Development of a Fully Depleted Back Illumination Sensor Based on - - PowerPoint PPT Presentation

development of a fully depleted back illumination sensor
SMART_READER_LITE
LIVE PREVIEW

Development of a Fully Depleted Back Illumination Sensor Based on - - PowerPoint PPT Presentation

Hideaki Matsumura (Kyoto University) S. Nakashima, T. G. Tsuru, T. Tanaka, A. Takeda, Y. Arai, T. Miyoshi, R. Ichimiya, T. Imamura, T. Ohmoto, A. Iwata Development of a Fully Depleted Back Illumination Sensor Based on SOI CMOS Technology for


slide-1
SLIDE 1

Development of a Fully Depleted Back Illumination Sensor Based on SOI CMOS Technology for Future X-Ray Astronomy Satellites

Hideaki Matsumura (Kyoto University)

  • S. Nakashima, T. G. Tsuru, T. Tanaka,
  • A. Takeda, Y. Arai, T. Miyoshi, R. Ichimiya,
  • T. Imamura, T. Ohmoto, A. Iwata
slide-2
SLIDE 2

Motivation in X-ray Astronomy

2

Standard Detector = X-ray CCDs 45 mm CCD ~ 20 μm 2% FWHM@6 keV ~ sec 0.5 - 10 keV positional resolution Energy resolution Timing resolution

  • bservable

energy band Suzaku XIS ~1 Mpix

Weak Points

new imaging spectrometer capable of high speed reaout and low background

SOI pixel pixeldetector

< 30 μm 2% FWHM@6 keV < 10 µsec 0.5 - 40 keV

X-ray from field of view Cosmic-rays from outside (1 kHZ)

Active Shield

Anti-coincidence Trigger driven readout (100 kHz) “Our Goal” Poor time resolution of ~ sec Non X-ray background above 10 keV

  • due to cosmic rays in satellite orbit
slide-3
SLIDE 3

Concept of X-ray SOIPIX

Monolithic pixel sensor by Silicon on Insulator (SOI) Tech Si sensor + CMOS readout circuit with trigger

3

X-ray SOIPIX (cross-section)

  • ne pixel

pitch ~ 30 µm Back Bias Insulator (SiO2) Sensor layer (high resistivity) ~5 µm 200 nm >100 µm Hole Electron Time V_sig

X-ray

≈ CMOS layer (low resistivity)

Readout with Trigger Readout with Trigger Readout with Trigger

Si substrate

P+

Readout with Trigger

(not to scale)

Silicon On Insulator (SOI) Technology

slide-4
SLIDE 4

Achievements and Scope of This Work

4

Achievements

  • Trigger-driven readout of X-ray signals
  • Thick depletion layer (500 µm)
  • Readout noise ~ 60 e-
  • Energy resolution 700 eV FWHM at 8 keV

Scope of This Work

  • Spectroscopy with back illumination (BI) device
  • Investigation of charge collection efficiency
slide-5
SLIDE 5

Prototype: XRPIX1b

5

2.4 mm

32 x 32 format 30.6 µm□ pixel size

XRPIX1b-CZ-BI:

Back illumination sensor Sensor layer: 100 µm Resistivity: 0.7 kΩcm

XRPIX1b-FZ-FI:

Front illumination sensor Sensor layer: 500 µm Resistivity: 7 kΩcm

slide-6
SLIDE 6

3 x 3 Pixel Readout Method

120 120 20 40 60 80 100 120 20 40 60 80 100 120

  • utput signal

time base level Signal level X-ray signal

20 40 60 80 100 120 20 40 60 80 100 120 20 40 60 80 100 120 20 40 60 80 100 120 20 40 60 80 100 120 20 40 60 80 100 120 20 40 60 80 100 120 20 40 60 80 100 120 20 40 60 80 100 120 20 40 60 80 100 120 20 40 60 80 100 120 20 40 60 80 100 120

1 pixel event 2 pixel event

Charge sharing event

Charge often spreads over several pixels

We used special readout method to reduce readout noise

slide-7
SLIDE 7

7

Channel (ADU) 10 20 30 40 50 60 70 80 Counts 50 100 150 200 250 300

CMS Split1 Histgram

BI Spectrum

Al Kα (1.48 keV) Cu Kα (8.04 keV)

730 eV FWHM 420 eV FWHM

⇒ dead layer = 9.3 µm

Device: XRPIX1b-CZ-BI

Back Bias: 85 V Temperature: - 50 ℃ 1 hit events spectrum

10

2

10 10

2

10

3

10

Al_Kα (1.48 keV) Cu_Kα (8.04 keV) c

  • u

n t s

Quantum Efficiency at 1.48 keV = 30 %

20 40 60 channel 300 200 100 channel/counts 10 102

back bias [V]

10 102

fully depleted

> 30V

slide-8
SLIDE 8

Issues on BI Device

8

(1) Energy resolution at Al line (1.48 keV) BI: 730 eV FI (same device): 300 eV (2) Thick dead layer 9.3 µm >> our goal ~ 0.1 µm

We consider that this degradation is due to imperfection backside process. Improvement of backside treatment is ongoing.

slide-9
SLIDE 9

Spectral Shape

Device: XRPIX1b-FZ-FI

Back Bias: 30 V Temperature: - 50 ℃ 1 pixel and 2 pixel events spectrum

Channel (ADU) 20 40 60 80 100 120 140 Counts 5 10 15 20 25 30 35 40 Channel (ADU) 20 40 60 80 100 120 140 Counts 20 40 60 80 100 120 140

channel [ADU] counts / channel

1 pixel event 2 pixel event

Cu Kα (8.04 keV) Mo Kα (17.4 keV)

140 120 100 counts / channel 80 60 40 20 40 35 30 25 20 10 5 15

140 120 100

80 60 40 20

140 120 100

80 60 40 20

Peak shift Large tail components

slide-10
SLIDE 10

Spectral Shape

0.4 0.5 0.6 0.7 0.8 0.9 1 20 40 60 80 100 120 140

Center pixel pulse height / Sum Sum of pulse height of 2 pixels [ADU]

Mo Cu

slide-11
SLIDE 11

Comparison with Previous Device

Previous Device

XRPIX1-FZ-FI

Current Device

XRPIX1b-FZ-FI

1 pixel event 2 pixel event

Channel (ADU) 20 40 60 80 100 120 140 Counts 5 10 15 20 25 30 35 40 Channel (ADU) 20 40 60 80 100 120 140 Counts 20 40 60 80 100 120 140

counts / channel channel [ADU]

Cu Kα (8.04 keV) Mo Kα (17.4 keV)

140 120 100 80 60 40 20 counts / channel

140 120 100

80 60 40 20

140 120 100

80 60 40 20

40 35 30 25 20 10 5 15

Channel (ADU) 20 40 60 80 100 120 140 Counts 500 1000 1500 2000 2500 Channel (ADU) 20 40 60 80 100 120 140 Counts 200 400 600 800 1000 1200 1400 1600

channel [ADU] counts / channel

Mo Kα (17.4 keV) Cu Kα (8.04 keV)

No tail No peak shift

2000 1500 1000 500 2500

counts / channel

1400 1000 800 400 1600 1200 600 200 140 120 100

80 60 40 20

140 120 100

80 60 40 20

BPW size small BPW size large

1 pixel event 2 pixel event

slide-12
SLIDE 12

Our Hypothesis of Charge Loss

20.9 µm

Buried P-Well (BPW): Suppression of backgate effect

Previous Device

XRPIX1-FZ-FI

Current Device

XRPIX1b-FZ-FI

BPW A part of signal charge is lost at the interface region between the insulator and the sensor layer

⇒ We plan to test the hypothesis by irradiating the device

with a pencil X-ray beam

30.6 µm

14.0 µm

slide-13
SLIDE 13

Summary

13

  • We successfully obtained spectrum of Al lines using

Back Illumination SOIPIX for the first time.

  • Energy resolution is 730 eV for Al line (1.48 keV).
  • We think that a part of signal charge is lost at the

interface region between the insulator and the sensor layer.

  • We can test our hypothesis with a pencil beam

experiment.

slide-14
SLIDE 14

Thank You

14