Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Mater Chem. Author manuscript; available in PMC 2011 January 1.
Published in final edited form as:
J Mater Chem. 2010 January 1; 20(25): 5280–5293.
doi:  10.1039/C0JM00037J
PMCID: PMC2947801

Synthesis and characterization of highly photoresponsive fullerenyl dyads with a close chromophore antenna–C60 contact and effective photodynamic potential


We report the synthesis of a new class of photoresponsive C60–DCE–diphenylaminofluorene nanostructures and their intramolecular photoinduced energy and electron transfer phenomena. Structural modification was made by chemical conversion of the keto group in C60(>DPAF-Cn) to a stronger electron-withdrawing 1,1-dicyanoethylenyl (DCE) unit leading to C60(>CPAF-Cn) with an increased electronic polarization of the molecule. The modification also led to a large bathochromic shift of the major band in visible spectrum giving measureable absorption up to 600 nm and extended the photoresponsive capability of C60–DCE–DPAF nanostructures to longer red wavelengths than C60(>DPAF-Cn). Accordingly, C60(>CPAF-Cn) may allow 2γ-PDT using a light wavelength of 1000–1200 nm for enhanced tissue penetration depth. Production efficiency of singlet oxygen by closely related C60(>DPAF-C2M) was found to be comparable with that of tetraphenylporphyrin photosensitizer. Remarkably, the 1O2 quantum yield of C60(>CPAF-C2M) was found to be nearly 6-fold higher than that of C60(>DPAF-C2M), demonstrating the large light-harvesting enhancement of the CPAF-C2M moiety and leading to more efficient triplet state generation of the C60> cage moiety. This led to highly effective killing of HeLa cells by C60(>CPAF-C2M) via photodynamic therapy (200 J cm−2 white light). We interpret the phenomena in terms of the contributions by the extended π-conjugation and stronger electron-withdrawing capability associated with the 1,1-dicyanoethylenyl group compared to that of the keto group.


Single-photon excitation based photodynamic therapy (1γ-PDT) for conditions including basal-cell carcinoma, psoriasis, actinic keratosis, oesophageal dysplasia, restenosis and age-related macular degeneration (AMD), in light-accessible organs, such as the digestive tract, artery and eye has achieved success in the past decade.1 The approach provides an alternative to chemo- and radio-therapeutic treatments of cancer. Effective PDT method requires a combination of a convenient light source and the selection of photosensitizers capable of generating reactive oxygen species (ROS), such as 1O2, O2 and HO,2 that induce spatially confined cell death in the disease-affected area upon irradiation.

Two-photon excitation based photodynamic therapy3 (2γ-PDT) has an additional advantage of being able to deliver effective treatment deeper into the tissue at the focal area of irradiation, lessening damage to the healthy adjacent and overlying tissue. However, to make 2γ-PDT applications efficient, significantly larger two-photon absorption (2PA) cross-sections (σ2)4 of the photosensitizer are required that often involves the specific design of photoresponsive chromophores for this specific purpose.5 In addition, the wavelength range of 700–1200 nm generally used in the 2γ-PDT treatment enhances the penetration depth and fits well with the optical window for biological tissue.

Photoexcitation of C60 and its derivatives induces a singlet fullerenyl excited state that is transformed to the corresponding triplet excited state, via intersystem energy crossing, with nearly quantitative efficiency.6 Subsequent energy transfer from the triplet fullerene derivatives to molecular oxygen produces singlet molecular oxygen (1O2) in aerobic media. This photocatalytic effect is responsible for cytotoxicity involved in anticancer and antimicrobial photodynamic therapy using [60]fullerenyl derivatives as photosensitizers,79

We previously reported the ultrafast photoresponsive C60-antenna conjugate molecules C60(>DPAF-Cn)x (x = 1, 2 or 4) that exhibited simultaneous multiphoton excitation events with the observed significant nonlinear photophysical properties.10 These compounds showed large structure- and concentration-dependent two-photon absorption (2PA) cross-section (σ2) values and nonlinear laser intensity transmittance capability. For example, at concentrations between 10−4 or 10−2 M, both C60(>DPAF-C9)2 and C60(>DPAF-C9)4 exhibited a large fs 2PA cross-section value of 4300 or 80 GM and 8520 or 710 GM, respectively. These values were higher than 2190 or 30 GM, respectively, for the linear C60(>DPAF-C9) structure. High two-photon absorption cross-sections were correlated with increased electronic interactions and molecular polarization between the closely linked DPAF-Cn donor antenna and the C60 acceptor cage. This facilitated ultrafast energy transfer from photoexcited C60(>1DPAF*-Cn) to the C60 moiety forming the excited singlet 1C60*(>DPAF-Cn) and subsequently the triplet 3C60*(>DPAF-Cn), via intersystem crossing. These intramolecular electron and energy events are similar to those occurring with single-photon excitation. One particular application of materials exhibiting nonlinear optical (NLO) effects is the development of efficient light intensity attenuators for sensor development and personnel protection. The feasibility of this goal was demonstrated by the detection of nonlinear optical transmittance reduction responses in the femtosecond (fs) region with a low transmittance value (35–40%) for C60(>DPAF-C9)x (x = 1, 2 or 4) at the laser irradiance intensity level above 600 to 850 GW cm−2.10

In our continuing efforts to extend the maximum range of the optical absorption of DPAF-Cn moieties (centered at ~400 nm) and their corresponding subsequent fluorescence emissions from 1DPAF*-Cn state at 450–460 nm (in toluene), to longer wave-lengths in the red region for both NLO and PDT, we modified the structure of C60(>DPAF-Cn) in order to allow light harvesting across a greater visible spectrum. Accordingly, we replaced the keto functional group of C60–keto–DPAF by a highly electron-withdrawing 1,1-dicyanoethylenyl (DCE) bridging group forming C60–DCE–DPAF structures that were analogous to C60(>CPAF-Cn) derivatives.11 Interestingly, this chemical modification resulted in a large bathochromic shift of ground-state absorption of CPAF-Cn moieties beyond 400–600 nm. The solution of C60(>CPAF-Cn) adducts were found to be dark burgundy-red in color. In the molecular structure of new C60–DCE–DPAF assemblies, intramolecular electronic interactions between the 1,1-dicyanoethylenyl bridging group of the CPAF-Cn moiety and the C60 cage, via through-space periconjugation, are expected to be significantly more active than those of C60(>DPAF-Cn). This was due to the π-conjugation bonding consideration of both entities and a close inter-space contact of the cyano group to the C60 cage in only roughly 2.0–3.0 Å of the separation distance. Upon such structural variation, one particular effect observed upon the change of photophysical properties is the large enhancement of singlet oxygen (1O2) production efficiency of C60(>CPAF-Cn) in toluene in comparison with those of C60(>DPAF-Cn).

In this report, we describe the synthesis, structural characterization, and several photophysical properties of 7-(1,2-dihydro-1,2-methano[60]fullerene-61-{1,1-dicyanoethylenyl})-9,9-diethyl-2-diphenylaminofluorene C60(>CPAF-C2) and its related 9,9-di(3,5,5-trimethylhexyl) derivative C60(>CPAF-C9), in terms of intramolecular energy and electron transfer processes studied by nanosecond transient measurements. The results were correlated to highly effective killing of HeLa cells under 1γ-PDT conditions.

Materials and methods


Reagents 2-bromofluorene, malononitrile, tris(dibenzylideneacetone) dipalladium(0) and rac-2,2′-bis(diphenylphosphino)-1,1′-binaphthyl (BINAP) were purchased from Aldrich Chemicals. All other chemicals and solvents, including toluene (TN), o-dichlorobenzene (o-DCB), benzonitrile (PhCN) and CS2 were purchased from Acros Ltd. Pure C60 (99.5%) was purchased from NeoTech Product Company, Russia and used as received.

UV-vis spectra were recorded on either a Hitachi U-3410 UV spectrometer or Hewlett-Packard diode-array model 8452A UV-visible spectrophotometer. Fluorescence spectra were recorded on a Perkin-Elmer LS55 luminescence spectrometer. Infrared spectra were recorded as KBr pellets on a Nicolet 750 series FT-IR spectrometer. 1 NMR, 13C NMR and COSY spectra were recorded on either a Bruker Avance Spectrospin-400 or Bruker AC-300 spectrometer. HMQC spectra were recorded on Bruker-500 SB FT-NMR spectrometer. Mass spectroscopic measurements were performed by the use of positive ion matrix-assisted laser desorption ionization (MALDI–TOF) technique on a micromass M@LDI-LR mass spectrometer. Negative ion desorption chemical ionization (DCI) mass spectra were collected using a direct probe on a JEOL JMS-SX 102A mass spectrometer. Positive ion fast atom bombardment (FAB+) mass spectra were collected using a direct probe on a JEOL SX-102A mass spectrometer. Elemental analyses of fullerene derivatives and X-ray single crystal analysis, performed on NONIUS KAPPA CCD diffractometer, were carried out at National Instrumentation Center located at National Taiwan University, Taiwan.

Synthesis of (1,2-methano[60]fullerene-61-carbonyl)benzene, C60(>CO-Ph), 1

To the mixture of C60 (1.0 g 1.38 mmol) and phenacyl bromide (276 mg, 1.38 mmol) in toluene (700 mL) was added 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU, 0.210 ml, 1.41 mmol) under nitrogen atmosphere. After stirring at room temperature for a period of 5 h, the reaction mixture was filtered and concentrated to a 10% volume. Crude products were precipitated by the addition of methanol. The precipitates were isolated by centrifugation and purified by column chromatography (silica gel, hexane–toluene (8:2)). A chromatographic band corresponding to Rf = ~0.4 on an analytical thin-layer chromatographic plate (TLC, hexane–toluene (8:2)) was collected to afford (1,2-methano[60]fullerene-61-carbonyl)benzene 1, C60(>CO-Ph), as a brown solid (450 mg) in 52% yield (based on recovered C60); DCI-MS: calc. for 12C681616O1 m/z 838; found, 838.0. UV-vis (CHCl3, 2.0 × 10−5 M) λmax/nm (ε/L mol−1 cm−1): 260 (1.2 × 105) and 327 (3.9 × 104).

Synthesis of the model compound (1,2-methano[60]fullerene-61-{1,1-dicyanoethylene})benzene, C60(>DCE-Ph), 2

(1,2-Methano[60]fullerene-61-carbonyl)benzene 1, C60(>CO-Ph), (500 mg, 0.6 mmol) and malononitrile (78 mg, 1.2 mmol) were dissolved in dry toluene (~250 mL) under nitrogen atmosphere. Pyridine (188 mg, 0.2 mL, 2.4 mmol) was added to the solution with stirring. To the resulting mixture was then added titanium tetrachloride (560 mg) in an excess. After stirring at room temperature for a period of 5 min, the reaction mixture was quenched with water (100 mL). The organic layer was washed several times with water, dried over magnesium sulfate, and concentrated in vacuum to afford orange–red solids. Crude products were purified by preparative thin-layer chromatography (TLC, silica gel, hexane–toluene (3:2)). A chromatographic band corresponding to Rf = ~0.4 on analytical TLC (hexane–toluene (8:2)) was collected to afford (1,2-methano[60]fullerene-61-{1,1-dicyanoethylene})benzene 2, C60(>DCE-Ph), as orange–brown solids (340 mg) in 66% yield; DCI-MS: calc. for 12C711614N2 m/z 886; found, 886.0; UV-vis (CHCl3, 2.0 × 10−5 M) λmax/nm (ε/L mol−1 cm−1): 260 (1.3 × 105) and 326 (5.2 × 104); 1 NMR (400 MHz, CDCl3, ppm) δ 8.07 (m, 2H), 7.61 (m, 3H) and 5.48 (s, α-proton, 1H). 13C NMR (400 MHz, CDCl3, ppm) δ 169.0, 147.3, 145.5, 145.4, 145.3 (×3), 145.1, 144.9, 144.8, 144.7, 144.6, 143.8, 143.0 (×2), 142.4, 142.1 (×2), 142.0 (×2), 141.2, 137.4, 137.0, 134.7, 133.5, 129.7, 128.4, 112.6 (C[equivalent]N), 112.5 (C[equivalent]N), 90.4, 72.2 (CF1, CF2) and 41.0 (C61).

Synthesis of 7-[1-(1,1-dicyanoethylene)-2-bromoethyl]-9,9-diethyl-2-diphenylaminofluorene 6-C2 (BrCPAF-C2)

In a reaction flask, 7-bromoacetyl-9,9-diethyl-2-diphenylaminofluorene 5-C2 (BrDPAF-C2, 150 mg, 0.3 mmol) and malononitrile (38 mg, 0.6 mmol) were added followed by the addition of dry toluene (25 mL) under nitrogen atmospheric pressure to give a clear solution. Pyridine (92 mg, 1.2 mmol) and an excess amount of titanium tetrachloride were then added with stirring. After keeping for a period of 5.0 min at ambient temperature, the reaction mixture was quenched with water (30 mL). The resulting organic layer was washed several times with water, dried over magnesium sulfate, and concentrated in vacuo to give dark bright red solids. The crude product was purified by preparative thin-layer chromatography (TLC, silica gel) using a solvent mixture of chloroform–hexane (1:1) as eluent. A chromatographic fraction, corresponding to Rf 0.6 on an analytical TLC plate using the same eluent, was isolated to afford 7-[1-(1,1-dicyanoethylene)-2-bromoethyl]-9,9-diethyl-2-diphenylaminofluorene 6-C2 (BrCPAF-C2) as dark red solids in a yield of 67% (110 mg).

Spectroscopic data for 5-C2: FAB+-MS: calc. for 12C3112879.9Br114N116O1 m/z 510; found, m/z 509 and 511; Anal. Calc. for C31H28BrNO: C, 72.94; H, 5.49; N, 2.74. Found: C, 73.22; H, 5.62; N, 2.54%. UV-vis (CHCl3, 2.0 × 10−5 M) λmax/nm (ε/L mol−1 cm−1): 299 (2.4 × 104) and 406 (2.7 × 104); FT-IR (KBr) νmax 3037 (w), 2966 (s), 2928 (m), 2878 (w), 1674 (s), 1595 (vs), 1491 (s), 1281 (vs), 754 (s) and 698 (s) cm−1; 1H NMR (400 MHz, CDCl3, ppm) δ 7.95 (dd, J = 8 Hz, J = 1.6 Hz, 1H), 7.92 (d, J = 1.4 Hz, 1H), 7.65 (d, J = 8 Hz, 1H), 7.60 (d, J = 8 Hz, 1H), 7.28–7.09 (m, 10H), 7.05–7.02 (m, 2H), 4.49 (s, 2H), 2.05–1.84 (m, 4H) and 0.35 (t, J = 7.3 Hz, 6H); 13C NMR (400 MHz, CDCl3, ppm) δ 191.0, 152.8, 150.3, 148.9, 147.3, 134.3, 131.6, 129.3, 129.2, 128.9, 124.4, 123.1, 122.8, 121.6, 118.8, 118.1, 56.2, 32.4, 31.2 and 8.5.

Spectroscopic data for 6-C2: Anal. Calc. for C34H28BrN3: C, 73.71; H, 5.01; N, 7.52. Found: C, 74.03; H, 5.47; N, 6.91%. FAB+-MS: calc. for 12C3412879.9Br114N3 m/z 558; found, m/z 557 and 559; UV-vis (CHCl3) λmax/nm (ε/L mol−1 cm−1): 316 (2.8 × 104) and 492 (1.8 × 104); FT-IR (KBr) νmax 3031 (w), 2955 (s), 2921 (s), 2873 (m), 2846 (m), 2225 (C[equivalent]N), 1592 (vs), 1541 (m), 1489 (vs), 1463 (s), 1344 (s), 1313 (m), 1280 (vs), 817 (s), 752 (s) and 691 cm−1; 1H NMR (400 MHz, CDCl3, ppm) δ 7.69 (d, J = 8 Hz, 1H), 7.63 (d, J = 1.6 Hz, 1H), 7.62 (d, J = 8 Hz, 1H), 7.57 (d, J = 8 Hz, 1H), 7.28–7.23 (m, 4H), 7.12–7.10 (m, 4H), 7.06–7.01 (m, 4H), 4.59 (s, 2H), 2.00–1.85 (m, 4H) and 0.35 (t, J = 7.2 Hz, 6H). 13C NMR (400 MHz, CDCl3, ppm) δ 171.1, 152.7, 150.8, 149.3, 147.6, 147.1, 133.8, 131.9, 130.1, 129.4, 127.5, 124.7, 123.4, 122.9, 122.7, 121.7, 119.5, 117.8, 113.2, 112.4, 84.1, 56.5, 32.4, 28.7 and 8.5.

General procedure of the fullerene adduct preparation

Synthesis of 7-(1,2-dihydro-1,2-methanofullerene[60]-61-carbonyl)-9,9-dialkyl-2-diphenylaminofluorene monoadduct 7-Cn, C60(>DPAF-Cn)

Preparative procedures for the synthesis of 7-bromoacetyl-9,9-di(3,5,5-trimethylhexyl)-2-diphenylaminofluorene 5-C9 (BrDPAF-C9) follow those reported previously. 12 The reagent compound 5-C9 (0.97 g, 1.38 mmol) and C60 (1.0 g, 1.38 mmol) were dissolved in toluene (700 mL) under an atmospheric pressure of nitrogen. To this mixture was added 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU, 0.2 ml, 1.38 mmol) and stirred at room temperature for a period of 5 h. At the end of stirring, suspended solids of the reaction mixture were filtered off and the filtrate was concentrated to a 10% volume. Methanol (100 mL) was then added to the liquid to cause precipitation of the crude product, which was isolated by centrifugation. Isolation of the monoadduct 7-C9 was made by column chromatography (silica gel) using a solvent mixture of hexane–toluene (3:2) as eluent. The chromatographic band at Rf 0.45 on the thin-layer chromatographic (TLC, SiO2) plate using hexane–toluene (3:2) as eluent afforded C60(>DPAF-C9) 7-C9, 7-(1,2-dihydro-1,2-methanofullerene[60]-61-carbonyl)-9,9-di(3,5,5-trime-thylhexyl)-2-diphenylaminofluorene, as greenish brown solids (960 mg, 70% yield based on recovered C60).

Spectroscopic data for C60(>DPAF-C2) 7-C2: Anal. Calc. for C91H27NO: C, 95.03; H, 2.34; N, 1.2. Found: C, 94.58; H, 2.63; N, 1.00%. FAB+-MS: calc. for 12C9112714N116O1 m/z 1149; found, m/z 1150; DCI-MS: calc. for 12C9112714N116O1 m/z 1149; found, m/z 1149; UV-vis (CHCl3, 2.0 × 10−5 M) λmax/nm (ε/L mol−1 cm−1): 260 (1.5 × 105), 326 (6.5 × 104) and 410 nm (4.8 × 104); FT-IR (KBr) νmax 3029 (w), 2963 (s), 2921 (m), 2875 (w), 2853 (w), 1677 (C=O, s), 1591 (vs), 1492 (s), 1276 (s), 750 (s), 695 (s) and 524 (s) cm−1; 1H NMR (400 MHz, CDCl3, ppm) δ 8.48 (dd, J = 8 Hz, J = 1.6 Hz, 1H), 8.32 (d, J = 1.6 Hz, 1H), 7.83 (d, J = 8 Hz, 1H), 7.66 (d, J = 8 Hz, 1H), 7.29–7.11 (m, 10H), 7.07–7.03 (m, 2H), 5.69 (s, 1H), 2.13–1.89 (m, 4H) and 0.40 (t, J = 8 Hz, 6H); 13C NMR (400 MHz, CDCl3, ppm) δ 189.6 (C=O), 153.0, 150.8, 149.1, 148.2, 147.9, 147.6, 146.9, 145.6, 145.4, 145.3, 145.1, 145.0, 144.9, 144.7, 144.6 (×2), 144.3, 143.9, 143.7, 143.3, 143.1, 143.0 (×2), 142.9, 142.8, 142.5, 142.2 (×2), 142.1, 141.2, 140.9, 139.7, 136.6, 134.2, 133.5, 129.3, 129.0, 124.5, 123.2, 123.1, 122.8, 121.8, 119.2, 118.1, 72.7 (×2, CF1, CF2), 56.4, 44.4 (C61), 32.5 and 8.7.

General synthetic procedure of 7-(1,2-dihydro-1,2-methanofullerene[60]-61-(1,1-dicyanoethylene))-9,9-dialkyl-2-diphenylaminofluorene monoadduct 8-Cn, C60(>CPAF-Cn)

Method A

In a reaction flask, to a mixture of 7-(1,2-dihydro-1,2-methanofullerene[60]-61-carbonyl)-9,9-diethyl-2-diphenylaminofluorene C60(>DPAF-C2) (7-C2, 300 mg, 0.26 mmol) and malononitrile (34 mg, 0.52 mmol) dissolved in dry toluene (25 ml) under nitrogen atmosphere was added pyridine (82 mg, 1.04 mmol) with stirring. An excess amount of titanium tetra-chloride was then added. After stirring for a period of 5.0 min at ambient temperature, the reaction mixture was quenched with water (30 mL). The organic layer was subsequently washed several times with water, dried over magnesium sulfate, and concentrated in vacuo to afford orange–red solids. This crude product was purified by preparative thin-layer chromatography (TLC, silica gel) using a solvent mixture of hexane–toluene (3:2) as eluent. A chromatographic fraction, corresponding to Rf 0.3 on an analytical TLC plate using the same eluent, was isolated to give 7-(1,2-dihydro-1,2-methanofullerene[60]-61-{1,1-dicyanoethylene})-9,9-diethyl-2-diphenylaminofluorene 8-C2, C60(>CPAF-C2), as orange–red solids in 67% yield (210 mg).

Method B

To C60 (1.0 g, 1.38 mmol) dissolved in dry toluene (700 mL) was added 7-[1-(1,1-dicyanoethylene)-2-bromoethyl]-9,9-diethyl-2-diphenylaminofluorene 6-C2 (BrCPAF-C2, 770 mg, 1.38 mmol) in dry toluene (30 mL) under nitrogen. 1,8-Dia-zabicyclo[5.4.0]undec-7-ene (DBU, 0.205 mL, 1.38 mmol) was then added to this mixture. After stirring for a period of 5 h at room temperature, the reaction mixture was filtered and concentrated to a 10% volume. The crude product was precipitated by the addition of methanol and isolated by centrifugation. The solid was further purified by column chromatography (silica gel, hexane–toluene (3:2) as eluent). A chromatographic fraction, corresponding to Rf 0.3 on an analytical TLC plate using the same eluent, was isolated to afford 7-(1,2-dihydro-1,2-methanofullerene[60]-61-{1,1-dicyanoethylene})-9,9-diethyl-2-diphenylaminofluorene 8-C2, C60(>CPAF-C2), as orange–red solids in 52% yield (870 mg) or 72% based on recovered C60.

Spectroscopic data for C60(>CPAF-C2) 8-C2: Anal. Calc. for C94H27N3: C, 94.23; H, 2.25; N, 3.50. Found: C, 94.09; H, 2.55; N, 3.49%; FAB+-MS: calc. for 12C9412714N3 m/z 1197; found, 1198 (MH+), 1197 (M+), 768, 745, 720 (C60), 696, 672 and 613; DCI-MS 1197 (M+), 900, 730 (C60) and 479 (MH2+–C60); UV-vis (CHCl3, 2.0 × 10−5 M) λmax/nm (ε/L mol−1 cm−1): 260 (1.7 × 105), 327 (8.2 × 104), 503 (2.9 × 104); λmax (DMF, ε) 261 (1.2 × 105), 326 (4.4 × 104) and 471 (5.9 × 103); FT-IR (KBr) νmax 3033 (w), 2963 (s), 2923 (m), 2875 (w), 2853 (w), 2224 (C[equivalent]N), 1592 (s), 1492 (s), 1278 (s), 750 (s), 697 (s) and 524 (s) cm−1; 1H NMR (400 MHz, CDCl3, ppm) δ 8.11 (dd, J = 8 Hz, J = 1.2 Hz, 1H), 7.99 (d, J = 1.6 Hz, 1H), 7.78 (d, J = 8 Hz, 1H), 7.58 (d, J = 8 Hz, 1H), 7.27–2.08 (m, 10H), 7.03–7.00 (m, 2H), 5.52 (s, 1H), 2.05–1.88 (m, 4H) and 0.29 (t, J = 7.2 Hz, 6H); 13C NMR (400MHz, CDCl3, ppm) δ 168.8, 152.7, 150.9, 149.4, 147.5, 147.4, 145.8, 145.4, 145.3, 145.2, 145.1, 144.8 (×2), 144.7, 144.5, 144.3, 143.8 (×2), 143.0, 142.4, 142.0 (×2), 141.4, 141.0, 137.5, 137.1, 133.8, 132.2, 129.4, 128.2, 124.6, 123.4, 123.0, 122.7, 121.9, 119.8, 117.7, 113.3 (×2, C[equivalent]N), 87.7, 72.5 (×2, CF1, CF2), 56.6, 41.3 (C61), 32.6 and 8.6.

Spectroscopic data for C60(>CPAF-C9) 8-C9: Anal. Calc. for C108H55N3: C, 93.04; H, 3.95; N, 3.02. Found: C, 93.13; H, 3.92; N, 2.82; FAB+-MS: calc. for 12C10815514N3 m/z 1393; found, 1395 (MH2+), 1394 (MH+), 1155, 1141, 1128, 768, 745, 720 (C60) and 696; UV-vis (CHCl3, 2.0 × 10−5 M) λmax/nm (ε/L mol−1 cm−1): 261 (1.3 × 105), 326 (5.3 × 104) and 499 nm (2.0 × 104); FT-IR (KBr) νmax 3029 (w), 2946 (s), 2860 (m), 2222 (C[equivalent]N), 1591 (vs), 1490 (s), 1464 (s), 1425 (s), 1277 (s), 1184 (s), 751 (s), 694 (s) and 525 (s) cm−1; 1H NMR (400 MHz, CDCl3, ppm) δ 8.13 (d, J = 8 Hz, 1H), 8.03 (t, 1H), 7.75 (d, J = 8 Hz, 1H), 7.59 (d, J = 8 Hz, 1H), 7.28–7.01 (m, 12H), 5.54 (s, J = 4 Hz, 1H), 2.1–1.8 (m, 4H) and 1.28–0.4 (m, 34H). 13C NMR (500 MHz, CS2–CDCl3, ppm) δ 168.0, 154.0, 152.0, 149.8, 148.0, 147.8, 147.4, 146.5, 146.4, 145.9, 145.8, 145.7, 145.3, 145.1, 145.0, 144.8, 144.3, 144.2 (×2), 143.5, 143.0, 142.6, 142.5, 141.9, 141.6 (×2), 137.9, 137.8, 137.5, 133.9, 132.6, 129.9, 128.7, 125.3 (×2), 125.2, 124.0, 123.2, 122.4, 122.3, 120.3, 118.0, 117.9, 113.6 (C[equivalent]N), 113.5 (C[equivalent]N), 88.5, 73.0 (×2, CF1, CF2), 55.8, 51.5, 51.2 (×2), 42.0 (C61), 33.7, 31.2, 30.6, 30.4, 30.5, 30.0 and 23.1.

Crystal data for 8-C2

Empirical formula, C98.25H36N3O1.5S0.50; formula weight, 1298.33; crystal size 0.25 × 0.20 × 0.15 mm; crystal system, monoclinic; space group, P21/c; T = 295(2) K; unit cell dimensions a = 22.0080(2), b = 9.9830(10), c = 28.1780(3) Å, β = 93.4330(10)°; V = 6179.76(11) Å3 with Z = 4; Dc = 1.385 Mg m−3; independent reflections = 10878 (Rint = 0.0636); reflections collected = 34057; R1 = 0.0831 and wR2 = 0.1932 for I > 2σ(I), R1 = 0.1488 and wR2 = 0.2336 for all data.

Physical measurements

Optical and electrochemical instrumentation

Optical absorption and transmission spectra in correlation to the transient measurements in the range of 180–1200 nm were recorded using a 1-mm thick quartz sample cuvette on a Model UV 3600 UV-vis-NIR spectrophotometer (Shimadzu, Kyoto, Japan).

Electrochemical redox potential values were measured using a BAS CV-50W Voltammetric Analyzer. A conventional three-electrode system was used with a platinum disk electrode (1 mm in diameter) serving as the working electrode with a platinum wire and an Ag/AgCl electrode serving as the counter and reference electrodes, respectively. All measurements were carried out in deaerated benzonitrile and o-dichlorobenzene containing tetrabutylammonium perchlorate [(n-Bu)4N+ClO4] (0.1 M) as a supporting electrolyte in a scan rate of 0.1 V s−1. The potentials were referenced to a ferrocene/ferrocenium (Fc/Fc+) redox couple as an internal standard.

Emission and nanosecond transient absorption measurements

Emission spectra were measured using a Perkin-Elmer model LS 50B fluorometer. Time-correlated single-photon counting (Edinburgh Instruments OB 920 Spectrometer) was utilized to determine singlet state lifetimes (data given in Table 1). The sample was pumped with a 70-ps laser diode at 401 nm. Emission was detected on a cooled microchannel plate PMT. Data were analyzed by using a reconvolution software package provided by Edinburgh Instruments. Nanosecond transient absorption measurements were carried out using 486 nm of a 355 nm optical parametric oscillator (OPO) of a Q-switched Nd:YAG laser (Quantel Brilliant, pulse width ca. 5 ns). Pulse fluences of up to 4 mJ cm−2 at the excitation wavelength were typically used.

Table 1
Photophysical properties of C60(>CPAF-Cn) and related model compounds in toluene, including the lifetimes τs of transient 1C60*(>CPAF-Cn) state and τt of transient 3C60*(>CPAF-Cn) state

Time-resolved singlet oxygen production measurements

Time-resolved singlet oxygen measurements were performed by direct detection of the near-IR luminescence emission of oxygen (1Δg3Σg transition) at 1270 nm corresponding to a singlet–triplet transition state. Both the fullerene derivatives were excited by a frequency-doubled Nd:YLF laser (QG-523-500; Crystalaser Inc, Reno, NV) at 523 nm (λmax). The pulse duration was 10 ns and the pulse repetition rate was approximately 3.0 kHz. The singlet oxygen luminescence was detected by a PMT detector (R5509-42; Hamamatsu Corp, Bridgewater, NJ) with high sensitivity in the near-IR region. Five bandpass filters (1210, 1240, 1270, 1300 and 1330 nm) were placed sequentially in front of the photodetector to sample the luminescence spectrum. All measurements were performed using 10 or 50 μM concentrations of the fullerene derivatives in toluene. The time-resolved kinetics of the singlet oxygen luminescence in terms of the-concentration of 1O2 at time t, S(t) were expressed and fitted with the following equation.13


where C is a constant, τt and τs are triplet state and singlet oxygen lifetime (ms), respectively.

Cell culture

A human cervical cancer cell line, HeLa was obtained from ATCC (Manassas, VA). The cells were cultured in RPMI medium with L-glutamine and NaHCO3 supplemented with 10% heat-inactivated fetal bovine serum, penicillin (100 U mL−1) (Sigma, St. Louis, MO) at 37 °C in 5% CO2-humidified atmosphere in flasks (75 cm2, Falcon, Invitrogen, Carlsbad, CA). When the cell reached 80% confluence, they were washed with phosphate–buffered saline (PBS) and harvested with the trypsin–EDTA solution (0.25%, 2.0 mL, Sigma). Cells were then centrifuged and counted in trypan blue to ensure viability and plated at a density of 5000/well in flat-bottom 96-well plates (Fisher Scientific, Pittsburgh, PA).

In vitro PDT studies

We used a broad-band white-light source (400–700 nm band pass filter, LumaCare, Newport Beach, CA) to deliver 200 J cm−2 at an irradiance of 100 mW cm−2 as measured with a power meter (Model DMM 199 with 201 Standard head, Coherent, Santa Clara, CA). Illumination took 33 min to complete.

Fullerenyl monoadduct derivatives were dissolved in dimethylacetamide (DMA) to form stock solutions of 5.0 mM. After cells had grown for 24 h, dilutions of fullerene derivatives were prepared in complete RPMI medium containing 10% serum, penicillin (100 U/mL), and streptomycin (100 μg mL−1). Fullerene derivatives were added to the cells at 1.0–5.0 μM concentrations for an additional 3.0 h incubation. The highest DMA concentration in the medium did not exceed 0.1%. The medium was replaced with fresh complete medium and illumination (200 J cm−2) with white light was delivered. The light spot covered four wells, which were considered as one experimental group illuminated at the same time. Control groups were as follows: no treatment, light alone and fullerene derivative alone in the dark (at the same dilutions used for PDT). Following PDT treatment, the cells were returned to the incubator overnight and a 4-h MTT assay was carried out the next day and read at 562 nm using a microplate spectrophotometer (Spectra Max 340 PC, Molecular Devices, Sunnyvale, CA). Each experiment was repeated 3 times.

Results and discussion

Each chemical functionalization by alkyl or aryl addition to a fullerenyl olefin effectively changes the orbital configuration of two linking fullerenyl carbons from sp2 to sp3 and makes the direct π-conjugation between the C60 cage and attached organic chromophores impossible. Our recent approach to circumventing this issue and enhancing the electronic interactions of light-harvesting photoactive donor antenna with π-orbitals of the C60 cage was to apply a bridging π-functional group for connecting these two moieties.14 This provides the possibility of an increasing amount of partial to full periconjugation. The first bridging functional group used for this purpose was a keto group in the molecular construction of C60–keto–DPAF-Cn assemblies, where DPAF-Cn represents 9,9-dialkyl-2-diphenylaminofluorene. As an example, the solutions of C60(>DPAF-Cn) homologs were found to show large enhancements of two-photon absorption (2PA) cross sections in the femtosecond region as compared to the sum of the values of individual chromophore components.10,15 These nonlinear photonic responsive data together with the nanosecond transient absorption measurements based on the linear photoexcitation of the same series of fullerenyl adducts provided strong evidence of the effectiveness of through-space periconjugation interactions. These interactions were proposed to facilitate the high efficiency of intramolecular electron and energy transfer processes from the DPAF-Cn donor antenna to the C60 acceptor cage upon photoexcitation.16,17 In this study, we further extended our investigation of the periconjugation approach by functional conversion of C60(>DPAF-Cn) dyad materials to produce molecules with broadband absorption characteristics.

Accordingly, we carried out chemical modification of the keto group to the 1,1-dicyanoethylenyl (DCE) group leading to the resulting C60–DCE–DPAF-Cn assemblies, C60(>CPAF-Cn). The higher electron-withdrawing ability of the DCE group compared with the keto group should increase the degree of molecular polarization with negative charges being localized in the region of the fullerene cage. The modification was also expected to decrease the through-space contact distance between the C60 cage and the bridging DCE functional group. Owing to a limiting space available in the region, this should result in closer and more effective π-electron interactions between these two moieties.

Synthesis and spectroscopic characterization of C60(>CPAF-Cn)

Chemical modification of a keto group to a DCE group can be made by using malononitrile as a reagent. However, a close contact of DCE group with the C60 cage could also be accompanied by the possibility of high steric hindrance at the intermediate reaction center when bonding DCE in a limited space adjacent to the C60. Synthetically, a recently reported reaction sequence for the preparation of 7-bromoacetyl-9,9-dialkyl-2-diphenylaminofluorene 5-Cn, BrDPAF-Cn, was applied as the key intermediate precursor molecule,12 as shown in Scheme 1. Addition of a DPAF-Cn subunit to C60 leading to the formation of dyads 7-C2, C60(>DPAF-C2) and 7-C9, C60(>DPAF-C9), was accomplished by the treatment of α-bromoacetylfluorene derivative 5-C2 and 5-C9, respectively, with C60 in toluene in the presence of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU, 1.0 equiv.) at ambient temperature for 5.0 h. By the control of an equivalent quantity of DPAF-Cn vs. C60, the yield of the monoadduct was maximized with the bisadduct in a low yield and other higher adducts in an insignificant amount.

Scheme 1
Synthetic route for the preparation of C60(>CPAF-C2) and C60(>CPAF-C9). Reagents and conditions: i, α-bromoacetyl bromide, AlCl3, ClCH2CH2Cl, 0 °C to rt, 4.0 h; ii. C60, DBU, toluene, rt, 5.0 h; iii, malononitrile, pyridine, ...

Conversion of the keto group of 7-C2 and 7-C9 to the corresponding 1,1-dicyanoethylenyl group was effected by the reaction using malononitrile as a reagent, pyridine as a base, and titanium tetrachloride as a deoxygenation agent in dry toluene at ambient temperature for a short period of 5 min. The reaction afforded 7-(1,2-dihydro-1,2-methanofullerene[60]-61{1,1-dicyanoethylene})-9,9-diethyl-2-diphenylaminofluorene 8-C2, C60(>CPAF-C2), as orange–red solids in 67% yield, based on recovered C60, after purification. Alternatively, C60(>CPAF-C2) can be synthesized from the reaction of C60 with 7-[1-(1,1-dicyanoethylene)-2-bromoethyl]-9,9-diethyl-2-diphenylamino-fluorene 6-C2, BrCPAF-C2, in dry toluene in the presence of DBU (1.0 equiv.) at ambient temperature for a period of 5.0 h, as shown in Scheme 1. The reaction resulted in the identical products in a slightly higher yield of 52 or 72% based on recovered C60.

Evidence of the CPAF moiety attached to a methano[60]fullerene cage by a 1,1-dicyanoethylenyl bridging group can be seen clearly from the infrared spectra of 8-C2 and 8-C9 where both showed a strong absorption band of cyano (–C[equivalent]N) stretching vibration centered at 2222–2224 cm−1 with the complete disappearance of carbonyl stretching vibration at 1677 cm−1 for both 7-C2 and 7-C9. There were also three typical fullerenyl signals at 751, 694 and 525 cm−1 for the unfunctionalized half-cage sphere of C60. The former cyano absorption band matches well with that of the model compound C60(>DCE-Ph) 2 at 2224 cm−1 and the BrCPAF-C2 precursor, at 2225 cm−1, confirming successful conversion of the keto to corresponding DCE moiety. Interestingly, this chemical modification led to a large bathochromic shift of the long-wavelength absorption band of 8-C2 to 503 nm (ε = 2.9 × 104 L mol−1 cm−1) in CHCl3 (2.0 × 10−5 M) which was red-shifted nearly 93 nm compared to that of C60(>DPAF-C2) (7-C2) or C60(>DPAF-C9) (7-C9) centered at 410 nm (λmax, Fig. 1d), as shown in UV-vis spectra of Fig. 1. This band was accompanied with two other absorption bands with λmax centered at 260 (ε = 1.7 × 105) and 327 (ε = 8.2 × 104 L mol−1 cm−1), matching approximately with those of the model compound C60(>CO-Ph) (Fig. 1a) containing mainly an optically active fullerene moiety. The absorption profiles also agree well with those of allowed 1T1u1Ag transition bands of pristine C60.18 The comparison provided the confirmation of these two absorption bands being attributed to the C60 cage. Similarly, the compound 8-C9 displayed three absorption bands with λmax centered at 261 (ε = 1.3 × 105), 326 (ε = 5.3 × 104) and 499 nm (ε = 2.0 × 104 L mol−1 cm−1). Further comparison of the latter band with the main optical absorption of BrCPAF-C2 (Fig. 1c) centered at λmax 492 nm (ε = 1.8 × 104 L mol−1 cm−1) also showed a good match with a slight red-shift of absorption characteristics of DCE–diphenylaminofluorene moiety, along with an additional aryl absorption band at 316 nm (ε = 2.8 × 104 L mol−1 cm−1). More importantly, the structural change of C60(>CO-Ph) (Fig. 1a) to C60(>DCE-Ph) (Fig. 1b) does not greatly alter its absorption spectrum, indicating the close π-conjugation linkage of DCE being associated with the fluorene moiety, not with the C60 cage. Thus, the 503 nm band should be attributed to the transition from the HOMO level to the LUMO level of the CPAF-Cn moiety. A red-shift of 8–12 nm from that of BrCPAF-C2 might also imply the contribution of fullerene cage, perhaps, via through-space electronic interactions. When the spectrum C60(>CPAF-C2) (8-C2) (Fig. 1e) was superimposed onto the combined spectrum of Fig. 1c (6-C2) and either Fig. 1a (1) or Fig. 1b (2) (two independent chromophores DCE–diphenylaminofluorene and fullerene cage, respectively), the close resemblance suggested either no detectable or negligible ground-state interaction between these two moieties present in the molecule of 8-C2 and 8-C9. In the long-wavelength absorption region beyond 650 nm, a very weak characteristic steady-state absorption band of the methano[60]fullerene (C60>) moiety became visible at 690 nm only at a high concentration of 4.4 × 10−4 M.

Fig. 1
UV-vis spectra of (a) C60(>CO-Ph) 1, (b) C60(>DCE-Ph) 2, (c) BrCPAF-C2 6-C2, (d) C60(>DPAF-C2) 7-C2 and (e) C60(>CPAF-C2) 8-C2 in CHCl3 at 2.0 × 10−5 M and (a′) the absorption peak of the C60 moiety ...

A solvent-dependent effect on the absorption spectrum of C60(>CPAF-C2) was observed with a trend towards increasing peak intensity going from DMF, benzonitrile (PhCN), toluene (TN) and chloroform, to o-dichlorobenzene (o-DCB), as shown in the inset of Fig. 1, roughly correlated to increasing solvent power and decreasing solvent polarity. In the case of DMF as a polar solvent, a blue-shift of λmax to 471 nm with reduced peak intensity (ε = 5.9 × 103 L mol−1 cm−1) was detected. Characteristics of steady-state fluorescence emission of 8-C2 and 8-C9 were also found to be solvent-dependent. This variation was due to photophysical events involving different degrees of either intra-molecular energy- or electron-transfer from the photoexcited CPAF-Cn moiety to the C60 cage in a competitive process, and the fact that the energy-transfer process is favorable in the less polar solvent. Both processes significantly reduce the intensity of 1CPAF*-Cn emission. Therefore, unlike the large fluorescence emission of the model compound CPAF-C9 showing maximum intensity centered roughly at 550 nm (Fig. 2a) in toluene or 620 nm (Fig. 2b) in THF upon excitation at 486 nm, the fluorenyl fluorescence emission of C60(>CPAF-C2) and C60(>CPAF-C9) in toluene was sharply reduced in the region of 500–650 nm, as shown in Fig. 2c and d, respectively. Similarly, a large decrease in fluorescence emission was also detected in polar PhCN. This clearly indicated an efficient fluorenyl fluorescence quenching effect by the C60> cage in the molecular system of C60(>CPAF-Cn). In toluene, the quenching effect was accompanied with weak fluorescence emissions at 700–780 nm, as shown in Fig. 2c and d. These were comparable to the fluorescence of the model compound C60(>CO-Ph) 1 showing nearly identical emission peak profiles and intensity (Fig. 2e) in the same region. Since the compound 1 contains a C60> cage as the sole photoactive component, thus, the fluorescence peak can be correlated to the lowest excited singlet energy of 1C60*(>CPAF-C2) which was estimated to be 1.75 eV in TN and CHCl3. A recent report proposed that the quenching could be due to an electron-transfer process from either C60[>(DCE) δ-(DPAF)+δ-C2] or C60[>1 (CPAF)*-C2] states to the corresponding C60 (>CPAF +-C2) state in addition to the alternative energy-transfer process from C60[>1(CPAF)*-C2] to the C60> moiety yielding 1C60*(>CPAF-C2).11 Formation of the former intra-moiety partial charge-transfer state is proposed to be feasible because of the red-shift of absorption peak of CPAF-C9 1 to 620 nm in THF (Fig. 2b) with a more intense peak profile than that in toluene (TN) (Fig. 2a). Significant optical absorption characteristics change of CPAF-C9 in these two solvents was explained by the increasing tendency to form the partial charge-transfer state, (DCE) δ-(DPAF)+δ-C2, in THF (more polar than TN) that enhances the absorption at longer wavelengths. Furthermore, in nonpolar solvents, the intensity of the long-wavelength emission band of 1(C60>)* was found to increase as the intramolecular energy-transfer processes became more favorable. This resulted in absorption at 700–740 nm that was more intense in TN and CS2, compared with low or negligible intensity in o-DCB and PhCN.

Fig. 2
Steady-state fluorescence spectra of the model compound CPAF-C9 in (a) TN and (b) THF; fluorescence spectra of (c) C60(>CPAF-C2) 8-C2, (d) C60(>CPAF-C9) 8-C9 and (e) the reference C60(>CO-Ph) 1 in TN with the excitation wavelength ...

In the composition and structural characterization, the molecular ion mass of C60(>CPAF-C2) 8-C2 was clearly detected in both negative ion desorption chemical ionization mass spectrum (DCI-MS), by a group of mass peaks with a maximum peak intensity centered at m/z 1197 (M+), and in positive ion fast atom bombardment mass spectrum (FAB+-MS). The latter spectrum displayed sharp mass ion peaks at m/z 1198 (MH+) and 1197 (M+), shown in Fig. 3a. In the case of C60(>CPAF-C9) 8-C9, the positive ion FAB+ mass spectrum (Fig. 3b) showed a group of mass ion peaks at m/z 1395 (MH2+) and 1394 (MH+). It was followed by a group of mass peaks centered at m/z 1155, 1141, 1128 and 720 (C60) matching well with the fragmented mass ion of 8-C9 showing the loss of two 3,5,5-trimethylhexyl (C9) subunits and the CPAF-C9 moiety. No other mass fragmentation peaks were found in the upper mass region in the range of m/z 800–1190 or 800–1390 giving a relatively simple spectrum of Fig. 3a or b, respectively. This revealed high stability of the aromatic diphenylaminofluorene moiety under mass spectroscopic measurement conditions. The fragmentation pattern fits well with the bond cleavage occurring mostly at the cyclopropanyl carbon conjunction bonds bridging the fullerene cage and CPAF-Cn moieties. These data clearly provided evidence of the mass composition of fullerene–fluorene dyads C60(>CPAF-Cn) as 8-C2 and 8-C9.

Fig. 3
Positive ion fast atom bombardment (FAB+) mass spectra of (a) C60(>CPAF-C2) 8-C2 and (b) C60(>CPAF-C9) 8-C9.

In addition to IR spectrum analysis, chemical conversion of a keto group to the corresponding 1,1-dicyanoethylenyl group of C60(>CPAF-C2) was also confirmed by the disappearance of key carbonyl carbon peak at δ 190 in 13C NMR spectrum of C60(>DPAF-C2) and a clear upfield shift of the α-proton (Hα next to 1,1-dicyanoethylene group) chemical shift to δ 5.52, appearing as a singlet peak. The shift is 0.17 ppm upfield from δ 5.69 for Hα (next to carbonyl group) of C60(>DPAF-C2), as shown in Fig. 4c and d. It was also characterized by a large downfield shift of 0.93 ppm from δ 4.59 for the chemical shift of Hα peak as a singlet in the spectrum of CPAF-C2 (Fig. 4b). Detailed information of all aromatic protons of C60(>CPAF-C2) and the peak assignments were given by the analysis of 11 2D COSY NMR spectrum (Fig. 5a). Assignment of H6 proton peak was made by its most downfield chemical shift at δ 8.11 (dd, J = 8 Hz, J = 1.2 Hz) in the spectrum of Fig. 4d due to through-space close contact with the fullerene cage. Its proton coupling with the adjacent H5 at δ 7.78 (d, J = 8 Hz) was then revealed in the 2D spectrum along with the long-range proton coupling with H8 at δ 7.99 (d, J = 1.6 Hz). Chemical shifts of these phenyl protons at C6, C8 and C5 of the fluorene ring were shifted downfield from δ 7.69 (d, J = 8 Hz), 7.63 (d, J = 1.6 Hz) and 7.62 (d, J = 8 Hz), respectively, in the spectrum of CPAF-C2 (Fig. 4b) due to the influence of the fullerenyl cage current. This suggested the successful cyclopropanation reaction of CPAF-C2 (6-C2) with C60. The degree of downfield chemical shift is apparently less than the corresponding fluorenyl protons of C60(>DPAF-C2) (7-C2, Fig. 4c) at δ 8.48 (dd, J = 8 Hz, J = 1.6 Hz), 8.32 (d, J = 1.6 Hz) and 7.83 (d, J = 8 Hz), respectively. Most importantly, large downfield chemical shifts of three fluorenyl protons located in close vicinity of C60> clearly revealed strong electronic interactions between CPAF-Cn antenna and C60> moieties.

Fig. 4
1 NMR spectra (CDCl3) of (a) DPAF-C2, (b) CPAF-C2, (c) C60(>DPAF-C2) and (d) C60(>CPAF-C2) showing downfield shift of fluorenyl proton peaks adjacent to the C60 cage.
Fig. 5
(a) 11 2D COSY NMR and (b) 113C 2D HMQC NMR spectra (CDCl3) of C60(>CPAF-C2) 8-C2.

According to the assignment of fluorenyl protons of the dyad C60(>CPAF-C2), chemical shifts of the corresponding carbons, C6, C8 and C5, were assigned at δ 128, 123 and 120, respectively, based on the analysis of 113C 2D HMQC NMR spectrum, as shown in Fig. 5b. In the same spectrum, chemical shifts of the cyclopropanyl or methanofullerene carbon C61, alkyl carbon C10 and methyl carbon of 8-C2 were determined to be δ 41, 33 and 8.6, respectively, with the remaining alkyl carbon peak at δ 57 being assigned to the ring junction carbon C9. The 2D correlation of fluorenyl carbons to their corresponding aryl protons allowed us to differentiate aromatic carbon peaks in Fig. 6 into three regions as marked with chemical shifts of three aminoaryl carbons in CPAF-C2 moiety being assigned as δ 153, 151 and 149. Chemical shifts of the three types of functional carbons, –C=C(CN)2, –C[equivalent]N and =C(CN)2, in the 1,1-dicyanoethylenyl moiety of C60(>CPAF-C2) were found to be δ 169, 113 and 88, respectively, in Fig. 6c. The assignment was made by comparison with the spectrum of one reference compound iso-propylidenemalononitrile. They also agree well with those of the model compound C60(>DCE-Ph) 2 showing the corresponding carbon peaks (Fig. 6a) at δ 169, 113 and 90, respectively. A simple phenyl group of 2 makes it easy to identify its four aryl carbons in the region of δ 128–135, with the remainder of fullerenyl sp2 carbon peaks located within δ 136–148, that were used as the reference for the differentiation of fullerene cage carbon peaks derived from C60(>DPAF-C2) and C60(>CPAF-C2). Accordingly, the chemical shift of two fullerenyl sp3 carbons, CF1 and CF2, are close to each other as a peak at δ 73. A total of 28 peak signals in the region of 135 to 148 ppm, as marked in Fig. 6, are accounted for by the remaining 58 sp2 fullerenyl carbons that fits well with a C2 molecular symmetry.

Fig. 6
13C NMR spectra of (a) C60(>DPAF-C9) 7-C9 and (b) C60(>CPAF-C9) 8-C9.

To unambiguously determine the structure of C60(>CPAF-C2), single crystals of 8-C2 were grown by the slow diffusion method in a H-tube using a co-solvent mixture of CS2–ethanol–acetone. X-Ray crystallographic analysis of one single crystal effectively resolved and confirmed the molecular structure of 8-C2. Collection of crystallographic data and the structural refinement gave the results provided in the Experimental section (see also ESI). Accordingly, a perspective 3-D ORTEP view and its molecular unit cell packing are displayed in Fig. 7a and b, respectively. The empirical formula of the single crystal was reconstituted to a composition of (C94H27N3)1.0(C3H6O)1.0-(C2H6O)0.5(CS2)0.25 as a mixture of (8-C2)1.0(acetone)1.0(ethanol)0.5(CS2)0.25. Interestingly, a highly ordered array of the fullerene cages along the b axis of the unit cell was observed with the planar DCE-diphenylaminofluorene (CPAF-C2) rings located on the top to each other in the ac plane of the unit cell to maximize π–π interaction between CPAF-C2 rings (Fig. 7b). This showed the intermolecular forces involved in the molecular assembly of C60(>CPAF-C2) that were dominated by the strong fullerenyl π-interactions. Two single-bond lengths connecting the methanocarbon and two fullerenyl carbons in the cyclopropanyl ring, CF1–C61 and CF2–C61, were found to be 1.502 and 1.495 Å. The distance between two fullerenyl carbons in the cyclopropanyl ring CF1–CF2 is 1.614 A, longer than the normal single C–C bond length of ~1.54 Å, that leads to the three angles of the cyclopropanyl ring as 65.2(2), 57.2(2) and 57.6 (2)° with the former largest angle located at the methanocarbon. Interestingly, the closest distance between one –C[equivalent]N group and the fullerene cage having the DCE moiety in a co-planar structure of fluorene ring was estimated to be ~2.6 Å (Fig. 7c) shorter than the estimated distance (~3.5 Å) between the –C=O group and the fullerene cage by the single-crystal structural analysis of C60(>DPAF-C2) 7-C2.12

Fig. 7
(a) Perspective ORTEP view and (b) the unit molecular packing of C60(>CPAF-C2) in the crystal with solvent molecules and hydrogen atoms omitted for clarity and (c) a 3-D model of 7-C2 or 8-C2 to show the closest estimated distance between C=O ...

Electrochemical and time-resolved fluorescence decay measurements of C60(>CPAF-Cn)

Electrochemical redox potential values of C60(>CPAF-Cn) were measured using cyclic voltammetry (CV) carried out in either deaerated PhCN or o-DCB containing tetra(n-butyl)ammonium perchlorate salt, (n-Bu)4N+ClO4, as a supporting electrolyte. The recorded electrochemical potential values were referenced to the ferrocene–ferrocenium (Fc/Fc+) redox couple as an internal standard. Previously reported CV measurements on the sample of C60(>DPAF-C2) in deaerated PhCN (Fig. 8a)16 were used as the reference in this study for comparison. A different solvent was used to maximize the reversibility of electrochemical redoxwaves of either the C60> or CPAF-Cn moiety measured. In the case of C60(>CPAF-C2) in deaerated PhCN, its CV profile (Fig. 8b) showed two reversible redox waves with the first E1ox and second E2ox values of the CPAF-C2 moiety observed at +0.58 and +0.70 V vs. Fc/Fc+, respectively, in close resemblance to those reported.11 It was accompanied with the corresponding re-reductive waves centered at +0.63 V for E2red and ~+0.45 V for E2red. The value of E1ox is in a close range to the E1ox value (+0.62 V) of the model compound CPAF-C2 in PhCN. In a separate measurement on the electrochemical reduction processes of 8-C2 in o-DCB (Fig. 8c), three reversible redox waves were detected with the first E1red, second E2red and third E3red values of the C60 cage moiety at −0.90, −1.17 and −1.40 V vs. Fc/Fc+, respectively, with the corresponding re-oxidative waves centered at −1.30 V for E3ox, −0.93 V for E2ox and −0.61 V for E1ox. The E1red potential value is slightly negative to that of pristine C60 (E1red = −0.93 vs. Fc/Fc+).19 In comparing the electrochemical characteristics between C60(>DPAF-C2) and C60(>CPAF-C2), we observed that changing the keto moiety to a DCE group increases the electron uptake capability of the fullerene cage in 8-C2, based on the detected first and second reduction potentials changing from −1.03 and −1.45 V (Fig. 8a), respectively, to −0.90 and −1.17 V (Fig. 8c), respectively. It also decreases the electron donating ability of diphenylaminofluorene moiety of 8-C2 with an increase of the E1ox value from +0.48 V for 7-C2 to +0.58 V. Since the redox potentials in covalent-bound donor–acceptor systems are essential for the evaluation of these optoelectronic materials in the photocell applications, the overall increase of molecular electronegativity of 8-C2 may alter the photoinduced transient electron-transfer characteristics and the decay time of the charge-separated transient states from those of C60(>DPAF-C2) in a solvent-dependent manner.

Fig. 8
Cyclic voltammetry of (a) C60(>DPAF-C2) in PhCN and C60(>CPAF-C2) (1.0 × 10−4 M) in (b) PhCN showing redox waves of the CPAF-C2 moiety and (c) o-DCB showing redox waves of the C60 cage moiety.

Photophysical properties of the C60–DCE–DPAF assemblies involve the primary photoexcitation events of either the fullerene moiety at UV wavelengths or the DPAF-Cn moiety at both UV and visible wavelengths up to 600 nm (Fig. 1). Much higher optical absorption capability of DPAF-Cn than the C60> cage in visible wavelengths enables the former moiety to serve as a light-harvesting antenna. Accordingly, formation of the photoexcited 1(DPAF)*-C2 moiety should be considered as the early event in the photophysical process. Alteration of the keto group of 7-C2 to the 1,1-dicyanoethylenyl group of C60(>CPAF-C2) effectively extended its photoresponsive region to longer red wavelengths. Photoexcitation processes of 7-Cn and 8-Cn pump an electron from their highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO). In the case of the similar structurally analogous C60(>CPAF-C2M) 8-C2M in the molecular orbital calculation and energy minimization, the majority of the HOMO electron density was reported to be delocalized over the diphenylaminofluorene (DPAF) moiety, whereas the LUMO electron density was located on the C60 spheroid, and therefore C60 (>CPAF+-C2M) was suggested as the most stable charge-separated (CS) state.11 The same argument can be applied for the formation of C60 (>CPAF +-C2) and C60 (>CPAF +-C9) states owing to close similarity of the core structure. These charge-separated states may be generated by photoinduced intramolecular electron-transfer between the diphenylaminofluorene donor and C60> acceptor moieties. The process effectively quenches fluorenyl fluorescence that can be observed in the most of 7-Cn and 8-Cn monoadducts.

By using the same arguments reported for 8-C2M,11 other plausible fluorescence quenching pathways of photoexcited C60[>1(CPAF)*-C2] may involve ultrafast intramolecular energy-transfer processes going from 1(CPAF)*-C2 to the C60> moiety yielding the excited 1C60*(>CPAF-C2) state that resulted in fluorenyl fluorescence suppression. The energy-transfer route was observed predominantly in nonpolar solvents as reported previously.16 It was indicated by and associated with a weak 1(C60>)*-derived fullerenyl fluorescence. The intensity of this emission band at 710 nm was found to be highly sensitive to the solvent polarity. Variation of the solvent polarity from that of TN and CS2 increases the chance of a competitive electron-transfer process occurring on the same compound. Identification of photoinduced transient states can be made by performing time-resolved picosecond fluorescence emission measurements with photoexcitation mainly at the DPAF-Cn or CPAF-Cn moiety. It is immediately followed by the detection of transient fluorescence bands at either 600 or 700 nm at a different short-time scale for the indication of C60[>1(CPAF)*-C2] or 1C60*(>CPAF-C2) transient state formation, respectively. Progressive fluorescence intensity decrease and decay–time profiles of these two bands are then monitored for understanding the quenching mechanism and following the kinetics of photo-induced processes. By using the C60(>CPAF-C2M) sample in TN as an example, the transient lifetime of the 1(CPAF)*-C2M moiety was evaluated based on the decay-time profile of the band at 600 nm and reported to be ~53 ps.11 A short decay time reveals efficient fluorescence quenching by the C60> cage yielding the corresponding excited 1C60*(>CPAF-C2M) state. That allowed the detection of a new emission band of the 1(C60>)* at 715–720 nm appearing immediately after considerable quenching of the C60[>1(CPAF)*-C2M] state in nonpolar TN and CS2.

When the fluorescence emission of 1(C60>)* in TN reached the steady state, the lifetimes (τs) of 1C60*(>CPAF-C2) and 1C60*(>CPAF-C9) were evaluated to be 1521 and 1510 ps, respectively, as shown in Table 1. These values were found to be in a similar range as those of C60(>DPAF-C9) (1650 ps)16 and C60.20 They are slightly longer than the lifetime of 1C60*(>CO-Ph) (1350 ps) measured at 708 nm. It is also interesting to note that, unlike the τs value of the model keto-compound Br-1DPAF*-C9 (2125 ps), the fluorescence lifetime of the model DCE-compound 1CPAF*-C9 is rather short (241 ps). This may be indicative of a facile photoinduced intramolecular charge polarization process forming the corresponding (DCE) –DPAF +-C9 charge-separated state, consistent with the recently proposed photoinduced partial inter-moiety electron-transfer (DCE)−δ–DPAF-C2M state.11

Nanosecond transient spectroscopic measurements of C60(>CPAF-Cn)

Confirmation of the generation of either charge-separated states or fullerenyl triplet transient states of C60(>CPAF-Cn) was made by the measurement of nanosecond transient absorption spectra, followed by monitoring their charge-recombination processes or triplet energy transfer, respectively, in TN. Prior to the study, we determined the optical absorption wavelength (λmax) of the cation radical CPAF +-C2 reference species by the detection of a new near-IR band appearing at 875 nm in the spectrum upon the treatment of the model compound CPAF-C2 with a strong oxidizing agent FeCl3 in solution. In the presence of the charge-transfer mode of C60(>CPAF-C2), this band should accompanied by an additional broad absorption band of (C60>) radical-ion pairs centered at 1020 nm. Therefore, by the detection of these transient absorption bands, the formation of photoinduced charge-separated transient state C60 (>CPAF +-C2) can be confirmed. In other word, the appearance of a new broad transient band centered at ~700–740 nm, corresponding to the T1–Tn absorption of 3(C60>)*, should reveal clearly the formation of 3C60*(>CPAF-Cn) in the photoprocesses that can be correlated to the occurrence of intramolecular energy-transfer events.

In the current transient experiments, laser flash photolysis of the samples was carried out by irradiation at 486 nm (λex). All samples were prepared in the same O.D. of 0.2 at the excitation wavelength to allow direct comparison of the triplet excited state formation with the indication of a relative quantity. A model compound C60(>CO-Ph) 1 was used for the correlation of spectral profiles of all C60-antenna samples. Upon photoexcitation of 1 in TN at 486 nm in the nanosecond region, the detected transient absorption spectral profile (Fig. 9a) showed a broad peak centered at 710 nm, which is attributed to the absorption of the 3(C60>)* moiety since the C60> cage is the only chromophore moiety in 1. Interestingly, the same photoexcitation process carried out on samples of C60(>CPAF-C9) 8-C9 and C60(>CPAF-C2) 8-C2 in TN also displayed a similar transient absorption spectral profile with a major peak centered at 720 nm, as shown in Fig. 9b and c, respectively. The lack of two characteristic bands arising from moieties of CPAF +-Cn and (C60>) radical-ion pairs clearly ruled out the possible formation of charge-separated C60 (>CPAF +-C2)and C60 (>CPAF +-C9) states. Apparently, photoexcitation performed on the fluorenyl CPAF-Cn ring moiety of 8-Cn induced the formation of 3C60*(>CPAF-Cn) transient state during the subsequent photo-processes. It is worthwhile to note that the quantity of triplet excited state formed for C60(>CO-Ph) was more than the other two C60(>CPAF-Cn) systems, indicating non-quantitative occurrence of the energy-transfer process in the latter two materials. This is consistent with the steady-state fluorescence data. Evidently, the reduced peak intensity at 720 nm of Fig. 9b and c from that of Fig. 9a represents the corresponding similar extent of C60 (>CPAF +-Cn) formation if there is no energy loss during the photoevents. Finally, the lifetime of 3C60*(>CPAF-C2) and 3C60*(>CPAF-C9) transient states was found to be 34 and 39 μs, respectively, as shown in Table 1, in a similar range as that of 3C60*(>DPAF-C9) (35 μs). These values are longer than that (24 μs) of 3C60*(>CO-Ph) as measured by the intensity decay of the 720 nm band.

Fig. 9
Transient absorption spectra obtained by flash laser photolysis of (a) the model compound C60(>CO-Ph) 1, (b) C60(>CPAF-C9) 8-C9 and (c) C60(>CPAF-C2) 8-C2 at the excitation wavelength (λex) of 486 nm in toluene.

Efficient singlet oxygen generation of C60(>CPAF-C2) in toluene

Predominant generation of 3C60*(>CPAF-Cn) as detected by laser excitation of C60(>CPAF-Cn) in TN prompted us to investigate the photosensitized production of singlet oxygen (1O2) in the presence of molecular oxygen. Formation of 1O2 becomes possible via intermolecular triplet energy transfer from 3C60*(>CPAF-Cn) to O2. Our recent examples of the efficient singlet oxygen production in biological medium using hydrophilic molecular micelle-like C60 derivatives (FC4S), in the form of nanosphere structures,9b as photosensitizers allowed the demonstration of cytotoxicity after single-photon excitation based PDT (1γ-PDT) against cancer cells in vitro and tumors in vivo.9a Direct detection of 1O2 was accomplished by observing its fluorescence emission at 1270 nm during the quenching process. The production of 1O2 is also accompanied with a considerable increase in the decay rate of 3C60*(>CPAF-C2) at 710–740 nm in the presence of O2. As a result, a significantly strong fluorescence emission of singlet oxygen upon photoexcitation of C60(>CPAF-C2) and C60(>CPAF-C2M) in TN suggested its effective quenching of the 3(C60>)* moiety in a 1O2 production yield close to that of pristine C60.11 Synthesis of C60(>CPAF-C2M) was made by the malononitrilation of 7-C2M in the presence of titanium tetrachloride and pyridine, as shown in Scheme 1.

By the application of time-resolved fluorescence emission measurement of 1O2 at 1270 nm upon photoexcitation using a laser light source at 523 nm (λex) with a power of ~10.4 mW, we were able to compare the relative singlet oxygen production efficiency between C60(>CPAF-Cn) and its precursor C60(>DPAF-Cn). The intensity varied as a function of the concentration, the number of laser pulses, and the laser intensity. As an example, we recorded the emission intensity of 1O2 fluorescence arising from C60(>CPAF-C2M) 8-C2M (Fig. 10a) and C60(>DPAF-C2M) 7-C2M (Fig. 10b) to be 3.0 × 105 and 5.0 × 104 M−1 cm−1, respectively, at a concentration of 1.0 × 10−5 M in TN. The latter value of C60(>DPAF-C2M) was found to be comparable to the value found for tetraphenylporphyrin used as a photosensitizer (Fig. 10c). Remarkably, the 1O2 production yield of C60(>CPAF-C2M) was found to be nearly 6-fold higher than its keto analogous 7-C2M, indicating the large light-harvesting enhancement by the CPAF-C2M moiety and subsequent efficient triplet state generation of the C60> cage moiety. The lifetimes of 1O2 generated from C60(>CPAF-C2M), C60(>DPAF-C2M) and tetraphenylporphyrin in TN were estimated to be 32, 31 and 31 μs, respectively. The observation revealed the population mechanism of the 1C60*(>CPAF-C2M) state arising from the energy-transfer process since the direct photoexcitation of the C60> moiety to 1(C60>)* is unlikely to occur with the 523 nm light source. Furthermore, the fluorescence intensity of 1C60*(>DPAF-C2M) and 1C60*(>CPAF-C2M) was significantly quenched by increasing the solvent polarity from TN to PhCN and DMF that led to a nearly disappearance of 1O2 luminescence in DMF (inset of Fig. 10). The observation revealed the domination of the C60(>DPAF +-C2M) charge-separated state in polar solvents, such as PhCN and DMF.

Fig. 10
Time-resolved fluorescence emission and decay of 1O2 upon photoexcitation of (a) C60(>CPAF-C2M) and (b) C60(>DPAF-C2M) in comparison with that of (c) tetraphenylporphyrin in TN at a concentration of 1.0 × 10−5 M with the ...

Photodynamic therapy killing of human cancer cells

We tested C60(>DPAF-C2M) (7-C2M) and C60(>CPAF-C2M) (8-C2M) as photosensitizers to kill cancer cells after illumination. We used the human cervical cancer cell line, HeLa and incubated the cells for 3.0 h with increasing concentrations of 7-C2M and 8-C2M dissolved in DMA. Broad-band white light irradiation (200 J cm−2) was delivered and the cells were returned to the incubator overnight. A 4-h MTT assay was then carried out for mitochondrial reductase activity as a surrogate measure of the cell viability.

Comparison of survival fractions for HeLa cells treated with 7-C2M and 8-C2M (1.0–5.0 μM) by either keeping in the dark or illuminated with 200 J cm−2 white light is depicted in Fig. 11. A significant, dose-dependent loss of viability in the presence of light was observed using 8-C2M as the photosensitizer. The phenomena were not seen either with 8-C2M in the dark or with 7-C2M in either light or dark conditions. These data clearly show the markedly enhanced photocatalytic activity of C60(>CPAF-C2M) compared to that of C60(>DPAF-C2M), due to its higher absorption intensity in visible wavelengths and an increased singlet oxygen quantum yield of the former compound.

Fig. 11
Concentration and light-dependent killing of human cancer cells (HeLa) by C60(>CPAF-C2M) and C60(>DPAF-C2M). The cells were incubated for 3.0 h with specified concentrations of the fullerene derivative diluted from DMA into complete medium ...


A new class of photoresponsive C60–DCE–diphenylaminofluorene (DPAF) nanostructures, C60(>CPAF-Cn), was synthesized and characterized. Structural modification was made by chemical conversion of the keto group in C60(>DPAF-Cn) to a strong electron-withdrawing DCE subunit that resulted in significant increase of electronic polarization in the resulting CPAF-Cn moiety The modification also produces a large bathochromic shift of the main optical absorption band in the visible spectrum up to 600 nm that extends the photoresponse capability of C60–DCE–DPAF nanostructures to longer red wavelengths than C60(>DPAF-Cn). Accordingly, C60(>CPAF-Cn) should allow the operation of 2γ-PDT at the light wavelength of 1000–1200 nm for the enhanced tissue penetration depth.

Based on the evaluation of fluorescence emission and nanosecond transient absorption data in toluene, we proposed a rapid quenching pathway of the photoexcited C60[>1(CPAF)*-Cn] transient state leading to a high population of 1C60*(>CPAF-Cn) state and subsequently the formation of 3C60*(>CPAF-Cn) transient state via intersystem-crossing processes. The latter process leads to an efficient production of singlet oxygen by C60(>DPAF-C2M), which is comparable to that of tetraphenylporphyrin. Remarkably, the 1O2 quantum yield of C60(>CPAF-C2M) was found to be nearly 6-fold higher than that of its keto analogous C60(>DPAF-C2M), demonstrating the large light-harvesting enhancement of the CPAF-C2M moiety and subsequent efficient triplet state generation of the C60> cage moiety. We also demonstrated the feasibility of efficient photosensitizing generation of cytotoxic 1O2 via intermolecular triplet energy transfer of 3C60*(>CPAF-Cn) state with molecular oxygen, as a crucial mechanism in the field of photodynamic therapy. This led to highly effective killing of HeLa cells by C60(>CPAF-C2M) via 1γ-PDT. Finally, C60(>CPAF-Cn) exhibits large two-photon absorption cross-sections at both 780 and 1000 nm enabling its potential uses as the 2γ-PDT photosensitizer in the biological window of 800–1100 nm. This serves as a significant advantage over tetraphenylporphyrin-related PDT agents for 2γ-PDT applications.

Supplementary Material

LYC J Mat ChemS1

LYC J Mat ChemS2


We thank financial support of National Institute of Health under the contract number 1R01CA137108 and Air Force Office of Scientific Research under the contract number FA9550-09-1-0183. We thank Yi-Hung Liu of National Taiwan University Instrumentation Center for X-ray single-crystal structural analysis.


Electronic supplementary information (ESI) available: Fig. S1: 1H NMR spectrum of C60(>DCE-Ph), Fig. S2: Infrared spectra of (a) C60(>DPAF-C9) and (b) C60(>CPAF-C9), Fig. S3: Infrared spectra of (a) C60(>DPAF-C2) and (b) C60(>CPAF-C2) for comparison, Fig. S4: 1H NMR spectrum of C60(>CPAF-C2), Fig. S5: 13C NMR spectra of (a) C60(>CPAF-C2) and (b) expanded aromatic carbon peaks region, Fig. S6: 13C NMR spectra of (a) C60(>DPAF-C9) 7-C9 and (b) C60(>CPAF-C9) 8-C9. Fig. S7: Negative ion desorption chemical ionization mass spectrum (DCI-MS) of C60(>CPAF-C2). CCDC reference number 762136. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c0jm00037j


1. For recent reviews of PDT and some photosensitizers, see: (a) Dougherty TJ. Photochem Photobiol. 1993;58:895–900. [PubMed] (b) Bonnett R. Chem Soc Rev. 1995;24:19–33. (c) Dougherty TJ, Gomer CJ, Henderson BW, Jori G, Kessel D, Korbelik M, Moan J, Peng Q. J Natl Cancer Inst. 1998;90:889–905. [PubMed] (d) Ali H, van Lier JE. Chem Rev. 1999;99:2379–2450. [PubMed] (e) Josefsen LB, Boyle RW. Met-Based Drugs. 2008;2008:1–24. [PMC free article] [PubMed]
2. (a) Henderson BW, Dougherty TJ. Photochem Photobiol. 1992;55:145–157. [PubMed] (b) Sharman WM, Allen CM, van Lier JE. Methods Enzymol. 2000;319:376–400. [PubMed] (c) Chen B, Pogue BW, Hoopes PJ, Hasan T. Crit Rev Eukaryot Gene Expression. 2006;16:279–305. [PubMed]
3. (a) Bhawalkar JD, Kumar ND, Zhao CF, Prasad PN. J Clin Laser Med Surg. 1997;15:201–204. [PubMed] (b) Liu J, Zhao YW, Zhao JQ, Xia AD, Jiang LJ, Wu S, Ma L, Dong YQ, Gu YH. J Photochem Photobiol, B. 2002;68:156–164. [PubMed] (c) Karotki A, Khurana M, Lepock JR, Wilson BC. Photochem Photobiol. 2006;82:443–452. [PubMed] (d) Starkey JR, Rebane AK, Drobizhev MA, Meng F, Gong A, Elliott A, McInnerney K, Spangler CW. Clin Cancer Res. 2008;14:6564–6573. [PubMed]
4. (a) Reinhardt BA, Brott LL, Clarson SJ, Dillard AG, Bhatt JC, Kannan R, Yuan L, He GS, Prasad PN. Chem Mater. 1998;10:1863–1874. (b) Albota M, Beljonne D, Bredas JL, Ehrlich JE, Fu JY, Heikal AA, Hess SE, Kogej T, Levin MD, Marder SR, McCord-Maughon D, Perry JW, Rockel H, Rumi M, Subramaniam G, Webb WW, Wu XL, Xu C. Science. 1998;281:1653–1656. [PubMed]
5. For a recent review: Pawlicki M, Collins HA, Denning RG, Anderson HL. Angew Chem, Int Ed. 2009;48:3244–3266. [PubMed]
6. For a review: Guldi DM, Prato M. Acc Chem Res. 2000;33:695–703. [PubMed]
7. (a) Tabata Y, Murakami Y, Ikada Y. Jpn J Cancer Res. 1997;88:1108–1116. [PubMed] (b) Nitta N, Seko A, Sonoda A, Ohta S, Tanaka T, Takahashi M, Murata K, Takemura S, Sakamoto T, Tabata Y. CardioVasc Interventional Radiol. 2008;31:359–366. [PubMed]
8. (a) Tegos GP, Demidova TN, Arcila-Lopez D, Lee H, Wharton T, Gali H, Hamblin MR. Chem Biol. 2005;12:1127–1135. [PubMed] (b) Mroz P, Tegos GP, Gali H, Wharton T, Sarna T, Hamblin MR. Photochem Photobiol Sci. 2007;6:1139–1149. [PubMed] (c) Mroz P, Pawlak A, Satti M, Lee H, Wharton T, Gali H, Sarna T, Hamblin MR. Free Radical Biol Med. 2007;43:711–719. [PubMed]
9. (a) Yu C, Canteenwala T, Chen HC, Jeng US, Lin TL, Chiang LY. Free Radical Scavenging and Photodynamic Functions of Micelle-like Hydrophilic Hexa(sulfobutyl)fullerene (FC4S) In: Osawa E, editor. Perspectives of Fullerene Nanotechnology. Kluwer Academic Publisher; London: 2002. pp. 165–183. (b) Yu C, Canteenwala T, El-Khouly ME, Araki Y, Pritzker K, Ito O, Wilson BC, Chiang LY. J Mater Chem. 2005;15:1857–1964. (c) Yu C, Canteenwala T, Chiang LY, Wilson Brian, Pritzker K. Synth Met. 2005;153:37–40. (d) So G, Karotki A, Verma S, Hauck TS, Wilson BC, Pritzker K, Chiang LY. Proc SPIE. 2005;5969:59690D.
10. Elim HI, Anandakathir R, Jakubiak R, Chiang LY, Ji W, Tan LS. J Mater Chem. 2007;17:1826–1838.
11. El-Khouly ME, Padmawar P, Araki Y, Verma S, Chiang LY, Ito O. J Phys Chem A. 2006;110:884–891. [PubMed]
12. Padmawar PA, Canteenwala T, Tan LS, Chiang LY. J Mater Chem. 2006;16:1366–1378.
13. Niedre M, Paterson MS, Wilson BC. Photochem Photobiol. 2002;75:382–391. [PubMed]
14. (a) Chiang LY, Padmawar PA, Canteewala T, Tan LS, He CS, Kanna R, Vaia R, Lin TC, Zheng Q, Prasad PN. Chem Commun. 2002:1854–1855. [PubMed] (b) Padmawar PA, Canteenwala T, Verma S, Tan LS, Chiang LY. J Macromol Sci, Part A: Pure Appl Chem. 2004;41:1387–1400. (c) Padmawar PA, Canteenwala T, Verma S, Tan LS, He GS, Prasad PN, Chiang LY. Synth Met. 2005;154:185–188.
15. Padmawar PA, Rogers JO, He GS, Chiang LY, Canteenwala T, Tan LS, Zheng Q, Lu C, Slagle JE, Danilov E, McLean DG, Fleitz PA, Prasad PN. Chem Mater. 2006;18:4065–4074.
16. Luo H, Fujitsuka M, Ito O, Padmawar P, Chiang LY. J Phys Chem B. 2003;107:9312–9318.
17. El-Khouly ME, Anandakathir R, Ito O, Chiang LY. J Phys Chem A. 2007;111:6938–6944. [PubMed]
18. (a) Saito S, Oshiyama A. Phys Rev Lett. 1991;66:2637–2640. [PubMed] (b) Bertsch GB, Bulgac A, Tomanek D, Wang Y. Phys Rev Lett. 1991;67:2690–2693. [PubMed]
19. Allemand PM, Koch A, Wudl F, Rubin Y, Diederich F, Alvarez MM, Anz SJ, Whetten RL. J Am Chem Soc. 1991;113:1050–1051.
20. (a) Arbogast JW, Foote CS, Kao M. J Am Chem Soc. 1992;114:2277–2279. (b) Steren CA, von Willigen H, Biczok L, Gupta N, Linschitz H. J Phys Chem. 1996;100:8920–8926.