Just Because It’s Small Doesn’t Mean It’s Well Mixed: Ensuring Good Mixing in Mesoscale Reactors J. F. Hall, †,‡ M. Barigou, † M. J. H. Simmons,* ,† and E. H. Stitt § Centre for Formulation Engineering, Department of Chemical Engineering, University of Birmingham, Edgbaston, Birmingham B15 2TT, U.K., and Johnson Matthey Catalysts, P. O. Box 1, Belasis Avenue, Billingham, Cleveland TS23 1LB, U.K. The advent of high throughput experimentation (HTE) for molecular discovery and rapid screening of new catalyst formulations has led to interest in the mixing characteristics of small stirred vessels at scales below those previously studied. In this paper, particle image velocimetry (PIV) is used to obtain macromixing characteristics for single phase fluids of two different viscosities (µ ) 0.001 and 0.433 Pa s) and for two-phase air-water mixtures in a 45 mm diameter vessel using a 6 blade up-pumping pitched-blade turbine. Eccentric agitation is examined as a means of improving the mixing performance since the vessels are used without baffles, and the global circulation of the fluid was comparable with that of a conventional baffled configuration for single-phase mixing at a constant power input per unit mass (0.168-5.5 W kg -1 ). For high viscosity fluids, the characteristic flow patterns resembled those for a radial device. For two- phase mixing at gassing rates of 0.25 and 0.5 vvm, two sparger configurations were used with the sparger being placed either beneath the impeller axis (R) or away from the impeller (). The  configuration appeared to be a better choice because of the smaller size of bubbles generated compared with the R configuration and the satisfactory levels of global mixing observed in the vessel. 1. Introduction The control of the degree and rapidity of mixing is essential for the successful operation of any industrial process. The complex nature of the flow field within most industrial equipment, in particular, the ubiquitous stirred tank, has led to considerable research effort to obtain an understanding of the mixing length-scales and time-scales of both single and multiphase mixtures. Several books have been written on the subject. 1-3 The general approach taken for the translation of a devel- opmental process from the initial laboratory scale to the industrial scale of production is to attempt to mimic the operational conditions in the pilot vessel at the produc- tion scale. This has been a perfectly valid assumption for scenarios involving the development and testing of a process at vessel scales on the order of 10 -3 to 10 -2 m 3 upward, as the large number of previous studies (both academic and industrial) have proved. However, the advent of revolutionary high throughput experimentation (HTE) techniques requires a reexami- nation of this approach. HTE offers the potential to dramatically reduce the time-scales currently required for the screening of novel molecules and catalysts, and as such, significant benefits are anticipated via the implementation of the HTE protocol in the worldwide chemicals industry. Many commercially available HTE units are based upon small agitated stirred vessels with typical volumes of 10 -5 to 10 -4 m 3 , an order of magni- tude below existing well-researched lab-scale mixers. The mixing performance of these units is not optimal because of two primary factors: 4-6 First, the small scale of the reactors precludes the generation of high Rey- nolds numbers, even at high agitation speeds. Second, the HTE reaction vessels are generally unbaffled, to facilitate automated loading and cleaning cycles by robotic server units, as well as to prevent fouling/con- tamination and excessive particle attrition. Without baf- fles to break up the dominant tangential flow, efficient mixing cannot be achieved in conventional unbaffled vessels where the impeller shaft is placed at the vessel axis. Since decisions regarding the viability of a certain process or operating condition requirements are often made based on the information generated by the HTE unit, it is vital that the hydrodynamic behavior and fluid mixing performance of these small vessels are fully quantified. Without this information, there is no way for the operator to discern the parametric sensitivity of the process under investigation and differentiation be- tween mass transfer effects and/or kinetic effects be- comes impossible. Studies by previous authors based upon conventional vessel configurations at the laboratory and pilot scale (H ) T, C ) 1 / 3 H, D ) 1 / 3 T, fully baffled conditions) for both radial and axial impeller types using flow visual- ization techniques and computational fluid dynamics have characterized the macroflow behavior in terms of mean flow patterns, 7-12 the microflow qualities in terms of the turbulence parameters, 13-18 and the mixing per- formance in terms of the concentration fields. 19-23 Laser Doppler velocimetry (LDV) and particle image veloci- metry (PIV), in particular, have enabled researchers to obtain high-resolution velocity fields and, hence, develop a thorough understanding of both the global flow patterns and the behaviour of the fluid within the highly energetic regions, such as the impeller discharge jet. * Corresponding author. Dr. Mark Simmons. Tel.: +44 (0) 121 4145371. E-mail: m.j.simmons@bham.ac.uk. † University of Birmingham. ‡ Current address: Johnson Matthey Catalysts, P. O. Box 1, Belasis Ave, Billingham TS23 1LB, U.K. § Johnson Matthey Catalysts. 9695 Ind. Eng. Chem. Res. 2005, 44, 9695-9704 10.1021/ie050224w CCC: $30.25 © 2005 American Chemical Society Published on Web 07/08/2005