Maurizio Boscardin FBK-irst,...

Maurizio BoscardinLegnaro 2007

3D detectors3D detectors

Maurizio Maurizio BoscardinBoscardinFBKFBK--irstirst, , TrentoTrento

boscardiboscardi@@itcitc..itit


Dividero’ la mia presentazione in due parti:

• rivelatore 3D “standard” basandomi

essenzialmente sul lavoro di Parker e Kenney e

sulla presentazioni di C. Da Via a Vertex 2006;

• rivelatore 3D “semplificato” Single-Type Column

workshop sui 3D nel febbraio 2006 a Trento: http://tredi.itc.it/


3D standardPrincipio di funzionamento

Processo di fabbricazione

Caratteristiche

area morta

resistenza al danno da radiazione

velocita’


“Standard” 3D detectors - concept[Parker et al. NIMA395 (1997)]

ionizing particlen-columns p-columns wafer surface

n-type substrate

Distance between n and p electrodes can be made very shortextremely radiation hard detector(low full depl. volt. and high CCE even at very high fluences)

Drawbacks: - electrodes are dead regions (or partially)- feasibility of large scale production still to be verified

carriers collectedat the same time


3Dplanare

Q per MIP e’ la stessa (a parita’ di spessore)aumento della capacita’Vdepl da 70V a 10Vcolonne come area mortabordo morto da ~spessore fetta a ~distanza tra colonne


3D altre possibilita’ tecnologiche

Connessioni elettrichefronte/retro

Come esempio:rivelazione di neutroni tramite 6Litalk di Uher univ. Praga a http://iworid-8.df.unipi.it/

PassantePassante ilil polisiliciopolisilicio

Riempire i fori con scintillatore




Caratteristiche

1. area morta

2. resistenza al danno da radiazione

3. velocita’


“Standard” 3D detectors

Kenney et al. IEEE TNS, vol. 46, n. 4 (1999)

1) wafer bondingsupport wafer

detector wafer

2) n+ hole definitionand etching

resist

oxide

3) hole doping andfilling

n+ polysilicon

4) p+ hole definitionand etching

resist

5) hole doping andfilling

p+ polysilicon

6) Metal depositionand definition

metal

Rather challenging process for mass production!


Keys to the technology

i. Deep RIE: capacita’ di fori profondi

ii. Wafer bonding: controllo dimensioniuscita del foro e rivelatori edgless

iii. Drogaggio tramite stato solido o gas: B2H6 (diborane), PH3(phosphine)

iv. Riempimento con polisilicio dei fori: processo LPCVD con SiH4

v. Planarizzazione della superficie

~200 micron


D- RIE

Deep reactive-ion etching (DRIE) is a highly anisotropic etch process

Aspect ratio of 20:1 or more.

The Bosch process alternates repeatedly between two modes to achieve nearly vertical structures.

1. A standard, nearly isotropic plasma etch SF6

2. Deposition of a chemically inert passivation layer C4F8


Wafer BondingCapacita di “saldare” tra loro due fette

una su cui realizzare il dispositivouna di supporto

varie tecniche disponibili

Per un 3D lo scopo e’: controllo foro uscitasupporto per rivelatori edgeless

Fetta di supporto

fetta processata

rivelatore


0.7μm0.8μm1μmPoly0.6μm0.7μm1μmTEOS

bottonTopSurface

hole

Deposizioni ossido (TEOS) e polisilicio

Come depositare/drogare il foro ?Drogaggio da stato solidoDeposizione di ossido/polisilicio tramite LPCVD• uniformita’ di deposizione• tempo richiesto per riempire il foro

Foro da 5 micron

Poly + SiO2 deposti

supe

rfic

ietop botton


3D different technological approach

• SLAC & SINTEF (Sherwood Parker) doppia colonna , riempita con poly, fori passanti

• University of Glasgowdoppia colonna con diodi Schottky

• VTT realizzato singola colonna, drogata boro, profondita’ di colonna 150-200μm

• ITC-irst •realizzato singola colonna, drogata fosforo, profondita’ di colonna 150-200μm;•In fase di realizzasione doppia colonna non passante alternata

•CNM : In fase di realizzazione doppia colonna non passante alternata




Caratteristiche

area morta


velocita’


3D silicon detectors were proposed in 1995 by S. Parker, and active edges in 1997 by C. Kenney.

Combine traditional VLSI processing andMEMS (Micro Electro Mechanical Systems)technology.

Both Electrodes are processed inside the detector bulk instead of being implanted on the wafer's surface.

The edge is an electrode! Dead volume at the Edge < 5 microns!

1. NIMA 395 (1997) 328 2. IEEE Trans Nucl Sci 46 (1999) 12243. IEEE Trans Nucl Sci 48 (2001) 1894. IEEE Trans Nucl Sci 485 (2001) 1629 5. IEEE Trans Nucl Sci 48 6 (2001) 2405 6. CERN Courier, Vol 43, Jan 2003, pp 23-267. NIMA 509 (2003) 86-918. NIMA 524 (2004) 236

3D silicon sensor fabricated at Stanford by J. Hasi (Brunel)

and C. Kenney (MCS)

C inz

iaD

aV i

a’ -V e

r te x

06

-P e

r ug i

a-

S ep t

e mb e

r 2 0

0 6




Caratteristiche

area morta


velocita’


Reasons for dead borders on standard planar technology sensors

a

b

c

d

a. space for guard ringsb. sawed edges connecting top and bottom are conductorsc. chips and cracks are also conducting and can reach inside the edgesd. the field lines bulge out, and should be kept away from b and cs


Scribe line = DRIE & drogateUtilizzabile sia su rivelatori planari che 3DRichiede un substrato di supporto → wafer bonding

Processo “a bordo attivo”

Come limitare l’area morta laterale ?

support wafer oxide

sensor wafer

p n n

poxide

support wafer oxide

p n

p

n


3D edge sensitivity using3D edge sensitivity using 13 13 keVkeV XX--rays at Berkeleyrays at Berkeley

Electrodes ~ 1.8% of total area

MeasurementPerformed using a2 μm beam

J. Hasi, C. Kenney,J. Morse, S. Parker

X-ray

10-90% < 5μm

X-ray micro-beam scan, in 2 µm steps, of a 3D, n bulk and edges, 181 µm thick sensor. The left electrodes are p-typeEfficiency measured in test beam ~98%

Cinz

iaD

aV i

a’ -V e

r te x

06

-P e

r ug i

a-

S ep t

e mb e

r 2 0

0 6


Efficiency: p and n electrodes responseEfficiency: p and n electrodes responseElectrodes area ~1.8% of total areaElectrodes area ~1.8% of total area

40% reduction in count efficiency at p40% reduction in count efficiency at p--electrodeelectrode

Cell study using 120GeVmuons (Cern X5), TelescopePrecision ~4μm.

Electrode response using 12KeV X-ray beam (ALS), beam size~ 2μm

n

n

n

n

p

n

50μm

100μm

A. Kok PhD thesis

J. Hasi, PhD thesis

Cinz

iaD

aV i

a’ -V e

r te x

06

-P e

r ug i

a-

S ep t

e mb e

r 2 0

0 6




Caratteristiche

area morta


velocita’


0

50

100

150

200

0 1 2 3 4

Fluence = 1015 protons cm-2

Eff

ectiv

e D

rift

Len

gth

(mic

rons

)

Electric Field ( Volt/micron )

Electrons

Holes

T = -20 oC

Trapping times from Kramberger et al. NIMA 481 (2002) 100 Simulations CDV and S.Watts NIM A 501(2003) 138 (Vertex 2001)

Short collection distance (50-70 μm)High average e-field per applied VbiasParallel charge collectionAlways use full substrate thickness (MIP ~80 e-/mm)

Why is 3D radiation hardWhy is 3D radiation hard3D planar

Leff = vdriftx τtrap

Ottaviani, Canali et al.

e-

h+

Cinz

iaD

aV i

a’ -VS

TD6

--

S ep t

e mb e

r 2 0

0 6


0

2 10 -5

4 10 -5

6 10 -5

8 10 -5

0 .0001

0 2 10 15 4 10 1 5 6 10 1 5 8 10 1 5 1 10 16

y = -2 .0 23 4e-6 + 1 .32 9e -20x R = 0 .99 243

Cur

rent

at 2

0C fu

ll de

plet

ion

[A]

F lu en ce [n /cm 2]

α = 6 x 10-17 A /cm

Radiation hardness: macroscopic Radiation hardness: macroscopic parametersparameters and signal efficienciesand signal efficiencies

0

20

40

60

80

100

120

140

160

0 2 1015 4 1015 6 1015 8 1015 1 1016

Volta

ge F

ull D

eple

tion

[V]

Fuence [n /cm 2]

n-type startingmaterial

expected typeinversion point

-0 .01

-0 .008

-0 .006

-0 .004

-0 .002

0

0.002

-3 10 -8 -2 10 -8 -1 10 -8 0 1 10 -8 2 10 -8 3 10 -8

Am

plitu

de [V

]

T im e [s ]

8 .6 e 15 n /cm 2

5 .98e 1 5 n /cm 2

3.7e 15 n /cm 2

N O N IR R A D IA TE DC . D aV ia e t a l M arch 06

Average of~1000 pulses

IR Laser1060nm

T=-10C

bias-0 .0 0 6

-0 .0 0 5

-0 .0 0 4

-0 .0 0 3

-0 .0 0 2

-0 .0 0 1

0

0 .0 0 1

0 .0 0 2

-3 1 0 -7 -2 1 0 -7 -1 1 0 -7 0 1 1 0 -7 2 1 0 -7 3 1 0 -7

Sign

al [V

]

T im e [s ]

Gain = ~1000

Oscilloscope

20 oCno ben. ann.

Cinz

iaD

aV i

a’ -VS

TD6

--

S ep t

e mb e

r 2 0

0 6


C inz

iaD

aV i

a’ -VS

TD6

--

S ep t

e mb e

r 2 0

0 6


Detectors Parameters

-10541210022400900280ATLAS strip

-10481094022800600285CMS pixel

2073985513500500500Diamond

-10771448018801602353D

T [C]Signal after 10years LHC at 4cm [%]

Signal after 10years LHC at 4cm [e-]

MIP charge[e-]

V bias[V]

Thickness[μm]

C inz

iaD

aV i

a’ -V e

r te x

06

-P e

r ug i

a-

S ep t

e mb e

r 2 0

0 6




Caratteristiche

area morta


velocita’


Short collection distance (50-70 μm)High average e-field at low VbiasParallel charge collection

speedspeed

3D planar

Cinz

iaD

aV i

a’ -VS

TD6

--

S ep t

e mb e

r 2 0

0 6 rt~1ns

5ns T=300K

rt 1.5ns≈

oscilloscope trace

3D Tests in progress with a 0.13 3D Tests in progress with a 0.13 μμm CMOS m CMOS Amplifier chipAmplifier chip (designed by (designed by DepeisseDepeisse--AnelliAnelli--CERN MIC)CERN MIC)

3D Inter-electrodedistance = 50 μm


Attivita’ IRST

rivelatori 3D a singola colonnaprincipio di funzionamento

Tecnologia Single Type Column

Misure

Sviluppi futuri


Rivelatore 3D a singola colonna

primo passo verso un 3D

semplificare il processocolonne di un solo tipo

fori non passanti


Single-Type-Column 3D detectors - concept[C. Piemonte et al NIMA 541 (2205)]

electrons are swept away by the transversal field

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

Fabrication process is much simpler:• column etching and doping performed only once• holes not etched all through the wafer

n+ electrodes

Uniform p+ layer

p-type substrate

…on the way to a fully 3D device: 3D-STC

ionizing particle

n+ n+

…BUT collection mechanism is not very efficient

(see slides on signal formation)


StaticStatic devicedevice simulationssimulations

null field lines!!

1) Vbias=0V 2) Vbias=2V

3) Vbias=5V 4) Vbias=20V

Depletion mechanismpitch = 80μmhole depth = 150μmsubst. hole conc. = 5e12cm-3

=>lateral full dep. volt. ~ 5Vvertical full dep. volt ~ 40V


Pote

ntia

l

High subst. dopant conc. implies smaller null field region and higherelectric field.

⇒ for p-type subst. the detector works better after irradiation

Na=1e12 1/cm3

Na=5e12 1/cm3

Na=1e13 1/cm3

Na=1e12 1/cm3

Na=5e12 1/cm3

Na=1e13 1/cm3

StaticStatic devicedevice simulationssimulations (2)(2)El

ectri

c fie

ld

Profiles along the cutline


Full Full chargecharge collectioncollection timetime

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)

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

Same Vbias, different impact point

charge collectedis ¼ for interactionin the middle point

e h

250μ

m50

μm

First phaseTransversal movement

Second phaseHole vertical movement


Attivita’ IRST

RIVELATORI 3D a singola colonnaprincipio di funzionamento


Misure

Sviluppi futuri


Processo 3D STC

ossidazionedefinizione e drogaggio p-stopdrogaggio contatti di substrato (retro)ossidazione


definizione “colonne”attacco DEEP-RIE

Processo 3D STC


Drogaggio “colonne”“allargamento” colonna Diffusione da sorgente solidaDeposizione polisilicio e successivo drogaggio

Ossidazione ColonnePossibile deposizione ossido/nitruroOssido termico

Processo 3D STC


Definizione e apertura contattiDeposizione metal fronteDefinizione metal fronteDeposizione metal retro

Processo 3D STC


3D process3D process

n+ diffusion

contact

metaloxide

hole

Hole etching with Deep-RIE technologyWide superficial n+ diffusion in which the contact is locatedPassivation of holes with oxide

hole

hole metal strip

Si High Resistivity, p-type, <100>Surface isolation: p-stop or p-sprayHoles are “empty”

performed @ IRST

Hole depth 120-150μHole diameter ~10μm


ColumnColumn etchingetchingSo far, 3 runs have been fabricated:2 runs holes etched by CNM (Barcelona);1 run holes etched by a company providing microtechnology services;

Important, in both cases:- same column parameters;- extremely good process yield (leakage current).

Furthermore, etching test have been performed with a third provider, again, with good results.

column etching and treatment are not

critical200μ

m

220μ

m

Etched through

waferthickness

D-RIE equipment will be available at IRST in september 2007


MaskMask layoutlayout

“Short” strip detectors~ 0.8x5mm2 64 strips10 col./strip

Planar and 3D teststructures

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

2. Strip detector = “easy” to (electrically) test

“Long” strip-like detectors~ 2x0.5cm2 64strips~230 col./strip


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

Strip Strip DetectorsDetectors –– layoutlayout


Full Full depletiondepletion evaluationevaluation in 3Din 3D--stcstc

From 1/C2 curves one can determine:• full depletion between columns (in this case ~5V for 80μm col. pitch)• full depletion of the bottom region (~35V for col depth of 150μm)

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

1/Cback2 characteristic

Phase 1

Phase 2

• high Cback• ~ zero Cint

• max Cint• slowly dec.

Cback

undepleted Si

undepleted Si

Phas

e 1 Phase 2

matrix of10x10 holes

guard ring

3D diode

BackCbackTot CintTot

f=10kHz


LeakageLeakage currentcurrent & & yieldyield

• Low leakage current• Good process yield

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

Measured more than 100 devices 90% showingcharacteristics similar to those reported in the plot

Measurement on “long” strips (area about 1cm2)

Examples of IV curves

• Number of columns per detector: 12000 – 15000

⇒ Average leakage current < 1pA/column

• Increase of current caused by surface effects. No guard rings were implemented.


Attivita’ IRST



Misure

Sviluppi futuri


On On goinggoing activityactivity

SCIPP (USA): CCE measurements on large strips

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

University of Freiburg (D); measurements on short strips

Ljubljana: TCT and neutron irradiation


MeasuredMeasured currentcurrent signalsignal in 3Din 3D--stcstc

DEVICES: small strip detectors

SETUP:• IR laser (m.i.p. simulation) – beam diameter in the silicon FWHM~7 μm• Width of light pulses ~ 1ns , repetition rate 100 Hz • 3 independent channels – fast current amplifiers 1kHz-2GHz

Study performed in Ljubljana. See Kramberger’s talk at 8th RD50 workshop: http://rd50.web.cern.ch/rd50/

3D-stc DC coupled detector(64 x 10 columns)80 μm pitch80 μm between holes10 μm hole diameter


MeasuredMeasured currentcurrent signalsignal in 3Din 3D--stc (2)stc (2)

beam position

signal induced on the central strip

beam position

signal induced on the central strip

50ns0 50ns0

very fast component +long tail due to driftof holes

bipolar with high fast component (non collecting electrode)

Measurements well reproduce the simulations previously reported!More work has to be done, above all on irradiated detectors.

Many data available! Two examples shown below.


RadiationRadiation damagedamage studiesstudies 11

Irradiation: neutrons at TRIGA research reactor in Ljubljana;6 fluences between 5e13n/cm2 and 5e15n/cm2

Annealing: 15 days at RT (~ minimum depletion voltage).

Measurements: IV and CV (series model @10kHz) @ 23C

Aim: study of the depletion characteristic (at the moment)

[performed in collaboration with V. Cindro, Ljubljana]

Devices: 3D diodes, p-type FZ 525μm thick substrate,p-stop isolation& planar diodes with same subst. characteristics.



1.E-11

1.E-10

1.E-09

1.E-08

1.E-07

1.E-06

1.E-05

1.E-04

0 50 100 150 200 250 300Vrev (V)

Idio

de (A

)

n-irrad 3D diodes (D2)

before irradiation

F1…F6 1) 3D Diode current(80μm pitch)

Normal current behavior:current increases with fluence

2) CV measurementsdifficult measurement as it depends ofrequency and model (series/parallel)⇒ we look only for kinks in the

CV related to full lateral depletion

Conc ~ 1/d(1/Cback2)/dV)1.0E+12

1.0E+13

1.0E+14

1.0E+15

1.0E+16

0 50 100 150 200Voltage [V]

Con

c [c

m-3

]

n-irradiated 3D diodesn-irradiated 3D diodes

F1F2

F3F4

F5 F6

80μm pitch100μm pitch



Simulating the full lateral depletion voltage with Nsubestimated from equation (*) we obtain values comparable with those reported on the plot.

ΔNeff=β*Φβ=0.021cm-1

see Cindro’s talk at 8th RD50 workshop: http://rd50.web.cern.ch/rd50/

(*)

40/50μm 40/50μm1

10

100

1000

10000

1.0E+13 1.0E+14 1.0E+15 1.0E+16Fluence (n/cm 2̂)

Dep

letio

n V

olta

ge (V

)

col. pitch = 80um

col. pitch = 100um

expected from planardiode300μm

lateral depletion

each column depletes half col. Pitchthe lateral depletion voltage is very low


CCE CCE withwith β β particlesparticlesINFN Firenze & SCIPPINFN Firenze & SCIPP

Charge Collection Efficiency

3D diode, 80 µm pitch between columnsFZ substrate 500 µm thick

Measurement system:• β- source with scintillator/PMT trigger• AMPTEK read-out electronics:

- Shaping time: 250ns - ENC ≈ 500 e-

Good correlation between CCE and C-V measurements

0 10 20 30 40 500,2

0,4

0,6

0,8

1,0

1,2

norm

CC

E an

d 1/

C

Reverse Voltage [V]

CCE 1/C

3D diode, 80 µm pitch between columnsCZ substrate 300 µm thick (VFD ~ 35V)

http://scipp.ucsc.edu/STD6/abstracts/sadrozinski.pdf


Laser Resultsunirradiated

channel 262

channel 261Lateral depletion around 12V

x

y

Penetration depth @ 982nm ≈ 100µmLength of pulse ≈ 1-2nsMicroscope to focus optically → laser spot Ø ≈ 4–5µmx-y stages with µm resolution

Simon Eckert, Friburg University presented at VCI Vienna 2007


Post-Irradiation: CCE @ 130V

Sum of both strips:

Irradiated with 26MeV protons dose 1015Neq/cm2

low CCE under p-stops – as before irradiation

Low field regionor trapping?

Sign

al in

mVlef

t srip

right s

trip



Post-IrradiationUsing a smaller step-size you can nicely see how the depleted region is growing with increasing bias voltage

Vbias = 130V Vbias = 65V

Vbias = 32V

Sign

al in

mV

Sign

al in

mV

Sign

al in

mV



Attivita’ IRST



Misure

Sviluppi futuri


Next technological steps

First Process• p-type Si• DRIE ~ 200mm• no hole filling• single column• single side

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


New Layout

First Layoutmicrostrip

New LayoutPixel like

p-diff n-diff

bump regionmetal


• 3D p-on-n• ALICE and MEDIPIX pixel layout

• Strip detectors

• Test structures

•3D n-on-p• ATLAS and CMS pixel layout

• Strip detectors

• Test structures

On going 3D-DTC processes

Bulk contact junction columns


3D – dtc227.02.2007

6 ATLA

S pixel detectors

planar test structures (8)

3D diodes (stc&dtc; 80&100µm)

single-columnstest strucures (8)

6 CMS pixel detectors

16 ATLA

S pixel detectors

13 microstrip

detectors (stc& dtc)

+ 1 strip CAP test

3D diodes (stc&dtc; 80&100µm)

CMS “small” pixel detectors (8)


Conclusioni

• Rivelatori 3D hanno dimostrato caratteristiche promettentiper un utilizzo nei rivelatori di vertice - in particolare per quantoriguarda la resistenza al danno da radiazione -

E’ importante ora continuare questo sviluppo sia• Tecnologico (yield, …)• Caratterizzazione ( dipendenza del segnale dalla posizione,..)

Interesse da parte di ATLAS per uno sviluppo di rivelatori 3D in vista di una sostituzione ( e up-grading) del b-layerhttp://test-3dsensor.web.cern.ch/test-3dsensor/Default.htm


Date post:	19-Oct-2020
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times

Maurizio Boscardin FBK-irst,...

Documents