When Arinna announced its $4 million seed round to build ultrathin solar cells for spacecraft, the most interesting line in the press wasn't the dollar amount or the investor list. It was a six-syllable phrase that almost no one outside a materials science department would recognize: transition metal dichalcogenides, or TMDs.
If TMDs deliver on their promise, they could reshape not only how we power satellites and orbital data centers, but also how we think about flexible electronics, solar windows, and a long list of devices on Earth. Here's a plain-English guide to what they actually are.
What Is a Transition Metal Dichalcogenide?
Strip away the chemistry-class vocabulary and a TMD is just a sandwich. Each "slice" of the sandwich is a single atomic layer:
- A layer of a transition metal, such as molybdenum (Mo) or tungsten (W)
- Sandwiched between two layers of a chalcogen element, such as sulfur (S) or selenium (Se)
Common examples include molybdenum disulfide (MoS₂), tungsten diselenide (WSe₂), and molybdenum diselenide (MoSe₂). Each three-atom-thick layer behaves as a self-contained semiconductor — meaning a single TMD layer is roughly 10,000 times thinner than a human hair, but still capable of doing the job that a much thicker piece of silicon does in a conventional solar cell.
[HIGHLIGHT]A single TMD layer is approximately three atoms thick — yet it can absorb light, generate charge carriers, and conduct electricity. Stacks of these layers can be engineered into photovoltaic cells that are mechanically flexible, mass-light, and radiation-tolerant in ways that conventional rigid panels cannot match.[/HIGHLIGHT]
How TMDs Compare to the Incumbent Materials
To appreciate what's different about TMDs, it helps to look at the materials they would replace in space:
[STATS]
Silicon | Cheap, mature, but degrades in radiation
Gallium Arsenide | Highly efficient, expensive, rigid
TMDs (Arinna) | Atomically thin, flexible, claims 32% efficiency uplift
[/STATS]
**Silicon** is the workhorse of terrestrial solar. It's well understood, manufacturable at extraordinary scale, and increasingly being adapted for low-cost satellites. Its weaknesses in space are well documented: comparatively low specific power (watts per kilogram), and accelerated degradation under high-energy particle radiation.
**Gallium arsenide multi-junction cells** are the gold standard for traditional satellites. They achieve efficiencies above 30% and tolerate radiation well, but they are rigid, heavy, expensive, and constrained by the rare-earth supply chain. They were designed for an era when satellites were bespoke, billion-dollar science projects rather than mass-produced commodities.
**TMDs** are a categorical departure from both. Because each layer is just three atoms thick, the same surface area can accommodate stacks of differently tuned semiconductors — each absorbing a different slice of the solar spectrum — at a fraction of the mass and rigidity of a multi-junction GaAs cell.
Why TMDs Are Hard
If TMDs are so promising, why aren't they already in every spacecraft? The honest answer: making them at scale has historically been brutal. Three challenges have kept TMDs in the lab for most of the last two decades:
1. **Wafer-scale uniformity.** Producing a single, uninterrupted atomic layer across a useful area is fundamentally harder than depositing a thicker conventional film. Defects that would be invisible in silicon become performance-killing problems in TMDs.
2. **Layer transfer and stacking.** Real devices need multiple TMD layers stacked with precise alignment. Doing that without introducing contamination or mechanical damage is a delicate engineering problem.
3. **Encapsulation and durability.** TMDs are sensitive to oxygen, moisture, and certain processing chemistries. Producing cells that survive long enough to be commercially useful requires solving an entire encapsulation stack.
Arinna's pitch is essentially that the founders' Stanford research has moved each of those problems from "intractable" to "engineerable" — and that the company can wrap them into a roll-to-roll manufacturing process by 2028.
What TMDs Could Unlock in Space
If TMD photovoltaics survive Arinna's planned 2026 on-orbit qualification campaign and reach megawatt-scale production in 2028, the implications go well beyond a slightly better satellite power bus:
- **Conformal and structural solar.** Because the cells are mechanically flexible, they can be wrapped around or embedded into spacecraft structures, freeing up surface area for other instruments.
- **Roll-out arrays.** A roll-to-roll process means deployable arrays could be lighter, cheaper, and easier to fold for launch.
- **Power-hungry payloads.** Orbital data centers, high-power radar, deep-space radios, and electric propulsion all benefit when watts-per-kilogram improves by even a few percent. A 32% efficiency uplift is a generational change.
- **Lunar and Mars surface power.** Flexible, lightweight cells that ship in weeks rather than years are a much better fit for the surface logistics of off-world infrastructure than traditional rigid panels.
The Honest Caveat
None of this is guaranteed. TMDs have been "five years away" for most of a decade, and many promising materials-science companies have failed in the gap between lab demo and flight qualification. Cosmic radiation, thermal cycling between sunlight and shadow, micro-meteoroid impact, and the simple reality of operating without maintenance for 15 years are an unforgiving combination.
What makes Arinna interesting is that they have reached the point of building real qualification panels on a real customer's spacecraft — moving the question from "could TMDs work in space?" to "do these specific TMD cells survive their first year on orbit?" That's a meaningful step closer to the answer that the entire space-power supply chain has been waiting for.