|http://video.google. com/ videoplay?docid=-2022170440316254003. The pump sorts acetylene molecules from raw feedstock molecules before they enter the main machinery of the nanofactory. " width="300" height="302">|
www.nanoengineer-1.com. " width="300" height="335">
Mark Sims, CEO
Bloomfield Hills, Mich.
The future of technology is, in some ways, easy to predict. Materials will become stronger, medicine will cure more diseases, and computers will become faster, more powerful, and smaller yet. At some point, however, it will no longer be feasible to drive technological advance by miniaturizing macroscale products and processes. After all, they can only get so small. Advanced nanotechnology will then come into play. Instead of manufacturing equipment altering material from the top down, nanoscale "machines" will build products from the bottom up.
CAD molecular-modeling engineering software is an important tool in making this happen. Chemists and biologists have long used molecular modeling programs to test ideas before they enter the lab. New software that combines molecular modeling with a familiar CAD interface now lets mechanical engineers begin designing the products of tomorrow, today. Necessary understanding will come, in part, through the computational models such software provides, which describe nanomachines at the molecular level, in precise atomic detail.
Before discussing how the software works, it may be helpful to explain some underlying principles of chemistry, biology, and physics and how they relate to molecular manufacturing.
Conventional manufacturing methods arrange atoms quite crudely. For example, even the finest commercial microchips are grossly irregular at the atomic scale. Chemists, in contrast, can create molecules that are precisely defined by particular arrangements of atoms, always with the same numbers, kinds, and linkages. Chemistry shows how structures form when reactive molecules meet. Similar structures can be built at larger scales using molecular-machine systems to guide reactions.
Living cells, for example, contain such systems, called ribosomes. They use digital data, that is, genes in DNA, to guide the assembly of molecular objects or proteins that serve as parts of molecular machines. Thus, DNA stores the genetic information necessary for the rest of the machinery to, for instance, build copies of cells, and eventually, macroscale structures such as trees and human beings.
Advanced manufacturing techniques will likewise use stored data to guide construction work done by molecular machines. These will put reactive molecules together using machines to position them, building macroscale products with atomic precision. Such products will be stronger, tougher, and more capable than the delicate structures found in living cells. Applied physics aided by computer modeling shows products could range from a host of useful gadgets to computers, motors, and even entire factories.
An example of CAD molecular modeling software is the NanoEngineer-1 package from Nanorex Inc., Bloomfield Hills, Mich. The software combines features of high-end solid-modeling packages with the chemistry, biology, and physics capabilities of advanced molecular-modeling programs.
From a modeling perspective, atoms can be thought of as building blocks, like Legos. Each atom type has a unique size, shape, and connectivity geometry. NanoEngineer-1 lets users sketch molecular fragments by depositing and connecting atoms together to build-up entire parts. Additionally, its solid-modeling capabilities let users model large molecular structures quickly. The software also allows displaying large components comprising many thousands or millions of atoms at higher levels of abstraction, such as smooth shapes or coarsegrained faceted objects, for quick rendering of structures.
As with typical parts designed in 3D CAD programs, engineers design molecular parts using features that provide higher-level operations. Thus, users sketch a molecular fragment representing a repeating unit of a part that has the required profile and size, and then apply features to build a larger part. An Extrude feature, for instance, lets users take a molecular fragment and add copies of it, extending the length of the structure. This makes it easy to create common parts of molecular machines such as casings, rods, shafts, and cylinders. A Revolve feature works by adding copies of a sketched fragment positioned around an axis. This lets users rapidly design hollow cylinders and ring structures. Other useful features for nanomechanical modeling include Pattern, Mirror, Shell, and Combine Bodies.
The software also includes structure generators. These make it easy to quickly model nanostructures such as DNA, dendrimers, graphene sheets, and nanotubes. Users can, for instance, specify length and helical twist and the software generates a carbon nanotube containing thousands of atoms. Users can edit dimensions interactively when the nanotube's diameter or length requires modification.
MOLECULAR MODELING DIFFERENCES
Although modeling in Nano-Engineer-1 is similar to traditional 3D CAD, there are some differences worthy of note. The software models molecules as a series of atoms connected by chemical bonds. Atoms must be periodically adjusted to make the geometry of a part more accurate. This is accomplished by invoking the Minimize command on the structure, which moves the atoms to positions where calculated forces on them balance. Called "energy minimization" or "geometry optimization," this method is common to all conventional molecular modeling packages.
Another difference: In molecular modeling, atomic positions matter in combining pieces to make a larger part. Atomic positions on facing surfaces must be properly aligned or it can be difficult to combine pieces. For example, when embedding a molecular ring in the wall of a molecular casing, it would be difficult to combine the bodies with their different atomic patterns unless enough atoms on the outer surface of the ring have positions that match atoms in the casing wall. Thus, ensuring proper arrangements of surface atoms is an essential design aspect of molecular modeling.
Also, in conventional molecular modeling packages, users must have a firm grasp of chemistry. There are many ways to link atoms into larger structures that are quite stable, producing structures as strong as silicon carbide. But it's all too easy to construct a model that contains unstable groups of atoms that can split apart, rearrange, or combine with other atoms. Designers may not even recognize unstable groups as problematic for a practical design. Thus, NanoEngineer-1 includes a structure checker. Like a spelling checker, which highlights misspelled words by underlining them in red, the structure checker flags unstable groups of atoms so they can be revised.
SIMULATING MOLECULAR MACHINES
Complete 3D designs created by the software supply virtual models for simulation and analysis. The software lets users simulate the motion of nanoscale parts using what molecular scientists call "molecular mechanics models." To calculate the location of each atom in space would normally require going to the quantum level. But for modeling molecular machines, this is overly exact and demands excessive-computational time.
Molecular mechanics models, embodied in a simulation engine, describe the forces between atoms using easy-to-compute formulas. These treat bonds like springs, and provide a good approximation of the predictions of quantum mechanics. The Minimize command uses these forces to find the equilibrium atom positions. The dynamics simulator computes how the atoms will move using these forces and the familiar equation from classical physics, F = ma.
The program also provides "motor jigs" for analysis of assemblies with moving parts. A motor jig lets the molecular dynamics simulator know which parts are driving the assembly and the parts remaining fixed. Users attach the jig to carefully selected atoms simulating the prime mover, such as a rotary motor for a pinion gear, and type in parameters that determine the motor's direction, torque, and speed. Users can also fix atoms in space with "anchor jigs" that keep components, such as a casing, held in place during a simulation run.
Once the simulation is set up, it is launched and runs until it is complete or interrupted. Users can play back results as an interactive 3D movie viewable from any angle and zoom factor. The simulation also writes important data including the torque and speed of the rotary motor to a file. The data can be plotted and analyzed to derive useful information for validating a design or indicating areas for improvement and revision.
Molecular machines currently exist only in computer modeling and theory. But mechanosynthesis, which guides specific chemical reactions using machines to position them, has already been demonstrated in the lab with the deposition of single silicon atoms on a surface and hydrogen atom abstraction, two fundamental steps for advanced nanotechnology. Progress in nanotechnology will be similar to that of aircraft. Fighter planes and 747s weren't possible until the Wright brothers built a flying machine with a twocycle engine, a propeller, and wings made of canvas and wood. Advanced nanotechnology will likewise go through several generations of development.
Thus, the first nanosystems will not look anything like some of the advanced models the software can provide. But such models help guide research by providing reliable predictions and reducing experimental failures by indicating unworkable ideas. Engineers can explore designs and examine potential targets for product development well in advance of physical feasibility. Our prediction is that advanced nanosystems will exist in about 20 years.
The software does not currently output data that could drive such machines as, for instance, conventional CAD/CAM software outputs data to drive CNCs. However, we are developing a feature of the program that will do this for specific types of structures. For example, we are working with a research group at the California Institute of Technology in Pasadena, Calif. that has come up with a technique for synthesizing arbitrary 2D shapes out of DNA. A Fall release of NanoEngineer-1 will let engineers design and model custom DNA shapes that can actually be synthesized. NanoEngineer-1 will create a special file format that users can send to a company to order the individual strands of DNA needed to create the final structure. Once received, the materials are then combined for a reaction that creates a custom DNA structure.
The author thanks K. Eric Drexler for his contribution to the article.
Engines of creation
K. Eric Drexler introduced the term " nanotechnology" to a broad audience in his 1986 book, Engines of Creation. He used the term to refer to an idea described by physicist Richard Feynman in a 1959 talk, "There's Plenty of Room at the Bottom." Feynman theorized the development of "productive nanosystems," or nanoscale machinery for building atomically precise products under digital control. He drew his inspiration from biology, generalizing nanomachinery that exists within living systems and applying it to a broad set of productive capabilities. Drexler explains nanosystems in technical detail in his 1992 book, Nanosystems: Molecular Machinery, Manufacturing and Computation.
Carbon nanotubes illustrate the promise of nanotechnology and how far nanoscience still needs to progress. Since their discovery in 1991, nanotubes have been the focus of a great deal of research and speculation. They have been proposed as a miracle material for everything from interconnects for ICs to a cable for a space elevator. The reality today, though, is scientists are still working out how to engineer nanotubes for applications more precise than tennis rackets, and how to manufacture them in uniform, scalable quantities. However, scientists, entrepreneurs, and investors understand that real applications and products will follow the expected technological leaps. Thus, those with foresight are already preparing a business, financial, and manufacturing infrastructure for carbon nanotube applications which, so far, exist only in theory and computer models.
Nanorex Inc., nanorex.com