An that cannot give off its energy as light, or only very weakly – because its have the wrong to recombine, or sit at different places in the crystal’s momentum space. Dark excitons do not show up in ordinary measurements of the light a material emits, yet they can be the lowest-energy state, so they decide how long excitons live and where their energy goes. In tungsten-based layers such as WSe2 a dark exciton lies just below the bright one, which is why these layers glow less, not more, when they are cooled.
As the site uses it
Large spin–orbit coupling in both bands produces spin-dark and momentum-dark excitons that dominate low-temperature emission, so exciton–phonon coupling and phonon-assisted recombination must be modelled to explain spectra.
An exciton is dark when its electron and hole have opposite spins or sit in different valleys. Either way light alone cannot carry off its energy, so the dark state can be the lowest one and still go unseen.
Why some excitons cannot shine
When an electron drops back into the hole it left, it can hand its energy to light only if the two match: light barely changes an electron’s spin, and it carries almost no momentum. In a , splits both the valence and the in each into two sub-bands of opposite spin. In MoSe2 the lower conduction sub-band has the same spin as the top of the valence band, so the lowest exciton is bright; in WSe2 and WS2 it has the opposite spin, so the lowest exciton is dark.
An electron can also settle in another valley than its hole – at K′ while the hole sits at K, or in the Q valley between them. Such momentum-dark excitons need a to make up the difference before they can emit, and in many TMDCs they lie close to or below the bright state.
Making the dark visible
An in-plane magnetic field mixes the two spin sub-bands and lends the dark exciton some of the bright one’s ability to emit, so it appears as a new peak that grows with the field – in WSe2 about 40 meV below the bright exciton. The weak emission of the spin-dark exciton travels along the plane rather than out of it, so it also shows up when light is collected from the side, through a lens with a very wide aperture, or through surface on a nearby metal, which couple to exactly that direction. In clean, hBN- samples at low temperature the dark peak is sharp enough to see directly.
Why they matter
Lying lowest and living long, dark excitons act as a reservoir: bright excitons relax into them, which is why tungsten-based layers glow more weakly when cold – the trace that first gave them away in WSe2. Their long lives make them candidates for storing valley information, and dark excitons trapped at strained spots are one of the explanations proposed for the found in WSe2.
For specialists
An exciton whose radiative recombination is forbidden or strongly suppressed: spin-forbidden, with electron and hole in conduction and valence sub-bands of opposite spin in the same valley, or momentum-forbidden, with electron and hole in different valleys so that a phonon is needed. In W-based TMDC monolayers the spin-split conduction band puts the spin-dark state roughly 40 meV below the bright A exciton; it carries a weak out-of-plane dipole and lives far longer than the bright state. It is made visible by in-plane magnetic fields that mix the spins, by coupling to surface plasmons, or by collecting light emitted along the plane.