Solvent-free one-step covalent functionalization of graphene oxide and nanodiamond with amines

Abstract

We attempted amide functionalization of the carboxylic groups present in graphene oxide (GO) and nanodiamond (ND) by means of a solvent-free gas-phase treatment with two aromatic amines 1-aminopyrene (AP) and 2-aminofluorene (AF), and also with one aliphatic amine, 1-octadecylamine (ODA), for comparison. The procedure was carried out under moderate vacuum, at temperatures of 150–180 °C, using short reaction times of about 2 h. The amine treatment generally gave rise to changes in Fourier-transform infrared (FTIR), Raman, ultraviolet-visible, and X-ray photoelectron spectra (XPS), to a variable degree depending on the particular sample. Thermogravimetric analysis showed the highest amine content for GO and ND treated with ODA, which turned out to be roughly one order of magnitude higher than in the case of aromatic AF and AP. Considerable changes in sample morphology after functionalization were observed by transmission electron and atomic force microscopy. Based on the analysis of FTIR and XPS spectra, we concluded that amidation is the only possible route for ND functionalization with AF, AP and ODA, whereas we found spectroscopic evidence for an alternative reaction channel on GO sheets, which is amine addition onto epoxy bridges.

---

1. Introduction

The exploration of the covalent chemistry of both graphene oxide (GO) and nanodiamond (ND) is an important step in developing new carbon nanohybrid materials with novel properties and improved performance related to the design of drug delivery systems, field emission displays, composite materials, sensors, biosensors, nanocatalysts, etc.^1–4 GO is formed of stacked graphene-derived sheets with multiple oxygen-containing functionalities (e.g., hydroxyl, epoxide, carbonyl and carboxyl groups) distributed throughout the sheet edges and surfaces: this fact serves as an important prerequisite for successful chemical modification of graphene oxide.⁵

As regards ND, it exhibits a complex structure based on the diamond core made up of sp³-hybridized carbon atoms, a fullerene-like shell of sp² carbon atoms, and an outer surface where the C atoms form functional groups with hydrogen and oxygen atoms. The most common groups identified on the surface of detonation (hereafter pristine) ND are oxygen-containing groups, such as carboxyl, lactone and carbonyl functionalities.³

Due to the presence of similar surface functionalities in these two structurally different materials, similar strategies for their chemical functionalization can be applied. Among all possible functionalizing chemical species one should emphasize amine compounds, since the latter can facilitate a number of biomedical applications, due to strong electrostatic interactions between amine groups and biological components, capable of providing stable immobilization of biological compounds onto the carbon nanomaterial surfaces.

Therefore, it is not surprising that the search for convenient functionalization routes remain to be the subject of numerous studies focused on the preparation and applications of amino-functionalized derivatives of carbon nanomaterials, including GO and ND (e.g. (ref. 6–14) and references therein). The most common and widely used way is based on the reaction of carboxyls with thionyl chloride (SOCl₂) in order to generate acyl chloride groups C(=O)Cl on GO or ND, capable of reacting with amine compounds with the formation of amide group; duration of the whole procedure can be up to several days. As a typical example, the preparation of hydrophobic ND covalently functionalized with octadecylamine (ODA) was reported by Gogotsi et al.⁹ by means of refluxing ND (previously purified in acidic conditions) with excess SOCl₂ in N,N-dimethylformamide at 70 °C for 24 h, followed by stirring with ODA at 90–100 °C for 96 h. The entire protocol included the separating, washing and drying of solid acyl chloride derivative of ND, followed by the reaction with ODA. The excess of ODA was removed first by sonication-assisted washing with anhydrous methanol for 4–5 times, then by further purification of the functionalization product by 10-fold extraction with hot methanol in a Soxhlet apparatus. Purified ND–ODA samples obtained were dispersible in chloroform, and exhibited fluorescent properties potentially useful for optical applications. Similar synthetic steps were followed by Shanmugharaj et al.¹⁰ when preparing decylamine-, hexadecylamine-, and octadecylamine-grafted GO samples. Additional thermal treatment of the powders obtained resulted in the formation of superhydrophobic surfaces. Interestingly, alkylamines were found to chemically react with GO surface both through amidation reaction with chemically activated COOH functionalities and through nucleophilic substitution reactions with epoxy groups present in GO along with carboxyls. The effect of chain length of alkylamines employed on the superhydrophobic wetting properties of GO samples was investigated.

Other examples of liquid-phase preparation of similar functionalized GO and ND materials to mention are the report by Chang et al.,¹¹ who modified ND surface by grafting oligomers with COOH, NH₂ and aliphatic moieties by means of ultrasonication and microwave-initiated free radical copolymerization, as well as the study by Navaee and Salimi,¹² who synthesized aminated GO through the Bucherer reaction between ammonia and graphene oxide, catalyzed by sodium bisulfite. The reaction was performed in an autoclave tube at 170 °C for about 10 h. The resulting material was washed via centrifugation and re-dispersion in fresh water 3 times to remove catalyst and non-reacted ammonia. The samples obtained displayed superior electrocatalytic activity towards the oxygen reduction reactions. In a recent report by Saptal and coworkers,¹³ an amine-functionalized GO hybrid was obtained via the use of 3-aminopropyltrimethoxysilane (APTMS), and the reaction was carried out by pouring APTMS into ethanol containing GO under sonication for 30 min and then refluxing for 6 h at 80 °C. The functionalization procedure was followed by filtration, washing and drying of resulting amine-functionalized GO under vacuum overnight. The catalytic performance of obtained samples was investigated for the chemical fixation of CO₂.

Thus, the conventional routes of GO and ND functionalization with amines are usually carried out in an organic solvent medium, and therefore such auxiliary procedures as ultrasonication, filtration, washing and drying are inevitable, which dramatically raises the material cost, not to mention bringing on the generation of excessive chemical wastes.

We proposed an alternative one-step solvent-free approach for the amidation of COOH groups present on oxidized carbon nanomaterials with amine molecules of variable structure.^14,15 Its distinctive feature is that the formation of amide bonds takes place due to thermal activation at 150–180 °C, as successfully tested for the functionalization of oxidized carbon nanotubes and ND with several aliphatic amines, and thus the synthetic procedure does not require any additional chemical activation (for example, with toxic SOCl₂). The elevated temperature serves not only to make the amidation possible but additionally to remove unreacted amine from the nanomaterial, which is a very important advantage from the point of view of possible biomedical and electronic applications of resulting functionalized materials.¹⁶ Ecological and economical benefits cannot be overestimated either.

The main goal of the present study was to explore the efficiency of solvent-free amidation of GO with one aliphatic and two aromatic amines, which are 1-octadecylamine (ODA), 2-aminofluorene (AF), 1-aminopyrene (AP) (Fig. 1a–c), by carrying out detailed comparative characterization of spectroscopic, microscopic, thermal and dispersibility properties of the nanohybrids obtained. In addition, we applied the same derivatization procedure to ND, which is another carbon nanomaterial bearing COOH functionalities. The latter was done mainly for the sake of comparison with GO; nevertheless, amide functionalization of ND represents independent interest for tuning its physicochemical properties (dispersibility in different solvent media, affinity to organic polymers and biomolecules, etc.).

Fig. 1 Chemical structures of 2-aminofluorene (a), 1-aminopyrene (b) and 1-octadecylamine (c); suggested general schemes for the amidation of COOH groups of GO (d) and ND (e). R = aryl or alkyl.

2. Experimental

2.1. Materials

ND powder (98% purity; average particle size of 6 nm and density of 3.5 g mL⁻¹ at 25 °C; oxygen content of 6–7 wt%) from Nanostructured and Amorphous Materials, Inc., and GO powder (>99 wt% purity; sheet size of 300–800 nm and thickness of 0.7–1.2 nm; oxygen content of 45–55 wt%) from CheapTubes, Inc., were used. For their covalent functionalization, 1-octadecylamine (CH₃(CH₂)₁₇NH₂), 1-aminopyrene (C₁₆H₁₁N) and 2-aminofluorene (C₁₃H₁₁N) from Sigma-Aldrich were employed as received. Isopropanol from Femont (analytical grade) was used for dispersibility tests.

2.2. Solvent-free functionalization

For the covalent functionalization of ND and GO we used an experimental setup and optimized reaction conditions described earlier (ref. 14–16 and references therein). The functionalization procedure was carried out in a Pyrex glass reactor, in which graphene oxide or nanodiamond powder was placed together with amine reagent, at 1 : 1 w/w ratio. Before starting the reaction, the components mixed were degassed for 1 h at 100 °C under constant evacuation at about 10⁻² Torr. After that, the reactor bottom containing amine and carbon nanomaterial was heated at 180 °C with constant evacuation during 2 h, for the functionalization with AP and AF, and at 150 °C under static vacuum for 2 h, in the case of functionalization with ODA. (Increasing the temperature by 10–20 °C and treatment duration by several hours did not cause any significant changes in spectral characteristics and/or morphology of the functionalized materials.) After completing the reaction, the excess of amines (which condensed above the reaction zone) was removed by additional constant evacuation during 1 h and heating at 150 °C. The functionalized samples obtained hereafter are referred to as GO+AP, GO+AF, GO+ODA, ND+AP, ND+AF and ND+ODA.

2.3. Characterization

Fourier-transform infrared (FTIR) spectra were acquired using a Nexus 670 FTIR Thermo-Nicolet instrument, equipped with an Olympus BX52 microscope, under room temperature and atmospheric pressure; the samples were deposited onto ZnSe windows. Raman spectra were recorded on a Thermo-Nicolet Almega Dispersive Raman instrument (λ = 532 nm). Ultraviolet-visible (UV-vis) spectra were recorded in a range of 200–800 nm using a Cary 5000 UV-Vis-NIR spectrophotometer from Agilent; the samples were prepared by dispersing/dissolving in ethanol.

For X-ray photoelectron spectroscopy (XPS) studies, we used a SPECS GmbH custom made X-ray photoelectron spectrometer microprobe, equipped with a PHOIBOS 150 WAL hemispherical analyzer and a monochromated Al Kα X-ray source (μ-FOCUS 500) with energy of 1486.6 eV. XPS survey spectra were obtained for a wide binding energy range with a 1 eV step size, while high-resolution energy regions with a range of 30 eV were selected for all elements of interest (N 1s, C 1s and O 1s) using a 0.1 eV step size, dwell time of 0.2 s. Spectra are presented without smoothing. Charge referencing was done against adventitious carbon by setting the C 1s peak maximum at 284.7 eV.

Thermogravimetric analysis (TGA) curves were acquired by using a SDT-Q600 analyzer from TA Instruments, under air flow of 100 mL min⁻¹ and with a heating ramp of 10 °C min⁻¹.

For scanning electron microscopy (SEM) characterization of dry samples we used a JEOL JSM-6510LV instrument operating in low voltage mode at 5 kV. For transmission electron microscopy (TEM) observations, a JEOL 4000EX microscope was employed, operating at 200 kV. Atomic force microscopy (AFM) images were acquired using a JEOL JSPM-5200 instrument in tapping mode, for the samples deposited onto silicon wafers from isopropanol dispersions.

3. Results and discussion

3.1. Amine functionalization of graphene oxide

With the purpose of detecting changes in chemical structure of GO samples after the gas-phase amidation according to scheme shown in Fig. 1d, we carried out simple dispersibility tests in water and isopropanol for the samples before and after treatment with amines. As shown in Fig. 2, GO is well dispersed in both solvents giving brown solutions. The good GO dispersibility in water is due to a large number of oxygenated groups present on GO sheets. The aqueous dispersion of GO remains stable after more than 48 h. In contrast, all three functionalized samples (GO+AF, GO+AP and GO+ODA) completely precipitated in water; this behavior can be attributed to the introduction of hydrophobic groups, namely, long hydrocarbon chains in the case of GO+ODA and aromatic rings for GO+AF and GO+AP. On the other hand, all the samples formed equally homogeneous dispersions in isopropanol under ultrasonic bath treatment. After 24 h, non-functionalized GO precipitated considerably and GO+ODA sample completely precipitated, which can be explained by different phenomena. Due to a high content of oxygen-containing functionalities, GO has a better affinity to water than to isopropanol. After the treatment with ODA, a large number of highly hydrophobic, long and flexible octadecyl radicals appear on GO surface, which dramatically decrease the affinity of graphene sheets to the polar solvent. At the same time, they increase the possibility of sticking the sheets together due to hydrophobic interactions of ODA radicals belonging to adjacent sheets, resulting in the strong agglomeration of GO+ODA observed. In the case of GO+AF and GO+AP, the hydrophobic aromatic substituents are much shorter than C₁₈H₃₇ chains and therefore cannot cause such a strong agglomeration. Complete sedimentation of the latter two samples was observed only after two days.

Fig. 2 Comparative dispersibility test (1 mg of each sample per 2 mL of solvent) in water and isopropanol for GO samples before and after amine functionalization. Images were taken at elapsed time of 0 and 24 h after ultrasonic bath treatment for 5 min.

By using several spectroscopic techniques, we looked for more detailed information about the composition of GO samples. Comparison of their Raman spectra is presented in Fig. S1 (ESI†). Before treatment with amines, GO displayed typical spectral features, such as a prominent D band at 1338 cm⁻¹, attributed to a defect-induced breathing mode of sp² backbone, and a G band at 1575 cm⁻¹, corresponding to the first-order scattering of E_2g mode.¹⁷ In general, Raman spectra of all functionalized GO samples were featureless. However, a slight shift of G band from 1575 to 1555, 1565 and 1555 cm⁻¹ for GO+AF, GO+AP and GO+ODA, respectively, was observed, which can be attributed to the introduction of amine functional groups onto GO sheets.¹⁸ The ratio of intensity for D and G bands (I_D/I_G), which serves as a measure of defects present on graphene, increased only slightly from 0.91 for GO before functionalization to 0.95 for all amine-treated samples, indicating that amidation did not introduce more defects, consequently preserving basic GO structural properties. Weak and broad 2D and D + D′ overtone bands can be observed in the spectrum of pristine GO at 2655 and 2900 cm⁻¹, respectively, whereas become less distinguishable for the functionalized samples.

Dramatic changes due to GO amide functionalization were observed in FTIR spectra of GO samples (Fig. 3). GO before amine treatment displayed typical peaks corresponding to a variety of oxygen-containing functional groups, including a strong absorption band at 1726 cm⁻¹ due to C=O stretching vibrations of carboxyls, O–H bending absorption at 1628 cm⁻¹, a strong and broad band due to epoxide C–O–C bonds at 1227 cm⁻¹, and a broad band due to O–H stretching vibrations in carboxyl and alcohol functionalities around 3338 cm⁻¹ (with a possible contribution of water molecules adsorbed on the hydrophilic surface of GO).^12,19 The most important change observed after amine treatment is a strong decrease in intensity of the ν_C=O band of COOH groups along with appearance of strong absorption at 1570–1580 cm⁻¹. The formation of amide derivatives according to the scheme suggested in Fig. 1d would logically lead to the formation of two partially overlapping bands ‘amide I’ (ν_C=O) and ‘amide II’ (δ_NH) usually observed at 1600–1660 and 1520–1580 cm⁻¹. The fact that absorption in this region is unresolved and centered at 1570–1580 cm⁻¹ can be explained by the presence of large amounts of non-amide NH groups, whose δ_NH vibrations mask the ‘amide I’ component. A possible explanation is that AF, AP and ODA covalently bond not only to COOH groups, but also to epoxide functionalities, which are abundant in GO samples. Other changes due to the introduction of new functional groups onto GO sheets include C–H stretching modes for GO + ODA and GO + AP at 2847, 2927 (aliphatic) and 2900–3100 cm⁻¹ (aromatic), respectively. In the case of GO + AF and GO + AP, new strong signals were found within the region of 700–1300 cm⁻¹ which are characteristic for C–C stretching vibrations in the aromatic rings of AF and AP moieties. For GO + ODA, one should also mention strong CH₂ scissor mode at 1454 cm⁻¹.

Fig. 3 FTIR spectra of GO before and after functionalization with amines.

Thermogravimetric analysis was employed to characterize the degree of graphene oxide functionalization with amines and to reveal possible changes in thermal stability of GO (Fig. 4). TGA curve of GO before amine treatment exhibits three steps of weight loss, similarly to related TGA data reported by other groups.^10,20 Here, the first weight loss of ca. 8% occurs until about 120 °C due to physisorbed water molecules. The second weight loss to about 270 °C is of ca. 20% and corresponds to the pyrolysis of oxygen-containing groups of GO surface, where decarboxylation processes most likely dominate. Finally, the third, major weight loss of roughly 70% corresponds to the combustion of graphene lattice which ends at about 630 °C. The reaction of GO with amines gave rise to considerable changes in the shape of TGA curves. Unlike pristine GO, none of the amine-treated samples displayed the low-temperature weight loss until ca. 120 °C associated with adsorbed water molecules, which is consistent with the conclusion on enhanced hydrophobicity due to functionalization with AP, AF and ODA based on the dispersibility test. The content of amine species covalently bonded to GO surface was roughly estimated from the higher-temperature weight loss between approximately 200 and 400 °C, and it was found to be about 5%, 12% and 33%, for GO + AF, GO + AP and GO + ODA, respectively. In other words, the highest efficiency of functionalization is observed for linear aliphatic ODA, followed by aromatic species AP and AF.

Fig. 4 TGA curves for pristine GO, in comparison with the curves for amine-functionalized samples.

In addition, TGA results are indicative of some increase in thermal stability of GO backbone in functionalized samples as compared to pristine GO: the complete combustion of GO + AF, GO + AP and GO + ODA is observed at approximately 710, 730 and 680 °C, respectively, versus 630 °C for non-functionalized GO.

UV-visible absorption spectra (Fig. 5) reflect the appearance of new chemical groups due to functionalization as well. The spectrum of GO exhibits an absorption band at 290 nm attributable to n → π* transitions in C=O groups.²¹ For all amine-functionalized GO samples, the consistently higher absorption can be observed in the near UV region, as well as partially in the visible spectrum for GO + AP. The n → π* transition in C=O groups shows notable hypsochromic shift to 275–277 nm for GO + AF and GO + ODA, whereas for GO + AP it overlaps with a strong absorption in the region between 220 and 420 nm due to π → π* transitions in fused aromatic system of pyrene.

Fig. 5 UV-visible spectra of GO before and after functionalization with amines.

The chemical nature of GO samples was in more detail explored by means of XPS analysis; the results are presented in Fig. 6. As it was expected, the C 1s and O 1s peaks were found in both pristine and functionalized GO samples, whereas the N 1s signal around 400 eV appeared only after amine treatment. The most dramatic changes due to functionalization were observed for C 1s peak. Its deconvolution for pristine GO produced typical components with the binding energies of 289.2 (O–C=O of carboxyls), 286.9 (epoxy groups) and 284.7 eV (sp²/sp³ C–C bonds).^1,22 While the latter remained to be most intense for amine-treated GO samples, the carboxyl component became hardly detectable in the spectra of functionalized GO samples. Furthermore, the intensity of epoxide component drastically decreased as well, by several times, which makes us admit that amine addition onto epoxy groups might be a major route of attempted covalent functionalization of GO. This admission is further supported by the fact that a new peak characteristic for C–N bonds in amides, expected to appear at binding energies above 288.2 eV,²³ can be observed only in the C 1s spectrum of GO + ODA, for which the functionalization degree is much higher than for GO + AF and GO + AP. What is obvious, on the other hand, is that the spectra of all functionalized samples exhibit a new component at 285.6 eV due to C–N bonds, presumably in amine moieties added onto epoxy groups.

Fig. 6 Comparison of XPS spectra for GO samples before and after amine functionalization: survey spectra (top row) along with deconvolution of the corresponding C 1s, N 1s and O 1s lines. Raw spectra are shown in black, and sum, in red.

Nevertheless, the formation of amide bonds according to the scheme shown in Fig. 1d unambiguously follows from the deconvoluted N 1s spectra, where the component with a binding energy of 400.5–401.1 eV was found for each functionalized GO sample. Equally important is the detection of peaks at 399.2–399.6 eV attributable to secondary amine moieties resulting from the addition onto epoxy groups. The third component is evident in the spectrum of GO + AF only (at 402.0 eV) and almost invisible for GO + AP and GO + ODA: it was assigned to protonated NH₂ groups²³ forming salts with COOH functionalities instead of forming amide bonds.

We also attempted to analyzed O 1s peaks, but in a more simplistic way. As fairly noted Yang et al.,¹⁷ many assignments of the “carbon–oxygen” functional groups found in the literature are conflicting. There are different reasons for that. The first one is the fact that O 1s photoelectron kinetic energies are lower than those of the C 1s, the O 1s sampling depth is smaller, and therefore the O 1s spectra are more surface specific. The second is a large variety of oxygen-containing groups which contribute to the O 1s peak. The possibility of contamination with atmospheric oxygen should be mentioned as well. Therefore, we did not attempt to perform deconvolution of closely positioned components in the region between 531 and 533 eV, leaving unresolved the main contribution centered at about 532 eV due to the most characteristic bonds including C=O and C–O in carboxylic and epoxy groups.^17,24 Instead we were looking for changes appearing after treating GO with amines, which would be outside the range of 531–533 eV. The evident change we found is the appearance of a new component at 533.5–533.7 eV due to oxygen atoms in hydroxyls,^17,25 which form when epoxy rings open upon amine addition reaction. Its contribution into the O 1s peak is notably larger in the case of GO + AP and GO + ODA as compared to GO + AF; this correlates with the higher degree of functionalization with AP and ODA found from TGA measurements (Fig. 4). On the other hand, the signal due to O atoms in amide groups would fall into the region of 531–533 eV and is impossible to reliably discriminate. Finally, one should mention the deconvoluted peak at 529.8 eV apparently due to quinones, which was detected in the O 1s spectra of pristine GO and GO + ODA. As a whole, XPS results turned to be in a good agreement with those obtained by means of FTIR spectroscopy.

For comparative morphological characterization of pristine and functionalized GO samples we employed several microscopic techniques, namely, conventional SEM, TEM and AFM. SEM images presented in Fig. 7a and d for GO before functionalization show a typical morphology of graphene oxide aggregates, similarly to those reported elsewhere (see, for example (ref. 26–28)). It is a disordered array of sheets (monolayers), where it is possible to distinguish the edges of individual sheets (some of them with rolled edges), as well as corrugated areas. Amine-treated GO samples (Fig. 7c–h) acquire new morphological features. In all three cases, GO exhibit the same ‘spongy’ morphology, however the agglomerates appear to be more compact and dense, with more rolled individual sheets in which the edges became more difficult to distinguish. We attribute the above morphological changes to a distortion of GO sheets due to amine attachment onto their edges (where COOH groups prevail) and surfaces (where epoxy groups dominate). To observe the morphological changes due to functionalization in more detail, we used TEM technique. From Fig. 8a and b one can see that pristine GO monolayers exhibit relatively smooth edges and do not tend to rolling.²⁹ After functionalization with AP and ODA (Fig. 8e–h), the individual graphene sheets acquired a strongly irregular shape, with numerous wrinkles and corrugations, as well as showed a stronger trend to agglomeration. Similar changes in the case of GO + AF were visible but less significant (Fig. 8c and d) than for GO + AP and GO + ODA, which correlates with the lower organic content found from TGA measurements (Fig. 4).

Fig. 7 Representative SEM images at different magnification for GO samples before (a, b) and after functionalization with AF (c, d), AP (e, f) and ODA (g, h). Scale bar: (a, c, e, g) 10 μm; (b, d, f, h) 1 μm.

Fig. 8 Representative TEM images at different magnification for GO samples before (a, b) and after functionalization with AF (c, d), AP (e, f) and ODA (g, h).

According to AFM imaging, the principal difference between GO before (Fig. 9a and b) and after functionalization (Fig. 9c–h) is the degree of sheet agglomeration and rolling, which is in line with both SEM and TEM results. In some images (Fig. 9d–g) one can also observe the appearance of grainy texture, similarly to that previously found by AFM in PEGylated GO³⁰ as well as in graphene oxide coordination-functionalized with nickel(II) tetraazamacrocyclic complexes.²⁸ As mentioned above, the above morphological changes can be explained by the distortion of individual GO sheets due to amine attachment onto their edges (amidation of COOH functionalities) and surfaces (addition onto epoxy groups). Furthermore, the factor directly causing agglomeration can be the interaction between functionalizing amine moieties attached to adjacent sheets, which is π–π stacking between aromatic ring systems in the case of AF and AP, and hydrophobic interaction between long hydrocarbon chains in the case of ODA.

Fig. 9 Representative AFM topography images at different magnification for GO samples before (a, b) and after functionalization with AF (c, d), AP (e, f) and ODA (g, h). z-Scale in nm.

3.2. Amine functionalization of nanodiamond

After testing the one-step solvent-free amine functionalization of GO, which is a two-dimensional carbon nanomaterial, we attempted direct amidation with the same amine compounds of pristine ND, which is a three-dimensional material, according to the general scheme shown in Fig. 1e. Besides the dimensionality, an important difference between these two carbon form is that epoxy bridges are not found among most abundant groups on ND,³¹ contrary to GO.

Comparative dispersibility tests for pristine and functionalized ND samples were carried out in water and isopropanol (Fig. 10). Pristine ND was very well dispersed in both solvents, however, its solution in isopropanol was less stable showing notable precipitation after 24 h. The functionalization with ODA turned pristine ND into a material which was impossible to disperse in water, meanwhile ND + AF dispersion showed signs of precipitation after almost one day. ND + AP is an intermediate case, which was poorly dispersed from the very beginning and precipitated completely after 24 h. The behavior observed can be explained by introducing highly hydrophobic amine species into the surface layer of ND particles; the better (as compared to ND + ODA and ND + AP) dispersibility of ND + AF can serve as an indication of a lower degree of functionalization. At the same time, pristine and all chemically modified ND samples formed homogeneous dispersions in isopropanol upon ultrasonication, and partially precipitated after 24 h. In the case of ND + ODA, the dispersion was somewhat more stable as compared to aromatic amine-treated NDs, which is likely due to the fact that aliphatic octadecyl chains have a higher affinity to aliphatic alcohol than aromatic substituents do. In addition, this behavior can be indicative of a higher degree of functionalization with ODA.

Fig. 10 Comparative dispersibility test (1 mg of each sample per 2 mL of solvent) in water and isopropanol for ND samples before and after amine functionalization. Images were taken at elapsed time of 0 and 24 h after ultrasonic bath treatment for 5 min.

Raman spectra of ND samples are expected to display three characteristic peaks around 1320 cm⁻¹ attributed to sp³-hybridized carbon atoms of diamond lattice, at 1590 cm⁻¹ resulting from sp² C atoms of the shell, and the peak at 2930 cm⁻¹ due to C–H bonds.^14,32 No bands are observed in the spectra of pristine ND and ND + ODA (Fig. S2 of ESI†), which would have higher intensity than the above three features, whereas ND samples functionalized with AF and AP exhibit a series of much stronger signals between 1000 and 1700 cm⁻¹ corresponding to the aromatic groups. Similar Raman spectral features were described in the case of amine functionalization of buckypaper fabricated from acid-oxidized carbon nanotubes³³ and of pristine multi-walled carbon nanotubes.³⁴

Comparison of FTIR spectra of ND before and after amine treatment are presented in Fig. 11. Pristine ND shows a broad O–H stretching absorption around 3400 cm⁻¹ and δ_O–H bending mode at 1639 cm⁻¹ due to adsorbed water molecules. The most intense band at 1098 cm⁻¹ and its higher frequency shoulders can be assigned to ν_C–O vibrations in hydrogen bonded alcohol and COOH groups, respectively. Low-intensity ν_C–H absorption due to C–H bonds present on pristine ND surface can also be seen around 2900 cm⁻¹.^14,35 The spectral feature at 1726 cm⁻¹ is characteristic of H-bonded carboxyls and correspond to ν_C=O vibrations. It does not disappear completely after amine treatment: one can suggest that the amidation reaction apparently takes place mainly on external surfaces of ND agglomerates, since amine molecules cannot easily penetrate into the interstitial spaces between nanodiamond particles. At the same time, a number of new absorption bands appear due to the presence of new organic species. For ND + AF and ND + AP, a series of bands below 1300 cm⁻¹ are characteristic for C–C stretching vibrations in the aromatic rings,³⁴ whereas the band at 717 cm⁻¹ in the spectrum of ND + ODA corresponds to CH₃ rocking vibrations. The bands at 1410–1459 cm⁻¹ are due to δ_C–H vibrations in hydrocarbon radicals (for example, CH₂ scissor mode at 1459 cm⁻¹ for ND + ODA). The latter are complemented by ν_C–H modes in the region of 2800–3200 cm⁻¹ (aliphatic and aromatic) in AF, 2900–3200 cm⁻¹ (aromatic) in AP, at 2847 and 2920 cm⁻¹ (symmetric and asymmetric aliphatic ν_C–H, respectively) in ODA moieties. Since adsorbed water was mostly removed during functionalization, the broad ν_OH absorption around 3400 cm⁻¹ was substituted by narrower ν_NH bands at 3445 cm⁻¹ especially clearly seen for ND + AF and ND + AP. For the same reason, the spectral features found between 1500 and 1650 cm⁻¹ now corresponded to newly formed amide derivatives, where ‘amide I’ (ν_C=O) and ‘amide II’ (δ_NH) components cannot be resolved.

Fig. 11 FTIR spectra of ND before and after functionalization with amines.

The degree of covalent functionalization was roughly estimated from TGA curves shown in Fig. 12. Pristine ND behaved in a typical way^14,36 losing about 3–4% before 120 °C due to removal of adsorbed water, with a main weight loss between 500 and 660 °C due to decomposition of carbon network. None of functionalized samples exhibited the lower temperature weight loss due to adsorbed water, but instead showed decomposition of covalently-bonded amine species at 230–450 °C.^{14–16,34,37} Its content, however, strongly depends on whether amine is aliphatic (ODA) or aromatic (AF and AP).³⁴ For ND + ODA, the contribution of organics was found to be very high, of more than 30%, whereas it barely reached 3–4% in the case of ND + AF and ND + AP. The final combustion of all functionalized samples occurs at slightly lower temperatures with respect to pristine ND, before 550 °C. The data obtained are in agreement with our previous results on a higher reactivity of aliphatic amines as compared to aromatic ones in the amidation reactions of ND,^14,33 as well as with the results presented in this paper for GO (Fig. 4).

Fig. 12 TGA curves for pristine and amine-functionalized ND samples.

Bearing in mind that the oxygen content in ND and GO is 6–7% and 45–55%, respectively (see Experimental section), a question arises why the degree of functionalization with octadecylamine for GO + ODA is insignificantly higher as compared to that for ND + ODA? At least, this could be expected for functionalization in solution, where GO is well dispersed, and amines can graft onto individual GO sheets. However, in the present case of solvent-free functionalization, the sheets of dry GO form stacks, and the contact between them is much closer than between quasi-spherical primary ND particles. This makes large ODA molecules more difficult to access reactive groups within GO stacks than in ND agglomerates. AF and AP molecules are more compact, and for them the functionalization degree is notably higher for GO than for ND.

Thus, according to TGA, the degree of GO and ND functionalization with ODA turned to be very similar, of 33 and 30%, respectively. In view of this fact, another question arises: why the stability of GO + ODA (Fig. 2) and ND + ODA (Fig. 10) dispersions in isopropanol differs so significantly? As we suggested in Section 3.1., after the treatment with ODA, a large number of highly hydrophobic, long and flexible octadecyl radicals appear on GO surface, which dramatically decrease the affinity of graphene sheets to the polar solvent. At the same time, they increase the possibility of sticking the sheets together due to hydrophobic interactions of ODA radicals belonging to adjacent sheets, resulting in the strong agglomeration of GO + ODA observed (Fig. 2). This effect was discussed in a recent report by Mungse et al.,³⁸ who studied grafting n-alkylamines with different chain length (including ODA) onto GO under reflux in ethanol. A major fraction of n-alkylamines covalently interacted with the epoxy groups via nucleophilic substitution reaction (like it was observed in the present study), whereas others formed H-bonds with OH groups and salts with acidic functionalities of GO. Regardless of the type of attachment, the terminal CH₃ and CH₂ moieties of the n-alkylamines grafted onto GO were interdigitated with the counter layer and afforded a double-layer structure of alkyl chain-functionalized GO. In the case of ND functionalization with ODA, the same phenomenon can be expected, in principle. Nevertheless, the average size of primary ND particles (about 6 nm) is incomparably smaller than that of GO sheets (10² to 10³ nm), in addition to a spherical shape of the former vs. a quasi-flat morphology of the latter: these two circumstances drastically reduce the interaction between octadecyl radicals attached to adjacent ND particles, and, correspondingly, the degree of ND + ODA agglomeration in isopropanol (Fig. 10).

Changes in the UV-visible spectra for functionalized ND samples (Fig. 13) are very similar to those found for GO (Fig. 5). Pristine ND exhibited a weak n → π* transition due to C=O groups at 280 nm.³⁹ All amine-functionalized samples had stronger absorbance in the entire UV and visible regions studied, with the n → π* transition at approximately the same wavelength (277 nm). For nanodiamond functionalized with aromatic amines, new absorption bands can be observed in the range of 240–410 nm for ND + AP and at 319 nm for ND + AF, which are associated with π → π* transitions in aromatic systems of AF and AP.

Fig. 13 UV-visible spectra for pristine and amine-functionalized ND samples.

XPS spectra presented in Fig. 14 are very difficult to interpret unambiguously, especially for N 1s peaks. In this particular case, the reason is that pristine ND already contains nitrogen impurities, which overlap with signals expected to result from functionalization. There are three components in all C 1s spectra, which can be assigned to O–C=O of carboxyls (287.7–288.8 eV), sp² C–C bonds of the shell (284.7 eV), with a dominating component at about 286–287 eV due to sp³ bonded carbon of the main framework. The changes observed due to functionalization are a decrease in intensity for the carboxyl peak, an increase in the intensity of sp² component for ND + AP (where amine moieties do not contain sp³ C atoms) as well as of sp³ component for ND + AF and ND + ODA (where sp³ carbons are present in both ODA and AF). The contribution of C–N bonds in amides formed can be expected to appear at binding energies above 288.2 eV,²³ which would be problematic to reliably discriminate from the carboxyl component. Also, similarly to the case of graphene oxide, we can give only a very general assignment for the components of O 1s peaks, for the same reasons as explained in Section 3.1. What is obvious here is that the lowest-energy component around 532 eV is due to C=O bonds, whereas the one with binding energy of 533.4–534.2 eV likely corresponds to OH groups of the nanodiamond shell. One could suggest that the component at 534.4–535.3 eV can be due to adsorbed water: indeed, it strongly decreases for ND + AF and ND + AP. But then it is difficult to explain its high intensity in the case ND + ODA, which is most hydrophobic of all the samples.

Fig. 14 Comparison of XPS spectra for ND samples before and after amine functionalization: survey spectra (top row) along with deconvolution of the corresponding C 1s, N 1s and O 1s lines. Raw spectra are shown in black, and sum, in red.

Morphology of ND agglomerates, as observed by SEM (Fig. 15), remains without significant changes after functionalization. As usually,¹⁴ in addition to large micrometer-sized agglomerates in pristine ND, is was also possible to identify globular primary structures of 20–30 nm size. The commercial pristine ND powder used in this work consisted of primary particles with the size of about 5–6 nm. However, the presence of numerous polar oxygenated functionalities at their surfaces is responsible for strong inter-particle interactions, which make primary particles, as well as their primary agglomerates, stick together. The resulting formations reaching the sizes of tens of micrometers are very stable and very difficult to disperse,⁴⁰ which limits a broader use of this important class of carbon nanomaterials. Amine functionalization does not alter in an evident way the above morphology for ND + AF and ND + AP (Fig. 15c–f). Only for ND + ODA (Fig. 15g and h) we observed that dominating agglomerate size became larger than 1 μm, where smaller particles were present in insignificant numbers. This can be explained by stronger inter-particle interactions due to longer hydrophobic radicals of ODA and their higher content, appreciated from the dispersibility test in water (Fig. 10), FTIR spectrum (Fig. 11) and TGA curve (Fig. 12).

Fig. 15 Representative SEM images at different magnification for ND samples before (a, b) and after functionalization with AF (c, d), AP (e, f) and ODA (g, h). Scale bar: (a, c, e, g) 10 μm; (b, d, f, h) 1 μm.

It is important to note that SEM observations were performed by using as-received commercial pristine ND and as-synthesized amino-functionalized samples, without any attempt to destroy the agglomerates. On the contrary, for AFM imaging employed in order to provide finer morphological details, we deposited the samples onto silicon supports after ultrasonic bath treatment. As one can see from Fig. 16a and b, micrometer-sized and larger agglomerates were typically observed in pristine ND, where it was possible to discern primary aggregates. On the contrary, in the topography images of ND + AP and ND + ODA (Fig. 16e–h) 10¹ nm sized primary clusters dominated. The case of ND + AF was intermediate, where neither micrometer-sized agglomerates nor 10¹ nm sized primary clusters were common, but instead 10² nm sized secondary agglomerates were observed.

Fig. 16 Representative AFM topography images at different magnification for ND samples before (a, b) and after functionalization with AF (c, d), AP (e, f) and ODA (g, h). z-Scale in nm.

4. Conclusions

The results obtained demonstrate that two rather different carbon nanomaterials, GO and ND, can be covalently functionalized through gas-phase treatment with aromatic (AF and AP) and aliphatic (ODA) amines, which is carried out under moderate vacuum, at temperatures of 150–180 °C, using short reaction times of about 2 h. The amine treatment generally gives rise to changes in FTIR, Raman, UV-visible, and XPS spectra, to a variable degree depending on particular sample. The most important conclusion in terms of functionalization chemistry, based on the analysis of FTIR and XPS spectra, is that, while amidation is the only possible route for ND functionalization with AF, AP and ODA, there is an alternative reaction channel on GO sheets, which is amine addition onto epoxy bridges. Therefore, the amide functionalization scheme presented originally in Fig. 1d should be corrected in the way shown in Fig. 17.

Fig. 17 Scheme of GO functionalization accounting for the covalent attachment of amine molecules to both carboxylic and epoxy groups.

The dispersibility in water for both GO and ND decreased after functionalization, whereas the stability of GO + AF and GO + AP dispersions in isopropanol increased as compared to non-functionalized GO. The content of amine species covalently bonded to GO sheets, roughly estimated from TGA curves, was about 5%, 12% and 33%, for GO + AF, GO + AP and GO + ODA, respectively. In the case of ND functionalization, it barely reached 3–4% in the case of ND + AF and ND + AP, whereas for ND + ODA, the contribution of organics was higher than 30%. In other words, the quantitative trends for GO and ND functionalization efficiency are similar: a higher efficiency for aliphatic amine as compared to its aromatic analogues.

Considerable changes in sample morphology due to functionalization were observed by microscopy techniques. In particular, according to TEM, pristine GO monolayers exhibit relatively smooth edges and do not tend to rolling, whereas after functionalization with AP and ODA (and to a lower degree with AF), the individual graphene sheets acquired a strongly irregular shape, with numerous wrinkles and corrugations, as well as showed a stronger trend to agglomeration. Similarly, the principal difference between GO before and after functionalization according to AFM imaging is the degree of sheet agglomeration and rolling. As regards ND, micrometer-sized and larger agglomerates were typically observed in pristine ND, where it was possible to discern primary aggregates. On the contrary, in the topography images of ND + AP and ND + ODA 10¹ nm sized primary clusters dominated. The case of ND + AF was intermediate, where neither micrometer-sized agglomerates nor 10¹ nm sized primary clusters were common, but instead 10² nm sized secondary agglomerates were observed.

As a whole, the results obtained further broaden the spectrum of oxidized carbon nanomaterials which can be covalently functionalized with amine molecules by using the facil and fast one-step solvent-free procedure under reduced pressure and elevated temperature.

Acknowledgements

Financial support from the National Autonomous University of Mexico (grant DGAPA-IN100815, synthetic part and spectroscopic characterization, and DGAPA-IN200516, microscopy imaging) and from the National Council of Science and Technology of Mexico (CONACYT, grants 250655 and 270242) is greatly appreciated. N. A.-C. acknowledges CONACYT for a PhD fellowship. The authors are also grateful to María Cristina Zorrilla-Cangas (Instituto de Física, UNAM) for Raman spectral measurements, to David A. Domínguez (Centro de Nanociencias y Nanotecnología, UNAM) for XPS measurements, and to Dr Edgar Alvarez-Zauco (Facultad de Ciencias, UNAM) for the opportunity to use TGA equipment.

References

1. D. Chen, H. Feng and J. Li, Graphene oxide: Preparation, functionalization, and electrochemical applications, Chem. Rev., 2012, 112, 6027–6053.
2. K. P. Loh, Q. Bao, G. Eda and M. Chhowalla, Graphene oxide as a chemically tunable platform for optical applications, Nat. Chem., 2010, 2, 1015–1024.
3. V. N. Mochalin, O. Shenderova, D. Ho and Y. Gogotsi, The properties and applications of nanodiamonds, Nat. Nanotechnol., 2012, 7, 11–23.
4. A. Krueger, New carbon nanomaterials: Biological applications of functionalized nanodiamond materials, Chem.–Eur. J., 2007, 1382–1390.
5. D. R. Dreyer, S. Park, C. W. Bielawski and R. S. Ruoff, The chemistry of graphene oxide, Chem. Soc. Rev., 2010, 39, 228–240.
6. W. Zhang, J. Ma, D. Gao, Y. Zhou, C. Li, J. Zha and J. Zhang, Preparation of amino-functionalized graphene oxide by Hoffman rearrangement and its performances on polyacrylate coating latex, Prog. Org. Coat., 2016, 94, 9–17.
7. Y. Liu, Z. Gu, J. L. Margrave and V. N. Khabashesku, Functionalization of nanoscale diamond powder:  fluoro-, alkyl-, amino-, and amino acid-nanodiamond derivatives, Chem. Mater., 2004, 16, 3924–3930.
8. A. Krueger and D. Lang, Functionality is key: Recent progress in the surface modification of nanodiamond, Adv. Funct. Mater., 2012, 22, 890–906.
9. V. N. Mochalin and Y. Gogotsi, Wet chemistry route to hydrophobic blue fluorescent nanodiamond, J. Am. Chem. Soc., 2009, 131, 4594–4595.
10. A. M. Shanmugharaj, J. H. Yoon, W. J. Yang and S. H. Ryu, Synthesis, characterization, and surface wettability properties of amine functionalized graphene oxide films with varying amine chain lengths, J. Colloid Interface Sci., 2013, 401, 148–154.
11. I. P. Chang, K. C. Hwang, J. A. Ho, C. C. Lin, R. J. R. Hwu and J. C. Horng, Facile surface functionalization of nanodiamonds, Langmuir, 2010, 26, 3685–3690.
12. A. Navaee and A. Salimi, Efficient amine functionalization of graphene oxide through the Bucherer reaction: an extraordinary metal-free electrocatalyst for the oxygen reduction reaction, RSC Adv., 2015, 5, 59874–59880.
13. V. B. Saptal, T. Sasaki, K. Harada and D. Nishio-Hamane, Hybrid amine-functionalized graphene oxide as a robust bifunctional catalyst for atmospheric pressure fixation of carbon dioxide using cyclic carbonates, ChemSusChem, 2016, 9, 644–650.
14. (a) E. V. Basiuk, A. Santamaría-Bonfil, V. Meza-Laguna, T. Y. Gromovoy, E. Alvares-Zauco, F. F. Contreras-Torres, J. Rizo, G. Zavala and V. A. Basiuk, Solvent-free covalent functionalization of nanodiamond with amines, Appl. Surf. Sci., 2013, 275, 324–334; (b) E. V. Basiuk, V. A. Basiuk, V. Meza-Laguna, F. F. Contreras-Torres, M. Martínez, A. Rojas-Aguilar, M. Salerno, G. Zavala, A. Falqui and R. Brescia, Solvent-free covalent functionalization of multi-walled carbon nanotubes and nanodiamond with diamines: looking for cross-linking effects, Appl. Surf. Sci., 2012, 259, 465–476.
15. E. V. Basiuk, V. A. Basiuk, J.-G. Bañuelos, J. M. Saniger-Blesa, V. A. Pokrovskiy, T. Yu Gromovoy, A. V. Mischanchuk and B. G. Mischanchuk, Interaction of oxidized single-walled carbon nanotubes with vaporous aliphatic amines, J. Phys. Chem. B, 2002, 106, 1588–1597.
16. (a) E. V. Basiuk and V. A. Basiuk, Green chemistry of carbon nanomaterials, J. Nanosci. Nanotechnol., 2014, 14, 644–672; (b) E. V. Basiuk and V. A. Basiuk, Solvent-free functionalization of carbon nanomaterials, in Green Processes for Nanotechnology: From Inorganic to Bioinspired Nanomaterials, ed. V. A. Basiuk and E. V. Basiuk, Springer International Publishing, New York, 2015, pp. 163–205.
17. D. Yang, A. Velamakanni, G. Bozoklu, S. Park, M. Stoller, R. D. Piner, S. Stankovich, I. Jung, D. A. Field, C. A. Ventrice Jr and R. S. Ruoff, Chemical analysis of graphene oxide films after heat and chemical treatments by X-ray photoelectron and micro-Raman spectroscopy, Carbon, 2009, 47, 145–152.
18. W. Li, X.-Z. Tang, H.-B. Zhang, Z.-G. Jiang, Z.-Z. Yu, X.-S. Du and Y.-W. Mai, Simultaneous surface functionalization and reduction of graphene oxide with octadecylamine for electrically conductive polystyrene composites, Carbon, 2011, 49, 4724–4730.
19. B. T. McGrail, B. J. Rodier and E. Pentzer, Rapid functionalization of graphene oxide in water, Chem. Mater., 2014, 26, 5806–5811.
20. B. Yuan, C. Bao, L. Song, N. Hong, K. M. Liew and Y. Hu, Preparation of functionalized graphene oxide/polypropylene nanocomposite with significantly improved thermal stability and studies on the crystallization behavior and mechanical properties, Chem. Eng. J., 2014, 237, 411–420.
21. S. K. Singh, M. K. Singh, P. P. Kulkarni, V. K. Sonkar, J. J. A. Grácio and D. Dash, Amine-modified graphene: thrombo-protective safer alternative to graphene oxide for biomedical applications, ACS Nano, 2012, 6, 2731–2740.
22. (a) W. Zhang, Q. Zhao, T. Liu, Y. Gao, Y. Li, G. Zhang, F. Zhang and X. Fan, Phosphotungstic acid immobilized on amine-grafted graphene oxide as acid/base bifunctional catalyst for one-pot tandem reaction, Ind. Eng. Chem. Res., 2014, 53, 1437–1441; (b) S. Barazzouk and C. Daneault, Amino acid and peptide immobilization on oxidized nanocellulose: spectroscopic characterization, Nanomaterials, 2012, 2, 187–205; (c) C. Goel, H. Bhunia and P. K. Bajpai, Synthesis of nitrogen doped mesoporous carbons for carbon dioxide capture, RSC Adv., 2015, 5, 46568–46582; (d) F. A. Permatasari, A. H. Aimon, F. Iskandar, T. Ogi and K. Okuyama, Role of C–N configurations in the photoluminescence of graphene quantum dots synthesized by a hydrothermal route, Sci. Rep., 2016, 6, 21042.
23. (a) B. Wang, B. Luo, M. Liang, A. Wang, J. Wang, Y. Fang, Y. Chang and L. Zhi, Chemical amination of graphene oxides and their extraordinary properties in the detection of lead ions, Nanoscale, 2011, 3, 5059–5066; (b) G. Zorn, L.-H. Liu, L. Árnadóttir, H. Wang, L. J. Gamble, D. G. Castner and M. Yan, X-Ray photoelectron spectroscopy investigation of the nitrogen species in photoactive perfluorophenylazide-modified surfaces, J. Phys. Chem. C, 2014, 118, 376–383.
24. A. Kolmakov, D. A. Dikin, L. J. Cote, J. Huang, M. K. Abyaneh, M. Amati, L. Gregoratti, S. Günther and M. Kiskinova, Graphene oxide windows for in situ environmental cell photoelectron spectroscopy, Nat. Nanotechnol., 2011, 6, 651–657.
25. (a) L. Roldán, A. M. Benito and E. García-Bordejé, Self-assembled graphene aerogel and nanodiamond hybrids as high performance catalysts in oxidative propane dehydrogenation, J. Mater. Chem. A, 2015, 3, 24379–24388; (b) B. Yu, X. Wang, X. Qian, W. Xing, H. Yang, L. Ma, Y. Lin, S. Jiang, L. Song, Y. Hu and S. Lo, Functionalized graphene oxide/phosphoramide oligomer hybrids flame retardant prepared via in situ polymerization for improving the fire safety of polypropylene, RSC Adv., 2014, 4, 31782–31794.
26. D. C. Marcano, D. V. Kosynkin, J. M. Berlin, A. Sinitskii, Z. Sun, A. Slesarev, L. B. Alemany, W. Lu and J. M. Tour, Improved synthesis of graphene oxide, ACS Nano, 2010, 4, 4806–4814.
27. K. Lellala, K. Namratha and K. Byrappa, Ultrasonication assisted mild solvothermal synthesis and morphology study of few-layered graphene by colloidal suspensions of pristine graphene oxide, Microporous Mesoporous Mater., 2016, 226, 522–529.
28. V. A. Basiuk, N. Alzate-Carvajal, L. V. Henao-Holguín, E. V. Rybak-Akimova and E. V. Basiuk, Coordination functionalization of graphene oxide with tetraazamacrocyclic complexes of nickel(II): generation of paramagnetic centers, Appl. Surf. Sci., 2016, 371, 16–27.
29. K. Krishnamoorthy, M. Veerapandian, K. Yun and S. J. Kim, The chemical and structural analysis of graphene oxide with different degrees of oxidation, Carbon, 2013, 53, 38–49.
30. X. Sun, Z. Liu, K. Welsher, J. T. Robinson, A. Goodwin, S. Zaric and H. Dai, Nano-graphene oxide for cellular imaging and drug delivery, Nano Res., 2008, 1, 203–212.
31. L. Schmidlin, V. Pichot, M. Comet, S. Josset, P. Rabu and D. Spitzer, Identification, quantification and modification of detonation nanodiamond functional groups, Diamond Relat. Mater., 2012, 22, 113–117.
32. V. Mochalin, S. Osswald and Y. Gogotsi, Contribution of functional groups to the Raman spectrum of nanodiamond powders, Chem. Mater., 2009, 21, 273–279.
33. E. V. Basiuk, I. J. Ramírez-Calera, V. Meza-Laguna, E. Abarca-Morales, L. A. Pérez-Rey, M. Re, P. Prete, N. Lovergine, E. Álvarez-Zauco and V. A. Basiuk, Solvent-free functionalization of carbon nanotube buckypaper with amines, Appl. Surf. Sci., 2015, 357, 1355–1368.
34. I. J. Ramírez-Calera, V. Meza-Laguna, T. Y. Gromovoy, M. I. Chávez-Uribe, V. A. Basiuk and E. V. Basiuk, Solvent-free functionalization of fullerene C60 and pristine multi-walled carbon nanotubes with aromatic amines, Appl. Surf. Sci., 2015, 328, 45–62.
35. G. A. Inel, E.-M. Ungureau, T. S. Varley, M. Hirani and K. B. Holt, Solvent–surface interactions between nanodiamond and ethanol studied with in situ infrared spectroscopy, Diamond Relat. Mater., 2016, 61, 7–13.
36. H. Huang, E. Pierstorff, E. Osawa and D. Ho, Protein-mediated assembly of nanodiamond hydrogels into a biocompatible and biofunctional multilayer nanofilm, ACS Nano, 2008, 2, 203–212.
37. E. V. Basiuk, O. Ochoa-Olmos, F. F. Contreras-Torres, V. Meza-Laguna, E. Alvarez-Zauco, I. Puente-Lee and V. A. Basiuk, “Green” functionalization of pristine multi-walled carbon nanotubes with long-chain aliphatic amines, J. Nanosci. Nanotechnol., 2011, 11, 5546–5554.
38. H. P. Mungse, R. Singh, H. Sugimura, N. Kumar and O. P. Khatri, Molecular pillar supported graphene oxide framework: conformational heterogeneity and tunable d-spacing, Phys. Chem. Chem. Phys., 2015, 17, 20822–20829.
39. H. Huang, L. Dai, D. H. Wang, L.-S. Tan and E. Osawa, Large-scale self-assembly of dispersed nanodiamonds, J. Mater. Chem., 2008, 18, 1347–1352.
40. M. Y. Shalaginov, G. V. Naik, S. Ishii, M. N. Slipchenko, A. Boltasseva, J. X. Cheng, A. N. Smolyaninov, E. Kochman and V. M. Shalaev, Characterization of nanodiamonds for metamaterial applications, Appl. Phys. B, 2011, 105, 191–195.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra22658b

---