Primitive Molecular Manufacturing Chris Phoenix Tihamer Toth-Fejel NIAC Fellows Meeting, March 15-16
Small is Beautiful ÎScaling laws: throughput ~ size^-4 Î10 cm tool: 3 GY. 100 nm tool: 100 sec.
ÎPower density ~ size^-1 ÎDigital logic likes to be small ÎAccess atoms Æ Digital construction ÎLow wear, high strength, superlubricity ÎDigital design bottom-up, like computers
Molecular Manufacturing ÎBuilding precise, useful nanoscale structures with nanoscale tools ÎEngineering approach to nanoscale ÎMolecular fabrication ÎAssembly and operations ÎInformation handling ÎReliability
ÎEventual goal: Large, high-performance products including manufacturing tools
A desktop nanofactory manufactures a Filtomat M102C high performance self-cleaning water filter
The product is extruded at a surface that assembles nanoblocks one layer at a time
Ridge joints interlock to connect nanoblocks
A robot arm with thiophene-based molecular actuators adds a nanoblock to the extruded product
Outline ÎFoundational approach ÎUltimate goal ÎExisting capabilities ÎRoadmap
Information Delivery ÎExisting manufacturing technologies: A few KB per second
ÎGoal: A few GB per second
ÎThis requires: ÎVery small tools (or massively parallel medium-small tools)
ÎReliable operations
Small is Beautiful ÎScaling laws: throughput ~ size^-4 Î10 cm tool: 3 GY. 100 nm tool: 100 sec.
ÎPower density ~ size^-1 ÎDigital logic likes to be small ÎAccess atoms Æ Digital construction ÎLow wear, high strength, superlubricity ÎDigital design bottom-up, like computers
Rapid Design ÎOnce you get a set of reliable tools… You can combine them to make new tools ÎOnce you can build lots of variants… You can do evolutionary design ÎWith high performance and small parts… You can afford to use levels of abstraction
Planar Assembly ÎScaling laws indicate that deposition speed is independent of block/robot size ÎRobots can be sub-micron size (Human cell = 20 micron)
ÎRapid joining should allow cm/sec deposition Æ Build large products directly from submicron blocks (Build sub-micron blocks directly from atoms…)
Molecular fabrication ÎToday: Bulk chemistry ÎEnvironmentally controlled polymer fabrication ÎSequence of operations ÎLight (“DNA Chip”)
ÎGoal: Mechanically controlled reactions ÎSequence ÎPosition
Mechanosynthesis: Just a Tool ÎCan be either “solution phase” or “machine phase” ÎMechanical control of reactions: More a style than a restriction ÎControl can be direct or indirect ÎImplies precise nanoscale tools – which we want anyway
Error Handling in Generational Manufacturing ÎIn molecular devices, errors are digital ÎError rates vary by many orders of magnitude ÎIn small self-contained devices, you just need an error rate less than the number of operations ÎIn large devices, you will have errors ÎYou need to work around them, not pass them on
Nanoscale Toolkit ÎWe want tools that are: ÎReliable ÎDependable ÎEasy to characterize ÎVery flexible
ÎShort-term tools Actuators; Binding/recognition; Conductors; Sensors; Bonding; Structure
ÎLong-term tools Digital logic; Bearings; Mechanosynthesis
The Ultimate Goal ÎTabletop nanofactoy ÎProduces its own mass in ~1 hour ÎUses simple feedstock ÎDirect computer control; Rapid prototyping ÎPowerful nanoscale toolkit ÎVery high performance products ÎEasy product design
ÎBurch/Drexler animation (with real chemistry)
Where To Start ÎSimple 2D array of block-placing equipment ÎA “membrane” separating loose blocks from product ÎSelective placement of blocks, optionally followed by strengthening connections ÎBlocks (5-20 nm?) are pre-made ÎChemistry ÎSelf assembly ÎTemplating, mechanosynthesis
Building blocks DNA RNA Protein Schafmeister’s polymers Other…
Binding and Bonding ÎCovalent ÎMetal ligation ÎDative ÎHydrogen ÎCharge/Dipole ÎVan der Waals (“surface forces”)
Bonding Blocks ÎBy light ÎBy zinc ÎOther…
Smart Silkscreen
Simplest Nanofactory
First Improvement
Improved nanofactory
Roadmap ÎSilkscreen membrane ÎInkjet membrane ÎAdd local digital logic ÎImprove error handling ÎImprove the mechanicals ÎAdd intermediate block-assembly stages ÎAdd mechanosynthesis ÎAdd directed internal transport ÎRemove solvent ÎAdd mill-style synthesis
Goal…
ÎJohn Burch, Lizard Fire Studios
Local digital logic ÎHP’s “crossbar” claims storage, amplification, logic ÎMechanical logic? Nanoscale machines move at GHz speeds…
Improve error handling ÎWith primitive membrane and small area, you can handle errors from the top ÎWith internal function, you need redundancy ÎLocal error handling reduces data flow ÎEspecially the difficult back-channel
Improve the mechanicals ÎActuators ÎDNA (tweezers, rotary motors) ÎATP (ATP synthase, kinesin, etc.) ÎIon (linear rotaxane dimers, contractile polymers) ÎRedox (rotaxane, VPL, etc.) ÎPhoto or electrochemical (pseudorotaxanes) ÎElectrostatic
ÎBearings ÎSingle bond ÎCompliant (squishy) ÎSuperlubricity
Intermediate block assembly ÎCombine molecules into blocks ÎImprove throughput ÎEnhance flexibility ÎAdd “fins” or other structure on the feedstock side to supply the assembly plane
Mechanosynthesis ÎSimpler, cheaper feedstock ÎSmaller, less error-prone feedstock intake ÎWider range of components ÎLow level programmability (faster R&D) ÎSmaller features? ÎHigher bond density?
Directed internal transport ÎShuffle product components within the factory ÎSpecialization and optimization of factory equipment ÎEfficient handling of product voids ÎEfficient handling of errors (redundancy)
Remove solvent ÎLess drag; more throughput; higher efficiency ÎFeedstock can still be delivered in solution ÎLimits choice of actuators ÎMay require different synthetic reactions ÎThis may be a good thing! ÎBetter materials ÎMore extreme reactions
Mill-style synthesis ÎLess flexible but a lot faster than robotstyle synthesis ÎEasier to recover reaction energy (stiffer equipment, smoother motion) ÎLess computation per operation ÎLess beneficial for component assembly, but may still be useful
Conclusion ÎToday: We can build and join precise molecular parts. ÎWith this NIAC project, we have begun design of a system that can do precise, programmed, actuated assembly of molecular parts. ÎFrom there, incremental steps lead to desktop nanofactory.
Who we are ÎCenter for Responsible Nanotechnology: Research and education on advanced nanotech capabilities and implications ÎResearchers: ÎChris Phoenix, CRN Director of Research cphoenix@crnano.org ÎTihamer Toth-Fejel