Surface molecular system

Basically surfaces with molecules atop. These are higly relevant for early mechanosynthesis and force applying mechanosynthesis along the direct path.

Some major of 3D covalent bulk material classes are:

• Metallic: typically relatively noble metals aka "coinage metals" are used (Au, Ag, Cu, Pt, Ir, …)
  reactive and/or low melting materials are typically avoided
• Semiconducting: Often III-V or II-VI semiconductors with diamondoid structure (cubic zinkblende structure or hexagonal wurtzite structure)
• big class of nonmetal gems like e.g. rutile
• salts or simple cubic oxides

2D covalent layered materials include:

• single crystalline graphite aka HOPG (highly ordered pyrolytic graphite)
• sp2 hexagonal boron nitride (isostructural to HOPG)
• MoS₂ and more …

The latter class often can be deposited as monolayer or few layers
on bulk materials fully pr partly covering the bulk below.

Garden of molecules

Due to its easy handling in cleaning and preparation,
quite large fraction of experimental research had been directed towards gold as a bulk base surface.
One particular 2015 paper by Leo gross et.al. (see references) shows
a larger number of at deep cryo temperatures co-deposited molecular species,
that at higher temperatures would be very volatile, diffusing to agglomerates or even desorbing.

Graphene nanoribbons (GNRs)

As time of writing (early 2026) there has been significant progress in
synthesis and subatomic resolution imaging of flat nanographene covalent bonding structures
in particular including graphene nanoribbons with partial positional atomic precision. Pure and doped.

The key technology making this possible on the imaging side was:
Stiff cantilever AFM with pickup of a CO molecule for imaging.
Scanning tunneling microscopy (STM) images electronic state orbitals that are typically larger and
rarely match the mechanical bonding structure in detail.

Synthesis falls apart in …
– conventional chemistry pre-synthesis and
– on surface synthesis with some aspects of 2D self assembly (self folding and self finding).

Barely any work on less flat structures - thus big focus on GNRs

Stiff cantilever AFM has exceptionally high height sensitivity.
This is a double edged sword.
– While in makes for awesome contrast
– it also makes for an extremely limited range of imageable depth.

Crude optical analogy would be "depth of field" but here it is not things getting blurry but
– either too far away and nothing shows up at
– or too close and the tip structure gets damaged (possibly irreversibly)

Stiff cantilever AFM:

• Imaging sp2 alkane tydrocarbon chains has been achived (smapped between pieces of graphene. (wiki-TODO: add reference to relevant paper)
• Imaging DNA did not yield good structure depicting results. (wiki-TODO: add reference to relevant paper)
• Imaging of nanotubes has been demonstrated but the results are likely deceptive. (wiki-TODO: add reference to relevant paper)
• Imaging of polyadamantanes hads been achieved (wiki-TODO: add reference to relevant paper)

Imaging of polyadamantanes may serve as one of the strongest indications that
mechnosyntehsi high up in 3D might eventually actually work
even with macsoscale scanning probe microscopy systems.

GNR synthesis: On-surface-cyclodehydrogenation (and dehalogenatio)

– Liquid phase pre syntehsis of aromatic ring rich hydrogacbon chains. Then purification.
– Deposition onto typically gold samples
– Controlled hearing profile for the rings to desorb as H2 molecules hydrogens and join up to larger nanographitic structures.
– Sometimes halogens are included that desorb earlier (as Cl2 Br2 I2) to some degree controlledly. Giving more options.

= GNRs under termination control i.e. graphene 2D crystolecules

Most graphene nanoribbons are only internally positionallly atomic precise.
Since they have no termination condition in the lengthwise direction they overall
can't be considered positionally atomic precise. They are statistic in their length distribution.

There have been attempts go get fully atomically precise structures.
– either by just making small flakes instead of non-terminating nanoribbons
– or by step-by-step iterative pre-sysntehsis (wiki-TODO: Add reference to relevant pager examples.)
The latter may still have small finite number of possible outcomes due to binary rotation options (ignoring damage error states here).
If the desired ones can picked out in post processing then this can be considered atomically precise.

There have been attempts on …
– dragging small (more or less atomically precise) graphene flakes controlledly around
– fusing small flakes edge-to-egde by halogen split-off
– tip-based local on surface cyclodehydrogenation (away from the gold surface indicating abence of dependence on caralysis)
(wiki-TODO: Add reference to relevant pager examples for these.)

External links

Wikipedia:

• Single-layer materials
• Silicene
• MXenes
• Graphene nanoribbons
• Graphene

---

Gross, Leo & Schuler, Bruno & Mohn, Fabian & Moll, Nikolaj & Repp, Jascha & Meyer, Gerhard. (2015). Atomic Resolution on Molecules with Functionalized Tips. NanoScience and Technology. 97. 223-246. 10.1007/978-3-319-15588-3_12.

https://link.springer.com/chapter/10.1007/978-3-319-15588-3_12 on reseachgate Fig12 on researchgate