Comment on “Ultrafast Dynamics of Polarons in Conductive Polyaniline: Comparison of Primary and Secondary Doped Forms”

Abstract

Kim, Park, and Scherer reported the resultsof femtosecond pump-probe investigations of primary andsecondary doped polyaniline (PD and SD PANI). They mea-sured changes of transmission at selected wavelengths from 650to 1025 nm after illumination of the sample by a 30 fs laserpulse with a central wavelength of 800 nm. The main featuresof photoinduced spectra were interpreted as a result of excitationof valence electrons into the polaron band, their subsequentcooling and, finally, relaxation via an intermediate state.Recently, we have performed analogous experiments on PDPANI using femtosecond pump pulses at two wavelengthss400and 800 nm. Inside the 650-1025 nm interval photoinducedspectra were similar to those measured in ref 1. However, dueto the broader spectral range in our measurements (from 400 to1050 nm), it became possible to observe new features. A detaileddescription of our results will be published elsewhere. In thiscomment, we discuss several arguable statements in ref 1 andpresent some new experimental facts that are inconsistent withthe model of PD PANI charge carrier dynamics developed bythe authors of ref 1.(i) The main features of the photoinduced response of thesample according to ref 1 are PA1, PA2, and PA3. PA1 isphotoinduced absorption near time zero in the 900-1025 nmrange due to “hot” charge carriers. As time delay varies from-100 to +100 fs, the detected signal changes sign several times.In this case, pump and probe pulses overlap in the time domain,and the resulting photoinduced response is usually referred toas a coherent artifact. Its temporal shape strongly depends onthe parameters (both spectral and temporal) of pump and probepulses, optical properties of the sample, and photodetectorspectral characteristics. It is very important to take into accountthat the change of ∆OD value near time zero (when pump andprobe pulse overlap) does not necessarily reflect the change ofpopulation of energy levels. Indeed, a typical alternatingtemporal shape recorded in ref 1 can be observed even if acentral wavelength of the pump pulse falls inside the transpar-ency range of the sample. Then, all transitions to higher lyingenergy states are virtual. Therefore, a proper treatment of thesignal near time zero must include an analysis of the way theelectrical field of the pump pulse affects energy levels of PANImolecules. A similar approach is used in the theory of thealternating current (ac) Stark effect and third-order nonlinearinteractions.(ii) As for PA2 and PA3, the authors of ref 1 ascribe them tothe “cooling” of free charge carriers excited by the pump pulsefrom valence to polaron band. While PA1 reflects the possibilityof transitions of “hot” population to higher lying states insidethe polaron band induced by the probe pulse, PA2 and PA3stand for the same process for “cold” charge carriers.This model implies that there exists a well-defined range ofstates inside the polaron band which are final for transitionsinvolving light quanta of the pump pulse and states in thevalence band. In this case, photoinduced spectra must stronglydepend on the central wavelength of the pump pulse. Becauseradiation at 400 nm promotes electrons to higher lying statesthan pulses at 800 nm, one can expect spectral positions of PA1and of ground state bleaching to change significantly in thiscase. Neither was detected in our experiments. Figure 1 showsphotoinduced spectra of PD PANI obtained using 40 fs pulsesat 800 and 400 nm with excitation fluence of about 10