Article pubs.acs.org/Macromolecules
Electrical, Mechanical, and Structural Characterization of SelfAssembly in Poly(3-hexylthiophene) Organogel Networks Gregory M. Newbloom,† Katie M. Weigandt,† and Danilo C. Pozzo* Department of Chemical Engineering, University of Washington, Box 351750, Seattle, Washington 98195-1750, United States S Supporting Information *
ABSTRACT: An electrically percolated network structure of conjugated polymers is critical to the development of organic electronics. Herein, we investigate the potential to rationally design an interconnected network of conjugated polymers using the gelation of poly(3-hexylthiophene) (P3HT) as a model system. The three-dimensional network structure is evaluated through small-angle neutron scattering (SANS) and ultrasmall-angle neutron scattering (USANS). The analytical models used for data fitting provide relevant structural parameters over multiple length scales. Structural parameters include the fiber cross section (height and width), the specific surface area, and the network density (i.e., fractal dimension). Simultaneous rheological and conductivity measurements also provide insight into the development of the mechanical and electrical properties of organogels and allow us to propose a detailed gelation mechanism for P3HT. The fiber shape is found to be relatively independent of the solvent type, but P3HT organogels show distinct differences in conductivity, which can be directly linked to differences in the branching network structures. These results suggest that the gelation of fiber-forming conjugated polymers offers an excellent platform for designing electrically percolated networks that can be used for structural optimization in organic electronic devices.
1. INTRODUCTION Conjugated polymers are a keystone material for developing the next generation of economical electronic devices. The potential for low-temperature and solution-based processing of semiconducting polymers makes it possible to manufacture inexpensive, large area devices directly on a variety of substrates.1,2 These unique features are embraced in the design of organic light-emitting diodes (OLEDs) that are now being incorporated into high-definition electronic displays and organic field-effect transistors (OFETs).3 Organic photovoltaic (OPVs) devices, also based on conjugated polymers, offer a lower-cost alternative to traditional solar cell technologies. OPVs are also economically well-positioned to capture a significant fraction of the available solar energy (120 000 TW/year globally).4,5 However, the overall efficiency and stability of these devices must still be improved before they can make a significant impact.2,6 Engineering a network structure that is interconnected and spans the gap between electrodes is one of the grand challenges of designing efficient electronic devices from conjugated polymers.2,3,7,8 In the case of OPVs, the structure must satisfy important constraints such as being interconnected for effective charge transport, having adequate thickness for photon absorption and having domains that are smaller than the exciton diffusion distance.6,9−11 Unfortunately, the morphology of conjugated polymers is sometimes loosely characterized in the literature because of difficulties involved with probing the bulk structure of opaque thin films and because of their multiscale nature. Therefore, instead of designing the structure to achieve good performance, it is more common to optimize © 2012 American Chemical Society
device properties through changes to the processing parameters. This approach leads to a lack of accurate structure− property relationships that describe the correlation between network structure and device performance.12 Achieving a specific network morphology for use in organic electronics also becomes increasingly difficult when structural development is strongly coupled to processing conditions. OPVs, for example, are typically produced by coating a solution of dissolved p-type and n-type materials onto a conductive substrate.13 This leads to a film that may or may not phase separate into the desired structure depending on parameters related to coating conditions and post-processing annealing (e.g., temperature, solvent vapor pressure, coating speed, etc.).1,14−16 Given the operating constraints of OPVs, the ideal structure of the photoactive layer should be a bicontinuous network that vertically bridges the anode and cathode, with nanoscale phase segregation of p-type and n-type materials.1,6−8,10,15,17,18 It is difficult to corroborate that the ideal structure has been achieved because there are few analytical techniques that can directly and quantitatively probe the multiscale structural features of these materials. The top surface of thin films is commonly studied with microscopy techniques, such as atomic force microscopy (AFM), even though it is unlikely that this surface structure will match the bulk morphology underneath.16,19−21 Received: November 23, 2011 Revised: February 28, 2012 Published: April 6, 2012 3452
dx.doi.org/10.1021/ma202564k | Macromolecules 2012, 45, 3452−3462
Macromolecules
Article
2. EXPERIMENTAL METHODS
An alternative to adjusting the morphology of the network via post-processing is to directly build the desired network structure prior to coating. The self-assembly of conjugated polymers prior to deposition could ensure the formation of a threedimensional, stable, and interconnected network that should be beneficial for organic electronic devices. Gelation of conjugated polymers in organic solvents is a very common phenomenon. Poly(3-alkylthiophenes) (P3ATs), poly(9,9-dioctylfluorene-2,7diyl) (PFO), and poly[2-methoxy-5-(2-ethylhexyloxy)-1,4-henylenevinylene)] (MEH-PPV) have all been shown to form organogels in organic solvents.22−34 The propensity for π-stacking among conjugated polymers frequently leads to growth of long, semicrystalline fiber structures that can extend for several micrometers.2,22,24 In particular, the fibrillar structure of P3HT has been shown to be beneficial for charge transport.12,22 Malik and coworkers further demonstrated that the interconnectivity of P3HT fibers in the gel phase results in increased charge mobility when compared to traditional spin-cast films with the same composition.29 The ability to engineer the network phase of fiber-forming conjugated polymers could provide a route toward achieving optimized active layer morphologies. In the past, conjugated polymer gelation has been largely regarded as a hurdle to overcome for the benefit of storing and processing electronic inks.22,28,35 The elastic nature of interconnected networks makes them difficult to process using traditional coating techniques. To overcome this problem, our lab recently demonstrated the possibility of processing P3HT organogels directly from aqueous dispersions through the preparation of microgel particles. The feasibility of this approach was demonstrated by fabricating prototype solar cells.36 Other researchers have also demonstrated the feasibility of using conjugated polymer organogels in OPVs through freeze-drying methods and through organogel sonication to form homogeneous solutions.25,33 These advances in gel processing now make it possible to utilize conjugated polymer gel networks in real applications. Additionally, fundamental research in the field of small-molecule organogels suggests that network structures can be effectively tuned by controlling the solubility of the gelator molecules.37,38 Here we present a systematic, multiscale structural analysis of P3HT organogels formed at different concentrations and in various aromatic solvents for the purpose of informing the development of percolated networks for organic electronics. Utilizing multiscale analysis techniques is critical for the development of robust structure−property relationships because structural differences can occur simultaneously on multiple length scales.12 Small-angle scattering of X-rays and neutrons are promising techniques to nondestructively probe morphology in these materials.12,14,20,23,36,39−41 Here, we use small-angle neutron scattering (SANS) and ultra small-angle neutron scattering (USANS) to evaluate the structure of P3HT gels over length scales of 1− 10 000 nm. Scanning transmission electron microscopy (sTEM) is also utilized to complement the scattering analysis and to directly observe the network structures. Simultaneous rheology and conductivity measurements as well as UV−vis spectroscopy are utilized to further understand the complex self-assembly mechanism and to develop structure−property relationships. Conductivity measurements confirm that structural variations can significantly impact charge transport characteristics in organogels. These results demonstrate that it is possible to engineer a network structure for specific organic electronic applications using gelation as a general platform.
2.1. Materials. Two lots of Sepiolid P200 poly(3-hexylthiophene) (P3HT) were used as received from Rieke Metals (Lincoln, NE). The molecular weight (Mw) and polydispersity index (PDI) of each batch were characterized using gel permeation chromatography (GPC) relative to polystyrene standards. Batch 1 (Lot # 31-10-2008) had Mw = 24 060 g/mol and PDI = 2.0. Batch 2 (Lot # 22.10.2009) had Mw = 16 770 g/mol and PDI = 1.6. The regioregularity of both batches was reported by the manufacturer to be greater than 98%. The results presented in this paper are primarily from batch 2 samples, unless otherwise indicated. Some results from batch 1 are also included to illustrate effects originating from batch-to-batch variations. Hydrogenated solvents p-xylene, toluene, and benzene were purchased from Sigma-Aldrich (St. Louis, MO) and used as received. Deuterated solvents were used to increase the scattering contrast and to decrease the incoherent scattering background in SANS and USANS experiments. d10-p-Xylene (D > 98%), d8-toluene (D > 99.5%), and d6benzene (D > 99.5%) were purchased from Cambridge Isotopes (Andover, MA) and were used as received. 2.2. Sample Preparation. Samples were prepared by adding 5− 50 mg/mL P3HT to p-xylene, benzene, or toluene and then heating to 80 °C until the polymer was fully dissolved. The polymer is considered fully dissolved when the solution is bright orange, and there are no visible signs of solid polymer (black). The polymer solution was then removed from heat and allowed to gel at a lower temperature for 24 h unless otherwise noted. 2.3. UV−vis Spectroscopy. The fraction of solids in the gels was determined via UV−vis spectroscopy after allowing gels to selfassemble for a specific length of time. At this point, the gels were fractured through vigorous shaking in the presence of a small controlled amount of extra solvent (∼2 mL). The fractured P3HT gels were then centrifuged at 21100g for 1 min, and the polymer concentration in the supernatant was measured via UV−vis spectroscopy. UV−vis absorbance spectra were obtained using a Thermo Evolution 300 UV−vis spectrophotometer at wavelengths between 350 and 750 nm. The concentration of the dissolved polymer fraction was determined from the maximum absorption at 450 nm using Beer’s law with experimentally determined extinction coefficients of 50.3, 50.2, and 51.6 mL/mg cm for dissolved P3HT in p-xylene, toluene, and benzene, respectively. The fraction of polymer in the solid phase was calculated from a simple mass balance. 2.4. Rheology and Conductivity. Rheological measurements were made using an Anton Paar Physica MCR301 stress-controlled rheometer. A parallel plate geometry with a 50 mm diameter and 0.4 mm gap was preheated to 50 °C prior to loading the hot polymer solution in the liquid state. Once loaded, the solution was rapidly cooled to induce gelation. After the solution had solidified (∼1 min) a fluorinated nonconductive oleophobic liquid, Fomblin Y 25/6, was added to the outer edge of the sample geometry to fully suppress the evaporation of the organic solvent. We have experimentally confirmed that the addition of Fomblin does not impact the rheology of P3HT organogels. The gel was then reheated to 80 °C for 2 min to completely redissolve the gel before the experiments were performed. Small-amplitude oscillatory strain measurements (γ = 0.25% and f = 1 Hz) were used to track changes in mechanical properties during gelation. Simultaneous conductivity measurements were also made for some experiments. This was accomplished through the use of a special rheometer cell designed for simultaneous dielectric measurements. This cell is electrically isolated from the frame and the electronics of the rheometer. Electrical contact is created between a gold wire and the upper tool geometry (stainless steel) and also from the bottom tool (stainless steel). The conductivity and other electrical properties are then measured across the sample gap with the parallel plates of the rheometer also acting as electrodes. The use of stainless steel plates ensures that corrosion does not impact the electrical measurements. The ac conductivity measurements were made using an Agilent e4980a Precision LCR meter at a frequency of 2 kHz and a small perturbation voltage (200 mV) in the linear impedance region. Although the 3453
dx.doi.org/10.1021/ma202564k | Macromolecules 2012, 45, 3452−3462
Macromolecules
Article
contact between the upper tool and the gold wire causes a constant frictional resistance to rotation, its effect is only noticeable when the samples are in the fluid dissolved state. Experiments were reproduced both with and without the gold wire to fully characterize its effect on the measured mechanical properties. 2.5. Scanning Transmission Electron Microscopy. Small quantities of the organogels were transferred to copper TEM grids with Formvar support films by briefly placing the grids in contact with the surface of the gel samples without applying pressure. In this way, only a thin-film of organogel is transferred to the TEM grid. Because of the very thin nature of the samples, the aromatic solvents evaporate under ambient conditions. TEM images were obtained with a FEI Tecnai G2 F20 transmission electron microscope operating at 120 kV in scanning TEM (sTEM) mode. 2.6. Small-Angle Scattering. Ultra small-angle neutron scattering (USANS) and small-angle neutron scattering (SANS) experiments were performed at the NIST Center for Neutron Research in Gaithersburg, MD. SANS measurements were performed on the NG3 and NG7 instruments using standard configurations to cover a broad q-range (0.001 < q < 0.5 Å−1).42 Dissolved polymer solutions were loaded into demountable cells with quartz windows (1 mm path length) and allowed to gel for a minimum of 24 h prior to measurement. The 2D scattering profiles of the sample were placed on an absolute scale by measuring the empty beam flux and correcting for background and empty cell scattering.43 The 2D scattering profiles were averaged about the azimuthal angle, and the scattering intensity (I) was plotted against the scattering vector (q). USANS measurements were performed at NIST on the BT5 perfect crystal diffractometer extending the q-range to 4 × 10−5 Å−1.44 The data were reduced and desmeared using the NIST Igor-based algorithms.43 Model fitting was performed using the NIST Igor-based analysis algorithms as well as the DANSE SansView software.43,45
Figure 1. Temperature-dependent oscillatory rheology of a 30 mg/mL P3HT organogel formed and redissolved in p-xylene. The temperature sweep rate is 2 °C/min. When the temperature reached 20 °C, gelation was allowed to proceed to completion for 2 h prior to the heating ramp. Inset: frequency sweep of the fully developed P3HT organogel.
electrical properties during the gelation process. Figure 2a tracks the gelation and dissolution of a sample with 30 mg/mL of P3HT dissolved in p-xylene with simultaneous mechanical and electrical property measurements. For clarity, the mechanical data in Figure 2a are shown in terms of the complex modulus (G*), which is related to the elastic and viscous moduli (G′ and G″) by the equation
3. RESULTS 3.1. Gelation of P3HT. Poly(3-hexylthiophene) (P3HT) can be solubilized in many aromatic solvents when the temperature is increased. When the solvent quality is lowered, often through temperature reduction, the polymer spontaneously selfassembles into semicrystalline nanofibers.12,22,46 When the polymer is in dilute concentrations, fibers grow individually or in discrete fiber bundles (colloidal networks). On the other hand, if the concentration is larger than ∼5 mg/mL, fiber formation at low temperatures leads to gelation. Rheology was utilized to track the mechanical evolution of a 30 mg/mL P3HT solution in p-xylene during the sol−gel transition (both gelation and redissolution). Figure 1 shows the temperature-dependent elastic (G′) and viscous modulus (G″) of a P3HT organogel formed in p-xylene. As the temperature of the polymer solution is reduced from 80 to 20 °C (at a rate of 2 °C/min), the sample remains a viscous liquid (no elastic component) until about 26 °C when a clear sol−gel transition occurs. At this temperature the elastic response due to the formation of the percolating network overwhelms the viscous response, which is indicated by the crossover (G′ = G″) of the elastic and viscous moduli. Although gelation occurs rapidly, when the gel is held at 20 °C, the gel continues to stiffen as a function of time for several hours. This steady increase in the elastic modulus indicates that there is ongoing gelation and slow development of the fiber network. The inset frequency sweep in Figure 1 confirms that P3HT is a gel across a wide range of frequencies. Upon reheating (at a rate of 2 °C/min), P3HT shows significant thermal hysteresis prior to full dissolution at about 61 °C. The thermal hysteresis is likely due to the existence of a different dissolution mechanism with slower kinetics for fiber breakup. Simultaneous rheology and conductivity measurements allow for direct correlation of the evolution of the mechanical and
G* =
(G′)2 + (G″)2
(1)
The complex modulus is effectively equivalent to the elastic modulus for P3HT in the gel phase because G′ ≫ G″. However, it is also important to note that the complex modulus prior to gelation (i.e., in the soluble phase) is significantly affected by the friction between the upper geometry shaft and the gold contact that is necessary for the electrical connection, depicted in Figure 2b. The data shown in Figure 2a suggest that the increase in conductivity develops prior to the increase in G*. However, this could also be due to the background friction that makes the measurement less sensitive to the initial changes in the elastic modulus. Nevertheless, it is clear that gelation occurs around 26 °C and that it corresponds to an increase of more than 2 orders of magnitude in the conductivity in addition to the observed increase in elasticity (G*). When self-assembly is allowed to proceed at 20 °C for an additional 2 h, both the elastic modulus and the conductivity continue to increase, indicating a slow evolution of the organogel structure. A figure highlighting this increase can be found in the Supporting Information (Figure S4). Figure 2a also shows that upon reheating there is significant thermal hysteresis in the mechanical and electronic properties of the gel prior to complete dissolution. This is particularly evident for the conductivity, which shows a significant increase in magnitude upon reheating. This response is due to the Arrhenius behavior of conductivity that is described by eq 2: σ = Ae[−EA / k bT ]
(2)
Here, σ is the conductivity, A is a proportionality constant, EA is the activation energy, kb is Boltzmann’s constant, and T is the 3454
dx.doi.org/10.1021/ma202564k | Macromolecules 2012, 45, 3452−3462
Macromolecules
Article
Figure 2. (a) Temperature-dependent, simultaneous oscillatory rheology and ac conductivity of a 30 mg/mL P3HT organogel formed and redissolved in p-xylene. The temperature sweep rate is 2 °C/min. When the temperature reached 20 °C, gelation was allowed to proceed to completion for 2 h prior to the heating ramp. (b) Schematic of the experimental configuration for simultaneous rheology and ac conductivity measurements.
temperature. When the initial increase in conductivity is fit to the Arrhenius equation, we find that EA = 300 ± 79 meV. This value is very similar to what others have reported for solid P3HT films.47 3.2. Properties of P3HT Organogels. Time-dependent simultaneous rheology and conductivity measurements were also performed to probe the development of mechanical and electrical percolation in networks. Rheology is only sensitive to fiber interactions that contribute to the transduction of mechanical stress between the two plates of the rheometer. Meanwhile, conductivity measurements are only sensitive to fiber paths that electronically bridge the electrodes (Figure 3a).
Thus, a single network can be characterized by two fractal dimensions resulting from these two different macroscopic properties, and we will refer to them as DfE for electronic percolation and DfM for mechanical percolation. Figure 3b depicts an example of two different network structures that might have very similar mechanical properties but would have drastically different electrical properties. Liu and co-workers have demonstrated the use of rheology to determine the fractal dimension (Df) of branching fibrillar organogels by using the Avrami theory to characterize the kinetics of gelation.48−50 This powerful analysis relies on some basic assumptions including a constant fiber growth rate and branching frequency. Under isothermal conditions, the total extent of gelation is characterized by XG XG(t ) = 1 − exp( −k(t − tgel)Df )
(3)
where k is a constant proportional to the rate of fiber growth, t is the elapsed time, and tgel is the initial gelation time. The extent of gelation (XG) is related to measurable properties such as the complex modulus (G*) and the conductivity (σ) using XG(t ) =
* G*(t ) − G bkg σ(t ) − σ bkg or XG(t ) = * − G bkg * σmax − σ bkg Gmax
(4)
where G*(t) is the time-dependent complex modulus, Gbkg * is the complex modulus prior to gelation, Gmax * is the maximum complex modulus associated with kinetic equilibrium, σ(t) is the time-dependent conductivity, σbkg is the conductivity of the solution prior to gelation, and σmax is the maximum conductivity of the gel. To ensure that gelation was monitored at a constant temperature and from identical initial states, the fully dissolved polymer was rapidly cooled from 80 to 20 °C at a rate of 22.5 °C/min. The gelation process was then monitored as a function of time when the test solution reached 20 °C. The time required to reach the test temperature (
