J. Phys. Chem. B 2004, 108, 11473-11479
11473
Influences of Structural Properties on Stability of Fullerenols Gengmei Xing, Jun Zhang, Yuliang Zhao,* Jun Tang, Bo Zhang, Xinfa Gao, Hui Yuan, Li Qu, Wenbing Cao, Zhifang Chai, Kurash Ibrahim, and Rui Su Lab for Nanoscale Materials & Their Bio-EnVironmental Health Sciences, Institute of High Energy Physics, Chinese Academy of Sciences, Beijing100039, China ReceiVed: March 18, 2004; In Final Form: May 25, 2004
Influences of structural properties on the stability of fullerenols are studied using experimental techniques including laser-induced dissociation associated with a time-of-flight measurement, synchrotron radiation XPS, and FT-IR spectroscopy. Stabilities of a family of fullerenols (C60(OH)42, C60(OH)44, C60(OH)30, C60(OH)30, C60(OH)32, and C60(OH)36) as functions of structural parameterssthe hydroxyl number, intensity of the impure group, and the ratio of the carbonyl to hydroxyl groupssare investigated. It is found that the molecular stability largely depends on the quantity of impure groups, especially the highly oxygenated carbons in fullerenols, but less on the hydroxyl number. This is different from the previous consideration that the stability of fullerenols largely depends on the hydroxyl number. Previously, to gain the larger solubility required by practical applications, it was suggested to increase the number of the hydroxyl groups. This idea needs to be restudied, because in highly hydroxylated fullerenol molecules, the coinstantaneous formation of a large amount of impure groups is observed. The use of C60(OH)n>36 in practical applications should proceed with caution, since these could lead to unstable open-cage structures. The results reveal a way of controlling the formation of impure groups to gain fullerenols of high stability.
Introduction Fullerenols, one class of the most attractive fullerenic derivatives, have drawn extensive attention because of their promising applications in many fields such as solar energy conversion and storage,1 fuel cells,2,3 macromolecular materials,4,5 and biomedical and life sciences.6-15 For instance, Rincon et al. first found their functions in solar energy conversion and storage. Hinokuma et al. revealed the proton conductivity of C60(OH)12 in the temperature range from 253 to 343 K.2,3 Because of the strong electrophilic property of the fullerene cage, fullerenols easily release protons and thus provide a new family of proton conductors. The biological properties of fullerenols have been studied by many researchers. It has been found that fullerenols exhibit an efficient free redical (superoxide radicals O2-) scavenging activity in biological systems; they are able to reduce the concentration of free radicals in pathological blood and can inhibit the growth of abnormal or ailing cells.15,16 More biological activities reported include inhibition of the human immunodeficiency virus (HIV) protease because of the C60 spherical cage fitting well into the hydrophobic cavity of the enzyme active site,17,20 cytotoxicity against tumor cells and DNA cleavage under visible light irradiation,21,22 etc. Fullerenols were reported to be an excellent material for medicinal chemistry,23,24 and gadolinium fullerenols25,26 were found to be the best candidate for a new generation of novel magnetic resonance imaging (MRI) contrast agents. Applications require a good stability and solubility of fullerenols, which strongly influence functionalized properties mentioned above. Theoretically, the structural stabilities of fullerenols were studied using quantum-chemical semiempirical computations.27,28 A structure analysis of the calculations * To whom correspondence should be addressed. E-mail: zhaoyuliang@ ihep.ac.cn. Tel/Fax: +86-10-8823-3191.
indicated that the relative stability of fullerenols largely depended on the number of hydroxyl groups,29 and the thermodynamic stability of fullerenols could be measured through a partial dissociation heat.28 However, because the introduction of several tens of hydroxyl groups onto the surface of a fullerene cage leads to the high dimensionality of electronic and nuclear problems, experimental analyses of either structural or thermal stability become complicated and quite difficult. So far, no experimental result on these important issues was reported. In this paper, six C60(OH)n with n ) 42, 44, 30, 30, 32, and 36 are studied as functions of M[OH-/C60] (the molar ratio of NaOH to C60 in the synthesis process), the hydroxyl number, the quantity of impure groups, and the structural parameters of fullerenols. Direct correlations between the stability of the functionalized fullerenes and the structural parameters studied are revealed. Experimental Section Chiang et al.5,7-9,15,16 have established the synthesis methods for the efficient preparation of fullerenols and related derivatives. Six C60(OH)n used in the present work were synthesized using the alkaline reaction.24,25 Reactant concentrations were varied to obtain a group of fullerenol molecules of different hydroxyl numbers. A C60 toluene solution (30 mL, 1.5 mg/mL) was transferred into each of six beakers, then 1 mL of catalyst of 40% TBAH (tetrabutylammonium hydroxide) and 2 mL of aqueous solution containing 0.27, 0.55, 1.11, 2.22, 3.33, and 4.44 g of NaOH (corresponding to M[OH-/C60] ) 0.09, 0.18, 0.38, 0.76, 1.14, and 1.52) were added into beakers marked (a), (b), (c), (d), (e), and (f), respectively. The mixture solutions were vigorously stirred at room temperature for 24 h, the color of the solution in each beaker changed from the originally deep violet to colorless, while a brown sludge precipitated on the bottom of the beaker. After the reaction, the organic phase was
10.1021/jp0487962 CCC: $27.50 © 2004 American Chemical Society Published on Web 07/14/2004
11474 J. Phys. Chem. B, Vol. 108, No. 31, 2004 removed and the aqueous phase was evaporated using vacuumevaporation under a pressure of 0.882 Pa at 45 °C for complete removal of the residual organic solution. Each sample was washed by 50 mL of MeOH, which was then removed by the vacuum-evaporation system. This manipulation was repeated several times for further removal of the remnant TBAH and NaOH. Deionized water (10 mL) was added into each of the sludge samples with continuous stirring for 24 h until the solution color became a clear reddish brown. After completion of reaction, additional deionized water (20 mL) was added into the sample, which was then purified by Sephadex G-25 column chromatography (5 × 50 cm2) with an eluent of neutralized water. The remaining trace catalyst and Na+ ions were completely removed in this step. Finally, six C60(OH)n products (a, b, c, d, e, f) of different hydroxyl groups were obtained. (The C6030 used in the experiment was prepared by the Kra¨tschmer-Huffman method31 and isolated using a highperformance liquid chromatograph (HPLC, LC908-C60, Japan Analytical Industry Co) coupled with Cosmosil 5PPB and Buckyprep preparation columns (Nacalai Co. Japan). The purified C60 was identified by a matrix assistant time-of-flight mass spectrometer (MADLI-TOF-MS, AutoFlex, Bruker Co., Germany). The purity was about 99.9%.) The hydroxyl number of the fullerenol was measured by an elemental analyzer (EA) (Hash EA112, Italy). The relative stabilities of C60(OH)n products a-f formed at different M[OH-/ C60] were investigated using a direct laser-induced dissociation technique. The nitrogen laser wavelength was 337 nm, with a laser power of about 4 µJ, which was further attenuated to be 23% in the measurement. The fragmental species were analyzed by a time-of-flight mass spectrometer (AutoFlex, Bruker Co., Germany) with negative ions and a reflection mode. Before the first measurement, fullerenol samples were placed into the vacuumizing target chamber of ∼10-8 Torr vacuum for two weeks to remove the possibly absorbed air gas. After the first measurement, the C60(OH)n samples were placed into the vacuumizing chamber and measured at 3, 6, and 8 months. The structural properties of C60(OH)n formed at different M[OH-/ C60] were analyzed by Fourier transform infrared spectroscopy on a Nicolet Magna-IR750 FT-IR spectrometer with a Nic-plan IR microscope. The IR spectra were collected from the clear reddish brown aqueous solution of C60(OH)n in the sample cells. The electronic properties of C60(OH)n a-f were studied using X-ray photoelectron spectroscopy (XPS) at the photoelectron station of Beijing Synchrotron Radiation Facility, the Chinese Academy of Sciences. The samples were deposited onto the high-purity golden substrates to obtain thin films for the XPS measurements, which were carried out at an ultravacuum chamber with background pressure of ∼8 × 10-10 and ∼1 × 10-9 Torr during the measurement. The photon with energy hν ) 400.0 eV from the synchrotron radiation was used as the excitation source. The experimental resolution was estimated to be ∼0.5 eV. To inspect the contamination, XPS survey scans on the surface were performed before and after measurements. Results and Discussion The hydroxyl number determined by the EA method for C60(OH)n, a, b, c, d, e, f, formed with M[OH-/C60] ) 0.09, 0.18, 0.38, 0.76, 1.14, and 1.52 was n ) 42 ( 2, 44 ( 2, 30 ( 2, 30 ( 2, 32 ( 2, and 36 ( 2, respectively. The highest hydroxyl number was obtained at M[OH-/C60] ) 0.18. Results of laser-induced dissociation of C60(OH)n are given in Figures 1a-f. The dissociation of the fullerenol was primarily via a successive elimination of C2 moieties. The dissociation
Xing et al. behaviors of fullerenols from Figuress 1a-f exhibit a surprising systematic feature. To examine the reproduction of the data, the same samples were repeatedly measured at a time interval of 3, 6, and 8 months, respectively. The results indicated a very good reproducibility of data. C60(OH)42 (a) yielded intense fragments from C60, C58, C56, C54 C52, C50, to C48, while C60(OH)44 (b) dissociated into fragments from C60, C58, ..., C48 down to C46. These seem to suggest that the relative stability of fullerenols depends on their hydroxyl number. But when studying experimental data of a wider range from a to f, one found that C60(OH)30 (c and d) yielded more intense dissociation than C60(OH)32 (e) and C60(OH)36 (f) did. Therefore, the systematics in both the fragmental category and the intensity diminution (from Figures 1a-f) suggest that the relative stability of fullerenols is less dependent on the hydroxyl number. There should exist other factors that play crucial roles in the stability of fullerenols. The impure groups in the fullerenol were analyzed. To remove the influence of oxygen being absorbed by C60(OH)n samples, they were degassed by placing them into a vacuumizing sample chamber of ∼10-8 Torr vacuum for 8 months. They were measured after 3, 6, and 8 months. The mass spectra obtained in these durations were the same as those shown in Figure 1. The highly oxygenated species C60O and C60OH4 were observed (in the mass region greater than m/z ) 720) in fullerenols a-e but not in f. They show systematic features with the change of NaOH. The fragments in the mass region between C60 and C60O are the hydrogenated fullerene. For easier understanding, the fragmental mass spectra in the heavier mass region were enlarged (Figure 2). Three paths are possible that result in these species. The first is the reaction of fragmental ions C60* with the absorbed oxygen in fullerenol samples, but this was excluded by the Degas manipulation in the vacuumizing sample chamber for as long as 8 months. The second path is the coalescence reaction between fragmental ions C60* and OH* in the flight tube of the time-of-flight mass spectrometer. The third one is the impure groups that originally exist in the fullerenol. If C60O and C60OH4 originated from the second path, their intensities should be sensitive to the laser power used. However, a change in the laser power hardly affected their intensities. Moreover, their production yield should be proportional to the OH* concentration (in the flight tube of the mass spectrometer) that is directly proportional to the hydroxyl number of the fullerenol molecule. The fullerenols having higher hydroxyl numbers should yield higher intensities of C60O and C60OH4, but they were also not observed. In Figure 3a,b, intensities of the C60O and C60OH4 ions as a function of the hydroxyl number of the fullerenol are studied. For a comparable study of different fullerenols (a-f), the intensity was normalized to that of the corresponding C60 ion. Up to n ) 36, it exhibits a clear tendency of decreasing impurity content as the hydroxyl number increases. In more highly hydroxylated fullerenes with n > 36, this tendency is reversed, suggesting that in these molecules other reaction paths (which will be further discussed later on) are possible. To study if the C60O and C60OH4 originated from impure groups originally existing in the cage of the fullerenol, their mass intensities were plotted as a function of M[OH-/C60] in Figure 4a,b, respectively. A nearly linear correlation between M[OH-/C60] and intensities of the C60O and C60OH4 ions is indicated, suggesting close relations between them and fullerenols, because their formations in the fullerenol were found to be essentially dominated by M[OH-/C60] in the synthesis process.
Influences of Structural Properties on Fullerenols
Figure 1. Time-of-flight mass spectra of a direct laser-induced dissociation of C60(OH)42 (a), C60(OH)44 (b), C60(OH)30 (c), C60(OH)30 (d), C60(OH)32 (e), and C60(OH)36 (f). All were measured with the negative ion and a reflect modes (without matrix) under the same experimental conditions.
To clarify the factors that dominate the stability of the fullerenol molecule, the degree of dissociation (DD, stability parameter) was extracted from the experimental data and defined by the quantity of fragmental species. For instance, C60(OH)42 (a) dissociated into seven fragmental species, C60, C58, C56, C54 C52, C50, to C48, and its degree of dissociation is defined to be 7, while C60(OH)36 (f) dissociated into only C60, its DD being 1. Here, the larger the DD value is, the lower the stability of the fullerenol is. DD was then studied as functions of the hydroxyl number n (Figure 5) and M[OH-/C60] (Figure 6). In Figure 5, results do not show a functional relation over all the data, implying less correlation between the stability and hydroxyl number of C60(OH)n. Up to n ) 36, there is a clear tendency of increasing stability as the hydroxyl number increases, but this tendency is reversed in more highly hydroxylated fullerenes. On the other hand, a good correlation between DD and M[OH-/ C60] is observed in Figure 6. An important finding here is that the DD of C60(OH)n shows a nearly linear decrease with the increase of M[OH-/C60], suggesting that a high stability of the fullerenol can be gained by controlling the molar ratio M[OH-/ C60]. However, at the conditions (M[OH-/C60] ) 0.09 and 0.18) rendering the highest hydroxylated fullerenes (n > 36), it deviates the most from the linear relationship, in close agreement with the behaviors found in Figures 3 and 5. The DD was further studied as a function of the intensity of C60O (Figure 7a) and C60OH4 (Figure 7b).The results indicate that a larger amount of impure groups in the fullerenol led to the larger instability of the fullerenol molecule. This analysis
J. Phys. Chem. B, Vol. 108, No. 31, 2004 11475
Figure 2. Time-of-flight mass spectral measurements of the highly oxygenated carbons in the direct laser-induced dissociation processes of C60(OH)42 (a), C60(OH)44 (b), C60(OH)30 (c), C60(OH)30 (d), C60(OH)32 (e), and C60(OH)36 (f). The others are the same as those described in Figure 1.
Figure 3. Mass intensities of C60O (a) and C60OH4 (b) (normalized to C60) as a function of the hydroxyl number of C60(OH)n.
suggests that the stability of C60(OH)n directly depends on the number of impure groups formed in the fullerenol cage. As the IR characteristic absorptions help to understand some of the features observed in Figure 2, particularly the source of the hydrogenated fullerenes in the mass range between C60 and C60O, all the six fullerenols were studied by FT-IR spectroscopy. FT-IR spectra fullerenols a-f are displayed in Figure 8a-f, respectively. The absorption wavenumbers are listed in Figure 8a, unless otherwise indicated, and the wavenumbers in other portions are constant within experimental error. The C-O
11476 J. Phys. Chem. B, Vol. 108, No. 31, 2004
Xing et al.
Figure 4. Mass intensities of C60O (a) and C60OH4 (b) (normalized to C60) as a function of M[OH-/C60]. Figure 7. DD as a function of the mass intensity of C60O (a) and C60OH4 (b) (normalized to C60).
Figure 5. Degree of dissociation (DD) of C60(OH)n as a function of the hydroxyl number (n).
Figure 6. DD as a function of M[OH-/C60].
stretching absorption at about 1088 cm-1, an intense broad O-H band at about 3410 cm-1, and a very strong O-H vibration frequency at 1412 cm-1 indicated the -C-O-H structure of the fullerenol. The IR spectrum shows an intense band at 1593 cm-1 for the undestroyed CdC bond. These characteristic values of IR absorptions are consistent with results reported by Chiang et al.5 More important information obtained from the IR spectral measurements of C60(OH)n is the absorption frequency for the impure groups. The C-H absorption band at 2855 cm-1 was observed in a, b, and c (formed at the relatively lower M[OH-/ C60] ) 0.09, 0.18, and 0.38). They exhibit the tendency to clear up as the amount of NaOH increases. As seen in Figure 2, the fragmental species of the laser-induced dissociation of C60(OH)n hydrogenated fullerenes were also observed in a, b, and c. Thus, the alkyl hydrogen observed by IR spectra explained the origin of the fragments (hydrogenated fullerene) observed in the mass region between C60 and C60O. In addition, a weak hemiketal
Figure 8. FT-IR spectra for C60(OH)42 (a), C60(OH)44 (b), C60(OH)30 (c), C60(OH)30 (d), C60(OH)32 (e), and C60(OH)36 (f). The M[OH-/C60] used in the synthesis process is indicated inside the plot. The characteristic absorptions of C60(OH)n are marked by the value of wavenumbers without arrows. Arrows indicate absorption bands of impure groups.
band centered at 1658 cm-1, a weak lactone at about 2927 cm-1, and the carbonyl stretching absorption band at 1710 cm-1 were observed. With a further increase of M[OH-/C60], IR spectra
Influences of Structural Properties on Fullerenols of C60(OH)n formed at the relatively higher NaOH concentration (M[OH-/C60] ) 0.76, 1.14, and 1.52) became clean. On the other hand, formations of these impure groups strongly depend on the alkalinity of the synthesis conditions. This becomes more explicit when comparing the data in Figure 8c-e. Despite the similar hydroxy number of products c, d, and e, absorption bands of impure groups became weaker and weaker from c, to d, to e. This clearly suggests that the use of a higher molar ratio M[OH-/C60] or excess alkali in the synthesis process of C60(OH)n is propitious to obtain the fullerenol with fewer impure groups. This conclusion is consistent with that drawn from the results in Figures 4, 6, and 7. Because the strong carbonyl absorption was observed from the fullerenol after an acidic treatment,7 Chiang et al. proposed a mechanism of conversion of the vicinal hydroxyl group to the ketone structure (pinacol rearrangement) to account for the formation of the carbonyl group in fullerenols. The present reaction and all chemical processes took place under alkaline conditions without the presence of the acid; that is, it was not a chemical condition allowing the occurrence of the pinacol rearrangement. In fact, before the acidic treatment there still existed a weak band at 1720 cm-1 in the IR spectrum of ref 7, and after acidic treatment this absorption was strongly enhanced. This suggests that there should exist other reaction mechanisms that can also lead to formation of impure groups such as ketone structures and alkyl hydrogens in fullerenols. In fact, Rinco´n et al. have drawn similar conclusions from their work studying low hydroxylated fullerenes.1 Recently, Wudl, Hummelen, and Ptator observed a similar frequency (1720 cm-1) for a characterized cage-opened ketone structure formed by oxidation of an N-methoxyethoxymethylazahomo-[60]fullerene.32 More recently, Taylor et al. reported the same carbonyl absorption band observed from a cage-opened bisepoxide fullerenol which was formed via 1,3-tautomeric shift of the hydroxyl hydrogen in fullerenols.33 This process took place without the acidic environment similar to the present reaction conditions. Thus, we assume that in the alkaline synthesis reaction of fullerenols the tautomeric shift of the hydroxyl hydrogen can take place and lead to formation of the carbonyl and C-H structures. In the 1,3-tautomeric shift of the hydroxyl hydrogen, the reaction equilibrium is held by the H+ of the solution, which accepts the electron from the reactive center of high electronic density in the fullerenol molecule and is the source of the alkyl hydrogen. As NaOH increases, it lowers the H+ content of the system and hence inhibits the 1,3tautomerism of the hydroxyl hydrogen from shifting to the side of the alkyl hydrogen. This consequently prohibits the formation of the carbonyl and C-H groups. Moreover, the mechanism of 1,3-tautomerism shift is consistent with observed phenomena that formations of impure groups were inhibited and the stability of the fullerenol was decreased by increasing NaOH. Note that the existence of the tautomerism process in the alkaline synthesis process of fullerenols is not in contradiction with the process of the pinacol rearrangement forming a ketone structure in fullerenols under acidic conditions, because they were observed in fullerenols formed under different chemical processes. The electronic structures of fullerenols were analyzed by measuring binding energy spectra of C 1s electrons. The results are given in Figure 9a-f for C60(OH)n a-f, respectively. The main line of XPS spectra shows a broad and asymmetric shape. This is different from the symmetric main line of C60, which is well described by a true Voigt function with a Gaussian dispersion.34 The Gaussian analysis of XPS data showed the presence of three components (indicated by broken, dotted, and
J. Phys. Chem. B, Vol. 108, No. 31, 2004 11477 broken-dotted lines, respectively). The C 1s binding energies observed for sp2 nonfunctionalized carbons are centered at 284.7 ( 0.2 eV, in good agreement with the value observed from C60.34 Those for hydroxylated (C-OH) and highly oxygenated (CdO) carbons are centered at 285.8 ( 0.3 and 288.3 ( 0.3 eV, respectively. The carbonyl group was observed in all fullerenols a-f. Compared with the IR measurement, the higher sensitivity of the XPS methodology revealed more detailed information that highly oxygenated carbons of impure groups were contemporarily formed in fullerenols at any alkalinity of M[OH-/C60]. So far, three methods have been developed for the synthesis of the fullerenol; oxygen takes part in the reaction and plays crucial roles in any of these processes.7,35,36 In these oxidative processes of cage-functionalization of fullerenes, the contemporary formation of the highly oxygenated carbons might not be prevented. The intensities of C 1s components in the fullerenol were estimated from integration of the peak area under each broken line. The relative intensities for sp2 nonfunctionalized carbon (ICdC), hydroxylated carbon (IC-O), and highly oxygenated carbon (ICdO) groups were obtained by normalizing them to the total area under the solid curve of Figure 9. The XPS intensities for the hydroxylated and highly oxygenated carbons are displayed in Figure 10 as a function of M[OH-/C60]. As IC-O mostly originates from hydroxylated carbon atoms (COH) in the fullerenol, its varying tendency similar to the EAmeasured number of hydroxyl groups indicates the reliability of the data analysis procedure. The results indicate that M[OH-/ C60] ) 0.1-0.2 is the best alkaline region for obtaining a fullerenol of the highest hydroxyl number. But on the other hand, the intensities of impure groups (solid diamonds) are much higher in this range than those at M[OH-/C60] > 0.2 and decline quickly as M[OH-/C60] increases. For instance, ICdO is about 18% at M[OH-/C60] ≈ 0.18, it decreases to about 5% at M[OH-/C60] ≈ 1.5. These results suggest that the formation of the CdO structure in the fullerenol can be repressed through increasing alkalinity of the reaction conditions. This is consistent with the results of the FT-IR measurement and laser-induced dissociation experiments. In Figure 11, the intensity ratio of highly oxygenated carbons relative to hydroxylated carbons (R[C)O/C-OH]) is studied as a function of DD. The former represents the structural parameter, and the latter stands for the parameter of stability. The DD increased with an increase of R[C)O/C-OH]. This revealed an interesting fact that the fullerenol became more unstable when the content of highly oxygenated carbons increases. In ref 33 Taylor et al. found that formation of highly oxygenated carbons could result in a coinstantaneous formation of a cage-opened structure in C60Me5O3H. Similar to their finding, formations of highly oxygenated carbons in fullerenols may result in a coinstantaneous formation of a cageopened structure in the fullerenol cage. It is hence understood that the stability of the fullerenol sensitively depends on the quantity of highly oxygenated carbons, because the formation of a cage-opened structure leads to a large instability of the fullerenol molecule. Conclusion The stability of six fullerenol molecules (C60(OH)42, C60(OH)44, C60(OH)30, C60(OH)30, C60(OH)32, C60(OH)36) was investigated by laser-induced dissociation associated with a TOF-MS measurement, XPS, and FT-IR spectroscopy. Intensities for fragmental species and for nonfunctionalized (sp2) carbons (CdC), hydroxylated carbons (C-OH), and highly oxygenated carbons (CdO) of the fullerenol molecule were
11478 J. Phys. Chem. B, Vol. 108, No. 31, 2004
Xing et al.
Figure 10. C 1s XPS intensities of hydroxylated carbons (C-OH) and highly oxygenated carbons (CdO) (normalized to that of undestroyed sp2 CdC carbons) as a function of M[OH-/C60].
Figure 11. DD as a function of R[C)O/C-OH], the quantity ratio of Cd O and C-OH, measured by the C 1s XPS.
Figure 9. C 1s XPS spectra for C60(OH)42 (a), C60(OH)44 (b), C60(OH)30 (c), C60(OH)30 (d), C60(OH)32 (e), and C60(OH)36 (f). The M[OH-/C60] used in the synthesis process is indicated inside each plot. The spectra for sp2 nonfunctionalized carbons (CdC), the hydroxylated carbons (C-OH), and highly oxygenated carbons (CdO) are indicated.
analyzed as functions of the hydroxyl number (n), the degree of dissociation (DD) of the fullerenol, and molar ratio M[OH-/ C60]. The stability of the fullerenol was then studied as functions of the structural parameter, hydroxyl number, intensity of impure group, and R[C)O/C-OH]. Under the same laser power, C60(OH)42 decomposed into fragmental species C60, C58, C56, C54, C52, C50, and C48; C60(OH)44 into C60, C58, C56, C54, C52, C50, C48, and C46; C60(OH)30 into C60, C58, C56, C54, and C52; but C60(OH)32 into C60 and C58; and C60(OH)36 decomposed into only C60. A careful analysis of intensities of observed C60O and C60OH4 ions as functions of n and M[OH-/C60] revealed that they originated from impure
groups of the fullerenol molecule and revealed a nearly linear relationship between the intensity of impure groups and M[OH-/ C60]. A weaker correlation between the stability and hydroxyl number, but a strong relationship between the stability and the intensity of impure groups of C60(OH)n were observed. The FTIR and XPS results indicated that coinstantaneous formations of highly oxygenated carbons such as a ketone structure in the fullerene cage could not be turned away from formation processes of fullerenols, even if they were prepared using the alkaline reaction instead of strong acidic or oxidative processes. The impure groups, especially the highly oxygenated carbons that may lead to a cage-opened structure, can largely lower the stability of the fullerenol molecule. The fullerenol molecules formed at the higher M[OH-/C60] were much more stable than those formed at the lower M[OH-/C60]. When the quantity ratio of highly oxygenated carbons (CdO) to hydroxylated carbons (C-OH) in a fullerenol, R[C)O/C-OH], became smaller, the fullerenol became more stable. Intensities of both the mass and XPS of highly oxygenated carbons were quickly decreased for fullerenols formed at a higher M[OH-/C60]. As the solubility of C60(OH)n depends on the hydroxyl number, a conventional consideration was to increase the number of hydroxyls to gain a larger solubility in practical applications. But the present results indicate that this idea needs to be restudied, because the fullerenol molecule of C60(OH)n>36 could lead to unstable opencage structures. The use of M[OH-/C60] >0.3 for fullerenol synthesis could achieve relatively pure fullerenols of less impure groups. Fullerenols having a structure of R[C)O/C-OH] < 0.2 were mostly stable, indicating the importance of the structural balance between the quantity of hydroxyl and impure groups. Acknowledgment. We are grateful to Prof. Y. F. Liu of Peking University for many discussions. We thank Prof. Long Wei and Dr. Chuang Xing Ma for experimental assistance with vacuum molecular spattering. Y.L.Z. acknowledges the fund support from the Ministry of Science & Technology (2001CCA03800), a major project of the National Natural
Influences of Structural Properties on Fullerenols Science Foundation of China (10490180), and State Key Laboratory of Spectroscopy and Atomic and Molecular Physics (T152302). K. Ibrahim acknowledges the fund supports from NSFC (10274084). A part of the experiments were carried out at Beijing Synchrotron Radiation Facility. References and Notes (1) Rincon, M. E.; Hu, H.; Campos, J.; Ruiz-Garcia, J. J. Phys. Chem. B 2003, 107, 4111. (2) Hinokuma, K.; Ata, M. Chem. Phys. Lett. 2001, 341, 442. (3) Li, Y. M.; Hinokuma, K. Solid State Ionics 2002, 150, 309. (4) Gosivami, T. H.; Nandan, B.; Alam, S.; Mathur, G. N. Polymer 2003, 44, 3209. (5) Chiang, L. Y.; Upasani, R. B.; Swirczewski, J. W. J. Am. Chem. Soc. 1992, 114, 10154. (6) Naim, A.; Shevlin, P. B. Tetrahedron Lett. 1992, 33, 7097. (7) Chiang, L. Y.; Upasani, R. B., Swirczewski, J. W.; Soled, S. J. Am. Chem. Soc. 1993, 115, 5453. (8) Chiang, L. Y.; Wang, L. Y.; Tseng, S. M.; Wu, J. S.; Hsieh, K. H. J. Chem. Soc., Chem. Commun. 1994, 2675. (9) Chiang, L. Y.; Wang, L. Y.; Tseng, S. M.; Wu, J. S.; Hsieh, K. H. Synth Met. 1995, 70, 1477. (10) Chiang, L. Y. In The Chemistry of Fullerenes; Taylor, R., Ed.; World Scientific: Singapore, 1995; p 67. (11) Slanina, Z.; Lee, S.-L.; Adamowicz, L.; Chiang, L. Y. In The Chemistry and Physics of Fullerenes and Related Materials; Kadish, K. M., Ruoff, R. S., Eds.; Electrochemical Society: Pennington, NJ, 1996; Vol. 3, p 987. (12) Trajmar, S.; Kanik, I. Chem. Phys. Lett. 1996, 262, 241. (13) Wang, D.-C.; Cheng, C.-Y. J Mol. Struct. (THEOCHEM) 1997, 391, 179. (14) Slanina, Z.; Sugiki, T.; Chiang, L. Y.; Osawa, E. In The Chemistry and Physics of Fullerenes and Related Materials; Kadish, K. M., Ruoff, R. S., Eds.; Electrochemical Society: Pennington, NJ, 1997; Vol. 4, p 721. (15) Chiang, L. Y.; Lu, F.-J.; Lin, J.-T. J. Chem. Soc., Chem. Commun. 1994, 1283. (16) Chiang, L. Y.; Lu, F.-J.; Lin, J.-T. In Science and Technology of Fullerene Materials; Bernier, P., Bethune, D. S., Chiang, L. Y., Ebbesen, T. W., Metzger, R. M., Mintmire, J. W., Eds.; Materials Research Society: Pittsburgh, 1995; p 327.
J. Phys. Chem. B, Vol. 108, No. 31, 2004 11479 (17) Friedman, S. H.; DeCamp, D. L.; Sijbesma, R. P.; Srdanov, G.; Wudl, F.; Kenyon, G. L. J. Am. Chem. Soc. 1993, 115, 6506. (18) Friedman, S. H.; Ganapathi, P. S.; Rubin, Y.; Kenyon, G. L. J. Med. Chem. 1998, 41, 2424. (19) Schuster, D.; Wilson, S. R.; Schinazi, R. F. Bioorg. Med. Chem. Lett. 1996, 6 (11), 1253. (20) Marcorin, G. L.; Ros, T. D.; Castellano, S.; Stefancich, G.; Bonin, I.; Miertus, S.; Prato, M. Org. Lett. 2000, 2 (25), 3955. (21) Tokuyama, H.; Yamago, S.; Nakamura, E.; Shiraki, T.; Sugiura, Y. J. Am. Chem. Soc. 1993, 115, 7918. (22) Boutorine, A. S.; Tokuyama, H.; Taskasugi, M.; Isobe, H.; Nakamura, E.; Helene, C. Angew. Chem., Int. Ed. Engl. 1994, 33, 2462 (23) Okada, E.; Komazawa, Y.; Kurihara, M.; Inoue, H.; Miyata, N.; Okuda, H.; Tsuchiya, T.; Yamakoshi, Y. Tetrahedron Lett. 2004, 3, 527. (24) Ros, D. T.; Prato, M. Chem. Commun. 1999, 663. (25) Mikawa, M.; Kato, H.; Okumura, M.; Narazaki, M.; Kanazawa, Y.; Miwa, N.; Shinohara, H. Bioconjugate Chem. 2001, 12, 510. (26) Bolskar, R. D.; Benedetto, A. F.; Husebo, L. O.; Price, R. E.; Jackson, E. F.; Wallace, S.; Wilson, L. J.; Alford, J. M. J. Am. Chem. Soc. 2003, 125, 5471. (27) Slanina, Z.; Zhao, X.; Chiang, L. Y.; Osawa, E. Int. J. Quantum Chem. 1999, 74, 343. (28) Wang, B. C.; Cheng, Ch. Y. J. Molecular Structure (THEOCHEM) 1997, 391, 179. (29) Wang, B. C.; Wang, H. W.; Tso, H. C.; Chen, T. L.; Chou, Y. M. J. Mol. Struct. (THEOCHEM) 2002, 581, 177. (30) Kroto, H. W.; Heath, J. R.; O’Brien, S. C.; Curl, R. F.; Smalley, R. E. Nature 1985, 318, 162. (31) Kratschmer, W.; Lamb, L. D.; Fostiropoulos, K.; Huffiman, D. R. Nature 1990, 347, 354. (32) Hummelen, J. C.; Ptator, M.; Wudl, F. J. Am. Chem. Soc. 1995, 117, 7003. (33) Al-Matar, H.; Abdul-Sada, A. K.; Avent, A. G.; Taylor, R. Org. Lett. 2001, 3 (11), 1669. (34) Leiro, J. A.; Heinonen, M. H.; Laiho, T.; Batirev, I. G. J. Electron Spectrosc. Relat. Phenom. 2003, 128, 205. (35) Iezzi, E. B.; Cromer, F.; Stevenson, P. Synth. Met. 2002, 128, 289. (36) Li, T.-B.; Huang, K.-X.; Li, X.-H.; Jiang, H.-Y.; Li, J. Chem. J. China UniV. 1998, 6, 858.
