1 The Detector Seminar at CERN Recent Progresses of Visible Light Image Sensors February 23, 2018 Nobukazu Teranishi University of Hyogo / Shizuoka University ©2018 N. Teranishi 2 Contents 1. Basics of Visible Light Image Sensors 2. Dark Current and Blemish 3. Pinned Photodiode (PPD) Recent Progresses 4. Photon Counting Image Sensor 5. Stack 6. Near InfraRed (NIR) 7. ToF (Time of Flight) 8. Polarization Image Sensor 9. Monocular 3D Image Sensor 10. Queen Elizabeth Prize for Engineering ©2018 N. Teranishi 3 What is Image Sensors? - Semiconductor device, which converts light image to electric signal. - Used in cameras. Examples of applications Smart- Movie Automobile Endoscope Iris verification Security phone Light Image Sensor Microlens Color filter ● ● Photodiode Electron Image sensor Pixel array DSC ©2018 N. Teranishi Pixel cross-section Image Sensor (IS) Market 4 - IS sales amount has grown by camera phone. Camcoder PMP - IS spreads into various applications, DSLR Medical Smart glasses - “Others” includes scientific, Compact DSC Broadcast industrial, … Game Automobile Others Surveillance Bpcs PC/WEB camera 4 CMOS image sensor CCD image sensor Tablet 3 2 Camera phone 1 Sales amount 0 02 04 06 08 10 12 14 Application in amount (2015) 01 03 05 07 09 11 13 15 Year ©2018 N. Teranishi (Source: TSR) 5 Pixel Structure microlens (ML) Color filter (CF) Amplifier (SF) Sotrage Pinning layer Transfer gate (TG) + P+ N+ + P e- P e- Silicon Ne- P Detective capacitance (Floating diffusion (FD)) Cross-section of frontside illumination type CMOS image sensor pixel ©2018 N. Teranishi Sensitivity Expansion to Invisible 6 Inner Shell Excitation Large depletion is needed. Small surface dead layer is X-ray Near UV needed. Visible light ©2018 N. Teranishi Based on (Piet De Moor (IMEC)) 7 Contents 1. Basics of Visible Light Image Sensors 2. Dark Current and Blemish 3. Pinned Photodiode (PPD) Recent Progresses 4. Photon Counting Image Sensor 5. Stack 6. Near InfraRed (NIR) 7. ToF (Time of Flight) 8. Polarization Image Sensor 9. Monocular 3D Image Sensor 10. Queen Elizabeth Prize for Engineering ©2018 N. Teranishi Classification of Dark Current and Blemish 8 Dark current FPN Average: Dark current (Dark current pixel-by-pixel fluctuation) ・Dark current shot noise ・Black level shift by temperature Micro white blemish (Mid-range) White blemishes (Large-range) Pixel Counts Pixel 3σ Very large level blemish by radiation damage Dark Signal Level ©2018 N. Teranishi 9 Metal Contamination (1) ・Metals, especially transition metals, form mid-gap levels, which become GR centers. ・Generate dark current based on Shockley-Read-Hall (SRH) process 2 pn ni U: Recombination Rate U vthNt Et Ei (Sze: “Semiconductor Devices” n p 2ni cosh Chap. 1 Eq.(59)) kT Assuming the cross sections, n p Conduction band Nt: GR center (trap) density ・Assuming depletion condition, then, n, p = 0, Et and Et=Ei because U becomes the largest; v n N U th i t (Unit : 1/cm3s) 2 Valence band v n ・One GR center generates U th i , (Unit : 1/s) Dark current by 2 mid-gap level ©2018 N. Teranishi Metal Contamination (2) 10 400 275 K a a a 20 Frames Two series of specific and periodic 0 1 60 s integration 300 b b b peaks, labeled as “a” and “b”. Poisson distribution 200 Assuming metals are distributed 2 as Poisson distribution, metal 100 Number of counts Number 3 density per pixel can be derived. 0 0 1000 2000 3000 4000 5000 Signal (e-) σ s are derived from the pitches σ(cｍ2) Density(1/pixel) Density(cm-3) using the previous slide formula. a 1.8E-15 1.25 1.3E9 b 5.4E-16 0.15 1.5E8 ・ “Dark Current Spectroscopy” identifies metals and estimates concentrations. ・ It can measure very small amounts of metal contamination. (D. McGrath et al. (TI); “Counting of Deep-Level Traps Using a Charge-Coupled Devices”) ©2018 N. Teranishi 11 Contents 1. Basics of Visible Light Image Sensors 2. Dark Current and Blemish 3. Pinned Photodiode (PPD) Recent Progresses 4. Photon Counting Image Sensor 5. Stack 6. Near InfraRed (NIR) 7. ToF (Time of Flight) 8. Polarization Image Sensor 9. Monocular 3D Image Sensor 10. Queen Elizabeth Prize for Engineering ©2018 N. Teranishi PPD (Pinned PD) Structure and Advantages 12 TG FD Pinning 1. Grounded P+ pinning layer Layer x x x+ xx prevents interface to be + P N + P N P depleted, and stabilizes PD electrically. P-Well ・Low dark current Storage N-Substrate ・Large saturation ・High blue sensitivity x: GR (generation-recombination) center 2. Complete electron transfer 0V OFF ・No image lag, ・No transfer noise ON Signal Shallow P+ pinning layer Potential (Low energy implantation) ⇒ ©2018 N. Teranishi Good electron transfer 13 Dark Current Reduction by PPD - Theoretical dark current reduction ratio using Shockley-Read-Hall Process: U Not depleted v N n 2 p th t i ~ 10 7 2 U Depleted vth Nt ni p 2ni Assuming hole density (p) at the P+ pinning layer is 1017 h+ cm-3 - Actual dark current reduction ratio: Non-PPD (1982) PPD (2012) Unit Scheme CCD FSI CMOS Pixel size 23 x13.5 1.12 x 1.12 μm Dark current 1,300 5.6 e-/s/μm2 at 0.4 % 60℃ Still big reduction! ©2018 N. Teranishi Example of Dark Current Reduction by PPD 14 Conventional PD PPD If Dark current is reduced, dark current FPN, and dark current shot noise are also reduced. Dark current FPN is suppressed, then, picture quality is much improved. ©2018 N. Teranishi Saturation is Increased by PPD Saturation becoms12 times larger by PPD. ・ PN junction area is increased. ・ Acceptor density at pinning layer >> Donor density at storage ) 2 Pinning layer 10 /cm 11 8 10 x xx - x x P + 6 + N+ + (e 12.5 times PPD P e- P 4 - e N - e 2 Non-PPD Storage P region 0 Saturation 0 1 フォトダイオード電位2 3 4 PPD pixel cross-section Storage potential （V) Saturation by simulation 1984 B. C. Burkey et al.（Kodak) ©2018 N. Teranishi Pixel Shrinkage Trend 16 ・ Shrinkage has been carried out steadily. ・ Lens volume and weight become 1/1000 when pixel size becomes 1/100, assuming that F-number, view angle and pixel number are constant, (um2) CCD 24 years CMOS 100 1/100 10 50％ shrinkage Minimum Pixel Area in 3.5 years 1 80 85 90 95 00 05 10 15 Mass Production Year ©2018 N. Teranishi 17 Contents 1. Basics of Visible Light Image Sensors 2. Dark Current and Blemish 3. Pinned Photodiode (PPD) Recent Progresses 4. Photon Counting Image Sensor 5. Stack 6. Near InfraRed (NIR) 7. ToF (Time of Flight) 8. Polarization Image Sensor 9. Monocular 3D Image Sensor 10. Queen Elizabeth Prize for Engineering ©2018 N. Teranishi 18 Photon Counting Image Sensor SPAD 4-Tr CMOS + High conversion gain + (Single photon avalanche diode) CMS (Correlated multiple sampling) n=1 n=2 n=3 n=0 n=4 128 samplings n=5 n=6 n=7 (N. Dutton et al., VLSI Symposium 2014) ・ QVGA SPAD, 20 fps, room temperature (MW. Seo, S. Kawahito et al., IEEE EDL 2015) ・ In 2015, several organization reported ・ High avalanche gain makes following low noise < 0.3 e- rms. circuit noise negligible. ・ DEPFET (Max Plank) ・ Large dark count. Small fill factor ©2018 N. Teranishi Photon Counting (2) 19 (Seo, Kawahito et al, IEEE EDL, Dec. 2015) ©2018 N. Teranishi 20 Contents 1. Basics of Visible Light Image Sensors 2. Dark Current and Blemish 3. Pinned Photodiode (PPD) Recent Progresses 4. Photon Counting Image Sensor 5. Stack 6. Near InfraRed (NIR) 7. ToF (Time of Flight) 8. Polarization Image Sensor 9. Monocular 3D Image Sensor 10. Queen Elizabeth Prize for Engineering ©2018 N. Teranishi Stack (1) 21 Motivation of stack image sensors Sensor - Non-silicon material for sensor part Bump - Finer process technology for logic part - Smaller footprint - More flexibility of wiring Readout circuit - Isolation between analog and digital circuits Bump bonding Approach - Bump bonding (In, solder Au, …) - SOI - Wafer-to-wafer oxide bonding + TSV - Wafer-to-wafer hybrid bonding SOI pixel detector (KEK Arai et al.) ©2018 N. Teranishi 22 Stack (2). Wafer-to-Wafer Oxide Bonding Wafer-to-wafer oxide bonding + TSV(Through silicon via) at peripheral ・Smaller chip size → Smaller camera, lower cost ・Optimal processes for both sensor and logic ・More functions can be integrated. Suitable for large volume applications Conceptual Diagram SEM Cross-section ©2018 N. Teranishi (S. Sukegawa et al., ISSCC 2013) 23 Stack (4). Wafer-to-Wafer Hybrid Direct Bonding ・ Oxide bonding ・ Hybrid bonding + TSV(Through silicon via) Interconnection can be located Interconnections are at pixel array also. limited at only peripheral. Cu2Cu optimized CMP < 150 ℃ 400 ℃ BEOL CMP K. Shiraishi et al. (Toshiba), ISSCC 2016 CMP optimization Post-bond annealing ・ 2 um pitch interconnection Y. Kagawa et al. (Sony), IEDM 2016 E. Beyne (IMEC), 2018 ©2018 N. Teranishi 24 Contents 1. Basics of Visible Light Image Sensors 2. Dark Current and Blemish 3. Pinned Photodiode (PPD) Recent Progresses 4. Photon Counting Image Sensor 5. Stack 6. Near InfraRed (NIR) 7. ToF (Time of Flight) 8. Polarization Image Sensor 9. Monocular 3D Image Sensor 10. Queen Elizabeth Prize for Engineering ©2018 N. Teranishi 25 NIR --- Motivation １．To cover shortage of visible light and to increase sensitivity eg. Most illuminations have NIR. Night glow ・ Security use ２．Invisible for mankind and animal. ・ Gesture interface, ToF, ・ Animal observation at night ３．Eye safe laser (1.4-2.6um) ・ Longer than Si cutoff wavelength ４．Specific wavelength for each application ・ Window to living organisms ・ Fluorescent protein ・ Hemoglobin (Hb) ・ Bill validator (NIR and UV inks are used.) 5. Smaller sunlight background ©2018 N. Teranishi Nightglow (Airglow) ・ Emission in the upper atmosphere - Photoionization by the sun - Luminescence by cosmic ray - Chemiluminescence ・ Light intensity at near IR is larger than that of visible light Intensity of night glow Night glow from ISS http://en.wikipedia.org/wiki/Nightglow ©2018 N.