Abstract: The electrogenerated chemiluminescence (ECL) of anthracene is characterized by emission at the frequency of anthracene fluorescence and also at longer wavelengths. One longer wavelength component is shown to be caused by emission from anthranol produced by decomposition of the cation radical of anthracene and prbbably excited by energy transfer from excited anthracene. Another component, arising from ECL of anthranol itself, is also observed. revious reports of electrogenerated chemiluminescence (ECL) from anthracene solutions in N,Ndimethylformamide (DMF) have noted that the emission spectrum comprises two or more components. The spectral distribution of the component shortest in wavelength is similar to that of anthracene fluorescence, but the others are broad, structureless, and located toward the red with respect to the first component.2s3 This general behavior is common among several polycyclic hydrocarbons and their derivatives. Presently four alternatives are available to explain the longwavelength emission from these systems. Chandross, Longworth, and Viscoz have proposed the formation of an anthracene excited state dimer (excimer), which radiates to produce the low energy emission. A similar explanation is the formation of an anthracene excited state complex with some other species (exciplex). Both the excimer and the exciplex dissociate into component ground state molecules upon deactivation. Zweig, Maricle, Brinen, and Maurer have suggested that solution phosphorescence may be responsible for the longwavelength emission in some of these systems4 Finally, wes have previously pointed out the possibility of emission from an excited state of a product formed during the reaction of the electrogenerated radical ions with their environment. Anthracene has been chosen for study because it is representative of the class of hydrocarbons exhibiting this behavior and because it is available and easily purified. We have performed a number of experiments designed to aid in identifying the emitting species in the anthracene-DMF system, and to help illuminate the means of exciting the species in so lu t ion which do emit. Experimental Section The anthracene used in all experiments was produced by Matheson Coleman and Bell (mp 215-217”). It was purified by triple recrystallization from Baker Spectroquality benzene and Baker Reagent Grade methanol according to a modification of a procedure available in the l i terat~re .~ A portion of the triply recrystallized (1) (a) National Science Fcundation Predoctoral Fellow; (b) to whom correspondence and requests for reprints should be directed. (2) E. A. Chandross, J. W. Longworth, and R. E. Visco, J . A m . Chem. Soc., 87, 3259 (1965). (3 ) A . J. Bard, I<. S . V. Santhanam, S . A. Cruser, and L. R. Faulkner in “Fluorescence,” G. G. Guilbault, Ed., Marcel Dekker, Inc., New York, N. Y . , 1967, Chapter 14. (4) A. Zweig, D. L. Maricle, J. S . Brinen, and A. H. Maurer, J . Am. Chem. Soc., 89, 473 (1967). material was also resublimed twice in DCICUO. No differences in behavior .were found between the material which had been doubly resublimed after recrystallization and that which had merely been recrystallized thrice. For this reason, most subsequent experiments used only the triply recrystallized anthracene. Fluorescence analysis of cyclohexane solutions of the purified anthracene showed no luminescence bands other than those directly attributable to anthracene.6 Maxima in fluorescence intensity were found at 378, 397, 420, 447, and ca. 475 mp. In particular tetracene was shown by absorption spectroscopy and by fluorescence measurements to be present in amounts less than 0.1 z, since none was detectable by these methods. The solvent used in every case was N,N-dimethylformamide which was also supplied by Matheson Coleman and Bell (bp 152154”). The solvent was further purified,by two methods. Method A involved storing the solvent over anhydrous cupric sulfate for several days to complex water and dimethylamine. The solvent was then decanted and distilled at a reflux ratio of 5 from a glass bead packed column 100 cm high under a nitrogen pressure of 20 mm. The middle fraction was retained for use. Method B also involved storage over anhydrous cupric sulfate For a period of several days. The distillation which followed was under the same conditions as above except that the reflux ratio was unity. Following this distillation, the solvent was stored over Linde Type 4A Molecular Sieves for a period of 48 hr. Then the material was decanted and redistilled using a reflux ratio of .l. Once again, only the middle fraction rvas taken. The solvent was stored under an inert helium atmosphere. Neither solvent batch showed fluorescence bands, even under the most sensitive conditiuris. The supporting electrolyte used in all experiments was tetra-rzbutylammonium perchlorate (TBAP), Polarographic grade, supplied by Southwestern Analytical Chemicals, Austin, Texas. The TBAP was used without further purification, but was dried in a vacuum oven for 48 hr at a temperature of 100” and then stored in a desiccator over magnesium perchlorate. The TBAP contained no fluorescent impurities. The electrolysis cell used for ECL emission measurements consisted of two platinum helices inserted through graded seals into the Pyrex wall of a 14/35 standard taper joint, as shown in Figure 1. An adapter was provided so that the cell could be evacuated easily. The electrodes were 2-5 mm apart. It was generally found that greatest emission intensities were incident upon the monochromator entrance slits when the slits and the two electrodes were arranged colinearly. This arrangement was used uniformly in the experiments. Immediately after loading the cell, it was degassed on a vacuum line similar to that described previously7 using two freeze-pumpthaw cycles. Minimum pressure over the frozen solution on the second cycle was at most 10-4 torr in every case. The voltage applied to the cell was simply the 60-cycle sinusoidally alternating line voltage which was reduced from 110 V root mean square to any (5) T. Takeuchi and M . Furusawa, Kogpo Kagaku Zasshi, 68, 474 ( 6 ) I. Berlman, “Handbook of Fluorescence Spectra of Organic (7) K. S. V. Santhanam and A. J. Bard, J . A m . Chem. Soc., 88, 2669 (1965); Chem. Abstr., 63, 4060e (1965). Molecules,” Academic Press, New York, N . Y., 1965.