TEOM part 1


of 8
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.
Chemical Engineering Science 59 (2004) 1131 – 1138 www.elsevier.com/locate/ces Adsorption and desorption kinetics of hydrocarbons in FCC catalysts studied using a tapered element oscillating microbalance (TEOM). Part 1: experimental measurements Chi Keng Lee, Sunil Ashtekar, Lynn F. Gladden, Patrick J. Barrie∗ Department of Chemical Engineering, University of Cambridge, Pembroke Street, Cambridge CB2 3RA, UK Received 15 October 2003; received in revised form 19 December 2003; accepted 6 Januar
  Chemical Engineering Science 59 (2004) 1131–1138www.elsevier.com/locate/ces Adsorption and desorption kinetics of hydrocarbons in FCC catalystsstudied using a tapered element oscillating microbalance (TEOM).Part 1: experimental measurements Chi Keng Lee, Sunil Ashtekar, Lynn F. Gladden, Patrick J. Barrie ∗ Department of Chemical Engineering, University of Cambridge, Pembroke Street, Cambridge CB2 3RA, UK  Received 15 October 2003; received in revised form 19 December 2003; accepted 6 January 2004 Abstract An experimental procedure to measure the adsorption and desorption kinetics of hydrocarbons in uid catalytic cracking (FCC) catalystsusing a tapered element oscillating microbalance is described. It enables adsorption rates to be measured on a timescale of about 1 s.Experiments using n -hexane, n -heptane, n -octane, toluene and p -xylene were performed on both a commercial FCC catalyst and a pure rare-earth exchanged zeolite Y sample under non-reacting conditions (temperatures of 373–473 K). Heats of adsorption for thesehydrocarbons are reported. The overall adsorption and desorption kinetics are found to depend on carrier gas owrate in cases whenadsorption is strong indicating that the length of the catalyst bed can have a signiÿcant inuence on the observed kinetics. However, athigh carrier gas owrates the overall adsorption and desorption kinetics do not depend on the amount of catalyst present. It is found thatthe rates of adsorption and desorption of the hydrocarbons studied are the same for the FCC catalyst as for the pure zeolite Y sample.This means that mass transport in the matrix component of the FCC catalyst is rapid and not a limiting step in the adsorption process forthe hydrocarbons studied in this work. ? 2004 Elsevier Ltd. All rights reserved. Keywords: Adsorption; Diusion; Zeolites; FCC catalysts; Microbalance; TEOM 1. Introduction The uid catalytic cracking (FCC) process in an oil re-ÿnery is a key unit operation in which heavy hydrocarbonmolecules are cracked to form smaller more valuable prod-ucts. The FCC catalyst particles employed contain zeolitecrystallites embedded in a matrix that contains alumina, clayand sometimes silica. The most common zeolite used israre-earth exchanged zeolite Y, and this component is re-sponsible for the high activity and selectivity of FCC cata-lysts. The matrix material has some catalytic activity (e.g.in cracking molecules too large to enter the zeolite) and provides attrition resistance. It also acts as a heat sink andcan reduce the extent of heavy-metal poisoning of the zeo-lite component (Venuto and Habib, 1979;Satterÿeld, 1980; Scherzer, 1991). ∗ Corresponding author. Tel.: +44-1223-331864;fax: +44-1223-334796. E-mail address: patrick barrie@cheng.cam.ac.uk(P.J. Barrie). There have been a large number of studies on the catalytic performance of zeolites for cracking reactions. For instance,these have explored the inuence of pore structure and Si/Alratio,characterisedthedierenttypesofacidsitethatmaybe present, investigated coking and deactivation, and proposedreaction mechanisms (Corma, 1993;Rabo and Gajda, 1989; Farneth and Gorte, 1995;Bhatia et al., 1989;Karge, 1991; Cumming and Wojciechowski, 1996;Corma and OrchillÃes, 2000). Molecular transport within zeolites has also been ex-tensively studied because it can inuence overall reactionrates, and also because of the use of zeolites in separations processes (Karger and Ruthven, 1992;Chen et al., 1994). There have, however, been very few published studies onadsorption and transport properties in commercial FCC cat-alyst particles that contain matrix material as well as zeolitecrystallites.While many workers believe that the cracking reaction inan FCC unit is not limited by diusional eects (e.g.Chenet al., 1994), there have been suggestions that pore diusionwithin zeolite Y does inuence the observed activity, even 0009-2509/$-see front matter ? 2004 Elsevier Ltd. All rights reserved.doi:10.1016/j.ces.2004.01.005  1132 C.K. Lee et al./Chemical Engineering Science 59 (2004) 1131–1138 for molecules as small as n -hexane (Williams et al., 1999;Kung et al., 2000). For this reason, it is important to measurethe molecular transport rates of hydrocarbons in FCC cata-lysts. Knowledge of the adsorption and desorption kineticsis important in the FCC process not only in understandingthe chemical reactions occurring, but also in the design of the riser (in which adsorption of heavy compounds could in principle be a limiting step) and in the design of the strip- per (in which desorption of hydrocarbons of value is desired before the catalyst is sent to the regenerator).While it is expected that the eective diusion coecient,  D e , in the zeolite component is smaller than that in the ma-trix component, the particle radius, a , of the zeolite crystal-lites is about one hundred times smaller than the radius of the overall FCC catalyst particle. Because the eective rateof transport depends on the value of  D e =a 2 for each com- ponent present, it is therefore important to see whether thelimiting step for molecular transport in FCC catalyst parti-cles occurs in the matrix or zeolite component. Another un-certainty when considering molecular transport in FCC cat-alysts is that there remains signiÿcant disagreement in theliterature over the correct value of diusion coecient forhydrocarbons in zeolite Y as values diering by two ordersof magnitude have been obtained using dierent experimen-tal techniques for this particular zeolite structure (Kargerand Ruthven, 1989;Ruthven, 2000). In this paper, an experimental conÿguration suitable formeasuring fairly rapid rates of adsorption and desorptionfrom FCC catalyst particles is described. The experimentswere performed using a tapered element oscillating mi-crobalance (TEOM). This is a form of inertial balance inwhich mass changes can be found by measuring the naturalresonance frequency of a tapered quartz element containingthe sample of interest (Patashnick and Rupprecht, 1986). It allows accurate mass measurements to be made rapidlyat elevated temperatures under conditions in which the ad-sorbate vapour passes through the catalyst bed. The owconditions used, and the fact that the amount of catalystused is small (typically 60 mg), means that external massand heat transfer limitations are far less signiÿcant thanwhen using other microbalance techniques. Further, it is possible to measure mass in a TEOM every 0 : 1 s and sorapid changes can be investigated.While the TEOM was initially developed to measure par-ticulate concentration in a gas (Patashnick and Rupprecht,1991), its features make it particularly attractive for study-ing many phenomena in catalysts. It has thus been used tomeasure equilibrium adsorption isotherms (Zhu et al., 1998,2001a;Giaya and Thompson, 2002), mass changes during reduction treatment (Rekoske and Barteau, 1997), the kinet- ics of coking (Liu et al., 1997a;Petkovic and Larsen, 1999; Chen et al., 2000), and the inuence of coke on catalystactivity and selectivity (Chen et al., 1996a, 1997, 1999a;Liu et al., 1997b;Petkovic and Larsen, 2000;van Donk et al., 2002). There have also been attempts to use TEOMresults to consider the combined eects of adsorption, dif-fusion, reaction and coking (Chen et al., 1999b;van Donk et al., 2001). Despite this work, there have only been a few published studies so far that have taken advantage of thefast-time resolution of the TEOM to measure adsorption ordesorption kinetics (Hershkowitz and Madiara, 1993;Chen et al., 1996b, 1999c;Rebo et al., 1997;Zhu et al., 2001b; Alpay et al., 2003)which is the subject of this work.In this paper experimental TEOM results are reported onthe adsorption and desorption kinetics of some simple hy-drocarbon molecules within a commercial FCC catalyst anda pure zeolite Y sample under non-reacting conditions (tem- peratures not exceeding 473 K). The aliphatic hydrocarbonsinvestigated were n -hexane, n -heptane and n -octane; theseallow the eects of chain length to be studied. The aromatichydrocarbons investigated were toluene and p -xylene; theseallow comparison with the results on the straight-chain alka-nes to be made. In this paper, the experimental conÿgurationis characterised, equilibrium adsorption results are analysed,and qualitative conclusions are drawn from the transient ki-netic measurements. Detailed numerical simulations of therates of adsorption and desorption will be the topic of anaccompanying paper. 2. Experimental The catalyst samples were supplied by BP Oil. The com-mercial FCC catalyst contains rare-earth exchanged zeoliteY in a matrix, and consists of spherical particles of approx-imately 70  m diameter. A pure rare-earth exchanged zeo-lite Y was also used for comparison purposes, and this has particle sizes of 0.5–1  m. Both samples were steamed at1089 K for 5 h in order to cause “ultrastabilisation”. This process causes dealumination of the zeolite component, butenhances the overall stability of the catalyst (Satterÿeld,1980;McLean and Moorehead, 1991). It was performed in order to give a catalyst with a composition and structurecomparable to that found in FCC units, as the steaming mim-ics the actual conditions experienced in an FCC unit. Sur-face area measurements were made by analysing nitrogenadsorption data (Gregg and Sing, 1982) after the steamingtreatment to check the structural integrity of the samples.The steamed FCC catalyst sample has a BET surface areaof 183 m 2 = g, of which t  -plot analysis estimates 157 m 2 = g isdue to micropores. The steamed pure zeolite Y sample hasa BET surface area of 377 m 2 = g, of which t  -plot analysisestimates 343 m 2 = g is due to micropores. In this paper, thelabel FCC denotes the FCC catalyst sample, while the labelREUSY denotes the pure rare-earth exchanged ultrastablezeolite Y sample.A schematic diagram of the experimental conÿgurationused is shown in Fig.1.The catalyst was placed at the bot- tom of the quartz element of a Rupprecht and PatashnickPMA 1500 TEOM. The amount of catalyst used was upto 64 mg for the FCC sample and 32 mg for the REUSYsample. In some cases the catalyst was diluted with quartz  C.K. Lee et al./Chemical Engineering Science 59 (2004) 1131–1138 1133 valve 1to GC PTTC to GC or MSto ventHe(pretreat)He(carrier)He(purge)hydrocarbonsaturator TC TEOMto vent Fig. 1. Diagram of the experimental conÿguration used.  particles. During a typical experiment to measure adsorp-tion and desorption kinetics, the catalyst in the TEOM isÿrst activated by heating at 623 K in a dry stream of helium pretreat gas. The sample is then cooled to the temperatureof interest (373–473 K in this work) while maintaining thehelium ow, and the balance allowed to equilibrate at thistemperature until a stable baseline is obtained. Pneumaticvalve 1 is then switched to allow the helium carrier streamcontaining hydrocarbon vapour to reach the sample. Onceequilibrium has been attained (typically after a 60 s dura-tion), the desorption step is performed by switching valve 1 back to its srcinal position so that the sample is once againexposed only to helium pretreat gas. The element resonancefrequency, and thus mass, is monitored at all stages of this process, with a time resolution of 0 : 11 s per data point.All gas ows in the apparatus are controlled by automaticmass ow controllers. The helium owrates used in all ex- periments was 200 ml = min at STP conditions unless other-wise indicated. The function of the helium purge stream isto ensure that any hydrocarbons desorbing from the catalystare removed from the balance. The pipes after the satura-tor were heat traced to ensure that no hydrocarbon conden-sation took place before the balance. The concentration of the hydrocarbon in the feed gas can be adjusted by chang-ing the temperature of the saturator. A gas chromatograph(GC) equipped with a ame ionisation detector is used tomeasure the composition of the feed gas. Experiments were performed using dierent hydrocarbon partial pressures inorder to measure equilibrium adsorption isotherms on thecatalysts.An important part of the experimental protocol is to en-sure that the needle valve on the vent outlet from valve 1 isadjusted to ensure that there is no pressure imbalance in thesystem when valve 1 is switched. Any pressure imbalancecould be detected by the pressure transducer at the inlet tothe TEOM. Pressure imbalances need to be avoided duringkinetic measurements as mass changes due to variations inthe density of gas owing through the TEOM can be compa-rable to the mass changes that occur during adsorption anddesorption, and a pressure imbalance can take many sec-onds to stabilise. For instance, when measuring the kineticsof adsorption of alkanes in silicalite, Zhu and co-workersignored the ÿrst 5–10 s of data because it was aected bya pressure imbalance (Zhu et al., 2001b). This meant that they were unable to measure rapid kinetic changes usingtheir experimental conÿguration. Such a limitation is not a problem using our experimental protocol as is demonstrated by the results below.The experimental conÿguration allows products from themicrobalance to be examined by gas chromatography ormass spectrometry (MS). These showed that there was neg-ligible reaction for the hydrocarbons studied under the con-ditions employed (373–473 K). In order to test for repro-ducibility, and to rule out the possibility of any coke formedaecting the kinetic measurements, sequences of 5–10 sep-arate adsorption/desorption experiments were performed onthe same catalyst with no regeneration between each step;identical results were obtained in each case.Experimentswerealsoperformedusingparticlesofquartzin the balance in order to check that the mass proÿles ob-tained reect adsorption within the catalyst rather than justchanges in the gas density in the interparticle space. In prin-ciple, it is necessary to subtract the data obtained usingquartz from that obtained using catalyst under identical con-ditions in order to achieve pure adsorption results. However,under the conditions employed it was found that the massadsorbed was far greater than the mass in the interparticlespace and so this correction was not necessary for the resultsreported in this paper.The apparatus was arranged in such a way that the lengthof pipe (diameter 1 = 16 in) between valve 1 and the mi-crobalance is as small as can practically be achieved in orderto minimise dead volume. In order to characterise the deadvolume, which is necessary for detailed interpretation of theadsorption kinetics, quartz particles were used in the bal-ance and the carrier gas switched from helium to nitrogenwhile the pressure and temperature were maintained. 3. Results 3.1. Characterisation of dead volume Fig.2shows TEOM results when the carrier gas wasswitched from helium to nitrogen with 60 mg of quartz par-ticles (diameter approximately 70  m) present in the reac-tor. An increase in mass is observed due to the higher massof nitrogen compared to helium. At time t  = 0, there is asmall peak due to vibrations detected by the balance when  1134 C.K. Lee et al./Chemical Engineering Science 59 (2004) 1131–1138 - 0 5 10 15Time (s)    M  a  s  s   (  m  g   ) Fig. 2. TEOM results with 60 mg quartz present and a switch fromhelium gas to nitrogen gas at time t  =0 and owrate 200 ml = min (at STPconditions). The TEOM temperature was 473 K, the line temperature423 K, and the pressure was 1 : 33 bar. The solid line is a ÿt to the datato characterise the dead volume present (see text).  pneumatic valve 1 was switched. However, there is no indi-cation of any long-term pressure imbalance when the valveis switched, and it is clear that reliable adsorption kinet-ics measurements can be measured on timescales of the or-der of 1 s using our experimental conÿguration. The totalmass change observed under these conditions is 0 : 723 mg.This corresponds to the balance detecting gas in a volumeof 0 : 89 cm 3 . Ideal step behaviour is not observed due tothe dead volume in the apparatus between valve 1 and that part of the element in which mass is detected. This vol-ume comprises pipework, the valve to the GC, the pressuretransducer, and volume at the top of the TEOM. Under theconditions of this experiment, it can be seen that there is atime lag of 0 : 8 s before any change in mass is detected, af-ter which the shape of the curve follows a curve that can be ÿtted with an exponential function with time constant0 : 96 s − 1 . This behaviour corresponds to the apparatus be-having as if there was a region of plug ow after valve 1with volume 3 : 1 cm 3 , followed by a well-mixed region withvolume 4 : 2 cm 3 , before gas reaches that part of the elementin which mass is detected. The solid line in Fig.2is thecurve that uses these volumes to model the behaviour. Thevolumes obtained can then be used to predict what actualgas concentrations will be in the absence of adsorption atother temperature and pressure conditions. This characteri-sation means that ideal step behaviour need not necessarily be assumed when modelling TEOM results for the uptakeof an adsorbate into a catalyst. 3.2. Measurement of equilibrium adsorption isotherms The partial pressure of hydrocarbon in the carrier streamwas varied by changing the temperature of the saturator, andthe mass adsorbed in the catalyst at equilibrium measured by the TEOM. Experimental results for p -xylene in the FCCcatalyst at ÿve dierent temperatures are shown in Fig.3. Partial pressure of p  -xylene (bar)    M  a  s  s  c   h  a  n  g  e  p  e  r  g  r  a  m   o   f  c  a   t  a   l  y  s   t   (  g   /  g   ) 473 K423 K448 K398 K373 K Fig. 3. Equilibrium adsorption isotherms for p -xylene in FCC catalyst.The curves are ÿts to the Langmuir isotherm expression. The curves are ÿts of the experimental data to the Langmuiradsorption isotherm  M  = M  max  K   L  P  HC  1 +  K   L  P  HC  ; (1)where M  is the mass adsorbed, M  max is the amount adsorbedat high pressure, K   L is the Langmuir constant, and P  HC  isthe partial pressure of the hydrocarbon. The ÿtted curvesshow that this isotherm is a reasonable model of the amountadsorbed at equilibrium for this catalyst. Similar qualitydata and ÿts were obtained for all the hydrocarbons stud-ied ( n -hexane, n -heptane, n -octane, toluene and p -xylene)on both the FCC catalyst and the REUSY sample. In mostcases, it was possible to obtain reasonable ÿts to the data by constraining M  max to be independent of temperature forthe systems studied, though this is not necessarily a require-ment for hydrocarbon adsorption in zeolites. It can be seenthat TEOM experiments can readily be performed in boththe linear region of the isotherm (e.g. at high temperature orlow partial pressure) or in the non-linear region where sat-uration can be approached (e.g. at low temperature or high partial pressure).Heats of adsorption, −   H  ads , in the low coverage limitwere calculated from the isotherms using the equation(Ruthven, 1984):dln  K   H  d T  =  H  ads  RT  2 ; (2)where K   H  is the Henry constant (and thus equals the gra-dient of the isotherm plots at low coverage). Values of theheats of adsorption, together with the Langmuir parame-ters obtained, are shown in Table1.It should be mentioned that there is a reasonably large uncertainty ( ± 3 kJ mol − 1 )in the heats of adsorption obtained from our TEOM mea-surements as our experimental conÿguration is optimisedfor the measurement of transient kinetics at moderate pres-sures rather than the amount adsorbed at very low pres-sures. The values obtained on the linear alkanes are in goodagreement with values obtained in the literature on zeo-lite Y samples in other cation-exchanged forms (Eder and
Related Search
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