Reversible transformation of hydrophobicity and hydrophilicity of aligned carbon nanotube arrays and buckypapers by dry processes

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Reversible transformation of hydrophobicityand hydrophilicity of aligned carbon nanotube arraysand buckypapers by dry processes

H.Z. Wang a, Z.P. Huang b, Q.J. Cai c, K. Kulkarni b, C.-L. Chen c, D. Carnahan b, Z.F. Ren a,*

a Department of Physics, Boston College, Chestnut Hill, MA 02467, USAb NanoLab Inc., Newton, MA 02458, USAc Teledyne Scientific & Imaging, LLC, Thousand Oaks, CA 91360, USA

A R T I C L E I N F O

Article history:

Received 20 July 2009

Accepted 27 October 2009

Available online 1 November 2009

A B S T R A C T

Dry treatment using a combination of UV and ozone can readily change the surface of ver-

tically aligned carbon nanotubes from superhydrophobic to superhydrophilic. This treat-

ment is also effective for buckypapers. Heating in a vacuum at an elevated temperature

(650–750 �C) can reverse the surface state from superhydrophilic to superhydrophobic.

The UV & ozone treatment causes the least amount of damage to the stripe-like carbon

nanotube patterns. The effect of rough surface on apparent contact angles of CNT forests

was discussed to explain the origin of superhydrophilicity and superhydrophobicity.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Surface modification of carbon nanotubes (CNTs) has recently

attracted a great deal of attention [1–6], because the surface

characteristics considerably affect applications of CNTs in

the fields of biomedical applications [7–10], biosensors [11–

13], catalysts supports [14], and composites [15–22]. Wettabil-

ity of CNTs by liquids is one of the most important surface

properties, which normally expressed by contact angle [4–

6,23,24]. It is believed that the contact angle of CNTs is deter-

mined by chemical composition and surface roughness [6,24–

26]. Due to the nanosized diameter of CNTs, microscopically,

almost all of the pristine CNTs should inevitably provide a

rough surface. Accordingly, most of the surface modifications

of CNTs have been focused on tailoring the chemical compo-

sitions [1,27–29].

Normally, as-grown CNTs by chemical vapor deposition

(CVD) are insoluble in most solvents, which seriously hinders

their applications [28]. There are several methods to induce

the transition of CNTs surface from superhydrophobic (con-

tact angle >150�) [2,4] or hydrophobic (contact angle >90�) to

hydrophilic (contact angle <90�) or superhydrophilic (contact

angle <5�) [3], such as acid treatments [28,30], microwave

treatment [6], oxygen plasma etching [5], and incorporation

of heteroatoms on the surface of CNTs [3,29,31,32]. However,

in general, strong acid treatments can significantly make oxi-

dative damages to the tips and sidewalls of CNTs, introduce

new oxygenated groups to the CNTs [33], and decrease the

electrical and thermal conductivities and mechanical

strength [31]. Although the microwave treatment can modify

the wettability in dry conditions, it may seriously weaken the

adhesion of CNTs to the substrates on which CNTs are grown

[1]. Due to the fact that most of the nanosized catalyst (Fe)

particles remain in the interfacial region between the bottom

of CNTs and the substrates [34,35], the microwave radiation

may melt or oxidize the catalyst nanoparticles. Oxygen plas-

ma etching may also modify the wettability of CNTs, but the

hydrophobic-to-hydrophilic transition can only occur on the

upper portion (top layer) of the CNTs film [5]. On the other

hand, all of the above mentioned treatments are concentrated

on the transition from hydrophobic to hydrophilic, only a few

studies could be found regarding CNTs surface transition

0008-6223/$ - see front matter � 2009 Elsevier Ltd. All rights reserved.doi:10.1016/j.carbon.2009.10.041

* Corresponding author: Fax: +1 617 552 8478.E-mail address: [email protected] (Z.F. Ren).

C A R B O N 4 8 ( 2 0 1 0 ) 8 6 8 – 8 7 5

avai lab le at www.sc iencedi rec t .com

journal homepage: www.elsev ier .com/ locate /carbon

Author's personal copy

from hydrophilic to hydrophobic, which is required in some

cases for CNT applications.

Recently, ultraviolet assisted ozone (UV & ozone) treat-

ment has been employed to functionalize single-walled CNTs

[36,37] and multi-walled CNTs [38,39] , or to induce superhy-

drophobic to superhydrophilic transition on biomimetic

nanostructured surfaces [40] and a-Fe2O3 nanoflakes film

[41]. In this paper, we report for the first time the transition

from superhydrophobicity to superhydrophilicity and the

mechanism of vertically aligned CNTs forest by UV & ozone

treatment and the reverse by heating-in-vacuum. Further-

more, the method is also applicable to buckypapers [2,42].

The controllable wettability of CNTs will significantly facili-

tate the fabrication of composites containing CNTs.

2. Experimental

A polished silicon wafer, with orientation (100), thickness

�0.4 mm, p-type (boron doped), and thermal oxide thickness

2 lm on the surface, was cut into 10 · 10 mm2 pieces, fol-

lowed by (1) cleaning (immersed in acetone with ultra-sonica-

tion for 5 min) and blowing dry with a nitrogen gas; (2)

deposition of catalyst layers of Fe 3 nm/Al 3 nm/Fe 3 nm by

sputtering; (3) growth of vertically aligned multi-walled CNTs

by CVD with gases of ethylene (C2H4, 110 sccm) and hydrogen

(H2, 100 sccm) at 745 �C and 760 Torr for 30 min [43]. To obtain

stripe-like CNTs patterns (width � 100 lm and spac-

ing � 50 lm), two additional steps were needed before and

after the catalyst deposition, respectively. One was the fabri-

cation of stripe-like photoresist (Shipley 1818) pattern on

cleaned silicon substrate by photolithography, and another

was the removal of the photoresist with acetone. The as-

grown CNTs were examined by scanning electron microscope

(SEM, JEOL JSM-6340F) and transmission electron microscope

(TEM, JEOL JSM-2010F).

Two kinds of dry surface treatments were employed for the

reversible superhydrophobic and superhydrophilic transition.

In order to convert the CNT film from superhydrophobic to

superhydrophilic, an UV & ozone Dry Stripper (Samco, Model

UV-1) was employed with oxygen gas supply at 0.2 L/min,

working at 50 �C. For the transition from superhydrophilic

back to superhydrophobic, a tube furnace (Thermolyne,

21100) was used to heat the samples from 500 �C to 750 �Cfor 10 min. In this study, the wettability of CNTs, which di-

rectly related to superhydrophobicity or superhydrophilicity,

was revealed by contact angles (with water droplets on CNTs

film) and recorded by a digital camera.

3. Results and discussion

3.1. Effect of rough surface on apparent contact angles

Microscopically, the top surface of vertically aligned CNTs

(length � 500 lm and diameter � 15 nm) can be regarded as

extremely rough, with many tips (ends) oriented vertically.

It has been reported that microscaled roughness plays an

important role on the wettability of the surface [44]. When li-

quid can fill into the grooves of the protrutions on the surface

[45], Wenzel model [46] may be applied as the following,

cos hwa ¼ r � cos h ð1Þ

where hwa and h are apparent contact angle on rough surface

and ideal contact angle on flat surface, respectively, and the

surface roughness factor r (P1) is the ratio of actual surface

area to the projected (flat) area. It is noted that the r in Eq.

(1) is an amplification factor of the cosh, which can either

make hwa � h in the case of a wetting surface (h < 90�), or

hwa � h in the case of an unwetting surface (h > 90�).

If the liquid can not penetrate into the grooves of the pro-

trutions on the surface [45], it has been supposed that a drop-

let on the surface is supported by both the protrutions and air

pocket [47]. In this case, Cassie–Baxter model [48] can be used

as the following,

cos hca ¼ �1þ /sðcos hþ 1Þ ð2Þ

where hca and h are apparent contact angle on rough surface

and ideal contact angle on flat surface, respectively, and /s

is the solid fraction on the whole supporting surface. Some

researchers [49,50] pointed out that Eq. (2) should hold for

substrates either superhydrophobic or very large roughness

factor, whereas Eq. (1) can hold for all kinds of hydrophilic

and slight hydrophobic surfaces (h just above 90�) [47]. Based

on theoretical analysis and experimental results, however,

Bico et al. [51] believed that Wenzel model can be used in

the hydrophilic regions (h < 90�) and Cassie–Baxter model

can be used in hydrophobic region (h > 90�).In order to obtain more details from the above two classi-

cal models, a simplified rough surface was proposed in the in-

set in Fig. 1a, with spacing (D), height (H), and diameter (d) of

the protrutions. It is assumed that each protrution locates in

the center of the square region (edge length D + d). Based on

these assumptions, Eq. (1) can be modified as,

cos hwa ¼ 1þ pH=d

ð1þ D=dÞ2

" #cos h ð3Þ

That is; the roughness factor r ¼ 1þ pH=d

ð1þ D=dÞ2ð4Þ

Supposed the ratio H/d = 5, 20, and 100, respectively, the ef-

fect of D/d ratio on the roughness factor is illustrated in

Fig. 1a. It is shown that ratio H/d plays an important role on

the roughness factor when the ratio D/d is less than 5.

For hydrophobic surface, Eq. (2) can be rewritten as,

cos hca ¼ �1þ pð1þ cos hÞ

4ð1þ D=dÞ2ð5Þ

It is reasonable that Eq. (5) is independent of the height (H)

of the protrutions, since a droplet may completely hold on the

top tips of the protrutions. Fig. 1b reveals that the ideal con-

tact angles (h = 91�, 95�, 100�, see the inset in Fig. 1b) is insen-

sitive to hca in all the range of the ratio (D/d). Once the ratio (D/

d) reaches more than 5, it should be easy to obtain a superhy-

drophobic surface no matter how much the ideal contact an-

gle (h > 90�) is.

It is known that the ideal contact angle for flat graphite is

h = 84–86� [52,53], which is close to the critical angle (h = 90�)[54]. By surface modifications [54,55], accordingly, it is not dif-

ficult to change the surface from hydrophilic to hydrophobic.

Regarding the effect of surface roughness on the apparent

contact angles (see Fig. 1), it is understandable that the wetta-

C A R B O N 4 8 ( 2 0 1 0 ) 8 6 8 – 8 7 5 869

Author's personal copy

bility of CNT surface can be switched between superhydro-

philic and superhydrophobic due to the nanoscaled diameter

of CNTs.

3.2. UV & ozone treatment for superhydrophilicity

When a drop of water (10–15 lL) was dripped on the as-grown

CNTs film, due to its pristine superhydrophobic property, the

water droplet can be completely supported on the surface

without any wetting, with a contact angle around 158�, as

shown by the image (0 s) in Fig. 2a. It is shown in Fig. 2a that

the UV & ozone treatment can very rapidly change the chem-

ical composition and the CNTs surface from superhydropho-

bic to superhydrophilic in less than 10 s. During the UV &

ozone treatment, the internally generated ozone (O3) can

decompose into oxygen gas (O2) and highly reactive oxygen

atom (O) under the radiation of UV (wavelength 200–300 nm)

light. Simultaneously, on the surface (tips and sidewalls) of

each CNT, discontinuous spots, imperfections, and dangling

bonds such as –H, –OH, –COOH, @O, and @CO [27,33,38,56]

are exposed to the UV, and then excited or dissociated by

the UV radiation. The extremely reactive O atoms and fresh

O2 molecules can readily react or combine with the defects

and dangling bonds on CNTs. Moreover, it is reported that

some undecomposed O3 molecules can directly bind to the

carbon atoms on the outer shell of CNTs [36,37], which is sim-

ilar to the ozonolysis treatment [2,33]. Herein, the oxidation

process can quickly finish a sigmoidal wetting transition [40]

on the CNTs surface. In Fig. 2a, the contact angle did not

change too much at the first 6 s, which suggested that there

is an incubation period for the surface oxidation. Once the

surface returns to hydrophilic state (h changes from 90�+ to

90��) in the next 3 s (6–9 s in Fig. 2a), the apparent contact an-

gle quickly decreases to around 5� due to the effect of rough

surface.

It is suggested that Wenzel model should hold on this

superhydrophilic CNT surface (hwa ¼ 5�). In this case the

roughness factor is around r1 = 11.43 by Eq. (1) with h = 85�

Fig. 1 – Diagrams illustrating simplified models of rough

surfaces and their applications: (a) effect of D/d ratio on

roughness factor r for Wenzel model, and (b) effect of D/d

ratio on the value of cos hca for Cassie–Baxter model.

Fig. 2 – Contact angle dependence of UV & ozone treatment

time (a) and of heating-in-vacuum treatment temperature.

The SEM image in (a) shows the vertically aligned CNTs and

TEM image in (b) shows the diameter of the CNTs.

870 C A R B O N 4 8 ( 2 0 1 0 ) 8 6 8 – 8 7 5

Author's personal copy

[52,53]. However, based on the inset in Fig. 1a, H = 500 lm,

d = 15 nm (TEM image in Fig. 2b), and D = 50 nm (D + d = 30–

100 from the CNT site density 1010–1011 cm2 [57]), the rough-

ness factor r2 = 5577.80 by Eq. (4). The extremely big difference

between r1 and r2 could be explained by the following two ma-

jor reasons. Firstly, there is a limit to the roughness factor in

Wenzel model, which is rmax ¼ cos hwa = cos h ¼ cos 0�=

cos 85� ¼ 11:47. It is noted that r1 is already close to the rmax.

When r > rmax, the roughness factor can not enhance the

reduction of the hwa any more (just like r2) since the contact an-

gle has reached the limit (hwa ¼ 0�). Secondly, the CNTs in the

aligned arrays are very much tilted, entangled, and over-

lapped with each other (see the SEM image in Fig. 2a), so

the effective height of CNTs should be much smaller than

the value H = 500 lm which directly resulted in the huge value

of r2 by Eq. (4). Regarding the D/d = 3.33 and r1 = 11.43, in

Fig. 1a the effective height of CNTs should be Heff < 100d =

1500 nm. In addition, at present we are not clear about the

accuracy of Wenzel model at the study of nanoscaled wetta-

bility since the model was originally obtained from micro-

scaled characteristics of wetting.

3.3. Heating-in-vacuum treatment for superhydrophobicity

Reversibly, a facile and controllable method on the transition

was developed to modify the surface property from superhy-

drophilic to superhydrophobic. When the superhydrophilic

CNTs film was heated in a CVD tube furnace [43] with a pres-

sure �0.02 Torr for 10 min, at 500–650 �C, the contact angle in-

creased from �5� (superhydrophilic) to >150�(superhydrophobic). It was also observed that further heating

at higher temperature (>650 �C) could not significantly in-

crease the magnitude of the apparent contact angle, sup-

ported by the fact that the angle at 650 �C is very close to

that at 750 �C, as shown in Fig. 2b. Heating-in-vacuum at less

than 500 �C will not significantly affect the superhydrophilic-

ity of the CNTs.

CNTs grown by CVD are often heated in air at around

600 �C to remove adsorbates from the surface [58]. This clean-

ing process is also a method to change the CNTs film from

superhydrophobic to superhydrophilic [27]. Apart from the

in situ TEM work at 1652 �C [58], little literature has been

found on heating CNTs in vacuum at high temperature. As

shown in Fig. 2b, when the superhydrophilic sample was

heated in vacuum (�0.02 Torr) at 500 �C for 10 min, the appar-

ent contact angle increased from 5� to 18�. It is suggested that

the heating can evaporate most physically adsorbed mole-

cules, such as O, O2, O3 (from UV & ozone treatment), H2O

and CO2 from air, which causes the contact angle increases.

When heating at 600–750 �C for 10 min, it was reported that

the chemically bonded O3 can be removed [36], and most or-

ganic composition be burned out. It is supposed that the

ozone removal is one of the dominant processes in the incre-

ment of contact angle at the heating-in-vacuum. Once the

heating makes the contact angle (h) passing the critical value

(h = 90� [54]), Cassie–Baxter model can be used to reveal the ef-

fect of rough surface. As illustrated in Fig. 1b, at D/d = 3.33, the

rough surface can significantly reduce the value of cos hca and

produce a superhydrophobic surface regardless of how much

the h passing through the critical value (90�).

3.4. Treatments for buckypapers

In addition, it was also found that the UV & ozone treatment

is also effective to modify buckypapers from superhydropho-

bic to superhydrophilic. When the buckypaper (�20 mm in

diameter, and �100 lm in thickness) was treated by UV &

ozone for 5 min (5 min instead of 10 s is explained below) at

50 �C, superhydrophilic surface was achieved. As shown in

Fig. 3a, a water droplet can completely spread on the whole

top surface of the buckypaper with a contact angle <5� (super-

hydrophilic). When the treated buckypaper is put on water, it

rapidly sinks into water owing to its excellent wettability, as

shown in Fig. 3b. It is known that the surface of buckypaper

(Fig. 4a) has much fewer tips (ends) than that of the as-grown

CNTs film (Fig. 4b), and the tips of CNTs are very reactive and

easy to be modified due to the high density of dangling bonds

and defects.

Accordingly, it is reasonable that the surface modification

of buckypapers is much more difficult than that of the as-

grown CNTs film. That’s why it takes 5 min (rather than 10 s

in Fig. 2a) for UV & ozone treatment to complete the superhy-

drophobic to superhydrophilic transition for buckypaper

(Fig. 3a).

However, the above heating-in-vacuum operation can not

fulfill the surface modification of buckypapers from superhy-

drophilic to superhydrophobic at 745 �C for 30 min. It is inter-

esting to find that a CVD CNTs growth operation

(C2H4 � 110 sccm, H2 � 100 sccm, 745 �C, and 760 Torr) [43]

for 10 min can fulfill the reverse transformation from super-

hydrophilic to superhydrophobic, as shown in Fig. 3c. This

phenomenon can be explained by the following two reasons.

One is that the 5 min UV & ozone treatment may have intro-

duced much more structural defects on the CNTs walls than

the 10 s operation, and X-ray photoelectron spectroscopy

(XPS) and Raman spectra confirmed that UV & ozone treat-

ment could functionalize CNTs surface more aggressively

than the microwave irradiation [39]. Another reason is that

the mending of structural defects on CNTs surface. With en-

ough carbon source provided by thermal CVD operation at

745 �C, it is suggested that some carbon atoms, which decom-

posed from ethylene (C2H4), can fix the defect sites by cova-

lent bonding to restore the original surface state of the

CNTs. It is noted that such a growth procedure does not yield

any new CNTs on buckypapers since there is no catalyst. If

the treated buckypaper was put into the bottom of the water,

it can immediately float up to the water surface (Fig. 3d) due

to its superhydrophobicity.

3.5. Applications of UV & ozone treatment

Apart from being facile, rapid, and dry, the UV & ozone treat-

ment is nondestructive. For a stripe-like CNTs pattern with

width of �100 lm, spacing �50 lm, and depth of trench

(length of CNTs) �300 lm (see Fig. 5a), UV & ozone treatment

can readily modify the pristine CNTs from superhydrophobic

to superhydrophilic, and keep the pattern intact without con-

tamination. As a comparison, Fig. 5b shows that the liquid

treatment (soaked into acid solution, rinsed by water, and

then dried in air) resulted in (1) collapse or tilt of some CNTs

fins; (2) breakdown of a few CNTs fins (inset in Fig. 5b); and

C A R B O N 4 8 ( 2 0 1 0 ) 8 6 8 – 8 7 5 871

Author's personal copy

(3) contaminations (many debris in a trench in Fig. 5b). It is be-

lieved that the acid treatment will inevitably cause imperfec-

tions to the CNTs pattern.

Fig. 3 – Wettability of buckypapers. (a) Hydrophilic: water spreading on buckypaper, (b) hydrophilic: buckypaper sinking into

water, (c) hydrophobic: a water droplet on buckypaper, (d) hydrophobic: buckypaper floating on water.

Fig. 4 – SEM surface micrographs of CNTs. (a) as-prepared

buckypapers, (b) as-grown CNTs films (top view).

Fig. 5 – SEM micrographs of stripe-like CNTs patterns after

superhydrophobic-to-superhydrophilic treatments. (a) UV &

ozone treatment (50 �C for 4 min); and (b) acid treatment

(immerged in HNO3 26% for 5 min, rinsed with deionized

water, and dried in air).

872 C A R B O N 4 8 ( 2 0 1 0 ) 8 6 8 – 8 7 5

Author's personal copy

For large area CNTs film (without pattern), UV & ozone treat-

ment is also better, proved by the results shown in Fig. 6. The

same CNTs samples were treated by baking-in-air (600 �C for

30 min, Fig. 6a), acid treatment (immerged into 2% HCl solution

for 9 min, and then rinsed by deionized water, Fig. 6b), and UV

& ozone treatment (50 �C for 4 min, Fig. 6c). After the superhy-

drophilic treatments, all of the samples were dripped with a

few water drops from the top surface till the sample is soaked

with water, and then dried in air. SEM micrographs shown in

Fig. 6 revealed the morphologies of the dried CNTs films. Due

to the effect of surface tension of water during the drying, a

lot of CNTswere peeled off the substrate and aggregated in sep-

arate domains, which resulted in many huge cracks shown in

Fig. 6a. It is proposed that the baking-in-air operation damaged

the weak adhesion of CNTs to the substrate, and made most of

CNTs easily lift off. As for the acid treatment, Fig. 6b shows a

few small cracks. For UV & ozone treated samples, Fig. 5c dis-

plays that the whole surface was preserved much better than

Fig. 6a and b with very limited cracks.

4. Conclusions

Under UV radiation at 50 �C, ozone (O3) and its decomposed

components (O2, O) vigorously react with CNTs by oxidation

and ozonolysis in dry condition, and rapidly transform the

surface state of CNTs from superhydrophobicity to superhy-

drophilicity. This treatment also works for buckypapers. A re-

verse transformation, from superhydrophilic to

superhydrophobic, was also achieved by heating the CNTs

in vacuum at high temperature. It is believed that the rough

surface of CNT forest plays a very important role in all the

surface wettability transitions. However, buckypaper needs

CVD CNTs growth conditions to reverse its transformation

from superhydrophilicity to superhydrophobicity due to the

requirement of mending the structural defects on CNT walls.

Compared with acid treatment and baking-in-air method, UV

& ozone treatment is much better in the sense that it is a fac-

ile, fast, clean, and nondestructive.

Acknowledgment

This research is sponsored by DARPA/MTO under contract

#N66001-08-C-2009 with Dr. Tom Kenny as the program man-

ager. The views, opinions, and/or findings contained in this

article/presentation are those of the author/presenter and

should not be interpreted as representing the official views

or policies, either expressed or implied, of the Defense Ad-

vanced Research Projects Agency or the Department of

Defense.

R E F E R E N C E S

[1] Pastine SJ, Okawa D, Kessler B, Rolandi M, Llorente M, Zettl A,et al. A facile and patternable method for the surfacemodification of carbon nanotube forests usingperfluoroarylazides. J Am Chem Soc 2008;130(13):4238–9.

[2] Kakade B, Mehta R, Durge A, Kulkarni S, Pillai V. Electric fieldinduces, superhydrophobic to superhydrophilic switching inmultiwalled carbon nanotube papers. Nano Lett2008;8(9):2693–6.

[3] Han JT, Kim SY, Woo JS, Lee G-W. Transparent, conductive,and superhydrophobic films from stabilized carbonnanotube/silane sol mixture solution. Adv Mater2008;20:3724–7.

Fig. 6 – SEM micrographs of CNTs film after different

superhydrophobic to superhydrophilic treatments. (a) baked

in air at 600 �C for 30 min, dripped with water drops on the

surface, and then dried in air; (b) acid treatment (immerged

into HCl 2% for 9 min), rinsed with deionized water, and

then dried in air; (c) UV & ozone treatment for 4 min, dripped

with water drops on the surface, and then dried in air.

C A R B O N 4 8 ( 2 0 1 0 ) 8 6 8 – 8 7 5 873

Author's personal copy

[4] Hosono E, Fujihara S, Honma I, Zhou HS. Superhydrophobicperpendicular nanopin film by the bottom–up process. J AmChem Soc 2005;127(39):13458–9.

[5] Li PH, Lim XD, Zhu YW, Yu T, Ong C-K, Shen ZX, et al.Tailoring wettability change on aligned and patterned carbonnanotubes film for selective assembly. J Phys Chem B2007;111(7):1672–878.

[6] Kakade BA, Pillai VK. Tuning the wetting properties ofmultiwalled carbon nanotubes by surface functionalization. JPhys Chem C 2008;112(9):3183–6.

[7] Guo Y, Shi DL, Cho H, Dong ZY, Kulkarni A, Pauletti GM, et al.In vivo imaging and drug storage by quantum-dot-conjugatedcarbon nanotubes. Adv Funct Mater 2008;18:2489–97.

[8] Cai D, Mataraza JM, Qin Z-H, Huang ZP, Huang JY, Chiles TC,et al. Highly efficient molecular delivery into mammaliancells using carbon nanotubes spear. Nat Methods2005;2(6):449–54.

[9] Nednoor P, Chopra N, Gavalas V, Bachas LG, Hinds BJ.Reversible biochemical switching of ionic transport throughaligned carbon nanotubes membranes. Chem Mater2005;17(14):3595–9.

[10] Rojas-Chapana JA, Correa-Duarte MA, Ren ZF, Kempa K,Giersig M. Enhanced introduction of gold nanoparticles intovital acidothiobacillus ferrooxidans by carbon nanotube-based microwave electroporation. Nano Lett 2004;4(5):985–8.

[11] Wang J, Musameh M. Carbon nanotube/teflon compositeelectrochemical sensors and biosensors. Anal Chem2003;75(9):2075–9.

[12] Lin YH, Lu F, Tu Y, Ren ZF. Glucose biosensors based oncarbon nanotube nanoelectrode ensembles. Nano Lett2004;4(2):191–5.

[13] Zhang MG, Smith A, Gorski W. Carbon nanotube-chitosansystem for electrochemical sensing based on dehydrogenaseenzymes. Anal Chem 2004;76(17):5045–50.

[14] Serp P, Corrias M, Kalck P. Carbon nanotubes and nanofibersin catalysis. Appl Catal A 2003;253:337–58.

[15] Liu ZM, Han BX. Synthesis of carbon-nanotube compositesusing supercritical fluids and their potential applications.Adv Mater 2008;20:1–5.

[16] Yu C, Kim YS, Kim D, Grunlan JC. Thermoelectric behavior ofsegregated-network polymer nanocomposites. Nano Lett2008;8(12):4428–32.

[17] Luo C, Zuo XL, Wang L, Wang E, Song SP, Wang J, et al.Flexible carbon nanotube-polymer composites films withhigh conductivity and superhydrophobicity made by solutionprocess. Nano Lett 2008;8(12):4454–8.

[18] Biercuk MJ, Llaguno MC, Radosavljevic M, Hyun JK, JohnsonAT, Fischer JE. Appl Phys Lett 2002;80(15):2767–9.

[19] Qian D, Dickey EC, Andrews R, Rantell T. Load transfer anddeformation mechanisms in carbon nanotubes-polystyrenecomposites. Appl Phys Lett 2000;76(20):2868–70.

[20] Dujardin E, Ebbesen TW, Hiura H, Tanigaki K. Capillarity andwetting of carbon nanotubes. Science 1994;265(5180):1850–2.

[21] Wardle BL, Saito DS, Garcia EJ, Hart AJ, Villoria RGD,Verploegen EA. Fabrication and characterization of ultrahigh-volume-fraction aligned carbon nanotubes-polymercomposites. Adv Mater 2008;20:2707–14.

[22] Xie X-L, Mai Y-W, Zhou X-P. Dispersion and alignment ofcarbon nanotubes in polymer matrix: a review. Mater Sci EngR 2005;49:89–112.

[23] Lahann J, Mitragotri S, Tran T-N, Kaido H, Sundaram J, ChoiIS, et al. A reversibly switching surface. Science2003;299:371–4.

[24] Liu H, Zhai J, Jiang L. Wetting and anti-wetting on alignedcarbon nanotubes films. Soft Matter 2006;2:811–21.

[25] Blossey R. Self-cleaning surface-virtual realities. Nat Mater2003;2:301–6.

[26] Erbil HY, Demirel AL, Avci Y, Mert O. Transformation of asimple plastic into a superhydrophobic surface. Science2003;299:1377–80.

[27] Kanyo T, Konya Z, Kukovecz A, Berger F, Dekany I, Kiricsi I.Quantitative characterization of hydrophilic–hydrophobicproperties of MWNTs surface. Langmuir 2004;20(5):1656–61.

[28] Hu CG, Hu SS. Surface design of carbon nanotubes foroptimizing the adsorption and electrochemical response ofanalytes. Langmuir 2008;24(16):8890–7.

[29] Ebbesen TW, Hiura H, Bisher ME, Treacy MMJ, Shreeve-KeyerJL, Haushalter RC. Decoration of carbon nanotubes. AdvMater 1996;8(2):155–7.

[30] Nguyen CV, Delzeit L, Cassell AM, Li J, Han J, Meyyappan M.Preparation of nucleic acid functionalized carbon nanotubearrays. Nano Lett 2002;2(10):1079–81.

[31] Hsin Y-L, Lai J-Y, Hwang KC, Lo S-C, Chen F-R, Kai JJ. Rapidsurface functionalization of iron-filled multi-walled carbonanotubes. Carbon 2006;44:3328–35.

[32] Sun TL, Song WL, Jiang L. Control over the responsivewettability of poly(N-isopropylacrylamide) film in a largeextent by introducing an irresponsive molecule. ChemCommun 2005;13:1723–5.

[33] Banerjee S, Hemraj-Benny T, Wong SS. Covalent surfacechemistry of single-walled carbon nanotubes. Adv Mater2005;17(1):17–29.

[34] Stadermann M, Sherlock SP, In J-B, Fornasiero F, Park HG,et al. Nano Lett 2009;9:738–44.

[35] Zhang RY, Amlani I, Baker J, Tresek J, Tsui RK, Fejes P.Chemical vapor deposition of single-walled carbonnanotubes using ultrathin Ni/Al film as catalyst. Nano Lett2003;3:731–5.

[36] Cai LT, Bahr JL, Yao YX, Tour JM. Ozonation of single-walledcarbon nanotubes and their assemblies on rigid self-assembled monolayers. Chem Mater 2002;14:4235–41.

[37] Ma RJ, Yoon D, Chun K-Y, Baik S. The effect of UV/ozonetreatment on the electrical transport behavior of single-walled carbon nanotubes arrays. Chem Phys Lett2009;474:158–61.

[38] Najafi E, Kim J-Y, Han S-H, Shin K. UV-ozone treatment ofmulti-walled carbon nanotubes for enhanced organic solventdispersion. Coll Surf A 2006;284–285:373–8.

[39] Moon K-S, Lin W, Jiang HJ, Ko H, Zhu LB, Wong CP. Surfacetreatment of MWCNT array and its polymer composites forTIM application. In: 2008 Electronic components andtechnology conference, Florida, May 27–30, 2007. p. 234–7.

[40] Han JT, Kim S, Karim A. UVO-tunable superhydrophobic tosuperhydrophilic wetting transition on biomimeticnanostructured surface. Langmuir 2007;23:2608–14.

[41] Yan B, Tao JG, Pang C, Zheng Z, Shen ZX, Huan CHA, Yu T.Reversible UV-induced ultrahydrophobic-to-ultrahydrophilictransition in a a-Fe2O3 nanoflakes film. Langmuir2008;24:10569–71.

[42] Endo M, Muramatsu H, Hayashi T, Kim YA, Terrones M,Dresselhaus MS. ‘Buckypaper’ from coaxial nanotubes.Nature 2005;433:476.

[43] Xiong G-Y, Wang DZ, Ren ZF. Aligned millimeter-longnanotubes arrays grown on single crystal magnesia. Carbon2006;44:969–73.

[44] Patankar NA. Mimicking the lotus effect: influence of doubleroughness structures and slender pillars. Langmuir2004;20(19):8209–13.

[45] Martines E, Seunarine K, Morgan H, Gadegaard N, WilkinsonCDW, Riehle MO. Superhydrophobicity andsuperhydrophilicity of regular nanopatterns. Nano Lett2005;5(10):2097–103.

[46] Wenzel RN. Resistance of solid surface to wetting by water.Ind Eng Chem 1936;28:988–94.

874 C A R B O N 4 8 ( 2 0 1 0 ) 8 6 8 – 8 7 5

Author's personal copy

[47] Lafuma A, Quere D. Superhydrophobic states. Nat Mater2003;2:457–60.

[48] Cassie ABD, Baxter S. Wettability of porous surface. TransFaraday Soc 1944;40:546–51.

[49] Bico J, Thiele U, Quere D. Wetting of textured surfaces.Colloids Surf A 2002;206:41–6.

[50] Patankar NA. On the modeling of hydrophobic contact angleson rough surfaces. Langmuir 2003;19:1249–53.

[51] Bico J, Marzolin C, Quere D. Pearl drops. Europhys Lett1999;47(2):220–6.

[52] Fowkes FM, Harkins WD. The state of monolayers adsorbedat the interface solid-aqueous solution. J Am Chem Soc1940;62:3377–86.

[53] Morcos I. Electrocapillary studies on partially immersedmercury-plates electrode in DMF–water solutions. J ChemPhys 1972;56(8):3996–4000.

[54] Lau KKS, Bico J, Teo KBK, Chhowalla M, Amaratunga GAJ,Milne WI, et al. Superhydrophobic carbon nanotubes forest.Nano Lett 2003;3(12):1701–5.

[55] Yan AH, Xiao XC, Kulaots I, Sheldon BW, Hurt RH. Controllingwater contact angle on carbon surfaces from 5� to 167�.Carbon 2006;44:3113–48.

[56] Hirsch A. Functionalization of single-walled carbonnanotubes. Angew Chem Int Ed 2002;41(11):1853–9.

[57] Qu LT, Dai LM, Stone M, Xia ZH, Wang ZL. Carbon nanotubesarrays with strong shear binding-on and easy normal lifting-off. Science 2008;322:238–42.

[58] Martin GL, Schwoebel PR. Transmission electron microscopyobservation of CNT morphology before and after heating invacuum. Surf Sci 2007;601:1705–8.

C A R B O N 4 8 ( 2 0 1 0 ) 8 6 8 – 8 7 5 875