From apm
Jump to: navigation, search
This article defines a novel term (that is hopefully sensibly chosen). The term is introduced to make a concept more concrete and understand its interrelationship with other topics related to atomically precise manufacturing. For details go to the page: Neologism.
Only one of very many possible microcomponent shapes. This is a space filling truncated octahedron.
A cube shaped subproduct block assembled from many microcomponents of about ~1µm size. The microcomponents displayed here all have truncated octahedral shape. Here 24x24x24 microcomponents are depicted. The on this wiki usually used factor of 32 between assembly level sizes would be a bit too big to make out the individual truncated octahedra. The whole block is with its 24µm in size just a bit below the human eye visibility limit which is at about 75µm.

Microcomponents are (re)composable functional units. (Re)composability is very important. Hence the "components" part in the name..
Assembled microcomponents make up gemstone based metamaterials and thus provide the basis for advanced atomically precise products.


  • are mainly built out of standard mass produced diamondoid molecular elements (greater variety than the mass produced crystolecules though)
  • are roughly in the size range from roughly 0.1µm to 5.0µm (see section "Limits to microcomponent sizes" below). Hence the "micro" part in the name.
    2µm will be the reference size here in this wiki.
    Their size constitutes a trade-off between re-usability and space usage efficiency.

Microcomponents are also mentioned on the "assembly levels" page and all over the place on this wiki.

Advantages of microcomponent products vs monolithic ones:

  • potential recyclability (see: Global microcomponent redistribution system)
  • potential repairability
  • product assembly is much more energy efficient when done from recycled parts
  • product assembly is way faster when done from recycled parts (eventually almost instantly by human perception)

Disadvantages of microcomponent products vs monolithic ones:

  • additional interface border constraints in system design
  • minor loss of material strength
  • minor loss of density of functionality

Limits to microcomponent sizes

Lower limit

A microcomponent should be exposable to air.
Just a convention here following from a focus on recycleability.

Even parts as small as a few nanometers in size can already be locked out of vacuum if
all the open bonds are already passivated and/or sealed into the interior.
Though it's a fully passivates crystolecule (or nanocomponent) then.

It's relatively easy to mix in a small amount of quite a bit smaller parts. See: Jumping assembly levels.
Doing too much of this assembly level skipping slows down the assembly process tough.
For size-steps of x32 (the branching factor usually assumed in this wiki)
skipping an assembly level entirely makes assembly 32x32x32 = 32000 times slower.
OUCH!! So this should be used in moderation.

Upper limit

Parts as big as 50 micrometers (=0.05mm) are in many cases still invisible for human eyes.
Bigger will give the products a visible texture. Like today's visible layers in plastic 3D prints.
Likely not desirable, but doable. This blurs into mesocomponents.


In the book "Radical Abundance" Eric Drexler introduces them as microblocks.
Since I want to especially point to the possibility of including functionality and the possibility of recycling and recomposing we'll call them microcomponents in this wiki. "Block" sounds more like a typically passive thing that may not typically be disassemblable and reusable.

Related page: Terminology for parts


Upper microcomponent size limit form choice of early vacuum lockout

In advanced nanofactories the size of a microcomponent is limited by the sizes of the building chambers that are void of any gas molecules (assembly level II).
Well, actually this size limit only applies if (clean but) reactive air is introduced at the soonest possible point (that is assembly level III) after covalent welding with many highly reactive open bonds is all done and finished.

Why breaking this size limit is not necessarily desirable

For the creation of bulk monolithic (non-recyclable) structures a nanofactory must be completely "filled" with vacuum (or noble gas). Not only PPV at the lowest levels. This is likely

  • more challenging as it requires large scale atomically tight gas seals and
  • not necessarily desirable as it promotes throwaway products.

Assuming large scale PPV is actually implemented despite these points:
The larger manipulators at (assembly level III) then too can fuse diamondoid surface interfaces together.
Microcomponents can then be more or less irreversibly joined by covalent welding.
They then devolve down into mere microblocks as they lose their recomposability.

Flawless macroscale single crystals are a challenging edge case of relatively minor practical use

Higher stages of convergent assembly of bulk monolithic crystals (from microblocks) may need to be avoided because
self alignment becomes more difficult without any aid of self centering.
See Nanosystems Fig 14.1. for a possible approach.

Design constraints on microcomponents from desired in air microcomponent recomposability

Since it can be desirable to operate microcomponents in a non vacuum environment (separation of assembly levels) and
since one should want to be able to recycle them, microcomponents

  • should have no exposed open bonds ( = chemical radicals) on their external surfaces
  • should preferably use reversible locking mechanisms
  • should be meaningfully tagged

Shapes of microcomponents

In the simplest case one could use a simple cube as delimiting base shape.
Stacking them then forms a simple cubic microcomponent crystal.
To get less anisotropic behavior of metamaterials one can make them have the shape of either of:

  • truncated octahedrons [1] [2] (the Wigner-Seitz cell of the body centered cubic system bcc);
    these preserve parts of the cubes <100> surface planes and expose much of the <111> octahedral planes which are
    conveniently normal to diamond bonds (when standard orientation is choosen for the majority of the internal crystalline components).
    Completely flat surfaces for Van der Waals bonding can be used since partly finished assemblies always have
    (in contrast to partly finished assemblies of simple cubes) dents that prevent side-ward sliding.
  • rhombic dodecaherdons (the Wigner-Seitz cell of the face centered cubic system fcc)
  • trapezo-rhombic dodecaherdons (the Wigner-Seitz cell of the hexagonal closed packed system hcp)

Base cells of more complicated crystal structures or even quasi-crystals will make geometric reasoning exceedingly hard and
will therefore probably only be considered if needed for a good reason. Some examples:

  • tetrahedrons and octahedrons ("geomag-spacefill" which is the octet truss)
  • space fills not derived from crystal structure base cells
  • space fills derived from crystal structure base cells

Deriving shapes for microcomponents from more complex crystal structures:
With the voroni cells around atoms in simple crystal structures of especial interest
one also gets space-filling sets of polyhedra that can be used at the much larger scale of microcomponents.
Spacefilling sets of polyhedra from quasicrystals may lead to especially desirable more isotropic mechanical properties.
Examples for crystal structure derived shapes:

  • Weaire–Phelan structure - structure with the least surface area (yet unproven)
  • base cells for quasicrystals with 5,7,9,10,11,... fold rotation symmetries; Symmetric rods with a single global rotation axis can be built.
    Spacefills can be generated either by straightforward projection from higher dimensional space by subdivision rules or by potentially difficult puzzling.
    They may have interesting mechanical properties.

Further shapes of practical interest are:

  • tubbing segments (like in tunnel construction work) for curved parts of several µm size
  • adapter cells from one space-fill to another
  • partly rounded cells for outer shells
  • special shapes required for a mechanical metamaterial function (interlinking slide-rolling platelets)

  • and many many more ...

Transitions between the most basic polyhedra (related to bcc, fcc, and hcp crystal structures):

Transitions from rhombic dodecahedra (fcc WS-cell) to trapezo-rhombic dodecahedra (hcp WS-cell) and vice-versa are trivial in the natural directions.
As with hcp one gains a single special z axis one looses symmetry and one needs to pick a specific direction for that z-axis direction in that transition.
Transitioning sideways to that z-axis will need adapter microcomponents. Unclear shape.
(wiki-TODO: to analyze if fcp hcp sideways connecting adapters are possible and what shape the would need to take).

Transitions from rhombic dodecahedra (fcc WS-cell) to trunctated octahedra are nontrivial. One needs to make a step in scale for that transition.
Also the corners of the rhombic dodecahedra become the centers of the smaller trunctated octahedra. So adapter pieces are definitely necessary.
(wiki-TODO: add the OpenSCAD screen-cap showing the relation between the fcc and bcc polyhedra)

Special sets of shapes:

Flat surfaces would be critically important for things like larger scale bearings like e.g. in infinitesimal bearings.
Generally it's desirable to start out with most natural shapes (like the ones described in the preceding section)
or shapes best matching a problem as a starting point an then add the munimum number of adapter part shapes and surface block shapes possible.

One possible small set of bodies that fill space and can create flat surfaces
is presented here by George W. Hart [3]

A word on the octet truss:
The rhombic dodecahedron shape is special in having all its faces normal to the struts of the octet truss (110 direction).
For details see page: Octet truss

sub-structure within microcomponents; the interior of microcomponents

Monolithic internals

The inside of a microcomponent ca be a more or less monolithic diamondoid gemstone based machine structure that is

Tradeoffs for finer grained reversibility in assembly

Sometimes even a lot of the sub-structure of micro-components may be quite reversibly assembled.
Think: Many fully passivated crystolecules that are not fused together but can be taken apart again.
This poses more of an overhead than reversibility at larger size scales as
nanoscale passivation layers take up a significant fraction of part dimenssions.
So this is not to be expected in ultra high performance applications like say e.g. moissanite based rocket engines or such.

In other words: Internal reversible joints are possible but may waste too much space in some high performance applications.
Simplemost use of Van der Waals joints (flat surface contact sticking) waste less space than more complicated interlocks, but one must take care.
The internal structure shouldn't be weaker bonded than the bonds between the microcomponents.
Accidental or intentional breaking of structures could then create a big mess of spilled internal parts.
See: Spill prevention guideline.

Encapsulating Van der Waals assemblies by shape locking seems to be a good choice if applicable.
Like with (sorts/types/letters) in a book printing press a mill style robotic mechanosyntesis cores could be assembled with a reversibly matter-hardcoded program.

super microcomponent structures

Some systems stretch over many microcomponents and thus can't be counted to the diamondoid metamaterials as a whole. They are makro-heterogenous.
Diamondoid heat pump systems are one example, nanofactories another.
Nuclear technology systems are all inherently macroscale systems (unless entirely surprising scientific discoveries crop up, soft SciFi speculations we do not want to entertain here).

Systems for macroscale active align-and-fuse connectors (operable by human hands, direct manipulation) will be pretty big and probably implemented around the size range of microcomponents and located at distances perceivable by human vision.


Microcomponent threshold

In the stack of convergent assembly the size of microcomponents is the point below which things change. Stages become less similar to each other and there's a change from freely programmable general purpose assembly to hard-coded factory style conveyor belt assembly.

External links:

  • Customizable Crystallographic Building Block [4]
  • One possible small set of bodies that fill space and can create flat surfaces by George W. Hart [5]


  • Close packing fcc and hcp do close packing but bcc does not
    fcp & hcp is easy to mix (across a common defined z-axis) in fact in natural moissanite crystal structures do form such a mix.
    Neither ABAB hcp hexagonal nor ABCABC bcc cubic but a mix.
    In diamond that would mean a mix with lonsdaleite but this does not occur naturally.