Letter
https://doi.org/10.1038/s41586-019-1504-9
Efficient molecular doping of polymeric
semiconductors driven by anion exchange
Yu Yamashita1,2,3, Junto tsurumi1,2,3, Masahiro Ohno1,2, ryo Fujimoto1,2, Shohei Kumagai1,2, tadanori Kurosawa1,2,
toshihiro Okamoto1,2,4,5, Jun takeya1,2,3,4 & Shun Watanabe1,2,4,5*
The efficiency with which polymeric semiconductors can be
chemically doped—and the charge carrier densities that can thereby
be achieved—is determined primarily by the electrochemical
redox potential between the π-conjugated polymer and the dopant
species^1 ,^2. Thus, matching the electron affinity of one with the
ionization potential of the other can allow effective doping^3 ,^4. Here
we describe a different process—which we term ‘anion exchange’—
that might offer improved doping levels. This process is mediated
by an ionic liquid solvent and can be pictured as the effective
instantaneous exchange of a conventional small p-type dopant anion
with a second anion provided by an ionic liquid. The introduction
of optimized ionic salt (the ionic liquid solvent) into a conventional
binary donor–acceptor system can overcome the redox potential
limitations described by Marcus theory^5 , and allows an anion-
exchange efficiency of nearly 100 per cent. As a result, doping levels
of up to almost one charge per monomer unit can be achieved. This
demonstration of increased doping levels, increased stability and
excellent transport properties shows that anion-exchange doping,
which can use an almost infinite selection of ionic salts, could be a
powerful tool for the realization of advanced molecular electronics.
Chemical doping of π-conjugated materials necessarily involves
redox reactions between the host and dopant^3. In this process, an inte-
ger number of electrons is transferred from the host to the dopant via
electron transfer in the ground state, which is well described by Marcus
theory^5. Because the driving force in donor–acceptor association is
dominated primarily by the electrochemical redox potential between
the π-conjugated material and the dopant, efficient doping occurs only
when charge transfer is energetically favourable. To achieve higher dop-
ing efficiency (for example, p-type doping), the electron affinity of the
acceptor (the dopant) should match or exceed the ionization potential
of the host material^2.
Various dopants and processes have been used to achieve the effi-
cient chemical doping of organic semiconductors. Despite the success-
ful tuning of the electron affinity of various conjugated molecules to
promote efficient doping, increasing the electron affinity often causes
chemical instability. This represents a major challenge to extending the
scope of potential molecular dopants, and there have been attempts
to use photo-assisted doping to mitigate this problem^6. In addition to
charge-transfer interactions based on redox reactions, Coulomb inter-
actions such as hole–hole and hole–counterion interactions in the case
of p-type doping^4 ,^7 are an important aspect of organic-semiconductor
doping, suggesting the possibility of optimizing molecular doping by
tuning Coulomb or ionic interactions^8 ,^9.
Here we demonstrate a general strategy for overcoming
charge-transfer limitations by using anion exchange. We focus on the
chemical doping of the well-studied thiophene-based conjugated pol-
ymer poly(2,5-bis(3-tetradecylthiophen-2-yl)thieno[3,2-b]thiophene)
(PBTTT)^10 in conjunction with tetrafluorotetracyanoquinodimethane
(F4TCNQ)^11 , as shown in Fig. 1a, introducing an additional anion to
the host–guest system (see Extended Data Fig. 1 and Supplementary
Information section 1.1). This combination of substances results in
spontaneous exchange of the F4TCNQ radical anion and the newly
introduced anion with near-unity exchange efficiency. The anion-
exchange doping process used here results in remarkable improve-
ments in both the doping level and the thermal durability of the doped
material.
Our anion-exchange doping process involves a host–guest system
that also includes a large excess of a salt (consisting of a cation, X+,
and an anion, Y−; Fig. 1b). As an example, an ionic liquid in which X is
1-ethyl-3-methylimidazolium (EMIM) and Y is bis(trifluoromethylsul-
fonyl)imide (TFSI) can be used. Surprisingly, in this scenario, the TFSI
anions are instantaneously exchanged with the F4TCNQ radical anions
that form the intermediate ion pair [PBTTT•+ F4TCNQ•−] (Fig. 1b).
Here, we realized anion-exchange doping by using the ionic liquid
EMIM-TFSI instead of n-butylacetate as a solvent for F4TCNQ (details
of this method are provided in Supplementary Information section
1.2). We confirmed hole doping of the polymer via charge transfer by
observing bleaching of the neutral absorption of PBTTT at 553 nm
(Fig. 1c). In contrast to a conventional binary system, a characteristic
doublet originating from the F4TCNQ radical anion was not seen in the
absorption spectrum of the doped film. We further verified the absence
of F4TCNQ radical anions in the PBTTT by Fourier transform infrared
(FTIR) spectroscopy (Fig. 1d and Extended Data Fig. 2). Specifically,
the FTIR spectrum of the anion-exchanged film does not show the
peak assigned to the C≡N stretching mode of the F4TCNQ radical
anion (2,190 cm−^1 ; ref.^12 ), indicating highly efficient anion exchange
(F4TCNQ•− → TFSI−).
We also confirmed the absence of the F4TNQ radical anion in the
PBTTT by Raman spectroscopy (Supplementary Information section
1.3) and electron spin resonance measurements. We estimated the
Curie spin concentration resulting from the F4TCNQ radical anions
to be 1.7 × 1019 cm−^3 (Extended Data Fig. 3), which is very low com-
pared with the actual doping level of roughly 1 × 1021 cm−^3 derived
from Hall effect measurements. Therefore, we conclude that the lower
limit of the anion-exchange efficiency for this process (F4TCNQ•− →
TFSI−) is 98%. This near-unity exchange efficiency suggests that the
present anion exchange is driven by the host–guest hybrid system seek-
ing to achieve a minimum free energy value. We note that hole doping
is never observed when a PBTTT thin film is immersed in a pure ionic
liquid (Extended Data Fig. 4), showing that the F4TCNQ has a vital
role by producing the initial hole doping, following which the F4TCNQ
radical anions are exchanged with Y−. Note that because a large excess
of TFSI− is introduced into the hybrid anion system, entropy gainaccording to − ∙−
kTB ln −[TFSI]
[F4TCNQ]can be expected, and will be discussedbelow (here, kB is the Boltzmann constant and T is temperature).
Ion exchange is applicable to various chemical processes, and oper-
ates not only on the principle of the binding preferences of ion pairs,(^1) Material Innovation Research Center (MIRC), University of Tokyo, Kashiwa, Japan. (^2) Department of Advanced Materials Science, Graduate School of Frontier Sciences, University of Tokyo, Kashiwa,
Japan.^3 International Center for Materials Nanoarchitectonics (WPI-MANA), National Institute for Materials Science (NIMS), Tsukuba, Japan.^4 AIST-UTokyo Advanced Operando-Measurement
Technology Open Innovation Laboratory (OPERANDO-OIL), National Institute of Advanced Industrial Science and Technology (AIST), Kashiwa, Japan.^5 Japan Science and Technology Agency (JST),
Precursory Research for Embryonic Science and Technology (PRESTO), Kawaguchi, Japan. *e-mail: [email protected]
634 | NAtUre | VOL 572 | 29 AUGUSt 2019