No abstract

No abstract

No abstract

We present measurements of the optical absorption and emission properties of poly(p-phenylene vinylene͒ ͑PPV͒ -related materials focusing on the differences between molecules isolated by dispersion in an inert host and concentrated molecular films. Optical absorption spectra, photoluminescence ͑PL͒ spectra, PL efficiency, and time-resolved PL spectra of dilute blends of PPV oligomers with 2-5 phenylene-phenyl rings are compared with those of dense oligomer and polymer films. In dilute oligomer-poly͑methyl methacrylate͒ ͑PMMA͒ blends with high PL efficiency, the PL decay is exponential, independent of both temperature and oligomer length. This implies that the fundamental radiative lifetime of PPV oligomers is essentially independent of oligomer length. Concentrated spin-cast oligomer films and polymers have a faster and strongly temperaturedependent PL decay that approaches that of the dilute oligomer results at low temperature. The differences in PL decay correspond to changes in PL efficiency. The efficiency of the oligomer-PMMA blend is high and only weakly temperature dependent, whereas that of concentrated films is lower and strongly temperature dependent, decreasing by more than a factor of 3 from 10 to 350 K. The quenching of the PL efficiency in concentrated films is due to migration to extrinsic, impurity related centers as opposed to an intrinsic intermolecular recombination process. The PL spectrum of a dilute oligomer blend redshifts substantially, both as the excitation energy is decreased and as the emission time increases. This spectral redshift is due to disorderinduced site-to-site variation and not to diffusion to lower-energy sites. In contrast, no spectral shift with excitation energy or emission time was observed for dense oligomer films. ͓S0163-1829͑96͒03132-3͔

ERRATAWe regret that an incorrect version of the Response of Mahler etal. to the Comment by Combescot was published [Phys. Rev. Lett. ^9, 1744]. The correct version follows. Mahler et al. Respond: (1) The nonuniform state of a binary (or multicomponent) system is governed to zero order by the Euler hydrodynamic equations supplemented by the thermodynamic equation of state. The first -order approximation (Enskog) 1 then leads to the irreversible processes of diffusion, heat conduction, and thermal diffusion (neglecting here internal friction) as phenomenologically treated by the Onsager linear transport equations. 2 ' 3 When applied to electronic transport phenomena in solids the binary system is built up by the electrons (and/or holes) and the lattice atoms (where any macroscopic drift of the latter is usually excluded). This is the basis for a large variety of transport effects ranging from Ohm's law to complicated cross effects with or without external fields, including our thermodiffusion model. 4 Alternatively one may try to model the electronhole transport process by a one-component Euler equation to which a phenomenological damping term has been added to simulate the coupling to the lattice. 5 In a way, the resulting equation resembles the Navier-Stokes equation, in which the internal damping term is replaced by an external one, though of different functional dependence.Obviously, these two approaches are by no means equivalent: Combescot 6 just succeeded in proving this for a special case. However, starting from the Euler equation it is certainly not possible to say anything about thermal diffusion, in general, or the sign of the resulting gradients, in particular. This renders Combescot's whole section (1) invalid.(2) First, the condition of vanishing heat current j Q =0 is not essential for obtaining our interesting confinement behavior. The dominant sources for j Q are Auger heating and lattice cooling (through which the bath temperature comes in), with the boundary condition j Q (?c -°°) =0. Only in extreme cases will this lead to a destruction of the confinement, as will be shown in a forthcoming paper.Second, it is textbook knowledge that there exist linear transformations for the currents and forces in the Onsager equations which leave the Onsager reciprocity relations unchanged. 3 The force V(~ MA 1 ), where M is the chemical potential and T the temperature, is the correct force when 3 q is taken to be the heat current while V(-\i) T enters with the so-called reduced heat current (for details see Ref.3).Third, for the {n, T) region we discuss, N =L qq L mm -L qm 2 is certainly positive. N becomes negative in the extreme quantum limit, where not only the present transport theory but any hydrodynamic approach breaks down.(3) Under actual experimental conditions the particle current j m (0) and the initial density n(0) are related to each other. In our approach, however, j m together with the isothermal diffusion constant D only defines the length scale of our profiles, as is easily seen from Eq. (3)...