WM

*ll’l

I
l

r

146
168

 

TH '

TEMPERATURE DEFENDER? LUMINESCENCE
OF CaW'CD‘4 AND CJW'O4

Thesis for flu Degree of M. S.
MICEEGAN STATE UNIVERSITY
Merrit L. Mallory
1961

ssssss gmflilflfl'mflllfll‘lfli'lfllflfl'lml‘fl'lflfll”WWII :

m 3 1293 01743 0277

 

LIBRARY '

Michigan State
University

 

wk z 0 29m

f: «NF

 

ABSTRACT

TEMPERATURE DEPENDENT LUMINESCENCE

OF CaWO4 AND CdWO4
by Merrit L. Mallory

The relative efficiencies and decay times of alpha particle induced
scintillations of CIaWO4 and CdWO4 were investigated as a function of
temperature in the range 100K to 3750K. Their behavior at intermediate
and high temperatures is in agreement with that expected from the Mott-

Seitz-Kroger configurational coordinate model. Values of E thermal

Q)
quenching energy, of O. 34 and 0. 31 ev were found for CaWO4 and CdWO4,
respectively. As the temperature was decreased below 600K, an increase

in the decay times and a decrease in the relative efficiencies were found.

This behavior can be explained qualitatively by assuming a trapping level.

TEMPERATURE DEPENDENT LUMINESCENCE

OF CaWO4 AND CdWO4

BY

MERRIT L. MALLORY

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Physics and Astronomy

1961

ii

ACKNOWLEDGEMENTS

I am most grateful to Dr. G. B. Beard and Dr. W. H. Kelly for
their help and encouragement throughout the execution of this work.

Iwish to thank Dr. F. J. Blatt for the fruitful suggestions, and
discussions on the interpretation of the data. The assistance of Dr. M.
M. Garber and Dr. H. A. Forstat on the low temperature measurements

is gratefully acknowledged.

iii

TABLE OF CONTENTS

Page
I. INTRODUCTION ................................ 1
II. EXPERIMENTAL ARRANGEMENT ................ 2
III. RESULTS AND INTERPRETATION ............... 6
BIBLIOGRAPHY ................................. 15
APPENDIX I ................................... 16

APPENDIX II . . . . . .............................. l9

Figure

LIST OF FIGURES

Cross sectional diagram of probe ..................

Reciprocals of efficiency (no/n) and decay time (10.5/7)
sec”1 plotted as a function of the reciprocal of the
absolute temperature for CaWO4. Upper diagram:
Entire temperature range. Lower diagram: High
temperature range expanded by a factor of 10. .......

' Reciprocals of efficiency (no/n) and decay time (10-5/7)

sec"l plotted as a function of the reciprocal of the
absolute temperature for CdWO4. Upper diagram:
Entire temperature range. Lower diagram: High
temperature range expanded by a factor of 10 .......

Configuration coordinate diagram ..................
White cathode follower preamplifier circuit ........

Simplified diagram of the photomultiplier anode
circuit and block diagram of detection system .......

iv

Page

10

17

18

LIST OF TABLES

Table Page

I. Results obtained from the CaWO4 and CdWO4
scintillation efficiencies and decay times . . . . . . 12

I. INTRODUCTION

Both CdWO4 and CaWO4 crystals have been known to be good
1

scintillators for some time. Because of their high densities and high
atomic numbers, the crystals have relatively high photo-efficiencies.
However, they have the disadvantage of a long scintillation decay time
which makes them poor for high counting rate experiments. It is only
recently that their scintillation properties have been studied in some
. . . . . . . . . . . 2, 3
detail in connection With their use for certain speCific investigations.
These investigations included a search for the natural alpha activities
of tungsten in CaWO4 and CdWO4 and the relative energy response of
CaWO4 to various sources of excitation.

I have undertaken a study of the effect of temperature on the
scintillation efficiencies and decay times of CaWO4 and CdWO4 crystals
primarily to ascertain whether the scintillation response could be
improved by operating in a particularly favorable temperature range.

4
Kroger has previously investigated the relative luminescent efficiencies
of CaWO4 and CdWO4 as a function of temperature using excitation from

. . . . o o
ultraViolet light (A = 2537 A) in the temperature region from 80 K to 480 K.

. . 210
In the experiment reported here alpha particles from P0 were used as
o o
the source of excitation in the temperature region from 10 K to 375 K.
Measurements were also made over a limited temperature range using

137
Cs gamma rays. ~

II. EXPERIMENTAL ARRANGE MEN T

Various crystals of CaWO4 and CdWO4 with dimensions of about
10 mm x 5 mm x 3 mm were used. The crystals were polished such that
one side was flat. This side was placed in direct contact withalight pipe.
On the side opposite the flat side, a drop of P0210 NO3 solution was
placed and evaporated to dryness. A counting rate of roughly 6000
counts/minute was used with all crystals. The crystal and the remaining
exposed end of the light pipe were covered with a thin layer of MgO,
aluminum foil, and black electrical tape. Using this arrangement, little
difficulty with light trapping is experienced. 5 A thermocouple and carbon
resistance thermometer were mounted in contact with the aluminum foil.
The light pipe and crystal were placed in the probe as indicated in Fig. l.

The temperature of the crystal was measured with a constantan-
copper thermocouple in the region from 300K to 3750K. Below 300K it
was measured with an Allen-Bradley 56 ohm, 1/10 watt, carbon resistance
thermometer.

The measurements were divided into five overlapping temperature
regions, corresponding to the appropriate temperature baths. In order to
obtain temperatures above those of the cooling baths a variable current
was passed through a Manganin heating wire wrapped around the probe.
The temperature was held to within a variation of 30K during the 10-minute

counting runs in the liquid helium region while for similar runs in the

0
other temperature regions the deviations were less than 1 K. For each

o Pholomulliplier

—>
.1

<— Quartz Light Pipe

[It

German Silver
Tubing ;

 

Brass

 

Heafinq

Coil ——>

W

 

 

 

 

L__

Copper-Constantan Carbon Resistance
Thermocouple Thermometer

Tungstate Crystal

  

 

 

 

F—W—i

Figure 1. Cross sectional diagram of probe.

4
point 30 minutes were allowed for temperature equilibrium to be reached
throughout the crystal before the data were taken.

To detect the scintillations the light pipe was mounted on RCA 6342
photomultiplier. The anode RC time constant was ~10.-3 sec, and the
preamplifier consisted of a conventional White cathode follower. As long
as the scintillation decay time is small compared to the anode time
constant, the decay time can readily be determined from the rise time
of the output pulse and the amplitude of the output pulse will be proportional
to the scintillation efficiency.* The photomultiplier temperature was moni-
tored and maintained constant to within 20C. For various runs light pipes
of lucite or quartz were used. The failure to detect any differences in the
relative scintillation efficiencies and decay times using the two different
light pipes makes it reasonable to assume that any wavelength shifts in the
detected luminescence radiation were relatively small. This conclusion is
in accord with the results of Kroger4 who found little change in the
emission spectrum in going as low as 800K using ultraviolet excitation.

As a further precaution an RCA 6903 photomultiplier (quartz window)
was also used for runs in the region of 100K to 900K. No noticeable change
was observed in the scintillations.

A Tektronix 531 oscilloscope was used to observe the output of the
preamplifier. The trace of the scope was photographed using exposure times
sufficiently long to determine a reliable average of the amplitude and decay

time. As a check on and to supplement these measurements a model A-6l

 

*Appendix I.

amplifier (modified for long rise time pulses) and a 256 channel pulse
height analyzer were also used to determine the average pulse height at
each temperature. The amplifier-analyzer response was measured
using a variable rise time pulser and a correction factor depending on
the rise time of the pulse was determined. This factor was used to
correct the pulse height data. Measurements were made on the crystal
at room temperature before and after each experimental run and were

found to agree.

III. RESULTS AND THEIR INTERPRETATION

The analyzer data were compared with the photographic data and
the results agreed to within 10% for the relative scintillation efficiencies.
Figures 2 and 3 are the results obtained for CaWO4 and CdWO4 with the

210 . .
Po alphas. The effiCiency data represent an average of the results
as determined by the two methods. The reciprocals of decay time and
relative efficiency are plotted against the reciprocal of the temperature
to facilitate the comparison with theory. The efficiency at 2730K is arbi-
trarily chosen as 100%. The uncertainties in the relative efficiencies and
decay times are estimated to be 51-10%.
Within the limits of the experimental accuracy, data obtained using
137 . . . . . .
Cs gamma rays as a source of exc1tation in a limited region above
~. . 210 _
and below room temperature agree With the results usmg Po alphas in
the same temperature range. These results are also in agreement with
5
those of Gillette. The absolute scintillation efficiency is greater for

!

gamma ray excitation. Although artificial crystals of CaWO4 were

used in this work, previous experience has shown that one obtains similar

scintillations using a natural CaWO4 (scheelite) crystal at room temperature.
Relatively little theoretical work has been done on the problem of

luminescence in pure (unactivated) crystals. The most recent work on

temperature quenching dealing with crystals of the type used here appears

4 6
to have been done by Kroger and P. T. Botden. The model advanced to

explain their work on temperature quenching using ultraviolet excitation

also is in agreement with the results reported here for CaWO4 and CdWO4

J. -|
T (deg K)

 

 

 

 

 

0.l0 0.05 0
IO __ I I T I I I j I I T I I
’ Ca W04- Low Temp. :
5 ..
3 ~ -
n
and 2 ~ ..
IO'S
,1, (sea. I) E
|.0 E n I
r I :
1. / —1
0.5 : |O_5 // —1
T
_ / i
//
O 2 _ / r
f /
/
1 1 l 1 1 /4 1 1 1 1 1 1
: / :
.. / .1
11° * / Hl h Tom . ‘
w 5.0 - 9 p . _
and r / ‘
l(Ls-(seen) " l -5 I ‘
'r - IO
20 s ‘T —1
LC E 1?- E
_ 1' :
0.5 l l l l l l M l l l l l
0.0l0 0.005 0

_|_ -1
T (deg K)

. - -1 ,
Figure 2. Reciprocals of efficiency (no/n) and decay time (10 5/-r) sec plotted as afunction
of the reciprocal of the absolute temperature for CaWO4. Upper diagram: Entire
temperature range. Lower diagram: High temperature range expanded by a fac-
tor of 10.

l0
5
l.
and 2
I0'5 -1
—'r (sec. )
lo
05
02
n.
T 5.0
and
Krs
—,;—(sea. I)
20
ID
05

_I_ -1
T (deg K)

(105

 

IIII

I

I

IIIII

 

CdW04- Low Temp.

I

 

 

 

 

 

 

_ /

_ ‘//

E / / High Temp.
i a

*H i

r

 

 

 

0.005
.L -I
T (deg K)

. -5 -1
Figure 3. Reciprocals of efficiency (no/n) and decay time (10 /'r) sec plotted as a function
of the reciprocal of the absolute temperature for CdWO4. Upper diagram: Entire

temperature range.
tor of 10.

Lower diagram: High temperature range expanded by a fac-

9
o
in the temperature range above about 60 K. For this temperature region
Kroger uses the picture of configuration coordinates as applied to
. . 7 . . . .
luminescence by Seitz, With the modification as proposed by Gurney
and Mott. Figure 4 is a configurational coordinate diagram shOWing the
ground state and only one excited state of a luminescent center. The
ordinate of the curves is the total energy of the system, including both
ionic and electronic terms. The abscissa is a configuration coordinate
which specifies the configuration of the ions around the center. The
equilibrium position of the ground state in Figure 4 is at A. If the center
is excited, it is raised to the excited state at B. A new equilibrium is
obtained at C, with the energy difference between B and C given up as
phonon emission. The center then decays from C to D by photon emission
and again the energy difference between D and A is given up as phonon
emission. The decay from C to D is assumed to be temperature inde-
pendent. Gurney and Mott proposed that an alternate return to the ground
state could occur by a non-radiative transition at E if the excited state at
C is given sufficient thermal energy, E . Thus, the photon is not emitted
Q
. . . . 9
and thermal quenching results. This leads to the followmg equations for

the luminescence efficiency and decay time. ’i‘

n =[1 + s/PL exp <-EQ/k"r) 1"1 (1)
%_::—;J:SexpL-EQ/I{T) (2)

where n is the efficiency for luminescence, S is a constant, PL is the

 

=:<
Appendix II.

Total Energy

 

 

Figure 4.

Excited State

0

Ground State

 

A

a

 

Configuration Coordinate, r

Configuration coordinate diagram.

E

__T_

“I:

""7

AE

Trapping
Level

V

11
probability of luminescence with no thermal quenching and equal to l/TL,
EQ is the energy difference between states C and E of the excited state,
and “r is the measured decay time. It is seen from the data presented in
Table I that the value of EQ determined from the decay time data and
scintillation efficiency data agree very well. The value obtained for
EQ = 0. 34 ev for CaWO4 agrees with that obtained by Botden6 using ultra-
violet light (x = 2537 81) as the source of excitation. It is interesting to
note that the EQ values obtained for CaWO4 and CdWO4 are approximately

equal. Table I contains also the values TL and S obtained for the two
tungstate crystals.

The decay time of CaWO4 at room temperature, T = 5 microseconds
is found to agree with the value given by Dixon and Aitken. 3 The decay
time of CdWO4 at room temperature is 7.1 microseconds.

The model used here to picture the luminescence behavior above ~
600K does not describe the small dip noted in the luminescence curve for

CaWO4 at = 0. 004 (deg. K)—1. This dip does not have a corresponding

_1_
T
variation in the decay time data and does not appear in the CdWO4 data.

In the discussion above it is assumed that both the radiative
transition probability and the proportion of the absorbed exciting radiation
actually absorbed in fluorescent centers are constant and independent of
temperature. According to this picture the luminescence decay times and
efficiencies should remain constant below the temperature quenching

region. This is in obvious disagreement with the experimental results.

However, by a relatively simple modification one can obtain qualitative

12

TABLE I. Results obtained from the CaWO4 and CdWO4 scintillation
efficiencies and decay times

 

 

E from
Q 1
. . . ‘r , sec S, sec
Scmtillation . L
, , Decay time
effiCiency
data, ev
data, ev

 

6 10

CaWO (o. 34:“0. 03) (0. 34:0. 03) (6. 72‘0. 7) x 10" (1. 6I0. 8) x 10

6 10

cawo4 (o.3ofo.o3) (o.32f0.03) (7.8I0.8)x10- (0.8io.8)xio

 

l3

agreement between the model and experiment for both the high and low
temperature regions.

The proposed modification is as follows:

Assume there exists a metastable level F lying an energy AE below
C and that this level is preferentially excited from B. See Figure 4. The
center may be thermally excited to C from F with a probability proportional
to exp(-AE/kT), where it may decay from C to D by photon emission. Also
the center may be de-excited from F to D by an unobserved transition whose
rate may or may not be a function of temperature. It should also be
assumed that the probability of the de-excitation by the unobserved transi-
tion is small compared to that of the thermal excitation to C at temperatures
above 600K. From this picture one would expect the decay time of the
transition from C to D to increase and the scintillation efficiency to
decrease (approximately as exponentials) as the temperature is decreased
below AE/k; this is the trend that is observed. If one assumes that the
de-excitation of the trapping level by the nonradiative transition is
independent of the temperature then one obtains relations for the lumines-
cence efficiency and decay times that are somewhat similar to Eqs. (1) and
(2). If one assumes that the state C is de-excited only by a luminescent
transition in this temperature region, then the luminescence efficiency is

given by the relation

17 = [l + PT/S' exp (AE/kT)]-1 (3)

where PT is the probability per unit time for the excitation of the trapping

level via the unobserved transition, and S' is a constant.

14

Since the decay leading to luminescence goes by a cascade of levels,
the luminescence decay is not a simple exponential in time and hence the
decay time data are not easily compared with the formulas. Applying
equation (3) to the data, one obtains AE as, 0023 ev for CaWO4 and AB 3
.0026 ev for cciwo4 with PT/S' z. 010 and .016 for the two crystals
respecfively.

Assuming the decay time from the trapping level is temperature
independent, an order of magnitude for the decay time from the trapping
level by the non-radiative transition can be estimated to be 2. 50 micro-

seconds.

(1)

(Z)
(3)

(4)

(5)
(6)
(7)
(8)

(9)

15

BIBLIOGRAPHY

J. A. Birks, Scintillation Counters (McGraw-Hill Book Co. , Inc. ,
New York, 1953), ch. 5.

 

G. B. Beard and W. H. Kelly, Nuclear Physics 16, 591 (1960).

W. R. Dixon and J. H. Aitken, Nuclear Instruments and Methods
2, 219 (1960).

F. A. Kroger, Some Aspects of the Luminescence Solids (Elsevier,
New York-Amsterdam-London-Brussels, 1948), ch. 3, 6.

 

R. H. Gillette, Review of Scientific Instruments _2_l, 294 (1950).
F. J. Botden, Philips Research Reports _6_, 425 (1951).
F. Seitz, Transactions of the Faraday Society _3_5_, 79 (1939).

R. W. Gurney and N. F. Mott, Transactions of the Faraday
Society 35, 69 (1939).

C. C. Klick and J. H. Schulman, "Luminescence in Solids, " in
Solid State Physics _5_, edited by F. Seitz and D. Turnbull (Academic
Press Inc., Publishers, New York, 1957), 97.

 

16
APPENDIX I

Time Constant and Decay Time Relations

Figure 5 is the schematic diagram of the preamplifier. Figure 6
is a simplified diagram of the photomultiplier anode circuit and block
diagram of the electronics. Where R is the photomultiplier anode
resistance to ground and C is the distributed capactiance to ground. If

the scintillation decay is assumed to be exponential,

/T

I(t) = A/Te_t (4)

where I(t) is the current, A is a constant, 1' is the decay time, and t

is time. By simple circuit analysis,
I(t) = V/R + CdV/dt (5)

where V is the voltage across the anode resistance. Equating (4) and
(5) and solving for V(t), using the initial condition that V(t) = 0 at t = 0,
equation (6) is obtained.

V(t) = A/c (——R§Cf THe't/RC - e't/T)

(6)
If the anode circuit is made to integrate, RC >> 'r, then

V(t) = A/cu-e't/T) (7)

In the experiment, one measure V(t) C/A as a function of t and is

thus able to obtain 'r.

17

$50.30 uoflwfimgmoum .83023 060330 33>? .m onswwm.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

la -
05mm w - i
2
vamm . _.
.1. Mn N
x. U ..
>803- 06% s
. .4
_O ._._< m
m
H . N
m
_o.
I m
3 .
Yam. o.
28%
1. :

 

18

£33686 Ems: 66:3 36:85-63 2: :3? .3538... .. .964:
.omoomofiflomo :omoom: .Hoflfimzhmoum Hoe/020m opofido 03:3 moflwcmwm : .QEmoHnH: .539?
cofloooop mo Shawna Moog pom. 3.90.30 opoca nofimfladgopoam 05 mo Eduwmwp poflfimcfim .o oudmfm

 

 

wmm

 

 

 

 

 

02,4 _wl<

 

 

 

maOom

 

 

 

 

6.25%

mm?
muses: 2051a

 

 

 

 

 

 

 

 

 

 

 

 

 

 

19

APPENDIX II
Efficiency and Decay Time Relation

The efficiency for luminescence n in the high temperature region
is given by,

n = PL/1P + PQ) (8)

L

where PL is the probability per unit time of a luminescence transition

and P is the probability per unit time for thermal quenching. P and

Q L

P are two competing processes. PL is assumed to be independent of

Q

the temperature and P to be temperature dependent. Gurney and Mott

Q

proposed that P could be represented by Sexp(-EQ/kT). 8 Substituting

Q

this into equation (8), the following results are obtained.
_ -l
n — [1 + s/PL exp(-EQ/kT)] (1)

The decay time T is given by,

1/7 = 1/TL +1/TQ (9)

where TL = l/PL is the decay time for luminescence independent of

temperature and 1' = l/PQ is the decay time for temperature quenching.

Q

Using the assumption of Gurney and Mott, equation (2) is obtained.
(1/7— 1/1-L) =Sexp(-EQ/kT) (2)

In the experiment, TL is assumed to be the decay time in the tempera-

ture region 600K to 2500K.

“11111111“ 11111111) 111111111 Iliiliil i

 

430277