When customers tour the workshop, someone often points at the reactor and asks: "What does the platinum actually look like inside there?" It's a good question — because "what it looks like" directly determines how well it works. Platinum's form is not a single concept but a continuous spectrum, from the individual atom to the macroscopic metal lump. This article lays that spectrum out.

1. One Premise to Keep in Mind: Only Surface Atoms Do the Work

When we calculated TON in Article No. 6, the word "dispersion" came up once. Here we expand on it.

In a lump of metallic platinum, the atoms buried in the core never contact the substrate — money spent for nothing. So the central engineering goal for any catalyst has always been one sentence: spread the same amount of platinum as widely as possible.

The degree of spreading is measured by dispersion — the fraction of surface atoms out of the total. This number changes sharply with particle size: for particles tens of nanometres across, dispersion is only a few percent; particles a few nanometres across reach thirty to forty percent; at cluster level it can exceed seventy percent; and for single-atom platinum, dispersion is simply 100%.

Understand that curve and the four forms below fall into order easily.

2. Platinum Black: the Most Primitive Highly Dispersed Form

Platinum black is platinum reduced to an extremely fine powder (typically nanometre to sub-micrometre scale) — black, loose, and with a very large specific surface area.

Its active sites are the atoms on the metallic platinum surface, especially at steps, edges and kinks — positions that are coordinatively unsaturated and therefore the strongest at grabbing molecules. In the platinum black spontaneous ignition experiment described in Article No. 2, what burns is precisely the freshly dissociated hydrogen atoms on the surface.

Its characteristics are clear-cut: high activity, no ligand required, relatively low price — but poor thermal stability. At high temperature the small particles agglomerate and grow (by Ostwald ripening, and by particle migration and coalescence); dispersion declines over time, and activity falls with it. On the other hand it is a solid, separated by filtration, so the residue risk is lower than with homogeneous catalysts.

Applications: laboratory hydrogenation, parts of fine chemicals production, and as the starting raw material for preparing other platinum catalysts (Article No. 9 covered where it comes from).

3. Nanoparticles: the Workhorse of Supported and Heterogeneous Catalysis

The mainstream form of industrial heterogeneous catalyst is platinum particles a few nanometres across, anchored on a support — activated carbon, alumina, silica, titania, zeolites.

The active sites are still metal surface atoms, but the support is no mere "rack." It does at least three jobs:

• Physical isolation — preventing particles from contacting and agglomerating with one another, holding dispersion steady;
• Metal–support interaction — the electronic character of the support fine-tunes platinum's d-band centre (Article No. 2), affecting adsorption strength;
• Diffusion pathways — pore structure and pore size distribution determine whether the substrate can reach the sites at all.

The engineering trade-off is stark: make the particles small and activity is high but sintering is easy; make them large and they are stable but platinum is wasted. The entry "dispersion typically in the tens of percent" in the table of Article No. 7 refers to exactly this tier. The platinum, rhodium and palladium in an automotive three-way catalyst take this route — it must survive several hundred degrees Celsius in the exhaust for well over 100,000 kilometres, so stability carries far more weight than single-point activity.

4. Platinum Nanoclusters: the Critical State of a Few to a Few Dozen Atoms

When particle size drops below 1–2 nm and the atom count is only tens to a few hundred, the properties begin to depart from those of "small-particle metal."

The reason is the quantum size effect: platinum's energy bands are no longer continuous but split into discrete energy levels. The surface atom fraction jumps to seventy percent or higher, and the coordination environment of those atoms differs markedly from that of bulk metal.

The result is that activity often rises significantly, but stability becomes harder to achieve — a cluster's specific surface energy is extremely high, and thermodynamically it is strongly driven to grow. Engineering-wise, it takes a strongly interacting support, a confined space (for example, inside zeolite channels) or protective ligands to hold it down.

At present, cluster catalysts remain largely in research and in high-end fine chemicals; they are not used much in large-scale continuous industrial units, and the core bottleneck is lifetime.

5. Single-Atom Platinum: One Atom, One Active Site

At the other end of the spectrum, platinum is broken apart completely — every platinum atom anchored independently on the support, not touching its neighbours, existing in an ionic or quasi-ionic state.

It captures the advantages of both sides at once, which is the "middle ground" mentioned at the end of Article No. 7:

• Dispersion of 100% — platinum utilization is maximized, making this the core direction of the low-platinum route;
• Well-defined, highly uniform active sites — there is no "surface atoms differ from core atoms" problem;
• Heterogeneous in form — separation and recovery are as simple as with conventional supported catalysts.

The costs are just as real:

First, there is no continuous platinum surface. Reactions requiring the cooperation of several adjacent platinum atoms — typically homolytic dissociation of the hydrogen molecule, or C–C bond cleavage — cannot be carried out on a single atom, so it is no universal substitute.

Second, the coordination environment depends entirely on the support. A single atom must form strong coordination bonds with defect sites, oxygen vacancies or nitrogen doping sites on the support; otherwise, on heating it migrates and agglomerates into particles, and performance instantly reverts to that of an ordinary supported catalyst. The stability of the anchoring sites is the Achilles heel of single-atom catalysts.

Third, characterization and quality control are difficult. Atoms have to be inspected one by one with aberration-corrected electron microscopy; conventional methods struggle to confirm that "these really are all single atoms, with no clusters mixed in."

In cathodic oxygen reduction in fuel cells, low-temperature CO oxidation and selective hydrogenation, research on single-atom platinum is advancing fast. Our judgement is that it is one of the most promising technical routes for cutting platinum consumption — but the move from laboratory to large-scale production units still has to clear two hurdles: stability and consistency.

6. One More Class: the Molecular Homogeneous Active Site

All four forms above are "metallic platinum surfaces." Don't forget the other route covered in Article No. 7 — the active site of a homogeneous catalyst is not a metal particle at all, but a well-defined platinum complex molecule.

In Karstedt's catalyst, platinum is coordinated to vinylsiloxane in the zero oxidation state; every platinum atom is wrapped in ligands and dispersed independently in solution — "100% dispersion" by nature. Its active site structure can be resolved directly by NMR and IR spectroscopy, and its selectivity can be finely tuned by swapping ligands. That is the state single-atom catalysts are still chasing; homogeneous systems have had it all along.

What it buys that with is difficult separation and weaker thermal stability (Article No. 7). Engineering has always been about trade-offs.

7. A Quick-Reference Table

8. How to Choose: Go Back to the Reaction Itself

Choosing a form is not about picking "the most advanced"; it is about what kind of site your reaction needs:

• Needs a continuous metal surface (hydrogen dissociation, C–C cleavage) → platinum black or supported particles; single atoms are out;
• Only single-site activation needed, with extreme selectivity demanded → homogeneous complex or single atoms;
• High temperature, long-cycle continuous operation → supported nanoparticles, stability first;
• Platinum consumption is the core cost pressure, and the reaction type permits it → the single-atom / cluster route is worth evaluating;
• Batch kettle, ambient to moderate temperature, ppm-level loading → homogeneous — which is the drum we ship most of.