Dynamics of Additive Migration to Form Cathodic Interlayers in

Subscriber access provided by University of Rochester | River Campus & Miner Libraries

Article

On the Dynamics of Additive Migration to Form Cathodic Interlayers in Organic Solar Cells Jane Vinokur, Stas Obuchovsky, Igal Deckman, Lishai Shoham, Tom Mates, Michael L. Chabinyc, and Gitti L. Frey ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b06793 • Publication Date (Web): 11 Aug 2017 Downloaded from http://pubs.acs.org on August 13, 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.

ACS Applied Materials & Interfaces 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 40

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

ACS Applied Materials & Interfaces

On the Dynamics of Additive Migration to Form Cathodic Interlayers in Organic Solar Cells Jane Vinokur,1 Stas Obuchovsky, 1 Igal Deckman, 1 Lishai Shoham, 1 Tom Mates,2 Michael L. Chabinyc, 2 and Gitti L. Frey 1* 1

Department of Materials Science and Engineering, Technion – Israel Institute of Technology, Haifa 3200003, Israel. 2

Materials Department, University of California, Santa Barbara, CA 93106-5050

*Corresponding author E-mail: [email protected]

ABSTRACT

Migration of additives to organic/metal interfaces can be used to self-generate interlayers in organic electronic devices. To generalize this approach for different additives, metals and organic electronic devices it is first necessary to study the dynamics of additive migration from the bulk to the top organic/metal interface. In this study we focus a known cathode interlayer material, polyethylene glycol (PEG), as additive in P3HT:PC71BM blends and study its migration to the blend/Al interface during metal deposition, and its effect on organic solar cell (OSC) performance. Using DSIMS depth profiles and XPS surface analysis we quantitatively correlate the initial concentration of PEG in the blend and sequence of thermal annealing/metal deposition

ACS Paragon Plus Environment

1


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 40

processes, with the organic/Al interfacial composition. We find that PEG is initially distributed within the film according to the kinetics of the spin coating process, i.e. the majority of PEG accumulates at the bottom substrate, while the minority resides in the film. During electrode evaporation, PEG molecules kinetically "trapped" near the film surface migrate to the organic/Al interface to reduce the interfacial energy. This diffusion-limited process is enhanced with the initial concertation of PEG in the solution and with thermal annealing after metal deposition. In contrast, annealing the film before metal deposition stalls PEG migration. This mechanism is supported by corresponding OSC devices showing that Voc increases with PEG concertation at the interface, up to a saturation value associated with the formation of a continuous PEG interlayer. Presence of a continues interlayer excludes the driving force for further migration of PEG to the interface. Reveling this mechanism provides practical insight for judicious selection of additives and processing conditions for interfacial engineering of spontaneously generated interlayers.

Keywords: OSC, additives, interlayer, annealing, PEG, PCBM, P3HT, XPS, DSIMS

INTRODUCTION Tremendous efforts are invested in controlling the bulk heterojunction (BHJ) morphology of the active layer in organic solar cells.1-2 Nevertheless, even the most optimized morphology would never result in an efficient cell without proper contacts. A Schottky barrier at the electrode-active layer interface, i.e. metal-organic semiconductor junction, increases the series resistance and reduces the built-in potential of the cell. These are directly translated to lower open circuit voltage (Voc) and lower fill-factor (FF).3-7 Furthermore, poor selectivity of carriers at the contact results in significant recombination losses that reduce the short circuit photocurrent


2

Page 3 of 40

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


density (Jsc).3 Because the device efficiency is directly proportional to FF, Voc and Jsc, reduction of these parameters deteriorates the device performance. Consequently, highly charge selective ohmic contacts are essential for efficient OSC devices. Ohmic contacts require adjusting the electronic band alignment at the metal-semiconductor junction to allow efficient charge transfer across the interface.8-10 Namely, the cathode should attain a work function suitable for electron injection/extraction to/from the LUMO level of the acceptor, and the anode work function should allow hole injection/extraction to/from the HOMO level of the donor.11 Therefore, cathodes are usually metals with low work functions, inherently imposing poor environmental stability due to their low reduction potential.12 Cathodes with higher work functions, although more stable, impose barriers for charge injection and poorselectivity for electrons, which limit device performance.11 Therefore, interlayers introduced between the active layer and the electrode can energetically tune the interface to overcome the two major drawbacks of OSCs: performance and stability. For example, interlayers which exhibit an overall dipole moment induce a strong local electric field at the organic/metal interface that shifts the effective work function of the metal, tuning the energetics at the interface.13-14 Alternatively, charge selective interlayers are used to block one type of charge carrier reducing charge recombination.15-17 Interlayers are generally deposited in a discrete processing step, such as thermal evaporation in vacuum or spin coating from an orthogonal solvent. These methods are suitable for laboratory research, nevertheless, technologically demanding with large area roll-to-roll (R2R) processing.18 Recently, spontaneous segregation of additives from the BHJ to the bottom/top of the active layer was harnessed to self-direct desired interlayers.19 This methodology reduces the number of processing steps, allows the formation of interlayers at buried interfaces that are not directly


3


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 40

accessible during processing, and is specifically beneficial for the growing printed OSC industry.20-21 For example, additives with high surface energy spontaneously segregate to the ITO or PEDOT:PSS underlying substrate during film processing. Similarly, additives with low surface energy migrate to the BHJ surface during processing. Alternatively, we recently established that additives with affinity to the electrode metal, for example O-H to Al or S-H to Ag, spontaneously migrate to the BHJ/metal interface during the metal deposition. In addition to improving the performance of the devices, these interlayers also improved their stability and lifetime.22-23 The simplicity and versatility of this methodology, with respect to types of additives, BHJ systems and metal contacts, make it promising for utilization in fabrication of printed, large area applications. However, the utilization of this technique requires understanding the kinetic and thermodynamic considerations that direct and limit additive migration to the deposited metal. Our working hypothesis is, therefore, to select materials that will thermodynamically form the technically desirable BHJ:additive phase separation, and apply processing and annealing protocol treatments that will drive the system to the thermodynamic structure. To gain control over the vertical phase separation of an additive, we investigate the effect of additive migration dynamics to the organic/metal interface on the interfacial composition and device performance. We demonstrate that we can enhance the migration process and facilitate the formation of a continues interlayer at the top contact, or alternatively, enhance segregation to the substrate to form the interlayer at the bottom contact. The selected system for this study is the OSC working-horse donor:acceptor blend: Poly(3-hexylthiophene-2,5-diyl)(P3HT):Phenyl-C71butyric acid methyl ester (PC71BM). The associated device performances are highly reproducible and fairly predictable, which makes this system suitable for our study. We used polyethylene


4

Page 5 of 40

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


glycol (PEG) as an additive because it was shown to induce a dipole moment at the cathode/organic semiconductor junction that reduces the energy barrier for charge injection, effectively enhancing Voc. Furthermore, PEG strongly interacts with Al and hence should segregate to the organic/Al interface during Al deposition.22, 24-26 By varying PEG concentration in the P3HT:PC71BM:PEG blend and applying thermal treatments before or after Al cathode evaporation we tailor the organic/Al interfacial composition. The metal/organic interfacial composition is determined by DSIMS and XPS techniques, and correlated with PEG concentration in the blend and the device parameters. By doing so, we gain practical insight on additive migration dynamics which we translate to directive tools for practical implementation for self-generated interlayers in OSC and other organic electronic devices.

EXPERIMENTAL

MATERIALS P3HT (Sepiolid P100, regioregularity >95%) was purchased from Rieke Metals and used as received. PC71BM was purchased from Solenne and used as received. PEDOT:PSS was acquired from Haraeus (Clevios PVP AL 4083) and filtered through a 0.45 µm poly(tetrafluoroethylene) (PTFE) filter before use. PEG of Mn=200 was purchased from Fluka and used as received. Deuterated Poly(ethylene glycol) Dihydroxy Terminated (D-PEG) of Mn=350 was acquired from Polymer Source (custom synthesis).

FILM DEPOSITION AND DEVICE FABRICATION


5


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 40

ITO-covered glass substrates were cleaned by sonication in acetone, methanol and 2-propanol, followed by 15 min of UV-ozone treatment. PEDOT:PSS was spin coated at 3600 rpm onto ITO/glass and dried at 200 °C for 5 min in ambient conditions. 1,2-Dichlorobenzene solutions of P3HT (60 mg/ml), PC71BM (60 mg/ml), and PEG (60 mg/ml) were prepared and mixed to obtain solutions of P3HT:PC71BM:PEG 20:20:0 mg/ml (reference); 20:20:0.0625 mg/ml; 20:20:0.125 mg/ml; 20:20:0.25 mg/ml; 20:20:0.5 mg/ml; 20:20:1 mg/ml; 20:20:3 mg/ml and 20:20:6 mg/ml. P3HT:PC71BM:PEG blend films were deposited by spin coating at 1100 rpm for 20 sec. onto ITO/PEDOT:PSS substrates (or onto ITO substrates for inverted cells), and slowly dried in a petri dish. Thermal evaporation of a top Al layer of ~120nm (or ~10nm of MoO3 and 120nm of Al layer for inverted cells) was conducted through a shadow mask at a system pressure of ~10-6 Torr. Device area was 3 mm2. For the XPS and DSIMS measurements the thicknesses of the Al layer was ~3nm and ~30nm, respectively. Thermal annealing of direct devices was performed using a hot plate at 110°C for 30 min under N2 atmosphere. The inverted devices where thermally annealed at 115°C for 1 hour before MoO3/Al deposition.

CHARACTERIZATION X-ray Photoelectron Spectroscopy (XPS) was performed in a Thermo VG Scientific Sigma Probe fitted with a monochromatic Al Kα (1486.6 eV) source. A 100W X-ray beam of 400 µm in diameter was used for high energy resolution scans of the C1s spectra with pass energy of 30eV. Line-shape analysis was done using the XPSPEAK4.1 software after a Shirley-type background subtraction. The binding energy scale calibration of the C1s spectra was done by referencing the C-C/C-H bond signal to 285eV. For all samples, the C1s spectra were measured in the standard and bulk-sensitive modes, i.e. with the angle between the direction of the analyzer


6

Page 7 of 40

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


and the specimen normal in the range of 53o±30o or 30.5o±7.5o, respectively. The C1s spectra of the metal-covered blend films were line-fitted to three main peaks: 285 eV (C-C/C-H), 286.6 eV (C-O) and 288.6 eV (O-C=O), and the corresponding fitting parameters are summarized in the Supplementary Information section (Table S1). Based on inelastic mean free path estimations of the metal/blend system, the information depth of C1s electrons includes a ~5nm thick organic film beneath the Al layer.18 For quantitative XPS analysis of the surface composition, we compared only samples with identical metal coverage, i.e. samples that were together in the evaporation chamber. For such a set of samples we normalized the intensity of the C1s spectra to the area under the peak of the metal (Al 2p). The area under the metal peak represents the thickness of the metal capping layer and is constant for each set of samples. Dynamic Secondary Ion Mass Spectroscopy (DSIMS) was performed with a Cameca IMS 7fAuto system using an oxygen (O2+) primary ion beam at an impact energy of 2kV. Typical background pressure in the system was 1x10-9 mbar. The crater size was 175 µm, and secondary ions were monitored from a 63 µm-diameter circle in the center, to avoid crater-edge effects. Various secondary ion species were monitored, and data were recorded at intervals of approximately 1 nm depth for each element. Depth scales were quantified using a contact profilometer. Device characterization was performed in inert atmosphere under 100 mW/cm2 AM1.5G class A sun simulator (Science Tech Inc. ss150 solar simulator) with a Keithley 2400 source meter. The reported J-V curves and performance values are average values over at least 16 similar devices.

RESULTS AND DISCUSSION


7


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 40

The segregation of species in a blend during metal deposition depends on its initial distribution in the deposited film, attraction to the metal as well as to the underlying substrate, and its diffusion coefficient in the organic matrix.19, 27-28 Additive distribution in the spun film, in turn, is governed by its surface energy ( γ ), relative solubility compared other blend components, the centrifugal forces during spin coating, and the solvent evaporation rate.29-31 Therefore, even in cases where additive phase separation would actually reduce the overall free energy of the system, low diffusivity at the room temperature could result in a metastable kinetic arrangement, as illustrated schematically in Figure 1a. Under such conditions, the device performance will reflect the kinetic, and not thermodynamic, structure.32 Driving the system towards the thermodynamically-favored diffusion process requires energy to overcome the diffusion barrier. This energy can be supplied by intentional thermal treatments or unintentional processes during device preparation, storage and/or operation. Therefore, to follow the dynamics of additive separation towards the energetically favorable arrangement it is necessary to characterize the system at each step of preparation, i.e. after the active layer deposition, electrode deposition, and post-operation. During the active layer deposition, thermodynamic considerations dictate that the lowest surface energy component will enrich the liquid/air interface to reduce the overall free energy of the system. In contrast, the highest surface energy component will migrate away from the surface, and phase separate from other components in the blend.33 When the substrates’ surface energy is high, if the film is thin enough and the additives’ surface energy is significantly higher than those of coexisting phases, the additive will accumulate on the substrate to form a distinct layered phase.31 This surface energy-directed phase separation occurs in the liquid blend solution instantaneously when applied onto a surface.33 However, rapid volatile solvent evaporation


8

Page 9 of 40

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


during spin coating leads to non-equilibrium morphologies driven by kinetic factors such as relative solubility of the components and solvent evaporation rate.19 Therefore, anticipating the thermodynamically-directed phase separation requires knowledge of not only the surface energies, but also the relative adhesion of the components to each other and the substrate. The surface energy of PEG oligomers (43 mJ/m2),34 is higher than that of P3HT (27 mJ/m2) or PC71BM (38 mJ/m2), and therefore PEG is not expected to enrich the films surface.35-36 Hence, the distribution of PEG inside the active layer is governed by the complex interplay between the kinetics of film deposition, the thermodynamically induced phase separation from the other blend components, and the attraction to the underlying substrate. To analyze the composition of the surface of the P3HT:PC71BM:PEG film we performed contact angle measurements of water on the blend and the separate blend components. The optical images and extracted contact angles of water drops on the blend and blend components are presented in Figure 1b-1f. The water/P3HT contact angle is ~110°, while water/PC71BM contact angle is ~81°. Thus, PC71BM clearly exhibits a higher surface energy than P3HT, in agreement with the literature values mentioned above. Water/P3HT:PC71BM and water/P3HT:PC71BM:PEG contact angles are both nearly similar to that of water/P3HT, suggesting that the surface consists predominately of P3HT. The surface enrichment of P3HT in P3HT:PC71BM blends is also known and documented.28, 37


9


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 40

Figure 1. (a) Schematic illustration of the thermodynamic/kinetic phase separation of an additive with high surface energy inside a thin film organic blend, and optical images and contact angles measurements of water drops on: (b) P3HT, (c) P3HT:PC71BM, (d) PC71BM, (e) P3HT:PEG, (f) P3HT:PC71BM:PEG. To identify the attraction of PEG to the PEDOT:PSS substrate relatively to its attraction to the other components, i.e. P3HT and PC71BM, we also measured contact angles between a liquid drop of the PEG additive and solid films of PEDOT:PSS, P3HT and PC71BM. The comparison of PEG/P3HT and PEG/PC71BM contact angles with PEG/PEDOT:PSS contact angle, Figure 2, reveals that PEG is attracted to the PEDOT:PSS substrate more than to the other film components, and hence PEG segregation toward the PEDOT:PSS substrate is energetically preferred.31 Therefore, the surface energy considerations dictate that when a P3HT:PC71BM:PEG


10

Page 11 of 40

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


blend is spun on a PEDOT:PSS substrate, the PEG phase separation will most likely result in a distinct PEG layer in the vicinity of the bottom PEDOT:PSS substrate rather than PEG domains inside the BHJ.

Figure 2. Optical images and contact angles measurements of PEG drops on: (a) PC71BM, (b) P3HT, and (c) PEDOT:PSS films. After the active layer is deposited by spin coating, the top electrode is thermally evaporated onto the active layer. When an Al electrode is evaporated onto the spun P3HT:PC71BM:PEG film, the dangling bonds of the metal are very high in energy and can affect the blends metastable structure. Namely, the potential of the PEG-Al chemical interaction can induce a driving force for the migration of PEG toward the Al interface. This migration can generate a cathodic interlayer at the organic/Al interface, as recently shown by us.19, 22, 24, 27, 32 However, the Al-PEG attraction decays rapidly with depth, and hence is expected to be sufficient only for attracting PEG molecules that were “kinetically trapped” in vicinity of the surface during the film formation. Therefore, we expect that thermal treatments will significantly affect the interlayer formation. To study the effect of thermal treatments on PEG segregation during Al deposition, we analyzed the chemical composition of the organic/Al interface of P3HT:PC71BM:PEG blends that were either not annealed, or annealed after the Al deposition, by X-ray Photoelectron Spectroscopy (XPS). The samples for XPS were covered by thermally evaporated ~3nm Al stripes to allow the collection of XPS signal from ~5nm depth underneath the Al. PEGs XPS


11


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 40

finger-print is a C-O peak at 286.6 eV in the C1s spectrum. The XPS spectra of films that were not-annealed (black lines) or annealed after Al deposition (red lines) are presented in Figure 3 a, b and c for P3HT:PC71BM:PEG blends that contain 0, 2 or 5 mg/ml PEG, respectively. The spectra of the reference P3HT:PCB71M samples that were either annealed or not annealed, Figure 3a, are identical. Namely, the C-O characteristic peak is not affected by the thermal annealing, suggesting that thermal annealing does not induce C-O-Al complexation. Therefore, the C-O/CC peak area ratios are directly associated with the amount of PEG at the Al/organic interface. The C-O/C-C ratios are summarized in Table 1. For the P3HT:PC71BM blends with no PEG (Figure 3a) the C-O/C-C ratio is very low, ~16% regardless of thermal treatment. This C-O peak is associated with carbon-based contamination on the Al contact, and possibly a small contribution from the terminal methoxy of PCBM. Previously, similar C-O contamination levels were found for all samples that were thermally annealed before metal deposition regardless of PEG content, indicating that thermal annealing before cathode deposition depresses PEG migration to the organic/metal interface.32 In not-annealed films (black curves in Figure 3), the C-O/C-C peak ratio depends on PEG concentration in the blend, and approaches 25% and 28% for blends containing 2 mg/ml and 5 mg/ml respectively. In films annealed after Al deposition (red lines in Figure 3) the C-O/C-C ratio approaches saturation of ~34% for both PEG concentrations.


12

Page 13 of 40

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


Figure 3. The XPS spectra of Al-covered P3HT:PC71BM:PEG films that were not-annealed (black lines) or annealed after Al deposition (red lines). (a) 0 mg/ml PEG, (b) 2 mg/ml PEG, and (c) 5 mg/ml PEG.


13


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 40

Table 1. The C-O/C-C XPS peak area ratios calculated from Figure 3. P3HT:PC71BM:PEG films

0 mg/ml PEG

2 mg/ml PEG

5 mg/ml PEG

Not annealed

16%

25%

28%

Annealed after Al deposition

16%

34%

34%

The XPS spectra provide quantitative information of PEG content at the organic/Al interface which is directly related to PEG’s migration to the interface. For films that were not annealed, interfacial PEG content depends on PEG concentration in the film, while annealed films show significantly higher PEG content, which is constant for both PEG concentrations. Namely, PEG migration does require substantial thermal activation energy. We suggest that PEG molecules that are kinetically “trapped” near the organic/metal interface interact with the depositing Al. The potential of this interaction decays rapidly with distance from the electrode. Therefore, only PEG molecules that are within a region close to the interface are attracted to the Al. The concentration of PEG molecules trapped within the characteristic diffusion length depends on PEG’s initial concentration in the blend. The combination of the rapid potential decay and short characteristic diffusion length at room temperature, limit the depth for PEG migration. Therefore, in notannealed films PEG interlayer formation is actually diffusion dependent. In films that were annealed after the metal deposition there were two stages of PEG migration to the interface. The first, during metal deposition, similar to the not annealed films. The second during the thermal annealing after metal deposition. The annealing provides activation energy for PEG migration from increased depth and facilitates full coverage of the metal by the additive to form a fully continues interlayer. Hence, the interfacial PEG content after metal deposition and thermal


14

Page 15 of 40

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


annealing is constant regardless of PEG concertation in the blend reflecting a continuous PEG interlayer. To study the effect of additive migration dynamics to the metal/organic interface on device performance, we tested devices with active layers that were either not annealed, thermally annealed before-, or after- Al deposition, as schematically shown in Figure 4 a, b and c, respectively. The devices were prepared by spin coating 1:1 blends of P3HT:PC71BM with different concentrations of PEG onto ITO/PEDOT:PSS substrates, topped with a thermally evaporated Al cathode. The dark J–V characteristics of devices with varying concentrations of PEG either not-annealed or annealed before or after Al deposition are presented in Figure 5 a, b and c, respectively. The dark characteristics of the devices that were not annealed (Figure 5a) reveal that turn-on voltage increases with increasing PEG concentration. The turn-on voltage is closely related to the built-in potential, and therefore we can infer that adding PEG to the blend increases the built-potential in the device. Notably, Figure 5a clearly shows that the device with the highest PEG concentration has a unique current behavior (brown line in Figure 5a). While all other devices show an exponential current at low voltages, high PEG concertation results in an ohmic current at low voltages. In ideal devices, there are few carriers injected from the contacts to the BHJ at low voltages (below turn on voltage), so currents, if present, are associated with ohmic leakage.3 Above turn-on voltage, carriers are injected from the contacts to the BHJ leading to a predominantly exponential rise in current with increasing voltage. The dark currents in Figure 5a, except for the device with high PEG concentration, are exponential at low voltages indicating parasitic charge injection before the diode opens. In contrast, the device with high PEG concertation shows a more ideal behavior with ohmic and exponential currents below and


15


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 40

above turn on voltage, respectively. Namely, introducing 6 mg/ml PEG effectively reduces the parasitic injection into the film under forward bias.

Figure 4. Schematic illustration of the device preparation and thermal annealing/cathode evaporation sequence performed in this study. Devices with (a) non-annealed films; (b) films thermally annealed before Al deposition; and (c) films thermally annealed after Al deposition. The reduction of parasitic injection by PEG interlayer is realized by comparing the dark currents of devices that were annealed either before or after the Al deposition. Parasitic injection dominates the dark currents of all devices that were annealed prior to metal deposition, regardless of PEG presence or concertation (Figure 5b). In contrast, thermally annealing the samples after metal deposition improvers the ability of PEG to suppress parasitic injection, as clearly seen in Figure 5c. Annealing before cathode deposition suppresses the formation of the PEG interlayer and hence all devices in Figure 5b exhibit significant leakage and poor current rectification implying poor charge selectivity at the contacts.11 In contrast, annealing after Al deposition enhances the formation of the PEG interlayer. As a result, the leakage currents in Figure 5c are reduced by nearly an order of magnitude in comparison to reference P3HT:PC71BM devices. Therefore, based on the dark J-V measurements we conclude that the generated PEG interlayer has three significant contributions: (1) increased built-in potential, (2) reduced parasitic injection currents, and (3) reduced leakage currents.


16

Page 17 of 40

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


Figure 5. J-V curves in the dark (a), (b), (c) and under illumination (d), (e), (f) of P3HT:PC71BM:PEG devices with 0, 0.0625, 0.125, 0.25, 0.5, 1, 3, 6 mg/ml PEG concentration. For each composition the films are either not annealed (a) and (d); annealed at 110°C for 30 min before (b) and (e), or after (c) and (f), Al deposition. The J-V curves under illumination of the same devices clearly show that the sequence of processing steps, i.e. annealing before or after the cathode deposition, significantly affects the performance of the devices, as shown in Figure 5d-f and Table 2. We first examine the P3HT:PC71BM:PEG films that were not-annealed, Figure 5d and Table 2. The Voc, Jsc and FF gradually increase with PEG content, from ~0.34V, ~7.7mA and 47% (no PEG) to ~0.61V, ~9mA and 64% for devices with highest PEG concentration. The device parameters enhancement results in a gradual efficiency improvement with PEG concentration from 1.22% to


17


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 40

3.49%. This behavior is in a good agreement with the XPS results (Figure 3 and Table 1), confirming that PEG segregates to the organic/metal interface during the metal deposition to form an interlayer that effectively enhances Voc. The slight Jsc enhancement might be associated with combined effects of higher built-in potential within the cell due to PEG interlayer formation, and a bumpy surface morphology induced by PEG near the substrate.27,

38

The

enhancement of the FF, in turn, is mostly associated with Voc enhancement but also includes an increase of Rsh and decrease of Rs induced by the PEG interlayer (Supplementary Information Figures S1 and S2). Importantly, the correlation of these values with PEG’s initial concentration indicates that when films are not annealed, the amount of PEG at the Al/organic interface depends on PEG’s concentration in the blend. In contrast, when the films were thermally annealed before the Al deposition, Figures 5e and Table 2, we find that for low PEG concentrations the Voc, Jsc and FF values are independent of PEG concentration and comparable to that obtained for the P3HT:PC71BM devices with no PEG. On the other hand, for high PEG concentrations, the Voc and FF decay with PEG content. Interestingly, at high PEG concentrations the device suffers substantial leakage, effectively losing diode behavior and functionality almost completely. Again, these results are in a good agreement with our previous studies, confirming that annealing before Al deposition suppresses PEG migration toward the metal essentially preventing the formation of a PEG interlayer. Notably, the significant decrease in Voc with high PEG concentrations could suggest that PEG aligns at the underlying ITO/PEDOT:PSS substrate forming a dipole layer at the organic/anode interface which reduces the effective work function of the anode The reduced effective anode work function results in injection barrier for holes at the anode, which is also evident from the increased Rs (see Table 2 and Supplementary Information Figure S2). Indeed, Fong-Yi Cao et al.


18

Page 19 of 40

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


utilized PEG as cathodic interlayer at the ITO/organic interface in inverted cells.39 We will demonstrate spontaneous PEG interlayers formation in inverted cells and address this hypothesis further in the text. Finally, when thermal annealing is applied after Al deposition, Figures 5f and Table 2, Voc and FF initially increase with PEG content, up to 0.5mg/ml, and then saturate at ~0.6V and ~60%. Interestingly, the Rs follows the same trend, i.e. decreases with PEG concentration and saturates at 0.5 mg/ml PEG (Supplementary Information Figure S2). This suggests that reduction in Rs is mostly associated with Shottky barrier reduction, i.e. decrease in contact resistance. The Jsc values are all similar and higher than that of the reference device. Assuming the extent of migration is not diffusion limited (a legitimate assumption for diffusion distances under 200nm in soft films), the saturation at ~0.5 mg/ml concentration might be associated with the amount of PEG required to create a fully continues layer of PEG at the Al interface. We hypothesize that thermal annealing after Al deposition further enhances PEG diffusion to the organic/Al interface and hence a continuous interlayer of PEG is obtained at lower concentrations (>0.5mg/ml) in contrast to films that were not annealed, where a continuous interlayer is obtained only for PEG concentrations above 6 mg/ml.


19


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 40

Table 2. Device characteristics of P3HT:PC71BM:PEG films. Each parameter represents an average of 16 similar devices. PEG Conc. [mg/ml]

0

Treatment

Jsc [mA/cm²]

Voc [V]

FF

Efficiency [%]

Rsh [ Ω ]

Rs [ Ω ]

not-annealed

7.74±0.8

0.336±0.009

0.47±0.01

1.22±0.2

18018±3200

496±35

annealed before Al

7.73±0.9

0.348±0.007

0.465±0.012

1.25±0.16

4307±1250

498±50

annealed after Al

7.58±0.8

0.31±0.03

0.44±0.02

1.03±0.18

10582±1430

520±36

not-annealed

7.76±0.9

0.337±0.008

0.47±0.015

1.22±0.16

14065±3560

514±48

annealed after Al

8.19±0.9

0.447±0.01

0.52±0.02

1.9±0.2

28490±1425

481±55

not-annealed

8±1

0.342±0.009

0.48±0.01

1.23±0.15

15015±3120

478±43

annealed before Al

7.3±0.8

0.333±0.01

0.482±0.007

1.17±0.13

19724±2245

454±45

annealed after Al

8.65±1

0.512±0.01

0.56±0.015

2.47±0.28

31153±3240

414±40

not-annealed

8.23±0.8

0.35±0.004

0.47±0.015

1.36±0.13

17921±2310

469±38

annealed before Al

7.9±1.2

0.37±0.01

0.409±0.025

1.19±0.23

17180±1780

607±95

annealed after Al

8.23±0.7

0.55±0.007

0.57±0.014

2.6±0.2

29762±2510

399±25

not-annealed

8.63±1

0.374±0.008

0.48±0.01

1.57±0.4

18622±1925

467±55

annealed before Al

8±0.85

0.378±0.01

0.458±0.016

1.38±0.16

14684±2260

556±45

annealed after Al

8.5±1.2

0.601±0.02

0.59±0.014

3.03±0.28

21645±3015

324±47

not-annealed

8.5±1

0.471±0.007

0.53±0.01

2.12±0.23

33033±3240

490±60

annealed before Al

7.83±1.1

0.375±0.01

0.472±0.01

1.39±0.17

19048±3560

518±65

annealed after Al

8.14±0.9

0.615±0.005

0.6±0.01

3±0.34

18257±2450

320±30

not-annealed

7.67±0.9

0.559±0.012

0.59±0.013

2.53±0.36

35804±3070

404±43

annealed before Al

7.9±0.85

0.294±0.014

0.432±0.023

1±0.17

8251±2580

474±63

annealed after Al

7.8±0.9

0.609±0.008

0.6±0.013

2.86±0.38

22310±1760

313±35

not-annealed

8.94±1.2

0.614±0.006

0.64±0.04

3.49±0.46

19724±2750

267±35

annealed before Al

7.5±1.2

0.154±0.09

0.317±0.06

0.45±0.38

572±1540

390±90

annealed after Al

8.8±1.4

0.623±0.004

0.6±0.03

3.28±0.43

15504±3200

338±30

0.0625

0.125

0.25

0.5

1

3

6


20

Page 21 of 40

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


To study the dynamics of PEG segregation during Al deposition and thermal treatments, we performed Dynamic Secondary Ion Mass Spectroscopy (DSIMS) depth profiling of PEGcontaining films, as a function of annealing condition before, or after, Al deposition. This technique provided a quantitative measure of the PEG concertation in the bulk and at the organic/Al interface with the processing sequence and device parameters. For these experiments, we used backbone-deuterated PEG (D-PEG) and determined the depth profile using oxygen ion etching. The normalized intensity of deuterium in the depth profile is directly proportional to the amount of PEG in that depth. Films were spun on silicon substrates, followed by thermal evaporation of ~30 nm Al on half of the sample area. DSIMS profiles were measured on the bare and Al covered areas, Figure 6a and 6b, respectively. The samples were either not-annealed (black lines), annealed before (blue lines) or after (red lines) the Al deposition. The selected P3HT:PC71BM:D-PEG blend composition was 20:20:2.5mg/ml, above the amount of PEG required to obtain a continuous PEG interlayer by annealing after Al deposition (as realized from Figures 3b and Table 1).


21


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 22 of 40

Figure 6. DSIMS depth profiles of (a) Si/P3HT:PC71BM:D-PEG, and (b) Si/P3HT:PC71BM:DPEG/Al samples, either not-annealed (black lines); annealed before (blue lines) or after (red lines) Al deposition. The insets show the areas measured on the samples: (a) bare area, and (b) Al contact area. The inset table in (b) shows the D-PEG area fraction calculated at the organic/Al interface in each sample (c) Schematic illustration of PEG distribution within the tested layers. The distribution of PEG within the not-annealed layers and layers annealed before or after Al deposition (on top of the left side of each film) is schematically illustrated in Figure 6c. The normalized DSIMS depth profiles under the bare side of all films (Figure 6a), regardless of thermal treatments, show accumulation of D-PEG mainly at the bottom silicon substrate. The doublet feature of the D-PEG peak is associated with a PC71BM-rich composition near the substrate (see Supplementary Information Figure S3). Thermal annealing before Al deposition further promotes segregation of D-PEG to the silicon substrate as evident from an increase of the peak area (blue line vs black line in Figure 6a). Thermal annealing after deposition of the adjacent Al layer, however, increases the distribution of D-PEG in the bulk under the bare surface (Figure 6a, red line). This indicates that even PEG molecules that are not directly under the contact are attracted to Al during the electrode deposition. This long-distance interaction is somewhat surprising and warrants further investigation. In contrast to the bare films, the DSIMS


22

Page 23 of 40

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


depth profiles measured under the Al electrodes show presence of D-PEG at the organic/Al interface (Figure 6b). These profiles corroborate unambiguously that Al deposition induces segregation of D-PEG to the organic/metal interface. Furthermore, the intensity of the DSIMS peak at the interface clearly indicates that thermal annealing before Al deposition slows D-PEG migration toward Al, while thermal annealing after Al deposition enhances it. Because all films were prepared from the same solution and under the same conditions, the concentration of D-PEG in all films is identical. Therefore, we can use the DSIMS profiles to quantitatively estimate the amount of PEG at the Al/organic interface. To do so we integrate the deuterium signal peak at the Al/organic interface (up to ~30 nm from the interface), and relate it to the integral of the full depth profile, for each thermal annealing condition. For the sample annealed after Al deposition (Figure 6b red line), the calculated fraction of D-PEG at the interface is ~22%. Namely, 78% of the D-PEG molecules remain close to the bottom silicon substrate, while 22%, corresponding to 0.55 mg/ml (of the initial 2.5 mg/ml concentration), segregated to the organic/Al interface. We suggest that this is the value required to form a continuous D-PEG interlayer. Similarly, the calculated fraction of D-PEG at the Al interface in the film that was not-annealed (Figure 6b black line) is ~7%, which is translated to 0.175 mg/ml of D-PEG. In this case, there was no thermal annealing and hence the segregation of D-PEG molecules to Al is limited by diffusion and the interfacial D-PEG concentration is below saturation. Figure 7 correlates the initial concertation of PEG in the solution with its segregation to the interface during Al deposition/annealing and the device performance. Because device performance in our study, i.e. FF and

PCE, are correlated with Voc (and not Jsc, see

Supplementary Information S1) we will confine our further discussion to the effect of PEG


23


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 24 of 40

interlayer formation on Voc. For devices that were not annealed, the Voc values (Table 2) and PEG interfacial fraction (Table 1) are marked as black circles and black asterisk, respectively. For devices that were annealed after the Al deposition, the Voc values (Table 2) and PEG interfacial fraction (Table 1) are marked as red circles and red asterisk, respectively. The black circles and black asterisks show a similar trend indicating a steady increase in both Voc and PEG interfacial concentration with the concentration of PEG in the processing solution. Importantly, this result demonstrates that the Voc is directly correlated with the amount of PEG at the interface. Furthermore, this trend also confirms that in films that were not annealed the kinetically “trapped” PEG molecules are attracted to the depositing Al, and the interlayer formation is diffusion dependent. A completely different trend is observed for devices that were annealed after Al deposition (red circles and asterisk). Under such conditions, the Voc and PEG interfacial fraction initially increase with the concentration of PEG in the processing solution, but then reach a saturation level. We suggest that saturation is reached when the PEG interlayer is continuous and hence the driving force for segregation is terminated. Under such conditions, the obtained Voc is associated with a complete interlayer and hence is the maximum value possible for this system. Therefore, the self-limiting character of the metal-induced segregation process ensures that optimal conditions for interlayers: full coverage yet minimal thickness. For devices that are annealed prior to the Al, PEG is nearly absent from the interface (Figure 6b, blue line), while the Voc decreases significantly with increasing PEG concentration (blue circles in Figure 7). The reduction of Voc suggests that during annealing PEG molecules not only segregate closer to the bottom substrate, as realized by DSIMS depth profiles (Figure 6), but probably adsorb to the


24

Page 25 of 40

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


ITO/PEDOT:PSS electrode to form an aligned dipole layer which reduces the effective work function of the anode.39-41 Introducing the DSIMS results to Figure 7 allows further quantitative correlation of Voc with PEG content at the interface. From the DSIMS results above we calculated the concentration required to form a continuous D-PEG interlayer: 0.55 mg/ml (see discussion above of Figure 6b). Marking this value (star) on the red line of Figure 7, which represents full segregation due to annealing after Al deposition, confirms that this is the concentration required to achieve a complete interlayer and hence the Voc saturation value. Furthermore, Figure 7 also shows that Voc of device with 2.5 mg/ml that was not-annealed (intersection of dashed and black line) is ~0.53V, is very similar to that expected for a blend with 0.175mg/ml that is annealed after the Al deposition (intersection of dotted and red line). Namely, for a processing solution with 2.5 mg/ml, 0.175 mg/ml segregates to the organic/metal interface during Al deposition to form discontinuous interlayer; consequent annealing induces further segregation to form a complete continuous interlayer, 0.55 mg/ml. Access PEG, ~ 2 mg/ml, are suspended mainly close to the bottom Si substrate (Figure 6b). To summarize the understanding revealed in Figure 7: Voc is directly correlated with the concentration of D-PEG at the organic/Al interface, i.e. the Voc increases with interlayer coverage and saturates at full coverage. Moreover, these results provide quantitative insight towards the design of interlayers in organic electronic devices. For example, equivalent unsaturated PEG interlayers can be achieved following 2 different paths: introducing 2.5 mg/ml without annealing the film, or, introducing 0.175 mg/ml and annealing the film after the Al deposition to ensure full segregation.


25


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 26 of 40

Figure 7. Voc (circles) and PEG interfacial fraction (asterisk) as a function of PEG concentration in the processing solution, for not-annealed devices (black), devices annealed prior to Al deposition (blue), and devices annealed after Al deposition (red). The PEG interfacial fraction is extracted from XPS measurements (Figure 3 and Table 1). The dashed line represents PEG concentrations used for the D-SIMS measurements: 2.5 mg/ml PEG. The star highlights the Voc saturation point at ~0.55 mg/ml PEG. The dotted lines guide the eye following analysis in the text.

To complete our study on the dynamics of PEG migration in BHJs, we investigated the reversibility of the process. Namely, we applied a second annealing treatment (110°C for 30 min), after Al deposition, on devices that were annealed prior to Al deposition. The second annealing process is intended to encourage migration PEG to the organic/Al interface. This migration was initially suppressed by the first annealing process that was performed before the Al deposition. The effect of the additional thermal treatment on the device performance is presented in Figure 8a and b for P3HT:PC71BM (20:20 mg/ml) with 0 or 6 mg/ml PEG, respectively. The device with no PEG shows reasonable performance when annealed prior to Al


26

Page 27 of 40

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


deposition (blue line in Figure 8a), and the second thermal treatment only slightly deteriorates the performance (green line in Figure 8a). In contrast, the device that was annealed prior to the Al deposition is shorted, possibly due to the degradation of the ITO/PEDOT:PSS contact by PEG (blue line in Figure 8b). However, a significant improvement is exhibited after the second annealing, ~2.1% PCE (green line in Figure 8b). We attribute this improvement to the formation of a PEG cathodic interlayer at the organic/Al interface during the second annealing process. Interestingly, this performance value is lower than that obtained for devices that were annealed only after the Al deposition, even with PEG concentrations under the approximated saturation value (0.55 mg/ml). We assign the reduced Jsc and Voc to non-reversible PEG adsorption at the bottom organic/anode interface during the first thermal treatment.39-43


27


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 28 of 40

Figure 8. J-V curves under illumination of P3HT:PC71BM devices with: (a) 0 mg/ml PEG; (b) 6 mg/ml PEG; that were annealed either before Al evaporation (blue curves), or before and after Al evaporation (green curves). Finally, realizing that annealing PEG-containing films before Al evaporation induces its alignment at the bottom substrate suggests using this process to generate a cathodic PEG interlayer in devices with inverted geometry (i.e. ITO cathode). To demonstrate this, we spun P3HT:PC71BM:PEG blends, with various PEG concentrations, onto ITO substrates and thermally


28

Page 29 of 40

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


annealed the films at 115Ԩ for 1hour to enable the alignment of PEG at the ITO substrates, and to prevent the migration of PEG towards the top contact in the following processing steps. The films were thermally covered by 10nm MoOx anodic interlayer and 120nm Al electrode. The J-V curves under illumination and in the dark of inverted P3HT:PC71BM:PEG devices with 0, 0.5, 1, 3, and 6 mg/ml of PEG, annealed before MoOx/Al deposition are presented in Figure 9a and b, respectively. The related performance parameters are summarized in Table 3.

Figure 9. J-V curves (a) in the dark, and (b) under illumination, of inverted ITO/P3HT:PC71BM:PEG/MoO3/Al devices with 0, 0.5, 1, 3, 6 mg/ml PEG. The films are


29


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 30 of 40

annealed at 115°C for 60 min before MoO3/Al deposition. Each parameter represents an average of 16 similar devices. The J-V curves in Figure 9 clearly show that introducing PEG into inverted devices and applying thermal annealing procedure before anode evaporation results in a gradual increase in turn-on voltage (Figure 9a), and a gradual enhancement of Voc and FF (Figure 9b and Table 3) with PEG concentration. Here too, the FF enhancement is contributed by the increase of Rsh and a decrease of Rs with PEG concentration. These trends are assigned to the gradual formation of a PEG interlayer at the ITO/organic interface that enhances the built-in potential as evident from the increase of turn-on voltage in the dark J-V characteristics (Figure 9a). Furthermore, the insertion of PEG interlayer reduces the exponential diffusion currents at low voltages, similar to that observed in the direct devices (Figure 5), corroborating that PEG reduces parasitic injection currents at the cathode.

Table 3. Inverted device characteristics of P3HT:PC71BM:PEG blends, annealed before the Al contact deposition, with PEG concentrations of 0, 0.5, 1, 3, 6 mg/ml. Each parameter represents an average value on 16 similar devices. PEG concentration [mg/ml]

Jsc [mA/cm²]

Voc [V]

FF

Efficiency [%]

Rsh [ Ω ]

Rs [ Ω ]

0

10.3±0.9

0.39±0.01

0.42±0.014

1.7±0.18

5659±520

486±34

0.5

11±1

0.42±0.01

0.44±0.009

2±0.2

5848±490

488±38

1

11±1.3

0.45±0.01

0.44±0.005

2.2±0.23

5910±500

486±32

3

11±1.2

0.5±0.01

0.47±0.007

2.6±0.3

6653±540

438±27

6

11±1.1

0.54±0.01

0.5±0.01

3±0.2

8150±560

414±18


30

Page 31 of 40

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


CONCLUSIONS

To summarize, in this research we studied the dynamics of metal-induced additive segregation to form cathodic interlayers in organic solar cells. To do so, we blended P3HT:PC71BM with PEG additive and showed that PEG concentration and the sequence of thermal treatments/metal evaporation tailors the top and bottom organic/electrode interfaces. XPS, DSIMS and device analysis were used to establish the correlation between PEG concentration in the initial solution, its amount at organic/Al interface, and the device parameters. We found that when the blends are not annealed, PEG is distributed within the film according to the kinetics of the spin coating process, i.e. the majority of PEG accumulates at the bottom substrate, while the minority resides in the film. During the electrode evaporation, Al-PEG interactions provide the driving force for segregation of PEG molecules that were kinetically "trapped" in the film towards the organic/Al interface. Under such conditions, the generation of the PEG interlayer is diffusion limited. On the other hand, annealing the film after Al evaporation enhances PEG migration to the organic/metal interface until full coverage is obtained. A saturation in Voc, with respect to PEG concentration, is obtained upon formation of a fully continuous interlayer. In the conditions of this study, the concentration of PEG required to form a continues interlayer is ~0.55 mg/ml. The saturation of Rs at the same concentration of PEG suggests the establishment of an ohmic contact associated with continues interlayer formation. At lower PEG concentrations, all PEG molecules migrate to the Al when annealed after the Al deposition, resulting in a partial coverage of the Al electrode. At PEG/Al areas, the local dipole moment induces local modification of the effective work function triggering a partial shift of the Voc parameter. Therefore, the shift in Voc represents a statistical value of modified and un-modified areas of the Al electrode. In contrast, annealing the film before the Al evaporation drives the thermodynamically favored PEG vertical


31


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 32 of 40

separation toward the ITO or PEDOT:PSS substrate significantly reducing the number of PEG molecules kinetically "trapped" in the film. Under such conditions, the distance between Al and PEG is too large for the PEG-Al interaction potential to induce migration of PEG to Al. PEG segregation to the bottom substrate affects the bottom electrode effective work function and can be harnessed for Voc enhancement in inverted cells. In conclusion, we demonstrated how the organic/metal interfacial composition, and hence electronic properties, can be "spontaneously" tuned by applying judiciously selected processing protocols based on kinetic and thermodynamic considerations. Although further study is required to generalize our method for other BHJ materials and additives, the rule of thumb we provided here is expected to hold for similar polymer-based BHJ systems. Because they are extremely soft and thin, the diffusion coefficients are sufficient to allow full activation of the diffusion process at temperatures as low as ~100°C. This implies, that under moderate thermal annealing, a full metal coverage by oligomer additive is expected, regardless of the specific diffusion coefficient value. Therefore, for other BHJ polymer-based systems with film thickness similar to those in this study, a similar PEG wt% is expected to obtain an ohmic contact and maximal Voc. If another additive is used, this saturation concentration can also be valid under condition that the chosen additive is not too branched and does not interact with the BHJ matrix. Furthermore, we speculate that if the magnitude of the additives’ dipole moment is sufficient for local fermi level pinning, the shift of the metal effective work function will be determined by the general dipole moment direction of the adsorbed additive and the extent of surface coverage. Therefore, this study provided practical insight for the selection of additives and its quantity with respect to the processing conditions. These results emphasize the necessity of understanding the thermodynamic and kinetic considerations that take place during the device processing in order to gain control over the


32

Page 33 of 40

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


device performance. Importantly, organic/metal interfaces are inherent in all organic electronic devices and hence these findings can impact the processing of all contact-limited organic devices.

ASSOCIATED CONTENT Fitting parameters of XPS profiles of Al-covered P3HT:PC71BM:PEG films that were notannealed or annealed after Al deposition. Figures displaying FF, Series resistance (Rs) and Shunt resistance (Rsh) as a function of initial PEG concentration for devices that were not-annealed, annealed before Al deposition, or annealed after Al deposition. The FF trends are consistent with Voc trends. Dynamic Secondary Ion Mass Spectroscopy (DSIMS) depth profiling of a P3HT:DPEG film with no PC71BM. The normalized depth profile shows a single peak of accumulated DPEG near the silicon substrate. The absence of the D-PEG doublet observed for all P3HT:PC71BM:D-PEG films (Figure 6 in manuscript) suggests a PC71BM rich phase is responsible of the peak doubling.

ACKNOWLEDGMENTS We thank Prof Nir Tessler and Dr. Drugesh from The Department of Electrical Engineering, Technion, for helpful discussions. We acknowledge partial funding of GIF, the German-Israeli Foundation for Scientific Research and Development Research Grant I-1251-302.5. J.V wishes to thanks RBNI, Russel Berrie Nanotechnology Institute, for support. MLC acknowledges Alstria for support of research on thin film solar cells at UCSB. The MRL Shared Experimental


33


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 34 of 40

Facilities are supported by the MRSEC Program of the NSF under Award No. DMR 1121053; a member of the NSF-funded Materials Research Facilities Net-work.

REFERENCES 1.

Liu, F.; Gu, Y.; Jung, J. W.; Jo, W. H.; Russell, T. P. On the Morphology of Polymer-

Based Photovoltaics. J. Polym. Sci., Part B: Polym. Phys. 2012, 50, 1018-1044. 2.

Liu, F.; Gu, Y.; Shen, X.; Ferdous, S.; Wang, H.-W.; Russell, T. P. Characterization of

the Morphology of Solution-Processed Bulk Heterojunction Organic Photovoltaics. Prog. Polym. Sci. 2013, 38, 1990-2052. 3.

Brabec, C. J. Semiconductor Aspects of Organic Bulk Heterojunction Solar Cells. In

Organic Photovoltaics: Concepts and Realization, Brabec, C. J.; Dyakonov, V.; Parisi, J.; Sariciftci, N. S., Ed.; Springer: Berlin Heidelberg, 2003; p 159-248. 4.

Hwang, J.; Wan, A.; Kahn, A. Energetics of metal–organic interfaces: New Experiments

and Assessment of the Field. Mater. Sci. Eng., R. 2009, 64, 1-31. 5.

Steim, R.; Kogler, F. R.; Brabec, C. J. Interface Materials for Organic Solar Cells. J.

Mater. Chem. 2010, 20, 2499-2512. 6.

Spies, A.; List, M.; Sarkar, T.; Würfel, U. On the Impact of Contact Selectivity and

Charge Transport on the Open-Circuit Voltage of Organic Solar Cells. Adv. Energy Mater. 2017, 7, 1601750 7.

Braun, S.; Salaneck, W. R.; Fahlman, M. Energy-Level Alignment at Organic/Metal and

Organic/Organic Interfaces. Advanced Materials 2009, 21, 1450-1472.


34

Page 35 of 40

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


8.

Endres, J.; Pelczer, I.; Rand, B. P.; Kahn, A. Determination of Energy Level Alignment

within an Energy Cascade Organic Solar Cell. Chem. Mater. 2016, 28, 794-801. 9.

Koch, N. Organic Electronic Devices and Their Functional Interfaces. ChemPhysChem

2007, 8, 1438-1455. 10. Opitz, A.; Wilke, A.; Amsalem, P.; Oehzelt, M.; Blum, R.-P.; Rabe, J. P.; Mizokuro, T.; Hörmann, U.; Hansson, R.; Moons, E.; Koch, N. Organic Heterojunctions: Contact-Induced Molecular Reorientation, Interface States, and Charge Re-distribution. Sci. Rep. 2016, 6, 21291. 11. Zhou, Y.; Fuentes-Hernandez, C.; Shim, J.; Meyer, J.; Giordano, A. J.; Li, H.; Winget, P.; Papadopoulos, T.; Cheun, H.; Kim, J.; Fenoll, M.; Dindar, A.; Haske, W.; Najafabadi, E.; Khan, T. M.; Sojoudi, H.; Barlow, S.; Graham, S.; Brédas, J.-L.; Marder, S. R.; Kahn, A.; Kippelen, B. A Universal Method to Produce Low–Work Function Electrodes for Organic Electronics. Science 2012, 336, 327-332. 12. Jørgensen, M.; Norrman, K.; Krebs, F. C. Stability/Degradation of Polymer Solar Cells. Sol. Energy Mater. Sol. Cells 2008, 92, 686-714. 13. Natan, A.; Kronik, L.; Haick, H.; Tung, R. T. Electrostatic Properties of Ideal and Nonideal Polar Organic Monolayers: Implications for Electronic Devices. Advanced Materials 2007, 19, 4103-4117. 14. Page, Z. A.; Liu, Y.; Duzhko, V. V.; Russell, T. P.; Emrick, T. Fulleropyrrolidine Interlayers: Tailoring Electrodes to Raise Organic Solar Cell Efficiency. Science 2014, 346, 441444.


35


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 36 of 40

15. Qi, B.; Zhang, Z.-G.; Wang, J. Uncovering the Role of Cathode Buffer Layer in Organic Solar Cells. Sci. Rep. 2015, 5, 7803. 16. Ratcliff, E. L.; Zacher, B.; Armstrong, N. R. Selective Interlayers and Contacts in Organic Photovoltaic Cells. J. Phys. Chem. Lett. 2011, 2, 1337-1350. 17. Park, J. H.; Lee, T.-W.; Chin, B.-D.; Wang, D. H.; Park, O. O. Roles of Interlayers in Efficient Organic Photovoltaic Devices. Macromol. Rapid Commun. 2010, 31, 2095-2108. 18. Lucera, L.; Kubis, P.; Fecher, F. W.; Bronnbauer, C.; Turbiez, M.; Forberich, K.; Ameri, T.; Egelhaaf, H.-J.; Brabec, C. J. Guidelines for Closing the Efficiency Gap between Hero Solar Cells and Roll-To-Roll Printed Modules. Energy Technol. 2015, 3, 373-384. 19. Vinokur, J.; Shamieh, B.; Deckman, I.; Singhal, A.; Frey, G. L. Mechanisms for Spontaneous Generation of Interlayers in Organic Solar Cells. Chem. Mater. 2016, 28, 88518870. 20. Zhang, H.; Tan, W.-Y.; Fladischer, S.; Ke, L.; Ameri, T.; Li, N.; Turbiez, M.; Spiecker, E.; Zhu, X.-H.; Cao, Y.; Brabec, C. J. Roll to Roll Compatible Fabrication of Inverted Organic Solar Cells with a Self-Organized Charge Selective Cathode Interfacial Layer. J. Mater. Chem. A 2016, 4, 5032-5038. 21. Kim, S.; Kang, H.; Hong, S.; Lee, J.; Lee, S.; Park, B.; Kim, J.; Lee, K. A Versatile SelfOrganization Printing Method for Simplified Tandem Organic Photovoltaics. Adv. Funct. Mater. 2016, 26, 3563-3569. 22. Dekman, I.; Brener, R.; Frey, G. L. Thermal Metal Deposition Induces Segregation in Polymer Thin Films: a Demonstration on OPVs. J. Mater. Chem. C 2013, 1, 6522-6525.


36

Page 37 of 40

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


23. Shamieh, B.; Obuchovsky, S.; Frey, G. L. Spontaneous Generation of Interlayers in OPVs with Silver Cathodes: Enhancing Voc and Lifetime. J. Mater. Chem. C 2016, 4, 18211828. 24. Deckman, I.; Obuchovsky, S.; Moshonov, M.; Frey, G. L. Chemical Composition of Additives That Spontaneously Form Cathode Interlayers in OPVs. Langmuir 2015, 31, 67216728. 25. Fisher, G. L.; Walker, A. V.; Hooper, A. E.; Tighe, T. B.; Bahnck, K. B.; Skriba, H. T.; Reinard, M. D.; Haynie, B. C.; Opila, R. L.; Winograd, N.; Allara, D. L. Bond Insertion, Complexation, and Penetration Pathways of Vapor-Deposited Aluminum Atoms with HO- and CH3O-Terminated Organic Monolayers. J. Am. Chem. Soc. 2002, 124, 5528-5541. 26. Fisher, G. L.; Hooper, A. E.; Opila, R. L.; Allara, D. L.; Winograd, N. The Interaction of Vapor-Deposited Al Atoms with CO2H Groups at the Surface of a Self-Assembled Alkanethiolate Monolayer on Gold. J. Phys. Chem. B 2000, 104, 3267-3273. 27. Deckman, I.; Moshonov, M.; Obuchovsky, S.; Brener, R.; Frey, G. L. Spontaneous Interlayer Formation in OPVs by Additive Migration Due to Additive-Metal Interactions. J. Mater. Chem. A 2014, 2, 16746-16754. 28. Germack, D. S.; Chan, C. K.; Hamadani, B. H.; Richter, L. J.; Fischer, D. A.; Gundlach, D. J.; DeLongchamp, D. M. Substrate-Dependent Interface Composition and Charge Transport in Films for Organic Photovoltaics. Appl. Phys. Lett. 2009, 94, 233303. 29. Yao, Y.; Hou, J.; Xu, Z.; Li, G.; Yang, Y. Effects of Solvent Mixtures on the Nanoscale Phase Separation in Polymer Solar Cells. Adv. Funct. Mater. 2008, 18, 1783-1789.


37


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 38 of 40

30. Susarova, D. K.; Troshin, P. A.; Moskvin, Y. L.; Babenko, S. D.; Razumov, V. F. Vertical Concentration Gradients in Bulk Heterojunction Solar Cells Induced by Differential Material Solubility. Thin Solid Films 2011, 519, 4132-4135. 31. Krausch, G. Surface Induced Self Assembly in Thin Polymer Films. Mater. Sci. Eng., R. 1995, 14, v-94. 32. Vinokur, J.; Deckman, I.; Obuchovsky, S.; Weinfeld, K.; Frey, G. L. In The Effect of Thermal Annealing on Additive Migration to the Organic/Metal Interface in OPVs, 2015; p 95670X-95670X-8. 33. Israelachvili, J. N. Adhesion and Wetting Phenomena. In Intermolecular and Surface Forces (Third Edition), Ed.; Elsevier, San Diego, 2011; p 415-467. 34. Uglea, C. V. Oligoethers. In Oligomer Technology and Applications, Ed.; Taylor & Francis, 1998, p 249-298. 35. Germack, D. S.; Chan, C. K.; Kline, R. J.; Fischer, D. A.; Gundlach, D. J.; Toney, M. F.; Richter, L. J.; DeLongchamp, D. M. Interfacial Segregation in Polymer/Fullerene Blend Films for Photovoltaic Devices. Macromolecules 2010, 43, 3828-3836. 36. DeLongchamp, D. M.; Kline, R. J.; Herzing, A. Nanoscale Structure Measurements for Polymer-Fullerene Pshotovoltaics. Energy Environ. Sci. 2012, 5, 5980-5993. 37. Clark, M. D.; Jespersen, M. L.; Patel, R. J.; Leever, B. J. Predicting Vertical Phase Segregation in Polymer-Fullerene Bulk Heterojunction Solar Cells by Free Energy Analysis. ACS Appl. Mater. Interfaces 2013, 5, 4799-4807.


38

Page 39 of 40

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


38. Yang, H. Y.; Kang, N. S.; Hong, J.-M.; Song, Y.-W.; Kim T. W.; Lim, J. A. Efficient Bilayer Heterojunction Polymer Solar Cells With Bumpy Donor–Acceptor Interface Formed by Facile Polymer Blend. Organic Electronics 2012, 13, 2688-2695. 39. Cao, F.-Y.; Lai, Y.-Y.; Chen, Y.-L.; Cheng, Y.-J. Self-Assembled Tri-, Tetra- and PentaEthylene Glycols as Easy, Expedited and Universal Interfacial Cathode-Modifiers for Inverted Polymer Solar Cells. J. Mater. Chem. A 2016, 4, 8707-8715. 40. Kang, H.; Kee, S.; Yu, K.; Lee, J.; Kim, G.; Kim, J.; Kim, J.-R.; Kong, J.; Lee, K. Simplified Tandem Polymer Solar Cells with an Ideal Self-Organized Recombination Layer. Advanced Materials 2015, 27, 1408-1413. 41. Kang, H.; Hong, S.; Lee, J.; Lee, K. Electrostatically Self-Assembled Nonconjugated Polyelectrolytes as an Ideal Interfacial Layer for Inverted Polymer Solar Cells. Advanced Materials 2012, 24, 3005-3009. 42. Chien, S.-C.; Chen, F.-C.; Chung, M.-K.; Hsu, C.-S. Self-Assembled Poly(ethylene glycol) Buffer Layers in Polymer Solar Cells: Toward Superior Stability and Efficiency. J. Phys. Chem. C 2012, 116, 1354-1360. 43. Ma, D.; Lv, M.; Lei, M.; Zhu, J.; Wang, H.; Chen, X. Self-Organization of Amine-Based Cathode Interfacial Materials in Inverted Polymer Solar Cells. ACS Nano 2014, 8, 1601-1608.


39


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 40 of 40

*Corresponding author E-mail: [email protected]

GRAPHICAL ABSTRACT


40

Dynamics of Additive Migration to Form Cathodic Interlayers in

Recommend Documents