This article was downloaded by: [103.49.147.

5]
On: 27 May 2015, At: 03:56
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Nuclear Science and

Technology
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/tnst20

Silicon Surface Barrier Detector

a

Seiichiro GOTOH & Zinzaburo TAKAGI

NAIG Nuclear Research Loboratory , Nippon Atomic Industry Group

Co. , Kawasaki-shi, Kanagawa-ken
Published online: 15 Mar 2012.

To cite this article: Seiichiro GOTOH & Zinzaburo TAKAGI (1964) Silicon Surface Barrier Detector,
Journal of Nuclear Science and Technology, 1:8, 311-315
To link to this article: http://dx.doi.org/10.1080/18811248.1964.9732130

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the
Content) contained in the publications on our platform. However, Taylor & Francis, our
agents, and our licensors make no representations or warranties whatsoever as to the
accuracy, completeness, or suitability for any purpose of the Content. Any opinions and
views expressed in this publication are the opinions and views of the authors, and are
not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not
be relied upon and should be independently verified with primary sources of information.
Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands,
costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising
directly or indirectly in connection with, in relation to or arising out of the use of the
Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions

Journal of NUCLEAR SCIENCE and TECHNOLOGY, 1, No.8, p. 311-315 (November, 1964).

311

Silicon Surface Barrier Detector

Seiichiro GOTOH* and Zinzaburo TAKAGI*
Received April

2,

1964

A method has been established for producing silicon surface barrier detectors having an energy
resolution of 0.7% for 5.3 MeV a particles, and which are stable for long periods of time. The
fabricated detector has been proved to possess a depletion layer as thick as 200 fL and a surface layer
with energy loss less than 15 keV for 5.3 MeV a particles. The stability of the detector was also
tested and it was found to be stable for at least 2-3 months.

Downloaded by [103.49.147.5] at 03:56 27 May 2015

I.

INTRODUCTION

Semiconductor radiation detectors are now

widely used for the detection of heavy charg
ed particles such as protons, a particles and
fission fragments. Many studies have already
been reported in such fields as nuclear and
reactor physics. A large number of reports
have been published concerning the design
and manufacture of both p-n junctions and
surface barrier silicon detectors(l)-cs>. Nevertheless, semiconductor detectors with good
resolution and high resistivity still appear
difficult to produce. The authors have endeavored to establish a reliable method of
obtaining detectors with good energy resolution and stability for long periods of the time.
Detailed description of the operating principles and the advantages of the semiconductor
detectors are found in a previous paper'6'.
The surface barrier detector has some advantages over the p-n junction detector: As
no diffusion process requiring an electric
furnace is involved, the former is fabricated
at a lower temperature than the latter. This
fact is advantageous for the preservation of
a long minority carrier lifetime. Another
advantage is the fact that the energy loss in
the window layer can be much lower in the
former. The authors' chief aim was to obtain
a detector with thin window layer for the
detection of low energy particles.

n.

CoNSTRUCTION

METHODS

by our method as follows. Care must be

taken to keep the room always clean. Thick
n-type silicon wafers, 21 mm diameter and
360 fL thick, lapped with 1,200 mesh carborundum
was used. The resistivity p of the material
is 2 kO -em. The wafer is first rinsed carefully
with deionized water, acetone and trichloroethlene, and then dried. The clean wafer is
etched in an ice-cooled beaker containing CP4A '7' which is stirred during the process.
After about 10 min of etching, a mirror-like
surface is obtained on the wafer which is
reduced to a thickness of about 200 fL. Then
the solution is amply diluted with deionized
water. The exposure of the wafer to air
must be strictly avoided in the presence of
the etchant, as it produces a brownish stain
on the surface. The wafer is washed 5-6
times with deionized water, using a fresh
beaker at each washing.
It is then dried and nickel-plated by
electroless nickel plating method'8'.
The central part of the rear surface of
the wafer is protected with Apiezon Wax
and the unprotected nickel in the remaining
surface is dissolved in nitric acid. After
careful rinsing with deionized water, the
protection is washed off with trichloroethylene
and acetone and the wafer is dried. Then
it is heated in an electric furnace to about
600C in vacuum for 10 min.
This procedure is believed to tighten the
bonding between nickel and silicon, and it
has proved to reduce contact noise and keep
the detector stable for longer periods of the

*
The surface barrier detectors are produced
-33-

NAIG Nuclear Research Loboratory, Nippon Atomic

Industry Group Co., Kawasaki-shi, Kanagawa-ken.

Downloaded by [103.49.147.5] at 03:56 27 May 2015

312

J. Nucl. Sci. Tech.

time. After cooling the wafer, it is again

nickel plated in the same manner. A tightly
adherent contact of thick nickel layer is
formed on the disk-shaped original deposit
of nickel. The remaining nickel deposit is
stripped off as before, and then a copper wire
is soldered to the nickel part to form the nside contact lead. The contact part and the
copper wire is then protected with a coating
of Apiezon Wax, and the wafer is immersed
in an ice-cooled solution of 40% hydrofluoric
acid and concentrated nitric acid till all surface
contaminants are completely etched off. The
mirror-finish is thus completed, and it is
again rinsed carefully as described before
and the protection is washed off with trichloroethylene.
Then, on the front surface, pure gold is
vacuum-evaporated to a thickness of 40-100
g/cm 2 , using a cut-out mask.
At the same
time gold is evaporated onto a thin glassplate
under similar conditions to monitor the amount
of the gold evaporated. Care must be taken
to prevent the gold from being evaporated
onto the wafer edge. A platinum wire of
0.1 mm diameter is attached to the gold with
silver paste to form the p-side contact lead.
The bare detector is completed by keeping
it overnight in a dry atmosphere to obtain
stable characteristics. When edge protection
:s necessary the bare detector is placed in an
acryl resin case as shown in Fig.1 and fixed

to it with Araldite.
Forced oxidation of the surface was tried,
but it did not lead to marked improvement.
For developing a suitably thick surface oxidation layer, it is sufficient to leave the detector
overnight in a dry atmosphere.

m.
1.

VARIOUS CHARACTERISTICS

Diode Characteristics

For measurements with high energy resolution detector noise must be kept small
under reverse biased conditions. The reverse
leakage current which is the main source of
noise, must be kept as small as possible: For
ordinary purposes the reverse leakage current
should be less than 1 JLA under operating
conditions. If the surface leakage current is
small and the diffusion current is much smaller than the generation current, the reverse
leakage current is given by Eq.( 1 )<9>,
I Reve =38X . (py)t/
To

where p is the resistivity of the n-type silicon

wafer, vis the applied reverse bias voltage
and To is the minority carrier life-time.
Figure 2 shows the reverse biased diode
characteristics of detectors with various diameters and thicknesses. As the figure shows,
the reverse current is not always proportional
to V'' 2 as might be expected from the Eq.( 1 ).
This means that the surface leakage current
should also be taken into account. It is obvious
from the figure, however, that the detector
can be used successfully with a current
smaller than 1 JLA under a bias of at least
40-50V.
10"s

3 s

!il

3
2

!0"'

o Effec~ive ~ Spa diL, 20~# w~fer

Ellective area 10 ... dia. 300# wafer
a llettive area s,., dia. 180# wo!e7

.1

,...tOj.ov

1---'
,......

.......

v
~

_........

ll
v

:JI

.............
10

100

Applied rever.. bias (V)

Fig. I

1)

Fig.2 Reverse Biased Diode Characteristics

of the Device

Assembly and Dedails of

Device Coostruction

-M-

1,000

318

Vol. 1 , No. 8 (Nov. 1964)

2.

Linearity and Re11olution

The linear response of pulse height to

incident particle energy has been confirmed
by using 1~3MeV protons and 4~9MeV a
particles. This is shown in Fig. 3.
9

3.

Downloaded by [103.49.147.5] at 03:56 27 May 2015

The energy resolution of the detector was

examined for 5.3 MeV a particles of 210 Po and
2 MeV protons, the spectra of which are shown
in Fig. 4.
In both cases the full width at half
maximum is about 36 keV, which means a
resolution of 0.7% for 5.3 MeV a particles
including the effect of the instrumentation
noise. The protons used were accelerated
by an ion Van de Graaf installed in the NAIG
Nuclear Research Laboratory.

3
2

10

Energy (MeV)

Fig. 3 Pulse Height Response of

the Detector to Incident
Charged Particles

150

IIIII

50

Depletion Layer

When a charged particle penetrates into

silicon, it produces free electrons and holes.
If these electron-hole pairs are created in a
region called depletion layer where an electric
field exists, they will drift in the electric
field and will finally induce net charges on
the electrodes. The electron-hole pairs created in the depletion layer give electric signals
proportional to the energy given by the
incident charged particles in this region, while
the pairs created outside this region are
thought to give no signals.
Consequently, for a particle having a path
range R longer than the thickness D of the
depletion layer, the linearity is broken. If
we can determine experimentally the pulse
height-energy curve and know the energy
point where the linearity is broken, we can
then estimate the value of D, since the breaking point of the pulse height-energy curve
indicates the condition where the path range
of a particle is equal to depletion layer
thickness.
Experimental values of the depletion layer
thickness obtained in the way mentioned
above are shown graphically as a function of
detector bias voltage.
Theoretically, the depletion layer thickness
is given by
D ex: (pV) I/2

Cha.nnel number

(b)

ChaMel number

(a)

Measured spectrum of 5.3MeV

a-rays with the device
(b) Measured spectrum of 2.05MeV
protons with the device
Fig.4

(a)

and for given values of p we can easily obtain

the values of D from Blankenship's nomographs00. These values are also given in Fig. 5
by the solid lines for p =1, 3, 4 and 5 kO. -em.
In Fig. 5 it is obvious that the experimental
value of p is much larger than the resistivity
of the wafer which was used for the fabrica-35-

314

J. Nucl. Sci. Tech.

10-',...------------------,

tion of the detector.

- - just after preparation

- a f t e r 7days
- x - after 31 days
~after4ldays
---0--

after 85 days

10-6

~
~

Appli~d

reverse bias(V)

--&-~-Measured points using protons

--+--Measured points using

210

10-'o

Po

Downloaded by [103.49.147.5] at 03:56 27 May 2015

Fig.5 Thickness of the Depletion Laver

as a Function of Peverse Bias
4.

10

20

30
40
50
Applied reverse bias (V)

60

70

Fig.6 Change of the Diode

Characteristic with Time

Surface Layer

The surface barrier detector is considered

advantageous for low energy charged particle
measurements compared with the p-tz junction
detector because the energy loss by the surface layer is generally made less than the
latter. The energy loss of a detector with
100 J.Lg/cm' thick gold layer was measured to
be about 25 keY for 5.3 MeV a particles, which
corresponds approximately to the energy loss
through the gold layer only.
Thus, the
surface layer of the surface barrier detector
can be considered to consist only of the gold
layer.
This consideration is based on measurements of the pulse height difference between
two incident beams entering the front surface
of the detector at two different angles. The
surface energy loss can be made less than
10 keY for 5.3 MeV a particles by controlling
the amount of gold to be evaporated.

w.

fiO days after prepatation

900
800
700
600

]soo
~

j 400

'-'

300
200
100

180

190
Channel number

200

Fig.7 Change of the Resolution with

Time for 210 Po
The detector used had 36
keV FWHM when new.

AGING EFFECTS OF
AcKNOWLEDGEMENT

CHARACTERISTICS

For the actual use stability during long

periods of time is desirable. Figure 6 shows
the change of the diode characteristics during
a lapse of 2-3 months. It may also be seen
from Fig. 7 that neither the diode characterisitics nor the energy resolution changes
markedly during this period. These results
thus indicate that degradation of the surfacebarrier detector with time is negligible for
at least several months.

The authors are indebted to Mr. R. Wakabayashi, Manager of the Research Department, Mr. E. Inaba, Chief of the Reactor
Engineering Section, and Dr. H. Shimojima,
Chief of the Nuclear Chemistry Section for
their strong support. We also wish to thank
Dr. T. Kawai, Chief of the Nuclear Physics
Section and his co-workers for their kind
help in experiments with the Van de Graaff
and for many useful discussions.

-36-

315

Vol. 1 , No. 8 (Nov. 1964)

MAYER,J.W.: J. Appl. Phys., 30, 1937 (1959).

MILLER, G. L.: BNL Rept., No. 4662.
JACKSON, R. S.: I.R.E. Trans., NS-8, No.1, 29
(1961).
(4) LOVE,T.A.: ibid., 91 (1961).
(5) SLRODE, A. C.: ibid., 59 (1961).

Downloaded by [103.49.147.5] at 03:56 27 May 2015

(1)
(2)
(3)

MURPHY,J.F.: UCRL-6505, (1961).

BRIGERS, H.E.,SCAFF, J.H., SHIVE, J.N. :"Transistor
Technology", Vol. I, p. 354 (1958), D. Van
Nostrand.
(8) BioNDI. F. J.: ibid., Vol. m, p. 173 (1958).
(9) GOULDING, F. S., HANSEN, W. L.: Nucl. lnstr.
Methods, 12, 249 (1961).
(1Q BLANKENSHIP, J. L., BoRKOWSKI, C. J.: l.R.E.
Trans., NS-7, No.2, 3, 190 (1960).

(6)
(7)

-REFERENCES--

-37-