MNT Weekly Delta — 2026-06-14

Molecular nanotechnology · lookback window Dec 2025 – 14 Jun 2026 · technology-intelligence delta

First run. No nano_seen_items.md existed in the working folder, so this is the initial MNT briefing — every item below is reported for the first time and seeded into memory. Subsequent runs will be true deltas. Evidence tiers: Demonstrated peer-reviewed/replicated · Reported preprint/single-source/vendor · Projected roadmap.

1 · Summary

The window's headline is additive mechanosynthesis. CBN Nano Technologies posted a preprint claiming positionally controlled C–C bond formation — donating carbon dimers and building short polyyne chains on hydrogenated Si(100) by inverted-mode STM. If replicated, it is the field's most significant APM result in decades. Reported, single group, not yet peer-reviewed — and accompanied by visible promotional activity, so treat as major-if-true.
De novo protein design was the strongest sub-area, with three Nature papers from the Baker lab / IPD: one- and two-component quasisymmetric protein nanocages, and miniprotein switches targeting GPCRs (with in vivo data). Demonstrated.
A new spinout, Skape Bio, launched to commercialize the IPD GPCR-miniprotein platform — the window's clearest commercial signal alongside CBN.
DNA nanotech delivered solid "manufacturing science": design rules for faster, higher-yield origami folding (UT Austin) and a computational framework for micron-scale crisscross megastructures.
Small-molecule synthetic molecular machines were quiet in-window; the strongest item (a motor directing rotaxane synthesis, Kathan lab) landed just before the window in Nov 2025.

2 · New technical developments

Atomically precise manufacturing & mechanosynthesis

Atomically precise mechanosynthesis of carbon structures on H–Si(100) by inverted-mode STM Reported — arXiv preprint, ~26 May 2026; CBN Nano Technologies (Freitas, Merkle, Allis et al.). Reports single-site C₂ donation, spatially patterned multi-site donation, and stepwise polyyne-chain assembly — i.e. adding atomically precise structure rather than only subtractive H-abstraction. Hype check: what was shown is C₂ donation and short polyynes under UHV; no productive nanosystem or assembler. Single-group, non-peer-reviewed, with a same-day Wikipedia edit by a co-author — promotion is running ahead of review. Independent replication is the open question.

De novo protein design & engineered protein nanomachines

One- and two-component quasisymmetric protein nanocages Demonstrated — two Nature papers, ~May 2026; Baker lab / IPD (Lee, Wang, Veesler et al.). Engineers viral-capsid quasisymmetry from scratch — programming local curvature so pentagonal defects emerge in a hexagonal lattice — to build larger, tunable cages for genetic-medicine cargo. A step toward designed self-assembling nanomachinery, not just static scaffolds.
De novo miniproteins targeting GPCRs Demonstrated — Nature, 21 May 2026 (DOI 10.1038/s41586-026-10656-8); Baker lab / IPD. Designed miniprotein agonists/antagonists with on/off control of GPCR signaling; one antagonist mobilized hematopoietic stem cells in vivo comparably to a clinical drug with fewer side effects.
Review: "The past, present and future of de novo protein design" Demonstrated — Nature 652:1139–1152, 2026 (DOI 10.1038/s41586-026-10328-7). Authoritative field-direction marker: frames the shift from directed evolution to deep-learning-driven intentional design, and flags integrated nanomachines (binding + conformational change + catalysis) as the next frontier.

Structural DNA/RNA nanotechnology

Design principles for accurate, fast DNA-origami folding Demonstrated — Small / UT Austin, ~21 May 2026 (Marras lab). Fewer scaffold crossovers give more cooperative folding; over-strengthening strand binding backfires entropically. A 1–2 h anneal raised yields up to ~17% and produced millions of structures faster — directly addressing the field's yield/throughput bottleneck.
Computational framework for micron-scale crisscross DNA megastructures Reported — bioRxiv, 23 Jan 2026 (Wyss/Harvard-associated). Evolutionary + graph-based handle assignment suppresses parasitic binding, pushing reliable programmable self-assembly from nanoscale toward microscale (prior demos reached ~5.4 GDa, ~2 μm).

Molecular machines, motors & switches

ML-designed RNA riboswitches via Restricted Boltzmann Machines Demonstrated — Nature Communications, ~Feb 2026 (CNRS / ENS Paris et al.). Generative ML learns natural riboswitch "design rules" and produces novel functional RNA switches — bridges AI molecular design and functional RNA nanotechnology.
Tuneable autonomous DNA-origami engine (RNA-fueled, RNase H–activated) Reported — Nature Communications, ~Feb 2026. A self-resetting molecular actuator driving cyclic microscale motion. Caveat: confirmed only via the Nature subject feed; article record not independently opened — verify before relying on specifics.
Context (borderline pre-window): a molecular motor directing rotaxane synthesis Demonstrated — Angew. Chem. Int. Ed. (DOI 10.1002/anie.202520085), ~Nov 2025, Kathan lab (HU Berlin). Flagged as just before the window.

Enabling tools — imaging, characterization, AI design

Atomic-resolution cryo-EM at 200 keV (1.24 Å) Reported→Demonstrated — bioRxiv 9 Jan 2026 → IUCr journal 2026. True atomic resolution on a cheaper 200 kV instrument (cold FEG + energy filter), lowering the cost barrier for atomic-resolution structure determination of designed nanostructures.
Riemannian denoising model for molecular structure optimization Demonstrated — Nature Computational Science, ~Feb 2026. AI that judges molecular stability and generates valid 3D structures ~10× faster (≈100 vs ≈1000 steps) — design substrate for atomic-precision targets. Related generative-flow work (e.g. PropMolFlow) is Reported.
Localization-AFM extensions for flexible proteins Demonstrated — PLOS Comp. Biol. 2025–2026. Unsupervised deep learning registers/clusters AFM images by conformation, extending single-amino-acid-resolution AFM to dynamic proteins. Freshness flag: the core LAFM method is 2021; only incremental tooling is in-window.

3 · New commercial activity

Company	What they do	Stage / funding	This window's update	Tier
CBN Nano Technologies	Mechanosynthesis / APM (carbon structures by STM)	Backed by Canada's Strategic Innovation Fund & Canadian Bank Note Co.	Posted the additive-mechanosynthesis preprint — the window's central MNT claim	Reported
Skape Bio	AI-designed GPCR biotherapeutics (IPD miniprotein platform)	New spinout, launched 21 May 2026	Public launch alongside the Nature GPCR-miniprotein paper	Reported

Institute for Protein Design (UW) remains the dominant productive engine in designed protein nanomachines (three Nature outputs in-window). Foresight Institute continues bio/nano fast grants and APM advocacy; no new in-window roadmap or large grant verified. Zyvex Labs' ZyvexLitho1 remains the commercial APM-lithography reference but had no fresh 2026 announcement (latest hard news 2022).

4 · New institutional / policy items

No new MNT-specific government program, NNI action, or major grant was verified in-window. The most consequential institutional signal is funding-adjacent rather than policy: CBN's continued state-linked backing and the IPD→Skape Bio spinout. No MNT M&A verified this window.

5 · Quiet areas

Small-molecule synthetic molecular motors/machines (Feringa/Leigh/Stoddart-style) — no clearly in-window flagship; strongest item (Kathan rotaxane) is Nov 2025.
Supramolecular chemistry & foldamers — no notable in-window primary result surfaced.
On-surface synthesis / single-atom manipulation (outside the CBN work) — freshest strong primary paper is 2025; nothing distinct in-window.
Policy / government programs — nothing MNT-specific in-window.
Zyvex / SPM hydrogen-lithography commercialization — quiet (latest hard news 2022).
M&A — none verified in MNT proper.

6 · Sources

Mechanosynthesis preprint — arxiv.org/abs/2605.27250 (26 May 2026); context somewhereville.com (27 May 2026)
Quasisymmetric nanocages — bakerlab.org (22 May 2026); hhmi.org
GPCR miniproteins — nature.com/articles/s41586-026-10656-8 (21 May 2026); spinout Skape Bio launch (21 May 2026)
De novo design review — nature.com/articles/s41586-026-10328-7 (2026)
DNA-origami folding — news.utexas.edu (21 May 2026); phys.org
Crisscross megastructure framework — bioRxiv (23 Jan 2026)
ML riboswitches — phys.org (Feb 2026)
Cryo-EM 200 keV — bioRxiv (9 Jan 2026); IUCr
AI stable-structure model — phys.org (Feb 2026); generative design nyu.edu (Jan 2026)
Rotaxane-by-machine (borderline) — Angew. Chem. (Nov 2025)
Foresight bio/nano grants — events.foresight.org

Confidence note: Source quality is mixed-to-good — the protein-design items rest on peer-reviewed Nature papers (high confidence); the headline mechanosynthesis result and several DNA items are preprints/single-source (treat as Reported, contested until replicated). The RNA-fueled DNA engine is feed-confirmed only. Lab and author self-summaries were used for context but the underlying Nature/arXiv records are the primary sources. report.css could not be inlined this run (outside mounted folder); a self-contained fallback style was used.