The miniaturization of nanometer-sized multicolor fluorescent features is of continuous significance for counterfeit security features, data storage, and sensors. Recent advances in engineering of stimuli-responsive supramolecular polymeric materials that respond upon exposure to heat or mechanical force by changing their fluorescence characteristics open new opportunities as functional lithographic resists. Here, we demonstrate the patterning of a thermochromic supramolecular material by thermal scanning probe lithography (t-SPL), an emerging nanofabrication technique, which allows for ultrafast indentation with a heated probe, resulting in both fluorescent and topographic nanofeatures. t-SPL indentation reveals a linear relationship between the temperature at which material softening occurs and the indentation force in the range from 200 to 500 nN. The softening temperature decreases as the heating time increases from 4 μs to 1 ms, following time-temperature superposition behavior. Our results herein confirm that the fluorescence contrast, perceivable as a shift from red to green, was obtained by kinetic trapping of the dissociated state due to ultrarapid cooling when the probe is removed. We use t-SPL to create highly customized fluorescence patterns up to 40 × 40 μm2 in size with a spatial resolution of 86 nm and change the pitch size to modify the fluorescence intensity when observed by fluorescence microscopy. As an application, multifaceted security features with nanometer resolution are explored.
The miniaturization of nanometer-sized multicolor fluorescent features is of continuous significance for counterfeit security features, data storage, and sensors. Recent advances in engineering of stimuli-responsive supramolecular polymeric materials that respond upon exposure to heat or mechanical force by changing their fluorescence characteristics open new opportunities as functional lithographic resists. Here, we demonstrate the patterning of a thermochromic supramolecular material by thermal scanning probe lithography (t-SPL), an emerging nanofabrication technique, which allows for ultrafast indentation with a heated probe, resulting in both fluorescent and topographic nanofeatures. t-SPL indentation reveals a linear relationship between the temperature at which material softening occurs and the indentation force in the range from 200 to 500 nN. The softening temperature decreases as the heating time increases from 4 μs to 1 ms, following time-temperature superposition behavior. Our results herein confirm that the fluorescence contrast, perceivable as a shift from red to green, was obtained by kinetic trapping of the dissociated state due to ultrarapid cooling when the probe is removed. We use t-SPL to create highly customized fluorescence patterns up to 40 × 40 μm2 in size with a spatial resolution of 86 nm and change the pitch size to modify the fluorescence intensity when observed by fluorescence microscopy. As an application, multifaceted security features with nanometer resolution are explored.
Thermal scanning probe lithography (t-SPL)
is a method that enables nanoscale patterning through the interaction
of a heated probe in contact with a substrate material.[1] The heat transferred from the tip to the substrate
induces physical and/or chemical modifications in a target material,
which are delimited to the tip–substrate contact area, typically
on the order of a few tens of nanometers in diameter. While the technology
was initially developed for data storage, t-SPL has become an emerging
lithography method for nanopatterning of temperature-sensitive and
functional materials, such as polymers,[2,3] graphene oxide,[4] molecular glasses,[5] metal alloys,[6] and semiconductors.[7−9] Closed-loop lithography, a key feature of t-SPL, enables thermal
patterning of a substrate material with the tip being heated and in
situ recording of the surface topography with the tip being at room
temperature. This enables rapid prototyping of three-dimensional (3D)
patterns into temperature-sensitive resists,[2,3] which
can subsequently be transferred into silicon or used for lift-off.[10−12] Chemical reactions, triggered by the heated probe in contact with
the substrate surface, enable nanoscale functional patterns, such
as electrically conducting reduced graphene oxide[4] or semiconducting photoluminescent poly(p-phenylenevinylene) nanostructures.[7,8] t-SPL is not
limited to organic materials and has been applied on GeTe phase change
materials to create crystalline, electrically conductive, and optically
absorptive nanoscale patterns from an amorphous, transparent, and
electrically insulating substrate.[9] In
comparison to laser-assisted nanoimprint lithography[13] or ultrafast thermal nanoimprint lithography,[14] t-SPL is a digital patterning technique which
does not require a mask.Supramolecular polymer glasses are
an emerging category of stimuli-responsive materials that exhibit
a high elastic modulus at room temperature in the solid state.[15−17] These materials can be converted into low-viscosity liquids when
heated above their glass transition temperature (Tg), on account of the thermally driven, reversible disassembly
into the parent low-molecular-weight building blocks.[18] These materials find applications in the field of sensing[19] and self-healing.[20] The concept of supramolecular polymer glasses has been recently
extended to mechanoresponsive luminescent (MRL) and thermoresponsive
luminescent (TRL) supramolecular polymers by synthesizing a MRL/TRL
dye with supramolecular binding motifs. Self-complementary hydrogen-bonding
ureido-4-pyrimidinone (UPy) groups were used to promote the self-assembly[21] of an excimer-forming cyano-substituted oligo(p-phenylenevinylene) (cyano-OPV) derivative[22−24] into a supramolecular polymer (UPy–OPV–UPy, Figure a).[25] This material has the following characteristics: (i) it
can be readily molded into self-supporting objects of various shapes,
(ii) it displays the thermomechanical characteristics of a supramolecular
polymer glass, (iii) it offers three different emission colors in
the solid state, and (iv) it exhibits both MRL and TRL behavior on
account of stimuli-induced modification of the molecular packing,
which in turn impacts the material’s emission characteristics
(Figure b).[25] While the ability to control the fluorescence
color of blends of cyano-OPVs and glassy amorphous polymers via the
extent of dye aggregation was exploited before in three-dimensional
optical data storage systems, heat-transfer processes did not permit
a voxels with dimensions of less than a few micrometers.[26] The resolution of two-photon laser-induced aggregate
switching is limited to a few micrometers due to the diffraction-limited
spot size and lateral heat transfer during exposure. In a more recent
publication, a quaternary data storage system using thermoresponsive
cyano-OPV and a photoresponsive dye in a polymer matrix was proposed
to create a material selectively responsive to heat and light.[27] While this material allowed four different colors
to be obtained under activation with light, heat, or a combination
of both, the resolution was limited to 200–300 μm. To
overcome such limitations in resolution, we exploit the capability
of t-SPL to apply extremely fast heating rates on the order of 107 K/s. In addition, we exert simultaneously well-controlled
mechanical forces in the range of 200–500 nN at the probe–sample
contact point. This concept is demonstrated by reversibly switching
the state of assembly and therewith the fluorescence characteristics
of UPy–OPV–UPy under extremely short heat and force
pulses created by a heated t-SPL probe (Figure c–f). It is hypothesized that kinetic
trapping of the green fluorescent high-temperature OPV dissociated
state is facilitated due to short contact and fast heat dissipation
after the heated probe is removed. We present results on atomic force
microscopy (AFM) force spectroscopy of UPy–OPV–UPy films,
nanoindentation and nanopatterning by t-SPL, and the maximum achieved
resolution, and demonstrate topographic/fluorescent images as an example
of anticounterfeiting nanosystems.
Figure 1
Nanolithography with a color-switchable
fluorescent supramolecular polymer resist. (a) The chemical structure
of the UPy–OPV–UPy building block is shown, with the
excimer-forming fluorescent OPV core and the supramolecular UPy binding
motifs highlighted. (b) Fluorescence spectra of a bulk sample of UPy–OPV–UPy
at 180 and 25 °C. The inset shows pictures of the fluorescing
sample under UV illumination. (c) Schematic representation of the
single-step fabrication of multicolor patterns using thermal scanning
probe lithography. (d) Illustration of the polymer film topography
and molecular assembly, with the OPV cores aggregated in excimer-forming
stacks and the UPy binding motifs dimerized. (e) Indentation with
a hot t-SPL probe locally liquefies the material and disassembles
the building blocks, which results in a green fluorescence. (f) Upon
rapid probe removal, the high-temperature state is quenched so that
a green fluorescing area is retained.
Nanolithography with a color-switchable
fluorescent supramolecular polymer resist. (a) The chemical structure
of the UPy–OPV–UPy building block is shown, with the
excimer-forming fluorescent OPV core and the supramolecular UPy binding
motifs highlighted. (b) Fluorescence spectra of a bulk sample of UPy–OPV–UPy
at 180 and 25 °C. The inset shows pictures of the fluorescing
sample under UV illumination. (c) Schematic representation of the
single-step fabrication of multicolor patterns using thermal scanning
probe lithography. (d) Illustration of the polymer film topography
and molecular assembly, with the OPV cores aggregated in excimer-forming
stacks and the UPy binding motifs dimerized. (e) Indentation with
a hot t-SPL probe locally liquefies the material and disassembles
the building blocks, which results in a green fluorescence. (f) Upon
rapid probe removal, the high-temperature state is quenched so that
a green fluorescing area is retained.
Materials and Methods
UPy–OPV–UPy
was synthesized as previously described and an in-depth structural
characterization of the molecule is recently reported.[9] Films for our studies were fabricated by melting UPy–OPV–UPy
in powder form on a silicon chip mounted on a hot plate at 180 °C
and doctor-blading the hot melt over the substrate. The samples were
then cooled under ambient conditions (1.3 °C/s) to room temperature,
thus allowing sufficient time for the cyano-OPV moieties to aggregate
into their thermodynamically favored stacked arrangement. As expected,
cooling was accompanied by a fluorescence shift from green to red
owing to the aggregation of cyano-OPV as depicted in the spectra and
images in Figure b.
The thickness of the resulting films was 20 ± 10 μm, as
measured by a mechanical profilometer (Alpha-Step 500, KLA-Tencore,
USA). The large deviation in film thickness between different films
originates from the manual fabrication and could be improved by using
a hot-press with pressure control at the expense of the simplicity
of the manual coating. Despite the occurrence of thickness variations
on the order of hundreds of nanometers over the entire sample, the
surface of the films was smooth over an area relevant for t-SPL. A
typical root-mean-square surface roughness of 5.7 ± 3.0 nm was
measured over an area of 50 × 50 μm2.Before t-SPL experiments are performed and to determine the thermomechanical
properties of UPy–OPV–UPy films, we performed AFM force
spectroscopy on heated samples using a Nano Wizard II AFM tool (JPK
Instruments AG, Germany) equipped with a cantilever for nanoindentation
(DNISP, Bruker AFM Probes, USA) and a heating stage (HTHS, JPK Instruments
AG, Germany). The samples were heated from 20 to 75 °C as monitored
with an Optris PI infrared camera. Between each temperature change,
enough time was given for the UPy–OPV–UPy film to thermally
equilibrate until the temperature reached a steady state. An indentation
depth of 30 nm was targeted for the AFM force spectroscopy measurements
on the bulk-heated sample so that it matches typical values used hereafter
in t-SPL. The force–distance curves were analyzed (JPK Software)
and all force spectra were corrected for cantilever deflection. The
unloading part of the force–distance curves were fitted (upper
50% of the curve) and analyzed according to the Hertz theory to obtain
the elastic modulus.After the nanoindentation
experiments with a cold probe and the sample on a hot plate, we now
shift to the t-SPL tool and perform experiments with a heated probe
on a sample stage that is kept at room temperature. t-SPL was carried
out using a commercial tool (NanoFrazor, SwissLitho AG, Switzerland),
which controls a silicon cantilever (Frazor-HPL1, Swisslitho AG, Switzerland)
with an integrated heater in close proximity to the tip for thermal
patterning. UPy–OPV–UPy-coated silicon chips were mechanically
clamped to the X–Y piezo positioning stage of the tool with
a metallic pin, which served as an electrical contact required for
electrostatic actuation of the cantilever to exert the local force
between probe and substrate. The governing patterning parameters for
the t-SPL process, temperature, force, and heating time were investigated
for UPy–OPV–UPy by consecutive variation of a single
parameter in a series of experiments as shown in Figure . The material response was
determined by scanning the patterned surface and evaluating the depth
of the indent and the height of the rim formed around the indent (pileup).
A script was written that computes the average depth of the indents
and the height of the pileup from the AFM topography scan (Supporting Information). The t-SPL tool was operated
in pulsed heating mode, whereby the heating of the tip occurred only
shortly before and during contact with the substrate; this procedure
increases the tip lifetime and also the thermal patterning reliability
on the supramolecular glass. Typical settings for heat pulses and
electrostatic force pulses were 5–20 and 5–10 μs,
respectively. The patterning speed was varied between 200 and 800
μm/s. The temperature at the tip–sample contact is not
a priori known because of a reduced heat transport from the resistive
heater to the conical tip, which is due to radiative and convective
losses, phonon scattering, and an unknown thermal contact resistance
at the interface.[28] To estimate the temperature
at the tip–sample interface, we used a heating efficiency of
0.25 between the heater and the substrate (Tcontact = 0.25(Theater – TRT) + TRT).[28,29] The heater temperature is computed by the software of the t-SPL
tool by fitting the knee point of the measured current–voltage
curve with the theoretical value for doped silicon.[30] The indentation force of the t-SPL probe, which was electrically
actuated, was calibrated by a procedure described in the Supporting Information. Optical fluorescence
microscopy images were acquired with an Olympus BX51 microscope, equipped
with a mercury lamp at a wavelength of 365 ± 15 nm (I-line) as
an illumination source and the emission was detected at wavelengths
larger than 420 nm. The illumination intensity of the fluorescence
microscope was carefully adjusted to a level where no bleaching occurred.
The optical contrast of fluorescence images was normalized by means
of an editing program (GIMP, www.gimp.org), for better visibility. All written t-SPL patterns were labeled
by an adjacent pattern to be clearly visible by optical microscopy
for identification during fluorescence microscopy (Figure S3). It has to be noted that the fluorescence imaging
is diffraction-limited (to about 310 nm), and patterns written with
the t-SPL, which are smaller than 310 nm can only be resolved in detail
by techniques such as AFM.
Figure 2
Thermomechanical indentation of the TRL supramolecular
polymer. (a) The elastic modulus of 20 ± 10 μm thick UPy–OPV–UPy
films obtained from AFM-based force spectroscopy is plotted against
the sample temperature. The inset depicts the experimental setup.
(b) AFM topography images obtained from t-SPL indentation by varying
the tip temperature between 130 and 310 °C and the indentation
force between 200 and 500 nN (tip actuation potential 6.0–9.0
V). The colored bars indicate whether an indent was formed (red) or
a pileup (blue). (c) The indentation depth and height of the pileup
around the indent from thermomechanical indentation with a hot probe
are plotted against the tip temperature. (d) The threshold temperature
above which an indent or a pileup is formed is plotted against the
indentation force and fitted with a linear function. (e) The temperature
threshold to form an indent is plotted against the heating time and
fitted using the Williams–Landel–Ferry equation for
a constant tip actuation potential of 7.8 V.
Thermomechanical indentation of the TRL supramolecular
polymer. (a) The elastic modulus of 20 ± 10 μm thick UPy–OPV–UPy
films obtained from AFM-based force spectroscopy is plotted against
the sample temperature. The inset depicts the experimental setup.
(b) AFM topography images obtained from t-SPL indentation by varying
the tip temperature between 130 and 310 °C and the indentation
force between 200 and 500 nN (tip actuation potential 6.0–9.0
V). The colored bars indicate whether an indent was formed (red) or
a pileup (blue). (c) The indentation depth and height of the pileup
around the indent from thermomechanical indentation with a hot probe
are plotted against the tip temperature. (d) The threshold temperature
above which an indent or a pileup is formed is plotted against the
indentation force and fitted with a linear function. (e) The temperature
threshold to form an indent is plotted against the heating time and
fitted using the Williams–Landel–Ferry equation for
a constant tip actuation potential of 7.8 V.
Results and Discussion
While thermomechanical properties,
such as the temperature-dependent elastic modulus and the glass-transition
temperature, for relatively thick (2 mm) UPy–OPV–UPy
films have been discussed in a previous publication,[25] no thermomechanical data is available for nanoindentation
of melt-processed films and fast heating rates as achievable in t-SPL.
Since the tip temperature in t-SPL is not a priori known and the tip
geometry is not well-defined, we perform both conventional AFM force
spectroscopy and t-SPL as a comparison to determine the thermomechanical
properties of the supramolecular polymer. The elastic modulus of UPy–OPV–UPy
films measured by AFM-based force spectroscopy is shown in Figure a as a function of
the substrate temperature. An elastic modulus of 1.5 ± 0.2 GPa
is measured at 20 °C. The modulus linearly decreases with increasing
substrate temperature until 68 °C is reached, where it sharply
drops by a factor of 20 to 10.5 ± 2.5 MPa. Above 73 °C,
indentation experiments were no longer possible due to adhesion of
the viscous melt to the tip. The data mirror the results of measurements
on bulk UPy–OPV–UPy films, for which a similar elastic
modulus of 1.5 ± 0.06 GPa was determined by a three-point bending
test and a similar mechanical transition was observed above 60 °C
by means of dynamic mechanical analysis.[25]In addition to the above-mentioned experiments with a cold
probe and a heated substrate, we used a heated t-SPL probe on a sample
at room temperature to test the thermomechanical properties of UPy–OPV–UPy.
In particular, the response of UPy–OPV–UPy was tested
using the t-SPL’s capability to apply heat on a sample area
of a few tens to hundreds of square nanometers at extremely high heating
rates ranging from 105 to 108 K/s. In a series
of experiments, we measure the indentation depth as a function of
probe temperature, indentation force, and heating time. Figure b shows a grid of AFM topography
images obtained from t-SPL indentation, which were created at tip–sample
contact temperatures in the range of 130–310 °C and a
tip actuation potential ranging from 6.0 to 9.0 V (indentation force
100–500 nN, see Supporting Information). Two main observations can be made from Figure b: first, to thermally modify the supramolecular
polymer with the heated probe, a minimum temperature (threshold temperature)
is necessary at a given tip actuation potential; second, if the tip
actuation potential is below 7.8 V, a pileup is formed, whereas above
an additional indent is formed. A small colored bar at the bottom
of each topography image in Figure b indicates whether an indent (red) or a pileup (blue)
is formed. The formation of a pileup without an indent can be attributed
to a sufficient high temperature to melt the material but the force
is too low to penetrate the material. In Figure c, the average indent depth and height of
the pileup from 16 indents are plotted as a function of the tip temperature
at a tip actuation potential of 7.8 V and a heating time of 8 μs,
which is a trade-off value between patterning speed and temperature
as will be explained later. The indentation depth increases with temperature
above 212 °C and as a consequence the pile-up height increases
due to displacement of the molten material. To systematically detect
this temperature by AFM, we define a threshold value for the indent
depth and pile-up height of 4 times the root-mean-square roughness
of the sample topography, which is usually in the range between 3
and 6 nm. The threshold value is chosen sufficiently high to avoid
false positive detection of an indent or pileup. The indentation experiment
was performed ten times with the same tip on different samples. All
patterns were clearly visible and the indents well reproduced. Three
experiments out of them were discarded due to surface topography anomalies,
which stem from the manual sample fabrication and caused large variations
in the data points. These surface irregularities can be avoided in
future experiments by using compression molding of UPy–OPV–UPy
on a substrate instead of doctor-blading of the melt. In Figure d, the average threshold
temperature from the seven consecutive indentation experiments is
plotted as a function of the indentation force. The temperature necessary
to form a modification in the material TThres decreases linearly with the indentation force F: TThres = T0 + mF with T0 = 380 °C and m = −0.42 °C/nN.
A linear dependence between threshold temperature and indentation
force has been observed for other polymers such as poly(methyl methacrylate)
(PMMA) and SU-8.[3] To determine the effect
of heating time on the threshold temperature, the tip–substrate
contact time was varied over 3 orders of magnitude as shown in Figure e. As expected, the
temperature required to form an indent decreases with the heating
time, known as time–temperature superposition, which is a typical
behavior found in polymers. We found that also in supramolecular polymers,
which differ considerably in their molecular assembly from regular
polymers, the temperature T at which material softening
occurs during thermal indentation can be described by the empirical
Williams–Landel–Ferry (WLF) equation: ln(tT/tr) = −c1(T – Tr)/(c2 + (T – Tr)), with a heating time tT and a reference time tr at a
reference temperature Tr. The curve was
best described with c1 = 13.0 ± 0.4, c2 = 13.1 ± 3 °C, and Tr = 88.0 ± 4 °C, whereby similar values have
been found for thermal indentation with PMMA and SU-8.[3] By extrapolation of the transition temperature using WLF
to long time scales, for example, 1 s, which is comparable to the
ones used in AFM spectroscopy, a threshold temperature of 88 °C
is obtained. This value is on the same order of magnitude as the glass-transition
temperature of UPy–OPV–UPy (73 °C).Another
essential benchmark for UPy–OPV–UPy as a thermal resist
is the spatial resolution, which was determined by patterning structures,
composed of bars ranging from 1.92 μm in width down to a single
line (SL) by consecutive indentation with the heated probe. The maximum
resolution is governed by the lateral tip apex diameter (27.7 ±
9 nm), the opening angle (∼60°), and the indentation depth
(10–30 nm). Figure a shows an AFM topography image of the pattern and one can
see that a topographic contrast is created between patterned and unpatterned
areas. For a line of consecutive indents, a fwhm resolution of 86
± 6 nm is obtained as shown in the topographical profiles in Figure c (black curve).
As a comparison, in two-photon laser lithography the voxel size was
previously limited to a volume of 3 × 3 × 6 μm3 due to heat diffusion from the heated spot.[26] In Figure c, two noticeable features in the pattern profiles are a trench and
a rim that are formed at the left and right side of each structure,
respectively. Our observations tell us that this is caused by displacement
of molten material through consecutive indentation with the heated
probe in the fast-scan direction when the pitch size (20 nm) is smaller
than the tip apex diameter. The following mechanism could explain
the surface topography reflected by the graph shown in Figure c: Starting from the left side,
a first indent is formed with the hot probe, and due to volume conservation
a rim is formed around the indent. When the pitch size is smaller
than the tip, during the consecutive indentation material is displaced
into the previous indent and hence no significant depth change can
be observed (Figure S4). The pileup of
material at the edges is more dominant on the right side than on the
left, independent of the width of the structures, which can be explained
by the patterning direction of the tip from left to right and the
tilt of the cantilever (6° with respect to the substrate surface).
After patterning and subsequent AFM microscopy, the sample was analyzed
by fluorescence microscopy. In Figure b, the corresponding fluorescence microscope image
of the t-SPL pattern from Figure a is shown. In this image the red fluorescence from
the background was suppressed by using a short-pass filter to obtain
a better contrast (for an image without filtering see Figure S5). A direct comparison of the resolution
between the topography and fluorescence intensity profiles from fluorescence
microscopy is not possible due to the diffraction limit of optical
microscopes (Figure c).
Figure 3
Comparison of the resolution achieved by t-SPL patterning of the
color-switchable fluorescent UPy–OPV–UPy and optical
readout with a fluorescence microscope. (a) AFM topography image of
patterns that were written with t-SPL with features ranging from a
single line (86 ± 6 nm) to 1920 nm in width. (b) Fluorescence
microscopy image of the same patterns acquired under irradiation with
UV light. (c) Profiles extracted from the surface topography image
(solid black) and the fluorescence intensity image (dashed red) along
the fast scan direction in (a) and (b), respectively.
Comparison of the resolution achieved by t-SPL patterning of the
color-switchable fluorescent UPy–OPV–UPy and optical
readout with a fluorescence microscope. (a) AFM topography image of
patterns that were written with t-SPL with features ranging from a
single line (86 ± 6 nm) to 1920 nm in width. (b) Fluorescence
microscopy image of the same patterns acquired under irradiation with
UV light. (c) Profiles extracted from the surface topography image
(solid black) and the fluorescence intensity image (dashed red) along
the fast scan direction in (a) and (b), respectively.On the basis of the findings above, and to analyze
further the achievable limits of the thermal patterning of UPy–OPV–UPy,
the correlation between pitch size and pattern quality was investigated.
Furthermore, pitch modulation can be used to regulate the green luminescence
intensity. Areas of the supramolecular polymer film were written at
a pitch size ranging from 20 to 2.56 μm and subsequently imaged
directly by the NanoFrazor’s probe as shown in Figure a. For a pitch size below 80
nm, a trench is formed on the left and a pileup is formed on the right
side of the pattern due to the displacement of the melt as already
discussed in connection with the data shown in Figure a. Above a pitch size of 80 nm, individual
indents are distinguishable by AFM and the material displaced during
the indentation remains around each indent. The fwhm sizes of the
indents are 93 ± 12 and 46 ± 6 nm in the fast-scan direction
(horizontal) and in the slow-scan direction (vertical), respectively.
A contribution to the asymmetry can be attributed to the tool-specific
patterning mechanism of the thermal scanning probe system, whereby
the piezo-actuated sample stage is continuously moving in the fast-scan
direction at a speed of 0.2 mm/s, while the cantilever is actuated
toward the substrate for short indentation times on the order of microseconds.
A comparison between the topography and fluorescence image (Figure b) reveals that the
green fluorescence intensity can be tuned over a broad range by varying
the indent density, but the indentation pattern itself is not visible
because the t-SPL resolution is higher than the diffraction limit
of optical microscopy. In Figure c, the normalized fluorescence intensity is plotted
as a function of the pitch size, whereby the fluorescence intensity
sharply decreases with increasing pitch. For pitch sizes larger than
the indent length, the intensity is inverse proportional to the square
of the pitch size, but as soon as the indents start to overlap, the
normalized intensity approaches 1. The inset in Figure c shows a magnification of the surface topography
with four indents at a pitch size of 320 and 160 nm and a sketch of
the model used to fit the fluorescence intensity. In the model it
is assumed that the extent of the thermally modified polymer can be
approximated with a semiellipsoid with a uniform fluorescence density.
The normalized intensity perceivable by fluorescence microscopy is
then the ratio between thermally modified volume and the square of
the pitch size times the depth of the indent. Details about the model
used to fit the fluorescence intensity can be found in the Supporting Information. We fit the measured intensity
as a function of pitch size as shown in Figure c, whereby only the major axis of the ellipse
was used as a fit parameter. The aspect ratio of the indents (ra = 2.2 ± 0.6) can be determined from the
surface topography and should be similar for both, the topological
indent and the fluorescent area, due to isotropic spreading of heat.
We obtain an indent length of 189 ± 12 nm from the fitted fluorescence
curve, which is in agreement with the value obtained by measuring
the peak-to-peak indent length from the topography (184 ± 31
nm). From these measurements, we conclude that the disaggregation
of the fluorophores due to the heat provided by the tip is confined
to only a few nanometers around the indent and the resolution is mainly
limited by the tip apex size.
Figure 4
Correlation between fluorescence intensity and
pitch size. (a) AFM topography image for different indent pitch sizes
varying from 20 nm to 2.56 μm (scale bar 2 μm). (b) Corresponding
fluorescence microscopy images acquired under irradiation with UV
light (scale bar 2 μm). (c) Plot of the normalized fluorescence
intensity as a function of the pitch size. The inset shows the surface
topography of indents at a pitch size of 320 and 160 nm and the corresponding
model used to fit the green emission per unit area (scale bar 200
nm).
Correlation between fluorescence intensity and
pitch size. (a) AFM topography image for different indent pitch sizes
varying from 20 nm to 2.56 μm (scale bar 2 μm). (b) Corresponding
fluorescence microscopy images acquired under irradiation with UV
light (scale bar 2 μm). (c) Plot of the normalized fluorescence
intensity as a function of the pitch size. The inset shows the surface
topography of indents at a pitch size of 320 and 160 nm and the corresponding
model used to fit the green emission per unit area (scale bar 200
nm).The combination of t-SPL with
a color-switchable fluorescent polymer provides the ability to pattern
multifaceted nanostructures, which are optically hidden due to the
limited contrast or resolution of most optical microscopy systems
but topologically detectable with AFM. To demonstrate this, a series
of patterns were fabricated with distinct information in the topographic
and fluorescent states by way of consecutive patterning steps and
readout using both AFM and fluorescence microscopy. First, a square
of 30 × 30 μm2 was patterned with a pitch size
of 60 nm (Figure a).
Subsequently, a second pattern was produced within the square, which
can be distinguished from the background by AFM (Figure b). However, as shown in Figure c, the second pattern
is virtually indistinguishable from the background when measured with
fluorescence microscopy. To obtain truly hidden features, it is important
to avoid deep indents and pileup (Figure S5). To visualize the features with a fluorescence microscope, we choose
the hidden t-SPL patterns to be larger than 1 μm, which makes
them easily detectable by optical microscopy, whereas in a real security
feature the information can be stored in the form of a binary pattern
composed of indents and unpatterned areas. For these patterns, we
expect an extended lifetime because at room temperature UPy–OPV–UPy
is in the glassy state and therefore the particular stacking of the
OPV cores—that lead to either green or red fluorescence—is
frozen in.
Figure 5
Hidden in plain sight. Thermal patterning of nanoscale features
with distinct topography but the same fluorescence intensity. (a)
T-SPL patterning of the square of 30 × 30 μm2 at a pitch size of 60 nm. (b) Overlay of a second t-SPL pattern
within the prepatterned square, containing the logos of the two institutes.
(c) Fluorescence microscopy image of the patterned area acquired under
irradiation with UV light, showing that the features programmed in
topology mode are virtually invisible in fluorescence microscopy mode.
Hidden in plain sight. Thermal patterning of nanoscale features
with distinct topography but the same fluorescence intensity. (a)
T-SPL patterning of the square of 30 × 30 μm2 at a pitch size of 60 nm. (b) Overlay of a second t-SPL pattern
within the prepatterned square, containing the logos of the two institutes.
(c) Fluorescence microscopy image of the patterned area acquired under
irradiation with UV light, showing that the features programmed in
topology mode are virtually invisible in fluorescence microscopy mode.
Conclusions
We have demonstrated
nanoscale thermal patterning with a heated probe of a novel thermochromic
luminescent supramolecular polymer. t-SPL nanoindentation of thin
coatings of UPy–OPV–UPy resulted in a fluorescent as
well as a topographical pattern at the nanometer scale on account
of a low viscosity above Tg and the ability
to kinetically trap a disassembled monomeric state that resulted in
a hypochromic shift. In-depth studies of the patterning conditions
showed that altering the nanoscale indentation pitch modulated the
green fluorescence intensity, material pileup, and overall patterning
parameters. This novel coating material combined with the writing
parameters could be interesting for application as nano- to micrometer-scale
anticounterfeiting features in environments with temperatures not
exceeding 180 °C with at least three degrees of security: fluorescence
intensity, topography, and the concept of hidden features. Recent
advances in laser-assisted and ultrafast nanoimprint lithography could
enable mix and match lithography with t-SPL to fabricate large-area
patterns in UPy–OPV–UPy which can be locally modified
in a subsequent step with a heated probe to create highly customized
features that are unique and difficult to reproduce. Furthermore,
we anticipate that the presented novel polymer has a broad application
beyond patterning by thermal probes such as in temperature, and strain
sensing.
Authors: Armin W Knoll; David Pires; Olivier Coulembier; Philippe Dubois; James L Hedrick; Jane Frommer; Urs Duerig Journal: Adv Mater Date: 2010-08-17 Impact factor: 30.849
Authors: Keith M Carroll; Xi Lu; Suenne Kim; Yang Gao; Hoe-Joon Kim; Suhas Somnath; Laura Polloni; Roman Sordan; William P King; Jennifer E Curtis; Elisa Riedo Journal: Nanoscale Date: 2014 Impact factor: 7.790
Authors: Adrian Podpirka; Woo-Kyung Lee; Jed I Ziegler; Todd H Brintlinger; Jonathan R Felts; Blake S Simpkins; Nabil D Bassim; Arnaldo R Laracuente; Paul E Sheehan; Laura B Ruppalt Journal: Nanoscale Date: 2017-06-29 Impact factor: 7.790
Authors: E Albisetti; D Petti; M Pancaldi; M Madami; S Tacchi; J Curtis; W P King; A Papp; G Csaba; W Porod; P Vavassori; E Riedo; R Bertacco Journal: Nat Nanotechnol Date: 2016-03-07 Impact factor: 39.213
Authors: Anna Lavrenova; Diederik W R Balkenende; Yoshimitsu Sagara; Stephen Schrettl; Yoan C Simon; Christoph Weder Journal: J Am Chem Soc Date: 2017-03-21 Impact factor: 15.419