INTERACTION OF RESIDUAL HYDROGEN WITH RADIATION

DEFECTS IN SILICON PARTICLE DETECTORS

L.F. Makarenko*, F.P. Korshunov**, S.B. Lastovski**, N.M. Kazuchits*, M.S. Rusetsky*,

E. Fretwurst***, G. Lindström***, M. Moll****, I. Pintilie*****, N.I. Zamiatin******

* Belarusian State University, Minsk, Belarus

**Institute of Solid State and Semiconductor Physics, Minsk, Belarus

***Hamburg University, Hamburg, Germany

****CERN, Geneva, Switzerland

*****National Institute of Materials Physics, Bucharest-Magurele, Romania

******Joint Institute for Nuclear Research, Dubna, Russia

1. Introduction

It is well-known that hydrogen can easily penetrate into silicon crystals at various stages of p-n structures manu-

facturing. Therefore possibly all silicon particle detectors contain hydrogen which remained in these structures after

fabrication. There are some indications on interaction of residual hydrogen with defects in irradiated Si detectors. It is

expected that the increase of hydrogen content will lead to the increase of radiation tolerance of devices. It may be

due to, first, hydrogen passivation of radiation defects, and, second, acceleration of oxygen diffusion by promoting

the formation of oxygen dimer О2.

Both high mobility and reaction ability of hydrogen result in its redistribution during subsequent technological

processes occurring even at rather low temperatures 200-400 °С. Therefore one expects that concentration of hydro-

gen in various areas of detectors will depend substantially on sequence of technological operations of their manufac-

turing and now there is not enough information on hydrogen in ready detector structures. It is the aim of this work to

get experimental data on hydrogen content and distribution in p+-n-n+ structures made of detector grade silicon and

influence of hydrogen on elimination of electrically-active radiation defects.

2. Experimental

Detector structures were manufactured from standard n-type Wacker Si by the following producers: 1) CiS Insti-

tute for Microsensors, Erfurt, Germany (CA-diodes); 2) ST Microelectronics, Catania, Italy (W336-diodes); 3)

ELMA, Zelenograd, Russia (Z-diodes). Simple p+-n-n+ structures with a guard rings were employed. The Z diodes

were prepared are by the same technology as for the CMS preshower sensors.

Table 1. Parameters of samples used for experiments

Device

acronym

Thickness,

µm

Starting resistivity,

KΩ⋅cm

Orientation Rear contact

CA 285 4 <111> grid

W336 290 2 <111> grid

Z 330 4 <111> block

Hydrogenation was performed using treatment in hydrogen plasma at T=300 °C during 0.5-4 hours. Glow dis-

charge plasma of hydrogen was created using a discharge chamber of the standard facility for the ionic etching . The

discharge dc voltage was 500 V. The density of the plasma current was 3-5 µA/cm2 that corresponded to density of

ion current about 2×1012 cm-2⋅s-1. Samples were arranged on a surface of a heater located in the discharge chamber.

The control of temperature was carried out with the help of the thermocouple which was in contact to a control sam-

ple of the same thickness.

1 – H+ plasma

2 – anode

3 – sample

4 – thermocouple

5 – heat shield

6 – heater in a bulb

Irradiation with electrons (E=3.5 or 6 MeV) was done using accelerators at the Institute of Solid State and Semi-

conductor Physics (Minsk). The irradiated samples were subjected to 30 min isochronal annealing in the temperature

range 50-350 °C with temperature increments of 50 °C.

Defect reactions have been studied using capacitance deep level transient spectroscopy (DLTS) (Tmeas=77-325 K)

and high frequency capacitance-voltage (C-V) measurements at room temperature. The experimental equipment con-

sisted of a capacitance meter, a HF-generator (1 MHz), a DC source and a pulse generator, and personal computers,

which controlled the pulse generator and performed the data acquisition. Possible variations of the reverse bias and

the rate window are up to 19 V and the rate window 4-19000 s-1, respectively. Normally both the reverse bias and the

pulse voltage were 5 V, and the rate window setting was 190 s-1.

Capacitance-voltage measurements are carried out with a digital capacitance bridge. The capacitance has been

measured at different reverse bias values by superimposing an ac voltage on the dc voltage. The reverse bias is typi-

cally varied in the range VR = 0-150 V in steps of 0.045-0:135 V. The amplitude and the frequency of the ac voltage

were 25 mV and 1 MHz, respectively. The measurements are controlled by home-developed Pascal program with

custom made instrument interface.

Table 1. Parameters of dominant peaks

Peak

number

Tmax, K

(at en=190 s-1)

A, s-1K-2 σA, cm2 EA, eV Peak origin

E1 94 3.9 107 6 10-15 0.174 VO+CiCs

E2 133 1.9 107 3.1 10-15 0.246 VV=

E3 ∼192 1.3 107 2 10-15(?) 0.358(?) Stable up 100-150 °C

E4a 230 7.5 106 1.1 10-15 0.423 VV−

E5 174 9.6 106 1.5 10-15 0.314 VOH

E6 111 1.9 105 2.9 10-17 0.163

E7 146 8.2 105 1.2 10-16(?) 0.238(?) ?

E8 198 5.9 106 0.9 10-15 0.356 ?

E9 186 9.1 104 1.4 10-17(?) 0.269(?)

E10 276 3.8 105 6.1 10-17(?) 0.51(?)

Fig.1a. Development of DLTSspectra for standard FZ diodes(CA-sample) after electron irra-diation at room temperature andupon 30-min isochronal an-nealing with temperature in-crements of 50 °C. Dose of ir-radiation was 2×1012 cm-2

(Ee=3.5 MeV). Measurementsettings wereen = 190 s-1,bias –5 → 0 V, andpulse duration =10 ms.

80 100 120 140 160 180 200 220 240 260 2800

50

100

150

200

250

300

350

400

450

500

550

600

as-irradiated 250 oC anneal 300 oC anneal

S,

fF

T, K

x1/3 aE1

E2

E3

E4

Fig.1b. Development of DLTS spec-tra for standard FZ diodes (W-sample) after electron irradiation atroom temperature and upon 30-minisochronal annealing with tempera-ture increments of 50 °C. Dose of ir-radiation was 2×1012 cm-2 (Ee=6MeV). Measurement settings wereen = 190 s-1,bias –5 → 0 V, andpulse duration =10 ms

80 100 120 140 160 180 200 220 240 260 2800

40

80

120

160

200

240

280

320

360

400

440

b as-irradiated 250 oC anneal 300 oC anneal

T, K

S, fF

E1

E2

E3

E4

E5

Fig.1c. Development of DLTSspectra for standard FZ diodes(ZA-sample) after electron irra-diation at room temperature andupon 30-min isochronal an-nealing with temperature in-crements of 50 °C. Dose of ir-radiation was 1×1012 cm-2 (Ee=6MeV). Measurement settingswereen = 190 s-1,bias –5 → 0 V, andpulse duration =10 ms

80 100 120 140 160 180 200 220 240 2600

1

2

3

4

5

6

7

8

9

10

11

12

13

14 E1

as-irradiated 250 oC anneal 300 oC anneal

∆C

/C0,

0.1

%

T, K

c

E2

E3

E4E5

Fig. 2. DLTS spectra for different

diodes made of standard FZ sili-

con after isochronal annealing at

temperature 350 °C during 30

min. Measurement settings were

en = 190 s-1,

bias –5 → 0 V, and

pulse duration (tp) - 10 ms

80 100 120 140 160 180 200 220 240 260 280 3000

40

80

120

160

200

240

280

320

360

E6 CA0419an350 W336#-8an350 ZamA4an350

S,

fF

T, K

E1

E5

E8E9

E10

Fig.3. DLTS spectra for stan-

dard FZ diode (CA-sample) af-

ter isochronal annealing at tem-

perature 350 °C. Measurement

settings were en = 19 s-1, bias –5

→ 0 V, and pulse duration (tp):

0.1 ms, 1 ms, 10 ms, 100 ms.

80 100 120 140 160 180 200 220 240 260 280 3000

40

80

120

160

200

240

280

320

360

400

440

480

tp=1e2 tp=1e3 tp=1e4 tp=1e5

S,

fF

T, K

E1

E6

E5

E8E7 E10

hidden peak

Fig.4. Annealing behavior of

divacancy (trap E2) in different

diodes irradiated with electrons

or gamma-rays (W336 diodes).

50 100 150 200 250 300 350 4000.0

0.2

0.4

0.6

0.8

1.0

ZA4 e ZB2 e W336#10 γ W336#8 e Ca0434 e

[V

V]/[V

V]as

-irr

Tann, oC

Fig.5. DLTS spectra for stan-

dard FZ diode (CA-sample)

after irradiation with 3.5 MeV

electrons at room temperature

and upon 30-min annealing at

350 °C.. Dose of irradiation

was 3×1012 cm-2. Measure-

ment settings were: rate win-

dow of 190 s-1, pulse duration

of 10 ms, bias: a) -5 → 0 V, b)

−10→ −5 V and c) −19→ −15

V. The peak amplitudes are

normalized assuming constant

concentration of E8 trap.80 100 120 140 160 180 200 220 240

0

40

80

120

160

200

240

280

320

360

V=-5 V (filling pulse 5v) V=-10 V (filling pulse 5v) V=-19 V (filling pulse 4v)

S,

fF

T, K

E1 E6

E5

E8

Fig 6. Carrier depth profile for a

CA sample determined from C-

V characteristics.

Vertical lines mark inverse

voltage which is necessary to

obtain corresponded width of

depletion region.

0 40 80 120 160 200 240 280

8.0x1011

1.2x1012

1.6x1012

2.0x1012

2.4x1012

10 V

Nd(

x), c

m-3

depth, µm

1 - 100 oC anneal2 - 300 oC anneal3 - 350 oC anneal

5 V 15 V20 V0 V

1

2

3

Fig. 7. Development of DLTS

spectra for hydrogenated standard

FZ diode(CA-sample) after irradia-

tion with 3.5 MeV electrons at room

temperature and upon 30-min iso-

chronal annealing with temperature

increments of 50 °C. Dose of irra-

diation was 3×1012 cm-2. Measure-

ment settings were en = 190 s-1, bias

–5 → -0 V, and pulse duration 10

ms.

120 140 160 180 200 220 240 2600

40

80

120

160

200

240

280

320

360

400

asIRR ann100 ann150 ann250 ann300S,

fF

T, K

Fig.8. DLTS spectra for hydro-

genated standard FZ diode(CA-

sample) after irradiation with 3.5

MeV electrons at room tempera-

ture and upon 30-min annealing

at 300 °C. Dose of irradiation

was 3×1012 cm-2. Measurement

settings were en = 190 s-1, pulse

duration 10 ms bias: a) -5 → 0 V,

b) −10→ −5 V, b) −15→ −10 V.

All spectra normalized the E5

peak (VOH) amplitude.

80 100 120 140 160 180 200 220 240 260 280 3000

50

100

150

200

250

300

350

-5 V ->0 V -10 V ->5 V -15 V ->10 V

S,

fF

T,K

Fig.9. Carrier depth profile for a

CA sample determined from C-V

characteristics. Vertical lines mark

inverse voltage which is necessary

to obtain indicated width of deple-

tion region.

0 20 40 60 80 100 120 140 160 180 200 220 240 260 2801.0x1012

1.1x1012

1.2x1012

1.3x1012

1.4x1012

1.5x1012

1.6x1012

1.7x1012

1.8x1012

1.9x1012

2.0x1012

CA0419asPREP CA0419ann250 Ca0419an300

Nd(

x), c

m-3

depth, µm

5 V10 V

15 V

20 V0 V

Fig.10. Diffusion length for dif-

ferent hydrogen forms

(monoatomic H and H dimer) in

silicon calculated for 30 min

time period. Points shows our

estimations from the depth of

donor formation in nonhydro-

genated (Fig.6) and hydrogen-

ated (Fig.9) samples.

200 240 280 320 360 4001

10

100

from Fig.6

Newman TD H1 (vanWieringen&Warmoltz) H2 (Markevich)

L di

if, µm

temperature, oC

H1

H2

from Fig.9

Conclusions

Hydrogen may exist in different regions of silicon particle detector.When it is present in the base region of a diode then it interacts with radiation defects alreadyat temperatures of 100-150 °C.

When hydrogen is present at Si-SiO2 boundary or in strongly doped regions then its activationtemperature is in the range of 250-350 °C. At these temperatures hydrogen can penetrate inthe base region at the depth of 50-100 µm during 30 min and its penetration depth depends onthe form of diffusing hydrogen particles (H or H2)

It is a procedure of diode processing that influence on hydrogen distribution in silicon detec-tors of particles.