Molecular Doping and Trap Filling in Organic Semiconductor Host

Publication Date (Web): March 13, 2017 ... We investigate conductivity and mobility of different hosts mixed with different electron-withdrawing guest...
0 downloads 0 Views 2MB Size
Subscriber access provided by University of Newcastle, Australia

Article

Molecular Doping and Trap Filling in Organic Semiconductor Host-Guest Systems Guangzheng Zuo, Zhaojun Li, Olof Andersson, Hassan Abdalla, Ergang Wang, and Martijn Kemerink J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b01758 • Publication Date (Web): 13 Mar 2017 Downloaded from http://pubs.acs.org on March 15, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Molecular Doping and Trap Filling in Organic Semiconductor Host-Guest Systems

Guangzheng Zuo a, Zhaojun Li b, Olof Andersson a, Hassan Abdalla a, Ergang Wang b, and Martijn Kemerink a*

a

Complex Materials and Devices, Department of Physics, Chemistry and Biology, Linköping University, 58183 Linköping, Sweden

b

Department of Chemistry and Chemical Engineering, Chalmers University of Technology,

41296 Göteborg, Sweden

Abstract

We investigate conductivity and mobility of different hosts mixed with different electronwithdrawing guests, in concentrations ranging from ultralow to high. The effect of the guest material on mobility and conductivity of the host material varies systematically with the guests’ LUMO energy relative to the host HOMO, in quantitative agreement with a recently developed model. For guests with a LUMO within ∼0.5 eV of the host HOMO the dominant process governing transport is the competition between the formation of a deep tail in the host DOS and state filling. In other cases, the interaction with the host is dominated by any polar side groups on the guest and changes in the host 1

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

morphology. For relatively amorphous hosts the latter interaction can lead to a suppression of deep traps, causing a surprising mobility increase by 1-2 orders of magnitude. In order to analyze our data, we developed a simple method to diagnose both the presence and filling of traps.

1. Introduction Conjugated polymers (CPs) have been studied intensively as promising materials in organic semiconductors. In recent decades, hundreds of CPs have been synthesized and applied in light emitting diodes, organic field-effect transistors, solar cells and so on.1–5 In many applications of CPs, which can vary strongly in terms of energy levels and degree of crystallinity, are combined with another material to achieve high performance. For instance, F4TCNQ is commonly used as dopant in CPs to improve mobility and conductivity;6–9 PCBM is used as acceptor in solar cells with CP absorbers to achieve high power conversion efficiency.10–13 Triplet emitters are commonly employed to yield more favorable spin statistics in lighting applications.14–16 In what follows, we shall loosely refer to all such systems as host-guest systems, with the guest being the non-CP component. The primary question we shall address is the effect the guest material has on the charge transport in the host. The key parameter that is systematically evaluated in this work, is the energy level offset between the host HOMO and guest LUMO. F4TCNQ with its low (w.r.t. the vacuum level) LUMO is the most common p-type dopant for CPs that reportedly can improve mobility and conductivity in several material systems.6–9,17–20 However, that is for CPs like P3HT and PBTTT7–9,17, which all have a HOMO level above or close to the LUMO of F4TCNQ. On the other hand, reports about the influence of F4TCNQ on CPs with lower HOMO are rare. Furthermore, in most reports only blends with high fractions of F4TCNQ were examined. It is thus unclear what that the effect of F4TCNQ on various CPs in a large part of the available parameter space is. Generalizing the above, since effective molecular doping usually means the LUMO of the guest is close to or lower than the HOMO of the host material, studies regarding guests with higher 2

ACS Paragon Plus Environment

Page 2 of 26

Page 3 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

lying LUMO (or low lying HOMO in the case of n-type doping), and their impact on transport in the host are rare. This hiatus is of particular importance if one intents to use low concentrations of guests to selectively fill deep-lying traps and improve transport.21–24 When, in the case of p-type doping, an electron transfers from the host material to the guest dopant, the Coulomb potential of the ionized dopant acts as trap, diminishing the charge carrier mobility - Pingel and Neher experimentally demonstrated this effect in F4TCNQ-doped P3HT.7,8 Several analytic models have been suggested to describe this effect.25–28 In our previous work we developed a simple yet quantitative model based on the assumption of a doping-induced deep tail formed in the intrinsic Gaussian density of states (DOS) and benchmarked it against a numerical kinetic Monte Carlo model and experiments performed on the F4TCNQ:P3HT system.29 The effect of charge carriers not managing to escape the counter ion is directly related to the so-called doping efficiency, i.e. the fraction of dopants that produces a free charge that contributes to charge transport. Often this number is used as an empirical parameter characterizing a certain host-guest combination.19,30,31 Although Mityashin et al. proposed a mechanism for the escape of the charge carrier from its counter ion that allows quantitative predictions,32 their atomistic model is hard to implement and leaves a clear need for a simpler framework. Here, we present a systematic study to charge transport in host-guest systems where we varied both the host HOMO energy as well as the guest LUMO energy to arrive at a holistic picture that can be transferred to other material systems. F4TCNQ, with the lowest LUMO, acts as p-type dopant and in the low doping regime the mobilities of all hosts decrease with increasing F4TCNQ fraction due to the formation of (Coulomb) trapping potentials by F4TCNQ ions, in quantitative agreement with our recently developed model.29 At higher doping fractions, F4TCNQ aggregation causes either an apparent increase or decrease in mobility, depending on the host HOMO level. For PCBM, with the highest LUMO, the effect on mobility seems entirely morphological and depends on the crystallinity of the host. When mixed in P3HT (most crystalline) a systematic reduction in mobility was found,

3

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

whereas for TQ1 (most amorphous) a surprising and strong mobility increase was found that could be assigned to the removal of traps. For NDI-CN, the effect on mobilities can be rationalized as a competition between weak doping and trap filling on the one hand, and trap formation due to NDICN’s strong electronegative cyano group that induces multipole potentials on the other.

2.

Materials and Methods

Materials: Regioregular-poly(3-hexylthiophene-2,5-diyl) ( rr-P3HT, Mr=166.3/ repeat unit); [6,6]Phenyl-C71-butyric acid methyl ester(PCBM, Mr=1030.99), and 2,3,5,6-Tetrafluoro-7,7,8,8tetracyanoquinodimethane (F4TCNQ, Mr=276.15) were purchased from Sigma-Aldrich. Poly[2,6-(4,4bis-(2-ethylhexyl)-4H-cyclopenta [2,1-b;3,4-b’]dithiophene)-alt-4,7(2,1,3-benzothiadiazole)] (PCPDTBT, Mr=534.84/ repeat unit) was purchased from 1-Material Inc.. Poly [2,3-bis(3octyloxyphenyl)-5,8-quinoxalinediyl]-2,5-thiophenediyl] (TQ1, Mr=618.84/repeat unit) and corecyanated naphthalene diimide (NDI-CN, Mr=877.29) were synthesized at Chalmers University of Technology. F4TCNQ was dissolved in ortho-dichlorobenzene (o-DCB), NDI-CN and PCBM were dissolved in chloroform (CF), to make different solutions with concentrations from 10 mg/ml to 10-4 mg/ml, respectively. P3HT, PCPDTBT, and TQ1 were dissolved in CF. Solutions of different concentrations of F4TCNQ, NDI-CN, or PCBM were added in the appropriate ratio to the CP solution to get a doping range from 10-5 to 10-1 or, in some cases, 1 (molar ratio). At the same time, we changed the concentration of P3HT, PCPDTBT or TQ1 in CF and the spin-coating speed used for the deposition of the active layer to get different thicknesses for the active layer. Device Fabrication and measurement: Hole only devices were fabricated with the following structure: ITO/PEDOT:PSS (4083, 40nm) / active layer /MoO3 (5nm) / Al (90nm). The active layers were spin-coated on the top of the PEDOT:PSS film to reach a thickness of 200 nm to 400 nm as measured with a Dektak surface profilometer. All samples were thermally annealed at 120 °C for 10 minutes after spin-coating. For P3HT this led to significant conductivity increases, whereas for

4

ACS Paragon Plus Environment

Page 4 of 26

Page 5 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

PCPDTBT and TQ1 no such changes were found. In passing we note that the latter rules out that any significant F4TCNQ evaporation occurred during the annealing step. After that the Al-contact was evaporated under a pressure of 1x10-6 mbar. The current-density vs. voltage (J-V) curves of all devices were measured in ambient air with a Keithley 2400. Drift-diffusion simulations. Drift-diffusion simulations of unipolar devices were carried out by numerically solving the coupled Poison and drift-diffusion equations on an equidistant mesh using Scharfetter-Gummel interpolation.33 Boundary conditions for the charge density  at the contacts were calculated as Boltzmann factors  =  exp− ⁄  with the injection barrier. For the Ohmic contacts of our devices the standard condition =0 is used; using small nonzero values makes no difference. The density- and field-dependent mobility that results from Gaussian disorder was implemented using the parametrization of Pasveer.34 An additional Gaussian trap level of density Nt at a depth Et was implemented as described by Paasch and Scheinert.35 Doping was accounted for in lowest order through the inclusion of a homogenous and immobile distribution of counterions (e.g. anions for p-type doping) with density Ndope.

3. Results and Discussion Here we selected various, well-studied conjugated polymers with different HOMO levels and different degrees of crystallinity36–38 as host materials: P3HT (most crystalline), PCPDTBT and TQ1 (most amorphous). As guest materials, F4TCNQ, NDI-CN and PCBM were chosen. Full names of all compounds are given in the experimental section, structures are shown in the SI (Figure S1). The corresponding approximate energy levels as determined by cyclic voltammetry (CV) are shown in Figure 1a. Apart from varying in energy levels, the guests also vary in polar character – F4TCNQ has no dipole moment owing to its symmetric shape, whereas PCBM and NDI-CN have 4.01 D39 and 3.1 D respectively. The different host materials were mixed with one guest material at a time in molar

5

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 26

concentrations ranging from ultralow (10-5) to high (10-1). In order to diminish the effect from charges diffusing from the contacts,29,40 all film thicknesses were above 200 nm. The Ohmic conductivity  and Ohmic mobility  were measured in hole-only devices by fitting the expressions  =    = 

(1)

in the slope = 1 region of the j-V curves on double-log scale. Here,  = /, where  and  are the applied voltage and film thickness, respectively. Further, j is the current density and  =   with  the total density of states that is taken 1.7×1026 m-3 and  the (known) molar doping concentration. Note that by using this definition of n the problem of having to know the concentration of ‘free’ charges in the device is circumvented: n is the density of dopants and, since each dopant can in principle contribute one charge carrier, n is also the density of charges that is potentially available for charge transport. As such, the Ohmic mobility should not be identified with an intrinsic mobility of free charges. Moreover, space charge limited mobilities  are extracted from fits to the slope ≈ 2 region of the j-V curves using the Mott-Gurney law  =   !"#"

$% #&

(2)

where   is the dielectric constant of the material, with  taken to be 3.6. Generally, values for  that are extracted from this procedure should not be over-interpreted;41 however, we will only be interested in trends, rather than in absolute values.

6

ACS Paragon Plus Environment

Page 7 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 1. (a) Host and guest materials and their energy levels. Dashed lines indicate the combinations investigated in this work. Chemical structures are given in the SI. (b) Schematic illustration of the effect of doping on the density of states (DOS). Due to the attractive Coulomb potentials of the ionized dopants the original DOS (black line) is slightly broadened and gets an exponential lowenergy tail. (c) Schematic illustration of trap ‘neutralization’ encountered in this work. Top: the (assumed Gaussian) trap level gets filled by doping-induced charges; Bottom: the trap level vanishes due to the addition of a guest. 3.1 Mobility and Conductivity 3.1.1. F4TCNQ

7

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

We measured the J-V characteristics of hole-only devices based on P3HT, PCPDTBT and TQ1 blended with F4TCNQ in molar fraction from 10-5 to 10-1. The derived Ohmic conductivities and mobilities are shown in Figure 2 for molar fractions up to 10-3. At higher loadings a sudden increase was observed for all materials except P3HT that we attribute to factors that are beyond the present study that focusses on the energetics; the full data set is displayed in the SI. We found that the conductivity of P3HT increased approximately linearly with increasing F4TCNQ fraction from 10-5 to 10-3, with values that are quite similar to those reported by Pingel and Neher.8 In contrast, the conductivity of PCPDTBT and TQ1 just slightly increased in the same range. These trends are reflected in Figure 2b where the Ohmic mobility of P3HT is largely unaffected by the F4TCNQ doping in the range from low to moderate, 10-5 to 10-3, but the Ohmic mobilities of PCPDTBT and TQ1 gradually decrease with increasing F4TCNQ concentration at this range. In the following paragraphs we will

discuss and rationalize this behavior. Figure 2. Dependence of (a) the Ohmic conductivity and (b) the hole mobility derived from that for different polymers with various F4TCNQ doping concentration. Symbols represent experimental data of individual devices of TQ1 (black down-triangle, thicknesses 255 nm and 327 nm), PCPDTBT (red rhombus, thicknesses 230 nm and 260 nm ), P3HT (blue up-triangle, thicknesses 265 nm and 360 nm), and TQ1:PCBM (pentagon, 1:1wt). Dashed lines in (a) show the conductivity of pure CPs with 8

ACS Paragon Plus Environment

Page 8 of 26

Page 9 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

the same color. Solid lines represent analytical data from the model with the following parameters: P3HT (blue): ΔE = 0.24 eV; PCPDTBT (red): ΔE = -0.1 eV; TQ1 (black): ΔE = -0.23 eV; TQ1+PCBM (olive): ΔE = -0.16 eV. For all materials ' = 5e9 s-1 and σDOS = 0.075eV were used.

Focusing on low doping concentrations, a dopant molecule turns into dopant ion when an electron is transferred between the host material and the dopant. The resulting Coulomb potential acts as a trap for nearby charges. In transport modeling it is common to account for these traps through the introduction of a deep tail in the (Gaussian) DOS.25,29 At low doping concentrations, the effect of these traps is that a significant fraction of charges is effectively immobilized. Intuitively, it may also be expected that the transfer from the guest molecule to the host matrix gets more difficult with increasing difference between the guest LUMO and host HOMO. The experimental data in Figure 2 indeed show these behaviors: the Ohmic conductivity increases sublinearly with doping concentration and, concomitantly, the Ohmic mobility decreases. This effect gets stronger when -./. #3/. going from P3HT to TQ1, i.e. towards more negative Δ = )*+, − 012+, . Note that a naive state

filling model, neglecting trap formation and differences in Δ , would predict a super-linear increase in conductivity, and a concomitant monotonous increase in Ohmic mobility, that is the same for all studied CP. Note also that the low values for the Ohmic mobility for especially TQ1 only reflect an essentially constant Ohmic conductivity being divided by an increasing dopant concentration n, and should not be confused with e.g. a space charge limited mobility of ‘free’ charge. In order to get a more quantitative, and predictive understanding of the dominant processes, we developed a simple analytical model for the conductivity of doped CP that is based on the energetics of the system, taking the doping concentration c as independent input – note that this model is completely independent from the drift-diffusion model that is used below. It was extensively described and benchmarked against kinetic Monte Carlo simulations in a previous work.29 In brief, it accounts for the Coulomb-trapping effect through introduction of a dopant-induced deep tail in an

9

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

otherwise Gaussian DOS as illustrated in Figure 1b. The effect of any offset between the guest LUMO and host HOMO is mapped on the depth of this tail by replacing, within a radius equal to the intersite distance, the original semiconductor by an (effective) dopant, consisting of an energy level at #3/. 012+, and an ion. For simplicity, the disorder in the dopant site energies is assumed to be equal to

that of the semiconductor. Charge transport is modeled as either a variable range or nearest neighbor hopping process. For consistency with earlier work we chose to use the latter. Using percolation theory, the conductivity is assumed to be dominated by a critical hop between the Fermi energy 4 and a transport energy

 ∗ .42 As the charge carrier density is already accounted for in the position of 4 , the Boltzmann factor exp−  ∗ − 4 ⁄  is proportional to the conductivity. As such, not only the positive effect of state filling on charge carrier motion is included, but also the problem of having to explicitly calculate a fraction of ‘free’ charges, to be multiplied with a ‘free’ mobility, is circumvented. In the standard parameter set that is used unless indicated otherwise, the inter-site distance aNN = 1.8 nm, corresponding to a total site density N0 = 1.715×1026 m-3; the temperature T = 300 K, and the ioncountercharge binding energy 6789780 = 0.5 eV. Other input parameters are the attempt to hop -./. #3/. frequency ' , the Gaussian disorder σDOS and the energy difference Δ = )*+, − 012+, that is

calculated from the numbers in Figure 1. The model implicitly assumes that dopants are homogeneously dispersed. If dopants instead sit in certain parts of the CP only, e.g. in (non)aggregated domains,43 this gives rise to a small (linear) correction on the logarithmic concentration axes that would not affect any of our results. In summary, the model calculates the Ohmic conductivity  , taking the energy difference Δ and the dopant concentration c as main inputs. Given the complexity of the problem at hand, the agreement of the model calculations, shown as solid lines in Figure 2, with the experimental data is quite good for doping concentrations up to 10-3. This indicates that our model, despite its simplicity, captures the key physics of the problem, i.e. the 10

ACS Paragon Plus Environment

Page 10 of 26

Page 11 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

balance between dopant-induced deep trap states and state filling. It is important that the agreement with experiment in Figure 2 is obtained by only varying Δ , while otherwise using a single set of transport parameters that are consistent with previous works on these and similar materials.44– 46

This highlights two important aspects. First, the parameter Δ is the dominant factor determining

the conductivity increase for a certain host-dopant combination. Second, the energetic disorder σDOS in the pristine host is only of minor importance as transport in the doped system is governed by the dopant-induced tail states that are the same for all materials (except for the influence of Δ ), as shown in Fig. 1b. Importantly, our model does not have a parameter to account for doping efficiency; rather, this is a quantity that follows from the conductivity model.29 Hence, our results suggest that mobility, conductivity and doping efficiency are all dominated by the electrostatics of the problem. In fact, doping efficiency is a somewhat ill-defined quantity as in any energetically disordered hopping system the mobility is an ensemble average over individual charges with massively varying trapping times. There exists, therefore, a degree of arbitrariness in selecting particles that do and do not contribute to transport and the number of ‘free’ particles is not a good observable. Instead, conductivity, and the Ohmic mobility as defined through Equation 1, is a good and experimentally accessible observable. At higher concentrations the model predicts a smooth upturn due to the positive effects of state filling becoming dominant over the negative effects of increased disorder due to Coulomb trapping. The smooth upturn is not present in the experiments and a deviation occurs when the F4TCNQ fraction is above ∼10-2 as shown in the SI. 3.1.2. NDI-CN In principle, as the energy difference Δ between P3HT and NDI-CN is similar to that between TQ1 and F4TCNQ, the conductivity of P3HT blended with NDI-CN should behave similar to TQ1 with F4TCNQ. The Δ between TQ1 and NDI-CN is over -1 eV, which is beyond thermal excitation and NDI-

11

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

CN should have no effect on conductivity and mobility though a doping mechanism. In the experimental results shown in Fig.3, we found that the Ohmic conductivity of crystalline P3HT with NDI-CN increased slightly with small concentrations of NDI-CN added, to remain almost constant up to 10-2, followed by a steep decrease at 10-1. The latter drop is similar to the trend we observed in P3HT with F4TCNQ (SI) and therefore tentatively attributed to the destruction of the crystalline morphology that is characteristic of regio-regular P3HT after annealing.47 The plateau between 10-5 and 10-2 reflects the anticipated inefficient doping of P3HT by NDI-CN. Surprisingly, for (amorphous) TQ1 with NDI-CN, the Ohmic conductivity and the SCLC mobility gradually decreased with increasing NDI-CN fraction over the whole studied range. We will come back to this behavior in section 3.2.

Figure 3. Dependence of experimental conductivity (black lines) and SCLC hole mobility (blue lines) on NDI-CN fraction 3.1.3. PCBM 12

ACS Paragon Plus Environment

Page 12 of 26

Page 13 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

On basis of the very large (and negative) Δ no doping effect of PCBM as guest in any of the three host materials is expected. The measured Ohmic conductivity and SCLC mobility vs. PCBM concentration are shown in Figure 4. For each host the trend in Ohmic conductivity is equal to the trend in SCLC mobility, indicating that the observed trends are due to mobility changes and not to changes in free carrier concentration.

Figure 4. Dependence of experimental conductivity (black lines) and SCLC hole mobility (blue lines) and on PCBM fraction. Blending PCBM with P3HT decreases the hole mobility at higher PCBM fractions, which we attribute to the degeneration of the crystalline morphology as discussed above for F4TCNQ and NDICN.48 We attribute the low value !"#" ≈ 2×10-12 m2/Vs to the fact that the 1:1 molar ratio corresponds to a 1:6 weight ratio, which is significantly beyond the 1:1 weight ratio that is typically used in P3HT:PCBM solar cells and that still gives reasonable hole mobilities; a mobility around 10-12 13

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

m2/Vs as found here is actually not uncommon for a poor percolating network for hole transport in P3HT.49 For the less crystalline, more amorphous hosts PCPDTBT50,51 and TQ1 a surprising, essentially monotonous increase in mobility with PCBM concentration is found. In particular, at high PCBM fractions that are relevant to solar cell applications, we find that the hole mobility is one (PCPDTBT) to two (TQ1) orders of magnitude larger than in the pure host material. The increase in hole mobility for the amorphous CPs is similar to the result reported by the Blom group for the MDMO-PPV:PCBM system.52 Blom et al. speculated that the ring-like conformation of amorphous MDMO-PPV53 would have a detrimental effect on mobility and would be suppressed by a large PCBM content. McGehee et al. concluded that PCBM intercalation in blends with MDMO-PPV is responsible for the increase,54 whereas Savenije explained the dramatic increase in mobility by formation of PCBM-rich aggregates.55 For the present case, the role of PCBM-rich aggregates can be ruled out as the mobility upswing is already observed for molar ratios that are far below the solubility limit. To further test whether the PCBM blends with increased hole mobility behave as ‘normal’ materials, we also measured the mobility of a 1:1 wt. blend of TQ1 with PCBM doped by F4TCNQ. As shown by the green pentagons in Figure 2, this is indeed the case and the TQ1:PCBM blend follows exactly the same trend as the other host materials. 3.2 Trap formation and filling In order to substantiate the interpretation of the data above and to answer the open questions, i.e. the unexpected mobility decrease observed in the TQ1:NDI-CN system and the mobility upswing in the TQ1:PCBM and PCPDTBT:PCBM systems for F4TCNQ concentrations above 10-3, we developed a novel method to more easily identify various processes, and in particular trapping and trap filling, involved in charge transport. The method is based on the inspection of the logarithmic slope of unipolar J-V curves, viz. ; =

9?@A B 9?@A $

(3) 14

ACS Paragon Plus Environment

Page 14 of 26

Page 15 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Interpretation of the obtained traces was facilitated by using a simple 1D drift-diffusion model that accounts for Gaussian disorder through a density- and field-dependent mobility, trapping in an additional Gaussian trap level (assumed to be of equal width as the HOMO DOS) and doping; the corresponding energetics are illustrated in Figure 1c. Further details are given in the experimental section.56

Figure 5. Calculated slope vs. bias curves from the drift-diffusion model with constant mobility (solid black line), Gaussian disorder (solid red line), Gaussian disorder+trap (solid blue line), Gaussian disorder+trap +doping (dashed blue line) and constant mobility+doping (dashed black line). The insert shows the corresponding J-V curves. The parameters used are: aNN = 1.8 nm, ' = 1e11 s-1, and σDOS = 0.08 eV corresponding to µ0 = 6e-8 m2/Vs, Nt = 1e23 m-3, Et = 0.35 eV, Ndope = 1e23 m-3, and T = 300K.

15

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Based on this model, we calculated the illustrative J-V curves and slopes shown in Figure 5. In absence of traps, doping and Gaussian disorder, the well-known transition from Ohmic conduction (slope = 1) at low bias to space charge limited (slope = 2) is reproduced (solid black lines). Introduction of Gaussian disorder causes an upswing in slope at high bias due to the mobility increase with field and density (solid red). Inclusion of a Gaussian trap level produces a peak at the transition between the trap-limited and trap-filled regimes; the height and position of this peak depend on trap depth Et and trap density Nt, respectively (solid blue). Adding a significant amount of doping has the effect of extending the Ohmic region (dashed black) and suppressing the peak associated with trap filling (dashed blue). The latter is explained by the doping-induced charges filling the trap level, see Figure 1c top panel. 3.2.1. F4TCNQ

Figure 6. Slope vs. bias curves for P3HT (a), PCPDTBT (b) and TQ1 (c) hosts with different F4TCNQ guest molar fractions as indicated in the legend (solid lines). Dashed lines are model fits with σDOS = 0.08eV and material-specific parameters for PCPDTBT: pure (Nt = 5e22 m-3 and Et = 0.27 eV), and doped (Nt = 1e23 m-3, Et = 0.27 eV, and Nd = 1e23 m-3); for TQ1: pure (Nt = 5e22 m-3 and Et = 0.43 eV), and doped (Nt = 1e23 m-3, Et = 0.43 eV, and Nd = 1e23 m-3). The inset to panel b shows the corresponding J-V curves. Figure 6 shows the voltage dependence of the slope obtained from J-V curves of the host-F4TCNQ systems discussed above; as an example, the raw J-V curves for PCPDTBT are shown in the inset of panel b; for the other hosts the curves are shown in the SI. For pristine P3HT the slope does not show 16

ACS Paragon Plus Environment

Page 16 of 26

Page 17 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

any obvious peaks and the shape is similar to that of the constant mobility, no trap case in Figure 5; the upswing due to the Gaussian disorder is just visible above ∼3 V. Upon adding F4TCNQ, the Ohmic region (slope = 1) is extended to higher bias, as expected on basis of Figure 5. For pure PCPDTBT and TQ1, an obvious peak atop a slowly increasing background is observed. On basis of the discussion above, these are attributed to the existence of traps in the two CPs. The more pronounced background slope as compared to P3HT indicates a higher disorder, in agreement with the σDOS values used in Figure 2. Upon fitting these peaks with our model, we estimate trap depths of 0.27 eV and 0.43 eV, both at a density of ∼1023 m-3. In combination with the HOMO energies of these materials, 5.3 eV and 5.6 eV, these numbers do not suggest the existence of a universal hole trap in these materials.57 The minor deviations between fit and experiment regarding the peak shape are attributable to the assumption that the trap DOS is of equal width as the HOMO DOS, which is not important for the present discussion. Note that the parameters above cannot be compared to those used in the modeling of Figure 2 since the former characterize a trap that is not accounted for in the conductivity model of Figure 2. When adding a very low fraction F4TCNQ (10-5 to 10-4), we found the height of the peak actually increased and we suspect this results from deeper traps caused by F4TCNQ, in line with the interpretation of the decreasing Ohmic mobility in Figure 2. Likewise, the shift to slightly higher voltages for 10-3 can be attributed to a larger number of traps with increasing F4TCNQ fraction. It is clear that the trap filling peaks for PCPDTBT and TQ1 vanished at high F4TCNQ doping concentrations of 10-2 and above. These concentrations exactly coincide with the conductivity upswing, suggesting the latter is at least in part due to the effective removal of the trap by filling with doping-induced charges. Fitting these curves using our model with a dopant density of 1e23 m-3 (dashed lines in Figure 6) corroborates this interpretation. In contrast to the conductivity model used in Figure 2, the drift-diffusion model assumes that all dopants are ionized, hence this is an effective number representing a lower limit for what would actually be needed in an experiment.

17

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

3.2.2 NDI-CN

Figure 7. Slope vs. bias curves for TQ1 hosts with different NDI-CN guest molar fractions as indicated in the legend. The slope derived from J-V curves based on devices with TQ1 as host material mixed with different NDI-CN fractions are shown in Figure 7 – the less interesting curves for P3HT are shown in the SI (Figure S4). For TQ1 with NDI-CN, the trap peak around 1 V increased in height and shifted to higher voltages with NDI-CN fraction increasing from 10-5 to 10-1, indicating that more and deeper traps were introduced by NDI-CN. As NDI-CN cannot act as dopant in TQ1 because of the large negative Δ, these traps cannot simply be due to ionized NDI-CN. Instead, we attribute the trapping to higher order moments, i.e. dipole, quadrupole etc., associated with the two strong electronegative cyano groups in NDI-CN. Importantly, the increased trapping coincides with, and rationalizes, the gradual decrease in mobility as visible in Figure 3. 18

ACS Paragon Plus Environment

Page 18 of 26

Page 19 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

3.2.3 PCBM

Figure 8. Slope vs. bias curves for PCPDTBT (a) and TQ1 (b) hosts with different PCBM guest molar fractions as indicated in the legend. Dashed lines are model fits with parameters for PCPDTBT: Nt = 5e22 m-3 and Et = 0.30 eV; for TQ1 Nt = 5e22 m-3 and Et = 0.43 eV. These curves cannot be directly compared to Figure 5 as a different CP batch was used (PCPDTBT) and devices were somewhat thinner (TQ1). Figure 8 shows the voltage dependence of the slope for both amorphous CP hosts with PCBM as guest in concentrations ranging from 10-5 to 10-1 - for P3HT with PCBM, shown in the SI (Figure S5), no distinct conclusion can be drawn from the slope observations. For both PCPDTBT and TQ1, it is obvious that the trap peak diminishes and shifts to lower voltages with increasing PCBM fraction, indicating that the traps that are present in the pure hosts are somehow removed by mixing in PCBM, as sketched in Figure 1c (lower panel). This is fully consistent with the observation in Figure 4 that the mobilities of these materials increase with PCBM loading. Considering the high LUMO of PCBM, any doping effects, i.e. an electron transferring to the PCBM, filling the host trap, can be ruled out. Hence, we are led to the conclusion that the mobility of TQ1, and to a lesser degree of PCPDTBT, increased due to a morphological improvement58 as previously suggested for MDMO-PPV,52,53 or,

19

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

alternatively, to the removal of water from nanometre-sized voids within the microstructure of the pristine polymer by filling of these voids by PCBM.59 We should finally remark that the absence of a PCBM-induced trap peak, despite the fact that PCBM has a larger dipole moment than NDI-CN (4.01 D vs. 3.1 D, respectively), suggests it is not necessarily the guest dipole moment that leads to such a peak in Figure 7. Instead, we attribute the trap formation to the strongly electronegative CN group in NDI-CN. In F4TCNQ the effect of the 4 CN groups will be overwhelmed by the much stronger (monopole) Coulomb potential that results from charge transfer.

4. Conclusions We systematically investigated the mobility and conductivity of polymer host materials with different HOMO levels doped by guest materials with different LUMO levels. For systems where the guest LUMO is within ∼0.5 eV of the host HOMO the dominant interaction is doping, that is charge transfer between the host and guest. The resulting conductivity is set by a competition between the formation of deep trap states due to the Coulomb potential of the dopant ion, the offset between the host HOMO and donor LUMO, and state filling, in agreement with our recent theoretical model. NDI-CN which has two strongly polar sidegroups is found to induce additional traps that lead to a decreased hole mobility. The electrically inert guest PCBM has the distinct effect of causing the removal of hole traps in amorphous host materials, leading to an increase in hole mobility for TQ1 and PCPDTBT; a decrease was observed for (poly)crystalline P3HT that showed no trapping in its pure form. To arrive at our conclusions we developed a simple method to reveal trap formation and trap filling/removal by following peaks at intermediate bias in the logarithmic slope of unipolar currentvoltage curves. The method was corroborated by drift-diffusion modeling.

ASSOCIATED CONTENT

20

ACS Paragon Plus Environment

Page 20 of 26

Page 21 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Supporting Information.

The Supporting Information is available free of charge on the ACS Publication website at DOI:

Molecular structures of host and guest materials used in this work; Full dataset of F4TCNQ-doped hosts; Cyclic voltammetry (CV) curve and energy level of NDI-CN; Slope vs. bias curves for P3HT host with different NDI and PCBM guest varied with molar fractions, and J-V curves. (PDF)

AUTHOR INFORMATION

Corresponding Author *Tel: +46 700895377; Fax: -; E-mail: [email protected]

Notes

Any additional relevant notes should be placed here. ACKNOWLEDGMENTS The research by G.Z. is supported by the Chinese Scholarship Council (CSC). O.A. and H.A gratefully acknowledge funding by the Knut och Alice Wallenbergs stiftelse (project ‘Tail of the Sun’). E.W. thanks the Swedish Research Council, the Swedish Research Council Formas and Chalmers Area of Advance Energy for financial support. We are grateful to Olle Inganäs for the use of technical infrastructure.

REFERENCES (1) (2)

Nielsen, C. B.; Holliday, S.; Chen, H.-Y.; Cryer, S. J.; McCulloch, I. Non-Fullerene Electron Acceptors for Use in Organic Solar Cells. Acc. Chem. Res. 2015, 48, 2803–2812. Kola, S.; Sinha, J.; Katz, H. E. Organic Transistors in the New Decade: Toward n-Channel, Printed, and Stabilized Devices. J. Polym. Sci. Part B Polym. Phys. 2012, 50, 1090–1120. 21

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(3) (4)

(5)

(6)

(7) (8) (9)

(10) (11)

(12) (13)

(14) (15)

(16)

(17)

(18)

Zhao, Y.; Guo, Y.; Liu, Y. 25th Anniversary Article: Recent Advances in n-Type and Ambipolar Organic Field-Effect Transistors. Adv. Mater. 2013, 25, 5372–5391. Grimsdale, A. C.; Leok Chan, K.; Martin, R. E.; Jokisz, P. G.; Holmes, A. B. Synthesis of LightEmitting Conjugated Polymers for Applications in Electroluminescent Devices. Chem. Rev. 2009, 109, 897–1091. Ye, L.; Zhang, S.; Huo, L.; Zhang, M.; Hou, J. Molecular Design toward Highly Efficient Photovoltaic Polymers Based on Two-Dimensional Conjugated Benzodithiophene. Acc. Chem. Res. 2014, 47, 1595–1603. Gao, W.; Kahn, A. Controlled P-Doping of Zinc Phthalocyanine by Coevaporation with Tetrafluorotetracyanoquinodimethane: A Direct and Inverse Photoemission Study. Appl. Phys. Lett. 2001, 79, 4040–4042. Pingel, P.; Schwarzl, R.; Neher, D. Effect of Molecular p-Doping on Hole Density and Mobility in poly(3-Hexylthiophene). Appl. Phys. Lett. 2012, 100, 143303. Pingel, P.; Neher, D. Comprehensive Picture of p-Type Doping of P3HT with the Molecular Acceptor F4TCNQ. Phys. Rev. B 2013, 87, 115209. Cochran, J. E.; Junk, M. J. N.; Glaudell, A. M.; Miller, P. L.; Cowart, J. S.; Toney, M. F.; Hawker, C. J.; Chmelka, B. F.; Chabinyc, M. L. Molecular Interactions and Ordering in Electrically Doped Polymers: Blends of PBTTT and F4TCNQ. Macromolecules 2014, 47, 6836–6846. Dang, M. T.; Hirsch, L.; Wantz, G. P3HT:PCBM, Best Seller in Polymer Photovoltaic Research. Adv. Mater. 2011, 23, 3597–3602. Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J. Polymer Photovoltiac Cells: Enhanced Efficiencies via a Network of Internal Donor-Acceptor Heterojunctions. Science 1995, 270, 1789–1791. Thompson, B. C.; Fréchet, J. M. J. Polymer–Fullerene Composite Solar Cells. Angew. Chem. Int. Ed. 2008, 47, 58–77. Wang, M.; Hu, X.; Liu, P.; Li, W.; Gong, X.; Huang, F.; Cao, Y. Donor–Acceptor Conjugated Polymer Based on Naphtho[1,2-c:5,6-c]bis[1,2,5]thiadiazole for High-Performance Polymer Solar Cells. J. Am. Chem. Soc. 2011, 133, 9638–9641. Gong, X.; Robinson, M. R.; Ostrowski, J. C.; Moses, D.; Bazan, G. C.; Heeger, A. J. HighEfficiency Polymer-Based Electrophosphorescent Devices. Adv. Mater. 2002, 14, 581–585. Su, Y.-J.; Huang, H.-L.; Li, C.-L.; Chien, C.-H.; Tao, Y.-T.; Chou, P.-T.; Datta, S.; Liu, R.-S. Highly Efficient Red Electrophosphorescent Devices Based on Iridium Isoquinoline Complexes: Remarkable External Quantum Efficiency Over a Wide Range of Current. Adv. Mater. 2003, 15, 884–888. Lo, S.-C.; Male, N. A. H.; Markham, J. P. J.; Magennis, S. W.; Burn, P. l.; Salata, O. V.; Samuel, I. D. W. Green Phosphorescent Dendrimer for Light-Emitting Diodes. Adv. Mater. 2002, 14, 975– 979. Yim, K.-H.; Whiting, G. L.; Murphy, C. E.; Halls, J. J. M.; Burroughes, J. H.; Friend, R. H.; Kim, J.-S. Controlling Electrical Properties of Conjugated Polymers via a Solution-Based P-Type Doping. Adv. Mater. 2008, 20, 3319–3324. Rangger, G. M.; Hofmann, O. T.; Romaner, L.; Heimel, G.; Bröker, B.; Blum, R.-P.; Johnson, R. L.; Koch, N.; Zojer, E. F4TCNQ on Cu, Ag, and Au as Prototypical Example for a Strong Organic Acceptor on Coinage Metals. Phys. Rev. B 2009, 79, 165306. 22

ACS Paragon Plus Environment

Page 22 of 26

Page 23 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(19) (20)

(21)

(22)

(23) (24)

(25) (26) (27)

(28) (29) (30)

(31)

(32)

(33) (34)

(35)

Wang, C.; Duong, D. T.; Vandewal, K.; Rivnay, J.; Salleo, A. Optical Measurement of Doping Efficiency in poly(3-Hexylthiophene) Solutions and Thin Films. Phys. Rev. B 2015, 91, 085205. Olthof, S.; Tress, W.; Meerheim, R.; Lüssem, B.; Leo, K. Photoelectron Spectroscopy Study of Systematically Varied Doping Concentrations in an Organic Semiconductor Layer Using a Molecular p-Dopant. J. Appl. Phys. 2009, 106, 103711. Zhang, Y.; Zhou, H.; Seifter, J.; Ying, L.; Mikhailovsky, A.; Heeger, A. J.; Bazan, G. C.; Nguyen, T.Q. Molecular Doping Enhances Photoconductivity in Polymer Bulk Heterojunction Solar Cells. Adv. Mater. 2013, 25, 7038–7044. Olthof, S.; Mehraeen, S.; Mohapatra, S. K.; Barlow, S.; Coropceanu, V.; Brédas, J.-L.; Marder, S. R.; Kahn, A. Ultralow Doping in Organic Semiconductors: Evidence of Trap Filling. Phys. Rev. Lett. 2012, 109, 176601. Zhang, Y.; de Boer, B.; Blom, P. W. M. Trap-Free Electron Transport in Poly(p-Phenylene Vinylene) by Deactivation of Traps with n-Type Doping. Phys. Rev. B 2010, 81, 085201. Yan, H.; Manion, J. G.; Yuan, M.; García de Arquer, F. P.; McKeown, G. R.; Beaupré, S.; Leclerc, M.; Sargent, E. H.; Seferos, D. S. Increasing Polymer Solar Cell Fill Factor by Trap-Filling with F4-TCNQ at Parts Per Thousand Concentration. Adv. Mater. 2016, 28, 6491–6496. Arkhipov, V. I.; Emelianova, E. V.; Heremans, P.; Bässler, H. Analytic Model of Carrier Mobility in Doped Disordered Organic Semiconductors. Phys. Rev. B 2005, 72, 235202. Tietze, M. L.; Pahner, P.; Schmidt, K.; Leo, K.; Lüssem, B. Doped Organic Semiconductors: TrapFilling, Impurity Saturation, and Reserve Regimes. Adv. Funct. Mater. 2015, 25, 2701–2707. Tietze, M. L.; Burtone, L.; Riede, M.; Lüssem, B.; Leo, K. Fermi Level Shift and Doping Efficiency in p-Doped Small Molecule Organic Semiconductors: A Photoelectron Spectroscopy and Theoretical Study. Phys. Rev. B 2012, 86, 035320. Schmechel, R. Hopping Transport in Doped Organic Semiconductors: A Theoretical Approach and Its Application to p-Doped Zinc-Phthalocyanine. J. Appl. Phys. 2003, 93, 4653–4660. Zuo, G.; Abdalla, H.; Kemerink, M. Impact of Doping on the Density of States and the Mobility in Organic Semiconductors. Phys. Rev. B 2016, 93, 235203. Di Nuzzo, D.; Fontanesi, C.; Jones, R.; Allard, S.; Dumsch, I.; Scherf, U.; von Hauff, E.; Schumacher, S.; Da Como, E. How Intermolecular Geometrical Disorder Affects the Molecular Doping of Donor–acceptor Copolymers. Nat. Commun. 2015, 6, 6460. Tietze, M. L.; Burtone, L.; Riede, M.; Lüssem, B.; Leo, K. Fermi Level Shift and Doping Efficiency in p-Doped Small Molecule Organic Semiconductors: A Photoelectron Spectroscopy and Theoretical Study. Phys. Rev. B 2012, 86, 035320. Mityashin, A.; Olivier, Y.; Van Regemorter, T.; Rolin, C.; Verlaak, S.; Martinelli, N. G.; Beljonne, D.; Cornil, J.; Genoe, J.; Heremans, P. Unraveling the Mechanism of Molecular Doping in Organic Semiconductors. Adv. Mater. 2012, 24, 1535–1539. Scharfetter, D. L.; Gummel, H. K. Large-Signal Analysis of a Silicon Read Diode Oscillator. IEEE Trans. Electron Devices 1969, 16, 64–77. Pasveer, W. F.; Cottaar, J.; Tanase, C.; Coehoorn, R.; Bobbert, P. A.; Blom, P. W. M.; de Leeuw, D. M.; Michels, M. A. J. Unified Description of Charge-Carrier Mobilities in Disordered Semiconducting Polymers. Phys. Rev. Lett. 2005, 94, 206601. Paasch, G.; Scheinert, S. Charge Carrier Density of Organics with Gaussian Density of States: Analytical Approximation for the Gauss–Fermi Integral. J. Appl. Phys. 2010, 107, 104501. 23

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(36)

(37) (38)

(39)

(40) (41)

(42) (43) (44)

(45) (46)

(47)

(48)

(49)

(50)

(51)

Irwin, M. D.; Buchholz, D. B.; Hains, A. W.; Chang, R. P. H.; Marks, T. J. p-Type Semiconducting Nickel Oxide as an Efficiency-Enhancing Anode Interfacial Layer in Polymer BulkHeterojunction Solar Cells. Proc. Natl. Acad. Sci. 2008, 105, 2783–2787. Mühlbacher, D.; Scharber, M.; Morana, M.; Zhu, Z.; Waller, D.; Gaudiana, R.; Brabec, C. High Photovoltaic Performance of a Low-Bandgap Polymer. Adv. Mater. 2006, 18, 2884–2889. Wang, E.; Hou, L.; Wang, Z.; Hellström, S.; Zhang, F.; Inganäs, O.; Andersson, M. R. An Easily Synthesized Blue Polymer for High-Performance Polymer Solar Cells. Adv. Mater. 2010, 22, 5240–5244. de Gier, H. D.; Jahani, F.; Broer, R.; Hummelen, J. C.; Havenith, R. W. A. Promising Strategy To Improve Charge Separation in Organic Photovoltaics: Installing Permanent Dipoles in PCBM Analogues. J. Phys. Chem. A 2016, 120, 4664–4671. Craciun, N. I.; Brondijk, J. J.; Blom, P. W. M. Diffusion-Enhanced Hole Transport in Thin Polymer Light-Emitting Diodes. Phys. Rev. B 2008, 77, 035206. Blakesley, J. C.; Castro, F. A.; Kylberg, W.; Dibb, G. F. A.; Arantes, C.; Valaski, R.; Cremona, M.; Kim, J. S.; Kim, J.-S. Towards Reliable Charge-Mobility Benchmark Measurements for Organic Semiconductors. Org. Electron. 2014, 15, 1263–1272. Cottaar, J.; Koster, L. J. A.; Coehoorn, R.; Bobbert, P. A. Scaling Theory for Percolative Charge Transport in Disordered Molecular Semiconductors. Phys. Rev. Lett. 2011, 107, 136601. Duong, D. T.; Wang, C.; Antono, E.; Toney, M. F.; Salleo, A. The Chemical and Structural Origin of Efficient p-Type Doping in P3HT. Org. Electron. 2013, 14, 1330–1336. Gao, F.; Himmelberger, S.; Andersson, M.; Hanifi, D.; Xia, Y.; Zhang, S.; Wang, J.; Hou, J.; Salleo, A.; Inganäs, O. The Effect of Processing Additives on Energetic Disorder in Highly Efficient Organic Photovoltaics: A Case Study on PBDTTT-C-T:PC71BM. Adv. Mater. 2015, 27, 3868– 3873. Mauer, R.; Kastler, M.; Laquai, F. The Impact of Polymer Regioregularity on Charge Transport and Efficiency of P3HT:PCBM Photovoltaic Devices. Adv. Funct. Mater. 2010, 20, 2085–2092. Melianas, A.; Pranculis, V.; Devižis, A.; Gulbinas, V.; Inganäs, O.; Kemerink, M. DispersionDominated Photocurrent in Polymer:Fullerene Solar Cells. Adv. Funct. Mater. 2014, 24, 4507– 4514. Gao, J.; Roehling, J. D.; Li, Y.; Guo, H.; Moulé, A. J.; Grey, J. K. The Effect of 2,3,5,6-Tetrafluoro7,7,8,8-Tetracyanoquinodimethane Charge Transfer Dopants on the Conformation and Aggregation of poly(3-Hexylthiophene). J. Mater. Chem. C 2013, 1, 5638–5646. Chen, F.-C.; Ko, C.-J.; Wu, J.-L.; Chen, W.-C. Morphological Study of P3HT:PCBM Blend Films Prepared through Solvent Annealing for Solar Cell Applications. Sol. Energy Mater. Sol. Cells 2010, 94, 2426–2430. Mihailetchi, V. D.; Xie, H. X.; de Boer, B.; Koster, L. J. A.; Blom, P. W. M. Charge Transport and Photocurrent Generation in Poly(3-Hexylthiophene): Methanofullerene Bulk-Heterojunction Solar Cells. Adv. Funct. Mater. 2006, 16, 699–708. Azazi, A.; Mabrouk, A.; Chemek, M.; Kreher, D.; Alimi, K. DFT Modeling of Conjugated Copolymers Photophysical Properties: Towards Organic Solar Cell Application. Synth. Met. 2014, 198, 314–322. Chen, H.-Y.; Hou, J.; Hayden, A. E.; Yang, H.; Houk, K. N.; Yang, Y. Silicon Atom Substitution Enhances Interchain Packing in a Thiophene-Based Polymer System. Adv. Mater. 2010, 22, 371–375. 24

ACS Paragon Plus Environment

Page 24 of 26

Page 25 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(52)

(53)

(54)

(55) (56) (57)

(58)

(59)

Melzer, C.; Koop, E. J.; Mihailetchi, V. D.; Blom, P. W. M. Hole Transport in Poly(phenylene Vinylene)/Methanofullerene Bulk-Heterojunction Solar Cells. Adv. Funct. Mater. 2004, 14, 865–870. Kemerink, M.; van Duren, J. K. J.; Jonkheijm, P.; Pasveer, W. F.; Koenraad, P. M.; Janssen, R. A. J.; Salemink, H. W. M.; Wolter, J. H. Relating Substitution to Single-Chain Conformation and Aggregation in Poly(p-Phenylene Vinylene) Films. Nano Lett. 2003, 3, 1191–1196. Cates, N. C.; Gysel, R.; Dahl, J. E. P.; Sellinger, A.; McGehee, M. D. Effects of Intercalation on the Hole Mobility of Amorphous Semiconducting Polymer Blends. Chem. Mater. 2010, 22, 3543–3548. Savenije, T. J.; Kroeze, J. E.; Wienk, M. M.; Kroon, J. M.; Warman, J. M. Mobility and Decay Kinetics of Charge Carriers in Photoexcited PCBM/PPV Blends. Phys. Rev. B 2004, 69, 155205. Nicolai, H. T.; Mandoc, M. M.; Blom, P. W. M. Electron Traps in Semiconducting Polymers: Exponential versus Gaussian Trap Distribution. Phys. Rev. B 2011, 83, 195204. Nicolai, H. T.; Kuik, M.; Wetzelaer, G. a. H.; de Boer, B.; Campbell, C.; Risko, C.; Bredas, J. L.; Blom, P. W. M. Unification of Trap-Limited Electron Transport in Semiconducting Polymers. Nat. Mater. 2012, 11, 882–887. Kim, Y.; Yeom, H. R.; Kim, J. Y.; Yang, C. High-Efficiency Polymer Solar Cells with a CostEffective Quinoxaline Polymer through Nanoscale Morphology Control Induced by Practical Processing Additives. Energy Environ. Sci. 2013, 6, 1909–1916. Nikolka, M.; Nasrallah, I.; Rose, B.; Ravva, M. K.; Broch, K.; Sadhanala, A.; Harkin, D.; Charmet, J.; Hurhangee, M.; Brown, A.; et al. High Operational and Environmental Stability of HighMobility Conjugated Polymer Field-Effect Transistors through the Use of Molecular Additives. Nat. Mater. 2017, 16, 356–362.

25

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table of Contents Graphic

26

ACS Paragon Plus Environment

Page 26 of 26