Proof of Picking and Placing Carbon Dimer Molecules

Molecular Nanotechnology has worked for over a decade. I, Brian Wang, invested in the startup that has been moving and placing molecules using tooltips for over a decade. Over 180 successful atomically precise movements of carbon dimers are described in a paper.

When some of the debates about the feasibility of molecular nanotechnology were ongoing, I knew that a few years later there was progress being made.

Jeremy Barton had described some of the work publicly. Ralph Merkle and Robert Freitas have led the work towards diamondoid MNT. CBN Nano Technologies (aka Nanofactory Collaboration) is the funded company doing a lot of the work and they are publishing papers. Jeremy worked for 15 years at CNB Nano. Jeremy founded the Nano Dynamics Institute, a new nonprofit/public charity (501c3) to pursue molecular nanotechnology openly and interdisciplinarily. The focus is systems engineering for nanomachines (design, construction, assembly, actuation, and control), remaining system-agnostic. It aims to combine approaches like protein engineering, MEMS/NEMS, positional chemistry, DNA/RNA nanotechnology, and nanoelectronics (e.g., hybrid systems using single-electron transistors or atomically precise qubits for control/sensing). Goals include building tech trees, open-sourcing methods, and enabling community/market-driven progress toward functional nanomachines and eventual molecular manufacturing. He views MNT as “the longest lever anybody has ever thought of to move the world” and the physical basis of the singularity.

Lead/corresponding authors and key contributors for one of the key papers is Megan Cowie (first author) and a large team from CBN Nano Technologies, Inc. (CBNNT, Ottawa), including Robert A. Freitas Jr. and Ralph C. Merkle (prominent theorists and proponents of diamondoid MNT and nanofactories), plus many experimentalists (e.g., Chris Deimert, Ryan Groome, etc.).CBNNT is the company executing the experimental work tied to the long-running Nanofactory Collaboration (originally involving Freitas, Merkle, and others. It evolved from Nanofactory Corporation, with patents on mechanosynthesis systems, tips, build sequences, and related mechanical computing assigned to CBNNT or predecessors).

What the Paper Demonstrates

Using inverted-mode scanning tunneling microscopy (IM-STM) at 4 K on a hydrogen-passivated Si(100) surface.

Custom molecular tools (EAOGe-C₂I: a Ge-substituted adamantane derivative with a C₂ group and iodine cap) serve dual roles as imaging probes and reagents. They anchor upright via legs, allowing precise positioning.

Dangling bonds (DBs) are patterned on the Si surface via bias pulsing (creating reactive sites in inter-row or other configurations).

The tool is de-iodinated (exposing a C₂ radical), precisely aligned over the DB site, and approached/retracted in controlled depth-sampling cycles. Mechanical force drives the reaction: initial Si–C bond formation, followed by Ge–C cleavage that transfers the C₂ unit to the surface, forming a stable IR-C₂ (inter-row carbon dimer).
This is repeated for multi-site patterning (e.g., multiple IR-C₂ units, including complex shapes like an ‘X’ with 9 C₂ / 18 carbon atoms).

Extension via C–C bond formation has an existing surface-bound IR-C₂ reacts with another de-iodinated tool to form longer chains like IR-C₄ (a novel polyyne-like structure), confirmed as the lowest-energy configuration by over 2 eV in simulations.

Key results

High reproducibility and yields: ~93% for single IR-C₂ (184/197 attempts), 97% for adjacent pairs, 92% for extension to C₄, 84% for more complex patterns.

Very low off-target reactions (~3% H-abstraction, rare Si abstraction).

Structures are stable for days/weeks under imaging conditions.

Mechanisms validated by STM imaging + simulations (STM simulations match experiment; QM/MM and DFT show downhill energy landscapes driven by mechanical positioning/work, with geometry favoring desired bonds/cleavage over alternatives).

Simultaneous spatial (precise site selection via DB pre-patterning and tool positioning) and chemical control (specific C₂ donation and C–C extension).

The ability to build atomically precise structures on surfaces with complete control over both atomic placement and chemical bonding remains a central challenge in nanoscale fabrication. There is published proof demonstrating simultaneous spatial and chemical control over the mechanosynthetic fabrication of carbon structures. Using inverted-mode STM, C2 units are donated from surface-deposited molecules to pre-patterned reactive sites on a hydrogen-passivated Si(100) surface. They demonstrate single-site C2 donation, spatially patterned multi-site C2 donation, and the stepwise assembly of polyyne structures through successive C-C bond formation. Together, these results establish controlled mechanosynthetic donation as a foundational capability for programmable atomically precise fabrication.

What Was Previously Claimed Impossible But Now Has Experimental Evidence

Nobel Prize winner in Chemistry Richard Smalley in the early 2000s Drexler–Smalley debate) argued that precise positional molecular assembly (“mechanosynthesis”) was fundamentally impossible due to

“Fat fingers” problem – Tools too bulky to position reactive groups precisely without steric interference.

“Sticky fingers” problem – Difficulty attaching feedstock and releasing it cleanly without unwanted bonding or contamination.

There is an updated 540 page paper on the roadmap to molecular nanotechnology.

Roadmap Structure and “Workstation” Concept

The plan has 11 phases progressing from validation to full production:Early phases (I–II): Macroscale Fabrication Workstation (MFW) and Macroscale Manufacturing Workstation (MMW) — Use scanning probe microscopy (SPM)-based systems (e.g., multi-tip picopositioners with 3–5 degrees of freedom) in UHV/cryogenic conditions to validate and scale DMS reactions. Milestones include demonstrating core reactions (H abstraction/donation, C₂ placement), fabricating simple hydrocarbon structures (hundreds to tens of thousands of atoms), and assembling nanoparts.

Mid phases (III–VI): Transition to nanoscale workstations — First-generation Nanoscale Fabrication Workstation (NFW), Nanoscale Assembly Workstation (NAW), and Nanoscale Manufacturing Workstation (NMW). These use Stewart-platform-style manipulators on multi-electrode surfaces, with bulk feedstock systems. Focus shifts to self-contained nanoscale operation and replication.

Later phases (VII–XI): Exponential scaling and full nanofactories — Plate-to-plate replication, massively parallel arrays (10¹⁴+ workstations), micro-assembler arrays, and first-/second-generation nanofactories (FGN/SGN) capable of macroscale products at useful throughput (0.01 kg/hour or more).

What Has Been Done or Shown Publicly (as of mid-2026)

Public progress aligns most closely with early roadmap phases (especially Phase I validation of DMS reactions using positional control).

Major public advances

Theoretical and design work — Extensively documented and updated. This includes tooltip designs, reaction sequences, build paths, mechanical computing systems, and component designs (the 2025 mFET — molecular field-effect transistor with ~7,694 atoms, IMM Report 56). Patents cover mechanosynthesis systems, tips, and sequences (many by the CBN Nano team including Merkle and Freitas).

Experimental validation of core DMS primitives (biggest recent step, matching Phase I) — The May 2026 arXiv paper (2605.27250) by the CBN Nano Technologies team (co-authored by Robert A. Freitas Jr., Ralph C. Merkle, Jeremy Barton, and many others) demonstrates controlled mechanosynthetic C₂ donation and extension on a hydrogen-passivated Si(100) surface using inverted-mode STM at 4 K.