1/16

Charge Explosion MeasurementsJ. Becker , D. Eckstein, R. Klanner, G. Steinbrück

University of HamburgDetector laboratory

1. Introduction and TCT-principles

2. Set-up available for measurement

3. Measurements on pad diodes

4. Measurements on strip detectors

5. Effects in (200µm)² pixels

6. Simulations (UHH)

7. Summary and outlook

XDAC Meeting 28.4.09, Hamburg

2/16

Introduction

under investigation by Uni-HH

• radiation damage due to very high intensities (up to 1012γ/bunch -> 105 γ/pixel)

-> integrated dose up to 1 GGy

• instantaneous charge deposition-> charge explosion

The multi channel Transient Current Technique (TCT) setup presented here allows studies of the charge explosion in a controlled laboratory environment

For the experiments at the XFEL new detectors have to be build. The AGIPD consortium is building a detector using a hybrid pixel array and commonly available silicon.

talk last session

this talk

3/16

TCT – Principles+-

electronhole

1 keV γ

+-

12 keV γ

electrons and holes drifting

holes drifting

electrons drifting

peak due to holes recombining

XFEL photons can be replaced by laser light with identical attenuation length

1 keV => 3µm => 660nm12 keV => 250 µm => 1015nm

at some XFEL experiments soft X-rays are used (SASE 3)

main XFEL X-ray energy (SASE 1)

charge carrier drift pathelectric field line

zero density limit

4/16

TCT – Principles – Plasma+-

many1 keV γ‘s

• charge carrier influence on electric field no longer negligible, dominating effect for high densities

• field free regions inside the plasma exist -> ambipolar diffusion

• plasma dissolved by diffusive processes -> time consuming -> effect on pulse shape-> charge carrier interactions (repulsion) increases spread of carriers

• at densities O(1020cm-3) (not reached at XFEL) electron hole recombination will occur

When many photons are absorbed:charge carrier density > bulk doping O(1012 cm-3)

-> e,h plasmas are created resulting in:

electronhole

charge carrier drift path?electric field line

5/16

Setup available for measurements

optics

high intensity sub-ns laser

2.5 GHz Scope

linear tables

defined set-up

32 channels

automated control (PC)

Main goals:• Determination of the pulse shape of individual pixels with XFEL type irradiation• Agreement of experimental reference data with simulations (WIAS-Berlin)

Key features:• laser pulses

~ 100ps duration~ up to 105 12 keV γ equivalent~ ≤ 3µm spot size~ 100µm focus depth

• 660nm, 1015nm, 1052nm wavelength*• multi GHz bandwidth electronics

(<100 ps risetime)• 0.1 µm position control• 32 channels

(4 simultaneously)• temperature control

(-35°C to 50°C)• front and backside injection possible

* 660 nm -> 3 µm absorption length ~ 1 keV γ1015 nm -> 250 µm absorption length ~ 12 keV γ1052 nm -> 900 µm absorption length ~ mips

6/16

Measurements on pad diodes

not to scale

not to scale

~ 0.5 µm

~ 0.5 µm

• frontside and backside injection of 660nm and 1015 nm light (~ 1 keV/12 keV γ )

• backside bias, Udep ~ 45 V

|E|

7/16

Measurements on pad diodesspotsize ~2.5µm, scaled to same integral (charge)

backside injection

frontsideinjection

1 keV 12 keV

junctionside injection favoredUbias=100V

8/16

Measurements on strip sensors

injection of 660nm and 1015nm light from backside

pn junction on front side (low field at injection side)

position scan with spotsize ~3 µm

Front

backside bias

structure (strip)

660nm light fixed intensity

x position

readout

attenuation length 3 µm(zone of injection)

strip sensor

same wafer as pad diode

thickness 280 µm

Udep ~ 63 V

pitch 80 µm

width 20 µm

electronhole

charge carrier drift path

9/16

Position sensitive transients 1 keV

9.1 x 106

e,h pairs-> 32’760 1 keV γγγγ

0.8 x 106

e,h pairs-> 2’880 1 keV γγγγ

central hit on center strip neighboring strip

0

10/16

Fitting (black lines) done with gaussian charge carrier distribution and box integration along strip pitch

Collected charge vs. position 1 keV

0.8 x 106 e,h pairs -> 2’880 1 keV γγγγ

increased number of photons

intensity dependence of charge spread

9.1 x 106 e,h pairs -> 32’760 1 keV γγγγ

p

x µ

σ

11/16

Position sensitive transients 12 keV

3 x 109 e,hpairs-> 90’000 12 keV γγγγ

1.57 x 106

e,h pairs-> 470 12 keV γγγγ

central hit on center strip neighboring strip

0

12/16

Collected charge vs. position 12 keVFitting (black lines) done with gaussian charge carrier distribution and box integration along strip pitch, apparent boost due to reflection of laser at metallization

1.57 x 106 e,h pairs -> 470 12 keV γγγγ

p

x µ

σ

increased number of photons

intensity dependence of charge spread

3 x 109 e,h pairs -> 90’000 12 keV γγγγ

13/16

increased number of photons

660nm -> 1 keV γγγγ

8 mA0.24 mApeak current @500V (<1ns)

54 µm / ?

55µm / 28µm

sigma of charge spread @max. intensity 100/500V

20ns / 5.5ns20ns / 5.5nsexpected time (no plasma effects) @100/500V

none15 nsplasma delay (@100V)

>200ns / 22ns

100ns / 11ns

maximum/minimum time (@max int. 100/500V)

55 - 9000030 - 32760investigated intensities (absorbed γ equivalent)

1015nm(12 keV γγγγ)

660 nm(1 keV γγγγ)

Investigated in a 280µm strip detector-> no guaranty for behavior in a 500µm detector

Results of transient evaluation

time [ns]cu

rren

t

plasma delay (660 nm)

measured transients have been distributed to AGIPD ASIC designers

14/16

Effects in (200 µm)² pixels

∫ ∫−+−−=

right

left

top

bottom

pixel dxdyyyxx

N )2

)()(exp(

2

102

20

20

2

5

σπσ

sigma = 50 µm

sigma = 25 µm

91‘113

51.8

2170 24‘997

1x10-4

1.6

99‘988

1x10-4

3.1 25‘000

4x10-26

3x10-11

central hit edge hitestimation of total collected charge for 105 12 keV γ

1x10-4

2x10-6

15/16

Simulations from UHHSimple pseudo-analytical simulation program for pad-, strip- and pixel geometry availableFeatures:

•timing structure of input considered

•exponential charge carrier distribution along depth

•field and temperature dependent mobility model

•trapping time taken into account

pulses for a centrally hit (200µm)2 pixellow photon limit, continuous e,h distribution

Limits:

•linear fields only

•applicable only above depletion

•no charge carrier interaction model (repulsion, recombination, diffusion, etc)-> zero density limit only

•1D simulation code, only simplified geometries

•uses mirror charge method

sensorreadout

photons

16/16

Summary• Better suppression of plasma effects if injection is done at the junction

– n-in-n technology recommended for AGIPD• High bias voltage counteracts plasma effects

– aim for 1000V@AGIPD• Charge spread measured with 660nm and 1015nm light at strip detector

(280µm thickness / 80µm pitch)– sigma of charge cloud up to ~50 µm measured– charge spread >25µm will render neighboring pixels incapable of detecting single

photons• Measured and simulated transients (current pulses) have been distributed to

the AGIPD ASIC designers

Outlook• detailed comparisons to simulations (WIAS-Berlin)• ongoing investigations with 1015 nm laser

(absorption length ~250 µm ~12 keV photons)• reference data for sophisticated detector simulations (WIAS-Berlin)

17/16

Backup

18/16

Simulations from WIAS BerlinSimulations using numerical solutions to the van Roosbroecksystem of (partial) differential transport equations on a Delaunay grid

Major features:

•arbitrary definition of input e/hdistribution possible (space and time)

•diffusive transport model

•field dependent mobility model

•charge carrier interaction models (repulsion, recombination, etc)

•full 3D simulation code capable of modeling elaborate structures

Open points:•so far no simulation of electronics-> done with external circuit simulation (SPICE)

•comparison to measurement data for non-zero density

19/16

Position sensitive transients

9.1 x 106

e,h pairs-> 32’760 1 keV γ

0.8 x 106

e,h pairs-> 2’880 1 keV γ

central hit on readout strip neighbor strip charge on neighbor

20/16

M-TCT ceramics

13x26 mm2 space for d.u.t

32 Pads -> individual channels

2x 16 landing zones for wire bonds

separate pads for ground contact

separate pads for HV contact

90° rotation symmetry

flip symmetry (vias at pads) and central hole for backside injection

21/16

Laser beam characteristics

minimum spotsize

Beam profile colored lines -> measurementblack lines -> gaussian fit

•gaussian spot with minimum sigma of ~2 µm(for 660nm light)

•depth of focus of ~100 µm due to special optics with 75mm focal length

22/16

System calibration

50Ω

Amp

CScopeV

system to calibrateinstead of detector

Injection of a step function (V) over a known capacitance (C) into the system instead of detector. When no charge is lost: is valid

50Ω

∫=× dtRtUVC oscosc /)(

984.0=

=

avg

inj

meas

K

QQK

measinj QQ =

Measurement yields:

23/16

Some words on electronics

Element Bandwidth

Attenuator: DC - 4 GHz

Oscilloscope: DC - 2.5 GHz

Amplifier*: 10 kHz -1.0 GHz

2.5 m Cable ?

Stray L’s, C’s, R’s ?

Pulse shape is electronically smeared!

=> SPICE Model

* only necessary with low intensity injection

actual detector

24/16

Comparison to simple Sim. (HH)

Laser timing structure taken into account

exponential injection decrease taken into account

non focused laser beam

no carrier interactions taken into account

no diffusion taken into account

frontside (electron) injection

25/16

Comparison to simple Sim. (HH)

Laser timing structure taken into account

exponential injection decrease taken into account

non focused laser beam

no carrier interactions taken into account

no diffusion taken into account

backside (hole) injection

26/16

Mobility in simple Sim. (HH)

α

βµ

µµµµ

µµβ

)1(

))(1(

min0min

*0

*0

1*0

ref

eff

sat

C

N

vE

+−+=

+=

for holes µ0* was multiplied by 1.075 to produce transients of the right time

electric field was assumed linear and independent of charge carriers

( )( )( )

( )( )( )( )

( )( )

( )( ) 17.0

1

66.0

12

3126,

1557,

45.0

300min,

45.0

300min,

18.2

300,0

0.2

300,0

2.3

30017

,

2.3

30017

,

065.0

300

46.0

1057.2

tanh1005.9

tanh1045.1

45

80

460

1430

1023.2

1012.1

72.0

2

2

2

2

KT

h

KT

e

TK

hsat

TK

esat

KT

Vscm

h

KT

Vscm

e

KT

Vscm

h

KT

Vscm

e

KT

href

KT

eref

KT

v

v

C

C

=

×=

×=

×=

=

=

=

=

×=

×=

=

−

−

−

−

−

β

β

µ

µ

µ

µ

α

Jacoboni

Selberherr

Jacoboni

27/16

Timing structure of 660 nm laser660 nm linear

0,00 mW

100,00 mW

200,00 mW

300,00 mW

400,00 mW

500,00 mW

600,00 mW

1 1,1 1,2 1,3 1,4 1,5 1,6 1,7 1,8 1,9 2 2,1 2,2 2,3 2,4 2,5

time [ns]

opt.

pow

er

11,51 mW

5,62 mW

2,50 mW

1,77 mW

provided by manufacturer

660 nm log

0,01 mW

0,10 mW

1,00 mW

10,00 mW

100,00 mW

1000,00 mW

0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5

time [ns]

opt.

pow

er

11,51 mW

5,62 mW

2,50 mW

1,77 mW