Landscape of Charge Carrier Transport in Doped Poly(3

Jun 18, 2018 - ACS eBooks; C&EN Global Enterprise .... Landscape of Charge Carrier Transport in Doped Poly(3-hexylthophene): Non-Contact Approach ...
0 downloads 0 Views 1MB Size
Letter Cite This: J. Phys. Chem. Lett. 2018, 9, 3639−3645

pubs.acs.org/JPCL

Landscape of Charge Carrier Transport in Doped Poly(3hexylthiophene): Noncontact Approach Using Ternary Combined Dielectric, Paramagnetic, and Optical Spectroscopies Yusuke Tsutsui,† Haruka Okamoto,† Daisuke Sakamaki,*,† Kazunori Sugiyasu,‡ Masayuki Takeuchi,‡ and Shu Seki*,† †

Department of Molecular Engineering, Graduate School of Engineering, Kyoto University, Nishikyo-ku, Kyoto 615-8510, Japan Molecular Design & Function Group, National Institute for Materials Science (NIMS), 1-2-1 Sengen, Tsukuba 305-0047, Japan

Downloaded via UNIV OF TOLEDO on June 20, 2018 at 11:37:30 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.



S Supporting Information *

ABSTRACT: We report on a comprehensive measurement system for mobility and energy states of charge carriers in matter under dynamic chemical doping. The temporal evolution of the iodine doping process of poly(3-hexylthiophene) (P3HT) was monitored directly through electron paramagnetic resonance (EPR) and optical absorption spectroscopy, as well as differential electrical conductivity by the microwave conductivity measurement. The increase in conductivity was observed after the EPR intensity reached a maximum and declined thereafter, and the conductivity finally reached ∼80 S cm−1. The carrier species changed from a paramagnetic polaron with an estimated mobility of μP+ ≈ 2 × 10−3 cm2 V−1 s−1 to an antiferromagnetic polaron pair with μPP+ ≈ 0.6 cm2 V−1 s−1. The technique presented here can be a ubiquitous method for rapid and direct observation of charge carrier mobility and energy states in p-type semiconducting materials as a completely noncontact, experimental, and quantitative technique.

C

intrinsic mobility, extensive efforts and improvements have been made in the conventional mobility assessment techniques to minimize extrinsic factors hampering ideal charge carrier transport: optimization of devices based on a single crystal,1−4 four-point probe FET measurement,5,6 and the application of a gated van der Pauw configuration.7 AC Hall measurements also provide intrinsic information on low-mobility materials in solid states.8,9 TRMC has also been one of the limited choices for measurement of the intrinsic property of various polymers in the solution/solid state as well as π-stacked materials. Radiation-induced ionization of the materials in pulseradiolysis TRMC (PR-TRMC)10−14 provides the most versatile and quantitative estimates of charge carrier yields, depending only on the density of electrons in the materials. This quantitative estimation of the yields enables instantaneous determination of the mobility of the radiolytically injected charge carriers. The main disadvantages of PR-TRMC are the limited availability of high-energy pulsed radiation sources and significantly low overall yield (per volume) of charge carriers compared to the other ionization techniques, making it difficult to assess the intrinsic mobility of charge carriers injected into thin solid films. Flash-photolysis TRMC (FP-TRMC)15−17 utilizes a pulsed laser for the injection of photocarriers into the material to provide information on charge recombination

harge carrier mobilitythe velocity of the charge carrier in a material under a unit electric fieldis one of the key factors for evaluation of electrical and photophysical properties of materials. Direct extraction of the mobility from electrical conductivity has been carried out by using current-based techniques such as field-effect transistor (FET), time-of-flight (TOF), space-charge-limited current (SCLC), and impedance spectroscopy (IS). These methods provide the proper way to measure the overall electrical performances of the actual operating devices because the current flows macroscopically in the whole device, which is important for the research and improvement on device applications. Compared to the macroscopic conductivity measurement techniques, timeresolved microwave conductivity (TRMC) is a complementary technique using electromagnetic waves as a probe for measurement of electrical conductivity and permittivity in the materials, which is different from the conventional currentbased techniques from the perspective of its noncontact nature. Probing the microwave absorption enables such noncontact conductivity measurement that is important for rapid screening of materials, especially for organic molecular and macromolecular materials with huge variety in their molecular and agglomerate structures. Fabrication of devices is unavoidable in electrical current-based techniques such as the aforementioned FET, SCLC, TOF, or IS, which hampers rapid assessment of electrical conductivity in newly designed and synthesized materials due to the requisite optimization of the device structures to lead to reliable values. Particularly, to extract their © XXXX American Chemical Society

Received: May 9, 2018 Accepted: June 18, 2018

3639

DOI: 10.1021/acs.jpclett.8b01465 J. Phys. Chem. Lett. 2018, 9, 3639−3645

Letter

The Journal of Physical Chemistry Letters

Figure 1. (a) Experimental setup for iodine doping. a: quartz tube; b: joint; c: isolation valve; d: plastic vial with iodine; e: sample; and f: holder. (b) Time evolution of the spin density in the P3HT film. (c) EPR signal of the P3HT film upon iodine doping. (d) Time evolution of peak-to-peak width (ΔHPP) of the EPR spectra.

to the iodine source with an isolation valve. Once the valve was opened, iodine vapor diffused into the quartz tube and the P3HT film, gradually leading to the formation of radical cations (holes). In this work, the doping process was roughly categorized into three stages: (I) beginning of doping (10−30 min), (II) formation of the polaron pair (PP, 30−50 min), as discussed below, and (III) the negative first derivative of spin density (>50 min). As shown in Figure 1b, the temporal change of the spin density in the P3HT film increases monotonically from 10 min in stage I. This response time matches well with the expected time of the iodine vapor to reach the sample. On the basis of the diffusion law (see the Supporting Information for the details), the iodine concentration at z = 17.5 cm will be 6.0% (18.3%) of the saturated value at 10 min (20 min) considering the diffusion coefficient of iodine D = 0.072 cm2 s−1 in air at room temperature (Figure S4). After reaching a maximum value of 3.5 × 1020 cm−3 at 60 min, the spin density decreases monotonically in the later stage (stage III). At the point of the maximum spin density, one spin is presumed to be injected for every 10 thiophene rings (10% doping relative to thiophene units) from the volume of the film. The decrease in the spin density in stage III indicates a transition from a polaron (P) to PP, in which two polarons are antiferromagnetically coupled.21 This assignment cannot be done solely from EPR measurement and supported by the absorption measurement where no significant absorption feature of the bipolaron (BP) was observed even at the highest doping level, as shown below. Theoretical and experimental works also support the favorable formation of a PP rather than a BP in the long polymer chain.21−25 Carriers would be homogeneously distributed in the film considering thin enough film thickness (35 nm), and 80% (90%) of iodine completely penetrates the film to reach the quartz substrate within 5 s (15 s), deduced from the diffusion constant of iodine in bulk P3HT (D = 2.5 × 10−11 cm2 s−1).26

dynamics. Designing photoionization systems in the target materials facilitates mobility assessment even in thin-film systems. The mobility values are given by a combination with the measurement of photocarrier generation yield from transient absorption spectroscopy or photogenerated charge accumulation with electrodes. On the other hand, the spin states of charge carriers in the material have not been largely considered in the conventional TRMC techniques, despite the fact that electron paramagnetic resonance (EPR) is also a well-established microwave-based spectroscopic technique. Revealing the relationship between the microscopic spin state and conductivity will provide a comprehensive understanding of the micro/macroscopic electrical conductivity and dynamics of the charge carriers as well as advanced insights into spintronics. In some representative and pioneering works, the spin state in operating devices such as metal−insulator−semiconductors (MISs) or FETs has been studied by using the field-induced electron spin resonance (FI-ESR) technique18−20 that allows investigation of the spatial extent of carrier wave functions19 and charge carrier distributions in the device.18 In this Letter, we propose a unique TRMC technique that monitors (1) microwave conductivity, (2) EPR, and (3) UV−vis−NIR absorption under the chemical doping process of the target material to obtain microscopic information on charge carriers. The feasibility of the present chemical-doping TRMC (CDTRMC) measurement is demonstrated clearly for a benchmark conjugated polymer, poly(3-hexylthiophene) (P3HT), as a target material, and the total landscape of the charge transport mechanism under doping by vapor diffusion of iodine was elucidated multilaterally by comprehensive analysis of the CDTRMC system. Doping of a P3HT film was preceded by diffusion of iodine vapor with the apparatus shown in Figure 1a. The spin-coated P3HT film (5 × 5 mm) was loaded in a quartz tube connected 3640

DOI: 10.1021/acs.jpclett.8b01465 J. Phys. Chem. Lett. 2018, 9, 3639−3645

Letter

The Journal of Physical Chemistry Letters

Figure 2. Differential absorption spectra of the P3HT film (colored lines) with the absorption spectrum of a pristine P3HT film (black line) in the energy regime of 1.16−3.5 eV (a) at the beginning of the doping (0−20 min, stage I−II) and (b) in the whole time range (0−1360 min, stage I− III) taken by a single detector setup. (c) Differential absorption spectra of the P3HT film in the energy regime of 0.74−3.5 eV in the whole time range (0−1410 min, stage I−III) taken by a dual detector setup.

Figure 3. (a) Block diagram of the microwave circuit and the chemical structure of P3HT. G: microwave generator; I: isolator; C: circulator; and S: sample. The change in (b) microwave reflectance spectra and (c) microwave conductivity of two different P3HT samples (red and blue) upon iodine doping.

conversion into BPs. The dominant charge carrier species in pristine P3HT is also considered to change from P (and πdimer) at the low doping level to PP with increasing doping and then to BP at the high level. However, no quantitative contribution has been deduced for the inter- and intrachain interaction between the spins of Ps.47 Figure 1c represents time-dependent EPR spectra of the P3HT film upon iodine doping. The signal intensity monotonically increased in stage I with a g value of 2.0021. On the other hand, the intensity monotonically decreased and the line shape became broader in stage II. The peak-to-peak width ΔHpp decreased slightly from 0.20 to 0.17 mT in 30 min, and this could be ascribed to the motional narrowing of Ps on PT chains (Figure 1d). After reaching the minimum at 30 min, ΔHpp increased up to 2 mT at 400 min, suggesting that the spin−spin relaxation time became shorter upon injecting Ps at high density, i.e., beginning of the formation of PPs. The EPR line shape was reproduced by the superposition of Lorentzian and Gaussian components, where Lorentzians and Gaussians can be considered as free and trapped carriers, respectively.29 The contribution of the Gaussian was less than 20% (Figure S5). Figure 2 shows time-dependent differential absorption spectra of the P3HT film. The data in Figure 2a,b were taken by a single multichannel detector. In the case of Figure 2c, the transmitted light was separated and fed into two detectors (one is for NIR, 0.75−1.24 eV, and another is for UV−vis, 1.24−3.6 eV), resulting in a worse S/N ratio than those in (a) and (b). In the entire range of doping, the appearance and successive growth of three transitions (below

The spin states of the doped carriers in polythiophenes (PTs) have been investigated in previous works.27−34 According to recent works, prerequisite charge carrier species in PTs have been considered and categorized as P, PP,21−24,35,36 and BP35−38 on a single isolated chain of PT. The existence of the π-dimer, which is a pair of interacting Ps on two different chains in the solution39−41 and condensed states,42,43 has also been confirmed. To simplify the complicated charge dynamics in the solid-state polymer with a kinetic mixture of crystallites and disordered phases, the carrier species in oligothiophenes (OTs) have been well studied as model systems both experimentally and theoretically. Janssen et al. reported that the dominant carrier species in sexithiophene switched from Ps to BPs upon doping, whereas the longer OTs are in an equilibrium between singly charged Ps and doubly charged species (π-dimer and/or PP) even at the lower level of charge carrier doping.40,44,45 For a chain with a distinct length of a 12-mer, PPs are predominantly formed rather than BPs, even in the heavily doped system.21−25 This was also confirmed by a theoretical study by taking electron correlation into account where a singlet PP would be in the ground state rather than a BP on the longer isolated OT.23 As for the carrier species in the solid state, the mechanism of the interchain interaction between charge carriers has not been fully understood to date.41 To obtain clearer insight into the charge states, “insulated” PT, whose main chain is caged by electrically inactive side chains,35,36,46 was studied. Because the interchain interaction was sterically hampered, they successfully observed the conversion of carrier species on isolated PT chains from Ps to PPs, followed by 3641

DOI: 10.1021/acs.jpclett.8b01465 J. Phys. Chem. Lett. 2018, 9, 3639−3645

Letter

The Journal of Physical Chemistry Letters

position of E2max. Figure S2 illustrates Emax and E2max values in the cavity (x = y = 0). Both Emax and E2max became smaller upon increasing the conductivity of the material up to σ = 103 S m−1. This is due to the dielectric loss of the microwave energy stored in the cavity by electrons in the conducting material, which is reflected by the decrease in the Q factor (Q) of the cavity (Figure S3). The filling factor was calculated from eq 3 numerically using the finite element method. The constant value of ∂F/∂d = 5.3 × 10−6 (μm)−1 was obtained in the whole σ range and adopted for further calculation of the conductivity. Figure 3b represents time-dependent reflection spectra of microwave from the cavity. By exposing the sample to iodine vapor, the resonant spectra became broader/shallower with a slight shift toward the lower resonant frequency. The broadening/shallowing corresponds to the change in the imaginary part of the complex permittivity of the P3HT film, and the real part is reflected by the resonant frequency shift. The change in the conductivity (Δσ) was evaluated using eq 2 and is plotted in Figure 3c. Iodine vapor reaches the sample position in 10 min (stage I), and in fact, the microwave conductivity of the film gradually increases from 10−2 S cm−1 from 20 min in the double-logarithmic plot. The abrupt rise of Δσ over 3 orders of magnitude at stages II−III in the semilogarithmic plot suggests that the mobility of carriers at the lower doping level is low and negligible compared with that of the carriers under higher doping. From the double-logarithmic plot, it can be observed that the conductivity increases from 99%, respectively. The samples were prepared by spin-coating (3500 rpm, 60 s) 10 μL of chlorobenzene solution of 1 wt % P3HT onto a quartz substrate (5 × 5 × 1 mm). The film thickness was determined with the surface profilometer (Dektak 150, Veeco) to be ∼35 nm. Chemical doping of P3HT was carried out by vapor diffusion of iodine, as shown in Figure 1a. Iodine (17.5 mg) was put into a vial and connected to a top-end closed and bottom-end opened quartz tube via a valve. The sample was loaded in the quartz tube (inner and outer diameters of 5.8 and 7.0 mm, respectively) at a height of 17.5 cm from the iodine source. Then, the quartz tube was loaded inside of the microwave cavity to evaluate the microwave conductivity or EPR upon iodine doping. Microwave generated from a synthesizer (SMF100A, Rohde & Schwarz) in the frequency modulation mode with a triangular wave was fed into an isolator, circulator, and resonant cavity with a resonant frequency at around 9 GHz using rectangular X-band waveguides. The coupling between the waveguide and the cavity was adjusted to critical without a sample. The power of the reflected microwave was detected with a Schottky diode (DZM124ABP, Herotek) connected to the oscilloscope (MDO 3024, Tektronix). The power of the injected microwave was adjusted to ca. 1 mW. Time-dependent UV− vis−NIR spectra were recorded in situ in a single-beam setup with a spectrometer (AvaSpec-2048 for >1.16 eV (Avantes) and Flame-NIR for