Responsible Nanotechnology

Solid-State Quantum Computer

It's not nanotech, except in the sense that anything small and interesting enough counts as nanotech. But it's a significant milestone toward a game-changing technology.

An electronic quantum processor with two qubits, each made of about a billion aluminum atoms, has been created at Yale university. They can get the qubits to maintain their state for about a microsecond. And they have a "quantum bus" that can pass information between the bits.

I don't understand all the cool things you can do with a quantum computer, but apparently it can be useful for simulating quantum chemistry. I suspect we'll have covalent-solid molecular manufacturing before this quantum computer technology develops to the point where it can help, but I've been surprised before by how fast technologies can advance.

Chris Phoenix

Today's Tech: Affordable Carbon Capture

Molecular manufacturing will give us the ability to do planet-scale engineering. We may need it.

Polar ice is melting faster than expected, and permafrost is releasing methane. Science tells us that large-scale reduction in atmospheric carbon dioxide may be necessary to avoid major, destructive, near-irreversible climate shifts.

According to a recent CNN news story on carbon capture technology, it may be possible to remove carbon dioxide economically with today's technology. A machine that costs as much as a car could remove the carbon emitted by 20 cars. The power required to remove and liquefy CO2 is 1/5 the power generated by burning coal to produce it.

Of course, removing CO2 is only part of the puzzle. There remains the question of what to do with it. There are several possible answers to that, and it remains to be seen which one is best.

I'm enough of a pessimist to think that this technology will not be adopted until it's too late for it to make much difference. Thus, instead of (for example) a 5% tax on cars to offset their carbon emissions, we'll need a massive crash program. Today's technology will just be a reminder that we don't use technology wisely.

When molecular manufacturing was proposed in the 1980's, there were many applications that could not be achieved by any other technology. As time goes on, some of those applications are being addressed by advances in other fields of technology... at least potentially. Understanding the uses, misuses, and non-uses of today's technologies gives us some clues about how molecular manufacturing may be used and abused.

Chris Phoenix

Progress Acknowledged Toward "Nanobots"

A research team at Singapore's Institute of Materials Research and Engineering has made a 1.2-nm gear that can be directly and precisely controlled. (Edit: ScienceDaily has an article with the same text plus a cool picture of the gear.)

This is cool. But what really impressed me is the quote from the Executive Director of IMRE, Dr. Lim Khiang Wee: "Christian and his team's discovery shows that it may one day be possible to create and manipulate molecular-level machines. Such machines may, for example, walk on DNA tracks in the future to deliver therapeutics to heal and cure. There already exists at least one international roadmap for creating such productive nanosystems. As we push the frontiers of nanotechnology, we increase our understanding of new phenomena at the nanoscale. This paper is a valuable step on the long road to applying this understanding for discoveries and breakthroughs in nanotechnology and bring to reality the tiny nanobots and nanomachines from science fiction movies."

Let's look at that quote closely. Dr .Wee references Drexler's roadmap for productive nanosystems. He uses Drexler's term: "productive nanosystems," meaning nanomachines that not only move, but build stuff. Drexler has invented several terms that didn't fly, such as "zettatechnology," and others that were mutated beyond recognition, such as "nanotechnology," so it's nice to see this one adopted.

He has included some qualifiers: "may one day be possible," "a valuable step on the long road."

He cites DNA as a possible building block for nanomachines. Recent breakthroughs have made me hopeful that DNA-based productive nanosystems are getting closer fast - so fast that I hesitate to estimate when the first one will be built.

Finally, and perhaps most importantly, he's not afraid to acknowledge that research into tiny controllable gears is leading toward nanoscale machines that are considered science-fictional today. 

Hat tip to Paul Suliin and Sander Olson.

Chris Phoenix

Industrial Nano Liability?

Although CRN focuses mainly on molecular manufacturing, since it is the most perilous form of nanotechnology in the long run, this story on nanoscale technologies deserves a mention.

The Investor Environmental Health Network has issued a report on eight loopholes that companies can use to avoid disclosing dangerous-chemical risk to shareholders - in some cases, right up to the point that the company goes bankrupt. It makes the interesting point that, not only do current regulations encourage companies not to disclose risk, but a company that chooses not to disclose risk is better off not even studying risk - if they don't know, then it's harder to fault them for not telling.

I found this report when someone (thanks, Paul!) sent me a link to a story about it. The story is interesting in its own right. It opens with the claim that the report "warns that billions of dollars in potential asbestos-like litigation risks for nanotechnology companies and investors are now hidden due to weak regulations governing disclosures of liabilities."

I read the report, and then searched for the word "billion" to make sure:  I did not find any dollar value, or any survey that would support a dollar value, for risks tied to nanotechnology. The report certainly shows how such risks might be hidden. And it shows how such risks were hidden for asbestos.
And nanoparticles of various types are discussed as potential health hazards. But the report does not give any risk numbers for nanotech - not in the summary, not in the body, and not in the appendix.

My takeaway: It's quite possible - one might even say likely - that health risks related to nanoparticles are being under-studied and under-reported. The report argues that such under-reporting is encouraged by the current regulatory structure - I'm not a lawyer, but the argument (with comparison to history) seems fairly strong. But at least one story about the report went too far, extrapolating claims that the report simply didn't make. When reading about nanotech, you can't always trust summaries and first impressions.

Chris Phoenix

Planar Assembly Scale

Jim Moore asked how small it would be useful to go with planar assembly.

A bit of background: When Eric Drexler proposed, back in 1992, that molecular manufacturing should be done in tabletop factories rather than vats with floating robots, his proposal had nanoscale sub-components being made and then put together in larger and larger assemblies by converging assembly lines. This design made a lot of sense, and variations of it were studied over the next decade.

Then, just a few years ago, John Burch and Eric Drexler came out with a new nanofactory architecture: one that made medium-small components, and then stuck them directly onto the surface of a product under construction, building it up block by block. This works because, as long as the machinery handling the block can scale down in proportion to the block (and scale up accordingly in operation speed), the linear deposition rate does not change.

So how small can planar assembly go? Well, the machinery to handle feedstock molecules will probably be quite a bit bigger than the molecules. So planar assembly probably doesn't work all the way down.

But it can go pretty far. The smallest number Jim asked about is 100 nm - the size of a smallish bacterium. Well, you can fit a lot of machinery into a block of that size. In my Primitive Nanofactory paper, I suggested 200 nm blocks - only eight times the volume.

It seems to me that the lower bound on size is set by the interfaces between blocks. If the blocks are intended to be functional - to include nanomachinery - then there will have to be interconnections between each block. If the interconnection systems are 10 nm thick, then almost 50% of a 100-nm block will be used up in connection hardware. A 200 nm block would use only a bit over 25% of its volume for interconnects.

There's not necessarily any upper bound on block size, but I don't see any reason to make them bigger than 200 nm. One nice feature about 200 nm blocks is that there's a fairly low chance per block of background radiation damage. You can pretty much design the machinery in a single block without worrying about fault-tolerance, and do your fault-tolerant design at a higher level. If your block was a micron or bigger, you'd have to assume that something inside it would probably be hit by background radiation in less than a year, so each block design would have to include some fault tolerance.

Jim's question included the suggestion that planar assembly might be used at small scales to make parts that could be assembled into the final product. To a first approximation, I don't think this is necessary, since the final product can be made directly with the small blocks. Rather than building a part and then fitting it into place, just build the part by planar assembly in its final position. That means you'll be building lots of fractional parts in layers at the same time, but I don't see a problem with that. The block-joining mechanisms should work the same in either case.

One more advantage to building the entire product directly out of small blocks: you never have to handle parts that are big enough to be heavy. In other words, you don't have to worry about gravity.

So planar assembly is an extremely flexible approach to building products, and should save a lot of volume and design complexity in nanofactories.

Chris Phoenix

Breakthrough! Particle Assembly Technique

Stick nanoparticles together only when you want to! Researchers at NYU have announced a way to use DNA strands to make nanoparticles sticky only under certain conditions.

If you can make nanoparticles stick on command, you're a big step closer to building a simple nanofactory.

In the study that I did with Tee Toth-Fejel for NIAC a few years ago on Large-Product General-Purpose Design and Manufacturing Using Nanoscale Modules, I proposed that a primitive sort of nanofactory might be built with molecular building blocks that would bind together when pressed together, but otherwise float around in solution. The idea was that the blocks would bind weakly to the tip of a "tattoo needle" which would then press them into a workpiece, where they would stick. Given an appropriate set of building blocks, it should be possible to build arrays of tattoo needles.

DNA strands stick together when their sequences match up properly. Sticking nanoparticles together by coating them with matching DNA strands has been done for a while. But the innovation here is that a strand with the right sequence can fold back and stick to itself.

A particle coated with doubled-over DNA strands will not be very sticky. But heat can unstick the strands from themselves, making them able to stick to each other, linking the attached particles. The article implies that holding a particle in position long enough will allow the strands to randomly uncoil and stick - whereas particles that bumped into each other in solution would almost always drift away before this could happen.

Another interesting point in the article: by making use of reversible attachment and conditional attachment, it seems possible to make intricate building block arrangements that can template copies of themselves. This could allow exponential growth of the patterns, even without externally controlled mechanical guidance.

DNA is extremely versatile, and the strength of the attachment between strands can be fine-tuned. The main thing I don't know here is the time scales involved. If it takes an hour to deposit a molecular building block, then this may not be useful for construction. But if it takes a minute or less, then things get very interesting.

I don't know yet if this innovation will go down in history beside Rothemund staples. But it is certainly complementary to it (no pun intended) since suitable molecular building blocks can be very easily built out of Rothemund-stapled DNA.

It wouldn't surprise me at all to see a computer-controlled DNA-based molecular manufacturing system within a few years. The material properties would not be all that great, and it might or might not have fast actuation, and it would require expensive feedstock and carefully controlled laboratory conditions, so it would not be revolutionary in an economic or military sense. But it would be an awesome proof of concept, leading people to wonder what other materials and binding systems might be used to build higher-performance nanofactories.

Chris Phoenix

Concise Summary of Molecular Manufacturing

Eric Drexler over on Metamodern has posted a summary of "The Physical Basis of Atomically Precise Manufacturing."

Most of what he says isn't new - it comes straight from his 1992 reference/analysis work Nanosystems. Small things work faster and have higher power density. Detailed and conservative analysis shows that molecular-scale objects can be built by molecular-scale objects. (Biology is an existence proof of this last point - but the kind of machines Drexler analyzed have fundamental performance advantages over biology.)

Drexler makes an interesting point about the difference between design for easy analysis and design for easy construction, and he provides links to earlier posts of his on things like alternative materials for machine-type nanoscale manufacturing systems.

He also discusses the amount of time required for fabrication, pointing out that this is proportional to operation speed, and thus can be expected to be faster (per machine mass) in smaller machines. Perhaps in a future post he'll go into more detail on nanofactory architecture and molecular fabrication vs. component assembly. (My nanofactory paper explores these issues for a particular desing of fabricator, but it's a lengthy read and the convergent-assembly design is now probably outdated - the Burch/Drexler planar assembly design seems better in almost every way.)

Though he didn't include this observation, scaling laws also describe the huge difference between handling molecules with big machines and handling them with small machines. Not only do the small machines work faster in proportion to their size, but their volume changes drastically. People's intuition correctly tells them that a desktop machine could never build a copy out of molecules - it would take billions of billions of years. But what their intuition won't tell them is that, if you shrink the machine by a factor of a million, it should be able to build a duplicate out of molecules in a few minutes.

Even if you're familiar with molecular manufacturing, it's worth reviewing all the cool and useful things that happen at the nanoscale - which means it's worth reading Drexler's article.

Chris Phoenix

Ions Smooth The Way For Protein On DNA

Although I generally prefer machine-like nanosystems to squishy nanosystems, I am capable of appreciating elegance in the squishy realm. The latest Physical Review Focus story, "The Electric Slide," explains how proteins can slide along DNA.

Some proteins perform operations on DNA at specific points - they recognize particular sequences of base pairs. For efficiency, this recognition has to happen in several steps: first, the protein snuggles up to the DNA at a random location; then, it slides back and forth; then, since it's close to the DNA, it can recognize the spot where it's supposed to act; finally, it does its job. So it has to be close to the DNA, but not too close.

This is not as simple as it sounds. Although some atomic surfaces - including stiff machine-like interfaces such as tungsten on graphene - can be almost friction-free, many similarly bumpy surfaces are not. In order to slide smoothly, the protein has to avoid binding too tightly to the DNA. It has to stay physically distant from the DNA - but not too distant: it doesn't just float randomly near the strand, but actually slides along it without touching it, like a mag-lev train.

The researchers simulated a simpler version of protein and DNA. Instead of simulating every atom in the molecules, they abstracted them as a long skinny cylinder for the DNA, and a cylinder with a notch in it for the protein. (They tried several other shapes, but that's the combination that worked.)

Protein and DNA are charged molecules; DNA has a negative charge, meaning that it will attract positive ions from salt water, and protein has a positive charge. So the two molecules will attract each other. But why don't they stick together too tightly to slide?

It turns out to be important that DNA has a stronger charge than protein. When protein and DNA get close, there is still a layer of positive ions between them to balance out the charge. And if they get too close, the ions will be too concentrated relative to the water. So the charge difference keeps a layer of ions between the DNA and protein, and osmotic pressure keeps a cushion of water around the ions, and the water holds the DNA and protein apart.

The researchers did some analysis of real DNA-binding proteins, and found that the proteins have on average 17% of the charge of DNA. Plugging this into their model, they found a separation of half a nanometer - which matches what's been observed, and is a suitable spacing for low-friction sliding.

Since this sliding mechanism relies on the motion of ions and water molecules, I suspect (though I haven't calculated) that it would only be low friction at low speeds. A graphene bearing would require less motion of its atoms, so it should be able to move considerably faster. So I still like machine-type bearings better. But as long as you're working under water anyway, you might as well use ion cushions.

Chris Phoenix

Hobbyist Molecular Manufacturing

Suppose that someone found a way to make a molecular manufacturing system not much more complex than a DNA synthesizer. Not a full-blown diamondoid nanofactory, but something cobbled together out of (for example) a DNA staple matrix, a video projector, several computer-controlled valves...

Suppose that this hypothetical system could build, in a day, a few square millimeters of stuff with the strength of thin plastic wrap... but with many millions of features, atomic or near-atomic precision, actuators (but only flex bearings) and electronics, and the ability to attach DNA-tagged molecules or nanoparticles at selected locations by self-assembly.

I'm not immediately sure what this would be most useful for. It would not make full products, but it would probably be useful for custom electronics, and might be bootstrappable to a somewhat more capable molecular manufacturing system.

At the recent Maker Faire, I was impressed by the amount of diversity and creativity I saw: people taking quite primitive building blocks and making really cool (and sometimes even useful) stuff. So I'm assuming that this basic capability would be useful for something, and at least a few hobbyists would build and use the system.

How fast would general-purpose software be developed to drive it?

How fast would new materials become available?

How fast would better systems evolve?

It'll be very interesting to see what RepRap and its cousins do over the next few years. That may give us useful hints about the potential of really primitive molecular manufacturing systems.

Chris Phoenix

So What About Microfactories?

I got an interesting question in my email earlier this week: with all the buzz about RepRap, is it plausible that microfactories could be developed soon, build many things cheaply, and then transition smoothly into nanofactories as the feedstock improves via nanotechnology?

The short answer is: yes and no.

Yes, because microfactories are already being developed. Open source designs such as RepRap are becoming competitive with commercial designs costing ten times as much. RepRap is designed to build (partial) copies of itself, driving down its cost. It uses inexpensive materials, and a rapidly expanding range of them.

It's possible to imagine an economically significant consumer-model rapid prototyping machine, sometime in the next 5-10 years, that can build a wide variety of products, using a variety of inexpensive solid, liquid, and melt-able inputs. It's possible to imagine that the product range will expand toward factory-manufactured quality, and meanwhile the inputs will become "smarter" as they increasingly incorporate nanotech.

But there are several reasons why a nanofactory will be better than a microfactory. There are multiple physics features that are very beneficial at the nanoscale but work less well at the micro-scale.

Perhaps the most important physics feature is operation frequency. The smaller you make something, the faster it works. The relationship is more or less linear, so a nanofactory's equipment could do a thousand times as many operations per minute as a microfactory. Combine this with the length-to-volume ratio, which means that you can fit a billion devices into the space of one by making them 1000 times smaller, and we see that a nanofactory could do a trillion times as many operations as a microfactory. (That's 10^12, or 12 orders of magnitude.)

So, while a microfactory was placing a tiny amorphous blob of plastic or droplet of ink, a nanofactory could be arranging a billion atoms into a highly functional nanoblock. Even if the microfactory were placing a pre-manufactured component, that component would not have a billion features - because nothing today can build an inexpensive device with a billion features. 

Making a late-night no-math guess about what objects might have a billion features, the first three that I thought of were the Space Shuttle, modern computer chips, and... a potato. Potatoes, of course, are made by nanotech, and they cost $1 per pound. Computer chips are made by microtech, and they cost many thousands of dollars per ounce. The Space Shuttle, of course, represents astronomical cost (no pun intended).

So it's possible that an advanced microfactory might use feedstock particles that incorporate dozens of features, and might place millions of them, and might build something with a billion features for a few hundred dollars. If this seems implausible, consider that molecules made by ordinary chemistry can have dozens of features, and an inkjet printer can print many millions of dots per minute and can print its own weight in ink in about a day. So in a sense, what I've described already exists - just not in a 3D factory version yet.

Another physics feature is that atoms are perfectly precise, and snap together with perfect precision. That's not true of microtech stuff. It might be hard to make a factory that could build duplicate factories without loss of precision out of amorphous feedstock. Of course, if the feedstock was, say, microtech structures built by lithography on million-dollar machines, then larger precise structures could be built, which might be cheap per feature, but they wouldn't be cheap per pound. (Think this is science fiction? Zyvex has been doing snap-together MEMS for years.)

It would be great, of course, if a microfactory could include enough lots of microtech fabrication systems, each system having been built by a microfactory. At that point, the feature count of products could go through the roof, at least by today's standards. I can think of a few possible approaches to try, but nothing that I'm sure would work, and nothing that could be developed quickly.

When I started writing this post, I was expecting to demonstrate that a microfactory couldn't work well enough - that its products would be so non-revolutionary, thanks to the less-cool physics at the microscale, that it wouldn't be worth engineering or selling. But the more I think about it, the more I start to suspect that there may be some interesting designs lurking somewhere in the design space. They'd be pretty different from today's gantry-crane melted-plastic fabbers, but that's an engineering challenge, not a criticism.

Chris Phoenix