Angewandte
Chemie
In a typical procedure, the nanoparticles were dispersed in
Owing to the reversibility of the nanoparticle aggregation,
thin, approximately 150 mm films of syndiotactic poly(methyl
all types of photopatterned images gradually self-erased as
the aggregates reverted to free NPs. When the films were left
on the benchtop and exposed to ambient laboratory light, the
erasure times were hours to several days depending on c; in
the dark, the images remained stable for over a week. When,
however, the films were heated or exposed to intense visible
light, erasure took only a few to tens of seconds. Once erased,
the films could be rewritten multiple times. Figure 3a shows a
series of images written into and erased from AuNP films. We
have verified that the quality of these images did not
deteriorate with the number of write/erase cycles for at
least several hundred cycles (Figure 3a,b). Finally, all photo-
patterned films were flexible, and the images remained intact
upon mechanical bending or twisting (Figure 3c).
[
17]
methacrylate) (sPMMA) organogel
laminated between
two flexible poly(vinyl chloride)-coated poly(ethylene ter-
ephthalate) sheets (up to 5 ꢀ 5 cm). Despite relatively low
concentrations of these NPs, but owing to their high
ꢁ
18
2
extinction cross-sections (4.2 ꢀ 10
m
for gold and 1.2 ꢀ
ꢁ17
2
[18]
1
0
m for silver ), the films were colored brightly: red
for AuNPs and yellow for AgNPs. Importantly, in the absence
of UV irradiation, the NPs in the gel had UV/Vis spectra
nearly identical to those of free NPs in toluene (l
max,Au
[
18]
ꢂ 525 nm, l
ꢂ 420 nm ), thus indicating that the par-
max,Ag
ticles were not aggregated. When, however, the films were
exposed to UV light, they changed color in the irradiated
regions, and the degree of this change depended on the
duration of UV irradiation (Figure 1c,d and Figure 2). In the
absence of irradiation, the images written into the films
gradually self-erased (Figures 2 and 3), with the erasure times
controlled by the composition of the mSAMs coating the NP
inks.
The times required to write (t ) and erase (t ) high-
w
e
contrast images depended on and could be controlled by the
intensity of light (UV for writing, Vis for erasure) and by the
fractional surface coverage c of the MUA ligands. Specifically,
tw decreased with increasing IUV and varied between approx-
ꢁ
2
The changes observed upon UV irradiation (365 nm,
imately 20 s for I = 0.7 mWcm
and 0.8 s for IUV =
UV
ꢁ2
ꢁ2
IUV = 0.7–10 mWcm ) were due to the rapid trans-to-cis
isomerization of the azobenzene groups of MUA. This
isomerization caused a significant increase of the dipole
moment (m ꢂ 5 debye for the cis form vs. m ꢂ 1 debye for the
10 mWcm . For a given value of IUV, tw decreased with
increasing c (Figure 4a). On the other hand, the times
required to erase the images decreased with increasing
intensity of the visible light (e.g., t ꢂ 24 h in IVis
=
e
[
19]
ꢁ2
ꢁ2
trans isomer ), which mediated attractive interactions
10 nWcm lighting vs. t ꢂ 20 s for I = 0.8 mWcm halo-
e
Vis
[
13]
between the NPs. The strength of these interactions
gen lamp) and with decreasing c (Figure 4b).
depended on the surface concentrations of the MUA tethered
to the NPs. For low coverage of MUA (c < 0.23) the dipole–
dipole forces between NPs were too weak to cause aggrega-
tion; for c > 0.34, irreversible aggregation occurred even in
the absence of irradiation. For 0.23 < c < 0.34, the NPs
aggregated reversibly into disordered, metastable aggregates
To explain these effects, we first observe that the
equilibrium solubility of NPs depends on the fractional
surface coverage of the cis-MUA isomer, ccis (0 < ccis < c, see
the Supporting Information). When ccis exceeds a critical
value (c * ꢂ 0.23) the NP dispersion becomes unstable, and
cis
NPs aggregate until equilibrium is reestablished between the
dispersed and the aggregated phases. The characteristic
duration of this process is controlled primarily by the trans-
to-cis isomerization of the MUA ligands on the NP surfaces.
(
ca. 150 nm in diameter by TEM), which disintegrated in the
absence of UV light (Figure 1c). This reversible NP aggrega-
tion was the basis of rewritable and self-erasing materials.
Formation of the metastable aggregates in the UV-
exposed regions manifested itself by the broadening and
red-shifting of the NPsꢁ SPR band and by concomitant color
changes. The degree of these changes reflected the proportion
of aggregated NPs (see the Supporting Information) and
depended on the nature of the NPs in the film and on the
irradiation dose: AuNP inks evolved gradually from red to
pale blue (Figure 1d, left) and AgNP inks from yellow to
violet (Figure 1d, right). Figure 2a shows examples of two-
color images created in AuNP and AgNP films by photomask
1
Upon UV irradiation, ccis evolves in time as ccis = ccis
[1ꢁexp(ꢁk t)], where k is the first-order rate constant for
tc
tc
19]
[
cis–trans isomerization (this rate increases approximately
linearly with the intensity of the incident UV light ), and
ccis is the cis-MUA coverage at the photostationary state
[
20]
1
1
(here, ccis ꢂ c) corresponding to all MUA being in the cis
form. Therefore, the characteristic writing time t is given by
w
ꢁ1
c (t ) ꢂ c * or t ꢂ k ln[c/(cꢁc *)]. For example, for
cis
w
cis
w
tc
cis
ꢁ
2
intensity
ꢂ 0.1 s ,
I
= 0.7 mWcm
the writing time in a gel containing NPs with
corresponding
to
k
tc
UV
ꢁ
1 [19,20]
ꢁ
2
exposure to IUV = 10 mWcm for approximately 0.8 s. In
c = 0.33 is estimated to be t ꢂ 15 s; for more intense
w
ꢁ
2
ꢁ1
Figure 2b, the violet text was written into the film using a UV
irradiation (I = 10 mWcm and k ꢂ 1 s ) the estimated
UV
tc
ꢁ
2
pen (I = 10 mWcm ) scanned above the surface at about
3
writing time in the same gel decreases to t ꢂ 1 s, in agreement
UV
ꢁ
w
1
mms . Remarkably, multicolor images could be created
with experiment. Note than since k / I , even faster writing
tc
UV
with one nanoparticle ink by varying the irradiation dose over
different regions of the film. For example, the flowers and the
Union Jack shown in Figure 2c were created in AuNP films by
should be possible with more intense irradiation.
We emphasize that the erasure is not due to the diffusion
of the aggregates, which would “smear” the images, but only
to their disassembly, which changes the color by weakening
the electrodynamic coupling between proximal NPs. The
diffusivities of single NPs and of the supraspheres in the
homogeneous sPMMA organogel may be estimated as D =
ꢁ2
irradiating (IUV = 10 mWcm ) the purple regions for 0.8 s,
purple-bluish regions for 4 s, and pale-blue regions for 10 s.
Figure 2d shows a multicolor AgNP film into which two
images were written consecutively, one of them onto a bent
film giving the overall illusion of curvature.
0
.75 [21]
D exp(ꢁaRf ), where D = kT/6pmR is the Stokes–Ein-
0
0
stein diffusivity, kT is the thermal energy, m is the solvent
Angew. Chem. Int. Ed. 2009, 48, 7035 –7039
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7037