• Google
    This Blog Web

October 2011

Sun Mon Tue Wed Thu Fri Sat
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          

RSS Feed

Bookmark and Share

Email Feed

  • Powered by FeedBlitz

« Molecular Manufacturing Video | Main | Assessing Levels of Risk »

September 22, 2006


Feed You can follow this conversation by subscribing to the comment feed for this post.

Richard Jones

A week or two ago you wrote, in your piece about why people had been slow to accept the idea of MM, that "MM people sometimes cited non-MM-focused research as evidence of MM's feasibility, and the researchers often objected to that, especially once MM became controversial." You were quite right, and you've just given us another example.

Protein folding is a self-assembly process, the archetype of the "soft" approach to nanotechnology. Protein machines only work because they are soft and because their non-bonding interactions are weak. They positively require a warm, wet environment with lots of Brownian motion - if you cool a protein machine down below the glass temperature at which it goes rigid it stops working. As such protein machines are fundamentally incompatible with the ideas of diamondoid molecular manufacturing, which is predicated on rigidity, absence of corrosive solvents like water, and design to minimise the effects of Brownian motion.

You also wrote "it will be interesting to see under what circumstances the paradigm shift is translated into action." Perhaps there is a paradigm shift going on, but if there is, its direction is away from the hard, mechanical engineering inspired vision of MM, and towards the "soft" paradigm inspired by biology, which exploits Brownian motion, strong surface forces, and lack of rigidity. Don't quote the successes of the new paradigm as support for the old one!


This discovery was posted on Brian Wang's blog weeks ago. Do you know at what stage this discovery is in? Has it been tested in simulations reliably?

As for Richard's comments:

I think that there is a shifting towards both organic and inorganic MM. Yes we are seeing huge changes in biotechnology design, understanding, and innovation. In many ways, biotech's advances are more obvious and exciting.

Mostly this is the case because biotech is still a couple of steps ahead of nanotech. Nature over the last few billion years has given us the molecular tools and the programming language to make our own atomically precise molecular structures.

In nanotech's case, the path to MM isn't quite as obvious to the naked eye. We are just barely playing with putting atoms together using slow and cumbersome tools like the Atomic Force Microscope. The trend towards MM is visible however. More companies are coming out with big-plan ideas of MM or creating incredibly complex self-assembling structures. Zyvex is my favorite example.

I think that the next decade will really start to convince people about the feasibility of inorganic MM. Of course, some people won't be convinced until it becomes a reality.

Keep up the great posts Chris.


I agree that progress toward the goal to MM is steady, even if those who are responsible for that progress don't agree. Perhaps whether or not something is a "show starter" for MM can only be determined after the fact. Only in retrospect do we see that the invention of the vacuum tube is the essential first step in the creation of the modern computer revolution. If one described in great detail life in the early 21st century to the inventor of the vacuum tube, no doubt he would be highly skeptical.

As for the wet-to-dry path towards MM, I have no idea how it would be achieved. I also don't really have a firm grasp of how our ancestors transitioned from stone knives and bear skins to metal tools and modern textiles. In fact, I think if you gave 100 material scientists the task using stone-age tools to build modern ones they would be hard pressed to figure out how. 100 anthropologist would fare much better I imagine.

Also, Richard, if I am not mistaken you once talked about how one could use wet nanotechnology to make things that could function out of a wet environment. (I may be mistaken about that) Of course, even if you did this it's still not clear if you could make a sufficiently rich set of parts to construct a new manufacturing device. It is also not clear how you would, in the absence of Brownian motion, assemble them.

I think the argument can be made that scientist, being human, are driven by their own desire to achieve a little bit of immortality. This can be hard to do if you devote yourself to something that is fundamentally speculative and theoretical, like molecular manufacturing. Nature provides a rich set of tools to make a name for yourself relatively quickly compared to MM. Ideally, scientist should behave in a completely logical manner; but they don't. It does not make sense for everyone to pile all their resources on one way of doing things, unfortunately that is what is happening. If we had a king of science dictating what people should be working on, he could tell more people to work on MM even if it seems like a longshot. This should be done, if for no other reason, than to avoid putting all your eggs in one basket. This would require a zen-like detachment from fame and addulation that few to none have. The plus side of this laser-like focus on biotechnology, is that after a while all the low hanging fruit will be picked, and the next generation of researchers will have to think outside of that paradigm to make names for themselves.

Finally, I must say that I don't know if MM will work. I lean toward saying yes; because I think that man-made solutions do, and will continue to have, an appearence that is very different to natural ones. A bird and airplane look completely differnt. I think Drexler's work will most likely be view the same way by future generations as is the work of Leonardo DaVinci. If you took Leonardo's sketches (And this has been done, by the way) to an engineer, they will be amazed how this man knew what would be possible centuries from now; but they will wince at the thought of actually building any of these designs, because as beautiful as they are, none of them actually work without considerable modification

Chris Phoenix, CRN

My current personal favorite for wet-to-dry path involves using positionally-controlled silicatein to build 3D silica shapes, which could include nanoscale machines.

Leonardo was only working conceptually; he did not apply any numbers, or he would have known that you can't build parachutes out of wooden beams. Leonardo was willing to think about what ought to be possible; then he sketched the closest thing to it in the materials he knew, without considering whether they would be sufficient. His drawings look like engineering drawings, but they are really artists' conceptions.

Drexler uses numbers whenever possible; in fact, there are not many places where he does not have numbers. I think that Drexler's work is better compared with Babbage's than with Da Vinci's. Babbage's Analytical Engine was superceded (after a century and a half), but it would have worked pretty much as designed.

I have little doubt that some of the atoms in Drexler's machine designs are misplaced; but I also have little doubt that something very similar would work as intended.

There will always be objections to Drexler's designs--because there will always be people who have a psychological need to think they know better. Similarly, a notable computer scientist recently objected that Babbage's design wouldn't have worked because entropy would build up--a complete and blatant misunderstanding of the design!

After decades of thought and criticism, anyone who attempts to disprove Drexler's approach with a quick argument is essentially guaranteed to be wrong.


Brian Wang


You have a different goal and paradigm - soft machines.

Others have the goal of molecularly precise manufacturing using a wide a range of atoms and molecules as possible. Diamond is a big part but not the only part of it.

Others have been writing about creating artificial ribosomes for over 20 years. So your attempt to mark your territory is too late.

Proteins and DNA are being extended and interacted with polymers, metals and carbon nanotubes and buckyballs. There is the possibility that there are interesting bootstrapping pathways that will come with greater control of proteins and DNA. There could also be interesting capabilities that could come by utlizing different materials. Also, one could mix techniques of self-assembly with laser and magnetic manipulation. If we have a toolbox then why do we have to only use the self-assembly hammer ?

Synthetic biology is looking at applying engineering principles to biology.


Brian I agree, I don't think it is helpful to look at this problem as only having two types of solutions; the first being the use of macro-scale engineering principles at the nano-scale, the other being using biology as a template. I think we will find many interesting things that don't fall into either category. Laser tweezers being a an excellent example, I don't think biology has ever used anything like that, and I know that you can't easily use lasers to manipulate macro-scale components.


Chris, we are already seeing tweaking to Drexler's designs. Merkle and Freitas's tooltip has deviated from Drexler's by reducing the number of atoms on all sides leaving a Phillips-head screwdriver appearance compared to the classic pyramid design of Drexler. This was done, to the best of my knowledge, to help reduce unwanted reactions. Also, I believe Drexler's carbon deposition did not stagger the reaction sites like was done in the Peng/Freitas et al paper. If the designs already, even in computer simulations, are being found lacking in some way, and are being changed already; what guarantee do we have that the bulk of "Nanosystems" will remain unchanged before the first nanofactory is built? There is no doubt that there is much more technical rigor in Drexler's work than in Leonardo's sketches, but that does not necessarily mean that it won't be changed to the same extent. I agree that the basic gist of it all will prove to be right.

I do not know what sort of claims Babbage and his supporters were making regarding the capabilities of his difference engine. I do know that some pretty spectacular claims have been made with respect to MNT. If you made claims that Babbage's machines could do anything you use your modern computer for, you would be wrong. This is because only a modern, integrated-circuit computer makes those applications practical. Some of the claims of MNT may also require the same amount of change to the underlying techology to make them a reality.

I also think researchers need to keep their eyes open to ideas that fall into the "other" category, not classically mechanical or biological. The final result may surprise us all.

jim moore

From the article:

"Rather than calculating the motions of a protein molecule step by step, as most simulations do, a team of Italian and French physicists studied the evolution of a molecule using variational principles. The technique allowed the physicists to evaluate all the possible paths that the molecule's parts would follow and then pick out the most likely one"

I am thinking that this technique should allow the development of some design rules for making 3-D nano-scale parts from a one dimensional polymer. If this technique is that general then chemists should be able to do what is now the toughest part of "bottom - up" manufacturing: making the smallest parts.

There may be a huge opportunity to use micro-fluidic systems for the step wise addition of monomers to growing nano-scale objects attached to particular points on a surface. (I am thinking of some process descended from the Merrifield method) You should be able to synthesize very precise objects that are hundreds to thousands of cubic nanometers in size. (In other words objects that would fit within a 20 nm per side cube.)

The mechanical fabrication step becomes taking that 3-D nano-part from a known location and transporting and attaching it to a precise location in the much larger object you are fabricating.
You have to be able to make the mechanical fabrication system out of the nano-scale objects that you have polymerized, for the system to scale up.

If this new insight really is generalizable to folding of all one dimensional polymers than it seems like a pretty sure bet sometime between 2010 and 2020 this will happen:
Some sort of 3-D, computer controlled, rapid prototyping like device with positional and compositional control down to the nanometer and capable of exponential manufacturing will be developed.

The comments to this entry are closed.