12. Benefits
• True rapid prototyping
• Development process becomes:
– Fail fast & often
– Early & rapid empirical feedback drives progress
• Dramatic reduction in:
– Opportunity cost to try new ideas
– Barrier to entry
• No cleanroom required
• Utilize full 3D volume
– Size reduction
– Parallel fabrication ➡ path to manufacturing
– Same tooling and materials for prototyping and
manufacturing
13. Barriers
• Feature sizes are in the millifluidic rather than microfluidic
regime
– Need features 100 µm or lose advantage of using small sample
and reagent volumes (~1 µL)
• Market pull
– Dental, custom jewelry, audiology
• Commercial 3D printers
– ~50 µm x-y resolution
– ~50 µm z layer thickness
• Proprietary commercial resins
– Viscosity (affects feature size)
– Lack of tailorable mechanical, optical, biocompatibility
properties
<
~
14. Overview
• Focus on our work over the last year
• Custom 3D printer
– High XY resolution: 7.6 µm
– UV light source: 385 nm LED
• Custom low-cost resin development
• Small channels
– 18 µm x 20 µm
• Valves and pumps
• Integrated mixer and pump with selectable
mixing ratio
– ~6.3 mm3 = (1.85 mm)3
15. Custom DLP-SLA 3D Printer
Optical Engine - Visitech
• 2560 x 1600 pixels
• 7.6 µm pixel pitch
• 1:1 lens system
• 19.5 x 12.2 mm2
• 385 nm LED
Mechanical System – Modified Solus
• Teflon film
• Tipping quartz window
• Typical build layer thickness:
5 – 10 µm
Software
• Custom in-house
developed 3D printer
control (Python)
• Open source 3D CAD
• Open source slicer
Gong et al., Lab Chip 17, 2899 (2017)
48. Take Aways
• Microfluidics ➙ It’s all about the voids
(interconnected network of microvoids)
• 3D Printer resolution specs ≠ achievable void
size
• Achievable void size - channels
– Z - resin optical properties:
– XY - projected image resolution:
• High density interconnects, valves, pumps
• Small devices ➙ parallel fabrication ➙
manufacturing with 3D printing
⇠ 2.3ha
4 pixels
49. Needs
• High resolution 3D printer
– UV LED source
– Complete control over operation of printer
• Resins
– Small ha
– Must be tailored to emission spectrum
– Open source
• Explore 3D structures
– Get away from conventional 2D thinking
• Components
• Layout
– Experiment-based exploration of parameter space and performance
optimization
• Automated design
– Library of standard components
– Automated layout
– Specify functional processes ➝ automated design generation