July 12, 2011
C 2011 American Chemical Society
Competition between Fullerene
Aggregation and Poly(3-hexylthiophene)
Crystallization upon Annealing of Bulk
Heterojunction Solar Cells
Wei-Ru Wu,†U-Ser Jeng,†,* Chun-Jen Su,†,* Kung-Hwa Wei,‡Ming-Shin Su,‡Mao-Yuan Chiu,‡
Chun-Yu Chen,†Wen-Bin Su,†Chiu-Hun Su,†and An-Chung Su§
†National Synchrotron Radiation Research Center, 101 Hsin-Ann Road, Hsinchu Science Park, Hsinchu 30076, Taiwan,‡Department of Materials Science and
Engineering, National Chiao Tung University, 1001 Ta Hsueh Road, Hsinchu 30050, Taiwan, and§Department of Chemical Engineering, National Tsing Hua University, Hsinchu 30013, Taiwan
M
orphology of the active layer ofbulk heterojunction (BHJ) thinfilms, comprising donor and acceptor components, is one of the critical factors in solar cell performance optimization.1,2 An ideal morphology for BHJ thin-film solar cells features in phase-separated nanodo-mains ca. 10 nm in size to facilitate exciton dissociation and charge transport;3also
re-levant is the connectivity of these nano-domains in the respective phases for charge transport.14In the past decade, BHJ thin-film processing parameters, such as compo-sition, annealing temperature/time, casting solvent, andfilm thickness, have been in-tensively studied. As a result, power conver-sion efficiency (PCE) in excess of 4% could be obtained with the popular BHJ blend of regioregular poly(3-hexylthiophene) (P3HT) and [6,6]-phenyl-C61-butyric acid methyl ester
(PCBM).57In general, optimized device per-formance of this BHJ blend can be achieved withfilms of PCBM/P3HT weight ratio c = 0.81.0 and ca. 100 nm in thickness after thermal annealing at 120160 C for 15 30 min; processing details determine that these parametersfluctuate in a certain range.7 In correlating performance and morphol-ogy, cross-sectional transmission electron microscopy (TEM)3and electron tomography8 revealed mutually intercalated nanodo-mains of the two constituted components in the P3HT/PCBM thin-film solar cells; cor-respondingly, grazing-incidence X-ray dif-fraction (GIXRD) indicated highly oriented P3HT lamellae (ca. 1020 nm in domain size), with chain stacking perpendicular to the film substrate.5,9 Compared to the relatively well-elucidated structural
characteristics of P3HT crystallites, PCBM aggregation behavior upon annealing in the compositefilm is much less clarified.10,11 This situation is partly due to the fact that the commonly used tool of X-ray diffraction revealed very limited structural information of noncrystalline PCBM aggregates; because of the contrast, PCBM aggregates could be clearly identified by TEM after full cry-stallinity development upon extensive thermal treatment beyond optimized film processing conditions. As a result, the pro-posed PCBM aggregate size for optimized
* Address correspondence to usjeng@nsrrc.org.tw, su.cj@nsrrc.org.tw.
Received for review March 22, 2011 and accepted July 12, 2011. Published online 10.1021/nn2010816
ABSTRACT Concomitant development of [6,6]-phenyl-C61-butyric acid methyl ester (PCBM)
aggregation and poly(3-hexylthiophene) (P3HT) crystallization in bulk heterojunction (BHJ) thin-film (ca. 85 nm) solar cells has been revealed using simultaneous grazing-incidence small-/wide-angle X-ray scattering (GISAXS/GIWAXS). With enhanced time and spatial resolutions (5 s/frame; minimum q ≈ 0.004 Å1
), synchrotron GISAXS has captured in detail the fast growth in size of PCBM aggregates from 7 to 18 nm within 100 s of annealing at 150C. Simultaneously observed is the enhanced crystallization of P3HT into lamellae oriented mainly perpendicular but also parallel to the substrate. An Avrami analysis of the observed structural evolution indicates that the faster PCBM aggregation follows a diffusion-controlled growth process (confined by P3HT segmental motion), whereas the slower development of crystalline P3HT nanograins is characterized by constant nucleation rate (determined by the degree of supercooling and PCBM demixing). These two competing kinetics result in local phase separation with space-filling PCBM and P3HT nanodomains less than 20 nm in size when annealing temperature is kept below 180C. Accompanying the morphological development is the synchronized increase in electron and hole mobilities of the BHJ thin-film solar cells, revealing the sensitivity of the carrier transport of the device on the structural features of PCBM and P3HT nanodomains. Optimized structural parameters, including the aggregate size and mean spacing of the PCBM aggregates, are quantitatively correlated to the device performance; a comprehensive network structure of the optimized BHJ thinfilm is presented.
KEYWORDS: GISAXS . GIWAXS . structural kinetics . bulk heterojunction solar cells .
charge mobility
P3HT/PCBM thin-film solar cells scattered in a wide range of 560 nm.9,1214
Very recently, structural kinetics of polymer/fuller-ene derivative thin films in thermal annealing has received increasingly more attention for better under-standing of the local phase separation and nanodo-main formation mechanism of the related solar cells. Using in situ GIXRD, Shin et al. showed that the P3HT crystallization process in a P3HT/PCBM blend was completed in 25 min during annealing at 140 C.15
Using the same methodology of in situ GIXRD, Verploe-gen et al. demonstrated that PCBM crystalline reflec-tions slowly appeared within 20 min of annealing above 150C.16In contrast, ex situ small-angle neutron
scattering (SANS) revealed a much faster morphology development for a similarfilm upon short-term ther-mal annealing of a few seconds.13Combining SANS, TEM, and GIXRD observations, Chen et al.13suggested that the phase separation length scale was incommen-surate with spinodal decomposition for polymer mixtures and hence more likely a result of competition between P3HT crystallization and PCBM diffusion during thermal annealing: the ratio of PCBM diffusion coefficient to P3HT crystallization rate conceivably determined the limited P3HT crystal size during local phase separation that led to bicontinuous P3HT- and PCBM-rich nanodomains in the BHJ thin-film solar cell. Retrieving the fast PCBM aggrega-tion kinetics before crystallizaaggrega-tion in such composite thin films, however, would require time-resolved (in seconds) grazing-incidence small-angle X-ray scattering (GISAXS) or SANS for large noncrystalline PCBM aggregates, which was not addressed in previous studies.
Basic understanding of the kinetics of PCBM aggre-gation and P3HT crystallization and their correlations in the morphological development of P3HT/PCBM com-posite films upon thermal annealing would provide hints to the optimal processing of future BHJ solar cells with varied components of fullerene derivatives or other conjugated polymers.13In this study, we employ simultaneous and time-resolved GISAXS and grazing-incidence wide-angle X-ray scattering (GIWAXS) to capture the kinetics of PCBM aggregation and P3HT crystallization in the corresponding BHJ thin films during in situ thermal annealing.15,16With improved
time and spatial resolutions (5 s/frame; minimum scattering wavevector q≈ 0.004 Å1), synchronized GISAXS and GIWAXS have captured the concomitant growth of the PCBM aggregates (from 7 to 18 nm in size) and the P3HT crystallites (from 7 to 12 nm), within 100 s of annealing at 150C. We compare the growth kinetics of the nanodomains of the two components based on the Avrami exponents and rate constants extracted from the evolutions of the corresponding structural parameters. Parallel measurements on the hole and electron mobilities of the corresponding BHJ thinfilms under similar annealing conditions further demonstrate synchronized developments of charge
mobility and nanodomain formation. Quantitative cor-relations drawn between the morphology and charge mobility are discussed in terms of the charge transport behavior and device performance of the BHJ solar cell. With the new approach of simultaneous/time-resolved GISAXS/GIWAXS and the structural analysis for kinetics parameters, we have made newfindings cen-tering on the critical aggregation behavior of PCBM in the P3HTPCBM blend; these include (1) the thermal equilibrium aggregation sizes and the growth kinetics of PCBM, (2) faster PCBM aggregation kinetics than that for P3HT crystallization (opposite to previous specula-tions), (3) a deduced activation energy for PCBM aggrega-tion that addresses the phase separaaggrega-tion mechanism of P3HT and PCBM (for the interacted or bicontinuous P3HT and PCBM nanodomains) in more appropriate terms of nucleation and growth, rather than the spontaneous spinodal decomposition (requiring no activation energy) that has been suspected for some time, and (4) closely and quantitatively correlated PCBM aggregation size and electron mobility that has long been postulated. The present results not only clarify the mechanistic origins of morphological development in P3HT/PCBM films but also provide insights to the performance optimization of polymer/fullerene organic thin-film devices in general. RESULTS AND DISCUSSION
Simultaneous GISAXS/GIWAXS. With the experimental setup shown in Figure 1, we could measure time-resolved GIWAXS and GISAXS patterns simultaneously for the P3HT/PCBM composite films (c = 1.0) in situ annealed at 150 C (cf. movie in the Supporting Information). The GISAXS and GIWAXS patterns, repre-sentatively shown in Figure 2, exhibited concomitant and substantial changes during the heating process (ca. 40 s) from an ambient temperature to 150C; the scattering patterns were quickly saturated within the subsequent 100 s of isothermal annealing, revealing fast morphology development of the composite film. Specifically, the GIWAXS patterns evidenced successively enhanced P3HT lamellar peaks up to the third order (Figure 2ae) in the vertical direction, corresponding
Figure 1. Schematic of the setup for synchronized GISAXS/ GIWAXS, with the beam incident angleR and the scattering anglesβ and Φ in the out-of-plane (qz) and in-plane
direc-tions (qx). The GIWAXS detector plane was normal to the incident beam; the detector was titled 45 out of the horizontal plane to cover scattering in theqzandqx
directions.
to edge-on P3HT lamellae with the lamellar stacking direction perpendicular to the substrate surface. The concomitant growth of the (100) peak in the in-plane direction (Figure 2ae) corresponded to the develop-ment of face-on lamellae domains with the lamellar stacking oriented parallel to the film surface. Mean-while, the emerged GIWAXS halo at q≈ 1.4 Å1during the annealing process revealed formation of a short-range packing of PCBM after aggregation.9The PCBM aggregates also resulted in prompt and drastic in-crease of scattering intensity in the low-q region (0.0040.04 Å1) of the corresponding GISAXS
pat-terns simultaneously observed. Note that scattering in the low-q region was mainly dominated by large PCBM aggregates as elucidated in our previous study.10 Dur-ing the subsequently prolonged annealDur-ing over 1800 s, both GISAXS and GIWAXS images changed marginally and resembled those measured for ex situ annealed (at 150C for 900 s) samples. With the time resolution of 5 s, these time-resolved GIWAXS/GISAXS data revealed rich structural information, especially the mutually confined growth kinetics of P3HT and PCBM nanodo-mains in the composite film as elucidated below.
Growth of PCBM Aggregates. Figure 3a presents the GISAXS profiles of the P3HT/PCBM (c = 1.0) film selec-tively extracted from the corresponding 2D patterns along the qxdirection (at the specular beam position
qz≈ 0.018 Å1),17revealing the fast scattering
devel-opment in the low-q region 0.0040.04 Å1within the
first 60 s of thermal annealing. Furthermore, the selected GISAXS profiles (Figure 3c) for 60, 600, and 1800 s at 150C overlapped roughly, revealing quickly saturated PCBM aggregation. A broad interference shoulder at qx≈ 0.025 Å1shaped during the annealing,
corresponding to formation of a liquid-like or distorted face-centered-cubic-like packing of PCBM aggregates with a mean spacing of ca. 25 nm (detailed below with model fitting).18 For comparison, the GISAXS profile
similarly measured for a pristine P3HT film annealed at 150C for 1800 s contributed only marginally in this monitored q region, as illustrated in Figure 3c.
To obtain detailed structural evolution of the PCBM aggregates, we used the sphere form factor P(q) with the Schultz size distribution for GISAXS datafitting.14
An approximated structure factor S(q) taken from the effective one-component system of hard spheres was also included in the data fitting to account for the scattering shoulder around qx≈ 0.025 Å1from loosely
packed PCBM aggregates. Illustrated in Figure 3a are the decent datafitting results. Note that the form factor of ellipsoids or rods could notfit the GISAXS data, as well. Hence, we focus on the structural information, especially the size evolution, given by the model of polydisperse spheres. Shown in Figure 3b,d are the size distributions (in general, 2030%) and the size evolu-tion during annealing, indicating that PCBM aggre-gates grew slightly from a diameter of ca. 7 nm (as-cast film) to 10 nm during heating toward 150 C. In the subsequent 60 s of isothermal annealing, the PCBM aggregate size increased quickly from 10 to 17.5 nm, then remained about the same in the prolonged thermal annealing over 1800 s. In all of the GISAXS data fitting, we had to include a scattering term described by the DebyeBuche correlation function,19
I(qx) = (1þ qx2ζ2)2, to account for the relatively sharp
upturn scattering in the very low-q region 0.004 0.007 Å1. The fitted values for the DebyeBuche correlation lengthζ were about the same (60 ( 10 nm) for all sets of data; hence, the structural origin of this correlation length might be irrelevant to PCBM aggrega-tion. Possibly, the correlation length might associate better with the reported P3HT aggregates of nanowhis-ker or protofibril structure already formed in solutions.20 Using a model of polydisperse rods,21with a mean rod length of 69 nm, a rod diameter of 11.2 nm, and 75% polydispersity in rod length, we couldfit decently the
Figure 2. Representativefive sets of the GIWAXS (top row) and GISAXS (bottom row) images simultaneously measured for the P3HT/PCBMfilm with c = 1.0. From left to right are the images collected (a) at room temperature, (b) during the heating process (106C), and (ce) after 0, 96, and 1842 s of annealing at 150 C. The lamellar peaks (100), (200), and (300) of the P3HT crystallites are marked in (d) by thin arrows, whereas the PCBM halo (q ≈ 1.4 Å1) is indicated with a thick arrow. The arrows in the GISAXS patterns mark the successively enhanced scattering in the low-qxregion.
GISAXS profile of the pristine P3HT annealed at 150 C (Figure 3c). These results suggest that P3HT forms/ enhances lamellar stacking within thefibril-like aggre-gates upon thermal annealing, as to be delineated below. Growth Behavior of P3HT Crystallites. Respectively shown in Figure 4a,b are the evolutions of the two P3HT (100) peaks for the edge-on and face-on lamellae of the P3HT/PCBM blend film (c = 1.0) during 150C annealing. These profiles were extracted from the corresponding GIWAXS patterns along the normal-to-plane (qz) and in-plane (qx) directions. During the
heating process to 150C, the two (100) peaks shar-pened concomitantly and shifted successively from qz = 0.388 to 0.355 Å1 owing largely to thermal
expansion of the lamellae; the peaks could shift back closely to 0.374 Å1when the sample was cooled to ambient temperature after thermal annealing, as also observed for a pristine P3HT film in a similar annealing process. After the thermal annealing, the enhanced orientation of P3HT lamellae along the surface normal
direction, however, revealed straightened lamellar stacking with a slightly larger lamellar spacing of 16.8 Å than that (16.2 Å) before annealing.15 Using the Scherrer's equation,10we could obtain the lamellar domain size evolution from the corresponding (100) peak widths. The growth behaviors shown in Figure 4c indicate that the P3HT edge-on and face-on lamellae developed with comparably fast kinetics and were largely saturated within the first 100 s of 150C thermal annealing. The slightly larger domain size of 14.5 nm for the face-on lamellae than that (11.5 nm) for the edge-on lamellae is cedge-onsistent with that shown in a previous study;15the edge-on lamellae, however, dominated the P3HT crystallinity development in the composite film, as revealed by their much higher (100) peak intensity (by 8-fold) as compared to that of the face-on lamellae (cf. Figure S1 in Supporting Information).
Formation Mechanism of the Nanograins. To better un-derstand the growth behavior of the PCBM aggregates and P3HT crystallites from the structural evolutions
Figure 3. (a) Selected GISAXS profiles measured for the P3HT/PCBM film (c = 1.0) during the heating process to 150 C and the subsequent isothermal annealing within 60 s. The data arefitted (solid curves) using polydisperse spheres with the Schultz size distributions shown in (b). (c) Approximately overlapped GISAXS data collected after 60, 600, and 1800 s of thermal annealing. The data for 1800 s arefitted (solid curve) with the size distribution shown in (b). For comparison, the GISAXS data for a pristine P3HTfilm annealed at 150 C for 1800 s are also shown; the data are fitted (dashed curve) with polydisperse (in rod length) rods. (d). Corresponding mean size (D) evolution of PCBM aggregates.
revealed by the time-resolved GISAXS/GIWAXS data, we adopted the AvramiErofeev expression22,23
R(t) ¼ 1 exp[ (kt)n] (1)
where the extent of the phase volume developmentR(t) at time t was characterized by the Avrami exponent n and rate constant k, assuming no induction time in the system. For PCBM aggregation kinetics,R(t) was taken from the scattering invariant Qinv =
R
I(q)q2dq in the GISAXS q region where PCBM aggregates dominated the scattering contribution. For P3HT crystallization kinetics, R(t) was deduced from the integrated intensity of the (100) peak I100(t) normalized by its maximum (saturated) value.
Shown in Figure 5 are the obtainedR(t) evolutions for the PCBM aggregation and the P3HT crystallization (edge-on and face-(edge-on lamellae, respectively) of the P3HT/PCBM (c = 1.0) composite film (cf. Figures 3 and 4). TheseR(t) profiles could be fitted reasonably well (Figure 5a) using eq 1, with the best-fitted n and k values listed in Table 1, or presented graphically using the SharpHancock form23
lnfln[1 R(t)]1g ¼ n ln t þ n ln k (2) as given in Figure 5b.
As shown in Table 1, PCBM aggregation and P3HT crystallization have similar Avrami exponents of n≈ 1. Nevertheless, since the Avrami exponent n is contrib-uted by both primary nucleation and growth, forma-tion mechanisms of the two types of nanodomains in the composite film may not necessarily be the same.24,25 In view of the quick growth (D t0.3) of the PCBM aggregates illustrated in Figure 3d, we neglected the nucleation factor and attributed the n ≈ 1 value largely to a three-dimensional growth (D3 t0.9) of PCBM aggregates in the BHJ thinfilm.24,25 Indeed, isotropic, diffusion-controlled type of growth fits adequately the local demixing process of PCBM from the well-dispersed P3HT/PCBM background. For ideal diffusion-limited aggregation of existing particles, phase dimension grows with D t1/2and the
corre-sponding Avrami exponent contributed by size growth alone is presumably n = d/2, where d is the Euclidean dimension between 1 and 3.25The smaller growth expo-nent to the annealing time (D t0.3) of PCBM aggrega-tion than that for the ideal diffusion-limited growth process might be attributed to more restricted, hetero-geneous diffusion routes out of the mixed phase.24
Figure 4. Evolutions of the P3HT (100) peak in the (a) normal-to-planeqzand (b) in-planeqxdirections extracted from the
corresponding GIWAXS images for the P3HT/PCBM (c = 1.0) blend film during 150 C annealing. The profiles are shifted vertically in intensity for clarity. (c) Corresponding size evolutions of the respective lamellar domains in the two perpendicular orientations.
In contrast, the size growth of P3HT crystallites (L t0.06)
extracted from the data measured during the 150C annealing (Figure 4c) was more strongly hindered, espe-cially for the predominating edge-on lamellae. Hence, we attribute the Avrami exponent n≈ 1 obtained mainly as a result of constant nucleation rate of crystalline P3HT nanograins in the compositefilm.24,25
Despite the similar Avrami exponent of n≈ 1, PCBM aggregation kinetics exhibited a significantly higher Avrami rate constant of k = 0.036 s1(Table 1) than those of P3HT crystallization for edge-on (k = 0.017 s1 along thefilm normal direction) and face-on lamellae (k = 0.016 s1along thefilm surface direction). It should be emphasized that the kinetics detailed here for PCBM aggregation is much faster than that reported for PCBM crystallization.16In the absence of PCBM, crystallization
in a pristine P3HTfilm annealed at 150 C (Figure 5) exhibited a 4-fold higher rate constant of k = 0.071 s1 for the edge-on P3HT lamellae (Table 1) and a similar Avrami exponent of n = 1 as that for the P3HT lamellae in the compositefilm. These results suggest that the presence of PCBM molecules seriously hinders the nuc-leation rate (in terms of k) of P3HT lamellae but does not change the basic scheme of nucleation-controlled crystallization (in terms of n ≈ 1, i.e., constant
nucleation rate) in the P3HT-rich region upon segrega-tion. Aggregates formed by densely segregated PCBM molecules then serve to confine the growth space and orientation of P3HT, leading to better aligned (space-filling) P3HT lamellae than that in the pristine P3HT film, as revealed from the significantly sharpened (100) peak of the compositefilm (cf. Figure S1). The con-current/consequential development of P3HT crystal-line domains upon PCBM segregation then serve in turn to confine further development of PCBM aggregates. The competition between the mutually interacting growth mechanisms, rather than spinodal decomposi-tion,13,26
of the two components leads to concomitant saturation (cf. Figures 3d and 4c) at mutually limited nanodomain sizes below ca. 20 nm.
We also conducted GISAXS/GIWAXS for the compo-sitefilms isothermally annealed at different tempera-tures. It was found that the rate constant of PCBM aggregation increased strongly from k = 0.014 s1at 120C to 0.036 s1at 150C, whereas the k value for P3HT lamellae grew moderately from 0.009 s1 at 120C to 0.016 s1at 150C (Supporting Information Figure S2). From the three Avrami rate constants for PCBM aggregation at 120, 150, and 160C, we could extract an activation energy of Ea= 64.0( 12.3 kJ/mol
(15.3( 3.0 kcal/mol) for PCBM aggregation from the Arrhenius plot25in Figure 6. Interestingly, the Eavalue for
PCBM aggregation is close to the barrier of∼15 kcal/mol theoretically calculated for rotation of neighboring thio-phene rings from coplanar to orthogonal conformations (related to the coil-to-rod transformation)27along the
regioregular poly(3-propylthiophene) (P3PT) back-bone. This Eavalue, however, is significantly smaller
than the activation energy 37.5 kcal/mol for P3HT crystallization.27 These observations suggest that chain motion of P3HT controls the diffusion and aggre-gation of PCBM, but its own (nucleation-dominated)
Figure 5. (a) Values ofR(t) for PCBM aggregation and P3HT crystallization (edge-on and face-on lamellae along qzandqx, respectively) of the compositefilm with c = 1.0, at 150 C annealing. Also shown is the R(t) for the edge-on lamellae in a pristine P3HTfilm in situ annealed similarly. The data are fitted (solid curves) using eq 1. (b) Corresponding SharpHancock presentations of the data, linearlyfitted (solid lines) using the Avrami exponents (slopes) and rate constants listed in Table 1.
TABLE 1.Best-Fit Values of the Avrami Exponents n and Rate Constants k of PCBM Aggregation and P3HT Crystallization for the Edge-On and Face-On Lamellae in the Composite Film of PCBM/P3HT with c = 1.0 at 150 C (Also Shown Are the Kinetics Parameters for a Pristine P3HT Film Similarly Annealed)
PCBM aggregates P3HT face-on lamellae P3HT edge-on lamellae pure P3HT edge-on lamellae n 0.9( 0.1 0.9( 0.1 0.9( 0.1 1.0( 0.1 k (s1) 0.036( 0.015 0.017 ( 0.009 0.016 ( 0.006 0.071( 0.030
ARTICLE
crystallization is controlled by, most reasonably, the supercooling from its equilibrium melting temperature) after PCBM demixed. The mutually confined growths of PCBM aggregates and P3HT crystallites may therefore be modulated by the choice of annealing temperature. Segregation and subsequent aggregation of PCBM are expectedly better allowed at higher temperatures, leading to more efficiently increased (but not over-grown) PCBM aggregate size (D) and P3HT lamellar domain size (L), as indeed observed and summarized in Tables 2 and 3.
We have also conducted temperature-dependent GISAXS for a P3HT/PCBM compositefilm (c = 0.6) at successively elevated temperature from 30 to 180C (in 10C increments) with annealing time of 600 s for each temperature. Results shown in Figure 7a indicate gradual but identifiable increases of the PCBM aggre-gate size (ca. 8 nm in the as-cast state) at annealing temperatures above 100C (close to the glass transi-tion temperature Tgof PCBM reported previously16) to
ca. 16 nm at 170C. Upon further annealing at 180 C, however, there is a stronger increase of the PCBM aggregation size to ca. 20 nm, as also evidenced by
atomic force microscopic (AFM) images in Figure 7c. This sharper increase may be attributed to the reorga-nization (partial melting and recrystallization) of P3HT lamellae as the melting temperature of P3HT (ca. 195 C)16
is approached, which results in the relaxation of confinement on PCBM aggregates. This is supported by the emergence of the crystalline reflection at qz=
1.37 Å1 from the broad halo of PCBM aggregates (Supporting Information Figure S3). On the other hand, GISAXS results for the P3HT/PCBMfilms with varied PCBM concentrations of c = 0.6, 0.8, and 1.0 revealed that the PCBM aggregate size (17.5( 1.0 nm) after annealing at the same temperature of 150 C was independent of the PCBM concentration. We also notice that a certain fraction of PCBM molecules remain in the mixed phase in an optimized structure (cf. Tables 2 and 3). All of the present GISAXS results (as summarized in Tables 2 and 3 and Table S1 in Supporting Information) indicate that, within the stu-died composition and temperature ranges, PCBM ag-gregation size is determined mainly by kinetics of structural evolution.
MorphologyCharge Mobility Correlation. To correlate the morphological development of the P3HT/PCBM thin films to the corresponding solar cell performance, we have further monitored the changes of electron and hole mobilities for a BHJ film with c = 1.0 during annealing at 150C. Shown in Figure 8a is the uprising electron mobility,μe, observed within the first 100 s of
thermal annealing, which becomes saturated subse-quently. This synchronizes well with the development ofR(t) for the PCBM aggregation (Figure 8a). Similarly observed is the accompanied development of the hole mobility,μh, that saturated slightly later than the
elec-tron mobility (Figure 8b). Correspondingly, the devel-opment of hole mobility synchronized with theR(t) for the development of P3HT lamellae, as illustrated in Figure 8b. These sensitive responses indicate that charge mobilities in the BHJ solar cell and conse-quently the short-circuit current density Jsc depend
strongly on the developments of the P3HT and PCBM nanodomains upon heat treatment, whereas the in-crease of optical absorption owing to enhanced P3HT crystallization observed previously contributes only partly to Jsc.14,28Note that the values of the electron
and hole mobilities obtained might be subject to device preparation conditions to some extent, and we believe that the relative values, hence the growth trends, of the charge mobilities shown in Figure 8 should be reliable.9,29 In Table 3, the bulk structural
characteristics extracted from GISAXS/GIWAXS for the P3HT/PCBM thin-film solar cells annealed at various temperatures are correlated to the corresponding device performance parameters extracted, with the corresponding current densityvoltage curves shown in Supporting Information (Figure S4); accordingly, a cartoon in Figure 9 depicts the morphological features.
Figure 6. Arrhenius plot for the temperature-dependent Avrami rate constantk measured at 120, 150, and 160 C for PCBM aggregation. The activation energy of PCBM aggre-gation can be obtained from the slope (Ea/R) of the fitted
line.
TABLE 2. Structural Parameters Extracted for the In Situ Annealed P3HT/PCBM Thin Films (c = 1.0; PCBM Volume Fractionφo= 41 vol %), Including the Edge-On Lamellar
Domain Size L of P3HT Crystallites and the Size D, Mean Distance between Aggregates d, Wall-to-Wall Distance Δ (=d D), Polydispersity p, and Surface-to-Volume Ratio S/V of the PCBM Aggregates, Together with the D/L Valuea
c = 1.0 (41 vol %) L (nm) D (nm) d (nm) Δ (nm) P (%) φ (%) φm (%) S/V (nm1) D/L as-cast 7.5 7.9 b b 33 b 0.76 1.1 120C 9.6 13.2 18.8 5.6 44 18 23 0.45 1.4 150C 11.5 17.4 24.0 6.6 25 20 21 0.34 1.5
aThe PCBM volume fraction in aggregation isφ (fitted based on the hard sphere
structure factor) and that in the mixing phaseφm(=φo φ) with P3HT. bInformation could not be determined.
Interestingly, the structure model is very much in line with the predictions based on the temperature composition phase diagram of the binary P3HTPCBM blend previously given by Kim and Frisbie,30 in that
below 200C and within the concentration range of ∼3050 wt % (or c = 0.51.0), PCBM is expected to dissolve in P3HT, in the form of dispersed molecules or noncrystalline aggregates, forming a metastable phase.
Note that the power conversion efficiency of a BHJ thin-film solar cell is collectively determined by many factors, although the bulk morphology of the BHJfilms
featured by intercalated nanograins is shown here to play a critical role on charge mobility, hence, Jsc. In view
of the fast kinetics and fast saturation of the bulk morphology, the reported slight improvement of the P3HT/PCBM solar cell efficiency upon prolonged ther-mal annealing over several hundred seconds may associate more with surface/interface morphological features revealed by neutron reflectivity31,32 or the slower interdiffusion behavior of P3HT/PCBM in the film in-depth direction.33
Presumably, interfacial struc-tures have slower thermal responses (kinetics) because TABLE 3.Structural Parameters Extracted for the P3HT/PCBM (c = 0.8 or φo= 35%) Thin Films Annealed at 120, 150, and
180C for 900 sa c = 0.8 (35 vol %) L (nm) D (nm) d (nm) Δ (nm) p (%) φ (%) φm(%) S/V (nm1) D/L Jsc(mA/cm2) Voc(V) FF (%) η (%) as-cast 7.7 8.3 b b 31 5 30 0.72 1.1 6.2 0.34 35 0.9 120C 10.8 13.8 20.9 7.1 20 15 20 0.43 1.3 8.7 0.47 55 2.8 150C 11.8 17.6 24.7 7.1 23 19 16 0.34 1.5 9.5 0.62 59 4.4 180C 15.9 19.4 26.7 7.3 44 20 15 0.31 1.2 8.7 0.66 53 3.8
aThe notations for the structure parameters are the same as those in Table 2. The corresponding device performance parameters are short-circuit current densityJ
sc, open-circuit
voltageVoc,fill factor FF, and power conversion efficiency η.bInformation could not be determined.
Figure 7. (a) GISAXS revealed growth of PCBM aggregates with the annealing temperature for a P3HT/PCBM compositefilm withc = 0.6. (b,c) Phase contrast AFM images of the P3HT/PCBM films annealed, respectively, at 150 and 180 C for 900 s. The scale bars correspond to 200 nm.
Figure 8. Strongly correlated developments of (a) the electron mobility andR(t) of PCBM aggregation, and (b) the hole mobility andR(t) of P3HT lamellae, of P3HT/PCBM composite films (c = 1.0) at 150 C annealing.
of larger substrate/surface restrictions or con fine-ments. The quantitative correlation between charge mobility andfilm morphology established here may serve as a valuable reference in improving the algo-rithms and/or input parameters used in Monte Carlo model simulations of charge transport behavior.4,34 Previously, in correlating hole mobility to film mor-phology for a conjugate MEHPPV film, we have adopted the Gaussian disorder model (GDM) for the description of charge transport in organic semicon-ductors.35In GDM, charge mobility is modeled by two structure-related parameters, energy disorderσ and position disorderΣ, describing respectively, the spread of energy levels associated with charge transport and thefluctuations of intersite distance, thus the disper-sion of site coupling.36 Given the detailed structural information obtained here, it would be interesting to draw physical correlations between the disorder para-meters σ (via a proper combination of molecular dynamics and quantum mechanical calculations)37,38 andΣ to the structural information extracted, including the loosely packed nanograins (ca. 18 nm) of PCBM aggregates with a mean spacing of ca. 2025 nm, the orientation preference and lamellar crystalline size (ca. 1215 nm) of P3HT, and the volume fractions of respective phases. This, however, awaits further efforts in the near future.
Very recently, a related study39of the same material
system focusing on afixed annealing temperature of
140C up to ca. 1 h came to our notice. Combining real-time GIWAXS/ellipsometry/carrier mobility measure-ments during annealing and results of room-tempera-ture device performance, Agostinelli et al.39 have demonstrated that the microstructural evolution in P3HT/PCBM thinfilms involves two stages: in the first 5 min duration, crystallization of P3HT correlates with a major increase of photocurrent, which is followed by postulated aggregation of PCBM with increasing com-pactness upon which device performance remained unchanged. The present GISAXS/GIWAXS results en-compass a broader temperature range (from 120 to 180C), additional details in the PCBM aggregation process, and thorough kinetic analysis of the morpho-logical evolution during the early stage (thefirst 100 s) of annealing during which the device performance is improved most significantly. We demonstrate that it is the competition between crystallization of P3HT nano-grains and aggregation of PCBM that dictates the morphological development and hence the device performance. The two studies are, nevertheless, com-plementary in certain aspects.
CONCLUSIONS
With synchronized GIWAXS and GISAXS of enhanced spatial/time resolutions, we have captured the com-peting kinetics of PCBM aggregation and P3HT crystal-lization of the corresponding BHJ thin-film solar cells. Within thefirst 100 s of thermal annealing at 150 C, PCBM aggregation size in the P3HT/PCBM composite films grew quickly from 7 to 18 nm then saturated; meanwhile, the majority P3HT lamellae increased from a size of 7 to 12 nm. Both the kinetics of PCBM aggregation and P3HT crystallization could be charac-terized by similar Avrami exponents close to unity; the faster PCBM aggregation, however, has a 2-fold higher Avrami rate constant. The mutually confined growths led to comparable nanograin sizes of PCBM and P3HT below 20 nm, when annealing temperature was kept below 180C. The developments of PCBM and P3HT nanodomains could influence strongly and sensitively the electron and hole mobilities of the BHJ thin-film solar cells. The illustrated kinetics of structural evolu-tion and its correlaevolu-tion to changes in charge mobility of P3HT/PCBM compositefilms may bear relevance to the selection of alternative polymer/fullerene derivative combinations and in the optimization of processing conditions for future BHJ thin-film solar cells.
EXPERIMENTAL METHODS
Regioregular P3HT (Mw= 35 kDa; received from Rieke Metals)
was dissolved in chlorobenzene (15 mg L1) and mixed with a chlorobenzene solution (12 mg L1) of PCBM (Nano-C, Inc.) for different PCBM/P3HT weight ratios of c = 0, 0.6, 0.8, and 1.0; the corresponding PCBM volume fractions were 0, 29, 35,
and 41%, respectively. Thinfilms (ca. 85 nm in thickness, char-acterized using X-ray reflectivity) were spun-cast from the mixtures onto Si wafers of dimensions of 2.0 cm 2.0 cm and dried subsequently at ambient temperatures under N2gasflow.
Simultaneous GISAXS/GIWAXS measurements were per-formed at the BL23A endstation of the National Synchrotron Radiation Research Center (NSRRC), Taiwan. Details of the Figure 9. Cartoon for a thermally annealed P3HT/PCBM
film, according to the structure characteristics shown in Table 3 (150C case). The intercalated PCBM aggregates (large spheres) and P3HT crystallites (blocks) are immersed in the matrix of P3HT amorphous chains (thin wires) and dispersed PCBM molecules (small spheres).
instrument were reported previously.40Sample thinfilms were placed horizontally on a hot stage enclosed in an airtight chamber with thin (8μm) Kapton windows for X-rays. To reduce sample degradation over prolonged irradiation more than 1800 s at 150C annealing, the sample chamber was evacuated then purged with N2 gas. With a heating rate of 4C/s, sample
temperature could reach 150C within 40 s. With an 8 keV (wavelengthλ = 1.550 Å) beam and an incident angle 0.2, time-resolved (5 s/frame), simultaneous GISAXS/GIWAXS were con-ducted using two area detectors triggered by the same signals for synchronized data collection. The detector system included (i) a CMOSflat panel X-ray detector C9728DK (52.8 mm square)41 situated 7.2 cm from the sample position for GIWAXS, covering thefirst three diffraction peaks of P3HT lamellar crystallites, and (ii) a MAR165 CCD detector (165 mm in diameter), 300 cm from the sample position, allowing GISAXS data collection in the q region from 0.004 to 0.15 Å1. The scattering wavevector, defined as q = 4πλ1sinθ (with 2θ the scattering angle), was
calibrated using silver behenate, sodalite, and silicon powders, respectively.40GISAXS/GIWAXS profiles were extracted from the
respective 2D patterns along qx(with qzfixed at the specular
beam position)42 and q
z in the in-plane and out-of-plane
scattering directions, respectively, as illustrated in Figure 1. Experimental reproducibility was generally confirmed by re-peated runs using fresh specimens.
For parallel measurements of charge mobility, hole-only devices were fabricated by spin-coating (600 rpm, 60 s) the same P3HT/PCBM mixture with c = 1.0 on top of the precoated PEDOT/PSS (Clevios PVP AI4083) layer (ca. 20 nm) of ITO substrates. Subsequently, a Au electrode (ca. 100 nm) was deposited on the P3HT/PCBM layer under high vacuum (ca. 107Torr). Similarly, electron-only devices were fabricated by spin-coating an active layer of P3HT/PCBM (c = 1.0) on top of glass/Al (ca. 100 nm) substrates, followed by Ag evaporation (ca. 100 nm) for the top electrode. The thickness of each deposited layer was monitored using Veeco Dektak 150 surface profile meter. These hole-only and electron-only devices, eight sam-ples for each type, were annealed at 150C for different lengths of time ranging from 60 to 900 s and measured subsequently at ambient temperature for characteristic responses of current density to the applied voltage. Electron or hole mobility was deduced from the measured current density based on the space charge-limited current model.43,44
Data Analysis. GISAXS intensity profiles in the lower-q region contributed by PCBM aggregates were modeled by polydis-perse spheres using
I(q)¼ ÆnpæÆP(q)æS(q) (3)
with the averaged form factorÆP(q)æ and structure factor S(q).45
The number density of the scattering particles np(r) =Ænpæf(r) is
defined by the mean number densityÆnpæ and the Schultz size
distribution function46 f (r) ¼ zþ 1 ra zþ1 rzexp zþ 1 ra r " # =Γ(z þ 1), z > 1 (4) with the mean radius ra, width parameter z, and polydispersity
p = (zþ1)1/2. The form factor for spheres P(q) = [3j
1(qr)/(qr)]2is
defined by the first-order spherical Bessel function j1(qr). For
spheres with small polydispersity, S(q) may be approximated by the structure factor of the effective one-component system45of hard spheres
S(q)¼ [1 npC(q)1 (5)
with the effective diameterσ and volume fraction φ.47,48Here,
C(q) = 4πσ3ξ6{R
oξ3(sinξ ξ cos ξ) þ βoξ2[2ξ sin ξ (ξ2 2)
cosξ 2] þ γ[(4ξ3 24ξ)sin ξ (ξ4 12ξ þ 24)cos ξ þ 24]} is
defined byξ = qσ, Ro= (1þ 2φ)2(1 φ)4,βo=6φ[1 þ
(φ/2)]2(1 φ)4, andγ = φR o/2.
Acknowledgment. We thank Dr. N. Yagi for the help in installa-tion of theflat panel X-ray detector, and the National Science Council forfinancial support (NSC 99-2112-M-213-002-MY3).
Supporting Information Available: Simultaneous GISAXS/GI-WAXS movie, GIGISAXS/GI-WAXS data, kinetics analysis for 120C data, current densityvoltage curves, structural parameters for the film with c = 0.6. This material is available free of charge via the Internet at http://pubs.acs.org.
REFERENCES AND NOTES
1. Frost, J. M.; Cheynis, F.; Tuladhar, S. M.; Nelson, J. Influence of Polymer-Blend Morphology on Charge Transport and Thotocurrent Generation in DdonorAcceptor Polymer Blends. Nano Lett. 2006, 6, 1674–1681.
2. Yang, X.; Loos, J. Toward High-Performance Polymer Solar Cells: The Importance of Morphology Control. Macromo-lecules 2007, 40, 1353–1362.
3. Moon, J. S.; Lee, J. K.; Cho, S.; Byun, J.; Heeger, A. J. “Column-like” Structure of the Cross-Sectional Morphology of Bulk Heterojunction Materials. Nano Lett. 2009, 9, 230–234. 4. Muller, C.; Ferenczi, T. A. M.; Campoy-Quiles, M.; Frost, J. M.;
Bradley, D. D. C.; Smith, P.; Stingelin-Stutzmann, T.; Nelson, J. Binary Organic Photovoltaic Blends: A Simple Rationale for Optimum Compositions. Adv. Mater. 2008, 20, 3510–3515.
5. Kim, Y.; Cook, S.; Tuladhar, S. M.; Choulis, S. A.; Nelson, J.; Durrant, J. R.; Bradley, D. D. C.; Giles, M.; Mcculloch, I.; Ha, C. S.; et al. A Strong Regioregularity Effect in Self-Orga-nizing Conjugated Polymer Films and High-Efficiency Polythiophene: Fullerene Solar Cells. Nat. Mater. 2006, 5, 197–203.
6. Li, G.; Shrotriya, V.; Huang, J.; Yao, Y.; Moriarty, T.; Emery, K.; Yang, Y. High-Efficiency Solution Processable Polymer Photovoltaic Cells by Self-Organization of Polymer Blends. Nat. Mater. 2005, 4, 864–868.
7. Ma, W.; Yang, C.; Gong, X.; Lee, K.; Heeger, A. J. Thermally Stable, Efficient Polymer Solar Cells with Nanoscale Con-trol of the Interpenetrating Network Morphology. Adv. Funct. Mater. 2005, 15, 1617–1622.
8. van Bavel, S. S.; Sourty, E.; de With, G.; Loos, J. Three-Dimensional Nanoscale Organization of Bulk Heterojunc-tion Polymer Solar Cells. Nano Lett. 2009, 9, 507–513. 9. Chiu, M. Y.; Jeng, U.; Su, M. S.; Wei, K. H. Morphologies of
Self-Organizing Regioregular Conjugated Polymer/Fuller-ene Aggregates in Thin Film Solar Cells. Macromolecules 2010, 43, 428–432.
10. Chiu, M. Y.; Jeng, U.; Su, C. H.; Liang, K. S.; Wei, K. H. Simultaneous Use of Small- and Wide-Angle X-ray Tech-niques To Analyze Nanometer Scale Phase Separation in Polymer Heterojunction Solar Cells. Adv. Mater. 2008, 20, 2573–2578.
11. Kennedy, R. D.; Ayzner, A. L.; Wanger, D. D.; Day, C. T.; Halim, M.; Khan, S. I.; Tolbert, S. H.; Schwartz, B. J.; Rubin, Y. Self-Assembling Fullerenes for Improved Bulk-Heterojunc-tion Photovoltaic Devices. J. Am. Chem. Soc. 2008, 130, 17290–17292.
12. Szarko, J. M.; Guo, J.; Liang, Y.; Lee, B.; Rolczynski, B. S.; Strzalka, J.; Xu, T.; Loser, S.; Marks, T. J.; Yu, L.; et al. When Function Follows Form: Effects of Donor Copolymer Side Chains on Film Morphology and BHJ Solar Cell Perfor-mance. Adv. Mater. 2010, 22, 5468–5472.
13. Chen, D.; Nakahara, A.; Wei, D.; Nordlund, D.; Russell, T. P. P3HT/PCBM Bulk Heterojunction Organic Photovoltaics: Correlating Efficiency and Morphology. Nano Lett. 2011, 11, 516–567.
14. Kiel, J. W.; Eberle, A. P. R.; Mackay, M. E. Nanoparticle Agglomeration in Polymer-Based Solar Cells. Phys. Rev. Lett. 2010, 105, 168701.
15. Shin, M.; Kim, H.; Park, J.; Nam, S.; Heo, K.; Ree, M.; Ha, C.-S.; Kim, Y. Abrupt Morphology Change upon Thermal Anneal-ing in Poly(3-Hexylthiophene)/Soluble Fullerene Blend Films for Polymer Solar Cells. Adv. Funct. Mater. 2010, 20, 748–754.
16. Verploegen, E.; Mondal, R.; Bettinger, C. J.; Sok, S.; Toney, M. F.; Bao, Z. Effects of Thermal Annealing upon the Morphology of PolymerFullerene Blends. Adv. Funct. Mater. 2010, 20, 3519–3529.
17. Renaud, G.; Lazzari, R.; Leroy, F. Probing Surface and Interface Morphology with Grazing Incidence Small Angle X-ray Scattering. Surf. Sci. Rep. 2009, 64, 255–380. 18. Chen, S. H.; Sheu, E. Y.; Kalus, J.; Hoffmann, H. Small-Angle
Neutron Scattering Investigation of Correlations in Charged Macromolecular and Supramolecular Solutions. J. Appl. Crystallogr. 1988, 21, 751–769.
19. Debye, P., Jr.; Anderson, H. R.; Brumberger, H. Scattering by an Inhomogeneous Solid. II. The Correlation Function and Its Application. J. Appl. Phys. 1957, 28, 679–683. 20. Chen, C.-Y.; Chan, S.-H.; Li, J.-Y.; Wu, K.-H.; Chen, H.-L.; Chen,
J.-H.; Huang, W.-Y.; Chen, S.-A. Formation and Ther-mally-Induced Disruption of Nanowhiskers in Poly(3-hexylthiophene)/Xylene Gel Studied by Small-Angle X-ray Scattering. Macromolecules 2010, 43, 7305–7311. 21. Lin, J.-M.; Lin, T.-L.; Jeng, U.; Zhong, Y.-J.; Yeh, C.-T.; Chen, T.-Y. Fractal Aggregates of the Pt Nanoparticles Synthesized by the Polyol Process and Poly(N-vinyl-2-pyrrolidone) Reduction. J. Appl. Crystallogr. 2007, 40, s540–543. 22. Milanesio, M.; Artioli, G.; Gualtieri, A. F.; Palin, L.; Lamberti,
C. Template Burning Inside TS-1 and Fe-MFI Molecular Sieves: An In Situ XRPD Study. J. Am. Chem. Soc. 2003, 125, 14549–14558.
23. Grandjean, D.; Beale, A. M.; Petukhov, A. V.; Weckhuysen, B. M. Unraveling the Crystallization Mechanism of CoAPO-5 Molecular Sieves under Hydrothermal Conditions. J. Am. Chem. Soc. 2005, 127, 14454–14465.
24. Bruna, P.; Crespo, D.; González-Cinca, R.; Pineda, E. On the Validity of Avrami Formalism in Primary Crystallization. J. Appl. Phys. 2006, 100, 054907–054917.
25. Allen, J. L.; Jow, T. R.; Wolfenstine, J. Kinetic Study of the Electrochemical FePO4to LiFePO4Phase Transition. Chem.
Mater. 2007, 19, 2108–2111.
26. Vaynzof, Y.; Kabra, D.; Zhao, L.; Chua, L. L.; Steiner, U.; Friend, R. H. Surface-Directed Spinodal Decomposition in Poly[3-hexylthiophene] and C61-Butyric Acid Methyl Ester Blends. ACS Nano 2011, 5, 329–336.
27. Malik, S.; Jana, T.; Nandi, A. K. Thermoreversible Gelation of Regioregular Poly(3-hexylthiophene) in Xylene. Macromo-lecules 2001, 34, 275–282.
28. Mihailetchi, V. D.; Xie, H.; de Boer, B.; Koster, L. J. A.; Blom, P. W. M. Charge Transport and Photocurrent Generation in Poly(3-hexylthiophene): Methanofullerene Bulk-Hetero-junction Solar Cells. Adv. Funct. Mater. 2006, 16, 699–708. 29. Shrotriya, V.; Yao, Y.; Li, G.; Yanga, Y. Effect of Self-Organi-zation in Polymer/Fullerene Bulk Heterojunctions on Solar Cell Performance. Appl. Phys. Lett. 2006, 89, 063505– 063507.
30. Kim, J. Y.; Frisbie, C. D. Correlation of Phase Behavior and Charge Transport in Conjugated Polymer/Fullerene Blends. J. Phys. Chem. C 2008, 112, 17726–17736. 31. Kiel, J. W.; Kirby, B. J.; Majkrzak, C. F.; Maranvillec, B. B.;
Mackay, M. E. Nanoparticle Concentration Profile in Poly-mer-Based Solar Cells. Soft Matter 2010, 6, 641–646. 32. Parnell, A. J.; Dunbar, A. D. F.; Pearson, A. J.; Staniec, P. A.;
Dennison, A. J. C.; Hamamatsu, H.; Skoda, M. W. A.; Lidzey, D. G.; Jones, R. A. L. Depletion of PCBM at the Cathode Interface in P3HT/PCBM Thin Films as Quantified via Neutron Reflectivity Measurements. Adv. Mater. 2010, 22, 2444–2447.
33. Treat, N. D.; Brady, M. A.; Smith, G.; Toney, M. F.; Kramer, E. J.; Hawker, C. J.; Chabinyc, M. L. Interdiffusion of PCBM and P3HT Reveals Miscibility in a Photovoltaically Active Blend. Adv. Energy Mater. 2011, 1, 82–89.
34. Watkins, P. K.; Walker, A. B.; Verschoor, G. L. B. Dynamical Monte Carlo Modelling of Organic Solar Cells: the Depen-dence of Internal Quantum Efficiency on Morphology. Nano Lett. 2005, 5, 1814–1818.
35. Inigo, A. R.; Chiu, H.-C.; Fann, W.; Huang, Y.-S.; Jeng, U.; Lin, T.-L.; Hsu, C.-H.; Peng, K.-Y.; Chen, S.-A. Disorder Controlled Hole Transport in MEH-PPV. Phys. Rev. B 2004, 69, 075201– 075208.
36. Bässler, H. Charge Transport in Disordered Organic Photo-conductors a Monte Carlo Simulation Study. Phys. Status Solidi B 1993, 175, 15–56.
37. Lan, Y.-K.; Huang, C.-I. A Theoretical Study of the Charge Transfer Behavior of the Highly Regioregular Poly(3-hexylthiophene) in the Ordered State. J. Phys. Chem. B 2008, 112, 14857–14862.
38. Lan, Y.-K.; Huang, C.-I. Charge Mobility and Transport Behavior in the Ordered and Disordered States of the Regioregular Poly(3-hexylthiophene). J. Phys. Chem. B 2009, 113, 14555–14564.
39. Agostinelli, T.; Lilliu, S.; Labram, J. G.; Campoy-Quiles, M.; Hampton, M.; Pires, E.; Rawle, J.; Bikondoa, O.; Bradley, D. D. C.; Anthopoulos, T. D.; et al. Real-Time Investigation of Crystallization and Phase-Segregation Dynamics in P3HT: PCBM Solar Cells During Thermal Annealing. Adv. Funct. Mater. 2011, 21, 1701–1709.
40. Jeng, U.; Su, C.-H.; Su, C.-J.; Liao, K.-F.; Chuang, W.-T.; Lai, Y.-H.; Chang, Y.-J.; Chen, Y.-J.; Huang, Y.-S.; Lee, M.-T.; et al. A Small/Wide-Angle X-ray Scattering Instrument for Structural Characterization of AirLiquid Interfaces, Thin Films and Bulk Specimens. J. Appl. Crystallogr. 2010, 43, 110–121. 41. Yagi, N.; Inoue, K. CMOS Flat Panel Detectors for SAXS/
WAXS Experiments. J. Appl. Crystallogr. 2007, 40, s439– s441.
42. Hsu, C.-H.; Jeng, U.; Lee, H.-Y.; Windover, D.; Lu, T.-M. Structural Study of a Low Dielectric Thin Film Using X-ray Reflectivity and Grazing Incidence Small Angle X-ray Scattering. Thin Solid Films 2005, 472, 323–327. 43. Mihailetchi, V. D.; Wildeman, J.; Blom, P.W. M.
Space-Charge Limited Photocurrent. Phys. Rev. Lett. 2005, 94, 126602–126605.
44. Melzer, C.; Koop, E. J.; Mihailetchi, V. D.; Blom, P. W. M. Hole Transport in Poly(phenylene vinylene)/Methanofullerene Bulk-Heterojunction Solar Cells. Adv. Funct. Mater. 2004, 14, 865–870.
45. Chen, S.-H.; Lin, T.-L. In Methods of Experimental Physics; Neutron Scattering in Condensed Matter Research; Sköld, K., Price, D. L., Eds.; Academic Press: New York, 1987; Vol. 23B, Chapter 16.
46. Sheu, E. Y. Polydispersity Analysis of Scattering Data from Self-Assembled Systems. Phys. Rev. A 1992, 45, 2428–2438. 47. Hayter, J. B.; Penfold, J. An Analytic Structure Factor for
Macroion Solutions. Mol. Phys. 1981, 42, 109–118. 48. Su, C. H.; Jeng, U.; Chen, S. H.; Lin, S. J.; Wu, W. R.; Chuang,
W.-T.; Tsai, J. C.; Su, A. C. Nanograin Evolution in Cold Crystallization of Syndiotactic Polystyrene as Illustrated via In-Situ Small/Wide-Angle X-ray Scattering and Dif-ferential Scanning Calorimetry. Macromolecules 2009, 42, 6656–6664.
