Gold-Hyperdoped Germanium with Room-Temperature Sub-Band-Gap Optoelectronic Response

Short-wavelength-infrared (SWIR; 1.4–3.0 µm) photodetection is important for various applications. Inducing a low-cost silicon-compatible material, such as germanium, to detect SWIR light would be advantageous for SWIR applications compared with using conventional (III-V or II-VI) SWIR materials. Here, we present a scalable nonequilibrium method for hyperdoping germanium with gold for dopant-mediated SWIR photodetection. Using ion implantation followed by nanosecond pulsed laser melting, we obtain a single-crystal material with a peak gold concentration of 3 × ${10}^{19}{\mathrm{cm}}^{-3}$ $({10}^3$ times the solubility limit). This hyperdoped germanium has fundamentally different optoelectronic properties from those of intrinsic and conventionally doped germanium. This material exhibits sub-band-gap absorption of light up to wavelengths of at least 3 µm, with a sub-band-gap optical absorption coefficient comparable to that of commercial SWIR photodetection materials. We show that germanium hyperdoped with gold exhibits sub-band-gap SWIR photodetection at room temperature, in contrast with previous doped-germanium photodetector studies, which only show a low-temperature response. This material is a potential pathway to low-cost room-temperature silicon-compatible SWIR photodetection.

Figure 1

Structural properties of hyperdoped $\mathrm{Ge}$. (a) Schematic of the hyperdoping process: (i) ion implantation introduces the desired dopant quantity into the host material, (ii) pulsed laser melts implantation-induced lattice damage, and (iii) epitaxial rapid solidification traps a supersaturated dopant concentration into single-crystalline phase. (b) RBS spectra of virgin $\mathrm{Ge}$ and implanted high-dose (6 × 10¹⁴ cm⁻²) $\mathrm{Ge}$ before and after PLM. All spectra shown are channeled measurements, except the random PLM spectra indicated and its simulation. (c) Inferred concentration-depth profiles of low- (10¹⁴ cm⁻²) and high-dose (6 × 10¹⁴ cm⁻² dose) hyperdoped $\mathrm{Ge}$ samples. (d) XTEM and high-resolution (HR) XTEM (inset) micrographs of the high-dose hyperdoped $\mathrm{Ge}$ sample.

Figure 2

Sub-band-gap absorptance of hyperdoped $\mathrm{Ge}.$ (a) Difference between the sub-band-gap absorptance A (absorptance = 1–transmittance–reflectance) of hyperdoped $\mathrm{Ge}$ samples (processed with a 0.49 J/cm² $\mathrm{Nd}:\mathrm{YAG}$ laser fluence) and that of the virgin $\mathrm{Ge}$ wafer, A_Ge:Au–A_Ge, measured with a UV-NIR spectrophotometer and FTIR. Crossover between spectrophotometer and FTIR spectrometer measurements is at 2.5 µm. Low-dose (10¹⁴ cm⁻², red line) sample shows no sub-band-gap absorptance, whereas high-dose (6 × 10¹⁴ cm⁻², blue line) sample shows significant sub-band-gap absorptance. (b) Calculated $\alpha d$ product (left axis) and absorption coefficient (right axis) for the high-dose hyperdoped sample. In estimating the absorption coefficient $\alpha$, the absorber layer thickness, d, is approximated as 50 nm.

Figure 3

Room-temperature sub-band-gap optoelectronic response of hyperdoped $\mathrm{Ge}$. (a) Schematic of the cw-laser photoconductivity measurement setup used to measure the optoelectronic response of the high-dose (6 × 10¹⁴ cm⁻²) hyperdoped $\mathrm{Ge}$ photodetectors at 2.0 and 2.44 µm. CH in the diagram stands for chopper. (b) Hyperdoped $\mathrm{Ge}$ photodetector’s dark I-V curve. (c) Difference between the hyperdoped $\mathrm{Ge}$ photodetector’s 2.0-µm-laser illuminated and dark I-V curves. (d) Measured room-temperature normalized photoresponse of a zero-biased hyperdoped $\mathrm{Ge}$:$\mathrm{Au}$ photodetector and $\mathrm{Ge}$:$\mathrm{Sb}$ control photodetector also made with implantation and PLM. Response is obtained from substituting each photodetector for the detector in a FTIR spectrometer and illuminating with Globar light. Response of each detector is normalized to the intensity spectrum of the FTIR Globar light.