On the Optoelectronic Mechanisms Ruling Ti‐hyperdoped Si Photodiodes

Abstract

room temperature to the short-wavelength infrared (SWIR) range (1.4-3 µm, i.e., 0.89-0.41 eV) has the potential to revolutionize silicon-based optoelectronic devices. The introduction of a complementary metaloxide semiconductor (CMOS) compatible process would enable the integration of optical and electronic functionality on a single chip. [1] Nowadays, several approaches are under intensive research to enhance Si SWIR photoresponse, such as the integration of III-V compound semiconductors with silicon, [2,3] photodetectors based on SiGe alloys [4,5] or HgCdTe/Si heterostructures. [6] Despite their high performances, these devices suffer from several drawbacks: they are based on nonabundant or contaminant materials, they usually require cryogenic temperatures to operate, and they are hardly integrated into the very mature Si-CMOS fabrication routes, adding considerable cost and complexity to photonic circuit manufacturing.An alternative path to increase the Si SWIR photoresponse that would avoid the hybrid integration of Si technology with unconventional materials is the direct modification of its electronic band structure. From this perspective it is possible to find interesting approaches. One is the use of a laser-crystallization process to induce an anisotropic tensile stress in silicon optical fibers. This way the bandgap can be reduced from 1.11 to 0.59 eV. [7] Unfortunately, the fiber structure hinders its incorporation into planar imaging arrays. A different method, that has attracted recently great attention, [8] is the extrinsic sub-bandgap photoresponse obtained from the incorporation of deep-level impurities at concentrations far above their equilibrium values. Single-crystalline silicon layers with transition metals (Ti, V, Au, and Ag) or chalcogens (S, Se, and Te) at concentrations above 10 19 cm -3 have been obtained by means of ion implantation and subsecond annealing techniques (in a process denominated as supersaturation or hyperdoping). We have shown that silicon hyperdoped with Ti [9,10] or V [11] presents outstanding properties such as a sub-bandgap optical absorption coefficient in the 10 4 cm -1 range [12] or a photoconductive response extended down to 0.2 eV (6.2 µm) at cryogenic temperatures. [13] Gold hyperdoped [14,15] or silver hyperdoped [16] This work deepens the understanding of the optoelectronic mechanisms ruling hyperdoped-based photodevices and shows the potential of Ti hyperdoped-Si as a fully complementary metal-oxide semiconductor compatible material for room-temperature infrared photodetection technologies. By the combination of ion implantation and laser-based methods, ≈20 nm thin hyperdoped single-crystal Si layers with a Ti concentration as high as 10 20 cm −3 are obtained. The Ti hyperdoped Si/p-Si photodiode shows a room temperature rectification factor at ±1 V of 509. Analysis of the temperature-dependent current-voltage characteristics shows that the transport is dominated by two mechanisms: a tunnel mechanism at low bias and a recombination process in the space charge...

Hyperdoped Crystalline Silicon for Infrared Photodetectors by Pulsed Laser Melting: A Review

“…The research on hyperdoped silicon fabricated by laser irradiation has a history of over 20 years, and many important research results have been made [15,22,33,39,55,[58][59][60][61][62][63][64][65] including the properties of nonequilibrium doping and the additional properties of femtosecond laser irradiation processing. In this section, we will go over the most important study findings from the past few years.…”

“…[15] Copyright 2011, AIP Publishing. Si:Ti [63] >10 21 -0.02% (1500; À) -of silicon-based detectors in the infrared communication wavelengths. [77] As mentioned earlier, sulfur is not the most suitable element for super-doped detectors, yet S-hyperdoped prototype photodetectors have outperformed commercial standards.…”

“…Only under very limited conditions, Eric García‐Hemme et al obtained Ti homogeneously distributed hyperdoped silicon by a 248 nm KrF excimer laser with a 20 ns pulse duration. [ 63 ] The key factors affecting the success or failure of nanosecond laser hyperdoping are still unknown. There is still controversy on this point.…”

High-quality single-crystalline epitaxial regrowth on pulsed laser melting of Ti implanted GaAs

“…Hyperdoping with deep level impurities has been proved to produce infrared photoresponse at room temperature for different systems based on Si. It was achieved by ion implantation followed by pulsed laser melting (PLM): Si:S [8], Si:Au [9], Si:Ti [10], Si:V [11], Si:Te [12] or flash lamp annealing Si:Se [13]. Solar cell prototypes based on Si hyperdoped with Ti have exhibited sub-bandgap quantum efficiency which would agree with the intermediate band formation [14].…”

“…Based on the results obtained on hyperdoping Si, Ge, and GaP [10,15,16], in this work we study the approach of hyperdoping GaAs with Ti by ion implantation followed by pulsed laser melting (PLM). The goal is to obtain high-quality single-crystalline epitaxially regrown layers, which is one of the main challenges faced by this technology [30][31][32][33].…”