Gamma radiation nose system based on In2O3/SiO thick-film sensors

Abstract

A prototype gamma radiation monitoring system based on In₂O₃/SiO thick-film sensors array was designed. Four sensors had an identical pn-heterojunction structure with different material compositions. These sensors were subjected to gamma radiation emitted by ¹³⁷Cs source with an activity of 370 kBq. Changes in their current-voltage characteristics were recorded and compared. The performance parameters of the devices, such as sensitivity to γ radiation exposure and working dose region, were found to be highly dependent on the composition of the materials used. To cover a wider range of radiation and improve the overall sensitivity, an approach of using sensor arrays was utilized. A dynamic selection of the multiple sensors of various sensitivities and working dose ranges was implemented by applying a pattern recognition analysis.

Full text

Page 1

380IEEE SENSORS JOURNAL, VOL. 6, NO. 2, APRIL 2006
Gamma Radiation Nose System Based
on In?O??SiO Thick-Film Sensors
Khalil I. Arshak, Member, IEEE, and Olga Korostynska
Abstract—A prototype gamma radiation monitoring system
based on In?O? SiO thick-film sensors array was designed. Four
sensorshadanidenticalpn-heterojunctionstructurewithdifferent
material compositions. These sensors were subjected to gamma
radiation emitted by???Cs source with an activity of 370 kBq.
Changes in their current-voltage characteristics were recorded
and compared. The performance parameters of the devices, such
as sensitivity to
radiation exposure and working dose region,
were found to be highly dependent on the composition of the ma-
terials used. To cover a wider range of radiation and improve the
overall sensitivity, an approach of using sensor arrays was utilized.
A dynamic selection of the multiple sensors of various sensitivities
and working dose ranges was implemented by applying a pattern
recognition analysis.
Index Terms—Metal oxides, pattern recognition (PR), sensor ar-
rays, thick films,
radiation.
I. INTRODUCTION
R
dustry,andmedicine(healthtreatmentandproductsterilization)
[1]. The choice of proper detector depends on application re-
quirements, expected working radiation dose range, sensitivity,
size, stability, and cost. Considerable research into new sen-
sors is underway, including efforts to enhance the sensors’ per-
formance through both material properties and manufacturing
technologies.
Metal oxides and their mixtures in different proportions are
considered as suitable materials for radiation sensing layers.
Mixing oxides was found to control the properties of semicon-
ductor films [2], [3]. The interaction of
mainly occurs by means of electronic excitation, electronic ion-
ization, and, primarily, atomic displacement of the orbital elec-
trons [4]. It is believed that ionizing radiation causes structural
defects (called color centers or oxygen vacancies in oxides)
leading to their density change on the exposure to
The influence of radiation on the material depends on dose rate
and the parameters of the films, including their thickness and
composition:Thedegradationismoresevereforthehigherdose
andthethinnerfilms[6]–[8].Radiationproducesachangeinthe
density of charge carriers in semiconducting material, which al-
tersitsproperties.Thismeasurablechangeprovidesinformation
on the dose absorbed by the sensing layer.
EAL-TIME radiation sensorsare essential in a wide range
of applications, including nuclear power production, in-
rays with materials
rays [5].
Manuscript received April 20, 2005; revised June 4, 2005. The associate ed-
itor coordinating the review of this paper and approving it for publication was
Prof. Eugenii Katz.
The authors are with the Electronic and Computer Engineering Depart-
ment, University of Limerick, Limerick, Ireland (e-mail: khalil.arshak@ul.ie;
olga.korostynska@ul.ie).
Digital Object Identifier 10.1109/JSEN.2005.860315
Fig. 1.
silicon wafer, 2 is the In O ?SiO layer, and 3 is the Ag electrodes.
Layout of manufactured pn-heterojunctions, where 1 is the p-type
In this paper, a number of pn-heterojunctions based on thick
filmswithfourdifferentmixturesofIn O andSiOcomponents
areincorporatedintoasingleradiationmonitoringsystem.These
sensors were exposed to gamma radiation in an identical way,
so one can compare the performance of these devices and trace
theinfluenceofcompositionontheresponsecharacteristics.The
valuesofradiationdamageinthemanufactureddeviceswerees-
timated from changes in their current-voltage characteristics. It
wasfoundthatthecompositionofthesensinglayersplaysavital
roleindeterminingthesensitivityandworkingdoserangeofthe
dosimeter. In most cases, the sensitivity and maximum level of
radiation that the device would sustain are in compromise. To
cover a wider range of radiation and to enhance the overall sen-
sitivity of radiation monitoring system, the approach of using
sensor arrays is utilized, where sensors differ in their material
composition.Adynamicselectionofmultiplesensorsofvarious
sensitivity and working dose range can be implemented by ap-
plying a pattern recognition (PR) analysis. The data-processing
unit of the presented radiation monitoring system receives in-
formation from each sensor and based on their reading, a sensor
with highest sensitivity in given dose range is chosen, providing
the most accurate estimation of the radiation dose.
II. EXPERIMENTAL PROCEDURE
Polymer pastes of In O and SiO mixtures in various propor-
tionsweremadeof92wt.%offunctionalmaterialand8wt.%of
polyvinyl-butyral (PVB), while ethylenglycolmonobutylether
was used as a solvent,following standard preparation procedure
for polymer thick-film pastes [9], [10]. Four various composi-
tions were used. These are only In O ; 75 wt.% of In O and
25wt.%ofSiO; 50wt.%of In O and 50wt.%ofSiO; 25wt.%
of In O and 75 wt.% of SiO [9]. These polymer pastes were
screen printed using DEK RS 1202 automatic screen printer
on p-type silicon wafers to form pn-heterojunctions. A silicon
wafer
had a dopant level of
and
cmon unpolished side. Afterward, printed films
were cured in Telco Model 6 laboratory oven at a temperature
of 100 C for 2 h. The active area of the diodes was
cmon polished side
mm ,
1530-437X/$20.00 © 2006 IEEE

Page 2

ARSHAK AND KOROSTYNSKA: GAMMA RADIATION NOSE SYSTEM381
Fig. 2.Dependence of normalized current ?I ? I ??I with ? dose under the applied voltage of ?2 V for In O ?Si diode.
Fig. 3. Dependence of current ?I ? I ??I with ? dose under the applied voltages of ?2 V for diode made with 75 wt.% of In O and 25 wt.% of SiO.
whereas all radiation-sensitive layers were 30 m in thickness.
Commercial DuPont 4929 silver paste was used to manufacture
the electrical contacts. The layout of manufactured pn-hetero-
junctions is shown in Fig. 1.
The
Cs (0.662 MeV) disk-type source was used to expose
the samples to
radiation (provided by AEA Technology QSA
GmbH as a standard reference gamma radiation source). The
radioactive gamma-emitting element (3.18
capsulatedinto2-mm-thickhigh-strengthepoxyresin(diameter
25mm)toshieldanyaccompanying
distanced 1 cm from the device under investigation at an angle
of incidenceof0 . Thedoserateofthesourcewas 5.7 Sv/min.
Asetofirradiationswereperformedchangingtheexposuretime
and, hence, the dose. The changes in –
samples were monitored after each dose to estimate its effect.
5 mm) was en-
radiation.Thesourcewas
characteristics for the
III. RESULTS AND DISCUSSION
A. In OSiO Thick-Film Structures as Radiation Sensors
Morphology and structural properties of four In O
were studied [9]. A fine-grained uniform surface of In O films
with flake-shaped particles measuring 0.2–0.5 m was revealed
with SEM analysis. The surface of the films made with 25 wt.%
of In O and 75 wt.% of SiO possessed large-scale uniformity.
The openings between the larger particles of SiO were filled
SiO
with much smaller particles of In O , leaving no big pores
[9]. Both the X-ray diffraction and Raman spectroscopy data
indicated the presence of indium–oxide crystal phase in all
compositions, whereas none of the silicon–oxide-based crys-
talline phases, such as quartz, stishovite, cristobolite, or quartz,
were identifiable. Furthermore, the intensity of the In O peaks
decreases and the intensity of the amorphous halo increases as
the amount of silicon oxide is increased. This indicates that SiO
must be present in the amorphous state [9].
Figs. 2–5 show dependences of the normalized current
(I–I )/I with
dose for various In O
the value of current of nonirradiated sample, and I is the value
of current of irradiated sample at the same applied voltage.
As one can see, all samples showed the most increase in the
values of current up to a dose of 114
samples (170 Sv). Beyond these doses, the values of normal-
ized current were found to be highly dependent on the material
composition. The threshold levels were:
•170 Sv for films made with 100 wt.% of In O ;
•578
Sv for films made with 75 wt.% of In O
25 wt.% of SiO;
•700
Sv for films made with 50 wt.% of In O
50 wt.% of SiO;
•2100 Sv for the films made with 25 wt.% of In O and
75 wt.% of SiO.
SiO diodes, where I is
Sv, except pure In O
and
and

Page 3

382IEEE SENSORS JOURNAL, VOL. 6, NO. 2, APRIL 2006
Fig. 4.Dependence of current ?I ? I ??I with ? dose under the applied voltages of ?4 V for diode made with 50 wt.% of In O and 50 wt.% of SiO.
Fig. 5.Dependence of current ?I ? I ??I with ? dose under the applied voltages of ?4 V for diode made with 25 wt.% of In O and 75 wt.% of SiO.
SamplesmadewithonlypureIn O arerecommendedforthe
detection of low levels of radiation. Counterpart samples made
with 25 wt.% of In O and 75 wt.% of SiO are recommended
for high-dose applications, as they sustained a higher dose of up
to 2100 Sv [11].
It is believed that in all tested pn-heterojunctions, the in-
crease in values of currents with radiation was caused by the
grain boundaries, which serve as current paths resulting in poor
leakage current characteristics. Another reason for worsening
the electrical properties is that the films were damaged by
the creation of radiation defects in the form of broken Si–O
and In–O bonds. It is reasonable to assume that the increase
in the leakage current after the influence of
partially attributed to the lowering of the barriers height at
Ag In O
SiO interface. Similar effects of
observed in the Al Ta O interface because of building up a
charge in Ta O near this contact [6]. During the irradiation
process, a modification of the Si In O
place as a result of the oxidation of the silicon wafer, leading
to an enlargement in the mixed-transition region, where SiO
and the intermediate oxidation states of Si coexist. The latent
defects, which are activated during irradiation, are in the form
of oxide traps and are also responsible for deterioration of the
device characteristics [6], [12].
radiation is
rays were
SiO interface takes
B. Radiation Nose System Design
To cover more than one energy or type of radiation, the
approach of using devices with a combined structure, such as
sensorarrays,canbeutilized,wherethesectionsoftheradiation
nose system could differ in material thickness or composition
[13]. The most important aspect of utilizing multiple radiation
sensors is choosing the most accurate one for a given radiation
dose.Additionally,thedetectionofdamagedsensorsisacritical
task necessary for ensuring the maximum possible accuracy in
measuring radiation doses. The characteristics of a radiation
sensor exposed to a radiation dose higher than its working dose
range will permanently change, making the sensor unreliable.
Therefore, damaged sensors should be excluded from further
usage. For example, a sensor based on pure In O is damaged
if exposed to a radiation dose level higher than 170 Sv.
Fig. 6 shows the dependences of normalized values of
currents versus radiation dose for four sensors with different
In O
SiO constituent, which are built in as sensor arrays
for radiation monitoring. The process of the detection always
starts from the analysis of the radiation dose readings from
sensor 4, which is capable of measuring highest radiation doses
without being damaged. If sensor 4 reports the detection of
radiation dose of more than 700
Sv, then sensors 1, 2, and 3

Page 4

ARSHAK AND KOROSTYNSKA: GAMMA RADIATION NOSE SYSTEM383
Fig. 6. Dependences of normalized values of currents versus radiation dose ffor four sensors with different In O ?SiO constituent.
TABLE I
COMPARISON OF RADIATION DOSE MEASUREMENT ACCURACY/SENSITIVITY
should be classified as damaged. When sensors 3 and 4 report
radiation dose higher than 578
damaged. Analysis starts from sensor 4 and if it does not detect
damage in sensor 3, then sensor 3 is used (as more accurate) to
detect whether sensors 1 or 2 are damaged. If sensor 3 reports
radiation dose higher than 578
considered as damaged. Otherwise, if sensor 3 reports radiation
dose higher than 170 Sv, then only sensor 1 is damaged.
Once it is known which sensors are operational at a given
dose, a selection of the most accurate/sensitive among them
should take place, as summarized in Table I.
The above-proposed radiation monitoring system has two
main advantages, e.g., wider working dose range and increased
overall sensitivity. This approach may be compared with
gas-sensing electronic nose system, which finds its application
not only in food industry, but also in healthcare, military, and
space exploration [14]. The analysis of gas mixtures requires
the use of multiple gas sensors, each one sensitive to a different
gases or gas concentration. Only then can the information be
properly evaluated to recognize a specific pattern. The main
goal of the PR algorithm is to classify set of input values into
one of available classes. In the case of the electronic nose,
input values consist of readings from each sensor in the array,
while theclass recognizedby thePRalgorithm reflects a typeof
odour [14].PR coversa veryrichfamily of different algorithms.
Performance of a particular algorithm can be measured in terms
of flexibility, efficiency, and reliability. The accuracy of mea-
surement classification by PR algorithms is the most common
problem in real applications of sensor arrays systems [14].
Sv, then sensors 1 and 2 are
Sv, then sensors 1 and 2 are
Theoretically, the proposed radiation monitoring system
design approach can be applied for more complex systems, for
mainly industrial applications, where few types of radiation
sources are used. These sources can have different parameters,
such as an activity and dose rate. Moreover, resultant radiation
exposure could be of a complex nature, e.g., it can include
,rays, etc. In this case, radiation-sensitive materials should
be carefully chosen. Similarly, an array of sensors can be
implemented, with each sensor having different sensitivity to
various radiation types or energies. Based on a set of the output
readings from these sensors, a specific PR algorithm should
be applied so that the radiation nose system will accurately
determine radiation dose and type.
,
IV. CONCLUSION
A novel approach to monitoring a radiation dose was pre-
sented.AnumberofIn O
SiO radiationsensorswithdifferent
compositions were fabricated using thick-film technology. The
sensitivity of these devices and their working dose range were
found to be highly dependent on the combination of the ma-
terials used. It was proposed to combine these sensors as an
array into the radiation monitoring system to ensure the highest
overall sensitivity and widest working dose range.
REFERENCES
[1] K. Arshak, O. Korostynska, and F. Fahim, “Various structures based on
Nickel oxide thick films as gamma radiation sensors,” Sensors, vol. 3,
pp. 176–186, 2003.
[2] K. I. Arshak, C. A. Hogarth, and M. Ilyas, “A study of electron spin
resonance and optical absorption edge in amorphous mixed films of SiO
and In O ,” J. Mater. Sci. Lett., vol. 3, pp. 1035–1038, 1984.
[3] K. Tominaga, T. Murayama, I. Mori, T. Okamoto, K. Hiruta, and T.
Moriga et al., “Conductive transparent films deposited by simultaneous
sputtering of zinc-oxide and indium-oxide targets,” Vacuum, vol. 59, pp.
546–552, Nov. 2000.
[4] R. El Mallawany, A. A. El Rahamani, A. Abousehly, and E. Yousef,
“Radiationg effect on the ultrasonic attenuation and internal friction
of tellurite glasses,” Mater. Chem. Phys., vol. 52, pp. 161–165, Feb.
1998.

Page 5

384IEEE SENSORS JOURNAL, VOL. 6, NO. 2, APRIL 2006
[5] R. Y. Zhu, “Radiation damage in scintillating crystals,” Nucl. Instrum.
Meth. A, vol. 413, pp. 297–311, Aug. 1998.
[6] E. Atanassova, A. Paskaleva, R. Konakova, D. Spassov, and V. F. Mitin,
“Influence of ?-radiation on thin Ta O -Si structures,” Microelectron.
J., vol. 32, pp. 553–562, July 2001.
[7] K. Arshak, O. Korostynska, and J. Harris, “?-radiation dosimetry using
screen printed nickel oxide thick films,” in Proc. MIEL, 2002, pp.
357–360.
[8] K. Arshak and O. Korostynska, “Influence of gamma radiation on the
electrical properties of MnO and MnO/TeO thin films,” Ann. Phys.,
vol. 13, pp. 87–89, 2004.
[9] K.Arshak,O.Korostynska,andJ.Henry,“Thickfilmpn-junctionsbased
on mixed oxides of indium and silicon as gamma radiation sensors,”
Microelectron. Int., vol. 21, pp. 19–27, 2004.
[10] M. Prudenziati, Handbook of Sensors and Actuators. Thick Film Sen-
sors.Amsterdam, The Netherlands: Elsevier, 1994.
[11] K. Arshak, O. Korostynska, L. Szumilas, J. Harris, and S. Clifford,
“Gamma radiation nose system based on In O ?SiO thick film
PN-junctions,” in Proc. IEEE Sensors, Vienna, Austria, 2004, pp.
1179–1182.
[12] K. Arshak and O. Korostynska, “Radiation-induced changes in thin film
structures,” Proc. Inst. Elect. Eng., vol. 150, pp. 361–366, Aug. 2003.
[13] K. R. Herminghuysen and T. E. Blue, “Development and evaluation of a
neutron-gammamixed-fielddosimetrysystembasedonasinglethermo-
luminescencedosimeter,”Nucl.Instrum.Meth.A,vol.353,pp.420–424,
Dec. 1994.
[14] K. Arshak, E. Moore, G. M. Lyons, J. Harris, and S. Clifford, “A review
ofgas sensorsemployedinelectronicnoseapplications,”Sens.Rev.,vol.
24, pp. 181–198, 2004.
Khalil I. Arshak (M’95) received the B.Sc. degree
from Basrah University, Iraq, in 1969, the M.Sc.
degree from Salford University, U.K., in 1979, and
the Ph.D. and D.Sc. degrees from Brunel University,
U.K., in 1986 and 1998, respectively.
He joined the University of Limerick, Limerick,
Ireland, in 1986, where he leads the Microelectronic
and Semiconductor Research Group. He is also a
Director of Advanced Manufacturing Technology
(AMT Ireland, Ltd.). He has authored more than
200 research papers in the area of microelectronics
and thin- and thick-film technology. His current research interests include
lithography process modeling, TSI processes characterization, mixed-oxide
thin- and thick-film sensor development, and application-specific integrated
circuit design.
Olga Korostynska received the B.Sc. and M.Sc.
degrees in biomedical electronics from the National
Technical University of Ukraine (KPI), Kyiv, in
1998 and 2000, respectively, and the Ph.D. degree
from the University of Limerick, Limerick, Ireland,
in 2003.
She is currently a Postdoctoral Research Fellow
with the Department of Electronic and Computer En-
gineering, University of Limerick. Her research in-
terests are in thin- and thick-film technology, mate-
rial properties characterization, personal gamma ra-
diation dosimeters, and pressure sensors for medical device applications.