Skip to main content

Full text of "DTIC AD0409481: NON-RADIATIVE ELECTRONIC ENERGY TRANSFER IN ORGANIC LIQUIDS AT HIGH DONOR CONCENTRATIONS"

See other formats


A n Q A A1 r r * * " J 


£ 3 - ¥- ^ 


00 

o 

i— k> 

¥p 


AFCRL-63-610 


NON-RADIATIVE ELECTRONIC ENERGY TRANSFER IN 
ORGANIC LIQUIDS AT HIGH "DONOR* CONCENTRATIONS 


Sanford Lipsky 

Dopartmont of Chemistry, University of Minnesota 

Contract No. AF 19 (604)-8356 
Project No. 6694 
Task No. 669403 

FINAL REPORT 
May, 1963 


AIR FORCE CAMBRIDGE RESEARCH LABORATORIES 
OFFICE OF AEROSPACE RESEARCH 
UNITED STATES AIR FORCE 
BEDFORD, MASSACHUSETTS 


D DC 

JUL12 1963 I 

I 

i 

LsvS'lbll U 

TISIA B 



Notices 


Requests for additional copies by Agencies of the Department of 
Defense, their contractors, and other Government agencies should be 
directed to the: 


Defense Documentation Center (DDC) 

Arlington Hall Station 
Arlington 12, Virginia 

Department of Defense contractors must be established for DDC services 
or have their "need-to-know" certified by the cognisant military agency of 
their project or contract. 

All other persons and organisations should apply to the: 

U. S. Department of Commerce 
Office of Technical Services 
Washington 25, D. C. 

When Government drawings, specifications, or other data are used 
for any purpose other than in connection with a definitely related Government 
procurement operation, the United States Government thereby incurs no 
responsibility nor any obligation whatsoever; and the fact that the Govern* 
ment may have formulated, furnished, or in any way supplied the said 
drawings, specifications, or other data, is not to be regarded by implica* 
tion or otherwise as in any manner licensing the holder or any other person 
or corporation, or conveying any rights or permission to manufacture, 
use, or sell any patented invention that may in any way be related thereto. 

The information furnished herewith is made available for study upon 
the understanding that the Government's proprietary interests in and 
relating thereto shall not be impaired. It is desired that the Staff Judge 
Advocate (ESJP) L. G. Hanscom Field, Bedford, Massachusetts, be 
promptly notified of any apparent conflict between the Government's 
proprietary interests and those of others. 



Table of Contents 


I. Abstract 

II. Scintillation Properties of Liquids 

a) The Dependence of the Scintillation Pulse on Solute 
Concentration 

b) Quenching Effects 

c) Mechanisms of Energy Transport 

d) Relative Scintillation Efficiency 

III. Description of the Fluorimeter 


IV.Personnel and Publications 



I. Abstract 


The efficiency of electronic energy transfer from the liquids benzene, 

toluene, p-xylene and mesitylene to p-terphenyl have been determined 

137 

both for ultraviolet excitation in the first absorption band and for Cs 
gamma excitation. In addition, the efficiency with which oxygen quenches 
the transport process has been determined for both modes of excitation. 
Both energy transfer and oxygen quenching constants are larger than 
would be predicted on the basis of a diffusion limited rate. Dilution of 
the aromatic liquids with hexane is found to significantly reduce the rate 
of energy transfer. In previous reports a similar effect on the quenching 
rate was noted. The participation of the bulk solvent in these processes 
seems to be implicated. Mechanisms of energy transfer utilizing the 
states of the bulk solvent are briefly reviewed. 

The relative scintillation efficiencies have been determined for 
solutions of p-terphenyl in benzene, toluene, p-xylene and mesitylene 
and compared with values approximately calculated from the optical 
transition probabilities and the efficiencies of internal conversion from 
upper electronic states to the transferring state - the first excited singlet. 

The construction of the fluorimeter is described and the emission 
spectrum of pure liquid benzene presented. 



II. Scintillation Properties of Liquids* 

INTRODUCTION 

A fraction of the energy dissipated by a high energy charged particle in its 

passage through a condensed phase may, under proper conditions, ultimately 
-8 -9 

appear (in ca 10 -10 sec) as a scintillation pulse with spectral distribution 

characteristic of the absorbing material. This phenomenon has stimulated con¬ 
siderable research activity in view of its application as a device for detection of 
nuclear particles. However, at the same time, research effort has been directed 
towards elucidating the mechanism responsible for the observed scintillation 
pulse. Both aspects of the field have been the subject of numerous reviews during 
the past ten years (l-5)« 

In pure liquids, the energy efficiency of the conversion is very low. In the 

most favorable cases ( e, g. , p-xylene) the efficiency is less than 0. 1%. In the 

least favorable cases, such as water and the simple aliphatic alcohols, it is a 

2 

factor of 50-100 lower (6-7). For the '•better 11 liquids, the efficiency can be 

improved by almost two orders of magnitude by dissolving in them small quantities 

of p-terphenyl, anthracene, diphenyloxazole or indeed most any substance with a 

high fluorescence efficiency and an absorption band red shifted with respect to the 

-2 

emission band of the solvent (4). A typical case is a 10 M solution of p-terphenyl 

137 

in benzene. If one excites this solution with Cs y -rays, one observes a 
luminescence with a spectral distribution almost completely characteristic of the 

* Supported by Air Force Cambridge Research Laboratories, USAF 
Contract No. AF 19(604)-8356. Presented at International Symposium on 
Physical Processes in Radiation Biology, May 5-8, 1963. 

2 w 
This is after correction for the contribution made by Cerenkov radiation. 



2 


•mission spectrum of p-terphenyl. The energy efficiency of this scintillation 
pulse has not yet been reliably measured. However, in a well degassed solution, 
the efficiency may be estimated to be about equal to that of a thin anthracene 
crystal for which values between 4-6% are currently quoted (8. 9. 10). Reducing 
the terphenyl concentration by a factor of ten, reduces the energy efficiency by 
about a factor of 1.5. Although the efficiency of the scintillation process is not 
high, it is clearly greater than could be expected from the probability of having the 
primary charged particle, or any of the electrons produced in its degradation, 
directly interact with p-terphenyl. Since, then, the initial energy absorption is 
predominately by the solvent, one must postulate a mechanism for the transfer of 
some of this energy from the solvent to the solute. 

In the following sections of this paper we consider and discuss some of the 
evidence bearing both on the mechanism of this energy transfer and on the relation¬ 
ship between scintillation efficiency and nature of the solvent. 

THE DEPENDENCE OF THE SCINTILLATION PULSE ON 
SOLUTE CONCENTRATION 

The energy ultimately transferred to the solvent by a fast electron of initial 
energy E may be considered to lead eventually to the production of a number Mg 
of solvent molecules capable of transferring electronic energy to the solute. The 
number of photons Ng emitted by the solute in such a system may then be written 

N E * <*> 

where refers to the fraction of solvent molecules that transfer their energy 
to the solute and <j>£ refers to the fraction of solute molecules that radiate this 
energy* It is, of course, implicitly ’ assumed here that all molecules are 
equivalent, at least to the extent of possessing the same probability for transferring 
their energy to the solute. This assumption, as will be shown later, seems to be 



3 


valid far the systems we consider, namely, aromatic solvents. 

In most scintillation experiments one normally measures a quantity Ig 
proportioned to Ng. For most experiments the proportionality constant is, unfor¬ 
tunately, not determined. However, for constant experimental conditions and for 
the same solute this "constant" is independent of solvent. Thus a comparison 
may be made of the scintillation efficiencies of a series of solvents all containing 
the same solute. This will be done in the last section of this paper. 

The dependence of Ig on solute concentration is contained implicitly in both 
and A possible method for separation of the observed concentration 
dependence of Ig into each of its two parts was first suggested by Kallmann and 
Furst on the basis of studies made with a large variety of solvent-solute combin¬ 
ations (11). The concentration dependence of ^ obtained in this fashion is identical 
to that obtained when the solute is directly excited optically in its first absorption 
band. It is given by 

4 = -- ( 2 ) 

1 1 + Kc 

where c is solute concentration, K is a constant characteristically known as a 
self-quenching constant and <^° is to be identified, at least in the optical experi¬ 
ments, with the emission quantum yield at infinite dilution. In the cases of 

p-terphenyl and diphenyloxazole, as solutes, K is less than 3 liter/mole. There- 

-2 

fore, for studies involving these solutes at concentrations less than 10 M, <j>^ may 
be regarded as constant and the concentration dependence of obtained directly 
from measurements of Ig. 

The dependence of Ig on solute concentration has been investigated for many 
aromatic hydrocarbon solutions and shown to be consistent with a transfer efficiency, 
proportional to 



1 + ac 


( 3 ) 



4 


where c is solute concentration and a a transfer constant characteristic of the 

particular solute-solvent pair chosen (11, 12). 

The proportionality constant cannot, of course, be specified since Mg is not 

known. The solvents that we have most extensively studied are benzene, toluene, 

p-xylene and mesitylene with p-terphenyl as solute. In all cases, unless otherwise 

specified, the solvents are spectroscopically clean and carefully degassed. In 

Ta’-jJe I are listed the transfer constants recently obtained for p-terphenyl in these 
137 

solvate using Cs excitation (13). Values of the same magnitude (i. e. , 1500- 
1800 llter/mole) have been reported for diphenyloxazole in toluene and p-xylene 
solutions (15, 16). 

Similar information pertinent to the mechanism of energy transport can be 
determined by studying the time dependence of the scintillation pulse (171 Making 
the same assumptions required for the validity of equation (1), we would expect 
that the decay of the pulse should be determined by two processes, the rate of 
energy transfer to solute and the rate at which the solute emits. Simple analysis 
of this leads to a decay law involving the sum of two exponentials. For the case of 
aromatic hydrocarbon solutions, this is experimentally verified at least within the 
precision of present day apparatus (18). However, again one must somehow 
determine which decay constant belongs to which physical process. This problem 
has been solved using methods analogous to those used in the similar probelm that 
occurred in intensity measurements. For the case of p-terphenyl in benzene, the 

_9 

solute emission decay time is identified as 2. 5 x 10 s (18) and the decay timfe 
'C corresponding to the lifetime of the transferring solvent species can be 
expressed as 

t 0 

r = —2— (4) 

1 + o c 

where a has the same significance and experimentally the same magnitude as 
before (15-20). The quantity in equation (4) is the decay time extrapolated to 



5 


zera solute concentration and has been experimentally identified reasonably well 
with the decay time of the emission from the pure solvent (16). 

It is also possible to sensitize a solute emission by exciting the solvent 
optically in its first absorption band. Furthermore, a comparison of the intensity 
of the sensitized emission to the intensity obtained by directly exciting the solute 
in its own first absorption band permits one to obtain a value for the transfer 
efficiency The dependence of this efficiency on solute concentration has been 
determined for a number of aromatic systems and shown to be consistent with the 
form: 


ac 

<{, ~ - 

1 + a c 


(5) 


where a is the transfer constant and c the solute concentration (13, 21,22), Within 

137 

experimental error, a is identical to the value obtained using Cs as an 
excitation source (see Table I.) It is important to note that equation (5) predicts 
unit transfer efficiency at infinite solute concentration. Consequently there must 
be no 11 instantaneous 11 quenching of these aromatic molecules in their first 
electronic singlet states, at least not at the low excitation densities employed in 
the usual optical experiments. Electronic energy transfer to p-terphenyl is, 
therefore, competitive with all other possible fates for the excited solvent. 

For optical excitation throughout the first solvent absorption band, equation 
(5) adequately describes the concentration dependence of with a retaining a 
constant value. For excitation in the second and third optical absorption bands, 
the transfer efficiency is reduced and its functional dependence on solute concen¬ 
tration altered to the form 


4 > = ( 6 ) 

t 1 + a c 

where p is a function only of exciting wavelength with the value 1, of course, for 
any wavelength within the first band and ^ 1 at all other wavelengths (13). 

Evidence for the form of equation (6) derives from the fact that l/^> fc depends 



6 


linearly on 1/c with Intercept and slope strongly dependent on exciting wavelength 

3 

but ratio of Intercept to slope, a, approximately constant at the value obtained for 

been 

first band excitation. The emission yields of the pure solvents have also/studled as 
a function of exciting wavelength and found to decrease with decreasing wavelength 
(23). Indeed, the solvent emission yield when normalized to unity for excitation In 
the first band, reproduces almost exactly the values, {3, obtained In the sensitized 
emission experiments. The following interpretation is suggested. Most, if not all, 
of the sensitization occurs via the lowest excited singlet state of the solvent. 
Furthermore, the transfer efficiency, (j> ( , in equation (6) is a product of two 
efficiencies, p - an internal conversion efficiency from an upper electronic state 
of the solvent to the sensitizing state, and y - m an e ^^ c ^ enc y energy trans¬ 
fer to p-terphenyl from this latter state. 

In view of the demonstrated invarience of a to mode of excitation (see Table 

I) and the absence of evidence for any appreciable transfer from upper electronic 
4 

states, we shall assume that the same mechanism of energy transport obtains in 

137 

both the scintillation process (induced by Cs y -rays) and the process induced 

by optical absorption in the first solvent band. 

OUENCHING EFFECTS 

Oxygen, bromobenzene or carbon tetrachloride, when present in small 
amounts in these two-component aromatic systems, quench the luminescence. 

In general, most of this quenching cannot be attributed to an interaction between 
the excited solute and the quencher. The major contribution seems to derive from 

3 

There seems to be some tendency for a to decrease with decreasing wave¬ 
length. For example in the case of benzene - p-terphenyl solutions a decreases 

o o 

from 1350 if m at 2600A to about 1100 i/m at 1610A. However, for the same wave¬ 
length variation, p changes from 1. 0 to 0. 3. 

4 

Measurements have only been reliably extended to the short wavelength side 

o 

of the third absorption band of the alkylbenzenes (i. e. # 1610A) (23). 



7 


the interference of the quencher with the rate of energy transfer to the solute (12, 
14, 24, 25). Quencher and solute act competitively for solvent energy and with 
comparable efficiencies. In the case of bromobenzene quenching of the scintillation 
process, Ng may be expressed as 


N E * M eV*£ 


1 1 + ac + yd 

where a, and c are as previously definedi d is the concentration of quencher and 

y is a quenching constant independent of ooth solute and solvent concentration. 

In the case of optical excitation, a similar relationship between transfer 

efficiency 4>. f and bromobenzene concentration is obtained with the same value for 

137 

a but a somewhat lower value for y (12, 24). Using Cs excitation 

y s= 500 ^ 60 i/m for bromobenzene quenching of p-terphenyl-benzene solutions. 

o 

Using 2137A excitation y = 350 ^ 120 i/m for the same system (24). A similar 

disparity has been noted for carbon tetrachloride quenching of p-terphenyl - 

toluene solutions (14) with again the higher quenching efficiency for the scintillation 

process. In the case of oxygen, the luminescence from degassed and air-equili- 

-3 

brated solutions of 2 x 10 M p-terphenyl in benzene, toluene, p-xylene and 

mesitylene have been compared (13)* This has been done both for UV excitation 

137 

in the first band of the solvent and for Cs excitation. The results are 


presented in Table II. The values for y have been corrected for the measured 
effect of C >2 on the fluorescnece yield of terphenyl (<p^). This contributes uniformly 
in all solvents about 10% to the reduced intensity. With this correction, Y was 
then calculated assuming the validity of equation (8). This, however, has not yet 
been reliably demonstrated for oxygen quenching. 

In view of our assumption that the energy transport process is independent 
of mode of excitation, differences in quenching efficiency must seek explanation in 



8 


5 

processes of the type: 

C C* 

C* + D —* C + D 1_ 

c C + + i" JS_ 

C + + l" C* 9 

Me. 

C + + i" + D -* C + D 10_ 

where C, C* and D refer respectively to solvent in gcund state, solvent in first 
excited state (transferring state), and quencher. It is suggested, therefore, that 
in the scintillation process the quencher may interfere not only by interaction 
(process 7) with the transferring state of the solvent (which it may do also in the 
optical process) out, in addition, by partially inhibiting the formation of such 
transferring states (process 10), 

MECHANISMS OF ENERGY TRANSPORT 

The fluorescence quantum yields of the pure liquid alkyl benzenes are not 
known accurately but are probably less than 0, 10. Values of 0, 01 and 0. 03 have 
been reported for toluene and o-xylene (26) and from relative measurements that 

we have made, even in the most favorable case of p-xylene the yield should not 

.3 

exceed 0.05. However, the energy transfer efficiencies at 2 x 10 M p-terphenyl 

are 0.72 (for toluene), 0.81 for p-xylene and extrapolate to unity at infinite 

terphenyl concentration. Clearly, the mechanism for energy transfer cannot 

involve radiative processes to any important extent for terphenyl concentrations 
-4 

above about 10 M, Further evidence for this point of view is obtained from the 
observation that the decay time of solvent emission is reduced by the addition of 

Not shown in this scheme is the transport process whereby C* transfers 
energy ultimately to the solute. 



9 


solute to the same extent as the intensity of solute emission is Increased (18,20). 
The transfer process we are considering, therefore, is predominantly noaradiative. 
One possible mechanism for such a process in a liquid is as follows. By optical 
absorption or ultimately from the energy dissipated by a fast electron, a specific 
solvent molecule is electronically excited. The molecule diffuses to within some 
distance R of the solute and then in a process which may involve perhaps an 
exchange interaction or some interaction of longer range, transfers energy to the 
solute. In terms of a conventional kinetic sequence we could write: 

C C* 0 


c* - 

—> 

C 


1 

C * +T - 

-* 

c + 

T * 

3 

T* - 


T 


4 

T* - 

-> 

T + 

hv 

5 


where C and T refer to solvent and solute respectively. With this scheme it is 
simple to show that the transfer efficiency, the transfer decay time and the 
quenching efficiency (if reaction 7 is involved) will all have the proper dependence 
on solute concentration. Furthermore, a, y and ^ (see equation 4) can now be 
identified as 


a » k^/kj 

y a k ? /kj (9) 

r 0 ~ 


To go one step further, if this model is correct, we should be able to estimate a 

lower limit for the encounter distance R using measured values of a, and and 

the appropriate solution to the diffusion equation (27). For values of 2"^ of the 
“8 

order of 10 sec, and reaction probability not too small at R nor diffusive dis¬ 
placements too large compared to R, ^ the transient term appearing in the solution 


^All of these conditions are reasonable for our problem. 



10 


to the diffusion equation may be neglected. Thus one obtains for the specific rate 
of reaction, k, 


4irNDR 

k ■ Toocr 


( 10 ) 


where N is Avogadro's number and D the mutual diffusion coefficient. The self 

diffusion constant of benzene and the diffusion constant of oxygen have been 

-52 -52 

determined as 2 # 15 x 10 cm /sec (28) and 5. 7 x 10 cm /sec (29) respectively. 

In the absence of similar information for the other solvents, a value of . 

D * 4. 30 x 10‘ 5 cm^/sec was used in all cases for calculation of the encounter 

-5 2 

distance for the enerby transfer reaction (R^) and a value of D = 7. 8 x 10 cm /sec 
used for the calculation of the encounter distance (R^ ) for the quenching reaction. 

In Table III, experimental values of (16) and calculated values of R q and 

R^ are presented. It will be noted that for both reactions, the values for R are 
somewhat larger than would be expected for geometrical encounter distances (29). 

It is perhaps not unexpected that the energy transfer process would have associated 
with it a larger cross-section. However, it is much more difficult to explain a 
similarly high value for oxygen quenching since 1 ‘resonance* 1 processes presum¬ 
ably cannot occur in this case. 

Attempts to determine experimentally the role of diffusion in these processes 

have taken two directions - one, a study of the effect of viscosity, the other a study 

of related processes in rigid plastic solutions. In the case of viscosity variation, 

there is evidence that the transfer efficiency is lowered as the viscosity increases 

.3 

(15, 31). For example, the transfer efficiency is 0.48 for a solution of 2.26 x 10 
M diphenyloxazole containing ten parts paraffin oil to one part toluene (viscosity 
10 cp) whereas a value of 0. 78 has been obtained for the same solution with 
cyclohexane used in place of the paraffin oil (viscosity 1 cp). It should be noted 
that in the absence of diphenyloxazole the decay time of toluene remained the same 
in the two systems. Thus diffusional processes seem to be implicated. However, 
the variation of transfer efficiency with change in viscosity is somewhat smaller 



11 


than one would predict for diffusion controlled processes. Furthermore, one 
obtains at viscosities at which all diffusional processes should be stopped, transfer 
efficiencies much higher than can oe accountef for by radiative processes (31). 

Measurements made in rigid plastic systems are also difficult to interpret 
in terms of their pertinence to the question of diffusional processes in liquid 
systems. In addition to this, there are experimented disagreements. Knau (32) 
finds the transfer process from benzene to anthracene to be similar to that from 
polystyrene to anthracene. A similar aosence of effect of solidification on transfer 
efficiency was reported earlier by Hardwick for transfer from naphthalene to 
anthracene (33). However, measurements from many other laooratories seem to 
indicate much lower transfer efficiencies in rigid plastics at solute concentrations 
comparable to those used in the liquid (19,22,34-37). Since in both plastic and 
liquid systems, ^ seems to exhibit the same type of dependence on solute concen¬ 
tration (see equation (3)), the difference in behavior can be expressed more 

quantitatively in terms of transfer constants. Comparing oenzene-> terphenyl 

transfer with polystyrene -* terphenyl transfer, the ratio a (solution)/a 

(plastic) » 30 (19,35). 

A long range "resonance" energy transfer of the Forster type (38) from some 

one initially excited solvent molecule to the solute can be shown, in the absence of 

diffusion, to be too inefficient to explain th^ observed values of (jy In the case of 

o 

transfer from p-xylene to p-terphenyl a value for Rq of about 20A can be calcu- 

7 

lated (for random orientation) assuming a p-xylene emission quantum yield of 
0.05. By suitably averaging the pair wise transfer probability over a random 
distribution of solute molecules about the excited energy donor molecule, Forster 

7 

Rq is the distance between excited p-xylene and p-terphenyl for which the 
transfer probability is unity within the lifetime of the p-xylene ( ^ = 20 n sec.). 



12 


has derived an expression for the transfer efficiency 

x 2 _ 

\ m T xe* 74 * [1 - erf(->~- x)] (11) 

-3 3 

where x = 2. 51 x 10 c Rq with c in mole/liter and Rq in Angstroms, and 

IT ^ 

erf (—x) is the error function (39). For Rq * 20A, and a solute concentration 

_2 

c * 10 M, equation (11) predicts <(> t = 0.26. Experimentally one obtains 
^ = 0.96. Similar discrepancies exist for the other solvents. 

In both previous models, material diffusion and long range transfer, 
electronic states of the bulk solvent are assumed not to participate directly in the 
transfer process. Alternatively we may consider a model in which energy moves 
from its initial absorption site to the solute (or quencher) exclusively via excited 
states of the bulk solvent. Two limiting cases (weak and strong coupling) are 
usually considered theoretically (40,41). Their distinction depends, approximately, 
upon the relative strengths of two electrostatic interactions - one, an interaction 
between an electroncally excited configuration of one solvent molecule "A" with 
a ground state configuration of a neighboring solvent molecule, the other an inter¬ 
action between electronic and nuclear motions within the molecdle "A". For the 
case in which the intermolecular coupling is relatively weak, an initial excitation 
of the solvent via photon or electron impact leads to an excited electronic state 
which may be considered as localized on a single solvent molecule. The subsequent 
motion of this excitation is diffusive with a diffusion length of the order of a 
molecular diameter. For energy transport in the alkyl benzenes, the weak coupling 
approximation predicts transfer efficiencies at least an order of magnitude too 
small (12,13). This conclusion is oased on a calculation of tie excitation diffusion 
coefficient using the procedure suggested by Forster (39). 

In the intermediate cases of relatively stronger intermolecular coupling, the 
initially excited state may be considered as delocalized over a region of the solvent 



13 


a 

large compared to the dimension of/single molecule. The description of the initial 
state is now best presented as a wave packet (or exciton state) and it is in terms of 
the motion of this packet that the energy ‘transport process is to t>e envisioned. In 
a liquid. Interactions of the exciton with nuclear motions tend to deflect it from the 
more rectilinear path characteristic of the stronger coupling cases that obtain in a 
low temperature solid. In the absence of any correlation between successive 
scattering events, the motion of the exciton may again be considered as diffusive 
but now with a diffusion length larger than a molecular diameter. Similar con¬ 
siderations have been applied to the description of energy transport processes in 
molecular crystals (42, 43). 

The efficiency of optically initiated energy transfer from toluene to anthracene 
has been studied as a function of degree of dilution of toluene in cyclohexane (21, 
44). The concentration of solute was kept constant. No variation in transfer 
efficiency was found for experiments carried to dilutions of 1:80 in degassed 
solutions and 1:400 in aerated solutions. The conclusion drawn was that transport 
via the bulk solvent need not oe invoked to explain results in the pure liquid. 

We have recently performed similar experiments involving p-xylene with 
hexane as diluent and anthracene as solute (13). The results are shown in Table IV 
together with recently reported values of the emission decay times of p-xylene in 
hexane solutions (30). About an 8% decrease in <j>^ may be noted over the entire 
range. This corresponds to a decrease of about 17% in the kinetically more 
significant parameter a. However, since the time, Z"q* available for transfer 
increases with dilution, there must be a decrease in rate of transfer (k^) by a 
factor of about 1.7. If transfer were exclusively by material diffusion, both the 
increase in ^ and the decrease in viscosity® on dilution would have predicted 
an increase in a about a factor of 3. Similar effects have been noted for p-xylene 
and benzene transfer to p-terphenyl (13). 

® *7 (p-xylene)/ ^ (hexane) = 2 



14 


It appears, therefore, that in the pure liquid, processes other than material 
diffusion and long range transfer contribute to the observed energy transport* How 
much is contributed cannot be determined at the present time* However, regard¬ 
less of the exact magnitude of this contribution by the bulk solvent, it cannot be 
derived from the concept of excitations initially localised on single molecular 
sites* 


RELATIVE SCINTILLATION EFFICIENCY 

For high speed electrons or gamma rays, the energy efficiencies of the 

-2 

aromatic solutions at about 10 M p-terphenyl concentration are probably between 

4-6%• Since energy transfer efficiencies are essentially unity at high p-terphenyl 

o 

concentrations and the terphenyl emission quantum yield is about 0, 5 (4)* these 
efficiencies are equivalent to the requirement of about 40 ev for the production of 
one excited solvent state capable of transferring energy* 

Values for W (energy/ion pr) are not known for the aromatic liquids* For 
molecular gases, W s' 25 ev (45)* Assuming this value applies to the liquid, and 
further assuming aa aromatic average ionization potential ct9ev and an aromatic 
average excitation potential — 7 ev, one derives that about 3« 5 primary excitations 
are produced per 40 ev absorbed* Therefore, approximately 70% of the primary 
excitations do not ultimately appear as states of the solvent capable of transferring 
energy* Discussions of the possible source for this inefficiency have appeared 
repeatedly in the literature (1,2,4,46), and gnerally involve postulating some form 
of rapid quenching reactions involving excited solvent states in regions of high 
LET. 

9 

The literature value refers to a p-terphenyl crystal* However, 0. 5 is also 
obtained from the measured emission decay time of p-terphenyl in solution and its 
integrated absorption spectrum (13). 



15 


Internal conversion efficiencies (P) from the second and third electronic 
states of pure alkyl benzene liquids to their lowest excited singlet states have 
recently been determined (23), It was decided, therefore, to investigate the role 
this might play in explaining relative scintillation efficiencies in these liquids. 

The dependence of Ig (equation (1)) on p-terphenyl concentration was determined 

1 ^7 

for Cs excitation* By extrapolating to infinite concentration! relative values 
of were determined (13)* Relative values of ^ were obtained optically* The 

results are presented in columns 2 and 3 of Table V* In column 4 are presented 
relative values, f, of the integrated optical transition probabilities (f^ £(v) dv)« 11 
These numbers are presumed to be approximately related to the total cross-section 
for excitations (47), In column 5 are presented relative values g of the integrated 
optical transition probabilities weighted by the measured internal conversion 
efficiencies to the first excited singlet (g £(y) gfv) dv). ** These numbers are 
presumed to be approximately related to the probability o£ producing first excited 
states via the electron impact induced excitations and are therefore to be compared 
with their experimental counterpart 

It will be noted that both the relative ordering of M E as well as the gap * 
between the value of for benzene and values for the other solvents are predicted 
by g* These trends are not apparent in the optical transition probabilities. How¬ 
ever, the spread in the experimentally determined values for is much less than 
predicted by g. Assuming there is some validity to these calculations, one possible 
explanation is that many of the transferring states arise from processes not 
involving the conversion of excitations. Recombination of electrons with positive 

ions might provide such a mechanism for the direct production of a transferring 
state and has already been suggested by the disparity between scintillation and 


^Corrected for differences in gamma-ray attenuation* 

**A11 integrals were terminated at about 7* 5 ev* Values for p are not yet 
determined above this energy. Values for the extinction coefficients £(2) were 
determined from dilute solution absorption spectra* 



16. 

optically induced quenching processes. 

III. Description of the Fluorimeter 

A schematic of the fluorimeter is shown in Figure 1. The source of 
illumination is a 900 watt Xe arc lamp. A. The light is collected by lens 
system B and made incident on a Bausch and Lomb 250 mm grating mono¬ 
chromator. The monochromator is normally operated at a spectral half 

0 

width of about 60-100A. The subsequent optical system comprised of lenses 
B and C and front surface mirrors D and E direct the monochromatic 
beam onto a 1 cm cylindrical quartz cell F. A concave front surface mirror 
G is used to collect the emission and focus it onto the entrance slit of a 
Beckman DU monochromator. Although not shown in Figure 1. the optical 
system is symmetrical about an axis passing through the entrance slit of the 
DU and the exit slit of the Bausch and Lomb monochromator. Thus a cell F 1 
not shown but placed symmetrically with respect to F may alternatively be 
illuminated via D 1 and E 1 (not shown). By rotating the platform containing 
D, D 1 and G. either cell F or F* may be illuminated and its emission 
collected and analyzed. After dispersion by the DU the fluorescence is 
made incident on the cathode of a 6256B EMI photomultiplier K, via a front 
surface mirror H and lens J. The output of the photomultiplier may either 
be monitored by a DC tip ammeter or alternatively, pulse amplified and dis¬ 
played on a scaler. In either mode of operation the photomultiplier is 
normally cooled to dry ice temperatures. The resulting noise level at a 
tube gain of about 10^ is of the order of 3 counts/min. 

All solutions are carefully degassed by successive freeze-thaw cycles 
on a high vacuum manifold and then connected to the fluorescence cell via a 
break seal. After evacuation of the fluorescence cell to ca. 10*^ mm, the 
first solution is admitted to the cell and retained therein by means of a 
solenoid operated ground glass float valve. At the conclusion of an experi- 



17 


ment this valve is opened and the solution drained into a reservoir. This 
procedure is similar to that previously employed and reported in an earlier 
publication (24). 

The equipment is sufficiently sensitive to measure accurately the 
emission from pure liquid benzene with quantum yield estimated at about 

O 

0.01. The spectral distribution of this emission excited at 2500A for a 
1 M benzene solution in cyclohexane and pure liquid benzene is shown in the 
upper and lower halves of Figure 2 respectively. The appearance of a 
long wavelength component in the emission apparent in the pure liquid 
spectrum has been previously reported (30). Its assignment is uncertain 
although an eximer origin has been suggested. Measurements to determine 
whether a difference exists in the quenching efficiency of the long and 
short wavelength components of the emission are currently being made. 



18 


IV. Personnel and Publications 

Sanford Lipsky, Assistant Professor 
William P. Helm an. Research Assistant 

a) S. Lipsky, W. P. Helman and J. F. Merklin, Quenching of Electronic 
Energy Transfer in Organic Liquids, In Luminescence of Organic 

and Inorganic Materials, p. 83-99 (H. P. Kallmann and G. M. Spruch, 
Editors) John Wiley and Sons, New York, 1962. 

b) S. Lipsky, Evidence for Triplet-Triplet Transfer from Benzene to 
Biacetyl in Cyclohexane Solution. J. Chem. Fhys. (accepted for 
publication, June, 1963). 

c) S. Lipsky, Scintillation Properties of Liquids, presented at 
International Symposium on Physical Processes in Radiation Biology 
and to be published as supplement in Advances in Radiation Biology, 
1963. 

d) S. Lipsky, Propagation of a One-Dimensional Exciton, Quarterly 
Status Report No. 6 to Air Force Command and Control Development 
Division, 1962 (to be submitted for publication). 



TABLE I 


Values of the Transfer Constant, o, 

£1 

in Units of liter/mole for Degassed p-Terphenyl Solutions 


Solvent 

Cs*^ Excitation 

UV Excitation 
in first solvent 
absorption band 

Benzene 

1400 + 75 

1350 + 75 

Toluene 

1250 + 75 b 

1350 + 75° 

p-Xylene 

2170 + 75 

2150 + 75 

Mesitylene 

580 + 15 

560 + 15 


Measurements were made at terphenyl concentrations ranging from 
0.5 x 10 -3 to 2 x 10" 3 M. 

b A value of 1290 i/mole can be obtained from data reported by Kilin, Kovyrzina 
and Rozman (14). 






TABLE II 


Values of the Quenching Constant, y , 
in Units of liter/mole for p-Terphenyl Solutions 4 


Sol vent 

1° 2 1 TSTr < 13 » 

137 

Cs Excitation 

UV Excitation 
in first solvent 
aosorpt ion band 

Benzene 

1.38 +.04 X 10" 3 

1800+ 150 

2000+ 150 

Toluene 

1.52 + .05 x 10" 3 

1650+ 100 

1200+ 100 

p-Xyl ene 

1. 32 + .10 x 10" 3 

2900 + 200 


Mesityl ene 

1. 13 + . 07 x 10" 3 

1150 + 100 

900+ 100 


Determined from measurements at 2 x 10 


-3 


M p-terphenyl. 





TABLE III 


Specific Rates and Diffusion Encounter Distances for Reactions 
Involving Energy Transfer to p-Terphenyl and Quenching by Oxygen 


Solvent a 

t Q D (n sec)(l 6) 

c 

kj (i/mole sec) 

R a (A) 

k^^(f/mole sec) R^ (A) 

Benzene 

17.7 

7. 8 x 10 10 

24 

11 x 10 10 19 

Toluene 

19.0 

6. 8 x 10 10 

21 

6. 3 x 10 10 11 

p-Xylene 

20,2 

H-* 

X 

w 

o 

o 

34 

14 x 10 10 24 


a There is no literature value currently available for the mesitylene decay time. 

^Values of 22 ns. 23ns and 20 ns have recently been reported for benzene, toluene 
and p-xylene respectively (30). 

c Values of a from columns 2 and 3 of Table I were averaged to calculate k^. 

^Value of y taken from column 4 of Table II, for benzene and toluene and 
from column 3 of Table II for p-xylene. 





TABLE IV 

Effect of Dilution on Transfer Efficiency 

_3 

from p-Xyleae to Anthracene (1.12 x 10 M). 


[p>xylene] (mole/liter) 

n 

a (liter/mole) 

tr Q (n sec) (30) 

8. 1 

0.63 

1500 

20 

0.81 

0. 62 

1450 

26 

0. 081 

0. 60 

1350 

28 


0.0081 


0. 58 


1250 


28 





TABLE V 


Comparison of Scintillation Efficiency 
and Internal Conversion Efficiency in Alkyl Benzene Liquids 


Solvent 

|>£ (relative) 

m e 

f 

g 

Benzene 

1.00 

1.00 

1.00 

1.00 

Toluene 


1. 36 a 

1.46 

1.83 

p-Xylene 

0.97 

1.48 

1. 58 

2. 69 

Mesitylene 

0. 94 

1.55 

1.29 

3.12 


& <)>f for toluene assumed to Oe 0, 98. 





Figure 1. Fluor ime ter Schematic 

Figure 2. Emission Spectrum of Liquid Benzene 




Figure I 









5 


I M benzene 



Figure 2 


REFERENCES 


I. J. B. Birks, Scintillation Counters , Pergamon Press, New York, 1953, 

2* S* C. Curran, Luminescence and the Scintillation Counter, Butterworths 

Scientific Publications, 1953. 

3. R, K. Swank, Characteristics of Scintillators. In Ann. Rev. Nucl, Sci. 4, 

111 (1954). 

4. F. D. Brooks, Organic Scintillators. In Progr. Nucl. Phys, 5, 252 (1956). 

5. C. G. Bell, Jr. and F. N. Hayes, Editors, Liquid Scintillation Counting , 

Pergamon Press, New York, 1958. 

6. L. I. Al'perovich, I. D. Sherbaf and R. Marupov, On the causes of 

luminescence of liquids under the action of hard radiation. Optics and 
Spectroscopy (Engl, transl.) 8, 132 (I960). 

7. K. A. Kovyrzina and I. M. Rozman, On the luminescence of certain solvents. 

Optics and Spectroscopy (Engl, transl.) 12, 133 (1962). 

8. R. W. Pringle, L. D. Black, B. L. tunt and S. Sobering, A new quenching 

effect in liquid scintillators. Phys. Rev. 92, 1582 (1953). 

9. G. T. Vvright, Absolute quantum efficiency of photofluorescence of anthracene 

crystals. Proc. Phys. Soc. B 68, 241 (1955). 

10. Z. A, Chizhikova, Luminescence and energy yield for y scintillation.in a 

stilbene crystal. Optics and Spectroscopy, (Engl, transl.) 7, 176 (1959). 

II. H. Kallmann and M. Furst, Fluorescence of solutions bombarded with high 

energy radiation (Energy transport in liquids) II. Phys. Rev. 81, 853 (1951). 

12. S. Lipsky and M. Burton, Comparison of high-energy and ultraviolet radiation 

induced luminescence in liquid systems. J. Chem. Phys. 31, 1221 (1959). 

13. C. Braun, S. Kato and S. Lipsky, The dependence on exciting wavelength of 

electronic energy transfer from solvent to solute in liquid solutions, to be 
published. 


14. S. F. Kllin, K. A. Kovyrzina and I. M. R os man. Luminescence of p-terphenyl 

in a mixture of toluene and carbon tetrachloride. Optics and Spectroscopy , 

11, 209 (1961). 

15. A. Weinreb, Effects of Temperature and Viscosity on Scintillation Decay 

Times of Solutions. In Luminescence of Organic and Inorganic Materials , 
p. 44 (H. P. Kallmann and G. M. Spruch, Editors) John Wiley and Sons, 

New York. 1962. 

16. I. B. Berlman, Ultraviolet and Lifetime Studies of Mechanisms in the 

Scintillation Process. In Luminescence of Organic and Inorganic Materials , 
p, 62. (H. P. Kallmann and G. M. Spruch, Editors) John Wiley and Sons, 

New York. 1962. 

17. H. Kallmann and G. J. Brucker, Decay times of fluorescent substances 

excited by high energy radiation. Phys. Rev. 108, 1122 (1957). 

18. M. Burton and H. Dreeskamp, Energy transfer and decay times in radiation- 

induced luminescence of oenzene solutions, Z. Elektrochem. 64, 165 (1960); 
M. Burton, A. Ghosh and J. Yguerabide, Decay times of high-energy induced 
luminescence. Rad. Res. Suppl. 2, 462 (1960). 

19. R. K. Swank, H. B. Phillips, W, C. Buck and L. J. Basile, Decay Times of 

Scintillators, IRE Trans. Nuclear Sci. NS-5, 183 (1958). 

20. T. B. Berlman, Efficiency of energy transfer in a solution of PPO in xylene. 

J. Ghem. Phys. 33, 1124 (I960). 

21. S. G. Cohen and A. Weinreb, Energy transfer from solvent to solute in 

liquid organic solutions under ultra-violet excitation. Proc. Phys. Soc. 
(London) B72, 53 (1958). 

22. F. H. Brown, M. Furst and H. Kallmann, Light and high energy induced 

energy transfer in liquid and rigid organic scintillators. Disc. Far. Soc. 

27, 43 (1959). 




23. S. Kato, S. Lipsky and C. L. Braun, Variation with exciting wavelengths ol 

the fluorescence efficiencies of some alkyl benzenes. J« Chem. Phys. 37, 

190 (1962). 

24. S. Lipsky, W. P. Helman and J. F. Merklin, Quenching of Electronic Energy 

Transfer in Organic Liquids. In Luminescence of Organic and Inorganic 
Materials, p« 83. (H. P. Kallmann and G. M. Spruch, Editors) John Wiley 
and Sons, New York. 1962. 

25. Ft H. Brown, M. Furst and H. P. Kallmann, A Summary of Quenching Studies 

of Energy Transfer in Organic Systems. In i -iuminescence of Organic and 
Inorganic Materials, p. 100. (H. P. Kallmann and G. M. Spruch, Editors) 

John Wiley and Sons, New York. 1962. 

26. T. P. Bellikova and M. O. Galanin, Photoluminescence with a solvent. Optics 

and Spectroscopy 1, 168 (1956); C. A. 50, I6403g (1956). 

27. F. C. Collins and G. E, Kimball, Diffusion controlled reaction rates. 

J. Colloid Sci. 4, 425 (1949). 

28. K. Graupner and E. R. S. Winter, Some measurements of the self-diffusion 

coefficients of liquids. J. Chem. Soc. 1145 (1952). 

29. W. R. iVare, Oxygen quenching of fluorescence in solutions. An experimental 

study of the diffusion process. J. Phys. Chem. 66, 455 (1962). 

30. T. V. Ivanova, G. A. Mokeeva and B. Ya Sveshnikov, Dependence of the 

fluorescence of benzene, toluene and p-xylene solutions on the concentration 
of fluorescent substance* Optics and Spectroscopy (Engl, transl.) 12, 325 
( 1962 ). 

31. A. Weinreo, Some effects of temperature and viscosity on fluorescence and 

energy transfer in solutions. J. Chem. Phys. 35, 91 (1961). 

32. H. Knau, Energieubertragung bei der fluoreszenz organischer losungen, 

Z. Naturforsciu 12A, 881 (1957). 




33. R. Hardwick, Role of collision transfer in fluorescent solutions. J. Chem. 

Phys, 26, 323 (1957). 

34. R. K. Swank and W. L. Buck, The scintillation process in plastic solid 

solutions. Phys. Rev. 91, 927 (1953). 

35. T. P. Bellikova aud M, P. Galanin, Mechanism of energy transfer in plastic 

scintillators. Izvest. Akad. Nauk. a. S. 5. R. (Engl, trans.) 22, 46 (1958). 

36. I. M. Rozman, Temperature dependence of the luminescence efficiency of 

polystyrene plastic scintillators. Izvest. Akad. Nauk. S. S. S. R. (Engl, 
transl.) 22, 48 (1958). 

37. L. J. Basile and A, Weinreb, Transfer of excitation energy in solid solutions 

of anthracene-polystyrene and 9-10-diphenylanthracene - polystyrene, 

J, Chem, Phys. 33, 1028 (I960). 

38. Th. Forster, Zwischenmolekulare energiewanderung und fluoreszenz, Ann. 

Physik. 2, 55 (1947). 

39. Th. Forster, Experimentelle und theoretische untersuchung des zwiscben- 

molekularen uoergangs von elektronenanregungsenergie. Z. Naturforsch, 

4a, 321 (1949). 

40. J. Franck and E. Teller, Migration and photochemical action of excitation 

energy in crystals. J. Chem. Phys. 6, 861 (1938). 

41. A. S. Davydov, Theory of Molecular Excitons (Engl, transl. by M. Kasha and 

M. Oppenheimer, Jr.) McGraw-Hill, New York. 1962. 

42. D. C. Northrop and O. Simpson, Electronic Properties of Aromatic 

Hydrocarbons. II. Fluorescence transfer in solid solutions. Proc. Roy. 
London A234, 136 (1956). 

43. I. Ya. Kucherov and A. N, Faidysh, Migration and transfer of electronic 

energy in anthracene and naphthacene crystals. Izvest. Akad. Nauk. 

S. S. S. R. (Engl, transl.) 22, 27 (1958). 




44. V. Bar and A, Weinreb, Transfer efficiency and fluorescence.of solutions, 

J, Chem. Phys, 29, 1412 (1958). 

45. R. L, Platzman, Ionization in Gases by High-Energy Particles, Chapter 5 in 

Penetration of Charged Particles in Matter. Nuclear Science Series Report 
No. 29, publication 752. National Academy of Sciences - National Research 
Council. Washington, O. C., (I960). 

46. I. M. Rozman, The mechanism of scintillations in organic substances. Optics 

and Spectroscopy (Engl, transl. ) 8, 277 (I960). 

47. H. St W. Massey, Electromscattering in solids. Advances in Electronics 4, 

1 (1952). 



UNCLASSIFIED UNCLASSIFIED