Let's talk
Contact us

Q&A: Rapid and lossless prototyping of HME formulations in the milligram to gram scale via Vacuum Compression Molding

Dr. Daniel Treffer, Melt Prep

Q: How did you put the diffusion barrier on the side on Page 26?

A: Above is slide 26 for reference. The three layers were produced individually during three VCM cycles. Then the three layers were stacked inside the VCM Disc Insert for a fourth VCM cycle, molding all three layers to form one solid sample. The process for this usage-case is visualized below.

Q: Is there any study on drug migration between the two layers in a multilayer tablet which can be produced using the VCM?
A: When starting by VCM prepared layers as shown by the IVR example, you will obtain sharp boundaries, as shown in the images. One has to keep in mind that diffusion will try to equalize concentration gradients, bringing the API, for example, from API rich region into a less saturated region of a different layer. To visually illustrate that we have performed experiments with color tracers. In this case, we used iron oxide with EVA. Very helpful in studying diffusion behavior as everything happens in basically 1d. One can think of cutting down individual layers with a microtome and do content analysis allowing the precise deviation of the diffusion coefficient in the material.

Q: Could you discuss a little bit about the energy input of VCM? Like thermal energy or specific mechanical energy?
A: The mechanical energy input is negligible. VCM does not have rotating or mechanical mixing elements. The heat required to melt the material comes in via thermal conduction from the pre-heated hot plate (170°C) and reaches at maximum the hot plate temperatures, which makes it suitable for setting precise temperature maximums in the system, without risking degradation of one of the rare sensitive materials.

Q: How well do the API and polymer powders need to be mixed to obtain a homogeneous amorphous disc? You mention the use of cryo-milling, is there a specific particle size needed to get this good mixing?
A: This question cannot be answered with one magic value. It pretty much depends on the material system. Some of the materials need no milling, and an amorphous system is obtainable directly with VCM starting from a physical mixture of API and carrier material. For example, the XRPD data shown on Slide 22 was obtained by physical mixtures indicating that full amorphization was achieved by VCM starting from physical powder blends.

When the API concentration approaches the solubility limit within the polymer, it might need the additional milling step to achieve a homogeneous amorphous result. Through cryo-milling, the particle size is reduced, but also, the mixing is improved to a homogeneous level on the µm scale. New particles can be formed while others break. The newly formed particle may consist of fragments of all components meaning the homogeneity is also improved on levels smaller than the particle size. Thus the particle size is not always a good measure to judge the mixing results. As a rule of thumb, when the main particle is below 50µm, the result will most likely be an extrusion-like result.

Q: Have you used the small cylinder molds to make solid dosage forms to be dosed for animal testing?
A: We are an equipment vendor, we are not conducting drug development. What I can tell from our customers’ interactions is, that we have some customers who have already done that. We have a customer who is working on 2 mm implants and has already implanted them in rabbits.

Q: How much time does it take to obtain a pallet?
A: As a rule of thumb, one can say less than 10 minutes. This depends on the material properties and what you would like to achieve. We know that customers have very short preparations of 2 mm samples in about 3 minutes per sample. 1:30 heating and 1:30 cooling. But there are also examples in which customers work with curing materials, which require times over several hours. It depends heavily on the material system.

Q: Is this disk easily milled using conventional mills?
A: This mostly depends on the carrier material properties. If the carrier is easy to break down and can be broken down with a coffee grinder or mortar and pestle, then yes. However, if you choose suitable carriers, you can also think of making abuse-resistant formulations. We have customers working on abuse-resistant formulations, and what we have heard from them is, they are working with PEO, rubbery materials or similar materials. There was an interesting formulation of Grünenthal at the European Meeting 2019 in Bologna, showing how they used our tools in one of their developments.

Q: What is the residence time of the material in the device?
A: The RTD is your choice and will depend on the material properties. There will be a minimum time, which depends on the starting material properties to melt and fuse them. As a rule of thumb, the preparation time usually takes less than 10 minutes. A typical temperature curve is shown below. If you are worried about degradation, you might benefit from having a vacuum in the chamber so that the risk of degradation by oxidation is not evident.

Q: Not much familiar with the device, but as you mentioned, the molted slug formed as a function of pressure. Did you see any effect of vacuum on drug release? How do you correlate that with extrusion?
A: We didn’t see any effects of using vacuum during sample preparation on drug release. The standard vacuum level applied during VCM Processing is less than 1/1000 of an atmosphere. But it can be adjusted to higher absolute pressures if low melting point APIs are prepared via VCM at high temperatures. An example could be the low melting point API, Fenofibrate, with a melting point of 80°C, processed with Soluplus at 170°C. In such a case, Fenofibrate would be at risk for evaporation. The results obtained via VCM mimic results obtained with hot-melt extrusion. Applying a vacuum is not new during the processing of thermoplastic materials. It is quite frequently used to dry material as a preconditioning step, or also hot melt extruders can be equipped with a vacuum zone when moisture or volatile components should be removed during processing.

Examples that good agreements between VCM and HME are obtained can be seen in the IVR publications mentioned on slide 26 or the screening comparison on slide 22. Thomas Kipping from Millipore Sigma showed in one of his presentations last year how well VCM samples correlate with hot-melt extruded samples. They have a similar pattern in the PXRD can, and all samples show full amorphization of the crystalline drug.

Q: Is this scalable. If so, what throughput have you achieved so far?
A: Yes, definitely! If higher quantities of one formulation are required, it is scalable by parallelization if we are talking about small batches. In my opinion, this makes sense for small scale productions of let’s say something like a 100 samples. The parallelization is achieved by using several VCM Tools. One is on the hot plate, one is on the cooling unit, and one is filled with the materials. The tools are in an alternating sequence at the individual stations. With an assumption of 10 minutes per sample, one could achieve 12 samples per hour and ~100 samples during one intense 8-hour working day.

If significantly higher quantities of one formulation (>100g) are requested, we do recommend for expensive material to determine their rheology properties on small scale VCM samples, predict the processing temperature on an Extruder and transfer them to a small scale extruder for continuous production of extrudates.

Q: What is the impact of residence time? How do you factor in residence time in extrusion?
A: A comparison of RTDs between HME and VCM has only limited meaning as the processing conditions are different (no mechanical mixing). It will yield a very similar result. For VCM, a certain time is required to allow complete warming up to a set temperature. After this, it can be cooled down again. The temperature seen by the sample can precisely be monitored with the supplied bluetooth temperature logger.

Q: I have only 1 mg of active can I use VCM?
A: Yes, you can. Please keep in mind that the material will be shaped in a lossless procedure into your sample. You just need the material to produce your specimen. If you desire to obtain a glassy amorphous system, it also depends a little on the material properties you are working with. If you work with formulations comprising different substances, it must be homogeneous and micronized that this is possible. Preferable in the small micrometer particle size range. But generally, if the density is ρ=1g/cm³, we can prepare samples as tiny as 0.5 or even 0.3 in diameter with material amounts between 20 and 100 μg. If the loading is 10%, 2 or 10 µg are required, respectively.

Q: Can you handle sensitive to oxygen or purge nitrogen?
A: Definitely, yes. As a rule of thumb, when you expect that the formulation can be processed via extrusion, then it will also give a good result with VCM, but it many cases, more can be achieved. Jörg Rosenberg (30 years of experience) from Abbvie stated in his slide at the ASD Masterclass in Ludwigshafen, that degradation of API is very rare when the extrusion is performed right. I mentioned the example of Soluplus during the live Q&A sessionWe did some investigations on how we were able to prepare neat Soluplus samples below the generally applied processing temperature for Soluplus, which usually lies in the range of 160-170°C (zero viscosity levels of 3000-5000 Pas) for the neat polymer. VCM samples were transparent and glassy down to temperature levels of 100°C (reaching in the 1 Million Pas Range, three orders of magnitude higher), which corresponds to a significant reduction in processing temperature lowering the risk for degradation. Also, a nitrogen purge is easily possible as it only requires a slight modification by the addition of a 3-way valve into the existing equipment and such tiny modifications are always possible when requested.

Q: Would we be able to simulate the HME via the VCM tool and create a predictive model of material properties?
A: Yes. Definitely. Please check out the publications on our page:

Q: How much stability data have you generated to date?
A: I think that is a question that will be of interest for a lot of people. We are a lab equipment provider; we do not conduct any research studies. But what we have seen in the work of our customers is, that the samples show very similar behavior to hot melt extruded materials. And some data is already published by our customers. Please check out the publications listed on our webpage for more information:

Q: How about the shear rate in extrusion? Your system does not apply shear! Is this really a good simulation for extrusion?
A: It’s true, the VCM process uses diffusion only to mix materials. This is working well when you use small particles, like micronized powder. If you have pellets, we recommend to cryo-mill them or do a solvent casting step before you use VCM. Our results compare exceptionally good to extruded samples. Also, consider that VCM is not having any shear heating and related risk of degradation caused by temperature increase by internal friction. Please see also question 9.

Q: Can liquid processing aids be screened through VCM to assess their impact on solubility enhancement?
A: VCM is in general not designed for working with low viscous liquids, the materials should have at least a viscosity of 100 Pas. However, there are always ways to test also liquid materials.

I will describe two frequently used methods:

1: Curing reactions: We have customers working on curing reactions, and they usually cure it to a viscosity of range of 10 – 100 Pas and load the sample than in the VCM Chamber for the final curing step. The optional low-pressure lid is having a lockout function for the compaction force, which is very helpful in this case. Thus, shaping via pressure can be added a level above 100 Pas, giving them a nice sample.

2: Pre-conditioning via milling or solvent casting. When liquid additives like Tween, Oils, or similar things are considered which are miscible with the carrier, one can think of putting them into the milling step and/or solvent casting step similar to the steps described on Slide 14. For proper distribution, they might be frozen before milling.

Q: This is a manual method, can it be automatized?
A: Our goal is to offer an easy and fast-forward tool for sample preparation. Let us look at the question from two angles.

1: When you already found your formulation, you can think of scaling-up by volume. Currently, VCM can make up to ~ 2-5 g per batch, depending on the bulk density of the material to be processed. If larger scales are requested, this is recommended by extrusion or a combination of extrusion and injection molding. We have built some custom tools which also enabled larger quantities, and this is something that we can do on request.

2: If you think of handling multiple formulations in an automated manner, this might be possible in the future, making VCM a large piece of equipment in comparison to what it is today. Reach out to us if you are interested in that direction to check if we can prepare something for your needs.

Q: Would content uniformity change during VCM processing, esp. migration of active to the surface?
A: This is a very good question. With micronized powders, it will be homogeneous, and we haven’t seen any API migration. With larger particles, you may see a low melting point API melting go through the porous network enriching the concentration at the bottom. With multilayer systems, you will see this diffusion/migration from API from fertile regions into lower concentration regions, compare the illustration shown in answer to question 2.

Q: Any commercial product using VCM technology?
A: This is something that I will answer with the necessary caution. I am not allowed to release sensitive information. We started to introduce our products in 2016. This is a very short time compared to the duration of a drug development process. What I can say is: Our tools are used by 4 out of the 10 biggest pharma companies worldwide. If you look up the publications, you will find names like Abbvie and Millipore Sigma mentioning VCM. And a lot of companies working on NuvaRing-like implants are using VCM to screen new formulations.

Q: Have you worked with Cannabis before and the dissolution of its oil in nanocrystalline?
A: We never used VCM with Cannabis so far, but I think this is something worth trying. Generally, we work with thermoplastic materials that have a viscosity of at least 100-pascal seconds, and if the carrier is in that range, it has good chances to work. We know that our other product, the VChamber, has attracted attention for gentle vacuum drying of herbs and other natural substances. Feel free to reach out to us in case you would like to try something.

Q: How’s the scale-up with VCM? And what are the challenges?
A: Once you found the right formulation and want to start to produce higher quantities, you can think about working with extrusion. VCM samples have extrusion-like properties, and scale-up can be performed by switching the formulation to a continuous twin-screw Extruder. You are having then a throughput of 200g/h on a 12 mm twin-screw for first clinical investigations and in a later stage a larger extruder, e.g., 27 mm with several kgs per hour. A proper procedure for the HME process development was described in a recent publication (open access), which might be a good read for you.

Q: As you know HME in general, is limited to the thermolabile molecules. What is your experience with these types of molecules, and more specifically, peptide therapeutics?
A: This is an excellent question. I answered this one in the live Q&A session for a small molecule by saying most APIs are not thermolabile, and this is one of the misconceptions which has been around for several decades. This is not only out of my experience. This was also discussed by Dr. Jörg Rosenberg from Abbvie at the APV ASD Masterclass a couple of years ago, and he stated that out of his 30 years of experience he has only seen a few which are not thermally stable. The majority of APIs were considered extrudable out of his experience.

However, when we look at larger molecules like proteins, peptides that are more sensitive to heat, the choice of a suitable carrier material to not to exceed the degradation temperature becomes more important. Carrier melting / softening points can be in a range of 50°C, which is also possible for VCM. The low-pressure lid helps to process low viscous materials and also has a lockout feature for turning off the compaction pressure – enabling work with curing carriers.

Q: Why cryo-milling prefer? Any way material used for VCM is thermophilic?
A: In general, lowering the temperature will make the material during milling more brittle and, therefore, more comfortable to reduce the particle size. We recommend cryo-milling or solvent casting to obtain small particles, to increase the homogeneity on the micro-scale. Diffusion is the primary mixing mechanism in VCM, and it can be empowered by making small length scales. Thus extrusion-like results are ensured. Please also take a look at questions four or the slides (13-14) in the presentation.

Q: How long does it take to make a single disc?
A: Usually, the sample preparation process for a single disc takes less than 10 minutes. But this is just a rule of thumb. It always depends on the material property. Some of our customers prepare samples in only some minutes, while others process their curing material for hours. Please have a look at question #6 for more information on that.

Q: As we see, some solubility difference between melt extrusion vs VCM. Once we develop the method using VCM, how can it be scalable to the lab & commercial scale?
A: Differences are minimal and, in most cases, negligible when cryo-milling is performed. To avoid stability issues, it is recommended to never design the formulation at its solubility limit as the risk for nucleation increases. Having some safety margin will allow a smooth transfer from a VCM developed formulation towards continuous production. There is a recent publication (open access) out there, which might be a good read for you. I am happy to help if you have any further questions.