Set pH deliberately, not by default. On TiO2 the point of zero charge is pH 6.0–6.5, so the surface is negative above pH 7 and repels anions. TEOA needs pH 8–11, sulfide/sulfite needs pH 12–13, and methanol works over pH 5–9. A pH move of one unit shifts the hydrogen potential by 59 mV and can change the rate by 20–50 %.
Three things move at once, and they pull in different directions, which is why a pH series rarely shows a single clean optimum.
First, surface charge. Below the point of zero charge the oxide surface carries net positive charge and adsorbs anions; above it, the surface is negative and adsorbs cations. For a cationic donor or a metal cation cocatalyst, raising pH past the PZC increases adsorption and can raise the rate. For an anionic donor such as S2- or SO32-, the opposite happens, and the electrostatic penalty is why those systems are run at high pH with a high ionic strength to screen it.
Second, the redox potential of the proton-coupled steps. Both hydrogen evolution and the donor oxidation step shift with pH; a one-unit rise lowers the H+/H2 potential by about 59 mV at 25 °C. If the conduction-band edge shifts by a similar amount, the driving force stays roughly constant and the rate barely moves — which is exactly why pH effects are often attributed to kinetics and adsorption rather than thermodynamics.
Third, the donor's own speciation. This is the largest effect in practice, and it is the one most often left uncontrolled.
| Donor | Working pH | Why that window | Failure mode outside it |
|---|---|---|---|
| Methanol, 10–20 vol% | 5–9 | Neutral molecule, no pKa in range; oxidation is hole-limited | Strong acid dissolves carbonate and changes the surface; strong base accelerates the Cannizzaro side path |
| TEOA, 5–10 vol% | 8–11 | pKa 7.8; the deprotonated amine is the active donor | Below pH 7 the amine is protonated and the H2 rate drops sharply |
| Na2S / Na2SO3, 0.1–0.35 M | 12–13 | Keeps S2- and SO32- deprotonated and suppresses H2S | Below pH 11, H2S forms, the rate falls and the reactor smells of sulfide |
| EDTA, 1–10 mM | 5–8 | Multi-charge anion; needs to be deprotonated to adsorb | Acidic conditions protonate it and cut adsorption; it also chelates dissolved metal |
Measure the pH after the donor is added, not before. Adding 10 vol% TEOA to water moves pH from 7 to about 10.5 on its own, so a solution described as “neutral with TEOA” is usually already at the top of its window.
Many buffers used in photocatalysis are themselves hole scavengers. Tris and EDTA both donate electrons to photoexcited TiO2, so buffering with them adds a second sacrificial agent and inflates the hydrogen rate. Phosphate adsorbs strongly on oxide surfaces and blocks active sites, shifting the apparent PZC and reducing the rate by 10–30 % in some systems.
Borax/boric acid (pH 8–9), carbonate (pH 9–10.5) and dilute NaOH are the safer choices when the buffer must not donate. For the sulfide system, the sulfide/sulfite couple is itself the buffer at pH 12–13, and adding a phosphate buffer only introduces a competing adsorbate.
Record buffer identity and concentration, because “pH 7 phosphate” and “pH 7 unbuffered” are different experiments.
Salt screens surface charge. Raising ionic strength from 1 mM to 0.1 M compresses the double layer from roughly 10 nm to 1 nm and can increase the adsorption of an oppositely charged donor, raising the rate by 20–40 % without any change in pH. Above 0.5 M, powders flocculate, the effective surface area drops and the rate decays during the run.
Choose one salt, use it across the whole series, and report the concentration in mol/L. NaCl is fine for most oxide systems; avoid chloride with silver-based cocatalysts, and avoid sulfate when the anion is also a donor. Keep ionic strength constant while varying pH, otherwise the two effects are inseparable and the pH trend you report is partly a salt trend.
Report pH, buffer, ionic strength and the drift together. A single “pH 7” in a methods section is not enough to reproduce the rate, and it is the first thing to check when another group reports a different number for the same catalyst.
It depends on the donor, not on the catalyst alone. Methanol works over pH 5–9, TEOA over 8–11 because its pKa is 7.8, and Na2S/Na2SO3 needs 12–13 to stay deprotonated. Pick the window from the donor, then optimise within it.
It shifts the band edges rather than the gap. Both the conduction and valence band edges move by roughly 59 mV per pH unit at 25 degrees Celsius, so the driving force for hydrogen evolution is nearly pH-independent while adsorption and surface charge are not.
Only with a buffer that does not donate holes. Borate and carbonate are acceptable; Tris and EDTA are sacrificial agents and will inflate the rate. Unbuffered runs are fine if you report the starting and ending pH, since TEOA and sulfide systems drift 0.5–1.5 units over a few hours.
At pH 3 the oxide surface is positively charged and repels anionic donors, and many oxide supports begin to dissolve. Check the filtrate by ICP for dissolved metal before attributing the drop to a change in mechanism.