Particle Formation

Granulation is a complex process in which many input variables influence many product properties. As Iveson et al. describe in a review paper (2001), the understanding of the fundamental processes that control granulation behavior and product properties have increased in recent years. This knowledge can be used during process design, in choosing the right formulation and operating conditions, and it can also be used to improve process control. Although many variables are set constant during process design, variations during production in input variables occur due to the variable nature of the powder feed. Even if all granule properties, except for size, are ignored for process control, a one dimensional granule size distribution can be constructed by multiple discrete output variables, in order to represent the shape of the distribution (these can be mean sizes (with coefficients of variation), percentile sizes, moments 4 or size bins). Model Predictive Control (MPC) is an effective method to control such multiple input, multiple output processes (García, et al., 1989). More details motivating the choice of MPC for granulation processes along with examples from the literature can be found in previous reports (IFPRI# ARR51-02 and IFPRI# ARR51-03).

Project Report

We report here on work performed on the project in its third year ending October, 2009. Two previous reports have been submitted to IFPRI in 2006 and 2008, respectively. The present research is focused on the study of Powder Mechanics and the ultimate goal is to develop a quantitative description of active flows of a wide variety of powders. The study is centered on the slow, frictional and the dense, “intermediate” regimes of flow where both frictional and inertial effects are important. The novelty of the project is the study of a large range of materials and flow geometries to gain meaningful insight.

Materials and Methods

We report on a series of materials from simple (round beds) to complex (fine, odd-shaped and elastic), used in an axial-flow Couette, high shear mixer, in a centripetal geometry characteristic of a “spheronizer and two hopper flows with a funnel and a flat bottom, respectively to measure stresses and their fluctuations as a function of geometry and shear rate.

Main Novelty

The main novelty during this year is the measurement of porosity (void fraction) and porosity distribution in the flowing material using a capacitance probe in addition to stress measurements.

General project outline and background

The structure of granules or agglomerates can be defined as the spatial arrangement of its basic components [1]. The basic components are primary particles, binder or liquid components and intra particle porosity. The quantification of granule structure is crucial for setting up processing maps and input data for simulations such as DEM and computational modelling [2]. Spatial statistics, also called stereology, provide well-defined measures to quantify the different aspects of powder structure. For infinite systems the covariance function (CVF) defines the distance relations of the particles and pores. In agglomerates a radial distribution function (RDF) quantifies shell or boundary structures. These statistical measures are well suited as a basis for stochastical property modelling. Several image analysis procedures provide the above mentioned data, e.g. morphological sieves, point sampling techniques or chord length measurements. The characteristic curves can also be correlated with stochastical processes e.g. Poisson processes. While structure measures have been determined in a number of applications, the fundamental combination of physical properties and structure is still immature. The parameter of the above mentioned stochastic models need to be linked to the key parameter of the structure generation process e.g. primary particle size, binder content, stress and growth history. Results of other simulation techniques like VOF (Volume of fluid) or Population Balance models can serve as input to generate the correlation between statistical parameter and process. The added value of a structure model would be its statistical reliability and ease of use for parameter variations. Some key powder properties are well understood physically [3-6] but lack systematic incorporation of structure measures above phase volume and particle size. Mathematical techniques to calculate structure dependent properties include cellular automata and convolution algorithms. The additional challenge is to convert local behavioural probabilities into bulk properties.

Executive Summary

During the first year of this project, a postdoctoral research was hired with background in DEM and started in January of 2013. One student was working on subjects that are close to this project. A second student was added in summer. A parallel project with Abbvie focuses on the modeling of compaction and strength of single type particles and leverages the effort of this project. In summary the following activities were undertaken:

• A literature survey was performed to identify prior efforts on DEM for powder compaction.
• A careful study was conducted to identify potential obstacles for the application of DEM on powder compaction.
• LIGGGHTS, an open source, highly parallelized code for granular phenomena was installed and used extensively.
• A preliminary model that takes into account contact interactions was implemented in LIGGGHTS and appears to perform better than prior efforts (Storakers model). Results from this work are being analyzed.
• Preliminary results for the compaction of two phase mixture were obtained by using the Storakers model. Despite its lack of accuracy, this work provides interesting insights in the problem in terms of relative magnitude of interparticle forces on different types of contacts.
• Ongoing work focuses on finalizing the selection of force-displacement model for high densities, introduction of different size and different properties effect in it and their cross interaction. Our work has also shown, there may be a need to introduce an elastic component in the force-displacement response, in order to understand some of the aspects of mechanical behavior of compacted powder mixtures.

Introduction

The development of strength during compaction of fine powder is a topic with immense importance for a variety of industrial applications. Results related to experimental evaluation of particle to particle interactions, particle-based computational simulation of compaction, and a renewed discussion about proper interpretation of traditional strength tests have been recently finding their way in the literature but have not yet compiled in a single document. In addition, the importance of the particle-to-particle interfaces has been rather understated. The development and evolution of interparticle cohesion during the process is central to the determination of the final compact properties but a quantitative understanding is not available.

This review attempts to contribute towards a more objective and coherent presentation on the subject, which will include the state-of-the-art in our understanding of the problem and will highlight areas where additional knowledge is needed. While experimental results will be incorporated in the discussion, this report is geared towards the discussion of the necessary models that analyze the various phenomena in compaction and post compaction mechanical properties. In such problems with multiple parameters and interactions, experiments offer correlations and often excellent insight but it is only when the physics is understood and modeled that predictive approaches become available and optimization becomes possible without the need for extensive experimentation.

PROJECT POWDER STRUCTURE CONTROL

The project powder structure control is concerned with the definition and calculation of mathematical descriptors of granule and powder structure. The additional goal is to give the structure descriptors a physical meaning by linking them to product properties e.g. dissolution behaviour and mechanical strength.

This report covers the last year (Dec. 2011 - Nov. 2012) of work on the powder structure control project. In this period the focus has been on the development of a software package for structure evaluation. The other main part of work was concerned with a literature review on stereological methods in practical application and on the relationship between process, structure and properties.

The software package includes first and second-order stereological methods. The volume fractions of the three phases and the interfaces between those have been calculated for a test image of spheres and for exemplary images of granules. Results are in good agreement with the known test image and with visual examination of the granule images. For the evaluation of the spatial arrangement of granules second-order methods are necessary. The chord length distribution, the volume weighted star volume and the covariance function deliver results that are plausible regarding the exemplary images. In the literature review publications of all disciplines have been screened and sorted into 3 different groups: description or derivation of stereological methods, practical application of stereological methods and relationship between process, structure and properties of granules. As the stereological investigation of structure is known for a long time a lot of review publications are available, especially for first-order stereology [9, 10, 12, 15, 21, 50]. Second-order stereology has been in the focus of recent publications, especially the normalized covariance function and the pair correlation function have been of interest.

Applications of these methods could be found in different fields of research and for different spatial materials like biological tissue, concretes and porous materials. Granular materials play an important role in industry and the control of their behaviour and structure is of interest. Different approaches can be made to correlate structure, process properties and product properties. The relationship between process and product properties is very common in literature. Process properties like impeller speed of a granulator or the recipe of the raw materials are varied and product properties like strength, porosity or dissolution behaviour are compared to each other [23, 37, 39]. Further publications are concerned with the relation between structure and granule properties [2, 5, 11]. Thereby porosity and primary particle size distribution has been a frequently used parameter.

Based on these findings the next step is a structure analysis that includes spatial arrangement of structure elements because first-order measures like porosity are not suitable for every issue. When the structure can be fully described by one or more of the mathematical descriptors it should become possible to predict product properties. Process parameters as well as the raw materials will be altered systematically to generate a variety of structures.

The investigation of the functional relationship between process, structure and properties shall be examined in the following year of this project. The implemented structure measures shall be tested and optimised by using experimental input for further interpretation and understanding. Different structure generation and different recipes are planned to achieve a good variability of internal structures of the granules.

Executive Summary

In this research work, nucleation and growth phenomena of organic nanoparticles from solution are investigated using an experimental approach. In particular, the effects of solution supersaturation and polymeric additives on submicron crystallization are studied. The ultimate goal is to gain fundamental understanding of the combination of process variables that will consistently produce submicron crystals during an antisolvent precipitation process.

Antisolvent precipitation experiments were carried out using naproxen, a poorly watersoluble drug, as an organic model compound. Naproxen solution in ethanol and water (antisolvent) were rapidly and homogeneously mixed in a static Y-mixer to generate high levels of supersaturation. The degree of supersaturation was varied by changing the flow rates of the solute and antisolvent streams respectively. Particle size distribution of naproxen particles obtained from the precipitation experiments were measured offline using dynamic light scattering technique. Preliminary results showed that the particle size ranged between 100−5500 nm. On increasing the initial supersaturation from 16 to 100, the z-average particle size decreased from 2100±1900 nm to 330±190 nm. This experimental result is in line with the classical nucleation theory, according to which primary homogeneous nucleation rates increase with the supersaturation levels, thereby resulting in smaller particles. On the other hand, the median particle size (d50) obtained in this supersaturation range was found to vary between 50−300 nm and did not show a clear trend with the changes in supersaturation. As observed under SEM, the primary naproxen particles obtained under the higher supersaturation conditions were mostly spherical in shape and were in reasonable agreement with the z-average particle size. The crystalline nature of these particles were confirmed using powder X-ray diffraction and from high resolution TEM images.

Executive Summary

In this research work, nucleation and growth phenomena of organic crystals from solution were investigated experimentally, with an objective to produce a dispersion of submicron particles. Specifically, the effects of solute concentration and polymeric additives on the precipitation kinetics of naproxen (organic model compound) were studied. The ultimate goal is to gain fundamental understanding of the combination of process variables that will consistently produce submicron crystals during an antisolvent precipitation process.

Naproxen solution and antisolvent (deionized water with trace amounts of polymeric additives dissolved) were rapidly and homogenously premixed in a Y-micromixer using two syringe pumps. The degree of supersaturation was varied by changing the solute concentration in ethanol. The outlet stream from the Y-mixer was directly fed into a glass vial and the solution turbidity monitored online using the Avantium Crystalline system. The rate of desupersaturation during precipitation process was determined by offline measurement of solute concentration using UV-Vis spectroscopy. Initial results show that in the presence of HPMC K86 polymer additive, induction period for nucleation of naproxen crystals increased significantly as compared to the pure system. On the other hand, in the presence of PVP K10 polymer additive, the induction time apparently decreased. In line with this trend, the rate of desupersaturation in the presence of HPMC decreased significantly as compared to the pure system. These kinetic data correlate well with the effectiveness of the polymeric additives in controlling particle size of naproxen crystals in the submicron range. From the experimentally determined induction periods for crystal nucleation as a function of supersaturation, the nucleation mechanism (viz., primary homogeneous and heterogeneous) controlling the precipitation process was determined.

Future work will focus on unraveling the mechanisms underpinning the effects of polymer additives on stabilization of colloidal crystal dispersion using both experimental techniques and molecular modeling.