Thickness dependence of the superconducting critical temperature in heavily doped Si:B epilayers

Abstract

We report on the superconducting properties of a series of heavily doped Si:B epilayers grown by gas immersion laser doping with boron content (n B ) ranging from ∼3 × 10 20 cm −3 to ∼6 × 10 21 cm −3 and thickness (d) varying between ∼20 nm and ∼210 nm. We show that superconductivity is only observed for n B values exceeding a threshold value (n c,S ) which scales as n c,S ∝ 1/d. The critical temperature (T c ) then rapidly increases with n B , largely exceeding the theoretical values which can be estimated by introducing the electron-phonon coupling constant (

Characterization of neutral boron-silicon clusters using infrared spectroscopy: The case of Si6B

“…Interestingly, similar effects have also been predicted for small B-doped silicon clusters [19]. Efforts have been made to increase Tc in B-doped silicon [16,20,21].…”

Superconductivity in doped semiconductors

“…However, the simulations failed to reproduce the values of TC when the free hole concentration nh was reduced down to nMIT, as well as the curvature of the experimental TC vs nh variations 34,35,39,40 . Rescaling the coupling and screening parameters and µ* according to power laws close to the critical concentration has been proposed for diamond 40 and silicon 41 but remains specific and unsatisfactory. More general descriptions 42,43 of the effect of approaching the MIT in disordered metals yield TC enhancements which have not been detected in the materials under study here.…”

“…Despite the notorious incapacity of BCS theory to predict new superconductors 54 , experimental evidence for superconductivity was gathered over the next few years, first in silicon 55 , then in silicon carbide 56 , both upon boron doping in the few at. % range, with TC soon 57 reaching 1.5 K in 3C-SiC:B and 6H-SiC:B (cubic 3C and hexagonal 6H are two of the many polytypes of silicon carbide), and more recently 41 0.7 K in Si:B. Ab initio calculations in the spirit detailed above were performed about superconductive Si:B, resulting in values which remained 41,55 below 0.45 even for 10 at.% B. Other dopants of silicon have also been investigated 58 , the calculations pointing out the detrimental influence on the electron-phonon coupling strength of the lattice contraction brought about by atomic substitution by impurities with a smaller covalent radius, so that substitutional doping (ntype) of silicon with aluminum was seen theoretically as a favorable (but notoriously unpractical) option.…”

“…As a matter of fact, such apparent densities have been found to be larger than the chemical concentration of the doping element by 50% in 9 SnTe, and by up to an order of magnitude in diamond 27,40 . A numerical application of equation 1 to the case of silicon (D = 640 K) using a boron concentration-dependent electron-phonon coupling parameter  evaluated by ab initio supercell calculations 41 is also given in fig. 1 to illustrate the shortcomings of such a simple approach when it comes to predicting TC variations over a wide range of concentrations.…”

Superconductivity observed in platinum-silicon interface

“…[1][2][3][4][5][6][7][8] It has been discussed theoretically that a superconducting state may exist in semiconductors since 1964. 9 Recent theoretical studies found that the lattice distortion may play an important role in the superconductivity of cubic silicon, 10 and that a dopant concentration threshold exists, above which B-doped Si becomes superconducting.…”