Transcript RD50_lowRstrips - Indico

Low Resistance Strip Sensors
– Funding request proposal –
CNM (Barcelona), SCIPP (Santa Cruz), IFIC (Valencia)
The problem
 In the scenario of a beam loss, a large
charge deposition in the sensor bulk can
lead to a local field collapse.
 An implant strip potential could reach a
significant voltage due to a conducting
path to backplane.
 AC-coupling capacitors can get damaged
by the voltage difference between the
implant and the readout strip (held to
ground by electronics).
They are typically qualified to 100 V
Miguel Ullán (CNM-Barcelona)
RD50 meeting (Liverpool) – May 2011
2
2
The current solution
 Punch-through (PT) effect is used at Strip end to develop low
impedance to the bias line in the case of voltages in excess of
the (PT) threshold
 Reduce implant distance to bias ring
 Placement of the resistor between the implant and bias rail
(“transistor effect”).
Y. Nobu, et al., NIMA A (2010)doi:10.1016/j.nima.2010.04.080
HPK
Micron
Miguel Ullán (CNM-Barcelona)
RD50 meeting (Liverpool) – May 2011
3
Punch Through effectiveness
Laser tests (SCIPP-Santa Cruz):
T. Dubbs et al., IEEE Trans. Nuclear Science 47, pp. 1902 – 1906, 2000.
 An IR cutting laser is used to deposit large amounts of charge
in the detector bulk
 Bias ring held to gnd and Vbias is swept
 Implant voltage is measured in DC pad through a pico-probe
Miguel Ullán (CNM-Barcelona)
RD50 meeting (Liverpool) – May 2011
4
Punch Through effectiveness II
 “Clamping” of the implant voltage due to PT effect. At a
certain value, the implant voltage saturates and do not follow
the applied bias.
 This is also present after irradiation.
Micron Data - Vnear vs Vbias - Laser Far, 22C
70
60
Vnear (-V)
50
40
30
2684-1 non-irr.
20
2551-7-3 1x10^15 neq neutrons
2551-2 3-3 2.4x10^14 neq pions
10
2551-1 3-4 1.2x10^15 neq protons
0
0
50
100
150
200
250
Vbias (-V)
Miguel Ullán (CNM-Barcelona)
RD50 meeting (Liverpool) – May 2011
5
Punch Through effectiveness III
 Measurements with a large charge injected by a
laser pulse showed that the strips can still be
damaged
 The voltages on the opposite end of the strip keep
rising well above the 100V objective
 The large value of the implant resistance
effectively isolates the “far” end of the strip from
the PT structure leading to the large voltages
C. Betancourt, et al.
“Updates on Punchthrough Protection”
ATLAS Upgrade week,
Oxford, March 31, 2011.
Miguel Ullán (CNM-Barcelona)
RD50 meeting (Liverpool) – May 2011
Near end, plateau
for PT structures
Opposite end, no plateau.
6
6
Newly proposed solution
 To reduce the resistance of the strips on the silicon sensor.
 A desired target value is 1.5 kOhm/cm (~ 1 order of magnitude reduction)
 Not possible to increase implant doping to significantly lower the resistance.
Solid solubility limit of the dopant in silicon + practical technological limits
(~ 1 x 1020 cm-3)
 Alternative: deposition of Aluminum on top of the implant
 A layer of high-quality oxide and nitride with metal strips on top to
implement the AC-coupled sensor readout.
Miguel Ullán (CNM-Barcelona)
RD50 meeting (Liverpool) – May 2011
7
7
Technology
 Metal layer deposition before the coupling capacitance is
defined
 MIM capacitors
 Low temperature deposited isolation (quality)
– PECVD (300-400 ºC)
– Use of an Oxide-Nitride-Oxide Layer to avoid pinholes
(~200 pF  ~1.3 µm )
 Alternatives
– Polysilicon + TaSi (tantalum silicide)
– Tungsten
 Experiments already performed at CNM on smaller
dimensions. More experiments to be performed.
Miguel Ullán (CNM-Barcelona)
RD50 meeting (Liverpool) – May 2011
8
Design
Automatic layout generation with GDSPY
• Python module for creating GDSII
stream files (GDS), published under
the GNU General Public License
(GPL).
 Requires Python 2.x (Python 2.x is
the status quo, Python 3.x is the
shiny new thing).
 Released by Lucas Heitzmann
Gabrielli in May 2009, currently in
version 0.2.6.
• Very usefull to automate mask
design:
 Boolean and slice operations.
 Uses NumPy (scientific
computing)
Miguel Ullán (CNM-Barcelona)
• We are using currently this to
develop the ATLAS-Upgrade
endcap prototypes
RD50 meeting (Liverpool) – May 2011
9
Timescale
1. Initial technological experiments (CNM)
(Ongoing)
2. Design definition and mask design and
fabrication (CNM, SCIPP)
(Started) 1-2 months
3. Device fabrication at CNM clear room facilities (CNM)
4-5 months
4. Electrical characterization of the test vehicles (CNM)
1months
5. Full detector characterization to check for device performance
degradation due to new technology, with Alibava System (CNM, IFIC) 2 months
6. Electrical characterization of the strip resistance and PT
structure performance (SCIPP)
1 months
7. Characterization of sensor protection against beam loss events by
implant voltage measurements under a laser pulse (SCIPP)
3 months
8. Irradiation of devices (neutron/protons) (CNM, SCIPP, IFIC).
4-5 months
9. Device full characterization after irradiation, with
Alibava System (CNM, IFIC)
2 months
10. Characterization of sensor protection against beam loss events
after irradiation (with laser pulse) (SCIPP)
3 months
Miguel Ullán (CNM-Barcelona)
RD50 meeting (Liverpool) – May 2011
10
Funding request
The total project costs are 21 k€ (27600 CHF) as detailed:
– Wafers : 1000 € (1300 CHF)
• 12 FZ HR wafers
– Mask Production : 8000€ (10500 CHF)
– Fabrication of 12 p-type wafers: 12.000€ (15800 CHF)
 The 3 participating institutes ask the RD50 funding for a total
of 15000 CHF (3 x 5000 CHF), and will share the extra
required budget of 12600 CHF to complete the project. The
costs of the required dicing, packaging or wire-bonding plus
the initial experiments will also be covered by the institutes.
Miguel Ullán (CNM-Barcelona)
RD50 meeting (Liverpool) – May 2011
11
Summary
 Low Resistance Strip Sensors Proposal
 CNM-Barcelona, SCIPP-Santa Cruz, IFIC-Valencia
 A run of detectors with the following features:
– Full p-type strip detectors (ATLAS Upgrade-like) (~8 masks)
– PT structures implemented
– Metal layer on top of the strip implant
– Automatic design (+ test ttructures)
 MIM coupling capacitors
 Tests to be performed at CERN and UCSC
 This could be a very innovative solution for new generations of long-strip
detectors and with great applications in future HEP experiments.
Miguel Ullán (CNM-Barcelona)
RD50 meeting (Liverpool) – May 2011
12
12