An efficient, soluble, and recyclable multiwalled carbon nanotubes-supported TEMPO for oxidation of alcohols

Abstract

The immobilization of homogeneous catalysts is a continuing goal for combining the advantages of both homogeneous and heterogeneous catalysis. However, a significant loss in catalytic activity is often found in the immobilization of a homogeneous catalyst. Herein, we report a novel strategy consisting of multiwalled carbon nanotubes (MWNTs) functionalized with homogeneous catalysts that are developed to combine the positive aspects of solid and soluble supports. Using the oxidation of alcohols as a model reaction, the supported catalysts (MWNTs–TEMPO) can be homogeneously dispersed in the reaction medium to conquer the mass transfer limitation, which leads to their catalytic activity being far superior to their heterogeneous counterpart and similar to their parent catalysts. In addition, they exhibit the additional advantages of characterization with solution-based techniques, easy separation and reutilization.

---

1. Introduction

The high activity and selectivity of homogeneous catalysts for organic transformations is of particular interest to chemists.¹ However, in homogeneous processes there is a perceived inability to separate the catalysts from the products readily. This has driven researchers to make long-term, unremitting efforts to tackle the problem.^2–4 Most of the approaches are devoted to tethering the catalyst on insoluble organic supports^5,6 and inorganic solids.^7,8 The utmost advantage associated with this strategy resides in the ready recovery and reuse of the catalysts through a simple filtration manipulation. Because most of the previous efforts generally suffered from mass transfer limitations, a substantial decrease in the activity of the immobilized catalysts was frequently observed.^9–11 Recent advances in metal nanoparticles^12–14 and especially semi-heterogeneous nanoparticles^15–18 as supports may mitigate the problems. Nevertheless, for these grafted catalyst systems, only a few examples exhibited equal or even superior activity relative to their counterparts. From the standpoint of achieving a comparable catalytic activity to their monomeric analogues, a feasible alternative is therefore the use of soluble polymeric supports and biphasic systems.^19–22 But the catalysts in the mixture were most commonly recovered by precipitation or extraction concerning the use of a large amount of solvent to achieve quantitative recovery. The use of excess solvents makes this process less convenient and also significantly detrimental to the environment. Hence, it is still desirable to develop new methods, if possible, combining the advantages of both immobilization of the solid and soluble support systems without their drawbacks.

Although carbon nanotubes have high chemical stability, large surface area and outstanding mechanical properties, only very limited literature concerns their use as a support for covalently bonded homogeneous catalysts.^23–27 Due to the limited solubility for these supported catalysts in the reaction mediums, these reported processes often focused on heterogeneous catalysis. However, if the homogeneous catalysts tethered on the carbon nanotubes were completely dissolved, they may mimic soluble polymeric supports to conquer the mass transfer limitations during the catalytic process. For that, one must overcome the critical issue of limited solubility to fully exploit the activities. It has been proposed that the functionalization of carbon nanotubes with organic molecules could improve the solubility of the nanotubes in aqueous or organic solvents.^28,29 Inspired by the high solubility of functionalized carbon nanotubes^30,31 we have ignited an effort to explore, when the carbon nanotube anchored homogeneous catalysts are soluble in the reaction medium,³² whether the catalytic activity is maintained in comparison with their homogeneous analogues.

Herein, we present the synthesis and characterization of 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO) attached on oxidized MWNTs, which are evaluated as heterogeneous and soluble catalysts for the oxidation of alcohols. To our delight, the soluble MWNTs–TEMPO catalyst proved to be effective and highly selective for the oxidation reaction of primary and secondary alcohols under standard Anelli's conditions.³³ Additionally, we further investigated the substrate scope and recycling of the catalyst, as well as the effect of the solubility of the functionalized MWNTs on the catalytic activity.

2. Results and discussion

2.1 Characterization by means of conventional techniques

The general strategy for the synthesis of immobilized TEMPO on multiwalled carbon nanotubes (MWNTs) is described in Scheme 1. Commercial MWNTs were firstly oxidized in concentrated HNO₃ solution, resulting in carboxylic acid-functionalized MWNTs. It is noteworthy to mention that the number and location of those addends depend on the oxidation conditions.^30,34–40 The oxidized MWNTs were converted into the acyl chlorides by refluxing in thionyl chloride, and by subsequent direct reaction with 4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl (HO–TEMPO). To prove the covalent linkage of the homogeneous catalysts, FT-IR was used to characterize the chemical composition (Fig. 1). The infrared spectra of pristine MWNTs, MWNTs–COOH and MWNTs–TEMPO are shown in Fig. 1. As shown in Fig. 1a, purified MWNTs exhibit three peaks in the range of 400–2000 cm⁻¹. After the oxidation of the tubes (Fig. 1b), the appearance of the peak situated at 1720 cm⁻¹ indicates the existence of carboxyl groups, and the peaks located at 1577 and 1228 cm⁻¹ are assigned to the C=C stretch of carbon nanotube backbones and the C–O stretch of the acid groups, respectively. After the acidification, the signals due to the stretching vibration of the carbonyl groups change from 1720 to 1730 cm⁻¹ (Fig. 1c). The new peak of the C–N stretching in TEMPO groups was also observed at 1350 cm⁻¹, indicating success in the functionalization of the MWNTs.

Fig. 1 Fourier transform infrared (FT-IR) spectra of pristine MWNTs (a), MWNTs–COOH (b) and MWNTs–TEMPO (c).

Scheme 1 Immobilization of TEMPO onto the MWNTs.

Raman spectroscopy was not only used to probe the structural integrity and electronic properties of the modified MWNTs materials, but also to gather information about the degree of nanotube functionalization. In Fig. 2, the D-band at about 1332 cm⁻¹, related to scattering from defects present in the hexagonal framework of the nanotube walls, and the G-band at about 1590 cm⁻¹, ascribed to the tangential C–C stretching vibrational mode,⁴¹ can be clearly observed for pristine MWNTs, MWNTs–COOH and MWNTs–TEMPO, indicating that the functionalized MWNTs still preserve their basic structure. A shoulder around 1600 cm⁻¹ assigned to the D′-band is ascribed to disorder and defects arising from a double resonance feature. Relative to the starting material, the D-band was observed to increase through oxidation (Fig. 2a, b), owing to the disruption of the graphene π-bonded electronic structure of the side walls. Also, we noted a slight increase of an intense I_D/I_G ratio after functionalization with TEMPO due to the influence of the grafted TEMPO on the electronic properties of the MWNT–COOH, as well as the production of field disturbance and physical strain in the graphite skeleton as reported.^42,43 The intensity enhancement of the D-band in MWNTs–TEMPO samples proves the covalent bonding of TEMPO to MWNTs, instead of physical adsorption with the graphene layer. The presence of TEMPO was further characterized by XPS and the detection signals of the nitrogen elements peaked at 400.5 eV (ESI Fig. 1†).

Fig. 2 Raman spectra of functionalized MWNTs: (a) pristine MWNTs, I_G/I_D = 0.66; (b) MWNTs–COOH, I_G/I_D = 0.78; and (c) MWNTs–TEMPO, I_G/I_D = 0.86.

Thermal gravimetric analysis (TGA) of the derivatized MWNTs materials was done in an inert environment to assess the amount of functional groups as shown in Fig. 3. Compared with MWNTs and MWNTs–COOH, TGA analyses of MWNT–TEMPO showed one major decomposition in the temperature range from 180 to 320 °C corresponding to TEMPO on the MWNTs. As expected, functionalization resulted in a higher degree of weight loss of 10.8% for TEMPO. This result is almost in agreement with ∼0.74 mmol g⁻¹ of nitroxyl radical in accordance with the element analysis.

Fig. 3 Thermogravimetric analysis (TGA) data under pure nitrogen at a heating rate of 10 °C min⁻¹ for (top, black) pristine MWNTs, (middle, blue) MWNTs–COOH and MWNTs–TEMPO (bottom, red).

2.2 Characterization with solution-based techniques

It is well known that homogeneous catalysts tethered on solid materials are insoluble, and difficult to explore using solution-based techniques. During a multi-steps sequential preparation, the difficulty of monitoring reactions often requires several time-consuming and costly elemental analyses, and final catalyst loadings are somewhat unpredictable. While MWNTs–TEMPO is readily soluble in water (3.7 mg ml⁻¹) and CH₂Cl₂ (0.23 mg ml⁻¹), forming an approximately homogeneous solution (Fig. 4), there was almost no sedimentation observed even after 30 days. The TEM image (Fig. 5) showed that the original MWNTs were piled up, but dispersed individually after immobilization of TEMPO because the TEMPO moieties on the MWNTs surfaces are effective in preventing aggregation. The solubility of these functionalized carbon nanotubes makes it possible characterize and study their properties using solution-based techniques. The solution phase UV-vis spectra (Fig. 6) gives evidence that the dependence of absorbance on solution concentration apparently obeys Beer's law in CH₂Cl₂, indicating optical behaviour associated with a homogeneous dispersion with individual bundles of tubes. Owing to the paramagnetism of the radicals for TEMPO, the chemical structure was also determined from solution ¹H NMR using TEMPOH. Due to the relatively high solubility of MWNTs–TEMPOH in CDCl₃, the chemical structure was also determined and shown in Fig. 7. The reaction of MWNTs–TEMPO with Na₂S₂O₄ yielded the expected MWNTs–TEMPOH.⁴⁴ In its ¹H NMR spectrum, the characteristic peaks of MWNTs–TEMPOH were clearly found. ¹H NMR was also a useful guide for optimizing the synthesis of supported catalysts: in situ¹H NMR increased the understanding of catalyst synthesis and confirmed the supported amount of TEMPOH under varying reaction conditions. All the evidence supports the MWNTs being covalently bound to TEMPO and not by adsorption.

Fig. 4 Photographs of (a) pristine MWNTs, (b) dispersions of MWNTs–COOH and (c) MWNTs–TEMPO in deionized water.

Fig. 5 TEM image of the original MWNTs (a), showing some bundled tubes. TEM image of the TEMPO functionalized MWNTs (b), showing debundling and individual tubes.

Fig. 6 UV-vis spectrum of MWNTs–TEMPO in CH₂Cl₂ at different concentrations (increasing solution concentration in the direction of the arrow). Shown in the inset is the linear dependence of absorption on solution concentration at 265 nm.

Fig. 7 ¹H NMR spectrum of MWNTs–TEMPOH.

2.3 Activity comparison for the oxidation of alcohols between the insoluble and soluble nanotubes functionalized with TEMPO

As a test reaction we chose the selective oxidation of alcohol to aldehyde, which is clear mechanisticly and, therefore, suitable for this purpose.³³ The oxidation began with benzyl alcohol as a prototypical substrate under Anelli's conditions, using NaOCl as the terminal oxidant and KBr as the co-catalyst (Table 1). When the content of nitroxyl radical is about 0.74 mmol g⁻¹, the supported catalysts can be almost dispersed in the reaction medium. In the absence of the bleach, the MWCNTs–TEMPO alone could hardly catalyze the oxidation of benzyl alcohol (entry 1). Regarding nitroxyl radicals supported on silica materials^45–47 and nanoparticals,^48,49 under Alleni's conditions, it is reported that the reaction time for oxidation of benzyl oxidation is beyond 30 min,^45,47 even 60 min.^46,49 In addition, TEMPO supported on polymer resins was reported at about 30 min⁵¹ and 150 min⁵⁰ for the complete oxidation of 1-octanol. The designed catalytic system afforded the desired efficiency in conversion and selectivity within 5 min (entry 2). It is worthy to note that the activity of the immobilized catalyst was similar with that of the nonsupported catalysts (entries 2 and 3). Hence, the catalysts immobilized on the soluble carbon nanotubes exhibited a higher activity than on the solid materials due to their homogeneous dispersion in the reaction medium.

Table 1 Activity comparison for the oxidation of alcohols between the insoluble and soluble nanotubes functionalized with TEMPO^a

Entry	Catalyst	Nitroxyl content (mmol g^−1)	Substrate	Conv.^b (%)	Sel. ^b (%)
a 0.8 mmol substrate, 8 μmol nitroxyl, 0.16 ml 0.5 M KBr (10%), 2 ml CH2Cl2, 2.86 ml, 0.35 M hypochlorite, 0.13 g KHCO3 (for pH 9.1), 0 °C, 5 min. b Conversions and selectivities are based on ^1H NMR. c Soluble supported catalysts. d Only MWNTs–TEMPO. e Heterogeneous supported catalysts.
1^c^d	MWNTs–TEMPO	0.74	benzyl alcohol	<2	—
2^c	MWNTs–TEMPO	0.74	benzyl alcohol	100	>99
3	TEMPO	—	benzyl alcohol	100	>99
4^e	MWNTs–TEMPO	0.21	benzyl alcohol	59.8	>99
5^e	MWNTs–TEMPO	0.21	1-phenylethanol	37.3	>99
6^c	MWNTs–TEMPO	0.74	1-phenylethanol	100	>99

---

To ascertain whether the homogeneously dispersed catalysts accelerated the rate of the oxidation or not, the catalytic activity was also investigated for insufficiently functionalized tubes as heterogeneous catalysts. With insufficient functionalization and the content of the nitroxyl radical at 0.21 mmol g⁻¹, most of the nanotubes were found as sedimentation in the bottom of the bottle. Though the catalysts are exposed to the same surrounding, the insoluble catalysts with the same catalyst loading displayed an obvious decrease in conversion (entry 4). To provide more evidence for the effect on the activity between soluble solid and heterogeneous catalysts, we further expanded the test to a secondary alcohol to determine the catalytic behaviour, because secondary alcohols often have relatively lower reactivity. Likewise, the heterogeneous catalysts also displayed lower conversion but complete conversions were still obtained with soluble solid catalysts for the oxidation of 1-phenylethanol within 15 min (entries 5, 6). Taken together, besides the reactants being easy access to catalysts on the surface of MWNTs, the solubility of supported catalysts is fully conclusive due to the supported catalysts being homogeneously soluble in the reaction medium. It can be said that, while oxidation continued, the soluble adducts could diffuse very quickly and experience the same solvation environment, rendering it easier for them to access the reactants. Similar observations were reported for dispersible nanoplatinum–carbon nanotube hybrids in solvent between homogeneous catalysis and heterogeneous catalysis.⁵²

2.4 Oxidation of primary and secondary alcohols

With the use of soluble MWNTs–TEMPO as the catalyst, we next extended the scope of the substrates, and the results were summarized in Table 2. The primary benzylic alcohols and aliphatic alcohols were smoothly oxidized into the corresponding aldehydes with high TOFs (turnover frequency), and selectivities above 99% (entries 1–8). It was found that electronic properties of the substituent on the benzene rings have a negligible influence on the reactivity in the cases of primary benzylic alcohols. Notably, in contrast to the unsupported TEMPO,³³ primary benzylic alcohols with electron-donating groups (–OCH₃) (entry 3), were more active with this catalytic system. Secondary alcohols are less reactive, and thus longer reaction times are needed to obtain complete conversions (entries 9–11).

Table 2 Soluble MWNTs–TEMPO catalyzed oxidation of primary and secondary alcohols to carbonyl compounds^a

Entry	Substrate	Time (min)	Conv.^b (%)	Sel.^b (%)	TOF (h^−1)
a 0.8 mmol substrate, 8 μmol nitroxyl, 0.16 ml 0.5 M KBr (10%), 2 ml CH2Cl2, 2.86 ml, 0.35 M hypochlorite, 0.13 g KHCO3 (for pH 9.1), 0 °C, ∼0.74 mmol g^−1 of nitroxyl radical. b Conversions and selectivities are based on ^1H NMR. c 40 μmol nitroxyl, other conditions are the same as with a.
1	benzyl alcohol	5	100	>99	1200
2	4-nitrobenzyl alcohol	5	100	>99	1200
3	4-methoxybenzyl alcohol	5	100	>99	1200
4	2-phenylethanol	5	100	>99	1200
5	1-octanol	5	100	>99	1200
6	3-methyl-1-butanol	5	100	>99	1200
7	1-undecanol	5	100	>99	1200
8	1-butanol	5	100	>99	1200
9	1-phenylethanol	15	100	>99	400
10	benzhydrol	15	100	>99	400
11^c	cyclohexene	30	100	>99	400

---

2.5 Catalyst recycling

Based on the findings mentioned above, we examined the reusability of the MWNTs-supported catalysts in order to demonstrate further advantages of this type of soluble support. After the completion of the oxidation, the catalysts were separated from the reaction mixture by filtration, thoroughly washed with water and CH₂Cl₂, and then reused in the next run under the same conditions. In comparison, the separation of soluble MWNTs–TEMPO from a reaction mixture is more straightforward than other soluble supports. The catalysts were recycled for six consecutive reactions without loss in catalytic activities (Table 3). Following the seventh run, the activity of the catalysts had decreased to a certain extent. The decrease may be due to either the gradual hydrolysis or decomposition of 7% of the total supported TEMPO according to the element analysis. This outcome suggested that MWNTs-supported TEMPO was stable enough to be recycled.

Table 3 Recovery and recycling of soluble MWNTs–TEMPO in the oxidation of benzyl alcohol^a

Run	Conv. (%)^b	Sel. (%)^b
a Except pH 8.6 and the reaction time (7 min, optimized time), the reaction conditions are the same as in Table 2. The catalyst was recovered by filtration and reused after washing twice both with water and CH2Cl2. b Conversions and selectivities are based on ^1H NMR.
1	100	>99
2	100	>99
3	100	>99
4	100	>99
5	100	>99
6	100	>99
7	90	>99

---

3. Conclusion

In summary, our findings suggest that the problems of difficult characterization and activity loss for the immobilization of homogeneous catalysts onto solid supports can be overcome by taking advantage of the functionalization of carbon nanotubes, which can be soluble in the reaction medium. The catalytic activity of the supported catalysts for homogeneous catalysis is far superior to heterogeneous catalysis, and these results confirmed that homogeneously dispersed MWNTs were a key factor for the effective catalytic performance. Besides that, this system also features easy recovery and solution-based characterization, and thereby opens up new prospects for combinatorial solution phase synthesis, as well as for the anchoring of recoverable reagents, scavenging agents or homogeneous catalysts. Extension of the application towards broad fields is currently underway.

4. Experimental section

4.1 Materials

The multiwalled carbon nanotubes (MWNTs) made from the chemical vapor deposition method were purchased from Chengdu Organic Chemicals Corporation. Tetrahydrofuran was distilled under N₂ atmosphere from sodium/benzophenone prior to use. All reagents were used as received, unless otherwise noted.

4.2 Techniques

¹H NMR spectra were recorded on a Varian INOVA-400 MHz type (¹H, 400 MHz) spectrometer. Their peak frequencies were referenced versus internal standard shifts (TMS) at 0 ppm for ¹H NMR and against the solvent. Fourier transform infrared spectroscopy (FT-IR) analysis was performed using a Nicolet NEXUS FT-IR spectrophotometer using a KBr pellet. The elemental analyses of the carbon nanotubes were carried out on a Vario El III, and the content of TEMPO was determined. Thermogravimetric analyses of all resulted nanotubes were measured on a Mettler-Toledo TGA/SDTA851e at a heating rate of 10 °C min⁻¹. The morphological structures of the samples were observed by means of transmission electron microscopy (TEM) (JEM-2000EX, JEOL Co., Japan). The samples for TEM observation were prepared by depositing a drop of the diluted solutions on carbon-coated copper grids. To avoid aggregation during the drying process, the TEM grids were lyophilized prior to the measurement. Raman spectra were obtained with an inVia microspectrometer (Renishaw) equipped with a He–Ne laser at 632.8 nm. UV-vis spectra from 190 to 1100 nm were collected from a Perkin-Elmer Lambda 35 UV-vis spectrophotometer. Background correction was performed using a cleaned aluminium substrate as a reflectance standard. XPS were conducted on a Kratos Amicus using Mg-Kα radiation (1486.6) at 180 W. The operating vacuum condition in the chamber was 1 × 10⁻⁶ Pa.

4.3 Synthesis of MWNTs–TEMPO

0.5 g of the crude MWNTs (1 g) were dispersed in concentrated HNO₃ solution (50.0 ml, 65%) in a 100 ml round bottom flask equipped with a condenser. The mixture was placed in an ultrasonic bath (40 kHz) for 30 min and then stirred for 24–48 h under reflux. The mixture was then vacuum-filtered through a Millipore polycarbonate membrane and subsequently washed with distilled water until the pH of the filtrate was ca. 7. The filtered solid was dried under vacuum overnight at 80 °C, yielding MWNTs–COOH. Then the generated MWNTs–COOH was suspended in SOCl₂ (50.0 ml) and stirred for 24 h at 65 °C. The solution was filtered, washed with anhydrous THF, and dried under vacuum at room temperature for 6 h, generating MWNT–COCl. MWNTs–COCl was mixed with 4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl (HO–TEMPO) in boiling THF (50.0 ml) and stirred for 48 h. The resulting solid was separated by vacuum-filtration using a Millipore polycarbonate membrane filter and subsequently washed with anhydrous THF. After repeated washing and filtration, the resulting solid was dried overnight in a vacuum, generating MWNTs–TEMPO.

4.4 General oxidation reaction conditions; recycling experiments

In a typical experiment, following Anelli's procedure, the oxidation was performed in a 25 ml flask at 0 °C. After MWNTs–TEMPO (0.8 μmol) was briefly sonicated to be soluble in the water (2.1 ml), aqueous NaOCl (0.63 ml) was added to dilute for 0.35 M and buffered by NaHCO₃ to adjust to pH 9.1 by combination with KBr (0.5 M, 0.16 ml). When the alcohol and dichloromethane (2.0 ml) were added into the reaction mixture, the oxidation reaction was carried out. The reaction was quenched with Na₂S₂O₃ solution (1 M, 1 ml). When the reaction finished, the catalyst was filtered with nylon membrane (0.22 μm). MWNTs–TEMPO was reused after washing twice both with water and CH₂Cl₂. The organic phase was isolated for GC analysis, and dried over sodium sulfate, with removal of the solvent to allow ¹H NMR analysis.

Acknowledgements

Gratitude is expressed to the National Natural Science Foundation of China (NSFC, Grant 21004007), the Scientific Research Foundation for Doctor of Liaoning Province (Grant 20101018). The authors also thank Prof. Xiao-Bing Lu for valuable discussion.

References

1. S. Bhaduri, D. Mukesh, Homogeneous Catalysis: Mechanisms and Industrial Applications, Wiley, New York, 2000, 1-11.
2. R. T. Baker, S. Kobayashi and W. Leitner, Adv. Synth. Catal., 2006, 348, 1337–1340.
3. D. J. Cole-Hamilton, Science, 2003, 299, 1702–1706.
4. J. A. Gladysz, Chem. Rev., 2002, 102, 3215–3216.
5. N. E. Leadbeater and M. Marco, Chem. Rev., 2002, 102, 3217–3274.
6. C. A. McNamara, M. J. Dixon and M. Bradley, Chem. Rev., 2002, 102, 3275–3300.
7. A. Corma and H. Garcia, Adv. Synth. Catal., 2006, 348, 1391–1412.
8. C. Li, Catal. Rev. Sci. Eng., 2004, 46, 419–492.
9. F. Marras, A. M. Kluwer, J. R. Siekierzycka, A. Vozza, A. M. Brouwer and J. N. H. Reek, Angew. Chem., Int. Ed., 2010, 49, 5480–5484.
10. W. Sebastian, A. Schatz, R. N. Grass, W. J. Stark and O. Reiser, Angew. Chem., Int. Ed., 2010, 49, 1867–1870.
11. K. E. Price, B. P. Mason, A. R. Bogdan, S. J. Broadwater, J. L. Steinbacher and D. T. McQuade, J. Am. Chem. Soc., 2006, 128, 10376–10377.
12. J. C. Love, L. A. Estroff, J. K. Kriebel, R. G. Nuzzo and G. M. Whitesides, Chem. Rev., 2005, 105, 1103–1170.
13. Z. Giovanni, M. Claudia, B. Renato, J. P. Leonard and S. Paolo, Chem.–Eur. J., 2011, 17, 4879–4889.
14. K. Hara, R. Akiyama, S. Takakusagi, K. Uosaki, T. Yoshino, H. Kagi and M. Sawamura, Angew. Chem., Int. Ed., 2008, 47, 5627–5630.
15. A. Schätz, O. Reiser and W. J. Stark, Chem.–Eur. J., 2010, 16, 8950–8967.
16. Q. M. Kainz, A. Schaetz, A. Zoepfl, W. J. Stark and O. Reiser, Chem. Mater., 2011, 23, 3606–3613.
17. A. Schaetz, M. Zeltner, T. D. Michl, M. Rossier, R. Fuhrer and W. J. Stark, Chem.–Eur. J., 2011, 17, 10566–10573.
18. B. Karimi and E. Farhangi, Chem.–Eur. J., 2011, 17, 6056–6060.
19. R. A. Sheldon, I. W. C. E. Arends, G. T. Brink and A. Dijksman, Acc. Chem. Res., 2002, 35, 774–781.
20. D. E. Bergbreiter, J. Tian and C. Hongfa, Chem. Rev., 2009, 109, 530–582.
21. R. T. Baker and W. Tumas, Science, 1999, 284, 1477–1479.
22. D. E. Bergbreiter and S. D. Sung, Adv. Synth. Catal., 2006, 348, 1352–1366.
23. C. Baleizão, B. Gigante, H. Garcia and A. Corma, J. Catal., 2004, 223, 106–113.
24. P. Serp, M. Corrias and P. Kalck, Appl. Catal., A, 2003, 253, 337–358.
25. M. Salavati-Niasari and M. Bazarganipour, J. Mol. Catal. A: Chem., 2007, 278, 173–180.
26. M. Salavati-Niasari, F. Davar and M. Bazarganipour, Dalton Trans., 2010, 39, 7330–7337.
27. S. Banerjee and S. S. Wong, J. Am. Chem. Soc., 2002, 124, 8940–8948.
28. N. Karousis and N. Tagmatarchis, Chem. Rev., 2010, 110, 5366–5397.
29. D. Tasis, N. Tagmatarchis, A. Bianco and M. Prato, Chem. Rev., 2006, 106, 1105–1036.
30. I. D. Rosca, F. Watari, M. Uo and T. Akasaka, Carbon, 2005, 43, 3124–3131 . The oxidation of multiwalled carbon nanotubes with high solubility (20–40 mg ml⁻¹) in water.
31. G. Ghini, L. Luconi, A. Rossin, C. Bianchini, G. Giambastiani, S. Cicchi, L. Lascialfari, A. Brandi and A. Giannasi, Chem. Commun., 2010, 46, 252–254 . This organic functionalization yields materials with a great solubility in DMF (close to 10 mg ml⁻¹ of DMF).
32. Nowadays, “soluble” and “dispersed” for functionalization of carbon nanotubes was found in a large amount papers, and even the terms were often interchanged. Most researchers reported that functionalized carbon nanotubes can be “soluble” in organic solvents or aqueous medium beyond 1 year, and are more stable than the colloidal systems. Hence, “soluble” for immobilization of carbon nanotubes with TEMPO was used in this paper.
33. P. L. Anelli, C. Biffi, F. Montanari and S. Quici, J. Org. Chem., 1987, 52, 2559–2562.
34. H. Hu, B. Zhao, M. E. Itkis and R. C. Haddon, J. Phys. Chem. B, 2003, 107, 13838–13842.
35. H. Peng, P. Reverdy, V. N. Khabashesku and J. L. Margrave, Chem. Commun., 2003, 362–363.
36. K. J. Ziegler, Z. Gu, H. Peng, E. L. Flor, R. H. Hauge and R. E. Smalley, J. Am. Chem. Soc., 2005, 127, 1541–1547.
37. Z. Chen, K. Kobashi, U. Rauwald, R. Booker, H. Fan, W. F. Hwang and J. M. Tour, J. Am. Chem. Soc., 2006, 128, 10568–10571.
38. M. N. Tchoul, W. T. Ford, G. Lolli, D. E. Resasco and S. Arepalli, Chem. Mater., 2007, 19, 5765–5772.
39. G. A. Forrest and A. J. Alexander, J. Phys. Chem. C, 2007, 111, 10792–10798.
40. J. Moskowitz, B. Shao, S. L. Deng, H. W. Liao, K. J. Gaskell and Y. H. Wang, J. Am. Chem. Soc., 2010, 132, 3932–3938.
41. A. Jorio, M. A. Pimenta, A. G. Souza Filho, R. Satio, G. Dresselhaus and M. S. Dresselhaus, New J. Phys., 2003, 5, 139.1.–139.17.
42. C. Gao, Y. Z. Jin, H. Kong, R. L. D. Whitby, S. F. A. Acquah, G. Y. Chen, H. Qian, A. Hartschuh, S. R. P. Silva, S. Henley, P. Fearon, H. W. Kroto and D. R. M. Walton, J. Phys. Chem. B, 2005, 109, 11925–11932.
43. V. A. Sinani, M. K. Gheith, A. A. Yaroslavov, A. A. Rakhnyanskaya, K. Sun, A. A. Mamedov, J. P. Wicksted and N. A. Kotov, J. Am. Chem. Soc., 2005, 127, 3463–3472.
44. K. S. Chan, X. Z. Li, W. I. Dzik and B. de Bruin, J. Am. Chem. Soc., 2008, 130, 2051–2061.
45. C. Bolm and T. Fey, Chem. Commun., 1999, 1795–1796.
46. T. Fey, H. Fischer, S. Bachmann, K. Albert and C. Bolm, J. Org. Chem., 2001, 66, 8154–8159.
47. A. Dijksman, I. W. C. E. Arends and R. A. Sheldon, Synlett, 2001, 1, 102–104.
48. A. Schätz, R. N. Gras, W. J. Stark and O. Reiser, Chem.–Eur. J., 2008, 14, 8262–8266.
49. A. K. Tucker-Schwartz and R. L. Garrell, Chem.–Eur. J., 2010, 16, 12718–12726.
50. M. Gilhespy, M. Lok and X. Baucherel, Catal. Today, 2006, 117, 114–119.
51. A. Gheorghe, A. Matsuno and O. Reiser, Adv. Synth. Catal., 2006, 348, 1016–1020.
52. Y. Chen, X. Zhang and S. Mitra, Chem. Commun., 2010, 46, 1652–1654.

---

Footnotes

† Electronic supplementary information (ESI) available: ¹H NMR spectra of the compounds produced and XPS N1s spectra for MWNTs–TEMPO. See DOI: 10.1039/c2ra21206d/

‡ Y.M.W conceived the experiments and wrote the manuscript, X.Y.S., S.H.S., F.L. and Y.M.W. performed the experiments, X.Y.S. and S.H.S contributed equally to this work. All authors discussed the results and commented on the manuscript.

---