3D Sensor Studies at New Mexico

1
3D Sensor Studies at New Mexico

• Sally Seidel for Martin Hoeferkamp,
• Igor Gorelov, Elena Vataga, and Jessica Metcalfe
• University of New Mexico

2
Introduction

• We have characterized 3D sensors of pitch 200 µm
  100 µm.
• We report probe station based studies of
  depletion voltage, leakage current, electrode
  capacitance, capture time, and signal rise time,
  supported by simulations
• The devices non-irradiated and irradiated (1014,
  21014, 1015 cm-2 55-MeV-p), unannealed, from
  Sherwood Parker.

3
Equipment

• Picoprobe Model 35 (26 GHz bandwidth, 14 pS rise
  time, 0.05 pF capacitance)
• Picoprobe Model 12 (500 MHz bandwidth, 0.8 nS
  rise time, 0.1 pF capacitance)
• Kentech APG1 Pulser (300 pS pulse width)
• Tektronix 7254B Oscilloscope (2.5GHz bandwidth)
• 1064 nm, 960 nm, 820 nm IR lasers, 12GHz
  Photoreceiver
• Cascade REL-6100 semiautomatic probestation
• Micromanipulator HC-1000 Thermal Chuck (-60C)
• Peltier Thermal Chuck (-20C)
• EichhornHausmann MX203 wafer thickness and
  flatness gauge

4
3D Sensor Configuration

• Configuration of the devices lt100gt p-type
  silicon.
• Alternating columns of n- and p-electrodes
• Most electrodes are connected together along each
  column
• Some electrodes are left isolated, to be
  contacted and measured individually

• Layout dimensions
• 200 mm x 100 mm spacing,
• 17 mm electrode diameter,
• 121 mm electrode length.

• Top view layout

5
Electrode Leakage Current

• Measured leakage current versus fluence

• Prior to irradiation, the n-electrodes are
  shorted together by a surface electron layer.

6
Electrode Depletion Voltage

• Pixel cell depletion voltage measured via LCR
  meter

• Pixel cell depletion voltage measured via pulse
  height

7
Array Depletion Voltage

• To test the entire device, we completely flood
  the 3D sensor with a uniform 1064 nm laser spot
  and scan the bias voltage above full depletion.
• Photo with IR filter of laser illuminating the
  sensor

8
Array Depletion Voltage

• Array depletion measured from signal efficiency
  (pulse height relative to the maximum for the
  non-irradiated device) versus bias

• Result very low values of depletion voltage for
  the entire sensor array,
• Vdepletion 15V for non-irradiated sensor
• Vdepletion 60V for sensor irradiated to 2x1014
• Vdepletion 130V for sensor irradiated to 1x1015

9
Electrode Capacitance

• Electrode capacitance using standard (HP4284A)
  LCR meter techniques

• Electrode capacitance versus fluence, type, and
  frequency
• Fluence (cm-2) Electrode Frequency (kHz)
• (55-MeV-p) type 10 100 1000
• 0 p 71 58 46
• 0 n 59 38 32
• 21014 p 96 69 53
• 21014 n 72 72 62
• 11015 p 98 70 55
• 11015 n 91 80 60

10
Electrode Capacitance

• Electrode capacitance versus temperature and
  frequency

11
Electrode Capacitance

• Direct measurement is checked by indirect
  measurement through signal decay time

• Indirect measurement using decay time of IR pulse
  on an isolated electrode.
• Electrode is grounded through input impedance of
  a Picoprobe 35.
• The IR laser induced charge is collected.
• When the laser is turned off the signal decay
  follows an exponential with a time constant
  R(CC3D) , referred to here as RC time constant.
• C3D is extracted from the decay time constant
  using values of probe resistance and capacitance.

12
Electrode Capacitance

• Performed at different bias voltages, using the
  procedure of Parker et al., Proc. IEEE Trans.
  Nucl. Sci., Oct 2001, p. 1635 Isolated electrode
  grounded through the 1.25 MO input impedance of
  the picoprobe. T 0 when the laser is turned
  off. After the light emission ends and the
  charge is collected, the pulse height follows an
  exponential of time constant 177 ns. Averaging
  the values for 50 V to 100 V gives a p-electrode
  capacitance of 91.6 fF.
• Irradiated 21014 cm-2 55-MeV-p sensor
  p-electrode

13
Electrode Capacitance

• A summary of the capacitance versus fluence for a
  p- and an n-electrode using the direct
  capacitance measurement technique and for the
  p-electrode using the indirect measurement
  technique. The indirect measurement gives about a
  50fF higher result.

14
Electrode Capacitance Calculation

• 3D electrostatic calculation (IES Coulomb)
• p electrode length 121 µm
• p electrode diameter 17 µm nominal
• Center electrode to nearest neighbors

Prediction for p electrode 28 fF
We are systematically varying the geometrical
parameters to understand the impact of each one
on capacitance. An example for electrode
diameter
Capacitance at 17 mm is 28 fF
15
Position Scans

• Scan the laser across one electrode cell to
  measure uniformity of signal collection

Y
X
16
Position Scans

• Signal collection versus position

• Non-irradiated 3D sensor, p-electrode

Y
X
17
Charge Collection

• Pulse the IR Laser as fast as possible and
  observe the rise time of the signal

• Measure the output rise time while reducing the
  laser pulse duration

18
Charge Collection

• Input 0.3 nS laser duration

• Output irradiated (1015) p electrode,
• 1.5 nS rise time

• Output non-irradiated p electrode,
• 2.5 nS rise time

NOTE The system isolation was improved, and a
broken cable shield replaced, after this
measurement was recorded. Revised graphs are in
preparation.
19
Capture Time

• For an irradiated 3D sensor, pulse the laser at a
  distance of 30 µm from the electrode and measure
  the output. Repeat with laser pulse at a
  distance of 90 µm.

20
Capture Time

• The 60 µm difference in laser position results in
  a collection time difference of 50.6 nS 47.4
  nS 3.2 nS

NOTE The system isolation was improved, and a
broken cable shield replaced, after this
measurement was recorded. Revised graphs are in
preparation.
21
Plans

• Plans for 2007-2008
• Repeat charge collection and capture time
  measurement with new low-noise system.
• Complete systematic simulation of full scope of
  geometrical options.
• Implement TCAD device simulation for improved
  capacitance and charge collection prediction.
• Systematics studies with 820 nm and 960 nm
  lasers.
• Irradiate ATLAS geometry devices at LANL and
  Sandia.
• Apply these measurement techniques to the ATLAS
  geometry devices.
• There is a larger range of measurements we would
  like to do additionally if a
• TurboDAQ system becomes available.
• We are 5 people available for testbeam staffing
  as well.

22
Budget for FY 2008

• 110,000 for electrical engineer, travel, and
  materials and supplies.