Digitalizing magnetic microdroplets for medical diagnostics using Lab-on-a-3D-Printer driven Ferrobots

Introduction: : Microfluidic technologies enable multi-step, parallel bioassays within small-volume droplets and microwells. However, most systems require specialized, high-cost equipment for manufacturing and operation, limiting their accessibility and scalability. To address these challenges, we introduce Lab-on-a-3D-Printer driven Ferrobots, a versatile, low-cost, programmable platform that integrates microfluidic automation into a repurposed benchtop 3D printer. This system replaces the traditional printer nozzle with a multifunctional manipulation head capable of executing a wide range of ferrobot-enabled microfluidic operations—including droplet generation, transport, splitting, merging, vortexing, heating, and imaging—through an intuitive graphical user interface (GUI).

Materials and

Methods: : The system modifies a commercial 3D printer by replacing the filament nozzle with a multifunctional manipulation head containing stacked magnets, heating elements, electrodes, or piezoelectric components, and a miniaturized autofocus camera. Ferrobots, biocompatible droplets containing 2.5–5% ferrofluid, are magnetically manipulated by the nozzle across custom-designed microfluidic chips—fabricated within 20 minutes using PET membranes, acrylic sheets, and double-sided tape—to execute designated assay workflows. Depending on the design and reagents included in the chip, the platform is compatible with a range of bioassays, including colorimetric, fluorescent, and magnetic bead-based workflows, with built-in support for post-assay separation and imaging.

Results, Conclusions, and Discussions:: The system demonstrated robust and precise performance across various microfluidic operations. Droplet transport speeds exceeded 15 mm/s with 2.5% ferrofluid and 60 mm/s with 5% ferrofluid for droplets ranging from 0.5 to 10 μL. Dispensing accuracy was consistent, with 0.5–9 μL droplets achieving a coefficient of variation (CV) of 2.44%, while buffer droplets ranging from 10–25 μL maintained a CV of 3.4%. The platform preserved the dynamic range and sensitivity of both colorimetric and fluorescent assays, even in ferrofluid backgrounds. Automated serial dilutions at 2X, 5X, and 10X closely matched manually prepared controls, confirming the system's reliability. Additionally, the platform successfully manipulated and separated droplets containing ~1 μm magnetic beads, achieving post-assay transparency with over 99.8% recovery—ideal for optical imaging.

The platform can be enhanced for specific research applications through several additions: a repurposed nuzzle serves as a precision pipette and liquid pump (1–3000 μL, 1–1000 μL/min, CV < 3.0%); an integrated compact autofocus camera (8K UHD) enables high-resolution imaging; a heated print bed allows rapid cancer biomarker detection (e.g., BRCA1 via loop-mediated isothermal amplification). Combining quantifications of NT-proBNP or other biomarkers with Ferrobots, the system automates cardiovascular ELISA-based assays at minimal cost.

The future integration of large language models (LLMs) promises further streamlined experimental design and data interpretation, significantly enhancing laboratory automation and biomedical screening capabilities.

Acknowledgements and/or References (Optional):: The research is supported by NSF ERC 1648451, NIH R01HL148182, NIH R01HL169695, and Veterans Affairs I01-CX001901. Shun Ye thanks the UCLA Bioengineering department and NSF PATHSUP Engineering Research Center for providing funding. Shun Ye also thanks collaborators and undergraduate mentees for their dedication.