ELEMENTS OF PHYSICS
ELEMENTS OF PHYSICS PHYS 111
U of L
Popular in Course
Popular in Physics and Astronomy
This 4 page Class Notes was uploaded by Mireille Terry on Friday October 23, 2015. The Class Notes belongs to PHYS 111 at University of Louisville taught by Staff in Fall. Since its upload, it has received 13 views. For similar materials see /class/228377/phys-111-university-of-louisville in Physics and Astronomy at University of Louisville.
Reviews for ELEMENTS OF PHYSICS
Report this Material
What is Karma?
Karma is the currency of StudySoup.
Date Created: 10/23/15
PHYSICAL REVIEW B VOLUIVIE 61 NUIVIBER 24 RAPID COl Il IUNI 15 JUNE 2000II Chemical functionalization of carbon nanotubes through energetic radical collisions Boris Ni and Susan B Sinnottquotlt The University of Kentucky Department of Chemical and Materials Engineering Lexington Kentucky 40506 Received 1 February 2000 Classical molecular dynamics simulations are used to model the bombardment of a bundle of single walled carbon nanotubes by CH3 radicals impacting with incident energies of 10 45 and 80 eV The simulations show that there is a high probability of adhesion of either the radicals or their fragments to the nanotube walls at all the incident energies considered They therefore predict a pathway to the chemical functionalization of the walls of carbon nanotubes The simulations also show how at 80 eV the incident radicals can induce cross 1inking between the nanotubes Carbon nanotubes are being considered for use as bers in the next generation of composite materials Sometimes bers are chemically functionalized with polymer chains to in crease their adhesion to the polymer matrix in a composite1 Creation of covalent nonplanar CC bonds to the walls of carbon nanotubes results in the breaking of the local Sp2 hybridization and the formation of 7777 conjugated bonds at the surface of the nanotube Calculations of the chemical functionalization of singlewalled nanotubes SWNT s pre dict that functionalization decreases the Young s modulus of the nanotubes by about 152 and can alter their electronic structure3 Recent experiments have succeeded in functional izing SWNT s at their open ends and at the walls 6 by using carbodiimide chemistry4 or mixing the nanotubes with an electrophilic reagent that adds to deactivated double bonds5 6 The objective of this work is to investigate a different route for chemical functionalization of SWNT walls by radi cal bombardment Simulations to study the creation of nan ogears through the collision of benzene radicals with a SWNT have been considered previously7 but only under idealized conditions to show that it was possible to attach a benzyl radical to a nanotube wall No extensive studies over a range of incident energies on more than one SWNT have yet been undertaken It is to be expected that radical colli sions at hyperthermal energies could also create defects in the walls of the SWNT s similar to those observed during electron irradiation of nanotubes8 Therefore the second goal of this work is to study the creation or removal of defects through radical collisions and determine their dependence on the radicals incident energy The impacts of energetic radi cals with the capped ends of the nanotubes are also consid ered to compare the reactions at the caps to those that occur at the walls Finally the effects of 57 defects already present within the nanotube walls on the results of the collisions are examined The approach in this study is classical molecular dynamics simulations The simulations use a thirdorder Nordsieck predictor cor rector routine9 to integrate Newton s equation of motion with a time step of 020 femtoseconds fs The forces on all the atoms are calculated using an analytic reactive empirical bondorder potential REBO developed by Brenner 11 coupled to a longrange LennardJones potential as described in detail elsewhere12 This manybody potential has been successfully applied to model the related processes of ion bombardment of polymer surfaces13 and thin lm growth 01631829200061241634341500 PRB through molecular and cluster beam deposition11 It has also been used extensively to study the mechanical properties of carbon nanotubes 16 In most cases this potential has been shown to provide reasonable predictions12 However as is the case for all empirical potentials there are cases where the quantitative accuracy is lacking even while the qualitative trends are correct For example Hase and coworkers17 have shown that the REBO potential predicts association poten tials for HCH3 and Hdiamond 111 that are signi cantly smaller than ab initio values because of the potential s shorter range Within the REBO potential cutoff the pre dicted association potentials are similar to the ab initio values17 Therefore this effect is not of signi cant concern in the present study because of the relatively high incident en ergies used that bring the radicals into close contact with the nanotube walls well within the potential cutoff prior to any reaction The system used in the simulations consists of a bundle of six 1010 SWNT s arranged in two layers as shown in Fig 1 Each SWNT is 50 A long and consists of 800 atoms Periodic boundary conditions9 are applied within the plane of the nanotubes perpendicular to the direction of collision onto the nanotube walls At the nanotube edges 20 atoms are held rigid not allowed to evolve in time throughout the 0C 01 FIG 1 Snapshot of the initial conditions of the impact simula tions The gray spheres represent nanotube carbon atoms black spheres represent carbon in the incident CH3 radicals and white spheres represent hydrogen in the CH3 radicals R16 343 2000 The American Physical Society RAPID COl IlIUNI R16 344 TABLE I Percentage of events taking place in collisions of CH3 with a bundle of 1010 singlewalled carbon nanotubes The data are the averages of the outcomes of 35 trajectories 105 collisions performed for each incident energy 10 eV 45 eV 80 eV Scattering of CH from nanotubes 67 13 Scattering of CH2 from nanotubes 93 27 Scattering H from nanotubes 536 663 Scattering of CH3 from nanotubes 240 80 40 Adsorption of C outside wall 267 Adsorption of C in inside wall 267 Adsorption of CH outside wall 280 107 Adsorption of CH2 outside wall 240 40 Adsorption of CH3 outside wall 760 53 Defects structures form 187 240 Adsorption of H outside wall 142 232 Adsorption of H inside wall 32 simulations Moving towards the center of the nanotubes 40 atoms have Langevin frictional forces9 applied to them to maintain the temperature at 300 K The other atoms in the system are allowed to evolve in time with no constraints Another con guration was also considered where none of the atoms are held rigid and 140 atoms at the nanotube ends have Langevin frictional forces applied to them The results from the two con gurations showed negligible differences Every trajectory involved the bombardment of three CH3 radicals initially positioned 10 12 and 14 A above the top three SWNT s as shown in Fig 1 The radicals were then given incident energies towards the nanotubes The trajecto ries ran until it was clear the results were not going to change with most lasting about 800 fs Three incident ener gies of 10 45 and 80 eV were considered and 35 trajectories 105 impacts were performed for each incident energy from slightly different starting conditions obtained by varying the positions of the incident CH3 radicals relative to the nano tubes Some of the starting conditions place the radicals di rectly above the center of the nanotubes as shown in Fig 1 while some position the radicals to impact the nanotubes along their sides The simulation results are presented in Table I At inci dent energies of 10 eV only two kinds of phenomena are predicted to occur The rst is the scattering of CH3 radicals from the nanotube bundle while the second is the adsorption of the fragments on the outer walls of the carbon nanotubes at the rst site they hit as shown in Fig 2a Adhesion is predicted to occur more often as most of the time the impact ooc o a b C FIG 2 Representative snapshots of collision outcomes for CH3 impacting at a 10 eV b 45 eV c 80 eV The color scheme is the same as in Fig 1 BORIS NI AND SUSAN B SINNOTT FREQ FIG 3 Examples of defects formed from CH3 impacting at a 45 eV and b 80 eV The color scheme is the same as in Fig 1 ing radicals impact onto or very near a C atom on the nano tube wall The nal length of the CC bond formed between the nanotube wall and the CH3 is about 155 A a value that is quite close to normal CC bond lengths in alkanes Before the collision the CC bond lengths in the nanotube wall are around 142 A while after adsorption the CC bond lengths in the nAanotube wall around the adsorption site are about 155 At incident energies of 45 eV only about 13 of the CH3 radicals remain intact after the collisions and only 8 scatter away intact Most of the radicals break apart on impact and the larger fragments such as CH and CH2 adsorb on the outer surface of the nanotubes in about 52 of the collisions In some cases fragments such as CH2 and CH create two or more rarely three covalent bonds with surface atoms as shown in Fig 2b Generally these fragments do not bond to the rst carbon atom they hit but rather move along the nanotube making contact with 2 3 carbon atoms before nally bonding to the nanotube wall Most of the H atoms knocked loose on impact scatter away from the nanotubes although some adsorb to the outer wall The collisions at 45 eV also cause defects to form in the nanotube walls However the nanotube atoms usually reform their original bonding con guration in the course of the simulation In about 2 of the collisions carbon atoms were knocked out of the nanotube walls by the CH3 creating per manent vacancies in the nanotubes It should be emphasized that each nanotube is only struck by one radical per trajec tory in the simulations and therefore the same ability to reconstruct might not occur during bombardment if the time between subsequent hits is less then the time of recreation Another defect observed in about 19 of collisions is the insertion of the carbon atom from the CH3 into the nanotube structure In these cases the defect takes the form of two conjugated heptagons that slightly deform the outer surface of the affected SWNT as shown in Fig 3a It has been recently shown that bent nanotubes have enhanced reactivity18 Therefore these defect centers could in their turn serve as centers of enhanced reactivity and functional ization for future impacts At 80 eV most CH3 radicals are completely broken apart and only 4 of them scatter away intact Furthermore the impacts at these energies are so severe that the amplitude of FREQ local deformation within the nanotubes can reach of the nanotube s diameter The largest fragments from the CH3 radicals adsorb on the nanotube walls and most of the hydro gen atoms scatter away from the nanotubes However in contrast to what was seen previously at 45 eV the fragments are mostly lone atoms see Fig 2c Again the fragments make contact with 2 3 carbon atoms in the nanotube wall before adhering to one of them About 15 of the time the incident CH3 radicals knock out one or more carbon atoms from the walls creating large holes that do not heal on the time scales on these simulations as shown in Fig 3b In cident C atoms can also insert into the nanotube stmcture creating various complexes of pentagons heptagons and oc tagons In contrast to what was seen at the lower incident ener gies about 27 of the carbon atoms from the CH3 radicals knock out other C atoms from the nanotube walls This usu ally happens when the impacting radical hits directly on or very close to an individual carbon atom on the wall In some cases the C from the CH3 substitutes for the knocked out atom In addition the knockedout atom sometimes reacts with the atoms on the interior far wall The knockedout atom can also knock out another atom on the far wall which then goes on to adhere to a nanotube in the second layer of the bundle None of the simulations predict that a C atom from an incident CH3 penetrates directly through the nano tube wall The interior bonding sites were predicted to be potential energy minima in Ref 19 The REBO potential predicts that adhesion of single C atoms to the exterior of a 1010 nano tube is more stable than adhesion to the interior by 0027 eVatom The simulations also predict crosslinking between the SWNT s in the bundle through bombardment at 80 eV as illustrated in Fig 2c This could toughen the nanotube bundle stmcture and stabilize it to shear in a manner that is analogous to the toughening of polymers by crosslinking the rst few layers through ion bombardment20 The CC bond lengthAs of the interconnecting segments vary from 155 to 172 To determine the effect of 57 defects on the results we considered the impact of CH3 radicals on nanotubes that al ready had 57 defects present in the walls Ten trajectories 30 impacts on or around the 57 defects on the nanotubes in the bundle were performed The most interesting effect pre dicted was the healing of the 57 defect when the CH3 radical impacted at incident energies of 45 eV in about 7 of the collisions and 80 eV in 10 of the collisions This defect healing is shown in Fig 4 for a collision at 45 eV and only happened for impacts that occurred directly on the defect Otherwise the outcomes of the collisions were very similar to those predicted for regular nanotubes Since experimental samples may not be perfectly aligned during bombardment we also studied the impact of CH3 radicals on the capped portions of capped 10 10 SWNT s Three trajectories were performed at every energy and each trajectory involved six CH3 radicals impacting six capped nanotubes from slightly different initial positions for a total of 18 collisions per incident energy The simulations predict that the cap ends are even more exible than the SWNT walls a result that is compatible with the results of previous simulations on nanotube indentation14 At incident energies CHEMICAL FUNCTIONALIZATION OF CARBON RAPID COl IMUNICATION S R16 345 a b FIG 4 Sequence showing an incident CH3 impacting a nano tube with a 57 defect a initial collision of the radical with the nanotube wall b temporary bonding of the C from the radical to atoms in the 57 defect 0 nal structureithe 57 defect has been removed and the incident C atom is bonded to the side of where the defect used to be The color scheme is the same as in Fig 1 of 10 45 and 80 eV 50 67 and 33 of impacting CH3 fragments respectively scatter away intact without affecting the bonding in the cap Several vacancies and defect struc tures are formed within the cap at impacts of 45 eV in about 17 of the collisions and 80 eV in about 83 of the col lisions because of the energy transferred to the impact site At 10 eV 50 of the collisions result in CH3 bonding to the outside of the cap which is in contrast to the much higher value of 76 for the nanotube wall At 45 eV 33 of the collisions result in CH fragments bonding to the outside of the cap which is close to the 28 that adhere to the nano tube wall However no CH2 or CH3 fragments bond to the cap in contrast to what occurs on nanotube walls At 80 eV C CH and CH2 bond to the outside of the nanotube caps in 17 17 and 33 of the collisions respectively In con trast these species adhere to the outside walls in about 27 11 and 4 of the collisions respectively In addition about 27 of the fragments adhere to the inside walls at incident energies of 80 eV Thus there are differences in reactivity at the cap and the wall These simulations predict that chemical functionalization of nanotube cap ends by en ergetic radical bombardment is less effective than chemically functionalizing the nanotube walls at the two lower incident energies To summarize atomistic simulations have been per formed to model the bombardment of a nanotube bundle with CH3 radicals at incident energies of 10 45 and 80 eV They show the chemical adhesion of radicals or heavy frag ments from the radicals such as CH2 CH or C to the SWNT s can occur at all the incident energies considered The results also predict that this method of functionalization is more effective on the nanotube walls than on the caps at incident energies of 10 and 45 eV Most of the hydrogen atoms that are knocked loose from the radicals on impact simply scatter away although some adhere to the nanotubes At incident energies of 45 and 80 eV the impacts can also create defects and vacancies within the nanotube walls When 57 defects are already present in the nanotube walls the impact of the incident radicals can remove the defects At 80 eV the simulations also predict that the nanotube bundle can be crosslinked in a manner similar to that seen in poly mers following ion bombardment RAPID COMMUNICATIONS R16 346 This work was supported by the NSF CHE9708049 and MRSEC DMR9809686 and by the NASA Ames Research Center NAG 21121 Correspondjng author 1 See for example R Lin R P Quirk J Kuang and L S Penn J Adhes Sci Technol 10 241 1996 2A Garg and S B Sinnott Chem Phys Lett 295 273 1998 3D W Brenner et al J Br Interplanet Soc 51 137 1998 4J Liu et al Science 280 1253 1998 5J Chen et al Science 282 95 1998 6S S Wong et al Nature London 394 52 1998 7J Han A Globus R Jaffe and G Deardorff Nanotechnology 8 95 1997 8P M Ajayan V Ravikumar and JC Charlier Phys Rev Lett 81 1437 1998 9M P Allen and D J Tildesley Computer Simulation of Liquids Oxford University Press New York 1987 10D w Brenner Phys Rev B 42 9458 1990 11S B Sinnott L Qi O A Shenderova and D W Brenner in BORIS N1 AND SUSAN B SINNOTT PRB g Molecular Dynamics of Clusters Surfaces Liquids and Inter faces edited by W Hase JAl Press Inc Stamford CT 1999 6 pp 172 12Z Mao A Garg and S B Sinnott Nanotechnology 10 273 1999 13 D W Brenner O A Shenderova and C B Parker Mater Res Soc Symp Proc 438 491 1997 14A Garg J Han and s B Sinnott Phys Rev Lett 81 2260 1998 15B 1 Yakobson C J Brabec and J Bemholc Phys Rev Lett 76 2511 1996 16c F Cornwell and L T Wille Solid State Commun 101 555 1997 17P De Sainte Claire K Son W L Hase and D W Brenner J Am Chem Soc 100 1761 1996 18D Srivastava et al J Phys Chem B 103 4330 1999 19A A Farajian et al J Chem Phys 111 2164 1999 20E H Lee G R Rao M B Lewis and L K Mansur Nucl Instrum Methods Phys Res B 74 326 1993