Solid-state multiple-pulse proton NMR studies of doped Poly( p-phenylene)

Abstract

Recently, it has been found that the conductivity of poly(p-phenylene) (PPP) increases by more than ten orders of magnitude upon doping with certain strong electron donors or acceptors [1]. Few studies of this doping mechanism have been reported [2]. By virtue of its sensitivity to the electronic environment of the nucleus, the solid-state NMR technique has provided important information on similar processes in the doping and conduct ion in polyacetylene [3, 4]. We report studies of the proton resonance in poly(p-phenylene) using multiple-pulse NMR techniques which indicate that these techniques may be useful for probing the environment of such doped conductive polyphenylenes. Many substances have been found to be good dopants for PPP [2], including electron donors such as metal naphthylides and the pure metals and alloys of lithium, sodium, and potassium, and electron acceptors such as arsenic pentafluoride, ant imony pentachloride, fluorosulfonic acid, iodine pentafluoride, and ferric chloride. Our studies include PPP doped with all of these except arsenic pentafluoride, the metal alloys, and lithium. Samples for s tudy by NMR spectroscopy were prepared in NMR tubes and sealed while attached to the vacuum manifold. For liquid dopants, PPP was exposed to the dopant directly. After reaction, excess dopant was removed by evacuation and mild heating, if necessary. Metal dopants were heated to melting to produce the doping reaction. Metal naphthylides were prepared by a standard technique and doping was accomplished as with the liquid dopants, the THF being removed by evacuation. Ferric chloride was dissolved in nitromethane for doping. Typical doping times were in the range of four to seven days. Solid-state proton NMR spectra were obtained using the 8-pulse sequence first demonstrated by Rhim e t al. [5]. The sequence greatly at tenuates dipole-dipole broadening of the resonance line, as shown in Fig. 1 for PPP. As can be seen, the use of the 8-pulse sequence reduces the linewidth by nearly two orders of magnitude. The important point to be made is that the lineshape of the multiple-pulse spectrum does not appear to be dominated by chemical-shift anisotropy. The line is approximately Gaussian