Photochemistry and Photobiology, 20**, **: *–* Photoisomerization of cis-1,2-di(1-Methyl-2-naphthyl)ethene at 77 K in Glassy Media † Christopher Redwood 1 , V. K. Ratheesh Kumar ‡1 , Stuart Hutchinson 1 , Frank B. Mallory 2 , Clelia W. Mallory 3 , Ronald J. Clark 1 , Olga Dmitrenko 4 and Jack Saltiel* 1 1 Department of Chemistry and Biochemistry, The Florida State University, Tallahassee, FL 2 Department of Chemistry, Bryn Mawr College, Bryn Mawr, PA 3 Department of Chemistry, University of Pennsylvania, Philadelphia, PA 4 Department of Chemistry and Biochemistry, University of Delaware, Newark, DE Received 21 August 2014, accepted 6 October 2014, DOI: 10.1111/php.12367 ABSTRACT cis-1,2-Di(1-methyl-2-naphthyl)ethene, c-1,1, undergoes photo- isomerization in methylcyclohexane, isopentane and diethyl ether/isopentane/ethanol glasses at 77 K. On 313 nm excitation the fluorescence of c-1,1 is replaced by fluorescence from t-1,1. Singular value decomposition reveals that the spectral matrices behave as two component systems suggesting conversion of a stable c-1,1 conformer to a stable t-1,1 conformer. However, the fluorescence spectra are k exc dependent. Analysis of global spec- tral matrices shows that c-1,1 is a mixture of two conformers, each of which gives one of four known t-1,1 conformers. The k exc dependence of the c-1,1 fluorescence spectrum is barely dis- cernible. Structure assignments to the resolved fluorescence spec- tra are based on the principle of least motion and on calculated geometries, energy differences and spectra of the conformers. The relative shift of the c-1,1 conformer spectra is consistent with the shift of the calculated absorption spectra. The calculated structure of the most stable conformer of c-1,1 agrees well with the X-ray crystal structure. Due to large deviations of the naph- thyl groups from the ethenic plane in the conformers of both c- and t-1,1 isomers, minimal motion of these bulky substituents accomplishes cis ? trans interconversion by rotation about the central bond. INTRODUCTION The cis ? trans photoisomerization of olefins in solution nor- mally proceeds by rotation about the isomerizing double bond, the one bond twist (OBT) mechanism (1–3). The Hula-twist mechanism (HT) involving concerted rotation about a double bond and an adjacent essential single bond (envisioned as equiv- alent to a 180° translocation of one CH unit) was proposed to account for cis ? trans photoisomerization in free volume restricted media (4,5). Specifically, in protein environments or glassy media it was postulated that the HT mechanism reduces the free volume requirements of the OBT mechanism by confin- ing motion to the vicinity of the atoms of the isomerizing double bond, while minimizing the motion of bulky substituents (4,5). Experimentally, the validity of the HT process was initially based on the interpretation of UV absorption spectra measured in the course of the photochemical sequence provitamin D ? previ- tamin D ? tachysterol (Pro ? Pre ? Tachy) at 92 K in an EPA glass (6). That this was a misinterpretation became clear when we monitored the same sequence at 77 K by fluorescence spectroscopy (7,8). However, that initial HT claim had motivated many investigators to seek evidence for HT photoreactions and many claims of HT photoisomerization in other molecular systems ensued (8–13). A case in point is the cis ? trans photo- isomerization of 1,2-di(2-naphthyl)ethene (DNE). Irradiation of c-DNE in a low T organic glass gives a t-DNE conformer mix- ture that differs substantially from the equilibrium composition. The Alfimov (14) and Fischer (15) groups interpreted such results to mean that the equilibrium conformer compositions of the two isomers differ. Complete photoisomerization of the equi- librium distribution of the c-DNE conformers in the glassy med- ium produces an unstable distribution of the trans conformers that on warming and recooling shifts to the thermodynamically stable trans conformer distribution. A necessary, albeit insuffi- cient, requirement of the HT process is that photoisomerization of a stable reactant conformer give an unstable photoproduct conformer and vice versa. Accordingly, it was proposed that the Alfimov–Fischer results could indicate that c-DNE undergoes HT photoisomerization in the glassy medium (10,16). We showed that at 77 K in glassy MCH the photoisomeriza- tions of cis-1-(2-naphthyl)-2-phenylethene (c-NPE, Scheme 1, R = R = H) that exists mainly as the B conformer (18) and cis-1-(3-methyl-2-naphthyl)-2-phenylethene (c-3-MPE) that should exist as the A conformer (19) (Scheme 1, bottom left, R = CH 3 , R = H, c-3-MPE A ) proceed as shown by the OBT labeled pathways c-NPE B ? t-NPE B and c-3-MPE A ? t-MPE A in Scheme 1. However, our conclusion that no HT is involved in those reactions was dismissed because HT photoisomerization about the benzylic CH would lead to the same result (14). For- mation of an unstable trans conformer starting from the o-tolyl analog of c-NPE (c-NTE, Scheme 1, R = H, R = CH 3 ) in EPA glass at 77 K was thought to confirm the benzylic HT pathway (17). On the basis of the known t-NPE A and t-NPE B absorption spectra (20), we pointed out that the correct assignment of the thermal relaxation of the photoproduct of c-NTE in EPA glass at *Corresponding author email: saltiel@chem.fsu.edu (Jack Saltiel) ‡Present address: Department of Chemistry, VTM NSS College, Trivandrum, Kerala, India †This manuscript is part of the Special Issue dedicated to the memory of Michael Kasha. © 2014 The American Society of Photobiology 1