If you work in drug formulation, you know this problem: a formulation looks fine on paper, but a small change in drug load suddenly collapses release. Or two nearly identical systems behave completely differently, and you don’t know why.
That’s when the most important step is not to run another experiment but asking the right question. One that points to a mechanism. If the answer reveals the cause, you know what to test next, and what not to waste time on.
This mindset is why most of our best ideas start with a client problem we can’t ignore. Over time, these real formulation questions have become the engine behind how we work at amofor, and they’re now a major input to our internal research roadmap.
From Academic Curiosity to Real-Formulation Problems
At its core, amofor is a modeling and science-driven company. Everything we do is anchored in physical laws, thermodynamics, and a strict scientific method. That foundation came from academia, and it still defines how we think today.
In academia, research often begins with curiosity. Scientists explore questions that appear intellectually interesting, elegant, or theoretically meaningful. This environment is extremely valuable because it allows fundamental ideas, methods, and theories to develop without the immediate pressure of application.
At the same time, the questions that drive academic research do not always originate from the specific constraints and challenges faced in industrial development. Bridging this gap between fundamental understanding and real-world formulation problems is where applied scientific work becomes particularly powerful.
When we left the university, we had a powerful simulation toolbox in our hands. But we faced the question: Does this actually solve the problems formulation scientists struggle with every day? In other words, we had a hammer. But we did not yet know whether the nail fit.
The first industry projects were eye-opening. We learned how formulations are developed in practice, where workflows break down, and which uncertainties truly slow projects. We saw how often development relied on empirical correlations, and trial-and-error approaches, not because scientists lacked skill, but because predictive tools were missing.
That transition marked a fundamental shift. Those questions forced us to look critically at our own tools and assumptions and adopt a deliberately problem-focused approach to research.
Client Questions as Our Most Important Research Input
Every client project at amofor is also a learning process for us. Each formulation challenge reveals which aspects of our models matter, which do not, and what scientists really need.
Drug formulators ask highly specific questions like:
- Why does this formulation crystallize after 2 months, while a nearly identical one does not?
- Why does drug release suddenly collapse above a certain drug load or without a certain surfactant?
- Which excipients actually matter for this molecule, and which are irrelevant?
This is why we rely on first-principles modeling wherever possible: it helps us identify the true drivers, critical molecule attributes and process parameters, early enough to support a Quality-by-Design workflow.
Sometimes this means expanding our models. A client may need predictions for specific compound classes, surfactants, or lipid systems that were not part of our original scope. When those components are relevant in real formulations, we integrate them. The model evolves because reality demands it.
And we are skipping things and focus on what actually matters. Shelf life is a good example. Many amorphous formulations fail due to crystallization, and for those systems, shelf-life prediction is critical. But not every molecule is a fast crystallizer. In some cases, crystallization risk is negligible, while dissolution behavior or release mechanisms dominate performance. In such situations, shelf life modeling adds little value.
Over time, question patterns emerge. When the same type of question comes up again and again across different projects, it signals a genuine gap in the industry’s toolkit. Those gaps define our research priorities. For drug formulators, this means tools that are built around real decision points, what to test next, what to ignore, and where risk actually comes from.
Examples of How Drug Formulation Questions Shape Our Problem-Driven Research
One of the examples of client-driven research evolution at amofor is our work on developing dissolution predictions.
For decades, the industry has relied on empirical approaches to describe release behavior. Dissolution curves are measured experimentally and then fitted with correlative equations that describe how concentration changes over time.
These equations can reproduce a single dissolution curve quite well under the specific conditions in which it was measured. However, they are typically descriptive rather than predictive. Once experimental conditions change, for example pH, formulation composition, or particle geometry, the fitted parameters lose their validity and the equation must be re-fitted to new data.
In other words, these models capture observed behavior under one set of conditions, but they provide limited guidance when phase changes, recrystallization, or transport limitations begin to dominate release.
Our approach is different.
We want to calculate how a formulation will behave based on intermolecular interactions. From that formulation state, we then want to understand what the dissolution curve will look like, while simultaneously accounting for phase changes within the formulation, including phase transitions, recrystallization, and particle size distribution.
This quickly becomes complex. The number of variables increases, and the degrees of freedom grow. But the advantage is that instead of testing blindly, we can directly screen virtual formulations and explore how they are expected to behave.
This is about predicting dissolution behavior mechanistically, and understanding how it can be adjusted and improved. In practice, this means simulating virtual formulations and evaluating how their dissolution curves evolve, which phase transitions occur, and whether additional effects, such as diffusion limitations, interfere and disrupt release.
That is the goal we are working toward.
We already have strong predictive capabilities for phase equilibria and solubilities (for example: Dissolution Mechanisms of Amorphous Solid Dispersions: Application of Ternary Phase Diagrams To Explain Release Behavior or Double action of HPMCAS as a dry binder and precipitation inhibitor in ASD tablet formulations of nifedipine prepared by hot-melt extrusion). The next step is to combine these strengths with other models. By linking them into a single simulation framework, we can understand which phenomena dominate at which time during dissolution.
Looking further ahead, we ultimately aim to build a digital twin that can mimic the entire life of a molecule: from manufacturing, through storage, to dissolution and performance. Not to eliminate experiments entirely, but to use the digital twin to replace blind screening with targeted, hypothesis-driven experiments.
Christian Lübbert, PhD
What also matters in research is validation, and validation depends on network. Over the years, amofor has built strong partnerships with industry clients and collaborators across Europe and North America. This gives us a broad view into formulation reality and chances to test models in real amorphous formulation systems. At the same time, we stay closely connected to the scientific community through publications and academic collaborations.
Bring Us Your Formulation Problems and Questions!
amofor has opened the door to applying a physical modeling approach to drug formulation, but it’s your questions that determine what we build next and why. If you are facing a formulation challenge, we encourage you to discuss it with Dr. Christian Lübbert.
In an initial discussion, we can walk you through a relevant case study and outline what a physics-based approach can (and cannot) clarify for your molecule, including the key drivers and the minimum set of targeted experiments needed for validation.
For best results, engage with us early in development. Early engagement gives the highest leverage: it allows to build a structured strategy that systematically de-risks your path to a robust formulation.
