Molecular Miscibility of Polymer-Fullerene Blends by wuyunqing



Molecular Miscibility of Polymer-Fullerene Blends
Brian A. Collins,† Eliot Gann,† Lewis Guignard,† Xiaoxi He,‡ Christopher R. McNeill,‡ and
Harald Ade*,†
 Department of Physics, North Carolina State University, Raleigh, North Carolina 27695, United States, and
 Cavendish Laboratory, University of Cambridge, J J Thomson Avenue, Cambridge, CB3 0HE, U.K.

    ABSTRACT The device function of polymer bulk heterojunction (BHJ) solar cells
    has been commonly interpreted to arise from charge separation at discrete interfaces
    between phase-separated materials and subsequent charge transport through these
    phases without consideration of phase purity. To probe composition, the miscibility of
    poly(3-hexylthiophene) (P3HT)and poly(2-methoxy-5-(30 ,70 -dimethyloctyloxy)-
    1,4-phenylenevinylene) (MDMO-PPV) with phenyl-C61-butyric acid methyl ester (PCBM)
    has been determined, while the effects of polymer crystallinity on miscibility are probed
    using P3HT grades of varying regioregularity. It is found that, while no intercalation
    occurs in P3HT crystals, amorphous portions of P3HT and MDMO-PPV contain
    significant concentrations of PCBM, calling into question models based on pure phases
    and discrete interfaces. Furthermore, depth profiles of P3HT/PCBM bilayers reveal that
    even short annealing causes significant interdiffusion of both materials, showing that
    under no conditions do pure amorphous phases exist in BHJ or annealed bilayer
    devices. These results suggest that current models of charge separation and transport
    must be refined.
    SECTION Macromolecules, Soft Matter

                                                                          performance of a device.9,10 In fact, initial studies of molecular

         hotovoltaic devices based on solution-processable or-
         ganic molecules and polymers hold immense poten-                 morphology have revealed that a fullerene intercalated into
         tial in providing a cheap, scalable, and renewable               polymer crystals can substantially improve device performance
source of energy using environmentally friendly materials                 over nonintercalated crystals.11 Other studies have suggested
and have thus been under intense investigation over the past              that amorphous blends do not contain pure phases12,13 even
several years. The payoff has been an orders of magnitude                 after annealing.14-16 Here we probe the thermodynamic forces
improvement of their power conversion efficiency since their              at play by allowing two key model polymers of poly(3-hexyl-
first demonstration,1,2 yet they are still in need of further             thiophene) (P3HT) and poly(2-methoxy-5-(30 ,70 -dimethyloctyl-
improvement before fully replacing current technology.3-6                 oxy)-1,4-phenylenevinylene) (MDMO-PPV) blended with phenyl-
Because these organic photovoltaic (OPV) devices operate via              C61-butyric acid methyl ester (PCBM) to achieve equilibrium
separation of a short-lived exciton at a donor/acceptor inter-            concentrations upon annealing and directly measure their
face, high interfacial area devices from blends of donor and              composition via near-edge X-ray absorption fine structure
acceptor materials in so-called bulk heterojunctions (BHJs)               (NEXAFS) spectroscopy, thus creating a miscibility phase
have resulted in much higher efficiencies than bilayer                    diagram for amorphous portions of the systems investigated.
constructions. Under the current paradigm of device operation,            To examine the influence of P3HTcrystallinity, we have studied
blends create a BHJ or bicontinuous network of phase sepa-                several grades of the polymer exhibiting different degrees of
rated materials, which provide a maximized area of discrete               regioregularity (RR) and characterized their crystallinity via
interfaces to separate the excitons into free charges and perco-          grazing-incidence wide-angle X-ray scattering (GIWAXS). In
lation pathways in which to transport those charges to the                addition, we have measured depth profiles of annealed P3HT/
appropriate electrodes.7 The phases are often implicitly or               PCBM bilayers of the two materials using dynamic secondary
explicitly assumed to be pure.4 Mixed phases are thought to               ion mass spectrometry (SIMS) to study their interdiffusion. The
be counterproductive to device performance, since isolated                results of this study show that within amorphous regions, pure
molecules could act as traps for separated charges and even-              phases are not thermodynamically favored. This has a direct
tually centers for charge recombination within the percolation            impact on how charge separation and transport is thought to
pathways.8 Short exciton diffusion lengths require phase sepa-            occur in OPV devices, and in light of this finding, a modified
ration to be on the order of 10 nm, making interface morphol-             paradigm of device operation may be required.
ogy and phase purity difficult to measure, yet this may be
critical as recent transient absorption spectroscopy measure-             Received Date: September 9, 2010
ments have suggested that the intimate molecular morphol-                 Accepted Date: October 7, 2010
ogy and ordering of the two materials may determine the                   Published on Web Date: October 19, 2010

    r 2010 American Chemical Society                               3160                  DOI: 10.1021/jz101276h |J. Phys. Chem. Lett. 2010, 1, 3160–3166

Figure 1. (a) STXM image of a high-RR P3HT:PCBM blended and annealed film acquired at 284.4 eV, an absorption peak of PCBM. Dark
regions are PCBM crystals. The arrow designates a typical area used to generate a spectrum. Each scan includes I0 data taken simultaneously
(right 30% of the line) to avoid milli-electron volt-scale drift of the energy from temperature drift in the beamline. (b) P3HT, (c) PCBM, and
(d) MDMO-PPV molecular structures. (e) Reference spectra used in fits and (f) ratio of P3HT reference spectra acquired at various incident
angles, showing the effect of crystallite orientation on absorption peaks of a linearly polarized photon beam. Angles posted are those the
photon electric field makes with the surface. Ratios displayed represent relative edge-on P3HT crystal orientation (see Figure 4f).

    For the P3HT-based system, three grades ; noncrystalliz-                    Once annealed, films were floated onto TEM grids and
ing regiorandom [Rieke 4007], 93-95% RR [Rieke 4002], and                   examined in the scanning transmission X-ray microscope
a >98% RR [Plextronics OS 1200] ; were used. PCBM was                       (STXM) at the 5.3.2 beamline of the Advanced Light Source at
obtained from Nano-C. MDMO-PPV was supplied by American                     Lawrence Berkeley National Laboratory.20 A nitrogen filter
Dye Source [ADS104RE]. With record power conversion effi-                   was used in the beamline (0.6 Torr for 1 m) to further enhance
ciencies of 3.3% for the latter17 and 5% for the former,18,19               spectral purity.21 Low- and high-resolution micrographs were
these systems represent some of the highest performing mate-                acquired with the STXM at key energies of the carbon 1s
rials that are commercially available whose morphologies are                excitation near 285 eV22,23 that selectively highlight the two
thought to be largely optimized.                                            materials. PCBM phase separations were easily identified
    Blended films were spincast onto poly(styrenesulfonate)                 from the mixture as shown in Figure 1a along with examples
(PSS)-coated glass substrates from chlorobenzene solutions in               from other films in the Supporting Information. Even though
a N2 glovebox containing <1 ppm O2 and ∼2 ppm H2O and                       amorphous and crystalline P3HT phases have subtly different
annealed on a covered hot plate. Excess PCBM in the blend                   NEXAFS spectra,24 they are not readily separable in a film that
phase separated into large crystals many micrometers in size                is thicker than the crystal size. In fact, no other distinguishable
in every film tested. PCBM depletion regions could be seen in a             phases could be identified other than PCBM crystals and the
visible light microscope growing with annealing time. The films             polymer-rich phase. This included any PCBM-rich phase in
were annealed until they had clearly reached equilibrium or                 either P3HT or MDMO-PPV, which after such a long anneal
PCBM depletion regions were more than 100 μm large (from                    should be visible with STXM imaging. Spectra were therefore
18 to 70 h) and subsequently quenched on a metal plate to                   acquired over relatively large areas of the sample (typically
room temperature. The temperature was measured by a                         0.5 Â 20 μm) in an energy range of 270-400 eV within P3HT
calibrated surface thermometer within the sample chamber                    amorphous and crystalline phases alike, while avoiding PCBM
on the hot plate. Initial blend ratios were varied, including 1:1,          crystals (see Figure 1). For each film a minimum of three
1:2, and 1:4 (polymer/fullerene by weight) to ensure the onset              spectra were measured in different regions approximately
of phase separation and universal equilibrium levels with initial           1 mm apart to check reproducibility. Gradual drift of the in-
concentration. Film thickness was also varied from 70 to 150                cident photon energy due to thermal drift in beamline optics
nm to ensure PCBM levels measured did not originate from                    was corrected to within 10 meV via common features in I0
interfacial or surface wetting layers. In all cases, the same               spectra taken simultaneously with absorption spectra (Figure 1a).
concentration of PCBM was measured at a particular tempera-                 Any detector nonlinearities were corrected through diagnostic
ture, regardless of initial blend ratio or film thickness.                  slit scans.

    r 2010 American Chemical Society                                 3161                  DOI: 10.1021/jz101276h |J. Phys. Chem. Lett. 2010, 1, 3160–3166

    The spectra were fit to reference spectra of each material/
grade (Figure 1e) taken from pure films normalized above the
edge (360-400 eV) to bare atom absorption cross sections of
the constituent atoms.25 Three parameters were used in least-
squares fitting: an overall scale or thickness parameter, a
relative composition parameter, and a P3HT crystal orienta-
tion parameter. Because the X-ray beam used was linearly
polarized, absorption peaks in P3HT were enhanced or
suppressed depending on crystallite preferential orientation
in or out of plane of the substrate and the orientation of the
excited-state electron orbitals with respect to that crystal. This
effect was included via a reference “spectrum” calculated by
taking the ratio of NEXAFS spectra at differing incident angles
of partially oriented P3HT thin film, utilizing angle-resolved
data from the Watts NEXAFS semiconducting polymer
database.26 Figure 1f shows this differential spectrum.
    GIWAXS data was acquired at beamline 7.3.3 of the
Advanced Light Source on samples prepared under identical
conditions as the NEXAFS data. The 10 keV beam was
incident on the sample at 0.20°, which maximized the
diffraction intensity detected by an ADSC Quantum 4 CCD.
At this angle, scattering from the PSS layer and glass substrate
were clearly visible, indicating a fully penetrating photon
beam. For dynamic SIMS measurements, bilayers were pre-
pared by floating 100 nm P3HT films onto 100 nm PCBM-                           Figure 2. Spectral fits to high-RR P3HT:PCBM mixtures before
coated Si substrates and left unannealed or were annealed for                   annealing (top) and after annealing (bottom) for 48 h at 180 °C.
                                                                                Yellow and blue curves are from reference P3HT and PCBM fllms,
5 min at 140 °C. Polyvinyl alcohol films (50 nm) were spincast                  respectively, and are plotted on an independent y-axis for com-
from water onto the bilayers as sacrificial overlayers for                      parison. Absorption data at the edge is expanded in energy for clarity.
improved SIMS analysis with 5 nm of Au evaporated on top
to reduce charging. During SIMS, a 0.96 kV Cs primary beam
was rastered over an area 180 μm on a side for sputtering
with C, S, Si, and O signals monitored from a 60 μm diameter
optically gated area. Three excavation sites were completed
on distant areas of each sample (bilayer and single layer
samples) to ensure reproducibility.
    Typical NEXAFS spectra from blends and their fits are
shown in Figure 2 for unannealed P3HT blend films and
P3HT:PCBM films annealed at 180 °C. Although the solutions
were stirred at 50 °C for several hours to ensure fully dissolved
components, fits of unannealed blends systematically showed
PCBM-rich concentrations (for example, 53-55 wt % in films                      Figure 3. Miscibility phase diagram for P3HT mixed with PCBM.
cast from 1:1 wt % solutions) when cast using 0.2 μm nylon                      Lines are guides for the eye. Uncertainties are from variation in
                                                                                localities on the film. The somewhat low regiorandom data point is
filters, whereas films cast without a filter resulted in fits agreeing          from data that suffered a temporary reduction in spectral energy
with solution concentrations (50-51% PCBM). This indicates                      resolution.
that P3HT is preferentially removed during filtering ; slightly
altering the ratio of the blend ; and points to the high sensitivity            Furthermore, slow growth of depletion regions was observed,
and accuracy of the NEXAFS measurement technique. The                           suggesting that these binary concentrations lie in a spinodal
resulting miscibility curves for the P3HT:PCBM system from                      region of the phase diagram, similar to the “metastable” regions
the NEXAFS fits are displayed in Figure 3. All grades show an                   seen in previous studies.27 A clean glass surface was shown to
increasing miscibility of PCBM with temperature, suggesting an                  produce far higher PCBM nucleation density and rates, due
upper critical solution temperature for the binary phase dia-                   potentially to a surface energy more attractive to PCBM mole-
gram. Both the middle and high-RR grades of P3HTexhibit very                    cules. Blends on glass slides were found to have the same
similar levels of PCBM starting at 3% and 4% concentration                      miscibility levels of PCBM as those cast on PSS after annealing,
at 120 °C, respectively, up to approximately 10% at 180 °C.                     reinforcing the conclusion that the universal bulk thermody-
Contrasting this is the regiorandom grade of P3HT with much                     namic equilibrium is measured here.
larger miscibility levels of 16-22%. For this grade, both initial                   GIWAXS data are displayed in Figure 4 and help to
blend ratios used (1:1 and 1:2 P3HT:PCBM) eventually achieved                   illuminate the influence of crystallinity on PCBM miscibility
the same equilibrium levels, although PCBM crystals did not                     levels between regiorandom and RR P3HT blends. Images
nucleate readily in these blends on PSS-coated glass substrates.                obtained with no sample (air scatter), glass substrate only, and

    r 2010 American Chemical Society                                     3162                   DOI: 10.1021/jz101276h |J. Phys. Chem. Lett. 2010, 1, 3160–3166

Figure 4. GIWAXS data on films annealed at 160 °C for 24 h. Data shown are from (a) mid-RR P3HT, (b) mid-RR blends, (c) high-RR P3HT,
and (d) high-RR blends. (e) Scattering profiles fully integrated along the arc of each data set at the (100) diffraction peak versus calculated
d-spacing. Intensities in panel e have been scaled for thickness of annealed films measured via ellipsometry. D-spacing values are 16.5 ( 0.1 Å for
high-RR P3HTand blends, while mid-RR P3HTand blends yield 16.2 ( 0.1 Å (vertical lines to guide the eye). (f,g) Diagrams defining edge-on
and face-on orientation of P3HT crystals, respectively.

substrate with PSS coating were used to remove background                     diffraction peaks with similar intensities, revealing compar-
scatter from the data. The resulting background subtracted                    able levels of crystallization. Although absorption and other
data exhibit the (100), (200), and (300) lamella reflections as               geometrical effects are known to attenuate the signal from in-
well as the (010) from P3HT π-π stacking. Figure 4b,d                         plane scattering,31 the similar relative orientation preference
additionally shows diffraction from PCBM crystals. The                        of the crystallites allows a rough comparison here. Addition-
different distributions of the (010) intensities along the arc                ally, the d-spacing of the diffraction peaks of pure P3HT films
clearly show a difference in P3HT crystallite preferential                    and those from blends are the same, indicating no intercala-
orientation with the blend films exhibiting a relative face-on                tion of PCBM in P3HT crystals. However, the d-spacing of
orientation compared with the edge-on orientation exhib-                      P3HTwith lower RR shows a systematically smaller d-spacing.
ited in pure P3HT films (see Figure 4f, g for schematics). This               The corresponding shifts are also observed in the higher order
makes necessary the polarization parameter in the NEXAFS                      peaks. While further studies would be required to illuminate
fit as shown in Figure 1f, resulting in increasingly negative                 the mechanism behind a d-spacing shift, the similar intensi-
polarization factors with increasing RR (0.2, -0.5, and                       ties of the two P3HT grades are likely due to the fact that both
-0.9 ( 0.2 for regiorandom, mid-RR, and high-RR P3HT,                         grades have relatively high RR and may not significantly
respectively) and corroborates the relative face-on orienta-                  impact the overall level of crystallization. Because the method
tion seen here. The mechanism of this difference is un-                       used to quantify miscibility averages the PCBM level across
known in our blended films as this is contrary to results from                both P3HT amorphous and crystalline regions, it is clear that
other studies.28,29 This phenomenon in our highly annealed                    PCBM miscibility in amorphous regions of RR-P3HTare actually
films is being investigated further. An increasingly face-on                  significantly higher than is measured here. We were not able
orientation of P3HT crystals in devices would dramatically                    to quantify the ratio of amorphous to crystalline regions
increase hole mobility30 and potentially power conversion                     using GIWAXS alone, but it is probable that the noncrystalliz-
efficiency if the morphology could be controlled. Contrary to                 ing regiorandom P3HT grade is a closer representative to the
RR-P3HT, GIWAXS from regiorandom P3HT (not shown) re-                         actual miscibility level of PCBM in amorphous P3HT. Recent
sulted only in relatively modest nearest-neighbor scattering,                 studies of pure P3HT32 and blends with PCBM33 indicate that
suggesting that no substantial P3HT crystallization occurs in                 P3HT crystalline regions comprise 40-60% of the polymer,
this material as expected.                                                    which would put the actual PCBM miscibility in amorphous
    Radial profiles integrated along the arc are plotted versus               portions of RR-P3HT between 10 and 20%, in the range of
d-spacing in Figure 4e. The two RR-P3HT grades exhibit (100)                  that exhibited in regiorandom P3HT.

    r 2010 American Chemical Society                                   3163                  DOI: 10.1021/jz101276h |J. Phys. Chem. Lett. 2010, 1, 3160–3166

                                                                            mechanism of OPV device operation. Clearly, pure phase
                                                                            separations do not occur in the MDMO-PPV:PCBM system nor
                                                                            the amorphous regions of the P3HT:PCBM system, as these
                                                                            films are cast from well-mixed solutions. Furthermore, if the
                                                                            photovoltaic effect in a device relies on phase separation into
                                                                            merely a “polymer-rich” and “fullerene-rich” phase, isolated
                                                                            minority molecules would be expected to impede transport
                                                                            through kinetic blocking of the percolation pathway, trapping of
                                                                            charges separated there, and eventually nongeminate recom-
                                                                            bination of those charges. Indeed experimental studies35 and
                                                                            Monte Carlo simulations8 of the PPV-based system have shown
                                                                            that when blended with small concentrations of PCBM up
                                                                            through 25 wt %, the resulting mobility values are signifi-
Figure 5. SC signal from dynamic SIMS of P3HT/PCBM bilayers                 cantly below even that of the pure polymer. Since misci-
compared with that of a pure P3HT film. Annealing is for 5 min at           bility levels for these two systems lie in this regime, it is
140 °C. Digging rates measured on pure single layers were linearly          unlikely that optimized devices, which have external quan-
interpolated based on the SC ratio to convert sputtering rate into
depth. Depth, therefore, is only approximate.                               tum efficiencies (electron out per incident photon) above
                                                                            50%, rely on phase separations at the thermodynamic
    Results of the dynamic SIMS measurements are shown in                   limit. This is supported by the fact that the highest-perform-
Figure 5, which represents a complementary measurement                      ing devices based on the PPV system are not annealed,
(diffusion into) to the demixing (diffusion out of) experiment              resulting in a seemingly fully mixed phase (close to 50-
described above. Here, P3HT levels were measured in the                     50 wt %) surrounding a pure PCBM phase.36 Greater mixing
bilayers by tracking the ratio of sulfur to carbon (S/C) signals.           on the molecular level could have a positive impact on
The green curve represents the ratio taken from a pure P3HT                 exciton disassociation as they would have less distance to
single layer, while the lowest level on the black curve corre-              travel before encountering a PCBM molecule but could also
sponds to that of pure PCBM for this technique. Via measuring               result in high rates of nongeminate recombination, which
pure single layers, the sputtering rate through P3HTwas found               have been reported in this system. This suggests that
to be 2.0 times that of PCBM. The sputtering time was then                  morphology on the molecular level is not optimized.
converted to depth through a linear interpolation of the two                    Contrasting this with the P3HT system, a nanometer-scale
sputtering rates versus S/C ratios from the pure films, which               entangled matrix of pure nanocrystallites helps holes percolate
resulted in the proper alignment of each interface in the                   to the electrode and keeps the size scale of phase separations
multilayer films. As can be observed in the figure, 5 min of                small. In this case, PCBM concentrations in films prepared for
annealing resulted in equilibrium in the P3HT-rich layer but                devices could very well be higher than the thermodynamic limit;
not in the PCBM-rich layer. This simple comparison clearly                  however, nongeminate recombination does not seem to be high
shows a swift interdiffusion of both species into the comple-               in this system. An interpretation could be that the concentration
mentary layer. The relatively constant S/C ratio in the P3HT                “frozen” into the amorphous phase during casting is beyond the
layer compared with the exponential decrease in the PCBM                    percolation threshold for each molecule, (i.e., it is likely that
layer of the annealed sample demonstrates that the PCBM has                 molecules of the same type lie within easy charge-hopping
the higher mobility, as expected. Recent work on P3HT/PCBM                  distance), and therefore device operation may rely on phase
bilayers devices questioned the need for BHJ morphologies                   separation less than an optimal molecular concentration for
and reported efficiencies approaching that of BHJ devices.34                percolation. Another possibility is that amorphous mixtures of
The conclusions were based on the bilayer persisting as pure                the two components reside between P3HT and PCBM nano-
materials. From our SIMS experiment, it is clear that even with             crystals, acting as an exciton diassociation layer. A recent study
initial conditions of pure components, thermodynamic forces                 of fullerene intercalation in the crystallizing polymer poly(2,5-
prevent the formation or preservation of pure phases if any                 bis(3-hexadecylthiophen-2-yl)thieno[3,2-b]thiophene (pBTTT)
thermal treatment is utilized. Rather, the high performance of              also challenges the paradigm of pure phases and discrete
such bilayer devices likely resulted from a partial interpene-              interfaces where devices with the intercalated polymer result
tration of the P3HT with PCBM during annealing without                      in substantially higher power conversion efficiency and ex-
destroying the overall integrity of the P3HT layer, thereby                 ternal quantum efficiency than those without intercalation.11
resembling a BHJ structure.                                                 While some ordering may occur in amorphous regions of the
    We have also quantified the miscibility levels for PCBM                 systems investigated here, intercalation of the side chains
with the noncrystallizing polymer MDMO-PPV (not shown).                     seems unlikely due to their significantly smaller spacing along
For this system, data was acquired from samples with various                the polymer backbone between side chains. Still, ordering
initial blend ratios and thickness annealed at 140 and 160 °C.              may not be an important aspect to device function. What part
Fits reveal an 8% PCBM miscibility in this polymer, which is                these mixed phases play in charge separation and transport is,
significantly lower than that observed for amorphous portions               therefore, an important question to be answered if OPV devices
of RR-P3HT, yet definitively nonzero.                                       are to be truly optimized.
    The result of a finite miscibility with PCBM at thermo-                     In summary, both the mixing and demixing experiments
dynamic equilibrium has substantial consequences for the                    reported here point to the conclusion that significant levels of

    r 2010 American Chemical Society                                 3164                  DOI: 10.1021/jz101276h |J. Phys. Chem. Lett. 2010, 1, 3160–3166

PCBM will be mixed within amorphous regions of P3HT and                                    from Interpenetrating Polymer Networks. Nature 1995, 376,
MDMO-PPVusing common device processing conditions. When                                    498–500.
accounting for the known high crystallinity in RR-P3HTand the                       (2)    Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J.
demonstration here that PCBM does not intercalate the crystals,                            Polymer Photovoltaic Cells: Enhanced Efficiencies via a Net-
the actual composition in the amorphous regions is quite high.                             work of Internal Donor-Acceptor Heterojunctions. Science
                                                                                           1995, 270, 1789–1791.
Thus, three phases were found in P3HT-based blends: pure
                                                                                    (3)                        u
                                                                                           Scharber, M. C.; M€hlbacher, D.; Koppe, M.; Denk, P.; Waldauf,
P3HT crystallites, pure PCBM crystallites, and an amorphous
                                                                                           C.; Heeger, A. J.; Brabec, C. J. Design Rules for Donors in Bulk-
polymer-rich phase. No PCBM-rich phase was found to exist in                               Heterojunction Solar Cells - Towards 10% Energy-Conversion
either polymer system. Since thermal annealing is the current                              Efficiency. Adv. Mater. 2006, 18, 789–794.
technique used to produce devices with the maximum perfor-                          (4)    Gunes, S.; Neugebauer, H.; Sariciftci, N. S. Conjugated
mance for this system, we would speculate that thermodynamic                               Polymer-Based Organic Solar Cells. Chem. Rev. 2007, 107,
phase separation would result in morphologies where PCBM                                   1324–1338.
molecules are surrounded by those of P3HT, which under the                          (5)    Dennler, G.; Scharber, M. C.; Brabec, C. J. Polymer-Fullerene
current paradigm of device function would result in charge                                 Bulk-Heterojunction Solar Cells. Adv. Mater. 2009, 21, 1323–
blocking, traps, and recombination centers within percolation                              1338.
pathways. This issue must be reconciled with the known high-                        (6)    Po, R.; Maggini, M.; Camaioni, N. Polymer Solar Cells: Recent
performance and external quantum efficiencies in these                                     Approaches and Achievements. J. Phys. Chem. C 2010, 114,
devices, potentially requiring a refinement of the physical                                695–706.
                                                                                    (7)    Ma, W.; Gopinathan, A.; Heeger, A. J. Nanostructure of the
model describing their function. Since BHJ OPVs are typi-
                                                                                           Interpenetrating Networks in Poly(3-hexylthiophene)/Fuller-
cally created through a mixture of donor and acceptor                                      ene Bulk Heterojunction Materials: Implications for Charge
components tailored to be codissolved in solution, pure                                    Transport. Adv. Mater. 2007, 19, 3656–3659.
phases in actual devices will be unlikely in most systems,                          (8)    Chatten, A. J.; Tuladhar, S. M.; Choulis, S. A.; Bradley, D. D. C.;
and our results may in fact prove rather general. Further                                  Nelson, J. Monte Carlo Modelling of Hole Transport in MDMO-
correlation and understanding of how molecular miscibility                                 PPV: PCBM Blends. J. Mater. Sci. 2005, 40, 1393–1398.
impacts nanoscale morphology and device performance is                              (9)    Guo, J.; Ohkita, H.; Benten, H.; Ito, S. Charge Generation and
therefore needed, particularly using other donor polymers,                                 Recombination Dynamics in Poly(3-hexylthiophene)/Fuller-
to illuminate the full mechanism of charge separation and                                  ene Blend Films with Different Regioregularities and Mor-
transport in these devices.                                                                phologies. J. Am. Chem. Soc. 2010, 132, 6154–6164.
                                                                                    (10)   Pensack, R. D.; Asbury, J. B. Beyond the Adiabatic Limit:
                                                                                           Charge Photogeneration in Organic Photovoltaic Materials.
SUPPORTING INFORMATION AVAILABLE A selection of                                            J. Phys. Chem. Lett. 2010, 1, 2255–2263.
STXM images of films used to probe miscibility is shown,                            (11)   Cates, N. C.; Gysel, R.; Beiley, Z.; Miller, C. E.; Toney, M. F.;
highlighting the large scale morphology of phase separation. This                          Heeney, M.; McCulloch, I.; McGehee, M. D. Tuning the Proper-
material is available free of charge via the Internet at http://pubs.                      ties of Polymer Bulk Heterojunction Solar Cells by Adjusting                                                                                   Fullerene Size to Control Intercalation. Nano Lett. 2009, 9,
AUTHOR INFORMATION                                                                  (12)    McNeill, C. R.; Watts, B.; Thomsen, L.; Belcher, W. J.; Greenham,
                                                                                            N. C.; Dastoor, P. C. Nanoscale Quantitative Chemical
Corresponding Author:                                                                       Mapping of Conjugated Polymer Blends. Nano Lett. 2006, 6,
*To whom correspondence should be addressed. E-mail: harald_ade@                            1202–1206.                                                                           (13)    Cates, N. C.; Gysel, R.; Dahl, J. E. P.; Sellinger, A.; McGehee,
                                                                                            M. D. Effects of Intercalation on the Hole Mobility of Amor-
                                                                                            phous Semiconducting Polymer Blends. Chem. Mater. 2010,
ACKNOWLEDGMENT The authors gratefully thank Fred Stevie                                     22, 3543–3548.
for dynamic SIMS measurements. Training and beamline support                        (14)    Watts, B.; Belcher, W. J.; Thomsen, L.; Ade, H.; Dastoor, P. C. A
were provided at beamlines 5.3.2 and 7.3.3. by David Kilcoyne and                           Quantitative Study of PCBM Diffusion during Annealing of
Cheng Wang, respectively. SIMS sample preparation was assisted                              P3HT:PCBM Blend Films. Macromolecules 2009, 42, 8392–
by Jaewook Seok. Work by NCSU was supported by the U.S.                                     8397.
Department of Energy, Office of Science, Basic Energy Science,                      (15)    McNeill, C. R.; Watts, B.; Thomsen, L.; Belcher, W.; Swaraj, S.;
Division of Materials Science and Engineering under Contract DE-                            Ade, H.; Dastoor, P. C. Evolution of the Nanomorphology of
FG02-98ER45737. Work at Cambridge University was funded by the                              Photovoltaic Polyfluorene Blends: Sub-100 nm Resolution
Engineering and Physical Sciences Research Council (EP/E051804/1,                           with X-ray Spectromicroscopy. Nanotechnology 2008, 19,
EP/G068356/1). STXM data were acquired at beamline 5.3.2 and                                424015.
GIWAXS at beamline 7.3.3. at the Advanced Light Source,                             (16)                                                 u
                                                                                            Swaraj, S.; Wang, C.; Yang, H.; Watts, B.; L€ning, J.; R.McNeill,
Berkeley, which is supported by the Director, Office of Science,                            C.; Ade, H. Nanomorphology of Bulk Heterojunction Photo-
Office of Basic Energy Sciences, of the U.S. Department of Energy                           voltaic Thin Films Probed with Resonant Soft X-ray Scatter-
under Contract No. DE-AC02-05CH11231.                                                       ing. Nano Lett. 2010, 10, 2863–2869.
                                                                                    (17)    Brabec, C. J.; Shaheen, S. E.; Winder, C.; Sariciftci, N. S.;
                                                                                            Denk, P. Effect of LiF/Metal Electrodes on the Performance of
                                                                                            Plastic Solar Cells. Appl. Phys. Lett. 2002, 80, 1288–1290.
(1)     Halls, J. J. M.; Walsh, C. A.; Greenham, N. C.; Marseglia, E. A.;           (18)    Ma, W.; Yang, C.; Gong, X.; Lee, K.; Heeger, A. J. Thermally
        Friend, R. H.; Moratti, S. C.; Holmes, A. B. Efficient Photodiodes                  Stable, Efficient Polymer Solar Cells with Nanoscale Control

      r 2010 American Chemical Society                                       3165                    DOI: 10.1021/jz101276h |J. Phys. Chem. Lett. 2010, 1, 3160–3166

         of the Interpenetrating Network Morphology. Adv. Funct.                              Fullerene Photovoltaics: The Role of Carrier Transport in
         Mater. 2005, 15, 1617–1622.                                                          All-Solution-Processed P3HT/PCBM Bilayer Solar Cells. J Phys.
(19)     Reyes-Reyes, M.; Kim, K.; Carroll, D. L. High-Efficiency Photo-                      Chem. C 2009, 113, 20050–20060.
         voltaic Devices Based on Annealed Poly(3-hexylthiophene) and                  (35)   Tuladhar, S. M.; Poplavskyy, D.; Choulis, S. A.; Durrant, J. R.;
         1-(3-Methoxycarbonyl)-propyl-1-phenyl-(6,6)C61 Blends. Appl.                         Bradley, D. D. C.; Nelson, J. Ambipolar Charge Transport in
         Phys. Lett. 2005, 87, 083506.                                                        Films of Methanofullerene and Poly(phenylenevinylene)/
(20)     Kilcoyne, A. L. D.; Tyliszczak, T.; Steele, W. F.; Fakra, S.;                        Methanofullerene Blends. Adv. Funct. Mater. 2005, 15,
         Hitchcock, P.; Franck, K.; Anderson, E.; Harteneck, B.; Rightor,                     1171–1182.
         E. G.; Mitchell, G. E.; et al. Interferometer-Controlled Scanning             (36)   McNeill, C. R.; Watts, B.; Thomsen, L.; Belcher, W. J.; Kilcoyne,
         Transmission X-ray Microscopes at the Advanced Light Source.                         A. L. D.; Greenham, N. C.; Dastoor, P. C. X-ray Spectromicro-
         J. Synchrotron Radiat. 2003, 10, 125–136.                                            scopy of Polymer/Fullerene Composites: Quantitative Che-
(21)     Warwick, T.; Ade, H.; Kilcoyne, A. L. D.; Kritscher, M.; Tylisczcak,                 mical Mapping. Small 2006, 2, 1432–1435.
         T.; Fakra, S.; Hitchcock, P.; Hitchcock, A. P.; Padmore, H. A. A
         New Bend-Magnet Beamline for Scanning Transmission X-Ray
         Microscopy at the Advanced Light Source. J. Synchrotron Radiat.
         2002, 9, 254–257.
(22)     Ade, H.; Zhang, X.; Cameron, S.; Costello, C.; Kirz, J.; Williams,
         S. Chemical Contrast in X-ray Microscopy and Spatially
         Resolved XANES Spectroscopy of Organic Specimens.
         Science 1992, 258, 972–975.
(23)     Ade, H.; Hitchcock, A. P. NEXAFS Microscopy, Resonant
         Scattering and Resonant Reflectivity: Composition and Or-
         ientation Probed in Real and Reciprocal Space. Polymer
         2008, 49, 643–675.
(24)     Ho, P. K.-H.; Chua, L.-L.; Dipankar, M.; Gao, X. Y.; Qi, D. C.;
         Wee, A. T.-S.; Chang, J.-F.; Friend, R. H. Solvent Effects on
         Chain Orientation and Interchain π-Interaction in Conju-
         gated Polymer Thin Films: Direct Measurements of the Air
         and Substrate Interfaces by Near-Edge X-ray Absorption
         Spectroscopy. Adv. Mater. 2007, 19, 215–221.
(25)     Henke, B. L.; Gullikson, E. M.; Davis, J. C. X-ray Interactions:
         Photoabsorption, Scattering, Transmission, and Reflection at
         E = 50-30 000 eV, Z = 1-92. At. Data Nucl. Data Tables
         1993, 54, 181–342.
(26)     Watts, B.; Swaraj, S.; Nordlund, D.; Luning, J.; Ade, H.,
         Calibrated NEXAFS Spectra of Common Conjugated Poly-
         mers. J. Chem. Phys., in press.
(27)     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.
(28)     Erb, T.; Zhokhavets, U.; Gobsch, G.; Raleva, S.; St€hn, B.;u
         Schilinsky, P.; Waldauf, C.; Brabec, C. J. Correlation Between
         Structural and Optical Properties of Composite Polymer/
         Fullerene Films for Organic Solar Cells. Adv. Funct. Mater.
         2005, 15, 1193–1196.
(29)                                                                   e
         Woo, C. H.; Thompson, B. C.; Kim, B. J.; Toney, M. F.; Frchet,
         J. M. J. The Influence of Poly(3-hexylthiophene) Regioregu-
         larity on Fullerene-Composite Solar Cell Performance. J. Am.
         Chem. Soc. 2008, 130, 16324–16329.
(30)     Sirringhaus, H.; Brown, P. J.; Friend, R. H.; Nielsen, M. M.;
         Bechgaard, K.; Langeveld-Voss, B. M. W.; Spiering, A. J. H.;
         Janssen, R. A. J.; Meijer, E. W.; Herwig, P.; et al. Two-Dimen-
         sional Charge Transport in Self-Organized, High-Mobility
         Conjugated Polymers. Nature 1999, 401, 685–688.
(31)     Warren, B. X-ray Diffraction; Dover: Mineola, NY, 1990.
(32)     Clark, J.; Chang, J.-F.; Spano, F. C.; Friend, R. H.; Silva, C.
         Determining Exciton Bandwidth and Film Microstructure in
         Polythiophene Films Using Linear Absorption Spectroscopy.
         Appl. Phys. Lett. 2009, 94, 163306.
(33)     Bavel, S. S. v.; Sourty, E.; With, G. d.; Loos, J. Three-Dimen-
         sional Nanoscale Organization of Bulk Heterojunction Poly-
         mer Solar Cells. Nano Lett. 2008, 9, 507–513.
(34)     Ayzner, A. L.; Tassone, C. J.; Tolbert, S. H.; Schwartz, B. J.
         Reappraising the Need for Bulk Heterojunctions in Polymer-

       r 2010 American Chemical Society                                         3166                   DOI: 10.1021/jz101276h |J. Phys. Chem. Lett. 2010, 1, 3160–3166

To top