Related Experiment Video
Updated: May 6, 2026
11:14
Designing Silk-silk Protein Alloy Materials for Biomedical Applications
Published on: August 13, 2014
17.7K
Microfluidic production of silk fibroin nanoparticles: Process optimization and modeling
Edoardo Bertania1, Angelo Modena1, Alessandro Caimi2
1University of Piemonte Orientale, Department of Pharmaceutical Sciences, Largo Donegani 2, 28100 Novara, Italy.
International Journal of Pharmaceutics
|March 2, 2026
Summary
We developed a framework combining experiments, machine learning, and CFD to optimize silk fibroin nanoparticle (SFN) production in microfluidics. This approach identifies optimal conditions for creating small, uniform SFNs for drug delivery applications.
Area of Science:
- Biomaterials Science
- Chemical Engineering
- Nanotechnology
Background:
- Silk fibroin is a promising biomaterial for drug delivery, forming nanoparticles via self-assembly.
- Microfluidic nanoprecipitation offers control but is difficult to optimize due to coupled parameters.
- Understanding structure-process relationships is key for reproducible silk fibroin nanoparticle (SFN) production.
Purpose of the Study:
- To develop and validate an integrated experimental-computational framework for optimizing microfluidic silk fibroin nanoparticle (SFN) formation.
- To elucidate the influence of operating conditions on SFN size and homogeneity.
- To provide a mechanistic understanding of fibroin self-assembly during microfluidic nanoprecipitation.
Main Methods:
- Systematic experimental investigation using Design of Experiments (DoE) for silk fibroin:acetone ratios and flow rates.
- Machine learning models (Linear, power-law, Random Forest) to predict nanoparticle characteristics.
- Computational Fluid Dynamics (CFD) simulations to analyze mixing, shear, and solvent distribution.
- Nanoparticle Tracking Analysis (NTA) for quantifying particle size and distribution.
Main Results:
- Optimal conditions for small, homogeneous SFNs identified at intermediate-to-high flow rates and higher acetone fractions.
- Machine learning models effectively captured non-linear parameter interactions.
- CFD simulations revealed key descriptors (e.g., Volume of Change, process efficiency) correlating with controlled desolvation and uniform SFN formation.
- The framework successfully elucidated structure-process relationships in microfluidic SFN production.
Conclusions:
- An integrated experimental-computational workflow enables rational optimization of microfluidic nanoprecipitation processes.
- CFD provides mechanistic insights into the complex interplay of fluid dynamics and self-assembly.
- This approach facilitates hypothesis-driven refinement for producing silk fibroin nanoparticles (SFNs) and similar systems.
