Nanotube Transistors Assembled with DNA
JohnFaith writes " Here's an article from Technology Research News about tools being developed for designing DNA structures with carbon nanotubes. The DNA assembles into a matrix and the semiconducting carbon nanotubes form transistors in the spaces between the DNA. The work is happening at Duke , with more papers and info at Chris Dwyer's page. Interesting to see research that may provide arbitrarily complex 2D structures versus homogeneous films."
September 4th, 2004 at 8:41 AM
One Step Closer
Perhaps this is a major step towards bootstrapping MNT through biological means. It might eventually be possible to use DNA to help create a nanoscale diamondoid robot arm, thus bootstrapping MNT. This might be done by creating building blocks large enough to be manipulated with current methods, creating parts for the arm, or creating structures that could somehow be used indirectly to create the arm.
September 10th, 2004 at 9:00 AM
Re:One Step Closer
If you look carefully at Dwyer's work and the TROIKA project you will see that the emphasis is on self-assembly. This is very different from classical biological mechanical motion (ATP synthase, myosin, kinesin, dynein, etc.). The problem with self-assembly is that its capabilities may be limited. There are probably molecular structures which simply cannot "self-assemble" (several of the complex structures designed by Drexler and Merkle may fall into this category). A nanoscale diamondoid robot arm almost certainly cannot self-assemble. The size of the classical nanoassembler arm is ~30x100nm which makes it much smaller than any parts which can be assembled using current mechanical methods. (Does anyone know the actual limits of Zyvex mechanical or perhaps light-manipulated component assembly?)
It is possible to design and manufacture molecules (enzymes, motors, etc.) that can manipulate other molecules to build and assemble molecular "parts". This is discussed in Protein Based Assembly of Nanoscale Parts. What is required however is significantly more individuals (at least 100 times, better 1,000x-10,000x, more people than the 10(s) currently involved in these topics). They need to be working in the areas of reaction libraries, enzyme design, mechanoassembly (not to be confused with self-assembly or mechanosynthesis), etc.
Even then it isn't clear that one can get a diamondoid nanoassembler out of the effort since to the best of my knowledge there has been no demonstration of larger 3D molecules (MW > ~4000 daltons) with high covalent bond density per unit volume (the distinguishing feature of diamondoid) that have been manufactured using biological methods.1 If this cannot be done then one is in for a long multi-stage development path as outlined in Table 16.1 of Nanosystems.
If the multi-stage (biological and derivative) development path, the "wet path", is not followed then the most probable development path will be via some method based on AFMs, the "dry path". IMO, that will likely follow a development timeline not significantly different from the that which has been the case in the semiconductor industry following (or being driven by) Moore's Law. That path may well take 40 years (if history is any guide). That is because that path will have to follow the same precision increase, error reduction, volume production, etc. paths the semiconductor industry followed as it evolved. In contrast the wet path already has significant elements of high precision, error checking and volume production though these would need to be generalized.
The question may be — "For the breakthroughs that MNT promises — do you want to bet the farm on any single production methodology (e.g. self-assembly, the 'wet path' or the 'dry path')?"
I would suggest that if self-assembly is going to create a "real" nanoassembly mechanosynthesis arm it is going to be very indirectly and require a lot of additional support from other areas of nanoscale science.
Robert
1. Using biology you can get molecules of much greater MW, the DNA in chromosomes for example, but these are essentially 2D structures (polymers), not 3D structures. Even the larger unique biological molecules (such as palytoxin) are largely 2D structures. This may be a side effect of the probability that it is much more difficult to evolve an enzyme that can grasp a complex 3-D structure than some subset of a 2-D structure to perform additional chemical reaction steps.