Kinetics of PolymerâFullerene Phase Separation during Solvent

Subscriber access provided by University of Newcastle, Australia

Article

Kinetics of polymer-fullerene phase separation during solvent annealing studied by table-top X-ray scattering Karol Vegso, Peter Siffalovic, Matej Jergel, Peter Nadazdy, Vojtech Nádaždy, and Eva Majkova ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b15167 • Publication Date (Web): 20 Feb 2017 Downloaded from http://pubs.acs.org on February 21, 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 24

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

Kinetics of polymer-fullerene phase separation during solvent annealing studied by table-top X-ray scattering Karol Vegso,†,‡ Peter Siffalovic,‡,* Matej Jergel,‡ Peter Nadazdy,‡ Vojtech Nadazdy,‡ and Eva Majkova‡ †

Japan Synchrotron Radiation Research Institute, 1-1-1 Kouto, 679-5198 Hyogo, Japan Institute of Physics SAS, Dubravska cesta 9, 84511 Bratislava, Slovakia *corresponding author: [email protected]

‡

Abstract Solvent annealing is an efficient way of phase separation in polymer-fullerene blends to optimize bulk heterojunction morphology of active layer in polymer solar cells. To track the process in real time across all relevant stages of solvent evaporation, laboratory-based in-situ small- and wide-angle X-ray scattering measurements were applied simultaneously to a model P3HT:PCBM blend dissolved in dichlorobenzene. The PCBM molecule agglomeration starts at ≈7 wt% concentration of solid content of the blend in solvent. Although PCBM agglomeration is sloweddown at ≈10 wt% of solid content, the rate constant of phase separation is not changed, suggesting agglomeration and re-ordering of P3HT molecular chains. Having the longest duration, this stage most affects BHJ morphology. Phase separation is accelerated rapidly at concentration ≈25 wt%, having the same rate constant as the growth of P3HT crystals. P3HT crystallization is driving force for phase separation at final stages before a complete solvent evaporation, having no visible temporal overlap with PCBM agglomeration. For the first time, such a study was done in laboratory demonstrating potential of the latest generation table-top high-brilliance X-ray source as a viable alternative before more sophisticated X-ray scattering experiments at synchrotron facilities are performed. Keywords: P3HT:PCBM, solar cell, solvent annealing, GI-SAXS, GI-WAXS

1 ACS Paragon Plus Environment


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 24

INTRODUCTION

One of the fundamental mechanisms having a significant impact on the power conversion efficiency of polymer solar cells1 is phase separation on the nanoscale between electron-donating polymer and electron-accepting fullerene. Such an active layer composed of bicontinuous, interpenetrating network of polymer and fullerene is referred as bulk heterojunction (BHJ)2. The nanoscale morphology of BHJ plays a key role in the conversion of solar energy into electrical power. The spatial correlation length of polymer:fullerene blends in BHJ can be directly related to the exciton diffusion length which, on the other hand, controls exciton dissociation and finally charge transport. Hence, much effort has been spent to elucidate phenomena of phase separation during BHJ formation3-16 in the last decade. A large number of experimental studies have been devoted to understanding BHJ formation from a generic blend based on poly-(3-hexylthiophene2,5-diyl) (P3HT) as electron acceptor and [6,6]-phenyl C61-butyric acid methylester (PCBM) as electron donor17-25. A widely accepted concept of P3HT:PCBM BHJ is based on a mixture of amorphous phase and crystalline P3HT domains intercalated with PCBM clusters26. Exciton generation, diffusion and dissociation takes place in the intercalated phase within a standard BHJ model27. The polymer:fullerene blend is applied from a solution by spin-casting or doctor-blade technique, BHJ morphology being affected by many factors. Already in 2005 Li and coworkers28 recognized that a controlled slow growth of BHJ in saturated solution vapors, known as solvent annealing, has a beneficial influence on the power conversion efficiency of solar cells. In particular, solvent annealing promotes P3HT crystallinity and improves nanoscale separation29. Moreover, slow formation of BHJ in solvent vapors favors a thermodynamically more stable final state which has a positive impact on the long-term stability of solar cells. The choice of experimental techniques allowing to track in-situ BHJ formation starting from liquid state is


Page 3 of 24

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


rather restricted. The in-situ spectral reflectometry30 accompanied by light scattering can monitor fullerene agglomeration and polymer crystallization but cannot yield any quantitative data on the nanoscale BHJ phase separation. On the other hand, simultaneous measurements of small- and wide-angle X-ray scattering18, 31-33 can provide a full picture of the real-time structural evolution on the scale ranging from ångstroms up to several hundreds of nanometers. Recent advanced in synchrotron instrumentation19, 21-22, 24 have shifted limits of temporal resolution into millisecond range, allowing thus true real-time studies of highly non-equilibrium phenomena taking place during BHJ formation after spin casting. Here, we report on an in-situ real-time study of BHJ formation by phase separation during solvent annealing of P3HT:PCBM blend dissolved in dichlorobenzene (DCB) which was realized by grazing-incidence small- and wide-angle X-ray scattering (GI-SAXS/WAXS). P3HT:PCBM blend is the first choice option for our case study, being one of the most studied polymer:fullerene systems for solar cells27. The solvent annealing was done precisely according to a common fabrication protocol of polymer solar cells. Recent study of W. Wang et al.34 selected the DCB solvent as the most suitable for solvent annealing. In particular, a freshly spincoated solution of the blend was covered by a Petri dish and GI-SAXS and GI-WAXS were measured simultaneously up to a complete solvent evaporation. While GI-SAXS provides information on agglomeration of PCBM molecules and phase separation between PCBM and P3HT resulting in particular BHJ morphology, GI-WAXS probes P3HT crystallization starting from P3HT clusters at a certain moment of solvent evaporation. In this way, we were able to identify all relevant stages of solvent annealing up to a complete solvent evaporation. To relate the solvent annealing time to a corresponding P3HT:PCBM blend concentration in DCB solution, a complementary in-situ white-light reflectometry (WLR)11,

20, 24

measurement was 3

ACS Paragon Plus Environment


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 24

done separately to track the thickness reduction of spin-coated film during solvent evaporation. In contrast to previous similar time-resolved GI-SAXS/WAXS studies completed exclusively at synchrotron facilities, the present work was performed in laboratory using a high-brilliance table-top X-ray source. This suggests new experimental possibilities for systematic laboratorybased in-situ studies of phase separation phenomena in soft matter in general and in polymer and hybrid solar cells to optimize BHJ morphology in particular. For example, the proposed method can help to improve control of phase separation in new low-band gap polymer:fullerene systems that require special additives to enhance miscibility of the components in order to refine the BHJ domain size24, 35-38.

■

RESULTS AND DISCUSSION

Before proceeding to results, we will present briefly some general aspects of the experimental methods used. (GI-)SAXS is sensitive to any electron density fluctuation and can thus reveal size of molecular complexes or clusters and their spatial correlations39. An effective experimental approach to assess average size of molecular complexes is based on a Guinier analysis (GA)40-41. In particular, GA provides an estimate of the average radius of gyration, Rg, based on a fit of the scattered intensity in low q-range, q being the scattering vector. In a particular case of spherical PCBM clusters, cluster radius R is given as = 5⁄3 . A fundamental prerequisite for application of GA is a so-called condition of dilute system in which scattering objects to be measured do not exhibit any spatial correlations in the course of measurement. In our case, the initial blend concentration of 2 wt% cannot guarantee fulfillment of this condition. Nevertheless,


Page 5 of 24

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


a number of experimental studies8-9,

18, 42-43

have shown that GA analysis of even dried and

thermally annealed P3HT:PCBM films can provide rational results on the size of PCBM clusters. A detailed analysis of small-angle neutron scattering44 indicated two slopes in the Guinier region, which correspond to large and small PCBM domains with ≈ 0.03 nm-1 being the threshold separating them. The reason can be found in a weak and broad interference function of weakly scattering PCBM clusters with a broad size dispersion. A detailed description of GA of our GISAXS data including an extended error analysis is given in the supplementary section. The Fig. 1 shows temporal evolution of Rg of PCBM clusters during solvent annealing. The auxiliary top horizontal axis shows solid content xs (in wt%) of P3HT:PCBM blend in the evaporating DCB solvent. Here, xs is defined as = BHJ ⁄BHJ + DCB where mDCB(t) is the mass of residual DCB solvent and mBHJ is the mass of solid P3HT:PCBM blend. mDCB(t) was determined by a time-resolved thickness measurement of the drying liquid film (Fig. S5a) using WLR technique20, 24, 30. The details are comprehensively described in the supplementary part. PCBM clusters are detectable above 7 wt% of solid content (Fig. 1), their diameter increasing steadily from an initial value of ≈6.5 nm. For toluene solvent, such a tendency to formation of large PCBM clusters was suggested by molecular dynamics simulations by Mortuza et al45. An amorphous spherical cluster of 6.5 nm diameter contains 150 randomly packed PCBM molecules with an effective diameter of 1.05 nm, packing fraction being ≈64%45-46. Crystallinity of PCBM clusters could not be checked due to a strong background coming from DCB solvent47. The growth of PCBM clusters starts to saturate at a concentration of ≈10 wt% of solid content. In order to provide a quantitative description of PCBM cluster growth kinetics, we adopted an equation derived within Johnson-Mehl-Avrami (JMA) theory48 of phase transformation in solids which predicts volume V(t) of the new phase to increase as = 1 −

!"#!#$ %

&. Here, k 5

ACS Paragon Plus Environment


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 24

is the growth rate constant, t0 is the initial time and n is a so-called Avrami parameter. If we assume a constant number of growing PCBM clusters after they emerge in GI-SAXS pattern, their volume increases as ∝ () . A JMA fit shown as the red line in Fig. 1 suggests the rate constant of PCBM cluster growth kPCBM=(33±2)×10-4 s-1 and Avrami exponent nPCBM=1.2±0.1. The threshold concentration PCBM for PCBM agglomeration (solubility limit) derived from JMA fit is 7.1±0.1 wt% of solid content at T=23°C. Binary phase diagrams of PCBM and P3HT solutions in DCB were experimentally measured by Schmidt-Hansberg et al47. The solubility limits for PCBM and P3HT were found at 3 wt% and 5 wt%, respectively, at T=25°C47, 49. Therefore for our P3HT:PCBM weight mixing ratio 1.5:1, the estimated solubility limit for PCBM should be PCBM ≈7.5 wt% assuming weakly interacting P3HT and PCBM molecules which is close to our experimentally found value PCBM =7.1±0.1 wt%. A small difference can be explained by a strong temperature dependence of the solubility limit. Therefore we can judge that PCBM agglomeration is not significantly affected by presence of P3HT. The observed Avrami exponent close to one can indicate a heterogeneous nucleation and onedimensional cluster growth16. However we should note, that the original JMA equation was derived under the assumption of isothermal system with a constant growth rate50 neglecting peculiarities of the molecular diffusion processes undergoing during the solvent annealing of P3HT:PCBM blend. The GI-WAXS pattern of BHJ after solvent annealing is shown in Fig. S6. The final dry state is composed of a crystalline P3HT lamellar phase having the edge-on orientation. The most distinct 100 diffraction peak was used to track polymer crystallization. Temporal evolution of its area and position obtained from Gaussian fit are shown in Fig. 2a and Fig. 2b, respectively, while a precise width evaluation is more complicated and is discussed in the supplementary part. The 6 ACS Paragon Plus Environment

Page 7 of 24

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


peak area proportional to the volume of P3HT crystalline phase can be fitted within JMA growth model with two growth rate constants. The first stage of fast P3HT crystal growth is completed by evaporation of DCB solvent, being characterized by a rate constant ,-P3HT =(10±2)×10-3 s-1 and Avrami exponent /-P3HT =2±0.4. The d100 lattice spacing of P3HT lamellar phase is reduced from 16.75 Å to 16.45 Å (-1.8%). The P3HT crystallization threshold visible in Fig. 2 is

P3HT =24.7±3 wt%. Pure P3HT in DCB solvent transforms to gel phase at ≈10 wt%47. Therefore in our P3HT:PCBM blend with the weight mixing ratio 1.5:1, the P3HT crystallization threshold is expected at P3HT ≈16.7 wt% of solid content which is lower than the value

P3HT =24.7±3 wt%. However, the P3HT 100 diffraction peak starts to emerge in GI-WAXS pattern already at the solid content P3HT ≈15 wt% (Fig. S7a) which fits the expected value of the gel phase transformation and is in agreement with previous GI-WAXS studies performed with intense synchrotron beams33, 47. Obviously, experimental detection limits of our laboratory X-ray source do not allow to track quantitatively P3HT crystallization from the very beginning. The Avrami exponent close to two indicates heterogeneous nucleation with two-dimensional crystallite growth16. The second stage of P3HT crystal growth is characterized by a much lower rate constant ,0P3HT =(20±7)×10-4 s-1 and Avrami exponent /0P3HT =1.2±0.7. This stage is dominated by a gradual slow release of residual DCB molecules leading to a 0.6% reduction of d100 lattice spacing. Such a small reduction is probably the reason why no measureable changes in the total layer thickness were observed at this growth stage using WLR19. In practice, this slow period can be shortened by thermal annealing51. Differently to P3HT, GI-WAXS cannot provide usable information on PCBM as its diffraction ring is overlaid by a strong diffraction signal coming from DCB solvent (Fig. S6). A comparison of Fig. 2a to Fig. 1 shows that agglomeration and crystallization of P3HT domains has no overlap with the preceding PCBM agglomeration. In 7 ACS Paragon Plus Environment


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 24

addition to PCBM clusters and P3HT crystal domains, amorphous P3HT:PCBM phase is always present in the final annealing product10,

26, 52

. However, its detailed analysis is beyond

possibilities of present X-ray techniques. Up to now, we have addressed PCBM agglomeration and P3HT crystallization independently. However, both processes affect percolated BHJ network composed of P3HT and PCBM domains. Their mean separation spacing can be easily tracked in time by an interference peak shift in GI-SAXS pattern32. Hence, being able to monitor temporal evolution of GI-SAXS intensity around the interference peak, we can relate proceeding phase separation during solvent annealing to PCBM agglomeration and P3HT crystallization. The Fig. 3 shows several GI-SAXS line profiles along the lateral component of scattering vector qy drawn in the form of Kratky plots39, 53. These were extracted from GI-SAXS patterns at the critical exit angle at selected instants of solvent annealing and the initial profile at time zero was subtracted. An interference peak at qy≈0.25 nm-1 visible from 900 s (xs=33 wt%) is a sign of separation between P3HT chains14, 54-55 with a mean value ∆≈ 23⁄ 456 being ≈25 nm. However, the interference peak area cannot be used directly to quantify temporal evolution of phase separation and BHJ formation. First, the interference peak in GI-SAXS pattern need not be resolved from the very beginning of this process. Second, the peak area is affected by a continuously changing electron density in the irradiated volume due to solvent evaporation. In order to tackle these problems, we introduce a quantity Φ which is proportional to the self-organized phase-separated volume fraction. In particular, we use the formalism of Porod invariant39, 56 and define the ratio Φ as

Φ =

8>? 9 : ;9: 9 : ;9

Kinetics of PolymerâFullerene Phase Separation during Solvent

Recommend Documents

Kinetics of PolymerâFullerene Phase Separation during Solvent