ARR - Annual Report

The presence of moisture in powders and grains drastically changes the mechanical properties of the materials. Inter-particle cohesion due to capillary bridges, causes grains to agglomerate into clumps of sizes much larger than the size of the constituent particles. Besides lacking a complete understanding of the effects of inter-particle cohesion on granular mechanics, the agglomeration of granules makes the processing of such materials challenging, in particular during drying under agitation processes. The methods of drying a mixture of particles and liquid affect the state of agglomeration of the final dried product, particularly through the influence of impacts and shear forces on the agglomerates.

The project is based on the development of model experiments to understand and model the mechanics, size distribution, and time evolution of particle agglomerates. By controlling cohesion and grain properties, we hope to shed some light on the mechanical behavior of agglomerates to develop models at agglomerate scales. In particular, we consider the breakup of model agglomerates, with an application to more efficient drying of powders. Besides causing trouble in the processing of such materials, agglomerates can also sustain humidity within them, reducing the overall ability to dry such bulk materials.

During the first year of the project, different experimental tools were developed and tested with model particles - spherical glass beads - and without considering heat exchange and drying. We developed an oscillating system consisting of a mechanical shaker and a quasi-2D transparent box allowing us to observe the agglomerates. Preliminary tests with model glass beads and water have been performed to probe the role of acceleration and amplitude of oscillation on the agglomerate sizes. A second system, relying on an agitation provided by a turbulent airflow has been designed and led to some first results. Finally, we initiated model experiments consisting of the impact of isolated agglomerates on a solid surface to probe their mechanical properties and fragmentation.

The focus of this project is to understand the physical mechanisms that lead to defect formation – pitting, cracking, and delamination – during pharmaceutical tableting. A leading hypothesis among IFPRI members is that trapped interstitial air leads to high pore pressures that tend to fracture adhered particle interfaces after removal of the confining pressure. The project objective is to explore this problem through coupled numerical methods including:

1. continuum mixture models
2. the discrete element method (DEM) coupled with a fluid solver.

The primary barrier to using these methods is that fact that the behavior of cohesive powders is not well understood, with neither a generally accepted constitutive relation nor contact model in existence.

To address this, the project has emphasized developing a reliable cohesive powder contact model for usage in DEM. This is the natural progression, since a powder DEM model will be indispensable in determining a constitutive relation for continuum simulations. In particular, we have concentrated on creating a mechanically-derived contact model for adhesive elastic-perfectly plastic particles.

In year one of the project, the majority of the theoretical framework for the contact model was developed, but a number of issues remained. The JKR-type adhesion of the contact model needed to be validated once significant plastic deformation had occurred and the scheme to respect plastic incompressibility required an overhaul. The contact model was limited to simple symmetric loadings of a single particle, necessitating adaptation to many-particle interactions. Additionally, the model, initially coded in Matlab, needed implementation in an established software like LIGGGHTS or LAMMPS, with a reliable fluid-solid coupling strategy.

In year two, the theoretical framework was completed by validating the adhesive model within the fully-plastic regime and correcting the plastic incompressibility scheme. The completed contact model was published in the premier solid mechanics journal, Journal of Mechanics and Physics of Solids, as a two part series. E↵orts have extended to modeling many-particle interactions, a necessary step to allow simulation of full-scale industrial applications. Implementation into LIGGGHTS is underway and preliminary simulations show promise in replicating compaction simulator data. In the upcoming phase, we are working directly with Sandia National Laboratories to create a LAMMPS implementation. In addition to greater computational e ciency and formal support for the model, LAMMPS provides the advantage of allowing immediate coupling with a multi-particle collision dynamics fluid solver, capable of simulating a compressible gas phase.

In summary, the project has made substantial progress on the front of creating a reliable powder DEM model and we are now poised at an exciting place where beginning to understand defect formation during pharmaceutical tableting is tangible. Because the model is mechanically-derived, it can also be trusted to assist in the future development of a continuum constitutive relation. Finally, because of the planned open-source nature of the implementation into LAMMPS, a familiar software to most IFPRI members, the powder DEM model can be used for industrial applications extending beyond tableting.

The following report is split into three chapters. Chapter 1 gives a high level update of the project, without detailed technical explanations and equations. Chapter 2 and 3 are both papers of the two part series published in the Journal of Mechanics and Physics of Solids. These papers contain detailed explanations regarding the theoretical background of the contact model in addition to validations made against finite element simulations. Throughout Chapter 1, references to the two papers are made directing the reader to pertinent information to supplement understanding, however, Chapter 1 was written with the intention that it could be understood without reference to the papers.

This annual report presents key advances made during year 2 moving towards year 3. Building upon the experimental study from the previous report period, comprehensive work was done towards validation of the contact mechanics based predictive models for the selection of flow aids (silica) type and amount. This significant component of the outcomes is reported as detailed in a recently published paper; “Selection of Silica Type and Amount for Flowability Enhancements via Dry Coating: Contact Mechanics Based Predictive Approach”, see Appendix A.

In the area of modeling, the previous year’s report reviewed the available particle contact models for smooth and rough particles. During the current reporting period, advancement was made by accounting for the intrinsic macro-roughness of the particle on its cohesion and subsequent reduction after dry coating. A Manuscript, in preparation, will be submitted in Spring 2024.

Another major accomplishment is a comprehensive study to account for the entire PSD of the powder sample via the size class dependent Bond number approach as a promising tool to account for different cohesion amongst fine powders with similar d50 but varying size distributions. That work can help develop a decision tool to identify which component of a powder blend to dry coat while minimizing cohesion yet using the lowest possible flow aid amounts in the blend. A manuscript in preparation will be ready and submitted by the end of Spring 2024.

Another effort towards fulfilling a major deliverable examined four industry relevant silicas, including three hydrophilic and two hydrophobic, detailed in a published Manuscript, “On predicting the performance of different silicas on key property enhancements of fine APIs, blends, and tablets”; see (Appendix B).

In this work, the predictability of the performance of different silica types on cohesion reduction is better explained by combining several models, some of which counter each other for more pragmatic understanding of the nuances involved:

• (1) the Multi-asperity contact model,
• (2) Deng’s stick and bounce model explain the aggregation tendency of nano flow aids (e.g., silica) on the host particle as a function of process intensity and guest-host sizes (Details in recently published paper is attached in Appendix C),
• and the interactive mixture model based on host-guest total surface energy differential (discussed in the paper included in Appendix B).

The results convey that while silica size and its extent are important, silica aggregation is another key parameter governing the performance of dry coating indicating what role is played by the process intensity and time. Interestingly, a major take home message of this work is that while coating of nano-sized flow aids requires high-intensity processing, subsequent processing of the blends in a conventional low intensity mixer may be adequate since it leads to synergistic transfer of flow aids that greatly enhances the blend flowability. This outcome has significant industrial implications in designing of powder blends and processing and will be one major focus of the future work beyond year three.

This synergy along with the effect of blend mixing time were further examined to uncover the effect of mixing time on blend bulk properties when a dry coated component is mixed; detailed in a published manuscript, see Appendix D.

Current work, which will continue through 2024 and beyond will examine applicability of industry relevant, potentially scalable approaches to dry coating. This includes assessing the dry coating performance from a batch device such as LabRAM against using low-intensity batch-mode V-Blender, as well as potentially continuous, higher intensity COMIL while varying the processing parameters. These outcomes will be included in a manuscript which will also form basis to propose additional work beyond year three.

Last, these significant advancements during the current reporting period would not have been possible without regular IFPRI member interactions gained through regular update meetings and additional tutorial style meetings as needed. Such activities will continue and a workshop on this topic will be offered in Fall 2024.

This project was a continuation of our previous work on developing atomization model. The goals of this work were to understand the breakup of ligament of highly viscous, rheologically complex fluid under given stretch rate and to predict the droplet size distribution from the ligament breakup. We divided the work in three stages: i) breakup of ligament, ii) breakup of ligament in air crossflow and iii) application of ligament breakup model to spray. The current report is on the first stage of the work.

We have completed a set of experiments on the breakup of polymeric ligaments by generating a liquid bridge between two rods. We found that the evolution of the liquid bridge can be studied in two stages, each with certain substages:

• Formation of a beads-on-string structure at equilibrium
  • First-generation beads form due to surface wave described by Rayleigh-Plateau instability.
  • Second-generation and higher generation beads form due to the pressure difference between the beads and string region.
  • Beads will slide along the string until they reach their equilibrium positions.
  • Unequal end drop sizes can create a liquid flow from small end drop to large one. This will generate a fresh, undisturbed ligament near small end drop. Later, a reverse liquid flow will be generated after the stopper bead formed at the large end drop side.
    • The initial liquid flow (from small to large end drop) delays the formation of first-generation beads.
    • The reverse liquid flow (from large to small end drop) will let a segment of the liquid bridge be absorbed by small end drop, while the rest is non-uniformly stretched.
  • Beads have multiple sizes at the equilibrium state.
• Breakup of liquid bridge from the equilibrium stage
  • Tension in strings starts to relax. This makes the string thinner and longer. The tension force in the string reduces. This process starts with two end strings.
  • Distances between the beads get closer, causing the smaller beads to move towards locally large beads and merge together.
  • Number of beads largely decreases at the time string breaks, generating multiple large beads.

Crystal morphologies are governed by growth conditions such as supersaturation and temperature as well as by the bonding structures of growth units within the crystal lattice. The vast majority of organic crystals of interest exhibit noncentrosymmetric growth units which feature anisotropic bonding interactions. Bonding anisotropy results in the presence of multiple distinct growth units within the unit cell which generate complex periodic bond chains and edge stability phenomena, presenting a significant challenge for contemporary morphology prediction tools. In this report, we consider the case of noncentrosymmetric molecules with two distinct growth units present in the unit cell (Z = 2).

Current Funding Period Advances

• Extended the use of kinetic Monte Carlo (kMC) simulations to predict the morphology of noncentrosymmetric organic crystals grown from solution.
• Expanded the development of a novel crystal growth model for asymmetric organic molecules with two molecules in the unit cell.

Spray drying is an advanced drying technology that is used in various industries. The development of the spray drying process is closely linked to the dairy industry and is require for longer shelf life of many food products. The origins of spray drying date back to the 1800s, but it was not until the 1850s that the process began to be used on an industrial scale. This method was further developed, and today a wide range of products are spraydried, with capacities ranging from a few kg/h to several dozen tons/h. Spray dryers are currently used for research and commercial purposes for drying agrochemical and biotechnological products, fine and heavy chemicals, dairy products, colorants, mineral concentrates and pharmaceuticals.

Spray-dried products

Spray-dried products can be divided into three groups, depending on the morphological structure of the particles (Walton & Mumford, 1999):

• skin-forming;
• porous;
• with crystalline structure.

As a rule of thumb, organic substances belong to the group of skin-forming materials, water-soluble inorganic substances to materials with crystalline structure, and water-soluble inorganic substances to porous materials. However, it is important to remember that this is only a general classification and that there are exceptions (Walton, 2000).

Skin-forming materials

Skin-forming materials are spherical and have a relatively smooth surface, which is usually filled with gas. With this type of material, drying initially takes place on the surface of the droplet, resulting in a thin, hard outer layer known as the "skin" or "shell". The thickness of the skin is usually between 50 and 130 µm. During the process, the skin thickens and a solid or hollow particle is formed, depending on the material being dried. Hollow particles tend to collapse after drying, while solid particles retain their shape. At a higher drying temperature, around 200 °C, the skin is quickly formed, whereupon the gas trapped inside the particle ruptures and causes it to collapse. At this temperature, the thickness of the "crust" is between 30 and 50 µm. In some materials, secondary bubbles can form in the original particle. This is caused by a certain amount of residual moisture inside the particle.

Skin-forming materials include: Sodium silicate, sodium dodecyl sulfate (SDS), potassium nitrate, gelatin, skim milk, chicken eggs (Walton & Mumford, 1999) or maltodextrin (Zbiciński & Kwapińska, 2003).

Porous materials

Porous materials, also known as agglomerates, consist of individual particles bound by submicron dust or a binder. The particles generally have a regular, spherical shape. Drying of this type of material is achieved by gradual evaporation of moisture from the interior of the particles. The highly porous structure allows water vapor and gasses to flow freely from the inside of the particles to their surface. This explains the high degree of sphericity of the particles and the rare occurrence of irregularities on their surface.

If the initial particle size of the suspension is much larger than 1 µm, the particles tend to form a solid structure upon drying, while the resulting particles are hollow if the initial particle size is less than 1 µm.

In contrast to skin-forming materials, the morphology of porous materials practically does not depend on the drying temperature. Only the drying speed changes with the temperature. Porous materials include, among others: Silica, colloidal carbon, cocoa and some detergents (Zbiciński & Kwapińska, 2003).

Materials with crystalline structure

Materials with crystalline structure are characterized by a highly ordered structure of their atoms or molecules. The solid phase is formed by the growth of crystals on nucleation centers on the droplet surface. In this type of material, the morphology of the particles depends largely on the type of substance. Sodium chloride, for example, forms large, cubic crystals, while sodium benzoate forms small, elongated crystals. Both solid and hollow particles can occur, and a relatively large shell thickness of the hollow particles is observed, namely 200-300 µm.

At a higher drying temperature, over 200 °C, a phenomenon analogous to that observed with skin-forming materials is observed, namely the disruption of the particle structure and the secondary formation of nucleation centers. In addition, materials dried at a higher temperature are characterized by much lower shell thickness of the empty particles, i.e. 50-100 µm.

Materials with crystalline structure include, among others: Sodium chloride, sodium carbonate, zinc sulfate, sodium pyrophosphate, sodium benzoate, sodium formate (Walton & Mumford, 1999).

Research program

The research program presented for IFPRI assumes finding the relationship between the rheological properties of the solution as well as the drying speed on the morphology of the particles obtained by the spray drying method. For this purpose, in the second year of the project, the following tasks were assigned:

• Carrying out spray drying experiments on the semi-industrial scale.
• Analysis of the physicochemical properties of the obtained powder samples.
• Preparation of a monodisperse droplet generator to construct devices to measure drying kinetics.

The objective of this research project is to assess and enhance the ability of a recently advanced high-fidelity modeling framework for spray formation to model complex liquid break-up and predict drop size distributions in high viscosity and non-Newtonian liquid atomization systems, such as found in spray drying applications. This framework, which has been developed by the PI’s research group, hinges on two key components:

1. a fully conservative Eulerian interface tracking technique with the ability to capture sub-grid scale liquid features such as thin films and thin ligaments, known to be of critical importance in the break-up of viscous and non-Newtonian fluids, and
2. simple physics-based break-up models to convert these thin liquid features into spray droplets that can be tracked in a Lagrangian fashion.

The focus of the work to date has been studying how complex fluid rheology (e.g., liquids with high viscosity and non-Newtonian behavior) alters atomization physics. Preliminary analysis of high viscosity liquid atomization in a pressure-swirl configuration has shown conventional liquid-gas interface techniques lead to break-up that is not physics-based (i.e., the break-up is caused by numerical errors) while our newly developed interface tracking method is able to maintain the thin-conical sheet that occurs during pressure-swirl atomization of high viscosity liquids. Additionally, non-Newtonian constitutive models have been implemented and tested in benchmark flow configurations with initial verification and validation matching published works. These models were implemented in a smaller scale flow configuration and were shown to have significant impact on liquid structure break-up, and for the case of high viscosity liquids, impact the mean droplet size.

Going forward, the focus will be to continue studying how complex rheology impacts liquid structure break-up and using those lessons and observations as guidance for how our current model can be adapted to account for complex liquid rheology. An updated model will then be implemented in large-scale, industrial-type atomization systems, which in turn, will allow for comparison with experimental measurements for drop size distributions.

The state of the art of capillary suspensions was described in a review paper published in Current Opinion in Colloid & Interface Science [1]. Our model system for capillary suspensions consists of fairly monodisperse, spherical, silica particles fluorescently labeled with rhodamine B isothiocyanate in a mixture of 1,2-cyclohexane dicarboxylic acid diisononyl ester (Hexamoll DINCH) and n-dodecane, with added aqueous glycerol. Capillary suspensions are prepared using a two-step ultrasonication: an emulsification of both liquids after which the particles are added and sonicated again. The three components are all index matched and the silica contact angle can be modified [2]. The attractive interaction strength can be modified by tuning the contact angle, and fraction of secondary liquid, and, by changing the particle roughness. This lets us access both granular-like systems with weak interactions and strong attractive gels using the same model system. During the project, we switched from using porous silica particles to nonporous particles as there were problems noted with adsorption of the secondary liquid into the pores. This means the particles are detected on the images as red rings rather than filled circles. To detect the particles, a self-coded particle detection algorithm based on edge detection and Hough transform was designed. The algorithm includes a simple graphical user interface for both local detection and manual addition or removal of missing or misdetected particles to improve the final detection efficiency. Using consecutive images on the rheoconfocal, the displacement of the particles can be detected as a function of the applied shear. AI tools have also been used to detect the particles.

Our initial goal of the project was to track microstructural changes in the network in response to external shear applied via a linear shear cell. These structural changes were correlated with the rheological response of the material. Application of external shear via the linear shear cell, however, was unsuitable. Due to the very low yield strain in capillary suspensions, the applied shear was often above the flow point and specific changes during yielding could not be adequately captured. Furthermore, the prior setup only allowed for the deformation profile to be captured in one shear plane. While this has provided valuable information, proving that capillary suspensions tend to undergo solid-body movement, where the rotation of particles around their respective bridges is resisted through both the structure of the network and the extra torque provided by the contact angle pinning and/or the contact angle hysteresis, full 3D tracking is necessary. Therefore, a rheometer was mounted onto a highspeed confocal microscope in 2022. The improved setup has allowed us to directly compare bulk, rheological changes with local, microscopic changes to the clusters and network, as discussed further in Section 2.1.

Experimental Approach to Investigate the Rheology of Powders

We propose an experimental approach to investigate the rheology of powders and their behavior during compaction and aeration processes. The first step is to develop protocols to synthetize and characterize model cohesive granular materials. The aim is to synthetize particles with tailored properties (stiffness and adhesion) using two technics (micro-polymer particles, or polymer coated silica particles).

Steps Involved

1. The first step involves developing protocols to synthetize and characterize model cohesive granular materials.
2. The second step involves developing tools to characterize particle properties and their bulk rheology.
3. The third step will involve studying different flow configurations encountered in packaging processes.
4. The final step concerns the coupling with air.

Executive summary

The main objective of this project is to apply a pneumatic nozzle design, the Air-Core-Liquid-Ring (ALCR)-nozzle, for spray-drying of highly viscous liquids and pastes. The project is divided into three main working packages (WP). WP 1 aims to validate the ACLR atomizer technology to enable spraying of highly viscous liquids, using both experimental measurements and CFD simulations. WP 2 aims to evaluate the impact of the composition and morphology of the atomized droplets on the drying kinetics, for highly concentrated feeds. WP 3 aims to join the results of both packages to investigate the applicability of the ACLR nozzle for spray-drying of highly viscous liquids. The followings findings were achieved in the present funding period.

WP 1: Atomization with the ACLR nozzle

• A model system of maltodextrin solutions was chosen and characterized for different dry-matter concentrations, which reaches viscosities up to 3 Pa·s at 20°C at 103 s-1.
• The experimental analysis of the ACLR nozzle internal flow instabilities was extended for viscosities up to 1.3 Pa·s.
• An automated algorithm for measuring the spray angle in real time was developed.
• Experimental droplet size distributions were measured for MD solutions with a dry-matter concentration of up to 57%. The distributions show that atomization is, in principle, possible. The bimodality of the DSD, the high number of large droplets and the apparent time-instability of the distribution still need to be further investigated.
• A CFD model was implemented to represent the internal flow of the ACLR, and has been validated for viscosities up to 0.14 Pa·s.

WP 2: Evaluation of the impact of the composition and morphology on the drying kinetics and model development by single droplet drying

• Hanging-droplet experimental setups were identified to be an appropriate method due to their advantages in measuring drying kinetics. CFD was employed to investigate the flow characteristics within the drying channel and to ensure an even and stable airflow. Pressurized air is utilized for drying, permitting volume flow rates of up to 850 cm³·min and temperatures of up to 200 °C. Droplets are generated using a syringe and are then transferred to the 0.3 mm thickness glass filament.
• Exploratory experiments were conducted to evaluate the experimental setups viability in determining drying kinetics, showing good results for solid contents of up to 20 wt.%.
• Particle morphologies and particle contour area were successfully captured using a high speed camera.

WP 3: Proof-of-concept of industrial applicability of the ACLR nozzle for spray-drying of highly viscous liquids

• This WP is planned to start on the funding year 2024-2025.