Introduction

Nanoparticles, fine particles having length ranging from 1 to 100 nanometres in two or three dimensions1, exhibit properties that are not observed for molecules or bulk counterparts2. They are widely used as building blocks for nanotechnology-derived applications such as single-electron devices, ultra dense recording media, bioelectronic devices and sensors, bioimaging, optoelectronic devices, catalysis and chemical sensors and energy conversion and storage2,3. Such ultrafine particles are usually prepared in a bottom-up manner by allowing molecules or atoms to assemble and build up into nanoparticles through chemical reactions in solutions or gas2,3.

Fine particles can also be obtained in a top-down manner by pulverising bulk solids4. When physical forces are applied to a solid, it undergoes plastic deformation to a breaking point, above which fracture results. The size of the solid is reduced as fracturing is repeated during pulverisation. However, as the size becomes smaller, the applied energy is rather dissipated as heat and the size reduction becomes increasingly difficult. Consequently, conventional milling devices typically produces particles with an average size no smaller than several ten micrometres5 and nanoparticles are not obtained unless very high energy is applied using a special device such as a high-energy ball mill6,7. We discovered entirely different size-reduction characteristics for bulk solids of fullerene C60.

Results

Solid C60 (1.5 mg) was placed between two glass microscope slides (76 mm × 26 mm, thickness 0.8 – 1.0 mm) and rubbed repeatedly between fingertips for a few minutes (Fig. 1a). Frictional resistance of C60 increased progressively as it was rubbed and coarse black particles (Fig. 1b, median diameter 129 μm)8 eventually turned to fine brownish powder and adhered to the glass surface.

Figure 1
figure 1

Formation of nanoparticles by rubbing solid C60 between glass slides.

(a) Experimental procedure. (b) SEM image of particles of as-received solid C60. Scale bar represents 100 μm. (c, d, e and f) SEM images of nanoparticles of C60 formed by rubbing bulks solids between glass slides. Scale bars represent 10 μm (c), 2 μm (d) and 200 nm (e and f).

Full size image

Examination by scanning electron microscopy (SEM) revealed that rubbed C60 had a bimodal size distribution, consisting of particles no larger than a few tens of micrometres and significantly smaller ones (Fig. 1c). Remarkably, detailed examination of the smaller particles showed that nanoparticles, some of which were even smaller than 100 nm, were generated by this mundane treatment. Such nanoparticles were mostly found associated on the surface of large particles (Figs. 1d and 1e), but agglomerates of nanoparticles were also observed around large particles (Fig. 1f).

Use of the glass slides was not critical. When solid C60 (ca. 2 mg) was rubbed between fingertips without using the glass slides for 1 minute (a polyethylene glove was used to avoid direct contact of C60 with skin), nanoparticles smaller than 100 nm (Fig. 2b) were also observed in rubbed C60 on the surface of the glove (Fig. 2a), although chances of finding such nanoparticles was much less compared with C60 rubbed with the glass slides.

Figure 2
figure 2

Formation of nanoparticles by rubbing solid C60 between fingertips.

(a) A photograph showing C60 on the surface of the glove after rubbing. (b) SEM images of rubbed C60 on the glove surface. Particles that are smaller than 100 nm are seen. Scale bar represents 500 nm.

Full size image

To quantify the size reduction by rubbing, C60 on the glass slides was dispersed in 2 mL of water containing 1 wt% of an anionic surfactant, sodium dodecyl sulphate (SDS). When the aqueous solution of SDS was poured onto rubbed C60 on the slides with a pipette, a brown and turbid dispersion was immediately formed. The dispersion was subjected to ultrasonic treatment for 5 min (Model 5510, 42 kHz output frequency, Branson Ultrasonic Corporation, Danbury, USA) and filtered with a membrane filter (nominal pore size, 5 μm). It should be noted here that ultrasonic treatment only helped disintegrating agglomerated nanoparticles and did not induce further size-reduction9.

The highly turbid and brown filtrate (Fig. 3a) was found to contain nanoparticles of C60, whose average diameter was 256.8 ± 1.1 nm by dynamic light scattering (Fig. 3b). The concentration of C60 in the dispersion (Fig. 3a), which was measured spectrophotometrically10, was 2.46 ± 0.17 × 10−4 M (177 ± 12 μg/mL), meaning that approximately 24 wt% of C60 solids was turned to particles smaller than 5 μm simply by rubbing them between fingertips. Examination by high-resolution transmission electron microscopy (HRTEM) revealed the presence of particles smaller than 20 nm in the dispersion (Figs. 3c and 3d). The particle in Fig. 3c had a dimension 9 nm × 12 nm and the nearly spherical particle in Fig. 3d was 14 nm in diameter. Mean fringe spacing of both particles was 0.50 nm and agreed well with the value for C60 crystal of (220) plane (0.50074 nm)11, showing that these particles are nano-sized crystals of C60 having an face-centred cubic (fcc) structure. Calculation using crystallographic data indicates that the spherical particle in Fig. 3d consists of approximately 2500 C60 molecules and 46% of them are exposed to the particle surface8. Rubbed C60 on the surface of the glove (Fig. 2a), on the other hand, could not be analysed in this way. C60 strongly adhered to the glove surface and could not be removed and dispersed in water containing SDS.

Figure 3
figure 3

Size distribution of nanoparticles produced by rubbing solid C60 between glass slides.

(a) An optical photograph of a dispersion of nanoparticles of C60 in water containing 1 wt% SDS. (b) Size distribution of the nanoparticles of C60 in the dispersion. (c and d) HRTEM images of nanoparticles of C60 found in the dispersion. Scale bars represent 5 nm.

Full size image

Several unconventional engineering can produce nanoparticles of C60 from bulk solids. These include sonication of solid C60 in water12, prolonged stirring of solid C60 in water for several weeks13, or hand-grinding of solid C60 with an agate mortar and pestle8,9,14,15. Nevertheless, formation of nanoparticles by such a mundane treatment as rubbing between fingertips is unprecedented and noteworthy. Moreover, efficiency of size-reduction by rubbing between fingertips appears to be comparable to that of hand-grinding with an agate motor and pestle, by which approximately 34 wt% of C60 was turned to particles smaller than 5 μm8,9.

Comparable size-reduction characteristics observed for both hand-ground and finger-rubbed C60 strongly suggests that top-down fabrication of nanoparticles of C60 by mechanical means require a surprisingly little effort. Far from it, we found that nanoparticles of C60 were generated even without deliberate engineering of any sort. Fig. 4a shows a mouth of a reagent bottle of C60. A white plastic bushing at the mouth was covered with brown powder of spilt C60, which had been compressed and sheared repeatedly by a screw cap whenever the bottle had been opened and closed. Examination by SEM revealed particles smaller than 100 nm in C60 that was collected from the bushing (Fig. 4b).

Figure 4
figure 4

Non-engineered nanoparticles of C60.

(a) A photograph showing a white bushing of the month of a reagent bottle of C60. Right-hand side of the bushing was cleaned before taking the picture to reveal the original white colour. (b) SEM image of C60 that was collected from the bushing. Nanoparticles that are smaller than 100 nm are clearly seen. Scale bar represents 500 nm.

Full size image

Discussion

Considering the efforts usually required to prepare the nanoparticle by pulverisation6,7, the size-reduction characteristics of C60 are truely anomalous. The size-reduction process of solid C60 is purely physical and is not associated with change in crystalline structure or chemical/mechanochemical reactions8,9. Formation of nanoparticles of C70 by pulverisation was also reported16, suggesting that anomalous size-reduction is a common characteristics among fullerenes. Detailed mechanisms behind the anomaly are not clear at present, but it could be ascribed to inherent properties of crystalline C60 such as fast isotropic rotation of the molecule17 or low cohesive energy (1.6 eV)8,9,18. Labille et al. used X-ray diffraction to study nanoparticles of C60 that were prepared by prolonged stirring of bulks solids in water and proposed that nanoparticles of C60 are formed by an erosion of large C60 crystals, which occurs preferentially via (111) lattice plane exfoliation19. The adhesion energy between smooth C60 surfaces is an order of magnitude lower than that of typical van der Waals solids20. This is ascribed to structural features of the C60 molecule including a large size, rigidity and smooth surface, which makes the C60 molecule behave like a macroscopic spherical object rather than a conventional molecule20. Shear stress exerted by fingertips or a screw cap may be sufficient enough to overcome the low adhesion energy to induce exfoliation.

Concerns have emerged on possible adverse effects of engineered nanomaterials21 and nanoparticles of C60 have attracted considerable scientific attention in this regard22,23,24. Pulverization has widely been used to prepare C60 nanoparticles for evaluating their possible environmental and health impacts25. Compared with conventional methods for producing C60 nanoparticles such as recrystallisation from organic solvents (toluene26 or tetrahydrofuran10), pulverisation has distinct advantages that the product is free from residual solvents and the procedure is simple to perform. A major drawback of pulverisation is a broad size distribution of the nanoparticles produced, but this can be circumvented by size-fractionation using filtration or centrifugation8,15. The size of pulverised nanoparticles can also be controlled when an automated milling device (wet grinding using a bead mill) is employed27.

Present study shows that anomalous size-reduction characteristics of solid C60 have immediate implications for assessing health risks of C60 nanoparticles. Assessing possible human exposure routes to nanoparticles is an important consideration21,28 and is among five grand challenges necessary toward safe handling of nanotechnology29. As nanoparticles are usually engineered by chemical reactions in gas phase or solutions2,3, possible exposure routes appear to be foreseeable. Inhalation of airborne nanoparticles is a major exposure scenario30 and an interim report on risk assessment of C60 proposed the acceptable exposure concentration of 0.8 mg/m3 as respirable dust in working environments31.

The supposition is not always true, however. Glover et al. showed nanoparticles were spontaneously generated from macroscopic silver and copper objects when they were simply in contact with surfaces in a humid air32. Proposed chemical mechanism involves surface oxidation with ambient oxygen and adsorbed water, diffusion of metal ions from the parent particle in the adsorbed water layer and nucleation via chemical and/or photochemical reduction of the ions. Their findings imply that macroscopic objects can be a potential source of incidental nanoparticles in the environment and that humans have long been in direct contact with these nanomaterials without being noticed32.

Our results show there exists a mechanical pathway for similar incidental generation of nanoparticles from bulk C60 solids and suggest highly unconventional exposure routes to nanoparticles of C60. For example, if one rubs a spillage of C60 solids on a lab bench with a bare fingertip, he/she may be exposed inadvertently to risk of dermal uptake and inhalation of C60 nanoparticles, even though he/she presumes handling bulk solids of C60 and does not foresee exposure to nanoparticles. Similar situations may be encountered commonly in research laboratories or manufacturing facilities that use C60, or during disposal of products that are deemed to contain solid C60 of macroscopic size. A possibility was also suggested that commercial solid C60 contains nanoparticles that are generated by friction of solid particles during production, storage, or transportation9.

Although formation of nanoparticles by simple pulverisation is not known for other materials to the best of our knowledge, it is worth mentioning that nanoparticles are produced by simple pulverisation when mechano-chemical reactions are involved. For example, Rao et al. demonstrated that nearly pure Ag9 quantum clusters were prepared by hand-grinding AgNO3 and mercaptosuccinic acid solids followed by reduction by NaBH4 using a mortar and pestle33. Thus, detailed understandings of underlying mechanisms behind the mechanical and/or mechano-chemical generation of nanoparticles form bulk solids, which still remain largely unclear, are crucial in characterizing environmental and health impacts of nanoparticles.

Methods

Materials

C60 (> 99.9% pure) was obtained from Tokyo Kasei, Co., Ltd. (Tokyo, Japan) and used as received. Sodium dodecyl sulphate (SDS) was purchased from Nacalai Tesque, Inc. (Kyoto, Japan). Millipore water was used throughout the work.

Scanning electron microscopy

After rubbing C60 between glass microscope slides, one slide was held, with the side with C60 down, above a conductive double-sided tape that was mounted on a brass stub. The upper side of the slide was tapped with a lab spoon so that fine particles of rubbed C60 fell on the tape. Solid C60 at the mouth of a reagent bottle was collected in a same manner. The specimens were coated with osmium (estimated coating thickness, < 10 nm) and examined on a JSM-6700F (JEOL, Tokyo, Japan).

High-resolution transmission electron microscopy

A drop of the dispersion of nanoparticles of C60 in water containing 1 wt% SDS was deposited on a carbon TEM grid. The surface of the carbon supporting film was subjected to hydrophilic treatment before the sample deposition. Crystals of SDS that precipitated after drying the specimen was removed with methanol. The specimen was air-dried again and was examined on a Hitachi HF-2000 operating at 200 kV incident beam energy. The observations were performed at Nissan-Arc, Ltd. (Yokosuka, Japan).

Dynamic light scattering

Average size of the C60 nanoparticle in the aqueous dispersion containing 1 wt% SDS was measured by dynamic light scattering on an FDLS-1200 (Otsuka Electronics Co., Ltd., Osaka, Japan) equipped with a solid-state laser (λ = 532 nm, 100 mW). The measurements were done at 25.0 ± 0.1°C and at a fixed scattering angle of 90°. Average hydrodynamic diameter was calculated by using a cumulant method34, while CONTIN35 was employed to obtain a size distribution. The dispersion was 100-fold diluted with water before the measurements.