CMOS Monolithic Pixel Sensor M. Battaglia 1,2 , D. Bisello 3 , D. - PowerPoint PPT Presentation

Development of a Radiation Hard CMOS Monolithic Pixel Sensor M. Battaglia 1,2 , D. Bisello 3 , D. Contarato 2 , P. Denes 2 , D. Doering 2 , P. Giubilato 2,3 , T.S. Kim 2 , Z. Lee 2 , S. Mattiazzo 3 , V. Radmilovic 2 1 University of California at Berkeley, Berkeley, CA, USA 2 Lawrence Berkeley National Laboratory, Berkeley, CA, USA 3 University of Padova & INFN Padova, Padova, IT, EU

Rad-Hard CMOS Monolithic Pixels for... 1) High energy Physics CMOS monolithic Pixel Sensors allow high-resolution, Expected doses at the position of the innermost layer low material budget, high area/price ratio sensors. of vertex trackers: hybrid pixels necessary with very high non-ionising doses Commercially available manufacturing technologies. Monolithic Pixels favoured when higher resolution CMOS Pixels that are tolerant to several ~MRad and smaller material budget are important ionising dose and to moderate non-ionising fluences (10 12 -10 13 n eq cm -2 ) could be employed in a wide range Monolithic Pixels resistant to O (10 13 ) n eq cm -2 yr -1 and of scientific applications, from tracking and vertexing O (MRad) could be used as high-resolution outer at HEP experiments to direct imaging in Electron layers, also for an LHC upgrade. Microscopy and beam monitoring. Luminosi Non Ionising Ionising Machine ty fluence dose [cm -2 yr -1 ] [n eq cm -2 yr -1 ] [kRad yr -1 ] 1.4 × 10 14 10 34 LHC 11300 1.4 × 10 15 10 35 Super LHC 71400 1.0 × 10 11 10 34 ILC (500 GeV) 50 1.0 × 10 11 10 34 -10 35 CLIC (3 TeV) 70 6.0 × 10 11 Super Belle 10 35 500 6.0 × 10 12 10 36 Super B O (1000) 8 × 10 27 3.0 × 10 12 STAR @ RHIC 80 2/14 KEK 2009 - Piero Giubilato – Development of a Radiation Hard CMOS MAPS

2) Electron Microscopy 3) Beam monitoring CMOS monolithic pixels are being developed as an Real time monitoring of hadron-therapy beam alternative to CCDs. Advantages: intensity and profile. Single electron sensitivity (direct detection) Secondary electron emission from thin Al foil Excellent Point Spread Function (thin collection Desired granularity of 1 mm at 10 kHz frame rate layer, see picture) ~10 7 -10 9 e - (20 keV) mm -2 s -1 High readout speed (over 100 frame/s desired) Exposures of ~minutes and ~1000 exposures/yr Improved radiation hardness (ELT layout) correspond to 5-10 MRad for typical annual usage Energies of interest ~80-300 keV 500 e - /pixel dynamic range in imaging mode, ~10 6 e - /pixel/s in diffraction mode 10 rad/pixel/s, corresponds to ~ 1 MRad ionizing dose for typical annual usage 200 keV e - in CMOS sensor Metal + Oxide e - beam Epi layer Focal plane detectors for PEEMs systems (10 keV e - ). ... 3/14 KEK 2009 - Piero Giubilato – Development of a Radiation Hard CMOS MAPS

Ionizing radiation effects Non ionizing radiation effects Charge trapping in field oxide and at the interfaces. Leakage current in diode from bulk damage. Sensitive region around the charge collecting diode Sensitive region: shallow depleted region in the (electron accumulation, charge losses). vicinity of the charge collecting diode, may lead to significant diode leakage current 3T cell Low-resistivity epitaxial layer: type inversion expected only at high fluencies. Transistor leakage current may be an important contribution, especially for the reset node. Displacement damage threshold for e - on Si is about 260 KeV for free defects production, 5 MeV for clusters generation. 4/14 KEK 2009 - Piero Giubilato – Development of a Radiation Hard CMOS MAPS

The LDRD2-RH chip Developed in the frame of LBNL Laboratory Directed Research & Development (LDRD) grant. Manufactured in AMS 0.35 μm CMOS-OPTO (optimized low leakage current, 5 metal layers)process, with 14 μm nominal epitaxial layer thickness. 96x96 pixels, 20x20 μm 2 , arrayed in several sub-sectors implementing different transistor layouts and different configurations of the charge collection PO NW GR diode. Simple 3-transistor (3T) pixel architecture. 2 sub-pixels with and without ELT layout. Serial readout, up to 25 MHz clock ~70 e - ENC frequency. GR layout NW layout PO layout n-well diode with n-well diode with p+ n-well diode with p+ enclosing p+ guard-ring guard-ring and thin rings, thin oxide on top oxide on top and polysilicon ring 5/14 KEK 2009 - Piero Giubilato – Development of a Radiation Hard CMOS MAPS

200 keV electrons irradiation Performed at the LBNL National Center for Electron Chip covered with standard EM Microscopy (NCEM). gold mesh (25 µm wire, 40 µm hole) during irradiation. 200 keV electrons are expected to cause only ionising damage in Si (thr. energy for DD is 260 keV). After irradiation, the increase of Electron flux of ~2300 e - μm -2 s -1 ~9x10 5 e - /pixel/s (e.g. leakage current in the exposed pixels gives a latent image of the diffraction mode). mesh wires. Irradiation performed in steps up to a total dose of Measurement of PSF ~30 μm 1.11 MRad . Dark levels monitored as dose function. possible, but e - scattering on Tests performed at 6.25 MHz readout frequency → mesh borders spoils the actual 737 μs integration time. figure. Leakage current vs. dose for pixels with ELT layout No n-well PO NW GR diode 20 ADC = 100fA 6/14 KEK 2009 - Piero Giubilato – Development of a Radiation Hard CMOS MAPS

Charged particle detection tests after 200 keV e - irradiation Cluster signal Cluster noise Pre irradiation Pre irradiation Post irradiadion Post irradiadion Irradiation Before After (5.57 ± 0.04) keV (5.54 ± 0.04) keV Energy deposited in a 3x3 pixel matrix measured 200 KeV e - Landau m.p.v. with 200 keV and 1.5 GeV e - (from LBNL ALS) before (1.19 ± 0.04) keV (0.98 ± 0.03) keV Gaussian noise and after 1.11 Mrad on the PO pixels. 3 × 3 matrix noise (1.22 ± 0.20) keV (1.11 ± 0.20) keV Energy spectrum fitted with Landau convoluted with Gaussian distribution to represent the noise. (3.05 ± 0.03) keV (3.37 ± 0.07) keV Landau m.p.v. 1.5 GeV e - Change in gain of ~1.35. (0.60 ± 0.06) keV (0.81 ± 0.07) keV Gaussian noise Good agreement between distributions after gain 3 × 3 matrix noise (0.74 ± 0.01) keV (0.88 ± 0.01) keV correction. 7/14 KEK 2009 - Piero Giubilato – Development of a Radiation Hard CMOS MAPS

29 MeV proton irradiation Leakage current Noise in PO pixels 29 MeV protons , flux of ~3x10 8 pcm -2 s -1 , irradiation Among all ELT designs, best performance is obtained in steps up to a total fluence of 8.5x10 12 p/cm 2 (~2 with PO pixels; noise of ELT cells only slightly MRad). increases at the highest doses while noise of standard cells doubles after ~1 MRad . Dark level monitored in-between irradiation steps: at equal doses, a larger leakage current is associated All irradiations performed at the BASE Facility of the with proton irradiation, hinting at a contribution LBNL 88-inch Cyclotron; all tests performed at room from displacement damage. temperature. 8/14 KEK 2009 - Piero Giubilato – Development of a Radiation Hard CMOS MAPS

14 MeV neutron irradiation 55 Fe spectrum in PO pixels Noise in PO pixels 55 Fe peak 1-14 MeV neutrons produced from 20 MeV No significant leakage current or noise increase after deuteron breakup on a thin target, flux of ~4x10 8 neutron irradiation: negligible bulk damage effects ncm -2 s -1 , up to total fluence of 1.2x10 13 n eq /cm 2 at the considered fluences. All irradiations performed at the BASE Facility of the LBNL 88-inch Cyclotron; all tests performed at room temperature. 9/14 KEK 2009 - Piero Giubilato – Development of a Radiation Hard CMOS MAPS

Calibration with 55 Fe: comparison between irradiations Charge-to-voltage conversion gain calibration and single pixel noise as measured from the position of the 5.9 keV peak from 55 Fe spectra before and after the various irradiations on the PO pixels. No significant change in noise observed after the three irradiation. Change in gain of ~1.3 observed after ionising irradiation; no gain change after neutron irradiation. Before Irradiation After Irradiation ( 130 ± 6) e - ENC ( 122 ± 6) e - ENC Noise e - (26.7 ± 0.6) e - /ADC (36.3 ± 0.8) e - /ADC Gain ( 128 ± 5) e - ENC ( 126 ± 5) e - ENC Noise P (25.9 ± 0.5) e - /ADC (34.9 ± 0.9) e - /ADC Gain ( 69 ± 3) e - ENC ( 64 ± 5) e - ENC Noise n (25.2 ± 0.3) e - /ADC (24.4 ± 0.4) e - /ADC Gain 10/14 KEK 2009 - Piero Giubilato – Development of a Radiation Hard CMOS MAPS

LDRD2-RH electrons imaging spatial resolution Electrons imaging Point Spread Function (PSF) measured on the Titan column with a thin 75 µm diameter gold wire taut over the detector. The use of the gold wire limited to a minimum the degradation due to the electron scattering at the borders. 75 µm Energy (KeV) 10 µm pixels 20 µm pixels 11.6 ± 1 µm 13.7 ± 2.3 µm 120 10.6 ± 1 µm 12.1 ± 2µm 200 8.1 ± 1.6 µm 10.9 ± 2.3 µm 300 11/14 KEK 2009 - Piero Giubilato – Development of a Radiation Hard CMOS MAPS

Latest results: atoms e - imaging with 1MPixel device TEAM 1K detector Si lattice – 2.5 ms integration time 0.35 AMS opto process 1M pixel, 9.5 um pixel pitch Rad-hard design 25 MHz readout speed 16 parallel analog outputs Up to 400 Frames/s Thinned down to 50um to reduce backscattering FFT 12/14 KEK 2009 - Piero Giubilato – Development of a Radiation Hard CMOS MAPS