Luminescent Emission of Excited Rydberg Excitons from Monolayer WSe₂

Abstract

We report the experimental observation of radiative recombination from Rydberg excitons in a two-dimensional semiconductor, monolayer WSe2, encapsulated in hexagonal boron nitride.Excitonic emission up to the 4s excited state is directly observed in photoluminescence spectroscopy in an out-of-plane magnetic field up to 31 Tesla. We confirm the progressively larger exciton size for higher energy excited states through diamagnetic shift measurements. This also enables us to estimate the 1s exciton binding energy to be about 170 meV, which is significantly smaller than most previous reports. The Zeeman shift of the 1s to 3s states, from both luminescence and absorption measurements, exhibits a monotonic increase of -factor, reflecting nontrivial magnetic-dipole-moment differences between ground and excited exciton states. This systematic evolution of magnetic dipole moments is theoretically explained from the spreading of the Rydberg states in momentum space.

Revealing exciton masses and dielectric properties of monolayer semiconductors with high magnetic fields

“…The exciton g -factors are all close to (see Fig. 1f), in line with earlier studies of the exciton in WS monolayers 31,33,34 and in accordance with the similar Zeeman splittings of all the Rydberg excitons that was recently reported in WSe monolayers 27,35 .…”

Exciton g factors of van der Waals heterostructures from first-principles calculations

“…However, the convergence is slow. In the literature on TMD L nk is sometimes divided into a contribution coming from the atomic orbital (ao) and one from the lattice (l) (or valley) L nk = L ao nk + L l nk and the two contributions are separately discussed [9,11,14,33]. However, this division is only of qualitative nature, as the projection of a Bloch state |nk onto atomiclike orbitals is nonunique and leads to contributions from multiple atomiclike orbitals.…”

“…For the g factors of A and B excitons Eqs. ( 12), (11), and (7) give g 1L A,B = 2g +K = 2( +K + L +K ), where +K and L +K are the difference of the spin and the orbital angular momentum expectation values between conduction and valence band, respectively. Figure 1(c) shows that circular polarized light couples valence and conduction band states with the same spin, consequently +K = 0 and only L +K matters.…”

“…Early experimental studies of the magnetic field dependence of excitons, i.e., optical excitations formed by bound electron-hole pairs, in monolayer MoSe 2 observed a Zeeman shift g ≈ −4, which has been attributed to the d-orbital character of the conduction and valence states involved in the excitonic transition [7,8]. However, subsequent studies in WSe 2 [9][10][11] and WS 2 [12] where excitons exhibit the same orbital character, observed slightly larger values, which pointed to possible corrections due to the angular momentum texture of the conduction and valence bands. This picture became even more puzzling when g factors of ≈ 9.5 were experimentally observed for dark exciton states in bilayer WSe 2 [13], and when interlayer excitons in heterobilayers of TMD where demonstrated to have g factors of ≈ 6.7 and ≈ −16 [14], which deviate even more from the value expected for ground state excitons in TMD.…”

Valley-selective chiral phonon replicas of dark excitons and trions in monolayer WSe2

“…They have direct band gap at two time-reversal valleys (K, K') [2,3], where spin-orbit coupling (SOC) splits each band into two subbands with opposite spins [4][5][6]. The electrons and holes can form tightly bound excitons at each valley [7][8][9][10][11][12][13][14]. If they come from bands with the same electron spin, they form bright excitons with efficient radiative recombination [ Fig.…”