-oSNL-TR-67-8623 June 1967

00 RADIOPHOTOLUMINESCENCE SPECTRA OF SILVER-ACTIVATEDOSIMETER GLASSES

g by* - K Becker*

D.W. McOuilling

P -3~-

~ AUG 24 1967

U.S. NAVA.L RADIOLOGICAL

DEFENSE LABORATORY

S A N FRANCISCO *CALIFORNIA- 94135

ThIs docunent hc, bv'n 0'1:roodror public rolcaro and salo; itsdistribution is unlimito,' " " ~~~~~~~~~~~~~Reproduced b/ thee,,,, ---. :.:-_- -- _,. . . .. .ICLEAR ING HOUSE

for Federal Scientific & TechnicalInformation Springfield Va 2215'

r1

-I-

ABSTRACT

The radiophotoluminescence- (RPL) spectra of silver-activatedphosphate glasses of different compositions under 365 mg excitationhave been measured in the 400 to 750 mg range. The apparatus,specially designed for this type of measurement, uses a red-sensitivephotomultiplier tube and permits comparison of the relative RPL inten-sity with the known output of a NBS-calibrated lamp for each wavelengthinterval.

Shape and maximum of the RPL spectrum depend on thc type ofphotomultiplier tube, if no corrections fcr the photomultiplier spectralresponse are made. After corrections, a steep RPL increase above ca.500 mg, spectral radiant intensity maxima between 615 (French C.E.C.glass) and 640 mil (DT-60 glass) depending on the glass composition,and a slow increase in the red have been observed. For a constantglass base composition, the Ag concentration of from 0.5% to 5% did notaffect the spectrum. Also the dcse level and the degree of the RPLbuild-up after exposure had only minor influence on the spectrum.Dur±ng thermal annealing, a 1{PL increase in the blue has 'been observed. -

After a megarad exposure, a permanent RPL which could not be annealedhas been found. Some further experiments are suggested.

i:

I - _ _ -

S2MARY

The absolute radiophotoluminescence spectra of several corcial.and experimental Ag-activated dosimeter glasses'have-been measured withP; speciallSy designed apparatus. Dependi mainly on the glass baseccumosition, maxima. between 615 and 640,1' and a slow decrease in thered .have-beenifTund. The effect of build-up, silver concentration anddose of the RPL- spectral shape is small. The influence of very highdoses and thermal annealingalso has-been measured. "

I _

I. INTR ODUCTION

Radiophotoluminescence (ITL) spectra of silver-activated sodium

chloride and a metaphosphate glass containing different mouts ofAgPO3 (2-16%) were first measured by Schulman, et aa.. I in the 420 to64o ng range. Maximum response was obtained at 630 Wa. In a laterpublication, 2 a -ML spectrum was given for the Navy DT-60 dosimeterglass cont~aining 8% Aglm) (4.,1 Ag). A poorly resolved minor peak wasobserved around 480 q1 w2t a main peak at 640 mI. Recently, Yokota

and Nakajima3 published a RFL spectrum for their low Z4 low predoseglass which peaked at 610 mg (50% of the maximum RPL at 545 and 685 mg).This spectrum had been measured in a spectrofluorometer wi'th atomaticcorrection for, the detector response characteristics designed byUehara., et al.4 Other authors (see, for instance, 5 and 6) measured"rrelative" MP spectra (uncorrected for the luminescence detectorresponse) and found maxima around 560-590 M, depnding on the red

~sensitiv-ity of the photomultiplier tube. The value of this type of~~measurements is rather limited...

| iA knowledge of the ITL spectra is of both practical and theoreti-I!cal interest. It allows a combination of optical filter and ITL

detector to be chosen that will give optimal sensitivity together withI i high background (predose) discrimination. There is also a correlation

betw en luminescence and the coordination of the activating ion in the! glazs. For instance, it has been concluded from the fluorescence

spectra rhat the Mh2+ ion has octahedral symmetry in phosphate and_ borate glasses (maximum ca. 600 m i in the relative spectrum) but is,! tetrahedrally coordinated in silicate glasses (maximum ca. 520 Wa).7

II. WE SPECTROFLUOROWMER

The fluorometer used in the measurements (Figs. 1 and 2) vacdeveloped for checking the constancy (i.e., in color and intensity) of

1t.

II

the fluorescence standards which are part of the U.S. Navy's phosphateglass :casualty-dosimeter system. In order to serve as an "absolute "t

device in a useful sense, this fluorometer makes use of a standadlight source which is reproducible within known limits in its intensityand spectral energy distribution. Barbrow9 has shown that a tungsten-filament lamp calibrated for luminous intensity and color temperaturecan be used as an approximate standard of spectral radiant intensity

J(S.R.I.). Lamps so calibrated can be obtained from the National BureauI of Standards, and readily yield values of S.R.I. at a selected plane inmicrowatts per cm2 per lO-millimicrons spectral interval, over a wave-length range of special interest for dosimeter-glass fluorescencemeasurements--namely, from 40 to approximately 750 mi11 1m crons.

Briefly, the fluorometer is designed to compare (i.e., in radiantirtensity for each specified wavelength interval) the fluorescencespectrum of a. sample of photoluminescent glass to the emission spectrmof the standard lamp. In order to rake such a comparison useful,certain conditions have been met:

(a) The photoluminescent-glass sample is excited by a known fl.uxof UV light, which has been chopped at 13 cycles-per-second to providean a-c signal from the photomultiplier (which sees the dispersedfluorescence) into a tuned, phase-sensitive amplifier. The UV light,from a 250-watt Hg-arc 'lamp, is essentially the 365-mix mercury line'passed by a 5-,in Co=ing Neo. 5860 glass filter. UV output is monitoredby a vacuum photocell and recorded by a second pen on the same recorder

(b) owing to the temperature dependence of its fluorescence, the

glass sample is held at the convenient temperature of 25 ± 10C whileits spectrum is being recorded. T1his is achieved by thermallyinsulating the sample from the UV lamp and filter, using a radiant-heat-absorbing water cell, and cooling the sample-holder by temperature-controlled water.

(c) The radiant intensity of the standard lamp, which provides thereference spectrum, is reduced by neutral density filters (or by aglass attenuating sphere) to a value comparable to that of the weakfluorescence of the glass. An opal-glass diffusion screen (SC) isadded, being in effect a neutral filter, and also providing a lightsource geometrically similar to the fluorescing glass.

The photomaltiplier scans the photoluminescent-glass spectrum! -- coming from the monochromator, which is provided with an automatic

wavelength drive. After amplification, the phoiomultiplier signal issent through a 4-step, low-pass filter for noise control, and is

2

I"

recorded on a strip-chart recorder. The attenuated spectrum of thestandard lamp is then superimposed on the same chart.

The essential steps in obtaining the spetral-energy output ofthe dosimeter glass are:

(a) Determining the S.R.I. of the standard lamp at the plane ofexit of the lamp beam from the attenuator (i.e., from the opal-gl4ssdiffuser). This requires reducing the non-attenuated lamp. value, foreach required wavelength interval, by applying the spectral trans-mittance factor of the "neutral" attenuator for that wvelengthinterval.

(b) Comparing the recorder-chart ordinates of the glass-fluorescence spectrum to the corresponding ordinates of the attenuatedlamp spectrnun. Owing to the geometrical design of the fluorometer, ,itis only necessary to move the diagonal plane mirror (M-i) from one tothe other of its pre-set positions in order to substitute the knownstandard-lamp output for the unknown output of the fluorescent-glasssample, as seen by the monochromator entrance slit. Thus, the spectralradiant intensity of the glass is obtf inable by a simple comparison ofordinates.

If the primary concern, as in the present case, is to determinethe shape of fluorsscence spectra, the relative spectral transmittanceof the attenuator is sufficient for obtaining spectra normalized topear vaLUe -i0.

The major known sources of error are the following:

(a) standard-lamp spectral data: Reference 9 gives the following:"Tie uncertainty of the spectral data computed by this procedure isprobably no greater than 3% but may be as much as 5%. This estimate oftotal uncertainty is based upon the following estimates of contributinguncertainties:" ..." 2% from uncertainty in the determination of colc rtemperature, and 1% from photometric uncertainty."

(b) The spectral transmittance of the "neutral" attenuator hasbeen obtained by a spectrophotometric measurement, and may add an errorof 2 or 3%.

(c) The overall linearity of the electronic circuits should

probably not be counted on to better than 2%.

(d) It was ne-essary to use 1.0-mm slits to obtain the desiredcombination of signal/noise and scanning speed. However, since

.4t

- °

fluorescence spectra are typically broad bands of wavelengths, a slit-width of this size is tolerable. The manufacturer of the monochromator(a Perkin-Klmer Model 98 single-pass Littrow instrument of effectiveaerture f/4.5 using the manufacturer's "standard" fused quartz prism)gives the resolution at the above slit-idth as follows: At 400ui,10.5 ui; at: 500 mg, 20 m; at 600 m, 31 mg; at 700 my41, 52 up.

As a result of these sources of uncertainty, a cumulative error asgreat as 10% could occur in the absolute values (i.e., in microwatts percm2 per 10 mp). However, for the present purpose, namely the measure-ment of changes in spectral distribution and intensity, by far thelarger pat -of the uncertainty in (a), (b) and (d) is common to allspectra recorded and should disappear from the final results, whichshould then be reliable to within 5%.

III. RESULTS

In order to demonstrate the effect of the photomiltiplier tuberesponse characteristics on the relative spectra and for the selectionof a proper detector, uncorrected spectra have been measured usingdifferent types of photomultiplier tubes. Three commercial dosimeter.asseg and one experimenta on.e iied in- the experiment have thefollowing composition:

TABLE I

Composition of Some Dosimeter Glasses

Glass type, author Composition (% wt)Ag Al Li P 0 Others

DlT-60 (Schulman, et al.)2 4.3 4.7 - 28.4 44.1 10.8 Ba7.7 K

Toshiba (Yokota, et al.)3 4.2 4.6 3.6 33.3 53.5 0.8 B

C.E.C. (Francois, et al.) 2.4 3.5 2.5 33.8 52.5 0.5 Be4.7 Na

11Experimental glass (Becker) 0.6 0.5 7.3 34.7 55.9 1.0 B

4

Glass locks 8 x 8 x 4.7 mm in size have been exposed to identicaldoses of CoW°' radiation under electron equilibrium. After stabilizationby a 20 min 150C heat treatment 5 ,12 for obtaining maximum RPL, therelative spectra in Fig. 3 have been obtained with a IP 21 photomulti-plier tube (S-4 photocathde) and the spectra in Fig. 4 with a IP 22(S-8 photocathode). Because the second tube was less sensitive, theglasses had to be irradiated with 1000 R instead of 200 R. As expected,the different blue-to-red-sensitivity ratios of the two-*otomutipliertube types causes ITL peaks at different wavelengths (560-570 mp and600-610 mj., respectively), different slopes of the intensity decrease inthe red, and different ratios between the RPL intensity in the blue andthe red. 9he different peak heights are in good agreement with theresults of sensitivity comparisons, which have shown about twice thesensitivity for the Yokote. than for the Schulman type glassA2 ,1 3 he IP22 tube was selected for further experiments because this tube permittedmeasurements with reasonable accuracy up to ca. 750 r41.

By correction for the photomultiplier tube sensitivity, "absolute"spectra (spectral radiant intensity as a function of wavelength.) havebeen obtained. For better comparison, the spectra are rormalized attheir peak intensity = 100 (FLg. 5).. It can be seen that there areconsiderable differences between the glasses in the intensity in theblue, the peak wavelength and the slope in the red, in particular,between the Schulman glass and the three other glasses. In some of ourexperimental glasses5 containing the same amount of Ag (ca. 4%), butdifferent phosphate glass bases, similar differences havebeen found

f(Fig. 6). phe avelengt of maximum intensity, however, is about 620 p_ .all _.- in- ,i.ll -anoher- p-.iiiai. t'e gl base was

kept constant and the Ag concentration varied (Fig. 7). For concen-trations between 0.5 and 5% Ag20, no differences in the RPL spectra havebeen found. %be 0.1% AgnO glass spectrum also may be more similar tothe others than it appears from Fig. 7, because the low absolute RPLintensity of this glass did not permit very accurate measurements in thelower wavelength regions.

In another experiment, the possible effect of the ganma dose levelon the RPL spectrum has been studied. The upper dose limit was givenby the possible discoloration of the glass, the lower dose limit by thesensitivity of the instrument. Poor accuracy because of nearness to thelower limit of measurability in the 50 R glass may in part be responsi-ble for the increase in the relative intensity below 500 no in theresults presented in Fig. 8. Above 520 mi, no difference in the RPLspectrum can be seen between glasses which had been exposed to 50, 500and 5000 R. Glasses with a low Ag content such as an experimental glasswith 0.6% Ag exhibit a considerable build-up after exposure. The RPLimmediately after a short time exposure may be only a few percent of the

5I

finl I'L fte stbilzaton 1 2jfinal after stabilization. If the ITL spectra of such a glassafter exposureand after stabilization are compared, only a smalldifference in the spectral shape can be seen in spite of the verydifferent absolute intensities (Fig. 9).

During thermal annealing, however, a relative increase of the low-wavelength part of the, RPL spectrum can be seen if compared with theA13 spectrum-of the seme glass after keeping it for 100 days at roomtemperature after exposure to 1000 R ga~ma radiation in order to obtain

4 the maximim RPL intensity (Fig. 10). There may be thermally-more-stableluminescence centers with a peak intensity below 400 mi. Because of theproximity of the 365 mu excitation line, no measurements in this region

could be made. The blueI IPL disappears during the final stages oft"hermal annealing.

5It has been demonstrated in previous experiments that afterextremely high gamma exposure a permanent increasg in predose whichcannot be annealed will result. Indeed, after 10 R and 2 hoursannealing of the discolored glass at 4000 , the luminescence of theglass shows -a peak -probably below 400 mg and extending intothe red (Fig. 11). Obviously, this effect cannot be omitted by a red

a' filter.

IV. RECOMMENDATIONS

It would be desirable to extend these experiments, in particular,to measure the spectral distribution below 400 and abo re 750 mg; tomeasure the spectra as a function of the excitation wavelength and toincrease the sensitivity in order to make measurements of the "predose"spectra of the modern low predose glassses and, may be, to find methodsfor a better discrimination of predose and luminescent surface con-tamination against the radiation induced effect.

The use of a UV laser for excitation, which has recently beensuggested by Kastner, et al.,l 4 help solve problems of intense and

sufficiently monochromatic excits+'on. The spectrum may also bedetermined during the ca. 3 Psec decay time of the RPL after the end ofa pulsed excitation, which has been found to be by ca. a factor of tenlonger than the predose decay time.

6

In reader design, it would be desirable to use ITL light detectorshaving a sensitivity extending to at least 700-800 Wp for the betterutilization of the RPL output in the red and for better discriminationagainst disturbing "spurious" luminescence and scattered excitationlight.

7

, I

fl, A.

---I i- -- - - a . _ **4-)

4

t$421s

C5 .1

00I

.00

''-4

weJ

7,1

Wn9

I!'

NROL 25"66

P.M. TUBE IP2.I (S-4) EXPERIMENTAL GLASS (BecAer)200 Ry. STABILIZED 20min. 15O *C

I 200-YOKOTA GLASS

Nz_ ' C.E.CGLASS •-' ,o. -/

I-U, /1ILM~l LS

Cr

0400 500 600 700

Fig. 3 Relative radiophotoluminescence (RPL) intensity of differentdosimeter glasses as a function of avelength', uncorrected for photo-multiplier tube IP 21 response.

" 10

I.

4

NRDL25-66i i I i iIi i I i i i I i i i -

P.M.TUBE 1P22 (S-8) EXPERIMENTAL GLASS (Becker)

1000Ry, STABILIZED 20min. 1500C -

200 YOKO M L GASS

= !N-

./ \ / SCMA GL-ASS "

- .

100 .0 /0 0

/ SCHULMANV GLASS

0500 0 700

Fig. 4 Relative radiophotoluminescence (RPL) intensity of differentdosimeter glasses as a function of wavelength, uncorrected for photo-multiplier tube IP 22 response.

-ll

11i

tn

00

/ 0

tn H

~ .1 ~ '

*14

Vr 0

/0

C4)

:1, ~0

1'U ADp52 o-I.)

(Ootnoadpazi-jo Alsuall JUIPO joi~ad

12)

0- -

r4 I______ __ .1

- to 0

I,.

AI 4>

(00

133

.43

It 0

0r 0

44

In r 0

0

C4~tn0

C.,d

0 t

o0

0O 0

(01=ood peziwxjou) Allsualul jumipo lojl~.ds

14

1 LI/ Li

L II4,

4)

.10

E

42

03

o;O

tl(54 ~ ja POIDJU AISOW UIP ias0

o 4,15

/ 0

*43

--

i

0

caS

coo

z 4.w 00

04

UO

4, go

(OO =oa pzitm u lluaullu~imjplad 0O

U)16

.in

+1~

N-1. 1j

//C0 C

E/

(01jodPZIDJO)AISOUIlopo /iod

17i

~00

00

Sr

0-'A

rd A

011

*18

1. J. H. Schulman, R. J. Ginther, and C. C. Xlick (RRL);R. S. Alger and R. A. Levy (MMD).- J. Appi. Phys. S, l1 79 (1951).

2. J. H. Schulman W. Shurcliff, R. J. Ginther and F. H. Attix,Nucleonics l, No. 10, 52 (1953).

3. R. Yokota and S. Nakajima, Health Phys. 13 241 (1965).

4. Y. Uehara, T. Kushida, K. Hanada and S. Kaneko, Toshiba Rev. 17856 (1962).

5. K. Becker, in "Personnel Dosimetry for Radiation Acidents," IAEA,Vienna (1965) p. 169.

6. R. Maushart and E. Piesch, Atompraxis 12 (1966), in press.

7. K. Bingham and S. Parke, Phys. and Chem. of Glasses 6 225 (1965).

8. D. W. cQuilling and R. S. Alger, "An Absolute Fluorometer for theU.S. Navy Phosphate Glass Dosimeter Standards System." USNRDI,Technical Report. In preparation.

9. Louis E. Barbrow, J. Optical Soc. Am. (Ltr), 49 1122 (1959).

10. H. Francois, et al., Int. Congr. Dosimetry Extern. Rad., Paris(1964).

11. K. Becker, Nuel. Instr. Meth. 36 323 (1965).

12. K. Becker, Proc. IAEA Symp. Solid-State and Chemical Dosimetry,Vienna (1966).

13. J. Cheka, Health Phys. 1 303 (1964),

14. J. Kastner, et al., Proc. IAEA SyMp. Solid-State and ChemicalDosimetry, Vienna (1966).

19

r<

SecuAt Classifcation

DOCUMENT CONTROL DATA - R & D(Selvrft C;s1fkde"o 8. had body tzac and 8&ISCZ zk-j Ac~ bentzd wbeoa r~lsvpi seassif.e)

I. ORIGIM.ZtNG ACTIVITY (Carp~o" A - 2.I REPORT SECURITY CLASS3lCATIO CU. S. Kwall Radiodlog~ica!1 Defense La.boratory_ UM KSS _- 1

f- OP0UJVfliESCE!C SPECTRA OF S DWTTIVMTD UA&Sr

4. OCSCRJPTIZV NOTES (27pe *f"wrZ ad iclu.lro dres)

S. AUTIORgS) (fitsI Meo. mide Waal. teSS sM)

)MAUS BYIOKERD~MAW W. YXQM

6-REPORT DATE 7A. fTTAL A30. CrPAGES 7Xh No. or MREFS

25 August 1967 29 I 181 CONTRACT OR GNANT NO. 90K. ORIGIAMTORS REPORT I.LOdDERI1S

US, DL-T-67-66b. PROJECT NO. !LVSHIPS, Su project SF 01

05 U-1, Task 0503C. U,. O 14ER REPORT NOtSI (Aq7 odMert =6", st.r y be 1szlowdih.is ,wpofl)

10. OISTRIUTION STATEMNT

THIS DOCUM HAS BM I A-PP E FOR IUBIC R=,-E AND SALE; ITS DIS=rI3TVC

$I. -UPPLEM7NTARY NOTES 12. SPONSORING MILITARY ACTIVITY

IAVAL SHIP SYS !S C OMEAIDWAShMGTON, D. C. 20360

13. A8STRACT

The radiopnotoluminescence (RPL) spectra of silver-a':tivated phosphate glassesof different compositions under 365 mj± excitation hare been measured in the 400to 750 rn± range. The apparatus, specially designed for this type of measure-ment, uses a red-sensitive photomaltiplier tube a-i permits comparison of therelative RPL intensity with the knn output of a MUJS-calibrated lamp for eachwavelength interval.

Shape and maximum of the RPL spectrum depend on the type of photomultipliertube, if no corrections for the photomultiplier spectral response are made.After corrections, a steep RPL increase above ca. 500 mg, spectral radia-.tintensity maxima between 615 (French C.E.C. glass) and 640 mg (DT-60 glass)depending on the glass composition, and a slow increase in the red have beenobserved. For a constant glass base composition, the Ag concentration offrom O.5% to 5% did not affect the spectrum. Also the dose level and thedegree of the RPL build-up after exposure had only minor influence on thespectrum. During thermal annealing, a RPL increase in the blue has beenobserved. After a megarad exposare, a permanent RPL which could not be an-nealed has been found. Some further experiments are suggested.

+t D ORM 7 (PAGE 1)D D, PA.GE UNCLASSIFIED

DSN 0101-807-6801 Security Classification

I