Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors,J.M. Klaiber-Lodewigs - Univ. Dortmund

Quality assurance

for the

radiation hard ATLAS pixel sensors

Contents: - Why is systematic QA necessary?

- Procedures and design features- Measurements before irradiation- Measurements after irradiation

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 2

Introduction

• large number of detector parts(2228 modules fitted with one sensortile and 16 front-end chips each)

• parts not easily accessible afterassembly (central position, coolingand radiation)

• every bad pixel degradesperformance

• ≈ 1.4·108 pixel channels in total

Why systematic quality assurance for theATLAS pixel sensor?

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 3

Important steps for QA

Setting technical specifications

Designing according to specifications

Identifying relevant qualities

Defining reliable measurement procedures

Calibrating involved test sites

Archiving measurement data in data base

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 4

Pixel design requirements

• pixel size 50µm x 400µm50 µm pitch12µm diameter bump connection

• total active area 2.3m2 (2228 modules)high yieldtestability

• 10 years operationfault tolerance

• harsh radiation environmentup to 1015 cm-2 (1 MeV neutron eq.)radiation hard technology and design

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 5

Isolation techniquesOptimizing performance before and

after type-inversion and ensuring testability

p-stop p-spray moderated p-spray

For p-spray testing of worst caseof breakdown before irradiation

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 6

Testability

I-V tests on test pixels using punch-through

current throughsingle pixel

current throughpunch-through array

Leakage current indicative for quality of every pixel

1.00E-04

1.00E-03

1.00E-02

1.00E-01

1.00E+00

1.00E+01

1.00E+02

0 20 40 60 80 100 120 140 160 180 200

bias voltage [V]

curr

ent

[nA

]

defective pixel

good pixel

1.00E-01

1.00E+00

1.00E+01

1.00E+02

1.00E+03

1.00E+04

1.00E+05

0 20 40 60 80 100 120 140 160 180 200

bias voltage [V]

curr

ent

[nA

]

good pixel matrix

matrix with defective pixels

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 7

Testability

• Punch-through effect across a biasgrid allows testing of all pixels usingonly two probes on wafer p-side

I-V tests before bonding using bias grid

groundp-side bias

punch-through

guard ring

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 8

Leakage current

• I-V measurements ofleakage current show pixel quality

• breakdown voltageindicates type of defect

• tile classification possible

Tile classification by pixel quality

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 9

Leakage currentYield analysis based on I-V curves

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 10

Sensor depletion

• Test diodes on production waferfor well defined capacitancemeasurements

• Full depletion visible by levellingout of C vs. V-1/2 curve (suppression of possible constantstray capacitances)

Diagnostic measurement bydiode capacitance

Defined n-Bulk area

p+

n+

guard ring

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 11

Oxide characteristicsDiagnostic measurements on

MOS and GCD

• oxide breakdown and capacitancemeasured in I-V and C-V curves onMOS pads

• interface generation currentmeasured on gate controlled diodesas I-V curve around flat-band casewith identical gate and diode voltage

• current step indicates chargedensity on interface

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 12

P-spray dose

• source - drainmeasurement on MOSFETon depleted bulk

• p-spray inverts at highergate voltage: threshold Vth

• source-drain current risesrapidly at Vth

backside voltage Vb

Vgate SourceDrain Vdrain= 0.1V I

n+ n+

p-sprayp+

n-Bulk

Method of measurement

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 13

P-spray dose

• after Vth is identified p-spraydose is given by

using flat-band voltage andoxide capacitance from MOSC-V

Analysis

φp-spray = Cox· (Vth-Vfb) / e

under depleted over depleted

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 14

Radiation hardness tests

Bulk damage testing• after irrad. with 3.1⋅1014cm-2

neutron equivalent protons

• after irrad. with 1015cm-2 neutronequivalent protons (design fluence)

Surface damage testing• after irrad. with 500 kGray lowenergetic electrons (design dose)

• depletion measurement on diode

• I-V measurements on mini chipand diode (small structures)

• interface generation currentmeasurement on GCD• p-spray measurement onMOSFET

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 15

Testing responsibilities

vendor• providing process data

• testing pixel quality on sensortiles on wafer level

• performing diagnostic tests onwafer level for depletion, oxidequality and capacitance, and p-spray dose

ATLAS institutes• checking process data

• testing pixel quality on sensorstiles, single chips and mini chips

• performing all diagnostic testson wafer level and on diced teststructures

• measuring irradiated structures

Pixel 2000, June 2000, Genova QA for rad. hard ATLAS pixel sensors, J.M. Klaiber-Lodewigs - Univ. Dortmund 16

Summary

• technical specifications completed and approved• sensor design completed and approved

• quality test for every pixel on sensor defined and demonstrated• diagnostic tests for relevant qualities defined and demonstrated

• cross calibration of test sites in progress

• data base for test results under construction

Progress of quality assurance process