[About the Foresight Institute] [Index of Foresight Conferences] | [Index of Papers] |

by

Guanghua Gao,
Tahir Cagin^{*}, and William A. Goddard III

Materials and Process Simulation Center, Beckman Institute (139-74)

California Institute of Technology, Pasadena, California 91125

This is a paper of a talk given at the Fifth Foresight Conference on Molecular Nanotechnology.

It has been submitted for publication in the special Conference issue of

In this paper, we present extensive molecular mechanics and molecular dynamics studies on the energy, structure, mechanical and vibrational properties of single wall carbon nanotubes. In our study we employed an accurate interaction potential derived from quantum mechanics. The simulation studies were primarily carried out using the MPSim Program.

We explored the stability domains of circular and collapsed cross-section
structures of armchair (n,n), zigzag (n,0), and
chiral (2n,n) isolated single-walled carbon nanotubes (SWNTs)
up to a circular cross section radius of 170 Å.
We have found three different stability regions
based on circular cross section radius. Below 10 Å radius
only the circular cross section tubules are stable.
Between 10 Å and 30 Å both
circular and collapsed forms are possible, however the circular cross section
SWNTs are energetically favorable. But, beyond 30 Å (cross-over radius)
the collapsed form becomes
favorable for all three types of SWNTs.
We report the behavior of the SWNTs with radii close to the cross-over radius
[(45,45),
(80,0), (70,35)] under uniaxial compressive and tensile loads.
Using classical thin plane approximation and variation of
strain energy as a function of curvature, we calculated the bending modulus
of the SWNTs. The calculated bending moduli are
k _{(n,n)} = 963.44(GPa),
k _{(n,0)} = 911.64(GPa), and
k_{(2n,n)} = 935.48(GPa).
We also calculated the interlayer spacing between the opposite sides of
the tubes and found
d_{(n,n)} = 3.38(Å),
d_{(2n,n)} = 3.39(Å),
and d_{(n,0)} = 3.41(Å)

Using an enthalpy optimization method,
we have determined the crystal structure and Young's modulus of
(10,10) armchair, (17,0) zig-zag and
(12,6) chiral forms [which have similar diameter as (10,10)]. They all
pack in a triangular pattern in 2-D. Calculated lattice parameters are
a_{(10,10)} = 16.78 Å, a_{(17,0)} = 16.52 Å and
a_{(12,6)} = 16.52 Å.
Using the second derivatives of potential we calculated the Young modulus
along the tube axis and found
Y_{(10,10)} = 640.30 GPa,
Y_{(17,0)} = 648.43 GPa, and
Y_{(12,6)} = 673.94 GPa.
Using the optimized structures of (10,10), (12,6) and (17,0), we determined
the vibrational modes and frequencies.
Here, we report the highest in plane mode, compression mode, breathing mode,
shearing mode and relevant cyclop mode frequencies.

Carbon nanotubes were discovered in 1991 by Iijima of NEC Corporation [1]. Since then, efforts in synthesis, characterization and theoretical investigation on nanotubes has grown exponentially. This is mostly due to their perceived novel mechanical and electronic properties and their tremendous potential for future technological applications. In 1993, the simplest kind of carbon nanotubes, single walled carbon nanotubes, were discovered independently by Iijima group [2] and an IBM team headed by Bethune [3]. These SWNTs can be regarded as a rolled-up graphite sheet in the cylindrical form. Some specific defect-free forms of these SWNT show remarkable mechanical properties and metalic behavior [4]. These materials present tremendous potential as components for use in nano-electronic and nano-mechanical device applications or as structural elements in various devices.

Thess and coworkers [4] later produced
crystalline "ropes" of metallic carbon nanotubes with 100 to 500
SWNTs bundled into a two-dimensional triangular lattice.
These tightly bundled linear "ropes" are expected to have remarkable
mechanical properties, as well as superior electronic and magnetic properties.
Various levels of studies performed on the properties of the SWNTs,
including use of classical molecular mechanics (MM), lattice dynamics (LD),
molecular dynamics(MD), tight binding and ab initio level Quantum Mechanical
(QM) methods [5-18].
Robertson* et. al.* carried first studies on the structure and energetics of
isolated small diameter [13].
More recent theoretical studies include, behavior of nanotubes under high
rate tensile strain and fracture [14], size dependent resonance raman
scattering [15] in nanotubes, mechanical behavior of SWNT under
compression using MD [16], and a pioneering study on the
structural flexibility of carbon nanotubes using HREM and atomistic
simulations addressing the role of van der Waals radii reported by
Tersoff and coworkers [17-18].

In this paper, we present a detailed study of the energetics, structures, and mechanical properties of single walled carbon nanotubes with different radius and chirality (armchair (n,n), chiral (2n,n), and zigzag (n,0)). We used an accurate force field, derived through quantum mechanical calculations, to represent the interactions between the carbon atoms [19]. These interaction potentials were used earlier in studying structure, mechanical and vibrational properties of graphite, various fullerenes and intercalated compounds of fullerenes [20] and nanotubes [21]. In our studies, we employed classical molecular dynamics and molecular mechanics methods as implemented in MPSim, (a massively parallel program for materials simulations) program [22]. MD runs are performed to anneal the structures, whereas molecular mechanics, energy and/or enthalpy minimization, are applied at the end of annealing cycle to obtain the final optimized structures. Using the analytical second derivatives of the potential energy, we also calculate the vibrational modes and frequencies of three kinds of nanotube bundles, (10,10) arm chair, (17,0) zig-zag and (12,6) chiral. These tubes have comparable cross section diameters, and are among the easiest to make.

In order to assess mechanical stability of various SWNTs, we created
three chiral forms [(n,n) armchair,
(n,0) zigzag, and (2n,n)] with various diameters.
For each form, we studied two sets of initial structures, perfect
circular cross section and elongated or collapsed cross section.
For the collapsed structures, the opposte walls in the middle section are
within van der Waals distance and the shape at the two ends is close to
circular diameter of D ~ 10.7 Å.
To mimic long isolated nanotube,
we imposed a periodic boundary condition in c-direction,(tube direction).
To eliminate the inter-tube interactions,
we set the cell parameter *a* and *b* as 50 times of the circular
tube diameter.
Energy and structural optimizations
were carried out using MPSim.
Figure 1. is the strain energy per carbon atom
versus radius of its circular form.
[See also, Suplementary Material]
We put the two sets (collapsed and circular)
with three chiral forms ( (n,n) armchair,
(n,0) zigzag, and (2n,n) ) on the same plot. For all three forms, there are
regions associated with two transition radii
(R_{1} and R_{2}). For tubes with "circular" radius
smaller than R_{1}, only the circular form is possible, the collapsed
initial structure transformed to the circular form during the structural
optimization. For tubes with 'circular' radius between
R_{1} and R_{2}, there are two stable structures with the
circular structure energetically more stable. For tubes with "circular"
radius larger than R_{2}, collapsed form becomes energetically favored
while circular form becomes metastable. The structures and radii
of the first transition are:

- (n,n) armchair, R
_{1}is between 10.77 A of (16,16) and 11.44 of (17,17). - (2n,n) chiral, R
_{1}is between 10.28 A of (20,10) and 11.31 of (22,11). - (n,0) zigzag, R
_{1}is between 10.49 A of (27,0) and 10.88 of (28,0).

- (n,n) armchair, R
_{2}is between 29.62 A of (45,45) and 30.30 of (46,46). - (2n,n) chiral, R
_{2}is between 29.82 A of (58,29) and 30.85 of (60,30). - (n,0) zigzag, R
_{2}is between 29.93 A of (77,0) and 30.32 of (78,0).

Looking at the collapsed structures of various radii along tube axis, we found that they all have two circular (or elliptical) ends of diameter D ~ 10.5 Å and flat middle section. The inner wall distances in the flat region are close to 3.4 Å, which is very similar to the inter layer spacing of adjacent sheets in graphite crystal. The end sections are highly strained compared to the circular form, thus cost energy. In addition to zero strain energy, the flat region is further stabilized by inter layer van der Waals attractions. The relative strength of these two opposite forces dictate the two structural transformations. Shown in Figure 2 are the optimized structures of armchair (n, n), tubes with collapsed initial structures viewed along c-axis. The zigzag (n,0) and chiral (2n, n) tubes are the same.

Based on the optimized structures and their energies of the circular form,
we can model the
basic energetics by approximating the tube as a membrane
with a curvature 1/R and bending modulus of k [23].
Assuming *a* as the thickness of
tube wall, the elastic energy
stored in a slab of width *L*, is given by
The per atom energy can be written as

where N is the number of carbon atoms per slab and
E_{o
} is energy per
carbon atom for tubes with 1/R ~ 0,
i.e. flat sheets.
Considering _{}
the number of carbon atoms per
unit area of tube wall, we have

Setting *a* as the spacing between two graphite sheets, 3.335(Å),
R_{0} = 1.410(Å)
as the C-C bond distance, we obtained
k_{(n,n)} = 963.44(GPa), k_{(n,0)} = 911.64(GPa), and
k_{(2n,n)} = 935.48(GPa).
These results are plotted against the
theoretical estimates in Figure 3.

The bending modulus of sheets with different chirality suggest that the transition radius depends on the chirality, with (2n, n) transition radius larger thab that of (n, 0) zigzag, but smaller than that of (n, n) armchair. That is what we expected, because the higher the bending modulus, the higher the strain energy. The (n, n) armchair has higher binding energy than that of (2n, n), while the (n, 0) zigzag has lower binding energy than that of (2n, n). However, by examining the collapsed structures closely, we also found different interlayer stacking in the collapsed region. Figure 4 is the side views of the two attracting layers for three cases. The interlayer stacking patterns are different due to different chirality.

**Figure 4**: Side views of interlayer stacking patterns of:
a) (30,30) armchair;
b) (50,25) chiral and
c) (50,0) zigzag SWNTs.

The inter layer distances also differ slightly, with
d_{(n,n)} = 3.38(Å),
d_{(2n,n)} = 3.39(Å),
and d_{(n,0)} = 3.41(Å).
Energetically,
inter-layer attraction of armchair is the best with per atom energy
E = 0.7336 (kcal/mol), and stacking of the opposite walls is almost identical
to the graphite stacking. The inter-layer attraction in the zigzag form is
the worst with per atom energy, E = 0.7439 (kcal/mol), since the carbon
atoms on different layers are lined up on top of each other. The attraction
energy per atom for the collapsed (2n,n) chiral nanotube is in between the two.
Overall, the two factors (bending modulus and van der Waals attraction)
cancels out for different chirality, so that in terms of transition
radius and the cross-over radius, the size of the circular
tube (radius) is the dominant factor in deciding the stable forms.

In order to assess the tensile and compressive strength of the SWNTs, we also
performed a series of compressive and tensile loading experiments by varying
the *c* lattice parameter, along tube axis. With respect to the optimum *c*
lattice parameter, c_{o}, the strain can be defined as

At cross-over radius, we calculated the energies of collapsed and circular tubes of (45,45), (80,0), and (70,35) under compression and tensions along tube axis. In Figure 5a, the armchair, Figure 5b, the zigzag, and Figure 5c, the chiral forms, we observe a nonlinear dependence on the strain.

Figure 5a:
Strain Energy vs Strain for (45,45) Armchair SWNT

Figure 5b: Strain Energy vs Strain for (80,0) Zigzag SWNT

Figure 5c: Strain Energy vs Strain for (70,35) Chiral SWNT

The tubes are softer under compression due to buckling effect. We present views along the tube axis and perpendicular to the tube axis in Figure 6. Based on those energy strain curves, we also calculated "elastic constants", second the derivative of energy with respect to applied strain, e, along the tube axis as

Among various conformations, the (10,10) armchair SWNT is the easiest to make.
We studied the mechanical properties of its bulk phase (10,10) tube bundles.
We also calculated the bulk properties of (17,0) zigzag and (12,6) chiral
tubes, with cross section radii close to that of (10,10) armchair tube.
In the bulk phase we determine the specific packing, density, lattice
parameters.
Molecular Dynamics and Molecular Mechanics
studies led to a triangular packing as the most stable form for all
three forms.
The triangular lattice parameter for (10,10) armchair is a = 16.78 Å
with a density d = 1.33 (g/cm^{3}). For the (17,0) zigzag, they are
a = 16.52 Å, and d = 1.34 (g/cm^{3}).
For the chiral form (12,6), they are a = 16.52 Å and
d = 1.40 (g/cm^{3}).
More importantly, we determined the Young's modulus along the tube axis
for triangular-packed SWNTs
using the second derivatives of the potential energy.
They are Y = 640.30 GPa,
Y = 648.43 GPa, and Y = 673.49 GPa, respectively. Normalized to carbon sheet
these values are within a few percent of the graphite bulk value.

Using the second derivatives of the potential energy surface we calculated vibrational modes of a series of (n,n) tubes. The uniform compression and highest inplane modes are listed in Table 2 for n=7 to 13.

(7,7) | (8,8) | (9,9) | (12,6) | (17,0) | (10,10) | (11,11) | (12,12) | (13,13) | |

Uniform Comp. | 261 | 231 | 207 | 202 | 188 | 186 | 168 | 152 | 138 |

Highest Mode | 1583 | 1584 | 1584 | 1585 | 1586 | 1584 | 1584 | 1584 | 1584 |

Among (n,n) tubes we focused on (10,10) tube crystal. We also calculated
the vibrational frequencies and modes for the
zigzag (17,0) and chiral (12,6) tube crystals due to their comparable tube
radius to (10,10).
We listed the breathing, shearing and cyclop modes for these
equivalent radius nanotubes in Table 3.
These
results may be used to differentiate chiral tubes with comparable diameters.
The uniform compression mode is also shown in Figure 7a occurs at
186 cm^{-1}
for (10,10), which is exactly the same as the experimental frequency[24]. In
the Table 3,
*B* denotes breathing mode as displayed in
Figure 7b,*S* stands for shearing mode as in
Figure 7c, and *C* stands for cyclopes as in
Figure 7d.

**Figure 7**: Selected vibrational Modes for (10,10) tube, a) Top left,
uniform compression mode; b) Bottom left, shearing along tube mode;
c) Top right, breathing mode
d) Bottom right, Cyclop mode.
Click on a Figure to Animate.

We presented a detailed study of structure, energetics and mechanical properties of SWNTs of varying size and chirality. We determined the stability domains for circular and collapsed cross section SWNTs. We reported the uniaxial strain-energy behavior of the SWNTs with cross-over radius in the range of -8% strain to 8% strain. The determined structure and lattice parameters for closed packed (10,10) like nanotubes are in close agreement with observations. We also determined vibrational modes and frequencies of bulk and isolated nanotubes using a highly accurate classical force field.

This research was funded by NASA (computational nanotechnology) and by NSF-GCAG (ASC 92-100368). The facilities of the MSC are also supported by grants from NSF (CHE 95-22179), DOE-ASCI, ARO/DURIP, BP Chemical, ARO-MURI, Exxon, Seiko-Epson, Asahi Chemical, Beckman Institute, Chevron Petroleum Technology Co., Chevron Chemical Co., NASA/JPL, ONR, Avery Dennison, and Chevron Research Technology Co.

- Iijima, S. (1991) "Helical microtubules of graphitic carbon,"
*Nature***354**, pages 56-58. - Iijima, S. and Ichlhashi, T. (1993)
"Single-shell carbon nanotubes of 1-nm diameter,'
*Nature*,**363**pages 603-605. - Bethune, D.S., Kiang, C. H., Devries, M. S., Gorman, G.,Savoy, R.,
Vazquez, J., Beyers, R. (1993),
"Cobalt-catalyzed growth of carbon nanotubes with single-atomic-layer walls,"
*Nature***363**, pages 605-607. - Thess, A., Lee, R., Nikolaev, P., Dai, H., Petit, P.,
Robert, J., Xu, C., Lee, Y. H., Kim, S. G., Rinzler, A. G., Colbert,
D. T., Scuseria, G. E., Tomanek, D., Fisher, J. E., and Smalley, R.
E. (1996)
"Crystalline ropes of metallic carbon nanotubes,"
*Science***273**, pages 483-487. - Krotov, Y.A., Lee D.-H., and Louie, S.G., (1997)
"Low-energy properties of (n, n) carbon nanotubes,"
*Phys. Rev. Lett.***78(22)**, pages 4245-4248. - Tuzun, R. E., Noid, D. W., Sumpter, B. G., and Merkle, R. C., (1996)
"Dynamics of fluid-flow inside carbon nanotubes,"
*Nanotechnology*,**7(3)**pages 241-246 - Ihara, S. Itoh, S., (1996)
"Molecular dynamics study of toroidal and helically coiled forms of graphitic
carbon,"
*Surf. Rev. Lett.***3(1)**pages 827-834; - Chopra, N. G., Benedict, L. X., Crespi, V. H., Cohen, M.L., Louie, S. G.
and Zettl, A. (1996)
"Fully collapsed carbon nanotubes,"
*Nature***377**, pages 135-138; - Menon, M, Richter, E. and Subbaswamy, K.R., (1996)
"Structural and vibrational properties of fullerenes and nanotubes in a
nonorthogonal tight-binding scheme, "
*J. Chem. Phys.***104(15)**pages 5875-5882. - Hamada, N., Sawada, S. and Oshiyama, A. (1992)
"New one-dimensional conductors - graphitic microtubules,"
*Phys. Rev. Lett.***78(10)**, pages 1579-1581; - Saito, R., Fujita, M., Dresselhaus, G. and
Dresselhaus, M.S., (1992)
"Electronic-structure of chiral graphene tubules,"
*Appl. Phys. Lett.***60**, pages 2204-2206; - Blase, X.,
Benedict, L.X., Shirley, E.L., and Louie, S.G.,
"Hybridization effects and metallicity in small radius carbon nanotubes,"
(1994)
*Phys. Rev. Lett.***72(12)**, pages 1878-1881. - Robertson, D. H., Brenner, D. W. and Mintmire, J. W. (1992)
"Energetics of Nanoscale graphitic tubules,"
*Phys. Rev. B***45 (21)**pages 12592-12595. - Yakobson, B.I., Campbell, M. P., Brabec, C. J. and Bernholc, J. (1997).
"High strain rate fracture and c-chain unraveling in Carbon Nanotubes",
(1997)
*Comp. Mater. Sci.***8(5)**, pp 341-348. - Richter, E. and Subbaswamy, K. R. (1997)
"Theory of size dependent resonance raman scattering from carbon nanotubes,"
*Phys. Rev. Lett.***79(14)**pages 2738-41. - Iijima, S., Brabec, C., Maiti, A. and Bernholc, J. (1996)
"Structural flexibility of carbon nanotubes,"
*J. Chem. Phys.***104(5)**pages 2089-2092 - Tersoff, J. and Ruoff, R. S. (1994)
"Structural properties of carbon nanotube crystals"
*Phys. Rev. Lett.***73(5)**pages 676-679 Article - Ruoff, R. S., Tersoff, J., Lorents, D. C. Subramoney, S., and
Chan, B. (1993)
"Radial deformation of carbon nanotubes by van der Waals Forces"
*Nature***364**, pages 514-516 - Guo, Y. J. (1992) Ph.D. Dissertation, California Institute of Technology.
- Guo, Y. J., Karasawa, N. and Goddard III, W. A. (1991)
"Prediction of Fullerene Packing in C
_{60}and C_{70}Crystals,"*Nature***351**, pages 464-467. - Gao, G., Cagin, T. and Goddard III, W.A. "Position of K atoms
in Doped Single Walled Carbon Nanotube Crystals",
*Phys. Rev. Lett.***80**, 5056 (1998). - Lim, K.T., Brunett, S., Iotov, M., McClurg, R.B., Vaidehi, N.,
Dasgupta, S., Taylor, S. and Goddard III, W.A. (1997)
"Molecular Dynamics for very large systems on massively-parallel computers
- MPSim Program,"
*J. Comp. Chem.***18**, pages 501-521. - Landau, L.D. and Lifshitz, E.M.,
*Elasticity Theory*(Pergamon, Oxford, 1986). - Rao, A. M., Richter, E., Bandow, S., Chase, B., Eklund, P.
C., Williams, K. A., Fang, S., Subbaswamy, K. R., Menon, M., Thess,
A., Smalley, R. E., Dresselhaus, G., and Dresselhaus, M. S. (1997)
"Diameter-selective raman scattering from vibrational modes in carbon
nanotubes,"
*Science***275**, pages 187-191.

Energies as a function of radii (R up to 170 A) for (n,n), (n,0), (2n,n) tubes

| Index of Papers | Conference
Home Page | Index of Foresight Institute
Conferences |

| About the Foresight Institute
| Foresight Institute Home Page |

Foresight materials on the Web are ©1986-1997 Foresight Institute. All rights reserved.

Last updated 30September97. The URL of this document is: http://www.foresight.org/Conferences/MNT05/Papers/Template/index.html

Send requests for information about Foresight Institute activities and membership to inform@foresight.org.

Send comments and questions about material on this web site and reports of errors to webmaster@foresight.org.