• 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

« Is CRN Playing Politics? | Main | That's What I'm Talkin' About »

June 14, 2004


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

jim moore

Sub question C
I think you made a typo, isn't the error rate one error in 10 to the 8th not one error in 108.

Sub question E - raw material
I think adamantane (molecular diamond C10 H16) might be a good choice. It has very interesting physical properties. It is a white powder with a very high melting point 269 degrees C but it also sublimes at room temp and pressure. (kind of like being a gas and a solid at the same time)
The big drawback to using adamantane is that it is an uncommon raw material. (at least for now)

Mike Treder, CRN

Yes, you're right about the error. Thanks for catching that -- it's been corrected.


Subquestion F:
How do the 250 kWh/kg compare to the energy required to produce 1 kg of, say, potatoes, just to get a feeling for the numbers?

In my opinion, the energy equivalent of 28 liters of gasoline is pretty much energy for 1 kg of product, which only holds true for 100% nanofactory efficiency anyway. Is there any product today that would by estimate consume 250 kWh/kg or more during its assembly process?

On the other hand, what would be a reasonable lower limit of energy consumption for products out of an advanced nanofactory at useful speeds? How much can the efficiency be improved?

Chris Phoenix, CRN

I've read that plants are about 1% efficient. So the nanofactory takes more energy per kg, but not by much. And the products are far more useful than potatoes. Look at the cost of that energy--$20 or so per kg--and compare that with the cost per kg of, say, semiconductors.

I'm not sure what you mean by "100% nanofactory efficiency." My estimate left room for lots of inefficiency. An efficient nanofactory (for example, one that made 99% of the product bonds with molecular mills) might use far less energy.

I don't know, even within an order of magnitude, how much efficiency can be improved. But I don't think it matters for most applications. When you consider the superior material strength (meaning, you need to produce much less mass for the same structural functionality), the nanofactory's products become competitive with today's bulk materials even at $20/kg.



With 100% efficiency I meant that if a gasoline-powered nanofactory could convert the full 100% of the 9 kWh/l of gas into work (i.e. assembling atoms), it would burn ~27 liters per kg of product at 250 kWh. So if it were 50% efficient at converting, it would need 54 liters for 250 kWh. If it were 1% efficient like potatoe plants, it would need 5400 liters per kg of product.

But maybe I got this wrong and this converting inefficiency was already factored in?

Even if, these numbers gave rise to my main concern regarding energy consumption (if it is worth of being a concern at all, hence my question about comparative values). It is less the question of economic competitiveness, more of the net enviromental effects.

If a product were competitive despite double the energy consumption (which is likely for MM products since the energy price by then is likely to drop due to cheap solar energy), there still remains the environmental cost. And demand for a both more affordable and technically superior but more energy-intensive (during manufacturing only) product is very likely to grow (judging from my own lay understanding of economics), if the environmental conscience of customers is not able to surpass the former aspects (price and superiority, which is NOT going to happen if grain-sized "Earth Simulators" are as cheap to buy as their weight in sand).

On the other hand, with MNT the global energy consumption -not to forget the raw material need for manufacturing per product, both in structure and waste- will drop in other places, like in large-scale shipping of goods around the world (which can mostly be abandoned), more efficient personal transportation, heating, refining, waste production and disposal etc., maybe making up for that potential increase in manufacturing energy consumption. Is it even possible to start analyzing the question which will outrun the other, energy consumption increase through efficiency decrease, or energy savings by efficiency increase in other places?

In the long run, manufacturing will probably be more efficient than this proposed nanofactory might suggest, since nanofactories will evolve too. So what do you think, what are possible and probable short/middle/long term environmental costs of MM? And how redundant is this question anyway, was I just too lazy to look it up in your preliminary questions?

Chris Phoenix, CRN

No, the converting efficiency wasn't factored in; I was just assuming electricity, which is convenient both to use and to price.

Will it be worth building energy-using products with MNT? Absolutely yes. Power density and computation density are insanely high--which means that it'll take tiny specks of material to replace today's motors and computers. The energy cost of building them will be far less than the energy cost of making silicon and copper. And MNT-built products should be very energy-efficient as well.

The bigger question is: what new products will we make, and how many of them, and how much energy will they use, and what will the environmental impact of all this be? We list this on our "Dangers" page and in Study #26 http://crnano.org/study26.htm



I am not in the field and dont really have any basis for Im about to say other than me reading a few articles, but I hope you like it regardless.

I read an article not too many days ago about a an atom being able to be suspended within a laser beam because of the gravitational force of the photon concentration (or something like that). yet I always am seeing people talk about nanofactories and using strictly chemical procceses or using tiny robotic arms. well if an atom can be held by 1 laser, could then an atom be moved along that one laser by an intersecting laser, and couldn't a pulse be aplied that contained a catlyst to bond atoms together, and then if you were to say multiply the amount of lasers to say 1000-10000 lasers and 1 really sweet computer program, be able to place and bond trillions of atoms/hour once the system was sped up to even todays top laser technology. Just a thought i had the other day, I even have grade school sketches, lol. Anything Im not quite grasping fully, please, refferance me to a rebuttal, or supply one yourself, cause this has really been eating at me.

jim moore

Check out a company called Arryx,
Their system works at a larger scale but it is really neat.


Thanks much jim, that is really neat. Im wondering if it their machine could be tuned even further, and a very fast method of bonding individual atoms could be introduced to the system, if it could handle the number of atoms required to make macroscopic products. although my origional thought of even 10,000 lasers may have to be raised to about a million, but still.

jim moore

I don't think a laser system can be precise as nano-machines or chemistry in positioning individual atoms or small molecules. To be atomically precise you need an accuracy of ~0.1nm this is deep in the X-ray region of the em spectrum. As the wavelength of light gets shorter and shorter, each photon packs more and more energy. Long before the time you have a laser that you can focus on a small enough spot you have a laser with a wavelength that will pump more than enough energy into the molecule to break the chemical bonds you were trying to make.

Now that being said, there might be some very interesting hybrid, laser-nanomachine or laser-chemisty systems.

Chris Phoenix, CRN

To hit a single atom with a photon, it's true that the wavelength has to be extremely small.

But in an optical trap, the object is forced to the brightest part of the beam(s) with sub-wavelength precision. I don't know whether this means that single atoms can be positioned precisely enough for eutactic chemistry. (It'd be especially hard to do this near a surface.) But I do know that optical tweezers applied to microspheres have been used to characterize the forces involved in single steps of biomotors (8 nm?)


Brett Bellmore

I'd say there's more potential for atomic optics, including holography; The effective wavelength of an atom is incredibly small, so theoretically you could achieve the required resolution. Might even be able to directly "image" some smaller components. It IS hard to imagine optical tweasers working at all close to a solid surface which wasn't designed for transparency.


For the moment scrapping my previous posts.

I was just running numbers for assembly of individual atoms, and came to the realization that I cannot comprehend the speed needed to assemble even 1kg of material. Then I went back to the idea or self replication, and for arguement sake, if 1 atom went through mearly 69 cycles of duplication you would have over 1kg of material.

Would it be more feasable then, to make an effecient self replicating component system rather than atom-by-atom assembly with one machine. And is that what is refered to when everyone talks about "building blocks". I was assuming that when building blocks were mentioned that an initial machine had to produce all of them individually and not the blocks replicating.

Also If you could tell me just how many atoms might be able to be moved and placed by a mini-robotic arm, also mentioned, and how many arms are you talking about.

Chris Phoenix, CRN

Mark, you're exactly right: positioning kilogram quantities of atoms appears to require exponential manufacturing. (Self-assembly is easier, but you have to build all the product's complexity into the molecules before you start.)

A 100-nm robot arm and associated equipment might require (very approximately) one billion atoms. Scaling laws indicate that it could place a million atoms (and/or molecular fragments) per second. So it would take a fraction of an hour to duplicate its mass.

"Building blocks" could refer to any of several things. My nanofactory design is based on "nanoblocks" which are small enough to be quickly produced by a single robot arm, but big enough to contain a single robot arm--or a simple CPU, or motors, etc. Molecular building blocks could mean a feedstock molecule like acetone.


Brett Bellmore

Really, I've been giving this some thought, and while I can see some use for the general purpose assembler CRN proposes, it's not really necessary or even desirable for most home use.

99 44/100ths percent of home needs could be met by a molecular mill churning out a limited number of types of "utility fog", and a MEMS chemical factory designed to produce reagent chemicals and a limited number of biomolecules to assemble synthetic food from. Power consumption per kg of product would be much lower than any system based on robot arms. Home design would require less knowlege, simpler software. And it wouldn't be capable of self reproduction.

Chris Phoenix, CRN

Utility fog would probably use a fair amount of power, and be hard to write software for to make it simulate products. Besides having rather poor "material properties."

General product-building could also use mills; a cubic nanometer of diamond has 176 atoms, so even sticking together small pieces would save most of the robot-arm work.

Without more detailed architectural designs, I'm not sure how much more subversion-resistant a UFog system would be than a nanofactory system. I suspect that premade nanoblocks would be about as hack-proof as UFog. (Only somewhat, in either case.)


Karl Gallagher

My assessment of the nanofactory's capability came out much lower. I've written up my take on it here and set up a place for comments and discussion here.

Chris Phoenix, CRN

I've answered Karl's critique on his site. Basically, where I left design options, he did a lot of things the hard way and assumed they couldn't be improved. And he complained about some things I'd already addressed.

There were also some things that I didn't explain fully, and I addressed those. But he has not yet made me worry about any aspect of my design.


The comments to this entry are closed.