UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A study of the spectra of pyrene Vilkos, Verna Vilma Beruthe 1971

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1972_A1 V54.pdf [ 7MB ]
Metadata
JSON: 831-1.0060137.json
JSON-LD: 831-1.0060137-ld.json
RDF/XML (Pretty): 831-1.0060137-rdf.xml
RDF/JSON: 831-1.0060137-rdf.json
Turtle: 831-1.0060137-turtle.txt
N-Triples: 831-1.0060137-rdf-ntriples.txt
Original Record: 831-1.0060137-source.json
Full Text
831-1.0060137-fulltext.txt
Citation
831-1.0060137.ris

Full Text

A STUDY OF THE SPECTRA OF PYRENE by VERNA VILMA BERUTHE VILKOS A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of CHEMISTRY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October, 1971 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make i t freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver 8, Canada Date <^o\K x r / l i -ABSTRACT The polarized absorption and fluorescence spectra of pyrene-h.^ and of pyrene-djQ as guests in biphenyl and fluorene single crystal matrices at about 10°K have been measured. Incomplete vibrational analyses of the ground and of the first two excited singlet electronic states are given in terms of these spectra and further information is obtained from low-temperature spectra in polycrystalline n-paraffin samples. Ground state vibrational data were also obtained from laser-Raman studies of a pyrene-h^ single crystal and of pyrene-h^Q and -d-^Q in solutions of benzene, carbon disulphide and carbon tetrachloride; as well, the fundamental frequencies of the two isotopic spec-ies were calculated using approximate force fields that have been applied successfully to similar molecules. The spectra are complex. In fluorescence, an exceptionally large number of modes appear which could not be accounted for in terms of a and b_ funda-. g 3g mentals. These lines either (a) are common to a l l matrices used and appear through Fermi-resonance or are either Franck-Condon or Herzberg-Teller allowed, or (b) appear in only one or two matrices being induced by the crystal forces of the host environment. The effect of Herzberg-Teller coupling is the intro-duction of intensity which does not arise from the *B_ -> *A transition; in ' 2u g contrast, the remaining interactions are similar in that they cause only a redistribution of the intensity of the pure electronic transition amongst various vibronic bands. Mechanical coupling in the lowered symmetry of the site in the host lattice causes a significant redefinition of the atomic dis-placements in the pyrene guest molecule and plays an unusually important part in these spectra. For example, the intensities of many lines assigned as combination bands were inconsistent with the Franck-Condon predictions. More-over, the sheer numbers of lines appearing in the fundamental region in one matrix or another suggest that almost all fundamentals may be present and that pyrene in its first excited electronic state is probably not situated at the inversion and mirror-plane sites in the biphenyl and fluorene lattices, respec-tively. Further, the pyrene molecule has probably caused a localized lattice distortion; a measurement of the absolute intensity of absorption indicates that the unexcited guest molecule causes a local expansion of the fluorene lat-tice in the ab_ plane. An upper estimate for these matrix induced interactions is about 10 cm 1 and probably arises from repulsive contributions, since for all the matrices used in this work, pyrene is somewhat bigger (usually wider) than the host molecule i t replaces. There is no mirror-image relationship between fluorescence and absorption and this is probably due to changes primarily in the off-diagonal elements of 1 1 the force constant matrix. The first transition is assigned *• A^ (short-axis polarized) and the second *B u^ +• *A (long-axis polarized). Vibrations belonging to the first excited electronic state and having b total symmetry provide a vibronic interaction with the second electronic state. The more nearly degenerate the interacting states become the more intense is the coup-ling and in the single crystal spectra where "solvent" shifts move the two excited electronic states close together in energy the usual Herzberg-Teller approach does not apply and vibronic basis functions must be considered. Thus, a fundamentals of the first electronic state which are very weak or g absent in the b-polarized spectrum appear in c_' polarization built on false origins (the more important being a b^ combination at 1423 cm 1 and a b^ fundamental at about 1500 cm"1 in pyrene-h1Q in biphenyl) in the energy region -iv-of the second electronic system. In the n-paraffin spectra, the combination corresponding to 1423 cm * could not be found while the fundamental was easily located (at 1564 cm * in n-heptane) . Thus, while the changes in shape of the pyrene molecule in going from the ground to the two excited states are different, the sets of normal coordinates belonging to the first and second excited electronic states are probably nearly parallel to each other but not to the ground state set. -V-TABLE OF CONTENTS CHAPTER 1. INTRODUCTION 1 1.1 Review of Previous Work 1 1.2 Some Remarks Concerning Spectral Analysis 5 1.3 Spectra of the Solid State 8 CHAPTER 2. EXPERIMENTAL 10 2.1 Purification of Materials 10 2.2 The Crystal Structures 13 2.2.1 Fluorene 13 2.2.2 Biphenyl 15 2.2.3 Pyrene 16 2.3 Mixed Crystal Preparation 17 2.4 Raman Spectra 17 2.5 Apparatus for the Low Temperature Spectra 18 2.6 Measurement of the Absolute Intensity of Absorption .. 20 2.7 Triplet-Triplet Absorption 21 CHAPTER 3. THE RAMAN SPECTRA 23 CHAPTER 4. CALCULATION OF THE FUNDAMENTALS 36 CHAPTER 5. THE FLUORESCENCE SPECTRA 46 5.1 Pyrene-h^ 46 5.1.1 The Spectrum in Outline 46 5.1.2 Intra- and Intermolecular Perturbations 49 5.2 Pyrene-d 70 -vi-CHAPTER 6. THE ABSORPTION SPECTRA 86 6.1 Room Temperature Spectrum 86 6.2 Low Temperature Spectra 89 6.2.1 Pyrene-h10 89 6.2.1(a) The first electronic transition 91 6.2.1(b) The second electronic transition .... 93 6.2.2 Pyrene-d1() 96 6.3 Triplet-Triplet Absorption 98 6.4 Lattice Modes 103 CHAPTER 7. CONCLUSION 106 REFERENCES 110 APPENDIX. The Fluorescence Spectrum of Pyrene-h1Q In Some n-Paraffin Matrices At About 10°K 114 - v i i -LIST OF TABLES 2.1 Squared direction cosines of the molecular axes with respect to the crystallographic axes in fluorene 14 2.2 Squared direction cosines of the molecular axes with respect to the crystallographic axes in biphenyl 15 3.1 Correlation table showing the selection rules for the isolated molecule and for the crystal 25 3.2 Oriented-gas predictions of the relative intensities of free-molecule Raman lines in various crystal configurations related to the frames a*, b_ and c* 25 3.3 The Raman spectrum of a single crystal of pyrene-h^ near the exciting line 31 3.4 The Raman spectrum of pyrene-h^^ 32 3.5 The Raman spectrum of pyrene-d^Q 35 4.1 Observed and calculated g-fundamentals of pyrene 42 4.2 Observed and calculated u-fundamentals of pyrene 44 5.1 Correlation table showing the selection rules for the pyrene molecule in the fluorene lattice 54 5.2 Correlation table showing the selection rules for the pyrene molecule in the biphenyl lattice 54 5.3 The fluorescence spectrum of pyrene-hlf) in fluorene and biphenyl at about 10°K 77 57 5.4 Weak fluorescence lines which cannot be accounted for in terms of known a and b, fundamentals 67 g 3g 5.5 The fluorescence spectrum of pyrene-d1f1 in fluorene and biphenyl at about 10°K 77 75 6.1 A correlation of the a g fundamentals of pyrene-h1Q and pyrene-d^ in the ground and excited electronic states 99 6.2 A correlation of the b3 g fundamentals of pyrene-h^ and pyrene-d^ in the ground and excited electronic states 100 The fluorescence spectrum of pyrene-hin in some n-paraffin matrices at about 10°K .7 119 - v i i i -LIST OF FIGURES 2.1 The crystal structure of fluorene showing four unit cells 14 2.2 The unit cell of the biphenyl crystal 15 2.3 (a) The unit cell of the pyrene crystal 16 (b) Orientation of molecular axes and optical directions in the ac plane 16 3.1 The Raman spectrum of pyrene-h^ 30 4.1 The internal coordinates of pyrene 39 5.1 The polarized fluorescence spectra of pyrene-h1f. in biphenyl at about 10°K TT 56 5.2 The polarized fluorescence spectra of pyrene-d1f) in biphenyl at about 10°K .7 74 6.1 The polarized absorption spectra of pyrene-h1f- in fluorene at 300°K TT 87 6.2 The polarized absorption spectra of pyrene-h^ at about 10°K 90 6.3 The polarized absorption spectra of pyrene-d^^ at about 10°K 97 6.4 The dimer absorption spectrum at about 10°K of (a) pyrene-h1Q in n-octane, and (b) pyrene-d^ in n-octane 101 6.5 The polarized triplet-triplet absorption spectra of pyrene-din in fluorene at about 77°K .: 102 The fluorescence spectrum of pyrene-h^^ in n-pentane at about 10°K. ... 115 The fluorescence spectrum of pyrene-h^^ in n-hexane at about 10°K 116 The fluorescence spectrum of pyrene-h^^ in n-heptane at about 10°K. ... 117 The fluorescence spectrum of pyrene-h1ft in n-octane at about 10°K 118 -ix-ACKNOWLEDGEMENTS I am deeply indebted to Professor Bree for the patient guidance, encour-agement and very considerable time he has given me both during the course of this work and to make this thesis possible. My sincerest appreciation to Dr. R. Zwarich for his friendship, valuable assistance with experimental procedures and for many helpful discussions. I am also grateful to Dr. R.A. Kydd for a sample of pyrene-h^ and for , his many helpful suggestions that were incorporated (with his further assis-tance) into the final draft of this thesis. I wish to thank Mr. Muenster for generously donating his time to re-crystallize a sample of pyrene-d^^. Finally, I wish to express my appreciation to my associates and the mem-bers of staff and faculty for their assistance and contributions to this work. CHAPTER 1 INTRODUCTION 1.1 Review of Previous Work A transition is characterized by three essential properties, experiment measures the wavelength of an absorption or emission band, its strength and polarization while theory calculates the transition energy, oscillator strength and symmetry. A number of theoretical studies and experimental investigations carried out to obtain this information for pyrene have produced results that are not entirely in agreement. While theoretical considerations Cl-43 are in general agreement that the fir s t and second electronic singlet states of pyrene are polarized along the short ) a n c* ^on& ^ B i u ^ m ° l e c u l a r axis, respectively^ (although one calculation [_S3 predicts the directions of the two lowest energy transitions to be the reverse of this), experimental evidence on the whole has been inconclusive. Williams C6H found that the two lowest electronic states were either both long- or both short-axis polarized by observing the polarization direction of fluorescence when polarized light was absorbed into these excited states of pyrene in a rigid glass at 195°K. The solution spectrum of pyrene was also examined at low temperature using photoselection techniques C7,8]. From these measurements of the relative directions of polarization i t was concluded that the first two transitions alternated between the short and long axis. -2-Using polarized light, Lyons and Morris C9j studied the spectral depend-ence of photocurrents in crystalline pyrene at room temperature. These results were consistent with the fir s t transition being short-axis polarized and the second more nearly long-axis than short-axis polarized. Ferguson ClO"! measured the absorption and fluorescence spectra of a pyrene crystal at room temperature and 77°K but was unable to make any assignments as both the polarization ratio and the measured Davydov splitting l a y between the values calculated for the *B^u and assignments. He observed a marked change in the polarization ratios of the vibronic bands of the lowest state indicating that vibrational-electronic interactions were occurring with a nearby state of different symmetry. Ferguson concluded that a level close in energy to the fir s t excited electronic state must be responsible for the interaction and so restricted the perturbing level to be one of the first three excited states. Tanaka E l l ] also examined the crystal spectrum and by comparison with calculated intensity ratios concluded that the polarizations of the first four transitions alternated between the short and long molecular axes in agreement with the theoretical predictions of Ham and Rudenberg L~2j. However, from an investigation of the absorption spectrum of pyrene substitutionally dissolved in phenanthrene and dibenzyl matrices, Hochstrasser Cl2lj assigned the two lowest-energy transitions as short-axis polarized. Fluorescence, phosphorescence and ultraviolet absorption spectra also provide information about the vibrational levels of the appropriate electronic states. Assuming that pyrene has covering symmetry in the low energy electronic states, the 72 fundamental vibrations may be given the following symmetry classification: 13 a . 5 a . 4 b, , 12 b. , 7 b„ , 12 b„ , 12 b_ ' g u lg lu 2g 2u' 3g and 7 b, . Of these, 36 are Raman active having symmetry a , b, , b- or 3u 6 g' lg' 2g -3-b_ , 31 are infrared active with symmetry b. , b„ or b_ and the 5 funda-3g / ' lu' 2u 3u mentals with a^ symmetry are inactive in both the infrared and Raman. Mecke and Klee [lZ~] studied the Raman spectrum of pyrene in solutions of carbon tetrachloride and carbon disulphide and were able to assign 8 a^ and 4 b^g fundamentals from depolarization ratios. In addition, they also examined the infrared spectrum in the vapour phase, in several solvents and with single crystals using polarized light incident on the ab_ face. Califano and Abbondaza Cl4] extended the infrared measurements down to 400 cm 1 for both pyrene and deutero-pyrene by studying oriented polycrystalline films. More recently, Neto and di Lauro [I5j calculated the planar vibrations for pyrene-h^ and -d^ Q using an approximate valence force field which had been adjusted to f i t the known vibrations of benzene, naphthalene and anthracene. They compared the calculated spectra with their experimental data and the previously reported vibrational assignments Cl3,14h. The fluorescence of pyrene from the pure crystal ClOU as well as from both concentrated solid Cl63 and liquid Cl7H solutions is broad and structureless, and provides no information about the vibrations of the ground electronic state. This emission is believed to originate from an excited dimeric species (or "excimer") formed through the association of a ground state molecule with a second molecule in its first excited singlet state [18j. The molecules in the crystal are arranged favourably for excimer formation since they lie in pairs with molecular planes parallel Cl93> Because the absorption is always charac-teristic of the monomer, the ground state of the dimer is unstable. Even at 77°K the spectra of organic molecules in n-paraffin matrices are remarkably sharp, although they exhibit a multiplet structure which depends upon the relative sizes of the solute and solvent molecules, the temperature -4-and the rate at which the sample is cooled [20,21]. Presumably the multiplet structure arises from solute molecules which occupy different environmental sites in the paraffin matrix. Recently Hochstrasser and Small [22] have pre-sented evidence which suggests that a pyrene molecule in a biphenyl matrix at 4°K may occupy one of three stabilized orientations close to the true substi-tutional position. Although the spectra in the polycrystalline n-paraffins do not provide information about the polarization of lines, the well-resolved structure is a valuable aid in making a vibrational analysis of large polyatomic molecules. Attempts to grow single crystals of n-hexane and of n-heptane doped with pyrene were unsuccessful in the present work; the pyrene precipitated as the solutions were slowly cooled during the crystal growing process. The most extensive study of the quasilinear fluorescence and absorption spectra of pyrene in an n-paraffin solvent was reported by Klimova [23]. A vibrational analysis of the fluorescence spectrum of pyrene in n-hexane at 4°K was given, as well as lists of the fundamentals found in the fi r s t and second excited states. The third absorption system consisted of diffuse bands and was not analyzed. Symmetry assignments of the 25 ground state fundamentals observed were based on the results of Mecke and Klee [13]. Eight a and four b ^ fundamentals were associated with lines which appeared in the infrared spectrum. Klimova assigned these latter vibrations to be either of symmetry a^, b 2g or on the assumption that they appeared in the fluorescence spectrum through vibrational-electronic interaction and in the infrared spectrum by a loss of D,^  symmetry, especially in the crystal. Pesteil et al. [24] have also reported the absorption and fluorescence spectrum of pyrene in n-pentane at 20°K, but an analysis was not sought. The fluorescence spectrum of pyrene has been measured in methylcyclo--5-pentane at 9°K by P e l l o i s and Ripoche L"25j and symmetry assignments of the 35 observed fundamentals were made usin g the Mecke and Klee Cl3j p o l a r i z a t i o n data. Here again e i g h t a and f o u r b_ fundamentals were recognized w h i l e g og f o u r t e e n l i n e s were c o r r e l a t e d w i t h i n f r a r e d bands. The present research was undertaken to r e s o l v e the d i f f e r e n c e s between experiment and theory concerning the symmetries of the two lowest e x c i t e d s i n g l e t e l e c t r o n i c s t a t e s by studying the p o l a r i z e d f l u o r e s c e n c e and a b s o r p t i o n s p e c t r a of pyrene-h^g and of p y r e n e - d 1 Q i n s u i t a b l e s i n g l e c r y s t a l m a t r i c e s . An attempt was a l s o made to remove the apparent i n c o m p a t i b i l i t i e s between the f l u o r e s c e n c e s p e c t r a [23,253 on the one hand and the Raman and i n f r a r e d s p e c t r a C13-153 on the other. 1.2 Some Remarks Concerning S p e c t r a l A n a l y s i s In the Born-Oppenheimer approximation C263 the v i b r o n i c wave f u n c t i o n can be separated i n t o e l e c t r o n i c and v i b r a t i o n a l p a r t s . Then the t r a n s i t i o n moment, M.. ., between two v i b r o n i c s t a t e s reduces t o -13,st' 3N-6 M . . f = < * , ( q , Q ) | M ( q ) | * . ( q , Q ) > II <o. (Q) | o . (Q)> i j , s r l j s,t=0 J 3N-6 M (OJ n <a • (Oj|a (0J> J s,t=0 J where the i n t e g r a t i o n i s c a r r i e d out f o r f i x e d n u c l e a r c o o r d i n a t e s , Q, u s u a l l y chosen as those of the e q u i l i b r i u m molecular c o n f i g u r a t i o n i n the s t a t e from which a b s o r p t i o n or emission o r i g i n a t e s . The product i n t e g r a l measures the overlap between the v i b r a t i o n a l wave f u n c t i o n , a(Q), of the ground and e x c i t e d s t a t e s and governs the d i s t r i b u t i o n of i n t e n s i t y w i t h i n the band system. The -6-square of this overlap integral is called the Franck-Condon factor and the 2 vibromc transition probability is proportional to M. . 1 j , s t Electronic transitions are symmetry allowed when M „ (Q) has a component which transforms according to the totally symmetric representation and the overlap factors also must be totally symmetric for the transition moment to be finite. At low temperatures the molecules are assumed to be in the vibra-tionless level of the ground state since kT is then very small compared with vibrational energies. Emission will originate from the vibrationless level of the lowest-energy excited singlet (fluorescence) and ground triplet state (phosphorescence) because vibrational relaxation occurs in a time short com-pared to the lifetime of the electronic state. The zeroth vibrational levels are totally symmetric so that the overlap factor and hence ML ^  g t will be non-zero only for those transitions occurring between totally symmetric vibrational levels. However, the coupling between electronic and nuclear motions may be appreciable, and in this case the Born-Oppenheimer approximation is no longer valid. Herzberg and Teller L"27j treated the nuclear coordinate dependence of the electronic Schrodinger wave equation as a perturbation in the form of small displacements from the equilibrium position. The variation in the electronic Hamiltonian, H , with the nuclear coordinates is given as a Taylor expansion in the normal coordinates about the appropriate equilibrium nuclear configura-tion (indicated by subscript zero) and the second term of the series, Hg(Q), is taken to be the perturbation , 3N-6 8 H -7-When the relative electronic energy differences are much greater than the vibrational spacings, the perturbed state corrected to first order is *,(q,Q) = i|,°(q,Q ) + I X (Q^°(q,Q ) J J O ..^ JK K 0 1 3 N - 6 9 H where X j k ( Q ) = < ^ ( q , Q o ) | y - ^ ) ^ | * ? ( q , Q o ) > k The coefficent, X., (Q), records the mixing of the pure electronic states j and Jk k induced by a perturbing vibration of appropriate symmetry. Since the effec-tiveness of this interaction is greatest when the two states are close in energy, the amount of mixing involving the ground state is assumed negligible because i t is usually isolated from other electronic states. The corrected wave functions are substituted into the transition moment integral and whether or not M.. l j ,sr. vanishes depends on the matrix element 3 N - 6 8 H <*J(q,Q 0 )| R J i ( T ^ 0 Q r l * ; ( q . Q 0 ) > Integration of the above matrix element is over electronic coordinates only and because He(Q) is invariant under all symmetry operations in total space i t fol-lows that ( 3 H / 8 ) must have the same symmetry properties in electron space as has in nuclear space. For pyrene, the first excited singlet state is of symmetry (see Chapter 4) and can mix with a nearby state through a vibration with symmetry bjg- The vibronic transition involving this b^^ frequency is called a false origin and normal Franck-Condon progressions may be built on i t . The polarization of these false origins will be the same as -8-the B. •*- A transition from which the intensity arises. Further, the inten-lu g J ' sity of totally symmetric vibrations derived from the Franck-Condon principle may be modified through this vibrational-electronic interaction L " 2 8 H . 1.3 Spectra of the Solid State Aromatic hydrocarbons form molecular crystals held together by weak Van der Waals forces. In the "oriented-gas" limit when the intermolecular forces vanish, the intensity of absorption along a particular crystallographic axis is proportional to the square of the projection of the corresponding molecular axis on to that axis of the crystal. In real crystals, however, intermolecular forces do alter the spectrum of an isolated molecule when i t is placed in a crystal field, the principal differences being: A splitting of a non-degenerate excited state of the molecule into a number of components equal to the number of molecules in a unit cell, displacement of the mean energy of these components from the free-molecule value and deviation from the polarization ratios calcu-lated from squared direction cosines. A theory which was capable of explaining these observed modifications in the spectra was developed by Davydov C29] based on the earlier ideas of Frenkel L~30H. In the method of mixed crystals first introduced by McClure C31j, the dis-persing of guest molecules in a host lattice minimizes the interactions between guest molecules thus avoiding many of the complications which arise in the pure crystal spectrum. Although a true oriented-gas situation is not realized, the method of mixed crystals provides a valuable technique for determining the sym-metry properties of vibronic transitions. Indeed by examining the polarized spectrum of naphthalene substitutionally dissolved in a matrix of durene, McClure was able to determine unambiguously the polarization of the first two -9-excited states of napththalene. The results of previous attempts C32] to mea-sure the polarization properties of naphthalene from the pure crystal spectrum were questionable because intermolecular interactions mix the lowest-energy excited singlet states sufficiently to cause complications in the spectrum E 3 3 ] . It is commonly found that impurity molecules substitute in lattices of mol-ecular crystals only when they have about the same mass and overall dimensions of the host molecule and when the binding energies of the pure guest and pure host crystals are nearly equal. In this work we make the further restriction that the host must be transparent in the spectral region where absorption and emission of the impurity is to be observed. CHAPTER 2 EXPERIMENTAL 2.1 Purification of Materials Pyrene-h^Q. A sample of purified pyrene-h^^ was kindly provided by Dr. W . G . Schneider of the National Research Council of Canada. In later experiments Eastman white label grade pyrene-h^^ was chromatographed on a column of silca gel saturated with petroleum ether (b.p. 60-110°) using an automatic apparatus described by Sangster C34j. The recovered material was sealed in an evacuated pyrex tube and zone-refined through 100 passes. Both samples gave identical spectra. Pyrene-d^Q. Pyrene-d^ supplied by Merck, Sharp and Dohme of Canada Limited was purified by chromatography on a column of silica gel saturated with pet-roleum ether (b.p. 60-110°). The chromatogram was developed with petroleum ether and the product eluted from the column with petroleum ether-benzene mix-tures varying from 4:1 to 1:4 v/v. A yellow impurity band remained at the column head. The pyrene-d^Q was recovered from a single fraction of eluate by evaporation and recrystallized twice using benzene and petroleum ether as the solvent pair. A high resolution mass spectrum of the purified pyrene-d1Q was obtained on an A.E.I. MS9 double focussing mass spectrometer and indicated that slightly less than 17% of the sample was C^DgHj. Fluorene. A greater effort had to be made in the purification of fluorene and -11-biphenyl, and in particular the purification of fluorene presented a serious problem. Anthracene, phenanthrene, benzf_f jindan and carbazole were the only impurities which could be identified. Removal of the impurities from fluorene was attempted using the following techniques; sublimation, recrystallization, chromatography (vapour-phase, column and thin-layer), irradiation with ultra-violet light, reaction with maleic anhydride, extraction with concentrated sulphuric acid and zone-refining. However, the elimination of one impurity often revealed the presence of another and sometimes created a new one. For example, the following purification method was based on the fact that under the action of ultraviolet light, the anthracene impurity in a carbon tetra-chloride solution is photo-oxidized or reacts with the solvent L~35j and hence is more easily separated from fluorene. A saturated solution of commercial fluorene in carbon tetrachloride was irradiated by a high-pressure quartz mercury lamp for about 48 hours during which time air was bubbled through the solution. Recovery of the fluorene from the orange-brown reaction mixture was effected by chromatography on a column of silica gel using carbon tetrachloride as both developer and eluant. A narrow blue-fluorescing impurity band (possibly unreacted anthracene) was eluted from the column just before the fluorene and an unresolved multi-coloured band of photo-reaction products remained at the column head. The fluorene was then sealed in an evacuated pyrex tube and zone-refined through about 100 passes. At about 10°K a new system consisting of sharp doublets appeared in the fluorescence and absorption spectra of this material, with c_-polarized origin bands at 29 505 and 29 519 cm *. A photo-reaction apparently involving fluorene had occurred and this method was discarded as a purification procedure. The photo-oxidation products of fluorene in a benzene solution in -12-the presence of oxygen and ultraviolet light are 9-fluorenylhydroperoxide and 9-fluorenone C363 but these impurities are held more strongly on the column than fluorene and probably remain at the column head. The most satisfactory method of purifying fluorene was the following. A solution of Eastman white label fluorene was refluxed with maleic anhydride in xylene for about 24 hours to remove anthracene C37]. The brown reaction mixture was filtered, diluted with xylene and then the anthracene adduct, and perhaps some 9-fluorenyl succinic anhydride (the reaction of fluorene with maleic anhydride at 210°C and 250°C has been reported [38H) were removed by repeated extractions with fresh portions of 5% sodium hydroxide until the aqueous layer remained colourless. After the organic layer was washed with water and the solvent removed, the recovered fluorene was dissolved in pet-roleum ether (b.p. 60-110°) and the saturated solution extracted repeatedly with fresh portions of 36N sulphuric acid until the acid layer containing any carbazole E 3 9 ] remained colourless. The organic layer was washed with water again, dried, concentrated and chromatographed on a column of si l i c a gel using petroleum ether as both developer and eluant. BenzCfllindan was eluted from the column just ahead of fluorene C40D. The fluorene fractions were combined, the solvent removed and the recovered material sealed in an evacuated pyrex tube and zone-refined through about 100 passes. The method described above was not entirely successful in purifying fluor-ene since a weak, long-lived green phosphorescence always persisted. In par-ticular, crystals of fluorene doped with pyrene-d^ phosphoresced green at 77°K i f the sample had been irradiated at room temperature prior to cooling, while a freshly-prepared sample gave a bright orange-red phosphorescence at 77°K. A sample which phosphoresced green could be converted to one which phosphoresced -13-orange at 77°K by dissolving away the surface layers with a tissue soaked in acetone. Clearly, fluorene, even as a solid, is quite photo-reactive and every effort was made in this work to avoid excessive irradiation of fluorene at room temperature; the optical systems, for example, were aligned after the fluorene sample had been cooled. Biphenyl. Eastman white label biphenyl was also freed of anthracene by reac-tion with maleic anhydride, then sealed in an evacuated pyrex tube and zone-refined through 62 passes. The purity of the final product was determined by examining the absorption and emission spectra of a crystal at about 10°K. No absorption was detected but a 2H hour exposure revealed very weak fluorescence from two impurities, one of which was identified as phenanthrene. The other emission originated in an energy region well within the absorption systems of pyrene. n-Paraffins. Spectroquality Reagent grade n-pentane, n-hexane and n-heptane (Matheson, Coleman and Bell) and 0.99 mole fraction minimum purity n-octane (Phillips Petroleum Company) were used without further purification. 2.2 The Crystal Structures 2.2.1 Fluorene The four fluorene molecules are arranged in an orthorhombic unit cell of space group p n a m ( D 2 h ^ s u c n that the molecular symmetry plane containing the short and normal axes is parallel to the crystallographic (001) mirror plane C41j. This orientation enables the unambiguous separation of short and long-axis polarized transitions. The (001) cleavage plane was identified from its known optical properties [42] using conoscopic techniques. In the present work, the (100) face of fluorene was used for all low-temperature polarization -14-measurements. Table 2.1 gives the squares of the direction cosines of the long (L), short (M) and normal (N) molecular axes with respect to the a_, b and c crystallographic axes. Figure 2.1. The crystal structure of fluorene showing four unit cells. Table 2.1. Squared direction cosines of the molecular axes with respect to the crystallographic axes in fluorene a b c L(z) 0 000 0 .000 1.000 M(y) 0 326 0 .674 0.000 N(x) 0 674 0 .326 0.000 -15-2.2.2 Biphenyl Biphenyl belongs to the space group P2^/aiC^) with the two molecules in each monoclinic unit cell occupying inversion sites L"43]. The crystallographic (201) plane is very nearly parallel to the molecular plane with the long molec-ular axis perpendicular to the b_ axis of the crystal. The optical properties are known C42] and allow the principal (001) cleavage plane to be distinguished from the less prominent (201) cleavage plane. Under conoscopic observation the (001) plane gives a single isogyre while the (201) plane presents a crossed figure. In-plane transitions were studied using the be* face of biphenyl. Direction cosine squares of the L, M and N molecular axes with respect to the a_, b and c_* crystallographic axes (where £' is perpendicular to the ab_ crystal face) are given in Table 2.2. Figure 2.2. The unit cell of the biphenyl crystal. i -16-2.2.3 Pyrene Pyrene crystallizes with four molecules in a monoclinic unit cell of space group P^/aCC^) L~19j as shown in Fig. 2.3. Although i t is possible that a small out-of-plane distortion occurs in the crystal, we shall assume that pyrene retains f u l l D^ ^ symmetry. From the known optical properties [44] the crystal faces belonging to the a*bc* frame were found using conoscopic techniques. The perpendicular axes a* and c* correspond to extinction directions with an angle 24.0 ± 0.5 degrees between a and a* and 34.3 ± 0.5 degrees between c_ and c* (see Fig. 2.3 where the axis c_', perpendicular to the ab plane is also shown). Figure 2.3. (a) The unit cell of the pyrene crystal. (b) Orientation of molecular axes and optical directions in the ac plane. -17-2.3 Mixed Crystal Preparation Mixed crystals were grown by slowly lowering the melt contained in an evacuated pyrex tube through a Bridgman furnace C 4 5 D . Pyrene-h^p and pyrene-d^ nominal concentrations ranged from about 0.0003 to 0.7 mole percent in fluorene. It was more difficult to grow single unstrained crystals of doped biphenyl and attempts to obtain any suitable samples from ingots which had melt concentra-tions greater than about 0.5 mole percent were always unsuccessful. For both fluorene and biphenyl, the pyrene was nonuniformly distributed through the solid host material with the impurity concentration greatest at the top of the ingot. The required crystal faces were found by locating the crystallographic axes from the conoscopic interference figures under the polarizing microscope. The single crystals were placed with the correctly aligned plane face down in a brass ring of the desired thickness and the ring fil l e d with plaster of Paris. When the plaster of Paris had set, the samples were ground to the thickness of the brass ring, polished on tissue saturated with acetone and removed from the plaster of Paris. The crystal path length was estimated by measuring the thickness of the plaster of Paris with micrometer calipers. Brass rings ranging in thickness from 0.5 to 3 mm were employed. The method used for aligning the optical directions of the prepared sam-ples with those of the Wollaston prism for the polarization measurements has been described elsewhere C46j. 2.4 Raman Spectra Raman spectra of pyrene-h^^ single crystals and of saturated solutions of pyrene-h and pyrene-d., dissolved in benzene, carbon disulphide and carbon -18-tetrachloride were recorded on a Cary model 81 Raman spectrophotometer fitted for coaxial excitation. The 6328 A line from a helium-neon laser operated at about 80 mW with a maximum output power of about 40 mW at the sample-to-hemi-spherical lens interface was used for exciting the spectra. The diameter of the solution cells was many times larger than the diameter of the laser beam and the cell was aligned so that reflection of the beam at the cell wall did not occur. This precaution was taken to minimize corrections in measured depolarization ratios (p). Single crystal samples were grown in a Bridgman furnace. The desired crystal faces were prepared by orienting a suitable reference face (usually the ab_ face) on a turntable using the polarizing microscope. The crystal was cut along the a*, b_ and c* directions to within one degree with a Lastec wire saw equipped with a diamond impregnated wire of diameter 0.008 inch. Samples were mounted on triangular blocks with 45 degree base angles so that the crystal extinction directions were aligned parallel and perpendicular to a sloping edge of the block. This 45 degree mounting was chosen to avoid grat-ing bias in the reflection of polarized light. Four spectra were run on each crystal face corresponding to the possible polarization combinations of inci-dent and scattered radiation. For example, a be* face gives rise to a spec-trum for each of the (bb), (c_*c*), (be*) and (c*b) configurations where the first and second terms enclosed in the parentheses refer to the polarization of the incident and scattered light, respectively. If the two "crossed" spectra did not agree, the crystal was remounted. 2.5 Apparatus for the Low Temperature Spectra A Hilger and Watts large Littrow and a Jarrell-Ash f/6.3 Czerny-Turner -19-spectrograph was used to photograph the spectra, which were calibrated from iron arc and neon lamp spectra. The continuum for absorption was provided by a PEK type X-75 xenon short arc lamp. Fluorescence and phosphorescence were excited by the group of lines at 3130 A isolated from a PEK type 110 high-pressure mercury lamp by means of a 5 cm path of aqueous K^CrO^ and Corning f i l t e r number CS7-54 f_47j. In order to minimize reabsorption at the fluores-cence origin, front-face excitation was used. The two liquid helium cryostats used in this work follow the basic design of Duerig and Mador C48]. Provision was made to allow a vertical displacement of the sample chamber relative to the outer windows through an arrangement incorporating a Sylphon bellows so that two samples could be studied in each experiment. The samples were cooled by conducting away their heat to the liquid helium through GE 7031 cement and a short length of copper or by im-mersion into helium exchange gas which transferred the heat by convection to the walls of the sample chamber in contact with the refrigerant. However, the samples themselves are thermal insulators and were not efficient at conducting away the heat that was directed on to them from the light source so that the sample temperature was dependent on the intensity of the incident radiation. Indeed, in the cryostat where the sample was mounted in a vacuum, fluorescence excited by the mercury lines isolated using an aqueous nickel sulphate f i l t e r in place of the potassium dichromate showed hot bands and somewhat broadened structure as also did crystals which had cracked badly on cooling or which had been improperly anchored to the metal. An upper limit to the temperature of the samples of about 10°K was estimated from the absence of a 24 cm 1 "hot band" (phonon) in the emission spectra from fluorene mixed crystals. Solutions in the paraffin solvents to be studied at 4°K were contained in -20-1 mm path length brass cells. The cells were thermally anchored to the sample holder by means of screws and the GE varnish. Samples were syringed into the cells through a hole that was sealed with a small lead ball compressed against the opening by a strip of metal secured with two screws. The linewidths in the low temperature pyrene spectra were about 10 cm * in fluorene and about 5 cm * in the paraffins and biphenyl, and were largely determined by the s l i t width for the latter matrices. Plates were enlarged by a factor of about ten on to Kodabromide A5 photographic paper and the meas-ured line positions, always quoted in vacuum wavenumbers, were reproducible to about 1 cm * for the sharper spectra. 2.6 Measurement of the Absolute Intensity of Absorption The molar extinction coefficients for the two lowest-energy absorption systems of pyrene-h^^ in a fluorene matrix at about 300°K were determined in the following way. With a hydrogen lamp providing the continuum for absorp-tion, optical densities were plotted point by point at a constant s l i t width of about 0.15 mm using a Beckman D.U. monochromator fitted with a 1P28 photo-multiplier at the exit s l i t , the signal being detected by a Keithley model 414 micro-microammeter. The path length of the crystal was measured both with micrometer calipers and using the microscope and a reticle calibrated in m i l l i -meters. Micrometer thickness readings were often somewhat larger when the sample surfaces were not quite parallel. In these cases, reticle/microscope values were accepted since they were measured at the crystal cross-section where the optical density was determined. The actual concentration of pyrene in the host material was determined from solution optical density measurements recorded on a Cary model 14 spectrophotometer. The single crystal was weighed -21-on a Mettler Micro Gram-atic balance, dissolved in 1.4 ml of Fisher Spectro-analyzed Reagent benzene which was dispensed from a 5 ml micro-burette into a 2 ml volumetric flask. The solution was syringed into a specially constructed brass cell with an effective path length of 10.14 cm. (That any measureable contribution to the observed optical density did not arise from reflection at the cell walls increasing the apparent path length was determined from measure-ments on a solution of benzene containing a known amount of pyrene. The molar extinction coefficient of the 3380 A pyrene band had been found already using commercial quartz cells.) Before weighing, the crystal was polished with ace-tone and examined under ultraviolet light and with the polarizing microscope to ensure that the surfaces were clean. This procedure was repeated for six o samples and the results averaged. The s l i t width at 3380 A was about 0.1 mm for a ll solution measurements. The most serious error arose from the uncertainty in estimating the base-lines for the crystal spectra and for the necessarily lower optical density solution spectra because of the close proximity of a very much stronger absorp-tion band due to b e n z C f l i n d a n present as an impurity in the fluorene. The error ranged from about 5 to 15% from this source and was much greater than errors arising from, for example, loss of solvent from evaporation, temperature effect on the density of benzene, volume measurement of a non-aqueous solvent using equipment calibrated for aqueous solutions and the effect on the optical densi-ties by any difference in s l i t width between the solution and crystal spectra. 2.7 Triplet-Triplet Absorption Polarized triplet-triplet absorption spectra were measured using the mod-ified Beckman D.U. monochromator. The crystals were mounted over a pin-hole -22-on a stage in an evacuated liquid nitrogen brass cell which was equipped with an arm connecting the stage to the outer cell permitting the crystal axes to be aligned in situ with respect to the Wollaston axes. A steady-state popu-lation of the lowest triplet state was maintained by oblique irradiation from a high-pressure mercury arc, and an iodine quartz projector lamp provided the continuum for absorption. One of the experimental difficulties encountered was providing illumin-ation sufficiently intense to maintain a steady-state concentration of the lowest triplet state so that absorption was measureable and at the same time prevent those mercury lines emitted in the triplet absorption region from entering the monochrometer. In early experiments, the signal from the photo-multiplier was measured by a Keithley electrometer equipped with a decade shunt, the output being connected to a Bristol strip-chart recorder. This arrangement was not entirely satisfactory since the filters necessary to remove unwanted mercury lines also decreased the intensity of the exciting light resulting in a substantial reduction in the triplet state concentra-tion. For later experiments a lock-in amplifier became available to us for a short time. Since the amplifier afforded phase-sensitive detection i t was possible by chopping only the light from the tungsten lamp which provided the background continuum to eliminate any interference from mercury lines. Triplet-triplet absorption of pyrene-d.^ in n-hexane at 77°K was photo-graphed in first order using the Jarrell-Ash f/6.3 spectrograph at a s l i t width of 0.08 mm. CHAPTER 3 THE RAMAN SPECTRA The polarized Raman spectra of a pyrene-h^^ single crystal is shown in Fig. 3.1, and the complete details of the spectra near the exciting line are given in Table 3.3. Crystal frequencies together with solution intervals and depolarization ratios are collected in Table 3.4. Information from pyrene-d^ solutions is listed in Table 3.5. Before inclusion in the tables, the depol-arization ratios were corrected by linear interpolation making use of the known depolarization ratio values for carbon tetrachloride and benzene bands E49U. Calibration curves for carbon disulphide (pyrene was the most soluble in CS^ but no depolarized solvent bands are observed) were constructed in a similar manner using the corrected values for pyrene bands already measured in the other solvents. The depolarization ratios have errors that depend on the line strengths and proximity to other lines. Each ratio was measured a number of times for a given solvent and the values ranged about 10 - 15% from the mean for a depolarized weak line. All polarization measurements include an error arising from an inaccuracy of about 5 degrees in aligning the half-wave plate with the analyzer. A neon lamp spectrum was used for wavelength calibration and each of the intervals in the tables has been corrected accordingly. The uncertainty in the measurement of the Raman intervals was about 2 cm 1 and the values obtained in a l l solvents agreed within this range. No correction was made for the change -24-of spectrophotometer sensitivity with wavelength but comparisons of line inten-sities at any wavelength amongst the crystal spectra are valid. Selection rules for the vibrations of the free molecule and of the crystal are correlated in Table 3.1. Each free-molecule state may split in the crystal into four zero wave-vector states; two of these crystal levels are infrared active and two Raman active. However, crystal splitting is usually very small for molecular crystals which are more often regarded as an "oriented gas". In this approximation, the intensity distribution of Raman lines in the crystal is obtained through a transformation relating the molecular and crystal coordinate systems and is given in Table 3.2. The Raman solution symmetry assignments were made as follows: Totally symmetric vibrations were identified on the basis of their low depolarization ratios (significantly less than 0.75) while Raman lines having depolarization ratios approximately equal to 0.75 were taken to mark non-totally symmetric modes. The following discussion will apply to both pyrene-h^g and pyrene-d^ wherever relevant with the entries in parenthe-ses referring to the perdeuterated molecule. On the basis of intensity the following totally symmetric modes were assigned as fundamentals: 3103 (2296), 3059 (2284), 3024 (2272), 1632 (1619), 1553 (1507), 1408 (1391), (1277), 1242 (1166), 1145 (878), 1067 (833), 802 (752), 592 (568) and 408 (400) cm"1. In the above l i s t , the pyrene-d10 Raman interval corresponding to 752 cm 1 is inactive and the assignment is taken from the fluorescence spectrum (see Chapter 5) where the fundamental appears with strong intensity. Using the observed fundamentals and assuming an average deviation of ±5 cm 1 from the harmonic frequencies, the theoretical product rule ratio of 0.177 for the totally symmetric symmetry block of pyrene places the final -25-Table 3. 1. Correlation table showing the selection rules for the isolated molecule and for the crystal. Molecular Point Group C^ Factor Group N Bases C ^ Site Group Bases n 13 5 xx, yy, zz a g a u a g aa, bb, cc, ac 6 4 xy lg a u A b 5 12 z lu 7 12 xz y bo 2g 2u J\ b g ab, be 6 12 7 yz Y b 7 3g V, b u a, c 4 / A b3u N is the number of molecular fundamentals and n the number of lattice frequencies with the wave vector k equal to zero. Table 3. 2. Oriented--gas predictions of the relative intensities of free-molecule Raman lines in various crystal configurations related to the frames a*, b and c*. I (a ) xx gJ I (a ) yy g I (a ) I (b, ) zz gJ xyv lg' !xzV V^V a*a* 0.187 0.315 0.000 0.972 0.018 0.023 Ibb 0.311 0.156 0.003 0.879 0.111 0.079 c*c* 0.000 0.002 0.879 0.002 0.040 0.171 Ja*b 0.241 0.221 0.001 0.021 0.005 0.053 Ibc* 0.006 0.018 0.047 0.050 0.490 0.314 a*c* 0.005 0.026 0.010 0.004 0.419 0.495 -26-a g fundamental at about 1413 ± 170 cm . However, both the force-field calcu-lation and the correspondence between the assigned pyrene-h.^ and pyrene-d^ fundamentals suggest that the missing frequency lies below 1400 cm The nearest polarized Raman lines within the indicated region appear at 1358, 1351 and 1327 cm 1 and the strongest of these, 1358 cm 1 (which probably is per-turbed as a result of mixing with 1351 cm 1 which is then taken to be a com-bination) is tentatively accepted as the required fundamental. The band at 1395 cm 1 was discounted as a candidate because i t more likely marks a combin-ation sufficiently close to the strongly Raman active 1408 cm 1 fundamental that some intensity is transferred to i t . Indeed, the abrupt termination of the spectrum beyond about 1700 cm-1 (excluding C-H and C-D vibrations) i l l u s -trates the intrinsic weakness of combination and overtone bands. In the crystal, the molecules are arranged in the lattice with their axes projecting appreciably on to a l l crystal axes so that the intensities of mol-ecular Raman lines are so scattered amongst the possible crystal configurations that non-totally symmetric lines could be assigned only with difficulty. The situation is made more complex because each pyrene molecule occupies a general site in the unit cell and the crystal field provides a means by which a l l free-molecule states can mix and the intensity distribution can shift away from that predicted by the oriented-gas model. These perturbations are expected to be most noticeable for lines with small tensor elements adjacent to strong lines (totally symmetric or lattice modes). Some entries in Table 3.4 whose symme-tries are not given actually exhibit a character being strongest in the (c*c*) spectrum and this probably only reflects the fact that this spectrum contains the strongest lines; in particular, the 693, 714 and 1317 cm 1 intervals appear only in (c_*c*) . For this reason, the true symmetries of many of the weaker -27-lines cannot be determined; for example, the best observed energy f i t for the 498 cm 1 interval is provided by the 493 cm-1 b^ fundamental C44H. A consid-erable simplification is afforded by the solution spectra which provides r e l i -able depolarization ratios, allows nearly degenerate molecular vibrations to be distinguished from factor-group components and where the problem of incom-plete separation of inter- and intramolecular motions is removed. It is not possible to distinguish between b^, b^ and b^ symmetries from solution data alone. Further, i t is extremely difficult to distinguish between b^^ and bj modes from their intensity distributions amongst the Raman spectra of the solid (see Table 3.2). However, b_ modes may be assigned using the polarization data derived from the fluorescence spectrum (see Chapter 5) and the reasonable assumption that depolarized Raman intervals greater than about 1000 cm 1 represent b, vibrations. In fact, in the following l i s t , the •jg 1174 cm interval appears in the crystal spectrum only in the (c*c*) polari-zation, and has been tentatively assigned as b^ from its depolarization ratio and its polarization in the fluorescence spectrum (see Chapter 5). Thus the following b fundamentals were identified: (2255), 1597 (1581), (1435), 1370, (1240), 1174, 1110, 737, 505 (465) and 458 (437) cm"1. The (c*c*) intensity shown by some of the b„ frequencies is presumably acquired through crystal-field mixing; for example, 458 mixing with 408 cm \ 736 probably with 805 cm 1, 1176 with 1144 and 1243 cm"1, and 1373 with 1243 and/or 1408 cm"1. The inter-vals 3049 and 3015 (2272) cm"1 may also mark b^ fundamentals. With the exception of the interval at 221 cm 1, b^ fundamentals were accordingly more difficult to assign. In particular, of the two tentatively accepted candidates, 773 and 970 cm 1 (see Table 3.3), the selection of the latter as a fundamental was guided in part by the force-field calculation -28-(see Chapter 4) which indicates that all b_ frequencies in the region below g about 1000 cm have been already assigned. Further, i t appeared likely that observed deviations from the oriented-gas predictions may be accounted for in terms of 773 and 970 cm-1 mixing with 592 and 1067 cm-1, respectively. The polarization characteristics of the 530 and 580 cm 1 intervals, although un-clear in the crystal, most closely correspond to b^g or symmetry, and the calculations (see Chapter 4) show that they are not b_ . In the crystal spec-g tra (see Fig. 3.1), 530 cm 1 is stronger than 580 cm 1 and i t is on this rather arbitrary basis that 530 cm 1 is tentatively accepted as a b 2g fundamental in Table 4.1. The interval at 256 cm 1 was the only b^g fundamental identified. By analogy with pyrene-h.^, the pyrene-d^^ bands at 206 and 232 cm 1 are assigned as b 2g and bj_g> respectively. Mecke and Klee C l 3 ] observed the interval 349 cm 1 in the Raman spectrum of pyrene-h^ in a carbon disulphide solution. We have found this line only for saturated solutions when small crystals are present; dilution causes our 353 cm 1 band to vanish while reducing the intensities of the other lines only slightly. Thus the 353 cm 1 band is crystal induced gaining intensity from the strong 408 cm 1 line. Energetically i t may be attributed to the 349 cm-"'' b 2 u fundamental [443 or the combination 126 (b 3 u) + 227 (b 2 g) = 353 cm-1 (b l u). Significantly, the 353 cm 1 interval is broad so that both vibrations may be weakly active. All the Raman active pyrene-h^^ lattice modes with k=0 have been identi-fied as a : 30, 46, 56, 76, 92, 127 cm"1, and b : 30, 41, 56, 67, 93, 126 cm 1. The intervals 169 and 170 cm 1 are taken to mark b and a factor-group g g components, respectively, of the lowest-energy a u molecular fundamental. -29-In the unit cell the four molecules may be divided into two pairs (see Fig. 2.3); the two molecules in each pair have their planes parallel and are interchanged through an inversion centre. The splitting between a and b phonon levels arises from the interaction between sets of pairs, and i t is clear that the observed splitting is small. This result is consistent with the view that pyrene may be regarded as made up of weakly interacting pairs ClO] leading to the possibility of excimer formation. -30-(a*Q*) eoo 1000 1200 woo W A V E N U M B E R (CM"') 3 Z O O Figure 3.1. The Raman spectrum of pyrene-h^. The polariza-tions of the incident and scattered radiation are indicated from left to right within the parenthesis. Broad background scattering with a maximum about 800"1 to the red of the ex-citing line was subtracted from the (c*c*) and (a*c*) spectra. -31-Table 3.3. The Raman spectrum of a single crystal of pyrene-h^^ near the exciting line. (a*a*) (bb) (c_*c*) (a*b) (be*) (a*c*) symmetry 30 w 30 vw 30 m a 30 vw 29 w b g 40 w 41 ms b g 46 m a g 56 s 55 vs 55 w 56 ms a 55 mw 56 w b g 67 w b g 76 w 76 m 77 w a g 92 vw 92 w 92 m a g 93 mw b g 127 w 127 vw 126 vw 127 vw a g 126 w b g 169 w 169 w b g 170 w a g t This line was not completely resolved so that the true value of the lattice frequency may be less than that given here. -32-Table 3.4. The Raman spectrum of pyrene-h^g, ^solution P A vcrystal symmetry"1" 223 (b ) ) 221 sh dp g b 231 (ag) ) l g 256 w 0.67 263 b, 353 see text 408 s 0.32 408 a g 458 w 0.9 458 b, ? 3g 490 498 see text 505 w 0.8 505 b, 3g 530 see text 580 see text 592 s 0.05 593 a g 693 see text 714 see text 737 vw 0.7 736 b_ 3g 750? 773 w 0.78 g y b 776 (aJ i K2g 783 (b ) g 802 mw 0.23a 805 a g 844 vw 0.6 844 b_ or b, 2g 3g 904 vwb dp 913 958 0.6 961 -33-A vsolution P A vcrystal symmetry1" 970 w 1006 vw 1040 vw 1067 m 1110 w 1145 mw 1174 vw 1192 w 1199 w 1212 sh 1233 sh 1242 vs 1327 w 1351 w 1358 w 1370 w 1395 sh 1408 vs 1426 w 1462 w 1504 vw 0.64 0.19 .£ 0.04 0.06 0.76 0.25 0.7 0.17 0.16 <> 0.3 P? 0.21 0.38 0.23 0.26 dp P 0.14 < 0.1 < 0.1 P 971 1005 1040 1067 1109 1144 1157? 1176 1191 1205 b„ or b_ 2g 3g 1243 1317 1328 1352 1362 1373 1393 1408 1427 1460 1502 3g 3g see text 3g -34-solution P Av . , crystal symmetry^ 1553 w 0.08 1552 a g 1562 vw < 0.1 1567 a g 1597 ms 0.70 1596 b 3 g 1632 ms 0.12 1630 a g 1647 mw 0.11 1644 a g 1665 sh 0.3 1668 a g 1697 vw 0.2 3015 a g 3024 b m < 0.3 3026 3049 a g 3059 s 0.2 3059 a g 3103 w < 0.3 3102 a g f The symmetry species of the crystal factor-group components are given in parentheses. This depolarization ratio was determined in a benzene solution and no correction has been applied for a small contribution (polarized) arising from the solvent itself. k This line is broad. -35-Table 3.5. The Raman spectrum of pyrene-d 10' Av solution symmetry Av solution symmetry 206 sh 232 w 400 s 437 w 465 w 498 vw 568 s 616 w 771 w 833 mw 849 w 878 m 898 w 939 vw 1041 vw 1069 vw 1134 w 1157 sh 1166 s dp? 0.75 0.32 0.78 0.74 < 0.3 0.05 0.69 0.7 0.44 0.8 0.03 < 0.1 dp P P P 0.18 2g 3g 53g 1199 w 1222 vw 1240 mw 1254 vw 1277 mw 1303 w 1345 vw3 1376 vw 1391 vs 1436 vw 1507 m 1583 ms 1604 mw 1619 ms 1645 vw 2255 w 2272 m 2284 m 2296 mw < 0.27 0.2 0.72 P? 0.30 0.28 0.2 dp? 0.15 0.7 0.08 0.72 0.05 0.13 P dp 0.5 P P 3g 3g 3g 3g 3g This line is broad. CHAPTER 4 CALCULATION OF THE FUNDAMENTALS The fundamental frequencies of the ground electronic state are accessible experimentally through infrared, Raman, fluorescence and phosphorescence mea-surements. Despite these many sources from which data may be selected, a complete vibrational analysis is not ensured solely through the experimental approach. In most cases, the analysis of these spectra relies on the assump-tions that the stronger lines mark the presence of fundamentals and that departures of observed polarization measurements from the oriented-gas model are not severe. However, there can be ambiguity in the assignment of some modes (i) when the selection of some fundamentals within a symmetry block may involve the choice between many lines of moderate strength that indicate the presence of either weak fundamentals or combinations, or (ii) when the tran-sition mechanism in emission spectra is not known. Moreover, some fundamentals may be too weak to be observed in any of the spectra. Thus the calculation of normal frequencies can provide a useful (and often necessary) additional tool in a detailed vibrational study. However, the reliability of the calculations is restricted by the validity of the necessarily simplified force fields (and the numerical values assigned to the force constants) that are used to approx-imate the very high complexity of molecular reality. In principle, observed frequencies are used to obtain information about the force field. Since the number of force constants for the most general -37-type of potential energy function usually exceeds the number of known funda-mental frequencies even for a molecule with high symmetry and with a l l isotopic species available, assumptions are made to simplify the potential function in order to reduce the number of constants. Some approximate force fields used in recent years have been modifications based on a simplified potential function called the valence force field which assumes that the change in potential energy when a molecule is distorted from its equilibrium position can be described in terms of changes in lengths of chemical bonds and changes in angles between chemical bonds. One valence force field which was extended to include off-diagonal constants was developed by Neto, Scrocco and Califano C50] using the combined observed in-plane frequen-cies of benzene, naphthalene, anthracene and some of their deuterated deriva-tives. In a sense, the correctness of the force field can be tested by trans-ferring the derived force constants to other similar molecules to predict approximate frequencies. In fact, this unified field has been applied to phenanthrene C5lD, and pyrene [15] and the agreement between observed and calculated values was satisfactory. In the present work a normal-coordinate analysis was carried out to assist the interpretation of the rather complex pyrene spectrum. The in-plane constants were those developed by Neto et al. C50H; CC stretching force con-stants used for pyrene were interpolated from a plot of the CC force constants quoted in C 5 0 3 against the corresponding CC bond lengths. The out-of-plane force constants were those found by Scully and Whiffen [52] for naphthalene. The internal coordinates are defined in Fig. 4.1 where CC bond stretches (designated as R), CH bond stretches (r) and angle bends (a) are shown. Where possible, the angle bends on the outside of the aromatic rings were redefined -38-as in-plane wags ( B ) of the CH bonds to reduce the order of the matrices to be handled. The CC bond torsions (4>) and out-of plane bends (y) were numbered in the same way as the R and a respectively and are not shown in Fig. 4.1. Those force constants unrelated by symmetry are listed below in an abbrev-iated nomenclature similar to that of Freeman and Ross [53]. The constants are named with reference to the internal coordinates and the units are mdyne/A for stretching constants, mdyne for stretch-bend interactions and mdyne A* for bend-ing constants. Diagonal elements: R ^ = 6.76, R 2 R 2 - = 6 , 3 0> R3 R3 = 5- 6 6> R 4 R 4 = 7.80, R 1 5 R 1 5 = Rj9 R29 = a l l r^r^ = 5.055; = a 2 a 2 = a()a(> = 0.934, a^Cj = a4 a4 = a24a24 = °- 6 1 9' a5 a5 = a23a23 = ° - 9 2 0 ; a 1 1 = 1 - 0 1 > a 1 1 Y ^ = 0.315; a l l (|>.<f>. = 0.057. T i l Off-diagonal elements: R,R_ = R R = R„RC = 0.750, R„R~ = R0R1r. = R_R1.. = 1 Z o 4 / o Z J Z l b o l b R15R19 = R15R18 = °' 4 5 7' R1 R13 = R3 R5 = -°' 3 1 6> R1R15 = R2R18 = R3 R19 = R4R15 = R 1 5R 1 6 = R lR 3 = R2R4 = -0.176, R ^ = R^R^ = R3 Ri8 = °' 1 5 6' R 2 R13 = 0.342, R L R L G = R 3R 1 6 = R 4R 1 9 = R ^ = R ^ = R ^ = R ^ = R ^ = 0.156, R„R 1 0 = R0R1<; = -0.160; R,a, = R,a0 = R„a, = 0.279, R0a_ = R0a_ = R_ac = R_a. 2 12 2 16 11 12 46 22 23 2 5 3 4 = R3 a5 = R3 a6 = R15 a3 = R15a4 = R15a23 = R15a24 = R19a24 = ° ' 3 9 7 ; Rl&2 = ~RlH = - R 2 B 2 = R 3 B 3 = -R483 = 0.174; = = = = a 3 a 2 3 = a 4 a 2 4 = a24a26 = - ° ' 0 4 5 ' a2°5 = a5 a6 = a3°24 = a24a28 = a4 a23 = 0 ' 1 1 1 ; r l r 2 = T3 r4 = 0.068; Y l Y 2 = Y 2 Y 5 = Y 5 Y 6 = Y 6 Y y = Y 5 Y 2 3 = Y 2 3 Y 2 7 = 0.013, y ^ = Y 2 Y 6 = Y 5 Y y = Y2Y22 = Y2 Y23 = Y5 Y21 = Y5 Y27 = Y6 Y23 = - 0 - 0 2 4 ' ^ 2 3 = Y2 Y21 = Y5 Y10 = Y 6 Y 2 ? = -0.019; * 1 + 2 = + 1 * 1 4 = «t> 2* 3 = <t> 2* 1 5 = 4>3<f>4 = = <t>15<t>18 = * 1 5 * = -0.020; * L Y L = - ^ Y 2 = <f»2Y2 = -<|»2Y5 = ^ 5 = " ^ 6 = + 4 Y 6 = "0.021. 19 There are some inaccuracies involved in setting up a potential field in this way. Thus, Neto and di Lauro f_15j have used a somewhat different set of -39-Figure 4.1. The internal coordinates of pyrene. -40-force constants for the in-plane problem, although they started from the same basic field E50D. We have chosen greater CC stretching constants (because we made use of the more recent crystal structure data C19H) and there are other (smaller) differences in the choice of angle bending and interaction constants since some interactions associated with pericondensed systems have no counter-part in the transferred field. However, despite these differences, the results of the two calculations are similar. This suggests that the modified valence force field described here can be used at least as a guide to the location of molecular fundamentals. The in-plane frequencies of pyrene were also calculated using force con-stants based on other modified valence force fields L53-56J; some force fields contained as many as forty-five different non-zero constants. In general, the frequencies calculated by a l l tr i a l sets of force constants were not start-lingly different, with the largest variation being displayed by the lowest-energy members of the a^ and b^g symmetry blocks. The calculations were performed on IBM 7040 or 360/67 computers using a somewhat modified version of a programme written by Schachtschneider [_S7~]. A comparison of the calculated gerade frequencies with the observed fundamentals assigned from the present work and reference C44H is presented in Table 4.1; Table 4.2 lists the calculated ungerade frequencies together with the infrared assignments taken from reference C44j. Tables 4.1 and 4.2 also contain the calculated in-plane frequencies from reference Cl5H. The agreement between the calculated fundamentals and the experimental assignment is quite good when i t is considered that the possibility of perturbations being present in the crystal spectra has been ignored and that the force field has been transferred directly from other molecules with no refinement -41-being attempted. The comparison can sensibly be made only for the a^, bj u> b 2 u and b^ u symmetry blocks where the experimental information is more com-plete. The f i t is worst in the CH in-plane bending region, particularly for the b^ species. No attempt was made to use the calculated frequencies as a guide to locate missing g-fundamentals from the unassigned Raman intervals since there seems to be l i t t l e experimental justification for any such assign-ment. However, for example, the calculation suggests that a weak pyrene-h^^ interval appearing at about 897 cm 1 in all of the fluorescence spectra (see Chapter 5) is an unlikely candidate for a b^g fundamental as otherwise might be supposed on the basis of the experimental evidence. -42-Table 4.1. Observed and calculated g-fundamentals of pyrene. > pyrene-h1Q pyrene-d1Q symmetry obs.a obs. calc. calc. a obs.a obs.'3 calc. calc. a 3102 3081 3052 2302 2294 2272 3059 3073 3050 2292 2287 2268 3026 3021 3046 2273 2242 2264 1630 1630 1656 1602 1615 1619 1644 1587 1552 1567 1504 1504 1517 1443 1409 1408 1407 1396 1388 1388 1358 1382 1349 1352 1300 1326 1275 1263 1281 1241 1243 1232 1231 1158 1164 1162 1153 1147 1144 1142 1128 875 841 834 1069 1067 1058 1059 833 820 813 805 737 809 752 703 769 594 593 607 565 563 564 579 541 408 408 393 388 398 399 383 378 899 739 811 653 512 463 263 251 235 224 971? 963 849? 808 947 770? 768 780 764 754 755 618 600 530? 579 489 . 469 414 227 272 205 254 -43-pyrene-h10 pyrene-d1Q symmetry obs.a obs. calc. calc. a obs.a obs.h calc. calc. a 3049? 3049 3047 2273? 2265 2266 3015? 3028 3045 2252 2245 2260 1597 1596 1601 1634 1587 1582 1590 1622 1509 1492 1433? 1451 1429 1372 1373 1383 1382 1373? 1313 1301 1357 1327 1239 1198 1217 1201 1220 1209 1036 1014 1176? 1160 1162 904 903 1113 1109 1077 1096 831 833 835 834 736 694 745 688 647 699 505 496 485 466 455 444 458 425 428 437 411 414 Data taken from reference L~15l]. Data taken from reference L~44]. -44-Table 4.2. Observed and calculated u-fundamentals of pyrene. symmetry lu 2u pyrene-h1Q pyrene-d1Q obs. calc. calc. u a obs. calc. calc. 942 789 897 736 660 592 319? 391 344 164 179 153 165 3098 3081 3052 2294 2292 2270 3080 3029 3046 2278 2247 2265 3040 3020 3045 2256 2241 2260 1585 1613 1558 1570? 1588 1527 1468 1432 1437 1427 1392 1415 1449 1418 1413 1366 1314 1308 1242 • 1255 1236 1188? 1026 1024 1095 1074 1081 841 954 972 1064 982 1006 821? 814 811 820? 767 842 755 711 769 651 659 626 638 493 503 487 461 468 454 3028 3073 3050 2242 2289 2269 2989? 3049 3047 2211 2264 2263 1599 1607 1584 1561 1564 1559 1487? 1472 1474 1461? 1425 1447 1432 1397 1419 1338 1382 1328 1310 1384 1326 1276 1271 1275 1272 1172 1187 1037 1016 1022 1204? 1169 1167 945 911 921 1184 1145 1126 903 833 830 -45-pyrene-h 10 pyrene-d 10 symmetry obs calc. calc. obs. calc, calc, 891 954 981 762 828 821 537 506 562 519 489 543 349 355 336 324 330 312 b, 963 3u 957 804 798 845 816 745 703 748 753 598 638 710 717 568 568 484 483 431 428 219 195 202 178 126 113 119 105 Data taken from reference C 4 4 ] . Data taken from reference C l 5 j . CHAPTER 5 THE FLUORESCENCE SPECTRA 5.1 Pyrene-h^ 5.1.1 The Spectrum in Outline The polarized fluorescence spectrum of pyrene-h^ in a biphenyl matrix at about 10°K is shown in Fig. 5.1 and an analysis of the observed frequencies in biphenyl and fluorene is given in Table 5.3. Lattice modes and low frequency internal vibrations are not included in Table 5.3 and are considered later (see Chapter 6). Unpolarized fluorescence data from pentane, hexane, heptane and octane solutions are collected in an appendix. The spectrum in biphenyl extending about 4500 cm 1 from the origin consists of about two hundred lines with slightly more than half this number in fluorene. In the paraffin matri-ces, there was a marked increase in the number of observed lines with more than two hundred crowding an interval of 3300 cm 1. The following discussion refers to the spectrum in the biphenyl matrix unless specifically indicated otherwise. The short-axis polarized band at 26 734 cm the only resonance line observed in the spectrum, is the electronic origin of the *B_. -*- *A transi-r 2u g tion. Other prominent bands 408, 456, 496, 596, 736, 801, 1063, 1111, 1408, 1552, 1597 and about 1240 cm 1 from the origin form a framework on which an analysis of the spectrum can be based since combination and overtone bands -47-involving the above fundamentals can be identified. The intervals 408, 596, 801, 1063, ^1240, 1408 and 1552 cm 1 clearly represent a fundamentals since they retain the same polarization as the origin band while the intervals 456, 496, 736, 1111 and 1597 cm 1 are taken to mark b_ fundamentals appearing in •jg single quanta as false origins by "stealing" intensity from a strongly allowed TT - ir* transition of opposite polarization at higher energy. The assignment of the weak intervals at 1144 and 1631 cm 1 as a^ fundamentals supports the information obtained from the Raman spectrum where they appear with somewhat greater strength. In agreement with the Raman assignment, the one remaining a^ fundamental of frequency less than 2000 cm 1 is taken to be the weak 1355 cm 1 fluorescence interval. The strongest interval in fluorescence, 1408 cm \ forms the longest and most prominent progression in the spectrum. Because only three quanta are observed in this progression, even at the longest exposure times, the change in molecular dimensions of pyrene in the electronic transition must be small. Furthermore, the absence of progressions in non-totally symmetric vibrations indicates that the molecular point group symmetry is retained in the first excited electronic state. Frequency differences measured from different plates (but the same matrix system) agreed to within 2 cm \ the error in measurement. The intervals for one matrix did not quite agree with those obtained in other matrices or with the Raman values. Although these discrepencies are small we feel that they are real and represent the effect on intramolecular motions due to the environment. A more obvious manifestation of the presence of crystal forces is to be found in the relative intensity variation of the components of the 1240/1246 cm 1 "doublet" found in al l matrices except fluorene where the greater bandwidth -48-precludes the resolution of lines so close in energy. That the doublet may be in fact two nearly degenerate a fundamentals is unlikely; i t is more likely the result of a resonance between an a fundamental and an overtone or combin-g ation with the same total symmetry. The "impure" modes in this instance are each a mixture of the two original unperturbed levels. In biphenyl where the splitting is about 6 cm 1 and the intensity of the 1240 cm 1 line is at least twice that of the 1246 cm 1 line, i t is reasonable to suppose that the per-turbed level at 1240 cm 1 more closely corresponds to the a^ fundamental. While the splittings in the paraffin solvents are essentially the same as in biphenyl, the relative intensities of the two lines show a considerable vari-ation. In pentane and octane the higher-energy component is the stronger member of the pair, in heptane they have almost equal intensity, while in hexane the intensity pattern reverts to that shown in biphenyl. This appears to be an instance of a Fermi-resonance which becomes further enhanced by intermolecular interactions. It is clear that a resonance such as this should be quite sensitive to small perturbations provided by the crystalline environ-ment. In fact, for some combinations involving the doublet in the paraffin matrices, the intensity of the higher-energy component was greater than that predicted by the Franck-Condon principle. The Raman spectrum of pyrene in solution (where a static crystal field does not exist) shows a strong band at 1239 cm 1 with a weak satellite at 1231 cm 1. _1 With the exception of the 596 cm mode, the distribution of intensity amongst the fundamental vibrations below 2000 cm 1 in the ground electronic state of pyrene is similar for a l l hosts used. The anomalous bandwidth of the 1634 cm 1 interval in pentane apparently arises because the expected a funda-mental at about 1628 cm"1 and the combination 408 + 1237 = 1641 cm 1 have not -49-been resolved. Interactions complicate the C-H fundamental region in a l l spectra although assignments of some of the latter fundamentals seem secure (see Table 5.3). The dramatic decrease in intensity of the 596 cm 1 interval and its 572 cm 1 ~^IQ counterpart in going from the single crystal matrices on the one hand (especially biphenyl) to the paraffin matrices on the other is a further illustration of crystal-field mixing. However, without knowledge of the intrinsic intensity of this fundamental, i t does not necessarily follow that the greater intensity in fluorene and biphenyl is a reflection of greater effectiveness by these crystal fields in mixing the pyrene vibronic mode with other states since a diminution in intensity could result from transition moment phase cancellation effects. Further matrix effects must also account for the abrupt decrease in the intensity of the combinations involving the 596 cm 1 interval with one or two quanta of the 1408 cm 1 fundamental in the biphenyl spectrum. However, apart from small matrix-induced disturbances in the intensity patterns of the combinations and overtones, normal Franck-Condon distributions _1 (admittedly always steep ones) are usually observed. The 801 cm a funda-mental provides a possible exception since, although i t is prominent, both the overtone and combinations with b, false origins are either absent or too weak 3g to be unambiguously detected. Thus the 801 cm 1 mode probably derives most or a l l of its intensity through vibrational-electronic coupling. That the 752 cm 1 pyrene-d^ counterpart behaves as a false origin could not be ascer-tained since the overtone and combinations with b_ fundamentals a l l f a l l in 3g regions masked by other stronger lines whose assignments were secure. 5.1.2 Intra- and Intermolecular Perturbations The appearance of a wealth of emission lines, particularly in the paraffin -50-solvents, which could not be accounted for in terms of a and b_ fundamentals g 3g must arise from effects other than a lowering of covering symmetry of the free molecule in the first excited state. These lines, some of which are collected in Table 5.4, either (a) are common to a l l matrices used and are more likely to be appearing through Fermi-resonance or are either Franck-Condon or Herzberg-Teller allowed, or (b) appear in only one or two matrices and be matrix induced. Also contained in Table 5.4 are lines inconsistent with an assignment as combination bands because of either an energy or intensity mis-match. The intervals are characteristic of the pyrene molecule and not some unidentified impurity (say) since (i) they are often in more than one matrix and (ii) the magnitudes of the intervals vary when the guest molecule is pyrene-djQ. Because of the large number of such lines, only representative examples will be individually dealt with in the following discussion. Due to the presence of multiple sites, the spectra in the paraffin matri-ces are often complex; since each site gives rise to a perturbed free-molecule spectrum (and the perturbation at each site is not necessarily the same), a vibration need not appear with the same relative intensity or with the same frequency in each site spectrum. However, for the most part (where exposure time permitted and where the expected multiplets were not coincident with or obscured by other bands already accounted for) the vibrational bands dupli-cated the multiplet structures seen at the electronic origins in intensity and spacing; the most notable exception was the 1105 cm 1 interval in heptane which increased by 4 cm 1 for the second strongest multiplet. Some of the paraffin matrix frequencies entered in Table 5.4 are coincident with intervals built on other multiplet origins but these lines appeared with enhanced inten-sity. -51-Each symmetry assignment in Table 5.4 was at best difficult to make and must necessarily be tentative because of the indistinguishable intra- and intermolecular routes through which intensity could be brought into an other-wise forbidden vibronic transition. Case (a). The effect of intramolecular perturbations Extension of the Herzberg-Teller method of vibrational-electronic coupling (see Chapter 1) to take into account second-order corrections either by includ-ing the quadratic term in the Taylor expansion of the electronic Hamiltonian in powers of normal coordinates or by correcting to second order the wave v. function for the perturbed electronic state would formally allow the appearance of false origins involving any two frequencies whose total symmetry was a or b_ . The large number of fundamentals of the appropriate symmetry that can •jg combine to give energies that f i t the observed frequencies prevents an unam-biguous assignment of the participating vibrations. Thus the interval at 897 cm 1 may mark a b, combination acting as a Herzberg-Teller false origin •'g with the symmetries of the components either x °2g' au X ^3u 0 r ^lu * ^2u' Alternately, as will be shown later, even though examples belonging to case (a) probably involve intra- rather than intermolecular mechanisms, 897 cm 1 could be an exception corresponding to a calculated a u, b^ or bj fundamental (see Table 5.4) and gaining intensity through crystal-field mixing. In fluorene the appearance of a b 2 u or b^ fundamental would require that pyrene has moved away from the ab mirror plane while in biphenyl the presence of an a y or b^ fundamental would indicate that the inversion site symmetry was lost. In addition to the Herzberg-Teller mechanism, (0,2) bands in non-totally symmetric vibrations can derive intensity from the origin through Franck-Condon effects [58] i f there is a pronounced change in the force.field along the -52-appropriate normal coordinate between the ground and first excited electronic states. Thus 1027 cm 1 may represent such a (0,2) band (e.g. 2 x 512 cm-1); in the pyrene-d1Q spectrum (see Table 5.5) the 503 and 1006 cm-1 intervals correspond to 512 and 1027 cm 1 above. However, the weakness of all the bands (precluding observation of any (0,4) transitions) and a lack of knowledge of the excited state frequencies makes i t difficult to confirm this hypothesis. An upper limit to the intensity of the above possible (0,2) band is 0.01 times that of the origin and this corresponds to a not unacceptable change of about 25% in the frequency of the fundamental in the upper electronic state. Some of the weak lines adjacent to the intense 1408 cm 1 interval may be examples of Fermi-resonance but have not been unambiguously located in some matrices due to coincidences with stronger multiplet components or lattice structure (see Chapter 6). The 1432 cm 1 interval which retains the same strength relative to 1408 cm 1 as i t displays in the Raman spectra may repre-sent such an example. Case (b). Crystal induced effects Most of the lines in Table 5.4 seem to belong to case (b) and appear in the spectra through a matrix perturbation. The symmetry selection rules, indi-cating which pyrene molecular modes may appear through the lowered symmetry of the substitutional site, are outlined below. Assuming exact replacement of host molecules in the fluorene lattice where the site symmetry is Cg, pyrene electronic states could only site mix with I I A (in C ) states of fluorene and/or pyrene during an a" (in Cg) vibration of the latter (see Table 5.1). Thus in the lowered symmetry of the site, pyrene molecular modes of symmetry a^, b^, and b ^ could be active along the £ axis while a , b, , b„ and b„ modes could be expected polarized in the ab g lg 2u 3u r r —_ -53-plane. In biphenyl, the CL site symmetry would allow B l u and B 2 u electronic states to site mix through pyrene molecular modes of gerade symmetry which could appear along a l l crystal axes (see Table 5.2), and in the polycrystalline paraffin samples where there is probably no site symmetry, al l pyrene molecular modes would be formally allowed. An order of magnitude estimate of the matrix element that causes the per-turbation can be made in the following way. Consider two vibrational states a b of pyrene o^ and a separated in energy by AE^ i n the isolated pyrene mole-cule such that a transition from the vibrationless excited electronic state to a b O p is allowed and to is forbidden. If the host matrix provides a perturba-a b tion V which mixes the states a and a then some of the allowed intensity p p appears in the transition that terminates on the now impure state o^. An upper limit for the matrix element H ^ may be determined from one of the stronger matrix-induced transitions that has a large energy separation from an allowed transition. Thus, i f we suppose that the b_ component of the depolarized inter-val at 512 cm 1 marks a vibration which mixes with the 408 cm 1 fundamental (and not, for example, with the 596 cm 1 fundamental) and has only about 0.01 -1 -1 ^ times the intensity of the 408 cm line, then H ^ is about (104 cm ) (0.01)2 - 1 or 10 cm To examine the nature of this interaction in more detail, consider a pyrene molecule surrounded by the host, say biphenyl, molecules a l l in the vibrationless state o\ , where i is an index that specifies a biphenyl molecule. Then in the absence of the perturbation the eigenfunctions of the system are 3. b 3- C a Ha. and a Ha.. The zeroth-order eigenfunctions o a., H a. do not contribute P i P i P 1 i ^ . 1 any important effects to the spectrum as is seen by the absence of any vibra-tions appearing with unshifted frequency in the pyrene-h^ and -d^ Q spectra, -54-Table 5.1. Correlation table showing the selection rules for the pyrene molecule in the fluorene lattice.* Molecular Point Group Site Group D„, C 2h s Factor Group D2h n y X A 1 g Ig B2u B3u i A f A g Ig 2u • B* 3u b a 3 3 2 2 z A u B l u B2g B3g I I A f A u I B2g 1 B3g c 3 2 3 3 * n is the number of lattice frequencies with the wave vector k = 0. Table 5.2. Correlation table showing the selection rules for the pyrene molecule in the biphenyl lattice.* Molecular Point Group Site Group D2h C i Factor Group C2h n A g B i g B2g B3g ) A 1 g A g B g 3 3 z y x A u lu Bo 2u 3u I 1 A 1 U Au B u b a,c 2 1 n is the number of lattice frequencies with the wave vector k_ = 0. -55-and hence characteristic of the host. The range of i will depend on the nature of V, and i f a short-range potential built up from interactions between pairs of atoms on different molecules is assumed, as other authors C59H have done, then i need run only over nearest neighbour molecules. It is not possible to estimate values for the integrals as V is very sensitive to small changes of the molecular orientation and the exact positions of the pyrene molecule in its first excited state and of the disturbed biphenyl molecules surrounding i t are not known. The molar extinction coefficient cal-culation of pyrene in a single crystal of fluorene (see Chapter 6) indicates that the dispositions of the host molecules are distorted from their equili-brium positions in the immediate neighbourhood of the guest. Because of the rather different molecular dimensions of solute and solvent, a similar local-ized distortion is also expected in the biphenyl lattice. Thus the perturbation described by the model that has been developed above, requiring that the potential between adjacent molecules in the lattice create an asymmetric distortion (in °^ ^ e ^ree-molecule force field so that the set c° are no longer the eigenfunctions in the matrix environment, must be especially large for pyrene in a l l the matrices used. In fact, the crystal mixing of vibrational states is probably sufficiently severe to upset the normal Franck-Condon intensity pattern. Thus, for example, the b, com-bination 596 + 736 may appear with unexpected intensity at 1332 cm 1 even to the extent that in the paraffin solvents the combination is stronger than the 596 cm 1 a fundamental, g It is difficult to reach any firm conclusion concerning the exact des-cription of how the guest molecule is placed in the distorted host lattice. For example, the fluorescence was not well polarized, although the polariza-ID t CD CM * 10 w in co in //b 8 9 2 m W<DS eg - O + o + <r co S co o „ ^CO - N «1 - <• + + in ^o . x co2 -cW w o S x 7 » • S • N m lO l*> • !2 • ilDCO N «i w m o c in co O 10 * CM \ o CM to o CO o Figure 5.1. The polarized fluorescence spectra of pyrene-h1Q in biphenyl at about 10°K. The electronic origin band appears at 26 734 cm 1. -57-Table 5.3. The fluorescence spectrum of pyrene-h^ in fluorene and biphenyl at about 10°K.+'* fluorene biphenyl //b //c //b / / C analysis 26 690 vs 410 s 597 ms 805 s 820b ms 459 m 500 mw 738 ms 894 w 26 734 vs 408 vs 512a w 549 vw 567* w 596 s 698a vw 753 vw 801 s 816 m 456 m 496 mw 576 mw 669 w 736 s 771 w 782 w 840 w 862 vw * 885 vw 897 mw 0-0; B. -*• A 2u g 408, a 456, b 496, b g 3g 3g 596, a g 736, b 3g 801, a g 2 * 408 408 +456-2 -58-fluorene biphenyl //b / / £ //b / / £ ' * 909 vw 959C vw 978C vw 1007 mw 1004 mw 408 + 596 1023 mw 1038 vw 1027 w 2 * 512 - 3? 1066 s 1063 s 1088 w 1063, a g 1116 ms 1111 s 1111, b 3 g 1146 mw 1144 mw 1146, a g 1146 w 1144 mw 408 + 736 1160 w 1155 mw 1177 vw 1176 mw 1176, b, ? 3g 2 x 596 1196* w 1192* w 1215 m 1208 m 408 +801-1 1246 s 1240 s 1246 ms ! ^ 1240, a 1335 w 1355 mw 1332 mw 596 + 736d 1355, a ? g 1370 w 1370, b_ ? 3g 1376 m 1383*emw 1409 vs 1408 vs 1432 m 1408, a g -59-fluorene biphenyl analysis //b //c //b //£' 1476 m 1471 m 408 + 1063 1511 w § 1511 w 1526 w 1539* mw 1519 mw 408 + 1111 1555 s 1552 s 1552, a g 1570b ms 1566 t m 1600 m 1614 w 1597 ms 1597, 1635 m ( 1631 1648 mw mw 1631, 408 + a g 1240 + 2 1657 " I 1672f w 1656 1672a 1686 1709 mw w vw w 1697* vw 408 + 456 + 1246 + 2; 1240/1246 596 + 1063 - 3 ± 1/5 1710 w 1729 1762 w vw 1709 § 1741 mw vw 596 + 408 + 1111 + 2 1355 - 1? 1805* w 1800* w 736 + 1063 + 1 1819 ms 1816 ms 408 + 1408 1839b m 1839 w 1839 mw 1845 mw 596 + 408 + 1240/1246 1432 - 1 ± 3; -60-fluorene biphenyl analysis //b //c //b //£' 1867 mw 1864 mw 456 + 1408 1871 mw 1864 mw 801 + 1063 1879 vw 2 x 408 + 1063 *b 1886 w 1917 mw 1879 1903? w vw 496 + 1408 - 1 1936 w 1944? vw 408 + 1539 - 3 1965 mw 1960 m 408 + 1552 1984b w 1974 vw 408 + 1566 1984 w 1979 mw 736 + 1240/1246 ± 3 2007 mw 2005 m 596 + 1408 + l g 2007 mw 2005 m 408 + 1597 2050 m 2044 mw 801 + 1240/1246 ± 3 *b 2066 w 2059* vw 2 x 408 + 1240/1246 ± 3 2070 vw 1063 + 1111 - 4 2090§ vw 2131 w 2126 mw 2 x 1063 2147 mw 2146 m 736 + 1408 + 2 2154 mw 2146 m 596 + 1552 - 2d \ 2180 vw 2176 w 1063 + 1111 + 2 2213 ms 2209 ms 801 + 1408 2231b mw 2224* w 2 x 408 + 1408 -61-fluorene biphenyl analysis //b //c //b / / £ 2271 w 408 + 801 + 1063 - 1 * * 2304 mw 2288 vw 736 + 1552 2311 mw 2306 mw 1063 + 1240/1246 ± 3 2359 w 2354 mw 801 + 1552 + 1 2359 mw 2354 mw 1111 + 1240/1246 ± 3 2368* vw 2 x 408 + 1552; 801 + 1566 + 1 2391 vw 2414a w 408 + 596 + 1408 + 2 2438 w * * , d 2460 w 2456 mw 408 + 801 + 1240/1246 + 7/1 2474 ms 2471 ms 1063 + 1408 *h { 2483* mw 2 x 1240 + 3; 1240 + 1246 - 3 2491 m \ * I 2494 w 2 x 1246 + 2; 1063 + 1432 + 1 2523 mw 2519 mw 1111 + 1408 2552 w 1144 + 1408 2552 w 408 + 736 + 1408 2563 w 1155 + 1408 2583 w 1176 + 1408 - 1 2603* w 2 x 5 96 + 1408 - 1 2621 mw 2616 m 408 + 801 + 1408 - 1; 1063 + 1552 + 1 2627* w ( 2648 ms 1240 + 1408 2652 ms { I 2654 m 1246 + 1408 -62-fluorene biphenyl analysis //b //c //b //c' 2674 w 1240/1246 + 1432 ± 2/4 2693 vw 2714 w 2742 w 408 + 1063 + 1240/1246 ± 3 1332 + 1408 + 2 2762 w 408 + 801 + 1552 + 1; 1355 + 1408 - 1 2801* m 2795 m 1240/1246 + 1552 ± 3 2814 s 2815 2839 s mw 2 x 1408 - 1 1408 + 1432 2840f mw 2879 § 2917 2947* w vw w 2839 mw 1240/1246 + 1597 ± 2/4 408 + 1063 + 1408 1408 + 1511 - 2 1408 + 1539 2960 m 2958 ms 1408 + 1552 - 2 2979*bm 2974* mw 1408 + 1566 3007 w 3003 mw 1408 + 1597 - 2 3025§ w 3022* w 3022, b, 408 + 1063 + 1552 - l d 3022* w 3025 mw j 3032 mw 3032, a g 3066§ w 3055 mw 3055, b_ ( 3055 mw 3055, a : 408 + 1240 +1408 - 1 g 408 + 1246 + 1408 + 2; 596 + 1063 + 1408 - 3 3066 mwI 3064* w 3091* vw -63-fluorene biphenyl analysis //b / / c //b //£' 3110 mw 3105 mw 2 x 1552 + 1; 801 + 1063 + 1240/1246 + 1/5 * d 3117 w 1552 + 1566 - 1 3142 vw 3154§ vw 1552 + 1597 + 5 3165 vw 3198 vw ^3198§ vw 408 + 1240/1246 + 1552 - 2/8h 3222 w 3222 mw 408 + 2 x 1408 - 2 3247 vw 3247 vw 596 + 1240/1246 + 1408 + 3; 408 + 1408 + 1432 - 1 32585 w 3247 w 408 + 1240/1246 + 1597 + 2/4d 3282 vw 3272 w 801 + 1063 + 1408 3272 w 456 + 2 x 1408 32886 vw 3324 w 1408 + 1917 - 1 3333? vw 3345 vw 1409 + 1936 3352? vw 3370 vw 3363 w 408 + 1408 + 1552 - 5 3387 vw 596 + 1240/1246 + 1552 - 1/7 3394 vw 3387 vw 736 + 1240/1246 + 1408 + 3 3412 vw 3410 w 596 + 2 x 1408 - 2 g 3412 vw 3410 w 408 + 1408 + 1597 - 3 3456 vw 3452 w 801 + 1240/1246 + 1408 + 3 - 6 4 -fluorene biphenyl analysis //b / / £ //b / / £ 3 4 6 5 * vw 4 0 8 + 3 0 5 5 + 2 3 5 1 5 vw 4 0 8 + 2 x 1 5 5 2 + 3 3 5 4 1 * vw 3 5 3 3 w 2 x 1 0 6 3 + 1 4 0 8 - 1 3 5 5 7 vw 3 5 5 0 w 5 9 6 + 1 4 0 8 + 1 5 5 2 - 6 d 3 5 5 7 vw 3 5 5 0 vw 7 3 6 + 2 x 1 4 0 8 ' - 2 3 5 9 8 vw 3 5 7 9 vw 1 0 6 3 + 1 1 1 1 + 1 4 0 8 - 3 3 5 9 8 vw 3 5 9 6 vw 8 0 1 + 1 2 4 0 / 1 2 4 6 + 1 5 5 2 ± 3 3 6 1 8 w 3 6 1 5 mw § 3 6 4 3 ? vw 8 0 1 + 2 x 1 4 0 8 - 2 3 6 7 8 vw 3 6 7 5 vw 4 0 8 + 8 0 1 + 1 0 6 3 + 1 4 0 8 - 5 d 3 7 1 7 vw 3 7 1 4 w 1 0 6 3 + 1 2 4 0 / 1 2 4 6 + 1 4 0 8 ± 3 3 7 6 6 vw 3 7 6 1 vw 8 0 1 + 1 4 0 8 + 1 5 5 2 3 7 6 6 vw 3 8 2 4 vw 3 7 6 1 vw 1 1 1 1 + 1 2 4 0 / 1 2 4 6 + 1 4 0 8 ± 2/4 4 0 8 + 5 9 6 + 2 x 1 4 0 8 + 4 ; 8 0 1 + 3 0 3 2 - 9 3 8 6 4 * vw 3 8 6 0 * w 1 0 6 3 + 1 2 4 0 / 1 2 4 6 + 1 5 5 2 ± 5 / 1 ; 4 0 8 + 8 0 1 + 1 2 4 0 / 1 2 4 6 + 1 4 0 8 ± 3 ; 8 0 1 + 3 0 5 5 + 4 d 3 8 8 1 w 3 8 7 5 mw 1 0 6 3 + 2 x 1 4 0 8 - 3 3 8 9 9 * vw 3 8 9 1 * 3 9 6 2 4 0 0 6 * w vw vw 3 9 2 4 3 9 9 2 vw vw 2 x 1 2 4 0 / 1 2 4 6 + 1 4 0 8 ± 3 / 9 ; 1 2 4 0 + 1 2 4 6 + 1 4 0 8 1 1 1 1 + 2 x 1 4 0 8 - 3 1 1 4 4 + 2 x 1 4 0 8 + 2 ; 1 1 5 5 + 2 x 1 4 0 8 - 9 1 1 7 6 + 2 x 1 4 0 8 - 6 5 -fluorene biphenyl analysis //b //b / / £ 4 0 2 7 * vw 4 0 2 0 w 1 0 6 3 + 1 4 0 8 + 1 5 5 2 - 3 ; 4 0 8 + 8 0 1 + 2 x 1 4 0 8 - 5 4 0 3 3 * vw 4 0 5 3 w 4 0 5 4 w 1 2 4 0 / 1 2 4 6 + 2 x 1 4 0 8 - 2/8 4 0 8 3 vw 1 2 4 0 / 1 2 4 6 + 1 4 0 8 + 1 4 3 2 ± 3 4 1 2 1 vw 4 0 8 + 1 0 6 3 + 1 2 4 0 / 1 2 4 6 + 1 4 0 8 + 2 / 4 4 1 4 8 ? vw 1 3 3 2 + 2 x 1 4 0 8 4 1 7 0 vw 1 0 6 3 + 2 x 1 5 5 2 + 3 ; 1 3 5 5 + 2 x 1 4 0 8 - 1 ; , 4 0 8 + 8 0 1 + 1 4 0 8 + 1 5 5 2 - 1 4 2 0 0 * vw 4 1 9 6 w 1 2 4 0 / 1 2 4 6 + 1 4 0 8 + 1 5 5 2 - 4 / 1 0 4 2 1 6 w 4 2 1 5 w 3 x 1 4 0 8 - 9 4 2 4 4 vw 2 x 1 4 0 8 + 1 4 3 2 - 4 4 2 4 4 vw 1 2 4 0 / 1 2 4 6 + 1 4 0 8 + 1 5 9 7 - 1/7 4 3 6 2 vw 4 3 5 8 w 2 x 1 4 0 8 + 1 5 5 2 - 10 4 3 7 8 * vw 2 x 1 4 0 8 + 1 5 6 6 - 4 4 4 0 9 vw 2 x 1 4 0 8 + 1 5 9 7 - 4 4 4 3 4 vw 1 4 0 8 + 3 0 2 2 + 4 ; 1 4 0 8 + 3 0 3 2 - 6 4 4 6 2 vw 1 4 0 8 + 3 0 5 5 - 1; 1 4 0 8 + 3 0 6 4 - 10 4 4 6 2 vw 1 4 0 8 + 3 0 5 5 - 1 4 5 0 8 vw 1 4 0 8 + 3 1 0 5 - 5 ; 1 4 0 8 + 3 1 1 7 - 13 -66-+ -1 The position of the origin is given in cm , and a l l other entries show differences from the origin. * Intervals with assignments involving lattice modes are not listed. * shoulder S broad This line is essentially depolarized. k Some of this b-polarized intensity arises from an a assignment involving a lattice mode. 0 This line is probably depolarized. d This Franck-Condon assignment does not account for all the line intensity. e The polarization of this line is uncertain. Some of the c-polarized intensity arises from an a assignment involving a lattice mode. ^ This Franck-Condon assignment appears to account for more than the line intensity. This assignment does not appear to account for the linewidth. Table 5.4. Weak fluorescence lines which cannot be accounted for in terms of known a and a g b_ fundamentals. , ^ observed in various matrices ground state fundamentals0 biphenyl fluorene MCP C-25j pentane hexane heptane octane (469) 484 b-> 2g b3u 488 488 493 (512) b l u b l g 512 526d 528?d 527? 537 b2u 549 565? (579) b2g ( 567 1 576 578 577? 574? (651) b i lu 626 621 616 615 614 (660) a u 669 678 678 676 710 3u 698? 714 709? 703 748 »3u 753 (764) 1 I" 1 779? 788? (811) Ig S 1 782 845 b T 3u 840 842? 828? ground state observed in various matrices fundamentals0 biphenyl fluorene MCPC[253 pentane hexane heptane octane 891 b-> 2u 885 873? (897) a u 8976 8946 8816 8836 881e 8816 8806 (899) Ig 909 959 911 906 963 903? 900 963 b3u j 978 996 998f £ 993 992 f , g (963) 1013 1006? 1005g ( 1027e 10236 10286 1031e 10266 10256 10266 1064 b 1 lu 1041h I 1038 1040 1046n 1040 1042 1095 b l u 1088 (1160) b3g I 1155 1160 1168 1163 (1169) b2u < ( 1176 1177 1184 1181 1172 1181 1173 1184 b2u (1220) b3g 1206 1202 1200 1272 b2u 1268 1304 1298? (1310) b2u , 1332?1 1335?1 13251 13271 13211 13231 13211 (1357) b3g i 1355? 1349 1358 1345 1353 I oo I , ^ ^ observed in various matrices ground state^ fundamentals biphenyl fluorene MCPCE25H pentane hexane heptane octane I 1370 1363 1368 1367 1366 1366 (1383) b g ( 1383 1376 1384 1386 1389 1389? 1387 Table 5.2 was terminated when i t was apparent that combination bands were becoming increas-ingly important. b Entries enclosed in parentheses are calculated values; observed u-fundamentals are taken from reference [44], c Methylcyclopentane Eg d This line may mark the overtone of 263 cm 1 (t^g)• See text for an alternative assignment. £ This line coincides in energy with the expected combination (408 + 596) but is anomalously strong. g There is a weak multiplet component 37 cm"''' to the blue of the origin in octane; these lines, having unexpected strength, may be "site-selected" components of the 1026 and 1041 cm 1 matrix-induced vibrations. h This line coincides in energy with the expected combination (456 + 596) but is anomalously strong. 1 This line coincides in energy with the expected combination (596 + 736) but is anomalously strong. i ON I -70-tion was more complete in the fluorene matrix than in biphenyl. This depolar-ization was certainly too large to arise from the misalignment of the crystal axes of the sample with respect to the polarizer. Again, in fluorene, the variation in the polarization ratios among the a fundamentals and in partic-ular the large £ component of 596 cm 1 suggests that these effects can not be reasonably accounted for solely on a geometrical basis, viz., a movement of the long axis of pyrene away from the £-crystallographic axis into the ac plane to account for the fact that the short-axis transitions have greater intensity along £ than the long-axis along b_. However, a rotation of this nature would lower the Cg site symmetry allowing a g and b^ g molecular modes to (site) mix. For biphenyl, the C\ site symmetry precludes any equivalent conclusion regarding the orientation of the guest in the unit cell on the basis of the polarization ratios alone. Table 5.4 also attempts to correlate the observed with the calculated frequencies. There is ambiguity in making this correlation but the comparison does suggest that the pyrene molecule is not located at an inversion centre in the biphenyl lattice. 5.2 Pyrene-d1Q The polarized fluorescence spectrum of pyrene-d^^ in biphenyl at about 10°K is shown in Fig. 5.2 and an analysis of the observed frequencies in biphenyl and fluorene is given in Table 5.5. Information obtained from spec-tra in the paraffin solvents will be referred to, although for brevity the intervals are not listed. Identification of the fundamentals is analogous to that described for pyrene-h1Q. Thus, the intervals 399, 572, 752, 833, 879, 1160, 1280, 1389, 1506 and 1620 cm"1 represent a fundamentals while 436, 464, 690, 833, 1241 and 1585 cm"1 mark b fundamentals. -71-A mass spectrum showed that the deuterated sample contained 17% of the various monoprotonated impurities. A pure electronic band arising from the monoprotonated molecules with about one fifth the intensity of the main origin was not observed either in absorption or fluorescence. Thus the red shift associated with the substitution of a single proton is less than 10 cm-1; this estimate was made from the fluorescence origin which was somewhat diffuse to the red. The vibrational analysis is complicated by the fact that some of the weaker lines in the spectrum may arise from the three isotopic species of ^16^9^1" ^ we^g^lt^nS according to the number of hydrogens in equivalent positions, 3%% of the monoprotonated impurity is the isotopic isomer with C 2 v symmetry and the remaining 13%% is divided equally between the two with symmetry Cg. Thus the energy region below 2000 cm 1 may include lines aris-ing from these isotopic species since in C g symmetry, 39 a' and 23 a" funda-mental vibrations are allowed to appear polarized along b_ and c_', respectively, while in C 2 v symmetry, 19 a^ and 20 b^ modes can be active along b_ and c_', respectively. The frequency shift for the isotopic impurity modes predominately involv-ing carbon atoms is small so that while these are more easily located they would be difficult to distinguish from those pyrene-d^ lines appearing through, for example, Fermi-resonance. Thus the distinct shoulders observed to the red of the intense fundamentals 752, 1160, 1389 and 1506 cm"1 (but not 399, -1 -1 572 or 690 cm ) and the medium intense 464 cm fundamental may be due to the monoprotonated species. Two of these shoulders, at 1404 and 1519 cm may derive most of their intensity through Fermi-resonance as these are probably the analogs of the pyrene-h1f1 intervals at 1432 and 1566 cm respectively. -72-The remaining possible isotopic impurity lines are not readily identifiable and include those that undergo an appreciable shift on deuteration or appear because of the lowered symmetry of the isotopic species. No effort was made to determine whether the unassigned weak lines in Table 5.5 arose from iso-topic impurities or were induced by crystal forces, although some have clear counterparts in the spectrum of pyrene-h^ (e.g. the 503, 536 and 613 cm 1 intervals correspond to the 512, 549 and 669 cm 1 intervals of pyrene-h1Q). Analogous to what is observed in the pyrene-h^ spectra, there are combinations whose intensities deviate from the Franck-Condon prediction and these are indicated in Table 5.5. No explanation is offered for the medium strength b-polarized line at 2564 cm 1 although i t probably derives its intensity from the medium-strong 1160 + 1389 combination at 2549 cm the interval corresponding to 2564 cm 1 is prominent in fluorene where the "donor" combination appears with appreciable strength while in the paraffin matrices where the latter appears only weakly, the "forbidden" combination is corres-pondingly weaker. The only fundamentals that yield an a combination with energy 2564 cm 1 are of or b^ symmetry. The multiplet-structured phosphorescence spectrum of pyrene-d^ in the n-paraffin matrices agreed well with the previously reported spectrum in biphenyl C22j and did not reveal the presence of any fundamentals not seen in fluorescence. Typically, in n-hexane, fundamentals were observed at 395 (vs), 828 (mw), 1161 (w), 1272 (vw), 1387 (s), 1582 (vw) and 1615 cm"1 (s) from the main origin at 17 056 cm 1 (vs). This transition is probably orbitally allow-ed since the 1582 cm 1 interval represents a b, fundamental and the origin of the reverse singlet-triplet absorption spectrum (seen at 16 850 cm 1 only in the pure pyrene-h.. crystal C60j) is in near resonance with the phosphor--73-escence origin. Only two lines, observed at 1154 (vw) and 1601 cm (mw) in n-hexane, could not be accounted for as combinations of the a and b_ fund-g 3g amentals. The 1154 cm 1 interval appeared in all the n-paraffin spectra and probably derives its intensity through a Fermi-resonance with the 1161 a fundamental. While the 1601 cm 1 interval most likely also appears through a Fermi-resonance with the nearby 1615 cm 1 a g fundamental, coincident multi-plets in the other n-paraffin spectra make i t difficult to distinguish in this case between intramolecular and matrix-induced effects. Some support for these Fermi-resonance assignments is obtained from the Raman solution spectrum where intervals corresponding to 1154 and 1601 cm 1 with the same intensities relative to their associated fundamentals are observed. In fluorescence, the weakness of the 1620 cm 1 a g fundamental and the proximity of a line whose assignment is secure masking the region to the blue of the 1160 cm 1 fundamental prevents the observation of any such weak combination bands. In fluorene the spectrum is broad even at about 15°K. Figure 5.2. The polarized fluorescence spectra of pyrene-d1Q in biphenyl at about 10°K. The electronic origin band appears at 26 811 cm 1. -75-Table 5.5. The fluorescence spectrum of pyrene-d^Q in fluorene and biphenyl at about 10°K. + , t fluorene biphenyl analysis //b //c //b //c' 26 765 vs 26 811 vs 0-0; V -+ 1A 2u g 401 s 435a w 399 vs 436 mw 399, a g 436, b_ 465 w 464 m 464, b_ H * 476 vw 503b w 536 vw * 548 vw 572 m 572 ms § 612 vw 572, a g 690 m 690 ms 690, b, * 725 vw * 739 w 752 s 752 s 752, a g * 769 mw 801 w 798 mw * 803 w 2 x 399 836 w 833 mw 833, a g 836 w 833 m 833, b_ 3g 844* vw 864 vw 399 +464+1 -76-f luorene biphenyl //b //c //b //C analysis 884 w 972 w 1012 w 1163 s 1023 w 1038 w 1089 w 1109 w 867 w 879 mw 901C vw 923 vw 935b vw 948 vw 972 w 1006 w 1019* vw 1053 vw 1080 vw 1121 vw * 1140 mw 1160 s 1174* m 1208 w 1232 mw 916 vw 980 vw 1023 vw 1036 vw 1089 w 1113 w 879, a g 399 +572+1 2 x 503? 464 + 572 399 + 690 2 x 572 - 4 1160, a g 399 + 833 -77-fluorene biphenyl //b / / c //b //C analysis 1283 w 1324 w 1389 vs 1506 s 1563 w 1619 m 1241 m 1272 w 1300 w 1583 m 1280 mw 1315 w 1324 w 1336 vw 1350 vw * 1376 mw 1389 vs 1404* m 1474° w 1506 s 1519* m 1559 m 1570 mw 1620 mw 1630* w 1232 mw 1241 m 1267 mw 1296 w 1421 mw 1437 w 1540? vw 1585 ms 399 + 833 1241, b_ 3g 572 +690+5 1280, a ; 399 + 879 + 2 g 464 + 833 - l d 572 + 752 1389, a g 690 + 752 - 5? 1506, a g 399 + 1160 1585, b 3g 1620, a g -78-fluorene biphenyl analysis //b / / £ //b / / £ 1640 vw 399 + 1241 1665? vw 399 + 1267 - 1? 1680 vw 399 + 1280 + l d 1692b vw 1712 vw 833 + 879d 1732 w 572 + 1160 1745 vw 464 + 1280 + l d 1775* vw 1789 m 1788 ms 399 + 1389 * 1801 w 399 + 1404 - 2 1812 vw 572 + 1241 - 1 1825 vw 436 + 1389 1855 w 1851 mw 690 + 1160 + 1; 464 + 1389 - 2 1871 vw 1895* vw § I 1904 m 399 + 1506 - 1 1912 m I 1912 m 752 + 1160 1922* vw 1942 vw 436 + 1506? 1961 w 1961 mw 572 + 1389 1985 w 1984 mw 399 + 1585 1993 vw 833 + 1160d 1993 mw 833 + 1160 -79-fluorene biphenyl analysis //b / / £ //b / / £ 2019 vw 399 + 1620 2037§ vw 2078 w 2079 mw 572 + 1506 + 1 2078 w 2090* 2113C vw vw 2079 mw 690 833 + 1389 + 1280? 2130*°vw 399 + 572 + 1160 - 1? 2141 m 2141 2159* 2189 ms vw vw 2159* vw 752 752 572 2 x + 1389 + 1404 + 3 + 1585 - 2 399 + 1389 + 2 2195 w 2196 vw 803 + 1389 + 4; 690 + 1506 2225 w 2223 mw 833 + 1389 + 1 2225 w 2226§ w 833 + 1389 + 4d 2253 w 2250§ mw ^2250, a ; 752 + 1506 - 8 g 2274 w § 2271 mw 879 + 1389 + 3; 2271, a g 2274 w 2271 mw 2271, b_ 3g 2299 w 2289§ 2312* w w 2289, a g 2 x 572 + 1160 + 8?; 399 + 752 + 1160 + l d 2326 W } 2321 mw 2 x 1160 + 1 ( 2333* w 1160 + 1174 - 1; 833 + 1506 -- 6d 2339* vw 752 + 1585 + 2?; 833 + 1506 -80-fluorene biphenyl analysis //b //c //b //c' 2358 vw 399 + 572 + 1389 - 2 2373* vw 752 + 1620 + 1; 867 + 1506d? 2385* vw 879 + 1506 2394 vw 399 + 833 + 1006 + 1389 1160 - 1 + 2; 2394* vw 399 + 833 + 1160 + 2 2402 vw 1160 + 1241 + l d 2424 vw 1036 + 1389 - i ; 1160 + 1267 - 3 2444 vw 1160 + 1280 + 4? 2479 vw 1160 + 1315 399 + 572 + 572 + 752 + + 4; 1506 1160 + 2; - 5 2479 vw 399 + 690 + 1389 + 1 2505 vw 1113 + 1389 + 3 2511 vw 1121 + 1389 + i ; 1006 + 1506 -• l d 2552 m 2549 ms 1160 + 1389 *e 2562 w 2564 2602 m vw see footnote 1208 + 1389 f + 5? 2621 vw 399 + 833 + 1389d 2621* vw 399 + 833 + 1389 2629 w 2630 2656* w vw 1241 + 1389 1267 + 1389 2668 w 2666 m 1160 + 1506; 1280 + 1389 - 3 2677* w 2683* vw 1160 + 1519 1296 + 1389 - 2; - 2 1174 + 1506 - 3 -81-fluorene biphenyl lib lie lib //C analysis 2746 w 2776 m 2894 m 2972 w 3011 w 3172 w 2717 vw 2776 ms 2791* w 2818§ vw 2846?Cvw 2863*°vw 2894 ms 2907* w 2948 w 2960 w 3010 m 3022* w 3070§ vw 3123 vw 3134 vw 3175 w 3193§ vw 2746 vw 2818 vw 2974 mw 3088 vw 1315 + 1389 + 13: 572 + 752 + 1389 + 4; 399 + 2 x H60 - 2; 1232 + 1506 - 11; 1160 + 1570 - 13 1160 + 1585 + 1; 1241 + 1506 - l d 2 x 1389 - 2 1389 + 1404 - 2 g 572 + 2250 - 4 1389 + 1421 + 8; 1389 + 1437 - 8 572 + 2271 + 3 1389 + 1474 1389 + 1506 - 1 1389 + 1519 - 1; 1404 + 1506 - 3 399 + 1160 + 1389 1389 + 1570 + 1; 399 + 2564 - 3 1389 + 1585 1389 + 1620 + 1; 2 x 1506 - 2d 1389 + 1630 + 3; 1506 + 1519 - 3 399 + 1160 + 1506 +5; 1506 + 1570 - 6 1506 + 1585 - 3 572 + 1160 + 1389 + 2; 1506 + 1620 - 3 572 + 2564 - 2 399 + 2 x 1389 - 2 399 + 2791 + 3 - 8 2 -f luorene biphenyl //b //c //b //C analysis 3 3 0 5 5 w 3 4 0 7 w 3 5 2 7 w 3 2 3 9 vw 3 2 9 7 w 3 3 1 4 * vw 3 3 4 8 vw 3 4 1 0 vw 3 4 2 5 vw 3 4 5 0 vw 3 4 6 6 vw 3 4 7 8 * vw 3 5 2 9 w 3 5 8 2 vw 3 1 9 3 vw 3 2 0 5 vw 3 2 3 9 vw 3 2 5 7 * vw 3 3 5 3 vw 3 3 7 3 vw 3 3 8 2 vw 3 4 3 1 vw 3 4 6 6 vw 1 5 8 5 + 1 6 2 0 5 7 2 + 1 1 6 0 + 1 5 0 6 + 1 6 9 0 + 1 1 6 0 + 1 3 8 9 ; 4 6 4 + 2 x 1 3 8 9 - 3 3 9 9 + 1 3 8 9 + 1 5 0 6 + 3 ; 7 5 2 + 1 1 6 0 + 1 3 8 9 - 4 7 5 2 + 2 5 6 4 - 2 5 7 2 + 2 x 1 3 8 9 - 2 6 9 0 + 1 1 6 0 + 1 5 0 6 - 3 : 4 6 4 + 1 3 8 9 + 1 5 0 6 -"6d 3 9 9 + 1 3 8 9 + 1 5 8 5 8 3 3 + 1 1 6 0 + 1 3 8 9 1 1 6 0 + 2 2 5 0 ; 3 9 9 + 2 x 1 5 0 6 752 + 1 1 6 0 + 1 5 0 6 + 7 ; 1 1 6 0 + 2 2 7 1 - 6 1 1 6 0 + 2 2 7 1 1 1 6 0 + 2 2 8 9 + 1 5 7 2 + 1 3 8 9 + 1 5 0 6 - 1 6 9 0 + 2 x 1 3 8 9 - 2 1 1 6 0 + 2 3 1 2 + 6 ? ; 1 3 8 9 + 2 0 9 0 - 1? 7 5 2 + 2 x 1 3 8 9 - l d 2 x 3 9 9 + 2 x 1 3 8 9 + 4 ; 5 7 2 + 2 x 1 5 0 6 - 2 - 1 -83-f luorene biphenyl //b //c //b //C analysis 3639 w 3620 vw 3639 vw 3660 vw 3680 vw 3701* vw 3710 vw 3723 vw 3757 vw 3780 vw 3796 vw 3814 vw 3826 vw *b 3845 vw 3874§ vw 3582 vw 3608 vw 3660 vw 3701? vw 3723 vw 3790§ vw 3864? vw 690 + 1389 + 1506 - 3 833 + 2 x 1389 - 3 1389 + 2227 + 4 1389 + 2250 1389 + 2271 1389 + 2271 1389 + 2289 + 2 1389 + 2312 690 + 3010 + 1 2 x H60 + 1389 + 1 1389 + 2333 + 1; 1160 + 2564 833 + 1389 + 1506 - 5 d 1506 + 2250 + 1 1506 + 2271 + 3 1506 + 2271 + 13; , 1160 + 1241 + 1389 + 1 1506 + 2289 + 1 1506 + 2312 - 4 2 x H60 + 1506 833 + 2 x 1506 399 + 572 + 1389 + 1506 + 8; 572 + 752 + 1160 + 1389 + 1 3896 vw -84-fluorene biphenyl analysis //b //c //b //c' 3945§ w | 3935 w 1160 + 2 x 1389 - 3 3953 vw 4017 vw 1389 1241 + 2564 + 2 x 1389 - 2 4060§ w ! 4054 w 1160 + 1389 + 1506 - 1 1 4066 vw 1160 1506 + 1389 + 1519 -+ 2564 - 4 2; 4118§ vw 1232 + 1389 + 1506 - 9? 4133 vw 1160 + 1389 + 1585 -1241 + 1389 + 1506 -4160 w 4161 w 3 x 1389 - 6 *e 4182 w 4180* vw 1160 + 2 x 1506 + 8; 1389 + 2791 4278 w 4280 4292* w vw 2 x 2 x 1389 + 1506 - 4 1389 + 1519 - 5 4343§ vw 399 + 1160 + 2 x 1389 + 6; 2 x 1389 + 1570 - 5; 399 + 1389 + 2564 - 9 4360 vw 2 x 1389 + 1585 - 3 4385* vw 4396 w 4397 vw 2 x 1389 1389 + 1620 - 1; + 2 x 1506 - 4 4408* vw 1389 1404 + 1506 + 1519 -+ 2 x 1506 - 8 6; 44595 vw 399 1389 + 1160 + 1389 + + 1506 + 1570 -1506 + 5; 6 4516S vw 572 + 1160 + 2 x 1389 + 6; 1389 + 1506 + 1620 + i ; 3 x 1506 - 2 -85-The position of the origin is given in cm and all other entries show differences from the origin, t Intervals with assignments involving lattice modes are not listed. * shoulder S broad Some of the c-polarized intensity arises from an a assignment involving a lattice mode. b This line is essentially depolarized. The polarization of this line is uncertain. d The Franck-Condon assignment does not account for al l the line intensity. Some of this b-polarized intensity arises from an a assignment involving a lattice mode, f An estimate of the frequency shift expected as the result of an interaction of the magnitude observed here can be made as follows. Assuming that the in-tensity of the 2564 cm * interval (a..) arises from mixing only with the Franck--1 Condon allowed combination at 2549 cm (a ), then a1 corrected to first order H12 is a\ = o° + . c 0 a° . Since the ratio of the observed intensities (I.,/I,.) is i i 12 about 1/6 in the biphenyl spectrum, the frequency correction for o1 to second 1 -1 order is (?AE° ), i.e. about 3 cm and falls within the combined errors of o 1Z. measurement. g This assignment does not appear to account for the linewidth. CHAPTER 6 THE ABSORPTION SPECTRA 6.1 Room Temperature Spectrum The room temperature spectrum of pyrene-h^ in fluorene displayed in Fig. 6.1 shows that the two lowest-energy singlet transitions of pyrene have opposite polarization. If pyrene occupies a nearly substitutional site in the fluorene lattice, then the first and second electronic transitions are assigned ^B„ •*- *A short-axis polarized and B^, -«- ^ A long-axis polarized, & 2u g r lu g 6 r respectively. This result was confirmed in the absorption spectrum of pyrene in biphenyl where the first transition was b_ polarized and the second c_' polarized. In the oriented-gas model, the molar extinction coefficient at the origin -1 of the long-axis polarized second transition of pyrene is 142 000 I mole cm (3 e , ); the observed molar extinction coefficient of pyrene in fluorene solution" r J -1 -1 -1 at 28 840 cm in c_ polarization was 100 000 ± 15 000 % mole cm . Since the bandshape is essentially the same in the liquid and solid solutions i t is evident that there is an appreciable reduction of intensity in the fluorene matrix which apparently does not act solely as an orienting medium; the host molecules must perturb the pyrene guest molecules in such a way that intensity is transferred from the guest to the host absorption systems. The transition moment, M, to the upper state of the guest in the mixed crystal is given by \ f — S l i t w i d t h lOOOO • r -Z UJ o LL LU o CJ z o H u z r-X LU cr < o 8000 6000-4000-2000-0 0 I 27000 28000 29000 30000 W A V E N U M B E R (cm-1) 31000 Figure 6.1. The polarized absorption spectra of pyrene-h^g in fluorene at 300°K; full line //c, dotted line //b and broken line //a. -88-the expression deduced by Craig and Thirunamachandran [_602 for the deep trap limit , ( M* + .E'H\ .MS } I Aw£ - Aw* i where and are free-molecule transition moments to the r t b and s t b excited states of the guest (molecule 1) and host (molecule i), respectively, r s AWg - AWj_j is the difference between the free-molecule transition energies, is a dipole-dipole interaction energy and c is a normalizing constant. The calculation suggested by the above equation was carried out to see i f the theory predicts a diminution of intensity of the size observed. The origin band at 29 950 cm 1 of the second pyrene absorption system in benzene was treated as an independent transition having a dipole length of 0.657 A directed along the long molecular axis. The first and second absorption systems of fluorene, both long-axis polarized L~6lD, were taken at 33 050 cm 1 -1 o and 38 000 cm with transition dipole lengths of 0.372 and 0.882 A, respec-O-tively. Point dipole sums were taken over a sphere of radius 40 A; extending the radius limit to 150 A made l i t t l e difference to the sums. Then the ratio of the calculated oscillator strength to the corresponding oriented-gas value is 0.106 and the second origin of pyrene in fluorene should have a molar extinction coefficient of about 15 000 Z mole 1 cm 1. Thus, an intensity decrease is indeed expected although the theory appears to overestimate i t . It was assumed in the above calculation that pyrene occupies a substitu-tional site with its centre of mass at the position once occupied by the fluorene molecule and with its molecular axes exactly parallel to those of the undisturbed translationally-equivalent host molecules. Further the host -89-molecules adjacent to the guest were considered not to be displaced with res-pect to translation or rotation from their normal lattice sites. However, while pyrene has the same length as fluorene, i t is 1.7 A wider. If the nearest neighbour fluorene molecules in the ab_ plane (i.e. those within 9 A of the guest) are shifted 1 A away from the pyrene molecules along the line joining the molecular centres, the next nearest neighbours shifted 0.5 A with the positions of the remaining host molecules in the lattice unchanged and the calculation repeated then the molecular extinction coefficient in c_ pol-arization is estimated to increase from 15 000 to 52 400 £ mole - 1 cm"1. The localized distortion pictured above has involved two of the translationally-inequivalent sets of molecules; only a very small change in the calculated extinction coefficient is expected from displacements of fluorene molecules belonging to the other two sets. It is clear that a model such as this can account for the observed extinction coefficient. Since there are probably many sets of distortion parameters which give acceptable agreement with experiment, the details of the model were not sought. 6.2 Low Temperature Spectra 6.2.1 Pyrene-h.^ Microdensitometer tracings of the polarized absorption spectra of pyrene-h^g in biphenyl and fluorene matrices at about 10°K are shown in Fig. 6.2 and a partial vibrational analysis is given on the tracings. Unpol-arized information obtained from the n-paraffin spectra will be referred to although the intervals are not listed. Lattice modes are discussed later together with those observed in the ground state spectra. The polarization is more complete than in fluorescence to the extent that only the strongest Figure 6.2. The polarized absorption spectra of pyrene-h^g at about 10°K; upper curve //b_ in biphenyl, middle curve //c 1 in biphenyl and lower curve //c in fluorene. -91-long-axis transitions had very weak components along b_. In fluorene, the fraction of the short-axis transition intensity projected on to c_ indicates that the orientation of the ground state pyrene molecule is different from that of the host molecules enabling geometrically produced depolarization effects to be further enhanced through crystal-field mixing. As noted in Chapter 5, the biphenyl C^  site symmetry precludes any similar conclusion purely on the basis of observed polarization ratios. The frequency intervals referred to in the following discussions are taken from the spectrum in biphenyl or in n-heptane unless specifically indicated otherwise. 6.2.1(a) The first electronic transition The short-axis polarization of the pure electronic transition at 26 734 cm * confirms the assignment as ^B 2 u *Ag. That there can be no great change in the equilibrium geometries of the two electronic states is evidenced by the appearance (within reasonable exposure times) of intervals involving no more than two or three quanta of the strongest totally symmetric fundamentals and the absence of progressions in non-totally symmetric vibra-tions in both absorption and fluorescence, i.e. the Q matrices must be sensi-bly the same. However, there is a startlingly different distribution of intensity amongst the a g fundamentals of the excited electronic state from that of the ground state. For example, the analogues of the 408 and 1408 cm intervals prominent in the fluorescence spectrum are very weak or absent in absorption; in fact, in contrast to fluorescence, only two a intervals stand out, 583 and 1452 cm-1 in biphenyl and fluorene, and 573 and 779 cm 1 in the n-paraffins. Thus the fixed change in equilibrium geometry projects very differently on to the normal coordinates of the two electronic states; the -92-normal modes of motion are very different, a difference that must be reflected in the F matrices. Totally symmetric fundamentals of the first excited electronic state are marked by the intervals 400, 583, 782, 1030, 1249, 1339 and 1452 cm"1. B fundamentals are observed at 456, 496, about 721, 1111, about 1564 and perhaps 1162 and 1394 cm 1; these vibronic transitions acquire intensity by coupling 1 1 -1 with the B, A electronic transition at 28 404 cm . As the vibronic lu g state becomes more nearly degenerate with the pure electronic state with which it mixes, the coupling becomes very strong and the identification of b, fund-g amentals becomes less clear. For example, in the paraffin matrix spectra where the energy separation between the first two electronic transitions is greater than in biphenyl and fluorene, the most prominent line in the first absorption system lies 1564 cm 1 to the blue of the origin in ri-heptane; the corresponding interval is depressed to 1551 cm 1 in fluorene and to about 1500 cm 1 in biphenyl as the decreasing energy denominator in this series of matrices causes the vibronic interaction to become increasingly severe. The pair of lines near 720 cm 1 represent a possible example of Fermi-resonance. In the paraffin matrices, the lower-energy member has by far the greater intensity and is taken to represent the b_ fundamental; typically, in n-heptane the intensity of the 719 cm 1 interval is at least four times that -1 -1 of the 733 cm component. The fundamental is observed at 723 cm in fluorene where the presence of any weaker combination band would be obscured by phonon bands. Increased mixing of the two vibrational states appears to have occurred in the biphenyl matrix where the two lines have equal intensity. Because of the proximity of the *B^u electronic state in the single crystal spectra, combinations of a^ with b^ fundamentals are unexpectedly -93-strong and the usual Franck-Condon arguments that relate the intensity of the combination to its parent false origin no longer apply and the analysis becomes increasingly complex. Most of the fairly strong lines in the long-axis spectrum in the biphenyl and fluorene matrices from about 1423 and 1430 cm respectively, to the origin of the second absorption system prob-ably mark such combination bands; there are no such strong lines in the paraffin matrix spectra. Thus, a detailed vibrational analysis of the £'-polarized lines above 1111 cm * was considered impractical. In the energy region below about 1100 cm lines that are more probably matrix induced are located by comparing the absorption spectra in many sol-vents; for example, a very weak interval of 511 cm * appeared only in the n-heptane matrix, and a 635 cm * interval was fairly prominent in n-heptane and n-pentane yet absent in n-octane. This type of behaviour is similar to that observed in fluorescence. 6.2.1(b) The second electronic transition The origin of the second electronic transition of pyrene is located somewhere amidst the strong absorption in cj polarization of the spectrum in biphenyl (see Fig. 6.2), or in £ polarization in the fluorene matrix (see Fig. 6.1). Thus the transition is * B ^ u "*" * A g - The strongest feature in the biphenyl spectrum at 28 404 cm * was arbitrarily taken as the origin. The *B2u s y s t e m carries no intrinsic long-axis intensity, a l l the observed inten-sity in the £'-polarized spectrum being derived by vibronic interactions. However, these interactions continue to energies well into the ^B^u system, and thus absorption above 28 404 cm * records either transitions to Franck-Condon allowed vibronic levels of the *B^u electronic state or to Herzberg-Teller vibronic levels of the *B electronic state; matrix-induced lines -94-probably also appear in this region. When the energy difference between the interacting electronic levels is smaller than the vibrational spacing, the usual Herzberg-Teller theory C27H which estimates the mixing of electronic wave functions induced by a displac-ment of the nuclei along some chosen normal coordinate is no longer appropri-ate and instead in the present approach the interaction between vibronic wave functions is considered. Using iJ>?(q,Q ) as the zero-order function for the i t b excited electronic th state, a. .(CJ and al .(OJ for the j quantum vibrational states having symme-r J r J try a (unprimed) and b, (primed) respectively, the matrix element that § induces mixing between the basis states a n d ^2°2102O * S 3 H < 0 l l l«'21>«'ill < »l- ' JoX*?ltTTr'.l*2> It has been further assumed for simplicity that the normal coordinates Q are parallel in the two electronic states. The important consequence of this result lies in the appearance of the overlap factor <\°n\a2i^' Only those vibrational states need be considered that have an appreciable overlap with the vibrationless ground electronic state. Then the integral <\an\021^ selects the corresponding vibration of IJJ° and transfers some of the intensity from the transition that terminates on ,l'2a21a20 t C > t b e "^orbidden" transition that terminates on -^cj^ • It is in this way that the presence of the a fundamentals 400 and 1133 cm 1 belonging to the first electronic system are explained in the energy region of the second electronic system, even though these modes are either very weak or entirely absent in the spec-trum before the onset of the second system. Indeed, the overtone of the -95-400 cm fundamental is also observed. These modes appear with such unex-pected strength in the biphenyl spectrum because the transitions to the cor-responding fundamentals 393 and 1128 cm * of the second electronic system are strongly Franck-Condon allowed and the energy difference between zero-order states is probably very small. The following picture emerges from the above analysis: The sets of normal coordinates belonging to the first and second excited electronic states of pyrene are very nearly parallel to each other but not to the ground state set and, further, the changes in geometry between the ground state and the two lowest excited states must be different. Although no detailed analysis was attempted for the bands' of the first system between 1394 and 1512 cm * or near the second electronic origin, i t is clear that their intensity pattern repeats itself at intervals that correspond to the energies of totally symmetric modes. This is also a consequence of the presence of the integral <s°ijI a2i^ ^ n t n e matrix element of the perturbation operator. In fact, the analysis shown on Fig. 6.2 was carried through by relating the successive appearances of sharp lines to the 1423 cm * band. The broad features about 1472 and 1593 cm 1 to the blue of the second origin are probably too strong to be accounted for as combination bands alone, and as indicated on Fig. 6.2 two more fundamentals of the second electronic state may be located somewhere in these regions, the exact positions being obscured by the increasing complexity of the spectrum. The second absorption system was not well resolved in the fluorene spec-trum and this may be due to the activity of phonons allowed by the Franck-Condon principle. Although no analysis of this system was sought, the overall intensity distribution closely resembles that in the biphenyl spectrum. -96-6.2.2 Pyrene-d1Q Microdensitometer tracings of the polarized absorption spectra of pyrene-d^ in biphenyl and fluorene matrices at about 10°K together with a partial vibrational analysis are shown in Fig. 6.3. The general appearance of each spectrum is similar to that for the protonated molecule. 1 1 Transitions to the B„ and electronic states occur at 26 811 and 2u lu at about 28 472 cm" in biphenyl, and at 26 765 and about 28 757 cm-1 in fluorene. A fundamentals of the *B„ state are observed at 388, 554, 856, g 2u 1282 and 1444 cm"1, and b_ fundamentals at 429, 460, about 680, 829 and 3g 1231 cm one other a fundamental appears at 735 cm 1 in the b-polarized fluorene spectrum. In the region of the second electronic absorption system, a g fundamentals of the ^%2u. s t a t e appear at about 393, 828 and 1438 cm of course, these latter frequencies may be considerably perturbed through vib-ronic interactions (see, for example, the behaviour of the 1438 cm * line). The vibrational analysis of the c_' spectrum was not attempted near the origin of the ^^^u electronic transition. However, a probable b^ fundamental at 1545 cm * in octane appears at 1534 cm 1 in fluorene and at 1489 cm * in biphenyl, a shift brought about by the decreasing energy separ-ation between the two electronic transitions. In fluorene (as in the paraffin matrices) two lines appear in c_ polar-ization at 668 and 683 cm * with the higher-energy line having the greater intensity. However, in biphenyl these lines are not resolved and give rise to an uncharacteristic broad line at about 680 cm This behaviour is analogous to that already noted in the spectra of the protonated molecule. The analysis of the second system is complicated in the region 800 -1200 cm * to the blue of the origin by the weakness of the lines and in the Figure 6.3. The polarized absorption spectra of pyrene-d10 at about 10°K; upper curve //b in biphenyl, middle curve //c_' in biphenyl and lower curve //c_ in fluorene. -98-region 1400 - 1600 cm * by the large number of modes possible. Thus, the analysis is deliberately left incomplete. Some arrows placed above the b-polarized biphenyl spectrum indicate possible b^ fundamentals of the *B^u state which appear through a vibronic interaction with a yet higher-energy *B2u state; these bands do not arise from impurities since they do not appear at the same frequencies in the spectra of the protonated (Fig. 6.2) and deuterated molecule (Fig. 6.3). The sharp line superimposed on the broad feature in Fig. 6.3 represents the overtone of the 1444 cm * funda-mental . Correlations amongst the fundamentals of the two isotopic pyrene mole-cules in the ground and excited electronic states are shown in Tables 6.1 and 6.2. These correlations are necessarily approximate since the related frequencies do not always have similar intensities in absorption and fluor-escence so that some normal motions appreciably change (and so lose their identity) in different electronic states. In n-octane, broad absorption bands about 210 and 200 cm * to the red of the principal monomer origin of pyrene-h^Q (26 870 cm *) and of pyrene-d^^ (26 949 cm *), respectively, are thought to mark the origins of a transition arising from a dimeric or higher aggregate species. Broad vibrational struc-ture also underlies the monomer spectra (see Fig. 6.4). Although absorption attributed to a dimer has not been observed previously (see Chapter 1), presumably in the present case certain aggregate configurations of pyrene are stabilized by the arrangement of the surrounding octane molecules. 6.3 Triplet-Triplet Absorption The triplet-triplet absorption spectrum of pyrene-d1Q in fluorene at -99-Table 6.1. A correlation of the fundamentals of pyrene-h.^ and pyrene-d^ in the ground and excited electronic states. pyrene-h1Q pyrene-d1Q '''A state . g *B_ state 2u 1B 1 state lu ''"A state g B^,, state 2u '''B, state lu 408 vs 400 vw 393 m 399 vs 388 vw 388 m 596 s 583 ms 572 ms 554 ms 801 s 782 w 683? vw 752 s 735a 1063 s 1030 mw .... 833 m (828)b 812? vw 1144 mw (1133)b 1128 w 879 m 856 mw 1240 s 1249 mw 1237? vw 1160 s .... 1355? mw 1339 mw .... 1280 mw 1282 w 1408 vs .... 1434? mw 1389 vs .... .... 1552 s 1452 ms 1472? w 1506 s 1444 ms .... 1631 mw 1593? mw 1620 m This entry is taken from the fluorene spectrum; all other entries are from the biphenyl spectrum. b This interval appears only in the region of overlap between the first and second systems and so has been subjected to a vibronic perturbation. -100-Table 6.2. A correlation of the b^ fundamentals of pyrene-h^ and pyrene-d in the ground and excited electronic states. pyrene-h. pyrene-d A^ state g ^B„ state 2u *A state g state 2u 456 m 456 w 436 m 429 vw 496 mw 496 vw 464 m 460 vw 736 s 721 mw 690 ms 680 w 1111 s 1111 s 833 m 829 w 1176? mw 1162? w .... .... 1241 m 1231 mw .... 1394? mw .... 1597 ms 1564a vs 1585 ms 1545a vs This entry is taken from a spectrum in a paraffin matrix; a l l other entries are from the biphenyl spectrum. -101-ENERGY Figure 6.4. The dimer absorption spectrum at about 10°K of (a) pyrene-h.^ in n-octane, and (b) pyrene-d^^ in n-octane. The sharp features mark the overlying absorption and fluorescence spectra of the monomer. The origins of the monomer absorption systems have been made coincident. -102-77°K is shown in Fig. 6.5. It is clear that there are two distinct absorption systems, one at about 18 400 cm ^ polarized along the short molecular axis and a second at about 23 850 cm * polarized parallel to the long molecular axis. This confirms the result of Pavlopoulos [63] who determined these polarizations using a photoselection method. 0.9 1 6 O O 0 n 18000 2COOO 22COO 2 4 0 0 0 2 6 0 0 0 j/Ccrrr1) Figure 6.5. The polarized triplet-triplet absorption spectra of pyrene-d^ in fluorene at about 77°K. The presence of the small peak at 17 650 cm 1 is not understood. It is too strong to be a hot band, and the overall intensity [64], shape and polari-zation of the system is not consistent with the possibility that the transition is intrinsically weak or forbidden gaining intensity through vibronic inter-actions. Since this weak band did not appear in the spectrum in an n-hexane matrix or in the previous spectra [63-65j i t is probably in some way charac--103-teristic of the matrix chosen. The spectrum in both the single crystal and n-paraffin matrices did not appreciably sharpen on cooling the samples to about 15°K. The symmetry of the lowest triplet state has not been established so that no firm assignment of the excited triplet states can be given. An analysis C66H of the e.p.r. spectrum suggests that the lowest triplet state has B^u symmetry and the results of two recent calculations C67H agree on the symmetry assignment of the first two allowed triplet-triplet transitions of pyrene and are consistent with the available experimental evidence. Thus the two systems 3 3 observed probably mark the transitions B^ ^ •*- B^u (short-axis polarized) and 3 3 Ag B^u (long-axis polarized) . 6.4 Lattice Modes In the absorption spectrum of pyrene in fluorene, fairly sharp lattice modes are observed at (values for pyrene-d^^ are given parenthetically) 24 s (20 s) and 52 cm 1 m (48 m); the intensity scale corresponds to an origin intensity vs. The fluorescence lines are diffuse to the red and i t is difficult to distinguish maxima. The main electronic origin appears at 26 690 cm 1 but a second very weak origin, seen only in fluorescence when excitation is from the front face, appears at 26 724 cm 1 and is taken to mark some pyrene molecules at a second site in the lattice; the stronger vibronic bands also have very weak companion-lines 34 cm 1 to the blue. Phonons of 17 s (15 s), 31 ms (26 ms), 243 mw (222 mw), 243 + 17 w (222 + 15 w) and (257 cm-1 w) are associated with the principal origin. The intervals greater than 200 cm-1 probably mark internal modes, but since intermolecular and low frequency intramolecular motions are not independent they are con--104-sidered here with the lattice modes. All these satellite lines have the same polarization as the pure electronic transition. The phonons are weakly active in the spectra because the excited and unexcited pyrene molecules occupy different equilibrium positions in the fluorene lattice. Because the dependence of the electronic wave function for the pyrene molecule on the intermolecular (repulsive) potential is slight, the phonon lines are weak. Since a change of the Tr-electrons of pyrene only are involved, the phonons active in the spectra must represent molecular motions which bring together carbon atoms of pyrene and hydrogen atoms of adjacent fluorene molecules and hence which are restricted to the ab plane. It is difficult to be more explicit than this. The1 low frequency a^ modes of the ordered fluorene lattice are not known E68H but as the carbazole structure is very similar C 6 9 ] , all three a modes (k = 0) probably lie above 40 cm • Of course, the pyrene molecule is situated in a disturbed region of the lattice where the lattice modes may be localized and the quantum number k_ does not apply. In any event such low frequency phonons as observed here probably connect with the acoustic branches well away from the origin of the Brillouin zone, a conclusion also reached by Zahlan [70], and Pawley and Yeats C7lU when considering the spectra of pure organic crystals. The role played by phonons in the spectra of pyrene in biphenyl is more obscure. No lattice modes are observed in absorption (although the complex-ity of the spectra in the region of the pure electronic *B l u •*- 1A g transition may be due in part to phonons) while in fluorescence an exceptionally large number appear at 17 ms, 34 ms (34 ms), 47 sh (47 sh), 67 m (67 m), 93 sh (93 sh), 107 m (108 m), 136 m (136 m) , 144 m (145 m), 156 mw (154 mw), 175 mw (177 mw), 208 w, 217 vw (214 vw), 243 vw (247 vw), 262 w (266 w), 279 vw and -105-(333 cm vw). These lines are associated only with the electronic origin. However, the 34 cm 1 interval produced very weak satellites to the red of the more intense lines in the spectrum, such as 408, 596, 801, 1408 and 1552 cm 1. Unlike what is observed in the fluorene matrix, the lower-energy intervals at least do not depend on the isotopic impurity molecule. However, a l l the lines (except 156 and 279 cm 1 which are essentially depolarized) retain the same polarization characteristics as the origin. No attempt was made to analyze this structure which was not understood. CHAPTER 7 CONCLUSION This work was undertaken in the first instance to determine from experi-ment the correct symmetry assignment of the first excited singlet state of pyrene. A knowledge of these electronic energy levels provides a very impor-tant contribution to our understanding of chemical bonding in aromatic com-pounds. Little attention had been paid to pericondensed aromatic molecules, and what l i t t l e experimental and theoretical work that was available for pyrene showed many inconsistencies. In an effort to locate (and assign) the second pure electronic transition, a vibrational analysis had to be carried through for the first system. This led to a second area of uncertainty: There were too many bands in the fluor-escence (and probably absorption) spectra in the n-paraffin solvents to be readily accounted for by the Franck-Condon principle, even allowing for the presence of some b_ fundamentals acting as false origins. The occurrence of •^g these lines in paraffin matrices is confirmed, and an insight into how they arise is obtained by an examination of the spectra of pyrene in single crystal matrices. Vibrational analyses of pyrene had already been reported [13,141). A more complete analysis has been attempted here making use of the results of laser-Raman experiments and frequencies calculated using an approximate force field. -107-Solid solutions are generally formed when both host and guest molecules have similar physical dimensions. For all the matrices used in this work, pyrene is somewhat bigger (usually wider) than the host molecule i t replaces so that the lattice must suffer a localized distortion. Concentrations of the pyrene dopant were kept low to avoid the formation of pyrene crystallites (and typical broad excimer fluorescence), although there was evidence that aggregation may have begun to occur in the octane matrix. However, the effect of this compression is to enhance the exchange repulsion between adjacent molecules to such an extent that intramolecular vibrations can mix. If the nature of the distortion was known, the symmetries of the intramolecular modes that could mix could be found; the reverse process, using the knowledge of which vibrations are crystal induced to define the distortion geometry is not practical because the large number of modes induced does not allow a unique path for the intensity transfer to be determined. However, the sheer number of lines observed in the spectra allows a rather startling conclusion to be drawn, viz., in one matrix or another used in this work, every fundamental with energy less than about 1000 cm 1 may be appearing in the fluorescence spectrum; above 1000 cm 1 some induced lines more often may be accounted for energetically in terms of combinations. This s t i l l does not allow the sym-metries of fundamentals to be assigned (even below 500 cm *) since expected errors in the calculated frequencies are such that the assignments of vib-rations adjacent in energy could well be interchanged. Crystal-induced effects, although more evident in fluorescence, were also seen in the absorp-tion spectra. The first electronic transition provides intensity to the b-polarized spectrum, where the intensity distribution is not the mirror image of that -108-observed in fluorescence. Since there is no long progression in any frequency, the force fields in the two electronic states are different. However, the important perturbation observed in absorption was Herzberg-Teller coupling. The intensities in c_< polarization increased sharply just before the onset of the second transition. This region of the spectrum is complex and the electronic origin of the second transition could not be located. An analysis of the prominent features of the c_* spectrum was achieved by noting that the two absorption systems suffer different solvent shifts so that the strength of the Herzberg-Teller interaction was modulated by the changing energy denominator. For example, a b fundamental of the •^g •^ B^  electronic state appearing at 1564 cm * in n-heptane was found at about 1500 cm * in biphenyl, while a b^ combination that was prominent at 1423 cm * in the biphenyl matrix could not be located in the paraffins. Of course, the frequency of the vibration and the intensity with which i t appears in the spectrum depends on the strength of the interaction. A final unexpected result is that the intensities of the a fundamentals F g built on b„ false origins of the first system are accounted for in terms of 3g 6 Franck-Condon overlap factors connecting the two excited electronic states. The similarity between the intensities of the a fundamentals built on false ' g origins of the first transition to those built on the second electronic ori-gin itself indicate that the sets of normal coordinates lie roughly parallel in the two excited electronic states (but they are not parallel to the set in the ground electronic state). The third electronic transition was not observed in any of these exper-iments although some b-polarized bands at high energy were taken to represent b„ vibrations of the second absorption system. This suggests that the third -109-electronic transition is B„ •*- A , as has been found already in photoselec-2u g' J f tion experiments L~7]. REFERENCES [ l l J. Piatt, J. Chem. Phys. 22_, 1448 (1954). [2]. N.S. Ham and K. Ruedenberg, J. Chem. Phys. 25_, 13 (1956). [3D. G.J. Hoijtink, N. Velthorst and P. Zandstra, Mol. Phys. 3, 533 (1960). [4]. R.L. Hummel and K. Ruedenberg, J. Chem. Phys. 66, 2334 (1962). [5]. E.M. Evleth and W.L. Peticolas, J. Chem. Phys. 41_, 1400 (1964). [6]. R. Williams, J. Chem. Phys. 26, 1186 (1956). [7]. G.J. Hoijtink and P.J. Zandstra, Mol. Phys. 3, 371 (1960). [8]. H. Zimmerman and N. Joop, Z. Elektrochem. 65, 138 (1961). [9]. L.E. Lyons and G.C. Morris, J. Chem. Soc. 3661 (1957). [10]. J. Ferguson, J. Chem. Phys. 28, 765 (1958). [ l l ] . J. Tanaka, Bull. Chem. Soc. Jap. 38_, 86 (1965). [12]. R.M. Hochstrasser, J. Chem. Phys. 33_, 459 (1960). [13]. R. Mecke and W.E. Klee, Z. Elektrochem. 6_5, 327 (1961). [14]. S. Califano and G. Abbondanza, J. Chem. Phys. 39_, 1016 (1963). [15]. N. Neto and C. di Lauro, Spectrochim. Acta A26, 1175 (1970). [16]. A. Schmillen, J. Phys. Chem. 69, 751 (1965). [17]. T. FBrster and K. Kasper, Z. Elektrochem. 59, 977 (1955). [18]. J.N. Murrel and J. Tanaka, Mol. Phys. 7_, 363 (1964). [19]. A. Camerman and J. Trotter, Acta Crystallogr. 18_, 636 (1965). [20]. E.V. Shpol'skii, A.A. Il'ina and L.A. Klimova, Dokl. Akad. Nauk SSSR 87_, 935 (1952). [21]. E.V. Shpol'skii, Sov. Phys.-Usp. _3, 372 (1960). [22]. R.M. Hochstrasser and G.J. Small, J. Chem. Phys. 48_, 3612 (1968). [23]. L.A. Klimova, Opt. Spectrosc. (USSR) 15_, 185 (1963). -111-[24]. L. Pesteil, R. Troisplis and P. Pesteil, J. Chim. Phys. 60, 1294 (1963). [25]. A. Pellois and J. Ripoche, Chem. Phys. Lett. _3, 280 (1969). [26]. M. Born and R. Oppenheimer, Ann. Physik. 84_, 457 (1927). [27]. G. Herzberg and E. Teller, Z. Physik. Chem. B21, 410 (1933). [28]. D.P. Craig and R.D. Gordon, Proc. Roy. Soc. (London) A288, 69 (1965). [29]. A.S. Davydov, Zh. Eksp. Teor. Fiz. 18, 210 (1948). [30]. J. Frenkel, Phys. Rev. 37, 17, 1276 (1931). [31]. D.S. McClure, J. Chem. Phys. 22, 1668 (1954). [32]. D.P. Craig and L.E. Lyons, J. Chem. Phys. 20_, 1499 (1952); 0. Schnepp and D.S. McClure, J. Chem. Phys. 21_, 959 (1953). [33]. D.S. McClure and 0. Schnepp, J. Chem. Phys. 23, 1575 (1955); D.P. Craig and S.H. Walmsley, Mol. Phys. 4_, 113 (1961). [34]. R.G. Sangster and J.W. Irvine, Jr., J. Chem. Phys. 24_, 670 (1956). [35]. E.J. Bowen and K.K. Rohatgi, Discuss. Faraday Soc. 14_, 146 (1953). [36]. H. Hock, S. Lang and G. Knauel, Chem. Ber. 83, 229 (1950). [37]. J. Feldman, P. Pantages and M. Orchin, J. Amer. Chem. Soc. 7_3, 4341 (1951); D.D. Phillips, Org. Synth. 34, 31 (1954). [38]. E. Clar, C.A. 41_:6553h; K. Alder, F. Pascher and H. Vagt, Ber. Deut. Chem. Ges. B75, 1501 (1942). [39]. G.N. Lewis and M. Kasha, J. Amer. Chem. Soc. 66_, 2100 (1944). [40]. E.A. Johnson, J. Chem. Soc. 994 (1962). [41]. D.M. Burns and J. Iball, Proc. Roy. Soc. (London) A227, 200 (1955). [42]. A.N. Winchell, The Optical Properties of Organic Compounds, Academic Press, New York, 1954. [43]. J. Trotter, Acta Crystallogr. 14_, 1135 (1961). [44]. A. Bree, R.A. Kydd, T.N. Misra and V.V.B. Vilkos, to be published. [45]. P.W. Bridgman, Proc. Amer. Acad. Arts Sci. 60, 303 (1925). [46]. A. Bree and T. Thirunamachandran, Mol. Phys. 5_, 397 (1962). -112-[47]. M. Kasha, J. Opt. Soc. Amer. 38, 929 (1948). [48]. W.H. Duerig and I.L. Mador, Rev. Sci. Instrum. 23_, 421 (1952). [49]. W.R. Hess, H. Hacker, H.W. SchrStter and J. Brandmiiller, Z. Angew. Phys. 27, 233 (1969). [50]. N. Netto, M. Scrocco and S. Califano, Spectrochim. Acta 22, 1981 (1966). [51]. V. Schettino, N. Netto and S. Califano, J. Chem. Phys. 44, 2724 (1966); V. Schettino, J. Chem. Phys. 46_, 302 (1967). [52]. D.B. Scully and D.H. Whiffen, J. Mol. Spectrosc. 1, 257 (1957). [53]. D.E. Freeman and I.G. Ross, Spectrochim. Acta 16_, 1393 (1960). [54]. D.H. Whiffen, Phil. Trans. Roy. Soc. (London) A248, 17 (1955). [55]. D.B. Scully and D.H. Whiffen, Spectrochim. Acta 16>, 1409 (1960). [56]. D. Steele, Spectrochim. Acta 22_, 1275 (1966). [57]. J.H. Schachtschneider, Tech. Report No. 57-65, Shell Development Co., California, U.S.A. [58]. J.B. Coon, R.E. DeWames and CM. Loyd, J. Mol. Spectrosc. 8_, 285 (1962); W.L. Smith, Proc. Phys. Soc. (London) 89_, 1021 (1966). [59]. See, for example, N. Rich and D.A. Dows, Mol. Cryst. 5_, 111 (1968) and other references contained therein. [60]. P. Avakian and E. Abramson, J. Chem. Phys. 43_, 821 (1965). [61]. D.P. Craig and T. Thirunamachandran, Proc. Roy. Soc. (London) A271, 207 (1963). [62]. A. Bree and R. Zwarich, J. Chem. Phys. 51_, 903 (1969). [63]. T.G. Pavlopoulos, J. Chem. Phys. 52_, 3307 (1970). [64]. W. Heinzelmann and H. Labhart, Chem. Phys. Lett. 4_, 20 (1969). [65]. D.P. Craig and I.G. Ross, J. Chem. Soc. 1589 (1954); G. Porter and M.W. Windsor, Proc. Roy. Soc. (London) A245, 238 (1958). [66]. O.H. Griffith, J. Phys. Chem. 69, 1429 (1965); S.W. Charles, P.H.H. Fischer and C.A. McDowell, Mol. Phys. 9_, 517 (1965). [67]. M.K. Orloff, J. Chem. Phys. £7, 235 (1967); D. Lavalette, J. Chim. Phys. 66, 1845 (1969). - u s -ees]. A. Bree and R. Zwarich, J. Chem. Phys. 5_1, 912 (1969). [69]. A. Bree and R. Zwarich, J. Chem. Phys. 49_, 3344 (1968). [70]. A.B. Zahlan, Excitons, Magnons and Phonons in Molecular Crystals, ed. A.B. Zahlan, Cambridge University Press, 1968, p. 153. [71]. G.S. Pawley and E.A. Yeats, Solid State Commun. 7, 385 (1969). APPENDIX - S I T --9XT-The fluorescence spectrum of pyrene-h in n-octane at about 10°K. -119-Th e fluorescence spectrum of pyrene-hin in some n-paraffin matrices at about 10°K. n-pentane n-hexane n-heptane n-octane analysis -98 m -111 vw -75 vw -78 vw * -71 vw -71 m -58 vw -60 ms -58 ms These bands are located to the -50 mw -48 vw blue of the principal origin -43 mw and this is indicated by the -37 vw -33 w -38 vw -37 mw minus sign before each entry. -27 m -21 m -22 vw -6 m * -17 vw 26 916 s + 26 836 s + 26 868 s + 8 m 26 870 t s 0-0; 1B. -+ 1k 2u g 13 m 17 m 20 mw 30§ m 27 § mw 25 m 33 mw *§ 43 mw * § 48 vw 65 mw * § 79 vw 84 111 m w 127 vw 123§ vw 146 vw 149§ vw 171 w 182§ vw -120-n-pentane n-hexane n-heptane n-octane analysis 198 vw >• 206 vw 204 vw 226 mw 238 vw * 244 vw 254§ vw 252 vw 253 vw 268 vw 261 vw 260* 318 337 vw vw vw 408 s 404 s 405 s 405 s 405, a g 456 ms 454 ms 455 ms 454 ms 454, 3g 488 m 499 m 497 m 498 m 497 m 497, 3g 526 vw 528 541 w vw * 527 § 555 w vw 565 vw 559 vw 577 vw 574 vw 590 m 586 w 588 mw 587 w 587, a g 621 w 616 vw 615 w 614 w 656 vw 678 ms 676s mw 714 m 709 w § 703 vw 737 s 734 s 735 s 734 755 s mw 734, b* 3g -121-n-pentane n-hexane n-heptane n-octane analysis 777 vw 779 m 788* mw 802 815 vs m 800 vs 809* m 801 809 vs m 800 809 828 vs m mw 800. a g 2 x 405 - 1 § 842 mw § 833 mw § 861 w 857§ w 859 vw 405 + 454 883 m 881 mw 881 m 880 m * 895 w 408 + 488 - l f * 895 § 906 w vw 897 vw 903* vw 900 w | 405 +497-2 § 918 vw 928* vw § 924 vw § 924 vw § 943 vw § 941 vw § 943 vw 963 w § 966 vw § 961 vw 998 mw 991 w 993 mw 992 mw 405 + 587f 1013 vw 1006 m 1007 m 1005 w 1031 mw 1026 mw 1025 mw 1026 mw 1046 mw 1040* m 1042 w 1041 mw 454 + 587f 1069 ms 1068 mw 1065 ms 1065 1097* ms w 1065, a g 1108 vs 1105 s 1105 s 1104 vs 1104, b_ 3g * 1110 m 1144 m 1140 m 1140 m 1138 m 1138, a ; 405 + 734 - 1 g -122-n-pentane n-hexane n-heptane n-octane analysis * 1152 w * 1148 w 1168 w 1163 w 1181 w 1172 m 1181 w 1173 w 2 x 587 - l f 1188* vw 408 +779+1 1210 ra 1204 w 1206 mw 1204 m 405 +800-1 1233 m 1231 m 1232 ms 1228 ms ) 1241 ms 1238 mw 1239 ms 1237 ms } ^ 1233, a 1268 mw 1257* vw 1291* vw 1287 vw 1284 vw 405 + 880 - 1? 1304§ w 1299§ vw 1298§ vw 1327 m 1321 mw 1323 m 1321 m 587 + 734f 1349 mw 1345 w 1358 mw 1353 w 1368 m 1367 mw 1366 ms 1366 mw 1386 m 1382* w 1392 m 1389 m 1389 m 1387 m 587 + 800f 1407 vs 1405 s 1405 vs 1403 vs 1403, a g 1423§ m 1423 vw 1422 m 1422 m 1454§ mw 1459 vw 1450§ w § 1452 w 1472 vw 1470£ mw 1468 mw 405 + 1065 - 2 1477 m * * 1 1482 vw 1475 mw 1475 mw 1504 m 1501 m 1496§ m 1499 m 1516 mw 1510* vw 1509 mw 1508 mw 405 + 1104 - 1 -123-n-pentane n-hexane n-heptane n-octane analysis 1523 mw 1521 vw 1519 mw 1517 mw 454 + 1065 - 2 £ 1549* mw * 1544 mw * 1542 mw 405 + 1138 - 1; 2 x 405 + 734 -• 2 f 1554 ms 1551 m 1550 ms 1550 ms 1550, a g 1566 m 1561 mw 1559 m 1559 m 1574 mw § 1576 mw 1572 mw 1571 w 1599 m 1595 mw 1594§ m 1593 m 1593, b_ 3g 1619§ vw 1618 vw 1611 vw 2 x 405 + 800 + 1? 1628 mw 1628 mw 1627 mw 1627, a 1634 m I g 1635 mw 1637 m 1634 mw 405 + 1228 + 1 1649f m 1642f mw 1644 m 1642£ m 405 + 1237 1658* vw * 1651 vw 1653 vw 1652 vw 587 + 1065f 1668§ w 1663 vw 1661 mw 1675 mw 1675 w 1672 mw 1687 w 1680 mw 1686* w * 1682 mw 454 + 1228 1698 m 1692 mw 1694 m 1692 m 454 + 1237 + l f 1709 mw 1711 vw 1708 w 1705 mw 1730 mw 1729 vw 1728 w 1726 mw 1741 mw 1738 w 1736 mw 1757 w 1752 w 1752 w 1768 vw 1770 vw § 1765 vw 1760* vw 1783 mw 1782 vw 1783 vw * 1775 vw 1793 mw 1794 mw 1783 w 1806 m 1802* m 1802 m 1799 m 734 + 1065f -124-n-pentane n-hexane n-heptane n-octane analysis 1815 ms 1809 ms 1810 ms 1808 ms 405 + 1403 1823* mw 1827 vw 1818 w * 1818 w 587 + 1228 + 3 £ 1831 mw 1826 w 1827 w 587 + 1237 + 3; 405 + 1422f 1855 m 1857§ ( ( m 1856 m 1863 m 1859 m 1 I 456 + 1407 i 1867 mw 1866§ ( 800 + 1068 - l 1871 m < 1874 w m j 1866 • i * 1881 vw 1889* vw 1878* vw 1880* vw 405 + 1475? 1895 w 488 + 1407f 1908 w | 1902§ w 1899 vw 1901 w 497 678 + + 1403 + 1233 -l 3 1922§ mw | 1922* vw 1919§ w 1921 mw 678 + 1241 + 3 1939 w 1937§ 1943§ 1932 vw * w < vw < * 1955 mw 1946 w 880 + 1065 + l f 1961 mw 1955 w 1955 w 1954 mw 405 + 1550 - 1 1970* mw 1965 mw 1966 w 1963 mw 734 + 1228 + i ; 405 + 1559 - 1 1978 m 1973 m 1974 mw 1972 m 734 + 1237 + i f 1997 mw 1990* w 1996§ I 1990 mw 587 + 1403f mw < 2005 mw 1999 w i 1998 mw 405 + 1593 2010* w 2005* vw 2006* vw 2002* mw 454 + 1550 - 2 f 2025 vw 2022* vw 2022* vw 2017 vw 614 + 1403 2035 mw 2031 mw 2033 mw 2029 mw 800 + 1228 2043 m 2039 m 2040 m 2038 m 800 + 1237 + i ; 2 x 405 + 1228?f -125-n-pentane n-hexane n-heptane n-octane analysis 2057 vw 2046* vw 2046* vw 2 x 405 + 1237 - 1? 2067 vw 2068§ vw 2061§ vw 2085 mw 2081§ vw 678 + 1407 2094* vw 2082§ vw 2103 vw 2096 vw 405 + 1692 - 1? 2123 mw 2112 -1 2109 vw 2107 vw { 703 + 1403 + 1 880 + 1228 + 1 £ 2121 vw 2120 vw 2120 w 880 + 1237 + 3 ; 800 + 1321 -• 1 ? £ 2139* w 2139 "I 2133* w 2132* w 2 x 1065 + 2 2144 m 1 2140 m 2137 m 734 + 1403 * 2151 vw * 2147 vw 2175 mw 2172 w 2171 w 2170 2178* mw vw 1065 + 1104 + 1 £ 2187 mw 2192* w 788 + 1403 + 1 2210 ms 2206 ms 2206 ms 2204 ms 800 + 1403 + 1 ( 2211* w 2215* w 2212* w 2 x 405 + 1403 - 1 2224 mw 2227§ I w 2221 2229 w vw 800 + 1422 - 1 828 + 1403 - 2? 2334§ w 2239* vw ) 2252* vw § 2252 vw 2253* vw | see footnote g 2267§ w 2263 vw 2261 w J 2290 mw 2285 mw § 2291 \ 2283 mw 880 + 1403; 734 + 1550 - 1 * mw \ f 2302 w 2292 w I 2292 w 1065 + 1228 - 1 2310 mw 2303 w 2303 mw 2302 mw 1065 + 1237£ -126-n-pentane n-hexane n-heptane n-octane analysis * 2314 vw 2310* vw 2328? vw 2329* vw 2326* vw 405 + 1921? 2340 mw 2337 mw 2336 mw 2334* mw 1104 + 1228 + 2 2349 m 2343 mw 2344 m 2342 m 1104 + 1237 + l f 2356* w 2351* vw 2351* w 2350* mw 800 + 1550 2369 w 2361 vw 2360 w 2360 w 800 + 1559 + 1?; 2 x 405 + 1550f 2384 w 2375 vw § 2375 w 2375 mw 1138 + 1237; 405 + 734 + 1237 - l f 2396 VW 1 2394§ 2392? vw vw 2391 w 1065 + 1321 + 5; 992 + 1403 - 4 2406 vw ) 2420 vw 2413 w 2405 vw 1005 + 1403 - 3 f 2437 w 2432 vw 2430 w 2429 mw 1026 + 1403f 2452 w 2446* w 2445 w 2443 mw 1041 + 1403 - l f 2460* vw 2456 w 2453 w 2476 m 2472 mw 2470 m 2468 m 1065 + 1403 2493 mw | 2487 vw 2487 vw 2499* vw 408 + 678 + 1407 1097 + 1403 - 1 2516 ms 2510 m 2510 m 2508 2515* m vw 1104 + 1403 + 1 1110 + 1403 - 2 2551 mw 2545 w 2545 w 2542 mw 1138 + 1403 + 1; 405 + 734 + 1403 2569 w 2565 vw 2563 vw 2559 w 1237 + 1321 + l f 2594§ vw 2580§ vw 2610* w 2604* vw 2603* vw 1237 + 1366f -127-n-pentane n-hexane n-heptane n-octane analysis 2620§ ( 2608 vw 26135 ( 2608* mw 405 + 800 + 1403 mw \ * mw < * I 2617 vw i 2615 mw 1065 + 1550; 1228 + 1387 2633* w 2626* vw 2628* vw 2624* mw 1065 + 1559; 1237 + 1387f 2640 mw 2636 mw 2637 m 2632 mw . 1228 + 1403 + 1 2648 m 2643£ mw 2644 m 2641£ m 1237 + 1403 + 1 2662 w 2654 vw 2657 w 1104 + 1550 + 3 2676 vw | 2668* vw 1268 + 1407 + 1 2703§ vw 2693§ w 1138 + 1550 + 5; 1065 + 1627 + 1; 405 + 1065 + 1228 - 5 2716 vw 2707§ w 1228/1237 + 1475 ± 4/5; 405 + 1065 + 1237 2735 w 2728 w 2730 w 2725 w 1321 + 1403 + 1; 1228 + 1499 - 2 2743 w 2738 w 2737 w 1237 + 1499 + 1 £ 2756 vw 2747 w 1345 + 1403 .- l f 2765 vw 2756 w 1353 + 1403 f 2775 w 2771 vw 2772 vw 2769 w 1366 + 1403 2793§ ( 2781 vw 2782 mw 2778 w 1228 + 1550 mw •< I 2791 vw 2789 mw 2789 mw 1237 + 1550 + 2; 1387 + 1403 - 1; 1228/1237 + 1559 ± 2/7?f 2814 ms 2808 m 2809 ms 2806 ms 2 x 1403 2833§ mw 2827 vw 2830§ mw 2827 mw 1403 + 1422 +2; f 1228/1237 + 1593 ± 6/3 2873 vw 2869§ ( 2864* vw 1237 + 1627f vw { ( ( 2877 vw 2873 w 405 + 1065 + 1403 2883 w * * ( 2883 vw 2878 w 1403 + 1475 -128-n-pentane n-hexane n-heptane n-octane analysis 2911 w 2906 vw 2901 vw 2902 w 1403 + 1499 2923f vw 2915 2927 vw vw 2912 2920 w w 405 + 1104 + 1403 1403 + 1517 2938 w 2938§ vw 2935 vw 2930 w 2953* vw 2949* mw 2945* w 1403 + 1542 2960 m 2953 w 2954 m 2951 m 1403 + 1550 - 2 2973 mw 2965 w 2963 mw 2962 mw 1403 + 1559 2981* vw 29855 vw 2975 vw 2974 w 1403 + 1571 3004 w * ! 3013§ 2995* w 1403 + 1593 - 1 3001 vw w < 3015 mw I ( 3010 mw 3023* mw ) 3019§ 3018 w mw 3019 mw see footnote h 3030 mw ) 3047 mw 3044§ w 3043§ mw 3038* mw 1403 + 1627 + 8?; 405 + 1228 + 1403 1237 + + 2f 1799 + 2; 3055 mw 3045 mw 405 + 1237 + 1403 f i 3059§ w • 3060 §f 3054* vw 3068 vw < w 3064 vw 1403 + 16611 3080 w 3076 vw 1403 + 1672 + l f 3094^  w 3088*^ 3086* vw 3086* vw 1228 + 1856 + 2 3105^  mw 3096^  vw 3098§ mw 3095 mw 1237 + 1856 + 2; 2 x 1550 - 5 1228 + 1866 + 1; * 3114 jw 3111§;'vw 3109§ mw 3107 mw 1237 + 1866 + 4; 1403 + 1705 - 1 * 3121 vw 3117* vw 3137 vw 3132 vw 3128 vw 1403 + 1726 - 1 -129-n-pentane n-hexane n-heptane n-octane analysis 3149 vw 3140 vw 1403 + 1736 + 1 3165 vw 3155 vw 1403 + 1752; 1237 + 1921 - 3 f * 3184 vw 3179 vw 3175* vw 1550 + 1627 - 2 f 3195§ vw) 3189 vw 3189§ vw 3185 w 1403 + 1783 - lf 1407 + 1793 - 5 f 3218§ mw 3209§ w § 3211 mw 3207 mw 1403 + 1799 + 5; 405 + 2 x 1403 - 4 * shoulder S broad t This line probably appears somewhat weaker because of reabsorption. 3. While an effort was made to "normalize" the relative line strengths, a line having medium, medium-weak or weak intensity cannot always be reliably compared amongst the spectra. b The table was terminated when the number of lines had increased to the point where they could not be easily resolved. All the spectra exhibit multiplet structure; only the multiplet components at the electronic origins have been listed. Some of the omitted intervals may well contain a small intensity contribution from weaker bands associated with the principal origin. Some entries coincide with a different assignment belonging to another multiplet spectra; these intervals and the fraction of intensity contributed by the coinciding multiplet can be determined by comparing the in-tensities of the electronic origins and their energy difference. d The frequency of the principal electronic origin is given in cm * and a l l other entries show differences built on that origin. e The analysis is based on the spectrum in n-octane whenever possible. -130-Th e line intensity of this interval is not consistent with the Franck-Condon assignment. a -1 6 There are some broad bands to the red of the 800 cm fundamental and these features complicate the analysis here. h 1 1 The region extending from about 3000 cm to about 3100 cm appears to be complicated by frequency and intensity perturbations arising from interactions between CH fundamentals and combination bands so that probable CH fundamentals have not been assigned in the table. 1 This combination, pertaining to the n-octane spectrum only, does not account for the observed intensity of the 3064 cm * interval. -* The corresponding assignments for the n-pentane and n-hexane spectra do not account for their observed line intensities. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0060137/manifest

Comment

Related Items