RESULTS 1. Irradiations

Batches of BULT118 transistors have been irradiated up to doses of 25 kGy in different laboratories and with different types of radiation sources. Measurements of the parameter T were done in Bologna. The changes of T with irradiation dose have been plotted as A(1/T) = 1/T - 1/T0 and are shown in Fig. 2. The response of the transistor is linear with absorbed dose up to 5 kGy, giving a linearity correlation coefficient r2 = 0.998 for 12 MeV electrons, r2 = 0.993 for 4 MeV electrons, r2 = 0.996 for 2.5 MeV electrons and r2 = 0.994 for gamma irradiations. At higher doses the response is sublinear and follows a second order polynomial fitting. To check the stability of the lifetime changes produced by irradiation, the irradiated devices, together with some blanks, were left on the shelf of a cabinet without any special precaution. They have been measured over a year period showing no significant variation of the parameter T. This is in agreement with the fact that transistors of the same type used for this study and electron irradiated, subjected to reliability tests ("high temperature reverse bias" HTRB) at 150°C for 1000 h, did not show any significant variations of the lifetime [11]. The results of these tests are reported in Fig. 3.

5.00E-02

4.00E-02

3.00E-02

2.00E-02

l.OOE-02

O.OOE+00

7.00E-03

6.00E-03

l.OOE-03

O.OOE+00

MeV e

2.5 MeV e

12 MeV 6'

j;amma

5000 10000 15000 20000 25000 30000

Gy in water

500 1000 1500

Gy in water

2000 2500

FIG. 2. Plot ofl/T -I/To vs. dose for yand electron irradiations performed in Bologna, Seibersdorf, Strasbourg, Budapest and Washington.

oo o

<L>6 (fi

=L UJ2

ID

1.2

1.1

1.0

0.9

0.9 1.0 1.1

LIFETIME (n sec.) at 0 h

1.2

F/G. 3. Comparison of electron irradiated transistors after 1000 h annealing at 150°C and soon after irradiation.

3.2. DLTS measurements

Deep level transient spectroscopy (DLTS) measurements have been carried out on the irradiated devices in order to characterize the recombination centres, introduced by irradiation, that affect the carrier lifetime. The results of these measurements are reported in Fig. 4 and in Fig. 5. In Fig. 4 three peaks, labelled as EI, E2 and E3, are clearly distinguishable in the samples irradiated with electrons.

They correspond to the three main electron traps identified as the oxygen-vacancy complex (A centre), the double negative (V-V)⁼ and single negative (V-V)~ charge state of the di vacancy, respectively. These three levels are all active recombination centres that can account for the changes of the lifetime, and consequently of the related electrical parameters, observed with irradiation [7]. Their thermal stability up to 150°C (the maximum working temperature of the devices), very important for safe operation of the device, is well known from previous study [12] and the results, reported in Fig. 3, are a direct consequence of that. The samples irradiated with y rays, instead, show a clear oxygen-vacancy peak and a broad DLTS signal in the temperature range 120-300 K. This broad band is determined by the presence of a complex structure of defects in high concentration which contributes to its amplitude, thus resulting comparable with that of the related oxygen-vacancy peak, differently from what happens in the electron irradiated samples where the amplitude of the oxygen-vacancy peak is always much larger than those of the divacancy. This fact can explain the larger reduction of the lifetime obtained in the transistors with gamma rays in comparison with 2.5 MeV electron irradiation, even though the production rate of the oxygen-vacancy is lower in gamma- than in 2.5 MeV electron-irradiated samples.

4. CONCLUSIONS

The work on the characterization of these transistors as possible routine dosimeters is not yet completed. In fact the possible influence of different environmental conditions has not yet been evaluated. A large spread of lifetime values after irradiation has been observed in the low dose range (up to 150 Gy) for y rays, 2.5 and 4 MeV electrons, nevertheless the results obtained so far, are satisfactory and promising. The bipolar transistor's advantages are its small size, low cost, ease of use, good sensitivity, possibility of immediate reuse and its ability to record dose history (dose information is

03 .1S3 (X!

H

e' 2.5 MeV

E3

El

50 100 150 200 250 300 350

Temperature (K)

FIG. 4. DLTS spectra from transistors irradiated with yrays, 2.5, 4 and 12 MeV electrons at a dose of 1.5 kGy. Rate window = 575 s'^!.

25

20

O

=0 15

10

OP:

/ A center e" 12 MeV

9

</°

/ ' A center e" 5 MeV / £

A center e" 2. 5 MeV

__, - " "O A center 7 / /

'°' - 1(C Divacancy e 12 MeV '' '

// V-'V-X Divacancy e'5 MeV

^ -JC.9KT-• • •' X ...-.-+ Divacancy e 2.5 MeV 0 2000 4000 6000 8000 10000 12000 14000

Dose (Gy) in water

FIG. 5. Plot of concentration of A centre and divacancy vs. dose from irradiated transistors.

not lost during retrieval process). Moreover dose can be ascertained within minutes after irradiation with an inexpensive support equipment. All that brings us to conclude that such device may be a suitable dosimeter for the day-to-day monitoring of radiation process and its possible use should be taken into consideration. Particularly for 12 MeV electron irradiation the behaviour of the transistors is good; the spread of A(1/T) values is < 10% for doses up to 100 Gy while it goes down to 3-4 % for the dose range 0.15-20 kGy.

The silicon devices are made from a mixture of plastic, copper and silicon, and it is difficult to evaluate their radiation absorption characteristics. The small size of the devices relative to the range of the secondary electrons makes the mass collision stopping power the most important parameter, and the ratio of mass collision stopping powers is not changing much with energy, but below 1-2 MeV the ratios differ significantly. If the devices are calibrated and used in radiation fields with significant differences in radiation energy spectra, differences in response may be expected. Anyway the results obtained by placing four sets of three transistors inside the product boxes together with ampoules containing ECB dosimeter, and irradiated with y rays at doses from 5 kGy up to 26 kGy (Fig. 2), during a production run, do not show any significant difference in their response with respect to the transistors irradiated in gamma-cells.

ACKNOWLEDGEMENT

The authors are indebted to Dr. K. Mehta and Mr. R. Girzikowsky, Dosimetry and Medical Radiation Physics Section, IAEA, Vienna, for the y-irradiations done in Seibersdorf, to Dr. W.

McLaughlin for the y-irradiations done at NIST, to Dr. F. Kuntz and Dr. A. Strasser, AERIAL, for carrying out irradiations of transistors with 2.5 MeV accelerator in Strasbourg and to the LINAC staff (Ing. A. Martelli, A. Monti and G. Mancini) of the FRAE Institute for their valuable technical assistance.

REFERENCES

[I] RIKNER, G., GRUSELL, E., Phys. Med. Biol. 32 (1987) 1109-1107 and references therein.

[2] For more information see: a) SCHARF, K., Health Phys. 13 (1967) 575-586; b) MULLER, A.C., "The "n" on "p" solar-cell dose-rate meter", Manual on Radiation Dosimetry (HOLM N.W., BERRY R.J., Eds), Marcel Dekker, New York (1970) 423-427; c) OSVAY, M., et al.,

"Silicon detectors for measurement of high exposure rate gamma rays", Biomedical Dosimetry, STI/PUB/401, IAEA, Vienna (1975) 623-632.

[3] MULLER, A.C., "The "p" on "n" solar cell integrating dosimeter", Manual on Radiation Dosimetry (HOLM N.W., BERRY R.J., Eds), Marcel Dekker, New York (1970) 429-433.

[4] HARTSHORN, A., et al, "Absorbed dose mapping in self-shielded irradiators using direct reading MOSFET dosimeters", (Proc. of the Health Physics Society), Annual Meeting, Boston, July 1995.

[5] GRUNEWALD, T., RUDOLF, M., Food Irradiation Newsletter, FAO/IAEA 11 (1987) 42-47.

[6] EHLERMANN, D.A.E., "Dose distributions and methods for its determination in bulk particulate food materials", Health Impact, Identification and Dosimetry of Irradiated Foods, Report of WHO Working Group, (BOGL, K.W., REGULLA, D.F., SUESS, M.J. Eds), ISH-Heft 125, Institut fur Strahlenhygiene des Bundesgesundheitsamt, Neuherberg (1988) 415-419.

[7] FUOCHI, P.G., Radiat. Phys. Chem. 44 (1994) 431-440.

[8] BIELLE-DASPET, D., Solid State Electron. 16 (1973) 1103-1123.

[9] MILLER, G. L., et al., Ann. Rev. Mater. Sci. 7 (1977) 377-448.

[10] GHEZZI, C., etal, Material Science Forum Vol. 10-12 (1986) 1213-1218.

[II] ARCORIA, G., etal., Radiat. Phys. Chem. 42 (1993) 1015-1018.

[12] BARBERIS, L., etal., Radiat. Phys. Chem. 26 (1985) 165-172.

IAEA-SM-356/28