Optical Dedoping Mechanism for P3HT:F4TCNQ Mixtures - The

Oct 12, 2016 - (9, 10) This method, called doping-induced solubility control (DISC), uses molecular dopants that charge the semiconducting polymer to ...
4 downloads 10 Views 3MB Size
Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Letter

Optical Dedoping Mechanism of P3HT/F4TCNQ Mixtures Jack Fuzell, Ian E. Jacobs, Sophia Ackling, Thomas F. Harrelson, David Mark Huang, Delmar S. Larsen, and Adam J. Moule J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.6b02048 • Publication Date (Web): 12 Oct 2016 Downloaded from http://pubs.acs.org on October 14, 2016

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 Letters 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 30

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 Letters

Optical Dedoping Mechanism for P3HT:F4TCNQ Mixtures Jack Fuzell,a# Ian E. Jacobs,b# Sophia Ackling,d Thomas F. Harrelson,c David M. Huang,d Delmar Larsen,a,* and Adam J. Moulé a,c,* a. Department of Chemistry, University of California, Davis, 1 Shields Ave., Davis CA 95616 U.S.A. b. Department of Materials Science, University of California, Davis, 1 Shields Ave., Davis, CA 95616 U.S.A c. Department of Chemical Engineering, University of California, Davis, 1 Shields Ave., Davis, CA 95616 U.S.A d. Department of Chemistry, School of Physical Sciences, The University of Adelaide, SA 5005, Australia

Corresponding Authors *Adam J Moule [email protected] and Delmar Larsen [email protected] #

These authors contributed equally to this manuscript

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

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 2 of 30

ABSTRACT Doping-induced solubility control (DISC) is a recently introduced photolithographic technique for semiconducting polymers, which utilizes reversible changes in polymer solubility upon doping to allow the polymer to function as its own photoresist. Central to this process is a wavelength sensitive optical dedoping reaction, which is poorly understood but generates sub-diffraction-limited topographic features, and provides optical control of the polymer doping level. Here, we examine the mechanism of optical dedoping in the semiconducting polymer poly-3-hexylthiophene (P3HT) doped by 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquinodimethane (F4TCNQ), via a combination of ultrafast and steady-state spectroscopy, ab initio calculations, and multi-dimensional NMR. A simple photoinduced back electron transfer mechanism from reduced F4TCNQ to oxidized P3HT does not explain the observed photophysics. Instead, photoexcited F4TCNQ* reacts with THF solvent molecules to form a neutral, non-doping, and highly soluble F4TCNQ-THF complex. Hence, ionized F4TCNQ is removed from the P3HT indirectly by depletion of the neutral F4TCNQ. Because the reaction involves only the dopant and similar photoreactivity would expected for most other dopant molecules, we expect optical DISC patterning should be generalizable to a wide range of polymer:dopant systems. TOC GRAPHICS

2 ACS Paragon Plus Environment

Page 3 of 30

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 Letters

KEYWORDS: electronic polymer, doping, patterning, P3HT, photoreaction, photoinduced Organic semiconductors have been heavily researched for applications involving sensors in medical devices, mobile gaming and computing, smart houses, and smart materials.1-3 . It is predicted that the market for sensors will grow from the current amount of three and a half billion devices in 2012 to over a trillion devices in ~2020.4-5 Organic semiconductors possess significant advantages over inorganic materials in these applications: they are mechanically flexible, low cost, and can be chemically tailored to specific applications. Organic semiconductors are currently used in LEDs and smartphone and television screens, but these devices are mostly made using vacuum thermal evaporation of small molecules or oligomers, a process that does not fully realize the potential price savings and prevents the use of polymeric materials.6 Vacuum deposition is the current industrial method of choice because electronic devices generally require a large number of layers to function.7 Solution and printing methods that deposit organic materials are unable to effectively coat and pattern multiple chemically controlled layers due to the mutual solubility and miscibility of organic materials. Rollto-roll printing of multiple functional layers using solution methods remains the greatest challenge for organic electronics technology.8 3 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

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 4 of 30

Recently, the Moulé group proposed a solution-based method to coat multiple layers of polymers from solution and to laterally pattern these layers.9-10 This method, called doping-induced solubility control (DISC), uses molecular dopants that charge the semiconducting polymer to render the polymer insoluble in typical processing solvents. Selective deposition or removal of the dopants returns solubility to the polymer locally, allowing for sub-300 nanometer scale patterning.9 The initial demonstration of this process used the hole-conducting polymer poly(3-hexylthiophene) (P3HT) doped with 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquinodimethane (F4TCNQ),9 but the reduction in polymer solubility upon doping appears to be general.9, 11 Remarkably, it was found that when P3HT:F4TCNQ films were exposed to 405 nm light while immersed in solvent, the film dedoped, causing dissolution of the exposed area.9-10 Shorter (UV) and longer (543 nm) wavelengths left the film intact, insoluble, and fully doped. Most recently, DISC patterning with 405 nm light was shown to produce sub-diffraction-limited topographic patterning and doping-level modulation in P3HT:F4TCNQ films.10 To the authors’ knowledge this is the first time that the optical removal of a dopant from a semiconductor has been shown. However, the mechanism that underlies this dedoping process has remained elusive, and is the subject of this letter.

P3HT is a semi-crystalline, semiconducting polymer widely studied for both organic photovoltaic (OPV) and organic field effect transistor (oFET) applications.12-13 Recently, several studies have focused on the p-type doping of P3HT with high electron affinity (EA) small molecules such as F4TCNQ in order to improve its charge transport properties.9,

11, 14-19

P-type doping in semiconducting polymers is similar to doping in

4 ACS Paragon Plus Environment

Page 5 of 30

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 Letters

inorganic semiconductors: in both cases, impurities are added to introduce extra holes in the valence band, resulting in an increase in conductivity of several orders of magnitude. However, for reasons that are not yet fully understood, the doping efficiency in organic semiconductors is quite low.20 As a result, doping levels in organic materials are typically on the order of percent, rather than the ppm levels in inorganic semiconductors. Unlike many molecular organic dopant pairs, when F4TCNQ dopes P3HT there is effectively a complete transfer of one electron from P3HT to the F4TCNQ molecule, forming P3HT+ and F4TCNQ-.16, 18, 20

Figure 1. a) UV/vis absorption spectra of P3HT (black) and P3HT:F4TCNQ (red) in film samples; also shown are neutral F4TCNQ (cyan) in acetonitrile and F4TCNQ- (blue) doped by KI in acetonitrile. b) Calculated molecular-orbital energy-level diagram using 5 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

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 30

DFT at the LRC-ωPBEh/6-31G* level of theory and implicit solvent (ε = 3.0) for a P3MT octamer, F4TCNQ, and P3MT-octamer/F4TCNQ dimer. Blue and red lines indicate the HOMO and LUMO of the CT state, respectively, and arrows indicate a possible photoinduced back electron transfer pathway.

Figure 1 contrasts the absorption spectra of a pristine P3HT film, a P3HT film doped with F4TCNQ (P3HT:F4TCNQ), neutral F4TCNQ, and reduced F4TCNQ in an acetonitrile (CH3CN) solution. The P3HT:F4TCNQ film spectrum shows contributions from neutral F4TCNQ, F4TCNQ-, neutral P3HT, and oxidized/doped P3HT (P3HT+, visible as a broad polaron absorption from 600 to 1100nm) as previously reported.16, 18 Examination of the spectra in Figure 1 shows that neutral F4TCNQ, F4TCNQ-, and P3HT all have significant absorbance features at 405 nm. All three of these of these species appear to be present in the P3HT:F4TCNQ spectrum, suggesting that absorption by one or more of these species may be responsible for dedoping. Although P3HT:F4TCNQ appears to undergo integer charge transfer,16,

18, 20

integer charge transfer is not a universal feature of doped organic semiconductors. Many if not most doping interactions between organic molecules result in fractionally charged and highly coupled interfacial states (referred to here as CT states) that give rise to a set of CT absorptions, which are not present in the neutral or integer charged species. It is possible that an excited CT state might dissociate into neutral species rather than back into the doped state, allowing for optical dedoping. Although the P3HT:F4TCNQ spectrum in Figure 1a does not appear to show any strong CT absorptions (because charge transfer is understood to be essentially integer), there may be a weak CT 6 ACS Paragon Plus Environment

Page 7 of 30

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 Letters

absorption near 405 nm which is not obviously visible due to the other overlapping absorption bands present. Figure 1b shows a DFT simulation F4TCNQ, poly-3-methylthiophene octamer, and the CT states generated by the combined P3MT-octamer/F4TCNQ in the face-on geometry using the LRC-ωPBEh/6-31G* functional in implicit solvent to model the surrounding P3HT chains in the polymer film (dielectric constant ε = 3.0). Supplemental Information Section 1 gives an in-depth discussion of this simulation and makes comparisons against another functional and in vacuo calculations. From time-dependent density functional theory (TD-DFT) calculations with both functionals and with and without solvent, there are several higher energy states that can be accessed with ~3.1 eV (405 nm) excitation that have lower electron affinity (higher energy) than the LUMO of neutral thiophene. Arrows on the energy-level diagram indicate a hypothetical route to photo-excited back charge transfer, with the vertical arrows indicating the molecular orbitals that contribute most significantly to the most intense transitions near 3.1 eV in the TD-DFT calculation.

7 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

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 8 of 30

Figure 2. Transient absorption spectra excited using 400 nm pulses comparing the photoinduced kinetics of film samples of (a) P3HT and (b) P3HT doped with F4TCNQ.

To test the hypothesis that optical dedoping results from dissociation of an excited state back to neutral species, we performed transient-absorption pump−probe spectroscopy using 400 nm pump pulses on films of P3HT:F4TCNQ and P3HT (Figure 2) and solutions of F4TCNQ and F4TCNQ- (Figures S5-S7). The P3HT sample (Figure 2a) shows a ground state bleach (GSB; negative signal) from 425 to 625 nm and a photoinduced absorption (PIA; positive signal) of the P3HT polaron below 650 nm.21 Both signals decay within 1 ns with the remaining signal corresponding to polarons in trap states (see Figure S3 and S4 for full kinetic analysis).21 By comparison, the P3HT:F4TCNQ mixed film (Figure 2b) also shows a somewhat blue-shifted P3HT GSB 8 ACS Paragon Plus Environment

Page 9 of 30

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 Letters

and P3HT polaron PIA features, but these features decay within 1 ps, most likely due to the high density of hole states in the sample.22 These features are blue shifted because (as previously reported) P3HT is more efficiently doped in crystalline domains,23 so the neutral P3HT features seen here are predominantly from amorphous P3HT domains.11 If 400 nm excitation resulted in direct charge back-transfer, we would expect to see a PIA signal at 390 nm and 520nm, corresponding to an increase in the concentration of neutral F4TCNQ and P3HT, respectively, but these signals are absent. Instead, what we see is an initial bleach of F4TCNQ- and P3HT at 390 nm and 510 nm, respectively, followed by a long-lived (>7 ns) product state showing an absorption at 450 nm and bleaches at both 390 nm and 500−800 nm. Constrained DFT simulations (Figure S2) show that neutral F4TCNQ is unstable in the presence of a P3MT octamer, indicating that even if a back charge transfer could occur, that the F4TCNQ would immediately re-dope.

9 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

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 10 of 30

Figure 3: Photoproduct formation: Steady-state UV/vis data of F4TCNQ samples exposed to light soaking. a) F4TCNQ and F4TCNQ- (reduced with KI) spectra before and after exposure to 405 nm light (approx. 150 mW/cm2) for 15 min. b) Difference spectra from samples in a) showing photoproduct formation in F4TCNQ samples. c) Spectra of F4TCNQ in acetonitrile with THF as a co-solvent. d) Spectra of samples from c) after exposure to 405 nm light, showing that the THF co-solvent has a significant effect on the probability of reaction.

10 ACS Paragon Plus Environment

Page 11 of 30

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 Letters

The observation of a long-lived product state in the transient absorption spectra raises the possibility that rather than electron back-transfer, a photochemical reaction may be responsible for optical dedoping. To test this theory, we first collected static absorption spectra of F4TCNQ and F4TCNQ- before and after illumination with 405 nm light to determine the signs of photoreaction (if any occurs). Figure 3a shows that F4TCNQ degrades upon light exposure, but that F4TCNQ- remains stable. The difference spectrum between the samples before and after illumination (Figure 3b) shows a photoproduct with absorbance maximum at 450 nm, the same as was seen in the P3HT/F4TCNQ transient absorbance experiment (Figure 2b). Figure S5 compares the transient spectra of F4TCNQ and F4TCNQ- in acetonitrile with the kinetics of the photoproduct formation. DISC patterning is typically conducted with the P3HT:F4TCNQ film immersed in THF, which is a good solvent for both neutral P3HT and F4TCNQ.9-10 Figure 3c shows the steady-state absorption spectra of F4TCNQ in CH3CN with a small amount of THF added (10:1 mole ratio THF:F4TCNQ). In the dark, the spectrum with the THF additive is identical to the spectrum of F4TCNQ in pure acetonitrile, indicating that no reaction occurs. However, after illumination with 405 nm light (Figure 3d), the THF additive solution shows complete conversion of F4TCNQ to photoproduct(s). This photoreaction is clearly much more rapid than the photoreaction in pure CH3CN, indicating that THF either directly reacts with F4TCNQ*, or catalyzes the conversion of F4TCNQ* to a photoproduct. Steady-state optical absorption spectra of F4TCNQ in pure THF before and after exposure to 405 nm light (~50 mW/cm2, 30 seconds) are shown in

11 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

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 12 of 30

Supplemental Figure S9. This product shows only a very weak absorption at 480 nm and slowly converts back to F4TCNQ on a timescale of hours.

Figure 4. NMR spectra of a 1:5 molar ratio F4TCNQ:THF solution in CD3CN exposed to 405 nm light (50 mW) for 15 minutes. A) 1H-13C HMBC, B) 1H, with integrals shown below each peak, and C)

13

C spectra. D) Structure assignment; bold text indicates 1H

shift, non-bold text indicates 13C shift.

To determine the structure of the F4TCNQ−THF reaction product, we collected NMR spectra of F4TCNQ and THF (1:5molar ratio) in CD3CN, before and after exposure to 405 nm light. Before exposure, the only signals visible are those expected for THF (1.81ppm and 3.65ppm 1H, 26.25ppm and 68.3ppm

13

C), residual water (2.14ppm 1H),

and the solvent (1.94ppm 1H, 1.3ppm and 118.3ppm 13C), as shown in Figures S10-S15. Carbon signals from F4TCNQ are not detectable due paramagnetic relaxation resulting from charge transfer with the solvent. To our knowledge, a

13

C NMR spectrum of

F4TCNQ has not been published.

12 ACS Paragon Plus Environment

Page 13 of 30

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 Letters

After collecting the initial non-light-exposed spectra, the sample was removed from the NMR spectrometer and exposed to a 50 mW 405 nm continuous laser through the bottom of the NMR tube for 15 minutes. Figure 4 shows 1H-13C heteronuclear multiple bond correlation (HMBC) (a), 1H (b), and 13C spectra (c), collected after 405 nm exposure. All three spectra show a number of new signals, which were not present prior to 405 nm exposure. Integration of the 1H signals in Figure 4b (relative areas shown below each peak) reveals that the new signals appearing after 405 nm exposure all have similar integration areas, indicating they likely originate from a single product. The number of new peaks indicates substitution of one of the THF alpha protons, as shown in Figure 4d, since the non-planarity of THF means that the protons are diasterotopic and therefore inequivalent after substitution. Each THF peak originates from one proton, so the integration areas indicate that the reaction is 1:1 (as shown in Figure 4d) and proceeds essentially to quantitative yield. The 1H-13C HMBC spectrum in Figure 4a confirms our structure assignment. The singlet peak at 5.87 ppm shows a 1JCH coupling to a carbon at 18.7 ppm, a long-range coupling to a fluorinated carbon (13C doublet at 145.9 ppm, identifiable by a characteristic 270 Hz 1JCF coupling), and a long-range coupling to another carbon at 110.9 ppm (a typical shift for a nitrile group). None of these carbons show couplings to other protonated carbons, indicating that the 5.87 ppm signal comes from a proton that is spatially separated from the remaining 7 protons on the THF ring. From these constraints we generated the structure shown in Figure 4d, which matches both the observed and calculated spectra (see Supplemental Information). This F4TCNQ−THF photoproduct

13 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

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 14 of 30

lacks the quinone structure that makes F4TCNQ a strong electron acceptor, and is not expected to dope P3HT. Although we have studied the reaction between F4TCNQ and THF in detail, similar photoreactions are likely to occur between nearly all neutral molecular dopants and solvent molecules. This is because, by design, dopant molecules have exceptionally deep LUMO levels. In their photo-excited state, these molecules contain two singly occupied orbitals corresponding to the ground state HOMO and LUMO, and should thereby be capable of accepting an electron near the energy of the ground state HOMO level. Therefore, in a p-type dopant, the acceptor strength of the molecule increases by approximately the optical bandgap, making these states reactive even towards fairly inert solvents. A similar argument can be made for n-type dopants. It should therefore be straightforward to engineer these types of reactions for a wide variety of dopants, allowing DISC to be applied to a wide range of polymers and dopants.

14 ACS Paragon Plus Environment

Page 15 of 30

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 Letters

Figure 5: UV/vis spectra. a) Absorption spectra of a dry P3HT:F4TCNQ film (black), the same film immersed in THF (red), and the THF solution in which the P3HT:F4TCNQ film was immersed (blue). b) Zoomed image of the dissolved-components spectrum (blue) and a fit to this spectrum (gray) composed of P3HT (dark yellow), F4TCNQ (green), and F4TCNQ- (magenta). c) Absorption spectra of the same film after exposure to a 50 mW 405 nm laser for 15 min. Since the film dissolves with light exposure, the P3HT in solution is observed. Inset: Sample geometry that allows simultaneous measurement of both film and solvent.

15 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

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 16 of 30

To monitor a film of P3HT:F4TCNQ under simulated photopatterning conditions, we immersed a film of P3HT:F4TCNQ in a cuvette of THF and made UV/VIS measurements both through the immersed film and orthogonally through only the THF solvent (see inset of Figure 5c for sample geometry). Figure 5a shows spectra through the film before and during immersion in THF, revealing a slight reduction in doping level upon solvent exposure, and a blue-shifted P3HT absorption resulting from solvent swelling of crystalline P3HT domains. Interestingly, although solvent exposure reduces the intensity of the P3HT+ polaron band (600−>1100 nm), the intensity of the F4TCNQbands at 850 and 760 nm do not significantly diminish. The corresponding spectrum of the solution (blue) confirms that most of the P3HT and F4TCNQ remains fixed in the film. Figure 5b shows a magnified image of the spectrum of the dissolved components in THF (same as the blue curve in Figure 5a) and a quantitative decomposition of the solvent spectrum composed of F4TCNQ, P3HT, and F4TCNQ-. In THF, a small amount of F4TCNQ- is present in solution, likely due to the slight solubility of P3HT+. By contrast, if the same measurement is done in CH3CN (Supplemental Figure S8), no P3HT dissolves, and no F4TCNQ- is observed in solution. Neutral F4TCNQ is present in both solvents, but in THF the concentration is quite low. This indicates that in the presence of a solvent-swollen P3HT film, the reaction rate of photoexcited F4TCNQ* and THF may be limited by the equilibrium of F4TCNQ- and F4TCNQ. Given the rapid de-doping and dissolution previously measured,10 we posit that the laser locally heats the swollen film and thereby shifts the equilibrium towards neutral F4TCNQ, thus enabling more of the total population to be photoexcited and to react with THF. This thermally activated

16 ACS Paragon Plus Environment

Page 17 of 30

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 Letters

mechanism is consistent with a recent report indicating that optical DISC patterning achieves sub-diffraction-limited resolution, resulting from a non-linear dependence of dissolution rate on light intensity.10 Figure 5c shows the solution spectrum of the same sample after exposure to a 50 mW 405 nm laser for 15 min. The observed spectra nearly match that of solvated P3HT, but is slightly red shifted and shows a peak at ~330 nm. Adding the photoproduct spectrum to the P3HT spectrum, we obtain a good fit for the observed spectrum, indicating that the only two species in solution are P3HT and the F4TCNQ−THF photoproduct. Supporting information figure S9 shows UV/vis spectra of F4TCNQ in THF, the F4TCNQ−THF photoproduct, and a series of spectra taken at 2 hour intervals after the photoconversion that shows recovery of the F4TCNQ. This recovery indicates that the F4TCNQ−THF is not thermodynamically stable and that the F4TCNQ could be recovered and recycled from the solution.

17 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

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 18 of 30

Scheme 1. The optical dedoping process. Scheme 1 is a pictorial depiction of the entire dedoping and dissolution process and allows for a summary of the results. When F4TCNQ is used to dope a P3HT film, there is an equilibrium between F4TCNQ- and F4TCNQ, but the majority of the dopant exists as F4TCNQ-. When the film is exposed to a solvent, the film remains insoluble but swells significantly. Because the polymer (and therefore the positive charges) are essentially insoluble, the F4TCNQ- counter-ions remain confined to the film, but the small amount of neutral F4TCNQ is able to dissolve. With exposure to 405 nm light, neutral F4TCNQ is excited to F4TCNQ*, which reacts with THF to form

18 ACS Paragon Plus Environment

Page 19 of 30

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 Letters

F4TCNQ−THF. This forward reaction is fast; F4TCNQ−THF is highly soluble and quickly dissolves. The removal of neutral F4TCNQ from the film shifts the equilibrium between F4TCNQ and F4TCNQ- towards the formation of F4TCNQ. In addition, heating by the laser further shifts the equilibrium towards neutral F4TCNQ.11 The F4TCNQ- is therefore removed from the film indirectly by reaction of the F4TCNQ to form F4TCNQ−THF complexes. Once the doping level of the polymer is sufficiently low, the polymer also dissolves into the solvent, thus efficiently photo-etching the film. Finally, the dissolved F4TCNQ−THF is unstable and will under room temperature conditions in the dark convert back to F4TCNQ and THF over hours to days. This reverse reaction will re-dope the unexposed areas of the film, preventing the reaction of F4TCNQ in the bulk solvent from slowly depleting the overall film doping level. This back conversion to neutral F4TCNQ also means that the dopant can be recycled and that the DISC process is “green chemistry”. In conclusion, we present the mechanism that explains optically induced dedoping and dissolution of P3HT/F4TCNQ. Contrary to expectation, this process is not a direct photo-induced back charge transfer reaction. Instead, the dedoping reaction is a two-step process whereby F4TCNQ is optically excited to F4TCNQ*, subsequently reacting with THF to form F4TCNQ−THF. The photoinduced reaction of trace quantities of neutral F4TCNQ in P3HT drives an equilibrium shift and subsequent dedoping of the P3HT until the polymer regains normal solubility. Since this reaction is specifically between photoexcited F4TCNQ and the solvent, the process is polymer independent. In addition, since nearly all neutral molecular dopants are expected to be reactive in their photoexcited states, it should be relatively simple to engineer photodedoping reactions in

19 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters

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 20 of 30

other dopants. Together, this suggests that high-resolution optical DISC patterning should be generalizable to a wide range of polymer:dopant systems.

Experimental P3HT (Plextronics, MW = 65k) was purchased from Sigma-Aldrich. F4TCNQ was purchased from TCI (98+%). P3HT solutions (10 mg/ml, CB) were heated to 60˚C and left to dissolve overnight. One-inch glass slides (Fisher Scientific) were cleaned by sequential sonication steps in acetone, 10% Mucasal:DI water, and DI water, dried with nitrogen, and UV-ozone treated for 30 minutes before use. All sample preparation was performed in a nitrogen glove- box (