RICERCA ITALIANA     

Research program

Multiscale modelling and development of process reactors for polymeric nanoparticle precipitation

University Co-ordinator

Università degli Studi di PALERMO - INGEGNERIA CHIMICA, DEI PROCESSI E DEI MATERIALI - ()

Research Unit Leader

Alberto Brucato

Description

The major aim of this research unit is the developement of a novel reactor for carrying out precipitation reactions of both inprganic and polymeric nanoparticles. The new reactor development and optmization will be assisted by suitable simulations of its fluid dynamics and of the complex chemico-physical processes that take place. The development of the models required for performing the simulations may be considered as an objective in itself, as the applicability of the models developed goes beyond the simulation needs of the novel reactor under development here.

The detailed research plan is:

1) Development of a robust and easily scalable precipitation reactor.

The reactor type that will be investigated first consists of two concentric cylinders, with the internal one rotating, known as “Couette cell”. In particular attention will be devoted especially to Couette cells characterized by very narrow gaps between the inner and outer cylinders, thus resulting in high and uniform shear rate throughout reactor volume. It is worth mentioning that the suitability of narrow gap, high-shear rate reactors to carry out mixing sensitive complex reactions has already been proven with a disk reactor (Rousseaux et al., 1999), also for the case of precipitation reactions (Rousseax et al., 2001). As a difference from the disk type reactor, the Couette type results in a perfectly uniform flow field
As it is well known, depending on cylinder sizes and rotation speed, this kind of geometry gives rise to different flow regimes, from laminar to turbulent, in either case with or without Taylor vortexes (Kataoka, 1986). The suitability of the above regimes for the precipitation of the nanoparticles of interest here, will be explored.

The activities planned for this section of the project are:

1a – reactor design, commissioning and assembly (to be completed by the first 4 months from project start);

1b – measurement of main fluid dynamic features: head losses, residence time distribution, mechanical power input in the liquid phase; these measurements will have to be repeated on all modifications of reactor design that will be devised along the project, and may therefore extend over the entire project life;

1c – conduction of a number of experimental runs aimed at precipitating Barium Sulphat ( a much simpler inorganic precipitation to be used as a test). The experiments will be conducted in strict collaboration with the research units of Bologna and Turin-Polytechnique for a first performance comparison with the precipitation reactor types there investigated (stirred vessel, impinging jet reactor, static mixer). Particles produced by the various units will be sent to the other units in order to compare them with exactly the same instrumentation operated by the same “hands”. The measurements that will be made in Palermo include particle imaging by scanning electron microscopy (SEM) for assessing particle morphology and size distribution. The measurements on the first reactor design are expected to be completed by the end of the first year.

1d – precipitation runs aimed at producing polymeric nanoparticles of potential pharmaceutical interest, according to the formulations indicated by the unit of Turin-University. This activity will take place during the second year of the project. Size and morphology of the particles obtained will be characterized by scanning electron microscopy (SEM). It may also be possible to get information on the amount and distribution of the co-precipitated drug by means of the X-ray dispersion probe for microanalysis installed on the SEM available in Palermo. In case of need, some measurements will also be made with an Atomic Force Microscope (AFM) also available in the department. This should allow the measurement of particle sizes below 100 nm and an assessment of nanoscale particle mechanical properties. Once again, similar measurements will be conducted on the particle samples produced by the other units, for comparing the performance of the various reactors with the same particles. For symmetrical reasons the particles produced in Palermo will be dispatched to the other units.


2) Reactor modeling and simulation

Reactor simulations will be performed with either fully three-dimensional CFD models or with simplified one-dimensional axial dispersion models. For the latter, values of the dispersion coefficient will be deduced from the previously quoted experimental retention time distribution (RTD) curves. For both modeling needs a strong support from the Turin-Polytechnique research unit will be needed, in order to exploit their large experience on subgrid models for the small scale processes that take place (nucleation, growth and agglomeration) as well as for the implementation of simplified population balances. The two modeling approaches can be further detailed as follows:

2a – Simplified one-dimensional models: these are likely to work well for the cases of turbulent motion, with or without Taylor vortices. In this case in fact the fast radial mixing caused by the radial velocity components will effectively destroy radial concentration gradients: Considering also that azimuthal gradients are cancelled out by the high rotational speed, it becomes clear that a simple one-dimensional flow model, namely the axial dispersion model, will probably be sufficient for such cases. For these simulations a simple, yet effective, technique already successfully tested for simulating axial dispersions [6] will be employed. The technique makes use of different time steps for the integration of the advective and diffusional terms, and is accurate and robust.
It is worth noting that the analysis of experiments conducted in conditions in which the one-dimensional model works well, is a suitable test bench for the nucleation, growth and agglomeration models developed in collaboration with the Turin-Polytechnique research unit, for both inorganic and polymeric nanoparticles.
The strong simplification involved in the adoption of a one-dimensional model will allow the adoption of non-approximated population balance implementations, in addition to the simplified approaches developed with the help of the Turin-Polytechnique, such as the Montecarlo or the DQMOM (Direct Quadrature Method of Moments). The comparison of results so obtained with those obtained by the former approach should therefore allow the validation of the simplified methods, for the class of problems of interest here.
This section of the research program will be started after the first experimental data will have been acquired, hence presumably by the second semester, and will continue up to the end of the project.

2b – Three-dimensional, CFD based models will probably be needed in the cases of laminar, or near laminar motion, where the flow field details will probably be affected by the viscosity gradients related to concentration variations. For such simulations, as far as the flow field is concerned, advantage will be taken from the past experience on the simulation of highly swirling flows [1, 11]. For the other phenomena advantage will be taken from the already mentioned collaboration with Turin-Polytechnique. CFD-RANS and CFD-LES simulation will also be conducted for the flow regimes discussed under 2a, and results will be compared with those from the simplified approaches. CFD simulations will be performed with the help of one of the commercial codes available to the group (CFX-4, CFX-10 and STAR-CD).
This part of the program will be carried out in the third and forth semesters of the project.