Poly(3-hexylthiophene) (P3HT):[6,6]-phenyl C61 butyric acid methyl ester (PCBM) is the most studied organic photovoltaic (OPV) system. However, up to the date this system is unable to perform consistently primarily due to lack of knowledge of optimized morphological structure.1’4 Morphological structure is important in every steps of OPV operation, from photon absorption to charge collection.5 Photon absorption is affected by conjugation length and inter-chain interaction; charge dissociation is influenced by domain size of donor and acceptor, charge transportation and collection by respective electrodes are impacted by percolation path ways, crystallinity/mesoscopic order as well as degree of phase separation and domain size of the donor and acceptor. From the device perspective the importance of morphology can be easily recognized as short circuit current (Jsc) and fill factor (FF) of OPV are directly related to morphology of the thin film.
Surface wetting of a blend thin film depends on surface energies of constituent components and substrate.6’8 Thus vertical profile of P3HT:PCBM binary mixture thin film depends on surface energies of P3HT (27 mJ/m2)9 , PCBM (38 mJ/m2)9 and the substrate. However, surface segregation in semi-crystalline polymers, such as, semicrystalline P3HT and PCBM nano-particles, is expected to be far complex than that of amorphous polymers due to crystal phase polymer chain interactions. Crystallization ordering decreases overall enthalpy of the system. This may balance the increase of the surface free energy as a result of migration of the low surface energy component to the bulk and the decrease in the entropy of mixing. In two separate studies both He et al.10 and Kiel et al.11’13 observed thin PCBM rich layer adjacent to the silicon substrate for the PCBM-P3HT spin cast films. It is believed that higher surface energy PCBM of the constituent components segregates to high free energy surface of silicon. When the nano-particle has higher or comparable surface energy to polymer in polymer:nano-particle thin films it is easily comprehended that self-assembly of nano-particles at the hard surface is entropically favorable as this process releases the polymer chains from the hard substrate forced constraints or loss of degrees of freedom. The whole philosophy of the particle migration can be summarized by the simple principle of total energy minimization of the system. From above discussion it is obvious that substrate modification can also control the vertical segregation of nano-particles in a polymer:nano particle thin film.10
In general the term ‘thin film’ is attributed to any film whose thickness is below 1000 nm.14 Structure and properties of polymer thin and ultrathin films can be different from thick films or bulk materials. The obvious difference between these two classes is that a thin film can be considered as under two dimensionally confined, which in turn may affect polymer crystallization behavior15,16 and crystal structures17, molecular ordering18, morphology19, etc. Physical properties of polymer of thin films, such as, glass transition temperature, Tg,20’27 and molecular mobility,28’31 are largely affected by the interaction of the polymer and the substrate. Since such properties greatly influence crystallization kinetics of polymers, substrate plays a vital role in the crystallization behavior of thin films.32’34
In the previous section we discussed the effect of the substrate imposed confinement on thin film morphology. In this section we shall illustrate other confinements, such as, confinement imposed by lithographic channel, pore, etc., and their implications on crystallization and crystal orientations. The crystallization of highly isotactic polypropylene confined in self-ordered nanoporous alumina with variable pore diameter was studied by Duran et al.35 Crystal nucleation changed primarily from heterogeneous to primarily homogeneous as the pore diameter falls below 65 nm. With the decrease in pore size crystallization decreases most probably due to constraint in chain mobility. The crystallization is completely suppressed when the pore diameter is 20 nm and below: indicative of the critical nucleus size. Such change in crystal formation behavior depends, in large, on the altered segmental dynamics of polymers due to the presence in confining hard anodic aluminum oxide.36,37 Other research groups also observed the change in crystal nucleation mechanism from heterogeneous to homogeneous for thin films of poly(vinylidene fluoride) when confined by nano-patterned trenchs38 and thin films of poly(ethylene oxide), when confined inside the porous anodic aluminum oxide as compared to bulk.39 The fast growth axis of poly(vinylidene fluoride) crystals is found to be along the direction of the nano-imprint lithographic trenches38 and of poly(ethylene oxide) crystal growth propagated along the cylinder length.39 However, in case of polyethylene heterogeneous surface nucleation is observed inside the porous anodic aluminum oxide.39
To achieve the right nano-structured bi-continuous morphology annealing of P3HT:PCBM blend film is required. But there is an inherent problem: during both thermal and solvent annealing PCBM forms several micron size rod-like crystals.40’55 Elongated single crystals are formed likely due to the asymmetric molecular architecture of bare fullerene by dissolvable pendent group.56 Formation of such large crystals causes inefficient charge separation,10,57 loss of surface area, loss of material, poor contact of active material with the electrodes and leaves the blend acceptor lean. This is also the probable cause of short circuit of OPV’s. P3HT crystal order formation during post treatment is prevented by the fast diffusion and aggregation of PCBM molecules.58 Diffusion and aggregation of PCBM in such BHJ OPV further limit its enduring stability.59 According to Yang et al. the kinetics of PCBM crystal growth in thin composite films is controlled by surface and interface dominated long-range diffusion of PCBM molecules through the film matrix to different crystal growth fronts and by local incorporation rate of PCBM molecules into the growing crystals.56 However, Watts et al.41 observed that diffusion of PCBM is bulk controlled.
The thickness of surface enriched layer in a thin film largely depends on the relative position of the equilibrium point with respect to the phase envelop.60 If the blend equilibrate away from the critical point of the blend into one phase region then the surface enrichment layer is thin. But if the blend equilibrates close to the critical point then the surface enrichment layer is thick. For a multi-component system the degree of difference in the surface energy of the constituent components entails relative enrichment of air interface by the low surface energy component(s):61
dG=’_i”’z^*’_(i ) ‘ ‘??_i ‘ (I)
where ??_i is the chemical potential and ‘z^*’_i is the surface excess of species i. Surface excess is defined by the following equation:62
z^*=’_0^”(??(z)-??_’ )dz (II)
where ??(z) is the fraction of material by volume at a depth, z and ??_’ is the bulk volume fraction at z=’. Substrate surface and confinement and their surface energy can have immense effect on in plane and vertical segregation of polymers. In particular, crystallizable thin films of P3HT:PCBM blends are in large dependent on the substrate. In many occasions an aluminum cathode is deposited on the top of active layers of OPV’s and subsequently annealed to optimize the structures of active layers. This procedure is known as post annealing.63,64 The active layer is confined between hole transport layer (PEDOT:PSS) and a cathode (aluminum). For this case an interfacial energy, rather than surface energy for the pre-annealed case (when sample is annealed prior to cathode deposition), directs the segregation of components to the cathode interface. Castro et al. observed higher open circuit current, open circuit voltage, fill factor and overall efficiency for post annealed samples as compared to pre-annealed samples for MEH-PPV/Fullerene system.65 It is interesting to observe that samples without annealing showed better efficiency compared to the pre-annealed samples. They believed that overgrowth of fullerene crystals causes crevice in the film and degradation of morphology is responsible in the reduction of efficiency due to annealing. However, in case of post annealing, in general, efficiency, fill factor, open circuit voltage, short circuit current as well as power conversion efficiency increases. Their correlation between the morphology of the active layer to the performance of OPV’s is far less than definitive. A clearer picture on how the performance of the post annealed OPV’s vary with the morphology of the active layer is described by Chen et al. for P3HT:PCBM system.66 They observed an increase in electron accepting PCBM layer just adjacent to cathode for post annealed device. However, they did not shed light on the role of suppression of PCBM micro-crystals on the performance of the device. They also observed the formation of face on orientation of P3HT crystals adjacent to the cathode and attributed this to the performance augmentation of the device. However, it has been reported that edge on crystal growth results in higher device performance.67 Also, it has been suggested that post annealed films under harsh cathode confinement might have unreleased stresses and sustain contact damage during annealing.68 As discussed earlier substrate surface energy plays a critical role in vertical segregation as well as interface composition of polymer blend thin films. Positive electrode ITO with adjacent PEDOT:PSS layer is wetted by acceptor PCBM and air is wetted by P3HT for pre-annealed samples.10,1 It is worth noting that thinner films are more influenced by the substrate.69 Germack et al. used two different hole transport layers: PEDOT:PSS which slightly enriched the buried interface with PCBM and low surface energy Nafion based hole transport layer which enriched the buried substrate with P3HT.70 They reported that two different types of surface enriched films produced similar device performances. To explain their contradicting results with that of Xu et al.71 and Liao et al.72, they explained that there will be greater gain in performance by engineering cathode interface than that of anode. Recently, Bulliard et al. were able to increase OPV efficiency by controlling the surface energy of buffer zinc oxide layer by self-assembled monolayer (SAM) on which P3HT:PCBM active layer was cast.73
In this research we studied post annealing process by utilizing surface energy controlled Dow Corning?? 5-8601 Fluorosilicone pad as top confinement during thermal annealing. We chose such a pad as a top confining medium due to a few its advantages: flexibility, ease in detachment, and easily tunable surface energy by UVO exposure like PDMS.74’78 Surface energy of silicon substrate (surface energy 70 mJ/m2) can be tuned by coating a self-assembled monolayer (SAM) of (Tridecafluoro-1,1,2,2-tetrahydrooctyl)-dimethylchlorosilane. Top confinement suppressed PCBM crystallization. However, top confinement did not have much effect on nanoscopic crystallization of P3HT and P3HT crystal orientation.
Effect of casting solvents on drop cast thin films of conductive conjugated polymers is largely studied by characterizing post processed films. However, the results have often been inconclusive due to the inability to study the in-situ evolution of structures. In this research we implement a novel in-situ grazing incidence wide angle X-ray scattering (GIWAXS) approach to extracting morphological evolution information during film formation in model Poly(3-hexylthiophene) (P3HT): [6,6]-phenyl C61-butyric acid methyl ester (PCBM) blend films that have otherwise been widely studied. Casting solvents include chloroform, benzene and tetrahydrothiophene (THT), carefully selected for their relative solubilities of P3HT and PCBM. We determined that the casting solvents’ solubility for P3HT and pure solvent boiling point, along with residual solvent content in the films have significant implications on final thin film morphology and crystallization of its constituent components. For example, we observed orientations of P3HT in P3HT:PCBM films, cast from different solvents, are largely affected by the individual solubilities of P3HT and PCBM, and substrate surface energy. On the other hand PCBM crystal growth from different PCBM solutions predominantly depends on the solubilities of PCBM in the solvents and boiling points of solvents. These results have important ramifications for controlling desired morphology for polymer electronics, such as organic photovoltaics (OPV), organic field effect transistor (OFET) and photo-detectors.
Poly(3-hexylthiophene) (P3HT) is utilized in different organic electronic applications, such as, organic field effect transistor1’4 (OFET), photo detectors5,6, organic photovoltaics7’9 (OPV), light emitting diode6 (LED), etc. Recently, Dang et al. have compiled a wide range of power conversion efficiencies of P3HT: [6,6]-phenyl C61-butyric acid methyl ester (PCBM) organic photovoltaics (OPV).10 Large variations of electronic properties of P3HT and P3HT:PCBM based thin films, is primarily due to lack of understanding of the correlation of different factors, such as post processing treatment,11,12 casting solvent,13,14 regioregularity,15,16 molecular weight and polydispersity17,18 of P3HT, active layer thickness,19 and underlying substrate properties1,20 to the morphology of the film. One of the key factors in optimizing P3HT orientation in neat films and P3HT:PCBM active layer morphology is the casting solvent. Furthermore recently substantial effort has been dedicated to study the effect of mixed solvents and solvent additives on the morphology of such thin films.21,22 In various OPV studies it is found that variation of choice of casting solvents influences domain size,23 optical properties,24 photovoltaic parameters25 (short circuit current, fill factor, power conversion efficiency, etc.), degree of P3HT crystallization,13 etc.
While the effect of casting solvents on the above factors are recognized, the underlying scientific understanding associated with film morphology evolution during drying process is missing. Similarly, even though the charge mobility of regioregular P3HT may vary by two orders of magnitude depending on the solvent used,26 the present understanding of effect of solvents on charge mobility in P3HT is confusing at best. Field effect mobility for P3HT is found to be substantially higher when the casting solvent is 1,2,4-trichlorobenzene as compared to that cast from chloroform.27 This is attributed to the higher boiling point of 1,2,4-trichlorobenzene (boiling point: 213??C28), that facilitates the crystallization of P3HT, as compared to those of chloroform (boiling point: 61.18??C28). Yang et al. conducted a more comprehensive study on field effect mobility as a function of solvents.4 Field effect mobility as a function of solvent, ??solvent, can be ordered as ??chloroform > ??tetrahydrofuran >> ??toluene > ??methylene chloride. Yet their boiling points are methylene chloride (39.75??C) < chloroform (61.18??C) < tetrahydrofuran (64.85??C), toluene (110.63??C). Thus, there is no correlation between boiling point and field effect mobility. From the above discussion it is difficult to correlate solvent properties and charge mobility in polymers. Furthermore, anisotropy in charge career mobility due to crystallization of P3HT makes degree of crystallization as well as crystal size and orientation of P3HT important factors in different electronic applications.29 Therefore, it is timely to understand the kinetics of P3HT and PCBM crystal growth in order to better appreciate device performance with the final morphology of thin films. Moreover, there is significant resemblance between the growth of thin film polymer crystals and growth of that of small organic molecules as well as metals.30’32 However, since the rate of crystallization of polymer is orders of magnitude slower, in many occasions it is a better candidate for experimental studies of crystal growth. However, effects of casting solvents on crystal growth kinetics in polymers are rather complex. Crystal growth rate for our OPV system largely depends on a number of variables including solubility of P3HT and PCBM, glass transition temperature of P3HT and PCBM and suppression of glass transition temperature due to presence of solvent, substrate surface energy and roughness effects, as well as boiling point, volatility and partial vapor pressure of the solvent under study.
Drop casting is widely used to make thick films on the order of a few microns for studying different organic electronic devices. For instance, P3HT and P3HT:PCBM films find applications in different organic electronic devices, such as, in organic field effect transistor (OFET),2’4,33 and photo-detectors.5 Unlike spin coating, drop casting allows for film formation under quiescent casting conditions. During the transition from solution to final film structure of drop cast organic electronic thin film, morphological evolution is key to understanding and controlling the internal structures of final films, e.g., by changing solvents. Therefore, in order to obtain a clearer and deeper understanding of the correlation between structure and performance of P3HT and P3HT:PCBM thin films, we need to understand the kinetics of structure evolution and crystallization of P3HT and PCBM during the solvent evaporation process. With this goal in mind we examine the basic understanding of transient crystal kinetics of drop cast active layer of P3HT:PCBM during film drying in relation to that of the individual P3HT and PCBM components. In-situ study enables us to capture the kinetics of morphology evolution frame by frame on the order of seconds by grazing incidence wide angle X-ray scattering (GIWAXS) images during the drying process of drop cast films. In this way P3HT and PCBM crystal growth kinetics during this process is directly probed. This understanding has direct repercussion for controlling structure evolution in film-drying process during mass production of organic electronics, such as, ones produced by spraying of solvents, for example, inkjet printing.34
Electronic grade, regioregular poly(3-hexylthiophene-2,5-diyl) (P3HT) (catalogue number #4002-EE) was purchased from Rieke Metals, Inc., with molecular weight (Mw) was 50K-70K with regioregularity of 91% to 94%. Also, 99.5% pure Phenyl-C61-butyric acid methyl ester (PCBM) was purchased from nano-c, Inc. Three solvents, chloroform, benzene and tetrahydrothiophene (THT), were carefully selected for this research based on their relative solubilities of P3HT and PCBM. P3HT and PCBM are both fully soluble in chloroform, whereas benzene is only a relatively good solvent for PCBM and THT for P3HT. Solutions of 0.5% P3HT, 0.5% PCBM, and 0.5%:0.5% P3HT:PCBM by weight were prepared in each solvents. Polymer solutions were sonicated in a 50??C hot water bath followed by shaking in a vortex shaker. GIWAXS experiments were conducted at the D1 beam line of Cornell High Energy Synchrotron Source (CHESS). Samples were placed inside a chamber with two Kapton windows for x-ray to pass through from source to the sample to detector (see Figure 4-1(a)). The Kapton films were 30 micron thick each with a density of 1.43 gm/cm3. The wavelength of the X-ray is 1.1555??. Given this data, the transmittance of x-ray is calculated to be 98%. Experiments were conducted slightly above the critical angle of the film (0.15??) at the first waveguide resonance angle value. Thus we get the information of the through films. Experimental log file data sorting was done by a ‘R’ statistical programming code to extract the relevant information for data analysis and image sequencing. GIWAXS data were processed using a matlab code that enables us to take into account the scattering volume of the polymer exposed to the X-ray beam. P3HT d-spacing and grain size was calculated from the (100) peak information (see Figure 4-1b) using Bragg’s law and Scherrer equation, respectively. Full width half max of the (100) peak was calculated using Voigt fit of OriginPro 9.0 software. Please note that two circular rings just above and below (100) are due to kapton window. Film topography was scanned by a Bruker Corporation’s ‘Dimension Icon’ atomic force microscope (AFM) in standard tapping mode. Samples were further analyzed by scanning electron microscopy (SEM) utilizing Japan Electron Optics Laboratory (JEOL)-7401 field emission scanning electron microscope. Thermal gravimetric analysis (TGA) was conducted on the films by scraping the films, formed during the drying process of drop cast solutions under ambient condition for two hours, in order to determine the residual solvents in the films by TA instrument’s Q50 model TGA machine with 10??C per minute ramping of temperature up to 700??C. Particle and domain sizes were determined by ImageJ software, developed at the National Institutes of Health (NIH).
Due to chemical incompatibility of the constituent polymer blocks in a block-copolymer, it can self-assemble in an ordered fashion into various morphologies with domain sizes as small as a few nanometers in thin films.1’5 Domain size, shape and structure can be exquisitely controlled by the constituent block structures. This gives rise to the potential applications of such systems in functional electronic devices such as solar cells, high-density magnetic storage, nano-capacitors, etc.6’8 For example, it has been proposed that cylinder forming block copolymer thin films, where the two different domain phases consist of electron donors and acceptors, can be used as model systems for efficient charge separation and transportation in ordered bulk heterojunction solar cells.9 In this regard, regular patterns of block copolymers can also be used as partitioning templates for electron accepting nanoparticles in ordered arrays.10’17 Likewise, it is also possible to incorporate additives such as functional nanoparticles into the block copolymer and have the capability to potentially controllably partition the nanoparticles into specific domains with the block copolymer, or remain uniformly dispersed and modify the overall film property such as its dielectric constant for high energy density block copolymer capacitors. Controllable dispersion of nanoparticles into polymer matrices is thus important to exploit the full potential of nano scale ordered polymer nanocomposite thin films. Cooperative and controllable interaction of block copolymer and nanoparticle often requires meticulous engineering and proper choices of nanoparticles and block copolymers.18’26 Therefore much effort has been expended on functionalizing and compatibilizing polymers and nanoparticels for selective segregation of nanoparticles into a preferential phase of a block copolymer.18 In this article we introduce a novel and facile mixed solvent strategy to preferentially incorporate nanoparticles overall into the systems, or as needed, into a selected block copolymer phase. This technique does not require modification of the polymer or nanoparticles that may influence the intended functionally of the composite structure.
One of the well-studied block copolymer systems is polystyrene-b-poly(ethylene oxide) (PS-b-PEO) due to its unique combination of a hydrophobic and a hydrophilic block. Notably, films of PS-b-PEO cast from solvent can readily produce robust vertically ordered cylindrical morphology6,19, potentially important for nanotechnology applications such as efficient membranes, capacitors and solar cells. In this paper, we critically examine various aspects related to the synergy of PS-b-PEO block copolymer self-assembly morphology when blended with an electron accepting nanoparticle, phenyl-C61-butyric acid methyl ester (PCBM), commonly used for polymer solar cells or modifying dielectric properties. Atomic force microscope (AFM) images are analyzed to examine several key morphological aspects of solvent cast PS-b-PEO block copolymer thin films filled with PCBM up to significantly high loading levels from 0 to 50 wt %. Loading of block copolymer films with electron accepting nanoparticles is particularly relevant for solar cell applications at 40-50 wt. % loading levels.27
Previous studies have reported that the relative radius of gyration of polymer versus the nanoparticle can affect dispersion and selectivity of nanoparticle into block copolymers.25,26 The radius of gyration, Rg of a polymer can be calculated for an unperturbed linear chain using the formula Rg2=’h^2 ‘_0/6, where ‘h^2 ‘_0 is mean square of end-to-end distance of an unperturbed polymer chain. Values of ‘h^2 ‘_0/M for poly(ethylene oxide) and polystyrene are 0.8 and 0.43 (A ) ??^2 mol/g respectively28, where M is the molecular weight in g/mol. Consequently, for poly(ethylene oxide) with M=6,500 g/mol the value of RgE is 29.4 A ?? and for polystyrene with M=20,000 g/mol the value of RgS is 37.9 A ??. The Rg of the PS-b-PEO block copolymer can be calculated by4 Rgb2= RgE2+ RgS2 which give an Rgb value of 48 A ??. Hence, the radius of gyration of the PCBM nano-particle ~ 1.06 nm (10.6 A ??) is much smaller compared to the radius of gyration of either of the blocks individually. In such case, the enthalpic contribution to the dispersion process due to creation of polymer-nanoparticle interface is less significant than the particle translational entropy.25
Transmission electron microscopy (TEM) measurements do not have sufficient contrast required to distinguish the PCBM molecules from either of the block copolymer components, and small angle neutron scattering (SANS) of multilayered stacked films would be potentially required to measure changes in molecular dimensions of the block copolymer, which is outside the scope of these studies. Our study shows that changes in molecular dimensions due to PCBM incorporation are equally reflected in changes in domain spacing, domain dimensions and inter-domain spacing. During the spin coating process the fast adsorption kinetics of the small molecule, PCBM, as compared to that of a slowly ordering macromolecular block copolymer, PS-b-PEO, results in its segregation at the substrate interface of the thin film. However, in this research utilizing a proper pair of mixed casting solvent in appropriate composition, we show that PCBM can be kinetically trapped in the PS phase. PCBM in excess of what can be incorporated through these molecular level changes of block copolymer domains are accommodated by changes in block copolymer film thickness or locate them outside of the film structure at the film substrate interface. Specifically, the contribution to changes of in-plane dimensions affect the cylinder diameter and cylinder-to-cylinder spacing, while PCBM amount in excess of what can be accommodated in increase of in-plane dimensions are expended as out of plane or increase of film thickness or as aggregates at top and bottom of film depending on solvent casting conditions. Thus, this approach leads to formation of self-adjusting nanocomposite ordered block copolymer thin film system, with the ability to accommodate high nanofiller loading levels without losing basic block copolymer ordered microstructure. In the present study, we rely upon atomic force microscope (AFM) and grazing-incidence small-angle X-ray scattering (GISAXS) to elucidate the morphology of the film structure and changes with introduction of the PCBM nanoparticles. These include measurements of changes in cylinder diameter, center-to-center distance of two cylinders, film roughness and fast Fourier transform (FFT) for analysis of long range order. Film thickness, obtained through AFM scratch tests, determined how nanoparticles contributed to in-plane dimensions versus out of plane thickness.