Photothermal CO2 hydrogenation runs at 200–450 °C and 1–30 bar, and the light must hold the bed at that temperature: 1 W/cm2 of full-spectrum xenon on a black absorber sustains roughly 100–200 K above ambient, reaching the set point in 30–120 s. Treat irradiance as a controlled heat flux and prove any non-thermal gain by a drop in apparent activation energy of 15–40 kJ/mol at matched surface temperature.
Absorbed flux leaves the bed by convection, conduction to the walls and radiation. At steady state ΔT = α·I / heff, where α is absorptance and heff lumps every loss path. In a gas-flowing metallic reactor heff runs 40–80 W/m2·K; in a vacuum-jacketed quartz cell, 10–20.
That line explains most surprises. Doubling irradiance does not double temperature once radiation dominates: at 700 °C a black surface radiates 1.4 W/cm2, so above about 500 °C radiative loss alone eats half of a 1 W/cm2 input. It is also why the same lamp takes one reactor to 350 °C and another to 180 °C.
| Irradiance at the bed | ΔT with heff = 60 W/m2·K | ΔT with heff = 150 W/m2·K (pressurised flow) | Typical regime |
|---|---|---|---|
| 0.2 W/cm2 | ~30 K | ~12 K | Photo-assisted only; furnace does the heating |
| 0.5 W/cm2 | ~75 K | ~30 K | Mild preheating, RWGS onset |
| 1 W/cm2 | ~150 K | ~60 K | Sabater methanation window with 200 °C furnace |
| 2 W/cm2 | ~300 K | ~120 K | Light-driven methanol synthesis, 220–280 °C |
| 3 W/cm2 and above | radiation-limited | ~180 K | Only useful in vacuum or concentrated optics |
Values assume α = 0.9, single-sided illumination, 25 °C ambient and no reaction heat: design numbers, not specifications. Recalibrate on your own cell with one thermocouple and three irradiance levels.
A 300 W xenon source puts roughly 40–55 % of its output between 700 and 1,100 nm, where most oxide catalysts absorb weakly. Photothermal beds therefore rely on an absorber: SiC (absorptance 0.85–0.95 from 400 to 2,500 nm), carbon or black TiO2, or a plasmonic metal resonant inside the lamp envelope.
Two hardware details decide whether the photons arrive. Borosilicate cuts off at 320 nm, standard quartz transmits to 2,500 nm, and above 10 bar you need sapphire. A water-cooled IR filter, standard in photocatalysis, removes exactly the photons a photothermal run wants: test with and without it before blaming the material.
In a bed of 0.5–1 mm pellets, incident light is extinguished within 0.5–2 mm, so a 3 mm bed runs an axial gradient: 380 °C at the front face, 240 °C at the back. Report where the thermocouple is, ideally three of them front, middle and back.
Two fixes are common. Dilute the catalyst with inert SiC or quartz of matched particle size, which flattens the profile at the cost of catalyst per unit volume. Or illuminate both faces of a 1–2 mm bed, doubling irradiated area and shortening the conduction path to the wall. Neither changes intrinsic kinetics; both change the temperature the catalyst sees.
Run the Arrhenius series twice: dark at a set of surface temperatures, then illuminated at the same surface temperatures with the furnace compensating the extra heat. Four temperatures spanning 40–60 K give a usable slope. If light only heats, the lines coincide and Ea is unchanged. A shift of 15–40 kJ/mol, with the illuminated line above and parallel or steeper, is the quantitative form of the claim.
Two biases ruin it. Using furnace temperature instead of surface temperature under illumination inflates the light-only rate, since the surface runs hotter than the block. A hot spot yields a rate averaged over a non-isothermal bed, which no Arrhenius fit rescues.
Methanol synthesis over Cu/ZnO catalysts wants 220–280 °C and 20–30 bar; Sabatier methanation over Ni runs at 300–400 °C and is far less pressure-sensitive. H2/CO2 of 3:1 to 4:1 suppresses reverse water-gas shift, and 3,000–12,000 h−1 keeps conversion differential. Light adds coking risk a furnace does not: the front face can run 100–150 K above set point, and past 450 °C carbon lays down on Ni fast, so watch ΔP with CH4 selectivity.
1–3 W/cm2 at the bed for a light-driven experiment, which is 10–30 times the 100 mW/cm2 used in photocatalysis. Below 0.5 W/cm2 the temperature rise is under 75 K and the furnace, not the light, sets the conditions.
Sometimes, and it must be shown rather than asserted. Compare dark and illuminated runs at matched surface temperature: an apparent activation energy lower by 15–40 kJ/mol, or a rate ratio above about 1.3 at identical temperature, is evidence of a non-thermal contribution.
Because the thermocouple sits in the gas stream or against the wall while the irradiated front face runs hotter. In a 1 mm-pellet bed the axial difference can reach 100–150 K. Use an IR camera or a contact probe on the front face and treat the block reading as a lower bound.
Only if you accept losing the heating photons. A 1 W/cm2 xenon beam loses roughly half its power to a typical IR cut filter, so the bed cools 50–100 K. Keep the filter for the photocatalysis control and remove it for the photothermal run.
