Light entering a photoredox flow reactor through the tube wall is absorbed by the photocatalyst as it travels inward, following the Beer–Lambert law:
I(x) = I₀ · exp(−ε·c·x)
x = depth from the illuminated wall
ε·c = molar absorptivity × catalyst concentration
1/e penetration depth δ = 1 / (ε·c)
Near the wall, catalyst molecules see full intensity and turn over quickly (bright particles below). Past a few penetration depths, intensity collapses exponentially and the reactor core sits in the dark — a "photon-starved" core radius r_c where I(r_c) has fallen to 10% of I₀:
r_c = R − ln(10)/(ε·c)
Illuminated area fraction = 1 − (r_c / R)² (r_c clamped ≥ 0)
- Channel diameter — a wider tube grows R faster than δ, so the dark core swallows a bigger share of the volume: this is why photoredox reactions don't scale by simply making the tube bigger.
- Catalyst absorptivity — a more absorbing catalyst shortens δ, which speeds up near-wall turnover but shrinks the illuminated shell faster too.
- Incident irradiance — sets the excitation (pulse) rate of catalyst molecules that do see light; it doesn't change the geometric penetration depth.
- Parallel channels — "numbering-up": running several thin channels side by side multiplies throughput (∝ channels × D²) without touching D, so δ and the illuminated fraction stay exactly as good as a single thin channel.
This is the physical reason process chemists scale photoredox flow chemistry by adding identical thin channels rather than by widening one — exactly the trade-off the on-screen throughput readout tracks against illuminated fraction.