brush-type layer having a thickness of only about 2.0 nm. Based
on this model, the hydrodynamic effect was about six times more
significant as compared to the effect of pore enlargement. There-
fore, instead of the change of the pore size, it is likely that the
rapidly increasing diffusivity is induced by the hydrophobicity of
the PNIPAAm-grafted surface. On a hydrophobic pore surface,
dextran molecules become less hydrogen-bonded, decreasing
the interaction with the PNIPAAm coated surface that can
enhance the effective partitioning of FITC–dextran into the
nanopore.10 To additionally confirm the effect of hydrophobic
coating on dextran diffusion, the diffusivity through the
nanopores modified with 4-tert-butylcatechol was also exam-
ined. The diffusion through PNIPAAm-grafted nanopores
was as fast as that through 4-tert-butylcatechol modified
nanopores at 42 1C (2.31 Â 10À8 cm2 sÀ1), which obviously
demonstrates that the hydrophobic transition of surface wett-
ability in response to the temperature change is responsible for
such dramatic increase in the diffusion of FITC–dextran
through the PNIPAAm-grafted nanopores.
We have described a facile method to thermally control the
surface wettability of nanopores by grafting catechol-tethered
PNIPAAm to the surface of alumina. The thermosensitive
change of surface wettability was used as a means of controlling
the diffusional permeability of a solute through nanopores. This
new underlying mechanism of grafting is distinct from the
previously reported systems based on the physical pore blocking
of entangled polymer chains.9 Although optimal conditions need
to be further explored, it is expected that the simple surface coating
for temperature-sensitive grafting of nanoporous materials can be
applied to chemical separation or biological sensors.
Fig. 3 Permeability of unmodified (blue circle), PNIPAAm-grafted
(red triangle) and 4-tert-butylcatechol (green diamond), 20 nm pore
AAO membrane measured using FITC–dextran at temperatures below
(a) and above (b) the LCST of PNIPAAm; (c) thermally responsive
permeability of the PNIPAAm-grafted, 20 nm pore AAO membrane
(eqn (S1) used for this plot is described in ESIw).
The measured diffusivities of FITC–dextran within the bare
membrane were 8.48 Â 10À9 cm2 sÀ1 at 15 1C and 1.02 Â
10À8 cm2 sÀ1 at 42 1C. The slightly higher diffusivity at 42 1C
seems to be caused by the increased thermal motion of
FITC–dextran. Within the nanopores grafted with PNIPAAm-ct,
the diffusivity at 15 1C was 7.94 Â 10À9 cm2 sÀ1, which is similar to
the value of the bare membrane though the actual pore size of the
PNIPAAm-grafted membrane was reduced to 64% of the size
of the bare pore. The diffusivity of FITC–dextran through the
nanopores was calculated using a ‘hindered diffusion’ model
equation to clarify the contribution of the partial unblocking
to the increased diffusivity. The effective diffusion coefficient is
determined by both the steric restriction resulted from the pore
blocking and the interaction between the surface of the pore
wall and solutes, as described in ESIw in more detail. The
calculated diffusivity in the polymer-grafted pores at 15 1C was
This study was financially supported by a grant of the
Korean Health Technology R&D Project, Ministry of Health
and Welfare (A040041 and A111552), Republic of Korea.
Notes and references
1 M. S. Yavuz, Y. Y. Cheng, J. Y. Chen, C. M. Cobley, Q. Zhang,
M. Rycenga, J. W. Xie, C. Kim, K. H. Song, A. G. Schwartz, L. H. V.
Wang and Y. N. Xia, Nat. Mater., 2009, 8, 935–939.
2 (a) M. Hesampour, T. Huuhilo, K. Makinen, M. Manttari and
M. Nystrom, J. Membr. Sci., 2008, 310, 85–92; (b) M. Yang,
L. Y. Chu, H. D. Wang, R. Xie, H. Song and C. H. Niu, Adv.
Funct. Mater., 2008, 18, 652–663.
3 B. H. Kim, D. H. Lee, J. Y. Kim, D. O. Shin, H. Y. Jeong,
S. Hong, J. M. Yun, C. M. Koo, H. Lee and S. O. Kim, Adv.
Mater., 2011, 23, 5618–5622.
4 J. H. Waite and M. L. Tanzer, Science, 1981, 212, 1038–1040.
5 (a) K. Kaleem, F. Chertok and S. Erhan, Nature, 1987, 325,
328–329; (b) H. Lee, K. Lee, I. K. Kim and T. G. Park, Adv.
Funct. Mater., 2009, 19, 1884–1890.
6.11 Â 10À9 cm2
s
À1, which was only 77.0% of the experi-
mental value, indicating that the hydrodynamic effect is more
significant than pore blocking in the hindered diffusion
through the PNIPAAm-grafted nanopores. Very interestingly,
the diffusivity of FITC–dextran at 42 1C greatly increased
to 2.03 Â 10À8 cm2 sÀ1, which is 2.5 times larger than that at
15 1C (Fig. 3). The dramatically increased diffusivity at 42 1C
is not only attributed to the reduced size or blocking of
nanopores, which is the reason widely adopted in the similar
systems.2,9 In our system, the extended length of PNIPAAm
used in this study was much smaller than the size of nano-
pores. The calculated grafting density of the PNIPAAm chain
on the AAO surface was 3.94 nm2 per chain (eqn (S3) in ESIw).
As its radius of gyration is ca. 1.0 nm based on a random
coil model, the grafted PNIPAAm chains seem to form a
6 R. Narain, M. Gonzales, A. S. Hoffman, P. S. Stayton and
K. M. Krishnan, Langmuir, 2007, 23, 6299–6304.
7 (a) C. W. Scales, A. J. Convertine and C. L. McCormick, Biomacro-
molecules, 2006, 7, 1389–1392; (b) J. W. Chan, B. Yu, C. E. Hoyle
and A. B. Lowe, Polymer, 2009, 50, 3158–3168.
8 M. S. Toprak, B. J. McKenna, J. H. Waite and G. D. Stucky,
Chem. Mater., 2007, 19, 4263–4269.
9 (a) P. F. Li, R. Xie, J. C. Jiang, T. Meng, M. Yang, X. J. Ju,
L. H. Yang and L. Y. Chu, J. Membr. Sci., 2009, 337, 310–317;
(b) O. Schepelina and I. Zharov, Langmuir, 2007, 23, 12704–12709.
10 L. F. Scatena, M. G. Brown and G. L. Richmond, Science, 2001,
292, 908–912.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 9227–9229 9229