First Synthesis and Characterization of CH4@C60

Abstract The endohedral fullerene CH4@C60, in which each C60 fullerene cage encapsulates a single methane molecule, has been synthesized for the first time. Methane is the first organic molecule, as well as the largest, to have been encapsulated in C60 to date. The key orifice contraction step, a photochemical desulfinylation of an open fullerene, was completed, even though it is inhibited by the endohedral molecule. The crystal structure of the nickel(II) octaethylporphyrin/ benzene solvate shows no significant distortion of the carbon cage, relative to the C60 analogue, and shows the methane hydrogens as a shell of electron density around the central carbon, indicative of the quantum nature of the methane. The 1H spin‐lattice relaxation times (T 1) for endohedral methane are similar to those observed in the gas phase, indicating that methane is freely rotating inside the C60 cage. The synthesis of CH4@C60 opens a route to endofullerenes incorporating large guest molecules and atoms.


S1.1 General methods
Reactions were conducted under an argon atmosphere using standard Schlenk and syringe techniques with freshly distilled solvents. All apparatus was dried in a hot oven (>140 °C, 12 h) before being cooled in a sealed desiccator over silica gel or assembled while hot and cooled under vacuum (0.1 mm Hg).
Toluene was freshly distilled from sodium benzophenone ketal under argon. Technical grade 1-chloronaphthalene (≥85%) was distilled under nitrogen and solutions in 1-chloronaphthalene were degassed under reduced pressure (<1 mm Hg) until evolution of gases had ceased. Triisopropyl phosphite was distilled over sodium at reduced pressure. Dimethyldioxirane [1] and di-(2-furyl)phenylphosphine [2] were prepared according to the published procedures. All other reagents, solvents or gases were used as received from commercial suppliers.
NMR spectra were recorded on a Bruker AVIIIHD500 FT-NMR spectrometer, or Bruker Ascend 700 NB magnet with Bruker AVANCE NEO console and Bruker TCI prodigy 5 mm liquids cryoprobe; in the indicated solvent at 295K. 1 H chemical shifts are reported as values in ppm referenced to residual solvent. 1  Bis(hemiketal) open-fullerene 4 [3] and sulfide open-fullerene 3 [4,5] were prepared according to the published methods.

S1.4 CH 4 @6
Intermediate CH 4 @4 was obtained using either procedure A or B:  transients and a delay of 10 s between scans. The peak corresponding to CH 4 @5 is found at −16.54 ppm and is shown in the expansion. The spectrum was processed using Lorentzian line broadening (full-width at half-maximum = 2 Hz).

S1.5 CH 4 @C 60
Triisopropyl phosphite (36 µL, 0.147 mmol) was added to a solution of CH 4 @6 (10.0 mg, 0.00920 mmol) in toluene (2 mL) and the resulting mixture was stirred at reflux for 21 h. After cooling to room temperature, the mixture was concentrated in vacuo and purified by column chromatography (SiO 2 eluted with toluene). [6] The fractions containing material with R f = 0.95 were collected and evaporated to dryness to afford a black solid which was taken into 1-chloronaphthalene (1 mL) and transferred to a Schlenk flask charged with N-phenyl maleimide (2 mg, 0.0113 mmol) and fitted with a straight condenser. The mixed solution was degassed and placed under an atmosphere of argon before stirring at reflux for 24 h. After cooling to room temperature, the mixture was flushed through a SiO 2 column packed with toluene, collecting a purple band. Purification by preparative HPLC on a Cosmosil™ Buckyprep column (eluted with toluene) gave the title compound CH 4 @C 60 , with 100.0 ± 0.3 % filling, as a black solid (7.1 mg, quant. yield).
Although this step gave a quantitative measured yield there is likely to be a significant error associated with the very small quantity of material.  [6] Figure S1.5 Experimental 13 C NMR spectrum of CH 4 @C 60 with 1 H WALTZ16 decoupling (nutation frequency = 14.2 kHz) of CH 4 @C 60 (4.5 mM in degassed 1,2-dichlorobenzene-d 4 ) acquired at 16.45 T ( 13 C nuclear Larmor frequency = 176 MHz) and 295 K with 4928 transients and a delay of 10 s between scans. The spectrum was processed using Lorentzian line broadening (full-width at half-maximum = 2 Hz). The three solvent peaks are found around 130 ppm. The 13 C CH 4 @C 60 peak is found at 143.34 ppm and the 13 CH 4 @C 60 peak at −13.63 ppm (marked with an asterisk); both resonances are shown in the expansions.

S3.2 13 C spin-lattice relaxation measurement using INEPT
In order to obtain 13 C T 1 measurements using a 4.5 mM solution of un-labelled CH 4 @C 60 in 1,2-dichlorobenzened 4 , the 1 H magnetisation was transferred to 13 C through a refocused INEPT sequence for an IS 4 system, [8] followed by a variable τ EV delay in which the 13 C magnetisation decays. The remaining 13 C magnetisation was transferred to the protons through a reversed INEPT sequence, generating an antiphase pattern for the 13 C satellites. Detection was made through the sensitive proton channel. The pulse sequence used is shown in Figure S3 intensities oscillate and are not always equal, however the maximum intensity for both satellites is seen at 1 ms τ EV delay and not 100 µs which is the shortest delay in the sequence. Figure S3.2 Pulse sequence used to monitor the decay of 13 C longitudinal spin order in 13 CH 4 @C 60 . The decay of 13 C longitudinal spin order is tracked by repeating the experiment for different values of the evolution period τ EV . The sequence uses a four-step phase cycle to remove residual proton magnetisation (ϕ 1 = (x,x,x,-x), ϕ 2 = (y,y,-y,-y), ϕ REC = (x,-x,-x,x)). A delay of 8 s was used between successive experiments. The experimental parameters were as follows: [8] Figure S3.2 with an evolution period τ EV = 1 ms. The spectrum was processed using Lorentzian line broadening (full-width at half-maximum = 1.5 Hz). The asterisk denotes the 12 CH 4 @C 60 proton peak removed by the application of the four-step phase cycle. The small residual peak (*) indicates a strong suppression of the 12 CH 4 @C 60 proton signal attributed to the implementation of the four-step phase cycle.

S5. Measurement of the relative yields for photochemical closure of CH 4 @5 vs H 2 O@5
We were unable to attribute the low yielding photochemical ring-contraction of CH 4 @5 to any factor other than the presence of endohedral methane and tested the proposal using partially filled 5 where we would expect a change in filling factor to be a sensitive test of the effect of the methane, i.e. if the closure of 'empty' 5 is higher yielding than that of CH 4 @5 we would expect the filling factor to be lower in the product 4, after closure.
Unfortunately, given the very fast entry of H 2 O into 5 under the partially aqueous conditions of the reaction, the competition is mostly between CH 4 @5 and H 2 O@5 (rather than between CH 4 @5 and empty 5) so we are unable to distinguish between an inhibitory effect of the CH 4 and a promoting effect of the H 2 O. The precise equilibrium filling of 5 with H 2 O under the reaction conditions is unknown, but closure of 'empty' 5 occurs to yield product 4 containing 80 ± 5% endohedral H 2 O in ~25% yield.

S5.1 Experimental method
A mixture of CH 4 @5 (>95% filling, 32 mg) and 5 (10 mg) was prepared in order to dilute the filled material. We assume that the % filling is unaffected in conversion of 4 to 6 since no change in % filling for conversion of H 2 O@4 to H 2 O@6 under identical conditions has been observed. [6] The possibility of release of the entrapped methane molecule during the photochemical ring-contraction is discounted by our observation that there was no measured change in % filling when the CH 4 @5 starting material has >95% filling. Estimation of the error in the calculated ratio CH 4 @4/H 2 O@4 assumes that the measurement of % filling from the experimental 1 H NMR spectrum has a standard deviation of ± 2.5%. The samples may contain a few % empty 6, but as this is the difference between the two measured incorporations and 100%, the uncertainty is large. We take the total of H 2 O@6 and empty 6 to be 43% for the calculation below.
In accordance with our observations, it would therefore be expected that the procedure which yields 25% of bis(hemiketal) H 2 O@4 from irradiation of sulfoxide 5, will furnish 4 -8% of CH 4 @4 from irradiation of CH 4 @5.
Supporting information S14 S6. X-Ray structure determination of CH 4 @C 60 Figure S6.1 Thermal ellipsoid plot of the asymmetric unit, ellipsoids drawn at the 50% probability level with hydrogen atoms depicted as spheres with an arbitrary radius. The hydrogen atoms of methane were attached at calculated positions and the rotation of the rigid group was refined until a local minimum in the shift/error was achieved. As described in the main paper, the visual depiction of hydrogen atoms has no physical significance.  Overlay of CH 4 @C 60 reported in this paper (blue) and empty C 60 (red) [9] structures; the r.m.s difference is 0.00876 (data collection temperatures differ by 8K).

S6.1 Experimental method
Single dark orange plate-shaped crystals were recrystallised from a solution of CH 4 @C 60 and nickel(II) octaethylporphyrin (1:1.6 molar ratio of CH 4 @C 60 :Ni(II)OEP) in benzene, by slow evaporation. [10] A suitable crystal of 0.10 × 0.10 × 0.02 mm 3 dimensions was selected and mounted on a MITIGEN holder with silicon oil on an Rigaku AFC12 FRE-VHF diffractometer. The crystal was kept at a steady T = 100(2) K during data collection.
The structure was solved with the ShelXT structure solution program using the intrinsic phasing methods solution method and by using Olex2 [11] as the graphical interface. The model was refined with version 2016/6 of ShelXL [12] using Least Squares minimisation.

Refinement:
A dark orange plate-shaped crystal with dimensions 0.10×0.10×0.02 mm 3 was mounted on a MITIGEN holder with silicon oil. X-ray diffraction data were collected using a Rigaku AFC12 FRE-VHF diffractometer equipped with an Oxford Cryosystems low-temperature device, operating at T = 100(2) K.
Data were measured using profile data from w-scans of 0.5 ° per frame for 15.0 s using MoKa radiation (Rotatinganode X-ray tube, 45.0 kV, 55.0 mA). The total number of runs and images was based on the strategy calculation from the program A multi-scan absorption correction was performed using CrysAlisPro 1.171.39.46 (Rigaku Oxford Diffraction, 2018) using spherical harmonics as implemented in SCALE3 ABSPACK. The absorption coefficient µ of this material is 0.363 mm -1 at this wavelength (l = 0.711Å) and the minimum and maximum transmissions are 0.727 and 1.000, respectively.
The structure was solved in the space group P-1 (# 2) by intrinsic phasing methods using the ShelXT structure solution program and refined by Least Squares using version 2016/6 of ShelXL. [12] All non-hydrogen atoms were refined anisotropically. Most hydrogen atom positions were calculated geometrically and refined using the riding model, but some hydrogen atoms were refined freely.
_refine_special_details: Hydrogen atoms of the CH 4 were generated from observed electron density peaks and then restrained to have tetrahedral geometry, the molecule will be freely rotating and the orientation represents one of many local minima in the LS refinement. There is a single molecule in the asymmetric unit, which is represented by the reported sum formula. In other words: Z is 2 and Z' is 1.