Toward Advanced Nanotechnology:

Nanomaterials (2)

by Eric Drexler on 2009/01/16

Head of a bull with a prominent horn, image fragment image (cc) Redskynight, Flickr
Beef or Horn?

As every mechanical engineer knows, the stiffness of a material — its elastic modulus — is often a critical property; likewise in nanomechanical engineering, though in part for a different reason. I’d like to say a few words about this, then discuss some materials of interest in implementing nanosystems. And there is something I must say about beef, too.

Stiffness matters

Stiffness determines how much a component will deform in response to a force, or equivalently, how much elastic energy will be stored in the process of distorting a component into a given shape. In nanomechanical engineering, the energy is of special importance, because the probability of a distortion declines exponentially with increasing elastic energy — in other words, stiffness limits the amplitude of thermal fluctuations. (This post was getting too long, so I moved further discussion of this topic to another page: Elastic Modulii, Stiffness, and Thermal Fluctuations.

The chart below compares the (Young’s) modulus of some very different materials.

Young's modulus of selected materials
Young’s Modulus (in GPa) of Selected Materials
(various sources)

Toward advanced nanosystems

These materials differ in stiffness by orders of magnitude, and this, together with related facts, will be important in charting paths toward advanced nanosystems. A few observations:

  • The range of modulus values for structural proteins is like that of engineering polymers: Some are like rubber; many are like epoxy, polystyrene, or polycarbonate; and silks can rival Kevlar.
  • DNA, although valuable as a medium for design and fabrication of large, atomically precise structures, is much softer than structural proteins.
  • In comparison to a typical structural protein, the stiffness of pyrite (FeS2, also known as fool’s gold), rivals that of diamond. A number of metal oxides have elastic moduli that are somewhat greater than pyrite.
  • Finally, except for steel and diamond, all these materials are readily synthesized at low temperatures from substances dissolved in water. The same is true of several metals and semiconductors, and of many silicates and metal oxides.

To me, these facts suggest several conclusions:

  • DNA will serve better as a means for organizing components than as a means for establishing tight control on geometrical relationships.
  • The greater stiffness of structural proteins is yet another property that complements the properties of DNA, suggesting that the two materials can play quite different mechanical roles in biomolecular composite nanosystems.
  • The ability to form complex, atomically precise structures of pyrite or of various metal oxides would be an enormous step forward in engineering stiff nanomechanical structures and machines.
  • The feasibility of water-based synthesis suggests that advances in the design of biomolecular composite nanosystems can open relatively direct paths toward this ability.

A multitude of organisms use crystal-binding proteins to direct the growth of stiff, inorganic, crystalline materials, and the principles of this process are being adapted for applications in nanotechnology. To quote from the abstract of a 2007 paper:

Nature has long used peptide- and protein-based manufacturing to create structures whose remarkable mechanical, transport, optical, and even magnetic properties are determined by a fine control of composition and architecture extending from the nanoscale to the macroscale…. In recent years, portable amino acid sequences selected from combinatorial libraries and supporting the assembly, nucleation, and geometrical organization of solid phases have emerged as attractive tools for bionanofabrication.

The solid phases include metals, semiconductors, and oxides.

Protein is not like meat

Head of a bull with a prominent horn, image fragment image (cc) Redskynight, Flickr
Horn, not beef!

When people think of protein, there’s a temptation to think of beef — to imagine a soft, tender material, and perhaps to salivate. This impression gives structural protein engineering a faintly ridiculous air.

When you imagine a protein structure, think not of meat, but of horn, a weapons-grade biomaterial made of keratin.

As the expanded graph below shows, confusing structural proteins with meat is is not just wrong, but enormously wrong — wrong not by a factor of ten, or a thousand, but by a factor of more than a million.

Young's modulus of selected materials, including meat
Muscle (“meat”) is softer than structural proteins
by a factor of more than a million

Muscle data from: “Young’s Modulus Measurements of Soft Tissues
with Application to Elasticity Imaging”
EJ Chen et al. IEEE Trans. Ultrason., Ferroelect., Freq. Contr.,
43(6), 191–194 (1996). [pdf]

It’s time to free the idea of protein engineering from this conceptual contamination.

Title updated 10 Feb 09

See also:

anon January 16, 2009 at 12:57 pm UTC

the links to the stiffness charts are missing a leading / I think

Phillip Huggan January 16, 2009 at 6:15 pm UTC

One of the earliest replicator metabolisms I can think of needs cheap plastic electronics and a LIMITED DURATION plastic power source (sail, solar, methane) as a precursor. WallE robots made of plastic bags could scavenge dumps.

Phillip Huggan January 16, 2009 at 6:20 pm UTC

Protein is soft enough that it can’t be used as a diamond AFM tool tip. I wonder if it can be the substrate of an UHV chamber’s walls; if can withstand vacuum pressure.

Chris Hibbert January 16, 2009 at 6:38 pm UTC

The charts are here and here.

Eric Drexler January 16, 2009 at 7:11 pm UTC

@ Chris and Anon: Thanks for the heads-up; I found the problem and fixed it. Please tell me if you see any more glitches like this.

Michael G.R. January 17, 2009 at 4:38 am UTC

I love these simple but insightful posts.

Are bull horns, f.ex., as tough as keratin can be, or would a pure man-made version be even harder?

Evil Rocks January 17, 2009 at 7:26 pm UTC

Ha, it’s great to hear someone in the field touting the possibilities of protein construction. I ran across a great paper on using dissolved wood pulp to make insanely strong paper-like material the other day:

I can’t imagine that this is particular news, but your readers might dig the link.

Eric Drexler January 17, 2009 at 9:20 pm UTC

Thanks for the link to the paper paper. Beyond just strength, their cellulose material is has high toughness, another engineering property of great value, though mostly for large structures under high stress. Many ceramics are stronger than steel, but they’re brittle, meaning that they absorb little energy in propagating a crack. It’s the fracture energy that defines ”toughness”, and really tough ceramics don’t exist. It’s basically the same story for materials that are easy vs. hard to tear.

Precise control of structure at the atomic level will enable the engineering of extremely tough materials from fibers of substances that are strong, but brittle at a macroscale. The principle is the same as that used to make (for example) glass fiber composites tough: Pulling fibers apart can absorb a lot of energy, letting the material stretch a bit without breaking, and bringing cracks to a halt.

With precisely tailored interfaces, the fibers can be made to start separating at just below their breaking stress, then lock up again before they pull apart entirely — thereby forcing other fibers, further out, to start sliding. What would otherwise be a crack would instead become an expanding zone of stretching, like that in a very ductile material. The resulting combination of high strength and high crack resistance would provide what engineers most want in making light-weight, reliable structures. Aerospace structures about 1/50 the mass of today’s aluminum should be quite feasible.

As is often the case, much of the advantage of really precise construction can be gained by more conventional means. How large a portion of the underlying physical potential can be achieved this way remains to be seen, but carbon nanotube based materials promise impressive results.

John Thompson January 18, 2009 at 9:08 pm UTC

Sounds like proteomics has quite a bit to do with nanotech. Pretty fascinating stuff.

I would like to know if molecular biology or proteomics will contribute more to the material understanding of our nanotech cook book.

What should we study to make the tools that make the things? I’m too dumb to put it all together. . . .

anon2 January 19, 2009 at 9:43 pm UTC
Erin January 25, 2009 at 1:58 am UTC

Very fascinating information. Question for you Eric Drexler and others: In Nanosystems and other nanotech resources, “Highly Crosslinked Polymers” are mentioned. If we were to hold a macroscopic piece of such material in our hands, say atomically precise, would it have a combination of hardness and toughness that would basically be a real life version of the famous science fiction material known as “plasteel” which is said to be hard or harder than steel, and flexible and light as plastic?

What other materials would fit such a role? Carbon nanotube based plastics? Diamondoid polymers?

Richard February 6, 2009 at 11:04 am UTC

Hi Eric,
I just came about this blog entry, which is related to some discussion posted here. You might be interested in taking a look:

Eric Drexler February 9, 2009 at 10:08 pm UTC

Thanks. The MIT paper on protein engineering looks interesting, and I’ll have a look at it.

amauri February 10, 2009 at 7:56 am UTC

Estou concluindo o curso de física
em Uberlândia-MG , BRASIL
e achei o material postado por você

{ 12 trackbacks }

Previous post:

Next post: