Introduction

The importance of helium is widely seen in its common use in helium balloons or respirator gas, as well as in its scientific use as a refrigerant for superconductive magnets and physical research at low temperature. Helium is well known as the smallest noble gas atom and it has refused to be directly observed by single crystal X-ray analysis, a powerful method to determine molecular structure. This is because neither helium nor a helium-incorporating material becomes crystalline solid under atmospheric pressure even in the vicinity of 0 K (ref. 1). Thus, it is crucial to embed a helium atom selectively into some crystalline material with a discrete structure if one wishes to observe it by X-ray analysis.

Because the strength of the X-ray diffraction of single crystal X-ray analyses depends on the total number of electrons of all atoms existing in the unit cell, it is usually difficult to observe atoms with poor electron density, such as hydrogen, after structural refinements. In contrast, if a proper single crystal is available, a helium atom should be observable by the single X-ray analysis because a helium atom possesses two electrons in its 1s electronic shell which is isoelectronic with the lithium cation widely observed in the literature2. Fullerene C60 is a good candidate to entrap a helium atom. Endohedral C60 and C70 encapsulating a helium atom, that is, 3He@C60 and 3He@C70 were first reported by Saunders et al.3 in 1994 in which 3He atom was forced into pristine C60 and C70 under drastic conditions, such as 650 °C under 3000, atm of helium gas3,4. However, the encapsulation level of the helium atom was quite low such as 0.15–0.4% (refs 5,6) because of the inefficiency of this physical method, although the inner 3He atom is useful for the clarification of the internal magnetic field of fullerenes5. On the other hand, we have developed rational synthetic methodology7,8,9,10,11,12,13,14,15, so-called ‘molecular surgery’16, and recently succeeded in the chemical syntheses of He@C60 and He@C70 with ca. 30% encapsulation levels in macroscopic scales. This method involves insertion of a helium atom into open-cage fullerene derivatives followed by restoration of the openings under mild conditions14. The encapsulation level of He@C60 thus obtained was enhanced to 95% by purification with recycling high-performance liquid chromatography (HPLC)14.

Here we report the single crystal X-ray observation of a helium atom inside C60. In addition, we also report the close contact of the helium atom with a nitrogen atom by placing a nitrogen atom inside He@C60 and He@C70 by the use of nitrogen radio-frequency (RF) plasma method17,18,19.

Results

Single crystal X-ray analysis of He@C60

Although C60 is a crystalline material, a C60 cage is known to rotate even in the solid state20. To prevent such a rotational or orientational disorder in the single crystal, co-crystals of the He@C60 and nickel(II) octaethylporphyrin (NiOEP) were prepared from the solution in o-xylene15,21. The X-ray diffraction data were collected using synchrotron radiation at 173 K at SPring-8. The least-square refinements gave the structure of He@C60·(NiOEP)2 without any disorders of the C60 cage (Supplementary Tables S1–S3). As shown in Fig. 1, the encapsulated helium atom was unambiguously observed at the centre of the C60 cage with 100% occupation level, probably owing to the enrichment during crystallization. It should be mentioned that the position of the helium is similar to that of krypton reported by the X-ray diffraction data of (Kr@C60/C60)·(NiOEP)·2C6H6, with the occupation level of krypton being only 9% (21). The relatively larger ellipsoid of the helium implies its translational motion at this temperature because there remains small space for the encapsulated helium atom inside C60 (estimated diameter of the hollow space is ca. 3.7 Å and the van der Waals diameter of helium is 2.80 Å (ref. 22)).

Figure 1: X-ray crystal structure of He@C60·(NiOEP)2.
figure 1

Thermal ellipsoids at the 50% probability level and hydrogen atoms are omitted for clarity. NiOEP, nickel(II) octaethylporphyrin.

Insertion of a nitrogen atom to He@C70 and He@C60

After confirming the existence of one helium atom inside C60, we wondered about the capability of this complex to further accommodate an additional guest species. It seemed intriguing to see if so-believed ‘inert’ helium could affect the properties of a neighbour atom in tightly confined space inside He@C60 or He@C70. For this purpose, a nitrogen atom was selected as the guest. It is reported that a sharp triplet electron spin resonance (ESR) signal is observed for the reaction products of C60 and C70 with a nitrogen atom source (see below), indicating the generation of N@C60 and N@C70 (23,24,25,26). The inner nitrogen atom of these endofullerenes retains its electron spin of 3/2 without any covalent bond formation with carbons of the C60 or C70 cage. These products are believed to be endohedral fullerenes, which are promising materials for quantum computers27,28.

Insertion of a nitrogen atom into He@C70 (ca. 30% encapsulation level), whose inner space (estimated long axis diameter is ca. 4.7 Å) is larger than that of He@C60, was conducted under the nitrogen RF plasma conditions17,18,19. As shown in Fig. 2a (also see Supplementary Fig. S1), the ESR spectrum of the resulting material showed a new triplet signal in addition to a larger triplet corresponding to N@C70 with the hyperfine coupling constant (hfcc) of 5.38 Gauss as reported27. In spite of a partial overlap with the known triplet corresponding to N@C70, a simulated spectrum revealed the presence of a new species with the hfcc of 5.59 Gauss. This result is most reasonably interpreted by the generation of C70, encapsulating both helium and nitrogen atoms, that is, HeN@C70. The existence of HeN@C70 was also confirmed by mass spectrometry of the enriched sample prepared by repeated recycling HPLC (Supplementary Fig. S2). Moreover, though the inner space of He@C60 seemed too small to accommodate a nitrogen atom as an additional guest, similarly overlapped ESR signals were obtained when He@C60 (ca. 30% encapsulation level) was used for the nitrogen insertion (Fig. 2b), indicating the generation of HeN@C60. The hfcc of HeN@C60 (5.99 Gauss) is slightly larger than that of N@C60 (5.67 Gauss). The ratio of the integrated peak areas of HeN@C60/N@C60 (1/9) was lower than that of HeN@C70/N@C70 (3/7) at room temperature. This difference implies that it is easier for a nitrogen atom to access to the larger inner space of He@C70 than that of He@C60.

Figure 2: Generation of the C70 and C60 incorporating a helium atom and a nitrogen atom.
figure 2

(a) Experimental ESR spectrum of HeN@C70 and N@C70 in CS2 at 220 K (black line), and simulated spectra of the mixture (green line), N@C70 (blue line) and HeN@C70 (red line), respectively. (b) Experimental ESR spectrum of HeN@C60 and N@C60 in CS2 at room temperature (black line), and simulated spectra of the mixture (green line), N@C60 (blue line) and HeN@C60 (red line), respectively.

Discussion

The inner nitrogen atoms of HeN@C60 and HeN@C70 still retain the electron spin state of 3/2, in spite of the existence of the tightly packed helium atom, showing the entire inertness of helium. However, the line widths of three signals for HeN@C60 and HeN@C70 are clearly broadened compared with those for N@C60 and N@C70. This is because placing a helium atom inside N@C60 and N@C70 causes lower symmetry of the resulting HeN@C60 and HeN@C70 compared with high symmetry of the starting materials. Such lower symmetry of the molecules breaks the degeneracy of the three 2p orbitals of the inner nitrogen atom, giving rise to appearance of the broad signals in which possible fine structures are about to appear. This is consistent with the reported observations on N@C60 derivatives with exohedral functionalizations29,30.

To gain insight into the structures and properties of HeN@C60, theoretical consideration was essential. The structures of the related molecules were fully optimized at the M06-2X31 method using basis sets of 6-31G(d)32 for C and 6-311+G(2df)33 for He and N atoms34 (Supplementary Tables S4–S6). It was shown that the encapsulation of both helium and nitrogen atoms into C60 is exothermic by −7.1 kcal mol−1 after correction of basis set superposition error. This stabilization energy of HeN@C60 is lower than that of N@C60 (–23.4 kcal mol−1), but higher than that of He@C60 (–3.2 kcal mol−1), indicating that the coexistence of the two atoms in a small space inside C60 could be reasonable. It is considered that these atoms are packed tightly inside the C60 cage because the distance between the helium and the nitrogen atoms (2.07 Å) was obviously shorter than that at the most stable position (3.75 Å) calculated at the CCSD(T)/6-311+G(2df) level of theory.

Using the optimized structures at the M06-2X method, the hfcc values were studied by the single-point calculations at MP235 method using basis sets of 6-31G(d) for C and 6-311++G(3df,3pd) for He and N atoms. The calculated hfcc values were found to highly depend on the basis sets. Among the tested basis sets, the use of 6-311++G(3df,3pd) gave the acceptable results for N@C60 as a benchmark compound; the calculated and experimental hfcc values are 5.44 and 5.67 Gauss, respectively. As shown in Fig. 3a, the calculated hfcc values reproduced the experimental tendency that the smaller space for a nitrogen atom results in the larger hfcc (5.57 Gauss for HeN@C60), most probably owing to the increased s electron density on the nucleus of the nitrogen. To clarify the contributing factors of this phenomenon, the He atom in the optimized structure of HeN@C60 was removed, and the hfcc of the remaining N atom located at off-centre inside C60 was calculated as the hypothetical molecule to afford the larger hfcc value of 5.80 Gauss (Fig. 3a: HeN@C60–He). These results indicated that the hfcc of the nitrogen is larger, as the distance between the N atom and the fullerene cage becomes shorter. On the contrary, the hfcc value of a nitrogen atom with a neighbouring helium atom in a distance of 2.07 Å (3.52 Gauss) was smaller than that with a helium atom at the optimized distance of 3.75 Å (3.66 Gauss) or without a helium atom (3.69 Gauss) (Fig. 3b). These results showed that the hfcc values decrease as the helium atom gets closer to the nitrogen atom. Thus, the experimentally observed increase in hfcc value for HeN@C60 compared with that of N@C60 is rationalized in terms of the compensation of the increase by the C60 cage and the decrease by the neighbouring helium atom.

Figure 3: Calculated hfccs of nitrogen atoms at the MP2 level.
figure 3

(a) The hfccs of N@C60, HeN@C60 and HeN@C60–He (the helium atom being removed from the optimized structure of HeN@C60). (b) The hfccs of a free nitrogen atom and the nitrogen atoms with a neighbouring helium atom at the distance of 3.75 Å (optimized distance) and 2.07 Å (extracted distance from HeN@C60).

To obtain more information on the inner encapsulated atoms of helium and nitrogen, the Atoms-in-Molecules (AIM) analysis36 was carried out at the MP2 level with the optimized structure at the M06-2X method described above, using the AIM 2000 package program (version 2.0). In the AIM analysis, if the bond path through the bond critical point (BCP) connects two atoms, it is defined to be an interaction existing between these atoms. As shown in Fig. 4, there is a bond path between the He atom and the N atom. There are also three He–C and four N–C bond paths in the C60 cage (Fig. 4a). These results imply that not only the C60 cage but also the He atom interact with the N atom. By the analysis at the BCP (Supplementary Fig. S3)36, the inner N atom is not covalently bound with the He atom nor C60 cage. Comprehensively, the He atom and the N atom interact with each other only through a van der Waals interaction inside the C60 cage, whose situation is similar to that of He2@C60 (ref. 37).

Figure 4: AIM analysis of HeN@C60 at the MP2 level.
figure 4

(a) Molecular depiction with displaying BCPs in red. (b) Cross-section contour map of electron density with bond paths in brown and BCPs in red. The contours [ea0–3] are at 2 l (l=±8, ±7,..., 0), and 0.0047 (heavy line).

The present result clearly showed that a single atom of helium in C60 is observable by means of single crystal X-ray analyses. Thus, the element of helium newly joined the libraries of X-ray crystallographic study38. The encapsulated helium was confirmed to be located at the centre inside the C60. In addition, the stepwise insertion of different atoms into fullerenes was achieved by the combined ‘molecular surgery’ and ‘nitrogen RF plasma’ methods. Even highly reactive nitrogen atom was found to coexist with a helium atom in a small space inside fullerenes. The ESR signals of the nitrogen atom were influenced by the helium coexisting in the same fullerene cages. Thus, these results are considered to be pieces of evidence for encapsulation of a nitrogen atom inside fullerenes. This technique is expected to create a variety of endohedral fullerenes, encapsulating heterogeneous atoms in the near future.

Methods

General

He@C60 and He@C70 were synthesized by following our previous report14. ESR spectra were measured by Bruker EMX and EMX plus spectrometers. The temperature was controlled with an Oxford ESR900 helium flow-type cryostat and an Oxford ITC503 temperature controller for the EMX plus. Simulation of the spectra was carried out on a WinSim program39. The atmospheric pressure chemical ionization mass spectrum was recorded on a Bruker micrOTOF-QII.

Single-crystal X-ray analysis of He@C60·(NiOEP)2

Single-crystal X-ray data were collected on a BL38B1 beamline in SPring-8 with the approval of the Japan Synchrotron Radiation Research Institute with a diffractometer equipped with an ADSC Quantum 315 CCD (charge-coupled device) detector. The collected diffraction data were processed with the HKL2000 software program. The structure solutions were obtained by direct method using the SHELXS-97 program40 and refined using the SHELXL-97 program40.

Experimental details for generation of HeN@C60 or HeN@C70

HeN@C60 or HeN@C70 was generated by a nitrogen RF plasma method17. He-encapsulating fullerene, He@C60 or He@C70, (~9 mg, ca. 30% encapsulation level) was put in a molybdenum boat and placed into the instrument. Under the reduced pressure of nitrogen gas (30 Pa), the material containing N-encapsulating fullerene was generated by sublimation through resistive heating in nitrogen plasma. After cooling to room temperature, the deposit (~3 mg) was dissolved in CS2, followed by filtration and evaporation of the solvent. The resulting material was dissolved in toluene and subjected to recycling HPLC equipped with the Cosmosil Buckyprep column (250 mm length, 20 mm inner diameter; mobile phase, toluene) for further purification. The major fraction was collected and evaporated. The residue was dissolved in CS2 and placed in an ESR tube, which was connected to a vacuum line. The solution was degassed by three freeze-pump thaw cycles under reduced pressure, and then sealed.

Enrichment of HeN@C70

After sublimation of He@C70/C70 (3/7) through resistive heating in nitrogen plasma, the resulting deposit (~3 mg) was subjected to the recycle HPLC equipped with the Cosmosil Buckyprep columns (two directly connected columns; 250 mm length, 20 mm inner diameter; mobile phase, toluene; 25 °C; flow rate was 8.0 ml min–1). The retention time of HeN@C70 was close to that of N@C70 (25). After two recycling, the latter portion of main peak was evaporated and further subjected to the same HPLC to enrich HeN@C70. The existence of HeN@C70 and N@C70 was monitored by the ESR spectra. After repeating this procedure five times, the resulting mixture in toluene solution (100 μl) was subjected to a Bruker micrOTOF-QII to give the mass spectrum shown in Supplementary Fig. S2.

Theoretical methods

Geometries were fully optimized with hybrid density functional theory at the M06-2X (31) level. The 6-31G(d) basis set32 was used for C, whereas the 6-311+G(2df) basis set33 was used for He and N. To calculate the hfccs, the M06-2X optimization was followed by single-point MP2 (second-order Møller–Plesset perturbation theory)35 calculations using 6-311++G(3df, 3pd) basis set for He and N, and 6-31G (d) basis set for C. Calculations were carried out using the GAUSSIAN 09 program34.

Additional information

Accession codes: The X-ray crystallographic coordinates for the structure reported in this Article has been deposited at the Cambridge Crystallographic Data Centre (CCDC), under deposition number CCDC 921390. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.

How to cite this article: Morinaka, Y. et al. X-ray observation of a helium atom and placing a nitrogen atom inside He@C60 and He@C70. Nat. Commun. 4:1554 doi: 10.1038/ncomms2574 (2013).