Agentbased modeling of fluidic selfassembly Massimo Mastrangeli Complex

Agent-based modeling of fluidic self-assembly Massimo Mastrangeli Complex Systems Summer School 2009 Santa Fe Institute

Context: electronics manufacturing �Success of microelectronics is largely driven by economy of scale � Higher production, lower price �Assembly and packaging costs do not scale! � Need for efficient techniques �Standard assembly done by robots… � Fast and precise at large scale �…do not downscale satisfactorily! � Hard to handle many small chips �More efficient techniques -> SA [Mastrangeli et al. , J. Micromech. Microeng. , in press] 07/03/2009 (c) Massimo Mastrangeli 2009 Slide 2

Challenge: Realistic models for passive SA �Components are pre-conditioned to assemble in a pre -designed collective configuration � Static SA, external drive required for mixing and transport � Lock-key driven by physical forces (capillarity, gravity, electro- magnetics, chemical) �Few analytical models exist � Rate equations, reaction-limited � Mean-field-like �How to include real-life? � Component’s excluded volume and momentum, finite assembly space, diffusion-limited reactions, dissipation and mixing, … 07/03/2009 (c) Massimo Mastrangeli 2009 Slide 3

Proposal: ABM for fluidic SA �Gas-in-a-box model in Net. Logo � Finite box and particles, temperature � Viscous drag in fluidic carrier � multiple particle types � Both 2 D and 3 D �Cooperative collision-driven assembly � Steric factors, redundancy, mixing, transport � Naturally accounts for fluctuations �Real-time statistics and interactions � Also offline in Behavior Space 07/03/2009 (c) Massimo Mastrangeli 2009 Slide 4

Perspective: tuning and exploration �Tune the model to reproduce data �Explore the parameter space � Maximize yield, minimize time-to-assembly �Encode physical interactions � Driving forces Zheng et al. PNAS 2006 � Analise environmental effects and constraints �Investigate cooperativity, selectivity, sequentiality �Work in progress… 07/03/2009 (c) Massimo Mastrangeli 2009 Slide 5