• 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

« NanoNews-Now: Special Issue | Main | Supplying Basic Human Needs: Good or Evil? »

March 08, 2006


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


The biggest problem here in my opinion is lack of evidence, it seems that when there are major issues facing Molecular manufacturing they're paved over with speculation and ones opinion that in theory fixes the problem but in reality there is no substance to the,, in simple terms it seems there is alot of "making up your own rules" going on.

Tom Mazanec

Might be interesting if you posted something like links to places where you addressed the three "practical problems" you gave in (1).


Chris: from the point of view of someone trying hard to understand your ideas its hard to separate the sci-fi from your real proposals when you say stuff like this. in 2004 you were saying that in a few years mm would allow villagers to forage for things which could be converted to oil and then this could be converted into products and bland food. sometimes its hard to work out what your basic proposals are on the big scale.

on the small scale it gets harder!! like your spms for building things. will it be an afm? is there a proposal for it (however early and incomplete) and if it doesn't use piezoelectric elements what does it use? how does it feel the thing its making?

you say that people keep criticising "off-topic proposals" and not "what's been proposed", but instead of rushing in with "its like this...." or "have a look at this design..." you are strangely silent! if your spm was a robot arm you could say that really quickly and everyone would understand what you are talking about. i can't speak for mark but he might have assumed you were talking about piezoelectric spms because you keep pointing to their research and saying its relevant. if you don't know how your completely new type of spm will work yet then its ok to admit it!

Brian Wang

Monty and others wondering about the proposals,

There are a series of proposals because this is a large and difficult challenge with many hurdles.

Some of the earlier tools along the series of tools and devices that are needed.

Have you read Design and Analysis of a Molecular Tool for Carbon Transfer in Mechanosynthesis”

The first seven papers and books

Hydrogen abstraction tool


The point in is start at the technical and science papers. Bring up technical objections against technical work. One blog article with social commentary out of 1000 blog articles is not the place to get more technical understanding.

Mark Wendman


I appreciate the effort and polite discourse.

To those not doing experimental nanotechnology related work (ie theory and supposition don't often have enough connection to the reality of what is actually doable - book reading alone does not cut it), I sense a repeated emphasis on conjecture from person dependent inexperienced (experimentally) perspectives. Even the wording is inconsistent nevermind almost obtuse on occasion.

It is nice to try to be a science policymaker but it does not make for innovation nor experimental results.

Reframing what I wrote, while helping to start a new thread, does not change the technical substance of what I write.

Some of what you interpreted in your summaries is technically confusing and unrelated to what exists, and might never actually be done.

My description of tips is accurate and clearly you don't understand this. Despite all kinds of obsession with fullerenes, they are NOT useful for tips for a whole host of reasons you miss.

There are some limited advantages in some classes of imaging when nanotube tips are used, but there are serious technical challenges to useful implementation. Any buckyball derivative is useless for any imaging or other probe based application.

You can refer to the 2nd earliest post to my blog about positional errors in imaging metrology with high aspect ratio tips - this is based on experiment and tip limited due to lateral flexing while scanning.

You can slow down but never eliminate the observed effect when a certain amount of high aspect ratio is seen in the tip structure. This is driven by adhesive forces overcoming vertical / lateral stiffness in the probe end.

You will benefit by reading if you want technical substance. My blog is linked, but here is the direct permanent link to the post


The subject matter is novel and would be publishable (even patentable), but I do not have the cash to spend on publishing independently, nor patenting.

I repeatedly state that Cambrios's BaISA (my name for it) IS a practical implementation of an MA system and is likely to be as good as it will ever get.

And everyone here has mostly ignored the factual status of this and its implications, trying to claim some otherwor(l)d(l)y definition (each with their own style of reinterpreting what I wrote, without adressing what I wrote point by point technically - not vague position paper philosophy).

Using a blunt tip results in huge positional errors. Using a nanotube tip even if closed end results in donor (carbon) emission imprecision. You cannot do better in positional placement accuracy than the sharpness of the tip for chem mechano atomic assembly by chemical emisson from a tip nor due to the tip shape change by emission induced and other sources of wear. It goes on and on.

Rigidity of the tip and its end matters.

As to scanners drift and the like - again what I write is factually correct and not understanding it relates to lack of experience.

Eiglers system at IBM (famous atomic quantum pick and place corral) is best in class acoustic isolation, best in class electronics, best inclass mechanics, I can wax poetic. Still drifts AS I described.

One can say the piezo is not the source of drift, and it is not huge in Eiglers setup small peizos and cryo cooled - but overall drift after easily a day of temp stabilization was not useful for non-imaged sub atomic positional placement accuracy ( ie without rereferencing) - I wrote this quite explicitly, and you still don't speak to this.

Nor has Brian responded tecnically to this point which directly contradicts his claim as to the ncessity for sub atomic positional accuracy for defining MA.

The ONLY non Tunneling (ie AFM) imaging which was subatomic IMAGING only was and to the best of my knowledge mostly remains so is Giessibl in Germany - Augsberg if my memory serves me right and his system is instructive in many aspects of techniques need for imaging stability but NOT sub atomic positional 3D MA.

You can handwave that there is something better, and that peizos behave as to your conjecture is, but I have my reasons based in facts of observation of experiments and apparatus that is not refuted by reworkding and reworded reinterpreting what I wrote.

Go back and rather than rephrasing what I wrote - examine each point and discuss THAT.

You will learn from it.

I describe technically the limits that will affect successful implementation of the dogma of arbitrary sub atomic precision "programmable" assemblers.

BaISA from Cambrios / Prof. Belcher is coming along quite nicely and it is not MA, but will work in a useful dedicated non adaptable non programmed manner for a limited class of "products".

I seem to have the impression that everyone wants to correct me by changing the wording /interpretation of my responses to questions but not addressing what I write poitn by point in response to requests.

And much later a few here undoubtedly will conveniently at some indeterminate point claim Cambrios's work as THE ASSEMBLER.

Trust me, I see it coming.

It does not fit the requirement for subatomic arbitrary position placement that so many here refer to as defining MA. Especially Brian being so emphatic about sub atomic precision in MA - read what he wrote.

BaISA is Bio Assisted Inorganic Self Assembly - and that is what it is. Viruses enslaved for mankind in inorganic [self] assembly.

Best Regards,

Mark Wendman
the fatigued nanotechnology experimentalist

Brian Wang


Do not claim that I am emphasizing something when it is one sentence in a typed discussion.

You state that you "do not have the cash to spend on publishing independently". Submit your "brilliant" and "ground breaking" work to a journal. This does not require money.

"No cash for patenting". You can file a provisional patent for $80.

Borrow a nolo press book from a public library.

Fill out the forms and paperwork and try to get a provisional patent good for one year. Then go and talk to the microscope and tool companies that so clearly need your "breakthroughs" and "experimental insights".

Don't complain about not getting rich from your brilliance. Go out and do it. If you are able to leverage my advice on this low cost approach, the only thanks I will need is not having to be asked to discuss everyone of your mind droppings point by point.


Brian: i looked at all the links but there wasnt anything about the thing i was asking about. most of them were about tooltips! the only ones about positional control were a stewart platform and some robot arms but these don't have spms in them. does this mean noone in mm has ever thought about it?

isn't that a fundamental flaw if you can't think of a way of making an Spm in less than 100nm!

i'm going to use chris's template for providing answer 2....!!

I don't see how you can do "SPM smaller than 100nm" (or be sure of "SPM smaller than 100nm") because "i looked at all brian's links and couldn't find an answer" so you have to anwer that."


Isn't the discussion about the pitfalls of current STM technology a bit of a strawman when considering the merits of the theory of molecular manufacturing? ie. http://www.molecularassembler.com/Papers/PathDiamMolMfg.htm

Or which elements of current STM technology would be relevant when considering how the proposed placement tool in the link above would operate?

I'm not a scientist!

Brian Wang


You looked at the links but you did not understand the content.

Although it might be nice to make a atomically precise positional control device that is smaller than a 100nm on a side cube. It is not required to achieve the benefits of molecular manufacturing.

So it is not a flaw in a proposal to not have that characteristic.

None of the proposed solutions have to have a scanning probe microscope or use any particular technology. They just need to be able to move and position molecules precisely for mechanochemistry.


Brian: i said a 100nm Spm becuase thats what chris said at the top of the post!

"Molecular manufacturing proposes to build a form of scanning probe microscope on a sub-micron scale"

"Since the time for an SPM to process its own mass in atoms decreases as roughly the fourth power of the size, a 100-nm SPM should in theory process its own mass in about 100 seconds."

it was his idea not mine!!

also he said; "This should be enough flexibility to implement nanoscale machines (such as SPM's) and to implement a fairly wide range of micro- and macro-scale material properties"

so has chris got it wrong?

Chris Phoenix, CRN

Mark - I have two reasons for rephrasing what you wrote. One is to make a coherent blog post. The other is to check if I understand you. If my rephrasing is incorrect or incomplete, please tell me where. I did read every word you wrote in the other thread, many passages twice or three times, and I don't have time to read it all again. Tell me what I missed and let's move forward.

Monty - The Stewart platform reference above has been proposed as the actuator part of a 100-nm scanning probe microscope. (I think Brian may be giving up too much; an SPM a lot bigger than 100 nm will be too slow processing its own mass of atoms.) The mechanism for feeling the workpiece hasn't been specified in detail, but it shouldn't have to do general-purpose imaging--just distinguish between a desired state and a few error/retry states. Some operations shouldn't require feedback at all.

Mark - I think we're talking past each other on SPM tips. My question about buckytube tips on graphite was only to ask about wear, not precision. I wasn't proposing buckytubes as the tip. Just asking whether a covalent structure (of which buckytubes are an example) could be wear-free on a clean covalent surface.

Likewise, I think you haven't understood that I understand that doing multiple deposition reactions from a single tip, without regenerating the tip, would be difficult or impossible. I am certainly not proposing to deposit carbon atoms from buckytubes! Instead I'm proposing to do one deposition from a sharp covalent structure on a relatively fat tip, then regenerate the structure, then do another deposition... Hence Brian's pointing you to papers on how to regenerate molecular tool tips (not to be confused with SPM tips).

For precise AFM, what about Oyabu's work? He not only imaged Si atoms, he pick-and-placed them from a Si crystal with an AFM.

As to thermal drift in Eigler's apparatus - His apparatus is many orders of magnitude larger, and many-many orders more massive, than the apparatus I'm expecting MM to use. Doesn't that mean that the thermal drift issues will be utterly different?

I'll take a look at BaISA, and probably write a post about it soon.


Mark Wendman


I understand. At least we can converse and you are pleasant.

Tip regeneration is how would I put it - speculative, esp at a near to atom scale.

You wear and it is gone. I have a wacky idea to use a nanotube SWNT as an atom dispense tube but that is so far fetched for me to try, even though it would be fascinating, again I have the skills but no $.

There'd have to be quite a bit of stuff set up to begin to test the crazy possbility and VERIFY it experimentally and then come up with a practical scheme for doing so with positionaL and dosing accuracy....

But as to tip regneration, the only larger scale I can beleive in, is analogous to a taylor cone FIB source but the LIMS - liquid metal ion source config is impractical for a broad class of real and call it necessary applications, never mind the high voltage requirement for sustaining a taylor cone in the LIMS.

At an atom scale, there is no such thing as what I interpret you are wishing for, at least to the best of my knowledge experimentally, and most of all practically.

I tend to avoid reading speculative conceptual theories - or conjecture, as real nano as I contend is technique limited and not imagination limited - so I focus on technique ....

I htink if you look over Cambrios / Prof Belcher BaISA it practically fullfills most of the realistic possiblities of something which is a very close approximation of MA within practical reason. Bu tmisses on the specifics of atomic positional placement accuracy for nano scale atmomic engineered nems, but it is useful for materials syntheis at ambient temps at low cost which is almost the spirit of MA if not the details. Exponentail properties of replication from in one case viral assistance / "quasi catalysis/ MA" ( my take at least )

I contended in the my letter to the editor of Chem Eng? when Smalley wrote his Editorial . Debate that MA is really in DNA already (it is about as general purpose in the context of life as one could conceive of) and now the Cambrios example of Viral assisted is a narrower capability but almost analogous in limited scope to DNA transcription - with different details of implementation.

Many will realize it for what Belchers work is and not call it an assembler but I guess Self Assembly is a limited context to varying degrees of a type of assembler. Cambrios is viral (not limited to virus) assisted self assembly of mostly inorganic materials almost for non specific structures except for chemical composition.

There is no need for sub atomic atom placement precision and there is not on any longer range except relative chemically driven bonding with the virus (in one case virus ) almost acting like a movable catalyst - the analogy is to the nanoparticle metal catalyst for SW C nanotube growth but the applications are considerably broader in materials scope and specifically low temperature synthesis enabled by virus driven self asseembly.

I think they will look at any biological assist but viruses so far seem to work well and can have DNA mods done fairly simply and replicate for parallelsim simply for overall fast aggregate growth rates ...(exponential scaling as some call it here re MA)

Anyways I am just mumbling here... but the exchange of ideas is intriguiging - I think we are getting through to one another in a bit of a decent manner. I'll read more later and followup more in a few days. This is a start.


chris: you are funny !! why are you calling it an spm then! it doesnt scan and its not a microscope. its a P!

you've invented a 100nm P!

"The mechanism for feeling the workpiece hasn't been specified in detail"

tell us more, don't be shy :-) it hasn't been specifeid in any way at all has it !! you've no idea how to do it!

the nanofactory won't work becuuse there's no way of telling if you got each atom in the right place!

Brian Wang


I think these links will be more helpful for you to reflect.


Chris Phoenix, CRN

Monty: Thanks for the chuckle.




Mark W mentioned a "wacky" idea of using a SWNT tube to feed an SPM, presumably to simplify the sensing problem (I would guess that sensing is most difficult in the "pick up" phase than in the "placement" phase of MM) We non-chemists (I'm comp sci!) types get a little simplistic in our ideas, but I once mentioned the (presumably wackier) idea of chaining components together, PCR-like, and feeding the chain to a tool tip (possibly with some components being instructions, but thats further down the track) via some tube. Is there any merit in such an approach? I remember reading about block components in your nasa work. Could they be chained? The invention of PCR was a huge step for scaling up bio, why not nano?

Chris Phoenix, CRN

Mark -

"You wear [a tip] and it is gone." I completely agree. In fact it's not just tips--any sliding surfaces won't last long *at all* if they wear (scaling laws), and there are a lot of sliding surfaces in Drexler-type machines. But along with superlubricity, a subset of covalent solids/structures should exhibit zero wear; the activation energy would be comfortably too high at room-temperature-plus-intended-motion. (I hope broadening the topic to mechanical interfaces won't cause confusion; tips are still on the table.)

As to tip regeneration mechanisms: Can we agree that *if* we had a programmably-positioned tip, and *if* we had a way to load in molecules from a feedstock reservoir, then work such as Merkle's "hydrocarbon metabolism" *might* point the way toward a class of technique that *might* regenerate a tip *in theory*?

I'll cheerfully agree that there's nothing like this in experiment today. My focus is on classes of systems that might be built someday. Your focus is different. If at some point you want to say "I can't prove you're wrong yet, but you're too speculative to interest me further," I'll count that as honorable.

I haven't had time to look over BaISA yet, but I will--maybe not till this weekend. Anyway, I'm planning a blog post on biopolymer approaches to MM, where the direct programmability (the main information->nanoscale link) is in the polymer sequence, and the structure happens by folding/self-assembly. It may be easier, at least at first, to control temporal position of reactions (monomer sequence) rather than spatial position of reactions (deposition->shape).

When I look at BaISA, I pretty much expect to find that they don't have a high-bandwidth information path to the nanoscale, and they may not have an easy engineering approach to function->design. If these expectations are wrong, I'll be pleasantly surprised, and I'll count BaISA as an enabling technology.


Ps. Monty, I'm realizing your final sentence deserves a reply: To tell if the atoms got to the right place, use any convenient combination of sensing and pre-planning. If all your mechanosynthetic reactions are extremely reliable (less than one error per picomole), then you don't need to sense anything. If a reaction fails 1% of the time but only has one failure mode and that mode is recoverable, then you only need to sense one condition. If a reaction is less reliable than that, design another reaction.


Chris Phoenix, CRN

Marko: Feeding a polymer down a buckytube has been proposed--someone proposed a polyyne/polyacetylene chain (single strand of carbons) several years ago. I think you'd have stiffness problems on any scheme that relied on non-covalent forces to attach the feedstock chain to the tool tip. Also, there's the problem of a mechanism to feed the chain forward, though clever design might allow some sort of self-assembly-like repositioning.

Mark, I know you'll want to remind me of your comments about how buckytube tips can't achieve subatomic resolution because of lateral forces. But if I'm only building a P instead of an SPM, I'm not sure this is a problem. Just plunk down the tip where you want it rather than scanning it in sideways.

This reminds me of another point: Unterminated diamond lattice is difficult because equivalent reaction sites are so close together. (If you can deprotect only the atoms you want to react, then this problem eases substantially; this would take longer but I don't see a problem with it in theory.) But there are other covalent-solid lattices such as silica where the equivalent reaction sites are farther apart. Silica isn't as versatile as carbon, but may be useful as a stepping stone (no pun intended).


Tom Craver

Just a thought: Could one apply the pantograph principle?

The far end of a scanning arm is used to scan a known pattern, and then position the tip to a specific position, accurate to 1 atom-diameter.

Perhaps 3/4th of the way back along the arm from the reference tip toward the base of the arm, a second probe tip, with independent Z-actuation, would then be positioned to about 1/4 atom-diameter precision.

I don't know how the independent Z actuation would work, but that doesn't sound impossible to figure out.

Since this second tip would not have been involved in the "registration scan", it won't have worn any prior to application to the workpiece.

Phillip Huggan

I'm still reading up on some stuff, I won't be posting any detailed ideas for a few more days. Just wanna say I think its wrong to exclude STMs from the discussion even though utilizing E-fields and other electricity induced effects isn't strictly Drexlerian. What Mark seems to be asking for is a near term path to some sort of atomic manipulations on/of a diamond surface. STMs are the best SPMS we have right now. Diamond can be doped to be conductive, and diamonds can be used to build solar panels. So MNT scale-up shouldn't be a problem if currents are involved. This part of the discussion was lacking in previous debates.

Mark Wendman

I'd rather be grateful if I was not paraphrased as in "Mark seems to be saying".

I have been doing enough writing here that I appreciate if you quote specifics and critique point by point at a technical level. Some of the "interpretations" have been bewildering to be kind, bordering on rarified philosophies.

Focus on the specifics of the technical details and then we will be more productive in addressing things of technical substance,


I have no particular affinity for carbon / diamond, in fact since there are most common characteristics of diamond from the wide gap, it can almost behave as a practical insulator, and as such the more common field assisted methods of "assembly" may not work particularly well in a hypothetical atomic scale nanostructure build up.

Diamond doping on a near atomic scale application of atom by atom directed build up is more fuel to the complexity fire (mechanism or otherwise). And diamond doping often does not easily add sufficient conductivity to be useful.

Emphasize physics and atomic scale properties and instrumentation / actual concepts and mechanisms and I suspect that will be far better than misinterpreted paraphrasing.

What Brian kept indicating is that there is a belief / prerequisite re the absolute necessity of subatomic placement precision in adatom buildup that I contend is not in the least possible for an assembly concept of any kind, which is what I have been breaking down into technical instrumentation specifics as best as I can, and while trying to avoid being sidetracked by philosophies masquerading as technical points.

Quite simple really, moreover I am also quite specific that BaISA / Cambrious is THE vaunted ASSEMBLER (as is DNA transcription) and they exist but don't fit all the preconceptions, and in the case of BaISA - it works and merely for a limited class of applications, and will only require pretty straightforward engineering - DNA / genetic engineering modifications to for example the virus (Virus as the prototypical ASSEMBLER with real world tolerance constraints) and accepting that arbitrary 3D atomic placement is largely irrelevant to the appearance of the ASSEMBLER.

The BaISA Viral Assisted Self Assembly will enable a sufficiently broad class of useful and COST EFFECTIVE new manufacturing processes ( humph I said that ! ) that it will be significant and permit new unusual devices / products / materials. Sounds like a molecular assembler to me.

By the specific criteria of exponential scalability, one can quite readily admit that most any biological replication (which a virus quite obviously meets the criteria of) will easily scale in a useful manner for production volumes.

Cheers and Good Nite and Good Luck.



Chris there is way too many contradictions,unanswered questions and far fetched ideas in your proposal for MM, in one article you propose that natural resources won't be mined anymore thanks to MM, then in another article you propose oil will be used in MM. Every article I've read says "feedstock" this "feedstock" that, but where does this feedstock come from? and what exactly is the feedstock? I know is "carbon", but carbon doesn't just pop out of the ground.

You make MM seem like this perfect system made by the gods, because everything will be made in perfect percision in less than 5 seconds for pennies, all in around 10-20 years...well automobiles have been around for 90 years and not only do some suck and need constant maintenance(again the car has been around for 90 years) they still use gasoline engines. So pardon me if I have a hard time believing humans will be able to make these perfect tools and machines, not in 20 years but 200 years, if ever.

If you want to make your proposal clear for MM, then write one whole article on it in simple and clear terms and then post it right on the front page, that way people understand your whole proposal.



DT says:

"Every article I've read says "feedstock" this "feedstock" that, but where does this feedstock come from? and what exactly is the feedstock? I know is "carbon", but carbon doesn't just pop out of the ground."

Looking out my window, I see a whole forest of carbon popping out of the ground.

Brian Wang


Just going to address the paragraph where you mention me.

You say I kept indicating it where I mentioned it once in the responses to your stuff.
I will also point out is Mark Wendman-hypocracy again. You tell everyone else to quote your stuff word for word and match up their comments to it. Yet you toss out paraphrases and interpretations of everyone else stuff. You also dismiss the peer reviewed and patent pending work without quoting the specific passages and breaking down your objections point by point. You do not get your work into the form of a clear and cohesive paper, but instead want people to pick through your ramblings word by word. Why don't you organize your stuff into journal format, then submit it for peer review? Then people who are paid to do so can go over it. You do not even site the Cambrios/Belcher work paper by paper or section by section.

You do not give anyone else courtesy but demand more of it from everyone else. Therefore, hypocrit. Yes, I repeated and emphasized that point.

Chris Phoenix, CRN

DT, I wrote an article on nanofactory design/architecture. It grew to about 80 pages. And that does not address mechanosynthesis or bootstrapping! I wish I could write an article covering molecular manufacturing in general, but it'd be about 500 pages long and very technical.

Now that I think of it, a 500-page technical book already exists. K. E. Drexler (1992), _Nanosystems_, Wiley/Interscience. And it's usually ignored by skeptics/critics. And it doesn't have space to answer all the questions. So I don't think it's at all possible to write a short article. You referred my "proposal" singular, but there are lots of different proposals.

To understand the field, I'd recommend starting from a different direction: Study the performance advantages of atomically precise nanoscale machinery, especially stiff dry machines. Then study the various ways in which molecular shapes can be built and molecular machinery can be controlled. But don't start with today's lab results--go to basic theory and generalize. Then start integrating all this, and you'll be able to criticize, extend, and even invent proposals.

The oil/feedstock thing isn't a contradiction, just an imprecision. Oil is one possible source of concentrated carbon. The mineral extraction I was talking about includes fossil fuels used for energy, and also most of the other minerals that we currently mine. We could continue limited mining, or we could extract material from seawater, or...

I haven't been more precise about what the feedstock is because I don't know. It could be acetone, acetylene, methanol, or something else carbon-rich--and that's just for the diamondoid proposals.


Ps. Brian, Mark has shown that he's willing to be civil. And he's taking time to visit with a bunch of crazy futurists; he's meeting us halfway. If he's picky on a point of style, let it slide, or at least don't insult him.

The comments to this entry are closed.