Experimental Project Report

on

Characterization of Single Sided Silicon

Microstrip Detector

Submitted by

Saurabh Sandilya

Under the guidance of

Dr. Gagan Mohanty

&

Prof. Kajari Mazumdar

Tata Institute of Fundamental Research

Mumbai 400 005

[10th Feb 2010]

2

3

Characterization of Single Sided Silicon Microstrip

Detector

Abstract

This report describes the different tests carried out on Single sided Silicon microstrip Detectors. We

have characterised different prototypes of the silicon microstrip detectors, which are fabricated at

Bharat Electronics Ltd (BEL), India with the feasibility of future upgradation in the Silicon Tracker of

the CMS experiment of CERN and of the Belle experiment at KEK. We have done the study of current-

voltage, capacitance-voltage characteristics, quality control testing like pin-hole test and also tested

them with IR LASER to get a feel of the position resolution.

1. Motivation

Silicon microstrip detectors played a crucial role in the discovery of top quark (third generation of

quarks), it continues to play a key role in the CMS and LHC experiments and will be important in the

future discoveries of the “most wanted” particle in high-energy physics, the Higgs boson. [1]

The CMS tracker at CERN is the largest silicon tracker ever built. The inner tracking system of the

CMS is designed to give precise and efficient measurement of the trajectories of charged particles

resulting from the proton-proton collision at LHC, therefore the detector needed to be precise and

should have fast timing response. The flux of the particle is also so intense that it can cause severe

radiation damage to the tracker. In order to fulfil these multiple requirements of granularity (position

resolution), fast timing response and radiation hardness silicon microstrip detectors are preferred. [2]

2. Introduction

Semiconductor detectors, and in particular silicon detectors have several advantages over other types

of radiation detectors – better energy resolution, linear response over a broad energy range, fast timing

response, flexibility of design, tolerance to high energy doses etc.

In semiconductor detector, the electron hole pairs are produced by the interaction of an ionizing

radiation and are collected by the application of an electric field to provide a signal proportional to the

total quantity of ionization and therefore to the absorbed energy.

To achieve this, the detecting medium should have the following properties:-

(a) Low values of the average energy to produce an electron-hole pair is desirable so that the

fluctuation in the total number of electron-hole pair produced is small.

(b) The charge carrier must move readily through the detector material, such that their collection

time should be smaller than the carrier life time to complete the collection of free charges.

(c) Leakage current resulting from the application of a fairly large electric field (~1000V.cm-1

, in

order to collect charges), must be very small so that the tiny signal from the transient current

can be measured. Therefore, high resistivity semiconductors are used. [3]

Silicon detectors fit very well to criteria for a detecting medium we had discussed above. We usually

use silicon because it has a band gap of ~1.1eV and can be used as a detector at room temperature,

while germanium with a band gap of ~0.7eV is required to be cooled (using liquid nitrogen) to bring

down the leakage current to an acceptable level. On the other hand since germanium has higher

4

atomic number, it is well suited for gamma-ray spectroscopy (as photoelectric cross-section varies as

Z5)

2.1 p-n junction as a semiconductor detector

In spite of the high degree of purity in which silicon or germanium is available, it cannot be used as a

detector, because the transient current of the signal will be completely masked by the noise current

(or, leakage current). So, it is essential to reduce the leakage current largely in the semiconductor

medium to use it as a detector. For this, it is required that the positive contact must not inject holes

and negative contact must not inject electrons into the bulk medium. [3]

In semiconductor detector it is achieved by providing “blocking contacts” to prevent free exchange of

holes and electrons between the semiconductors and the electrodes (to collect signal). A common

method of achieving this is to form p-n junction just beneath the surface of the detector and operated

with a reverse bias voltage, whereby the majority carriers cannot be injected into the detector material.

These detectors are produced by forming a heavily doped n-type layer (n+ layer) on the surface of p-

type material or a heavily doped p-type layer (p+ layer) on the surface of n-type material.

2.2 Silicon Microstrip detector

The position resolution of silicon detectors is achieved by dividing a large area p-n diode into many

small regions, pixel like or parallel strip like, and each pixel or strip like division act like an

independent electrode.

The detectors (single sided) that we have used are fabricated using n-type silicon wafers as bulk

material. The p+

strips are used to deplete the n-type bulk material and act as electrodes. When an

incident ionizing particle interacts with the detector medium (depleted region), it creates electron-hole

pairs, the holes drift towards the p+ (negative electrode) and the electrons drift towards the n

+ (positive

electrode) due to electric field in the depletion region. The collected charges produce a current pulse

which is read out. Hence, the strip which produces the signal gives the position of the ionizing

particle. Now the magnitude of the signal measured on a given strip depends on its position relative to

the site of charge formation.

Figure 1: Operational principle of a single-sided p

+ n Silicon detector. Figure shows reversed biased strip

diodes, operated at voltage sufficient to fully deplete the n-type bulk material.

Silicon microstrip detectors are generally fabricated by passivated planar techniques, which combine

the technique of ion implantation and photolithography.

In this process at first n-type silicon is taken as a substrate, it is then chemically cleaned and oxidised

by heating it in O2 atmosphere at around 1000oC to have whole surface passivated. Next, selected

5

areas of oxides are removed by using photolithography techniques. The opened areas now acts as a

window, enables doping of silicon in desired areas (by ion implantation techniques). Radiation

damage caused by ion implantation technique is removed by annealing. Finally surface of wafer is

metalized by Al and by proper masking desired pattern is generated. The oxide can be left to make a

capacitor (integrated coupling capacitors).

3. Specification of the detector

The detector that I have tested was single sided silicon microstrip detector (Detector # 2 and 8004-7),

whose mask designing was done at TIFR and was processed at BEL, India. Detector # 2 was having

11 sets of 32 strips (two of them set numbers 5 and 10 was having 30 strips) with different strip width

and pitch and detector # 8004-7 was having 1024 strips with fixed strip width and pitch.

Table 1: Specification of detector # 2

Wafer n type Silicon, 4inch Diameter, 300 micron

thickness

Resistivity

5 KΩ-cm

Number of independent

set of detectors

11

Type of implantation

for strips

p+

Number of strips

Per set

32

Table 2: specification of each set of detector # 2

Set number Strip length (μm) Width

(μm)

Pitch

(μm)

Number of strips

1 74734 12 65 32

2 74734 48 73 32

3 74734 12 80 32

4 74734 20 80 32

5 74734 35 80 30

6 74734 25 100 32

7 74734 35 100 32

8 74734 25 120 32

9 74734 35 120 32

10 74734 48 120 30

11 74734 25 135 32

Table 3: Specification of detector # 8004-7

Wafer n type Silicon, 4inch Diameter, 300 micron

thickness

Resistivity

2-4 KΩ-cm

Type of implantation

for strips

p+

Number of strips

Per set

1024

6

The actual area of the detector # 8004-7 is 79600μm x 28400μm (and the effective area of the detector

is 76800μm x 25600 μm). Strip width is 50 μm and pitch is 75 μm and length of each strip is 26500

μm.

Figure 2(a) Figure 2(b)

In the above figure 2(a) and 2(b), detector # 8004-7 and detector # 2 has been shown respectively.

Figure 3: Photograph of a corner of single sided silicon microstrip detector

The individual strips (p+ strips) are connected to “bias line” through polysilicon resistance such that

we can apply a same fixed potential in all strips. The resistance for bias resistance are kept high

enough to separate the strip from others.

7

The detector strips are surrounded by guard ring which prevents charge accumulations in the silicon

edges, and thus leakage current is prevented from being absorbed by the edge strips. It also degrades

the potential towards cut edge of silicon.

4. Reverse current-Voltage (I-V) Characterization

For measuring I-V characteristics of the detector, we applied reversed bias voltage and measured the

leakage current (or, reverse current). For a given bias voltage, reverse current depends to a certain

extent on environmental factors. So, we had carried out the experiment in “class 10000” room, where

temperature, humidity, dust and other environmental factors were controlled.

We used mechanically controlled biasing probe for giving negative potential to the bias pad by

looking in to the microscope. Positive potential was applied to the “chuck” on which detector rests

and fixed by using vacuum. The detector was placed in a box for minimising the effect of external

light. The voltage applied and the reverse current was measured by “Keithley 237”, and we have used

LabVIEW as interface.

The measurements were performed at room temperature, typically about 21oC and at a relative

humidity around 64%.

Figure 4: Schematic diagram of the experimental setup for measuring I-V characteristics.

We increased the reversed biased voltage in the step of 10V, (using Keithley 237) maximum up to

350 V or up to breakdown if observed before. We did this experiment for all the 11 sets of detector #

2 and for detector # 8004-7 and we obtained I-V characteristics as shown in figure 5 and figure 6

respectively.

Figure 5 shows block diagram of the connection made for obtaining I-V characterstics.

8

Figure 5: Reverse current as a function of biased voltage of detector # 8004-7. It shows the breakdown

voltage to be at 310V

Figure 6: Reverse current as a function of biased voltage of different sets of detector # 2

0

1

2

3

4

5

6

7

8

9

10

0 30 60 90 120 150 180 210 240 270 300 330

Rev

erse

Cu

rren

t in

(μ

A)

Reversed voltage applied (V)

I-V Characterstic of Detector # 8004-7

0

1000

2000

3000

4000

5000

6000

7000

8000

0 20 40 60 80 100 120 140 160 180 200 220 240 260 280

Re

vers

e C

urr

en

t (i

n n

A)

Reverse Voltage applied (in Volts)

I-V Characteristic of Detector # 2

Series1

Series2

Series3

Series4

Series5

Series6

Series7

Series8

Series9

Series10

Series11

9

Figure 7: Shows most of the sets (set numbers 1, 5, 6, 7, 8, 9, 10, and 11) were having breakdown voltage

between 210V to 260V

Result: The determination of I-V characteristic of the detector, gives an idea about its operating

voltage. The operating voltage should be high enough to get the signal and should be below to the

breakdown voltage. We found that some of the sets (2, 3 and 4) were having breakdown voltage

(below 50V) too low to be used as a detector.

5. Capacitance-Voltage (C-V) characteristics

In this experiment our aim was to look at the variation of bulk capacitance by varying reverse biased

voltage. The depletion region due to the fixed charges built on either side of the junction behaves as

charged capacitor. By increasing the reverse biased voltage the depletion width increases and hence

capacitance decreases. If a heavily doped p+ layer forms junction with n type layer the detector

capacitance per unit area is given as. [3]

C = 21.0 X 103 ρnV −1/2 . pf.cm-2

... (1)

Where, ρn is the resistivity of the n type Silicon (in detector #2 resistivity of n-type silicon wafer is

5000Ωcm), and V is the applied reverse bias voltage.

By putting the value of ρn = 5000Ωcm. Equation (1) becomes,

C = 2.97 X 102 V −1/2 ... (2)

Taking log of the equation (2).

ln 𝐶 = 5.694 − 0.5 ln 𝑉 . ... (3)

So, we will fit a straight line and we will get coefficients experimentally determined.

For applying reverse bias voltage we have used “Keithley 237”, its positive terminal was connected to

the chuck and negative terminal was connected to “bias pad” by using a mechanical probe. Similarly

for measuring capacitance we have used “HP 4284” LCR meter, its one end is connected to the chuck

(back plane) and other one was probed to bias pad.

0

50

100

150

200

250

300

1 2 3 4 5 6 7 8 9 10 11

Bre

akd

ow

n v

olt

age

(V

)

Set numbers

Comparison of breakdown voltage of all sets of det #2

10

Figure 8: Schematic diagram of the connections to be made for obtaining C-V characteristic.

We increased the reverse biased voltage (till stability in capacitance or breakdown whichever is early)

by using LabVIEW as interface to Keithely 237.Room temperature was at 20.6oC and humidity at

61.2%RH. We did our experiments on two sets, set number 1 and set number 3 of detector.

Figure 9(a) Figure 9(b)

Figure 10 (a) Figure 10(b)

Figure 9(a) and 10(a) shows the variation of bulk capacitance of detector#2 (of set1 and set3 respectively)

with the applied reverse biased voltage. In Figure 9(b) and 10(b) we have plotted the capacitance and

voltage in logarithmic scale.

0

20

40

60

80

100

120

0 50 100

Cap

acit

ance

(p

F)

Reversed voltage (V)

C-V Characteristics of set1 of det#2

0

50

100

150

200

250

300

0 10 20 30 40

Cap

acit

ance

(pF

)

Reverse Voltage(V)

C-V Characteristic of set3 of det#2

11

Result: From figure 9(b) and 10(b) we obtained the co-efficient of ln (V) in equation 3 (by fitting

straight line and getting its slope) experimentally, which is -0.46 ± 0.06 (in figure 9(b) in case of set1

of detector 2) and -0.65 ± 0.07 (in figure 10(b) in case of set1 of detector 2). The value of intercept

was found 5.5 ± 0.2 (in figure 9(b) in case of set1 of detector 2) and 5.9 ± 0.2 (in figure 10(b) in case

of set1 of detector 2). Which are in accordance with the equation 3, in that value of slope (co-efficient

of ln(V)) is -0.5 and intercept is 5.7.

6. Pinhole Tests

Pinhole test is a test of coupling capacitance integrated on the detector for readout from each strip.

Sometimes the coupling capacitors (SiO2 layer) between the implant and the aluminium readout strips

are broken, which allow excessive current to enter the readout chips which can spread over several

channels, these strips are referred as “pinholes” and makes strip unsuitable for particle tracking.

Pinhole is the direct contact between the implant and the readout strip through the SiO2 layer. [4]

To detect the pinhole in our detector (detector # 2) we applied 30V (DC, using Keithley 237) across

AC pad and DC pad by probe pins, and measured the current (using Keithley 237). Block diagram of

the connections made are shown in figure 12.

Figure 12: Block diagram showing connections to be made for the pinhole testing

Table 4: showing strips with pinhole defects

Set Number Number of

pinholes found

Strip in which

pinholes found

Current drawn

(μA)*

1 1 30 2.90

2 0 - -

3 0 - -

4 2 4 2.89

5 2.46

5 0 - -

6 2 7 2.81

26 2.75

7 1 8 2.80

8 1 28 2.80

9 0 - -

10 0 - -

11 0 - -

*Other strip (pinhole defect free) capacitors were giving current of the order of 0.01nA.

Result: We have checked each strips and found number of pinholes 1 in set1 (strip 30); 2 in set 4

(strip 4 and 5); 2 in set 6 (strip 7 and 26); 1 in set 7 (strip 8) and 1 in set 8 (strip 28) and rest strips

were found pinhole defect free.

12

7. Laser Test

To get a feel of position resolution of the single-sided silicon microstrip detector, we irradiated pulsed

LASER (1064nm wavelength, Q switched, 10ns pulse width) one of the strips of detector # 2 and we

observed the pulse height given by the different strips of the set on CRO. We used a filter in between

LASER and detector (NE 60B of thickness 1.3mm). We applied reverse biased voltage of 20V

(negative terminal probed to bias pad and positive terminal connected to the chuck). Then the signal is

taken from the AC pad of the strips to the CRO.

Figure 13(a) Figure 13(b)

Figure 13(a) is a photograph of our experimental setup of LASER test and figure 13(b) gives a closer

view.

Each time we changed the pin probing to the AC pad of strip and taken the pulse height.

Figure 14 is a snapshot of pulse in CRO

Then we have plotted Pulse height versus strip number which is shown in figure 15.

13

Figure 15: Gaussian is fitted to pulse height vs. Strip number graph.

Result: we have got a very nice Gaussian, peaked at strip number 8 (mean = 7.4 ± 0.1) and having σ ~

2.0 ± 0.1 strips (Because the LASER beam was broad enough). This gives us a realization of postion

resolution property of single sided silicon microstrip detector.

8. Discussions

While doing experiments we got a hands-on experience on single sided silicon microstrip detector.

We learnt and understood the operating principle of silicon microstrip detector as good position

resolution high energy particle detector. The significance of the quality control tests on the detectors

(det # 2 and det # 8004-7) was understood. Later a high energy simulation with a LASER was studied

to get a feel of position resolution property the detector.

9. Acknowledgement

I am highly indebted to Dr. Gagan B. Mohanty and Prof. Kajari Mazumdar, for their valuable

suggestions, and explaining me the role of the “Silicon Microstrip Detector” in CMS and Belle. They

were always available for discussion on the physics part of our project. I would like to thank Mrs.

Mandakini R. Patil for introducing the experiment, for her kind support and providing me study

materials. She kept me motivated throughout the work and guided me in the best possible manner.

I owe my sincerest gratitude to Prof. Tariq Aziz for discussing the experiment with me and for giving

valuable suggestions.

Last, but not the least, I thank the Subject Board of Physics, TIFR for giving me the opportunity to

work in the Silicon Detector Laboratory, of the EHEP group

14

References

1. “The Silicon Microstrip Detector”, A. M. Litke and A. S. Schwarz, Scientific American,

May 1995, p 76-81.

2. “The CMS experiment at the CERN LHC”, CMS collaboration. Institute of Physics

publishing and SISSA. JINST 0803:S08004, 2008

3. “Nuclear Radiation Detectors”, S.S. Kapoor, V. S. Ramamurthy. Wiley Eastern Limited.

4. “Integrity of the Coupling Capacitor on Prototype Hamamatsu Silicon Detectors”. T.

Bollinger et. al CDF/DOC/SEC_VTX/GROUP/3740. July 25, 1996

5. “Radiation Detection and Measurement” G. F. Knoll. Wiley India Edition. (Third

Edition.)

6. “Semiconductor Detector Systems”. H. Speiler. Oxford Science Publications.

7. “The CMS Silicon Tracker”, L. Feld, 1. Physikalisches Institut, RWTH Aachen GSI

Darmstadt, 18. 4. 2007

8. “Physics and Technology of Semiconductor Devices” A. S. Grove, John Wiley and Sons.