II, NUOVO CIMENTO VoL III , N. 1 1 o Gennaio 1956

Photographic Methods in y-Ray Scintillation Spectroscopy.

B. CHINAGLIA and F. DEMICHELIS

fstituto di Fisica Sperimentale del Politecnico - Torino

(ricevuto il 21 0ttobrc 1955)

Summary. We deal with an experimental research on the different possibilities to get efficient photographic methods of y-spectroscopy through scintillation crystals. The spectrum is always strongly disturbed by the Compton distribution, and the possibilities to reduce it by a fitted geometry of the crystals and an anticoincidence device, or by using a proper sized crystal, are discussed. In any case the resolving power is rather poor owing to the fluctuations in every stage of the process from the :t-ray to the recorded pulse. But strong improvements have been reached by a number of laboratories as well in the photonmltipliers as in the crystals and in their mounting. This is proved by the comparison up to date spectrograms with those which it was possible to obtain only a few year ago.

1 . - I n t r o d u c t i o n .

The purpose of this s tudy is to draw at tent ion on the actual possibilities

of scintillation methods used to obtain y-r~y spectra from r~dioactive sources,

~nd to present the results of our research with several of these methods.

The aim of "( spectroscopy is to obtain from these rays a spec t rum which

satisfies the conditions of resolution, absence of disturbances and uuivoc~l

correspondence between an hv v~lue (energy of the ~'-ray) ~nd ~ line of the

speotrum.

In the nuclear field this beco,~les quite complicated owing to :

1) the nmltiplici ty of factors affecting the efficiency of the apparatus

thus lessening the degree of resolution;

2) the number of electronic apparatuses necessary in order to record

the ou tpu t pulses from the photamuLipiiec;

3) the complexi ty of the interact ion of ~, radiation with the m~tter

52 B. CHINAGLIA and F. DEMICHELIS

However, within the last, few years, work in this field has made very notice- able progress; in fact, today the photographic spectrum of the oscillographic pulses coming from a photo.nultiplier enables a ready and easy solution of qualitative problems of spectral analysis of radioactive sources.

If conveniently calibrated, the apparatus permits quick measurement of the energy of the radiations.

We have applied the scintillation device in several experiments on radio- activity (~) and these researches led us to develop several reglizations in the field of the y-rays spectroscopy.

Moreover, the photographic method can also be used in order to kuow the intensity of the radiation; the densitometer analysis of the exposed emul- sion can give the distribution of the pulses on the film according to the usual methods of optical quantitative spectral ~nalysis. This leads to the measu- rement of the intensity of the radiations, in the limits o[ the normal exposures.

The experimental difficulties introduced by the photographic method in this case are analogous to those of the optical spectral analysis mentioned

above.

2. - E x p e r i m e n t a l m e t h o d s .

The detector is a lqaI(T1) crystal placed directly on a photo rnultiplier which , sees )) the scintillations produced in the crystal by the ,/ radiation.

In our research we have used Du Mont photomul~ipliers and cylindrical crystals of various sizes ranging between 20 mm in diameter, ]3 mm in thickness and 50 mm in diameter, 50 mm in thickness.

The bigger crystals were more suitable for obtaining the spectra, for reasons that shall be explained later, thus justifying their higher cost.

The output pulses from the photomultiplier are conveniently amplified and are sent directly to a wide-band oscilloscope.

Pulses from y-rays which lose the same amount of energy in the crystal

ought to produce identical patterns. The oscilloscope's screen is photographed with a time exposure of sufficient

length with the result that pulses of the sane height, adding up their photo- graphic impression, give rise to a well defined pattern on the photographic film.

T~entyfive superposedl inages of saccessive equal patterns corresponding to the sa~e loss of energy in the crystal, with the resolution and sweep speed of our apparatus, are sufficient to produce a quite well defined photographic

impression.

(1) F. DEMICHELIS a n d R. MALVANO: Nuovo Cimeuto, 11, 49 (1954); 12, 358 ( ]954) ;

F. DE~ICI~ELIS: Nuovo Cimento, 12, 407 (1954).

P n O T O G R A P I t I C M E T H O D S 1N "/ '-RAY S C I N T I L L A T I O N S P E C T R O S C O P Y ~ 3

The time exposure of a spectrogram is chosen so that the emulsion may give a clear photography. For instance, the spectrograms of Fig. 1 were ob- tained from ~ 100000 pulses.

a) The formation of spectrograms by means of the method indicated in a former article (1) proved to be of little value.

According to this method the pulse drives the vertical deflection of the electronic beam, while the horizontal deflection is driven by its derivative.

The various spedtral lines appear as closed curves of different size which lie on the same side of a straight line which is tangent of ~11 of them at the same point.

But this method has been given up both because of the lack of simnmtry of the lines, and because two lines generated by two beams of 7-rays with the same intensity, but having a differing l~v energy, do not produce equal photographic impressions.

b) The following method proved preferable: the tension pulses are sent directly to the vertical deflection plates of the oscilloscope, the horizontal axis acts as the t ime axis, and the sweep is triggered by the same recorded

y-ray. The simultaneous sweep of the t ime axis does not allow the scanning by

the microphotometer of the picture of the pulses, because the pulses have leading edge much more s~eep than the trailing edge and their oscillographic patterns were nbt fit for a microphotometric sca.nning.

c) But we obtained pulses of equal slope for both leading and trailing edge, by means of a lumped paramet~er delay line.

The resulting pat tern for every pulse has the shape of an error curve of Gauss, with a vertical symmetry axis passing through its top, and common to all patterns.

Furthermore, in this case the curve, near the top, is practically parallel to the time axis, and thus is traced by the electronic beam having a speed indi- pendent of its height.

Thereby a microphoto~mtrie scanning ~long the vertical symmetry axis produces satisfying results.

60 In Fig. I are shown some spectrograms of this type, respectively of .~7Co, 3 7 ~ 226 ~C~, ssRa (in equilibrium with its decay products) ~nd ~S0RdTh (in equi- librium with its decay products).

In Fig. 2 is shown the scanning by means of the microphotometer upon the ~37Cs spectrog'rau~. 55

In these figures it is possible to see how much information is available from the spectrogram.

We must emphasize that the apparatus necessary to obtain such a spectro..

54 B. CHIhrAGLIA and F. DEMICHELIS

gr~,~n is quite simple, bu t of course it requires tha t bo th crystal and the photo- mult iplier can give a good resolution, and moreover the oscilloscope screen be of fine grain so as not to weaken the resolution of tile apparatus .

d) Still another me thod can be followed; i t produces a speetrogra,~ in which the pulse pa t te rn is formed by a series of straight lines parallel to the

tinge axis. Such a spectrum~ which is analogous to an optical spectrum, is easily

obta ined in the following way. The pulses are stretched, so tha t af ter a very short rise t ime, they remain

at constant height for ~ 5.10 -6 s. On the oscillographic screen are seen straight li~es whose lengths depend

upon the constants of the pt~lse s t re tcher and whose heights ~re proport ional

to the -y-ray energies. In Fig. 3 are shown some spectrograms of this type. Since the pat terns are parallel to the t i ,ne axis, the sweep speed is uni-

form for all the pulses, and therefore for all energies. The scanning by the microphotometer is then possible along any line per-

pendicular to the t ime axis.

e) In order to obtain the measurement of the intensi ty of the spectral

lines without the use of a microphotometer , a neut ra l wedge may be used (2). The wedge must be placed on the screen of the oscillograph in such a

manner tha t isodensity lines be perpendicular to the spectral lines. In our case there is a transmission factor, ranging from 1/5 to 1/1000

(optical density 0.7 --3). In Fig. 4 are shown spectrograms obtained by means of the me thod of

spectral lines together with neutral wedge. We must no~.e however tha t the in t roduct ion of a lengthener of pulses re-

quires a greater lenght of t ime for each pulse. I f crowding occurs, it is pos- sible for one pulse to s tar t before the proceeding has co,ne to end; this would result in ~n overlapping of pulses and consequential clouding of the spect, ro-

gram.

3. - Dis turbance from Compton effect and its reduct ion.

F r o m the observat ion of the above spectrograms it is easy to recognize

tSe photoelectr ic lines and the continuous Compton distribution. The pair product ion line is appreciably present only in the spectrogram

of R2~SodTh due to the 2.62 MeV y-ray of ~S~Pb.

(2) D. MAED]~R: Helv. Phys. Acta, 20, 139 (1947); W. BERNSTEIN, R. L. CHASE and A. W. SCHARDT: Rev. Sci. Instr., 24, 437 (1953).

B. CI-IINAGLIA a n d F . DEI~[ICItELIS

~7Co, b) 55Cs, c) saRa, d)2~oSRdTh. Fig. 1. - Spectrograms of: a) 6o 137 2~

137 Fig. 2. - Microphotometr ic analysis of the 55t, s spectrogram of fig. 1.

228 - 27Co, b) ss~a, Fig. 3. Spectrograms of: a) 60 e2~n c) 9oRdTh (obta ined by means of a crystal

12 m m in diameter , 12 m m in thickness).

~7Co, b ) 5 ~ s , ssR~, Fig. 4. - Dens i tometc r analysis of the spectrograms of: a) 60 137~ c) 826 d) 2~RdTh, ob ta ined by means of the neu t ra l wedge method.

PHOTOGRAPHIC METHODS IN ~{-RAY SCINTILLATION SPECTROSCOPY 55

I n fact the Compton cont inuous dis t r ibut ion is a strong disturbnuce.

The technique of reducing this Compton dis turbance b y means of an anti-

coincidence me thod is not new, and has already been applied quite success-

fully. This me thod consists in surrounding the de tec tor crys ta l with another

(, guard >> crys ta l which cap tures the radia t ion scat tered in the fo rmer b y Compton

effect.

The ou tpu t pulses f rom the second crysta l are sent in anticoincidenee with

those f rom the first, so t ha t only the photoelectr ic lines r emain in the result ing

spect rogram.

This metho~l could be called primary spectroscopy, since it records only

the photoelectr ic pulses. We knew the work 1~. E. CONNALLu (a) did in ,;-rays spectroscopy b y means

of an anticoiacidence appara tus . Bu t when we wanted to muke use of this method, it seemed to us possible

to improve his appa ra tus in order to obta in even be t t e r results.

I n f ac t the efficiency of this me thod

depends s imply upon the geometr ical

dimensions of the guard crysta l : by using

a crystal t h a t would capture all the

scat tered ,;-rays, the Compton reduct ion

should result quite complete.

Therefore we resolved to Frepare an appa.ratus wi th a crysta l viewed f rom

the central crysta l with a solid angle of

about 4~: (see Fig. 5). I n the lup~e of t ime during the

ac tua l p repara t ion of the device, 1~. D. ALBERT (4) publ ished the results of his

own exper iments wi th the anticoinci-

dence me thod which were no tab ly be t t e r

t han the preceeding ones.

I 8 J I

p s Eli Ik

i I

Fig. 5. Anticoincidence y-rays spec- trometer with guard crystals viewing a solid angle near 4z. A) main detec- tor; B) guard crystals; P) photo-

multipliers.

Our own device has foreseen the use of a lit t le photonml t ip l ie r D u Mont

t y p e K 1193; t hanks to i ts small size, we can avoid the use of a light pipe;

however, the inavai labi l i ty of such a pho to tube up to date and the difficulty of

obtaining some specially shaped scintil lating crystals, have not ye t allowed us to prove exper imenta l ly to what point Compton dis turbance can be reduced

b y me~ns of our own device.

(3) R. E. CONNALLY: Rev. Sci. Instr., 24, 458 (1953). (4) R. D. ALBERT: l~ev. Sci. Instr., 24, 1096 (1953).

5 6 B . C H I N A G L I A a n d F . I ) E M I C H E L I S

A prel iminary appara tus already in use, owing to the lack of the above photo tube , is l imited only to the scheme of Fig. 6.

Fig. 6. - Anticoincidence y-rays spectro- meter used in this research. A) main detector; B) guard crystal; P) photo- multipliers; L) plexiglass light pipes.

A is a cylindrical crystal, 12 m m in

d iameter and 12 m m in thickness, B is

a hole crysta l 50 m m in diameter , 50 m m

in thickness, with a hole 25 m m deep and

25 m m in dia~meter.

In our appara tus we did not use any

electronic anticoincidence circuit, as it would have been required had we used

a differential analyser.

The oscilloscope alone works s imply as an anticoincidence device. As ~ m a t t e r of fact the pulses coming f rom A (see Fig. 6) drive vort ical

deflection in the usual way, while the pulses coming f rom B tr igger the b e a m

intens i ty till to cancel it.

Thus each t ime tha t the pulses come simultaneously f rom the two channels,

no p a t t e r n appears upon the screen.

Some no tewor thy conclusions m a y be drawn f rom this first tes t ing of our

appara tus . I n Fig. 7 is shown the expected reduct ion of the Compton dis t r ibut ion

theoret ical ly eva lua ted on the above geometr ical dimensions of B, assuming A as a point and the energy of the incident y- rays of 0.661 MeV.

Curve a) refers to rays com- ing f rom the hole side of the guard crystal, in which case the klsc

backsca t t e red photons, corre- sponding to higher pulses in

the scintil lation spect rum, are

not c a p t u r e d . . ~ Curve b) refers to rays co- _~

ruing f rom the opposi te side.

I n Fig. 8 are shown the expe- ~

r imenta l spect ra obta ined b y

means of our appara tus , using

the 0.661 MeV y-rays of ~37Cs ~' 55 '

and corresponding to curves a)

and b) of Fig. 7. I n apply ing the guard crysta l

technique we mu~t pay a t ten t ion

to the nuclides emi t t ing two or

more y - rays in cascade. As a m a t t e r of fact the

~"" ~ .(%b) j

(a) \ .

o'.I 0:2 o3 0.4 os Electron energy In MeV

Fig. 7. - Theoric Compton recoil electrons di- stribution for energy of incident y-rays of 0.661 MeV. Curve a) shows the reduction when y-rays are coming from the hole side; curve b) when the rays are coming from the oppo-

site side.

B. CHINAGLIA a nd F. DE~ICHELIS

137 Fig. 8. - a) Spectrogram of 55Cs. b) The same spectrogram obtained by means of the anticoincidence apparatus: the ,Frays come from the hole. c) The same spectrogram of fig. b: the ~-rays come from the opposite side. d) The same spectrogram of fig. a)

obtained by means of the bigger crystal.

Fig. 9. - Spectrogram of 2~]RdTh obtained by means of the 50 mm in diameter, 50 mm in thickness crystal. Note the reduction of Compton continuum compared

with the analogous spectrogram of fig. 3c).

PIIOTOGRAPHIC METHODS IN ~{-RAY SCINTILLATION SPECTROSCOPY 5 7

intensi ty of the photoelect r ic lines of y- rays in cascade is subject to lessening.

This is due to the fac t t ha t if one of the y- rays is cap tured b y photoelect r ic effect in A, and at the same t ime the other -l-ray in cascade is captured in B, the pulse genera ted in A is not recorded.

However this effect can be minimized by the use of a suf[iciently narrow b e a m of y - rays impinging on A.

I t is interest ing to n o t e t ha t the photoelect r ic capture in A of a pho ton

scat tered by B, does not cause the pulse recording, and therefore does not cause dis turbance.

We wished also to exper iment on the advan tage in the rat io of the n u m b e r

of photoelectr ic to the n u m b e r of Compton pulses, derived b y the use of the big guard crys ta l as ma in de tec to r (5).

Indeed a greater n u m b e r of scat tered quan ta are absorbed in the trysted itself.

By the use of crystals of appropr ia te sizes nearly all the scat tered photons

lose their energy in the crystal , with an a lmost to ta l reduct ion in the Compi on

distr ibution and an indirect photoelect r ic line enhauce,nent (see Fig. 8d).

A spec t rogram obta ined b y means of the bigger crys ta l is shown in Fig. 9.

The photoelectr ic lines of higher energy are not iceably more intense (see also Fig. 3d).

The reduct ion of the Compton dis t r ibut ion is comparable witht theft ob- ta inable by the use of the same crys ta l as guard detector .

Though more complicated f rom a theoret ical s tandpoint , the anticoinci-

dence m e t h o d is much more pract ical f rom an economical s tandpoint .

As regards the use of the big crystal, we mus t t ake into account its high price when used as main detector , since it mus t have all those propert ies which are necessary for a good resolution.

Poorer propert ies are required for a guard crystal . Since i~ does not require

a strict uniformity, it can be made up of small crystals, not even opt ical ly worked, because an homogeneous immers ion in silicone oil is sufficient.

As a low price plast ic scintil lator can be easily molded into any desired shape and size, we th ink t ha t it can be also successfully used, b u t at this date we dont dispose of it.

F r o m the very beginning of research in y- rays scintil lation spect roscopy (6)

and th rough the more recent exper iments up till the present day, most satisfying results have been and are still being obtained.

(5) p. R. BELL: Nucleonics, 12, 10, 15 (1954). (6) j . W. COLTMAN and F. H. MARSHALL: Phys. Rev., 72, 528 (1947); M. DEUTSCII:

Nucleonics, 2, 58 (1948); P. R. BELL: Phys. Rev., 73, 1405 (19_~8); R. HOFSTADTER: Phys. Rev., 75, 796 (1949); R. HOFSTADTER ~nd J. A. MCINTYRE: Phys. Rev., 78, 617 (1950); 79, 389 (1950).

5~ B. CHINAGLIA and F. DEMICHELIS

W e t h i n k t h a t our r e sea rch could h a v e p r o d u c e d s t i l l m o r e r e m a r k a b l e

r e s u l t s ; k u t t o o of ten our w o r k was r e q u i r i n g p a r t i c u l a r a p p a r a t u s e s n o t eas i ly

o b t a i n a b l e f r o m t h e f i rms or t o o expens ive .

W e wish t o t h a n k Prof . E . PERUCCA for his he lp fu l i n t e r e s t in our s t u d y

O u r e x p e r i m e n t a l se t u p was m a d e poss ibe b y a g e n e r o u s f in ianc ia l s u p p o r t

for which we w o u l d l ike to exp res s our t h a n k s to C.N.R.

R I A S S U N T O

Si prendono in esame i vari metodi J~otografici di spettroscopia di raggi "~, mediante eristalli scintillatori. Lo spettro ~ sempre per turbato dalla distribuzione dovuta ~ll'ef- ~etto Compton. Si discute la possibilits di eliminare questo disturbo sia mediante ada t t a sistemazione geometrica dei cristal]i ricevitori e un dispositivo ad anticoincidenze, sia utilizzando un cristallo di opportune dimensioni. La risoluzione 5 sempre piut tosto l imi ta ta a causa delle fluttuazioni nella catena, di processi dal quanto ,/ incidents a]la registrazione dell ' impulso. M a i l progresso raggiunto da vari costrut tori sia nei foto- molt ipl icatori sia nei cristalli scintillatori e nel loro montaggio si rivela nei forti miglio- rament i degli spet t rogrammi ot tenut i ora r ispetto a quelli realizzati anche solo qualche anno fa.