Skip main navigation

Fluid Bed Encapsulation – Part 1

Prof. Dr.-Ing. Erich J. Windhab (ETH Zürich)

Fluid Bed Encapsulation I - Aero-/hydrodynamic principle and encapsulation processing equipment
Hello, everybody, and welcome for the first course part on fluidized bed encapsulation in which I will focus on aero and hydrodynamic principles and the processing equipment. I will give you some subchapter information after introduction on the principles of fluidized bed technology and the hydro aerodynamics, then equipment for fluidized bed encapsulation or granulation, agglomeration, and coating, followed by particle motion and coupled fluid gas flow from an experimental and simulation perspective, and then summarise and conclude. Some nomenclature at the beginning, and let me start introducing, again, with the S-pro squared scheme, which means the coupling of process structure and properties, which is very helpful to follow because process make structure. Structure codes the property.
And from a consumer perspective, we are certainly interested to go backwards, starting with the properties, which are the preference, acceptance, and need, characteristics of the consumer. In our case of fluidized bed encapsulation, we want to have products at the end with controlled release of active components or protection of encapsulated material against unwanted reactions like with water or with oxygen during storage and so forth. But also techno-functional properties like powder flowability and so forth is of interest. This all relates to structure. So in our case on the macro scale, we have pellets, agglomerates, capsules, which are formed by mesoscalic to macroscalic particles but can also implement some emulsion drops or crystals.
And on the micro scale we have the molecules for binder materials and for the functional components. Processwise, we look for the encapsulation by either coating or agglomeration in the fluidized bed system.
Let me start to explain what fluidized bed means. And as shown on this first slide related to the process, so we can see that if we have a packed particle bed and insert some gas flow so we can, after reaching a certain volume flow rate of the gas, loosen the structure. That means we get a higher porosity and a higher bed volume, which is the particulate regime volume of this fluidized bed.
If we increase the volume flow rate, there may be some bubbles forming in this fluidized bed which can switch over to a slugging regime, then gets into turbulent flow, fast fluidization in turbulent flow, and at the very end some pneumatic conveying may make the particles move outside the reactor if you don’t hinder them to do so. So these are the different states of so-called aggregated fluidization. Most important for us will be the particulate regime, so the fluidized bed without strong motion to the bubbling regime, which is mostly interesting for what we want to do on the encapsulation.
The tables in the following slides show you characteristic velocities, superficial velocities of the gas in the system, starting with the minimum fluidized velocity, which is maybe most important in the context of what we want to look at, and then other characteristic velocities for the different states shown before, ending up with transition to turbulent fluidized and transport of the particles for pneumatic conveying, which is not in focus now but which is also certainly of interest from a processing perspective. Let’s look a bit further to the start of the system on the packed bad conditions. So when we have the flow through the packed beds, we may assume some cubic or other type of packing.
And flow through with a gas means we have flow through porous media. The so-called Ergun equation allows us to derive a friction coefficient for the flow through, which is derived from a false balance where buoyancy forces equal the drag forces. When we look at the Ergun equation, so we can also derive some critical velocity, minimum velocity flow for the fluidization, just taking the first term of the Ergun equation which is independent from the gas density. So we can derive this superficial velocity umf equals the porosity of the packed beds or the bed which starts to be fluidized times the velocity of the gas, the mean velocity of the gas in this situation.
So we can derive this equation, which can be further simplified by findings from Wen and Yu who have approximated this ratio between porosity and sphericity times porosity and found that this is equal or roughly equal to a value of 11, which leads us to the simplified equation for the superficial minimum fluidization velocity, as shown in equation number seven. This is valid for small Reynolds numbers. And what is interesting if we compare it is to the Stokes velocity for sedimenting or flotating sphere, so we can see that this is roughly a factor of 1/19 the Stokes flow velocity.
And this means that particles blown out of the fluidized bed by bubble bursting above the bed just above minimum fluidization returns to the bed since its terminal velocity is too low for it to be carried out of the bed. So the higher the solids volume friction, which is 1 over 1 minus porosity, the higher the friction coefficient will be. And this change of resistance with solids volume fraction gives rise to oscillatory behaviour of the fluidized bed.
For larger Reynolds numbers, the buoyancy equals drag equation, which we have shown before with the friction coefficient given by the Ergun equation, can be rewritten in terms of the so-called Archimedes dimensionless number, as shown in equation number nine, in which the Archimedes number is defined as shown in equation 10 and with the abbreviations of B, which is inserted in equation nine according to the approximation of Wen and Yu as given in equation 11. So from this we can express again for the higher Reynolds number range the superficial minimum fluidization velocity as given in equation 12. So this is not valid for Reynolds numbers larger than 1,000.
Some authors have done empirical refinements of these superficial velocity, minimum velocity for fluidization, and related this to particle sizes as shown in equation 13 and 14, so for particles smaller and larger than 100 microns. What we can see is the pressure loss in the fluidized bed is certainly increasing until it fluidizes from the packed-bed situation, from the static bed. We may have an overshoot due to some adhesion forces of the particles and then enter into the fluidized region. So for increasing bed-particle-size distribution, so we have a tendency of flattening of these curves, as shown in this slide.
Fluidized-bed fluid dynamics are a bit more overarchingly described from Molerus and Wirth by this diagram which is looking at the dimensionless relationships between the Archimedes and the omega number, Archimedes being kind of a dimensionless particle diameter and the omega number a dimensionless velocity. We can see for different porosities of the fluidized bed the curves along which the Reynolds number is changing. So this is the Reynolds number, the particle Reynolds number. When we look at the system parameters so we can find environmental-related ones, particle properties, and gas properties from which, by dimension analysis, dimentionless groups can be defined, which we will treat later on further.
When we go for the classification of the different types of particles, so it’s not only the particle size but particle interaction forces which play quite an important role. And so this is the famous diagram for bed fluidized from Geldart already published in 1973 and where the different groups of particles between aeratables and spoutable are subdivided. And the cohesive ones, C, which are the most difficult ones to be rated on the lower-left corner. The spoutable mostly related to very coarse particles. We can sort this also according to Arrhenius numbers where we have from 0.97 to 176,000, quite a wide range.
So as we can see, so the border lines between the different domains are not always power law or even linear type of relationships. But this has been frequently proved to be valid according to Geldart’s characterization. So in the next slide I give you a bit more information about what types of particles and particle interactions are leading to the different groupings. So group A, mostly small particles; group B, the larger particles; C, very fine particles, mostly less than 30 microns and strong adhesion forces; and the group D, very large ones or high-density-related particles, which is very difficult to fluidized because channel formation of the gas is to be expected.
For the very fine particles, because nano technologies have brought up a lot finer particles than usually treated in such fluidized bed, so there is further subdivision in modified Geldart classification diagram for fine particles down to the 1-micron size. So areas, domains of solid to fluid-like to elutriation behaviour is separated from solid to fluid-like to bubbling behaviour and then entering into the field, which we have seen before. Let’s have a look at equipment for the fluidized bed encapsulation. So equipmentwise, we have to deal with the three-phase system consisting of solid particles, liquid spray drops, and the continuous gas phase.
And what you can see is there is different, let’s say, spray directions into the fluidized bed like a top-spray granulator where we have counter flow of the liquid spray and the fluidized air. There is the opposite, bottom-spray granulator. These can be supported by a special tube to guide, let’s say, the particulate flow or so-called Wurster tube. D would be a rotor granulator with a rotating disc and with a tangential spraying into the system in order to have a rotational flow. And finally a spouted-bed granulator where we have the liquid binder coming from the bottom and without a Wurster tube having some circulation in the fluidized bed.
Next slide shows you a bit the technical forms of these types of different spray fluidized beds granulators. So the top spray, the bottom spray with the Wurster tube here also shown. The top spray also shows some continuous inflow and outflow. So we can apply this to batchwise on to continuous processes. Then we have the tangential-spray type with either rotating or nonrotating bottom part. Here also the comparison with drum coating as an alternative for larger pellets in particular. Spray granulation, high shear granulation, and drying are additional functions which can be taken in the fluidized bed systems. And schematically is shown here, including the structures, the more-compact agglomerate structures or more-loose type of structures which can be generated in the system.
Looking a bit more closely to one of the most successful systems, the so-called Wurster coating process where we have inflow from the bottom, so where the Wurster tube is kind of embedding a spray zone. The spray zone here denoted in green where the spray meets the particles. And then having a circulation of the particles through the Wurster tube and back at the walls, between the walls and the Wurster tube outside, back to the bottom, and recirculating. So we can subdivide the Wurster tube– the fountain region where the particles turn back to the bottom, the downbed region, and finally the horizontal transport region. Most important is the spray zone where the liquid droplets meet the particles in the flow.
Some more details are shown on the right-hand side. Important also this gap to have a recirculation made possible for the particles as demonstrated here on the right-hand side. So it’s not so easy to have experiments done and measure the circulation of the particles in such fluidized bed systems. But there is some interesting methodology which has been applied by Li and co-workers published in 2015 in the AIChE journal where they used the positron emission particle tracking, the PEPT technology. Means radioactive traces were with, let’s say, on an X-ray screen.
You can then follow the tracks, and you can see here some different tracks which comes out from the Wurster tube and comes back to the outside wall and down and some recirculation between or within the Wurster system. On the right-hand side you can see accumulation of different of these tracks as it looks, let’s say, with derived velocities, local velocities. It’s important to see that then residence time of particles in the different zones– the Wurster-tube zone, the fountain-region zone, the downbed region, and the horizontal-transport region– how they distribute. Most important, the residence time in the Wurster tube where we have the spraying zone because this is where the binder reaches the surfaces of the particles.
Most modern tool part or part of the toolbox is coupled discrete elements methods and computational fluid dynamics simulations, which is already pretty much advanced. And this allows us to do simulations of the particle flow and of the gas flow, like on the left-hand side in the figure showing the residence times of particles in the different zones. On the right-hand side we have from the work cited before we see the PEPT measurements, which are the roundly shaped symbols in blue, compared with the simulation results, which is quite satisfying. So simulation, DMC-CFD combined simulation is pretty close to experimental results.
As mentioned, the quite powerful DMC of these simulation allow us to also separate the flow between the particles and the gas because the relative motion of the particles and the gas decide about the mass transfer for the drying as soon as the binder has reached the surface of the particles in order to coat the particles or to form agglomerates, they have to be rapidly dried to run through a sticky type of domain very quickly. And this is where the relative motion of the gas and the particle motion is of most importance.
So DMC of these simulations is, together with the PEPT, parts of a new toolbox which will be further explored and which allows us to expect new insights also into local situations within such complex type of three-phase flows in the future. With this, let me just briefly summarise. We have seen encapsulation. Processing in fluidized bed relates to the entrapment of functional components into coated porous solid particles or the forming of agglomerates. The liquid sprayed into the fluidized bed as coating material or binder, depending on whether we coat or whether we form an agglomerate. The gas flow, the superficial velocity determines the degree of voidage in the packed bed and which enters into the fluidization state.
And from the buoyancy equals drag force balance in different flow regimes we can formulate dimensionless equations and characterise the fluidizability particle classes according to their Geldart diagram. The types of fluidized beds which we have seen are mainly different in the direction of particle motion and atomized fluid spray, which has an impact certainly on the efficiency of particle surface coating with the fluid and also the drying kinetics. Positron emission particle tracking and discrete element method-computational fluid dynamics coupled simulation are the new tools to step further into details of the fluidized bed encapsulation processing. With this, I would like to thank you for your attention and finalise this first course part on fluidized bed encapsulation. Thank you.
This article is from the free online

Drying and Encapsulation of Food Ingredients in Powder Form

Created by
FutureLearn - Learning For Life

Our purpose is to transform access to education.

We offer a diverse selection of courses from leading universities and cultural institutions from around the world. These are delivered one step at a time, and are accessible on mobile, tablet and desktop, so you can fit learning around your life.

We believe learning should be an enjoyable, social experience, so our courses offer the opportunity to discuss what you’re learning with others as you go, helping you make fresh discoveries and form new ideas.
You can unlock new opportunities with unlimited access to hundreds of online short courses for a year by subscribing to our Unlimited package. Build your knowledge with top universities and organisations.

Learn more about how FutureLearn is transforming access to education