Supplementary information for
Mechanical behavior of water filled C60
K. Min, A. Barati Farimani, N. R. Aluru1*
1
Department of Mechanical Science and Engineering, Beckman Institute for Advanced Science and
Technology, University of Illinois at Urbana-Champaign, Urbana, Illinois, 61802, USA
* E-mail: aluru@illinois.edu
1
In order to compare the mechanical behavior of water filled C60 under different loading conditions (i.e.
point load, and contact compressive loading), we considered two cases of C60 and H2O(1)C60 between
two plates to mimic the contact compressive loading condition. In these types of simulations (contact
compressive loading), two plates (made of graphene) are placed at a small distance of 2 Å from C60 (see
Figure.S1). MD simulation is performed using LAMMPS. To compare the effect of the choice of the
potential, we used two different potentials: 1.AIREBO (for C-C interaction) with water-carbon force-field
developed in our group and 2. ReaxFF potential (for C, H and O).1-3 The initial setup for MD is shown in
Fig. S1. In the case of water filled C60, water is placed in the center of C60. To compress the C60, a strain
increment of 0.0076 is applied to the plates, and then the energy minimization of the system is performed
while fixing the atomic coordinates of the plates in all the directions.
Fig. S1. System setup (top view (left) and side view (right)) for applying compression using two plates.
Strain Energy, (eV)
180
160
Point Load (C60)-DFT
Point Load (H2O(1)@C60)-DFT
140
2-Plate (C60)-AIREBO
2-Plate (H2O(1)@C60)-AIREBO
120
2-Plate (C60)-ReaxFF
2-Plate (H2O(1)@C60)-ReaxFF
0.0
100
0.15
0.3
0.45
0.5
0.15
0.3
0.45
0.6
80
60
40
20
0
0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Compressive Strain
Fig. S2. (Left) Variation of strain energy with compressive strain for C60 and H2O(1)@C60 using the two
plates approach (MD with AIREBO and ReaxFF) and the point load approach (DFT). (Right) Molecular
snapshots of C60 at each strain under compression using the two plates approach (top) with AIREBO
potential and the point load approach using DFT.
Based on the strain energy variation shown in Fig. S2, all three methods (point load, two plates with
AIREBO and two plates with Reaxff) are comparable up to around a strain of 0.25. Beyond that, the energy
variation depends on the method employed. It is notable that in the case of compression with two plates,
2
there is an extensive deformation distributed on the surface of sphere compared to the point load case (see,
for example, snapshots at 0.45 strain in Fig.S2). Under point load the applied load is more concentrated on
the middle of the sphere at 0.45 and 0.6 strains. It should be noted that the point load also deforms the
surface area but not as significantly as the two plates case (Fig. S3).
Fig. S3. Displacement vector of C60 at each applied strain with respect to its initial position.
Based on our observations, using the point load within DFT scheme can reasonably capture the mechanical
properties of water inside C60 due to following reasons: First, the potentials available for MD simulations
are not verified for mechanical properties, especially for interaction between water and C60. The watercarbon interaction force-field cannot capture the bond-formation and the bond-breaking which are key
factors in fracture of water-filled buckyball under deformation.2 In the case of ReaxFF potential, this has
not been adequately studied to understand if it can capture the interaction between buckyball and water,
and mechanical properties under various loading conditions. Second, it is difficult to perform loading/strain
with two plates using DFT because of the limited number of atoms DFT can deal with (the two plates
includes about 433 atoms). Finally, it is not clear if the tensile case can be studied easily using two plates,
hence the results in the main text contain only the point load case.
References
1
2
3
S. J. Stuart, A. B. Tutein, and J. A. Harrison, The Journal of Chemical Physics 112, 6472 (2000).
Y. Wu and N. R. Aluru, The Journal of Physical Chemistry B 117, 8802 (2013).
K. Chenoweth, A. C. van Duin, and W. A. Goddard, The Journal of Physical Chemistry A 112,
1040 (2008).
3