Impact of the layout on the electrical characteristics of - - PowerPoint PPT Presentation

Impact of the layout on the electrical characteristics of double-sided silicon 3D sensors fabricated at FBK

• M. Povoli1,2
• A. Bagolini3
• M. Boscardin3

G.-F. Dalla Betta1,2

• G. Giacomini3

F . Mattedi3

• E. Vianello3
• N. Zorzi3

1Dipartimento di Ingegneria e Scienza dell’Informazione

University of Trento, Italy

2INFN Sezione di Padova

Gruppo Collegato di Trento

3Centro per i Materiali e i Microsistemi

Fondazione Bruno Kessler (FBK), Trento, Italy

8th "Hiroshima" Symposium(HSTD-8) December 05 - 08, 2011

[Work supported by INFN CSN V, projects "TREDI" (2005-2008) and "TRIDEAS" (2009-2011), and INFN CSN I, project ATLAS]

Outline 3D detectors Full 3D and simplified approaches Current status at FBK (short summary) Electrical characterization of 3D test structures with different layouts Available layouts of 3D diodes I-V measurements with variable temperature Selected simulation results for different devices Layout effects on detector capacitance Layout effects on electrical quantities inside the devices Conclusions and outlook

2 / 16 Povoli et al.

3D silicon detectors

Original idea and simplified approaches

[S. Parker et. al. in NIMA 395 (1997), 328] Full 3D

• 1. Single side process
• 2. Passing through

columns

• 3. Active-edge

[E. Vianello et al., NSS11, Paper N10-6] [G. Pellegrini et al., NIMA 592 (2008) 38] Simpilfied approaches

• 1. Double side processes
• 2. Passing through columns (FBK),

non-passing through columns (CNM)

• 3. Slim-edge (FBK), 3D guard ring (CNM)
• 4. P-spray (FBK), p-stop (CNM)

3 / 16 Povoli et al.

Current status at FBK (short summary)

[E. Vianello et al., NSS11 Conference Record, Paper N10-6]

Considerations ◮ ATLAS07: bad currents and low breakdown (p-spray too high) ◮ ATLAS08: better current (less mechanical stress) still low breakdown ◮ ATLAS09: good currents and higher breakdown (p-spray adjusted) Investigation ◮ Try to gain insight into device behavior using TCAD ◮ Measured p-spray, N+ and P+ profiles ◮ Disentangle the effects of each component of the device ◮ Test performed on 3D diodes (two terminal devices, easy to test)

4 / 16 Povoli et al.

FE-I4 3D diode with Field Plate, Layout details

FRONT BACK

◮ Layout compatible with the ATLAS FE-I4 pixel ◮ Minimum N+ to P+ distance equal to 15µm ◮ WITH Field Plate (LFP = 4µm) ◮ Back side metal and P+ patterned ◮ Simulated cell: 25 × 62.5 × 230µm3

5 / 16 Povoli et al.

Other available 3D diodes FE-I4 3D diode

◮

Similar to the previous FE-I4 diode

◮

N+ columns connected only through metal

◮

Larger N+ to P+ distance

◮

NO Field Plate

CMS 3D diode

◮

Layout compatible with the CMS pixel detector

◮

N+ columns connected only through metal

◮

Minimum N+ to P+ distance equal to 18µm

◮

NO Field Plate

◮

Simulated cell: 50 × 75 × 230µm3

80µm 3D diode

◮

80µm pitch

◮

N+ and metal grid both on front and back sides

◮

Minimum N+ to P+ distance equal to 20µm

◮

WITH Field Plate (LFP = 5µm)

◮

Simulated cell: 40 × 40 × 230µm3 6 / 16 Povoli et al.

IV measurements with variable temperature

IV Curves

1e-08 2e-08 3e-08 4e-08 5e-08 6e-08 7e-08 8e-08 10 20 30 40 50 60 70 80 Currents [A] Reverse voltage [V] I-V 80BIG 35°C 30°C 25°C 20°C 15°C 10°C 5°C 0°C

• 5°C
• 10°C
• 15°C
• 20°C

1e-08 2e-08 3e-08 4e-08 5e-08 6e-08 7e-08 8e-08 10 20 30 40 50 60 70 80 Currents [A] Reverse voltage [V] I-V FEI4 35°C 30°C 25°C 20°C 15°C 10°C 5°C 0°C

• 5°C
• 10°C
• 15°C
• 20°C

1e-08 2e-08 3e-08 4e-08 5e-08 6e-08 7e-08 8e-08 10 20 30 40 50 60 70 80 Currents [A] Reverse voltage [V] I-V CMS 35°C 30°C 25°C 20°C 15°C 10°C 5°C 0°C

• 5°C
• 10°C
• 15°C
• 20°C

1e-08 2e-08 3e-08 4e-08 5e-08 6e-08 7e-08 8e-08 10 20 30 40 50 60 70 80 Currents [A] Reverse voltage [V] I-V FEI4-FP 35°C 30°C 25°C 20°C 15°C 10°C 5°C 0°C

• 5°C
• 10°C
• 15°C
• 20°C

Setup ◮ Devices coming from wafer W20 of the ATLAS09 batch ◮ Devices were diced and wire bonded on small PCBs ◮ Wire bonding contribution is negligible ◮ Temperature variation between −20◦C and 35◦C inside the climatic chamber ◮ Measurements performed with HP4145 Preliminary results ◮ Each type of device has its own characteristic behavior ◮ Breakdown voltages between 40 and 50V ◮ Different current slope for different devices ◮ More details in the next slide...

7 / 16 Povoli et al.

IV measurements with variable temperature

Breakdown voltage vs. temperature

35 40 45 50 55 60

• 20
• 10

10 20 30 40 Breakdown voltage [V] Temperature [°C] VBD= 48.50 mV/°C VBD= 76.82 mV/°C VBD= 50.72 mV/°C VBD= 55.42 mV/°C 80BIG FEI4-FP CMS FEI4

Breakdown voltages ◮ Devices breakdown between 40 and 50V ◮ Linear increase with temperature ◮ The increase is between ∼ 50 and ∼ 80mV/◦C ◮ In agreement with the expectation

[Crowell, C. R. and S. M. Sze, Appl. Phys.

• Lett. 9, 6 (1966) 242-244.]

8 / 16 Povoli et al.

IV measurements with variable temperature

Purpose

◮

Distinction between thermal generation and avalanche generation Equation I(T) = I(TR)

• T

TR 2 exp

• E

2kB

• 1

TR − 1 T

• ◮

TR = 293.15◦K

1e-11 1e-10 1e-09 1e-08 1e-07 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 Currents [A] 1000/T [°K-1] Arrhenius plot - 80BIG

• Meas. at 10V
• Calc. at 10V
• Meas. at 30V
• Calc. at 30V

1e-11 1e-10 1e-09 1e-08 1e-07 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 Currents [A] 1000/T [°K-1] Arrhenius plot - FEI4

• Meas. at 10V
• Calc. at 10V

Meas at 30V

• Calc. at 30V

1e-11 1e-10 1e-09 1e-08 1e-07 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 Currents [A] 1000/T [°K-1] Arrhenius plot - CMS

• Meas. at 10V
• Calc. at 10V
• Meas. at 30V
• Calc. at 30V

1e-11 1e-10 1e-09 1e-08 1e-07 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 Currents [A] 1000/T [°K-1] Arrhenius plot - FEI4-FP

• Meas. at 10V
• Calc. at 10V
• Meas. at 30V
• Calc. at 30V

Results

◮

The calculation was performed at different bias voltages for all the temperatures considered

◮

Good agreement at low biases (e.g 10V)

◮

Closer to breakdown the agreement is lost (e.g. 30V)

◮

Avalanche generation adds up to Shockley-Read-Hall (SRH) generation

◮

FEI4 diode without field plate seems to show breakdown behavior before the others!! 9 / 16 Povoli et al.

Parameters, models and data extraction

Structure parameters

◮

Thanks to symmetry only 1/4 of elementary cell is simulated

◮

Dimensions depending on the simulated layout

◮

Thickness: 230µm

◮

Measured xide charge: 3 × 1011cm−2

◮

Measured oxide thickness: 1µm

◮

Measured (SIMS) doping profiles for p-spray and N+ and P+ diffusions Models and simulation

◮

Mobility: Doping Dependent, High Field Saturation, E-Field normal

◮

Generation/Recombination: SRH, Avalanche

◮

Bias ramp applied on the P+ electrode from the back side Data analysis

◮

Monitor electrical quantities in different regions of the structure

◮

Understand where the breakdown is occurring

◮

2D slices at different depths

◮

Visualization of the most important quantities for our purpose

◮

1D cut extraction from 2D slices to look at the distribution of the electrical quantities 10 / 16 Povoli et al.

C-V measurements vs. C-V simulations

Only available for 80µm diode (Work in progress)

50 100 150 200 250 300 350 400 10 20 30 40 50 Capacitance [pF] Reverse voltage [V] Only columns Columns and p-spray with N+ diffusion with P+ diffusion Full structure Measured Results

◮

Electrodes contribution: ∼ 71pF (constant)

◮

p-spray causes an increase of ∼ 35pF (basically constant)

◮

P+ diffusion and metal on the back side does not cause much increase

◮

N+ diffusion and metal on the front side cause an increase of ∼ 41pF at a bias voltage of 20V

◮

At higher biases the contribution of front side saturates to a value similar to the one obtained only with electrodes and p-spray (∼ 111pF)

◮

Measured capacitance does not saturate at 40V: more voltage required

◮

Results for other devices available soon

Selected simulation results (summary)

◮ C-V simulations are in good agreement with the measured values and

help us to understand and possibly optimize the devices

◮ I-V simulations were also performed (more details in the next slides) ◮ Good agreement on reverse currents and breakdown voltages ◮ FEI4 simulated current is lower than expected (grid issues? checking...) ◮ CMS diode simulations still running (bigger structure, more points) ◮ Thanks to the simulator is possible to better understand what is

happening inside the devices

Diode type Imeas@10V Isim@10V VBD(meas) VBD(sim) [nA] [nA] [V] [V] 80BIG 6.53 4.39 51 48 FEI4 9.27 4.35 39.5 40 FEI4-FP 5.38 4.77 46 43 CMS 5.22 in progress 40 in progress

12 / 16 Povoli et al.

Selected simulation results - FE-4 diode (VB = −35V) E-field

• Elec. Potential

1D slices

Electric Field distribution

◮

High field at the N+/p-spray junction on the front side (2.65e5 V/cm)

◮

High field at the N+ column/p-spray junction on the back side (2.48e5 V/cm)

◮

E-field also increases under the front side metal (vertical and horizontal metal connections) Electrostatic potential

◮

P+ column brings the bias voltage directly on the front side p-spray

◮

Back side p-spray is also biased at high voltage

Selected simulation results - FE-4 FP diode (VB = −35V) E-field

• Elec. Potential

1D slices

Electric Field distribution

◮

High field at the N+/p-spray junction on the front side (2.11e5 V/cm)

◮

High field at the N+ column/p-spray junction on the back side (2.53e5 V/cm)

◮

The field-plate helps in redistributing and lowering the E-field on the front side Electrostatic potential

◮

The field-plate helps depleting part of the p-spray close to the N+ diffusion

◮

The potential of the p-spray now drops on a wider region

◮

Higher breakdown voltage (breakdown probably

• ccurs on the back side)

Conclusions and outlook

◮ We have studied the electrical behavior of different 3D diodes produced

at FBK and found it consistent with the expectations

◮ I-V measurements allow to find the intrinsic breakdown voltages of the

devices and to highlight differences related to device layout

◮ C-V measurements and simulations showed that for lower operating

voltages also the front and back side surfaces contribute to the overall capacitance of the device

◮ A good agreement was found between measurements and simulations,

confirming that TCAD is a reliable tool for device analysis and design (3D simulation compulsory)

◮ We will apply the same step-by-step procedure to I-V simulation in

• rder to optimize the sensor layout

◮ This study will then continue with irradiated devices (several 3D diodes

will be irradiated with 800 MeV protons at different fluences at Los Alamos later this month)

15 / 16 Povoli et al.

Thank You!

(...any questions?)

Acknowledgement: We would like to thank all members of the Processing Group within the ATLAS 3D Sensor Collaboration for fruitful discussions