Telegraph: Could nanobots destroy us?

The "grey-goo problem" was floated in 1986 by the futurist Eric Drexler in his book Engines of Creation. Three years ago, Bill Joy, co-founder of Sun Microsystems, remarked: "Gray goo would surely be a depressing ending to our human adventure on Earth... and one that could stem from a simple laboratory accident. Oops."

Earlier this year, the "incalculable risks" were spelt out by the Canadian group Etc in an extended polemic on the potential evils of nano-tech-nology, entitled 'The Big Down'. These warnings have now reached the prominent ears of Prince Charles.

Think small, really small. Nanotechnology is the ability to manipulate matter at scales of billionths of a metre. Because atoms are around one tenth of a nanometre across (there are around 200,000,000,000,000,000 in the full stop at the end of this sentence) building a nanobot would be akin to building a robot with Lego. Thus a nanobot would not have a smooth surface, like the ones we are familiar with, but an eggbox-like texture.

Dr Na-No lists the bot's components: it would require an assembler, some kind of device to select atoms and put them together to "breed" new nanobots. It would require propulsion, perhaps based on cilia, the waving fronds that row bacteria about. And it would need a "brain". Perhaps patterns of electronic charge flickering across conducting polymers could be used to issue the command to "duplicate me".

But it is a mistake to think that a nanobot would be a miniaturised version of the mostly awkward robots that engineers have already built, only made of atomic Lego. First, Lego is a lousy analogy. Real atoms are picky about the other atoms they bind to - the chemistry has to be right - and they stick together in many different ways.

At every-day temperatures, they also bounce about a lot as a result of heat energy. And the mist of electrons around one atom (which controls how it glues to others) can deform and change the electron clouds around nearby atoms, so these Lego bricks can alter each other's "stickiness".

At atomic dimensions, one also has to worry about Heisenberg's Uncertainty Principle - the more you know about a subatomic particle's position, the less you know about its momentum, and vice versa. How can you manipulate atoms if their component pieces are so hard to find?

Easy. In Dr Pike's laboratory, he can already boss about atoms with the tip of an extremely sharp needle in a device called an Atomic Force Microscope (AFM). This big, primitive assembler has to be cooled far below zero degrees C to stop the atoms it manipulates from jiggling about. And atomcraft is slow. It took an IBM team 22 hours non-stop to write the company trademark with 35 xenon atoms.

Another problem is to find a way for the nanobot to pick the right atoms to stick together to build a brother. Dr Pike uses a flash of light and analyses the way it is absorbed to identify an atom. Other approaches rely on an electronic interrogation of an atom stuck to the AFM's tip. Although not yet solved, this feat would require many more atoms in addition to the 10 million or so Dr Na-No would need for a basic mechanical assembler.

Nanobots also bother Prof Richard Smalley of Rice University, a top atom wrangler, having won the chemistry Nobel prize. Working furiously, a hypothetical nanorobot could forge many new bonds as it went about its assigned task, placing perhaps up to a billion new atoms in nanobots every second. "But as fast as it is, that rate would be virtually useless in running a nanofactory: generating even a tiny amount of a product would take a solitary nanobot millions of years."

Unless the nanobot works in a vacuum, it would be buffeted by air molecules, as well as those of whatever it is grazing on. The "fingers" of the assembler have to be rather nimble to grab the right atoms. This raises what Prof Smalley calls the "fat fingers" and "sticky fingers" problems.

Because the fingers must be made out of atoms, they have a certain irreducible size. Manipulator fingers on the hypothetical self-replicating nanobot are not only too fat; they are also too sticky for some atoms. So it will often be impossible to drop a nano building block in the right spot.

Perhaps nature can offer inspiration. She has assemblers, such as the ribosome, which helps turn our genes into proteins. Dr Venki Ramakrishnan of the Laboratory of Molecular Biology, Cambridge, likens the ribosome to a factory, consisting of three RNA molecules and 50 proteins - around 250,000 atoms in all. But it needs a host of protein factors to work in our cells - and this assembler is only capable of making part of itself, requiring the help of many other molecules to reproduce.

The complexity of nature's nanomachines also shows how hard it is to solve the sticky problem. To enable an oxygen molecule to be grabbed, transported and released, the haemoglobin in our blood uses 12,000 atoms.

Dr Na-No's 'bots also need energy. Perhaps they can obtain chemical energy as they trade high-energy chemical bonds for low-energy ones. "Diamond-like carbon is the favourite construction material," said Dr Pike. "But diamond has the strongest bonds of any material. Good food for such nanobots would be hard to come by, outside of a laboratory." Remember that the grey-goo scenario suggests the 'bots can dine on anything, rather than restrict themselves to a diet of high-energy carbon.

There is another problem. While biological systems can cope with manufacturing mistakes (mutations), mechanical robots break down. "You are going to be fighting entropy [the natural tendency for disorder] all the way," said Dr Pike.

To create a self replicating nanobot would take, he estimates, billions of atoms. The end result would not be a true nanobot but something bigger. In fact, nature teems with them: bacteria. At around 1,000 nanometres, they count as microbots, not nanobots – viruses are smaller, at 20-100 nanometres, but need living cells to multiply, that's too restrictive to be a grey-goo 'bot contender.

Even after 4.5 billion years of life on Earth, Mother Nature has not come up with omnivorous self-replicating nanobots. The microbots she has developed are fussy about what they eat. Some like flesh, a few even wet rock. Many graze in our stomachs. None has come close to consuming the world.

Removing his black eye patch and the white cat curled up in his lap, Dr Pike sighed and stepped out of his Dr Na-No persona. Rampant nanobots are firmly in the realm of science fiction. The reality is that nanotechnologists are working on new ways to release drugs, create sturdier building materials, fabricate stain-resistant clothes, make flexible computer displays and faster transistors. But not murderous nanobots. Prince Charles need not have any more grey-goo nightmares.

Telegraph | Connected | Could nanobots destroy us?


Post a Comment

Links to this post:

Create a Link

<< Home