4: Triplet Excitons (CIT)
In 1958 the paramagnetic resonance of the triplet state of naphthalene dissolved at low concentrations in durene crystals was reported[8]. This raised the question in my mind as to what the nature of triplet naphthalene molecules would be in pure naphthalene. Consequently Himan Sternlicht and I made a theoretical calculation of the expected resonances of a propagating triplet state, which I referred to as a triplet exciton (75, 79). Since I was quite confident in these calculations I assigned the task of observing these propagating excitons to a first-year graduate student, Hayes Griffith, who proceeded to spend the better part of a long, hot, miserable summer on the project, with no results. Now Hayes was a hard-working student and it began to bother me greatly that he could find no signal. I began to think more about the theoretical aspects of this project, and suddenly thought of triplet – triplet collisions and subsequent annihilation. Clearly it should be possible for two excitons to collide and produce a nonparamagnetic state, eg., a fluorescent naphthalene molecule. I immediately – the same day (September 1961) — asked Hayes to stop the project which I think he was more than happy to do. This was a big mistake on my part (see below). I was correct in assuming that exciton collisions or other annihilation processes reduced their concentration to a low level. My big mistake was to take Hayes off the project, and his big mistake was to follow my suggestion.
The theoretical calculations with Sternlicht were later experimentally verified by Schwoerer and Wolf[9] using isotopic mixtures of naphthalene which localizes excitations at liquid He temperatures on protonated naphthalene. For example, I predicted D = -0.00588 cm-1, E = 0.0478 cm-1 for a triplet exciton in naphthalene while Schwoerer and Wolf got D = -0.00585 cm-1, E = 0.0485 cm-1 for neighboring but differently oriented protonated naphthalene molecules (Nh8) in the unit cell. The crystals were Nh8 in excess perdeuterated naphthalene (Nd8) at 4.2 deg. K, at which temperature the triplet excitations are localized (trapped) on the protonated molecules[9].
Finally to complete my embarrassment, Haarer and Wolf reported an impressive study of the paramagnetic resonance of triplet excitons first in anthracene crystals[10], and then in naphthalene crystals[11]. And triplet exciton paramagnetic resonance was reported in tetracene crystals soon thereafter[12]. These authors attributed their triplet exciton formation to singlet exciton fission. Johnson and Merrifield demonstrated by optical methods the mutual annihilation of triplet excitons in anthracene crystals[13].
In retrospect, we believe that we failed to observe triplet exciton paramagnetic resonance in pure naphthalene crystals because of their low concentration and short lifetimes. We utilized a uv irradiation and ESR detection system similar to that reported in the pioneering experiments of Hutchison and Mangum[8]. We used a Varian V4502 ESR spectrometer with 100 Khz modulation of the dc magnetic field, a lock-in amplifier and phase sensitive detection to enhance the signal/noise ratio. Our experiments were carried out at liquid nitrogen and liquid helium temperatures. Haarer and Wolf used different conditions: They used a K band ESR spectrometer, and eventually a Q band ESR spectrometer to achieve higher magnetic fields and hence enhance the weak ESR signals. More importantly, they chopped the exciting light source and used phase sensitive detection at the chopping frequency (80 kHz). In this circumstance the short lifetime of the excitons is an advantage. D. Haarer found that he was then able to detect the excitons at room temperature but not at liquid nitrogen temperatures. (We are indebted to recent correspondence between D. Haarer and O. H. Griffith for experimental details.)
There is a lesson in this story. In October 1961 G. Wilse Robinson and collaborators in Chemistry at CIT reported the triplet state phosphorescence of protonated naphthalene in H, D isotopically mixed crystals[14]. At 4.2 ºK the lifetime of the phosphorescence was long, 2.6 sec. These were ideal conditions to observe triplet state paramagnetic resonance with our instrumentation, and to anticipate the Schwoerer-Wolf experiments by years. For some unknown reason we failed to do this experiment. The lesson here is that it is possible to let an important experiment slip through your fingers. There was not enough discussion of this subject amongst Robinson, Griffith and myself.
One of the predicted characteristics of triplet exciton paramagnetic resonance spectra is the absence of nuclear hyperfine structure; the only signal splitting expected is due to fine structure, the electron – electron dipolar interaction in the triplet state. I was thus intrigued to see a publication by Don Chesnut (my former graduate student) and Bill Phillips, both at Du Pont, in which they reported just such spectra – fine structure but no hyperfine structure[15]. These signals arose from crystals of charge transfer complexes (morpholinium TCNQ). Very much to my surprise these authors did not mention triplet excitons as the likely source of their signals, and I promptly published a paper pointing this out (79, 82). Based on this insight my lab was able to show the presence of these propagating triplet states in a familiar crystal of ionic free radicals, Wurster’s blue perchlorate (85, 88). In contrast to the excitons in naphthalene, the electron pairs in the ionic crystals are localized on adjacent molecules and are thermally excited to the triplet state.
Since triplet excitons in crystals had not been reported previously, students in my lab carried out a number of studies of these crystals, including phonon-coupled interactions between excitons (90), high pressure effects on exciton resonance (93), the effects of crystal phase transitions on the resonance spectra, and an attempt to detect x-ray scattering by excitons (104), as well as theoretical studies of their properties (105). My former graduate student Zoltan Soos (Princeton) has continued to work on the theoretical aspects of molecular crystals, and he would be the best source of current information in this area.
Chapter 5: Structure of the Methyl Radical (CIT)
