Mechanosynthetic Donation & Abstraction

• Mechanosynthetic donation and abstraction are atomically precise transformations driven by mechanical forces via IM-STM and tailored molecular tools.
• These processes enable programmable addition and removal of atoms on Si(100) with sub-ångström accuracy under ultra-high vacuum conditions.
• Optimized tool designs, such as EAOGe–C₂I and MAOC–C₂I, enhance selectivity and yield, paving the way for scalable, atom-by-atom fabrication.

Mechanosynthetic donation and abstraction are positionally controlled, mechanically driven atomic-scale transformations mediated by molecular tools and implemented using inverted-mode scanning tunneling microscopy (IM-STM). These processes enable the programmable addition (donation) and removal (abstraction) of individual atoms or small moieties—most notably carbon and silicon—directly onto or from atomically clean surfaces such as Si(100). The combination of deterministic site delivery, force-driven bond activation, and modular molecular tool design underpins a bottom-up approach to atomically precise fabrication, achieving sub-ångström spatial accuracy and high yields in ultra-high vacuum (UHV) at cryogenic temperatures (Cowie et al., 26 May 2026, Blue et al., 11 Jun 2026).

1. IM-STM Apparatus and Mechanistic Overview

Mechanosynthetic operations are performed in a commercial UHV STM configured in an inverted mode. The probe side consists of a silicon probe chip (SPC) with a nanoflat Si(100) mesa serving as the build site. The sample side is a Si(100) wafer sparsely functionalized with tripodal molecular tools, immobilized to minimize off-target motions. Radical feedstocks (e.g. EAOGe–C₂* adducts) are generated in situ via controlled deiodination and then engaged by precise vertical and lateral positioning of the SPC. Mechanosynthetic sequences are initiated with feedback-off vertical approaches (Δz up to 300 pm) and followed by programmed retraction, leveraging direct mechanical work to traverse activation barriers and drive specific covalent bond rearrangements (Blue et al., 11 Jun 2026).

2. Molecular Tools for Donation and Abstraction

Mechanosynthetic molecular tools use tripodal adamantane scaffolds, engineered for controlled presentation of the reactive feedstock:

• EAOGe–C₂I: Ethyl-anchored, Ge-bridgehead tool, delivers C₂ moieties through Ge–C bond homolysis upon approach. Iodine capping serves as a radical precursor. After remote deiodination at >3 V, imaging and reflected-probe signatures confirm radical state and spatial registration.
• MAOC–C₂I: Methylalcohol-anchored, C-bridgehead tool optimized for abstraction, featuring higher feedstock bond energies and minimized surface charges (Blue et al., 11 Jun 2026).

The choice of bridgehead atom (Ge vs. C), leg length, and feedstock protection directly influences the selectivity of donation vs. abstraction, the force profiles, and the error rate. The following table summarizes key attributes:

Tool	Bridgehead Atom	Feedstock BDE (eV)	Process	Selectivity/Yield
EAOGe–C₂I	Ge	5.213 (Ge–C)	Donation/Abstraction	54% (donation), 40–60% (abstraction)
MAOC–C₂I	C	5.488 (C–C)	Abstraction (optimized)	100% (abstraction @ 4/77 K)

Feedstock protection through I-capping enables clean radical generation, tripodal scaffolding enforces orientation, and steric shielding with bulk core elements restricts off-pathway reactivity (Blue et al., 11 Jun 2026).

3. Mechanisms and Energy Landscapes

3.1. Donation Pathway

The canonical process for carbon donation to H:Si(100) involves:

1. A de-iodinated EAOGe–C₂* tool positioned such that its distal carbon is directly above silicon dangling bonds at a reactive IR site.
2. Gradual approach until z = z₀ (onset of Si–C bonding), detected via a sudden drop in potential energy and imaging transition.
3. Continued retraction (Δz increase) builds tensile force across the newly formed Si–C bond and the Ge–C bond in the tool.
4. At Δz ≈ (z–z₀)₁ (typically ~+50 pm retraction), Ge–C bond homolysis transfers the C₂ fragment to the surface as a surface-bound radical.
5. Further retraction allows the C₂ fragment to relax to its thermodynamic minimum as an IR–C₂ unit (Cowie et al., 26 May 2026).

Energy and force characteristics are defined by bond dissociation energies and force–distance curves:

$$D_0(\mathrm{Ge}\!-\!C)\approx5.20~\mathrm{eV}, \quad D_0(\mathrm{Si}\!-\!C)\approx5.14~\mathrm{eV}, \quad F_{\rm th}\approx 1.2\text{–}1.6~\mathrm{nN}$$

MEP barriers are purely surmounted by mechanical displacement; thermal activation is not required. The accompanying potential energy surface E(Δz) displays net exothermic behavior, with the formation of the Si–C bond leading to a plateau drop, Ge–C dissociation corresponding to a tensile-force peak, and relaxation into the product state being barrierless (Cowie et al., 26 May 2026).

3.2. Abstraction Pathway

Abstraction is the formal reverse. The tip is approached toward a surface-bound C₂ (or longer chain), forming a new C–tip bond, then further displacement cleaves the surface Si–C bond. After retraction, the carbon fragment remains on the tip. The force and energy barriers (Si–C cleavage ≈5.1 eV) parallel those of donation, but with the energy profile initially uphill before bond transfer (Cowie et al., 26 May 2026, Blue et al., 11 Jun 2026).

The abstraction channel is further enhanced by optimized tools (e.g., MAOC–C₂I) achieving perfect yield at cryogenic temperatures, attributed to the higher C–C bridgehead bond energy, increased rigidity, and improved electrical stability.

4. Force–Distance Profiles and Activation

The mechanosynthetic transition states and yield rates are determined by the vertical force profile F(Δz). The force is obtained via:

$$F(\Delta z) = -\frac{\,\mathrm{d}E}{\mathrm{d}\Delta z}$$

For the approach regime (Δz < 0):

$$F(\Delta z)\approx A\,\exp(-\alpha\,|\Delta z|),\quad A\approx1.0~\mathrm{nN},\, \alpha\approx8~\mathrm{nm}^{-1}$$

The maximum tensile force (~1.5 nN for Ge–C dissociation) coincides with the critical point for bond rupture. The area under F(Δz) represents the work performed by the STM:

$W = \int_{z_0}^{z_1} F(z)~\mathrm{d}z$

Effective junction stiffness and vertical displacement parameters enable tuning of the force regime into the desired nanonewton regime. Lateral positioning tolerance for abstraction is more stringent than for donation due to the necessity of targeting a specific C atom of the adsorbate (Cowie et al., 26 May 2026).

5. Experimental Performance and Error Analysis

5.1. Yields and Error Rates

Performance parameters depend on both tool choice and STM trajectory. Key results include:

• Carbon donation: 107/197 attempts successful with EAOGe–C₂I (54 ± 7% yield), >90% site yield with optimized spatial registration and DB landmarks; error sources include lateral misalignment, tool degradation, and off-target H-abstraction (~3%) or Si-atom abstraction (0.3%).
• Abstraction: ~40–60% for EAOGe–C₂I depending on site (inter- vs. on-dimer visits); 100% for MAOC–C₂I across tested temperatures.
• Multi-step sequences: up to nine sequential IR–C₂ units donated (>80% yield in complex patterns); up to ten abstractions with a single MAOC–C₂I tool (Blue et al., 11 Jun 2026).

5.2. Mode of Control

Mechanosynthesis is driven by tip displacement and mechanical work, not by substrate temperature or bulk kinetics. Thermodynamic stability of IR–C₂ and IR–C₄ motifs ensures nanostructure persistence for days at 4 K. Error correction via abstraction tools (e.g., MAOC–C₂I) becomes feasible, enabling feedback-corrected fabrication cycles (Cowie et al., 26 May 2026, Blue et al., 11 Jun 2026).

6. Tool Design, Selectivity, and Future Prospects

Effective mechanosynthetic tool design must optimize bond dissociation enthalpy matching (e.g., tuning Ge–C vs. Si–C), radical protection (I-capping), mechanical rigidity (tripodal anchoring; leg length), steric shielding, and electronic stability (minimized charging, HOMO/LUMO engineering). The ability to steer the tool approach vector in multidimensional space is projected to further enhance selectivity, allowing off-vertical bond activation trajectories and extension beyond simple feedstocks (Blue et al., 11 Jun 2026).

Challenges remain with respect to small BDE differentials (e.g., 5.213 vs. 5.184 eV), surface dimer dynamics, multiplexing constraints, tool degradation, and throughput bottlenecks from RPI overlap. Prospective developments include expansion to room-temperature operation, in situ potential mapping, non-contact force characterization, and error correction protocols enabling atom-by-atom lithography (Blue et al., 11 Jun 2026).

7. Significance for Atomically Precise Fabrication

Mechanosynthetic donation and abstraction provide the experimental foundation for programmable, atomically precise construction and deconstruction of covalent nanostructures on silicon. High per-step yield, robust thermodynamic anchoring of products, and spatial registration against surface DB landmarks establish a reproducible path toward bottom-up fabrication of increasingly complex architectures. The modularity of molecular tool design and feedback-enabled error correction suggest a scalable trajectory toward atom-by-atom assembly, with implications for device miniaturization, quantum information platforms, and heterogeneous material integration (Cowie et al., 26 May 2026, Blue et al., 11 Jun 2026).