J807 Paper

Self-Assembling Fullerenes for Improved Bulk-Heterojunction Photovoltaic Devices Robert D. Kennedy, Alexander L. Ayzner, Darcy D. Wanger, Christopher T Day, Merissa Halim, Saeed I. Khan, Sarah H. Tolbert,* Benjamin J. Schwartz,* and Yves Rubin* Department of Chemistry and Biochemistry, University of California, Los Angeles, Los Angeles, CA 90095-1569 *Corresponding authors: rubin@chem.ucla.edu, schwartz@chem.ucla.edu, tolbert@chem.ucla.edu SUPPORTING INFORMATION ? EXPERIMENTAL SECTION 1) Experimental procedures 2) Spectroscopic data for compounds 1a and 1b 3) X-ray data and discussion for compounds 1a and 1b 4) Device Fabrication 1. Experimental Procedures General. All experiments were carried out under argon. THF was distilled from Na benzophenone ketyl. Anhydrous grade ODCB and CS 2 were used as purchased. Silica gel (60 Å porosity, 200 x 400 mesh) was purchased from Sorbent Technologies. All other reagents were commercially available and used as received. All 1 H (400 MHz) and 13 C{ 1 H} (100 MHz) spectra were recorded on a Bruker ARX400 spectrometer. Spectra are reported in parts per million from residual protons of the deuterated solvent (? 7.26 ppm for chloroform) for 1 H NMR and from solvent carbon (? 77.00 ppm for chloroform) for 13 C{ 1 H} NMR. Low-resolution mass spectra were measured using an Applied Biosystems Voyager DE-STR MALDI-TOF. High-resolution mass spectra were measured using an IonSpec Ultima 7T FT-ICR-MS with a MALDI ion source. IR spectra were recorded on a Jasco FT/IR-420 spectrometer. C 60 (4-t-butylphenyl) 5 H (1a): 1 To a mixture of magnesium turnings (800 mg, 32.9 mmol) and a single crystal of iodine in THF (5 mL) was added 4-t-butylbromobenzene (3.00 mL, 3.69 g, 17.3 mmol, 12.5 eq) in dry THF (15 mL) over 20 minutes. Local heating with a heat gun was used to initiate the reaction. The reaction mixture was maintained at reflux for a further 30 minutes, cooled to room temperature, then transferred by syringe, over 5 minutes, to a vigorously stirring suspension of CuBr?SMe 2 (3.60 g, 17.5 mmol, 12.6 eq) in THF (2 mL) at -78 °C. After 15 minutes, C 60 (1.00 g, 1.39 mmol, 1 eq) in anhydrous ODCB (60 mL) was added over 10 minutes. The mixture was allowed to warm to room temperature and stirred overnight. The mixture was quenched with saturated aqueous NH 4 Cl (1 mL), filtered through a short pad of silica gel with toluene, and the solvent was removed under reduced pressure. Purification by flash chromatography on silica gel (CS 2 ) and evaporation of the solvent gave 1.44 g (1.04 mmol, 75%) of the product as a bright red, amorphous solid: 1 H NMR (400 MHz, CS 2 /CDCl 3 9:1) ? (ppm) 1.27 (s, 3H), 1.31 (s, 6H), 1.37 (s, 6H), 5.18 (s, 1H), 7.09 (d, J = 8.8 Hz, 2H), 7.16 (d, J = 8.4 Hz, 4H), 7.23 (d, J = 8.4 Hz, 2H), 7.33 (d, J = 8.4 Hz, 4H) 7.45 (d, J = 8.8 Hz, 4H), 7.68 (d, J = 8.4 Hz, 4H); 13 C NMR (100 MHz, CS 2 /CDCl 3 9:1) ? (ppm) 31.08, 31.15, 31.20, 34.03, 34.13, 34.25, 58.35, 58.46, 60.48, 62.74, 125.20, 125.40, 125.54, 127.39, 127.78, 127.95, 136.55, 136.62, 142.59, 143.01, 143.78, 143.96, 143.99, 144.11, 144.18, 144.24, 144.44, 1 Matsuo, Y.; Tahara, K.; Morita, K.; Matsuo, K.; Nakamura, E. Angew. Chem. Int. Ed. 2007, 46, 2844- 2847. 145.32, 145.55, 145.82, 146.13, 146.73, 146.93, 147.04, 147.57, 147.76, 147.91, 148.08, 148.23, 148.51, 148.58, 149.56, 149.92, 150.23, 151.58, 152.14, 152.18, 156.26; FT-IR (KBr) ? (cm -1 ) 3086 (w), 3053 (w), 3030 (w), 2958 (s), 2900 (m), 2865 (m), 1900 (w), 1784 (w), 1608 (w), 1510 (m), 1458 (m), 1404 (w), 1394 (w), 1362 (m), 1269 (m), 1200 (w), 1111 (m), 1018 (m), 835 (m), 818 (m), 588 (m), 561 (m), 536 (m); MALDI-TOF- MS (-), anthracene, m/z 1386.5 [M - ]; MALDI-FTICR-HRMS (-), anthracene, calcd for C 110 H 65 : 1385.5086 [(M-H) - ]; found: 1385.5083. C 60 (3-tolyl) 5 H (1b): 1 To a mixture of magnesium turnings (400 mg, 16.5 mmol) and a single crystal of iodine in THF (3 mL) was added 3-bromotoluene (1.05 mL, 1.48 g, 8.67 mmol, 12.5 eq) in dry THF (8 mL) over 20 minutes. Local heating with a heat gun was used to initiate the reaction. The reaction mixture was maintained at reflux for a further 30 minutes, cooled to room temperature, then transferred by syringe, over 5 minutes, to a vigorously stirring suspension of CuBr?SMe 2 (1.80 g, 8.76 mmol, 12.6 eq) in THF (2 mL) at -78 °C. After 15 minutes, C 60 (500 mg, 0.694 mmol, 1 eq) in anhydrous ODCB (40 mL) was added over 10 minutes. The mixture was allowed to warm to room temperature and stirred overnight. The mixture was quenched with saturated aqueous NH 4 Cl (1 mL), filtered through a short pad of silica gel with toluene, and the solvent was removed under reduced pressure. Purification by flash chromatography on silica gel (CS 2 ) gave 653 mg (0.555 mmol, 80%) of the product as a deep red crystalline solid: 1 H NMR (400 MHz, CS 2 /CDCl 3 9:1) ? (ppm) 2.04 (s, 3H), 2.10 (s, 6H), 2.25 (s, 6H), 5.22 (s, 1H), 6.97-7.14 (m, 9H), 7.24 (t, J = 7.6 Hz, 2H), 7.26 (d, J = 7.2 Hz, 1H), 7.33 (s, 2H), 7.40 (d, J = 7.6 Hz, 2H), 7.52 (s, 2H), 7.58 (d, J = 7.6 Hz, 2H); 13 C NMR (100 MHz, CS 2 /CDCl 3 9:1) ? (ppm) 20.94, 21.02, 21.15, 58.49, 58.53, 60.57, 62.92, 124.23, 124.88, 124.98, 127.56, 127.81, 128.04, 128.21, 128.28, 128.46, 128.77, 128.91, 128.95, 137.78, 138.08, 139.34, 139.42, 142.91, 143.40, 143.80, 143.83, 143.90, 144.00, 144.12, 144.20, 145.08, 145.25, 145.41, 145.61, 145.81, 146.56, 146.74, 146.84, 147.40, 147.45, 147.74, 147.92, 148.05, 148.33, 148.38, 148.42, 151.22, 151.92, 152.05, 155.65; FT-IR (KBr) ? (cm -1 ) 3037 (m), 2949 (m) 2916 (m), 2856 (m), 1603 (s), 1585 (m), 1485 (s), 1460 (m), 1377 (m), 1286 (w), 1234 (w), 1173 (w), 1093 (w), 1036 (w), 771 (m), 754 (m), 696 (s), 542 (s), 436 (w); MALDI-TOF-MS (-) anthracene, m/z 1176.3 [M - ]; MALDI-FTICR-HRMS (-) anthracene, calcd for C 95 H 35 : 1175.2739 [(M-H) - ]; found: 1175.2725. 2) Spectroscopic data for compounds 1a and 1b 3) X-ray data and discussion for compounds 1a and 1b. X-Ray Structure Determination of 1a?(C 5 H 12 ) 3 , 1b?CHCl 3, 1b?CS 2, 1b?C 6 H 4 Cl 2, 1b?C 6 H 5 CH 3 , and 1b?(C 6 H 5 Cl) 1.875 . The X-ray intensity data were measured at 100K on a Bruker SMART APEX2 CCD- based X-ray diffractometer system equipped with a Mo-target X-ray tube (?= 0.71073 Å) operated at 2250 watts power. The detector was placed at a distance of 6.000 cm from the crystal. A total of 1080 frames were collected with a scan width of 0.3° in ?, with an exposure time ranging from 30 to 90 sec/frame. The data frames were collected using the program APEX2 and processed using the program SAINT routine within APEX2. 2 The data were corrected for absorption based on the multi-scan technique as implemented in SADABS. Structure solution and refinement were performed using Bruker SHELXTL (Version 6.12) Software Package. Mercury 1.4.2, Crystal Maker 7.2, and CrystalExplorer 2.0 were used to visualize the crystal structures. Crystal data and experimental parameters are summarized in Table S1. 2 APEX2 v2.2-0 and SAINT v7.34A data collection and data processing programs, respectively. Bruker Analytical X-ray Instruments, Inc., Madison, WI; SADABS v2008/1 semi-empirical absorption correction program. G.M. Sheldrick, University of Göttingen, Germany. Table S1. Crystal data and experimental parameters for 1a?(C 5 H 12 ) 3 , 1b?CHCl 3, 1b?CS 2, 1b?C 6 H 4 Cl 2, 1b?C 6 H 5 CH 3, and 1b?(C 6 H 5 Cl) 2 . 1a?(C 5 H 12 ) 3 1b?CHCl 3 1b?CS 2 1b?1,2-C 6 H 4 Cl 2 1b?C 6 H 5 CH 3 1b?(C 6 H 5 Cl) 1.875 crystal system orthorhombic orthorhombic orthorhombic monoclinic monoclinic monoclinic space group Pnma Pnma Pnma P2 1 /c P2 1 /c P2 1 /n a, Å 21.661(3) 18.298(5) 18.311(5) 16.549(2) 16.469(3) 14.199(1) b, Å 22.721(4) 17.955(5) 17.969(4) 18.529(2) 18.464(3) 18.842(2) c, Å 17.128(3) 17.352(5) 17.344(5) 38.904(4) 38.870(7) 47.797(5) ?, degree 90.00 90.00 90.00 90.00 90.00 90.00 ®, degree 90.00 90.00 90.00 93.1690(10) 92.545(2) 93.598(1) ?, degree 90.00 90.00 90.00 90.00 90.00 90.00 V, Å 3 8430(2) 5701(3) 5707(3) 11912(2) 11808(4) 12762(2) Z 4 4 4 8 8 4 T, K 100(2) 100(2) 100(2) 100(2) 100(2) 100(2) crystal description red platelet red platelet Red prism red block ?light red? platelet red prism crystal size, mm 0.30 x 0.20 x 0.20 0.40 x 0.10 x 0.10 0.20 x 0.20 x 0.05 0.50 x 0.30 x 0.10 0.20 x 0.10 x 0.10 0.40 x 0.20 x 0.10 2? min , 2? max , degree 7.60, 58.20 7.54, 56.6 7.74, 58.24 7.68, 59.96 7.68, 46.58 7.56, 58.20 no. of refl. (unique) 11521 7277 7858 33975 16895 33987 no. of refl. (I>2?(I)) 7222 4885 5934 24109 8323 23739 no. of parameters 719 476 475 1796 1817 2000 R1, wR2 (all data) 0.1241, 0.2391 0.1135, 0.2282 0.0838, 0.1892 0.1102, 0.2387 0.1632, 0.1643 0.1088, 0.2144 R, wR (I>2?(I)) 0.0758, 0.1961 0.0769, 0.2001 0.0629, 0.1705 0.0794, 0.2140 0.0640, 0.1243 0.0750, 0.1916 GOF on F 2 1.021 1.057 1.025 1.032 0.969 1.021 ?, e Å -3 0.410, -0.450 1.102, -1.022 1.073, -1.079 1.062, -1.088 0.405, -0.383 1.101, -0.975 Figure S1. (a) ORTEP representation of 1a?(C 5 H 12 ) 3 . Thermal ellipsoids at 50% probability. (b) Hirshfeld surface of 1a?(C 5 H 12 ) 3 , viewed along the a axis. (c) Antiparallel stacking motif in 1a?(C 5 H 12 ) 3 viewed along the bc plane. (d) Close-packed stacks and solvent channels in 1a?(C 5 H 12 ) 3 viewed along the a axis. Hydrogen atoms have been omitted for clarity in all cases, and selected solvent molecules have been shown in a space-filling style. The structure of compound 1a?(C 5 H 12 ) 3 is presented in Figure S1. Crystal data and selected experimental parameters are presented in Table S1. Compound 1a?(C 5 H 12 ) 3 has typical shuttlecock connectivity (Fig. S1a), with disorder in the 4-t-butylphenylene groups and solvent molecules. The cyclopentadiene hydrogen could not be unambiguously located. The phenylene groups are aligned approximately tangentially to a circle circumscribing their centroids, forming a bowl-shaped cavity. The Hirshfeld surface of a single shuttlecock is shown in Figure S1b. The body of an adjacent fullerene nestles in this cavity, producing a 1-dimensional stacking motif that extends along the a axis (Fig. S1c,d). The distance between the centroids of fullerene bodies within each stack is 10.832 Å, and the corresponding closest approach of carbon atoms is 4.035 Å. The stacks are roughly hexagonally packed, with adjacent stacks aligned in an antiparallel fashion (Fig. S1c-h). Three co-crystallized solvent molecules per fullerene fill the columnar voids between the fullerenes forming almost complete sheathes around each stack. The closest approaches between fullerene centroids in adjacent stacks are 14.854 and 15.857 Å, and the density of stacks, calculated by dividing the sum of unit cell dimensions, b and c, by the number of stacks in the unit cell, is 0.514 nm -2 . Figure S2. (a) ORTEP representation of 1b?CHCl 3 with thermal ellipsoids at 50% probability. (b) ac layer viewed along b axis. Solvent molecules are shown in a space- filling style. Close-approaching carbons are highlighted. (c) b-axis chain viewed along a axis. Close-approaching carbons are highlighted. (d) Close-contact network. Grey balls (drawn with an arbitrary radius) represent C 60 centroids. Bonds represent centroid- centroid distances under 10.5 Å. 1b?CHCl 3 has typical shuttlecock connectivity (Fig. S2a). One 3-tolyl group is disordered over two positions with equal occupancy (bottom of Fig. S2a) producing an apparent mirror plane which bisects the embedded cyclopentadiene ring. A disordered chloroform molecule resides in the shallow cavity formed by the five 3-tolyl groups. Each fullerene has four near-neighbors (distance between C 60 centroids < 10.5 Å), arranged in an approximately tetrahedral geometry. The fullerenes are organized into perpendicular zigzag chains which extend along the a and b axes. A single ac layer, viewed along the b axis, is shown in Fig. S2b. Chains, running parallel to the a axis, are defined by C 60 centroid-centroid distances of 10.436 Å, and two specific C?C close-contacts of 3.509 Å. Fig. S2c shows a perpendicular chain, aligned parallel to the b axis, viewed along the a axis. The C 60 centroid-centroid distance is 9.894 Å, and there are 10 specific C?C close contacts, corresponding to close-approaching, parallel hexagons, ranging between 3.412 and 3.551 Å. These b-axis chains connect adjacent ac layers, forming a diamond-like network of close contacts (Fig. S2d). The local and extended crystal structure of 1b?CS 2 is very similar to 1b?CHCl 3 , with a disordered CS 2 molecule in the place of disordered CHCl 3 . Selected crystallographic data is presented in Table S1. Figure S3. (a) ORTEP representation of 1b?1,2-C 6 H 4 Cl 2 . Thermal ellipsoids at 50% probability. (b) bc layers viewed along bc plane. (c) Single bc layer viewed along a axis. (d) Close-contact network. Yellow and blue balls (with an arbitrary radius) represent C 60 centroids. Bonds represent centroid-centroid distances under 10.5 Å. Purple balls represent the centroids of 1,2-C 6 H 4 Cl 2 molecules. (e) Single ac slice viewed along b axis. Hydrogens removed for clarity in all cases. Solvent molecules are shown in a space- filling style. 1b?1,2-C 6 H 4 Cl 2 (Fig. S3a) crystallizes with a solvent molecule (1,2-C 6 H 4 Cl 2 ) lying within the shallow cavity formed by the 3-tolyl groups. The extended crystal structure is lamellar, with a periodicity of 16.55 Å. Layers of solvent, encapsulated by the 3-tolyl groups, are sandwiched between adjacent layers composed of fullerene bodies (Fig. S3b). The fullerene bodies are densely and approximately hexagonally close-packed; each fullerene has five near neighbors (C 60 centroid-centroid distances are between 9.806 and 10.227 Å) forming a 2-dimensional network of close contacts (Fig. S3c,d). The shortest specific C?C close contacts range between 3.270 and 3.449 Å. Fig. S3e shows an ac slice, viewed along the b axis, describing the relationship between the bc layers. The local and extended crystal structure of 1b?C 6 H 5 CH 3 is very similar to 1b?1,2-C 6 H 4 Cl 2 , with a disordered toluene molecule in the place of disordered 1,2-C 6 H 4 Cl 2 . Selected crystallographic data is presented in Table S1. Figure S4. (a) ORTEP representation of 1b?1,2-C 6 H 4 Cl 2 . Thermal ellipsoids at 50% probability. (b) Single (10-1) layer, viewed perpendicular to the (10-1) plane. Selected solvent molecules are shown in the space- filling style. (c) Three (10-1) layers, viewed along the b axis. (d) Close-contact network, viewed parallel to the (10-1) plane. Balls (with an arbitrary radius) represent C 60 centroids. Bonds represent centroid-centroid distances under 10.5 Å. In all cases, hydrogen atoms are removed for clarity. Compound 1b?(C 6 H 5 Cl) 2 (Fig. S4) also crystallizes with a solvent molecule in the shallow cavity formed by the 3-tolyl groups. Each fullerene has 3 near-neighbors (C 60 centroid-centroid distances are between 10.095 and 10.318 Å), centroid forming a layered, puckered, honeycomb-like network of close-approaching fullerenes. Shortest C? C close contacts range between 3.513 and 3.616 Å. The layers lie parallel to the (10-1) plane, and co-crystallized chlorobenzene molecules occupy the interstitial voids. The extended crystal structure is shown in Fig. S4b-d. 4. Device Fabrication We fabricated bulk heterojunction photovoltaic devices according to well-established protocols. 3 Pre-cleaned indium-doped tin oxide (ITO) substrates were coated with a thin layer of pol(ethylenedioxythiophene-2,5-diyl):poly(styrenesulfonic acid) (Baytron V P). Regioregular P3HT (Rieke Metals) ? fullerene blend solutions were heated at 40 ?C overnight in a nitrogen atmosphere and further heated at 55 ?C for several hours to ensure complete dissolution of both components prior to spin-coating. All solutions were brought back to room temperature and passed through a 0.45 micron PTFE filter prior to active layer deposition. The active layer was spin-coated on top of the coated ITO slides at 700 rpm for 1.5 minutes out of a solution of P3HT (10 mg/mL) in o-dichlorobenzene blended with the shuttlecock fullerenes at a weight ratio of 1:0.45 P3HT:1a and the molar equivalent of compound 1b (1:0.35 P3HT:1b). A total of 10 nm of calcium followed by 40 nm of aluminum was then deposited under vacuum at a rate of 1 Å/sec, resulting in device areas of 6.52 mm 2 . Thermal annealing of completed devices was carried out on a digitally controlled hotplate covered with a glass Petri dish at 150°C for 20 minutes in an argon atmosphere. Photovoltaic device performance was tested in an argon atmosphere with a Keithley 2400 source meter. Radiation from a xenon arc lamp incident on the device at 100 mW/cm 2 was delivered through an Oriel liquid light guide coupled with an AM 1.5 filter. Error bars in the I-V plots represent one standard deviation. 3 Ma, W.; Yang, C.; Gong, X.; Lee, K.; Heeger, A. Adv. Funct. Mater. 2005, 15, 1617-1622. JP-U Microsoft Word - SC Paper SI Sept 23-RDKedit21.doc