Influence of Reaction Conditions on Catalyst Behavior during the Early Stages of Gas Phase Ethylene Homo- and Copolymerization Estevan Tioni, †,‡ Jean Pierre Broyer, † Vincent Monteil,* ,† and Timothy McKenna* ,† † Universite ́ de Lyon, Univ. Lyon 1, CPE Lyon, CNRS, UMR 5265 Laboratoire de Chimie Catalyse Polyme ̀ res et Proce ́ de ́ s (C2P2), LCPP team, Bat 308F, 43 Bd du 11 novembre 1918, F-69616 Villeurbanne, France ‡ Dutch Polymer Institute DPI, P.O. Box 902, 5600 AX Eindhoven, The Netherlands ABSTRACT: A packed bed stopped flow minireactor (3 mL) suitable for performing gas phase polymerizations of olefins has been used to study the initial phases of ethylene homo- and copolymerization with two supported metallocene catalysts. The reactor can be used to perform gas phase polymerizations at times as short as 100 ms under industrially relevant conditions. It has been used to follow the evolution of the rate of polymerization, the gas phase temperature (and indirectly the particle temperature), and the polymer properties (molecular weight distribution, melting temperature, and crystallinity) for the two catalysts. It is shown that polymerization activity during the first 2-5 s of reaction can be up to 20 times higher than what is measured at longer polymerization times. The main consequence is the release of a significant amount of heat due to the rapid reaction that has to be efficiently evacuated in order to avoid particle overheating and melting. It has been seen that insufficient heat removal can strongly influence the behavior of the active sites, eventually leading to uncontrolled transfer reactions and polymers with unusually broad molecular weight distributions (MWD). It is also observed that the kinetic behavior of the two types of catalyst is similar at short times. Finally, some influence of particle size on reaction rate and molecular weight is observed between the largest and smallest catalyst particle cuts. 1. INTRODUCTION The annual production of polyethylene (PE) and polypropy- lene (PP) in processes using supported catalysts is likely to be close to 90 million at the current time. 1 This process, which has been used commercially since the 1950s is clearly commercially significant and has been the object of many industrial and academic studies far too numerous to cite here. However, despite the intense research efforts that have been made in the past 5 decades, there remains much to understand about these processes. In particular, the events surrounding the trans- formation of the catalyst particle into a growing polymer particle still need to be better understood. 2 The most common types of catalyst supports used in olefin polymerization are magnesium dichloride for Ziegler-Natta (ZN) catalysts, and silica which is used to support chromium oxide and metallocene active sites (although hybrid supports of MgCl 2 on silica are used for certain types of ZN catalysts). Both types of support have a high surface area and porosity that allow for the deposition of a large number of active sites throughout the structure, and both types of support undergo physical transformations as soon as they are injected into the reactor. It can be said that the nature of the steps by which the particles of supported catalyst are transformed into growing polymer particles is certainly understood from a qualitative point of view: once the virgin catalyst particles are injected into the reactor, monomer diffuses from the bulk phase and begins to react at the active sites on the surface of the catalyst support. Polymer then quickly accumulates, generating pressure throughout the particle and provoking a local fragmentation of the support. Once the fragmentation step is complete, the resulting particle (now referred to as a polymer particle) will continue to grow as long as monomer arrives at the active sites. For a more in-depth discussion of the process of catalyst fragmentation and growth, the reader is referred to earlier reviews from our group, as well as the references therein. 2,3 The fragmentation step is quite short with respect to the average residence time of the reactor; fractions of a second up to several tens of seconds, depending on the type of support and reaction conditions, versus 1-3 h, respectively. Despite the almost negligible amount of time the fragmentation and initial growth phases take with respect to the length of an industrial reaction, it is at this point that the most extreme changes in particle morphology occur, that the production of fines due to an abrupt particle fragmentation is likely, and when severe overheating of the particles can be a significant problem. Given that the rates of mass transfer into the particles and heat transfer out of the particles are strongly dependent on the morphology (pore network, pore size, pore volume, and especially particle size), this can be also the time range in which the temperature and concentration values inside the catalyst particle can vary most abruptly. For instance, strong concentration gradients can lead to different parts of the particle core expanding more rapidly than others, with consequences as extreme as the production of hollow particles. 4,5 Excessive temperature gradients in the boundary layer (temperature profile inside the particle is known to be less important 6,7 ) can lead to polymer melting and reaction extinction. Received: June 25, 2012 Revised: October 14, 2012 Accepted: October 18, 2012 Published: October 18, 2012 Article pubs.acs.org/IECR © 2012 American Chemical Society 14673 dx.doi.org/10.1021/ie301682u | Ind. Eng. Chem. Res. 2012, 51, 14673-14684