Improved QE in CMOS image sensors with nano-black antireflection layer

Prest, Martin J.; Setälä, Olli; Stefanov, Konstantin D.; Vähänissi, Ville; Savin, Hele and Jordan, Douglas (2023). Improved QE in CMOS image sensors with nano-black antireflection layer. In: 2023 International Image Sensor Workshop, 21-25 May 2023, Crieff, Scotland.


A novel anti-reflection process is demonstrated which improves the quantum efficiency (QE) of a CMOS image sensor, with particular benefits at the ultraviolet (UV) and near infrared (NIR) ends of the electromagnetic spectrum. Also, the dark current and photoresponse non-uniformity (PRNU) were reduced to about 33% and 55%, respectively, of the values for a conventional control sensor. The nano-black anti-reflection layer was made using a reactive-ion-etch technique to form nano-scale spikes at the surface which greatly reduce the reflectivity of the surface, which has a matt-black appearance. The sensor used, a CIS115 from Teledyne-e2v, is a back-sideilluminated (BSI) device with ≈10 µm active silicon thickness and 2000 X 1504 pinned photodiode pixels with a pitch of 7 µm. The improved QE is most impressive at UV wavelengths, below 400 nm, where the QE increases towards 100%, although no correction was made for an increased electron generation rate, as this is not easily quantified. This high QE result is compared with a conventional antireflection (AR) coating which shows a steep drop in QE below 400 nm. There is also an improvement in QE in the NIR (from 700 nm to 1100 nm) for the nano-black sensor, and this is despite the approx. 1 µm thinning of the silicon by the etching process, which would normally reduce the QE. Some of the QE improvement may be the result of increased scattering of the incident light, which is supported by the reduced PRNU.

Viewing alternatives

Download history

Item Actions