in front of the lampfor visible excitation (Oriel Corp. filter 59472, % T   = 0 at400 nm).The  13 C NMR spectra were obtained using a wide-bore Bruker MSL-300 NMR spectrometer operating at75.470 MHz. Cyclohexane was loaded onto BaY ( ∼ 80 mg)and the sample was transferred to a 7.5-mm zirconia rotorin a glove bag containing oxygen. The sample was thenheated to 85 ◦ C for 1 h. A Chemagnetics double-channel7.5-mmpencilmagicanglespinning(MAS)probewasusedto spin the sample at  ∼ 5 kHz at the magic angle. Crosspolarization (CP) with high-power proton decoupling wasused for  13 C NMR signal acquisition with the following pa-rameters: CP, contact time 2.0 ms, recycle delay 2.0 s, 90 ◦ pulse length 5.1  µ s. All of the chemical shifts for  13 C arereported relative to TMS. Reagents and Gases Cyclohexane (Aldrich, 99% purity), cyclohexane- d  12 (Cambridge Isotopes, 99%), cyclohexanone (Aldrich), andcyclohexanol (Aldrich) were used in these experiments.Acetonitrile (HPLC grade) was obtained from Fisher. Allofthesecompoundswereusedwithoutfurtherpurification.O 2  (Air Products, 99.6% purity) was also used without fur-ther purification. RESULTS Effect of Cyclohexane Loading on the Thermal Oxidation Reaction in BaY  The cyclohexane loading in BaY was varied in order todetermine whether the cyclohexane loading level affectsthe thermal conversion of cyclohexane and oxygen to thethree observed products, cyclohexanone, cyclohexanol,and cyclohexyl hydroperoxide. BaY samples with cyclo-hexane stoichiometries varied from 0.2 to 5 cyclohexanemolecules/supercage were heated to 65 ◦ C for 1 h in anoxygen atmosphere and then the products were extractedfrom the zeolite and analyzed by GC. A graph of percent  CYCLOHEXANE OXIDATION IN ZEOLITE Y  443 FIG. 2.  Graph of the percent conversion of cyclohexane vs cyclohex-ane loading per supercage for the thermal oxidation of cyclohexane withmolecular oxygen in BaY. The reaction conditions were 65 ◦ C for 1 h. conversion vs cyclohexane loading per supercage is pre-sented in Fig. 2. The percent conversion is defined as thetotal moles of product (cyclohexanone  +  cyclohexanol  + cyclohexyl hydroperoxide) divided by the total moles of cyclohexane added times 100. The percent conversion of cyclohexane is approximately constant at  ∼ 17% up to aloadingof2moleculesofcyclohexane/supercage.Athigherloadings, the percent conversion decreases to almost 0at a level of 4 to 5 molecules of cyclohexane added persupercage. Kinetics of the Thermal Oxidation of Cyclohexaneand Oxygen in BaY  Thekineticsofthethermaloxidationofcyclohexaneandoxygen in BaY were monitored as a function of tem-perature. For each experiment, a 1:1 molar mixture of cyclohexane- h 12  and cyclohexane- d  12  was injected intoidentical vials containing activated BaY and oxygen. Thesamples were equilibrated to the desired temperature andwere then allowed to react for specific time periods beforebeing quenched. The products and unreacted cyclohex-ane were extracted in acetonitrile and analyzed by GC.Representative kinetic plots for the thermal oxidation of amixture of cyclohexane-H 12 , cyclohexane-D 12 , and oxygenin BaY at 65 ◦ C are shown in Figs. 3 and 4. The cyclohexaneloading is 1 cyclohexane molecule per supercage. Theproduct distributions for cyclohexanone, cyclohexanol,and cyclohexyl hydroperoxide formed versus time areshown in Fig. 3. The product distribution is defined asthe amount of each product divided by the total productsand multiplied by 100. In the first 10 min of the reac-tion, the amounts of cyclohexanol, cyclohexanone, andcyclohexyl hydroperoxide increase. The percentage of cyclohexanol remains approximately constant at 9% fortimes 10 min and greater. At times greater than 1 h, thepercentage of cyclohexanone increases as the percentage FIG. 3.  Product distribution (%) vs time for the oxidation of cy-clohexane (50% cyclohexane-D 12 , 50% cyclohexane-H 12 )  and oxygenin BaY. The reaction temperature was 65 ◦ C and the loading was1 cyclohexane/supercage. of cyclohexyl hydroperoxide decreases as would be ex-pected if cyclohexyl hydroperoxide was an intermediatethat decomposed to form cyclohexanone as shown inreaction [1].The percent conversion of cyclohexane-H 12  and cyclo-hexane-D 12  with time at 65 ◦ C at a loading of 1 cyclohexanemolecule per supercage is shown in Fig. 4. Two aspects of the competitive kinetics displayed in this figure are note-worthy. First, in any given time interval more hydrogen-containing products than deuterium-containing productsareobtained.Second,thedecreaseinthereactionratewithtime shows a common rate of decrease for both isotopiclabels.The first observation is a clear indication of a substan-tial deuterium kinetic isotope effect in this system. Us-ing the ratio of hydrogen- to deuterium-labeled products,at total conversions of less than 15%, the kinetic isotope FIG. 4.  Graph of the total percent conversion of cyclohexane (50%cyclohexane-D 12 , 50% cyclohexane-H 6 )  to cyclohexanone, cyclohexanol,and cyclohexylhydroperoxide vs time. The cyclohexane loading was1 cyclohexane/supercage and the reaction temperature was 65 ◦ C.  444  LARSEN ET AL. TABLE 1 Reaction conditions Kinetic isotope effectThermal, 35 ◦ C 5.3Thermal, 45 ◦ C 5.5Thermal, 55 ◦ C 5.5Thermal, 65 ◦ C 5.7Thermal, 75 ◦ C 5.4Photo, 400 nm 5.7 effect was calculated. The magnitude of the deuteriumkinetic isotope effect for the oxidation of cyclohexane withoxygen in BaY at 65 ◦ C is 5.5. The deuterium isotope ef-fects were measured at 35, 45, 55, 65, and 75 ◦ C and arelisted in Table 1. The results span the range from 5.3 to5.7 with an average of 5.5 ( ± 0.2). The presence of a sub-stantial deuterium kinetic isotope effect indicates that aC–H (C–D) bond is broken in the rate-determining stepof reaction [1]. The partitioning between the primary ver-sus secondary deuterium isotope effects will be discussedfurther under Discussion. The observation of the overallslowdown of the kinetics as the reaction proceeds is mostlikely due to the inhibition of the catalyst by the products.Since this inhibition depends only on the chemical identityof the products and not on their isotopic substitution, theresult is an overall slowing of the reaction kinetics commonto both isotopic species under these competitive reactionconditions.The reaction is further characterized by using thetemperature-dependent rate data to construct an Arrhe-nius plot, which is shown in Fig. 5. The initial rate at con-stant initial loading was determined using the linear rangeof the kinetic plots, which corresponded to conversions of less than 15%. Linear regression of the data yields the FIG. 5.  Arrhenius plot of 1 / T   vs ln(initial rate) for the thermal oxi-dation of cyclohexane and oxygen in BaY. The cyclohexane loading was1 cyclohexane/supercage and the reaction temperature was varied from35 to 75 ◦ C. FIG. 6.  13 C CP/MAS with proton decoupling NMR spectrum of BaYwithcyclohexaneandoxygenafterheatingto85 ◦ Cfor1h.Thecyclohexaneloading was approximately 1 cyclohexane/supercage. activation energy,  E  a , which is calculated from the slopeof the best-fit line. For these experiments, the loading wasconstant at 1 molecule of cyclohexane per supercage. Theactivation energy for the reaction of cyclohexane and oxy-gen and BaY is determined to be 62( ± 9) kJ/mol. Theactivation energy for the thermal oxidation of cyclohex-ane in NaY was also measured and was determined to be85 ( ± 3) kJ/mol.  Insitu solidstateNMRwasusedtodirectlymonitorprod-uct formation in the pores of BaY after the thermal oxida-tionofcyclohexane.Figure6showsthe 13 CsolidstateMASNMR spectrum of cyclohexane (natural abundance  13 C)andoxygenonBaYafterreactingfor1hat85 ◦ C.Thecyclo-hexane loading was 1 cyclohexane molecule per supercage. 13 C is the only isotope of carbon that possesses a nuclearspinanditsnaturalabundanceis1.1.Aftersignalaveragingfor approximately 10 h, the  13 C NMR spectrum (usingcross polarization and proton decoupling) shown in Fig. 6was obtained. As expected from the GC analysis, severalproductswerepresentinthezeolitepores.TheC-1carbonsof cyclohexanone, cyclohexanol, and cyclohexyl hydroper-oxide are easily identified by peaks at 231 ppm (C==O),74 ppm (C–OH), and 89 ppm (COOH), respectively (26).Althoughnotrigorouslyproportionaltoconcentrationdueto the cross polarization used for data acquisition, a com-parison of the C-2 carbon peaks of cyclohexanone (44 pm)and cyclohexanol (36 ppm) are consistent with the resultsof our  ex situ  GC analysis. These NMR results suggest thatthe higher ratio of cyclohexanol to cyclohexanone (0.5)found in our exsitu experiments relative to the ratio of Frei et al  . (10) in FTIR experiments (0.02) is not an artifactof the extraction procedure used in the  ex situ  analysis.It is possible that the cyclohexanol-to-cyclohexanonebranching ratio changed during the NMR experiment dueto thermal reactions of the hydroperoxide during the longsignal averaging required for the experiment. Substantialoverlap in the ring carbon region of the spectrum ( ∼ 25–33 ppm) makes interpretation of this region of the spec-trum difficult. A possible peak at 64 ppm is noted and iswithin the range expected for a carbon singly bonded to an
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!