Optimizing dot-in-a-well infrared detector architecture for achieving high optical and device efficiency corroborated with theoretically simulated model

Dot-in-a-well (DWELL) heterostructures have been extensively researched in quantum dot based infrared photo-detectors as it offers tuning of detection peak wavelength, low dark current and a higher operating temperature with optimized quantum well thickness. In this paper, we have correlated experimentally observed opto-electronic properties of three different InAs quantum dots in DWELL configuration with varying ternary capping (In0.15Ga0.85As - Strain reducing ternary alloy) thickness from 4 to 8 nm to the proposed simulation model for achieving high efficiency in device performance. Low temperature (8 K) photoluminescence (PL) spectra exhibits a blue shift of around 24 nm along with a decrease in intensity as the capping thickness increases above 6 nm. Decrease in full width at half maximum (FWHM) value for PL peak was observed as the capping thickness is increased from 6 to 8 nm. These are attributed to formation of InGaAs quantum wells via dots sublimation process. Presence of InGaAs wells for 8 nm capped sample was confirmed using low temperature photoluminescence measurement at 2.54 W/cm2 and photoluminescence excitation (PLE). Time resolved photoluminescence spectroscopy gave further insight into the carrier dynamics of the grown structures and confirms undesirable quantum well formation in 8 nm capped sample. Improved optical characteristics with formation of defect-free structure having larger quantum dot was achieved in 6 nm capped sample affirmed using transmission electron microscopy images. Low dark current density with high confinement energy was achieved from 6 nm capped DWELL structure. Dominant spectral response peak was obtained at 7.56 μm from all the samples. A concentration-dependent theoretical strain model using the Schrödinger equation was developed to calculate the potential, ground-state and inter-sub band energy-levels. It was validated by comparing experimentally achieved PL and PLE peaks along with spectral response peaks. The theoretical model was used to calculate the energy levels corresponding to conduction and valence bands for InGaAs well and indium concentration in the well was obtained to be around 30%. Similarly, concentration dependent 2D to 3D transition for InxGa1-xAs (0 ≤ x ≤ 1) quantum dots formation was modelled using People-Bean relation which matches with the proposed simulation.