Gold is weird. It’s one of the few metals that doesn’t really oxidize. Even silver and copper—from the same column of the periodic table—form weak oxides. Naively, you might expect that gold would tarnish just like silver. Gold also sits right next to platinum, but it has none of that metal’s catalytic properties.
Then came gold nanoparticles that acted like catalysts, and we were confused by their apparent willingness to take part in chemical reactions.
Now, a pair of scientists has explained that gold’s inertness isn’t inherent to the atom but rather to the surfaces that gold crystals form. Before we get to the results, let’s first take a look at the traditional explanation for gold’s inertness and why an inert material that has no catalytic activity suddenly acts as a catalyst when in its nanoparticle form.
Can’t play, won’t try
Atoms are made of a central nucleus surrounded by electrons. The electrons form a structure, pairing up and filling orbitals from lowest to highest energy, with each orbital requiring a different number of pairs.
These orbitals are not like planetary orbits but rather a kind of volume of influence—they’re sometimes round but have other shapes as well. The highest-energy orbitals are generally farther from the central nucleus, exposing them to the rest of the world.
An atom’s readiness to react is due to partially filled orbitals at the highest energy, but nothing is ever quite so neat. For very heavy atoms, the order in which orbitals are filled is complicated, and partially filled orbitals can end up closer to the nucleus. If that happens, those orbitals are shielded from the Universe by the outermost orbitals, which are full of happy electron pairs.
This is why, we are told, gold is inert: Its unhappy electrons are shielded by happy electrons.
Let me arrange a reaction
The discovery that gold nanoparticles could act as catalysts told us that this explanation was incomplete. It also made the lack of catalytic activity on bulk gold surfaces a bit of a mystery.
Generally speaking, a catalyst is a material that enables a reaction without being consumed by it (some catalysts are consumed, but let’s not sweat the details). Every reaction needs to overcome an energy barrier to start—we heat stuff up to set it on fire, for instance. A catalyst reduces that barrier, allowing a reaction to start at much lower energies.
Biological systems are incredibly good at making reactions run at low temperature and pressure through the use of catalysts. Industrial chemistry achieves speed, scale, and cost savings due to catalysts, enabling the production of everything from petrochemicals to drugs.
In catalysis, surface area and surface structure play hugely important roles. Essentially, a molecule first has to stick to the surface of the catalyst. Then, depending on where it sticks, the molecule bends and stretches out of shape. At extremes, this can cause it to fall apart—like, for instance, an oxygen molecule becoming two separate oxygen atoms, leaving both halves in a highly reactive state. The highly reactive halves can then react with other molecules to form new molecules at considerably lower temperatures than would have otherwise been required.
So a good catalyst has a large surface area with plenty of sites where the target species can stick and fall apart. For gold to behave as it does—active as a nanoparticle, non-catalytic in bulk—it has to present a surface that has catalytic sites when it is present as a nanoparticle, but those sites must disappear on bulk surfaces, even if you make them rough and irregular.
Hex is best (if you like jewelry)
Gold (and, indeed all metals in the solid state) forms a crystal. If you cut a crystal along different planes of atoms, you get different arrangements on the surface. In gold, some planes reveal a square lattice, while others have a hexagonal lattice. The researchers hypothesized that some of those surfaces would be more catalytically active than others.
To confirm this, the researchers studied the behavior of an oxygen molecule on each type of gold surface. They asked how much molecular oxygen would stick to the surface, and for the molecules that did stick, what energy is required to cause the oxygen molecule to split. They showed that the surface structure commonly observed in bulk gold—a hexagonal pattern—does not hold onto oxygen very strongly, and the oxygen’s structure is not deformed. That means it still takes a lot of energy to split the oxygen molecule into two atoms that are ready to react.
On the other hand, if the gold structure is a square pattern, oxygen molecules readily stick to the surface and are deformed to the point of splitting, leaving them available to react (indeed, under these conditions, gold will oxidize as well). The researchers estimate that the square lattice gold surface is as active as common catalytic metals, such as platinum.
Hiding your sensitive bits
Gold surfaces are also quite active in the sense that gold atoms will readily rearrange themselves on the surface. By shuffling around, they change an exposed flat square lattice into a slightly rougher inactive hexagonal lattice. But the change, called surface reconstruction, can’t happen in just any way. Instead, the atoms move to form a 2D repeating structure that covers the exposed face, and the area required to form a complete unit of the repeating structure is quite large. On a chunk of gold, this is not an issue because there are plenty of atoms to go around, so each surface ends up almost completely inert.
On nanoparticles, the story is different. The limited number of atoms means there are not enough atoms or space for surface reconstruction. So a material known for its inertness suddenly shows its true colors and starts to react and act as a catalyst.
These studies show just how intricate the details of surface chemistry and catalysis can be. Inert metals become active and then return to inertness simply due to a change in material volume. It also opens new avenues for research on catalysis, though I don’t imagine gold will become the catalyst of choice any time soon.
Physical Review Letters, 2026, DOI: 10.1103/g3bc-t1qv
