Auger lifetimes for excitons bound to neutral donors and acceptors in Si

Abstract

The lifetimes of excitons bound to the neutral donors Li, P, and As and neutral acceptors B, Al, Ga, and In in silicon are determined by time-resolved photoluminescence measurements a t 4.2 K and a strong dependence on the binding energy of the donor or acceptor is found. This dependence can be understood assuming a localized phononless Auger-process to be the main recombination mechanism. A theoretical calculation using approximate wavefunctions is presented and leads to a very good agreement with the experimental results.Die Lebensdauern gebundener Exzitonen an den neutralen Donatoren Li, P und As sowie den Akzeptoren B, AI, Ga und In in Silizium werden bei 4,2 K mit zeibufgeloster Photolumineszenzspektroskopie gemessen. Es ergibt sich eine starke Abhiingigkeit von der Bindungsenergie der entsprechenden Stdrstelle. Man kann diese Abhangigkeit verstehen, wenn man einen lokalisierten, phononenfreien Auger-Prom6 ale wesentlichen Rekombinationsmechanismus annimmt. Eine theoretische Berechnung der Lebensdauer unter Verwendiing angeniiherter Wellenfunktionen ergibt eine sehr gute Ubereinstimmung mit den experimentellen Ergebnissen.

Simultaneous Subsecond Hyperpolarization of the Nuclear and Electron Spins of Phosphorus in Silicon by Optical Pumping of Exciton Transitions

“…The results presented here were made possible by a newly grown crystal of 28 Si purposely doped with 31 P. This n-type sample allowed us to study the D 0 → D 0 X transitions without requiring any above-gap excitation, resulting in dramatically larger polarizations. An electron polarization of 90% and nuclear polarization of 76% were obtained simultaneously in less than a second, in a magnetic field and temperature regime where the equilibrium electron polarization was only ∼2%, and the nuclear polarization ∼ 3 × 10 −3 %.…”

“…Several methods for achieving quantum logic with spin states of the shallow neutral donor (D 0 ) 31 P in 28 Si have been proposed [1,2,3] and the manipulation of electron and nuclear spin coherences have been demonstrated [4], but unsolved challenges include the measurement of single spins and the initialization, or polarization, of these spins. Fortuitously, the isotopic enrichment of 28 Si has another dramatic effect: the linewidths of many optical transitions are drastically reduced [9,10,11,12,13], including those involving 31 P. These narrow transitions have been proposed both for measurement of single spins [12,13,14] and for preferentially populating specific spin states [12,13].…”

“…Preliminary attempts to demonstrate this mechanism yielded relatively small electron and nuclear polarizations [25], due to the fact that until now all samples of 28 Si with sufficiently high enrichment to resolve the donor bound exciton (D 0 X) hyperfine transitions were p-type, with residual boron acceptor concentrations typically ten times the 31 P concentration. At low temperature all donors were therefore ionized (D + ), precluding the observation of D 0 → D 0 X transitions, unless abovegap excitation provided photoneutralization.…”

“…It is based on the selective ionization of donors in a specific hyperfine state by optically pumping donor bound exciton hyperfine transitions, which can be spectrally resolved in 28 Si. Electron and nuclear polarizations of 90% and 76%, respectively, are obtained in less than a second, providing an initialization mechanism for qubits based on these spins, and enabling further ESR and NMR studies on dilute 31 P in 28 Si.…”

Determination of Free Exciton Capture Cross Section of Si:Al by Photoluminescence

“…The capture rate c 0 is calculated according to where vth is the average thermal velocity of FE, R0b/f(q, q') is the ratio of the peak heights of bound exciton (BE) and FE-related luminescence for diminishing excitation density, q and q' are the participating phonons in recombination and Wb/f(q, q') denotes, the ratio of radiative transition rates of BE relative to FE. The nonradiative Auger transition rate of an exciton bound to Al (τ = 76 ns) was taken from [7]. To calculate Wb/f(NP,TO) one needs the intrinsic full width at half maximum (FWHM) of the corresponding peaks.…”