surya.narada.systems

R5: in-space cell manufacturing, mapped

Sweep executed 18 August 2026. The finding: the direction is uniquely leveraged and unclaimed. Nobody makes photovoltaics in space, nobody is funded to, and the incumbent ISAM program that would have built the adjacent capability was cancelled at nearly $2B spent. The open critical path is a deposition-process demonstration at rate and quality on orbit — a 30-year-old molecular-beam precedent is the entire flight heritage.

Why this map exists

R5 is the long game on the directions page: deposit thin-film cells directly onto deployed membranes in orbit, escaping substrate, coverglass, and launch mass in one move. "Nobody else is motivated to build this" is a claim — this page checks it, prices the leverage, and names what is actually missing.

The leverage math

LineBasisSpecific power
Feedstock, launchedUltralight perovskite modules at 30.3 W/g demonstrated on Earth (Wu 2022) — cells are chemicals and microns of film~30,000 W/kg equivalent
State of practice arraysSOTA rigid/flexible arrays ~40 W/kg (Aerospace Corp 2018); Redwire ROSA roll-out 100–120 W/kg40–120 W/kg
Advanced projected arraysCost/mass modeling of advanced technologies (Gaddy 2017)to ~225 W/kg
The gapFeedstock-to-array mass leverage if the coater itself is cheap and reusable~100×

established Two orders of magnitude of mass leverage sit between launched feedstock and launched arrays. That is the largest single lever in the program's portfolio — larger than substrate liberation (R3), which fights for ~3× on the same axis. The catch is the coater: the leverage survives only if the deposition hardware is cheap, reusable, and does not need the array shipped back.

Flight heritage and the empty chair

  • Wake Shield Facility (1994–96) — The only in-space semiconductor deposition ever flown: 8 MBE/CBE source cells, 7 wafers, GaAs-family films grown on STS-60/69/80. Nearly carbon-free AlGaAs; ~10⁻¹⁰ Torr achieved behind the 3.7 m shield (theory promised 10⁻¹⁴). Never a complete device; the program ended with the Shuttle era.
  • IKAROS (2010) — The only ultralight cell-on-membrane ever flown: a-Si cells on 7.5 µm polyimide, 20 m diagonal sail, 307 kg, spin-deployed, ~500 W. Fabricated on Earth, survived the fairing — exactly what R5 wants to skip.
  • ISAM missions — OSAM-1 cancelled March 2024 at nearly $2B total cost; OSAM-2/Archinaut One concluded without flight. Actual in-space manufacturing heritage: polymer 3D printing, ZBLAN fiber draws (ISS, 2017–23), one Redwire ceramic demo. No PV anywhere in the portfolio.
  • Adjacent thermal processing — Space Forge generated >1000 °C plasma aboard ForgeStar-1 (June 2025) — high-temperature process control on orbit is being demonstrated, for semiconductors but not for deposition.

established The two halves of R5 have each flown separately, three decades apart: deposit a semiconductor film in orbit (Wake Shield, 1994–96) and fly a complete ultralight cell on a membrane (IKAROS, 2010). Nobody has combined them — grown the cell on the membrane in space. And nobody is trying: the ISAM portfolio that would naturally own the mission profile contains no PV manufacturing at all.

Environment and process constraints

  • Atomic oxygen (shapes the orbit) — Kapton erosion yield 3.0×10⁻²⁴ cm³/atom (MISSE PEACE). Unprotected micron-scale polymer membranes are short-lived in LEO ram — R5 lives at high LEO/GEO, in wake, or behind coatings.
  • Rate vs quality (shapes the process) — Record flexible CIGS on polymer (21.4%, Empa 2021) grows at 0.1–0.5 nm/s. A coater that must cover m²/minute accepts relaxed quality, or solves rate-quality co-optimization. The wedge is choosing the regime on purpose.
  • Space resilience of the product (supports the route) — Perovskite films stable through 10 months ISS exterior exposure (UC Merced 2023); quasi-2D ultrathin flexible perovskite cells radiation-stable (Putz 2026). The membrane-native cell materials tolerate the environment they'd be made in.

The economic bar

  • Launch cost — Falcon 9 ~$3,000/kg LEO; rideshare $7,000/kg ($350k/50 kg SSO); Starship aspiration $100–1,000/kg. Falling launch cost is the headwind: it shrinks the mass leverage that is R5's whole case.
  • Incumbent cell cost — 3J cells $50–100/W; a 20 W GaAs CubeSat panel runs €12,800 (~€640/W at panel level). Against that, Starlink flies COTS silicon at $0.20–0.50/W and builds ~70 satellites/week in Redmond — cost was never the cell for the volume player.
  • Regime dependence — R5 wins where mass and stowage dominate cost: very large arrays, GEO and beyond, deployables past fairing limits. It loses where launch is cheap and COTS silicon is good enough — which is Starlink's regime, and Starlink is eating the market.

The honest tension: R5's case strengthens with array size and distance (huge GEO platforms, deep-space, fairing-limited deployables) and weakens with every Starship milestone. The direction is a bet that very-large-structure economics arrive before very-cheap-launch economics eat them. The Starship headwind is real but bounded: at any launch price, shipping 1 kg of feedstock instead of 100 kg of array is a 99 kg saving — the question is whether the coater's amortized mass and operations give it back.

What is actually missing

established-with-gap The gap decomposes into four items, in dependency order: (1) a deposition-process demo — rate and quality in uncontrolled vacuum on a membrane-class substrate, on the ground first; the WSF precedent says vacuum quality is achievable and the rate-quality literature says the regime must be chosen on purpose; (2) contamination control — the coater and the host platform must not poison each other; WSF's own history (10⁻¹⁰ Torr achieved vs 10⁻¹⁴ promised) is the cautionary dataset; (3) the membrane environment answer — orbit selection or coatings that keep a micron-scale substrate alive against atomic oxygen; (4) the mission profile — repair and re-coat as a service, which is an ISAM rendezvous problem that no living program is staffed to fly.

What this changes

R5 stays far-term, but it is no longer a slogan: the leverage is quantified (~100×), the empty chair is verified (no funded in-space PV program exists), and the critical path is a ground deposition demo, not a flight. The cheap near-term moves: track Space Forge's plasma work and any Archinaut asset dispersal for acquirable capability; watch perovskite membrane flight results from the R2 perimeter (they double as R5 feedstock-material heritage); and keep the launch-cost trend on the watch list, because it is the one external variable that can kill the direction.

Perimeter and provenance

Four-track sweep (deposition precedents, membrane flight heritage, ISAM landscape, economics), 25 sources, admitted to the Surya epistemic graph as event ev-000000000011 with assessment surya-assessment-r5-gap-map-2026-08, judgment established-with-gap. Sweep-level gaps, flagged honestly: MicroLink and Redwire deposition-rate disclosures were not obtained; WSF contamination analyses beyond the project page were not reviewed; the ForgeStar-1 plasma result is a secondary source; launch prices are list prices, not contract prices. The sweep was executed in-session because the delegation-worker surface has no network access — recorded as method, not hidden.

Sources