‌
Vision · Diligence · Excellence
Grow with Light, Forge China's Instrument Brand
Home > Technology > Instrument Knowledge > Loading a Pt Co-Catalyst by Photodeposition: wt%, Method and Pitfalls
2026-11-22102

Loading a Pt Co-Catalyst by Photodeposition: wt%, Method and Pitfalls

Quick answer

Photodeposit 0.5–1.0 wt% Pt on most oxide photocatalysts. Dissolve H2PtCl6·6H2O, which is 37.7 wt% Pt, add 10–20 vol% methanol as hole scavenger, purge 30 min with argon, then irradiate at 100–300 mW/cm2 for 1–2 h. One wt% Pt usually raises the H2 rate 5–20 times; above 2 wt% the gain reverses.

wt% means metal mass, not precursor mass

Loading is platinum mass divided by catalyst mass: 1 wt% on 1.00 g of TiO2 means 10.0 mg of Pt. Because the hexachloroplatinic acid salt carries only 37.7 wt% Pt, reaching 10.0 mg of metal takes 26.5 mg of precursor. The most common error in the literature is weighing 10 mg of the salt, dissolving it and reporting “1 wt% Pt”, when the real loading is 0.38 wt%.

For a loading series, prepare one stock rather than weighing six times. 265 mg of precursor in 100 mL of water gives 1.0 mg Pt per mL, so a 0.2 / 0.5 / 1.0 / 2.0 wt% series on 1.00 g of catalyst is 2, 5, 10 and 20 mL of stock, each made up to the same final volume. Add the stock dropwise over 5 min with stirring; pouring it in creates a local concentration spike and larger particles.

Photodeposition versus the alternatives

MethodTypical Pt sizeWhere Pt sitsWorking rangeMain risk
Photodeposition, 10–20 vol% methanol, 1–2 h2–5 nmAt electron accumulation sites: defects, edges, specific facets0.1–2 wt%Incomplete reduction; dark adsorption before illumination
Incipient wetness + H2 at 300 °C3–10 nmSpread over the whole accessible surface0.5–5 wt%Sintering above 400 °C; chloride retention
NaBH4 reduction in slurry2–8 nmRandom, including nuclei formed in the liquid0.5–5 wt%Homogeneous nucleation wastes metal off-support
Commercial 1 wt% Pt/TiO22–4 nmVendor-definedFixedUnknown pretreatment and chloride history

The reason to choose photodeposition is placement, not size. Photoexcited electrons arrive at specific surface sites, so Pt grows where reduction actually happens, which shortens the charge-transfer path from the support to the metal. Above roughly 3 wt% the process becomes slow and uneven, and impregnation is the better tool.

A protocol that reproduces

  • Disperse the catalyst at the same g/L you will use in the activity test and keep solids loading constant across the series. Changing it changes light penetration and therefore the deposition pattern.
  • Add methanol to 10–20 vol%. Below 5 vol% reduction stalls; above 30 vol% the alcohol competes for photons and the suspension pH drifts.
  • Purge 30 min with argon or nitrogen while stirring in the dark, and log the dark contact time. Then illuminate without changing anything else in the vessel.
  • Irradiate at 100–300 mW/cm2 broadband or UV-A for 1–2 h. A 300 W xenon source at 20 cm behind a water filter sits in this range. The yellow PtCl62- colour fading to grey is a rough end point, not proof of completion.
  • Recover by centrifugation at 8000 g for 10 min, wash three times with water, dry at 60 °C under vacuum. Do not calcine unless you accept sintering.

Five ways the loading goes missing

Dark adsorption. PtCl62- is an anion and oxide surfaces are positive below pH 6.3, so 20–40 % of the platinum can leave solution within 30 min in the dark. Material adsorbed in the dark is not photodeposited even if it is reduced later, so run a dark control and report the adsorbed fraction separately.

Incomplete reduction. Check the supernatant by ICP-OES. More than 10 % of the nominal Pt still in solution means the run was too short, the donor was consumed, or light did not reach the whole slurry.

Chloride left behind. At 1 wt% on 1 g the precursor brings about 11 mg of chloride. Left in 100 mL that is over 100 ppm, enough to shift selectivity and attack stainless fittings. Three 100 mL washes bring it below 10 ppm.

Pt on the wrong surface. A grey mirror on the reactor wall or the stir bar is lost catalyst. Check the glass after each run and prefer quartz or a coated immersion well.

Aggregation above 2 wt%. Particles grow, platinum area per gram saturates, and at 3–5 wt% the metal begins absorbing incident light. Rates typically peak between 0.5 and 1.5 wt% and drop 20–50 % at 3 wt%.

Verify before you publish the number

Three measurements make a loading claim defensible: ICP-OES of the digested composite for actual wt%, ICP of the supernatant for reduction yield, and CO pulse chemisorption or STEM for dispersion. A nominal 1 wt% that measures 0.7 wt% by ICP is normal, and it should be reported as measured rather than nominal.

Frequently asked questions

What is the best Pt loading for photocatalytic hydrogen?

0.5 to 1.5 wt% covers most oxide photocatalysts, with the optimum usually near 1 wt%. Below 0.2 wt% there are too few active sites; above 2 wt% particles grow and screen light, and rates fall 20 to 50 % at 3 wt%.

Is photodeposition better than impregnation?

For placement, yes: it puts Pt where electrons arrive, typically 2 to 5 nm particles at defects. For loadings above 2 wt%, or for supports that cannot generate electrons, impregnation followed by H2 at 300 degrees Celsius is faster and more even.

Why did my Pt loading show no activity gain?

Check four things: actual Pt by ICP, chloride residue, metal deposited on the reactor wall, and whether the activity test used a donor at all. A nominal 1 wt% that measures 0.3 wt% explains a flat result without any mechanistic surprise.

Do I need a sacrificial agent for photodeposition?

Yes for oxides. Methanol at 10 to 20 vol% is standard, and ethanol, TEOA and Na2S/Na2SO3 also work. Without a donor, photogenerated holes oxidise the lattice or solvent slowly and reduction stays incomplete after two hours.

Related reading

Download
Chat Service
TOP