A silicon chip the size of a fingernail holds billions of tiny switches. But a quantum computer needs not switches but individual electrons trapped in microscopic pits. These pits are quantum dots. And here's the catch: two neighboring pits that look identical can hold an electron differently. One is slightly deeper, the other slightly shallower. Previously, such differences could only be guessed at.
Now there's a way to see them. Scientists scan the chip's surface with a fine needle and measure how strongly an electron responds to a magnetic field. This response, the g-factor, is unique to each electron, like a fingerprint. By scanning a region 40 by 400 nanometers, the researchers found that each quantum dot has two g-factor values. They correspond to two different electron states linked to valleys — special zones in the silicon crystal. The two valleys are like two floors in a building, and the electron can live on the first or the second.
The g-factor maps were compared with maps of another parameter — valley splitting, which shows the energy difference between the two floors. It turned out that both parameters vary along the same line — the path along which electrons are transported across the chip, like on a conveyor belt. This allowed them to calculate the coupling strength between valleys at each point along the path. This coupling is key to how an electron behaves during transport.
For quantum computing, it's important that all electrons are identical, like soldiers in formation. If one responds slightly differently to a magnetic field, it throws off the calculations. Now engineers will be able to spot problem areas on a chip in advance and fix them. It's like getting an X-ray before surgery, not after.
A quantum dot in silicon is so small that about ten thousand such dots would fit in a row across a human hair.