Citation
Investigation of the external heavy atom effect as a means of improving the sensitivity and selectivity of analytical phosphorimetry

Material Information

Title:
Investigation of the external heavy atom effect as a means of improving the sensitivity and selectivity of analytical phosphorimetry
Creator:
Hood, Lyal Van Sant, 1942-
Publisher:
[s.n.]
Publication Date:
Language:
English
Physical Description:
viii, 99 leaves. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Phosphorimetry ( lcsh )
Chemistry thesis Ph. D ( lcsh )
Dissertations, Academic -- Chemistry -- UF ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis - University of Florida.
Bibliography:
Bibliography: leaves 95-98.
General Note:
Manuscript copy.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Lyal Van Hood. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
021455709 ( ALEPH )
18066571 ( OCLC )

Downloads

This item has the following downloads:


Full Text
INVESTIGATION OF THE EXTERNAL HEAVY ATOM EFFECT AS A MEANS OF IMPROVING
THE SENSITIVITY AND SELECTIVITY OF
ANALYTICAL PHOSPHORIMETRY
By
LYAL VAN SANT HOOD
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA 1968




ACKNOWLEDGMENTS
The author will always be indebted to his research director, Dr. J. D. Winefordner, for his enthusiastic encouragement and guidance during the course of this work. Appreciation is also extended to the other members of his committee, Dr. R. B. Bennett, Dr. G. M. Schmid, Dr. S. P. Cram, and Dr. W. S. Brey.
The friendly assistance and good fellowship of his colleagues during four years of graduate study are also gratefully acknowledged.
Finally, this work is dedicated to Mary Eileen Gerling, who has been a source of encouragement and understanding during the past year and to his parents, Mr. and Mrs. K. C. Hood.
ii




TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS . . . . . . . . ii
LIST OF TABLES. v
LIST OF FIGURES vi
Chapter
I. INTRODUCTION. .. . . 1
II. PHOSPHORESCENCE AND TIHE HEAVY ATOM EFFECT . 6
Phosphorescence Theory . . ..... 6
Nature of the Heavy Atom Effect . . . 17
III. HEAVY ATOM EFFECTS: THE POLYNUCLEAR
AROMATIC HYDROCARBONS . . . . . 25
Introduction .. .. .. 23
Experimental Equipment. . . . . . 25
Experimental Procedure. . . . . . 27
Results and Discussion. . . . . . 30
IV. HEAVY ATOM EFFECTS: THE TRYPTOPHAN
METABOLITES ... 68
Introduction. .. . .. .. 68
Experimental Equipment. . . . .. 69 Experimental Procedure. . . . . . 70
Results and Discussion . . ...... 71
V SUMMARY . . . . . . . . .. 92
iii




Page BIBLIoGRAPHY . . 95
iv




LIST OF TABLES
Table Page
1. Phosphorescence Characteristics of
Polynuclear Aromatic Hydrocarbons in
Ethanol at 770K . . . . . . 31
2. Effect of Sodium Iodide and Ethyl Iodide on
Phosphorescence Intensity of Polynuclear
Aromatic Hydrocarbons ...... . . 48
3. Comparison of Phosphorescence Limits of
Detection for Polynuclear Aromatic
Hydrocarbons in Ethanol and in EEI at 770K. 55
4. Measurement of a Binary Mixture of
Triphenylene and 2,3-Benzfluorene . . 59
5. Phosphorescence Characteristics of
Tryptophan Metabolites in Ethanol at 770K 72
6. Effect of Ethyl Iodide on the Phosphorescence
Intensity of Tryptophan metabolites . . 84
7. Comparison of Phosphorescence Limits of
Detection for Tryptophan Metabolites in
Ethanol and in EEI at 77K. . . .. 85
8. Recovery of Kynurenic Acid from Urine . . 88
V




LIST OF FIGURES
Figure Page
1. Schematic representation of molecular energy
levels showing the ground singlet state, So ,
the first excited singlet state, S*, the
lowest triplet state, To, and the important
transitions which occur between them. . . 9
2-. Phosphorescence excitation and emission spectra
of naphthalene in ethanol at 770K . . . 33
3. Phosphorescence excitation and emission spectra
of phenanthrene in ethanol at 770K. . . 34
4. Phosphorescence excitation and emission spectra
of retene in ethanol at 770K. . .. . . 35
5. Phosphorescence excitation and emission spectra
of triphenylene in ethanol at 770K. . . 36
6. Phosphorescence excitation and emission spectra
of 1,2-benzanthracene in ethanol at 770K. . 57
7. Phosphorescence excitation and emission spectra
of chrysene in ethanol at 770K. . . . 38
8. Phosphorescence excitation and emission spectra
of 3,4-benzpyrene in ethanol at 770K. . . 59
9. Phosphorescence excitation and emission spectra
of 1,2,5,4-dibenzanthracene in ethanol at
770K 40
10. Phosphorescence excitation and emission spectra
of 1,2,5,6-dibenzanthracene in ethanol at
770K. .. 41
11. Phosphorescence excitation and emission spectra
of 1,2,7,8-dibenzphenanthrene in ethanol at
770K. . . 42
12. Phosphorescence excitation and emission spectra
of coronene in ethanol at 770K. . . . 45
vi




Figure Page
15. Phosphorescence excitation and emission spectra
of acenaphthene in ethanol at 770K. . . 44
14. Phosphorescence excitation and emission spectra
of 1,2-benzfluorene in ethanol at 770K. . 45
15. Phosphorescence excitation and emission spectra
of 2,5-benzfluorene in ethanol at 770K. . 46
16. Effect of ethyl iodide, A, and sodium iodide,
B, on the phosphorescence intensity of 1,2benzfluorene (10-4M) in ethanol-iodide
- solvent at 770K . 50
17. Absorption of ethanol-ethyl iodide, A,
ethanol-propyl iodide, B, ethanol-methyl
iodide, C, each 4/1, V/V, versus ethanol. . 52
18. Analytical curves for 1,2,3,4-dibenzanthracene in ethanol, A, and in EEI, B . 54
19. Phosphorescence emission spectrum of coronene
(10-4M) in ethanol,_, and in EEI,- -, at
770K # & # & 56
20. Phosphorescence emission spectrum of a mixture
of triphenylene, and 2,3-benzfluorene (both 10-4 M) in ethanol,_, and in EEI,- -, at
770K. & 58
21. Phosphorescence emission spectrum of 1,2-benzanthracene (10-M) in ethanol and in
EEI,- -, at 77K. ...... . . & 60
22. Phosphorescence emission spectrum of a mixture
of 1,2-benzanthracene, and 1,2,5,4-dibenzanthracene (both 10- M) in ethanol, and
in EEI,- -, at 770K. . . . . . 61
25. Phosphorescence emission spectrum of a mixture
of 1,2-benzfluorene, phenanthrene, and
acenaphthene (all 10-4M) in ethanol,-, and
in EEI,- -, at 770K . . . . . . 65
24. Semilogarithmic decay plot for 2,3-benzfluorene
in ethanol, A, and in ethanol which was
0.075 M in ethyl iodide, B . . . . 64 vii




Figure Page
25. Phosphorescence excitation and emission spectra
of quinaldic acid in ethanol at 770K. . . 73
26. Phosphorescence excitation and emission spectra
of kynurenic acid in ethanol at 770K. . . 74
27. Phosphorescence excitation and emission spectra
of xanthurenic acid in ethanol at 770K. . 75
28. Phosphorescence excitation and emission spectra
of indican in ethanol at 770K. . . . 76
29. Phosphorescence excitation and emission spectra
of indole-3-acetic acid in ethanol at 770K. 77
30. Phosphorescence excitation and emission spectra
of 3-indolepyruvic acid in ethanol at 7701K. 78
31. Phosphorescence excitation and emission spectra
of anthranilic acid in ethanol at 770K. . 79
32. Phosphorescence excitation and emission spectra
of 3-hydroxyanthranilic acid in ethanol at
770K. . . 80
33. Phosphorescence excitation and emission spectra
of nicotinic acid in ethanol at 770K. . . 81
34. Phosphorescence excitation and emission spectra
of quinolinic acid in ethanol at 770K . . 82
35. Phosphorescence emission spectrum of kynurenic
acid (10-4M) in ethanol, and in
EEI,- -, at 770K. 87
viii




CHAPTER I
INTRODUCTION
The use of phosphorescence emission spectra for identification of organic molecules was first suggested by Lewis and Kasha21 in 1944. In 1957, Keirs, Britt, and Wentworth20 evaluated phosphorimetry as a quantitative analytical technique. They reviewed the theory of phosphorescence, and discussed many of the factors which affect the intensity of the phosphorescence emission. Freed and Salmre12 in 1958 applied phosphorimetry to the measurement of various indole derivatives, and compared the sensitivities obtained with those of fluorimetry. Results indicated that phosphorimetry was ten times more sensitive for many of these biologically important compounds. In 1962, Parker and Hatchard32 considered the possibility of using phosphorescence measurements in chemical analysis. They concluded that phosphorimetry would be most useful when fluorescence at room temperature is insufficiently sensitive or specific.
The work described in this dissertation is a
continuation of basic studies in phosphorimetry initiated in this laboratory six years ago. In 1965, Winefordner
1




2
and Latz52 described construction of a spectrophosphorimeter from laboratory components, and applied this instrunent in the determination of aspirin in blood. The applicability of phosphorimetry to trace analysis in biological fluids was clearly demonstrated and efforts were directed toward improving techniques and perfecting new applications in this area. The availability of-commercial instrumentation* made it possible to determine the phosphorescence characteristics of a large number of corrmpounds, thus eliminating one of the major stumbling blocks to widespread acceptance of phospohorimetry as a routine metUhod--lac: of Oubl'ishe excitation and emission spectra and decay times. Much of this work in the field of applications has recently been
55
reviewed by Winefordner, McCarthy, and St. John.3
A restrictive factor in the practice of chosphorimetric analysis has been the availability of only a few suitable solvent media. Phoschorescence measurements must be made in a clear rigid matrix at low temperature, usually 770K. Some relaxation of these criteria may be realized using the rotating sample cell of Hollifield and Winefordner15 with which uniformly cracked low temperature
In most of the phosphorescence studies reported to
date, the Aminco-Bowman spectrophotoflucrometer with AmincoKeirs phosphoroscope attachment (American Instrument Co., Inc., Silver Spring, Maryland) has been used.




3
glasses can be measured reproducibly, Winefordner and
54k
St. John have reported a number of pure solvents and solvent mixtures which are suitable for phosphorimetry, but only ethanol, and EPA, a mixture of diethyl ether, isopentane, ethanol, 5/5/2, V/V, have been used extensively. Possibly because of this solvent restriction, the influence of environment on a solute molecule emitting phosphorescence has received little study in analytical phosphorimetry. It is interesting to note that in fluorimetry, the effects of environment have long been recognized, and investigations of pH dependence, quenching, and complexation, have contributed greatly to the versatility of this technique. Such environmental effects in fluorimetry have been recently reviewed by Wehry.50
The present research attempts to develop the
analytical potential of a specific environmental effect for application in phosphorimetry. Phosphorescence transitions are electronic transitions involving changes in multiplicity, and are forbidden quantum mechanically. Heavy atoms, halogens for example, introduced as molecular substituents, or contained in the solvent increase the transition probability of multiplicity-forbidden transitions in the perturbed molecule, thus enhancing phosphorescence emission. This perturbing action of heavy atoms is designated the heavy atom effect, and is categorized




4
as internal for heavy atom substituents, and external for heavy atoms in the solvent. This phenomenon has been widely studied by molecular spectroscopists, but little attempt has been made to exploit it in analytical procedures. Only the work of Zander,56 which appeared in the literature during the course of this investigation, has been analytically oriented. The emphasis in this dissertation is 6n achieving better sensitivity and selectivity in phosphorimetry through use of solvents which contain heavy atoms. The heavy atom species employed are the alkyl iodides and inorganic iodides which form a clear rigid glass when mixed with ethanol in certain proportions. Some experiments involving charge-transfer complexation are also reported because it is this type of interaction which is often invoked to explain the heavy atom effect. The acceptors, tetracyanoethylene, 1,3,5-trinitrobenzene, 2,LI,7trinitro-9-fluorenone, and 9(dicyanomethylene)2,4,7-trinitrofluorene, which form complexes with aromatic donors, are used as the perturbing species in this investigation.
The compounds selected for the perturbation studies were chosen for several reasons. The polynuclear atomatic hydrocarbons were initially investigated because these were previously used in most of the theoretical studies, and, for many, enhancement of sensitivity was required before phosphorimetry could be applied to measurement of




5
them at trace levels. Also, many of these compounds are carcinogens, and their determination is of biological interest. The tryptophan metabolites were chosen for their structural features, and for their biological importance.
The basis of the heavy atom effect is considered briefly in Chapter II. This chapter reviews some of the experimental studies which have elucidated the nature of the heavy atom effect, and suggested its application in phosphorimetry, as well as the theoretical aspects. In succeeding chapters, the experimental results obtained in analytical studies involving the polynuclear aromatic hydrocarbons (Chapter III) and the tryptophan metabolites (Chapter IV) are discussed. Emphasis in these investigations is on improvement of the measurement step through use of the heavy atom effect, rather than development of a complete analytical procedure for determination of these compounds in biological materials. However, in the metabolite study, estimation of kynurenic acid and xanthurenic acid in urine is considered. This latter work was undertaken to evaluate the comparability of phosphorimetry with heavy atom solvents and a single step thin-layer chromatographic separation.




CHAPTER II
PHOSPHORESCENCE AND THE HEAVY ATOM EFFECT
Phosphorescence Theory
Phosphorescence, as used in this dissertation, refers to a phenomenon in which photoluminescence is emitted by a molecule after absorption of light. A typical aromatic molecule has a singlet ground state (all electrons with paired spin) and several singlet and triplet excited states. In general, only the lowest excited state of a given multiplicity is capable of emission. Emission from the lowest excited singlet state is fluorescence; emission from the lowest triplet state is phosphorescence. Fluorescence may be directly excited by absorption of radiant energy of frequencies within the normal absorption band of the molecule. Similar direct excitation of phosphorescence is theoretically not possible due to restrictions imposed by the intercombination selection rule which forbids radiative transitions between states of different multiplicity. In order for phosphorescence to occur, the triplet state must be populated by radiationless transitions from the excited singlet state, a process called
6




7
intersystem crossing. This process is also forbidden quantum mechanically, and is only made possible through spin-orbit coupling, or through some perturbing influence.
The above relationships may be more readily
expressed in a simplified Jablonski17 term diagram, which describes the energy levels in a typical stable organic molecule (all electrons with paired spins). Such a diagram is shown in Figure 1. Each electronic state is assumed to contain numerous vibrational levels, but these are not shown for simplicity. By process 1, absorption of radiation, the electron is excited to the lowest electronic excited singlet state, S*, within a period of about 1015 seconds. If the electron were excited to a higher excited singlet state, rapid (10-13 to 10-11 seconds) radiationless deactivation would return the electron to the lowest excited singlet state. This process in which energy is dissipated as heat is called internal conversion. Vibrational relaxation occurring also in 10-13 to lO-ll seconds insures that any emission from the lowest excited state will originate from the lowest vibrational level of that state. Process 2, radiationless internal conversion, S*-So, is usually insignificant because the energy separation between the excited singlet and ground singlet is large. Process 3, fluorescence, is a radiational deactivation of the excited singlet, S*. The decay time of




Fig. 1.-Schematic representation of molecular energy
levels showing the ground singlet state, So,
the first excited singlet state, S*, the
lowest triplet state, To, and the important
transitions which occur between them.
Key to Figure
Process Rate Constant
1-Absorption of radiation k
a
2-Internal conversion, S*-S0 k
3-Fluorescence kf
4-Intersystem crossing, S*-T kic
5-Internal conversion, T -S k'
o o c
6-Phosphorescence k
p




s
4 To
S 0




10
fluorescence is generally 10-9 to lO-7 seconds, corresponding to the lifetime of the excited state. Process 4, intersystem crossing, S*-To, presents an alternative to fluorescence emission. This process involves vibrational coupling between the excited singlet state and the triplet state. The quantum mechanical multiplicity restriction on S*-T conversion renders intersystem crossing almost a million-fold less probable than normal vibrational relaxation or internal conversion. Consequently, the time required for intersystem crossing would be expected to be much longer than for vibrational relaxation. This time, lO-8 to 10-7 seconds, is approximately the same order of magnitude as the lifetime of the excited singlet state, and intersystem crossing can compete with fluorescence emission. Intersystem crossing results in population of the lowest vibrational level of the triplet state since internal conversion among the vibrational levels of the triplet manifold is rapid. Process 5, internal conversion, To-So, is a considerably more favorable process than S-So internal conversion for two reasons: (1) the energy separation between the triplet state and the ground singlet state is smaller, enhancing vibronic coupling; (2) the lifetime of the triplet state is long, favoring collisional




11
dissipation of energy.* In fact, for phosphorescence to be observed, highly viscous or rigid matrices must be employed to restrict collisional deactivation. Process 6, phosphorescence, is radiational deactivation of the triplet, T The decay time for phosphorescence may range from 10-4 to 10 seconds, or more, corresponding to the lifetime of the triplet state. Not shown is a thermal activation step, which can elevate an electron from the triplet state to the excited singlet state (intersystem crossing) resulting in delayed fluorescence. This process will not be important in a low temperature rigid matrix.
According to the approach of McCarthy and
Winefordner,23 an expression for the quantum efficiency of fluorescence, 0, and phosphorescence, Op, can be given in terms of the rate constants of Figure 1.
kf
S k, + k + k. ()
c ic
k k.
p c 1c+ f+ c
Strictly speaking, a radiationless transition involving a change in multiplicity is called intersystem crossing. It is convenient to group processes 2 and 5 under the heading of internal conversion since both result in radiationless return to the ground state.




12
The internal conversion and the intersystem crossing S*-T0 rate constants which determine the quantum efficiency depend on the temperature of the system. They are also influenced to a large extent by environmental factors, for example structure and viscosity of the solvent, and this will become important in the discussion of the heavy atom effect. In the absence of all quenching (k and k' are 0), the cc
relationships 0 F + 0P = 1 and F = kf/kic are obtained, which indicate the complimentary nature of fluorimetry and phosphorimetry.
Another parameter which is important in characterizing phosphorescence is the decay time. In terms of the experimentally observed signal,
I = Ioexp-t/-rp (3)
where 1 0is the intensity at time t = 0,1 is the intensity at time t after terminating the exciting light, and Tp is the phosphorescence decay time. In dilute, rigid media where To-S0 radiationless deactivation is the predominant quenching mode, Tp is given by the following equation.
Nk +k (4)
p C
This decay time should be distinguished from the natural phosphorescence decay time _o = 1/kp which would be observed in the absence of all quenching.




13
In order to apply phosphorescence measurements to
analysis, the intensity of the luminescence must be related to the concentration of the emitting species. This has been treated in detail by St. John, McCarthy, and
46D
Winefordner. Briefly, the radiant power of phosphorescence is given by
Pp = Pabsip (5)
where P is the radiant phosphorescence power emitted, P abs is the radiant power absorbed, and p is the power efficiency, the ratio of the radiant power emitted to the radiant power absorbed. The power efficiency is given by
1p O pex/Xem (6)
where p is the quantum efficiency and X ex and Xem are the peak wavelengths of the excitation and emission spectra,
respectively. The radiant power absorbed is 3iven by
Pabs = Po Pt = P0(1-exp-2.3abc)
where P is the radiant power incident on the absorbing sample, P is the radiant power transmitted through the sample absorbers, a is the molar absorptivity coefficient of the absorber, b is the path length over which absorption occurs, and c is the molar concentration of the absorbing sample. If the quantity (2.3abc) is less than 0.01 which is likely for dilute analytical solutions,




14
Pabs P02.3abc (8)
and
p = P 2.3ablp (9)
p o0
which gives the linear relationship between phosphorescence radiant power (intensity) and concentration.
It has been mentioned that spin-orbit coupling plays a predominant role in the phosphorescence process. Spinorbit coupling is significantly influenced by the heavy atom effect and for this reason is considered in some
de tail. 'Fr a more thorough discussion, see :be 2book 07 Condon and Shortly.5
In atoms, the interaction of the electrical and
magnetic fields of the different electrons determines the final pattern of energy levels. The motion of a negatively charged electron about a nucleus is governed by electrostatic attraction. This motion of the electron also
produces a magnetic field. Such fields are associated with the orbital motion of the electron. Similar fields are associated with its spin. In light atoms, t.he magnetic field resulting from the orbital motion of the electron about the nucleus is relatively small. In such cases, -he spin angular momenta of different electrons couple to form a total spin momentum designated S, and the orbital momenta couple to form a total orbital angular momentum designated




15
L, both by electrical mechanisms. Magnetic interaction of the spin and orbital momenta gives rise to a resultant total angular momentum, J, for the atom. This is the RussellSaunders coupling-scheme. In heavy atoms, the magnetic field which results from the orbital motion of the electron about the large nuclear charge becomes important. This magnetic field serves as a vehicle for interaction between the spin and orbital momenta to yield a resultant j for each electron. The individual j's then interact electrically to form a total J for the atom. This tyre p of coupling in called jj or spin-orbit coupling.
In molecules, as well as atoms, spin-orbit coupling
enhances the transition probability of sinrlet-triplet multiplicity-forbidden transitions through singlet-triplet mixing. Singlet-triplet mixing implies an "impurity" of state which can arise in the following way. A stable organic molecule in its ground state has all electrons, including the two of highest energy, arranged in orbitals with paired spins; that is, the multiplicity of the state is 1, a singlet state. In the excited singlet state, S*, the configuration of paired spins is maintained, one electron occupying its original orbital, the other electron occupying a higher formerly unoccupied orbital. In the triplet state, T, a similar situation prevails, but the
two electrons have parallel spins and the state is slightly




16
lower in energy than the excited singlet. The singlet state can be changed into a triplet state, or vice versa, if a force is applied which changes the direction of the spin magnetic moment of one of the electrons. The attractive nuclear field of atoms in the molecule can serve as the perturbing force which affects this change through the process of spin-orbit coupling. The result is singlet-triplet mixing, and a relaxation of the multiplicity restrictions for transitions between the perturbed states. This relaxation will make more permissive both non-radiational intersystem crossings, and radiative phosphorescence transitions, since both are multiplicity forbidden.
A quantitative relationship for the triplet-singlet 24
transition probability is given by McClure. The square of the fraction of singlet character in the triplet state is proportional to the probability of the intercombination transition. The fraction of singlet character introduced into the triplet state is also related to the spin-orbit coupling parameter, S-Values vary with the fourth power of the effective atomic number of an atom. The heaviest atom in a molecule can be used to estimate the magnitude of spin-orbit coupling for the molecule. As heavier atoms are introduced, electronic coupling gradually changes from LS to jj, and spin-orbit effects become significant.




17
Reviews which have discussed many of these aspects of phosphorescence theory in detail include those of Kasha and McGlynn19 and Lower and El-Sayed.22
Nature of the Heavy Atom Effect
The basis of the heavy atom effect has been described in the previous section dealing with spin-orbit coupling. An explanation of the effect of halogen substitution on phosphorescence transition probability follows directly from spin-orbit coupling theory. In 1949, McClure24 reported such a correlation for the series fluoro-, chloro-, bromo-, iodonaphthalene, and in addition noted considerable decrease in phosphorescence lifetimes for the same series. The halonaphthalenes have been the subject of extensive investigation by Pavlopoulos and El-Sayed.33 They found that the phosphorescence spectra of a halonaphthalene consists of two "subspectra" with one component similar to normal naphthalene in energy distribution and polarization and with a second component polarized perpendicularly to the first. This suggests that halogen substitution not only enhances spin-orbit coupling already present in the molecule, but also may introduce a new perturbation not present in the unsubstituted compound. This work has recently been reviewed by El-Sayed.11




18
Investigations of the external heavy atom effect have attempted to characterize the type of interaction required for the perturbing effect of the heavy atom solvent to be felt by a solute molecule. In 1952, Kasha18 observed that a binary solution of two colorless components, 1-chloronaphthalene and ethyl iodide, was yellow. This color was attributed to an enhanced SO-T0 absorption resulting from increased spin-orbit coupling in the 1-chloronaphthalene, induced through a simple collisional process. McGlynn and coworkers27 reported similar external heavy atom effects in phosphorescence studies and concluded that a collisional mechanism was unlikely in the rigid media employed. Instead, they suggested that charge-transfer interaction between an aromatic hydrocarbon and an alkyl halide might result in formation of a weak molecular complex. Graham-Bryce and Corkill13 made a similar suggestion. Equilibrium studies by McGlynn and coworkers28,35 have confirmed the charge-transfer complex hypothesis. NcGlynn, Sunseri, and Christodouleas28 were able to approximately correlate the magnitude of the heavy atom effect on a haloaromatic with the electron acceptor strength of a series of alkyl iodide perturbers. The order of electron acceptor tendency, CHI > CH3CH2I > CH3CH2CH2I > (CH3)2CHCH2I, was the same as the order of effectiveness of the perturbation. Recently, Eisenthal and El-Sayed, and Eisenthal9 studied




19
heavy atom effects on phosphorescence in charge-transfer complexes of a more conventional-nature. They investigated the complexes formed between haloaromatics and acceptors containing heavy atoms, for example the tetrahalophthalic anhydrides, and considered the importance of chargetransfer states in the singlet-triplet mixing process.
The spin-orbit coupling nature of the heavy atom effect has been well established, but exactly which rate constants are most significantly affected remains a question of controversy. The problem arises in trying to distinguish between effects on the rate constants kic, kp1 and k' (Figure 1) since all three involve multiplicityc
restricted processes. The increase in kic under heavy atom perturbation may be indirectly suggested by fluorescence quenching results, for example, the quenching of aromatic hydrocarbon fluorescence by potassium iodide.34 The implication is that under heavy atom perturbation, intersystem crossing competes more favorably with the fluorescence emission process. It is also likely that kic will be more sensitive to perturbation than k' on the basis of vic
brational overlap; the S*-T0 energy separation is usually less than the To-S0 separation, which facilitates crossover even in the absence of a perturber. Eisenthal and El-Sayed10 confirmed that the TO-S0 radiationless process is indeed not greatly affected by the heavy atom effect in a charge-transfer complex. McGlynn and coworkers have concluded that the




20
T o-S radiationless process is less sensitive to heavy atom perturbation than intersystem crossing, but more sensitive than phosphorescence. Siegel and Judeikis3 have reported a conflicting interpretation that the phosphorescence process is more sensitive to the heavy atom effect than either intersystem crossing or To-S radiationless deactivation.
Discrepancies in the interpretation of data from heavy atom experiments exist, but certain experimental results can be easily summarized. Addition of ethyl iodide to a solution of an aromatic hydrocarbon usually increases the quantum efficiency ratio OP/OF of the solute. The magnitude of an internal heavy atom effect and an external heavy atom effect are often comparable; the external effect may be more effective if the internal 22
effect is already present. Phosphorescence decay times are usually greatly decreased by the heavy atom effect. Decays of aromatics in alkyl halide solvents are also non-exponential. McGlynn and coworkers27 have explained this on the basis of charge-transfer complexation. A new species (complex) is formed which presumably has a short lifetime, and the resulting decay curve is the combination of the exponential decay of complexed and uncomplexed forms. The inhomogeneous nature of halide-containing glasses has also been offered as an explanation for the non-exponential




21
decay.43 In most heavy atom studies, phosphorescence spectra of perturbed compounds vary little from spectra of the unperturbed compounds. Graham-Bryce and Corkil113 did report shifts toward lower energy of 200 to 300 cm-1 for the phosphorescence emission of coumarin, acid fluorescein, and several nitronaphthalenes measured in an ethanol-ethyl iodide matrix. Similar red shifts have been described by McGlynn and coworkers.27 Eisenthal9 has recently noted significant changes in the vibrational character of donor phosphorescence for 1-chloronaphthalene when n-propyl iodide is introduced into the solvent.
Some comments should be made concerning the
luminescence of charge-transfer complexes formed between aromatic donors (n bases) and electron deficient acceptors (n acids) such as the aromatic hydrocarbon-trinitrobenzene system studied by Reid.36 Complexation affects the donors' luminescence in much the same way as the heavy atom perturbation. An increase in the quantum efficiency ratio OP/0F and a decrease in phosphorescence lifetime of the donor usually occurs on complexation. Czekalla et al.6
2
and Christodouleas and McGlynn ascribed the phosphorescence emission from complexes to a slightly perturbed emission of the donor. Recently charge-transfer triplet states characteristic of the molecular complex have been invoked to explain phosphorescence emission significantly different




22
from that of the donor component.16 Schenk and Radke41 have studied the quenching effect which tetracyanoethylene and other acceptors exert on the fluorescence of aromatic hydrocarbons. The fluorescence quenching mechanism appears to involve enhancement of intersystem crossing in the hydrocarbon. On the basis of these results, the use of charge-transfer acceptors as perturbing species in pho~phorimetry is suggested.




CHAPTER III
HEAVY ATOM EFFECTS: THE POLYNUCLEAR AROMATIC HYDROCARBONS
Introduction
Fluorimetry has been extensively applied in the analysis of polynuclear aromatic hydrocarbons.49 Phosphorimetry has been used only in a few instances. Since these two techniques are complimentary (0p + 0F 6 1), it is advantageous that the analyst use the procedure which might best be applied for a given determination. Considerable data are available in the literature concerning the fluorescence of aromatic hydrocarbons for analytical applications.49 Tabulations of phosphorescence excitation and emission wavelengths and decay times of a number of polynuclear aromatic hydrocarbons have also been reported.42 Much of this data is not useful analytically because of the large variation in experimental conditions used. The phosphorescence characteristics and limits of detection reported in this chapter are analytically useful information and suggest where phosphorimetry might be the preferred technique. The heavy atom studies show how phosphorescence measurement of several of the hydrocarbons 23




can be improved. In some cases, limits of detection may be lowered by a factor of 25. Changes which occur in the phosphorescence spectra of certain compounds in heavy atom solvents can be useful for identification purposes and also may improve the selectivity of measurement. These results extend the versatility of phosphorimetry for hydrocarbon analysis. In addition, the feasibility of using heavy atom perturbation techniques in phosphorimetry is clearly demonstrated. The analytical considerations important in this technique are developed to suggest the general applicability of such procedures in phosphorimetry. The effect of charge-transfer complexation not involving heavy atoms is also reported. This complexation does not appear to be useful in the same way as the heavy atom perturbation.
Mention should be made of the few phosphorimetric investigations of hydrocarbons which have appeared in the literature. Applications have been chiefly in the area of petrochemistry, or have involved determination of hydrocarbon pollutants in the environment. McGlynn, Neely, and Neely26 reported using phosphorimetry and low temperature fluorimetry to determine several hydrocarbons of petrochemical interest. Drushel and Sommers7 used phosphorimetry to characterize hydrocarbons in petroleum fractions. Zander55 reviewed the use of phosphorimetry for identification and quantitative determination of hydrocarbons in




25
coal-tar. The recent monograph by Zander57 includes a section on the use of phosphorimetry in the determination of impurities in polynuclear aromatic hydrocarbons. The heavy atom studies of Zander56 have already been mentioned and these will be discussed in detail later. Sawicki and Pfaff39 recorded phosphorescence spectra directly from thin-layer chromatograms to identify hydrocarbons in air pollution studies.
Experimental Equipment
Apparatus.-An Aminco-Bowman spectrophotofluorometer with an Aminco-Keirs phosphoroscope attachment, a 150-watt xenon arc lamp, and a potted RCA 1P28 multiplier phototube (American Instrument Co., Inc., Silver Spring, Maryland) was used for all phosphorescence measurements. Phosphorescence intensity readings were taken directly from the photometer unit supplied with the instrument. Phosphorescence excitation and emission spectra and decay times were recorded on an X-Y recorder. A fluorescence spectrophotometer with phosphoroscope attachment, Fluorispec, Model SF-1 (Baird-Atomic Inc., Cambridge, Massachusetts) was used for comparison purposes. A DK2 spectrophotometer (Beckman Instruments Inc., Richmond, California) was used to measure the absorbance of the heavy atom containing solvents.




26
Reagents and materials.-The hydrocarbons were
obtained from the following commercial sources: naphthalene, anthracene, phenanthrene (Distillation Products Industries, Rochester, New York); retene, triphenylene, 1,2,3,4-dibenzanthracene, 2,3,6,7-dibenzanthracene, 1,2,7,8-dibenzphenanthrene, coronene, perylene, 1,2,5,4dibenzpyrene, 3,4,8,9-dibenzpyrene, 1,2-benzfluorene, 2,5-benzfluorene (K and K Laboratories, Inc., Plainview, New York); chrysene, pyrene, naphthacene, acenaphthene (City Chemical Co., New York, New York); 1,2-benzanthracene, 3,4-benzpyrene, 1,2,5,6-dibenzanthracene, 20-methylcholanthrene (Nutritional Biochemicals Corp., Cleveland, Ohio). Charge-transfer acceptors were obtained from the following sources: tetracyanoethylene (City Chemical Co., New York, New York); 2,4,7-trinitro-9-fluorenone, 9(dicyanomethylene)2,4,7-trinitrofluorene, 1,3,5-trinitrobenzene (Distillation Products Industries, Rochester, New York). The majority of the above reagents were used as received. Alcohol solvent was purified by distillation using a five foot vacuum jacketed and silvered column with a reflux ratio of 20 to 1. Ethyl iodide (Fisher Scientific Co., Pittsburgh, Pennsylvania) and methyl and propyl iodide (Distillation Products Industries, Rochester, New York) were purified by passage through a column of activated silica gel in a dark room and were stored over copper away




27
from room light. Sodium iodide was reagent grade (J. T. Baker Chemical Co., Phillipsburg, New Jersey).
Experimental Procedure
Phosphorimetry of hydrocarbons.-Stock solutions of the hydrocarbons in ethanol were prepared in the concentration range 10-3M to 10- M depending on their solubilities. The stock solutions were stored at 200C, and were successively diluted with ethanol as required. Phosphorescence excitation and emission spectra were obtained for the ethanolic solutions of the hydrocarbons at 771K, and were uncorrected for instrumental response.51 Spectra were recorded at both high and low concentrations since phosphorescence excitation spectra are known to vary 110
somewhat with concentration. Various slit arrangements were used according to the manufacturer's literature.1 Spectra shown in this dissertation were obtained using the Aminco instrument and hydrocarbon solutions of intermediate concentration (l0-4M). In most cases the lmm slit was placed in front of the multiplier phototube to give suitable resolution Many spectra (not shown) were recorded on the high resolution Baird instrument with a spectral bandwidth of 2 mu for comparison.
Analytical curves (phosphorescence intensity signal versus sample concentration) were determined for each




28
hydrocarbon by measuring successive dilutions of the stock solutions. All precautions of cleanliness were observed as 48
previously described. Limits of detection in ug/ml were obtained using the slit program, 4,3,3,4,3, where the numbers indicate slit width in millimeters at various
1
positions in the instrument. The limit of detection was determined using a graphical extrapolation procedure.53 A sock solution was diluted well below the limit of detection, and a straight line was drawn through the resulting background points. The limit of detection was defined as that concentration corresponding to the intersection of the background line with the extension of the linear portion of the analytical curve. A compound was considered to be non-phosphorescent if a lO-3M solution gave less than a 50 per cent full scale meter deflection at the most sensitive instrumental settings.
Decay times were measured for hydrocarbon concentrations lying on the linear portion of the analytical curve. With a reading of approximately 90 units on the microphotometer, the exciting radiation was terminated using a manual shutter, and the phosphorescence decay was traced using the time base of the recorder. Decay times were determined. from semilogarithmic plots of relative intensity versus time. The response time of the recorder restricted decay time measurement to decay times greater than 0.5 seconds.




29
Heavy atom studies.-Spectra, analytical curves, and limits of detection were determined for each polynuclear aromatic hydrocarbon in ethanol-ethyl iodide mixtures of various proportions. Phosphorimetric measurements were made within one hour after adding the halide to the analytical solution. The effect of various concentrations of ethyl iodide on the phosphorescence of the hydrocarbons was measured to determine the halide concentration giving an optimum heavy atom effect. The enhancing effect of methyl iodide, ethyl iodide, propyl iodide, and sodium iodide were compared. Absorbances of the various solvent mixtures were also measured. Decay times were not determined for the compounds in ethanol-ethyl iodide solvent because the decays were generally non-exponential.2? Simple mixtures of hydrocarbons were studied to evaluate the selectivity of phosphorimetry with ethanol and ethanolethyl iodide solvent.
Charge-transfer complexes.-Stock solutions of tetracyanoethylene, 1,3,5-trinitrobenzene, 2,LJ,7-trinitro-9fluorenone, and 9(dicyanomethylene)2,4,7-trinitrofluorene were prepared in ethanol according to their solubility. The phosphorescence characteristics of the acceptors were determined. Phosphorescence spectra of mixtures of the acceptors with hydrocarbon donors were recorded. Little




30
useful information was obtained and no extensive investigation was undertaken.
Results and Discussion
Phosphorescence excitation and emission wavelengths, decay times, and limits of detection for the hydrocarbons in ethanol are reported in Table 1. These results in combination with the spectra shown in Figures 2 through 15 are useful in predicting whether phosphorimetry would be applicable to a particular problem in hydrocarbon analysis. Spectra shown are uncorrected for instrumental response,51 and are included mainly to indicate band shape and width which can be important in limiting the selectivity of phosphorescence measurements. From the wavelengths of the peaks, it should be possible to excite and to measure selectively many of the compounds in the presence of one or more interfering compounds. This selectivity may not be fully realized due to the broad bands of intensely phosphorescent interferents, such as triphenylene. Phosphorescence spectra of pyrene are not shown as this compound contained significant impurities. Problems encountered in the purification of pyrene have been discussed 4tf
at length by Srinivasan, Kinoshita, and McGlynn. It should be noted that naphthacene and 2,5,6,7-dibenzanthracene gave an intense phosphorescence emission which was the same




TABLE 1
PHOSPHORESCENCE CHARACTERISTICS OF POLYNUCLEAR AROMATIC HYDROCARBONS IN ETHANOL AT 770K
Hydrocarbon Excitation Emission Decay Timeblc Limit of Detectionb
Maximaa Maximaa
(mu) (mu) (seconds) (ug/ml)
Naphthalene 290,235 505,475,540 2.6 0.05
Phenanthrene 290 495,460,535 3.8 0.003
Retene 305 470,507,545 5.4 0.001
Triphenylene 290 457,455,490 16.2 0.0004
1,2-Benzanthracene 310,275 495,530 1.4 0.o07
Chrysene 270,310,525 500,540 2.2 0.009
Pyrene 332,255,300 595,610 0.5 0.4
3,4-Benzpyrene 330,285 510,550 2.3 1.5
1,2,3,4-Dibenzanthracene 295,330 560,600 0.9 0.4
1,2,5,6-Dibenzanthracene 305,330,340 550,590 1.5 0.01
1,2,7,8-Dibenzphenanthrene 290,530 500,535 2.5 0.002
Coronene 515,550 560,550 9.6 0.0009
Acenaphthene 305,235 475,510,550 3.0 0.05
1,2-Benzfluorene 315,275 495,535 2.7 0.04
H




Table 1 Cont'd.
Hydrocarbon Excitation Emission Decay Timeb'c Limit of-Detectionb
Maximaa Maximaa
(mu) (mu) (seconds) (ug/ml)
2,5-Benzfluorene 320,280 498,538 2.8 0.02
Naphthacenee.. 2,5,6,7-Dibenzanthracenee
Anthracene f... Perylene f... 20-Methylcholanthrene f... 1,2,3,4-Dibenzpyrene ....
f
3,4,8,9-Dibenzpyrene ...
aMost intense peak wavelengths are listed first; wavelengths are uncorrected for instrumental response;51 precision in wavelength observation is + 5 mu.
bMeasurements were made at most intense peak wavelength.
precision of measurement is + 0.1 second.
dApproximate value since sample contained impurities.
duct.3 ephosphorescence was measured, but is attributed to a photodecomposition profNo phosphorescence was measured.




I
z
Z
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 2.-Phosphorescence excitation and emission spectra of naphthalene in ethanol at 77OK.




LU
1
z
Lii
I! I I t I 1 I !
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mO)
Fig. 3.-Phosphorescence excitation and emission spectra of phenanthrene in ethanol at 770K.
p




Sf I 1 I III
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 4.-Phosphorescence excitation and emission spectra of retene in ethanol at
770K.




II I I I I I! I I
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 5.-Phosphorescence excitation and emission spectra of triphenylene in ethanol
at 77oK.




I I I 9I I 1 I 1
LIJ
I
z
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 6.-Phosphorescence excitation and emission spectra of 1,2-benzanthracene in ethanol at 770K.




uj
I--
z
LL
L
I 9 I I I 1 1 1 I
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mIl)
Fig. 7.-Phosphorescence excitation and emission spectra of chrysene in ethanol at 770K.
cO




Cf)
L 1 9 I I I I
z
uJ
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu) Fig. 8.-Phopphorescence excitation and emission spectra of 3,4-benzpyrene in ethanol at 770K.
.0




I
LU
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mI) Fig. 9.-Phosphorescence excitation and emission spectra of 1,2,3,4-dibenzanthracene in ethamol at 770K.
0




w
LU
260 250 300 350 400 450 500 550 600 650
WAVELENGTH (mJU)
Fig. 10.-Phosphorescence excitation and emission spectra- of l,2,5,6-dibenzanthracene in ethanol at 771K.




z
1 1 I I I I I
LU
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (m1u)
Fig. 11.-Phosphorescence excitation and emission spectra of 1,2,7,8-dibenzphenanthrene in ethanol at 770K.




I
m
a
I 1 I I i I I I
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 12.-Phosphorescence excitation and emission spectra of coronene in ethanol at 770K.




I I I I I I I I I
z
LU
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 13.-Phosphorescence excitation and emission spectra of acenaphthene in ethanol at 770K.




! I I I I f I I I
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (m)
Fig. 14.-Phosphorescence excitation and emission spectra of 1,2-benzfluorene in
ethanol at 770K.




p I I I 1 I I I
LI
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mj)
Fig. 15.-Phosphorescence excitation and emission spectra of 2,3-benzfluorene in ethanol at 770K.
ON




47
for both compounds. Clar and Zander have explained this naphthacene phosphorescence, which was also reported by Reid,37 in terms of emission from a photodecomposition product. It seems likely that 2,3,6,7-dibenzanthracene decomposes to give the same phosphorescent product. Phosphorescence decays for all compounds except 1,2-benzanthracene and pyrene were exponential. Significant differences exist among the decay times, for example, 16.2 seconds for triphenylene and 0.9 seconds for 1,2,3,4dibenzanthracene so that time-resolved phosphorimetryL might be applied to the measurement of such compounds in simple mixtures. Low limits of detection found for many of the hydrocarbons suggest that phosphorimetry could be used to measure these compounds at trace levels. Compounds which are rather insensitive by fluorimetry,49 such as triphenylene, retene, and phenanthrene, can be measured at much lower concentrations using phosphorimetry. Several intensely fluorescent hydrocarbons, for example, perylene, and 20-methylcholanthrene showed negligible phosphorescence. This gives phosphorimetry good selectivity for measurement of a species of interest in a solution containing these compounds as interferents.
The effect of sodium iodide and ethyl iodide on the phosphorescence intensity of the hydrocarbons is shown in Table 2. The phosphorescence intensity signal, Ro, was




48
TABLE 2
EFFECT OF SODIUM IODIDE AND ETHYL IODIDE ON PHOSPHORESCENCE
INTENSITY OF POLYNUCLEAR AROMATIC HYDROCARBONS
Hydrocarbona R/Rob
Sodium Iodide Ethyl lodidec
0.63M 0.63M 1.25M 2.50M
Naphthalene 4.0 0.29 0.21 0.12
Phenanthrene 1.5 0.64 0.40 0.22
Retene 1.0 1.1 1.1 0.77
Triphenylene 0.75 0.19 0.099 0.048
1,2-Benzanthracene 1.2 1.2 1.5 1.4
Chrysene 1.4 1.2 1.4 2.1
Pyrene 1.4 1.4 1.8 2.5
3,4-Benzpyrene 1.6 1.5 2.2 3.5
1,2,3,4-Dibenzanthracene 3.1 2.9 4.7 7.7
1,2,5,6-Dibenzanthracene 1.2 1.2 1.2 1.3
1,2,7,8-Dibenzphenanthrene 1.0 1.1 2.1 2.9
Coronene 3.0 2.7 3.9 6.0
Acenaphthene 3.4 0.91 0.86 0.74
1,2-Benzfluorene 5.6 6.4 9.8 13.
2,3-Benzfluorene 7.7 8.7 15. 25.
aHydrocarbon concentration was approximately 10-4M in ethanol-iodide solvent.
bRo is phosphorescence intensity of hydrocarbon in ethanol; R is phosphorescence intensity in ethanol-iodide solvent. Precision in intensity measurements is + 5 per cent. Wavelengths were adjusted to give maximum signal in each case.
CEthyl iodide concentrations expressed on a volume basis with ethanol at 20C00 are: 0.63M--19/1,V/V, ethanolethyl iodide; 1.25M--9/1,V/V, ethanol-ethyl iodide; 2.50NM-4/1,V/V, ethanol-eThyl iodide.




49
measured with the species in ethanol; the phosphorescence intensity signal, R, was measured with the species in the iodide solvent. The data in Table 2 were obtained by adjusting the excitation and emission wavelengths to give the maximum signal for each solution. The concentrations of ethyl iodide are expressed on a molar basis to permit comparison with the sodium iodide results. The increasing moldr concentrations of ethyl iodide correspond to 19/1, 9/1, 4/1, V/V, ethanol-ethyl iodide mixtures. Ethanol-ethyl iodide solutions in these proportions formed clear rigid glasses at 77'K. Concentrations of sodium iodide in ethanol which could be obtained were limited by solubility to less than 1 M. In addition, ethanol-sodium iodide solutions more concentrated than 0.63 M regularly formed cracked glasses when cooled to 771K. Figure 16 indicates the variation of the phosphorescence intensity of 1,2-benzfluorene as the ethyl iodide and sodium iodide concentrations are increased. The ethyl iodide plot was typical of those obtained for methyl iodide and propyl iodide, and indicated an optimum iodide concentration of approximately 2.5 1 (4/1, V/V, ethanol-ethyl iodide, henceforth designated EEI). Methyl iodide gave slightly greater enhancements of phosphorescence as suggested by McGlynn, 28
Sunseri, and Christodouleas, but solutions greater than
1 M in methyl iodide cracked consistently. There was




50
0

ww
~A C,i
zB
_J
0 1 2 3 4 5
CONCENTRATION OF IODIDE
(moles per liter)
Fig. 16.-Effect of ethyl iodide, A, and sodium iodide,
B, on the phosphorescence intensity of 1,2benzfluorene (10-4M) in ethanol-iodide solvent
at 770K.




51
essentially no difference between the enhancing effect of ethyl iodide and propyl iodide. Usable concentrations of sodium iodide were limited by solubility and cracking problems. The absorption characteristics of the three alkyl iodides were found to be similar as shown in Figure 17. All absorbed strongly below 3110 mu, and this explains the depression of phosphorescence noted for several cornpouzids (Table 2) which have strong excitation bands below 300 mu. At short wavelengths, the solvent competes favorably with the species of interest in absorbing the exciting radiation. Sodium iodide was found not to absorb the incident radiation, and some enhancement of phosphorescence was noted for almost all the hydrocarbons (Table 2). The disadvantages of using sodium iodide (solubility limitations and cracking) rendered it less useful than ethyl iodide except in a few instances. For example, a significant enhancement of naphthalene phosphorescence was obtained in ethanol-sodium iodide (Table 2) while no improvement in sensitivity was possible in ethanol-ethyl iodide. Iodide solutions are light sensitive, but no particular problem with photodecomposition was encountered if measurements were made within one hour after adding the iodide to the analytical solution.
Analytical curves for those hydrocarbons showing enhanced phosphorescence determined in EI were linear




0
0
0
C>1
300 30 40
WAEEGT mc
ZeR)




53
over ranges greater than or equal to those obtained in ethanol, and the limits of detection were lowered by an amount predictable on the basis of the enhancement ratio, R/Ro, shown in Table 2. Analytical curves are not shown for all the compounds because they were very similar. Figure 18 shows typical analytical curves for 1,2,3,4-dibenzanthracene in ethanol and in EEI. Limits of detection obtained in ethanol and in EEI are compared in Table 3. A considerable lowering of the limit of detection was observed for a number of the hydrocarbons which were not particularly low in ethanol. Excitation and emission wavelengths are also listed in Table 3 to indicate certain spectral changes brought about by addition of ethyl iodide to the solvent. The apparent red shift of the excitation wavelengths is primarily a result of the absorption of exciting radiation by the solvent. Any red shifts in the emission spectra were slight. Noticeable changes in vibrational structure of the emission spectra did occur, and this 'effect may often improve the selectivity of phosphorescence measurements.. For coronene and triphenylene measured in ethanolethyl iodide, the 0-0 emission band became much more intense at the expense of the longer wavelength bands. The change in the emission spectrum is shown for coronene in Figure 19. This appears to be an effect which is common to almost all the hydrocarbons but which is significant in only a few.




54k
z
w
>B A
CONCENTRATION (moles per liter)
Fig. 18.-Analytical curves for l,2,5,4-dibenzanthracene in ethanol, A, anda in EEI, B.




TABLE 3
COMPARISON OF PHOSPHORESCENCE LIMITS OF DETECTION FOR POLYNUCLEAR
AROMATIC HYDROCARBONS IN ETHANOL AND IN EEI AT 770K
Hydrocarbona Ethanol EEIb
Excitation Emission Limit of Excitation Emission Limit of Maximum Maximum Detection Maximum Maximum Detection
(mu) (mu) (ug/ml) (mu) (mu) (ug/ml)
1,2-Benzanthracene 310 500 0.07 327 505 0.06
Chrysene 308 506 0.009 328 507 0.003
Pyrene 332 592 0.4 335 595 0.1
3,4-Benzpyrene 325 508 1.5 332 510 0.3
1,2,3,4-Dibenzanthracene 300 564 0.4 335 564 0.06
1,2,5,6-Dibenzanthracene 305 550 0.01 336 550 0.008
1,2,7,8-Dibenzphenanthrene 290 503 0.002 335 503 0.0007
Coronene 315 560 0.0009 342 518 0.0002
1,2-Benzfluorene 315 500 0.04 320 501 0.003
2,3-Benzfluorene 325 500 0.04 325 500 0.002
a0nly hydrocarbons which show an enhancement of phosphorescence in EEI are listed.bEEI is the ethanol-ethyl iodide mixture, 4/1, V/V.
L.




56
e -X
Hi 'em, 3 20 -:iu
LW
CrI
400 40 500 550 60 65
WAVEENGT (M
Fig. 9.-Posphoescece eisso pcrm fcrnn (1 ~ I-L)i tao ,adi E I---,a
7?OII
Instrumental~~~ snitvtwareudbyafco
of~~~~~~~~~~~ aiI-o~aey3t bansetu nEI Exiato wvle ,t r-a,. 10m i -IIol n
320~~ mui E




57
Figure 20 shows how such changes can be useful. Triphenylene caused particular problems in attempting to measure less intensely phosphorescent compounds in mixtures. The intense emission of triphenylene was found to extend into the spectral region of emission of many other hydrocarbons. In EEI, the triphenylene emission spectrum was altered so that the maximum intensity was centered at 432 mu instead of at 457 mu, thus permitting measurement of compounds whose spectrum was not changed, for example, 2,3-benzfluorene. In Table 4, quantitative results are presented for measurement of triphenylene and 2,3-benzfluorene in ethanol and in EEI. For some hydrocarbons in EEI, spectral bands changed little in relative intensity, but fine structure became more or less well defined compared to the structure observed in ethanol. Figure 21 shows that the structure of 1,2-benzanthracene was considerably more prominent in EEI than in ethanol.
Use of the heavy atom effect to improve selectivity has already been mentioned for the case of triphenylene. Measurement of spectra in ethanol, and then in EEI was useful in characterizing mixtures of compounds. Changes in band structure and relative intensity provided an additional basis for identifying a compound not available when measurements were made only in ethanol. Figure 22 shows the phosphorescence emission spectrum of a mixture




2,3-benzf luorene
triphenylene
r
I ex 320 mu
w
w
400 450 5bO 5k 6bO 650
WAVELENGTH (m 1j)
20.-Phos-ohorescence e-n'
-Ission
of triphenylene, and 2,7;--benz'-1uo--enc ,,'--o-L,h
10-L M) in ethanol,
77ox- and in EEI at
Instrumental sensitivity was the same bot-h
snectjra. 3- -,cita--Lon 1.,ras ;-"D mu
in both e+,-hanol and




59
TABLE 4
MEASUREMENT OF A BINARY MIXTURE OF TRIPHENYLENE AND 2,3-BENZFLUORENE
Percent Error in Measured Concentration
Concentration Triphenylene 2,5-Benzfluorenec
Ethanol EEla Ethanol EEIa
2 x 10-4 -7 -4 +95 +2
8 x 10-6 -2 -6 +48 -3
aEEI is the ethanol-ethyl iodide mixture, 4/1, V/V.
bExcitation and emission maxima were 290 mu, 454 mu, in ethanol and 295 mu, 432 mu in EEI.
CExcitation and emission maxima were 320 mu, 498 mu, in ethanol and 325 mu, 498 mu in EEI.




60
II
%ex 308 mu1 ex 322 mu-I i
SI \
I I I
I I I I I I
400 450 500 550 600 650
WAVELENGTH (mil)
Fig. 21.-Phosphorescence emission spectrum of 12benzanthracene (10-4M) in ethanol,_, and
in EEI, - -, at 770K.
Instrumental sensitivity was reduced by a factor
of approximately 1.5 to obtain spectrum in EEI.
Excitation wavelength was 308 mu in ethanol and
322 mu in EEI.




61
t I I
1,2-benzanthracene
1Xex 318 mu
I'
1" 1,2,3,4-dibenzI anthracene
\ 1 I
I I
I / I
I
400 450 500 550 600 650
WAVELENGTH (m811)
Fig, 22.-Phosphorescence emission spectrum of a mixture
of 1,2-benzanthracene, and 1,2,5,4-dibenzanthracene (both 10- H) in ethanol, and in
EEI,- at 770Ko
Instrumental sensitivity was reduced by a factor
of approximately 1.5 to obtain spectrum in EEI.
Excitation wavelength was 318 mu in both ethanol
and EEI.




62
of 1,2-benzanthracene and 1,2,3,4-dibenzanthracene, recorded in ethanol and in EEI. Excitation was at 318 mu for both spectra. In EEI, the 1,2,3,4-dibenzanthracene emission peak appeared at 560 mu, and the band structure of the 1,2-benzanthracene became more pronounced. In Figure 23, the phosphorescence emission spectrum of a mixture of 1,2-benzfluorene, phenanthrene, and acenaphthene is shown. In ethanol, excitation at 320 mu gave a complex spectrum with contributions from the three compounds. In EEI, the spectrum of 1,2-benzfluorene could be clearly recorded. However, excitation at 290 mu in ethanol gave the phenanthrene emission spectrum alone.
Phosphorescence decays in ethanol-ethyl iodide
mixtures were non-exponential. This effect is shown in Figure 24 for the decay of 2,3-benzfluorene. The intensity of a phosphorescence signal could decrease if the decay time of a compound became short with respect to the exposure time of the phosphoroscope. O'Haver and Winefordner31 have measured the important phosphoroscope parameters which influence the measured phosphorescence intensity, and found that the exposure time for the Aminco phosphoroscope is around 10-3 seconds. Since the decay times of most of the hydrocarbons were found to be one second or longer, the decrease in decay time which occurs in heavy atom solvents should not adversely affect the




63
1,2-benzfluorene
%ex 320 mu
,,,I
,J phenanthrene I I 1
I.I
cc
1I
1 \ I 1
Fig. 2c.-Phosphorescence emission spectrum of a mixture of
1,2-benzfluorene, phenanthrene, and acenaphthene
(all 10-4) in ethanol, and in EEI,- at
770K.
Instrumental sensitivity was reduced by a factor
/
400 450 500 550 600 650
WAVELENGTH (mu)
Pig. 23.-Phosphorescence emission spectrum of a mixture of
1,2-benzfluorene, phenanthrene, and acenaphthene
(all 10- N) in ethanol,__, and in EEl,- -, at
77K.
Instrumental sensitivity was reduced by a factor
of approximately 15 to obtain spectrum in EEI.
Excitation wavelength was 320 mu in both ethanol
and EEI.




64
0
o
z
U
t I
0 12 3 4
TIME (seconds)
Fig. 24.-Semilogarithmic decVy plot for 2,5-benzfluorene in ethanol, A, and in ethanol
which was 0.075 M in. ethyl ioalde, B.




65
phosphorescence intensity for measurements made with this phosphoroscope. The decay time will rarely be reduced by a factor of 50 in halide solvents.27
The results presented in the above studies are very similar to those reported by Zander,56 who used EPA-methyl iodide, 10/1, V/V, as a perturbing solvent. Greater proportions of methyl iodide could not be used because of the 56
cracking problem. In Zander's studies, enhancements in phosphorescence intensity were achieved for such compounds as phenanthrene, because excitation was carried out at wavelengths lying in the long wavelength tail of the absorption band, 345 mu, where solvent absorption was not a significant factor. While more intense phosphorescence signals would be obtained in the halide solvent than in EPA, these signals would not exceed those in unperturbed media if excitation were performed at the excitation maximum of 290 mu. In other respects, the work reported by Zander56 confirms the usefulness of the heavy atom effect for phosphorescence measurement of hydrocarbons.
Investigation of the complexes formed between the hydrocarbons and various acceptors produced little useful information. The strongest acceptors, 2,4,7-trinitro-9fluorenone, and 9(dicyanomethylene)2,4,7-trinitrofluorene, which should exert the largest perturbing effect phosphoresced and were only slightly soluble in ethanol.




66
This made difficult the study of the phosphorescence of the donor components in solutions of.donor and acceptor. Tetracyanoethylene did not phosphoresce, but is a more selective complexing agent for simple aromatic compounds. The weakest of the charge-transfer acceptors, 1,3,5-trinitrobenzene, did not phosphoresce and was easiest to work with. An enhancement in phosphorescence intensity was observed for-solutions of pyrene which were 10-2 M in 1,3,5-trinitrobenzene. The magnitude of this effect was about onethird of that obtained with a 102 M pyrene solution in EEl. This enhancement was peculiar to pyrene and did not occur for any of the other hydrocarbon complexes of 1,3,5trinitrobenzene examined. The most significant problem encountered in working with charge-transfer complexes of this type was that the large concentration of acceptors needed for effective perturbation was difficult to achieve. In many cases, a crystalline complex would form when solutions were cooled to 770K.
The use of the heavy atom effect is simple and can
often improve the sensitivity and selectivity of phosphorescence measurements. Ethyl iodide and propyl iodide appear to be equally convenient for use in perturbing solvents. Sodium iodide has the advantage of not reducing the intensity of the incident light through absorption; however, solubility




67
limitations and cracking problems render it less useful than the alkyl iodides. The possibility of using heavy metal iodides is also governed by solubility limitations, and would offer no advantage. Ethyl iodide is expensive, but the small quantities needed to affect a large increase in phosphorescence intensity should justify the expense. It seems likely that heavy atom solvents could be used in those applications of hydrocarbon analysis where phosphorimetry has already been applied7'39'55 and that such use would greatly extend the versatility of the measurement procedure.




CHAPTER R IV
HEAVY ATOM EFFECTS: THE TRYPTOPHAN METABOLITES
Introduction
Phosphorimetry has been shown to be very useful in detecting certain indoles and related compounds at trace levels.12',45 This chapter presents the phosphorescence characteristics of a group of tryptophan metabolites, many of which have not been previously studied by phosaorimery. Interest in these compounds, especially kynurenic acid and xanthurenic acid, stems from their use as monitors in studying any malfunction in the metabolism of tryptophan. Low limits of detection are obtained for many of the metabolites. The limit of detection by phosphorimetry for kynurenic acid compares favorably with that obtained by fluorimetry.- Xanthurenic acid can be measured at much lower levels by fluorimetry, but exhibits a 10-fold enhancement of phosphorescence when measurements are made in EEL. Kynurenic acid and xanthurenic acid are well suited for heavy atom perturbation studies because their excitation wavelengths are near 550 mu where reduction of incident light intensity by the halide solvent is not significant.
68




69
The effect of heavy atoms on the phosphorescence of the metabolites differs in some ways from the effects observed in the hydrocarbon study, and these are discussed in this chapter.
Several methods have been used to determine kynurenic acid and xanthurenic acid in urine. Fluorimetric procedures have been used after a lengthy separation of the metabolites on columns of ion-exchange resin.38 One of the simplest methods has involved a thin-layer chromatography separation, 14
followed by measurement using absorption spectrophotometry.14 Phosphorimetry has been used successfully in combination with thin-layer chromatography for determination of pnitrophenol, the metabolic product of parathion in urine.29 The compatability of thin-layer chromatography and phosphorimetry in heavy atom solvents for determination of kynurenic acid and xanthurenic acid in urine is investigated in the present work. It is shown that adsorption on silica gel can render inoperative the heavy atom effect which is normally observed. Such an effect was encountered with xanthurenic acid.
Experimental Equipment
AEparatus.-The spectrophosphorimeter has been described in Chapter III. Commercial thin-layer chromatography equipment (Brinkmann Instruments, Inc., Westbury, New York)




70
was used for the separation studies. A 10 ul syringe (Hamilton Co., Inc., Whittier, California) was used to apply samples to the thin-layers. A long wavelength ultraviolet lamp (Blak-Ray Model XX15, Ultra-Violet Products, Inc., San Gabriel, California) was employed to locate fluorescent spots on the thin-layers.
Reagents and materials.-The tryptophan metabolites were used as received (Nutritional Biochemicals Corp., Cleveland, Ohio). Alcohol and ethyl iodide were purified as described in Chapter III. Solvents for thin-layer chromatography were reagent grade. Silica gel G for thin-layer chromatography (E. Merck A. G., Darmstadt, Germany) was heated at 7000C for 12 hours before use to reduce the luminescence background.30 Experimental Procedure
Phosphorimetry and heavy atom studies.-Procedures
described in Chapter III were followed in obtaining phosphorescence excitation and emission wavelengths, decay times, analytical curves, and limits of detection for the tryptophan metabolites. Heavy atom studies were restricted to the optimum ethanol-ethyl iodide mixture, EEI.
Thin-layer chromatography.-The thin-layer
chromatography separation has been described in detail by




14
Hill, Summer, and Roszel. Briefly, separation was achieved on silica gel G layers, 20 cm by 20 cm, and 250 u thick, activated for 30 minutes at 110C. A 20 ul sample, either urine containing added metabolite, or a standard ethanolic solution of metabolite was applied to the thin-layer using the 10 ul syringe in increments of 5 ul. Development by chloroform-methanol-acetic acid, 75/20/5, V/V, required 75 minutes. Fluorescent spots were located under ultra-violet light and the adsorbent removed. Ethanol extraction in centrifuge tubes followed by centrifugation yielded solutions ready for phosphorimetric analysis. During extraction, the solutions were frequently stirred and were maintained at 60*C in a water bath. Generally only a single extraction with 5 or 10 ml of ethanol was required. When ethyl iodide was used as a perturbing solvent, this was added to the ethanol directly before measurements were made.
Results and Discussion
Phosphorescence excitation and emission wavelengths, decay times, and limits of detection for the tryptophan metabolites determined in ethanol are listed in Table 5. Phosphorescence excitation and emission spectra are shown in Figures 25 -through 34. Significant differences in excitation and emission wavelengths make possible the




TABLE 5
PHOSPHORESCENCE CHARACTERISTICS OF TRYPTOPHAN METABOLITES IN ETHANOL AT 770K
Metabolite Excitation Emission Decay Timeb'c Limit of Detection
Maximaa Maximaa (seconds) (ug/ml)
(mu) (mu)
Quinaldic acid 505,245 510,475 1.1 0.015
Kynurenic acid 348,255 458 1.6 0.003
Xanthurenic acid 550,250 485 0.9 0.1
Indican 295,235 460 3.3 0.03
Indole-3-acetic
acid 285,250 435,410,455 7.0 0.008
3-Indolepyruvic
acid 502,265 425,410 0.6 0.003
Anthranilic acid 350,255 440 1.9 0.004
5-Hydroxyanthranilic acid 350 480 0.7 0.20
Nicotinic acid 262,250 400 0.6 0.015
Quinolinic acid 272,235 450 0.5 0.007
aThe most intense peak wavelength is listed first; wavelengths are uncorrected for instrumental response;51 precision in wavelength observation is + 5 mu.
b
bMeasurements were made at most intense peak wavelength.
precision of measurement is + 0.1 second.
I'.




L,.
I 19I I I I II
U
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mI)
Fig. 25.-Phosphorescence excitation and. emission spectra of quinaldic acid in
ethanol at 770K.




tI I I I I IF
4
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 26.-Phosphorescence excitation and emission spectra of kynurenic acid in
ethanol at 770K. -,




m
I I I I I I I I
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 27.-Phosphorescence excitation and emission spectra of xanthurenic acid in ethanol at 770K.
"i




I I 1 I I f I I I
LUJ
z
LL
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 28.-Phosphorescence excitation and emission spectra of indican in ethanol at 770K.
('




z
S
I I I I 1 I t I 1
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mij) Fig. 29.-Phosphorescence excitation and emission spectra of indole-3-acetic acid
in ethanol at 770K.




LL
I I I I I I 1 I f
IL
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 30.-Phosphorescence excitation and emission spectra of 3-indolepyruvic acid in ethanol at 770K.
00




m
LI
I I I I I I I I
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 31.-Phosphorescence excitation and emission spectra of anthranilic acid in ethanol at 770K.




r/
Cin
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mU)
Fig. 32.-Phosphorescence excitation and emission spectra of 3-hydroxyanthranilic acid in ethanol at 770K. CO
O




U
Lu
1=
I I, I I I I I
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (ml)
Fig. 33.-Phosphorescence excitation and emission spectra of nicotinic acid in ethanol at 770K.
OO




LI
1 I 1 I I I I I I
200 250 300 350 400 450 500 550 600 650
WAVELENGTH (mu)
Fig. 34.-Phosphorescence excitation and emission spectra of quinolinic acid in
ethanol at 770K.
N\)




83
measurement of many of these compounds in the presence of others. Most of the decay times-are quite similar, except for that of indole-3-acetic acid, and it is unlikely that time-resolved phosphorimetry could be easily applied to the analysis of mixtures.47
In Table 6, it is evident that for some of the
compounds, significant enhancement of the phosphorescence intensity occurred when ethyl iodide was added as a perturbing solvent. No attempt was made to measure those compounds having peak excitation wavelengths shorter than 300 mu, because of the problem of solvent absorption in this region. For xanthurenic acid and 3-hydroxyanthranilic acid which are relatively weakly phosphorescent in ethanol (Table 5), the limits of detection were lowered by nearly an order of magnitude when measurements were made in EEI. In Table 7, limits of detection for the metabolites in ethanol and in EEI are compared, and the wavelength maxima in the two solvents are given. For quinaldic acid, the analytical curve in EEI had a slightly greater slope than in ethanol, and for this reason, the limit of detection was the same in both solvents. Other analytical curves in EEI paralleled those in ethanol which was the behavior observed in the hydrocarbon study. The stability of metabolites in EEI was less than that observed for the hydrocarbons. It was essential that measurements of




84
TABLE 6
EFFECT OF ETHYL IODIDE ON THE PHOSPHORESCENCE
INTENSITY OF TRYPTOPHAN METABOLITES
Metabolitea R/Rob
Quinaldic acid 2.0
Kynurenic acid 5.7
Xanthurenic acid 9.5
Anthranilic acid 2.7
5-Hydroxyanthranilic acid 14.0
3-Indolepyruvic acid 1.0
aMetabolite concentration was approximately 10- M in EEI solvent.
bRo is phosphorescence intensity of metabolite in ethanol; R is phosphorescence intensity in EEI. Precision in intensity measurements is + 5 percent. Wavelengths were adjusted to give maximum signal in each case.




TABLE 7
COMPARISON OF PHOSPHORESCENCE LIMITS OF DETECTION FOR
TRYPTOPHAN METABOLITES IN ETHANOL AND IN EEI AT 770K
Metabolite a Ethanol EEIb
Excitation Emission Limit of Excitation Emission Limit of Maximum Maximum Detection Maximum Maximum Detection
(mu) (mu) (ug/ml) (mu) (mu) (ug/ml)
Quinaldic acid 305 510 0.015 328 517 O.015
Kynurenic acid 348 458 0.003 350 475 0.0015
Xanthurenic
acid 350 485 0.1 360 505 0.015
Anthranilic
acid 350 440 0.004 350 440 0.002
3-Hydroxyanthranilic
acid 350 480 0.2 350 480 0.015
3-Indolepyruvic
acid 302 425 0.003 312 425 0.003
aOnly those metabolites which show an enhancement of phosphorescence in EEI are listed.
bEEI is the ethanol-ethyl iodide mixture, 4/1, V/V.
0n




86
phosphorescence intensity be made immediately after addition of ethyl iodide to the analytical solution. If this procedure were adhered to, reproducible results were obtained.
Ethyl iodide also had a significant effect on the
emission spectra of several of the metabolites. Red shifts in the emission maximum not noted in the hydrocarbon studies were significant for quinaldic acid, kynurenic acid, and xant-hurenic acid. This red shift and an increase in structure in the kynurenic acid emission spectrum is shown in Figure 35. The emission spectrum of xanthurenic acid retained its shape in EEI but was shifted to lower energy. These shifts are in agreement with results reported by Graham-Bryce and Corkil113 for coumarin in ethanol-ethyl iodide mixtures. The use of such shifts and changes in spectral shape for characterization purposes has been pointed out in the hydrocarbon study.
Thin-layer chromatography followed by measurement of phosphorescence intensity in ethanol was found to be useful for the determination of kynurenic acid in urine. In Table 8, recoveries of kynurenic acid from urine after a simple thin-layer separation are given. A single extraction with 5 ml of ethanol gave better than 90 per cent recovery. Thin-layer chromatography required 75 minutes, and the extraction and measurement steps approximately 30 minutes. The precision indicated is for five complete




87
It
I Xex 348 mu
I\
L 'I
I
I I
I
I I
I I
, II II
400 450 500 550 600 650
WAVELENGTH (mj)
Fig. 35.-Phosphorescence emission spectrum of
kynurenic acid (10-4) in ethanol,
and in EEl,- -, at 77oK.
Instrumental sensitivity was reduced by a
factor of approximately 3 to obtain spectrum in EEI. Excitation wavelength was 348 mu in
both ethanol and EEI.




88
TABLE 8
RECOVERY OF KYNURENIC ACID FROM URINEa
Kynurenic Acid Concen- Percent
tration in Urine Recovery
(ug/ml)
200 94+3
150 92+2
100 95+5
50 91+3
10 94+8
aMeasurement was by phosphorimetry in ethanol after thin-layer chromatography; excitation and emission wavelengths were 348 mu, and 458 mu, respectfully. Precision is for five determinations.




89
determinations. Blanks obtained from urine chromatographed with the sample gave intensity readings only slightly higher than the ethanol background at the wavelengths used. Xanthurenic acid could not be satisfactorily measured in ethanol at the same concentrations as kynurenic acid. The Rf values for kynurenic acid and xanthurenic acid for the chromatographic system used are 0.20 and 0.29 respectively,14 and any overlap of kynurenic acid and xanthurenic acid on the thin-layer would not lead to errors in measurement of kynurenic acid using the above procedures. Thus, this method is very specific for kynurenic acid.
Xanthurenic acid extracted from silica gel as above could not be measured in EEI even though concentrations were well above the limit of detection. Apparently, adsorption on silica gel resulted in an irreversible effect on xanthurenic acid which rendered the extracted material different from pre-adsorbed compound. Adsorption on silica gel did not affect the fluorescence of xanthurenic acid, and the fluorescence intensity measured confirmed that the ethanol extraction of the compound was nearly complete. This apparent quenching of the heavy atom effect is not general for materials extracted from silica gel. The heavy atom effect normally found for the polynuclear aromatic hydrocarbons was observed after ethanol extraction of these compounds from thin-layers of silica gel. The heavy atom




90
effect was also observed in kynurenic acid extracts. Simply adding silica gel to a solution of xanthurenic acid did not quench the heavy atom effect. Once xanthurenic acid was adsorbed on silica gel, however, the heavy atom effect was inoperative even if the silica gel were heated at 1500C for several hours prior to extraction. The unusual results found for xanthurenic acid might be explained in terms of a strong interaction between the adsorbed metabolite and silica gel which blocks formation of the xanthurenic acidethyl iodide charge-transfer complex. Because xanthurenic acid can be an effective chelating agent, the suggestion of a strong interaction between it and silica gel may be reasonable. Clementi and Kasha have studied perturbation effects in dyes adsorbed on silica gel where the gel itself was the heavy atom species. Close contact between dye and gel, that is, a monolayer of adsorbed dye, was necessary for effective perturbation. By analogy, saturation of xanthurenic acid with silica gel could restrict the approach of ethyl iodide.
Application of heavy atom perturbation techniques to the determination of xanthurenic acid in urine was not practical using the simple thin-layer separation procedure. This result does not detract from the possible use of such techniques in measuring xanthurenic acid under different circumstances. Kynurenic acid can be adequately measured




91
by phosphorimetry in ethanol at the concentrations often encountered in urine after oral ingestion of tryptophan. Measurement of kynurenic acid by phosphorimetry is a more sensitive procedure than the absorption method of Hill, Summer, and Roszel.14 Even lower levels of kynurenic acid could be detected if measurements were made in the heavy atom solvent. These results indicate that the heavy atom effect in phosphorimetry is not limited to the polynuclear aromatic hydrocarbons, the group of compounds with which it is most often discussed. The results also point out that for more complicated molecules, such as xanthurenic acid, other environmental effects can influence heavy atom perturbation in unusual ways. Such effects of a highly selective nature may be of use in phosphorimetry in their own right, and suggest further research in this area.




CHAPTER V
SUMMARY
The possibility of capitalizing on environmental effects to improve the sensitivity, selectivity, and versatility of phosphorimetry has received little previous attention. In Chapter II, it is shown how theoretical considerations suggest the potential use of heavy atom perturbation in analytical phosphorimetry. The remainder of the dissertation has been concerned with a demonstration of this potential.
In Chapter III, it is shown that ethyl iodide added to the usual ethanol solvent is a convenient means of enhancing the phosphorescence intensity of many of the polynuclear aromatic hydrocarbons. A 25-fold enhancement was obtained for 2,3-benzfluorene. In addition, the changes in emission spectra which accompany heavy atom perturbation and changes in relative intensity of bands and in vibrational structure are shown to be useful for improving the selectivity of phosphorescence measurements. Investigation of simple charge-transfer complexes formed between the aromatic hydrocarbons and various acceptor compounds showed that, while theoretically this complexation 92




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ESVI5LS4R_QFURLR INGEST_TIME 2018-07-17T21:06:46Z PACKAGE AA00063469_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

INVESTIGATION OF THE EXTERNAL HEAVY ATOM EFFECT AS A MEANS OF IMPROVING THE SENSITIVITY AND SELECTIVITY OF ANALYTICAL PHOSPHORIMETRY By LYAL VAN SANT HOOD A DISSERTATION PRi:SENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN P AllTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHJLOSOPHY UNIVERSITY OF FLORIDA 1968

PAGE 2

ACKNOWLEDGMENTS The author will always be indebted to his research director, Dr. J. D. Winef ordner, for his enthusiastic encouragement and guidance during the course of this work. Appreciation is also extended to the other members of his committee, Dr. R. B. Bennett, Dr. G. M. Schmid, Dr. S. P. Cram, and Dr. W. S. Brey. The friendly assistance and good fellowship of his colleagues during four years of graduate study are also gratefully acknowledged Finally, this work is dedicated. to Mary Eileen Gerling, who has been a source of encouragement and understanding during the past year and to his parents, Yir. and Mrs. K C. Hood. ii

PAGE 3

TABLE OF CONTENTS Page ACKNOWLEDGM E NTS. . . . . . . LIST OF TABLES LIST OF FIGURES . Chapter I. II. III. IV. v. IN1'RODUCTION. . . PHOSPHORESCE N CE AND 'l'HE HEAVY ATOM EFFECT Phosphore s cence Theory Nature of the Heavy Atom Effect. HEAV Y ATO M E F FEC T S: THE POLY N UCL EA R ARO M A1l 1 I C HYD R OC ARB O N S Introductio n Experim e nt a l Equip m ent. . . . . Experim e nt a l Procedu r e Result s and Di s cu ss ion . . . HEAVY ATO M E PF 'EC'l'S: T HE T RYP 'l'O PHAN I'1E l'ABOLIT E S Introduction . . . Experi me ntal Equi pme nt Experi m ent a l Procedu r e . Result s and Discu ss ion. SU I'1I' 1ARY . . . . iii ii V vi 1 6 6 17 23 23 25 27 30 68 68 69 70 71 92

PAGE 4

Page BIBLIOGRAPHY 95 BIOGRAPHICAL SKETCH. 99 iv

PAGE 5

Table 1. LIS'l' OF TABLES Phosphorescence Characteri st ics of Polynuclear Aromatic Hydrocarbons in Ethanol at 77K Page 31 2. Effect of Sodium Iodide and Ethyl Iodide on Phosphorescence Intensity of Polynuclear Aromatic Hydrocarbons 48 3. Comparison of Phosphorescence Limits of Detection for Polynuclear Aromatic Hydrocarbons in Ethanol and in EEI at 77K. 55 4. Measurement of a Binary Mixture of Triphenylene and 2,3-Benzfluorene. 5. Phosphore sc ence Ch ara cteri sti cs of . 59 Tryptophan Metabolites in Ethanol at 77K. 72 6. Effect of Ethyl Iodide on the Pho sp horescence Intensity of Tryptophan metabolites. 84 7. 8. Compari s on of Phosphorescence Limits of Detection for Tryptophan Metabolites in Ethanol and in EEI at 77K Recovery of Kynurenic Acid from Urine V 85 88

PAGE 6

LIST OF FI-GURES Figure Page 1. Schematic representation of molecular energy levels showing the ground singlet state, S 0 the first excited singlet state, S*, the lowest triplet state, T 0 and the important transitions which occur between them. 9 6. Phosphorescence excitation and emission spectra of naphthalene in ethanol at 77K 33 3. Phosphorescence excitation and emission spectra of phenanthrene in ethanol at 77K 34 4. Phosphorescence excitation and emission spectra of retene in ethanol at 77K. 35 5. Phosphorescence excitation and emission spectra of triphenylene in ethanol at 77K 36 6. Phosphorescence excitation and emission spectra of 1,2-benzanthracene in ethanol at 77K 37 7. Phosphorescence excitation and emission spec t r a of chrysene in ethanol at 77K 38 8. Phosphorescence excitation and emis s ion spectra of 3,4-benzpyrene in ethanol at 77K 39 9. Pho s phorescence excitation and emis s ion spectra of 1,2,3,4-dibenzanthracene in ethanol at 77 OK. 40 10. Phosphorescence excitation and emission spectra of 1,2,5,6-dibenzanthracene in ethanol at 77 oK 41 11. Phosphorescence excitation and emission spectra of 1,2,7,8-dibenzphen a nthrene in ethanol at 770K. 42 12. Phosphorescence excitation and e m ission spectra of coronene in eth a nol at 77 K 43 vi

PAGE 7

Figure Page 13. Phosphorescence excitation and emission spectra of acenaphthene in ethanol at 77K 44 14. Phosphorescence excitation and emission spectra of 1,2-benzfluorene in ethanol at 77K 45 15. Phosphorescence excitation and emission spectra of 2,3-benzfluorene in ethanol at 77K 46 16. Effect of ethyl iodide, A, and sodium iodide, B, on the phosphorescence intensity of 1,2benzfluorene (104 M) in ethanol-iodide solvent at 77K 50 17. 18. 19. 20. 21. 22. 23. 24. Absorption of ethanol-ethyl iodide, A, ethanol-propyl iodide, B, ethanol-methyl iodide, C, each 4/1, V/V, versus ethanol. Analytical curves for 1,2,3,4-dibenz anthracene in ethanol, A, and in EEI, B. Phosphorescence emission spectrum of coronene (104 ~) in ethanol, __ and in EEI,-, at 77 o1c Phosphorescence emission spectrum of a mixture of triphenylene, and 2,3-benzfluorene (both 10: 4 ~) in ethanol,_, and in EEI,-, at 77 K. Phosphorescence emission spectrum of 1,2-benzanthracene (104 M) in ethanol~ and in o' __ EEI at 77 K. Phosphorescence emission spectrum of a mixture of 1,2-benza_~thracene, and 1,2,3,4-dibenz( -L~ ) anthracene both 10 in ethanol, __ and in EEI,-, at 77K Phosphorescence emission spectrum of a mixture of 1,2-benzfluorene, phenanthrene, and 4 acenaphthene (all 10~) in ethanol, __ and in EEI,-, at 77K Semilogarithmic decay plot for 2,3-benzfluorene in ethanol, A, and in ethanol which was 0.075 Min ethyl iodide, B vii 52 54 56 58 60 61 63 64

PAGE 8

Figure Page 25. Phosphorescence exci tatio:r;i. and emission spectra of quinaldic acid in ethanol at 77K. 73 26. Phosphorescence excitation and emission spectra of kynurenic acid in ethanol at 77K. 74 27. Phosphorescence excitation and emission spectra of xanthurenic acid in ethanol at 77K 75 28. Phosphorescence excitation and emission spectra of indican in ethanol at 77K 76 29. Phosphorescence excitation and emission spectra of indole-3-acetic acid in ethanol at 77K 77 30. Phosphorescence excitation and emi ss ion spectra of 3-indolepyruvic acid in ethano l at 77K 78 31. Phosphorescence excitation and emi s sion spectra of anthranilic acid in ethanol at 77K 79 32. Phosphore s cence excitation and emi s s ion spectra of 3-hydrox-yanthranilic acid in e t hanol at 770K 33. Phosphorescence excitation and emi ss ion of nicotinic acid in ethanol at 77K. 34. Phosphorescence excitation and emi ss ion of quinolinic acid in ethanol at 77K s p ectr a spectra 35. Phosphorescence emission spec t rum of kynurenic acid (10-L~I'.!) in ethanol, __ an d in EEI at 77 K viii 80 81 82 87

PAGE 9

CHAPTER I INTRODUCTION The use of pho sp horescence emission spectra for identification of organic molecules was fir s t suggested by ~ewis and Kasha 21 in 1944. In 1957, Keirs, Britt, and Wentworth 20 evaluated phosphorimetry as a quantitative analytical technique. They reviewed the theory of phosphorescence, and discussed many of the factors which i affect the intensity of the phosphorescence e m ission. Freed and Salm re 12 in 1958 applied phosphorimetry to the measure ment of variou s indole derivatives, and co mpared the sensitivities obtained with those of fluorimetry. Results indicated that pho sp horimetry was ten times more sensitive for many of these biolo g ically important compounds. In 1962, Parker and Hatchard 32 con sidered the possibility of using phosphorescence measurements in che mi c a l analysis. They concluded that phosphorimetry would be most u se ful when fluorescence at roo m temperature is insufficiently sensitive or specific. '1 1 he work described in this dis se rtation i s a continuation of ba s ic studies in pho sp horimetry initiat e d in this laboratory six ye ars ago. In 19 63 Winefordner 1

PAGE 10

2 and L atz 52 described construction of a spectrophosphorimeter from la b oratory co mpon ents, and applied this instrunent in the determination of aspirin in blood. The applicabili t ;s t of phosphorimetry to trace a..1J.alysis in biological fluids was clearly demonstrated and effor t s were directed toward i mp rovin g technique s a.'1.d perfecting new applications in this area. The availabi.l i ty of co :::nmer~ i aJ. instrumentation* made it pos sibl e to deter mine the :phosphorescence characteristics of a lar ge number of con r pou n ft.:::j, tLus eli. ::n.i D.ating one of the major stu mb ling blocks to wid.c s_?ree.d accept2 ~ nce of phosphori::netry as a rou-cine i':tethod--lac j : of puo::.i ::::c.e,:;. exci ta tion and emiss i on spectra a.nd decay ti:nes Mu~h of this work in the field of applications has recen t ly been reviewed by Winefordner, McCarthy, and St. John. 5 3 A restrictive factor in the pra.ctice of phos phor im etric analysis has been the availabilit7 of only a few suitable solvent media Phos phorescence ~easurem e n~s must be made in a clear rigid matrix at lo 'd te mpe r atu r e usually 77K. Some relaxation of these criteria ~ay be realized us ing the rotatin g sample cell of Hollifield a..>J d Winefordner 15 with which uniformly cracked l ow tewperature In most of the pho sph ore s cence studies re.;orted -:;o date the Aminco-3owma..'1. s De ctro oh oto.f luo ro me-r:;e r with L':l.inco Keirs phosphoroscope attac hr:lent-(Ame r ican Instru.::ien-ti Co ., Inc., Silver Sprin g rl arylar..d) has been used

PAGE 11

3 glasses can be measured reproducibly. Winefordner and 5l~ St. John have reported a number of pure solvents and solvent mixtures which are suitable for phosphorimetry, but only ethanol, and EPA, a mixture of diethyl ether, iso pentane, ethanol, 5/5/2, V/V, have been used extensively. Possibly because of this solvent restriction, the influence of environment on a solute molecule e mitt ing phosphorescence has received little study in analytical phosphorimetry. It is interesting to note that i n fluori m etry, the effects of environment have long been reco gn ized, and investigations of pH dependence, quenching, and comple xati on, have con tributed greatly to the versatility of this technique. Such environmental effects in fluorimetry have been recently reviewed by Wehry.50 The present re search attempts to develop the analytical potential of a specific environme ntal effect for application in phosphorimetry. Pho sphores cence transitions are electronic transitions involving ch anges in multiplicity, and are forbidden quantum mechanically. Heavy atoms, halogens for e xamp le, introduced as molecular sub stituents or cont a ined in the s olve nt increase the transition probability of multiplicity-forbidden tran sit ions in the perturbed molecule, thus enhancing phosphoresce n ce emission. This perturbing action of heavy ato ms i s de signated the he avy atom effect, and i s categorized

PAGE 12

4 as internal for heavy atom substituents, and external for heavy atoms in the solvent. Thiq phenomenon has been widely studied by molecular spectroscopists, but little atte mp t has been made to exploit it in analytical procedures. Only the work of Zander, 56 which appeared in the literature during the course of this investigation, has been analytically oriented. The emphasis in this dissertation is on achieving better sensitivity and selectivity in phosphorimetry throu gh use of solvents which contain heavy atoms. The heavy atom species employed are the alkyl iodides and inorganic iodides which form a cle a r rigid glass when mixed with ethanol in cer tain proportions. Some experiments involving char ge -tran sfer comple xation are also reported because it is t his type of interaction which is often invo ked to explain the heavy atom effect. The ac ceptors, tetracyanoethylene, 1,3,5-trinitrobenzene, 2,4,7trinitro-9-fluoreno ne and 9(dicyano m ethylene )2,4,7-tri nitrofluorene, which form co mplexes w:Lth aromatic donor s are used as the perturbing species in this inve stigati on. The compolu1 ds selected for the perturbation studies were chosen for several rea sons Th e polynuclear atomatic hydrocarbon s were initially i nvestigated because these were previou sly used in most of the theoretical studies, and, for many, enh ancement of sensitivity was required before phosphorimetry could be app li ed to measurement o f

PAGE 13

them at trace levels. Also, many of these compounds are carcinogens, and their determination is of biological interest. The tryptophan metabolites were chosen for their structural features, and for their biological importance. 5 The basis of the heavy atom effect is considered briefly in Chapter II. This chapter reviews some of the experimental studies which have elucidated the nature of the.heavy atom effect, and suggested it s application in phosphorimetry, as well as the theoretical aspects. In succeeding chapters, the experi mental results obt ained in analytical studies involving the polynuclear aromatic hydrocarbon s (Ch apter III) and the tryptophan metabolites (Chapter IV) are discussed. Emphasis in these investig ations is on improve ment of the measurement step through use of the he avy atom effect, rather than development of a complete analytical procedure for dete rmina tion of these compounds in biological materials. However, in the metabolite study, estimation of kynurenic acid and xanthurenic acid in urine is co nsidered T his latt er work wa s undertaken to evaluate the comp atability of phosphor imetry with heavy atom solvents and a single step thin-layer chromato graphi c separation.

PAGE 14

CHAPTER-II PHOSPHORESCENCE AND THE HEAVY ATOM EFFEC'l' Phosphorescence Theor~ Phosphorescence, as used in this dissertation, refers to a phenomenon in which photoluminescence is emitted by a molecule after absorption of light. A typical aromatic molecule has a singlet ground state (all electrons with paired spin) and several singlet and triplet excited states. In general, only the lowest excited state of a given multiplicity is capable of emission. Emission from the lowest excited singlet state is fluorescence; emission from the lowest triplet state is phosphorescence. Fluorescence may be directly excited by absorption of radiant energy of frequencies within the norm a l absorption band of the molecule. Similar direct excitation of phosphorescence is theoretically not possible due to restrictions imposed by the interco m bination selection rule which forbids radiative transitions between states of dif ferent multiplicity. In order for pho sph orescence to occur, the triplet state must be popul at ed by radiationless transitions from the excited singlet state, a process called 6

PAGE 15

intersystem crossing. This process is also forbidden quantum mechanically, and is only made possible through spin-orbit coupling, or throu g h some perturbing influence. The above relationships may be more readily expressed in a simplified Jablon sk i 17 term diagram, which describes the energy levels in a typical stable organic molecule (all electrons with paired spins). Such a diagram is shown in Figure 1. Each electronic state is assumed to contain numerous vibrational level s but these are not shown for simplicity. By process 1, absorption of radiation, the electron is e xc ited to the lowest electronic excited singlet state, S*, within a period of -15 about 10 seconds. If the electron were excited to a 7 higher e x cited si nglet state, ra pid (lo13 to 1011 seconds) radiationle ss deactivation would return the electron to the lowest excited singlet state. This process in which energy is dissipated as heat is called internal conver si o n. Vi bration al relaxation occurring also in 101 3 to 1011 seconds in sures that any emission from the lowest excited state will originate from the lowest vibration a l l evel o.f that state. Process 2, radiatio nless intern al conversion, S*-S 0 is usu a lly insignificant because the energy separa tion between the e xcit ed singlet and ground singlet is large. Proce ss 3, fluoresce nce, is a radiation.al deacti vation of the excited singlet, S The dec ay time of

PAGE 16

Fig. 1.-Schematic representation of molecular energy levels showing the ground singlet state, S, the first excited singlet state, S*, the 0 lowest triplet state, T, and the important 0 transitions which occur between them. Key to Figure Process Rate Constant 1-Absorption of radiation ka 2-Internal conversion, S*-S 0 kc 3-Fluorescence kf 4-Intersystem crossing, S*-T 0 k. ic 5-Intern a l conversion, T 0 -S 0 k~ 6-Phosphorescence kp

PAGE 17

S*-r-7-"~I To 3 So--'--

PAGE 18

10 fluorescence is generally 109 to 107 seconds, corres ponding to the lifetime of the excited state. Process~, intersystem crossing, S*-T 0 presents an alternative to fluorescence emission. This process involves vibrational coupling between the excited singlet state and the triplet state. The quantum mechanical multiplicity re stri ction on S*-T conversion renders intersystem cros sing almost a 0 miliion-fold less probable than normal vibration a l rela xation or internal conversion. Con sequently the time required for inter system cros sing would be expected to be much lon ger than for vibration a l relaxation. This time, 108 to 107 seconds, is approximately the same order of magnitude as the lifeti me of the excited singlet state, and inter sys tem cro ssing can co mpete with fluorescence emission. Intersystem cro ssing re sults in popul at ion of the lowest vibrational level of the triplet st at e since intern a l conver si on among the vibr ational level s of the triplet manifold is rapid. Process 5, internal con version T -S, i s a con s ider ably more favorable process than S -S 0 0 0 intern al conver si on for two reasons: (1) the ener gy sep arati on between the triplet state and the ground sing le t state is smaller, enhancing vibronic cou pling ; ( 2 ) the lifeti me of the triplet state i s lon g favoring collisional

PAGE 19

11 dissipation of energy.* In fact, for phosphor es cence to be observed, highly viscous or rigid matrices must be e m ploy e d to restrict c ollisional deactivation. Process 6, p h os phorescence, is radiational deactivation of the triplet, T 0 4The decay time for phosphorescence may r a nge fr om 1 0 to 10 seconds, or more, corresponding to the life t i m e o f th e triplet state. Not shown is a ther m al activation s tep, which can elevate an electron from the triplet st at e to th e excited singlet state (intersystem crossin g ) re s ul ting in delayed fluorescence This process will not be i mnor t a n t in a low temperature rigid matrix. According to the approach of McCarthy and \.Jinefordner 23 an ex p ression for the quantum e f fi cien c y o f fluorescence, 0F' and phosphorescence, 0p, can be g iv en i n ter m s of the rate constants of Fi g ure 1. 0F kf = K.+ K + K. C J.C (1) k k. 0p D lC = ~.;-:'2 k. k + k+ p C J.C f C (2) ---~ Strictly spea k ing, a radiationle s s tr a.. ~ s it i o n involvin g a c h m1 g e i!l multiplici t y is called in t er sys t e:n cro ss i ng It is con v e ni em:; to group proce ss es 2 a nd 5 un d e r the he ad i ng of inter na l conver s ion sin ce bo th resul t in r a dia t ion l es s return t o t h e g ro u n d st at e.

PAGE 20

12 The internal conversion and the intersystem crossing S*-T 0 rate constants which determine the quantum efficiency depend on the temperature of the system. They are also influenced to a large extent by environmental factors, for example structure and viscosity of the solvent, and this will become important in the discussion of the heavy atom effect. In the absence of all quenching (kc and k~ are O), the relationships 0F + p = 1 and 0F/0p = kf/kic are obtained, which indicate the complimentary nature of fluorimetry and phosphorimetry. Another parameter which is important in characterizing phosphorescence is the decay time. In ter ms of the experi mentally observed signal, I= I exp-t/lp (3) 0 where I 0 is the intensity at time t = O, I is the inten s ity at time t after terminating the exciting li g ht, and 7'-p is the pho sph ore s cence decay time. In dilute, rigid media where T 0 -S 0 radiationless deactivation i s the predominant quenching mode, 7'p is given by the foll01.ving e q u ati on. I 1 p = k + k' p C (4) This decay time should be distinguished fro m the n a tural phos ph orescence decay time 7'0 = 1/kp which would be observed in the absence of all quenching.

PAGE 21

13 In order to apply phosphoresce n ce m easur emen ts to analysis, the intensity of the luminescence must be rel a t e d to the concentration of the emitting speci e s. This ha s been treated in detail by St. John, McCarthy, and Winefordner. 46 Briefly, the radiant power of p h o s phorescence is given by (5) where Pp is the radiant phosphorescence po w er e mi t t ed, P abs is the rad~ant power absorbed, and ~ p is the pow e r efficiency, the ratio of the radian t power emitted to the radiant power absorbed. The power efficiency i s g i ven b y where 0p is the quantum efficiency and ~ex and ~e m are th e peak wa'V'elengths of the excitation and emission s p ectra, respectively. The radiant power absorbed is g i v en by (7) w here ? 0 is the radiant power :i nc:.dent on t he absorbing S allip le, Pt is the radiant power trans m itt~d t h rou g h the s am nle abs o rbers, a is the molar absorptivity c o efficie nt of t h e absorber, bis the path len g th over which a b sorp t ion occur s a n d c is the molar concentration of t h e ab s orbi ng sampl e T-" .L t h e q uant i ty (2.3a b c ) i s less th an O. O l w hic h is li~el y f o r dilute analyti c al s olu t ion s

PAGE 22

and = P 2. 3abc 0 Pp= P 2.3abc~ 0 p l'-1(8) (9) which gives th~ linear relationship betvrnen phosphorescence radiant power (intensity) and concentration. It has been mentioned that spin-orbit couplin g _plays a predominant role in the phosphorescence process. S p in orbit coupling i.s si g nificantly influenced by the heavy atom effect and for ti1is reason is considered in 3 ome clet.2. il. F9r a ~ore thorou g h discussion, see ~he C) 0 Ol :. n ~ r ..._ ._: Condon and Shortly.5 In atoms, the interaction of the electric a l and. magnetic fields of the different electrons d e t er s ines the final pattern of energy levels. The motion of a nega tively charged electron about a nucleus is g overned by electrostatic attraction. This motion of th e elec t ron als o produces a magnetic field. Such fields are asso ciated. wit h the orbital m.otion of the electron .Similar fields are associated with its spin. In light atom s the 2agnetic fielc resulting from the orbital motion of th e elec t ron about the nucleus is relatively small. In such cases, the spin angular I!lO I'.'.sn --:;a of different electrons cou:pl$ to .form a total spin momentum designated S, and t he oroi tal momenta couple to form a total orbital angular momentum designated

PAGE 23

15 L, both by electrical mechanisms. Magnetic interaction of the spin and orbital momenta gives rise to a resultant total a..ngular momentum, J, for the atom. This ir:i the Russell Saunders coupling-scheme. In heavy atoms, the ma g netic field which results from the orbital m otion of the electron about the large nuclear charge becomes im p or tant Thif, magnetic field serves as a vehicle for interactio n bet1:reen the spin and orbital momenta to yielci a resultant j for each electron. The individual j's then in te ract electrically to form a total J for the atom. This type of cou p lin g called jj or spin-orbit coupling. In molecules, as well as atoms, spin -orbi t coupling enhances the trax1.si tion probability of 2.inglet t ri :91.et multiplicity-forbidden transitions -chrou g h 3 ingle t -trip 1et raixing. Sing l et-triplet mixing implies an ffin p urj .t7 11 o f state which can ar:..se in the following way. A stable organic nolecule in its ground s-ca t e has all electro~s, including the two of highest energy, arra nged in orbitals with paired apins; that is, the multiplicity of ~he sta ~e is 1, a singlet state. In the excited s inglet state, ,..., v the configuration of paired spins is naintai~ed, one electron occupying its original orbital, the o~her electron occupying a higher formerly unoccupied orbital. In the triplet state, T 0 a similar situation prevails, but the two electrons have parallel spins aI'-d. the state is sli ght ly

PAGE 24

16 lower in energy than the excited singlet. The singlet state can be changed into a trip~et state, or vice versa, if a force is applied which changes the direction of the spin magnetic moment of one of the electrons. The attractive nuclear field of atoms in the molecule can serve as the perturbing force which affects this change through the process of spin-orbit coupling. The result is singlet-triplet mixing, and a relaxation of the multi plicity restrictions for transitions between the perturbed states. This relaxation will make more permissive both non-radiational intersystem crossings, and radiative phosphorescence transitions, since both are multiplicity forbidden. A quantitative relationship for the triplet-singlet transition probability is given by McClure. 2 ~ The square of the fraction of singlet character in the triplet state is proportional to the probability of the intercombination transition. The fraction of singlet character introduced into the triplet state is also related to the spin-orbit coupling parameter, S '3'-Values vary with the fourth power of the effective atomic number of an atom. The heaviest atom in a molecule can be used to estimate the magnitude of spin-orbit coupling for the molecule. As heavier atoms are introduced, electronic coupling gradually changes from LS to jj, and spin-orbit effects become significant.

PAGE 25

Reviews which have discussed many of these aspects of phosphorescence theory in detail include those of Kasha and McGlynn 19 and Lower and El-S~ed. 22 Nature of the Heavy Atom Effect 17 The basis of the heavy atom effect has been described in the previous section dealing with spin-orbit coupling. An explanation of the effect of halo gen substitution on phosphorescence transition probability follows directly 24 from spin-orbit couplin g theory. In 1949, McClure reported such a correlation for the series fluoro-, chloro-, bromo-, iodonaphthalene, and in addition noted con siderab le decrea s e in pho sph orescence lifetim es for the same series. The h al onaphthalenes have been the subject of extensive investi ga tion by Pavlo p oulos and El-Sayed. 33 They found that the pho sph ore s cence spectra of a halona phthalene con sists of two "sub spe ctra" with one co mp on ent similar to norm a l naphthalene in energy di stributi on and polarization and with a second component polarized perpendicularly to the first. This suggests that halogen substitution not only enhances spin-orbit coupling already present in t he molecule, but also may intro duc e a new perturbation not prese nt in the unsubstituted compound. T his work has 11 recently been reviewed by El-Sayed.

PAGE 26

Investigations of the external heavy atom effect have attempted to characterize the type of interaction required for the perturbing effect of the heavy atom solvent to be felt by a solute molecule. 18 In 1952, Kasha 18 observed that a binary solution of two colorless components, 1-chloronaphthalene and ethyl iodide, was yellow. This color was attributed to an enhanced S 0 -T 0 absorption re sulting from increased spin-orbit coupling in the 1-chloro naphthalene, induced throu g h a simple collisional process. l"IcGlynn and coworkers 27 reported similar external heavy atom effects in phosphorescence studies and concluded that a collisional mechanism was unlikely in the rigid media employed. Instead, they sug g ested th a t charge-transfer interaction between an aromatic hydroc a rbon and an. alkyl halide might result in formation of a weak molecular complex. Graham-Bryce and Corkil1 13 made a si m ilar sug g e st ion. Equilibrium studies by l"IcGlynn and coworkers 28 35 have con firmed the charge-transfer complex hypothesis. l"IcGlynn, Sunseri, and Christodouleas 28 were able to appro x imately correlate the magnitude of the heavy atom effect on a halo aromatic with the electron acceptor strength of a series of alkyl iodide perturbers. The order of electron acceptor tendency, CH 3 I > CH 3 CH 2 I > CH 3 cH 2 cH 2 I > (CH 3 ) 2 CHCH 2 I, w as the same as the order of effectiv eness of the perturbation. Recently, Eisenth a l and El-Sayed, lO a.nd Eisenthal 9 studied

PAGE 27

19 heavy atom effects on phosphorescence in charge-transfer complexes of a more conventional nature. They investigated the complexes formed between haloaromatics and acceptors containing heavy atoms, for example the tetrahalophthalic an.hydrides, and considered the importance of charge transfer states in the s inglet-triplet mixing process. The spin-orbit couplin g nature of the heavy atom effect has been well established, but exactly which rate constants are most significantly aff ecte d remains a question of controversy. The problem arises in trying to distingui s h between effects on the rate constant s kic' kp' and k; (Fi gur e 1) since all three involve multiplicity restricted processes. The increa se in kic under heavy atom perturbation may be indirectly suggested by fluor escenc e quenching results, for example, the quenching of ar o mati c hydroc arbon fluorescence by potassium iodide. 34 The i mp li cation is that under heavy ato m perturbation, intersystem crossing co mpet e s more favorably with the fluorescence emission process. It is also likely that k. will be more J.C sensitive to perturbation thank~ on the basis of vi bration al overlap; the S -T 0 energy separation is usually less than the T 0 -S 0 separation, which facilitates crossover even in the absence of a perturber. Eisenthal and El-Sayed 10 confir med th at the T -S radiationless process is indeed not 0 0 greatly affected by th e heavy atom effect in a charge-transfer co mp lex. McGlynn and co w o rkers 27 have concluded that the

PAGE 28

20 T -S radiationless process is less sensitive to heavy atom 0 0 perturbation than intersystem crossing, but more sensitive than phosphorescence. Siegel and Judeikis~ 3 have reported a conflicting interpretation that the phosphorescence process is more sensitive to the heavy atom effect than either intersystem crossing or T 0 -S 0 radiationless deactivation. Discrepancies in the interpretation of data from heavy atom experiments exist, but certain experimental results can be easily summarized. Addition of ethyl iodide to a solution of an aromatic hydroc ar bon u s u a lly increases the quantum efficiency ratio p/F of the solute. The magnitude of an internal heavy atom effect and an external heavy atom effect are often co mpara ble; the extern a l effect may be more effective if the intern a l 22 effect is already present. Phosphorescence decay times are usually greatly decreased by the heavy atom effect. Decays of aromatics in alkyl h a lide solvents are also non-exponential. McGlynn and coworkers 27 h a ve explained this on the basis of charge-tr ans fer co mp le xat ion. A new species (complex) is for me d which presumably has a short lifetime, and the resultin g decay curve i s the co mb i nat ion of the exponenti a l decay of complexed and uncornplexed forms. The inhomogeneous nature of hali decon taining g l asses h as also been offered as an explanation for the nonexp o nent i a l

PAGE 29

21 decay. 43 In most heavy atom studies, phosphorescence spectra of perturbed compounds vary little from spectra of the un perturbed compounds. Graham-Bryce and Corkil1 13 did report shifts toward lower energy of 200 to 300 cm-l for the phosphorescence emission of coumarin, acid fluorescein, and several nitronaphthalenes measured in an ethanol-ethyl iodide matrix. Similar red shifts have been described by McGlynn and coworkers. 27 Eisenthal 9 has recently noted significant changes in the vibrational character of donor phosphorescence for 1-chloronaphthalene when n-propyl iodide is introduced into the solvent. Some comments should be made concerning the luminescence of charge-transfer complexes formed between aromatic donors (rr bases) and electron deficient acceptors (rr acids) such as the aromatic hydrocarbon-trinitrobenzene system studied by Reid. 36 Complexation affects the donors' luminescence in much the same way as the heavy atom perturbation. An increase in the quantum efficiency ratio p/F and a decrease in pho s phorescence lifetime of the donor usually occurs on complexation. Czek a lla et al. 6 and Christodouleas and McGlynn 2 ascribed the phosphorescence emission from complexes to a slightly perturbed emission of the donor. Recently charge-transfer triplet states characteristic of the molecular complex have been invoked to explain phosphorescence emission si g nificantly diffe r ent

PAGE 30

from that of the donor component. 16 22 41 Schenk and Radke have studied the quenching effect whi9h tetracyanoethylene and other acceptors exert on the fluorescence of aromatic hydrocarbons. The fluorescence quenching mechanism appears to involve enhancement of intersystem crossing in the hydrocarbon. On the basis of these results, the use of charge-transfer acceptors as perturbing species in phosphorimetry is suggested.

PAGE 31

CHAPTER.III HEAVY ATOI'1 EFFECTS: 'l'HE POLYN1JCLEAR AROI'1A'l 1 IC HYDROCARBONS Introduction Fluorimetry has been exten s ively applied in the analysis of polynuclear aromatic hydrocarbon s 4 9 Phosphorimetry has been used only in a few instances. Since these two techniques are co mp liment ary (.0p + .0F 1), it is advantageous that the analyst use the procedure which mi gh t be s t be applied for a gi ven determin ation Consider a ble data are available in the liter ature concerni ng the fluorescence of aromatic hydrocarbons for analytical applications. 49 Tabulation s of phosphorescence excitation and emission wavelengths and decay tim es of a number of polynuclear aromatic hydroc arb ons have also been re ported.42 Much of this data is not useful analytically because of the lar ge v a riation in experimental co nditi ons used. The phosphorescence ch aracteristics and limits of detection re p orted in this ch apter are analytically u sef ul information and s u ggest where pho s phorimetry mi g ht be the preferred technique. The he a vy at01:i studies show how pho sph orescence measurement of several of the hydrocarbons 23

PAGE 32

24can be improved. In some cases, limits of detection may be lowered by a factor of 25. Changes which occur in the phosphorescence spectra of certain compounds in heavy atom solvents can be useful for identification purposes and also may improve the selectivity of measurement. These results extend the versatility of phosphorimetry for hydrocarbon analysis. In addition, the feasibility of using heavy atom perturbation techniques in phosphorimetry is clearly demonstrated. The analytical considerations important in this technique are developed to su gg est the general applicability of such procedures in phosphorimetry. The effect of charge-transfer complexation not involving heavy atoms is also reported. This complex a tion does not a ppea r to be useful in the sa me way as the heavy atom perturb at ion. Mention should be made of the few pho sp horimetric investigations of hydrocarbons which have appeared in the literature. Applications have been chiefly in the area of petrochemistry, or have involved determination of hydro carbon pollutants in the environment. McGlynn, Neely, and Neely 26 reported using pho sph ori metr y and low temperature fluorimetry to determine several hyd roca rbons of petro chemical intere s t. Drush e l and Somm ers 7 used phosphorimetry to ch ara cterize hydrocarbons in petroleum fractions. Zander 55 reviewed the use of phosphorimetry for identifi cation and quantitative deter mi n a tion of hydroc ar bons in

PAGE 33

coal-tar.. The recent monograph by Zander 57 includes a section on the use of phosphorimetry in the determination of impurities in polynuclear aromatic hydrocarbons. The heavy atom studies of Zander 56 have already been mentioned and these will be discussed in detail later. Sawicki and Pfafr 39 recorded phosphorescence spectra directly from thin-layer chromatograms to identify hydrocarbons in air poliution studies. Experimental Equipment 25 Apparatus.-An Aminco-Bowman spectrophotofluorometer with an Aminco-Keirs pho sp horo s cope attachment, a 150-watt xenon arc lamp, and a potted RCA 1P28 multiplier phototube (American Instrument Co., Inc., Silver Spring, Maryland) was u se d for all pho sp hore s cence measurements. Pho spho res cence intensity readin gs were taken directly from the photometer unit supplied with the instrum ent Pho sph ores cence excitation and emission spectra and decay ti mes were recorded on an X-Y recorder. A fluore s cence spectrophoto meter with pho s phoro s cope attachrnent, Fluorispec, Model SF-1 (Baird-Atomic Inc., Cambrid ge Massachusetts) was used for comparison purpo s es. A D K 2 spectrophotometer (Beck man Instruments Inc., Richmond, California) was u sed to measure the absorbance of the he a vy atom cont aining solvents.

PAGE 34

Reagents and materials.-The hydrocarbons were obtained from the following commercial sources: naphtha lene, anthracene, phenanthrene (Distillation Products Industries, Rochester, New York); retene, triphenylene, 1,2,3,4-dibenzanthracene, 2,3,6,7-dibenzanthracene, 1,2,7,8-dibenzphenarithrene, coronene, perylene, 1,2,3,4dibenzpyrene, 3,4,8,9-dibenzpyrene, 1,2-benzfluorene, 2,3~benzfluorene (Kand K Laboratories, Inc., Plainview, New York); chrysene, pyrene, naphthacene, acenaphthene 26 (City Chemical Co., New York, New York); 1,2-benzanthracene, 3,4-benzpyrene, 1,2,5,6-dibenzanthracene, 20-methyl cholanthrene (Nutritional Biochemicals Corp., Cleveland, Ohio). Charge-tran s fer acceptors were obtained from the following sources: tetracyanoethylene (City Chemical Co., New York, New York); 2,4,7-trinitro-9-fluorenone, 9(di cyanomethylene)2,4,7-trinitrofluorene, 1,3,5-trinitro benzene (Distillation Products Industries, Rochester, New York). The majority of the above reagents were used as received. Alcohol solvent was purified by distill at ion using a five foot vacuum jacketed and silvered column with a reflux ratio of 20 to 1. Ethyl iodide (Fi s her Scientific Co., Pittsburgh, Penn sy lvania) and methyl and propyl iodide (Distillation Products Indu s tries, Rochester, New York) were purified by passa g e throu gh a column of activated silica gel in a dark room and were st ored ov er co pper away

PAGE 35

from room light. Sodium iodide was reagent grade (J. T. Baker Chemical Co., Phillipsburg; New Jersey). Experimental Procedure Phosphorimetry of hydrocarbons.-Stock solutions of the hydrocarbons in ethanol were prepared in the concen tration range 103 ~ to 104 ~ depending on their solubilities. The stock s olutions were stored at 20C, and were successively diluted with ethanol as required. Phosphorescence excitation and emission spectra were obtained for the ethanolic solutions of the hydrocarbons at 77K, and were uncorrected for instrum ental respon s e. 51 Spectra were recorded at both hi gh and low concentr ati on s since phosphorescence e xcitati on spectra are known to vary somewhat with concentr ati on. 40 Variou s slit arrangements were us ed according to the m anufa ctur er 's li terature 1 Spectra s hown in this dissertation were obtaine d u sing the 27 Aminco instrument and hydrocarbon solut ion s of intermediate ( -4 ) concentration 10 ~. In most ca ses the 1 mm slit was placed in front of the multiplier phot otub e to give suitable resolution. Many spectra (not shown) were recorded on the high resolution Baird instru ment with a spectral bandwidth of 2 mu for compari s on. Analytic al curve s ( phosphorescence inte nsity si g nal versus sample co ncentrati on) were det e r mi ned for each

PAGE 36

28 hydrocarbon by measuring successive dilutions of the stock solutions. All precautions of cleanliness were observed as 1 d 48 L / 1 previous y escribed. imits of detection in ug m were obtained using the slit program, 4,3,3,4,3, where the numbers indicate slit width in millimeters at various positions in the instrument. 1 The limit of detection was determined using a graphical extrapolation procedure. 53 A stock solution was diluted well below the limit of detection, and a straight line was drawn through the resulting background points. The limit of detection was defined as that concentration correspondin g to the inter section of the background line with the extension of the linear portion of the analytical curve. A com p ound was con s idered to be non-phosphorescent if a lo3 M solution gave le ss than a 50 per cent full scale meter deflection at the most sensitive instrumental setting s Decay times were measured for hydrocarbon concen trations lying on the linear portion of the analytical curve. With a reading of approximately 90 units on the microphotometer, the exciting radiation was termin ated using a manual shutter, and the phosphorescence decay was traced u s ing the tim e b ase of the recorder. Decay times were determined from semilogarithmic plots of relative intensity versus time. The response time of the recorder restrict e d decay time measureme:n t to decay times greater than 0.5 seconds.

PAGE 37

Heavy atom studies.-Spectra, analytical curves, and limits of detection were determi~ed for each polynuclear aromatic hydrocarbon in ethanol-ethyl iodide mixture s of various proportions. Phosphorimetric measurements were made within one hour after adding the halide to the analytical solution. The effect of various concentrations of ethyl iodide on the pho sp horescence of the hydrocarbons was measured to deter m ine the halide concentr at ion giving an optimum he a vy atom effect. The enhancing effect of methyl iodid e ethyl io did e, propyl iodide, and sodium iodide were compared. Absorbances of the various solvent mixtures were also measured. Decay times were not deter mined for the compounds in ethanol-ethyl io dide solvent bec a use the decays were generally nonexp onenti a 1. 27 29 Simple mixtures of hydrocarbons were studied to evaluate the selectivity of phosphorimetry with ethanol and ethanol ethyl iodide solvent. Char ge-t r ansfer co m plexes .-S to c k s olutions of tetra cyanoethylene, 1, 3, 5-trini trobenze ne 2, LI 7-tri ni tro-9fluorenone, and 9(dicyanomethylene)2,4,7-trinitrofluorene were prepared in ethanol according to their solubility. The pho sph or escence ch a racteri sti cs of the acceptors were deter m ined. Pho sph o res ce nce spectra of mixtures of the acce pt ors with hydrocarbon donors were recorded. Little

PAGE 38

useful information was obtained and no extensive investi gation was undertaken. Results and Discussion 30 Phosphorescence excitation and emission wavelengths, decay times, and limits of detection for the hydrocarbons in ethanol are reported in Table 1. These results in combination with the spectra shown in Figures 2 throu g h 15 are useful in predicting whether phosphorimetry would be applicable to a particular problem in hydroc a rbon analysis. Spectra sho,m are uncorr e cted for instrument a l respon s e, 51 and are included mainly to indicate b and shape and width which can be important in limiting the selectivity of phosphorescence measurements. From the wavelengths of the peaks, it should be possible to excite and to measure selectively many of the compounds in the presence of one or more interfering compound s This selectivity may not be fully realized due to the bro a d bands of inten se ly phosphorescent interferents, such as triphenylene. Phosphore s cence spectra of pyrene are not sh own as this compound cont a ined significant impurities. Problems en countered in the purification of pyren e have be en discu ss ed at length by Srinivasan, Kino s hita, and McGlynn.~~ It should be not ed that n ap hth a cene and 2,3,6,7-dibenzanthracene gave an inten se phosphorescence e m i ssio n which was the same

PAGE 39

TABLE 1 PHOSPHOR E SCE N CE CHARACTERISTICS OF POLY NU CLEAR AROMATIC HY D R OC ARBONS IN ETHANO L AT 77K Hyd roc a r b o n Ex ci ta t i on Em i s sion Dec ay Timeb,c Limit of De t ectionb Maximaa Maximaa ( mu 2 ( mu 2 (secon d s2 (ugLml2 Naphtha l e ne 2 90 2 35 50 5, 4 7 5 ,540 2 .. 6 0 .. 05 P he n ant hrene 2 9 0 495 460 53 5 3 .8 0.003 Ret ene 305 47 0 50 7,545 3.4 0 .. 001 T ri phenyle ne 2 9 0 457, 435 ,4 9 0 1 6 .2 0.0004 1 2 Benzant hr a cene 310,275 4 95 ,5 3 0 1.4 0 .. 07 C hrysene 27 0 31 0,325 5 00 5 4 0 2.2 0.009 Py r ene 3 3 2 ,2 55 ,3 00 595 61 0 0 .. 5d 0,.4d 3 4 Be n zp yrene 330 285 51 0 5 5 0 2.3 1.5 1 2,3 4 D i ben zanthra c en e 295 33 0 56 0, 600 0.9 o.4 1 2,5 6 Dibenz an t h r acene 305 ,3 30 34 0 5 50 59 0 1.5 0.01 1 2 7,8 Dibenz p henanthren e 2 90,33 0 5 0 0 ,535 2.5 0.002 Co r o nene 3 1 5,350 560 530 9 .6 0.0009 Acenaphthen e 305 23 5 4 7 5 510 ,550 3 .0 0 .. 05 1, 2 Benzflu or e ne 3 1 5 275 495, 5 35 2 .. 7 0.04 \.>J !-'

PAGE 40

Table 1 Cont'd. Hydrocarbon 2,3-Benzfluorene Naphthacenee 2,3,6,7-Dibenza.Y).thracenee Anthracenef Perylene f 20-I"Iethylcholanthrenef 1,2,3,4-Dibenzpyrenef 3,4,8,9-Dibenzpyrenef Excitation I"Iaximaa (mu) 320,280 Emission I"Iaximaa (mu) 498,538 b c b Decay Time' Limit of-Detection (seconds) (ug/ml) 2.8 0.02 aI"Iost intense peak wavelengths are listed first; wavelengths are uncorrected for instrumental response;51 precision in wavelength observation is+ 5 mu. bl"Ieasurements were made at most intense peak wavelength. cPrecision of measurement is+ 0.1 second. dApproximate value since sample contained impurities. ePhosphorescence was measured, but is attributed to a photodecomposition pro duct.3 f No phosphorescence was measured.

PAGE 41

>1en z LLl 1z UJ > e( ..J LI.I c:: 200 250 300 350 400 450 500 650 WAVELENGTH (mJJ} Fig. 2.-Phosphorescence excitation and emission spectra of naphthalene in ethanol at 77K.

PAGE 42

UJ 1z I.LI 0::: 200 50 300 350 400 450 500 65 WAVELENGTH (mJ.J} Fig. 3.-Phosphorescence excit a tion and emission spectra of phenanthrene in ethanol at 77K.

PAGE 43

c:: 300 400 450 500 65 WAVE LENGTH ( mJJ) Fig. ~.-Phosphorescence excitation and emission spectra of retene in ethanol at 77K.

PAGE 44

300 600 650 WAVELENGTH emu) Fig. 5.-Phosphorescence excitation and emission spectra of triphenylene in ethanol at 77K.

PAGE 45

LLJ 1z 200 300 400 450 500 65 WAVE LENGTH (mu) Fig. 6.-Phosphorescence excitation and emission spectra of 1,2-benzanthracene in ethanol at 77K.

PAGE 46

300 400 WAVELENGTH 500 650 Fig. 7.-Phosphorescence excitation and emission spectra of chrysene in ethanol at 77K.

PAGE 47

en z UJ 1z 350 400 WAVELENGTH (mll) 550 600 Fig. 8.-Pho~phorescence excitation and emission spectra of 3,~-benzpyrene in ethanol at 77K. 650

PAGE 48

300 350 400 450 500 WAVELENGTH (mll) Fig. 9.-Phosphorescence excitation and emission spectra of 1,2,3,4-dibenz anthracene in ethamol at 77K. 650 .j::' 0

PAGE 49

UJ c::: 00 250 00 350 400 450 500 550 WAVELENGTH (mJ.J) Fig. 10.-Phosphorescence excitation and emission spectra of 1,2,5,6-dibenz anthracene in ethanol at 77K. 650

PAGE 50

,. 200 300 350 400 WAVE LENGTH 450 (mJJ) 500 Fig. 11.-Phosphorescence excitation and emission spectra of 1,2,7,8-dibenz phenanthrene in ethanol at 77K. 650

PAGE 51

200 250 300 350 400 450 550 650 W AVELE N GTH (mu) Fi g 12.-Pho s phorescence excit a tion and emission spectra of coronene in ethanol at 77 K \>I

PAGE 52

UJ 0:: 200 250 300 350 400 450 500 650 WAVELENGTH (mu) Fig. 13.-Phosphorescence excitation and emission spectra of acenaphthene in ethanol at 77K.

PAGE 53

er:: 300 400 450 500 650 WAVELENGTH (m) Fig. l~.-Phosphorescence excitation and emission spectra of 1,2-benzfluorene in ethanol at 77K.

PAGE 54

200 250 300 350 400 WAVELENGTH 450 (mJJ) 500 650 Fig. 15.-Phos ph orescence excitation and emission spectra of 2,3-benzfluorene in ethanol at 77K.

PAGE 55

for both co mpou.1.'lds Clar and Zan der 3 have expl a in e d th i s naphthacene pho sphorescence, which was als o re p orted by Reid, 37 in terms of emission from a p h otod e co mp osition product. It seems li kely th a t 2,3, 6 ,7-dib enzan thracene decomposes to give the same ph o sphores cent product. Phosphorescence deca ys for a ll co mpounds e xcept 1,2-benz anthr a cene and pyrene we re e xpon en t i a l. Significant differences exi st among the decay times, for ex a~p le, 16.2 seconds for triph enylene and 0 .9 seconds for 1,2,3,4d .b th ~ht 1 d h 4 7 i enzan racene so v a 01 m e-r es o~ve p~ osp orime v ry mi g ht be applied to the measurement of such co mpounds in simple mixtures. Lo w li m it s of dete ction found for many 47 of the hydrocarbon s su gg est that pbosp horim etry could te used to measure these co mp ounds at trace levels. Compounds w h ich are rat h er inse ns i tiv e by fluorime t ry, 49 such as triphenylene, retene, an d phenantb~en e, can be measured at much lower concen t ra t i ons usin g ph osphorimetry. Several intensely fluorescen t hydrocarbons for example, perylene, and 2O-~ethylcholanthrene s ncw ed negligible phosphorescence. This gives phosphorime try g ood selectivity for measurement of a species of interest in a solution containing these co mp oun ds as in terferenr,s The effect of sodium iodi de and ethyl iodide on the phosphorescence intensity of the hyd rocarbons is shown in Table 2. The pho sphores ce nce in tensity signal, R 0 was

PAGE 56

48 TABLE 2 EFFECT OF SODIUM IODIDE AND E'l 1 HYL IODIDE ON. PHOSPHORESCE N CE INTENSITY OF POLYNUCLEAR AROMATIC HYDROCARBONS Hydrocarbon a R/Rob Sodium Iodide Ethyl Iodidec 0.63!:! 0.63M 1.25M 2.50!:! Naphthalene 4.0 0.29 0.21 0.12 Phenanthrene 1.5 0.64 0.40 0.22 Retene 1.0 1.1 1.1 0.77 Triphenylene 0.75 0.19 0.099 0.048 1,2-Benzanthracene 1.2 1.2 1.3 1.4 Chrysene 1.4 1.2 1.4 2.1 Pyrene 1.4 1.4 1.8 2.5 3,4-Benzpyrene 1.6 1.5 2.2 3.5 1,2,3,4-Dibenzanthr a cene 3.1 2.9 4.7 7.7 1,2,5,6-Dibenzanthr a cene 1.2 1.2 1.2 1.3 1,2,7,8-Dibenzphenanthrene 1.0 1.1 2.1 2.9 Coronene 3.0 2.7 3.9 6.0 Acenaphthene 3.4 0.91 0.86 0.74 1,2-Benzfluorene 5.6 6.4 9.8 13. 2,3-Benzfluorene 7.7 8.7 15. 25. aHydrocarbon concentration was appro x imately 10-L~M in ethanol-iodi de solvent. bR 0 is ph o s phorescence inte nsity of hydrocarbon in ethanol; R is phosphore s cence intensity in ethanol-iodide solve n t. Precision in int ensity me as urements is+ 5 per cent. Wavelengths were adjusted. to give maximum signal in each c ase cEthyl iodide concentrations expressed on a volu me basis with ethanol at 20 are : 0 .. 63!:!-19/1,V/V, ethanol ethyl iodide; l.25!:!--9/1,V/V, ethanol-ethyl iodide; 2.50!:!-4/1, V /V, ethanol-ethyl iodide.

PAGE 57

measured with the species in ethanol; the phosphorescence intensity signal, R, was measured with the species in the iodide solvent. The data in Table 2 were obtained by adjusting the excitation and emission wavelengths to give the maximum signal for each solution. The concentrations 49 of ethyl iodide are expressed on a molar basis to permit comparison with the sodium iodide results. The increasing molar concentrations of ethyl iodide correspond to 19/1, 9/1, 4/1, V/V, ethanol-ethyl iodide mixtures. Ethanol-ethyl iodide solutions in these proportions formed clear rigid glasses at 77K. Concentrations of sodium iodide in ethanol which could be obtained were limited by solubility to less than 1 M. In addition, ethanol-sodium iodide solutions more concentrated than 0.63 regularly form e d crac k ed glas s es when cooled to 77K. Figure 16 indicates the variation of the phosphorescence intensity of 1,2-benz fluorene as the ethyl iodide and sodium iodide concen trations are increased. The ethyl iodide plot was typical of those obtained for methyl iodide and propyl iodide, and indicated an optimum iodide concentr a tion of approxi mately 2.5 (4/1, V/V, ethanol-ethyl iodide, henceforth designated EEI). Methyl iodide gave sli g htly greater enhance m ents of pho sph orescence as su gg ested by McGlynn, Sunseri, and Christodouleas, 28 but solutions greater than 1 Min methyl iodide cracked con s istently. There was

PAGE 58

50 o...-------r---------,.------r-------,----.... >..... en z UJ 0 .... ..... zo ..-t UJ > ..... ct _J UJ 0:: ..-41?..-..----L-----L ----~----'---------' 1 2 3 4 5 0 CO N CE N TR ATION O F IODIDE (moles per liter) Fig. 16.-Effect of ethyl iodide, A, and sodium i o dide B, on the phosphorescence inte nsity o f 1,2ben zflu o rene (104 n) in ethanol-iodide so lvent at 77K.

PAGE 59

essentially no difference between the enhancing effect of ethyl iodide and propyl iodide. Usable concentrations of sodium: iodide were l_imi ted by solubility and cracking problems. The absorption characteristics of the three alkyl iodides were found to be similar as shown in Figure 17. All absorbed strongly below 3l~Q mu, and this explains the depression of phosphorescence noted for several com pounds (Table 2) which have strong excitation bands below 51 300 mu. At short wavelengths, the solvent competes favorably with the species of interest in absorbing the exciting radiation. Sodium iodide was found not to absorb the inci dent radiation, and some enhancement of phosphorescence was noted for almost all the hydrocarbons ( 1 l 1 able 2). The dis advantages of using sodium iodide (solubility limitations and cracking) rendered it less useful than ethyl iodide except in a few instances. For example, a significant enhancement of naphthalene phosphorescence was obtained in ethanol-sodium iodide (Table 2) while no improvement in sensitivity was possible in ethanol-ethyl iodide. Iodide solutions are light sensitive, but no particular problem with photodecomposition was encountered if measurements were made within one hour after adding the iodide to the analytical solution. Analytical curves for those hydrocarbons showing enhanced phosphorescence determined in E E I were linear

PAGE 60

z 0 .... a. c:::: 0 en CD <( .... z UJ u 0:: L!.J a.. 0 Ci') 0 00 <.D 0 I!') 0 -=:t 0 M C'\I 0 ..-4 0 300 400 WAV EL ENG T H (ffiJJ) F i g 17.Abs o rp t i o n o f e than o l ethyl i o did e, A et h anol-propyl iodide, B ethan olme t hyl io dide C, e a c h 4/1, V/V, versus et h anol. V1 I\)

PAGE 61

over ranges greater than or equal to those obtained in ethanol, and the limits of detection were lowered by an amount predictable on the basis of the enhance m ent ratio, R/R 0 shown in Table 2. Analytical curves are not shown for all the compounds because they were very similar. 53 Figure 18 shows typical analytical curves for 1,2,3,4-di benzanthracene in ethanol and in EEI. Limits of detection obtained in ethanol and in EEI are compared in Table 3. A considerable lowering of the limit of detection was observed for a number of the hydrocarbons which were not particularly low in ethanol. Excitation and emission wavelengths are also listed in Table 3 to indicate certain spectral ch a nges brought about by addition of ethyl iodide to the solvent. The apparent red shift of the excitation wavelengths is primarily a result of the absorption of e x citing radiation by the solvent. Any red shifts in the emission spectra were slight. Noticeable changes in vibration a l structure of the emis si on spectra. did occur, and t h is effect may often improve the selectivity of pho sp horescence measure ments. For coronene and triphenylene measured in ethanol ethyl iodide, the 0-0 emission band became much mo're intense at the exp e nse of the longer wavelen gt h bands. The chan ge in the emission spectrum is shown for coronene in Fi g ure 19. This appears to be an effect which is co mm on to almost all the hydrocarbons but which is significant in only a few.

PAGE 62

> 1(/) z UJ .... z UJ > .... c( ,_J UJ CONCENTRATION (moles per liter) Fig. 18.-Analytical curves for 1,2,3,4-dibenz anthracene in ethanol, A, and in EEI, B. 54

PAGE 63

TABLE 3 COMPARISON OF PHOSPHORESCENCE LIMITS OF DETECTION FOR POLYNUCLEAR AROMA 1 l 1 IC HYDROCARBONS IN ETHANOL AND IN EEI AT 77K Hydrocarb on a Ethanol EEib Excitation Em ission Limit of Excitation Emissi on Maximum Maximum Detection Maximum Maximum (mu) (mu) (ug/ml) (mu) (mu) 1,2-Benzanthracene 310 500 0.07 327 505 Chrysene 308 506 0.009 328 507 Pyrene 332 592 0.4 335 595 3, l.J -Benzpyrene 325 508 1.5 332 510 1,2,3,4-Dibenzanthracene 300 564 0.4 335 564 1,2,5,6-Dibenzanthracene 305 550 0.01 336 550 1,2,7,8-Dibenzphenanthrene 290 503 0 .. 002 335 503 Coronene 315 560 0.0009 342 518 1,2-Benzfluorene 315 500 0.04 320 501 2,3 Benzfluorene 325 500 o.o4 325 500 aOnly hydrocarbons whi ch show an enhancement of phosphorescence in listed.b EEI is the ethanol-ethyl iodide mixture, 4/1, V/V. Limit of Detection (ug/ml) 0,.06 0.003 0.1 0.3 .0.06 0.008 0 .. 0007 0.0002 0 .. 003 0.002 EEI are \J1 \J1

PAGE 64

56 ,1 ,, I I I I I I I I I I t I I I I I I \ 310 Fi.U \. e:c I I 11 320 r: m 1 \ en I I I I I I I I I I I I I I I I I I I I I I /'., I I I I I I I I I I I I \ I / 400 450 5 0 0 550 650 WAV E L ENGTH (mu) Fi g 1 9 -P h o snhor e scence em i ss ion spec trum of coronene LL (1 0 M ) in ethan ol an d in EZ I ,-, at 7 7 K .Ins t r um e nta l sensit i v i ty was r edu ce d b y a f a c t or of a pp ~o x i nate l y 3 to o bta i n spect r um i n EEI Exci tat io n wav el en gt h we.s 310 ::n u in ethan ol and 3 20 r:1 u in EEI

PAGE 65

57 Figure 20 shows how such changes can be useful. Triphenylene caused particular problems in attempting to measure less intensely ~hosphorescent compounds in mixtures. The intense emission of triphenylene was found to extend into the spectral region of emission of many other hydrocarbon s In EEI, the triphenylene emission spectrum was altered so that the maximum intensity was centered at 432 mu instead of at 457 mu, thus permitting measurement of compounds whose spectrum was not changed, for example, 2,3-benzfluorene. In Table 4, quantitative results are presented for measure ment of triphenylene and 2,3-benzfluorene in eth an ol and in EEI. For some hydrocarbons in EEI, spectral b ands changed little in rel at ive intensity, but fine structure became more or les s well defined comp ared to the structure observed in ethanol. Fi gure 21 shows that the structure of 1,2-benzanthracene was considerably more prominent in EEI than in ethanol. Use of the h eavy atom effect to improv e selectivity has alre ady been mentioned for the c ase of triph eny lene. Measurem en t of spectra in ethanol, and then in EEI was useful in characterizin g mixtures of co mp ound s Chan ges in band structure and relative inten sity provided an addition a l basis for identifyin g a co mp ound not available when measurements were made only in ethanol. Fi gure 22 shows th e phosphorescence emission spectrum of a mixture

PAGE 66

LL! 0:::: >":1 t l g triphenylene (i ,, I I I I I I I I I I I I I I I I I \ I I I I I I I I l '1 I I I I I I I I I I I I I I I I \ I I l I I \ I I \.I I 450 2,3-be n z f I uo rene r I I I I 11 I I I I I I I I I I A ex 320 mu I I I I I I I I I I I I I I I I I : I I \ ,.. I I / I I I I I / I I I / I I \ I I I I I I \ I I \ I \ I I \I \ \ \ \ \ \ 500 550 -, \ \ 600 WAVELENGTH (m.U) 650 20 .P h o snhores ce n ce e :J.j_ss..:..on ,:;:lect r u:::i o: ::t TD. i:-:: t uI'e of tri pheny lene a...'1.d 2 3 oenzfluorene ( b o t h L~o, ) '""' -,T 10 1 I in e tha_r1 ol, and in i'.JD.L at 77K : -In s trum enta l sensi t i vity was the sa. u. e for bo t h s pe c t r a J:x ci tation ~-r a vele :ngt h w a s 3 20 :21. u in b o th e tha...'1.ol and ::2:I 58

PAGE 67

TABLE 4MEASUREMENT OF A BINARY MIX'l'URE OF TRIPHENYLENE AND 2,3-BENZFLUORENE 59 Percent Error in Measured Concentration Concentration Ct!) aEEI is the bExcitation in ethanol and 295 cExcitation in ethanol and 325 Triphenyleneb a Ethanol EEI 2,3-Benzfluorenec a Ethanol EEI -7 -4-2 -6 ethanol-ethyl iodide and emission maxima mu, 4-32 mu in EEI. and emission maxima mu, 4-98 mu in EEI. +95 +4-8 mixture, were 290 were 320 +2 -3 4-/1, V/V. mu, Lj 5LJ mu, mu, 4-98 mu,

PAGE 68

UJ 0:: 400 r, 1 e x 308 I "-e x I I 3 22 I \ I \ I I I I I I I I I I I I I I I I I I I I I \ I I \, I I I I I I I I I I f I 4 5 0 5 0 0 5 5 0 W A VEL ENGTH (mlJ) 60 mu mu 6 00 6 50 F ig 2 l e -Ph o sph o res c ence emissi o n spectrum o f 1, 2( -4be nzanthracene 1 0 M ) in ethanol, and in EEI ,-, at 7 7 K. -I nstrumental sensitivity was reduced by a factor o f appr o x imately 1. 5 to o btain spectrum in EEI Excitation wave l ength was 308 mu in eth a nol and 322 mu in EEI

PAGE 69

1,2-b en z a nth r ace n e I \ I I X. ex 3 18 mu I I \ I \ I \ I \ I \ I \ 1,2,3,4 di benzI \ anth ra ce n e I \ '\ I \ I I \ f I I \ I I \ I \ I I / I I \J I \ I \ I I I \J I I I I I \ I I \ \ I \ f \ I \ I \ I I \ I I I / 400 4 50 500 5 5 0 600 6 50 W A V E L ENGTH e mu > F ig 22. Ph o sph o rescence emission spectrum o f a mixture o f 1 2-benzanthr a cene and 1,2,3,4-dibenzanthra c ene ( b o th 104 ~) in ethanol, __ and in EEI at 7 7 K 61 Instrument a l sensitivity was reduced by a factor o f appr o ximately 1.5 to obt ain spectrum in EEI. Ex c itation wavelength was 318 mu in both ethanol and EEI.

PAGE 70

of 1,2-benzanthracene and 1,2,3,~-dibenzanthracene, re corded in ethanol and in EEI. Excitation was at 318 mu for both spectra. In EEI, the 1,2,3,~-dibenzanthracene emission peak appeared at 560 mu, and the band structure 62 of the 1,2-benzanthracene became more pronounced. In Figure 23, the phosphorescence emission spectrum of a mixture of 1,2-benzfluorene, phenanthrene, and acenaphthene is shown. In ethanol, excitation at 320 mu gave a complex spectrum with contributions from the three compounds. In EEI, the spectrum of 1,2-benzfluorene could be clearly re corded. However, excitation at 290 mu in ethanol gave the phenanthrene emission spectrum alone. Phosphorescence decays in ethanol-ethyl iodide mixtures were non-exponential. This effect is shown in Figure 2l.J for the decay of 2, 3-benzfl uorene. The intensity of a phosphorescence signal could decrease if the decay time of a compound became short with respect to the exposure time of the phosphoroscope. 0'Haver and. Wine ford.ner31 have measured the important phosphoro s cope parameters which influence the measured phosphorescence intensity, and found that the expo s ure time for the Aminco phosphoroscope is around 103 seconds. Since the decay times of most of the hydrocarbons were found to be one second or lon g er, the decrease in decay time which occurs in heavy atom solvents should. not adversely affect th e

PAGE 71

63 1, 2 b en z f I u o r e n e I "'ex 320 mu I I I I I I I I I \ I \ I \ I I ,1 I I f I I \ I I ac en ap ht hene I I I I phenanthrene I I I I LU I I I ex: I \ I \J I \ I I \ I \ I \ I \ I \ J -, / ./ \ ,. / 400 450 500 550 600 650 WAVELENGTH ( mJJ) Fig. 23.-Phosphorescence emissi o n spectrum o f a mixture o f 1 ,2-b enzflu or ene phenanthrene and acenaphthene (all l04 M) in e thanol and in EEI ,, at 77 K. ~In strwnental sensitivity was reduced by a factor o f approximately 15 to o btain spe ctr um in EEI Excit ati o n wavelength was 320 mu in both ethanol and EEI.

PAGE 72

> 1-U) z w 1z w > 1<( ...I UJ et:: 0.-------,------~----~---o ,-c 0 ,-c,__ ____ ._ ____ ....._ ____ ..1.___ ___.J 0 1 TIME 2 (seconds) 3 4 Fig. 24.-Semilogaritbmic decay plot for 2,3-benz fluorene in ethanol, A, and in ethanol which was 0.075 Min ethyl iodide, B. 64

PAGE 73

phosphorescence intensity for measurements made with this phosphoroscope. The decay time will rarely be reduced by a factor of 50 in halide solvents. 2 7 The results presented in the above studies are very similar to those reported by Zander, 56 who used EPA-methyl iodide, 10/1, V/V, as a perturbing solvent. Greater pro portions of methyl iodide could not be used because of the cracking problem. In Zander's studies, 56 enhancements in phosphorescence intensity were achieved for such compounds as phenanthrene, because excitation was carried out at wavelengths lying in the long wavelen gt h tail of the absorption band, 345 mu, where solvent absorption was not a significant factor. While more intense phosphorescence signals would be obtained in the halide solvent than in EPA, these signals would not exceed those in unperturbed media if excitation were performed at the excitation maximum of 290 mu. In other respects, the work reported by Zander 56 confirms the u s efulness of the heavy atom effect for pho sp horescence measurement of hydrocarbons. Investigation of the complexe s formed betwe e n the hydrocarbons and various acceptors produced little useful information. The strongest acceptors, 2,4,7-trini tr o-9fluorenone, and 9(dicyanom e thylene)2,4,7-trinitro f luorene, which should exert the largest perturbin g effect phosphoresced and were only slightly s oluble in eth an ol. 65

PAGE 74

66 This made difficult the study of the phosphorescence of the donor components in solutions of donor and acceptor. Tetra cyanoethylene did not phosphoresce, but is a more selective 41 complexing agent for simple aromatic compounds. The weakest of the charge-transfer acceptors, 1,3,5-trinitro benzene, did not phosphoresce and was easiest to work with. An enhancement in phosphorescence intensity was ob s erved forsolutions of pyrene which were 102 in 1,3,5-tri nitrobenzene. The magnitude of this effect was about one third of that obtained with a 102 M pyrene solution in EEI. This enhance ment was peculiar to pyrene and did not occur for any of the other hydrocarbon com p le xes of 1,3,5trinitrobenzene examined. The most significant problem encountered in working with char ge -tra ns fer complexes of this type was th at the lar ge concentr ati on of acceptors needed for effective perturbation was difficult to achieve. In many c ases a cry s talline complex would form when s olu tion s were cooled to 77K. The use of the heavy atom effect is simple and c a n often improve the sensitivity and selectivity of phosphores cence measurements. Ethyl io dide and propyl iodide appear to be equally conv enient for use in perturbing solvents. Sodium iodide has the adv antage of no t reducing the inten sity of the incident li gh t through absorption; however, solubility

PAGE 75

67 limitations and cracking problems render it less useful than the alkyl iodides. The possibility of using heavy metal iodides is also governed by solubility limitations, and would offer no advantage. Ethyl iodide is expensive, but the small quantities needed to affect a large increase in phosphorescence intensity should justify the expense. It seems likely that heavy atom solvents could be used in those applications of hydrocarbon an a lysis where phosphorimetry has already been applied 7 39 55 and that such use would greatly extend the versatility of the measurement procedure.

PAGE 76

CHAPTER IV HEAVY ATOM ~FFECTS: THE TRYPTOPH AN r1ETABOLITES Introduction Phosphorimetry has been shown to be very useful in detecting certain ind.oles and related compounds at t race 12 45 levels. This cha pt er p rese nts the phosphoresce n ce char a cteri?tics of a gr oup of tryptophan metabol i.tes, r,:a ~ 1 y of which have no t been previously studied by phoc.90.ori.: n e -;:;r:y Interest in these co mp ou..'1.ds, e specially kyn ur eni c acid. 3.11.:.1_ xan t hurenic acid, ste m s from their use as monitors in stud y ing arg malfunction in the metabolism o f tr y p~ophan Low limits of detection are obtained for ma ny 0f the metahol i. tes The limit of detection by pho s:pb. or i met:ry for h.7nu.::-enic acid compares favorably wi t h that obtained by A fluorimetry.-Xanthurenic acid can b e measured at nuch lower levels by fluorimetry, but exh i bi ts a 10-fold en haI1cement of phosphorescence when measure8ents are made in EEI. Kynurenic acid and xan.thurenic acid are well suited for heavy atom perturbation studies becau s e their excita tion wavelengths are near 350 mu where reduction of in c ident light intensity by the halide solvent is not si g nifica.'1.t. 68

PAGE 77

The effect of heavy atoms on the phosphorescence of the metabolites differs in some ways.from the effects observed in the hydrocarbon study, and these are discussed in this chapter. 69 Several methods have been used to determine kynurenic acid and xanthurenic acid in urine. Fluorimetric procedures have been used after a lengthy separation of the metabolites on columns of ion-exch a nge resin.3 8 One of the simplest methods has involved a thin-layer chrom at o graphy separation, followed by measurement using absorption spectrophoto me t~y. 14 Phosphorimetry has been used successfully in combin at ion with thin-layer chromato g raphy for determination of nitrophenol, the metabolic product of parathion in urine. 29 The compat ab ility of thin-layer chro mat o graphy and phosphorimetry in heavy atom solvents for determination of kynurenic acid and xanthurenic acid in urine is investi gated in the pre s ent work. It is shovm that adsorption on silica gel can render inoperative the heavy atom effect which is normally ob s erved. Such an effect was encountered with xanthurenic acid. Experim ental Equipment Appar a tu s .-The spectrophosphorimeter has been des cribed in Ch apt er III. Co mmer cial thin-layer chro matography equip men t (Bri nkmann In struments Inc., Westbury, New York)

PAGE 78

70 was used for the separation studies. A 10 ul syringe (Hamilton Co., Inc., Whittier, C 0 lifornia) was used to apply samples to the thin-layers. A long wavelength ultra violet lamp (Blak-Ray Model XX:15, Ultra-Violet Products, Inc., San Gabriel, California) was employed to locate fluorescent spots on the thin-layers. Reagents and materials.-The tryptophan metabolites were used as received (Nutritional Biochemicals Corp., Cleveland, Ohio). Alcohol and ethyl iodide were purified as described in Chapter III. Solvents for thin-layer chromatography were reagent grade. Silica gel G for thin-layer chromatogr ap hy (E. Merck A.G., Darm s t ad t, Germany) was heated at 700C for 12 hours before use to reduce the luminescence back g round.30 Experimental Procedure Phosphorimetry and heavy ato m studies ..:..Procedures described in Chapter III were folloNed in obt ain ing pho sp horescence excitation and emission wavelengths, decay time s analytical curves, and limits of detection for the tryptoph an metabolites. Heavy atom studies were restricted to the optimum ethanol-ethyl iodide mi xture EEI. Thin-layer chrom at o graphy .-The thin-layer chrom a tography se parati on has b een described i n det ail by

PAGE 79

Hill, Summer, and Roszel. 14 Briefly, separation was achieved on silica gel G layers, 20 cm by 20 cm, and 250 u thick, activated for 30 minutes at 110C. A 20 ul sample, either urine containing added metabolite, or a standard ethanolic solution of metabolite was applied to the thin-layer using the 10 ul syringe in increments of 5 ul. Development by chloroform-methanol-acetic acid, 75/20/5, V/V, required 75 minutes. Fluorescent spots were located under ultra-violet li gh t and the adsorbent removed. 71 Ethanol extraction in centrifuge tubes follo wed by centri fugation yielded solutions ready for phosphorimetric an a lysis. During extr a ction, the solutions were frequently stirred and were maintained at 60 in a water bath. Generally only a single extraction with 5 or 10 ml of ethanol was required. When ethyl iodide was used as a perturbin g solvent, this was added to the eth an ol directly before measurements were made. Results and Discu ssi o n Phosphore scen ce excitation and emission wavelengths, dec ay times, and limits of detection for the tryptophan metabolites deter mined in eth an ol are li sted in Table 5. Pho sph ore s cence excitation and emission spectra are shown in Figures 25 throu gh 34. Si gnifi c ant differences in excitation and emission wavelengths make possible the

PAGE 80

TABLE 5 PHOSPHORESCENCE CHARACTERISTICS OF TRYPTOPHAN METABOLITES tN ETHANOL AT 77K Metabolite Excitation Emission Decay Timeb,c Limit of Detectionb Maximaa I"Iaximaa (seconds) (ug/ml) mu mu Quinaldic acid 305,24-5 510, L~75 1.1 0.015 Kynurenic acid 34-8,255 4-58 1.6 0.003 Xanthurenic acid 350,250 4-85 0.9 0.1 Indican 295,235 4-60 3.3 0.03 Indole-3-acetic acid 285,230 4-35,4-10,4-55 7.0 0.008 3-Indolepyruvic acid 302,265 4-25,410 0.6 0.003 Anthranilic acid 350,255 4-4-0 1.9 0.0043-Hy d ro:xya _r1thranilic acid 350 4-80 0.7 0.20 Nicotinic acid 262,230 4-00 0.6 0.015 Quinolinic acid 272,235 4-30 0.5 0.007 aThe most intense peak wav~len g th is listed first; wavelengths are uncor rected for instrumental re s ponse)l precision in wavelength observation is+ 5 mu. b:Measure m ents were made at most intense peak wavelength. 0 Precision of measurement is+ 0.1 second.

PAGE 81

c::: 250 300 450 WAVE LENGTH ( mu) 500 600 Fig. 25.-Phosphorescence excitation and emission spectra of quinaldic acid in ethanol at 77K. 650

PAGE 82

200 250 300 350 400 WAVELENGTH 450 emu> 500 550 600 Fig. 26.-Phosphorescence excitation and emission spectra of kynurenic acid in 650 ethanol at 77K. -....J -F='

PAGE 83

200 250 300 350 650 WA VELENGTH cm.u) Fig. 27.-Phosphorescence excitation and emission spectra of xanthurenic acid in ethanol at 77K.

PAGE 84

LU 0:: 200 250 300 350 450 500 550 600 650 WAVELENGTH (mu) Fig, 28.-Phosphorescence excitation and emission spectra of indican in ethanol at 77K. CJ)

PAGE 85

200 250. 350 400 450 500 600 650 WA VELENGTH Cm11) Fig. 29.-Phos p horescence excitation and emission spectra of indole-3-acetic acid in ethanol at 77K.

PAGE 86

200 250 300 350 450 550 600 650 WA VELENGTH (mu) Fig. 30.-Phos p horescence excitation and emission spectra of 3-indolepyruvic acid in ethanol at 77K.

PAGE 87

0::: 350 450 500 600 650 WAVELENGTH ( m.u) Fig. 31.-Phosphorescence excitation and emi ss ion spectra of anthranilic acid in ethanol at 77K.

PAGE 88

200 250 300 350 400 450 650 WAVELENGTH (mJ.J) Fig. 32.-Phosphorescence excitation and emission spectra of 3-hydroxyanthranilic acid in ethanol at 77K. CP 0

PAGE 89

c::: 200 250 300 350 400 450 50 550 600 WAVELENGTH {mJ.J) Fig. 33.-Phosphorescence excitation and emission spectra of nicotinic acid in ethanol at 77K. 650

PAGE 90

200 250 600 WAVELENGTH (mu) Fig. 34-.'-Phosphorescence excitation and emission spectra of quinolinic acid in ethanol at 77K. 650 (X) [\)

PAGE 91

measurement of many of these compounds in the presence of others. Most of the decay times are quite similar, except for that of indole-3-acetic acid, and it is unlikely that time-resolved phosphorimetry could be easily applied to the analysis of mixtures.~ 7 83 In Table 6, it is evident that for some of the compounds, significant enhancement of the phosphorescence intensity occurred when ethyl iodide was added as a per turbing solvent. No attempt was made to measure those compounds having peak excitation wavelengths shorter than 300 mu, because of the problem of solvent absorption in this region. For xanthurenic acid and 3-hydroxyanthranilic acid which are relatively weakly pho sp horescent in ethanol (Table 5), the limits of detection were lowered by nearly an order of magnitude when measurements were made in EEI. In Table 7, limits of detection for the metabolites in ethanol and in EEI are compared, and the wavelength maxima in the two solvents are given. For quinaldic acid, the analytical curve in EEI had a slightly greater slope than in ethanol, and for this re as on, the limit of detection was the same in both solvents. Other analytical curves in EEI paralleled those in ethanol which was the beh a vior ob served in the hydrocarbon study. The stability of metabolites in EEI was less than that ob ser ved for the hydrocarbons. It was essential that measurements of

PAGE 92

'84 TABLE 6 EFFECT OF E'l'HYL IODIDE ON THE PHOSPHORESCENCE INT EN SI TY OF 1 l 1 RYP'l 1 0PHAN METABOLITES Metabolitea Quinaldic acid Kynurenic acid Xanthurenic acid Anthranilic acid 3-Hydro:xyanthr an ilic acid 3-Indolepyruvic acid aI1etaboli te concentr at ion was appr o x i mate ly -4 10 Min E EI solvent. R/Rob 2.0 9.5 2.7 14.0 1.0 bR: is phosphorescence inten s ity of met ab olit e in ethanol; R is pho sp hor es ce n ce inte nsit y in EEI. Pre cision in inten s ity me asu re men ts is+ 5 percent. Wavele ngt h s were adju st ed to give maximum sig n a l in each case.

PAGE 93

TABLE 7 COMPARISON OF PHOSPHORESCENCE LIMITS OF DE'l'ECTION FOR TRYPTOPHAN 1'1E'l 1 ABOLITES IN E'l'HANOL AND IN EEI AT 77K Metabolite a Quinaldic acid Kynurenic acid Xant:hurenic a cid Anthranilic acid 3-Hydroxy an.th ranilic acid 3-Indole py ruvic acid Ethanol Excitation Emission Maximum Maximum (mu) (mu) 305 348 350 350 350 302 510 458 485 440 480 425 Limit of Detection (ug/ml) 0.015 0.003 0.1 0.004 0.2 0.003 Excitation Emission Maximum Maximum (mu) (mu) 328 350 360 350 350 312 517 475 505 440 480 425 Limit of Detection (ug/ml) 0.015 0.0015 0.015 0.002 0.015 0.003 aOnly those metabolites which show an enhancement of phosphorescence in EEI are listed. bEEI is the ethanol-ethyl iodide mixture, 4/1, V/V. ()'.) V1

PAGE 94

86 phosphorescence intensity be made i mm ediately after addition of ethyl iodide to the analytical solution. If this pro cedure were adhered to, reproducible results were obtained. Ethyl iodide also had a significant effect on the emission spectra of several of the metabolites. Red shifts in the emission maximum not noted in the hydrocarbon studies were significant for quinaldic acid, kynurenic acid, and xanthurenic acid. This red shift and an increase in structure in the kynurenic acid emi s sion spectrum is shown in Figure 35. The emission spec t rum of xanthurenic acid retained its shape in EEI but was shift e d to lower ener g y. The s e shifts are in a g ree me nt with result s re p orted by Graham-Bryce a nd Corkil1 13 for cou m arin in ethanol-ethyl iodide mix t ures. The use of such s hift s and ch a n g e s in spectral sh a pe for char a cterization pu rp o s e s h a s been pointed out in the hydroc a r.bon study. Thin-layer chro ma to g ra p hy follo we d by m e a s ur eme nt of pho s phorescence inten s ity in ethanol w as found to be useful for the deter m ination of kynurenic a cid in urine. In Table 8 recoveri e s of kynurenic aci d from urine after a simple thin-layer separ a tion are giv e n. A sin g le ex traction with 5 ml of eth a nol g a ve better th an 90 p er c e nt recovery. Thin-lay e r chromato g r ap hy re q uired 75 minute s and the e x tr a ction a nd me as ure m ent s te ps a pp ro x i mate ly 30 minutes. The preci s ion indic a t e d is fo r five co mp l e te

PAGE 95

87 Ae x 34 8 mu /' / \ \ \ I \ I I I \ I \ I \ I \ I \ I \ I \ I \ I \ I \ I \ I \ I ...... 400 45 0 500 600 6 5 0 W A V ELENGTH (m J.J ) F ig 35 -Ph o sph o res c ence emission spectrum o f ky nur e n ic a c id ( 104 ~ ) in ethan o l, __ and i n EEI, at 77 K I nstrwnental sensitivity was reduced by a factor o f approximately 3 to obtain spe c trum i n EEI l!.ixci tat ion wavelen gth was 348 mu in bo th e t hanol and EEI.

PAGE 96

TABLE 8 RECOVERY OF KYNURENIC ACID FROM URINEa Kynurenic Acid Concen tration in Urine (ug/ml) 200 150 100 50 10 Percent Recovery 94 92 95 91 94 al"'Ie asur e ment w as by phosphorimetry in ethanol after thin-layer chro mat ogr aphy ; ex cit ati on and emission wavelengths we re 348 mu and 458 mu, respectfully. Precision is for five determin at ions. 88

PAGE 97

89 determinations. Blanks obtained from urine chromatographed with the sample gave intensity readin g s only slightly higher than the ethanol background at the wavelengths used. Xanthurenic acid could not be satisfactorily measured in ethanol at the same concentrations as kynurenic acid. The Rf values for kynurenic acid and xanthurenic acid for the chromatographic system used are 0,20 and 0.29 respectively, 14 and any overlap of kynurenic acid and xanthurenic acid on the thin-layer would not lead to errors in measure m ent of kynurenic acid using the above procedures. Thus, this method is very specific for kynurenic acid. Xanthurenic acid extracted fro m silica gel as above could not be measured in EEI even thou g h concentr ati ons were well above the limit of detection. Ap paren tly, ad sorption on silica gel resulted in an irreversible effect on xanthurenic acid which rendered the extracted material different from pre-adsorbed compound. Adso rpt ion on silica gel did not affect the fluoresce n ce of xanthurenic acid, and the fluorescence inten s ity measured confir me d that the ethanol extraction of the compound was nearly co mp lete. This apparent quenching of the heavy ato m effect is not general for materials ex tra cted from silica gel. The heavy atom effect normally found for the polynuclear aromatic hydroc a rbon s was observed after eth an ol extr a ction of these compounds fro m thin-l a yers of silica gel. The h eavy atom

PAGE 98

90 effect was also observed in kynurenic acid extracts. Simply adding silica gel to a solution of xanthurenic acid did not quench the heavy atom effect. Once xanthurenic acid was adsorbed on silica gel, however, the heavy atom effect was inoperative even if the silica gel were heated at 150C for several hours prior to extraction. The unusual results found for xanthurenic acid might be explained in terms of a strong interaction between the adsorbed metabolite and silica gel which blocks formation of the xanthurenic acid etbyl iodide char ge -tran sfer com p lex. Because xanthurenic acid can be an effective chelating agent, the sug g estion of a strong interaction between it and silica gel may be reasonable. Cl ement i and KashaL~ have studied pe rturbation effects in dyes adsorbed on silica gel where the ge l it self was the he avy atom species. Clo se co ntact between dye and gel, th at is, a monolayer of adsorbed dye, was necessary for effective perturbation. By analogy, satu ration of xanthurenic acid with silica gel could re stric t the ap pr oach of ethyl iodide. Applic at ion of heavy atom perturbation techniques to the det erminati on of xanthurenic acid in urine was not practical u s in g the simple thin-layer separati on procedure. This result does not det ra ct from the po ss ible u s e of such techniques in mea s uring xanthureni~ acid under different circ umstances Kynurenic acid c an be adequ at ely measured

PAGE 99

by phosphorimetry in ethanol at the concentrations often encountered in urine after oral ingestion of tryptophan. Measurement of kynurenic acid by phosphorimetry is a more sensitive procedure than the absorption method of Hill, ll~ Summer, and Roszel. Even lower levels of kynurenic acid could be detected if measurements were made in the heavy atom solvent. These results indicate that the heavy atom effect in phosphorimetry is not limited to the polynuclear aromatic hydrocarbons, the group of compounds with which it is most often discussed. The results also point out that for more complicated molecules, such as xanthurenic acid, other environmental effects can influence heavy atom perturbation in unusual ways. Such effects of a highly selective nature may be of use in phosphorimetry in their own right, and suggest further research in this area. 91

PAGE 100

CHAPTER V SUI'1I'1ARY The possibility of capitalizin g on environmental effects to improve the sensitivity, selectivity, and versatility of pho sp horimetry has received little previous attention. In Chapter II, it is shovm how theoretic a l considerations suggest the potential use of heavy atom perturbation in analytical phosphorimetry. The remainder of the dissertation has been concerned with a demonstration of this potential. In Chapter III, it is shown that ethyl iodi d e added to the u s ual eth an ol solvent is a convenient means of enh an cing the phosphorescence inte nsity of many of the polynuclear aromatic hydrocarbons. A 25-fold enhance ment was obtain ed for 2,3-benzfluorene. In addi ti o n the chan ges in emission spectra which accompany heavy atom perturbation and chan ges in relative intensity of b ands and in vibr ationa l structure are shown to be u sefu l for improving the selectivity of phosphorescence measurements. Investigation of simple char ge -tr ansfer complexes formed between the aromatic hydrocarbons and v arious acc e ptor compound s showed that, while theoretically this co mp l exat ion 92

PAGE 101

should result in effects similar to those observed for heavy atoms, practical considerations limited the use of this type of perturbation in phosphorimetry. It is sug gested that the heavy atom effect should considerably improve the phosphorescence measurement procedure in those analytical determinations of aromatic hydrocarbons where phosphorimetry has already been found u s eful. Phosphores cenee excitation and emis s ion wavelengths, decay times, and limits of detection of a number of polynuclear aro ma tic hydrocarbons are presented to aid such applications. In Chapter IV, the heavy atom effect is applied to the tryptophan metabolites, and these result s are co mpa red with those obtained in the hydrocarbo n study. Xanthurenic acid and 3-hydroxy ant hranilic acid which are relatively weakly phosphorescent in ethanol are shmm to have limits of detection nearly an order of ma g nitude lo v rnr in the perturbing solvent. A method is described for the simple determin at ion of kynurenic acid in urin e usi ng phosphor imetry and thin-layer chromato g raphy. Recoveries of kynurenic acid better than 90 per ce nt were ob tained using this procedure. Atte mpts to analyze xanthurenic a cid by a similar procedure with measure m ent in the ethyl io did e solvent failed due to a quenching of the heavy atom effect which occurred after adsorption of ;x:anthureni c acid on silica ge l. These re s ult s stress th a t the analyst should

PAGE 102

be always concerned with possible environ..~ental effects in phosphorimetry as these can greatly influence the success of an analytical procedure. 94

PAGE 103

B IBLI0GRA,1?HY 1 Am erican In strwnent C o ., Inc Ins tr uct ion s N o. 8 3 8 Silver S p ri. ng Maryla..11.d. 2 Chris t odoule as N ., M c Glynn, S. J Chem Ph ys 40 1 66(19 64) 3. Clar E., Z ande r, M. Chem Ber 89 74-9 (1 956) 4 Clemen t i, E. Kasha M ., J Chem Phys g 9 56( 1 95 7) 5. Co nd o n E. U Sho:ctly, G H "Th e Theory of Atomi c Spectra 11 Carnbrid g e at the Unive r sity Press, 1 95? Ch ap t e r s 7-10. 6 C z e k all a J ., :,Iag e r,_ X 1 z. Zl ek trochc m 6 0 :S 5(1 9S 2 ) 7. Drushel, H. 1 /., S o m...~ers A. L ., Anal Chem 38 10(19 66 ). 8 Du g gan, D E Bo ,r;:nan R. L ., Br o die B B. Uden f r iend, S ., } ,. rch Bi o chem 68 1(19 57 ). 9 E isent ha l, u3 J. C hem Phys 4 3, 1850(1 96 6) .. 1 0 E is e nthal K 3 El Sayed, ~,T A .' J Ch em Phys. J. J. 794 ( 1 965) 11 El 3aye~i. t"I ) .. cc ounts C h e;:n Res. 1 8(19 6 8). ., : l. 12 Fre ed S ., Saln r e V ., Sc i e nce 12 8 13 4 1(1 95 8) 13. Gr aham B r yc e I J C or ki ll, J M ., Na ture 18 6 965 (1 9 68) 14. H i l l, H ~ ., SuY....:ner S. X ., R oszel, N D Anal B ioc hem 1 6 SL~(J.9 66) 4-2 _, 15. H ol li fi e l d H C: \.fin2 f 0rdne r, J D ., A n al. Che m ., in p r ess 16 Iwa t a, S., T211 a.. 1--:a J N a gak u r a, s ,J Ch em Phys '+7 220 3 ( 1 967) 95

PAGE 104

17. Jablonski, A., Z. Physik 94-, 38(1935). 18. Kasha, M. J., Chem. Phys. 20; 71(1952). 19. Kasha, M., McGlynn, S. P., Ann. Rev. Phys. Chem. Z, 403(1956). 20. Keirs, R. J., Britt, R. D., Wentworth, W. E., Anal. Chem. 29, 202(1957). 21. Lewis, G. N., Kasha, M., J. Am. Chem. Soc. 66, 2100(194-4-). 96 22. Lower, S. K., El-Sayed, M. A., Chem. Rev. 66, 199(1966). 23. McCarthy~ W. J., Winefordner, J. D., J. Chem. Ed. 4-4-, 136(1967;. 24-. McClure, D.S., J. Chem. Phys. 17, 905(19 L~9) 25. McGlynn, S. P., Daigre, J., Smith, F. J., J. Chem. Phys. 39, 657(1963). 26. McGlynn, S. P., Neely, B. T., Neely, C., Anal. Chim. Acta 28, 4-72(1963). 27. McGlynn, S. P., Reynolds, M. J., D aig re, G. W., Christodouleas, N. D., J. Phys. Chem. 66, 2L~99(1962). 28. McGlynn, S. P., Sunseri, R., Christ;odouleas, N., J. Chem. Phys. 37, 1818(1962). 29. Moye, H. A., "The Analysis of Trace Quantities of Pesticides Utilizing Thin Layer Chrom at o graph y and Phosphorimetry," Ph.D. Di s sertation, University of Florida, 1965. 30. Moye, H. A., Winefor dne r, J. D., J. Agr. Food Ch em 13, 533(1965). 31. 0'Haver, T. C., Winefordner, J. D., Anal. Chem. 38, 602(1966). 32. Parker, C. A., Hatchard, C. G., An a lyst 87, 664-(1962). 33. Pavlopoulos, T., El-Sayed, M.A., J. Ch em Phys. 4-1, 1082(1964-).

PAGE 105

34. Pringsheim, P. "Fluorescence and Phosphorescence, 11 Interscience Publishers, Inc., New York, 1949, Chapter 4. 97 35. Ramakrishnan, V., Sunseri, R.' 1'1cGlynn, s. p.' J. Chem. Phys. 45, 1365(1966). 36. Reid, c. J. Chem. Phys. 20, 1212 ( 19 52) 37. Reid, C J. Chem. Phys. 20, 1214(1952). 38. Satoh, K., Price, J. 1'1., J. Biol. Chem. 230, 781(1958). 39. Sawicki, E., Pfaff, J. D., Anal. Chim. Acta 32, 521(1965). 40. Sawicki, E., Stanley, T. W., Pfaff, J. D., .Slbert, W. C., Anal. Chim. Acta 31, 359(196Lf.). l!.,l. Sc1,-, 0 n'(j,.., H R~d're c',nal rq,.,,::,.~1 -Z.'; .Li.v h., .1.~, J.ct -''-, ...... .1.vLl.~'-'- .;:;..;_, 42. Schmillen, A., Legler, R., "Landolt-B5rnstein, 11 N. S., Group II, Vol. 3, K. H. Hellwege, Ed., 3.9ringer-Verlag, New York, 1967 43. Siegel, S., Judeikis, H. S., J. Chem. Phys. 42, 3060(1965). 44-. Srinivasan, B. N., Kinoshita, I-I., :acGlynn, S. J. Chem. Phys. 47, 5090(1967). D 45. St. John, P. A., Brook, J. L., Biggs, R. H., Anal. Biochem. 18, 459(1967). 46. St. Jo:b.,n, P.A., McCarthy, W. J., Winefordner, :. D., iL~al. Chem. 38, 1828(1966). 47. St. John,?. A., Winefordner~ J. D., Anal. Chem. 39, 500(1967). L~8. :.1in, I-I., 11 Phosphorimetric Analysis of, Drugs in Blood and Urine," ?~.D. Dissertation, 'Jniversity of :C-lorida, ., I'"', ,.. /t ...!.. JO-r 49. Udenfriend, S., "Fluorescence Assay in Biology and I-Iedicine, 11 Academic Press, :New York, 1962. 50. Wehry, 3. L., in "Fluorescence, u G. Guilbault, Ed., I-Tare el Dekker, Inc# ~Tew Y ')rk, 19S?, Chapter 2.

PAGE 106

98 51. ./inefordner, J. D., in 11 :B'luorescence and Phosphorescence Analysis: Princip l es and App lic ations 11 D. M. Hercules Ed., Inter s cience Publishers, Inc., New Yor k 1966, Cha pt er 5. 52. Winefordner, J. D., Latz, H W., Anal. Che m 35, 1517(19 63 ) 53. 1 ,Jinefordn e r, J. D., McCarthy, W. J., St. Jo hn P A., in "Metho ds o f Biochemical Analysis," Vo l. 15, D. Glick, Ed ., Interscience Publi sh ers, Inc ., New Yor~, 19 6 ?. 54. Winefordner, J. D., St. John, P. A., Anal. C h9m 35, 2211(19 63 ). 55. Z and er, M ., An g ew. Chem. Intern. Ed. Eng. ::, 9 30 (19 65 ). 56. Zander 1 M ., Fre s enius' z Anal 8be m. ~26 251(1 967) 57. Zander, M "Pho sphorimetry 11 Acaderaic Press, Xew Yor : -c 19 68 Ch apter 4.

PAGE 107

BIOGRAPHICAL SKE'l'CH Lyal Van Sant Hood was born December 30, 1942, at Rahway, New Jersey. He attended public schools in Westfield, New Jersey, and Springfield, Pennsylvania, and was graduated from the Escola Americana do Rio de Janeiro, Rio de Janeiro, Brazil, in July, 1960. He entered Lafayette College, Easton, Pennsylvania, in September, 1960, and received his B.S. degree in chemistry in June, 1964. Since then he has attended the University of Florida, working toward the degree of Doctor of Philosophy. He is a member of the American Chemical Society and Alpha Chi Sigma. 99

PAGE 108

This dissertation was prepared under the direction of the chairman of the candidate's supervisory committee and has been approved by all members of that committee. It was submitted to the Dean of the College of Arts and Sciences and to the Graduate Council, and was approved as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August, 1968 Dean, Col Dean, Graduate School