Charge Carrier Lifetime and Trap States in Novel Pure-Boron-Based Ultrashallow <i>p</i> – <i>n</i> Junctions

Trap states and carrier transport parameters are critical to the performance of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">p</i>–<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">n</i>-junction-based photodetectors. This study investigates ultrashallow <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">p</i>–<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">n</i> junctions for ultraviolet and low-energy particle detection fabricated using chemical vapor deposition-based pure-boron deposition (CVD-PB) with a junction depth only tens of nanometers, in comparison, also using ion implantation (IMP), epitaxy (EPI), and vacuum evaporation (VEP). Deep-level photothermal spectroscopy (DLPTS) and homodyne photocarrier radiometry (HoPCR) were used to analyze charge carrier dynamics, including charge carrier lifetime and trap states. A phenomenological theoretical model for HoPCR signals was developed to explain trap-state-modulated carrier transport dynamics. CVD-PB junctions demonstrated long recombination lifetimes in the doping layer (7.09 ± 0.01 μs) and bulk (35.80 μs), comparable to high-energy ion-implanted (HE-IMP) junctions (5.68 ± 0.58 and 36.10 μs). EPI and VEP junctions have short carrier bulk recombination lifetimes (1.95 ± 0.17 and 0.63 ± 0.03 μs, respectively). DLPTS revealed an electron trap in the EPI junction with an activation energy of 0.09 eV and a capture cross section of 7.52 × 10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-<roman>20</roman></sup> ± 6.82 × 10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-<roman>21</roman></sup> cm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><roman>2</roman></sup>; no detectable traps were observed in CVD-PB or HE-IMP junctions. Simulations showed that EPI junctions had twice the dark current of CVD-PB and HE-IMP at 10 V, with a 15% reduction in photocurrent. When reversely biased at 5 V and without typical guarding ring designs generally used in optoelectronics, CVD-PB exhibited the lowest dark current (0.95 μA/cm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><roman>2</roman></sup>) when compared to HE-IMP (3.25 μA/cm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><roman>2</roman></sup>) and EPI (51.77 μA/cm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><roman>2</roman></sup>).