Fabrication of Carbon Nanoscrolls from Monolayer · PDF fileFabrication of Carbon Nanoscrolls from Monolayer Graphene ... multiwall CNTs. ... shell-composite nanostructures by the

2010

full papers

Carbon nanoscrolls

Fabrication of Carbon Nanoscrolls from Monolayer Graphene
Dan Xia, Qingzhong Xue,* Jie Xie, Huijuan Chen, Cheng Lv, Flemming Besenbacher, and Mingdong Dong*
A simple way of synthesizing carbon nanotube (CNT)/graphene (GN) nanoscroll core/shell nanostructures is demonstrated using molecular dynamics (MD) simulations. The simulations show that GN sheets can fully self-scroll onto CNTs when the CNT radius is larger than a threshold of about 10 Å, forming a stable core/shell structure. Increasing the length of the GN sheet results in multilayered carbon nanoscroll (CNS) shells that exhibit a tubular structure similar to that of multiwall CNTs. The distances between the CNT and the GN wall or adjacent GN walls are about 3.4 Å. It is found that the van der Waals force plays an important role in the formation of the CNT/GN nanoscroll core/shell-composite nanostructures. However, the chirality of the CNT and the GN sheet does not affect the self-scrolling process, which thus provides a simple way of controlling the chirality and physical properties of the resulting core/shell structure. It is expected that this preparation method of CNT/GN nanoscroll core/shell composites will lead to further development of a broad new class of carbon/carbon core/shell composites with enhanced properties and even introduce new functionalities to composite materials.

1. Introduction

Carbon nanotubes (CNTs) and graphene (GN) are

two of the most promising nanomaterials in the carbon

family due to their ideal 1- and 2D structures. Since their

discovery in 1991 [ 1 ] and 2004, [ 2 ] respectively, they have

© 2010 Wiley-VCH Vewileyonlinelibrary.com

DOI: 10.1002/smll.201000646

D. Xia , Prof. Q. Z. Xue , J. Xie , H. J. Chen , C. Lv College of Physics Science and TechnologyChina University of PetroleumDongying, Shandong 257061, P. R. China E-mail: [email protected]

F. Besenbacher , M. Dong Interdisciplinary Nanoscience Center (iNANO) and Department of Physics and AstronomyUniversity of AarhusDK-8000 Aarhus C, DenmarkE-mail: [email protected]. dk

attracted tremendous interest because of their unique prop-

erties [ 3–10 ] and promising applications. [ 11–18 ] The properties

of CNTs have been investigated in great detail during the

past few years and research into GN is rapidly catching up

with that of CNTs. Recently, folding and curling GN sheets

have been intensively researched and have been proposed

for the generation of novel nanostructures. [ 19–21 ] As a result,

another interesting carbon nanomaterial, called a carbon

nanoscroll (CNS), has arisen. The CNS structure has been

studied by rolling up a single GN sheet with varying diam-

eter and chirality, depending on the sheet size and roll direc-

tion. [ 22 , 23 ] CNSs are remarkable structures that share some

of the rich mechanical and electronic properties exhibited

by CNTs and GN but are also expected to exhibit novel

features. Some theoretical studies have predicted unusual

electronic [ 23 ] and optical properties [ 24 ] of CNSs because

of their unique topology. Unlike CNTs, the CNS diameter

can be easily varied because it is not a closed topological

structure. These properties can be exploited for a variety

rlag GmbH & Co. KGaA, Weinheim small 2010, 6, No.18, 2010–2019

Carbon Nanoscrolls from Graphene

Figure 1 . Snapshots of a GN sheet with a length of 53.379 Å self-scrolling onto a (8,8) CNT core at time intervals from 0 to 500 ps.

0 ps 1 ps 100 ps

150 ps 175 ps 500 ps

of technological applications, such as chemical doping, [ 22 ]

hydrogen storage, [ 25 ] and nanoactuators in nanomechanical

devices. [ 22 , 26 ]

There are several methods for fabricating CNSs, including

arc discharge, [ 27 ] high-energy ball milling of GN, [ 28 ] and the

chemical route. [ 29–31 ] However, these methods are not able

to produce high-purity, high-quality CNSs. Recently, Xu et al.

reported a simple and effi cient method for the in situ fabrica-

tion of high-quality CNSs and their direct incorporation into

devices. [ 32 ] Current routes for the synthesis of CNS structures

depend on the self-assembly of exfoliated graphite sheets

and lack suffi cient control over CNS diameter and chirality,

thus requiring a postsynthesis sorting processes. Therefore,

the development of new and alternative methods for synthe-

sizing CNSs with well defi ned diameter and chirality repre-

sents a great challenge in today’s nanotechnology.

By using molecular dynamics (MD) simulations, we inves-

tigate a simple method of synthesizing novel CNT/CNS core/

shell-composite nanostructures by the rolling up of single

GN sheets induced by CNTs. There are several reports of

core/shell structures with excellent properties [ 33–36 ] that make

them promising candidates for applications such as nanoelec-

tronic devices [ 35 ] or integrated circuits. [ 36 ] In this work, the

surface-adsorption stress of the CNTs, which results from the

van der Waals (vdW) force, is used to bend GN sheets to roll

up and cover the CNT surface. The process consists of two

basic steps: 1) placing GN sheets in close proximity to CNTs

and 2) the bending and rolling up of GN sheets into CNSs by

the CNTs’ surface stress. For a given CNT, the CNT size and

GN chirality determine the CNS diameter and chirality. Com-

pared with the stochastic growth process, this method con-

trols the diameter and chirality of the CNSs in a deterministic

manner. The CNT/CNS core/shell structures formed by these

two novel carbon materials have excellent properties, such as

high carrier mobility and high mechanical strength, and could

be used as microcircuit interconnects, nanoelectronic devices,

or nanosensors.

2. Results and Discussion

The self-scrolling of CNT/GN core/shell nanostructures

was simulated using MD. Figure 1 shows snapshots of the

interaction between an (8,8) CNT and a GN sheet with a

length of 53.379 Å. When the GN is placed beside the (8,8)

CNT, the CNT and the GN approach each other because of

attractive forces. During the approach, the GN atoms, which

are closer to the CNT than the other atoms, move faster

towards it and fi nally attach onto the CNT surface, as shown

in Figure 1 at t = 1 ps, because the vdW force acting on the

carbon atoms of the GN is stronger the closer they are to the

CNT surface. After the attachment of the nearest GN carbon

atoms onto the CNT, the GN begins to roll and gradually

wraps around the whole CNT (see Videos S1 and S2 in the

Supporting Information). At t = 175 ps, the wrapping is com-

plete and a tubular CNS is formed. After t = 175 ps, the CNT

and GN remain in a relatively stable state so the confi gura-

tions of the composite structure at t = 175 ps and t = 500 ps

appear more or less identical.

© 2010 Wiley-VCH Verlag GmbHsmall 2010, 6, No.18, 2010–2019

The interaction energy, refl ecting the adhesion between

the CNT and the GN, is defi ned as [ 33 ]

Einteraction � Etotal − (EGN + ECNT) ,

(1)

where Etotal is the energy of the system including the GN and

the CNT, EGN is the energy of the single GN sheet without the

CNT, and ECNT is the energy of the single CNT without the

GN. The deformation energy is defi ned as the difference

between the carbon structure’s initial energy and its energy

determined after a certain simulation step. In order to study

the self-scrolling in detail, we calculate the interaction energy

and deformation energy between the GN and CNT, as shown

in Figure 2 . From Figure 2 a, we can observe that the interac-

tion energy between the GN and the CNT shows a rapid ini-

tial increase, which is considerably slowed down at t ≈ 20 ps.

Finally, at t ≈ 175 ps, the interaction energy saturates. This indi-

cates that the system reaches a stable state after the GN has

completely wrapped onto the CNT. As shown in Figure 2 b,

the deformation energy of the GN shows a similar time

dependence as the interaction energy between the GN and

the CNT. In the case of the CNT, however, the deformation

2011 & Co. KGaA, Weinheim www.small-journal.com

D. Xia et al.

2012

full papers

Figure 2 . Change of the interaction energy and deformation energies with simulation time: a) the interaction energy between GN and CNT; b) the deformation energies of GN and CNT.

0 100 200 300 400 500

0

-200

-400

-600

-800

Inte

ract

ion

Ene

rgy

(Kca

l/mol

)

Simulation Time (ps)

(a) Interaction energy

0 100 200 300 400 500

550

600

650

700

750

Def

orm

atio

n E

ner

gy (

Kca

l/mol

)

Simulation Time (ps)

GN CNT (8,8)

(b) Deformation energy

energy saturates directly after the fi rst rapid increase and

only exhibits some minor fl uctuations for the rest of the simu-

lation. This indicates that the CNT reaches a relatively stable

state already at t ≈ 20 ps.

The process of self-scrolling a GN sheet onto a CNT

depends on the diameter of the CNT. When the diameter of

the CNT is < 10 Å, the CNT cannot induce the GN to wrap

it completely. As a result, we chose the (8,8) CNT for our

simulations, which is the least armchair-shaped CNT that can

induce a GN sheet even with a very long length to completely

wrap onto the CNT.

It is known that the chirality of CNTs and GN sheets

has a signifi cant effect on the properties of the CNTs and

the GN. A CNT is of metallic type when it exhibits armchair

chirality, whereas the CNT can be semiconducting type or

semimetallic type when its chirality is zigzag or armchair-

like. [ 37 ] Also, the chirality of a GN sheet can change its type

from metallic to semiconducting. [ 38 ] To investigate the effect

www.small-journal.com © 2010 Wiley-VCH Verlag Gm

Figure 3 . Schematics of the fi ve different chiral nanotubes with similar pin the MD simulations.

(8, 8) (9, 7) (10, 6) (12, 3)

of chirality on the mechanical properties of the composite

system, we simulated an armchair GN sheet interacting with

CNTs of different chiralities, as well as an (8,8) armchair

CNT interacting with GN sheets with different chiralities.

Five types of CNTs with different chiral angles, θ , ranging

from 0 to 30 ° were generated, as shown in Figure 3 . The cor-

responding chiral angle θ and diameter D n of the CNTs with

( n , m ) indices could be determined by using the rolling GN

model: [ 39 ]

2 � arctan

( √3m

2n + m

);

Dn �√

3

Bb√

(n2 + m2 + nm) (0 ≤ m ≤ n),

(2)

bH & Co. KGaA, Weinh

arameters utilized

(14, 0)

where b is the length of the C–C bond

(0.142 nm). The total number of atoms,

diameter, and length of each chiral nano-

tube are given in Table 1 . Figure 4 shows

the GN sheets with different chiralities

and the CNT/CNS core/shell-composite

nano structures formed after the MD sim-

ulations. The GNs are fi nite-size sheets

having special orthotropic behaviour and

positive Poisson ratios. [ 40 , 41 ] From Figure 4 ,

we can observe that different CNSs are

formed by rolling up GN sheets of different

chiralities and the CNSs formed exhibit the

single-wall CNT (SWCNT) structure with

different chiralities. To clarify the infl uence

of the chirality on the adhesion of this

CNT/CNS core/shell-composite structure,

we determine the saturation interaction

energies between the armchair GN and

the CNTs with different chiralities as well

as the saturation interaction energies per

unit surface between the (8,8) armchair

CNT and the GN sheets with different chi-

ralities, as shown in Figure 5 . From Figure 5 a,

eim small 2010, 6, No.18, 2010–2019



Table 1. Total number of atoms, diameter, and length of each chiral nanotube utilized in MD simulations.

Type of SWNTs

H atoms C atoms CNT diameter [Å]

CNT length [Å]

Chiral angle θ [deg]

(8,8) SWNT

armchair

32 672 10.85 51.65 30.00

(9,7) SWNT 32 672 10.88 51.54 25.87

(10,6) SWNT 32 672 10.96 51.12 21.79

(12,3) SWNT 30 672 10.76 52.06 10.89

(14,0) SWNT

zigzag

28 672 10.96 51.12 0.00

Figure 4 . Snapshots of GN sheets with different chiralities and the core/shCNT (the red highlighted structures represent the formed CNSs with differe

Figure 5 . a) Saturation interaction energies between the armchair GN shinteraction energies between the (8,8) CNTs and the different chiral GN sh

(8,8) (9,7) (10,6) (12,3) (14,0)0

-100

-200

-700

-800

-900

Inte

ract

ion

Ene

rgy

(Kca

l/mol

)

Chirality

(a)

we can observe that the CNT chirality has a negligible infl u-

ence on the adhesion. Similarly, the chirality of the GN sheets

affects the adhesion between the GN and the CNTs only

slightly, as shown in Figure 5 b.

The information on the interfacial region of the fi nal

structure can be characterized by the concentration profi le

of the combination consisting of the CNT core and the GN

shell. The concentration profi le is calculated for 3D peri-

odic structures by computing the atom-density profi le within

evenly spaced slices parallel to the bc , ca , and ab planes. In

practice, this is equivalent to taking the a , b , and c compo-

nents of the fractional coordinates of each atom and inde-

pendently generating a plot for each component. Figure 6 a,b


ell-composite nanostructures formed by the vdW force between GN and nt chiralities).

eets with a length of 53.379 Å and different chiral CNTs. b) Saturation eets.

0 5 10 15 20 25 30-0.38

-0.40

-0.42

-0.44

Inte

ract

ion

En

rgy

per

Un

it S

urfa

ce (

Kca

l/mol

•Å²)

Chiral Angle (°)

(b)

D. Xia et al.

2014 www.small-journal.com © 2010 Wiley-VCH Verlag Gm

full papers

Table 2 . Distances d1 to d4 between the CNT and the GN sheet and the a

Different chiral CNTs and the armchair GNs

d1 d2 d3 d4 da

(8,8)

armchair

3.993 4.054 3.525 3.044 3.521

(9,7) 2.972 3.544 2.986 3.484 3.147

(10,6) 3.016 3.505 3.509 3.509 3.345

(12,3) 3.016 2.521 3.525 3.474 3.338

(14,0) 3.525 3.001 3.503 3.002 3.343

Figure 6 . Concentration profi le of the fi nal structure including the (14,0) CNT and armchair GN in a) the X direction and b) the Y direction. c) Snapshot of the composite structure.

62 64 66 68 70 72 74 76 78 80 820

5

10

15

20

25

30C

(X

)

X (Å)

GN CNT (14,0)

d1 d2

(a)

64 66 68 70 72 74 76 78 80 82 840

5

10

15

20

25

C (

Y)

Y (Å)

GN CNT (14,0)

d3 d4

(b)

(c)

shows the concentration profi le of the fi nal structure of the

(14,0) zigzag CNT and the armchair GN sheet in the X and

Y directions, respectively. Here, we defi ned four distances, d1

to d4 , as shown in Figure 6 c. As marked in Figure 6 a,b, d1 is

3.525 Å, d2 is 3.001 Å, d3 is 3.503 Å, and d4 is 3.002 Å. All

of these distances are about or less than 3.4 Å, which is the

shortest distance of the graphite layer, and have thus almost

entered the strong-adhesive-binding region of the chemical

bond. The interaction energy between the armchair GN sheet

and the (14,0) zigzag CNT reaches ≈ − 777 Kcal mol − 1 . These

distances and the interaction energy indicate that the adhe-

sion between the GN sheet and the CNT is so strong that

they can hardly be separated again.

The distances d1 – d4 are determined from the concentra-

tion profi les for both the interaction between the armchair

GN sheet and the CNTs with different chiralities and the

interaction between the (8,8) armchair CNT and the GN

sheets with different chiralities. The results are shown in

Table 2 , which also gives the average distance, da , between

the CNT core and the GN shell. Because of the irregular

shape of the formed CNS in the X direction, the distance d2

has been omitted in the averaging. The average value, da , is

found to vary from 2.960 to 3.521 Å. Most of these values are

less than the shortest distance of the graphite layer, 3.4 Å.

This indicates that the above conclusion is valid for all cases

investigated in this study and that GN wrapping onto CNTs

results in stable core/shell structures with only little infl uence

of the chirality of both the GN sheet and the CNT.

Because the physical properties of the GN sheets and

the CNTs can be controlled by varying their chirality, it is

important to form different kinds of CNT/CNS core/shell

composites with different chiralities. Through this simple

self-scrolling, we can produce different types of heterojuc-

tion materials, including the semiconductor/semiconductor,

semiconductor/metal, and metal/metal types of CNT/CNS

core/shell composites, which are promising candidates for

various applications including nanomechanical devices or

nanocircuits.

In addition, we simulated GN sheets with different

lengths scrolling onto a CNT to investigate the size effect on

the adhesion. We chose six GN sheets with different lengths

of 30.340, 53.379, 74.394, 122.806, 202.664, and 255.127 Å,

respectively. It can be seen from Figure 7 that all the GN

sheets can completely self-scroll onto the CNT, forming

bH & Co. KGaA, Weinheim small 2010, 6, No.18, 2010–2019

verage distance, da .

Different chiral GNs and the (8,8) armchair CNTs

d1 d2 d3 d4 da

Zigzag 3.509 4.034 3.503 3.503 3.505

5 ° 3.509 3.509 3.498 3.503 3.504

10 ° 3.826 3.930 2.516 2.984 3.109

15 ° 3.878 3.509 2.501 2.501 2.960

20 ° 2.984 2.516 3.525 2.997 3.169

25 ° 2.984 2.989 3.529 3.484 3.332


Figure 7 . Size effect of the interaction between an (8,8) CNT and GN sheet with a width of 33.393 Å and different lengths.

30.340 Å 53.379 Å 493.47 Å

122.806 Å 202.664 Å 721.552 Å

a multilayered shell structure around the single-wall core

for sheet lengths > 54 Å. The saturated interaction energies

between the CNT and the GN sheets of various lengths, as

well as their saturated deformation energies are plotted in

Figure 8 . From Figure 8 a, we can observe that the interaction

energy between the CNT and the GN sheets increases rap-

idly until the GN length reaches 53.379 Å, which corresponds

to the length of a GN sheet that completely wraps the CNT

in one turn. A further increase of the GN length leads only

to a slight increase of the interaction energy. For very large

values of the GN length, a slight decrease of the interaction

energy is observed. This can be explained by the increasing

distance between the overlapped GN sheet and the CNT,

leading to a reduced interaction between the overlapped GN

sheet and the CNT. In addition, with the further increase of

the scrolled layer, the π – π interaction of the overlapped parts

decreases the total free energy of the GN sheet, [ 22 ] which also

leads to a somewhat-reduced interaction energy. In Figure 8 b,


Figure 8 . a) Interaction energies between an (8,8) CNT and GN sheets wsheets.

0 50 100 150 200 250-400

-500

-600

-700

-800

-900

Inte

ract

ion

Ene

rgy

(Kca

l/m

ol)

Lengthes of GNs (Å)

(a) Interaction energy

the deformation energy of the CNT remains almost constant,

indicating that the deformation of the CNT is independent

of the length of the GN. The deformation energy of the GN

sheet, however, increases rapidly with the GN length until

the CNT is completely wrapped. For large GN lengths, the

interaction energy decreases again dramatically. Again, this is

caused by the π – π interaction of the overlapped parts, which

decreases the total free energy of the GN.

To clarify the interaction between the long GN sheet (length

of 255.127 Å) and the CNT, we investigated this structure fur-

ther. The molecular model of the CNT/CNS core/shell-com-

posite structure formed by scrolling a GN sheet of 255.127-Å

length onto a (8,8) armchair CNT is shown in Figure 9 a,b in

side and top view, respectively. The concentration profi le of the

structure in the X direction is plotted in Figure 9 c. In a similar

manner to that described above, seven distances, d5 to d11 , are

defi ned that measure the distances between the different layers

of the core/shell structure (see Figure 9 b) and can be obtained


ith various lengths. b) Deformation energies of the CNT and the GN

0 50 100 150 200 250

400

500

600

700

800

Def

orm

atio

n E

ner

gy (

Kca

l/mol

)

Lengthes of GNs (Å)

CNT GN

(b) Deformation energy

D. Xia et al.

2016

full papers

Figure 9 . Molecular model of the CNT/CNS core/shell-composite structure formed by scrolling a GN sheet with a length of 255.127 Å onto an (8,8) armchair CNT in a) side view and b) top view. (c) Concentration profi le of the composite structure.

(a)(b)

130 135 140 145 150 155 160 1650

10

20

30

40

50

d11d10d9

d8d7d6

C (

X)

X (Å)

(8,8) CNT GN-255.127 Å

d5

(c)

from the separation distances between two adjacent peaks in

the concentration profi le given in Figure 9 c. From the concen-

tration-profi le analysis, we can obtain the values of the seven

distances as 3.283, 3.363, 3.363, 3.283, 3.371, 3.283, and 3.442

Å, respectively. Again, all these distances are about or even

less than the shortest distance of the graphite layer, 3.4 Å,

and have thus almost entered the strong-adhesive-binding

region of the chemical bond. Therefore, we can conclude that,

even for the multilayered GN shell, the core/shell structure is

stable. The CNS formed by the scrolled long GN sheet resem-

bles the multiwall CNT and the radius and the chirality of this

kind of “multiwall CNT” can be controlled by modulating

the radius of the CNT and chirality of the GN. This is a huge

improvement compared with stochastic growth, which cannot

control the diameter and chirality of the CNS in a determin-

istic way.


Figure 10 . Size effect of the interaction between a GN sheet of size 33.39

(8, 8) (9, 9) (10, 10)

The chirality of the CNS can be controlled by using dif-

ferent chiral GN sheets to wrap around the CNTs as dis-

cussed above. Furthermore, the diameter of the CNS can also

be controlled since it depends on the CNT diameter. Six types

of armchair CNTs with different diameters were chosen to

study the wrapping of a long GN sheet around the CNTs. The

armchair CNTs under investigation were (8,8), (9,9), (10,10),

(11,11), (12,12), and (13,13) with diameters of 10.85, 12.20,

13.56, 14.92, 16.27, and 17.63 Å, respectively. The MD simula-

tions show that the length of the GN sheet that can wrap these

CNTs in a single turn is 53.379, 57.582, 61.785, 65.988, 70.191,

and 74.394 Å, respectively. From the size-effect study and the

geometry-confi guration analysis, we can obtain the size rela-

tionship between the GN sheet and the CNT roughly as

L � B (D + 2d); D > 10o

A,

(3)


3 Å × 74.394 Å and armchair CNTs with different diameters.

(11, 11) (12, 12) (13, 13)


Figure 11 . a) Snapshot of the CNT put in close proximity to GN (33.393 Å × 53.379 Å) at places i and ii. b) The formed core/shell nanostructure; c) Snapshot of the CNT put in close proximity to GN (33.393 Å × 122.806 Å) at places I, II, and III, d) The core/shell nanostructure formed by the interaction between the CNT put at places I, II, and the GN. e) The core/shell nanostructure formed by the interaction between the CNT put at place III and the GN.

(a)

(b)

(d)(e)

(c)

where L is the length of the GN sheet, D is the diameter of

the CNT, and d is the average distance between the CNT

and the self-scrolled GN sheet. D > 10 Å accounts for the

minimum CNT diameter needed to induce the self-scolling

of the GN sheet (see above). When Equation 3 is satisfi ed, a

core/shell structure with single-layered shell can be produced.

When L > π ( D + 2 d ), there is an overlap of the GN shell, that

is, a multilayered shell is formed.

The interactions between a GN sheet of 74.394 Å in length

and CNTs with different diameters are shown in Figure 10 .

We can observe that the GN sheets can self-scroll onto all the

CNTs. Single-turn wrapping of the GN sheet is obtained for

the (13,13) armchair CNT, whereas smaller CNT diameters

lead to overlapping of the GN sheet.

Moreover, the location of the CNT placed in close prox-

imity to the GN has a great effect on the scrolling process. When

the GN length and CNT diameter size satisfy Equation 3 ,

the CNT can induce the self-scrolling of the GN wherever the

CNT is placed in close proximity to the GN. Figure 11 a shows

the (8,8) CNT induce the self-scrolling of GN (33.393 Å ×

53.379 Å) whether the CNT is put at position i or ii and the

fi nal confi guration formed is shown in Figure 11 b. When L >

π ( D + 2 d ), the GN can self-scroll onto the CNT and only

the perpendicular distance between the CNT and one

end of the GN is shorter than the value L = π ( D + 2 d ).

Figure 11 c shows a snapshot of the (8,8) CNT put in close

proximity to the GN (33.393 Å × 122.806 Å) at places I,

II, and III. The CNT at places I and II can induce the self-

scrolling of GN (as shown in Figure 11 d) but the CNT at

© 2010 Wiley-VCH Verlag Gmbsmall 2010, 6, No.18, 2010–2019

place III cannot induce the self-scrolling of GN like the mul-

tiwall CNT (as shown in Figure 11 e).

3. Conclusions

In summary, we have used MD simulations to study the

self-scrolling process of GN sheets onto CNTs. When the

diameter of the CNT exceeds a threshold of ≈ 10 Å, the CNT

can induce the GN sheet to scroll onto the CNT surface

by vdW interactions, thus forming a stable CNT/CNS core/

shell nanostructure. The simulations show that, by increasing

the length of the GN sheet, multilayered CNS shells can be

obtained that exhibit a tubular structure similar to that of

multiwall CNTs. The diameter and the chirality of the CNS

can be controlled through the dimensions and chirality of the

CNTs and the GN sheets. The chirality of both the CNT and

the GN sheet is found to have no strong infl uence on the self-

scrolling process, which thus enables the fabrication of a core/

shell structure that combines metal/semiconductor, metal/

metal, or semiconductor/semiconductor junctions. Since the

diameter of the CNS can be easily expanded by charge injec-

tion or intercalation, [ 22 ] it can be used as a nanoactuator. In

addition, this CNT/CNS core/shell-composite nanostruc-

ture can be used as nanocircuit interconnect since the cur-

rent density of the CNS can reach values as high as 5 × 10 7

A cm − 2 . [ 32 ] Thus, the results presented here may stimulate

further research to explore both the physical properties and

additional applications of CNT/CNS core/shell nanostruc-

tures and the further development of a broad new class of

materials with enhanced properties.

4. Experimental Section

MD simulations : MD simulations were implemented using the DISCOVER code in Materials Studio. The interatomic interactions are described by the force fi eld of the condensed-phase optimized molecular potential for atomistic simulation studies (COMPASS). [ 42 ] This is the fi rst ab initio force fi eld that has been parametrized and validated using condensed-phase properties in addition to various ab initio and empirical data and it has been shown to be appli-cable in describing the mechanical properties of CNTs. [ 43 , 44 ] The Andersen method [ 45 ] was employed in the thermostat to control the thermodynamic temperature and generate the correct statis-tical ensemble. As a temperature control, the thermodynamic tem-perature was kept constant by allowing the simulated system to exchange energy with a “heat bath”. [ 45 ] The force fi eld is expressed as a sum of valence (or bonding), cross-terms, and nonbonding interactions:

E total = E valence + E cross- term + E nonbonded (4)

E valence = ∑b

[K 2 (b − b0)2 + K 3 (b − b0)3 + K 4 (b − b0)4]

+∑2

[H2 (2−20)2 + H3 (2−20)3 + H4 (2− 20)4]

+∑N

[V1[1− cos(N−N 01)] + V2[1− cos(2N −N0

2)]

+ V3[1− cos(3N −N03)]]

+∑x

K xP 2 + E U B (5)

2017H & Co. KGaA, Weinheim www.small-journal.com

D. Xia et al.

2018

full papers

E cross-term � ∑b

∑b′

F bb′ (b − b0) b′ − b′0

+ ∑2

∑2′

F22 ′ (2 − 20) 2 ′− 2 ′0

+ ∑b

∑2

F b2 (b − b0) (2 − 20)

+∑b

∑N

F bN (b − b0) × [V1 cos(N) + V2 cos(2N) + V3 cos(3N)

]+ ∑

b′

∑N

F b′N b′ − b′0 b

′ − b′0

)× [

F 1 cos(N) + F 2 cos(2N) + F 3 cos(3N)]

+ ∑2

∑N

F2N (2 − 20) × [V1 cos(N) + V2 cos(2N) + V3 cos(3N)

]+∑

N

∑2

∑2 ′

K N22′ cos(N) (2 − 20) × 2 ′ − 2 ′0

)

(

( )

)

E non-bond �

∑i> j

[Ai j

r 9i j

− Bi j

r 6i j

]+

∑i> j

qi qj

gri j+ E H-bond

(7)

The valence energy, E valence , is generally accounted for by terms including bond stretching, valence-angle bending, dihedral -angle torsion, and inversion. The cross-term interaction energy, E cross-term , accounts for factors such as bond or angle distortions caused by nearby atoms to accurately reproduce the dynamic properties of molecules. The nonbonding interaction term, E non-bond , accounts for the interactions between nonbonded atoms and results mainly from vdW interactions. In Equation 1– 4 , q is the atomic charge, ε is the dielectric constant, and r ij is the i – j atomic separation dis-tance. b and b′ are the lengths of two adjacent bonds, θ is the two-bond angle, φ is the dihedral-torsion angle, and χ is the out-of-plane angle. b 0 , ki (i = 2–4), θ 0 , Hi ( i = 2–4), φ i

0 (i = 1–3), V i ( i = 1–3), F bb ′ , b 0 ′, F θ θ ′ , θ ′0 , F b θ , F b φ , F b ′ θ , F i ( i = 1–3), F θ φ , K φ θ θ ′ , A ij , and B ij were fi tted from quantum-mechanical calculations and imple-mented in the Discover module of Materials Studio.

MD simulations were performed in periodic boundary condi-tions in the range of 150 × 150 × 54.2205 Å 3 at 300 K. When the GN length was longer than 150 Å, the range of the periodic boundary condition was increased correspondingly. In this work, we considered different CNTs and GN sheets. As initial confi gura-tions, a series of GN sheets with the same width of 33.393 Å were aligned parallel to the CNTs with a separation of ≈ 5 Å. The models were put into an constant-volume/constant-temperature dynamics ( NVT ) ensemble simulation at 300 K. A time step of 1 fs was used and data was collected at intervals of 1 ps. The full-precision tra-jectory was then recorded and the results were analyzed.

Supporting Information

Supporting Information is available from the Wiley Online Library or from the author. It contains videos: Video S1: MD simulation showing a GN sheet with a length of 53.379 Å deforming and self-scrolling onto an (8,8) armchair CNT in cross-section; view Video S2: MD simulation showing a GN sheet with a length of 53.379 Å deforming and self-scrolling onto an (8,8) armchair CNT in side view.

(6)


[ 1 ] S. Iijima , Nature 1991 , 354 , 56 . [ 2 ] K. S. Novoselov , A. K. Geim , S. V. Morozov , D. Jiang , Y. Zhang ,

S. V. Dubonos , I. V. Grigorieva , A. A. Firsov , Science 2004 , 306 , 666 .

[ 3 ] M. M. J. Treacy , T. W. Ebbesen , J. M. Gibson , Nature 1996 , 381 , 678 .

[ 4 ] P. M. Ajayan , L. S. Schadler , C. Giannaris , A. Rubio , Adv. Mater. 2000 , 12 , 750 .

[ 5 ] A. B. Dalton , S. Collins , E. Muñoz , J. M. Razal , V. H. Ebron , J. P. Ferraris , J. N. Coleman , B. G. Kim , R. H. Baughman , Nature 2003 , 423 , 703 .

[ 6 ] L. Zhang , L. Balzano , D. E. Resasco , J. Phys. Chem. B 2005 , 109 , 14375 .

[ 7 ] K. S. Novoselov , Z. Jiang , Y. Zhang , S. V. Morozov , H. L. Stormer , U. Zeitler , J. C. Maan , G. S. Boebinger , P. Kim , A. K. Geim , Science 2007 , 315 , 1379 .

[ 8 ] J. N. Hu , X. L. Ruan , Y. P. Chen , Nano. Lett. 2009 , 9 , 2730 . [ 9 ] C. G. Lee , X. D. Wei , J. W. Kysar , J. Hone , Science 2008 , 321 ,

385 . [ 10 ] D. Prezzi , D. Varsano , A. Ruini , A. Marini , E. Molinari , Phys. Rev. B

2008 , 77 , 041404 . [ 11 ] T. Kawano , H. C. Chiamori , M. Suter , Q. Zhou , B. D. Sosnowchik ,

L. W. Lin , Nano Lett. 2007 , 7 , 3686 . [ 12 ] A. Star , Y. Lu , K. Bradley , G. Gru1ner , Nano Lett. 2004 , 4 ,

1587 . [ 13 ] J. Li , H. T. Ng , A. Cassell , W. Fan , H. Chen , Q. Ye , J. Koehne , J. Han ,

M. Meyyappan , Nano Lett. 2003 , 3 , 597 . [ 14 ] R. Ramasubramaniam , J. Chen , H. Y. Liu , Appl. Phys. Lett. 2003 ,

83 , 2928 . [ 15 ] F. Schedin , A. K. Geim , S. V. Morozov , E. W. Hill , P. Blake ,

M. I. Katsnelson , K. S. Novoselov , Nat. Mater. 2007 , 6 , 652 . [ 16 ] M. I. Katsnelson , Mater. Today 2007 , 10 , 20 . [ 17 ] L. A. Ponomarenko , F. Schedin , M. I. Katsnelson , R. Yang ,

E. W. Hill , K. S. Novoselov , A. K. Geim , Science 2008 , 320 , 356 . [ 18 ] K. S. Novoselov , A. K. Geim , S. V. Morozov , D. Jiang ,

M. I. Katsnelson , I. V. Grigorieva , S. V. Dubonos , A. A. Firsov , Nature 2005 , 438 , 197 .

[ 19 ] N. Patra , B. Y. Wang , P. Kral , Nano Lett. 2009 , 9 , 3766 . [ 20 ] S. Cranford , D. Sen , M. J. Buehler , Appl. Phys. Lett. 2009 , 95 ,

123121 . [ 21 ] M. M. Fogler , A. H. Castro Neto , F. Guinea , Phys. Rev. B 2010 , 81 ,

161408 . [ 22 ] S. F. Braga , V. R. Coluci , S. B. Legoas , R. Giro , D. S. Galvao ,

R. H. Baughman , Nano. Lett. 2004 , 4 , 881 . [ 23 ] Y. Chen , J. Lu , Z. X. Gao , J. Phys. Chem. C 2007 , 111 , 1625 . [ 24 ] H. Pan , Y. P. Feng , J. Y. Lin , Phys. Rev. B 2005 , 72 , 085415 . [ 25 ] G. Mpourmpakis , E. Tylianakis , G. E. Froudakis , Nano. Lett. 2007 ,

7 , 1893 . [ 26 ] R. Rurali , V. R. Coluci , D. S. Galvao , Phys. Rev. B 2006 , 74 ,

085414 . [ 27 ] R. Bacon , J. Appl. Phys. 1960 , 31 , 283 . [ 28 ] Y. Kaburagi , K. Hosoya , A. Yoshida , Y. Hishiyama , Carbon 2005 ,

43 , 2817 . [ 29 ] L. M. Viculis , J. J. Mack , R. B. Kaner , Science 2003 , 299 , 1361 .

Acknowledgements

This work is supported by the Cultivation Fund of the Key Scientifi c and Technical Innovation Project, Ministry of Education of China (708061), Natural Science Foundation of China (10974258), Pro-gram for New Century Excellent Talents in University (NCET-08-0844) and Postgraduate Innovation Fund of China University of Petroleum (SZ10-39).



[ 30 ] M. V. Savoskin , V. N. Mochalin , A. P. Yaroshenko , N. I. Lazareva , T. E. Konstantinova , I. V. Barsukov , I. G. Prokofi ev , Carbon 2007 , 45 , 2797 .

[ 31 ] H. Shioyama , T. Akita , Carbon 2003 , 41 , 179 . [ 32 ] X. Xie , L. Ju , X. F. Feng , Y. H. Sun , R. F. Zhou , K. Liu , S. S. Fan ,

Q. Q. Li , K. L. Jiang , Nano Lett. 2009 , 9 , 2565 . [ 33 ] K. Y. Yan , Q. Z. Xue , D. Xia , H. J. Chen , J. Xie , M. D. Dong , ACS Nano

2009 , 3 , 2235 . [ 34 ] Y. J. Qiu , J. Yu , G. Fang , H. Shi , X. S. Zhou , X. D. Bai , J. Phys. Chem.

C 2009 , 113 , 61 . [ 35 ] U. K. Gautam , X. S. Fang , Y. Bando , J. H. Zhan , D. Golberg , ACS

Nano 2008 , 2 , 1015 . [ 36 ] X. L. Li , Y. Q. Liu , L. Fu , L. C. Cao , D. C. Wei , Y. Wang , G. Yu , J. Phys.

Chem. C 2007 , 111 , 7661 . [ 37 ] C. W. Zhou , J. Kong , H. J. Dai , Phys. Rev. Lett. 2000 , 84 , 5604 . [ 38 ] X. T. Jia , M. Hofmann , V. Meunier , B. G. Sumpter ,

J. Campos-Delgado , J. M. Romo-Herrera , H. Son , Y.-P. Hsieh ,

© 2010 Wiley-VCH Verlag Gmbsmall 2010, 6, No.18, 2010–2019

A. Reina , J. Kong , M. Terrones , M. S. Dresselhaus , Science 2009 , 323 , 1701 .

[ 39 ] M. Al-Haik , M. Y. Hussaini , H. Garmestani , J. Appl. Phys. 2005 , 97 , 074306 .

[ 40 ] F. Scarpa , S. Adhikari , A. Srikantha Phani , Nanotechnology 2009 , 20 , 065709 .

[ 41 ] C. D. Reddy , S. Rajendran , K. M. Liew , Nanotechnology 2006 , 17 , 864 .

[ 42 ] H. Sun , J. Phys. Chem. B 1998 , 102 , 7338 . [ 43 ] Q. Wang , W. H. Duan , K. M. Liew , X. Q. He , Appl. Phys. Lett. 2007 ,

90 , 033110 . [ 44 ] Q. Wang , Nano. Lett. 2009 , 9 , 245 . [ 45 ] H. C. Andersen , J. Chem. Phys. 1980 , 72 , 2384 .

Received: April 17, 2010 Revised: June 14, 2010 Published online: August 16, 2010

2019H & Co. KGaA, Weinheim www.small-journal.com

Documents

Fabrication of Carbon Nanoscrolls from Monolayer · PDF fileFabrication of Carbon Nanoscrolls from Monolayer Graphene ... multiwall CNTs. ... shell-composite nanostructures by the