HAL Id: jpa-00214673
https://hal.archives-ouvertes.fr/jpa-00214673
Submitted on 1 Jan 1971
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
SPECTROSCOPIE ÉLECTRONIQUE DES
MOLÉCULES POLYATOMIQUES ET DES SOLIDES ORGANIQUESEXCITONS, AGGREGATES AND
DEFECTS IN ORGANIC SOLIDS
H. Wolf
To cite this version:
H. Wolf. SPECTROSCOPIE ÉLECTRONIQUE DES MOLÉCULES POLYATOMIQUES ET DES SOLIDES ORGANIQUESEXCITONS, AGGREGATES AND DEFECTS IN ORGANIC SOLIDS.
Journal de Physique Colloques, 1971, 32 (C5), pp.C5a-101-C5a-105. �10.1051/jphyscol:1971510�. �jpa-
00214673�
SPECTROSCOPIE ELECTRONIQUE DES MOLECULES POLYATOMIQUES ET DES SOLIDES ORGANIOUES
EXCITONS, AGGREGATE S AND DEFECT S IN ORGANIC SOLID S
H. C . WOLF
Physikalisches Institut der Universitat Stuttgart
Abstract.
-Spectroscopic studies using optical and ESR spectroscopy on organic molecular crystals like naphthalene and anthracene are discussed, which give the following information
:the magnitude of the energy transfer matrix element between two molecules, the mobility and inter- action of excitons, the incoherence of exciton motion, the properties of typical defects (so called X-traps) and the detectability of guest molecules using delayed fluorescence.
1. Introduction.
-Optical spectra of organic crys- tals are like most of the molecular crystal spectra mainly characterized by the close similarity between crystal spectra and spectra of the free molecules. The oriented gas model
-where the crystal only holds the molecules in fixed positions in space
-is a good first approximation. Therefore, organic solid state spectro- scopy is a useful method for studying problems of molecular physics.
In the following interactions between the molecules are discussed. These interactions are relatively weak, but they are responsible for phenomena such as conduction of charge and especially conduction of energy. These solid state properties make organic crystals interesting.
Spectroscopy is the method to study these solid state properties, especially energy transfer between mole- cules.
-The basic question is
:how far and how fast are excitons able to travel (i) in a perfect crystal and (ii) in a crystal containing defects.
-I also want to emphasize the importance of using spectroscopy at many dzfferent frequencies
-in fact between 10' and 10i5 NZ
-for such studies.
The problems, which are discussed in the following are
:1) Measurement of the energy transfer matrix element between two molecules by spectroscopy of pairs (or higher aggregates).
2) Spectroscopical measurements of the mobility and the interaction of free excitons.
3) Spectroscopic study of defects in crystals, and on their influence on energy transfer and trapping pro- cesses.
I shall discuss mostly triplet states and triplet excitons, with S
=1. Due to their longer lifetime, they are in several respects more interesting than the singlet
states. In addition, one can apply spin resonance techniques since they are paramagnetic. Most of the experiments are done with naphthalene and anthracene crystals, which are used as model substances. In the following, a resume of some of the work done in Stuttgart within the last years is given.
2. Pair spectra.
-The smallest possible crystal is the unit cell. There are crystal properties, which can be studied better by pair spectroscopy than by spectro- scopy of the crystal. The main difference between the spectra of molecules and of crystals are a shift and a splitting of the energy levels. The Davydov splitting is due to the fact that it is impossible to excite only one molecule. Interaction between translationally equiva- lent molecules shifts the levels, interaction between the different molecules in the unit cell splits the levels. The Davydov splitting is used to measure the matrix element for excitation exchange between inequivalent neighbors and for the velocity of energy travelling by hopping processes through the crystal.
Both the measurement and the interpretation of Davydov splittings is sometimes difficult. One has to interpret and to understand the whole exciton band structure, the line shape and the linewidth of the optical spectra. That this is not easy is illustrated by the anthracene spectrum (Fig. 1). Even in this very well investigated crystal the 0.0 transitions in absorption and fluorescence are not fully understood. Near 0.0, there is a broad absorption and a very sharp emission spectrum. This spectrum has been explained by self- trapping, but this seems erroneous. From this spectrum, one can determine the Davydov-splitting only with a high degree of uncertainty - somewhere between 100 and 400 cm-' [l], [2].
The elementary step of energy transfer is the transfer between two molecules, the two molecules in one unit
Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1971510
C5a-102 H. C. WOLF
cell of naphthalene. In order to measure this, a different spectroscopic method is used
:the spectroscopy of pairs of molecules, (AB)*-pairs.
FIG. 1. - Absorption (right) and fluorescence (left) spectrum of crystalline anthracene near 0.0 of the lowest singlet transition (l), (2). The absorption spectrum is given parallel to axes b (---) and a (.
. .).
The 0.0 line in emission at 25 097 cm-1 is not visible due to reabsorption. The first strong emission line is the 0.0 lineof the XI-series (2).
We investigate pairs of molecules in a single crystal matrix, for instance N-h,-molecules in a matrix of N-d,. Hanson [3] has measured such pairs in the singlet state. Here we present data on the triplet-singlet emission, (AB)*-triplet spectra
[4].In the emission spectrum of the mixed crystals, one observes at concentrations 0,2
;2
;5 and 10 % in addition to the monomer line also the pair of lines for dimers (Fig. 2,3). The direct result of this measurement is
:(i) the pairwise interaction matrix element is
=
1,2cm-l
;(ii) the interaction AB is
DAA or BB
-since no lines of AA or BB-pairs are visible. This means
:in the crystal, the exciton travels mainly
twodimensional,in the ab-plane
;(iii) the two dimer lines are differently polarized.
The value for P is in good agreement with the Davydov splitting
:the T,-band in naphthalene has a Davydov splitting of 10 cm-'. The Davydov splitting should be 8 P, if only nearest neighbors contribute.
Obviously the pair spectroscopy gives more infor- mation than the crystal spectroscopy
:the Davydov splitting is an integral value, one does not know whether the splitting energy is due to AB or to other interactions.
Pair spectroscopy dzfle~entiatesbetween these possibilities and shows in this particular case that the AB interaction is by far predominant.
These AB pairs can be studied also by ESR [5]. In mixed crystals N-h, in N-d,, at concentrations above 0,2 % one observes in the ESR spectra in addition t o the monomer lines (A and B-lines) dimer lines (M- lines). Their position and their line width are a result of a rapid exchange of the spin between the two sites.
FIG. 2. - 0.0-Line in the phosphorescence spectrum of CloHs in a CI0D8 host crystal. From top to bottom 0.2, 2, 5, lOand 10 m01 per cent CloHs. The top spectrum is for a crystal at 1.8 OK, the remainder at 4.2 OK. The isolated guest transition
energy is 21 208.9 cm-l (4).
FIG. 3. - Polarized 0.0 phosphorescence spectra at 4.2 OK for 3.5 mole % CloHs in C10D8. The isolated guest transition
occurs at vo.
Assuming an incoherent hopping process, one gets
the following expression for the line width of the M-
lines as a function of the distance between the two
monomer lines
:EXCITONS, AGGREGATES AND DEFECTS IN ORGANIC SOLIDS C5a-103
From the measurements, we find j3 in good agree- ment with the optical spectroscopy value.
Using ESR-spectroscopy, also higher aggregates can be observed, this is due to the line width perhaps not so easy with optical spectroscopy. At very low tempera- tures we find indications for a coherence of the mini- exciton motion
-one observes a broadening of the M-lines between 4.2 and 1.8 OK [6].
This kind of spectroscopy
-pair spectroscopy - can and will be applied in the future to study phase diagrams of organic mixed crystals. The question whether or not molecules are solved or distributed statistically in a mixed crystals can be studied very specifically using pair or aggregate spectroscopy.
3. Excitons. - Now we go back from the mini- excitons to real, free excitons in the pure crystal.
In ESR, the fine structure tensor, the number and position of lines, are equal for pairs and for excitons if interaction AB S AA, BB. The main difference is the line width, and this quantity is determined by averaging of the exciton spin between many molecules.
The mobility of the exciton shows up in the relaxa- tion times T , (the line width) and T , (from saturation measurements) [7]. The basic process to understand the relaxation times T , and T, is the modulation of the fine structure H,, when the triplet is jumping from site A to site B. H, becomes time dependent, because the triplet is jumping
-the hopping model is very appro- priate to describe the FS modulation process.
For a quantitative discussion, one needs the assump- tions
:(i) that the exciton is hopping statistically, incohe- rently
;(ii) that at every jumping process, the FS Hamilto- nian H, is modulated by A H
=H: - H! [7].
'5
l
0 100 15 0 200 250 300Temperature T[OK]
FIG. 4. - Line width of the phosphorescence spectrum of pure naphthalene crystals as a function of temperature [g].
Then one gets an expression for the relaxation times containing a correlation time
zc =4 th. Values for
zcare given in 171.
The good agreement between these correlation times and the p-value from the pair spectroscopy confirms that pairs are really
(<miniexcitons >>.
Now we go back to optical spectroscopy. The optical spectra of excitons are in several respects similar to those of pairs : the number of lines, the polarisation, and more or less the frequency are the same - at least for triplet excitons. The main diffe- rence is again the line width.
The line width of exciton emission at higher tempe- ratures is determined by exciton-phonon-interaction.
Since excitons are propagating through the crystal, exciton-phonon scattering takes place. The line width in the optical spectra measures directly the exciton- phonon-scattering time.
Triplet exciton emission in pure naphthalene at room temperature has a line width of 200 cm-', which corresponds to an exciton-phonon scattering time of 5
X10-l4
S.Since this time is much shorter than the hopping time, this experiment shows that the hopping process is really incoherent at room temperature. The apparent line width at low temperature is determined by trapping processes - at temperatures at or below 40K - but not above - we can expect coherent processes.
Still more spectacular than exciton-phonon interac- tion is exciton-exciton interaction, which shows up as delayed fluorescence. What happens, if two excitons meet each other
?with a good probability, they annihilate, and if they are triplet excitons, they give rise to delayed fluorescence, which is studied spectroscopi- cally.
One knows free-free and free-trap interaction. An example of free-trap interaction is again N-h, in N-d,, where by measuring the temperature dependence one can determine the trap depth AE and, still more important, the annihilation coefficient [9].
4. Defects. - In the last part of my talk, I want to emphasize the important role of defects in organic crystals, and I want to show that they can sometimes be analyzed even in extremely small concentrations. Since they can act as very effective traps for both excitons and charge carriers, they are able to influence funda- mentally the physical properties of the crystals.
Besides impurities, the most important and characte- ristic defects are disturbed exciton states, which we call X-traps, X for unknown.
These X-traps show up in optical spectra as series of lines with the vibronic intervals of the host, but at lower energies than the host lines [IO]. X-traps are host molecules in the vicinity of an impurity molecule or an imperfection, f. i. dislocation, which are - due to this small perturbation - a little bit lower in energy and act as trapping centers therefore, especially at low temperatures,
The trap depth is given directly as the difference
between the host 0.0 and the X-series 0.0 transition.
This trap depth is characteristic for the particular X-trap, for example
Singlet Triplet
trap trap
depth depth
- Thionaphthene in naph-
thalene 30 cm-' 45 cm-'
Unknown in anthra-
cene 247 cm-' 24 cm-'
Many X-traps have been observed so far by optical spectroscopy, but nothing is known on the relation between trap depth and specijic nature of the dis- turber [ l l].
These X-centers can be investigated also by ESR.
Quinoxaline in Naphthalene induces at 4 OK a X-trap.
From the ESR-spectrum we have evidence, that the molecule can be disturbed only very little, and however, that it is only one disturbed naphthalene molecule, which give rise to the X-trap, figure 5 [l l].
FIG. 5. - Hyperfine structure of one ESR-line (Am = 1 transi- tion) in the ESR spectrum of the triplet state of quinoxaline in perdeutero-naphthalene (left), and of the X-trap, which is induced by quinoxaline in naphthalene (right). The quinoxaline concen- tration is the same in both crystals. The hyperfine structure is characteristic for quinoxaline (left) and for naphthalene
(right) [l l].
a - d 8
+P&)
Those X-traps are always present in all crystals. It depends from the sensitivity of the particular experi- ment, whether or not they are detected.
03-h.3
+a:)
5. Sensitized Delayed Fluorescence.
-Finally I want discuss sensitized delayed fluorescence as a new method to detect guest molecules or traps in smallest concentrations. It is well known that sensitized fluo- rescence is extremely sensitive against impurities. Here I want to show that sensitized delayed fluorescence is still more sensitive.
Sensitized fluorescence is known from systems such as tetracene in anthracene. The lowest concentration, which can be detected, is in the order of I O - ~ , and the one
:one concentration is in the order of 10-S [ll].
Sensitized delayed fluorescence is observed, when a mobile triplet exciton collides with an triplet exciton trapped at an impurity or guest molecule
:they can annihilate, and the result is delayed fluorescence.
As mode1 system we used anthracene a s guest in naphthalene. The quantum ratio Q,/QN is measured for prompt and delayed fluorescence (Fig. 6) and compared with a simple kinetic scheme [l l].
ESR F
4,2 'K J
FIG. 6 .
-
Spectrum of prompt (broken line) and delayed (full line) fluorescence of a naphthalene crystal containing 2 x 10-7 parts of anthracene, at room temperature. Anthracene emissionis below 26 500 cm-1.
f
L
The kinetic analysis as given in [l l ] shows that using sensitized delayed fluorescence the lowest concentra- tion which can be detected is 10-l' molar, and the one
:one concentration is in the order of 1 0 - ~ molar.
In conclusion, I wanted to show with a few examples from our work kow optical spectroscopy of molecules, pairs, aggregates and defects is used to understand better the solid state physics of organic crystals, and how useful it is to use spectroscopy in a wide range of frequencies.
Whereas optical spectroscopy of crystals is an old method, new and promising trends in the field are visible. I think that
(i) the new method of pair spectroscopy
;I l l I 1 1 1 1 I
1 : 4 : 6 : 4
: l 2 a-Protons
4U-Protons
4(3-Protons
4P - Protons
(ii) the sensitized delayed fluorescence
;(iii) the combination of optical spectroscopy and HF spectroscopy
Quinoxaline
will be fruitful new methods in the future.
(Not resolved)
Naphthalene - h,
EXCITONS, AGGREGATES AND DEFECTS IN ORGANIC SOLIDS
