Abstract
We report on the IR sensitivity enhancement of back-illuminated CMOS Image Sensor (BI-CIS) with 2-dimensional diffractive inverted pyramid array structure (IPA) on crystalline silicon (c-Si) and deep trench isolation (DTI). FDTD simulations of semi-infinite thick c-Si having 2D IPAs on its surface whose pitches over 400 nm shows more than 30% improvement of light absorption at λ = 850 nm and the maximum enhancement of 43% with the 540 nm pitch at the wavelength is confirmed. A prototype BI-CIS sample with pixel size of 1.2 μm square containing 400 nm pitch IPAs shows 80% sensitivity enhancement at λ = 850 nm compared to the reference sample with flat surface. This is due to diffraction with the IPA and total reflection at the pixel boundary. The NIR images taken by the demo camera equip with a C-mount lens show 75% sensitivity enhancement in the λ = 700–1200 nm wavelength range with negligible spatial resolution degradation. Light trapping CIS pixel technology promises to improve NIR sensitivity and appears to be applicable to many different image sensor applications including security camera, personal authentication, and range finding Time-of-Flight camera with IR illuminations.
Highlights
In the past decade, crystalline silicon (c-Si) based CISs have rapidly replaced CCD image sensors in most commercial camera applications, due to their superiority in power consumption and higher flexibility readout
We report on the NIR sensitivity enhancement of back-illuminated CMOS Image Sensor (BI-CIS) with the inverted pyramid array (IPA)
Lateral crosstalk is suppressed with the use of low refractive index deep trench isolations (DTI) between pixels
Summary
C-Si based CISs have rapidly replaced CCD image sensors in most commercial camera applications, due to their superiority in power consumption and higher flexibility readout. The most straightforward approach to improve NIR sensitivity is to make the photo absorption layer thicker This degrades the visible image quality due to the lateral crosstalk especially for small pixel sizes. It is well known that black silicon, which has random needle-shaped surface, has very low reflectivity and high absorption efficiency[7] These structures have been applied to c-Si solar cells and image sensors to improve photo-electro conversion efficiencies and sensor sensitivities[8, 9]. This means that each pixel contains random but limited number of structures, and www.nature.com/scientificreports/ It results in large deviation of pixel characteristics such as PRNU (photo response non-uniformity), which is not good for high quality, uniform 2D imaging capabilities. We propose the novel pixel architecture of BI-CISs with IPA structure on c-Si surface plus deep trench isolations (DTI) between pixels to enhance NIR sensor sensitivity
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
