Public deliverable for the ATTRACT Final Conference Public deliverable for the ATTRACT Final Conference Superior gamma-detection and IR imaging via ALD-passivated germanium nanostructures (SUGER) Hele Savin 1* , Toni P. Pasanen 1 , Joonas Isometsä 1 , Kexun Chen 1 , Ville Vähänissi 1 , Rais Nurgalejevs 2 , Olga Savina 2 , Vladimir Gostilo 2 , Artto Aurola 3 1 Aalto University, School of Electrical Engineering, Department of Electronics and Nanoengineering, Tietotie 3, 02150 Espoo, Finland 2 Baltic Scientific Instruments Ltd., Ramulu str. 3, LV-1005, Riga, Latvia 3 Pixpolar Oy, Otakaari 5, 02150 Espoo, Finland *Corresponding author: hele.savin@aalto.fi ABSTRACT The SUGER is based on several innovative approaches that are implemented in photodiode manufacturing to reach record-high device performance in near-infrared and gamma-ray detection. The approaches mitigate electrical and optical losses and are fully CMOS compatible. The obtained results include external quantum efficiency >130 % in UV, surface recombination velocity <5 cm/s and absorbance >99 % in UV-NIR. In ATTRACT Phase 2, we plan to partner with experts from different parts of the detector value chain to scale up our technology to industrial system-level prototypes and even to commercial products. Keywords: near-infrared, gamma radiation, defect passivation, high absorption, high sensitivity 1. INTRODUCTION Near-infrared (NIR, wavelength 750–3000 nm) and gamma-ray sensors have a vital role in several application areas that are critical for our everyday life, including medical diagnostics, telecommunications and security. However, the current sensors have limited performance in terms of sensitivity, wavelength range and cost. There is hence an urgent need to develop sensor technologies that overcome these shortcomings. In the SUGER project, we have three innovative ideas for reaching ground-breaking sensor characteristics. First, instead of using externally doped pn-junctions that are expensive to fabricate and result in severe electrical losses, we propose a concept where the charge collection is realised via a dopant-free inversion layer. This kind of junction enables collection of the signal with very little electrical losses resulting in superior sensitivity. Furthermore, such concept brings a great asset for gamma-ray sensors that currently require complex multi- sided ion implantation for junction formation. Second, to address the major recombination losses present at the surface and interfaces of the state-of-the-art detectors, our idea is to control the presence of charge-carriers at the surfaces using charged dielectrics and take advantage of the properties of atomic layer deposition (ALD). Third, instead of depositing a separate antireflection coating that is only working for a single wavelength, our idea is to fabricate a graded refractive-index interface that is based on surface nanostructures. This should result in fully absorbing surface at wide range of wavelengths and acceptance angles as well as provide efficient light trapping paths inside the substrate. In this project we have applied the above ideas to CMOS compatible semiconductor substrates (Ge and Si). As a result, we have achieved record high (>99 %) absorbance in the whole UV-VIS-NIR spectrum up to 1600 nm wavelength using industrially up-scalable methods. Secondly, we have demonstrated that our novel idea of dopant-free junction is feasible and have achieved high performance induced junctions both in Ge and Si. Thirdly, we have succeeded in reducing the recombination losses drastically at the detector surfaces using ALD technology and as a result reached surface recombination velocities (SRV) below 1 cm/s and 5 cm/s in Si and Ge, respectively. Finally, after applying all these ideas into the final device, we achieved >130 % external quantum efficiency in deep UV, which is the highest performance ever achieved with a single photodiode. 2. STATE OF THE ART State of the art (s-o-t-a) NIR sensors are typically made of InGaAs. However, their fabrication is expensive, they have limited wavelength range and the material is not CMOS compatible. Hence, InGaAs has very limited potential to enable ground-breaking solutions for the