Universally dispersible carbon nanotubes

Article abstract of DOI:10.1021/ja309029n

We show that supramolecular chemistry provides a convenient tool to prepare carbone nanotubes (CNTs) that can be dispersed in solvents of any chemical nature, easily recovered and redispersed. Thymine-modified CNTs (CNT-Thy) can be dispersed in solution in the presence of diaminotriazine (DAT) end-functionalized polymers, through supramolecular Thy/DAT association. DAT-polymer chains are selected according to the solvent chemical nature: polystyrene (PS) for hydrophobic/low polarity solvents and a propylene oxide/ethylene oxide copolymer (predominantly propylene oxide based, PPO/PEO) for polar solvents or water. Long-term stable supramolecular CNT dispersions are reversibly aggregated by adding a few droplets of a selective dissociating agent of the Thy/DAT association (DMSO). CNT-Thy, simply recycled by centrifugation or filtration, can be redispersed in another solvent in presence of a suitable soluble DAT-polymer. Dispersion and aggregation can also be switched on and off by choosing a polymer for which a given solvent is close to Θ-conditions, e.g., PS in cyclohexane or PPO/PEO in water.

Full text of DOI:10.1021/ja309029n

Communication
pubs.acs.org/JACS
Universally Dispersible Carbon Nanotubes
Alexandre Prevoteau, Corinne Soulie-Ziakovic,* and Ludwik Leibler*
́
Matier
̀
e Molle et Chimie, UMR 7167 CNRS-ESPCI, Ecole Super
́
ieure de Physique et Chimie Industrielles de la Ville de Paris, 10 rue
Vauquelin 75005 Paris, France
S
* Supporting Information
approaches, polymers or surfactants adsorption or polymer
grafting, are irreversible, thus limiting recyclability and
redispersion possibilities.
ABSTRACT: We show that supramolecular chemistry
provides a convenient tool to prepare carbone nanotubes
(CNTs) that can be dispersed in solvents of any chemical
nature, easily recovered and redispersed. Thymine-
modified CNTs (CNT-Thy) can be dispersed in solution
in the presence of diaminotriazine (DAT) end-function-
alized polymers, through supramolecular Thy/DAT
association. DAT-polymer chains are selected according
to the solvent chemical nature: polystyrene (PS) for
hydrophobic/low polarity solvents and a propylene oxide/
ethylene oxide copolymer (predominantly propylene oxide
based, PPO/PEO) for polar solvents or water. Long-term
stable supramolecular CNT dispersions are reversibly
aggregated by adding a few droplets of a selective
dissociating agent of the Thy/DAT association (DMSO).
CNT-Thy, simply recycled by centrifugation or filtration,
can be redispersed in another solvent in presence of a
suitable soluble DAT-polymer. Dispersion and aggregation
can also be switched on and off by choosing a polymer for
which a given solvent is close to Θ-conditions, e.g., PS in
cyclohexane or PPO/PEO in water.
A seminal article by Quintana and Prato discloses how to
graft Thy units to CNTs and demonstrates that supramolecular
chemistry offers a versatile tool to control self-assembly of
CNTs.15a More rencently, Quintana et al. have shown that
supramolecular host/guest interaction between functional
polymers and thymine-modified CNTs can be used to prepare
CNT-polymer hybrids and to pattern functionalized polymeric
surfaces.15b
In a different supramolecular chemistry approach, Di
Crescenzo et al.16a and Llanes-Pallas et al.16b used pristine
rather than functionalized CNTs and showed that supra-
molecular polymers carrying the adequate side groups can
adsorb (wrap around surface) on CNTs and efficiently stabilize
dispersions. As for more classical systems, this approach can
work only for good solvents of side groups.
Inspired by these results and by physics of colloidal
stabilization, we address the question of efficient stabilization.
Here, we propose a method to achieve universal, reversible, and
controllable CNTs dispersions by using supramolecular
chemistry through host/guest association. Molecules acting as
hosts were grafted onto the CNT surface. Linear polymer
chains of different chemical composition containing a
complementary guest group at one end were synthesized.
Thanks to the host/guest interaction, the polymer chains attach
onto the CNTs by their end functional group. Thus, a layer of
polymer is formed on the surface of the CNTs and ensures
steric stabilization of dispersions in good solvents of the
polymer (Figure 1A). Using polymers with one functional end,
rather than polymers bearing multiple functional units, is
essential to avoid bridging by polymers and CNTs aggregation
at high concentrations.15b Adding a selective host/guest
dissociating additive breaks the supramolecular interaction.
While the guest-polymer chains remain in solution, the CNT-
host aggregates can be separated by centrifugation or filtration
(Figure 1B). CNT-host can be subsequently redispersed in
another solvent using adequate guest-polymers (Figure 1C). As
a consequence, such CNTs-host/guest-polymer systems are
universal since good dispersions can be achieved in any solvent
by appropriate choice of guest-polymer chains. Figure 1D
illustrates how aggregation/dispersion can be controlled by
changing temperature. Indeed, in a bad solvent, polymer chains
collapse, so steric stabilization is no longer effective and CNTs
tabilizing dispersions of carbon nanotubes (CNTs) is a
Smajor challenge since it is a prerequisite to high
performance CNT/polymer composites1 potentially useful in
a large field of applications, such as reinforced materials,2
conductive plastics or adhesives,3 electronics,4 and actuators.5
Because of strong van der Waals forces, CNTs are prone to
form entangled bundles6 and aggregate in solution. In general,
dispersion of CNTs involves two stages: first, mechanical
disentanglement, usually by sonication, and second, stabiliza-
tion by adsorbing an organic mediating molecule.1b,c,7 For
instance, because they partly absorb on the CNTs surface
through van der Waals hydrophobic forces or π−π-stacking,1a
block and/or amphiphilic polymers8 or surfactants9 can weaken
intertube van der Waals interactions and allow steric repulsions
between nonadsorbed moieties in good solvents.8a,9a,10 An
alternative way to stabilize CNTs is to chemically modify the
CNTs surface either by reacting surface defects (COOH or
OH)11 or by very reactive sp2 CNT sidewalls;12 such modified
CNTs are dispersible in adequate solvents. Other methods use
covalent grafting of polymers either by classical organic
reactions or more recently by click reactions.7a,12a,13 The
main disadvantage of these methods is that each modification is
specific of the solvent (polar/apolar, hydrophilic/hydrophobic)
and requires a particular registration of each newly modified
CNTs to environmental administrations.14 Both stabilization
Received: September 11, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja309029n | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
association, CNTs grafted with Thy (CNT-Thy) will
exclusively interact with DAT-ended polymer chains (Figure
1A).17 Purified CNTs (pristine, CNTp) were grafted in water
with a Thy derivative of 1,4-phenylenediamine in presence of
isoamylnitrite (Scheme 1a).18 Grafting of phenyleneamine-Thy
was monitored by UV/vis spectroscopy by following the K-
band of Thy at 280 nm. CNT-Thy were analyzed by TGA,
Raman, and infrared spectroscopies. Grafting was estimated to
be 1 Thy group per 50 carbon atoms (See Supporting
Information, SI). Two low molecular weight polymers were
used: poly(propylene oxide-ethylene oxide) (29/6 monomers,
PPO/PEO-DAT, Mw(PPO/PEO) = 2000 g/mol) and
polystyrene (PS-DAT, Mw(PS) = 3600 g/mol). PPO/PEO-
DAT was synthesized from a commercial monoamine PPO/
PEO-NH2 according to a procedure described elsewhere.19
Polystyrene oligomer PS-Br was synthesized by atom-transfer
radical polymerization (ATRP), converted to amine derivative
PS-NH2,20 and grafted with DAT unit according to the same
reaction as for PPO/PEO-NH2 (Scheme 1b). Syntheses and
characterizations of DAT-polymer chains are given in SI.
To evidence that supramolecular Thy/DAT interaction
favors stable dispersions of CNTs in solvents of any chemical
nature (Figure 1), we prepared by sonication (150W, 30 min)
solutions of 0.05 wt % CNT-Thy and 0.1 wt % DAT-polymer
in solvents of increasing polarity: cyclohexane, toluene,
chloroform, acetone, and water (Figure 2b). PS-DAT was
used in apolar solvents and PPO/PEO-DAT in polar solvents
and water.
For comparison, the same solutions were prepared with
CNTp alone (Figure 2a) and CNT-Thy alone (see Figure
S14). All solutions were allowed to stand 24 h and observed by
optical microscopy (Figure 2a,b). Particle size analysis was
performed with ImageJ. CNTs were considered as dispersed
when detected aggregates represent ≤1% of the total analyzed
surface with a mean projected area particle size <5 μm.
Furthermore, particle size distributions (PSD) revealed a
median diameter <2 μm and mode diameter comprised
between 1 and 2 μm (see SI).
CNT surface consists of sp2 hybridized carbon planes and is
mostly apolar and hydrophobic. CNTp used in this study have
been purified by a sulphuric acid treatment and present
carboxylic acid functionalities on the surface.21 Nevertheless,
their surface was neither hydrophobic/hydrophilic or apolar/
polar enough to ensure good dispersions in any studied solvent
(Figure 2a). Even for chloroform and acetone where CNTp
seemed “macroscopically” dispersed, large aggregates were
observed by optical microscopy (Figure 2a). All pure CNT-Thy
Figure 1. Universal, reversible, and controllable CNT dispersions
through supramolecular chemistry.
aggregate. By changing the temperature, the solvent quality can
be improved and CNTs redispersed.
The host/guest system used in this study is Thy and 2,6-
diamino-1,3,5-triazine (DAT) that associate through three
parallel H-bonds. Since Thy/DAT complementary association
is much stronger than either Thy-Thy or DAT-DAT self-
Scheme 1. (a) Chemical Grafting of Thymine Derivatives on Purified CNTs and (b) DAT End-Functionalization of Polymer
Chains
B
dx.doi.org/10.1021/ja309029n | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Figure 3. Universal dispersion and recycling of CNT-Thy: (a) 0.05 wt
% CNT-Thy/0.1 wt % DAT-PS in toluene; (b) CNT-Thy aggregation
triggered by two drops of DMSO; and (c) redispersion in a solution of
0.1 wt % PPO/PEO-DAT in acetone.
DAT in DMSO and water could explain these results: the more
Thy and DAT are soluble, the faster and complete is the
dissociative effect of the solvent.
To sum up, dispersibility and recycling of CNTs can be
controlled by supramolecular chemistry: CNT-Thy are
universally dispersed by adapting DAT-polymer nature to the
desired dispersing solvent and recycled by using a selective
dissociating additive of the supramolecular association. In
addition, some CNT-Thy dispersions can be controlled by
temperature (Figure 1D) since, in Θ-solvents, DAT-polymer
chains can swell or collapse.
For instance, cyclohexane is a Θ-solvent of PS with an upper
critical solution transition of 36 °C, and water is a Θ-solvent of
PPO/PEO with a lower critical solution temperature of 18 °C
for the used PPO/PEO-DAT (see SI). 0.05 wt % CNT-Thy/
0.1 wt % DAT-PS in cyclohexane gave stable dispersions at 50
°C. When cooled in an ice bath for 10 min, almost complete
aggregation was observed. Heating and stirring the aggregated
solution at 50 °C for 10 min restored a stable dispersion
(Figure 4a). 0.05 wt % CNT-Thy/1 wt % DAT-PPO/PEO in
Figure 2. 0.05 wt % CNTs solutions in cyclohexane (at 40 °C),
toluene, chloroform, acetone, and water and corresponding optical
microscopy: (a) purified CNTp and (b) Thy-grafted CNT-Thy/DAT-
polymer (PS or PPO/PEO).
solutions were also aggregated. Contrariwise, in all DAT-
polymer solutions, CNT-Thy were dispersed since large
aggregates were no longer observed by optical microscopy
(Figure 2b). Furthermore, in organic solvents, these dispersions
were stable over a few months. To confirm that dispersions are
driven by the Thy/DAT supramolecular interaction,15b,16 test
solutions in toluene of (i) CNT-Thy, (ii) CNTp/DAT-polymer
(PS or PPO/PEO), and (iii) CNT-Thy/H2N-polymer were
prepared according to the same protocol. All test solutions
showed large aggregates, indicating that polymer (PS or PPO/
PEO) derivatives were not good dispersants for CNTp or
CNT-Thy (see Figure S15).
Thus, Thy/DAT association allows universal dispersions of
CNTs in a large variety of solvents. Since Thy/DAT interaction
consists of three parallel hydrogen bonds (Figure 1A), this
supramolecular interaction can be broken by adding a
competitive molecule, i.e., a dissociating additive. CNT
dispersions can then be destabilized, and the CNTs recovered
by filtration or centrifugation (figure 1B).
Water, polar, and/or protic solvents, such as DMSO or
alcohols, are H-bond competitors that can break H-bonded
supramolecular associations.16,22 Indeed, adding two droplets of
DMSO to a stable dispersion of CNT-Thy/DAT-PS in toluene
led to a complete aggregation of CNT-Thy within 20 min
(Figure 3a,b). Aggregation by Thy/DAT disruption allowed
CNT-Thy recovery from the DAT-PS solution by centrifuga-
tion. By using a solution of 0.1 wt % DAT-PPO/PEO in
acetone, it was possible to redisperse the recovered CNT-Thy
(Figures 1C and 3c). In contrast to DMSO, a dispersion stable
for at least 24 h could be obtained in water with 1 wt % DAT-
PPO/PEO (Figure 2b). Differences in solubility of Thy and
Figure 4. Aggregation and redispersion 0.05 wt % of CNT-Thy/1 wt
% DAT-PPO/PEO in water (from left to right): stable dispersion at 5
°C, aggregation at 50 °C in 20 min, and redispersion at 5 °C under
stirring.
C
dx.doi.org/10.1021/ja309029n | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(7) (a) Sahoo, N. G.; Rana, S.; Cho, J. W.; Li, L.; Chan, S. H. Prog.
Polym. Sci. 2010, 35, 837−867. (b) Kim, S. W.; Kim, T.; Kim, Y. S.;
Choi, H. S.; Lim, H. J.; Yang, S. J.; Park, C. R. Carbon 2012, 50, 3−33.
(c) Di Crescenzo, A.; Demurtas, D.; Renzetti, A.; Siani, G.; De Maria,
P.; Meneghetti, M.; Prato, M.; Fontana, A. Soft Matter 2009, 5, 62−66.
(d) Di Crescenzo, A.; Aschi, M.; Del Canto, E.; Giordani, S.;
Demurtas, D.; Fontana, A. Phys. Chem. Chem. Phys. 2011, 13, 11373−
11383.
(8) (a) Nativ-Roth, E.; Schvartzman-Cohen, R.; Boubioux, C.;
Florent, M.; Zhang, D.; Szleifer, I.; Yerushalmi-Rozen, R. Macro-
molecules 2007, 40, 3676−3685. (b) Kang, Y.; Taton, T. A. J. Am.
Chem. Soc. 2003, 125, 5650−5651. (c) Shvartzman-Cohen, R.; Florent,
M.; Goldfarbn, D.; Szleifer, I.; Yerushalmi-Rozen, R. Langmuir 2008,
24, 4625−4632. (e) Di Meo, E. M.; Di Crescenzo, A.; Velluto, D.;
O’Neil, C. P.; Demurtas, D.; Hubbel, J. A.; Fontana, A. Macromolecules
2010, 43, 3429−3437. (d) Di Crescenzo, A.; Aschi, M.; Fontana, A.
Macromolecules 2012, 45, 8043−8050.
(9) (a) Moore, V. C.; Strano, M. S.; Haroz, E. H.; Hauge, R. H.;
Smalley, R. E.; Schmidt, J. Nano Lett. 2003, 3, 1379−1382. (b) Islam,
M. F.; Rojas, E.; Bergey, D. M.; Johnson, A. T.; Yodh, A. G. Nano Lett.
2003, 3, 269−273. (c) Di Crescenzo, A.; Germani, R.; Del Canto, E.;
Giordani, S.; Savelli, G.; Fontana, A. Eur. J. Org. Chem. 2011, 5641−
5648.
(10) (a) Star, A.; Stoddart, J. E.; Steuerman, D.; Diehl, M.; Boukai,
A.; Wong, E. W.; et al. Angew. Chem., Int. Ed. 2001, 40, 1721−1725.
(b) Shvartzman-Cohen, R.; Nativ-Roth, E.; Baskaran, E.; Levi-
Kalisman, Y.; Szleifer, I.; Yerushalmi-Rozen, R. J. Am. Chem. Soc.
2004, 126, 14850−14857. (c) Giulianini, M.; Waclawik, E. R.; Bell, J.
M.; De Crescenzi, M.; Castrucci, P.; Scarselli, M.; et al. Appl. Phys. Lett.
2009, 95, 1−3.
(11) (a) Liu, J.; Rinzler, A. G.; Dai, H.; Hafner, J. H.; Kelley, B. R.;
Boul, P. J.; et al. Science 1998, 280, 1253−1256. (b) Zhang, Y.; Shi, Z.;
Gu, Z.; Iijima, S. Carbon 2000, 38, 2055−2059.
water gave stable dispersions at 5 °C. When heated at 50 °C,
aggregation occurred after 2 min and was complete within 10
min. Cooling and stirring the aggregated solution in an ice bath
for 10 min restored a stable dispersion (Figure 4b and Figure
S11 for PSD analysis). These temperature cycles were repeated
three times and led to the same aggregation/dispersion process.
In summary, we have demonstrated that supramolecular
chemistry is a convenient way to control stable dispersions of
CNTs in any type of solvent. Thymine-grafted CNTs can be
stabilized thanks to a host/guest association using any DAT-
end-grafted polymer chains soluble in a given solvent. Steric
repulsions between polymer chains ensure dispersions stability.
Moreover, we have shown that CNT supramolecular
dispersions can be destabilized by adding a dissociating additive
selective of the host/guest association. In this manner, Thy-
grafted CNTs are recycled and can be redispersed in another
solvent in presence of DAT-end-polymer chains soluble in that
solvent. To conclude, such CNT-Thy/DAT-polymer com-
plexes are universally dispersible and recyclable. In addition,
possibility of reversible switch on and off dispersion/
aggregation by heating or cooling the solution offers desirable
opportunities. This concept could be applied to other particles
that are difficult to disperse and stabilize in solution.
ASSOCIATED CONTENT
* Supporting Information
Experimental details and characterization data. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
■
(12) (a) Tasis, D.; Tagmatarchis, N.; Bianco, A.; Prato, M. Chem. Rev.
2006, 106, 1105−1136. (b) Singh, P.; Campidelli, S.; Giordani, S.;
Bonifazi, D.; Bianco, A.; Prato, M. Chem. Soc. Rev. 2009, 38, 2214−
2230.
Corresponding Authors
ludwik.leibler@espci.fr
corinne.soulie@espci.fr
(13) (a) Li, H.; Cheng, F.; Duft, A. M.; Andronov, A. J. Am. Chem.
Soc. 2005, 127, 14518−14524. (b) Campidelli, S.; Ballesteros, B.;
Filoramo, A.; Diaz, D. D.; Torre, G.; Torres, T.; et al. J. Am. Chem. Soc.
2008, 130, 11503−11509.
(14) Registration, Evaluation and Authorisation of Chemicals, in
European Union; Environmental Protection Agency, in United States.
(15) (a) Quintana., M.; Prato, M. Chem. Commun. 2009, 6005−6007.
(b) Quintana, M.; Traboulsi, H.; Llanes-Pallas, A.; Marega, R.;
Bonifazi, D.; Prato, M. ACS Nano 2012, 6, 23−31.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We are grateful to Huntsman for providing Jeffamine M2005.
We are indebted to Renaud Nicolay for discussions and
■
̈
guidance in the PS-DAT synthesis. We thank Sylvie Tence-
Girault and Ilias Iliopoulos for helpful discussions.
́
(16) (a) Di Crescenzo, A.; Kopf, I.; Pieraccini, S.; Masiero, S.; Del
Canto, E.; Spada, G. P.; Girodani, S.; Fontana, A. Carbon 2012, 50,
4663−4672. (b) Llane-Pallas, A.; Yoosaf, K.; Traboulsi, H.; Mohanraj,
J.; Seldrum, T.; Dumont, J.; Minoia, A.; Lazzaroni, R.; Armaroli, N.;
Bonifazi, D. J. Am. Chem. Soc. 2011, 133, 15412−15424.
(17) KThy/DAT = 890, KThy‑Thy = 4.3, KDAT‑DAT = 2.2 M−1, as measured
in CDCl3 from Beijer, F. H.; Sijbesma, R. P.; Vekemans, J. A. J. M.;
Meijer, E. W.; Kooijman, H.; Speck, A. L. J. Org. Chem. 1996, 61,
6371−6380.
REFERENCES
■
(1) (a) Rahmat, M.; Hubert, P. Compos. Sci. Technol. 2011, 72, 72−
84. (b) Ma, P.-C.; Siddiqui, N. A.; Marom, G.; Kim, J. K. Composites
Part A 2010, 41, 1345−1367. (c) Bose, S.; Khare, R. A.; Moldenaers,
P. Polymer 2010, 51, 975−993.
(2) (a) Miaudet, P.; Badaire, S.; Maugey, M.; Derre, A.; Pichot, V.;
Launois, P.; Poulin, P.; Zakri, C. Nano Lett. 2005, 5, 2212−2215.
(b) Yang, B. X.; Shi, J. H.; Pramoda, K. P.; Goh, S. H. Nanotechnology
2007, 18, 1−7. (c) Kanagraj, S.; Varanda, F. R.; Zhiltsova, T. V.;
Oliveira, M. S. A.; Simoes, J. A. O. Compos. Sci. Technol. 2007, 67,
3071−3077.
(18) Dyke, C. A.; Tour, J. M. Chem.Eur. J. 2004, 10, 812−817.
(19) Cortese, J.; Soulie-Ziakovic, C.; Tence-Girault, S.; Leibler, L. J.
́
́
Am. Chem. Soc. 2012, 134, 3671−3674.
(20) Coessens, V.; Nakagawa, Y.; Matyjaszewski, K. Polym. Bull.
1998, 40, 135−142.
(3) Bauhofer, W.; Kovacs, J. Z. Compos. Sci. Technol. 2009, 69, 1486−
1498.
́
(21) Brosse, A.-C.; Tence-Girault, S.; Piccione, P. M.; Leibler, L.
(4) Sekitani, T.; Noguchi, Y.; Hata, K.; Fukushima, T.; Aida, T.;
Someya, T. Science 2008, 31, 1468−1472.
(5) (a) Mukai, K.; Asaka, K.; Sugino, T.; Kiyohara, K.; Takeuchi, I.;
Terasawa, N.; Futaba, D. N.; Fukushima, T.; Aida, T. Adv. Mater.
2009, 21, 1582−1585. (b) Ahir, S. V.; Huang, Y. Y.; Terentjev, E. M.
Polymer 2008, 49, 3841−3854.
(6) (a) Wilder, J. W.; Venema, L. C.; Rinzler, A. G.; Smalley, R. E.;
Dekker, C. Nature 1998, 391, 59−62. (b) Peters, M. J.; McNeil, L. E.;
Lu, J. P.; Kahn, D. Phys. Rev. B 2000, 61, 5939−5944.
Polymer 2008, 49, 4680−4686.
(22) Steed, J. W. ; Atwood, J. L. Supramolecular Chemistry, 2nd ed.;
Wiley: Chippenham, U.K., 2009.
NOTE ADDED AFTER ASAP PUBLICATION
Scheme 1 has been updated. The revised version was re-posted
on November 30, 2012.
■
D
dx.doi.org/10.1021/ja309029n | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX