Responsible Nanotechnology

Frequency-Specific Optical Nano-Actuators?

In order to control nanomachines via external signals, it will be necessary to use multiple actuators that can be activated independently.

Molecular structure is largely determined by where the electrons are. If the electrons can be made to move, the molecule may change shape. For some molecules, hitting them with light can make the electrons move. Thus, light-actuated molecular actuators have already been developed.

It would be nice if, instead of just "light," there was a whole palette of actuators, each activated by a different color. It would be nice if, instead of having to develop each one individually, a general approach could create a whole family of these actuators.

Nanotech solar cell research has explored ways that light energy can be transferred from one molecule to another - connect one molecule that absorbs the photon to another that creates electricity or hydrogen. So what if, instead of building an actuator molecule that both absorbs photons and moves, build one that doesn't absorb any photons - but can accept photon energy from a nanoparticle it's connected to?

It seems like this scheme would allow a variety of actuators to be built, sensitive to different frequencies of light, simply by swapping out the nanoparticles. Frequency-specific nanoparticles already exist, and the frequency can be tuned simply by adjusting the size of the nanoparticle when it's fabricated.

I'm not up to date on the relevant research, so this idea may already have been proposed elsewhere. And, of course, it may not work. But it seems plausible... and if it did work, it would provide a means of broadcast control of nanomachines, without needing physical macro/nano contact, and with multiple actuators individually controllable.

Chris Phoenix

Nanoscale Actuators Needed

Biology uses lots of techniques to make and transport and join the molecules it needs to build cells. There's self-assembly and wet chemistry, sure. But there's also mechanical transport along stiff tracks. There's mechanical linkages transmitting signals from one part of the cell to another. There's trusswork structures. Molecules under construction go through a series of machines, and not all of the transport happens by diffusion.

As our technology develops toward molecular manufacturing, which techniques will we bring online, in which order?

Today, the closest thing we've done to molecular manufacturing may be Ned Seeman's DNA-building machine, though there may be more advanced experiments that I haven't heard about yet. That machine is built of DNA, which is somewhat floppy and weak even when compared to protein. For actuation, it uses diffusion and self-assembly of large molecules, which is slow. Also, it has no mechanism for transporting the reactants to the machine - again, simple diffusion does the job.

More recently, Seeman put his machine into a framework of DNA assembled with Rothemund staples. This is a start toward the kind of spatial arrangement we find in cells, which can make metabolic processes faster and more efficient.

As I think about development pathways toward molecular manufacturing, it seems to me that faster actuation is a key capability. It's pretty annoying, and may seem pointless, to design a molecular manufacturing nanodevice that will take a month to test, and would take a year or more to actually build its own mass in product. But fast actuation could make that same device a lot more interesting.

Chris Phoenix

Early Molecular Manufacturing Products

During the past few years, several advances have been made in nanomachines and nanostructures. For example, Seeman's machine, built of DNA and programmed by DNA, which templates the construction of DNA strands, has now been mounted in a framework made of DNA by the Rothemund technique. Other systems of nanomachines, made with other types of molecules, have been proposed for guiding the construction of yet other types of molecules.

Someday, perhaps sooner than most people expect, nanotechnology will reach the point where nanomachines can build other nanomachines of equivalent complexity. At that point, nanotech will be able to make its own tools, and advanced manufacturing systems will soon become non-scarce.

Before that happens, nanomachines will be able to make products that are less complex than themselves, but still interesting from a research or even a commercial point of view.

Some of the nanomachine systems that are being developed are likely to be quite flexible in what they can assemble. For example, almost anything can be attached to a DNA strand. This raises the question: What would be interesting to assemble, in very small quantity, and with only moderate complexity? Suppose you could make a one-dimensional string or even a two-dimensional raft of molecules and nanoparticles, in programmed sequence and arrangement - what would you want to do with it?

Chris Phoenix

Synthetic Biology Pathway to Molecular Manufacturing?

I must begin by admitting that my thinking has been limited. When I think of biology, I think of complex poorly-understood feedback loops; designs that have been cobbled together by random chance over billions of years; systems that can be played with, but not really engineered.

When I think about human adaptation of biology, I think of slow improvement via genetic engineering or even breeding. And when I think of a cross between biology and nanotechnology, I think either of nanomedicine, or of the old tired arguments about how nothing can improve on biology so the best we can do is try to imitate nature's designs.

Now let's consider molecular manufacturing. A description of molecular manufacturing might go something like this:

You design machines at the molecular scale - machines that, working collectively, can build more of the same by atomically precise mechanically guided chemistry. You build a set of these machines, by hook or by crook, and fasten them together. They build duplicate machines. Exponential growth allows a tiny construction module, duplicated many times, to produce useful amounts of product. Then, you program the system to build a wide range of products.

There is a branch of synthetic biology that is working toward this same goal. The idea is to design an entire genome, to produce a minimal organism, that can duplicate itself, and eventually can be used for human benefit and profit.

This goal will take many years to achieve; the above-linked paper, which looks a lot like a roadmap, includes an impressive list of unknowns and problems to be solved. And the result will still face many of the limitations of biology: fluid drag from water, chemical transport by diffusion, limited material properties. And many of the components may not be completely understood even after it is built, since they will have been adapted from natural organisms.

But the creation of an engineered, constructed system that builds copies of itself from small molecules would, in my book, count as molecular manufacturing. It would not simply be duplicating existing life forms by means of synthesized chemicals (a project that is also underway). In fact, the roadmap discusses the possibility that the system might work better without membranes - and one of the reasons listed is the possibility of "spacial arraying for nanofabrication."

I don't know whether the first nanofactory will be produced by some other pathway before this synthetic biology project comes to completion. But I think this has to be considered a possible pathway to molecular manufacturing. There are now several pathways, each engaging lab researchers. One way or another, molecular manufacturing is coming.

(Thanks to Herman Salgado for asking the question that led to this post.)

Chris Phoenix

DNA Meets Buckytubes

I've talked recently about some of the cool stuff you can do with DNA. Here's one more: DNA can be used to sort single-walled carbon nanotubes.

Carbon nanotubes come in a variety of different styles. Imagine cutting a narrow strip of plaid cloth and wrapping it around a pencil. Depending on the angle of the wrap, the stripes would line up - or not - at various angles. A carbon nanotube can be thought of as a thin strip of graphite-like carbon, in which the atoms are lined up in regular rows, coiled up to form a tube.

Depending on the angle of the carbon rows - the so-called chirality of the tube - it can be a metal, insulator, or semiconductor. This makes it very interesting for computer applications. The trouble is that it's hard to synthesize just one kind of tube, and until now, it's been hard to separate out one kind from a mixture.

As reported in Nature, a group of researchers has searched among a massive number of DNA sequences to find sequences that will fold just right to wrap around tubes of only one particular chirality. This allows the wrapped tubes to be separated from the mixture. The DNA folds up into zig-zag sheets, which then wrap around the tubes that have just the right chirality.

This isn't the first time DNA has been attached to nanotubes, but it's cool that it is attaching to a single kind. And it sounds like the zone of attachment is large enough to form a reasonably strong connection. Although the researchers are mainly interested in electronic applications (at least, judging from the abstract of their paper), it may be that interesting nanomechanical applications will also be enabled by this work.

(Hat tip to Eric Drexler's Metamodern blog.)

Chris Phoenix

Robust Manufacturing Technology

Paul S commented on my last post: "...if regulation is necessary it will have to be applied before any robust [molecular manufacturing] technology is released to the public."

I agree with this - depending on just what is meant by "robust."

What makes a technology robust? Or, what would make a molecular manufacturing technology difficult to un-release to the public?

Consider first that molecular manufacturing development, along with other technologies, will not be standing still. A ten-year-old nanofactory would be more out of date than a ten-year-old computer. So if a branch of technology is left to stagnate for ten years, then it will effectively be un-released. It will still be able to do what it used to, but relative to the powers of modern technology, its products would be relatively unimportant. And if regulation is needed, ten years of progress in military and law enforcement technologies (including data mining) might make it difficult to use ten-year-old technologies without being spotted.

On the other hand, this analysis assumes that the governmental version of molecular manufacturing would continue developing rapidly. This is not necessarily a safe assumption. Without lots of minds working on it with the stimulus of economic competition, a technology could easily stagnate. So ten-year-old civilian tech vs. ten years of governmental progress might be a pretty even match. Just look at the U. S. space program.

The second question is whether a particular incarnation of technology is usable and useful. For commercial applications, this means it has to be reliable, predictable, and not too innovative. As a home appliance, it has to be thoroughly simple to use, safe and perceived as safe, and make finished products. For hobbyists, it has to be fairly inexpensive to work on. It's far from obvious that an embryonic molecular manufacturing technology would be adopted by any of these groups in a way that led to its rapid further development.

So if early molecular manufacturing was released to the public, it's not at all certain that this would lead to Pandora's box being opened a decade in the future. This is good news, because in a sense, early molecular manufacturing has already been released.

Chris Phoenix

Draconian Measures For Molecular Manufacturing?

A few days ago, I wrote a post implying that liberty in the U.S. may be at risk due to an ongoing state of near-war. I quoted Aldous Huxley: "Permanent crisis justifies permanent control of everybody and everything by the agencies of the central government."

A commenter named Baughn asked: "I wonder, however. Considering the rather draconian measures you believe would be required to control nanotechnology, do you think this is a bad thing?"

First, let me clarify (for any new readers) that "nanotechnology" here is used to mean molecular manufacturing - its original meaning - not all the newer stuff that has been grafted onto the word, such as nanoparticles. No one is suggesting that nanoparticles might need draconian control measures - though some kinds of nanoparticles might need a bit more control than they're currently getting.

So, molecular manufacturing: tiny nanotech machines, made out of precisely designed molecules, that can rapidly build more machines of equivalent precision and complexity. A manufacturing revolution: general-purpose manufacturing, using non-scarce equipment, of inexpensive and highly advanced products. And the manufacturing systems could be small, easily concealed, easily duplicated - very difficult to control, if an unrestricted system was ever in civilian hands.

Pretty revolutionary - which means disruptive - which means potentially destructive. So, does it require draconian control measures?

There is some argument that it should simply be allowed to be developed with minimal controls, in the expectation that the good will outweigh the bad, and problems will be outweighed by solutions. In my more optimistic moments, I have a lot of sympathy for this viewpoint. Computers have developed pretty much that way, and we - and our infrastructure and society - have so far managed to survive computer viruses, spam, and data-mining. On the other hand, if a computer virus could kill a person instead of just erasing their data, we might be a lot less sanguine.

If molecular manufacturing has to be controlled, how much of society needs to be controlled to accomplish that? The good news is that not much broad-based control may be required. In other words, it may be sufficient to keep control of a few key technological capabilities, to make it difficult or impossible for a private effort to develop molecular manufacturing until technology has advanced to the point that molecular manufacturing is no longer a big deal.

There may, of course, be paths to molecular manufacturing that, once conceptualized, turn out to be fairly simple recipes, accessible with technologies that are already widespread. That would be problematic. But without sophisticated tools and lots of R&D money, such recipes couldn't be developed and tested.

I'm not yet ready to say that broad-based control of society to avoid technological evils is definitely unnecessary and will always continue to be unnecessary. But I do think (at this moment, at least) that in terms of potential harm to humans from private development of high tech, there are other technologies that loom a lot larger. And I think this will continue to be the case until the first advanced molecular manufacturing system is not only developed, but released to the public in unrestricted form.

If it turns out to be necessary to restrict molecular manufacturing, then either limits on its development or technological limits on its implementation may well be sufficient. I don't see the need to restructure or oppress society to keep us safe from this technology.

My ideal future would have most of the limits be technological, and applied to limit the use of publicly available manufacturing systems. Technological limits would have to be carefully designed, because almost anything can be cracked given enough effort. But molecular manufacturing has enough potential for good that I'd like to see it available in some form that's appropriately restricted but still broadly useful.

Chris Phoenix

Liberty vs. War...

Today is Independence Day in the U. S., commemorating our Declaration of Independence from oppressive government (the British Empire).

A few weeks ago I ran across a quote about liberty, in Brave New World Revisited by Aldous Huxley, a noted futurist and science fiction writer. In 1958, he wrote:

"But liberty, as we all know, cannot flourish in a country that is permanently on a war footing, or even a near-war footing. Permanent crisis justifies permanent control of everybody and everything by the agencies of the central government."

As I write this, the current Homeland Security Threat Level is "Yellow - Elevated" with the airline sector at Orange. Any guesses when we will cease to be on a near-war footing?

Chris Phoenix

Graphene Ribbons Now Available

Precise graphene ribbons can be made by chemically unzipping carbon nanotubes.

This adds one more large molecule to the nanoscale construction toolbox. The ribbons may be useful in electronics, since a film of them should be even more conductive than a film of buckytubes. So they will likely be researched.

The article didn't say so, but I also speculate that these ribbons may be useful as reinforcements in composite materials (like a better kind of carbon fiber), since they seem to be water-soluble (which helps with processing) and might be easier to attach other molecules to than nanotubes are. The researchers estimate that they could be available in ton amounts in a couple of years, if there's demand for them.

In any case, I'm sure they'll be experimented on in many ways. They may turn out to be useful in mechanical nanosystems, though they'd probably be pretty floppy in comparison with buckytubes.

The unzipping process was discovered by Dmitry Kosynkin, a post-doctoral research associate at Rice University, who was studying oxidation of nanotubes. Yes, a lot of nanotechnology is accessible through chemistry.

(Hat tip to Mike Treder.)

Chris Phoenix