A methane photoconversion experiment succeeds or fails at quantification. Products split across phases, methanol and formate stay mainly in the liquid, CO leaves with the gas, and each phase has its own analytical method, calibration and error profile. Treating them as one "yield number" hides the hard parts.
Standard routes are headspace GC, gas chromatography of a liquid aliquot, and colorimetric assays after derivatization. At micromolar levels, contamination and blank subtraction dominate error: run reagent blanks every session and calibrate methanol standards freshly, it is volatile and hygroscopic, and stock concentrations drift.
Formate quantifies cleanly by ion chromatography, and by NMR with a formate-specific chemical shift or enzymatic/colorimetric assays. The main risk is oxidation during storage, analyze promptly or quench and refrigerate. Calibrate in a matrix matching the reaction liquid, since salts and catalyst residues bias some assays.
CO is measured by online detection or GC. It is reactive on metal surfaces, so inert flow paths and blank runs matter. If H2 co-evolves, verify detector selectivity, CO sensors often respond to H2 and vice versa.
Close the carbon balance: methane consumed versus methanol + formate + CO + CO2 found. Report detection limits per product and state explicitly when a product was below limit. Selectivity calculated on unmeasured products is not selectivity, it is optimism.
For coupled gas-liquid quantification, see Perfectlight's online gas analysis and photocatalysis solutions.
