DOI: 10.1002/adma.200701530
Surface Wrinkles for Smart Adhesion**
By Edwin P. Chan, Erica J. Smith, Ryan C. Hayward, and Alfred J. Crosby*
Nature has demonstrated that surface patterns can control
adhesion and release. For example, the attachment devices of
geckos
[1,2]
and some insects
[3,4]
are decorated with fibrillar
structures designed specifically for locomotion. Inspired by
the fibrillar arrays, there have been significant efforts in mim-
icking these materials
[5–12]
to develop synthetic analogs as
“smart” adhesives, that is, where the geometry of the patterns
can tailor the adhesion of the material. However, there are
currently two limitations in the design of patterned adhesives.
First, because of either materials selection or geometric de-
sign, the adhesive properties are not always reversible over
the course of multiple attachment–detachment cycles.
[5,6,11,12]
Second, as the current fabrication strategies are primarily
based on lithographic or soft-lithographic approaches, there
are issues related to the fabrication of these materials in an ef-
ficient and scaleable manner. These two limitations must be
addressed in order to successfully adopt these materials as
“smart” adhesives.
In this work, we present an alternative design strategy for a
reusable smart adhesive that uses surface wrinkles as patterns
to control the adhesion of a poly(n-butyl acrylate) (PnBA)
elastomer. The wrinkle patterning process is based on the
swelling of a laterally confined polymer film. We demonstrate
the ability to control the wrinkle dimensions as a simple ap-
proach to design a “smart” adhesive, where the control of ad-
hesion is determined by the wavelength of the wrinkles based
on a mechanism known as contact splitting. Our materials de-
sign offers several advantages over previous approaches in-
cluding: (i) enhanced control of adhesion provided by well-de-
fined surface wrinkle patterns, (ii) convenience and simplicity
of the fabrication process without expensive lithography for
patterning, and (iii) amenability to patterning a wide variety
of polymer systems.
Our wrinkled PnBA elastomer is fabricated using a pattern-
ing process based on surface wrinkling. The general prerequi-
site for surface wrinkling is the development of a critical com-
pressive stress; hence, the formation of surfaces wrinkles has
been observed for a variety of materials using different exter-
nal stimuli.
[13–20]
We use a process that is loosely based on a
wrinkling approach developed by Southern and Thomas on
the swelling of a laterally-confined elastomer (Fig. 1a).
[21]
We
begin by preparing a PnBA elastomeric film with defined film
thickness (h) by depositing a photocurable nBA monomer for-
mulation onto a glass substrate. Next, the same photocurable
nBA solution is deposited onto the PnBA film surface, which
swells the elastomer. As the PnBA film is pinned to the rigid
substrate, the lateral expansion of the polymer film is con-
fined. This osmotic stress coupled with lateral confinement
leads to the development of a net compressive force and the
formation of an elastic instability, that is, surface wrinkling.
Following the brief swelling process, the wrinkled film is irra-
diated with ultraviolet (UV) light, which photopolymerizes
the swelling agent and stabilizes the surface wrinkles to form
the final wrinkled PnBA elastomer (Fig. 1b).
The wavelength of the surface wrinkles is an important
length-scale that provides adhesion control. Based on wrink-
ling theory for a soft, polymer gel,
[16]
the wavelength (k)of
the wrinkles is directly proportional to the film thickness (h),
k
h, since this is the only relevant length-scale. We validate
this scaling in Figure 1c and show that k is directly propor-
tional to h. In addition, we demonstrate our ability to tune the
wavelength of the wrinkle patterns. Here, h of the film can be
easily defined by controlling the volume of nBA solution de-
posited onto the glass substrate.
Following fabrication of the wrinkled PnBA films, we quan-
tify their adhesion using a custom-built contact adhesion test
(CAT) instrument. The CAT is an axi-symmetric probe-type
test that measures the force (P), displacement, and interfacial
contact history by forming an interface (between the probe
surface and the adhesive) and subsequently separating this in-
terface. An example for the wrinkled PnBA is represented in
Figure 2a. Briefly, the test begins by forming an interface be-
tween the wrinkled adhesive and a flat-glass probe at a fixed
displacement rate (ca. 3 lms
–1
) that is controlled by a nano-
positioner stepper motor. The adhesive is continually com-
pressed until it reaches a maximum compressive force, P
c
,
which corresponds to the establishment of a maximum inter-
facial contact area. Upon reaching P
c
, the nanopositioner im-
mediately reverses direction at the same displacement rate
until complete interfacial separation occurs. To quantify the
adhesion for our samples, we use the separation strength (r
s
),
C
O
M
M
U
N
I
C
A
T
I
O
N
Adv. Mater. 2008, 20, 711–716 © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 711
–
[*] Prof. A. J. Crosby, E. P. Chan, Prof. R. C. Hayward
Polymer Science and Engineering Department
University of Massachusetts
Amherst, 01003 MA (USA)
E-mail: crosby@mail.pse.umass.edu
E. J. Smith
Department of Chemistry
Trinity College
Hartford, 06106 CT (USA)
[**] The authors thank Justin Turner and Steve Koback of Zygo Corpora-
tion for assistance and use of the Zygo NewView optical profiler in
characterizing the wrinkled surface. E.J.S. thanks the NSF-MRSEC
REU program for financial support. Funding for this research is pro-
vided by NSF CAREER Award DMR-0349078 and 3M Non-tenured
Faculty Research Award.

which is the tensile force of separation (P
s
) normalized by the
projected contact area (A = pa
p
2
) that is defined by the flat
probe of radius (a
p
). The separation strength is a relevant de-
scriptor of adhesion as it can be related to the materials’ de-
fined adhesion energy (G
c
) if the exact interfacial history can
be defined.
For P
c
below a critical value, P
s
depends upon P
c
. This ob-
servation is similar to the preload condition for fibrillar adhe-
sives.
[5,11]
Although this dependence provides additional
means to tailor adhesion and release, the advantages and
mechanisms of this compressive load-dependent adhesion
control will be discussed in a subsequent paper. Here, we fo-
cus on the adhesion of the wrinkled interface in the compres-
sive load-independent regime.
A summary of r
s
values obtained for our wrinkled PnBA is
presented in Figure 2b. The results demonstrate that the ad-
hesion can be easily tailored by designing wrinkle patterns of
specific wavelengths. Comparing the r
s
values between the
wrinkled adhesive with the smallest wavelength (k ≈ 325 lm)
to that of the largest wavelength (k ≈ 505 lm), we show a
threefold increase in adhesion simply by changing the wrinkle
wavelength.
Using the previously described proce-
dure, there are three possible contribu-
tions to the enhanced control of adhe-
sion. The first contribution is related to
the discretization of the interfacial con-
tact. This mechanism, termed contact
splitting, is determined by the pattern
geometry and will be discussed below.
The second contribution is associated
with changes in the PnBA material’s
properties. As our patterning approach
involves swelling of the PnBA elasto-
mer, the elastic modulus and adhesion
energy may differ from the non-swollen
PnBA. The third contribution is related
to the thickness of the adhesive. For a
film thickness significantly smaller than
its lateral dimensions, Crosby and co-
workers have demonstrated that
changes in the thickness of the adhesive
can play a significant role in the separa-
tion mechanism and ultimately the ad-
hesive properties of the material.
[22]
As
the wavelength is controlled by the film
thickness, the changes in thickness will
also contribute to the changes in adhe-
sion.
To isolate the effects of the wrinkle
patterns, we make replicates of the
wrinkled adhesives by micromolding
(details in the Experimental). This elim-
inates the adhesive contributions asso-
ciated with swelling and confinement
effects due to thickness change since
the materials’ properties and thickness of the replicated
PnBA are homogeneous. We again measure their adhesion
using CAT and compare the adhesion between the replicated
wrinkled PnBA against the smooth, non-wrinkled PnBA. The
results (Fig. 3a) are presented as a normalized separation
strength (r
s,n
, determined by the ratio of the wrinkle versus
the smooth) versus the wrinkle wavelength. We can make sev-
eral comments based on these results. First, the wrinkle pat-
terns do enhance adhesion related to the non-wrinkled ana-
log. In fact, for the wrinkled PnBA with the lowest
wavelength, r
s,n
≈ 1.5, indicating a 50% increase in adhesion
compared with the non-wrinkled system. Second, for highest
wavelength system, r
s,n
≈ 1, in other words, no enhancement
occurs. Third, the results indicate a trend and suggest that the
enhancement in separation force should scale inversely with
wavelength. In the following section, we discuss the mecha-
nism of enhancement and develop a scaling relationship to de-
fine the contributions of the wrinkle wavelength.
At one extreme, our wrinkled surface is analogous to a
polymer film with a characteristic surface roughness. The ef-
fects of surface roughness on adhesion have been a constant
theme for the past several decades. Previous work has shown
C
O
M
M
U
N
I
C
A
T
I
O
N
712 www.advmat.de © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2008, 20, 711–716
Figure 1. a) Procedure for fabricating a wrinkled adhesive by swelling a laterally confined PnBA film.
The right-hand side of the figure illustrates the mechanism of wrinkling formation. b) Optical pro-
file of a 1 × 1 mm
2
wrinkled PnBA adhesive. c) Tuning of wrinkle wavelength (k) by controlling the
initial thickness (h) of the PnBA film. We define the film thickness experimentally by depositing a
controlled volume of the nBA solution onto the substrate to form the PnBA film. The dashed line il-
lustrates the linear relationship between k and h.