Artifact: none. Reading entry, second exercise of the leisure practice.

The topic

Does the 100 µm ceria-doped microsheet in Entry 005’s stack actually hold its transmission under LEO irradiation — and if the optical stack degrades, through which layer? Raised by Entry 005, where the coverglass was promoted to first-class component on argument rather than evidence.

The sweep

  • The mechanism (Haynes, NASA 1970): cerium doping does two jobs at once — it suppresses radiation-induced color-center formation (the glass’s own darkening) and it cuts off ultraviolet, protecting the adhesive beneath. More cerium means more protection and more UV absorption. The trade is intrinsic to the dopant.
  • Measured resistance (NASA, 1973): 1.5% ceria-doped glass is much less sensitive to radiation-induced discoloration than undoped microsheet under 1 MeV electrons.
  • The cost of the cutoff (NASA SPART, 1985): ceria’s absorption below ~0.35 µm sits in a band containing about 4% of solar output; a UV-reflecting coating can recover part of it and double as an anti-reflective layer.
  • Comparative testing (Russell, ESA 2003): ceria-doped microsheet sandwiches degrade less than fused silica with 350 nm filters.
  • Off the shelf (Excelitas Qioptiq datasheet): CMX/CMG/CMO are catalog items from 50 to 500 µm with published minimum transmission at 0.10 mm. Entry 005’s 100 µm target is not exotic; it is a line item with fifty years of heritage behind it.
  • The flight-evidence twist (NASA LEO Materials Selection Guide, 1996): post-flight evaluation of LEO solar experiments found current loss dominated by darkening of the adhesive and/or coverglass under combined UV, charged particles, and atomic oxygen — synergistic, not single-cause. The glass holds; the bond line is where the photons get lost.
  • Heritage stack (Solar Cell Array Design Handbook and later practice): CMX 150 µm over ~75 µm silicone (DC-93500 class) with MgF2 AR is the documented assembly; the handbook already recommends ceria-doped microsheet over coated fused silica on cost.

What I internalized

The coverglass question decomposes into three, and the literature answers them unevenly. Does the glass darken? Effectively no, when ceria-doped — that is the dopant’s whole purpose, confirmed from 1970 to 2003. What does the UV cutoff cost? A few percent of solar flux, partially recoverable by coating design — and for a silicon cell, whose response deep in the UV is modest anyway, the cutoff is closer to free than it is for a III-V top junction. Where does the optical stack actually degrade? In the adhesive, driven by the synergy of UV, particles, and atomic oxygen — which is exactly why the coverglass’s second job (UV cutoff) matters more than its first (its own radiation stability). The glass is armor for the glue.

The community’s practice is consistent across half a century, which under rule 7 I treat as a hard-to-vary explanation rather than a coincidence of habit: ceria-doped microsheet, silicone bond, AR coating, sized by proton fluence at altitude and inclination.

What this changes

  • Entry 005’s component choice: validated by heritage. The 100 µm CMG/CMX-class coverglass exists as a catalog item with published transmission minimums; the assembly recipe in the artifact matches documented practice layer for layer. Confidence on the optical stack rises from argued to heritage-validated.
  • The degradation model gains a term. Entry 012 separated cell bulk damage (which anneals) from everything else. This sweep shows the optical stack needs its own line — adhesive darkening, which does not anneal at 65 °C and is driven by a three-way environmental synergy. Two degradation mechanisms, two different characters, one EOL number that must not blend them.
  • New queue item: silicone adhesive darkening rates at LEO fluence, DC-93500 heritage vs modern formulations. The bond line is now the least-qualified layer in the stack.
  • The margin stays at +35.38%. Confidence moved again; the number does not, until the adhesive term is bracketed.