Letter Cite This: J. Phys. Chem. Lett. 2019, 10, 171−175
pubs.acs.org/JPCL
Electron Transport with Mobility, μ > 86 cm2/(V s), in a 74 nm Long Polyfluorene Andrew R. Cook,*,† Sadayuki Asaoka,‡ Xiang Li,† and John R. Miller*,† †
Chemistry Department, Brookhaven National Laboratory, Upton, New York 11973, United States Department of Materials Science and Engineering, Kyoto Institute of Technology, Matsugasaki, Sakyo-ku, Kyoto 606-8585, Japan
‡
Downloaded via IOWA STATE UNIV on January 18, 2019 at 12:49:16 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
S Supporting Information *
ABSTRACT: The mobility of charges on conjugated polymers is a fundamentally important feature of these materials, but most fall far short of transport that might lead one to call them “molecular wires”. A commonly identified bottleneck is flexible dihedral angles between repeat units. Here we find a very high mobility, μ > 86 cm2/(V s), for electrons attached to polyfluorene polymers in isooctane, despite the presence of varied dihedral angles. The present data suggest that interactions with the surrounding medium may be a principal determinant of charge mobility.
S
Figure 1 shows transient absorption at the 580 nm peak11,12 of the radical anion, pF89−•, which is similar to spectra
ince their discovery and the 2000 Nobel Prize,1 conjugated polymers have held out the promise that they could serve as “molecular wires”. Indeed, pulse radiolysis transient microwave conductivity (PR-TRMC) experiments2−6 have demonstrated mobilities, μ, up to ∼1 cm2/(V s), with one report noting a spectacular value, μ = 600 cm2/(V s),7 along conjugated chains in solution. But while they play major roles in technologies,8,9 the use of conjugated polymers as wires has not come to fruition. Results reported below on a polyfluorene bearing end-trap groups find a very high mobility. They may point us to strategies with potential to enhance the use of polymers as “wires”. To explore ability of conjugated polymers to act as “wires”, electrons were attached to polyfluorene polymers in isooctane by pulse radiolysis experiments. The electrons then transported to chain end caps, which serve to trap electrons. The polymers used had length narrowed distributions, with average lengths in repeat units, n, denoted as pFn. Polymers without caps, pF89, were compared to those with anthraquinone (pF98AQ) and naphthylimide (pF83 NI) end-trap groups. Sample and experimental details are given in the Supporting Information, section S1. Scheme 1 shows polymer structures. Isooctane was chosen as the solvent because of the high mobility (μ = 6.6 cm2/(V s)) of electrons resulting in capture by solutes as fast as 2 × 1013 M−1 s−1,10 which enables astonishingly fast electron attachment here.
Figure 1. Transient absorption at 580 nm of pF89 in isooctane after pulse radiolysis collected with the optical fiber single-shot detection system (OFSS).
Scheme 1. Structures of Uncapped and AQ Capped Poly-(2butyloctyl)fluorenes Used in This Study, pF89 and pF98AQ
observed in films.13 Data in Figure 1 were recorded following pulse radiolysis of solutions containing 1−40 mM pF89. Concentrations are reported in polymer repeat units (PRU); thus average concentrations of pF chains are 89 times lower. The sudden “step” increases in transient absorption at 580 nm Received: October 17, 2018 Accepted: December 16, 2018 Published: December 16, 2018
© 2018 American Chemical Society
171
DOI: 10.1021/acs.jpclett.8b03185 J. Phys. Chem. Lett. 2019, 10, 171−175
Letter
The Journal of Physical Chemistry Letters are expected to be largely due to the production of pF89−• due to the high mobility of electrons and likely even faster capture prior to electron thermalization/localization,14 though it also contains contributions from pF+• and pF* discussed later. At ≤ 10 mM pF89, there is also a small fast growth of pF89−• due to capture of homogeneous electrons in competition with their ∼200 ps recombination with solvent cations. For the 2 mM trace the growth is 1.4 × 1010 s−1, giving a bimolecular rate constant of 7.0 × 1012 M−1 s−1 per repeat unit or 6.2 × 1014 M−1 s−1 per polymer molecule. At the highest [pF], all pF89−• is produced much faster than the 15 ps time resolution, providing the best possible time resolution for monitoring charge transport in capped polymers described below. Decays observed during the first ∼ 2 ns in Figure 1 are due to geminate recombination of pF89−• with solvent cations. When polymers with AQ end caps were examined at a series of concentrations, the transient absorption signals, still principally due to pF−•, were smaller by 27 ± 3% and showed faster decays, as seen in Figure 2. Nearly identical data for NI
end, we can determine the probability, Pcap, that a chain end has a cap is 0.29. The fractions of chains with 0, 1, or 2 caps are F0 = (1 − Pcap)2
F1 = 2Pcap(1 − Pcap)
F2 = Pcap2
(1)
Thus, for pF98AQ, F0 = 0.5, F1 = 0.41, and F2 = 0.09; only 9% of the chains have the intended two end-trap groups pictured in Scheme 1. The fact that 50% of the chains in Figure 2 have no end-cap groups means that those uncapped chains contribute an absorbance equal to 50% of the absorbance of pF at each concentration in Figure 1. To see the absorbance from capped chains alone, 50% of the pF89 absorbance was subtracted from pF98AQ and pF89 at each concentration to produce the comparisons in Figure 3 for the 40 mM (a) and 5 mM (b) PRU samples. In Figure 3, one can see that about half of the rapidly attached electrons make it to the AQ end caps in under the 15 ps experimental time resolution. Of the chains remaining after this subtraction, 17% have end traps at both ends and 83% have only one end trap.
Figure 2. Transient absorption at 580 nm of pF98AQ in isooctane under conditions similar to those in Figure 1.
capped polymers are given in Figure S2 of the Supporting Information. The more positive reduction potentials12 of the AQ and NI end caps predict that if an electron moving along a chain encounters an AQ cap, it will be trapped producing pF98AQ−•, which has only a weak, 13× smaller,12 absorption at 580 nm. Thus, the reduced absorbance at t = 0 and faster decay are indicative of electron transport to the AQ end caps, with at least part occurring in 70 nm lengths of the present molecules are large. With our 15 ps experimental time resolution and half the charges trapped at the end caps faster than we can detect, we estimate that these electrons reached the caps in 2.2 cm2/s. The Einstein relation, D = μkbT, then gives μ > 86 cm2/(V s). The present observations would be stronger if we also observed the arrival of the electrons on the end traps in
