development of 3d detectors and sipm @ itc-irstiworid-8.df.unipi.it/presentations/boscadin.pdf ·...

M. Boscardin IWORID-08

Development of 3D detectors and Development of 3D detectors and SiPMSiPM @@ ITCITC--irstirst

Maurizio Maurizio BoscardinBoscardin

[email protected]@itc.it


ITC-irst

ITC (Istituto Trentino di Cultura) is a public research institute in Trento mainly funded by the local government

Trento 250 researchers working on:

• information technology

• microsystems & physics

ITC-irst

An entire division (60) is working on silicon sensors


Silicon Radiation DetectorsR&D activity

“Standard” technologyFrom the specifications given by the “user”we design, produce, and (electrical) test the detector.• single/double-sided strip detectors• p-on-n/n-on-n pixel detector

R&D activitiesDevelopment in cooperation with the partners• very thin detectors• detectors made on radiation hard silicon substrates• 3D detectors• silicon photomultipliers

Development of 3D sensors and SiPM is being carried out in the framework of a collaboration between INFN and ITC-irst

TCAD simulationCAD design Fabrication Device testing


MT - LAB

MICROFAB. LAB.• Ion Implanter• Furnaces• Litho (Mask Aligner )• Dry&Wet Etching• Sputtering & Evaporator• On line inspection• Dicing and bonding

TEST LAB.• Automatic probe station• Manual probe station• Optical bench

Furnaces

We process4” wafers

Automatic probe station

(cassette-to-cassettedouble side testing)


Development of 3D detectors Development of 3D detectors @@ ITCITC--irstirst

3D 3D -- OutlineOutline

•“standard” 3D - concept

• 3D detectors: status

ITC-irst activity

• Single-Type Column 3D detector concept

• Simulation, Design, Process and First Characterization

• Future Activity


“Standard” 3D detectors - concept

Proposed by Parker et al. NIMA395 (1997)

n-columns p-columns

wafer surface

ionizing particle

Short distance between electrodes:• low full depletion voltage• short collection distance

more radiation tolerantthan planar detectors!!

n-type substrate


3D status

• SLAC (Sherwood Parker)

double columns filled with doped polisilicon, deep hole (entire wafer thickness)

• University of Glasgow

double columns Schottky & diffused diode, deep hole , more info on http://rd50.web.cern.ch/rd50/5th-workshop/

• VTT

Semi 3D: single column boron doped on n-type Si; limited depth (150-200µm)

• ITC-irst

Single Type Columns: single column phosphorus doped on p-type Si; limited depth (150-200µm).

• CNM

workshop on 3D in february 2006 at Trento : http://tredi.itc.it/

http://tredi.itc.it/


3D detectors @ ITC-irst

1.Simulations of 3D-STC detectors

2.Technology used in the first two fab. runs

3.Electrical characterization of first prototypes

4.Future Activity on 3D


Single-Type-Column 3D detectors - conceptNIM A 541 (2005) 441–448 “Development of 3D detectors ..” C. Piemonte et al

p-type substraten+ electrodes

Uniform p+ layer

electrons are swept away by the transversal field

n+ n+n+ n+n+ n+

ionizing particle

holes drift in the central region and diffuse towards p+ contact

Main feature of proposed 3D-STC:• column etching and doping performed

only once• holes not etched all through the wafer• bulk contact is provided by a backside

uniform p+ implant

Simplification of the fabrication process


Simulated potential distribution (1)

Potential distribution (vertical cross-section)50µm

300µ

m

Potential distribution(horizontal cross-section)

in scale

0Vnot in scalenull field lines

-10V

-5V

-15V


Simulated potential distribution (2)

Simulation of the electric field along a cut-line from the electrode to the center of the cell

Na=1e12 1/cm3

Na=5e12 1/cm3

Na=1e13 1/cm3 DRAWBACK:3D-stc: once full depletion is reached it is not possible to increase the electric field between the columns

To increase the electric field strength one can act on the substrate doping concentration


Capacitance simulations

1) Vbias=0V 2) Vbias=2V 3) Vbias=5V 4) Vbias=20V

0.0

1.0

2.0

3.0

4.0

5.0

0 10 20 30 40 50 60Bias Voltage (V)

C^-

2 (p

F-2)

pitch=80umpitch=80um simulation

2 34

Do not consider the hot spot in the pictures,it is the charge released by a particle.

The 1/C2 curve of the col-to-backcapacitance can be used to extract both the intercolumn as well as the col-to-back full depletion.


Full charge collection time

e h

250µ

m50

µm

First phaseTransversal movement

Same Vbias, different impact point

In the worst case of a track centered the central region, 50% of the charge is collected at t ~ 300ns

Outside this region, 50% of the charge is collected within 1ns.

1 2

34

(25,25)(20,20)(10,10)

Second phaseHole vertical movement

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1E-12 1E-11 1E-10 1E-09 1E-08 1E-07 1E-06 1E-05 1E-04Time (s)

Col

lect

ed c

harg

e (a

.u.)

250um_25-25250um_20-20250um_10-10

charge collectedis ¼ for interactionin the middle point


Mask layout

Small version of strip detectors

Planar and 3D teststructures

1. “Low density layout” to increase mechanical robustness of the wafer

2. Strip detector = “easy” to electrical test

“Large” strip-like detectors


Strip detectors - layout

metal

p-stop

hole

Contact opening

n+

Inner guard ring (bias line)

Different strip-detector layouts:• Number of columns from 12000 to 15000 • Inter-columns pitch 80-100 µm• Holes Ø 6 or 10 µm


3D process (1)

First Process• p-type Si• DRIE ~ 150µm• no hole filling• single column• single side

p-stop

n+ columnuniform p+ layer

hole

Hol

e de

pth

~ 12

0µm

hole metal strip


3D process (2)

n+ diffusion

contact

metaloxide

hole

Deep RIE performed at CNM, (we will

have the D-RIE in IRST within this year)

Wide superficial n+ diffusion around

the hole to assure good contact

No hole filling (with polysilicon)

Passivation of holes with oxide

Surface isolation: p-stop or p-spray


3D diode – layout:

Bulk

Guard ring

p-stop

10x10 holes matrix

p-stopIleak = 0.68 ± 0.2 pA/column @ 20V

p-sprayIleak = 0.59 ± 0.12 pA/column @ 20V

1.0E-12

1.0E-11

1.0E-10

1.0E-09

1.0E-08

1.0E-07

1.0E-06

1.0E-05

1.0E-04

0 10 20 30 40 50 60Vbias [V]

I leak

[A]

diodeguard ring

diode─ guard ring

1.0E-12

1.0E-11

1.0E-10

1.0E-09

1.0E-08

1.0E-07

1.0E-06

1.0E-05

1.0E-04

0 20 40 60 80 100Vbias [V]

I leak

[A]

1st punch through

2nd punch throughdiodeguard ring


3D diode – CV measurements p-stop

0.00

0.50

1.00

1.50

2.00

2.50

0 10 20 30 40 50 60Vbias [V]

C-2

[pF-

2 ]

Capacitance measurement versus back on a 300µm thick wafer with ~150µm deep columns, 100µm picth

Back

1/C2

0

2

4

6

8

10

12

0 10 20 30 40 50 60Vbias [V]

Cdi

ode

[pF]

1 2

region between columnsis not fully depleted⇒ large capacitance

full dep. between columns ~ 7V

Cd

1/C

d2[p

F-2 ]

Cd

[pF]

f=100kHz

region between col.is fully depleted⇒ depletion proceeds

only towards the backlike a planar diode

full depletion ~40Vdepletion width of ~150µm

Pha

se 1

Ph a

se 2


Strip detectors electrical characterization

1.0E-10

1.0E-09

1.0E-08

1.0E-07

1.0E-06

1.0E-05

0 50 100 150 200Vbias [V]

I leak

[A]

p-spray

p-stop

Bias line

Guard ring

Electrical Chacaterization• Leakage current: < 1pA/column• Single-strip “backplane” capacitance: <5pF• Inter-column capacitance range 12÷19 fF/column

Number of columns per detector: 12000 - 15000

Average leakage Leakage current < 1pA/column


Strip detectors – IV measurements

Current distribution @ 40V of 70 different devices

0

5

10

15

20

25

30

0 5 10 15 20 25 30 35 40 45 50 >50

I bias line [nA]

Det

ecto

rs c

ount

Good process yield

First production has proved the feasibility of 3D-stc detectors


on going activity

d4

d5

1.E-11

1.E-09

1.E-07

1.E-05

1.E-03

0 25 50 75 100

Reverse Bias [V]

Dio

de C

urre

nt [A

]before irradiation

after irradiation

University of Glasgow (UK): CCE measurements with α, β, γ on 3D diodes and short strips

SCIPP (USA):CCE measurements on large strips

INFN Florence (Italy): CCE measwith β,on 3D diodes;

University of Freiburg (D); measurements on short strips

Ljubljana: TCT and neutron irradiation

3D diode (80µm picth) irradiated at Liubliana at 5 different neutron fluences(from 5E13 to 5E15)


CCE measurements (preliminary)

0

0.2

0.4

0.6

0.8

1

0 20 40 60 80 100

Reverse Bias [V]

CC

E

Cz 300um Fz 500mm

d4

d5

tc

tw

a b

Thanks to Carlo Tosi, Mara Bruzzi, Antonio De SioINFN and University of Florence 100% CCE @ low voltages

The fast reaching (before full depletion) of a 100% efficiencysuggests that carriers generated in the undepleted region are effectively collected

CCE@0V ≈ tc/tw (for tc ~ 150µm)

due to the peculiar geometry of 3D detectors, a region as deep as the column is always sensitive


Next Run

New Process• n-type Si• DRIE ~ 250mm• no hole filling• double columns• double side

New Layout = Pixel• MEDIPIX1• ATLAS• ALICE


3D Conclusion

First production has proved the feasibility of 3D-stc detectors

3D-stc detector:• Advantage: “simple” fabrication process, extremely interesting device to tune the technology for the production of standard 3D detectors• Disadvantage: in those applications not requiring charge information in short time Very long full charge collection times, can be used

Next Step:• new process & new layout ( pixel detector ).


3D “technology”

3D active edge 3D “readout technology”

large area devices

large area imaging systems

planar detector + dopantdiffused in D-RIE etched edge then doped(C. Kenney 1997).

Back plane physicallyextends at the edge.

Active volume enclosed by an electrode: “active edge”

one pixel of imaging matrix

trench filled with doped polisilicon or metal

n+

p+ p+ p+ p+

Imaging pixel matrices with 3D readout; S. Eränen VTT finland

see at:http://tredi.itc.it/


Development of Development of SiPMSiPM@@ ITCITC--irstirst

SiPMSiPM –– OutlineOutlineIntroduction

•The Geiger-mode APD•The Silicon PhotoMultiplier

ITC-irst activity•First results of the electrical characterization of the SiPMs produced at ITC-irst.

http://sipm.itc.it/


Impact Ionization

1E-16

1E-15

1E-14

1E-13

1E-12

1E-11

1E-10

1E-09

1E-08

1E-07

1E-06

-30 -25 -20 -15 -10 -5 0Reverse Bias Voltage (V)

Cur

rent

(A)

diode structureSimulated diode reverse current

IMPACT IONIZATION ONIMPACT IONIZATION OFF

VBD VAPD

p+ n-

V

n

electric fieldin the reversedbias diode

depletion width

E

V < VAPD => photodiode 1 collected pair/generated pairVAPD < V <VBD => APD <M> collected pairs/generated pairV > VBD => Geiger-mode APD inf. collected pairs/generated pair


Geiger-mode APD

Diode biased at VD > VBD

p+

n-

VD

n

t

i

t0

i=iMAX

t1

t < t0 i=0 (if no free carriers in the depletion region)

t = t0 carrier initiates the avalanchet0 < t < t1 avalanche spreadingt > t1 self-sustaining current

In order to be able to detect another photon, quenching mechanism needed:

Two solutions:• large resistance:

passive quenching• analog circuit:

active quenching[extended literature from politecnicodi Milano (Cova et al.)]

t

i

t0 t1 t2

t

vD

vBD

VBIAS

VD


SiPM concept

GM-APD gives no information on light intensity

SiPM first proposed by Golovinand Sadygov in the mid ’90

A single GM-APD is segmented in tiny microdiodes connected in parallel, each with the quenching resistance.

Each element is independent and gives the same signal when fired by a photon

⇒ output signal is proportional to the numberof triggered cells that for PDE=1 isthe number of photons

Q = Q1 + Q2 = 2*Q1

substrate

metal


Dark count

Noise = false counts triggered by non photogenerated carriers

Sources of free of carriers: 1. SRH generation in the depleted region2. tunneling in high-field region3. diffusion from the highly-doped regions

⇒ Dark count rate depends on: - number of generation centres- temperature - overvoltage


Afterpulsing

AFTERPULSING: carriers are trapped during the avalanche and then released triggering an avalanche

If the carrier is released after the recovery time => increase dark count rate

If it is released within the recovery time => no/smaller pulse

Afterpulse depends on:- number of traps- number of carriers transiting during an avalanche

=> Ideally the recovery time should be long enough so that thetraps release the carrier within this time.


Optical cross-talk

During an avalanche discharge photons are emitted because of spontaneous direct carrier relaxation in the conduct. band

cell1 cell2Those photons can trigger the avalanche in an adjacent cell: optical cross-talk.

cell1 cell2

Solutions:- operate at low over-voltage => low gain => few photons emitted- optical isolation structure:


Photo-detection Efficiency

PDE = QE * Pt * Ae

Electrons should trigger the avalanche because of the higher ionization rate

In any case, the higher the overvoltage is the higher Pt is.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0.01 0.1 1 10Depth (um)

Ads

orbe

d lig

ht

350nm400nm420nm450nm500nm600nm

1) Internal quantum efficiency

2) Transmission efficiency of the coating

Dead area is given by the structures at the edges of the microcell (metal layers, trenches, resistor…)


Characteristics of first prototypes

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

300 350 400 450 500 550 600 650 700

Wavelength (nm)

Tran

smitt

ance

1. Substrate: p-type epitaxial

2. Very shallow junction to improve quantum efficiency at short wavelengths

3. Quenching resistance made of doped polysilicon

4. Anti-reflective coating optimized for λ~450nm

5. No structure for optical isolation

6. Geometry NOT optimized for maximum PDE

13

14

15

16

17

18

19

20

0 0.2 0.4 0.6 0.8 1 1.2 1.4depth (um)

Dop

ing

conc

. (10

^) [1

/cm

^3]

0E+00

1E+05

2E+05

3E+05

4E+05

5E+05

6E+05

7E+05

E fie

ld (V

/cm

)

DopingField

n+ p

Doping profiles and Electric Field

ARC transmittance


First prototypes

The wafer includes many structurediffering in geometrical details

The basic SiPM geometry iscomposed by 25x25 cells

Cell size: 40x40µm2

1mm

1mm


Electrical characterization

1.E-10

1.E-09

1.E-08

1.E-07

1.E-06

1.E-05

1.E-04

1.E-03

-40 -30 -20 -10 0Vbias [V]

Cur

rent

[A]

IV characteristics of 10 devices

T=22oC

Messages from the IV curve:

• Breakdown voltage 31V. Uniform all over the wafer surface.

• post-breakdown current very uniform (measured on 90 devices)only 20% of the devices show anomalous behavior.

position ofthe devices


Signal characteristics

-0.3

-0.25

-0.2

-0.15

-0.1

-0.05

0

0.05

-1.0E-07 0.0E+00 1.0E-07 2.0E-07 3.0E-07Time (s)

Volta

ge (V

)

VBIAS=35.5V

Dark signal

single cell signal

doublesignal

(optical cross-talk)

Rise time ~1ns (limited by read-out system)

Recovery time ~70ns

SiPM read-out by means of a wide-band voltage amplifier on a scope

-0.2

-0.15

-0.1

-0.05

0

0.05

-5 5 15 25 35 45 55 65 75 85 95Time (ns)

Volta

ge (V

)

Vbias=34VVbias=35.5VVbias=32.5V

T=22oC


Single electron spectrum

DARKSingle electron spectrum in dark conditionIntegration time = 10ns.

0

200

400

600

800

1000

1200

1400

-8E-10 -6E-10 -4E-10 -2E-10 0E+00QDC

Cou

nts

33.5V35.5V

double peak

For VBIAS=35.5V double peak counts = 1/20 single peak counts


Gain

DARK

0.0E+00

2.0E+05

4.0E+05

6.0E+05

8.0E+05

1.0E+06

1.2E+06

1.4E+06

1.6E+06

1.8E+06

2.0E+06

31 32 33 34 35 36Bias Voltage (V)

Gai

n

Gain vs Bias voltage

T=22oC

Linear dependence,as expected.

Q=Cmicrocell*(Vbias-Vbreakdown)

=> C = 80-90fF


Dark count

0.0E+00

5.0E+05

1.0E+06

1.5E+06

2.0E+06

2.5E+06

3.0E+06

3.5E+06

4.0E+06

31 32 33 34 35 36Voltage (V)

Dar

k C

ount

rate

(Hz)

Dark count vs Bias voltage

Dark count increases linearly with voltage.=> PDE should follow the same trend.


Response to light

QDC Channels50 100 150 200 250 300 350

histEntries 20000Mean 101.2RMS 61.39

QDC Channels50 100 150 200 250 300 350

Co

un

ts

0

2000

4000

6000

8000

10000


2_Si_PD 33.1V

1pe 2pe 3pe

∆V=1.5VT=22oC

Pulse charge spectrum from low-intensity light flashes (red LED)

Each peak corresponds toa different number of fired cells

Very good uniformity response from the micro-cells

QDC Channels0 200 400 600 800 1000 1200


QDC Channels0 200 400 600 800 1000 1200

Co

un

ts

0

200

400

600

800

1000

1200


2Si_PD_33_5V

T=22oC∆V=2V

1pe

2pe


Energy resolutionVERY PRELIMINARY

from Pisa (A. Del Guerra)

Very first measurement on one single device

• 1mmx1mmx10mm LSO crystal coupled to a SiPM• Data taken in coincidence with a 10mm diam, 5mm thick YAP crystal coupled to a PMT.

• 22Na source.• 2.5V overvoltage• 37% energy resolution

1) Optimizing the set-up andthe working conditions this value can be improved

2) Area efficiency has to beoptimized yet!


SiPM Conclusion

• Extremely encouraging results from the first productionof SiPMs at ITC-irst. Fully functional devices with:

- Gain ~ 106 (linear with VBIAS)- Dark count ~ MHz - Recovery time ~ 70ns- Good uniformity of the micro-pixels response- PDE measurement in progress,

encouraging first results

• Second production run just completed.- Implemented trenches for optical cross-talk isolation.- As for the first run, IV measurements indicate a

high production yield (80%)

• Next steps: SiPMs with lower dark count


Acknowledgements:

Thank to all the people involved in the ITC-irst & INFN project

DASIPM and TREDI


VACUUM

TECHNOLOGY

SOLID-STATE

TECHNOLOGY

PMT MCP-PMT HPD PN, PIN APD GM-APD

Blue 20 % 20 % 20 % 60 % 50 % 30%

60-70 % 50%

80 % 40%

few ns tens of ps

∼ 200 105 - 106

100-500V < 100 V

No sensitivity

No sensitivity

∼10 ph.e ∼1 ph.e

∼ 10 ps ∼ 100 ps

106 - 107 3 - 8x103

3 kV 20 kV

Axial magnetic field ∼ 2 T

Axial magnetic

field ∼ 4 T

1 ph.e 1 ph.e

compact sensible, bulky

Green-yellow 40 % 40 % 40 % 80-90 %

Photon detection efficiency

Red < 6 % < 6 %< 6 % 90-100 %

Timing / 10 ph.e ∼ 100 ps tens ns

Gain 106 - 107 1

Operation voltage 1 kV 10-100V

Operation in the magnetic field

< 10-3 T No sensitivity

Threshold sensitivity (S/N>>1)

1 ph.e ∼100 ph.e

Shape characteristics sensiblebulky

robust, compact, mechanically rugged


GAIN

The area of the current pulse represents the gain

GAIN = Area/q = IMAX x τ / q = C x (VBIAS-VBD)/q

The capacitance should be large in order to havea high gain. On the other hand, a large capacitance leads to longer recovery times.

VBIAS-VBD

C2>C1GAIN

C1


1.E+02

1.E+03

1.E+04

1.E+05

1.E+06

1.E+07

0 50 100 150 200 250threshold (mV)

dark

cou

nt ra

te (H

z)

Room temperature (~ 23°C)•1 p.e. dark count rate: ~ 3 MHz•3 p.e. dark count rate: ~ 1 kHz

trenches for the optical isolation between micro-cells were not implemented in the first run

SiPM dark count

32.0 V32.5 V

33.0 V33.5 V 34.0 V

34.5 V

0.0E+00

1.0E+06

2.0E+06

3.0E+06

4.0E+06

30 32 34 36Voltage (V)

Dar

k C

ount

rate

(Hz

Dark count rate: • linear variable with Vbias• increases with the temperature


Turn-on probability

1E-03

1E-02

1E-01

1E+00

1E+01

1E+02

1E+05 2E+05 3E+05 4E+05 5E+05 6E+05 7E+05E field (V/cm)

Ioni

zatio

n R

ates

(1/u

m)

P01 = turn-on probability = probability that a carrier traversing the high-field region triggers the avalanche.⇒ it affects the detection efficiency!

It is linked to the ionization rates.

electrons

holes

1) electrons higher ioniz. rate

⇒ first solution is better

2) ioniz. rates increase with field⇒ increased probability for

higher over-biasing

p+

n-

nn+

p-

p

he he


Strip detectors – CV measurementsCapacitance measurement between one strip and the two first neighboring p-stop; DC coupling; strip pitch=80µm

f=100kHz

0

1

2

3

4

5

6

7

8

0 10 20 30 40|Vback| [V]

Cin

t [pF

] Det1

Det3

Det4Det8Det3 Det1

93µm

230 columns per strip

80µm Typical inter-column capacitance range12÷19 fF/column


Det4Det8 Other capacitance measurement results

Typical coupling capacitance for AC detectors ~ 60pF

Typical single-strip “backplane” capacitance(after lateral full depletion)<5pF (~200 columns)

100µm80µm




3D 3D diodediode –– CCE CCE measurementsmeasurements((preliminarypreliminary))

Carlo Tosi, Mara Bruzzi, Antonio De SioINFN and University of Florence

0 25 50 75 100 125 150 175 2000

20

40

60

80

100

120

Cou

nts

Channels

Cz 300µm not irradiated p-type diode

Integratorcircuit

Amptek a225

SCINTILLATOR + PMT

Sr9

0

p+nn+

DAQ card

Trigger LINE

Shaping circuitOriginal system:β- sourceShaping time: 2.4µs ENC= (280+5.6C/pF)e-

Max inverse current= 1500pA

Actual System:Max inverse current=10µANoise=2000e-

Calibration on Cz, 300µm, p-typeplanar diode

development of 3d detectors and sipm @ itc-irstiworid-8.df.unipi.it/presentations/boscadin.pdf ·...

Documents