Copyright 2011 American Scientic Publishers All rights reserved Printed in the United States of America

Journal of Nanoscience and Nanotechnology Vol. 11, 14, 2011

Molecular Dynamics of Carbon Nanotube Bundles as Molecular Sieves

Jos Fernando P. Leal1 , David L. Azevedo2 , and Jordan Del Nero3 4
Ps-Graduao em Fsica, Universidade Federal do Par, Belm, PA 66075-110, Brazil Department de Fsica, Universidade Federal do Maranho, So Luis, MA 65080-040, Brazil 3 Diviso de Metrologia de Materiais, Instituto Nacional de Metrologia, Normalizao e Qualidade Industrial (INMETRO), Duque de Caxias, RJ 25250-020, Brazil 4 Department de Fsica, Universidade Federal do Par, Belm, PA 66075-110, Brazil
2 1

In this work we investigate a system composed by a bundle of carbon nanotube (BuCNT) confronted with fatty acid surrounded. The system consists of a rigid open nanotube and the oleic acid (OA) working as probe relaxing due the interaction with BuCNT. The OA has in his form an asymmetric shape with COOH termination provoking a close BuCNT interaction mainly by van der Waals (vdW) force eld. The simulations were performed by classical molecular dynamics coupled with quantum mechanics Hartree-Fock simulations with standard parameterizations. Our results show that these BuCNT are dynamically stable and it shows a preferential interaction position with OA resulting in two features: (i) when the OA terminated with CH2 is closer to open extremity of BuCNT, the OA is repelled; and (ii) when the OA terminated with COOH is closer to open extremity of BuCNT, the OA is encapsulated by them. These simulations can be utilized as a tool for the constructions of efcient molecular sieves.

Keywords: Bundles, Carbon Nanotube, Oleic Acid, Molecular Dynamics, Hartree-Fock.

RESEARCH ARTICLE

1. INTRODUCTION
Materials with intrinsic selectivity appear in nature in several forms as, for instance, in biological systems when ionic channels are opened or closed in specic conditions as pH, temperature among others.1 Within new technological advances have preclude to necessity of design nanodevices.2 An effort has been made to create devices with high selectivity beyond our present micromachining.3 4 Subashini and Pandurangan have prepared a synthesis route of mesoporous molecular sieves as catalytic template for the growth of carbon nanotubes.5 Molecular dynamics simulations performed using carbon nanotube including constriction show that entropic effect can be dominant when interacting with molecules with similar sizes concluding that it could useful for permeation selectivity.6 Also, dispersed oleic acid has been used as nanouid to stabilize colloidal nanoparticles showing direct evidence for efcient heat transport and it could be useful as a new generation of nanouids with tunable thermal conductivity7 as well the capacity for controlled encapsulation.8

In this communication, we have carried out molecular dynamics simulations grounded in classical mechanics with well-known molecular force eld9 10 (and references therein) including bond angle bend, bond stretch, diedral rotation, van der Waals terms and Hartree-Fock quantum mechanics calculations for a bundle of carbon nanotubes (BuCNT) with oleic acid (OA) (working as probe) representing a molecular sieves model. Figure 1 presents a summary of two possibilities of interaction between them: the head and tail of fatty acid in opposite positions to reach the BuCNT.

2. METHODOLOGY
In these simulations we have done calculations including structures containing up to 1973 atoms. This methodological feature has been guarantied to be very effective in the investigation of dynamic signature of CHON (carbon, hydrogen, oxygen, and nitrogen) structures.914 The convergence criteria for all calculations were based on previous works done and tested by us1214 as: The energy differences, maximum force, and maximum atomic displacement equal to 104 kcal/mol, 103 kcal/mol/, and 105 , respectively. Also, the root mean square deviations
doi:10.1166/jnn.2011.4141

Author to whom correspondence should be addressed.

J. Nanosci. Nanotechnol. 2011, Vol. 11, No. xx

1533-4880/2011/11/001/004

Molecular Dynamics of Carbon Nanotube Bundles as Molecular Sieves

Leal et al.

number up to 250 congurations. To compute the optical spectra, we take into account from the rst 40-UMOs (Unoccupied Molecular Orbitals) to the last 40-OMOs (Occupied Molecular Orbitals) and adopted the standard the Forsters-Mattaga-Nishimoto parameters.1719

3. RESULTS AND DISCUSSION

In Figure 2(a), we present the main features of the molecular dynamic simulations to the case presented in Figure 1(a) where the OA has the CH2 group interacting with the BuCNT. The potential and kinetic energies have a small otation due the interaction among the atoms of OA + BuCNT. Both quantities are in phase mode producing the same dynamical general features. The net effect of OA over the three external nanotubes could be seen as an oscillatory period of 6.2 fs. Corroborating this signature, we plot out the non-bond energy showing very low variation illustrated as background otation. Also, Figure 2(b) shows the
(a) 140
130 120 110 Potencial energy Kinetic energy Non-bond energy

Energy (kcal/mol)
Fig. 1. Pictogram representation of the proposed OA + BuCNT where (top) COOH and (bottom) CH2 molecular termination were simulated.

100 90 80 70 60 50 40 30 20 10 0 2 4 6 8 10 12 14 16 18 20 22 24

RESEARCH ARTICLE

and displacement as 103 kcal/mol/ and 105 , respectively. After minimization procedures selective canonical dynamics was permitted with time steps of 1 fs were accomplished for all calculations. The generation of structures were followed by a bundle of carbon nanotube (BuCNT) open at both ends acting as a multishell barrier and confronted with oleic acid surrounded. These carbon nanotube (CNT) that forms the BuCNT were separately optimized (including standards energy and geometric grounds) and reoptimized when put them together. After that we attribute several initial position for OA (as previous described) in order to initialize the full simulation. To calculate the spectroscopic absorption feature, we utilized the quantum mechanics Hartree-Fock derivative INDO/S-CI (Intermediate Neglect of Differential Overlap/SpectroscopicConguration Interaction) approach developed by Zener et al.15 16 where this methodology was specically parametrized to describe the ultraviolet-visible spectrum of organic compounds. All the calculations included single excitations, with an average 2

Time (ps) (b)

600 Temperature

500

Temperature (K)

400

300

200

100 0 2 4 6 8 10 12 14 16 18 20 22 24

Time (ps)
Fig. 2. (a) Potential, kinetic, and non-bonded energies, as well (b) temperature results from molecular dynamics simulations. These results refer to the case where OA is not captured by the BuCNT showing no dissipative effects with very small uctuation due the interaction between them.

J. Nanosci. Nanotechnol. 11, 14, 2011

Leal et al.

Molecular Dynamics of Carbon Nanotube Bundles as Molecular Sieves

dynamical aspects of temperature, and it can be described as a simple model of considering the motion of the centerof-mass due repulsion of the van der Waals forces that acting on. See complementary video presented in Ref. [20]. At this point, we could speculate that in opposite conguration OA (Fig. 1(b) with COOH group closer to BuCNT) the net effect could be the same remaining oscillatory frequencies due the huge repulsive force as signature of dynamic properties. These results are presented in Figure 3. Our ndings demonstrated that this repulsion effect disappears due the interactions between the nanotubes and OA inducing massive transfer momentum between them affecting the oscillatory behavior. Figure 3(a) shows the potential and kinetic energies presenting an huge otation and no sustained oscillatory regime. The non-bonded energy became important as well changing the signal as typical behavior of exchange interaction between OA and BuCNT (up to 15 ps). After that for values larger than 15 ps, the OA is captured inside
(a) 140
120 100 80 Potencial energy Kinetic energy Non-bond energy

Fig. 4. Simulated UV-visible absorption spectra of full system (OA + BuCNT) from INDO/S-CI calculations showing the main transitions.

60 40 20 0 20 40 60 80 0 2 4 6 8 10 12 14 16 18 20 22 24

Time (ps) (b)

600 Temperature

500

400

300

200

100 0 2 4 6 8 10 12 14 16 18 20 22 24

Time (ps)
Fig. 3. (a) Potential, kinetic, and non-bonded energies, as well (b) temperature results from molecular dynamics simulations. These results refer to the case where the OA presented encapsulation by the BuCNT. It clearly shows dissipative effects in energies and temperature when the interaction between both systems (OA + BuCNT) becomes strong.

a specic nanotube show oscillatory energies representing shelled and stable OA into BuCNT. See complementary video presented in Ref. [21]. The same oscillatory regime is followed for temperature (Fig. 3(b)) as non-periodic [periodic] behavior for time lower [greater] than 15 ps. A fundamental question is how to utilize this device to work as molecular sieves in a controllable way. This goal could be obtained by applying external electric elds and/or variable magnetic elds in the fatty acid to align in preferential direction. Also, it is possible to separate selectively BuCNT with/without fatty acid. For the calculation of the optical absorption spectra for BuCNT, OA, and BuCNT + OA, we identify the main optical transitions by their oscillator strengths (o.s.) and compute the corresponding theoretical UV-Vis spectra by broadening each of these principal transitions by a suitably weighted Gaussian function normalized to the calculated oscillator strengths.22 23 Figure 4 shows the resulting absorption spectra for these three systems. For BuCNT, the rst optical transition appears at 847 nm and it is composed by two transitions, HOMO to LUMO + 1 ( H L + 1 ) and HOMO 1 to LUMO ( H 1 L ) transitions with large oscillator strength. There are two more additional bands centered in 530 nm and 389 nm corresponding to mixing of several transitions (Also presented in Fig. 4). Differently, for OA the most important absorption line corresponds to the H L + 1 , H 1 L , and H 3 L transitions centered at 220 nm. Finally, the behavior of BuCNT + OA shown specic absorption lines and it could be useful to separate different features in terms of electronic behavior for OA encapsulated by BuCNT or only BuCNT.

Energy (kcal/mol)

RESEARCH ARTICLE

Temperature (K)

4. SUMMARY AND CONCLUSIONS

In principle, this molecular sieve device could work with specic experimental parametrization, although sometimes 3

J. Nanosci. Nanotechnol. 11, 14, 2011

Molecular Dynamics of Carbon Nanotube Bundles as Molecular Sieves

Leal et al.

a residual contamination probably will be present in the environment and this can generate serious problems with the bundle of nanotubes. It is important to rise up that difference values between ideal and real system could be reached but in order to prevent any undesirable interactions signature the main results in this work in grounded in the dynamics process of the interaction between these materials. We can summary our ndings as: (i) Energies (kinetic, and potential) are in phase and directly proportional to the temperature when repulsive action is more favorable (OA with CH2 group interacting with the BuCNT). (ii) Only the potential energy is in-phase and directly proportional to the temperature when attractive action is the main one (OA with COOH group interacting with the BuCNT) up to 15 ps. After that (OA encapsulated by BuCNT), the all system are stabilized showing the same qualitative feature as (i). The simulations presented here indicate that the asymmetric form and position of probe (OA) is related with device efciency due mainly the gradient of van der Waals potential acting in. Acknowledgments: JFPL is grateful to VALE/ FAPESPA fellowship. JDN and DLA would like to thank INCT/Nanomateriais de Carbono and CNPq/Rede Nanotubos de Carbono for nancial support. Also, DLA [JDN] would like to thank FAPEMA [CNPq, FAPERJ, and FAPESPA] for nancial support.

RESEARCH ARTICLE

References and Notes

1. C. Song and B. Corry, J. Phys. Chem. B 113, 7642 (2009). 2. L. B. Nie, H. S. Guo, Q. G. He, J. R. Chen, and Y. Q. Miao, J. Nanosci. Nanotechnol. 7, 560 (2007).

3. P. Vichchulada, L. D. Lipscomb, Q. H. Zhang, and M. D. Lay, J. Nanosci. Nanotechnol. 9, 2189 (2009). 4. M. Abe, K. Murata, T. Ataka, Y. Ifuku, and K. Matsumoto, J. Nanosci. Nanotechnol. 9, 1947 (2009). 5. D. Subashini and A. Pandurangan, Catal. Commun. 8, 1665 (2007). 6. G. Arora and S. I. Sandler, Fluid Phase Equilibria 259, 3 (2007). 7. J. Philip, P. D. Shima, and B. Raj, Nanotechnology 19, 305706 (2008). 8. S. Abraham and S. S. Narine, J. Nanosci. Nanotechnol. 9, 6326 (2009). 9. S. B. Legoas, V. R. Coluci, S. F. Braga, P. Z. Braga, P. Z. Coura, S. O. Dantas, and D. S. Galvo, Phys. Rev. Lett. 90, 55504 (2003). 10. Forcite Force Field is Available from Accelrys Inc. (http://www.accelrys.com) as part of Materials Studio program. 11. V. R. Coluci, N. M. Pugno, S. O. Dantas, D. S. Galvao, and A. Jorio, Nanotechnology 18, 335702 (2007). 12. J. Del Nero and A. M. J. C. Neto, J. Comput. Theor. Nanosci. 4, 606 (2007). 13. A. M. J. C. Neto and J. Del Nero, J. Comput. Theor. Nanosci. 4, 107 (2007). 14. J. S. Andrade, Jr., D. L. Azevedo, R. Correa Filho, and R. N. C. Filho, Nano Lett. 5, 1483 (2005). 15. J. D. Head and M. C. Zerner, Chem. Phys. Lett. 131, 359 (1986). 16. A. D. Bacon and M. C. Zerner, Theoret. Chim. Acta 53, 21 (1979). 17. K. Nishimoto and N. Mataga, Z. Phys. Chem. (Neue Folge) 12, 335 (1957). 18. K. Nishimoto and L. S. Forster, Theor. Chim. Acta 3, 407 (1965). 19. K. Nishimoto and L. S. Forster, Theor. Chim. Acta 4, 155 (1966). 20. See at the Following Homepage for Snapshot Movies, in AVI format, of simulations discussed in this work. The zipped le is reachable at (http://www.ufpa.br/jordan/n-entrou-C-bun-10-10.rar) for the case where the OA is reppeled by BuCNT. 21. See at the following homepage for snapshot movies, in AVI format, of simulations discussed in this work. For the case where BuCNT encapsulates the OA. The zipped les are reachable at http://www.ufpa.br/jordan/entrou-O-bun-10-10.rar for lateral view and a frontal view of same simulation can be found at http://www.ufpa.br/jordan/vista-frontal-entrou-O-bun-10-10.rar 22. D. B. Lima and J. Del Nero, J. Comput. Theor. Nanosci. 5, 1445 (2008). 23. M. A. L. Reis, A. Saraiva-Souza, and J. Del Nero, J. Comput. Theor. Nanosci. 6, 101 (2009).

Received: xx Xxxx xxxx. Accepted: xx Xxxx xxxx.

J. Nanosci. Nanotechnol. 11, 14, 2011